spec: prerelase 0.7 for testing hugetlb map for stack

Change-Id: I4997340cd984ca8915e45749b91b1d72c1de85af

.gitignore

@@ -13,6 +13,10 @@ old_timestamp
 CMakeFiles
 CMakeCache.txt
 Makefile
+!test/*/*/Makefile
+!test/signalonfork+wait/Makefile
+!test/perf_overflow/Makefile
+!test/*/*/*.cmd
 Kbuild
 cmake_install.cmake
 config.h
@@ -33,3 +37,8 @@ executer/user/libmcexec.a
 executer/user/libldump2mcdump.so
 executer/user/eclair
 tools/mcstat/mcstat
+/_CPack_Packages
+/CPackSourceConfig.cmake
+CPackConfig.cmake
+/build
+mckernel-*.tar.gz

.gitmodules

@@ -1,3 +1,6 @@
 [submodule "ihk"]
 	path = ihk
 	url = https://github.com/RIKEN-SysSoft/ihk.git
+[submodule "executer/user/lib/libdwarf/libdwarf"]
+	path = executer/user/lib/libdwarf/libdwarf
+	url = https://github.com/bgerofi/libdwarf.git

CMakeLists.txt

@@ -7,71 +7,157 @@ endif (NOT CMAKE_BUILD_TYPE)
 enable_language(C ASM)
 
 project(mckernel C ASM)
-set(MCKERNEL_VERSION "1.6.0")
+set(MCKERNEL_VERSION "1.7.0")
+
+# See "Fedora Packaging Guidlines -- Versioning"
+set(MCKERNEL_RELEASE "0.7")
 
 set(CMAKE_MODULE_PATH ${CMAKE_SOURCE_DIR}/cmake/modules)
 # for rpmbuild
 if(DEFINED SYSCONF_INSTALL_DIR)
 	set(CMAKE_INSTALL_SYSCONFDIR "${SYSCONF_INSTALL_DIR}")
 endif()
-include(GNUInstallDirs)
-include(CMakeParseArguments)
-include(Kbuild)
-include(Ksym)
-include(CheckCCompilerFlag)
-
-set(CFLAGS_WARNINGS "-Wall -Wextra -Wno-unused-parameter -Wno-sign-compare -Wno-unused-function")
-CHECK_C_COMPILER_FLAG(-Wno-implicit-fallthrough IMPLICIT_FALLTHROUGH)
-if(IMPLICIT_FALLTHROUGH)
-	set(CFLAGS_WARNINGS "${CFLAGS_WARNINGS} -Wno-implicit-fallthrough")
-endif(IMPLICIT_FALLTHROUGH)
-
-# C flags need to be set before enabling language?
-set(CMAKE_C_FLAGS_DEBUG "-g ${CFLAGS_WARNINGS}" CACHE STRING "Debug compiler flags")
-set(CMAKE_C_FLAGS_RELEASE "${CFLAGS_WARNINGS}" CACHE STRING "Release compiler flags")
-
-# build options
-option(ENABLE_WERROR "Enable -Werror" OFF)
-if (ENABLE_WERROR)
-	add_compile_options("-Werror")
-endif(ENABLE_WERROR)
 
 if (CMAKE_SYSTEM_PROCESSOR STREQUAL "x86_64")
 	set(BUILD_TARGET "smp-x86" CACHE STRING "Build target: smp-x86 | smp-arm64")
 elseif (CMAKE_SYSTEM_PROCESSOR STREQUAL "aarch64")
 	set(BUILD_TARGET "smp-arm64" CACHE STRING "Build target: smp-x86 | smp-arm64")
 endif()
+
 if (BUILD_TARGET STREQUAL "smp-x86")
 	set(ARCH "x86_64")
 elseif (BUILD_TARGET STREQUAL "smp-arm64")
 	set(ARCH "arm64")
+endif()
+
+include(GNUInstallDirs)
+include(CMakeParseArguments)
+include(Kbuild)
+include(CheckCCompilerFlag)
+include(AutoconfHelper)
+
+CHECK_C_COMPILER_FLAG(-Wno-implicit-fallthrough IMPLICIT_FALLTHROUGH)
+if(IMPLICIT_FALLTHROUGH)
+	set(EXTRA_WARNINGS "-Wno-implicit-fallthrough")
+endif(IMPLICIT_FALLTHROUGH)
+
+# build options
+set(CFLAGS_WARNING "-Wall" "-Wextra" "-Wno-unused-parameter" "-Wno-sign-compare" "-Wno-unused-function" ${EXTRA_WARNINGS} CACHE STRING "Warning flags")
+add_compile_options(${CFLAGS_WARNING})
+
+option(ENABLE_WERROR "Enable -Werror" OFF)
+if (ENABLE_WERROR)
+	add_compile_options("-Werror")
+endif(ENABLE_WERROR)
+
+option(ENABLE_LINUX_WORK_IRQ_FOR_IKC "Use Linux work IRQ for IKC IPI" ON)
+if (ENABLE_LINUX_WORK_IRQ_FOR_IKC)
+	set(KBUILD_C_FLAGS "${KBUILD_C_FLAGS} -DIHK_IKC_USE_LINUX_WORK_IRQ")
+	add_definitions(-DIHK_IKC_USE_LINUX_WORK_IRQ)
+endif()
+
+if (BUILD_TARGET STREQUAL "smp-arm64")
 	foreach(i RANGE 1 120)
 		add_definitions(-DPOSTK_DEBUG_ARCH_DEP_${i} -DPOSTK_DEBUG_TEMP_FIX_${i})
 		set(KBUILD_C_FLAGS "${KBUILD_C_FLAGS} -DPOSTK_DEBUG_ARCH_DEP_${i} -DPOSTK_DEBUG_TEMP_FIX_${i}")
 	endforeach()
-	add_definitions(-DCONFIG_ARM64_64K_PAGES -DCONFIG_ARM64_VA_BITS=48)
+
+	execute_process(COMMAND awk -F= "$1 == \"CONFIG_ARM64_64K_PAGES\" { print $2; exit; }" "${KERNEL_DIR}/.config"
+			OUTPUT_VARIABLE CONFIG_ARM64_64K_PAGES OUTPUT_STRIP_TRAILING_WHITESPACE)
+	execute_process(COMMAND awk -F= "$1 == \"CONFIG_ARM64_VA_BITS\" { print $2; exit; }" "${KERNEL_DIR}/.config"
+			OUTPUT_VARIABLE CONFIG_ARM64_VA_BITS OUTPUT_STRIP_TRAILING_WHITESPACE)
+
+	message("Host kernel CONFIG_ARM64_64K_PAGES=${CONFIG_ARM64_64K_PAGES}")
+	message("Host kernel CONFIG_ARM64_VA_BITS=${CONFIG_ARM64_VA_BITS}")
+
+	if(CONFIG_ARM64_64K_PAGES STREQUAL "y")
+		if(CONFIG_ARM64_VA_BITS STREQUAL 42)
+			add_definitions(-DCONFIG_ARM64_PGTABLE_LEVELS=2 -DCONFIG_ARM64_VA_BITS=42 -DCONFIG_ARM64_64K_PAGES)
+			set(LINKER_SCRIPT "smp-arm64_type3.lds")
+		elseif(CONFIG_ARM64_VA_BITS STREQUAL 48)
+			add_definitions(-DCONFIG_ARM64_PGTABLE_LEVELS=3 -DCONFIG_ARM64_VA_BITS=48 -DCONFIG_ARM64_64K_PAGES)
+			set(LINKER_SCRIPT "smp-arm64_type4.lds")
+		endif()
+	else(CONFIG_ARM64_64K_PAGES STREQUAL "y")
+		if(CONFIG_ARM64_VA_BITS STREQUAL 39)
+			add_definitions(-DCONFIG_ARM64_PGTABLE_LEVELS=3 -DCONFIG_ARM64_VA_BITS=39)
+			set(LINKER_SCRIPT "smp-arm64_type1.lds")
+		elseif(CONFIG_ARM64_VA_BITS STREQUAL 48)
+			add_definitions(-DCONFIG_ARM64_PGTABLE_LEVELS=4 -DCONFIG_ARM64_VA_BITS=48)
+			set(LINKER_SCRIPT "smp-arm64_type2.lds")
+		endif()
+	endif(CONFIG_ARM64_64K_PAGES STREQUAL "y")
 endif()
 set_property(CACHE BUILD_TARGET PROPERTY STRINGS smp-x86 smp-arm64)
 
+# define MAP_KERNEL_START
+
+set(tmpdir ${CMAKE_CURRENT_BINARY_DIR}/tmp.resolve_MODULES_END)
+file(REMOVE_RECURSE ${tmpdir})
+file(MAKE_DIRECTORY ${tmpdir})
+file(WRITE ${tmpdir}/driver.c "#include <linux/module.h>\n")
+file(APPEND ${tmpdir}/driver.c "unsigned long MAP_KERNEL_START = MODULES_END - (1UL << 23);\n")
+file(WRITE ${tmpdir}/Makefile "obj-m := driver.o\n")
+file(APPEND ${tmpdir}/Makefile "all:\n")
+file(APPEND ${tmpdir}/Makefile "\tmake ${KBUILD_MAKE_FLAGS_STR} -C ${KERNEL_DIR} M=${tmpdir} modules\n")
+
+execute_process(COMMAND make -C ${tmpdir})
+execute_process(COMMAND bash -c "offset=`readelf -S ${tmpdir}/driver.ko | grep .data | sed 's/.* //g'`; echo $((0x$offset))"
+  OUTPUT_VARIABLE MAP_KERNEL_START_OFFSET OUTPUT_STRIP_TRAILING_WHITESPACE)
+execute_process(COMMAND bash -c "dd if=${tmpdir}/driver.ko bs=1 skip=${MAP_KERNEL_START_OFFSET} count=8 2>/dev/null | od -tx8 -Ax | head -1 | sed 's|.* |0x|g'"
+  OUTPUT_VARIABLE MAP_KERNEL_START OUTPUT_STRIP_TRAILING_WHITESPACE)
+
+
 set(ENABLE_MEMDUMP ON)
 option(ENABLE_PERF "Enable perf support" ON)
 option(ENABLE_RUSAGE "Enable rusage support" ON)
-option(ENABLE_MCOVERLAYFS "Enable overlay filesystem" OFF)
 option(ENABLE_QLMPI "Enable qlmpi programs" OFF)
 option(ENABLE_UTI "Enable uti support" OFF)
 option(ENABLE_UBSAN "Enable undefined behaviour sanitizer on mckernel size" OFF)
+option(ENABLE_PER_CPU_ALLOC_CACHE "Enable per-CPU allocator cache (ThunderX2 workaround)" OFF)
+
+find_package(PkgConfig REQUIRED)
+set(PKG_CONFIG_USE_CMAKE_PREFIX_PATH ON)
 
 find_library(LIBRT rt)
+if (NOT LIBRT)
+	message(FATAL_ERROR "error: couldn't find librt")
+endif()
 find_library(LIBNUMA numa)
+if (NOT LIBNUMA)
+	message(FATAL_ERROR "error: couldn't find libnuma")
+endif()
 find_library(LIBBFD bfd)
+if (NOT LIBBFD)
+	message(FATAL_ERROR "error: couldn't find libbfd")
+endif()
 find_library(LIBIBERTY iberty)
+if (NOT LIBIBERTY)
+	message(FATAL_ERROR "error: couldn't find libiberty")
+endif()
+
+find_library(LIBDWARF dwarf)
+
+if (NOT LIBDWARF)
+	if (CMAKE_CROSSCOMPILING)
+		message(FATAL_ERROR "Could not find libdwarf.so, install libdwarf-devel to ${CMAKE_FIND_ROOT_PATH}")
+	endif()
+	message("WARNING: libdwarf will be compiled locally")
+	enable_language(CXX)
+else()
+	# Note that libdwarf-devel provides /usr/include/libdwarf/dwarf.h
+	# but elfutils-devel provides /usr/include/dwarf.h
+	# while mcinspect.c performs "#include <dwarf.h>"
+	find_path(DWARF_H dwarf.h PATH_SUFFIXES libdwarf)
+endif()
 
 if (ENABLE_QLMPI)
 	find_package(MPI REQUIRED)
 endif()
 
 if (ENABLE_UTI)
-	find_library(LIBSYSCALL_INTERCEPT syscall_intercept)
+	pkg_check_modules(LIBSYSCALL_INTERCEPT REQUIRED libsyscall_intercept)
+	link_directories(${LIBSYSCALL_INTERCEPT_LIBRARY_DIRS})
 endif()
 
 string(REGEX REPLACE "^([0-9]+)\\.([0-9]+)\\.([0-9]+)(-([0-9]+)(.*))?" "\\1;\\2;\\3;\\5;\\6" LINUX_VERSION ${UNAME_R})
@@ -81,29 +167,11 @@ list(GET LINUX_VERSION 2 LINUX_VERSION_PATCH)
 list(GET LINUX_VERSION 3 LINUX_VERSION_RELEASE)
 math(EXPR LINUX_VERSION_CODE "${LINUX_VERSION_MAJOR} * 65536 + ${LINUX_VERSION_MINOR} * 256 + ${LINUX_VERSION_PATCH}")
 
-ksym(sys_mount PREFIX MCCTRL_)
-ksym(sys_umount PREFIX MCCTRL_)
-ksym(sys_unshare PREFIX MCCTRL_)
-ksym(zap_page_range PREFIX MCCTRL_)
-ksym(vdso_image_64 PREFIX MCCTRL_)
-ksym(vdso_start PREFIX MCCTRL_)
-ksym(vdso_end PREFIX MCCTRL_)
-ksym(vdso_pages PREFIX MCCTRL_)
-ksym(__vvar_page PREFIX MCCTRL_)
-ksym(hpet_address PREFIX MCCTRL_)
-# POSTK_DEBUG_ARCH_DEP_50, add:find kernel symbol.
-ksym(vdso_spec PREFIX MCCTRL_)
-ksym(hv_clock PREFIX MCCTRL_)
-ksym(sys_readlink PREFIX MCCTRL_)
-ksym(walk_page_range PREFIX MCCTRL_)
-
-
 # compat with various install paths
-set(MCKERNEL_LIBDIR ${CMAKE_INSTALL_FULL_LIBDIR})
 set(BINDIR ${CMAKE_INSTALL_FULL_BINDIR})
 set(SBINDIR ${CMAKE_INSTALL_FULL_SBINDIR})
-set(ETCDIR ${CMAKE_INSTALL_FULL_SYSCONFDIR})
-set(ROOTFSDIR "${CMAKE_INSTALL_PREFIX}/rootfs")
+set(ETCDIR ${CMAKE_INSTALL_PREFIX}/etc)
+set(ROOTFSDIR "/rootfs")
 if (CMAKE_INSTALL_PREFIX STREQUAL "/usr")
 	set(KMODDIR "/lib/modules/${UNAME_R}/extra/mckernel")
 	set(MCKERNELDIR "${CMAKE_INSTALL_FULL_DATADIR}/mckernel/${BUILD_TARGET}")
@@ -138,23 +206,23 @@ configure_file(config.h.in config.h)
 
 # actual build section - just subdirs
 add_subdirectory(executer/kernel/mcctrl)
-if (ENABLE_MCOVERLAYFS)
-	add_subdirectory(executer/kernel/mcoverlayfs)
-endif()
 add_subdirectory(executer/user)
 add_subdirectory(kernel)
 add_subdirectory(tools/mcstat)
+add_subdirectory(tools/crash)
 
-configure_file(arch/x86_64/tools/mcreboot-smp-x86.sh.in mcreboot.sh @ONLY)
-configure_file(arch/x86_64/tools/mcstop+release-smp-x86.sh.in mcstop+release.sh @ONLY)
-configure_file(arch/x86_64/tools/mcreboot.1in mcreboot.1 @ONLY)
+configure_file(scripts/mcreboot-smp.sh.in mcreboot.sh @ONLY)
+configure_file(scripts/mcstop+release-smp.sh.in mcstop+release.sh @ONLY)
+configure_file(scripts/mcreboot.1in mcreboot.1 @ONLY)
+configure_file(scripts/eclair-dump-backtrace.exp.in eclair-dump-backtrace.exp @ONLY)
 install(PROGRAMS
 		"${CMAKE_CURRENT_BINARY_DIR}/mcreboot.sh"
 		"${CMAKE_CURRENT_BINARY_DIR}/mcstop+release.sh"
 	DESTINATION "${CMAKE_INSTALL_SBINDIR}")
-install(FILES
-		"arch/x86_64/tools/irqbalance_mck.service"
-		"arch/x86_64/tools/irqbalance_mck.in"
+install(PROGRAMS
+		"${CMAKE_CURRENT_BINARY_DIR}/eclair-dump-backtrace.exp"
+	DESTINATION "${CMAKE_INSTALL_BINDIR}")
+install(FILES "scripts/irqbalance_mck.in"
 	DESTINATION "${CMAKE_INSTALL_SYSCONFDIR}")
 install(FILES "${CMAKE_CURRENT_BINARY_DIR}/mcreboot.1"
 	DESTINATION "${CMAKE_INSTALL_MANDIR}/man1")
@@ -162,7 +230,7 @@ install(FILES "${CMAKE_CURRENT_BINARY_DIR}/mcreboot.1"
 
 configure_file(scripts/mckernel.spec.in scripts/mckernel.spec @ONLY)
 set(CPACK_SOURCE_PACKAGE_FILE_NAME "${CMAKE_PROJECT_NAME}-${MCKERNEL_VERSION}")
-set(CPACK_SOURCE_IGNORE_FILES "/.git$")
+set(CPACK_SOURCE_IGNORE_FILES "/.git/;/build;/CMakeCache.txt$;/CMakeFiles$;/Makefile$")
 set(CPACK_SOURCE_INSTALLED_DIRECTORIES "${CMAKE_SOURCE_DIR};/;${IHK_FULL_SOURCE_DIR};/ihk;${CMAKE_BINARY_DIR}/scripts;/scripts")
 set(CPACK_SOURCE_GENERATOR "TGZ")
 include(CPack)
@@ -181,12 +249,14 @@ message("KERNEL_DIR: ${KERNEL_DIR}")
 message("SYSTEM_MAP: ${SYSTEM_MAP}")
 message("VMLINUX: ${VMLINUX}")
 message("KBUILD_C_FLAGS: ${KBUILD_C_FLAGS}")
+message("MAP_KERNEL_START: ${MAP_KERNEL_START}")
 message("ENABLE_MEMDUMP: ${ENABLE_MEMDUMP}")
 message("ENABLE_PERF: ${ENABLE_PERF}")
 message("ENABLE_RUSAGE: ${ENABLE_RUSAGE}")
-message("ENABLE_MCOVERLAYFS: ${ENABLE_MCOVERLAYFS}")
 message("ENABLE_QLMPI: ${ENABLE_QLMPI}")
 message("ENABLE_UTI: ${ENABLE_UTI}")
 message("ENABLE_WERROR: ${ENABLE_WERROR}")
 message("ENABLE_UBSAN: ${ENABLE_UBSAN}")
+message("ENABLE_LINUX_WORK_IRQ_FOR_IKC: ${ENABLE_LINUX_WORK_IRQ_FOR_IKC}")
+message("ENABLE_PER_CPU_ALLOC_CACHE: ${ENABLE_PER_CPU_ALLOC_CACHE}")
 message("-------------------------------")

KNOWN_BUGS.md

@@ -0,0 +1,70 @@
+Linux crash when offlining CPU (el7, hardware-specific)
+=========================================================
+
+On some hardware with el7 kernel, linux can crash due to a bug in the
+irq handling when offlining CPUs (reserve cpu part of mcreboot)
+
+Example stack trace:
+```
+[ 4147.052753] BUG: unable to handle kernel NULL pointer dereference at 0000000000000040
+[ 4147.060677] IP: [<ffffffff8102ce26>] check_irq_vectors_for_cpu_disable+0x86/0x1c0
+[ 4147.068226] PGD 1057e44067 PUD 105f1e7067 PMD 0
+[ 4147.072935] Oops: 0000 [#1] SMP
+[ 4147.076230] Modules linked in: mcctrl(OE) ihk_smp_x86_64(OE) ihk(OE) xt_CHECKSUM ipt_MASQUERADE nf_nat_masquerade_ipv4 tun rpcsec_gss_krb5 nfsv4 dns_resolver nfs fscache ip6t_rpfilter ipt_REJECT nf_reject_ipv4 ip6t_REJECT nf_reject_ipv6 xt_conntrack ip_set nfnetlink ebtable_nat ebtable_broute bridge stp llc ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack iptable_mangle iptable_security iptable_raw ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter rpcrdma ib_isert iscsi_target_mod ib_iser libiscsi scsi_transport_iscsi ib_srpt target_core_mod ib_srp scsi_transport_srp scsi_tgt ib_ipoib rdma_ucm ib_ucm ib_uverbs ib_umad rdma_cm ib_cm iw_cm mlx4_ib ib_core
+[ 4147.148619]  dm_mirror dm_region_hash dm_log dm_mod sb_edac edac_core intel_powerclamp coretemp ext4 mbcache jbd2 intel_rapl iosf_mbi kvm_intel kvm irqbypass crc32_pclmul ghash_clmulni_intel aesni_intel lrw gf128mul ipmi_ssif glue_helper ablk_helper joydev iTCO_wdt iTCO_vendor_support cryptd ipmi_si ipmi_devintf ipmi_msghandler pcspkr wmi mei_me mei lpc_ich i2c_i801 sg ioatdma shpchp nfsd auth_rpcgss nfs_acl lockd grace sunrpc ip_tables xfs libcrc32c mlx4_en sd_mod crc_t10dif crct10dif_generic mgag200 drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops ttm isci igb drm mlx4_core libsas ahci libahci scsi_transport_sas libata crct10dif_pclmul ptp crct10dif_common pps_core crc32c_intel dca i2c_algo_bit i2c_core devlink [last unloaded: ihk]
+[ 4147.215370] CPU: 6 PID: 38 Comm: migration/6 Tainted: G           OE  ------------ T 3.10.0-693.2.2.el7.x86_64 #1
+[ 4147.225672] Hardware name: SGI.COM C1104G-RP5/X9DRG-HF, BIOS 3.0  10/25/2013
+[ 4147.232747] task: ffff880174689fa0 ti: ffff8801746ac000 task.ti: ffff8801746ac000
+[ 4147.240278] RIP: 0010:[<ffffffff8102ce26>]  [<ffffffff8102ce26>] check_irq_vectors_for_cpu_disable+0x86/0x1c0
+[ 4147.250275] RSP: 0018:ffff8801746afd30  EFLAGS: 00010046
+[ 4147.255608] RAX: 0000000000000000 RBX: 000000000000004e RCX: 0000000000000000
+[ 4147.262770] RDX: 0000000000000020 RSI: 000000000000005f RDI: 0000000000000023
+[ 4147.269936] RBP: ffff8801746afd58 R08: 0000000000000001 R09: ffff88017f800490
+[ 4147.277103] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000006
+[ 4147.284269] R13: 0000000000000000 R14: ffff88085ca82500 R15: 000000000000005f
+[ 4147.291429] FS:  0000000000000000(0000) GS:ffff88085fb80000(0000) knlGS:0000000000000000
+[ 4147.299556] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
+[ 4147.305326] CR2: 0000000000000040 CR3: 0000001059704000 CR4: 00000000001407e0
+[ 4147.312490] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
+[ 4147.319659] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
+[ 4147.326827] Stack:
+[ 4147.328857]  ffff8808f43078c8 ffff8808f4307850 0000000000000286 ffff8808f4307701
+[ 4147.336384]  0000000000000000 ffff8801746afd70 ffffffff81052a82 0000000200000000
+[ 4147.343915]  ffff8801746afd88 ffffffff81693ca3 0000000000000003 ffff8801746afdc0
+[ 4147.351447] Call Trace:
+[ 4147.353921]  [<ffffffff81052a82>] native_cpu_disable+0x12/0x40
+[ 4147.359795]  [<ffffffff81693ca3>] take_cpu_down+0x13/0x40
+[ 4147.365236]  [<ffffffff81116899>] multi_cpu_stop+0xd9/0x100
+[ 4147.370850]  [<ffffffff811167c0>] ? cpu_stop_should_run+0x50/0x50
+[ 4147.376983]  [<ffffffff81116ab7>] cpu_stopper_thread+0x97/0x150
+[ 4147.382942]  [<ffffffff816a8fad>] ? __schedule+0x39d/0x8b0
+[ 4147.388461]  [<ffffffff810b909f>] smpboot_thread_fn+0x12f/0x180
+[ 4147.394406]  [<ffffffff810b8f70>] ? lg_double_unlock+0x40/0x40
+[ 4147.400276]  [<ffffffff810b098f>] kthread+0xcf/0xe0
+[ 4147.405182]  [<ffffffff810b08c0>] ? insert_kthread_work+0x40/0x40
+[ 4147.411319]  [<ffffffff816b4f58>] ret_from_fork+0x58/0x90
+[ 4147.418893]  [<ffffffff810b08c0>] ? insert_kthread_work+0x40/0x40
+[ 4147.426524] Code: 81 fb 00 01 00 00 0f 84 8a 00 00 00 89 d8 65 44 8b 3c 85 20 c6 00 00 45 85 ff 78 e1 44 89 ff e8 91 31 10 00 48 63 15 7e 10 af 00 <48> 8b 70 40 48 c7 c7 80 71 cf 81 49 89 c6 48 83 c2 3f 48 c1 fa
+[ 4147.450352] RIP  [<ffffffff8102ce26>] check_irq_vectors_for_cpu_disable+0x86/0x1c0
+[ 4147.460135]  RSP <ffff8801746afd30>
+[ 4147.465154] CR2: 0000000000000040
+```
+
+This bug has been fixed upstream, but redhat will not backport the fixes.
+You can work around the problem with a kpatch by backporting the three
+following commits:
+
+x86: irq: Get correct available vectors for cpu disable
+https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=ac2a55395eddccd6e3e39532df9869d61e97b2ee
+
+x86/irq: Check for valid irq descriptor in check_irq_vectors_for_cpu_disable()
+https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d97eb8966c91f2c9d05f0a22eb89ed5b76d966d1
+
+x86/irq: Use proper locking in check_irq_vectors_for_cpu_disable()
+https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=cbb24dc761d95fe39a7a122bb1b298e9604cae15
+
+
+Alternatively, since it is related to the irq configuration, it might
+be possible to mitigate the issue by setting the irq affinities early
+on and making sure none of the cpus that will be offlined have any irq
+configured.

NEWS.md

@@ -0,0 +1,540 @@
+=============================================
+What's new in version 1.7.0rc4 (Apr 15, 2020)
+=============================================
+
+----------------------
+McKernel major updates
+----------------------
+1. arm64: Contiguous PTE support
+2. arm64: Scalable Vector Extension (SVE) support
+3. arm64: PMU overflow interrupt support
+4. xpmem: Support large page attachment
+5. arm64 port: Direct access to Mckernel memory from Linux
+6. arm64 port: utility thread offloading, which spawns thread onto
+   Linux CPU
+7. eclair: support for live debug
+8. Crash utility extension
+9. Replace mcoverlayfs with a soft userspace overlay
+10. Build system is switched to cmake
+11. Core dump includes thread information
+
+------------------------
+McKernel major bug fixes
+------------------------
+1. shmobj: Fix rusage counting for large page
+2. mcctrl control: task start_time changed to u64 nsec
+3. mcctrl: add handling for one more level of page tables
+4. Add kernel argument to turn on/off time sharing
+5. flatten_string/process env: realign env and clear trailing bits
+6. madvise: Add MADV_HUGEPAGE support
+8. mcctrl: remove in-kernel calls to syscalls
+9. arch_cpu_read_write_register: error return fix.
+10. set_cputime(): interrupt enable/disable fix.
+11. set_mempolicy(): Add mode check.
+12. mbind(): Fix memory_range_lock deadlock.
+13. ihk_ikc_recv: Record channel to packet for release
+14. Add set_cputime() kernel to kernel case and mode enum.
+15. execve: Call preempt_enable() before error-exit
+16. memory/x86_64: fix linux safe_kernel_map
+17. do_kill(): fix pids table when nr of threads is larger than num_processors
+18. shmget: Use transparent huge pages when page size isn't specified
+19. prctl: Add support for PR_SET_THP_DISABLE and PR_GET_THP_DISABLE
+20. monitor_init: fix undetected hang on highest numbered core
+21. init_process_stack: change premapped stack size based on arch
+22. x86 syscalls: add a bunch of XXat() delegated syscalls
+23. do_pageout: fix direct kernel-user access
+24. stack: add hwcap auxval
+25. perf counters: add arch-specific perf counters
+26. Added check of nohost to terminate_host().
+27. kmalloc: Fix address order in free list
+28. sysfs: use nr_cpu_ids for cpumasks (fixes libnuma parsing error on ARM)
+29. monitor_init: Use ihk_mc_cpu_info()
+30. Fix ThunderX2 write-combined PTE flag insanity
+31. ARM: eliminate zero page mapping (i.e, init_low_area())
+32. eliminate futex_cmpxchg_enabled check (not used and dereffed a NULL pointer)
+33. page_table: Fix return value of lookup_pte when ptl4 is blank
+34. sysfs: add missing symlinks for cpu/node
+35. Make Linux handler run when mmap to procfs.
+36. Separate mmap area from program loading (relocation) area
+37. move rusage into kernel ELF image (avoid dynamic alloc before NUMA init)
+38. arm: turn off cpu on panic
+39. page fault handler: protect thread accesses
+40. Register PPD and release_handler at the same time.
+41. fix to missing exclusive processing between terminate() and
+    finalize_process().
+42. perfctr_stop: add flags to no 'disable_intens'
+43. fileobj, shmobj: free pages in object destructor (as opposed to page_unmap())
+44. clear_range_l1, clear_range_middle: Fix handling contiguous PTE
+45. do_mmap: don't pre-populate the whole file when asked for smaller segment
+46. invalidate_one_page: Support shmobj and contiguous PTE
+47. ubsan: fix undefined shifts
+48. x86: disable zero mapping and add a boot pt for ap trampoline
+49. rusage: Don't count PF_PATCH change
+50. Fixed time processing.
+51. copy_user_pte: vmap area not owned by McKernel
+52. gencore: Zero-clear ELF header and memory range table
+53. rpm: ignore CMakeCache.txt in dist and relax BuildRequires on cross build
+54. gencore: Allocate ELF header to heap instead of stack
+55. nanosleep: add cpu_pause() in spinwait loop
+56. init_process: add missing initializations to proc struct
+57. rus_vm_fault: always use a packet on the stack
+58. process stack: use PAGE_SIZE in aux vector
+59. copy_user_pte: base memobj copy on range & VR_PRIVATE
+60. arm64: ptrace: Fix overwriting 1st argument with return value
+61. page fault: use cow for private device mappings
+62. reproductible builds: remove most install paths in c code
+63. page fault: clear writable bit for non-dirtying access to shared ranges
+64. mcreboot/mcstop+release: support for regular user execution
+65. irqbalance_mck: replace extra service with service drop-in
+66. do_mmap: give addr argument a chance even if not MAP_FIXED
+67. x86: fix xchg() and cmpxchg() macros
+68. IHK: support for using Linux work IRQ as IKC interrupt (optional)
+69. MCS: fix ARM64 issue by using smp_XXX() functions (i.e., barrier()s)
+70. procfs: add number of threads to stat and status
+71. memory_range_lock: Fix deadlock in procfs/sysfs handler
+72. flush instruction cache at context switch time if necessary
+73. arm64: Fix PMU related functions
+74. page_fault_process_memory_range: Disable COW for VM region with zeroobj
+75. extend_process_region: Fall back to demand paging when not contiguous
+76. munmap: fix deadlock with remote pagefault on vm range lock
+77. procfs: if memory_range_lock fails, process later
+78. migrate-cpu: Prevent migration target from calling schedule() twice
+79. sched_request_migrate(): fix race condition between migration req and IRQs
+80. get_one_cpu_topology: Renumber core_id (physical core id)
+81. bb7e140 procfs cpuinfo: use sequence number as processor
+82. set_host_vma(): do NOT read protect Linux VMA
+
+===========================================
+What's new in V1.6.0 (Nov 11, 2018)
+===========================================
+
+-----------------------------------------------
+McKernel new features, improvements and changes
+-----------------------------------------------
+1. McKernel and Linux share one unified kernel virtual address space.
+   That is, McKernel sections resides in Linux sections spared for
+   modules.  In this way, Linux can access the McKernel kernel memory
+   area.
+2. hugetlbfs support
+3. IHK is now included as a git submodule
+4. Debug messages are turned on/off in per souce file basis at run-time.
+5. It's prohibited for McKernel to access physical memory ranges which
+   Linux didn't give to McKernel.
+6. UTI (capability to spawn a thread on Linux CPU) improvement:
+   * System calls issued from the thread are hooked by modifying
+     binary in memory.
+
+---------------------------
+McKernel bug fixes (digest)
+---------------------------
+#<num> below corresponds to the redmine issue number
+(https://postpeta.pccluster.org/redmine/).
+
+1. #926: shmget: Hide object with IPC_RMID from shmget
+2. #1028: init_process: Inherit parent cpu_set
+3. #995: Fix shebang recorded in argv[0]
+4. #1024: Fix VMAP virtual address leak
+5. #1109: init_process_stack: Support "ulimit -s unlimited"
+6. x86 mem init: do not map identity mapping
+7. mcexec_wait_syscall: requeue potential request on interrupted wait
+8. mcctrl_ikc_send_wait: fix interrupt with do_frees == NULL
+9. pager_req_read: handle short read
+10. kprintf: only call eventfd() if it is safe to interrupt
+11. process_procfs_request: Add Pid to /proc/<PID>/status
+12. terminate: fix oversubscribe hang when waiting for other threads on same CPU to die
+13. mcexec: Do not close fd returned to mckernel side
+14. #976: execve: Clear sigaltstack and fp_regs
+15. #1002: perf_event: Specify counter by bit_mask on start/stop
+16. #1027: schedule: Don't reschedule immediately when wake up on migrate
+17. #mcctrl: lookup unexported symbols at runtime
+18. __sched_wakeup_thread: Notify interrupt_exit() of re-schedule
+19. futex_wait_queue_me: Spin-sleep when timeout and idle_halt is specified
+20. #1167: ihk_os_getperfevent,setperfevent: Timeout IKC sent by mcctrl
+21. devobj: fix object size (POSTK_DEBUG_TEMP_FIX_36)
+22. mcctrl: remove rus page cache
+23. #1021: procfs: Support multiple reads of e.g. /proc/*/maps
+24. #1006: wait: Delay wake-up parent within switch context
+25. #1164: mem: Check if phys-mem is within the range of McKernel memory
+26. #1039: page_fault_process_memory_range: Remove ihk_mc_map_virtual for CoW of device map
+27. partitioned execution: pass process rank to LWK
+28. process/vm: implement access_ok()
+29. spinlock: rewrite spinlock to use Linux ticket head/tail format
+30. #986: Fix deadlock involving mmap_sem and memory_range_lock
+31. Prevent one CPU from getting chosen by concurrent forks
+32. #1009: check_signal: system call restart is done only once
+33. #1176: syscall: the signal received during system call processing is not processed.
+34. #1036 syscall_time: Handle by McKernel
+35. #1165 do_syscall: Delegate system calls to the mcexec with the same pid
+36. #1194 execve: Fix calling ptrace_report_signal after preemption is disabled
+37. #1005 coredump: Exclude special areas
+38. #1018 procfs: Fix pread/pwrite to procfs fail when specified size is bigger than 4MB
+39. #1180 sched_setaffinity: Check migration after decrementing in_interrupt
+40. #771, #1179, #1143 ptrace supports threads
+41. #1189 procfs/do_fork: wait until procfs entries are registered
+42. #1114 procfs: add '/proc/pid/stat' to mckernel side and fix its comm
+43. #1116 mcctrl procfs: check entry was returned before using it
+44. #1167 ihk_os_getperfevent,setperfevent: Return -ETIME when IKC timeouts
+45. mcexec/execve: fix shebangs handling
+46. procfs: handle 'comm' on mckernel side
+47. ihk_os_setperfevent: Return number of registered events
+48. mcexec: fix terminating zero after readlink()
+
+===========================================
+What's new in V1.5.1 (July 9, 2018)
+===========================================
+
+-----------------------------------------------
+McKernel new features, improvements and changes
+-----------------------------------------------
+1. Watchdog timer to detect hang of McKernel
+   mcexec prints out the following line to its stderr when a hang of
+   McKernel is detected.
+
+       mcexec detected hang of McKernel
+
+   The watchdog timer is enabled by passing -i <timeout_in_sec> option
+   to mcreboot.sh. <timeout_in_sec> specifies the interval of checking
+   if McKernel is alive.
+   Example: mcreboot.sh -i 600: Detect the hang with 10 minutes interval
+
+   The detailed step of the hang detection is as follows.
+   (1) mcexec acquires eventfd for notification from IHK and perform
+       epoll() on it.
+   (2) A daemon called ihkmond monitors the state of McKernel periodically
+       with the interval specified by the -i option. It judges that
+       McKernel is hanging and notifies mcexec by the eventfd if its
+       state hasn't changed since the last check.
+
+2. Documentation
+   man page: Installed directory is changed to <install_dir>/share/man
+
+---------------------------
+McKernel bug fixes (digest)
+---------------------------
+1.  #1146: pager_req_map(): do not take mmap_sem if not needed
+2.  #1135: prepare_process_ranges_args_envs(): fix saving cmdline
+3.  #1144: fileobj/devobj: record path name
+4.  #1145: fileobj: use MCS locks for per-file page hash
+5.  #1076: mcctrl: refactor prepare_image into new generic ikc send&wait
+6.  #1072: execve: fix execve with oversubscribing
+7.  #1132: execve: use thread variable instead of cpu_local_var(current)
+8.  #1117: mprotect: do not set page table writable for cow pages
+9.  #1143: syscall wait4: add _WALL (POSTK_DEBUG_ARCH_DEP_44)
+10. #1064: rusage: Fix initialization of rusage->num_processors
+11. #1133: pager_req_unmap: Put per-process data at exit
+12. #731: do_fork: Propagate error code returned by mcexec
+13. #1149: execve: Reinitialize vm_regions's map area on execve
+14. #1065: procfs: Show file names in /proc/<PID>/maps
+15. #1112: mremap: Fix type of size arguments (from ssize_t to size_t)
+16. #1121: sched_getaffinity: Check arguments in the same order as in Linux
+17. #1137: mmap, mremap: Check arguments in the same order as in Linux
+18. #1122: fix return value of sched_getaffinity
+19. #732: fix: /proc/<PID>/maps outputs a unnecessary NULL character
+
+===================================
+What's new in V1.5.0 (Apr 5, 2018)
+===================================
+
+--------------------------------------
+McKernel new features and improvements
+--------------------------------------
+1. Aid for Linux version migration: Detect /proc, /sys format change
+   between two kernel verions
+2. Swap out
+   * Only swap-out anonymous pages for now
+3. Improve support of /proc/maps
+4. mcstat: Linux tool to show resource usage
+
+---------------------------
+McKernel bug fixes (digest)
+---------------------------
+1. #727: execve: Fix memory leak when receiving SIGKILL
+2. #829: perf_event_open: Support PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE
+3. #906: mcexec: Check return code of fork()
+4. #1038: mcexec: Timeout when incorrect value is given to -n option
+5. #943 #945 #946 #960 $961: mcexec: Support strace
+6. #1029: struct thread is not released with stress-test involving signal
+          and futex
+7. #863 #870: Respond immediately to terminating signal when
+              offloading system call
+8. #1119: translate_rva_to_rpa(): use 2MB blocks in 1GB pages on x86
+11. #898: Shutdown OS only after no in-flight IKC exist
+12. #882: release_handler: Destroy objects as the process which opened it
+13. #882: mcexec: Make child process exit if the parent is killed during
+          fork()
+14. #925: XPMEM: Don't destroy per-process object of the parent
+15. #885: ptrace: Support the case where a process attaches its child
+16. #1031: sigaction: Support SA_RESETHAND
+17. #923: rus_vm_fault: Return error when a thread not performing
+          system call offloading causes remote page fault
+18. #1032 #1033 #1034: getrusage: Fix ru_maxrss, RUSAGE_CHILDREN,
+                       ru_stime related bugs
+19. #1120: getrusage: Fix deadlock on thread->times_update
+20. #1123: Fix deadlock related to wait_queue_head_list_node
+21. #1124: Fix deadlock of calling terminate() from terminate()
+22. #1125: Fix deadlock related to thread status
+    * Related functions are: hold_thread(), do_kill() and terminate()
+23. #1126: uti: Fix uti thread on the McKernel side blocks others in do_syscall()
+24. #1066: procfs: Show Linux /proc/self/cgroup
+25. #1127: prepare_process_ranges_args_envs(): fix generating saved_cmdline to
+           avoid PF in strlen()
+26. #1128: ihk_mc_map/unmap_virtual(): do proper TLB invalidation
+27. #1043: terminate(): fix update_lock and threads_lock order to avoid deadlock
+28. #1129: mcreboot.sh: Save /proc/irq/*/smp_affinity to /tmp/mcreboot
+29. #1130: mcexec: drop READ_IMPLIES_EXEC from personality
+
+--------------------
+McKernel workarounds
+--------------------
+1. Forbid CPU oversubscription
+   * It can be turned on by mcreboot.sh -O option
+
+
+===================================
+What's new in V1.4.0 (Oct 30, 2017)
+===================================
+
+-----------------------------------------------------------
+Feature: Abstracted event type support in perf_event_open()
+-----------------------------------------------------------
+PERF_TYPE_HARDWARE and PERF_TYPE_CACHE types are supported.
+
+----------------------------------
+Clean-up: Direct user-space access
+----------------------------------
+Code lines using direct user-space access (e.g. passing user-space
+pointer to memcpy()) becomes more portable across processor
+architectures. The modification follows the following rules.
+
+1. Move the code section as it is to the architecture dependent
+   directory if it is a part of the critical-path.
+2. Otherwise, rewrite the code section by using the portable methods.
+   The methods include copy_from_user(), copy_to_user(),
+   pte_get_phys() and phys_to_virt().
+
+--------------------------------
+Test: MPI and OpenMP micro-bench
+--------------------------------
+The performance figures of MPI and OpenMP primitives are compared with
+those of Linux by using Intel MPI Benchmarks and EPCC OpenMP Micro
+Benchmark.
+
+
+===================================
+What's new in V1.3.0 (Sep 30, 2017)
+===================================
+
+--------------------
+Feature: Kernel dump
+--------------------
+1. A dump level of "only kernel memory" is added.
+
+The following two levels are available now:
+   0: Dump all
+  24: Dump only kernel memory
+
+The dump level can be set by -d option in ihkosctl or the argument
+for ihk_os_makedumpfile(), as shown in the following examples:
+
+   Command:		ihkosctl 0 dump -d 24
+   Function call:	ihk_os_makedumpfile(0, NULL, 24, 0);
+
+2. Dump file is created when Linux panics.
+
+The dump level can be set by dump_level kernel argument, as shown in the
+following example:
+
+   ihkosctl 0 kargs "hidos dump_level=24"
+
+The IHK dump function is registered to panic_notifier_list when creating
+/dev/mcdX and called when Linux panics.
+
+-----------------------------
+Feature: Quick Process Launch
+-----------------------------
+
+MPI process launch time and some of the initialization time can be
+reduced in application consisting of multiple MPI programs which are
+launched in turn in the job script.
+
+The following two steps should be performed to use this feature:
+1. Replace mpiexec with ql_mpiexec_start and add some lines for
+   ql_mpiexec_finalize in the job script
+2. Modify the app so that it can repeat calculations and wait for the
+   instructions from ql_mpiexec_{start,finalize} at the end of the
+   loop
+
+The first step is explained using an example. Assume the original job
+script looks like this:
+
+/* Execute ensamble simulation and then data assimilation, and repeat this
+   ten times */
+for i in {1..10}; do
+
+     /* Each ensamble simulation execution uses 100 nodes, launch ten of them
+        in parallel */
+    for j in {1..10}; do
+         mpiexec -n 100 -machinefile ./list1_$j p1.out a1 & pids[$i]=$!;
+    done
+
+    /* Wait until the ten ensamble simulation programs finish */
+    for j in {1..10}; do wait ${pids[$j]}; done
+
+    /* Launch one data assimilation program using 1000 nodes */
+    mpiexec -n 1000 -machinefile ./list2 p2.out a2
+done
+
+The job script should be modified like this:
+
+for i in {1..10}; do
+    for j in {1..10}; do
+        /*  Replace mpiexec with ql_mpiexec_start */
+        ql_mpiexec_start -n 100 -machinefile ./list1_$j p1.out a1 & pids[$j]=$!;
+    done
+
+    for j in {1..10}; do wait ${pids[$j]}; done
+
+    ql_mpiexec_start -n 1000 -machinefile ./list2 p2.out a2
+done
+
+/* p1.out and p2.out don't exit but are waiting for the next calculation.
+   So tell them to exit */
+for j in {1..10}; do
+   ql_mpiexec_finalize -machinefile ./list1_$i p1.out a1;
+done
+ql_mpiexec_finalize -machinefile ./list2 p2.out a2;
+
+
+The second step is explained using a pseudo-code.
+
+    MPI_Init();
+    Prepare data exchange with preceding / following MPI programs
+loop:
+    foreach Fortran module
+        Initialize data using command-line argments, parameter files,
+	environment variables
+    Input data from preceding MPI programs / Read snap-shot
+    Perform main calculation
+    Output data to following MPI programs / Write snap-shot
+    /* ql_client() waits for command of ql_mpiexec_{start,finish} */
+    if (ql_client() == QL_CONTINUE) { goto loop; }
+    MPI_Finalize();
+
+qlmpilib.h should be included in the code and libql{mpi,fort}.so
+should be linked to the executable file.
+
+
+========================
+Restrictions on McKernel
+========================
+
+ 1. Pseudo devices such as /dev/mem and /dev/zero are not mmap()ed
+    correctly even if the mmap() returns a success. An access of their
+    mapping receives the SIGSEGV signal.
+
+ 2. clone() supports only the following flags. All the other flags
+    cause clone() to return error or are simply ignored.
+
+    * CLONE_CHILD_CLEARTID
+    * CLONE_CHILD_SETTID
+    * CLONE_PARENT_SETTID
+    * CLONE_SETTLS
+    * CLONE_SIGHAND
+    * CLONE_VM
+
+ 3. PAPI has the following restriction.
+
+    * Number of counters a user can use at the same time is up to the
+      number of the physical counters in the processor.
+
+ 4. msync writes back only the modified pages mapped by the calling process.
+
+ 5. The following syscalls always return the ENOSYS error.
+
+    * migrate_pages()
+    * move_pages()
+    * set_robust_list()
+
+ 6. The following syscalls always return the EOPNOTSUPP error.
+
+    * arch_prctl(ARCH_SET_GS)
+    * signalfd()
+
+ 7. signalfd4() returns a fd, but signal is not notified through the
+    fd.
+
+ 8. set_rlimit sets the limit values but they are not enforced.
+
+ 9. Address randomization is not supported.
+
+10. brk() extends the heap more than requestd when -h
+    (--extend-heap-by=)<step> option of mcexec is used with the value
+    larger than 4 KiB.  syscall_pwrite02 of LTP would fail for this
+    reason. This is because the test expects that the end of the heap
+    is set to the same address as the argument of sbrk() and expects a
+    segmentation violation occurs when it tries to access the memory
+    area right next to the boundary. However, the optimization sets
+    the end to a value larger than the requested. Therefore, the
+    expected segmentation violation doesn't occur.
+
+11. setpriority()/getpriority() won't work. They might set/get the
+    priority of a random mcexec thread. This is because there's no
+    fixed correspondence between a McKernel thread which issues the
+    system call and a mcexec thread which handles the offload request.
+
+12. mbind() can set the policy but it is not used when allocating
+    physical pages.
+
+13. MPOL_F_RELATIVE_NODES and MPOL_INTERLEAVE flags for
+    set_mempolicy()/mbind() are not supported.
+
+14. The MPOL_BIND policy for set_mempolicy()/mbind() works as the same
+    as the MPOL_PREFERRED policy. That is, the physical page allocator
+    doesn't give up the allocation when the specified nodes are
+    running out of pages but continues to search pages in the other
+    nodes.
+
+15. Kernel dump on Linux panic requires Linux kernel CentOS-7.4 and
+    later. In addition, crash_kexec_post_notifiers kernel argument
+    must be given to Linux kernel.
+
+16. setfsuid()/setfsgid() cannot change the id of the calling thread.
+    Instead, it changes that of the mcexec worker thread which takes
+    the system-call offload request.
+
+17. mmap (hugeTLBfs): The physical pages corresponding to a map are
+    released when no McKernel process exist. The next map gets fresh
+    physical pages.
+
+18. Sticky bit on executable file has no effect.
+
+19. Linux (RHEL-7 for x86_64) could hang when offlining CPUs in the
+    process of booting McKernel due to the Linux bug, found in
+    Linux-3.10 and fixed in the later version.  One way to circumvent
+    this is to always assign the same CPU set to McKernel.
+
+20. madvise:
+    * MADV_HWPOISON and MADV_SOFT_OFFLINE always returns -EPERM.
+    * MADV_MERGEABLE and MADV_UNMERGEABLE always returns -EINVAL.
+    * MADV_HUGEPAGE and MADV_NOHUGEPAGE on file map returns -EINVAL
+      (It succeeds on RHEL-8 for aarch64).
+
+21. brk() and mmap() doesn't report out-of-memory through its return
+    value. Instead, page-fault reports the error.
+
+22. Anonymous mmap pre-maps requested number of pages when contiguous
+    pages are available. Demand paging is used when not available.
+
+23. Mixing page sizes in anonymous shared mapping is not allowed. mmap
+    creates vm_range with one page size. And munmap or mremap that
+    needs the reduced page size changes the sizes of all the pages of
+    the vm_range.
+
+24. ihk_os_getperfevent() could time-out when invoked from Fujitsu TCS
+    (job-scheduler).
+
+25. The behaviors of madvise and mbind are changed to do nothing and
+    report success as a workaround for Fugaku.
+
+26. mmap() allows unlimited overcommit. Note that it corresponds to
+    setting sysctl ``vm.overcommit_memory`` to 1.

README.md

@@ -10,7 +10,7 @@ IHK/McKernel is a light-weight multi-kernel operating system designed for high-e
 
 ## Contents
 
-- [Background] (#background)
+- [Background](#background-and-motivation)
 - [Architectural Overview](#architectural-overview)
 - [Installation](#installation)
 - [The Team](#the-team)
@@ -85,7 +85,7 @@ sudo reboot
 You will need the following packages installed:
 
 ~~~~
-sudo yum install kernel-devel binutils-devel libnuma-devel
+sudo yum install cmake kernel-devel binutils-devel systemd-devel numactl-devel gcc make nasm git
 ~~~~
 
 Grant read permission to the System.map file of your kernel version:
@@ -96,24 +96,51 @@ sudo chmod a+r /boot/System.map-`uname -r`
 
 ##### 4. Obtain sources and compile the kernel
 
-Clone the source code and set up ihk symlink (this is currently required):
+Clone the source code:
 
 ~~~~
 mkdir -p ~/src/ihk+mckernel/
 cd ~/src/ihk+mckernel/
-git clone -r git@github.com:RIKEN-SysSoft/mckernel.git
+git clone --recursive -b development https://github.com/RIKEN-SysSoft/mckernel.git
 ~~~~
 
+(Optional) Checkout to the specific branch or version:
+
+~~~~
+cd mckernel
+git checkout <pathspec>
+git submodule update
+~~~~
+
+Foe example, if you want to try the development branch, use "development" as the pathspec. If you want to try the prerelease version 1.7.0-0.2, use "1.7.0-0.2".
+
+###### 4.1 Install with cmake
+
 Configure and compile:
 
 ~~~~
 mkdir -p build && cd build
-cmake -DCMAKE_INSTALL_PREFIX=${HOME}/ihk+mckernel $HOME/src/mckernel
+cmake -DCMAKE_INSTALL_PREFIX=${HOME}/ihk+mckernel $HOME/src/ihk+mckernel/mckernel
 make -j install
 ~~~~
 
 The IHK kernel modules and McKernel kernel image should be installed under the **ihk+mckernel** folder in your home directory.
 
+###### 4.2 Install with rpm
+
+Configure, compile and build rpm:
+
+~~~~
+mkdir -p build && cd build
+cmake $HOME/src/ihk+mckernel/mckernel
+make dist
+cp mckernel-<version>.tar.gz <rpmbuild>/SOURCES
+rpm -ba scripts/mckernel.spec
+sudo rpm -ivh <rpmbuild>/RPMS/<arch>/mckernel-<version>-<release>_<linux_kernel_ver>_<dist>.<arch>.rpm
+~~~~
+
+The IHK kernel modules and McKernel kernel image are installed under the system directory.
+
 ##### 5. Boot McKernel
 
 A boot script called mcreboot.sh is provided under sbin in the install folder. To boot on logical CPU 1 with 512MB of memory, use the following invocation:
@@ -170,6 +197,71 @@ Finally, to shutdown McKernel and release CPU/memory resources back to Linux use
 sudo ./sbin/mcstop+release.sh
 ~~~~
 
+##### 7. Advanced: Enable Utility Thread offloading Interface (UTI)
+
+UTI enables a runtime such as MPI runtime to spawn utility threads such as MPI asynchronous progress threads to Linux cores.
+
+1. Install capstone
+
+Install EPEL capstone-devel:
+
+~~~~
+sudo yum install epel-release
+sudo yum install capstone-devel
+~~~~
+
+2. Install syscall_intercept
+
+~~~~
+git clone https://github.com/RIKEN-SysSoft/syscall_intercept.git
+cmake ../arch/aarch64 -DCMAKE_INSTALL_PREFIX=<syscall-intercept-install> -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=gcc -DTREAT_WARNINGS_AS_ERRORS=OFF
+~~~~
+
+3. Install UTI for McKernel
+
+Install:
+
+~~~~
+git clone https://github.com/RIKEN-SysSoft/uti.git
+mkdir build && cd build
+../uti/configure --prefix=<mckernel-install> --with-rm=mckernel
+make && make install
+~~~~
+
+4. Install McKernel
+
+~~~~
+CMAKE_PREFIX_PATH=<syscall-intercept-install> cmake -DCMAKE_INSTALL_PREFIX=${HOME}/ihk+mckernel -DENABLE_UTI=ON $HOME/src/ihk+mckernel/mckernel
+~~~~
+
+5. Run executable
+
+~~~~
+mcexec --enable-uti <command>
+~~~~
+
+6. Install UTI for Linux for performance comparison
+
+Install by make:
+
+~~~~
+git clone https://github.com/RIKEN-SysSoft/uti.git
+mkdir build && cd build
+../uti/configure --prefix=<uti-install> --with-rm=linux
+make && make install
+~~~~
+
+Install by rpm:
+
+~~~~
+git clone https://github.com/RIKEN-SysSoft/uti.git
+mkdir build && cd build
+../uti/configure --prefix=<uti-install> --with-rm=linux
+rm -f ~/rpmbuild/SOURCES/<version>.tar.gz
+rpmbuild -ba ./scripts/uti.spec
+rpm -Uvh uti-<version>-<release>-<arch>.rpm
+~~~~
+
 ## The Team
 
 The McKernel project was started at The University of Tokyo and currently it is mainly developed at RIKEN.
@@ -184,3 +276,10 @@ Some of our collaborators include:
 ## License
 
 McKernel is GPL licensed, as found in the LICENSE file.
+
+## Contact
+
+Please give your feedback to us via one of the following mailing lists. Subscription via [www.pccluster.org](http://www.pccluster.org/mailman/listinfo/mckernel-users) is needed.
+
+* English: mckernel-users@pccluster.org
+* Japanese: mckernel-users-jp@pccluster.org

arch/arm64/kernel/assert.c

@@ -1,4 +1,4 @@
-/* assert.c COPYRIGHT FUJITSU LIMITED 2015-2018 */
+/* assert.c COPYRIGHT FUJITSU LIMITED 2015-2019 */
 
 #include <process.h>
 #include <list.h>
@@ -53,4 +53,4 @@ STATIC_ASSERT(SVE_PT_FPSIMD_OFFSET == sizeof(struct user_sve_header));
 STATIC_ASSERT(SVE_PT_SVE_OFFSET == sizeof(struct user_sve_header));
 
 /* assert for struct arm64_cpu_local_thread member offset define */
-STATIC_ASSERT(offsetof(struct arm64_cpu_local_thread, panic_regs) == 160);
+STATIC_ASSERT(offsetof(struct arm64_cpu_local_thread, panic_regs) == 168);

arch/arm64/kernel/coredump.c

@@ -1,9 +1,15 @@
-/* coredump.c COPYRIGHT FUJITSU LIMITED 2015-2016 */
+/* coredump.c COPYRIGHT FUJITSU LIMITED 2015-2019 */
 #include <process.h>
 #include <elfcore.h>
 #include <string.h>
+#include <ptrace.h>
+#include <cls.h>
+#include <hwcap.h>
 
-void arch_fill_prstatus(struct elf_prstatus64 *prstatus, struct thread *thread, void *regs0)
+#define	align32(x) ((((x) + 3) / 4) * 4)
+
+void arch_fill_prstatus(struct elf_prstatus64 *prstatus,
+	struct thread *thread, void *regs0, int sig)
 {
 	struct pt_regs *regs = regs0;
 	struct elf_prstatus64 tmp_prstatus;
@@ -14,8 +20,6 @@ void arch_fill_prstatus(struct elf_prstatus64 *prstatus, struct thread *thread,
 	short int pr_cursig;
 	a8_uint64_t pr_sigpend;
 	a8_uint64_t pr_sighold;
-	pid_t pr_pid;
-	pid_t pr_ppid;
 	pid_t pr_pgrp;
 	pid_t pr_sid;
 	struct prstatus64_timeval pr_utime;
@@ -23,10 +27,66 @@ void arch_fill_prstatus(struct elf_prstatus64 *prstatus, struct thread *thread,
 	struct prstatus64_timeval pr_cutime;
 	struct prstatus64_timeval pr_cstime;
  */
+
 	/* copy x0-30, sp, pc, pstate */
 	memcpy(&tmp_prstatus.pr_reg, &regs->user_regs, sizeof(tmp_prstatus.pr_reg));
 	tmp_prstatus.pr_fpvalid = 0;	/* We assume no fp */
 
 	/* copy unaligned prstatus addr */
 	memcpy(prstatus, &tmp_prstatus, sizeof(*prstatus));
+
+	prstatus->pr_pid = thread->tid;
+	if (thread->proc->parent) {
+		prstatus->pr_ppid = thread->proc->parent->pid;
+	}
+
+	prstatus->pr_info.si_signo = sig;
+	prstatus->pr_cursig = sig;
+}
+
+int arch_get_thread_core_info_size(void)
+{
+	const struct user_regset_view *view = current_user_regset_view();
+	const struct user_regset *regset = find_regset(view, NT_ARM_SVE);
+
+	if (unlikely(!(elf_hwcap & HWCAP_SVE))) {
+		return 0;
+	}
+	return sizeof(struct note) + align32(sizeof("LINUX"))
+		+ regset_size(cpu_local_var(current), regset);
+}
+
+void arch_fill_thread_core_info(struct note *head,
+				struct thread *thread, void *regs)
+{
+	const struct user_regset_view *view = current_user_regset_view();
+	const struct user_regset *regset = find_regset(view, NT_ARM_SVE);
+
+	if (unlikely(!(elf_hwcap & HWCAP_SVE))) {
+		return;
+	}
+
+	/* pre saved registers */
+	save_fp_regs(thread);
+
+	if (regset->core_note_type && regset->get &&
+	    (!regset->active || regset->active(thread, regset))) {
+		int ret;
+		size_t size = regset_size(thread, regset);
+		void *namep;
+		void *descp;
+
+		namep = (void *) (head + 1);
+		descp = namep + align32(sizeof("LINUX"));
+
+		ret = regset->get(thread, regset, 0, size, descp, NULL);
+		if (ret) {
+			return;
+		}
+
+		head->namesz = sizeof("LINUX");
+		head->descsz = size;
+		head->type = NT_ARM_SVE;
+		memcpy(namep, "LINUX", sizeof("LINUX"));
+	}
 }

arch/arm64/kernel/cpu.c

@@ -1,6 +1,5 @@
-/* cpu.c COPYRIGHT FUJITSU LIMITED 2015-2018 */
+/* cpu.c COPYRIGHT FUJITSU LIMITED 2015-2019 */
 #include <ihk/cpu.h>
-#include <ihk/debug.h>
 #include <ihk/mm.h>
 #include <types.h>
 #include <errno.h>
@@ -30,9 +29,11 @@
 #include <debug-monitors.h>
 #include <sysreg.h>
 #include <cpufeature.h>
-#include <debug.h>
+#include <ihk/debug.h>
 #include <hwcap.h>
 #include <virt.h>
+#include <init.h>
+#include <bootparam.h>
 
 //#define DEBUG_PRINT_CPU
 
@@ -67,6 +68,7 @@ void (*gic_dist_init)(unsigned long dist_base_pa, unsigned long size);
 void (*gic_cpu_init)(unsigned long cpu_base_pa, unsigned long size);
 void (*gic_enable)(void);
 void (*arm64_issue_ipi)(unsigned int cpid, unsigned int vector);
+void (*arm64_issue_host_ipi)(unsigned int cpid, unsigned int vector);
 void (*handle_arch_irq)(struct pt_regs *);
 
 static void gic_init(void)
@@ -77,14 +79,18 @@ static void gic_init(void)
 		gic_cpu_init = gic_cpu_init_gicv3;
 		gic_enable = gic_enable_gicv3;
 		arm64_issue_ipi = arm64_issue_ipi_gicv3;
+		arm64_issue_host_ipi = arm64_issue_host_ipi_gicv3;
 		handle_arch_irq = handle_interrupt_gicv3;
+		kprintf("%: GICv3\n", __func__);
 	} else {
 		/* Setup functions for GICv2 */
 		gic_dist_init = gic_dist_init_gicv2;
 		gic_cpu_init = gic_cpu_init_gicv2;
 		gic_enable = gic_enable_gicv2;
 		arm64_issue_ipi = arm64_issue_ipi_gicv2;
+		arm64_issue_host_ipi = arm64_issue_host_ipi_gicv2;
 		handle_arch_irq = handle_interrupt_gicv2;
+		kprintf("%: GICv2\n", __func__);
 	}
 
 	gic_dist_init(ihk_param_gic_dist_base_pa, ihk_param_gic_dist_map_size);
@@ -114,42 +120,94 @@ static struct ihk_mc_interrupt_handler cpu_stop_handler = {
 };
 
 extern long freeze_thaw(void *nmi_ctx);
-static void multi_nm_interrupt_handler(void *priv)
+static void multi_interrupt_handler(void *priv)
 {
-	extern int nmi_mode;
-	struct pt_regs *regs = (struct pt_regs *)priv;
-	union arm64_cpu_local_variables *clv;
-
-	switch (nmi_mode) {
+	switch (multi_intr_mode) {
 	case 1:
-	case 2:
-		/* mode == 1or2, for FREEZER NMI */
-		dkprintf("%s: freeze mode NMI catch. (nmi_mode=%d)\n",
-			 __func__, nmi_mode);
+	case 2: /* mode == 1or2, for FREEZER intr */
+		dkprintf("%s: freeze mode intr catch. (multi_intr_mode=%d)\n",
+			 __func__, multi_intr_mode);
 		freeze_thaw(NULL);
 		break;
+	default:
+		ekprintf("%s: Unknown multi-intr-mode(%d) detected.\n",
+			 __func__, multi_intr_mode);
+		break;
+	}
+}
 
+void arch_save_panic_regs(void *irq_regs)
+{
+	struct pt_regs *regs = (struct pt_regs *)irq_regs;
+	union arm64_cpu_local_variables *clv;
+
+	clv = get_arm64_this_cpu_local();
+
+	/* For user-space, use saved kernel context */
+	if (regs->pc < USER_END) {
+		memset(clv->arm64_cpu_local_thread.panic_regs,
+				0, sizeof(clv->arm64_cpu_local_thread.panic_regs));
+		clv->arm64_cpu_local_thread.panic_regs[29] =
+			current_thread_info()->cpu_context.fp;
+		clv->arm64_cpu_local_thread.panic_regs[31] =
+			current_thread_info()->cpu_context.sp;
+		clv->arm64_cpu_local_thread.panic_regs[32] =
+			current_thread_info()->cpu_context.pc;
+		clv->arm64_cpu_local_thread.panic_regs[33] =
+			PSR_MODE_EL1h;
+	}
+	else {
+		memcpy(clv->arm64_cpu_local_thread.panic_regs,
+				regs->regs, sizeof(regs->regs));
+		clv->arm64_cpu_local_thread.panic_regs[31] = regs->sp;
+		clv->arm64_cpu_local_thread.panic_regs[32] = regs->pc;
+		clv->arm64_cpu_local_thread.panic_regs[33] =
+			regs->pstate;
+
+	}
+
+	clv->arm64_cpu_local_thread.paniced = 1;
+}
+
+void arch_clear_panic(void)
+{
+	union arm64_cpu_local_variables *clv;
+
+	clv = get_arm64_this_cpu_local();
+	clv->arm64_cpu_local_thread.paniced = 0;
+}
+
+static struct ihk_mc_interrupt_handler multi_intr_handler = {
+	.func = multi_interrupt_handler,
+	.priv = NULL,
+};
+
+static void multi_nm_interrupt_handler(void *irq_regs)
+{
+	extern int nmi_mode;
+
+	dkprintf("%s: ...\n", __func__);
+	switch (nmi_mode) {
 	case 0:
-		/* mode == 0,    for MEMDUMP NMI */
-		clv = get_arm64_this_cpu_local();
-
-		if (regs) {
-			memcpy(clv->arm64_cpu_local_thread.panic_regs,
-			       regs->regs, sizeof(regs->regs));
-			clv->arm64_cpu_local_thread.panic_regs[31] = regs->sp;
-			clv->arm64_cpu_local_thread.panic_regs[32] = regs->pc;
-			clv->arm64_cpu_local_thread.panic_regs[33] =
-				regs->pstate;
-		}
-		clv->arm64_cpu_local_thread.paniced = 1;
+		/* mode == 0, for MEMDUMP NMI */
+		arch_save_panic_regs(irq_regs);
 		ihk_mc_query_mem_areas();
 		/* memdump-nmi is halted McKernel, break is unnecessary. */
 		/* fall through */
 	case 3:
-		/* mode == 3,    for SHUTDOWN-WAIT NMI */
-		while (1) {
+		/* mode == 3, for SHUTDOWN-WAIT NMI */
+		kprintf("%s: STOP\n", __func__);
+		while (nmi_mode != 4)
 			cpu_halt();
+		break;
+
+	case 4:
+		/* mode == 4, continue NMI */
+		arch_clear_panic();
+		if (!ihk_mc_get_processor_id()) {
+			ihk_mc_clear_dump_page_completion();
 		}
+		kprintf("%s: RESUME, nmi_mode: %d\n", __func__, nmi_mode);
 		break;
 
 	default:
@@ -423,6 +481,8 @@ void ihk_mc_init_ap(void)
 
 	ihk_mc_register_interrupt_handler(INTRID_CPU_STOP, &cpu_stop_handler);
 	ihk_mc_register_interrupt_handler(INTRID_MULTI_NMI, &multi_nmi_handler);
+	ihk_mc_register_interrupt_handler(INTRID_MULTI_INTR,
+					  &multi_intr_handler);
 	ihk_mc_register_interrupt_handler(
 		ihk_mc_get_vector(IHK_TLB_FLUSH_IRQ_VECTOR_START),
 		&remote_tlb_flush_handler);
@@ -776,6 +836,21 @@ unsigned long cpu_disable_interrupt_save(void)
 	return flags;
 }
 
+/* save ICC_PMR_EL1 & enable interrupt (ICC_PMR_EL1 <= ICC_PMR_EL1_UNMASKED) */
+unsigned long cpu_enable_interrupt_save(void)
+{
+	unsigned long flags;
+	unsigned long masked = ICC_PMR_EL1_UNMASKED;
+
+	asm volatile(
+		"mrs_s  %0, " __stringify(ICC_PMR_EL1) "\n"
+		"msr_s  " __stringify(ICC_PMR_EL1) ",%1"
+		: "=&r" (flags)
+		: "r" (masked)
+		: "memory");
+	return flags;
+}
+
 #else /* defined(CONFIG_HAS_NMI) */
 
 /* @ref.impl arch/arm64/include/asm/irqflags.h::arch_local_irq_enable */
@@ -824,6 +899,20 @@ unsigned long cpu_disable_interrupt_save(void)
 		: "memory");
 	return flags;
 }
+
+/* save PSTATE.DAIF & enable interrupt (PSTATE.DAIF I bit set) */
+unsigned long cpu_enable_interrupt_save(void)
+{
+	unsigned long flags;
+
+	asm volatile(
+		"mrs    %0, daif	// arch_local_irq_save\n"
+		"msr    daifclr, #2"
+		: "=r" (flags)
+		:
+		: "memory");
+	return flags;
+}
 #endif /* defined(CONFIG_HAS_NMI) */
 
 /* we not have "pause" instruction, instead "yield" instruction */
@@ -951,7 +1040,7 @@ void ihk_mc_boot_cpu(int cpuid, unsigned long pc)
 		setup_cpu_features();
 	}
 
-	init_sve_vl();
+	sve_setup();
 }
 
 /* for ihk_mc_init_context() */
@@ -986,6 +1075,9 @@ void ihk_mc_init_context(ihk_mc_kernel_context_t *new_ctx,
 		/* branch in ret_from_fork */
 		new_ctx->thread->cpu_context.x19 = (unsigned long)next_function;
 
+		sp -= 16;
+		new_ctx->thread->cpu_context.fp = sp;
+
 		/* set stack_pointer */
 		new_ctx->thread->cpu_context.sp = sp - sizeof(ihk_mc_user_context_t);
 
@@ -1001,9 +1093,10 @@ void ihk_mc_init_context(ihk_mc_kernel_context_t *new_ctx,
 		const int lcpuid = ihk_mc_get_processor_id();
 		const unsigned long syscallno = current_pt_regs()->syscallno;
 #ifdef CONFIG_ARM64_SVE
-		const uint16_t orig_sve_vl = current_thread_info()->sve_vl;
-		const uint16_t orig_sve_vl_onexec = current_thread_info()->sve_vl_onexec;
-		const uint16_t orig_sve_flags = current_thread_info()->sve_flags;
+		struct thread_info *ti = current_thread_info();
+		const unsigned int orig_sve_vl = ti->sve_vl;
+		const unsigned int orig_sve_vl_onexec = ti->sve_vl_onexec;
+		const unsigned long orig_sve_flags = ti->sve_flags;
 #endif /* CONFIG_ARM64_SVE */
 
 		/* get kernel stack address */
@@ -1023,6 +1116,9 @@ void ihk_mc_init_context(ihk_mc_kernel_context_t *new_ctx,
 
 		/* set stack_pointer */
 		new_ctx->thread->cpu_context.sp = sp;
+		/* use the 16 bytes padding in ihk_mc_init_user_process()
+		 * as closing frame in the frame chain */
+		new_ctx->thread->cpu_context.fp = sp + sizeof(ihk_mc_user_context_t);
 
 		/* clear pt_regs area */
 		new_uctx = (ihk_mc_user_context_t *)new_ctx->thread->cpu_context.sp;
@@ -1183,7 +1279,7 @@ long ihk_mc_show_cpuinfo(char *buf, size_t buf_size, unsigned long read_off, int
 
 		/* generate strings */
 		loff += scnprintf(lbuf + loff, lbuf_size - loff,
-				  "processor\t: %d\n", cpuinfo->hwid);
+				  "processor\t: %d\n", i);
 		loff += scnprintf(lbuf + loff, lbuf_size - loff, "Features\t:");
 
 		for (j = 0; hwcap_str[j]; j++) {
@@ -1234,7 +1330,6 @@ err:
 }
 
 static int check_and_allocate_fp_regs(struct thread *thread);
-void save_fp_regs(struct thread *thread);
 
 void arch_clone_thread(struct thread *othread, unsigned long pc,
 			unsigned long sp, struct thread *nthread)
@@ -1346,11 +1441,15 @@ int ihk_mc_arch_get_special_register(enum ihk_asr_type type,
 }
 
 /*@
-  @ requires \valid_apicid(cpu);	// valid APIC ID or not
+  @ requires \valid_cpuid(cpu);     // valid CPU logical ID
   @ ensures \result == 0
   @*/
 int ihk_mc_interrupt_cpu(int cpu, int vector)
 {
+	if (cpu < 0 || cpu >= num_processors) {
+		kprintf("%s: invalid CPU id: %d\n", __func__, cpu);
+		return -1;
+	}
 	dkprintf("[%d] ihk_mc_interrupt_cpu: %d\n", ihk_mc_get_processor_id(), cpu);
 	(*arm64_issue_ipi)(cpu, vector);
 	return 0;
@@ -1398,6 +1497,19 @@ struct thread *arch_switch_context(struct thread *prev, struct thread *next)
 		}
 	}
 #endif /*ENABLE_PERF*/
+
+#ifdef PROFILE_ENABLE
+	if (prev && prev->profile && prev->profile_start_ts != 0) {
+		prev->profile_elapsed_ts +=
+			(rdtsc() - prev->profile_start_ts);
+		prev->profile_start_ts = 0;
+	}
+
+	if (next->profile && next->profile_start_ts == 0) {
+		next->profile_start_ts = rdtsc();
+	}
+#endif
+
 	if (likely(prev)) {
 		tls_thread_switch(prev, next);
 
@@ -1471,8 +1583,7 @@ check_and_allocate_fp_regs(struct thread *thread)
 
 		if (!thread->fp_regs) {
 			kprintf("error: allocating fp_regs pages\n");
-			result = 1;
-			panic("panic: error allocating fp_regs pages");
+			result = -ENOMEM;
 			goto out;
 		}
 
@@ -1481,37 +1592,51 @@ check_and_allocate_fp_regs(struct thread *thread)
 
 #ifdef CONFIG_ARM64_SVE
 	if (likely(elf_hwcap & HWCAP_SVE)) {
-		sve_alloc(thread);
+		result = sve_alloc(thread);
 	}
 #endif /* CONFIG_ARM64_SVE */
 out:
+	if (result) {
+		release_fp_regs(thread);
+	}
 	return result;
 }
 
 /*@
   @ requires \valid(thread);
   @*/
-void
+int
 save_fp_regs(struct thread *thread)
 {
+	int ret = 0;
 	if (thread == &cpu_local_var(idle)) {
-		return;
+		goto out;
 	}
 
 	if (likely(elf_hwcap & (HWCAP_FP | HWCAP_ASIMD))) {
-		if (check_and_allocate_fp_regs(thread) != 0) {
-			// alloc error.
-			return;
+		ret = check_and_allocate_fp_regs(thread);
+		if (ret) {
+			goto out;
 		}
 		thread_fpsimd_save(thread);
 	}
+out:
+	return ret;
 }
 
-void copy_fp_regs(struct thread *from, struct thread *to)
+int copy_fp_regs(struct thread *from, struct thread *to)
 {
-	if ((from->fp_regs != NULL) && (check_and_allocate_fp_regs(to) == 0)) {
-		memcpy(to->fp_regs, from->fp_regs, sizeof(fp_regs_struct));
+	int ret = 0;
+
+	if (from->fp_regs != NULL) {
+		ret = check_and_allocate_fp_regs(to);
+		if (!ret) {
+			memcpy(to->fp_regs,
+			       from->fp_regs,
+			       sizeof(fp_regs_struct));
+		}
 	}
+	return ret;
 }
 
 void clear_fp_regs(void)
@@ -1626,6 +1751,7 @@ static inline int arch_cpu_mrs(uint32_t sys_reg, uint64_t *val)
 		SYSREG_READ_S(IMP_PF_INJECTION_DISTANCE5_EL0);
 		SYSREG_READ_S(IMP_PF_INJECTION_DISTANCE6_EL0);
 		SYSREG_READ_S(IMP_PF_INJECTION_DISTANCE7_EL0);
+		SYSREG_READ_S(IMP_PF_PMUSERENR_EL0);
 		SYSREG_READ_S(IMP_BARRIER_CTRL_EL1);
 		SYSREG_READ_S(IMP_BARRIER_BST_BIT_EL1);
 		SYSREG_READ_S(IMP_BARRIER_INIT_SYNC_BB0_EL1);
@@ -1696,6 +1822,7 @@ static inline int arch_cpu_msr(uint32_t sys_reg, uint64_t val)
 	SYSREG_WRITE_S(IMP_PF_INJECTION_DISTANCE5_EL0);
 	SYSREG_WRITE_S(IMP_PF_INJECTION_DISTANCE6_EL0);
 	SYSREG_WRITE_S(IMP_PF_INJECTION_DISTANCE7_EL0);
+	SYSREG_WRITE_S(IMP_PF_PMUSERENR_EL0);
 	SYSREG_WRITE_S(IMP_BARRIER_CTRL_EL1);
 	SYSREG_WRITE_S(IMP_BARRIER_BST_BIT_EL1);
 	SYSREG_WRITE_S(IMP_BARRIER_INIT_SYNC_BB0_EL1);
@@ -1753,6 +1880,11 @@ int arch_cpu_read_write_register(
 		ret = -1;
 	}
 
+	dkprintf("%s: MCCTRL_OS_CPU_%s_REGISTER: reg: 0x%lx, val: 0x%lx\n",
+			__FUNCTION__,
+			(op == MCCTRL_OS_CPU_READ_REGISTER ? "READ" : "WRITE"),
+			desc->addr, desc->val);
+
 	return ret;
 }
 
@@ -1762,4 +1894,9 @@ int smp_call_func(cpu_set_t *__cpu_set, smp_func_t __func, void *__arg)
 	return -1;
 }
 
+void arch_flush_icache_all(void)
+{
+	asm("ic	ialluis");
+	dsb(ish);
+}
 /*** end of file ***/

arch/arm64/kernel/cpufeature.c

@@ -970,7 +970,7 @@ static const struct arm64_cpu_capabilities arm64_elf_hwcaps[] = {
 #ifdef CONFIG_ARM64_SVE
 	HWCAP_CAP(SYS_ID_AA64PFR0_EL1, ID_AA64PFR0_SVE_SHIFT, FTR_UNSIGNED, 1, CAP_HWCAP, HWCAP_SVE),
 #endif
-	{},
+	{ 0 },
 };
 
 /* @ref.impl arch/arm64/kernel/cpufeature.c */

arch/arm64/kernel/cputable.c

@@ -10,5 +10,5 @@ struct cpu_info cpu_table[] = {
 		.cpu_name	= "AArch64 Processor",
 		.cpu_setup	= __cpu_setup,
 	},
-	{ /* Empty */ },
+	{ 0 },
 };

arch/arm64/kernel/debug-monitors.c

@@ -2,7 +2,6 @@
 #include <cputype.h>
 #include <irqflags.h>
 #include <ihk/context.h>
-#include <ihk/debug.h>
 #include <signal.h>
 #include <errno.h>
 #include <debug-monitors.h>

arch/arm64/kernel/fpsimd.c

@@ -1,4 +1,4 @@
-/* fpsimd.c COPYRIGHT FUJITSU LIMITED 2016-2018 */
+/* fpsimd.c COPYRIGHT FUJITSU LIMITED 2016-2019 */
 #include <thread_info.h>
 #include <fpsimd.h>
 #include <cpuinfo.h>
@@ -9,8 +9,9 @@
 #include <prctl.h>
 #include <cpufeature.h>
 #include <kmalloc.h>
-#include <debug.h>
+#include <ihk/debug.h>
 #include <process.h>
+#include <bitmap.h>
 
 //#define DEBUG_PRINT_FPSIMD
 
@@ -21,11 +22,87 @@
 
 #ifdef CONFIG_ARM64_SVE
 
+/* Set of available vector lengths, as vq_to_bit(vq): */
+static DECLARE_BITMAP(sve_vq_map, SVE_VQ_MAX);
+
 /* Maximum supported vector length across all CPUs (initially poisoned) */
 int sve_max_vl = -1;
+
 /* Default VL for tasks that don't set it explicitly: */
 int sve_default_vl = -1;
 
+/*
+ * Helpers to translate bit indices in sve_vq_map to VQ values (and
+ * vice versa).  This allows find_next_bit() to be used to find the
+ * _maximum_ VQ not exceeding a certain value.
+ */
+
+static unsigned int vq_to_bit(unsigned int vq)
+{
+	return SVE_VQ_MAX - vq;
+}
+
+static unsigned int bit_to_vq(unsigned int bit)
+{
+	if (bit >= SVE_VQ_MAX) {
+		bit = SVE_VQ_MAX - 1;
+	}
+	return SVE_VQ_MAX - bit;
+}
+
+/*
+ * All vector length selection from userspace comes through here.
+ * We're on a slow path, so some sanity-checks are included.
+ * If things go wrong there's a bug somewhere, but try to fall back to a
+ * safe choice.
+ */
+static unsigned int find_supported_vector_length(unsigned int vl)
+{
+	int bit;
+	int max_vl = sve_max_vl;
+
+	if (!sve_vl_valid(vl)) {
+		vl = SVE_VL_MIN;
+	}
+
+	if (!sve_vl_valid(max_vl)) {
+		max_vl = SVE_VL_MIN;
+	}
+
+	if (vl > max_vl) {
+		vl = max_vl;
+	}
+
+	bit = find_next_bit(sve_vq_map, SVE_VQ_MAX,
+			    vq_to_bit(sve_vq_from_vl(vl)));
+	return sve_vl_from_vq(bit_to_vq(bit));
+}
+
+static void sve_probe_vqs(DECLARE_BITMAP(map, SVE_VQ_MAX))
+{
+	unsigned int vq, vl;
+	unsigned long zcr;
+
+	bitmap_zero(map, SVE_VQ_MAX);
+
+	zcr = ZCR_EL1_LEN_MASK;
+	zcr = read_sysreg_s(SYS_ZCR_EL1) & ~zcr;
+
+	for (vq = SVE_VQ_MAX; vq >= SVE_VQ_MIN; --vq) {
+		/* self-syncing */
+		write_sysreg_s(zcr | (vq - 1), SYS_ZCR_EL1);
+		vl = sve_get_vl();
+		/* skip intervening lengths */
+		vq = sve_vq_from_vl(vl);
+		set_bit(vq_to_bit(vq), map);
+	}
+}
+
+void sve_init_vq_map(void)
+{
+	sve_probe_vqs(sve_vq_map);
+}
+
 size_t sve_state_size(struct thread const *thread)
 {
 	unsigned int vl = thread->ctx.thread->sve_vl;
@@ -42,17 +119,19 @@ void sve_free(struct thread *thread)
 	}
 }
 
-void sve_alloc(struct thread *thread)
+int sve_alloc(struct thread *thread)
 {
 	if (thread->ctx.thread->sve_state) {
-		return;
+		return 0;
 	}
 
 	thread->ctx.thread->sve_state =
 		kmalloc(sve_state_size(thread), IHK_MC_AP_NOWAIT);
-	BUG_ON(!thread->ctx.thread->sve_state);
-
+	if (thread->ctx.thread->sve_state == NULL) {
+		return -ENOMEM;
+	}
 	memset(thread->ctx.thread->sve_state, 0, sve_state_size(thread));
+	return 0;
 }
 
 static int get_nr_threads(struct process *proc)
@@ -75,19 +154,7 @@ int sve_set_vector_length(struct thread *thread,
 {
 	struct thread_info *ti = thread->ctx.thread;
 
-	BUG_ON(thread == cpu_local_var(current) && cpu_local_var(no_preempt) == 0);
-
-	/*
-	 * To avoid accidents, forbid setting for individual threads of a
-	 * multithreaded process.  User code that knows what it's doing can
-	 * pass PR_SVE_SET_VL_THREAD to override this restriction:
-	 */
-	if (!(flags & PR_SVE_SET_VL_THREAD) && get_nr_threads(thread->proc) != 1) {
-		return -EINVAL;
-	}
-	flags &= ~(unsigned long)PR_SVE_SET_VL_THREAD;
-
-	if (flags & ~(unsigned long)(PR_SVE_SET_VL_INHERIT |
+	if (flags & ~(unsigned long)(PR_SVE_VL_INHERIT |
 				     PR_SVE_SET_VL_ONEXEC)) {
 		return -EINVAL;
 	}
@@ -96,13 +163,19 @@ int sve_set_vector_length(struct thread *thread,
 		return -EINVAL;
 	}
 
-	if (vl > sve_max_vl) {
-		BUG_ON(!sve_vl_valid(sve_max_vl));
-		vl = sve_max_vl;
+	/*
+	 * Clamp to the maximum vector length that VL-agnostic SVE code can
+	 * work with.  A flag may be assigned in the future to allow setting
+	 * of larger vector lengths without confusing older software.
+	 */
+	if (vl > SVE_VL_ARCH_MAX) {
+		vl = SVE_VL_ARCH_MAX;
 	}
 
-	if (flags & (PR_SVE_SET_VL_ONEXEC |
-		     PR_SVE_SET_VL_INHERIT)) {
+	vl = find_supported_vector_length(vl);
+
+	if (flags & (PR_SVE_VL_INHERIT |
+		     PR_SVE_SET_VL_ONEXEC)) {
 		ti->sve_vl_onexec = vl;
 	} else {
 		/* Reset VL to system default on next exec: */
@@ -114,39 +187,42 @@ int sve_set_vector_length(struct thread *thread,
 		goto out;
 	}
 
-	if (vl != ti->sve_vl) {
-		if ((elf_hwcap & HWCAP_SVE)) {
-			fp_regs_struct fp_regs;
-			memset(&fp_regs, 0, sizeof(fp_regs));
+	if (vl == ti->sve_vl) {
+		goto out;
+	}
 
-			/* for self at prctl syscall */
-			if (thread == cpu_local_var(current)) {
-				save_fp_regs(thread);
-				clear_fp_regs();
-				thread_sve_to_fpsimd(thread, &fp_regs);
-				sve_free(thread);
+	if ((elf_hwcap & HWCAP_SVE)) {
+		fp_regs_struct fp_regs;
 
-				ti->sve_vl = vl;
+		memset(&fp_regs, 0, sizeof(fp_regs));
 
-				sve_alloc(thread);
-				thread_fpsimd_to_sve(thread, &fp_regs);
-				restore_fp_regs(thread);
-			/* for target thread at ptrace */
-			} else {
-				thread_sve_to_fpsimd(thread, &fp_regs);
-				sve_free(thread);
+		/* for self at prctl syscall */
+		if (thread == cpu_local_var(current)) {
+			save_fp_regs(thread);
+			clear_fp_regs();
+			thread_sve_to_fpsimd(thread, &fp_regs);
+			sve_free(thread);
 
-				ti->sve_vl = vl;
+			ti->sve_vl = vl;
 
-				sve_alloc(thread);
-				thread_fpsimd_to_sve(thread, &fp_regs);
-			}
+			sve_alloc(thread);
+			thread_fpsimd_to_sve(thread, &fp_regs);
+			restore_fp_regs(thread);
+		/* for target thread at ptrace */
+		} else {
+			thread_sve_to_fpsimd(thread, &fp_regs);
+			sve_free(thread);
+
+			ti->sve_vl = vl;
+
+			sve_alloc(thread);
+			thread_fpsimd_to_sve(thread, &fp_regs);
 		}
 	}
 	ti->sve_vl = vl;
 
 out:
-	ti->sve_flags = flags & PR_SVE_SET_VL_INHERIT;
+	ti->sve_flags = flags & PR_SVE_VL_INHERIT;
 
 	return 0;
 }
@@ -156,44 +232,53 @@ out:
  * Encode the current vector length and flags for return.
  * This is only required for prctl(): ptrace has separate fields
  */
-static int sve_prctl_status(const struct thread_info *ti)
+static int sve_prctl_status(unsigned long flags)
 {
-	int ret = ti->sve_vl;
+	int ret;
+	struct thread_info *ti = cpu_local_var(current)->ctx.thread;
 
-	ret |= ti->sve_flags << 16;
+	if (flags & PR_SVE_SET_VL_ONEXEC) {
+		ret = ti->sve_vl_onexec;
+	}
+	else {
+		ret = ti->sve_vl;
+	}
 
+	if (ti->sve_flags & PR_SVE_VL_INHERIT) {
+		ret |= PR_SVE_VL_INHERIT;
+	}
 	return ret;
 }
 
 /* @ref.impl arch/arm64/kernel/fpsimd.c::sve_set_task_vl */
-int sve_set_thread_vl(struct thread *thread, const unsigned long vector_length,
-			const unsigned long flags)
+int sve_set_thread_vl(unsigned long arg)
 {
+	unsigned long vl, flags;
 	int ret;
 
-	if (!(elf_hwcap & HWCAP_SVE)) {
+	vl = arg & PR_SVE_VL_LEN_MASK;
+	flags = arg & ~vl;
+
+	/* Instead of system_supports_sve() */
+	if (unlikely(!(elf_hwcap & HWCAP_SVE))) {
 		return -EINVAL;
 	}
 
-	BUG_ON(thread != cpu_local_var(current));
-
-	preempt_disable();
-	ret = sve_set_vector_length(thread, vector_length, flags);
-	preempt_enable();
-
+	ret = sve_set_vector_length(cpu_local_var(current), vl, flags);
 	if (ret) {
 		return ret;
 	}
-	return sve_prctl_status(thread->ctx.thread);
+	return sve_prctl_status(flags);
 }
 
 /* @ref.impl arch/arm64/kernel/fpsimd.c::sve_get_ti_vl */
-int sve_get_thread_vl(const struct thread *thread)
+int sve_get_thread_vl(void)
 {
-	if (!(elf_hwcap & HWCAP_SVE)) {
+	/* Instead of system_supports_sve() */
+	if (unlikely(!(elf_hwcap & HWCAP_SVE))) {
 		return -EINVAL;
 	}
-	return sve_prctl_status(thread->ctx.thread);
+	return sve_prctl_status(0);
 }
 
 void do_sve_acc(unsigned int esr, struct pt_regs *regs)
@@ -203,25 +288,48 @@ void do_sve_acc(unsigned int esr, struct pt_regs *regs)
 	panic("");
 }
 
-void init_sve_vl(void)
+void sve_setup(void)
 {
 	extern unsigned long ihk_param_default_vl;
 	uint64_t zcr;
 
+	/* Instead of system_supports_sve() */
 	if (unlikely(!(elf_hwcap & HWCAP_SVE))) {
 		return;
 	}
 
-	zcr = read_system_reg(SYS_ZCR_EL1);
-	BUG_ON(((zcr & ZCR_EL1_LEN_MASK) + 1) * 16 > sve_max_vl);
+	/* init sve_vq_map bitmap */
+	sve_init_vq_map();
+
+	/*
+	 * The SVE architecture mandates support for 128-bit vectors,
+	 * so sve_vq_map must have at least SVE_VQ_MIN set.
+	 * If something went wrong, at least try to patch it up:
+	 */
+	if (!test_bit(vq_to_bit(SVE_VQ_MIN), sve_vq_map)) {
+		set_bit(vq_to_bit(SVE_VQ_MIN), sve_vq_map);
+	}
+
+	zcr = read_system_reg(SYS_ZCR_EL1);
+	sve_max_vl = sve_vl_from_vq((zcr & ZCR_EL1_LEN_MASK) + 1);
+
+	/*
+	 * Sanity-check that the max VL we determined through CPU features
+	 * corresponds properly to sve_vq_map.  If not, do our best:
+	 */
+	if (sve_max_vl != find_supported_vector_length(sve_max_vl)) {
+		sve_max_vl = find_supported_vector_length(sve_max_vl);
+	}
 
-	sve_max_vl = ((zcr & ZCR_EL1_LEN_MASK) + 1) * 16;
 	sve_default_vl = ihk_param_default_vl;
 
-	if (sve_default_vl == 0) {
-		kprintf("SVE: Getting default VL = 0 from HOST-Linux.\n");
-		sve_default_vl = sve_max_vl > 64 ? 64 : sve_max_vl;
-		kprintf("SVE: Using default vl(%d byte).\n", sve_default_vl);
+	if (ihk_param_default_vl !=
+		find_supported_vector_length(ihk_param_default_vl)) {
+		kprintf("SVE: Getting unsupported default VL = %d "
+			"from HOST-Linux.\n", sve_default_vl);
+		sve_default_vl = find_supported_vector_length(64);
+		kprintf("SVE: Using default vl(%d byte).\n",
+			sve_default_vl);
 	}
 
 	kprintf("SVE: maximum available vector length %u bytes per vector\n",
@@ -232,7 +340,7 @@ void init_sve_vl(void)
 
 #else /* CONFIG_ARM64_SVE */
 
-void init_sve_vl(void)
+void sve_setup(void)
 {
 	/* nothing to do. */
 }

arch/arm64/kernel/head.S

@@ -10,7 +10,7 @@
 #include <smp.h>
 #include <arm-gic-v3.h>
 
-#define KERNEL_RAM_VADDR	MAP_KERNEL_START
+/* KERNEL_RAM_VADDR is defined by cmake */
 
 //#ifndef CONFIG_SMP
 //# define PTE_FLAGS	PTE_TYPE_PAGE | PTE_AF

arch/arm64/kernel/include/arch-lock.h

@@ -255,90 +255,6 @@ static void __ihk_mc_spinlock_unlock(ihk_spinlock_t *lock, unsigned long flags)
 	cpu_restore_interrupt(flags);
 }
 
-/* An implementation of the Mellor-Crummey Scott (MCS) lock */
-typedef struct mcs_lock_node {
-	unsigned long locked;
-	struct mcs_lock_node *next;
-	unsigned long irqsave;
-#ifndef ENABLE_UBSAN
-} __aligned(64) mcs_lock_node_t;
-#else
-} mcs_lock_node_t;
-#endif
-
-typedef mcs_lock_node_t mcs_lock_t;
-
-static void mcs_lock_init(struct mcs_lock_node *node)
-{
-	node->locked = 0;
-	node->next = NULL;
-}
-
-static void __mcs_lock_lock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	struct mcs_lock_node *pred;
-
-	node->next = NULL;
-	node->locked = 0;
-	pred = xchg8(&(lock->next), node);
-
-	if (pred) {
-		node->locked = 1;
-		pred->next = node;
-		while (node->locked != 0) {
-			cpu_pause();
-		}
-	}
-}
-
-static void __mcs_lock_unlock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	if (node->next == NULL) {
-		struct mcs_lock_node *old = atomic_cmpxchg8(&(lock->next), node, 0);
-
-		if (old == node) {
-			return;
-		}
-
-		while (node->next == NULL) {
-			cpu_pause();
-		}
-	}
-
-	node->next->locked = 0;
-}
-
-static void mcs_lock_lock_noirq(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	preempt_disable();
-	__mcs_lock_lock(lock, node);
-}
-
-static void mcs_lock_unlock_noirq(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	__mcs_lock_unlock(lock, node);
-	preempt_enable();
-}
-
-static void mcs_lock_lock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	node->irqsave = cpu_disable_interrupt_save();
-	mcs_lock_lock_noirq(lock, node);
-}
-
-static void mcs_lock_unlock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	mcs_lock_unlock_noirq(lock, node);
-	cpu_restore_interrupt(node->irqsave);
-}
-
-
 #define SPINLOCK_IN_MCS_RWLOCK
 
 // reader/writer lock
@@ -743,5 +659,102 @@ static inline int irqflags_can_interrupt(unsigned long flags)
 }
 #endif /* CONFIG_HAS_NMI */
 
+struct ihk_rwlock {
+	unsigned int lock;
+};
 
+static inline void ihk_mc_rwlock_init(struct ihk_rwlock *rw)
+{
+	rw->lock = 0;
+}
+
+static inline void ihk_mc_read_lock(struct ihk_rwlock *rw)
+{
+	unsigned int tmp, tmp2;
+
+	asm volatile(
+	"       sevl\n"
+	"1:     wfe\n"
+	"2:     ldaxr   %w0, %2\n"
+	"       add     %w0, %w0, #1\n"
+	"       tbnz    %w0, #31, 1b\n"
+	"       stxr    %w1, %w0, %2\n"
+	"       cbnz    %w1, 2b\n"
+	: "=&r" (tmp), "=&r" (tmp2), "+Q" (rw->lock)
+	:
+	: "cc", "memory");
+}
+
+static inline int ihk_mc_read_trylock(struct ihk_rwlock *rw)
+{
+	unsigned int tmp, tmp2 = 1;
+
+	asm volatile(
+	"       ldaxr   %w0, %2\n"
+	"       add     %w0, %w0, #1\n"
+	"       tbnz    %w0, #31, 1f\n"
+	"       stxr    %w1, %w0, %2\n"
+	"1:\n"
+	: "=&r" (tmp), "+r" (tmp2), "+Q" (rw->lock)
+	:
+	: "cc", "memory");
+
+	return !tmp2;
+}
+
+static inline void ihk_mc_read_unlock(struct ihk_rwlock *rw)
+{
+	unsigned int tmp, tmp2;
+
+	asm volatile(
+	"1:     ldxr    %w0, %2\n"
+	"       sub     %w0, %w0, #1\n"
+	"       stlxr   %w1, %w0, %2\n"
+	"       cbnz    %w1, 1b\n"
+	: "=&r" (tmp), "=&r" (tmp2), "+Q" (rw->lock)
+	:
+	: "cc", "memory");
+}
+
+static inline void ihk_mc_write_lock(struct ihk_rwlock *rw)
+{
+	unsigned int tmp;
+
+	asm volatile(
+	"       sevl\n"
+	"1:     wfe\n"
+	"2:     ldaxr   %w0, %1\n"
+	"       cbnz    %w0, 1b\n"
+	"       stxr    %w0, %w2, %1\n"
+	"       cbnz    %w0, 2b\n"
+	: "=&r" (tmp), "+Q" (rw->lock)
+	: "r" (0x80000000)
+	: "cc", "memory");
+}
+
+static inline int ihk_mc_write_trylock(struct ihk_rwlock *rw)
+{
+	unsigned int tmp;
+
+	asm volatile(
+	"       ldaxr   %w0, %1\n"
+	"       cbnz    %w0, 1f\n"
+	"       stxr    %w0, %w2, %1\n"
+	"1:\n"
+	: "=&r" (tmp), "+Q" (rw->lock)
+	: "r" (0x80000000)
+	: "cc", "memory");
+
+	return !tmp;
+}
+
+static inline void ihk_mc_write_unlock(struct ihk_rwlock *rw)
+{
+	asm volatile(
+	"       stlr    %w1, %0\n"
+	: "=Q" (rw->lock) : "r" (0) : "memory");
+}
+
+#define ihk_mc_read_can_lock(rw) ((rw)->lock < 0x80000000)
+#define ihk_mc_write_can_lock(rw) ((rw)->lock == 0)
 #endif /* !__HEADER_ARM64_COMMON_ARCH_LOCK_H */

arch/arm64/kernel/include/arch-memory.h

@@ -34,7 +34,7 @@ void panic(const char *);
  */
 /* early alloc area address */
 /* START:_end, SIZE:512 pages */
-#define MAP_EARLY_ALLOC_SHIFT	9
+#define MAP_EARLY_ALLOC_SHIFT	5
 #define MAP_EARLY_ALLOC_SIZE	(UL(1) << (PAGE_SHIFT + MAP_EARLY_ALLOC_SHIFT))
 
 #ifndef __ASSEMBLY__
@@ -55,7 +55,11 @@ extern char _end[];
 # define MAP_BOOT_PARAM_END	(MAP_BOOT_PARAM + MAP_BOOT_PARAM_SIZE)
 #endif /* !__ASSEMBLY__ */
 
-#if (VA_BITS == 39 && GRANULE_SIZE == _SZ4KB)
+/*
+ * MAP_KERNEL_START is HOST MODULES_END - 8MiB.
+ * It's defined by cmake.
+ */
+#if (VA_BITS == 39 && GRANULE_SIZE == _SZ4KB) /* ARM64_MEMORY_LAYOUT=1 */
 #
 # define LD_TASK_UNMAPPED_BASE	UL(0x0000000400000000)
 # define TASK_UNMAPPED_BASE	UL(0x0000000800000000)
@@ -64,9 +68,8 @@ extern char _end[];
 # define MAP_VMAP_SIZE		UL(0x0000000100000000)
 # define MAP_FIXED_START	UL(0xffffffbffbdfd000)
 # define MAP_ST_START		UL(0xffffffc000000000)
-# define MAP_KERNEL_START	UL(0xffffffffff800000)
 #
-#elif (VA_BITS == 42 && GRANULE_SIZE == _SZ64KB)
+#elif (VA_BITS == 42 && GRANULE_SIZE == _SZ64KB) /* ARM64_MEMORY_LAYOUT=3 */
 #
 # define LD_TASK_UNMAPPED_BASE	UL(0x0000002000000000)
 # define TASK_UNMAPPED_BASE	UL(0x0000004000000000)
@@ -75,9 +78,8 @@ extern char _end[];
 # define MAP_VMAP_SIZE		UL(0x0000000100000000)
 # define MAP_FIXED_START	UL(0xfffffdfffbdd0000)
 # define MAP_ST_START		UL(0xfffffe0000000000)
-# define MAP_KERNEL_START	UL(0xffffffffe0000000)
 #
-#elif (VA_BITS == 48 && GRANULE_SIZE == _SZ4KB)
+#elif (VA_BITS == 48 && GRANULE_SIZE == _SZ4KB) /* ARM64_MEMORY_LAYOUT=2 */
 #
 # define LD_TASK_UNMAPPED_BASE	UL(0x0000080000000000)
 # define TASK_UNMAPPED_BASE	UL(0x0000100000000000)
@@ -86,9 +88,8 @@ extern char _end[];
 # define MAP_VMAP_SIZE		UL(0x0000000100000000)
 # define MAP_FIXED_START	UL(0xffff7ffffbdfd000)
 # define MAP_ST_START		UL(0xffff800000000000)
-# define MAP_KERNEL_START	UL(0xffffffffff800000)
 #
-#elif (VA_BITS == 48 && GRANULE_SIZE == _SZ64KB)
+#elif (VA_BITS == 48 && GRANULE_SIZE == _SZ64KB) /* ARM64_MEMORY_LAYOUT=4 */
 #
 # define LD_TASK_UNMAPPED_BASE	UL(0x0000080000000000)
 # define TASK_UNMAPPED_BASE	UL(0x0000100000000000)
@@ -97,7 +98,6 @@ extern char _end[];
 # define MAP_VMAP_SIZE		UL(0x0000000100000000)
 # define MAP_FIXED_START	UL(0xffff7ffffbdd0000)
 # define MAP_ST_START		UL(0xffff800000000000)
-# define MAP_KERNEL_START	UL(0xffffffffe0000000)
 #
 #else
 # error address space is not defined.
@@ -583,6 +583,40 @@ static inline int pgsize_to_tbllv(size_t pgsize)
 	return level;
 }
 
+static inline int pgsize_to_pgshift(size_t pgsize)
+{
+	/* We need to use if instead of switch because
+	 * sometimes PTLX_CONT_SIZE == PTLX_SIZE
+	 */
+	if (pgsize == PTL4_CONT_SIZE) {
+		if (CONFIG_ARM64_PGTABLE_LEVELS > 3) {
+			return PTL4_CONT_SHIFT;
+		}
+	} else if (pgsize == PTL4_SIZE) {
+		if (CONFIG_ARM64_PGTABLE_LEVELS > 3) {
+			return PTL4_SHIFT;
+		}
+	} else if (pgsize == PTL3_CONT_SIZE) {
+		if (CONFIG_ARM64_PGTABLE_LEVELS > 2) {
+			return PTL3_CONT_SHIFT;
+		}
+	} else if (pgsize == PTL3_SIZE) {
+		if (CONFIG_ARM64_PGTABLE_LEVELS > 2) {
+			return PTL3_SHIFT;
+		}
+	} else if (pgsize == PTL2_CONT_SIZE) {
+		return PTL2_CONT_SHIFT;
+	} else if (pgsize == PTL2_SIZE) {
+		return PTL2_SHIFT;
+	} else if (pgsize == PTL1_CONT_SIZE) {
+		return PTL1_CONT_SHIFT;
+	} else if (pgsize == PTL1_SIZE) {
+		return PTL1_SHIFT;
+	}
+
+	return -EINVAL;
+}
+
 static inline size_t tbllv_to_pgsize(int level)
 {
 	size_t pgsize = 0;

arch/arm64/kernel/include/arch-perfctr.h

@@ -20,17 +20,21 @@ struct arm_pmu {
 	void (*reset)(void*);
 	int (*enable_pmu)(void);
 	void (*disable_pmu)(void);
-	int (*enable_counter)(int);
-	int (*disable_counter)(int);
-	int (*enable_intens)(int);
-	int (*disable_intens)(int);
+	int (*enable_counter)(unsigned long counter_mask);
+	int (*disable_counter)(unsigned long counter_mask);
+	int (*enable_intens)(unsigned long counter_mask);
+	int (*disable_intens)(unsigned long counter_mask);
 	int (*set_event_filter)(unsigned long*, int);
 	void (*write_evtype)(int, uint32_t);
 	int (*get_event_idx)(int num_events, unsigned long used_mask,
 			     unsigned long config);
 	int (*map_event)(uint32_t, uint64_t);
+	int (*map_hw_event)(uint64_t config);
+	int (*map_cache_event)(uint64_t config);
+	int (*map_raw_event)(uint64_t config);
 	void (*enable_user_access_pmu_regs)(void);
 	void (*disable_user_access_pmu_regs)(void);
+	int (*counter_mask_valid)(unsigned long counter_mask);
 	struct per_cpu_arm_pmu *per_cpu;
 };
 

arch/arm64/kernel/include/arch/cpu.h

@@ -102,4 +102,6 @@ static inline void cpu_disable_nmi(void)
 
 #endif	/* __ASSEMBLY__ */
 
+void arch_flush_icache_all(void);
+
 #endif /* !__HEADER_ARM64_ARCH_CPU_H */

arch/arm64/kernel/include/arch/rusage.h

@@ -1,60 +0,0 @@
-#ifndef ARCH_RUSAGE_H_INCLUDED
-#define ARCH_RUSAGE_H_INCLUDED
-
-#define DEBUG_RUSAGE
-
-#define IHK_OS_PGSIZE_4KB 0
-#define IHK_OS_PGSIZE_2MB 1
-#define IHK_OS_PGSIZE_1GB 2
-
-extern struct ihk_os_monitor *monitor;
-
-extern int sprintf(char * buf, const char *fmt, ...);
-
-#define DEBUG_ARCH_RUSAGE
-#ifdef DEBUG_ARCH_RUSAGE
-#define	dprintf(...)											\
-	do {														\
-		char msg[1024];											\
-		sprintf(msg, __VA_ARGS__);								\
-		kprintf("%s,%s", __FUNCTION__, msg);					\
-	} while (0);
-#define	eprintf(...)											\
-	do {														\
-		char msg[1024];											\
-		sprintf(msg, __VA_ARGS__);								\
-		kprintf("%s,%s", __FUNCTION__, msg);					\
-	} while (0);
-#else
-#define dprintf(...) do {  } while (0)
-#define	eprintf(...)											\
-	do {														\
-		char msg[1024];											\
-		sprintf(msg, __VA_ARGS__);								\
-		kprintf("%s,%s", __FUNCTION__, msg);					\
-	} while (0);
-#endif
-
-static inline int rusage_pgsize_to_pgtype(size_t pgsize)
-{
-	int ret = IHK_OS_PGSIZE_4KB;
-#if 0 /* postk-TODO */
-	switch (pgsize) {
-	case PTL1_SIZE:
-		ret = IHK_OS_PGSIZE_4KB;
-		break;
-	case PTL2_SIZE:
-		ret = IHK_OS_PGSIZE_2MB;
-		break;
-	case PTL3_SIZE:
-		ret = IHK_OS_PGSIZE_1GB;
-		break;
-	default:
-		eprintf("unknown pgsize=%ld\n", pgsize);
-		break;
-	}
-#endif
-	return ret;
-}
-
-#endif /* !defined(ARCH_RUSAGE_H_INCLUDED) */

arch/arm64/kernel/include/arch_rusage.h

@@ -1,33 +0,0 @@
-#ifndef ARCH_RUSAGE_H_INCLUDED
-#define ARCH_RUSAGE_H_INCLUDED
-
-#include <arch-memory.h>
-
-#define DEBUG_RUSAGE
-
-#define IHK_OS_PGSIZE_4KB 0
-#define IHK_OS_PGSIZE_2MB 1
-#define IHK_OS_PGSIZE_1GB 2
-
-extern struct rusage_global rusage;
-
-static inline int rusage_pgsize_to_pgtype(size_t pgsize)
-{
-	int ret = IHK_OS_PGSIZE_4KB;
-
-	if (pgsize == PTL1_SIZE) {
-		ret = IHK_OS_PGSIZE_4KB;
-	}
-	else if (pgsize == PTL2_SIZE) {
-		ret = IHK_OS_PGSIZE_2MB;
-	}
-	else if (pgsize == PTL3_SIZE) {
-		ret = IHK_OS_PGSIZE_1GB;
-	}
-	else {
-		kprintf("%s: Error: Unknown pgsize=%ld\n", __FUNCTION__, pgsize);
-	}
-	return ret;
-}
-
-#endif /* !defined(ARCH_RUSAGE_H_INCLUDED) */

arch/arm64/kernel/include/asm-syscall.h

@@ -8,6 +8,7 @@
 #define SYSCALL_HANDLED(number, name)	DECLARATOR(number, name)
 #define SYSCALL_DELEGATED(number, name)	DECLARATOR(number, name)
 
+#include <config.h>
 #include <syscall_list.h>
 
 #undef DECLARATOR

arch/arm64/kernel/include/cpufeature.h

@@ -67,21 +67,12 @@ struct arm64_cpu_capabilities {
 	int def_scope;/* default scope */
 	int (*matches)(const struct arm64_cpu_capabilities *caps, int scope);
 	int (*enable)(void *);/* Called on all active CPUs */
-	union {
-		struct {/* To be used for erratum handling only */
-			uint32_t midr_model;
-			uint32_t midr_range_min, midr_range_max;
-		};
-
-		struct {/* Feature register checking */
-			uint32_t sys_reg;
-			uint8_t field_pos;
-			uint8_t min_field_value;
-			uint8_t hwcap_type;
-			int sign;
-			unsigned long hwcap;
-		};
-	};
+	uint32_t sys_reg;
+	uint8_t field_pos;
+	uint8_t min_field_value;
+	uint8_t hwcap_type;
+	int sign;
+	unsigned long hwcap;
 };
 
 /* @ref.impl include/linux/bitops.h */

arch/arm64/kernel/include/fpsimd.h

@@ -1,4 +1,4 @@
-/* fpsimd.h COPYRIGHT FUJITSU LIMITED 2016-2017 */
+/* fpsimd.h COPYRIGHT FUJITSU LIMITED 2016-2019 */
 #ifndef __HEADER_ARM64_COMMON_FPSIMD_H
 #define __HEADER_ARM64_COMMON_FPSIMD_H
 
@@ -42,16 +42,19 @@ extern void thread_sve_to_fpsimd(struct thread *thread, fp_regs_struct *fp_regs)
 
 extern size_t sve_state_size(struct thread const *thread);
 extern void sve_free(struct thread *thread);
-extern void sve_alloc(struct thread *thread);
+extern int sve_alloc(struct thread *thread);
 extern void sve_save_state(void *state, unsigned int *pfpsr);
 extern void sve_load_state(void const *state, unsigned int const *pfpsr, unsigned long vq_minus_1);
 extern unsigned int sve_get_vl(void);
-extern int sve_set_thread_vl(struct thread *thread, const unsigned long vector_length, const unsigned long flags);
-extern int sve_get_thread_vl(const struct thread *thread);
+extern int sve_set_thread_vl(unsigned long arg);
+extern int sve_get_thread_vl(void);
 extern int sve_set_vector_length(struct thread *thread, unsigned long vl, unsigned long flags);
 
-#define SVE_SET_VL(thread, vector_length, flags)	sve_set_thread_vl(thread, vector_length, flags)
-#define SVE_GET_VL(thread)				sve_get_thread_vl(thread)
+#define SVE_SET_VL(arg)	sve_set_thread_vl(arg)
+#define SVE_GET_VL()	sve_get_thread_vl()
+
+/* Maximum VL that SVE VL-agnostic software can transparently support */
+#define SVE_VL_ARCH_MAX 0x100
 
 #else /* CONFIG_ARM64_SVE */
 
@@ -80,12 +83,12 @@ static int sve_set_vector_length(struct thread *thread, unsigned long vl, unsign
 }
 
 /* for prctl syscall */
-#define SVE_SET_VL(a,b,c)	(-EINVAL)
-#define SVE_GET_VL(a)		(-EINVAL)
+#define SVE_SET_VL(a)	(-EINVAL)
+#define SVE_GET_VL()	(-EINVAL)
 
 #endif /* CONFIG_ARM64_SVE */
 
-extern void init_sve_vl(void);
+extern void sve_setup(void);
 extern void fpsimd_save_state(struct fpsimd_state *state);
 extern void fpsimd_load_state(struct fpsimd_state *state);
 extern void thread_fpsimd_save(struct thread *thread);

arch/arm64/kernel/include/ihk/atomic.h

@@ -124,7 +124,7 @@ static inline long ihk_atomic64_read(const ihk_atomic64_t *v)
 	return *(volatile long *)&(v)->counter64;
 }
 
-static inline void ihk_atomic64_set(ihk_atomic64_t *v, int i)
+static inline void ihk_atomic64_set(ihk_atomic64_t *v, long i)
 {
 	v->counter64 = i;
 }
@@ -147,6 +147,8 @@ static inline void ihk_atomic64_add(long i, ihk_atomic64_t *v)
 /* @ref.impl arch/arm64/include/asm/atomic.h::atomic64_inc */
 #define ihk_atomic64_inc(v)			ihk_atomic64_add(1LL, (v))
 
+#define ihk_atomic64_cmpxchg(p, o, n) cmpxchg(&((p)->counter64), o, n)
+
 /***********************************************************************
  * others
  */

arch/arm64/kernel/include/imp-sysreg.h

@@ -29,6 +29,7 @@
 #define IMP_PF_INJECTION_DISTANCE5_EL0		sys_reg(3, 3, 11, 7, 5)
 #define IMP_PF_INJECTION_DISTANCE6_EL0		sys_reg(3, 3, 11, 7, 6)
 #define IMP_PF_INJECTION_DISTANCE7_EL0		sys_reg(3, 3, 11, 7, 7)
+#define IMP_PF_PMUSERENR_EL0			sys_reg(3, 3, 9, 14, 0)
 #define IMP_BARRIER_CTRL_EL1			sys_reg(3, 0, 11, 12, 0)
 #define IMP_BARRIER_BST_BIT_EL1			sys_reg(3, 0, 11, 12, 4)
 #define IMP_BARRIER_INIT_SYNC_BB0_EL1		sys_reg(3, 0, 15, 13, 0)

arch/arm64/kernel/include/irq.h

@@ -1,4 +1,4 @@
-/* irq.h COPYRIGHT FUJITSU LIMITED 2015-2018 */
+/* irq.h COPYRIGHT FUJITSU LIMITED 2015-2019 */
 
 #ifndef __HEADER_ARM64_IRQ_H
 #define __HEADER_ARM64_IRQ_H
@@ -14,7 +14,8 @@
 #define INTRID_QUERY_FREE_MEM	2
 #define INTRID_CPU_STOP		3
 #define INTRID_TLB_FLUSH	4
-#define INTRID_STACK_TRACE	6
+#define INTRID_STACK_TRACE	5
+#define INTRID_MULTI_INTR	6
 #define INTRID_MULTI_NMI	7
 
 /* use PPI interrupt number */
@@ -29,6 +30,7 @@ extern void gic_dist_init_gicv2(unsigned long dist_base_pa, unsigned long size);
 extern void gic_cpu_init_gicv2(unsigned long cpu_base_pa, unsigned long size);
 extern void gic_enable_gicv2(void);
 extern void arm64_issue_ipi_gicv2(unsigned int cpuid, unsigned int vector);
+extern void arm64_issue_host_ipi_gicv2(uint32_t cpuid, uint32_t vector);
 extern void handle_interrupt_gicv2(struct pt_regs *regs);
 
 /* Functions for GICv3 */
@@ -36,6 +38,7 @@ extern void gic_dist_init_gicv3(unsigned long dist_base_pa, unsigned long size);
 extern void gic_cpu_init_gicv3(unsigned long cpu_base_pa, unsigned long size);
 extern void gic_enable_gicv3(void);
 extern void arm64_issue_ipi_gicv3(unsigned int cpuid, unsigned int vector);
+extern void arm64_issue_host_ipi_gicv3(uint32_t cpuid, uint32_t vector);
 extern void handle_interrupt_gicv3(struct pt_regs *regs);
 
 void handle_IPI(unsigned int vector, struct pt_regs *regs);

arch/arm64/kernel/include/prctl.h

@@ -1,4 +1,4 @@
-/* prctl.h COPYRIGHT FUJITSU LIMITED 2017 */
+/* prctl.h COPYRIGHT FUJITSU LIMITED 2017-2019 */
 #ifndef __HEADER_ARM64_COMMON_PRCTL_H
 #define __HEADER_ARM64_COMMON_PRCTL_H
 
@@ -6,15 +6,12 @@
 #define PR_GET_THP_DISABLE 42
 
 /* arm64 Scalable Vector Extension controls */
-#define PR_SVE_SET_VL		48			/* set task vector length */
-#define PR_SVE_SET_VL_THREAD	(1 << 1)		/* set just this thread */
-#define PR_SVE_SET_VL_INHERIT	(1 << 2)		/* inherit across exec */
-#define PR_SVE_SET_VL_ONEXEC	(1 << 3)		/* defer effect until exec */
-
-#define PR_SVE_GET_VL		49			/* get task vector length */
-/* Decode helpers for the return value from PR_SVE_GET_VL: */
-#define PR_SVE_GET_VL_LEN(ret)	((ret) & 0x3fff)	/* vector length */
-#define PR_SVE_GET_VL_INHERIT	(PR_SVE_SET_VL_INHERIT << 16)
-/* For conveinence, PR_SVE_SET_VL returns the result in the same encoding */
+/* Flag values must be kept in sync with ptrace NT_ARM_SVE interface */
+#define PR_SVE_SET_VL		50		/* set task vector length */
+# define PR_SVE_SET_VL_ONEXEC	(1 << 18)	/* defer effect until exec */
+#define PR_SVE_GET_VL		51		/* get task vector length */
+/* Bits common to PR_SVE_SET_VL and PR_SVE_GET_VL */
+# define PR_SVE_VL_LEN_MASK	0xffff
+# define PR_SVE_VL_INHERIT	(1 << 17)	/* inherit across exec */
 
 #endif /* !__HEADER_ARM64_COMMON_PRCTL_H */

arch/arm64/kernel/include/ptrace.h

@@ -1,4 +1,4 @@
-/* ptrace.h COPYRIGHT FUJITSU LIMITED 2015-2017 */
+/* ptrace.h COPYRIGHT FUJITSU LIMITED 2015-2019 */
 #ifndef __HEADER_ARM64_COMMON_PTRACE_H
 #define __HEADER_ARM64_COMMON_PTRACE_H
 
@@ -46,6 +46,7 @@
 
 #ifndef __ASSEMBLY__
 
+#include <lwk/compiler.h>
 #include <ihk/types.h>
 
 struct user_hwdebug_state {
@@ -78,6 +79,70 @@ struct user_sve_header {
 	uint16_t __reserved;
 };
 
+enum aarch64_regset {
+	REGSET_GPR,
+	REGSET_FPR,
+	REGSET_TLS,
+	REGSET_HW_BREAK,
+	REGSET_HW_WATCH,
+	REGSET_SYSTEM_CALL,
+#ifdef CONFIG_ARM64_SVE
+	REGSET_SVE,
+#endif /* CONFIG_ARM64_SVE */
+};
+
+struct thread;
+struct user_regset;
+
+typedef int user_regset_active_fn(struct thread *target,
+				  const struct user_regset *regset);
+
+typedef long user_regset_get_fn(struct thread *target,
+				const struct user_regset *regset,
+				unsigned int pos, unsigned int count,
+				void *kbuf, void __user *ubuf);
+
+typedef long user_regset_set_fn(struct thread *target,
+				const struct user_regset *regset,
+				unsigned int pos, unsigned int count,
+				const void *kbuf, const void __user *ubuf);
+
+typedef int user_regset_writeback_fn(struct thread *target,
+				     const struct user_regset *regset,
+				     int immediate);
+
+typedef unsigned int user_regset_get_size_fn(struct thread *target,
+					     const struct user_regset *regset);
+
+struct user_regset {
+	user_regset_get_fn		*get;
+	user_regset_set_fn		*set;
+	user_regset_active_fn		*active;
+	user_regset_writeback_fn	*writeback;
+	user_regset_get_size_fn		*get_size;
+	unsigned int			n;
+	unsigned int			size;
+	unsigned int			align;
+	unsigned int			bias;
+	unsigned int			core_note_type;
+};
+
+struct user_regset_view {
+	const char *name;
+	const struct user_regset *regsets;
+	unsigned int n;
+	uint32_t e_flags;
+	uint16_t e_machine;
+	uint8_t ei_osabi;
+};
+
+extern const struct user_regset_view *current_user_regset_view(void);
+extern const struct user_regset *find_regset(
+		const struct user_regset_view *view,
+		unsigned int type);
+extern unsigned int regset_size(struct thread *target,
+		const struct user_regset *regset);
+
 /* Definitions for user_sve_header.flags: */
 #define SVE_PT_REGS_MASK	(1 << 0)
 
@@ -85,7 +150,7 @@ struct user_sve_header {
 #define SVE_PT_REGS_SVE		SVE_PT_REGS_MASK
 
 #define SVE_PT_VL_THREAD	PR_SVE_SET_VL_THREAD
-#define SVE_PT_VL_INHERIT	PR_SVE_SET_VL_INHERIT
+#define SVE_PT_VL_INHERIT	PR_SVE_VL_INHERIT
 #define SVE_PT_VL_ONEXEC	PR_SVE_SET_VL_ONEXEC
 
 /*
@@ -99,7 +164,9 @@ struct user_sve_header {
  */
 
 /* Offset from the start of struct user_sve_header to the register data */
-#define SVE_PT_REGS_OFFSET	((sizeof(struct sve_context) + 15) / 16 * 16)
+#define SVE_PT_REGS_OFFSET					\
+	((sizeof(struct sve_context) + (SVE_VQ_BYTES - 1))	\
+		/ SVE_VQ_BYTES * SVE_VQ_BYTES)
 
 /*
  * The register data content and layout depends on the value of the
@@ -174,8 +241,10 @@ struct user_sve_header {
 #define SVE_PT_SVE_FFR_OFFSET(vq) \
 	__SVE_SIG_TO_PT(SVE_SIG_FFR_OFFSET(vq))
 
-#define SVE_PT_SVE_FPSR_OFFSET(vq) \
-	((SVE_PT_SVE_FFR_OFFSET(vq) + SVE_PT_SVE_FFR_SIZE(vq) + 15) / 16 * 16)
+#define SVE_PT_SVE_FPSR_OFFSET(vq)				\
+	((SVE_PT_SVE_FFR_OFFSET(vq) + SVE_PT_SVE_FFR_SIZE(vq) + \
+			(SVE_VQ_BYTES - 1))			\
+		/ SVE_VQ_BYTES * SVE_VQ_BYTES)
 #define SVE_PT_SVE_FPCR_OFFSET(vq) \
 	(SVE_PT_SVE_FPSR_OFFSET(vq) + SVE_PT_SVE_FPSR_SIZE)
 
@@ -184,9 +253,10 @@ struct user_sve_header {
  * 128-bit boundary.
  */
 
-#define SVE_PT_SVE_SIZE(vq, flags)				\
-	((SVE_PT_SVE_FPCR_OFFSET(vq) + SVE_PT_SVE_FPCR_SIZE -	\
-		SVE_PT_SVE_OFFSET + 15) / 16 * 16)
+#define SVE_PT_SVE_SIZE(vq, flags)					\
+	((SVE_PT_SVE_FPCR_OFFSET(vq) + SVE_PT_SVE_FPCR_SIZE		\
+			- SVE_PT_SVE_OFFSET + (SVE_VQ_BYTES - 1))	\
+		/ SVE_VQ_BYTES * SVE_VQ_BYTES)
 
 #define SVE_PT_SIZE(vq, flags)						\
 	(((flags) & SVE_PT_REGS_MASK) == SVE_PT_REGS_SVE ?		\

arch/arm64/kernel/include/rlimit.h

@@ -85,7 +85,11 @@ enum __rlimit_resource
   __RLIMIT_RTPRIO = 14,
 #define RLIMIT_RTPRIO __RLIMIT_RTPRIO
 
-  __RLIMIT_NLIMITS = 15,
+  /* timeout for RT tasks in us */
+  __RLIMIT_RTTIME = 15,
+#define RLIMIT_RTTIME __RLIMIT_RTTIME
+
+  __RLIMIT_NLIMITS = 16,
   __RLIM_NLIMITS = __RLIMIT_NLIMITS
 #define RLIMIT_NLIMITS __RLIMIT_NLIMITS
 #define RLIM_NLIMITS __RLIM_NLIMITS

arch/arm64/kernel/include/signal.h

@@ -1,4 +1,4 @@
-/* signal.h COPYRIGHT FUJITSU LIMITED 2015-2018 */
+/* signal.h COPYRIGHT FUJITSU LIMITED 2015-2019 */
 #ifndef __HEADER_ARM64_COMMON_SIGNAL_H
 #define __HEADER_ARM64_COMMON_SIGNAL_H
 
@@ -298,6 +298,7 @@ struct extra_context {
 	struct _aarch64_ctx head;
 	void *data;	/* 16-byte aligned pointer to the extra space */
 	uint32_t size;	/* size in bytes of the extra space */
+	uint32_t __reserved[3];
 };
 
 #define SVE_MAGIC	0x53564501
@@ -318,19 +319,25 @@ struct sve_context {
  * The SVE architecture leaves space for future expansion of the
  * vector length beyond its initial architectural limit of 2048 bits
  * (16 quadwords).
+ *
+ * See linux/Documentation/arm64/sve.txt for a description of the VL/VQ
+ * terminology.
  */
-#define SVE_VQ_MIN	1
-#define SVE_VQ_MAX	0x200
+#define SVE_VQ_BYTES	16	/* number of bytes per quadword */
 
-#define SVE_VL_MIN	(SVE_VQ_MIN * 0x10)
-#define SVE_VL_MAX	(SVE_VQ_MAX * 0x10)
+#define SVE_VQ_MIN	1
+#define SVE_VQ_MAX	512
+
+#define SVE_VL_MIN	(SVE_VQ_MIN * SVE_VQ_BYTES)
+#define SVE_VL_MAX	(SVE_VQ_MAX * SVE_VQ_BYTES)
 
 #define SVE_NUM_ZREGS	32
 #define SVE_NUM_PREGS	16
 
 #define sve_vl_valid(vl) \
-	((vl) % 0x10 == 0 && (vl) >= SVE_VL_MIN && (vl) <= SVE_VL_MAX)
-#define sve_vq_from_vl(vl)	((vl) / 0x10)
+	((vl) % SVE_VQ_BYTES == 0 && (vl) >= SVE_VL_MIN && (vl) <= SVE_VL_MAX)
+#define sve_vq_from_vl(vl)	((vl) / SVE_VQ_BYTES)
+#define sve_vl_from_vq(vq)	((vq) * SVE_VQ_BYTES)
 
 /*
  * The total size of meaningful data in the SVE context in bytes,
@@ -365,11 +372,13 @@ struct sve_context {
  * Additional data might be appended in the future.
  */
 
-#define SVE_SIG_ZREG_SIZE(vq)	((uint32_t)(vq) * 16)
-#define SVE_SIG_PREG_SIZE(vq)	((uint32_t)(vq) * 2)
+#define SVE_SIG_ZREG_SIZE(vq)	((uint32_t)(vq) * SVE_VQ_BYTES)
+#define SVE_SIG_PREG_SIZE(vq)	((uint32_t)(vq) * (SVE_VQ_BYTES / 8))
 #define SVE_SIG_FFR_SIZE(vq)	SVE_SIG_PREG_SIZE(vq)
 
-#define SVE_SIG_REGS_OFFSET	((sizeof(struct sve_context) + 15) / 16 * 16)
+#define SVE_SIG_REGS_OFFSET					\
+	((sizeof(struct sve_context) + (SVE_VQ_BYTES - 1))	\
+		/ SVE_VQ_BYTES * SVE_VQ_BYTES)
 
 #define SVE_SIG_ZREGS_OFFSET	SVE_SIG_REGS_OFFSET
 #define SVE_SIG_ZREG_OFFSET(vq, n) \

arch/arm64/kernel/include/syscall_list.h

@@ -2,7 +2,7 @@
 
 SYSCALL_DELEGATED(4, io_getevents)
 SYSCALL_DELEGATED(17, getcwd)
-SYSCALL_DELEGATED(22, epoll_pwait)
+SYSCALL_HANDLED(22, epoll_pwait)
 SYSCALL_DELEGATED(25, fcntl)
 SYSCALL_HANDLED(29, ioctl)
 SYSCALL_DELEGATED(35, unlinkat)
@@ -17,8 +17,8 @@ SYSCALL_DELEGATED(64, write)
 SYSCALL_DELEGATED(66, writev)
 SYSCALL_DELEGATED(67, pread64)
 SYSCALL_DELEGATED(68, pwrite64)
-SYSCALL_DELEGATED(72, pselect6)
-SYSCALL_DELEGATED(73, ppoll)
+SYSCALL_HANDLED(72, pselect6)
+SYSCALL_HANDLED(73, ppoll)
 SYSCALL_HANDLED(74, signalfd4)
 SYSCALL_DELEGATED(78, readlinkat)
 SYSCALL_DELEGATED(80, fstat)
@@ -83,6 +83,7 @@ SYSCALL_HANDLED(175, geteuid)
 SYSCALL_HANDLED(176, getgid)
 SYSCALL_HANDLED(177, getegid)
 SYSCALL_HANDLED(178, gettid)
+SYSCALL_HANDLED(179, sysinfo)
 SYSCALL_DELEGATED(188, msgrcv)
 SYSCALL_DELEGATED(189, msgsnd)
 SYSCALL_DELEGATED(192, semtimedop)
@@ -111,20 +112,16 @@ SYSCALL_HANDLED(236, get_mempolicy)
 SYSCALL_HANDLED(237, set_mempolicy)
 SYSCALL_HANDLED(238, migrate_pages)
 SYSCALL_HANDLED(239, move_pages)
-#ifdef PERF_ENABLE
+#ifdef ENABLE_PERF
 SYSCALL_HANDLED(241, perf_event_open)
 #else // PERF_ENABLE
 SYSCALL_DELEGATED(241, perf_event_open)
 #endif // PERF_ENABLE
 SYSCALL_HANDLED(260, wait4)
+SYSCALL_HANDLED(261, prlimit64)
 SYSCALL_HANDLED(270, process_vm_readv)
 SYSCALL_HANDLED(271, process_vm_writev)
-#ifdef PERF_ENABLE
-SYSCALL_HANDLED(601, pmc_init)
-SYSCALL_HANDLED(602, pmc_start)
-SYSCALL_HANDLED(603, pmc_stop)
-SYSCALL_HANDLED(604, pmc_reset)
-#endif // PERF_ENABLE
+SYSCALL_HANDLED(281, execveat)
 SYSCALL_HANDLED(700, get_cpu_id)
 #ifdef PROFILE_ENABLE
 SYSCALL_HANDLED(__NR_profile, profile)
@@ -132,6 +129,7 @@ SYSCALL_HANDLED(__NR_profile, profile)
 SYSCALL_HANDLED(730, util_migrate_inter_kernel)
 SYSCALL_HANDLED(731, util_indicate_clone)
 SYSCALL_HANDLED(732, get_system)
+SYSCALL_HANDLED(733, util_register_desc)
 
 /* McKernel Specific */
 SYSCALL_HANDLED(801, swapout)
@@ -146,3 +144,9 @@ SYSCALL_HANDLED(1045, signalfd)
 SYSCALL_DELEGATED(1049, stat)
 SYSCALL_DELEGATED(1060, getpgrp)
 SYSCALL_HANDLED(1062, time)
+SYSCALL_DELEGATED(1069, epoll_wait)
+
+/* Do not edit the lines including this comment and
+ * EOF just after it because those are used as a
+ * robust marker for the autotest patch.
+ */

arch/arm64/kernel/include/thread_info.h

@@ -1,4 +1,4 @@
-/* thread_info.h COPYRIGHT FUJITSU LIMITED 2015-2018 */
+/* thread_info.h COPYRIGHT FUJITSU LIMITED 2015-2019 */
 #ifndef __HEADER_ARM64_COMMON_THREAD_INFO_H
 #define __HEADER_ARM64_COMMON_THREAD_INFO_H
 
@@ -46,9 +46,9 @@ struct thread_info {
 	int			cpu;		/* cpu */
 	struct cpu_context	cpu_context;	/* kernel_context */
 	void			*sve_state;	/* SVE registers, if any */
-	uint16_t		sve_vl;		/* SVE vector length */
-	uint16_t		sve_vl_onexec;	/* SVE vl after next exec */
-	uint16_t		sve_flags;	/* SVE related flags */
+	unsigned int		sve_vl;		/* SVE vector length */
+	unsigned int		sve_vl_onexec;	/* SVE vl after next exec */
+	unsigned long		sve_flags;	/* SVE related flags */
 	unsigned long		fault_address;	/* fault info */
 	unsigned long		fault_code;	/* ESR_EL1 value */
 };
@@ -56,7 +56,7 @@ struct thread_info {
 /* Flags for sve_flags (intentionally defined to match the prctl flags) */
 
 /* Inherit sve_vl and sve_flags across execve(): */
-#define THREAD_VL_INHERIT	PR_SVE_SET_VL_INHERIT
+#define THREAD_VL_INHERIT	PR_SVE_VL_INHERIT
 
 struct arm64_cpu_local_thread {
 	struct thread_info thread_info;

arch/arm64/kernel/include/traps.h

@@ -4,6 +4,7 @@
 #define __ASM_TRAP_H
 
 #include <types.h>
+#include <arch-lock.h>
 
 struct pt_regs;
 

arch/arm64/kernel/irq-gic-v2.c

@@ -7,7 +7,7 @@
 #include <memory.h>
 #include <affinity.h>
 #include <syscall.h>
-#include <debug.h>
+#include <ihk/debug.h>
 #include <arch-timer.h>
 #include <cls.h>
 
@@ -31,10 +31,9 @@ void *cpu_base;
  * function, it is not necessary to perform the disable/enable
  * interrupts in this function as gic_raise_softirq() .
  */
-static void arm64_raise_sgi_gicv2(unsigned int cpuid, unsigned int vector)
+static void __arm64_raise_sgi_gicv2(unsigned int hw_cpuid, unsigned int vector)
 {
 	/* Build interrupt destination of the target cpu */
-	unsigned int hw_cpuid = ihk_mc_get_cpu_info()->hw_ids[cpuid];
 	uint8_t cpu_target_list = gic_hwid_to_affinity(hw_cpuid);
 
 	/*
@@ -50,6 +49,23 @@ static void arm64_raise_sgi_gicv2(unsigned int cpuid, unsigned int vector)
 	);
 }
 
+static void arm64_raise_sgi_gicv2(uint32_t cpuid, uint32_t vector)
+{
+	/* Build interrupt destination of the target CPU */
+	uint32_t hw_cpuid = ihk_mc_get_cpu_info()->hw_ids[cpuid];
+
+	__arm64_raise_sgi_gicv2(hw_cpuid, vector);
+}
+
+static void arm64_raise_sgi_to_host_gicv2(uint32_t cpuid, uint32_t vector)
+{
+	/* Build interrupt destination of the target Linux/host CPU */
+	uint32_t hw_cpuid = ihk_mc_get_apicid(cpuid);
+
+	__arm64_raise_sgi_gicv2(hw_cpuid, vector);
+}
+
+
 /**
  * arm64_raise_spi_gicv2
  * @ref.impl nothing.
@@ -77,6 +93,11 @@ static void arm64_raise_spi_gicv2(unsigned int cpuid, unsigned int vector)
 	);
 }
 
+void arm64_issue_host_ipi_gicv2(uint32_t cpuid, uint32_t vector)
+{
+	arm64_raise_sgi_to_host_gicv2(cpuid, vector);
+}
+
 /**
  * arm64_issue_ipi_gicv2
  * @param cpuid : hardware cpu id

arch/arm64/kernel/irq-gic-v3.c

@@ -6,7 +6,7 @@
 #include <cputype.h>
 #include <process.h>
 #include <syscall.h>
-#include <debug.h>
+#include <ihk/debug.h>
 #include <arch-timer.h>
 #include <cls.h>
 
@@ -195,15 +195,12 @@ static inline void gic_write_bpr1(uint32_t val)
 }
 #endif
 
-static void arm64_raise_sgi_gicv3(uint32_t cpuid, uint32_t vector)
+static void __arm64_raise_sgi_gicv3(uint32_t hw_cpuid, uint32_t vector)
 {
 	uint64_t mpidr, cluster_id;
 	uint16_t tlist;
 	uint64_t val;
 
-	/* Build interrupt destination of the target cpu */
-	uint32_t hw_cpuid = ihk_mc_get_cpu_info()->hw_ids[cpuid];
-
 	/*
 	 * Ensure that stores to Normal memory are visible to the
 	 * other CPUs before issuing the IPI.
@@ -239,6 +236,22 @@ static void arm64_raise_sgi_gicv3(uint32_t cpuid, uint32_t vector)
 	}
 }
 
+static void arm64_raise_sgi_gicv3(uint32_t cpuid, uint32_t vector)
+{
+	/* Build interrupt destination of the target CPU */
+	uint32_t hw_cpuid = ihk_mc_get_cpu_info()->hw_ids[cpuid];
+
+	__arm64_raise_sgi_gicv3(hw_cpuid, vector);
+}
+
+static void arm64_raise_sgi_to_host_gicv3(uint32_t cpuid, uint32_t vector)
+{
+	/* Build interrupt destination of the target Linux/host CPU */
+	uint32_t hw_cpuid = ihk_mc_get_apicid(cpuid);
+
+	__arm64_raise_sgi_gicv3(hw_cpuid, vector);
+}
+
 static void arm64_raise_spi_gicv3(uint32_t cpuid, uint32_t vector)
 {
 	uint64_t spi_reg_offset;
@@ -268,6 +281,11 @@ static void arm64_raise_lpi_gicv3(uint32_t cpuid, uint32_t vector)
 	ekprintf("%s called.\n", __func__);
 }
  
+void arm64_issue_host_ipi_gicv3(uint32_t cpuid, uint32_t vector)
+{
+	arm64_raise_sgi_to_host_gicv3(cpuid, vector);
+}
+
 void arm64_issue_ipi_gicv3(uint32_t cpuid, uint32_t vector)
 {
 	dkprintf("Send irq#%d to cpuid=%d\n", vector, cpuid);
@@ -292,6 +310,9 @@ void handle_interrupt_gicv3(struct pt_regs *regs)
 {
 	uint64_t irqnr;
 	const int from_user = interrupt_from_user(regs);
+	struct cpu_local_var *v = get_this_cpu_local_var();
+	//unsigned long irqflags;
+	int do_check = 0;
 
 	irqnr = gic_read_iar();
 	cpu_enable_nmi();
@@ -305,10 +326,18 @@ void handle_interrupt_gicv3(struct pt_regs *regs)
 	}
 	set_cputime(from_user ? CPUTIME_MODE_K2U : CPUTIME_MODE_K2K_OUT);
 
-	/* for migration by IPI */
-	if (get_this_cpu_local_var()->flags & CPU_FLAG_NEED_MIGRATE) {
-		schedule();
+	//irqflags = ihk_mc_spinlock_lock(&v->runq_lock);
+	/* For migration by IPI or by timesharing */
+	if (v->flags &
+	    (CPU_FLAG_NEED_MIGRATE | CPU_FLAG_NEED_RESCHED)) {
+		v->flags &= ~CPU_FLAG_NEED_RESCHED;
+		do_check = 1;
+	}
+	//ihk_mc_spinlock_unlock(&v->runq_lock, irqflags);
+
+	if (do_check) {
 		check_signal(0, regs, 0);
+		schedule();
 	}
 }
 
@@ -344,9 +373,11 @@ static void init_spi_routing(uint32_t irq, uint32_t linux_cpu)
 
 void gic_dist_init_gicv3(unsigned long dist_base_pa, unsigned long size)
 {
+#ifndef IHK_IKC_USE_LINUX_WORK_IRQ
 	extern int spi_table[];
 	extern int nr_spi_table;
 	int i;
+#endif // !IHK_IKC_USE_LINUX_WORK_IRQ
 
 	dist_base = map_fixed_area(dist_base_pa, size, 1 /*non chachable*/);
 
@@ -357,6 +388,7 @@ void gic_dist_init_gicv3(unsigned long dist_base_pa, unsigned long size)
 	}
 #endif
 
+#ifndef IHK_IKC_USE_LINUX_WORK_IRQ
 	/* initialize spi routing */
 	for (i = 0; i < nr_spi_table; i++) {
 		if (spi_table[i] == -1) {
@@ -364,6 +396,7 @@ void gic_dist_init_gicv3(unsigned long dist_base_pa, unsigned long size)
 		}
 		init_spi_routing(spi_table[i], i);
 	}
+#endif // !IHK_IKC_USE_LINUX_WORK_IRQ
 }
 
 void gic_cpu_init_gicv3(unsigned long cpu_base_pa, unsigned long size)

arch/arm64/kernel/memory.c

@@ -1,6 +1,5 @@
 /* memory.c COPYRIGHT FUJITSU LIMITED 2015-2018 */
 #include <ihk/cpu.h>
-#include <ihk/debug.h>
 #include <ihk/mm.h>
 #include <types.h>
 #include <memory.h>
@@ -14,7 +13,7 @@
 #include <context.h>
 #include <kmalloc.h>
 #include <vdso.h>
-#include <debug.h>
+#include <ihk/debug.h>
 #include <rusage_private.h>
 #include <cputype.h>
 
@@ -2672,17 +2671,28 @@ int set_range_l1(void *args0, pte_t *ptep, uintptr_t base, uintptr_t start,
 	}
 
 	phys = args->phys + (base - start);
-	if (__page_offset(base, PTL1_CONT_SIZE) == 0) { //check head pte
+
+	/* Check if we can begin / end a series of contiguous PTEs */
+	if (__page_offset(base, PTL1_CONT_SIZE) == 0) {
 		uintptr_t next_addr = base + PTL1_CONT_SIZE;
 
 		if (end < next_addr) {
 			next_addr = end;
 		}
 
-		// set contiguous bit until the next head pte
-		// if phys is aligned and range does not end early.
+		/* Begin the series if physical address is also aligned and
+		 * the range covers the series. Don't start or end it if
+		 * physical address is not aligned or the range ends early.
+		 */
 		if (__page_offset(phys | next_addr, PTL1_CONT_SIZE) == 0) {
 			args->attr[0] |= PTE_CONT;
+			if (rusage_memory_stat_add(args->range, phys,
+						   PTL1_CONT_SIZE,
+						   PTL1_CONT_SIZE)) {
+				dkprintf("%lx+,%s: calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
+					 phys, __func__, base, phys,
+					 PTL1_CONT_SIZE, PTL1_CONT_SIZE);
+			}
 		} else {
 			args->attr[0] &= ~PTE_CONT;
 		}
@@ -2692,12 +2702,13 @@ int set_range_l1(void *args0, pte_t *ptep, uintptr_t base, uintptr_t start,
 
 	error = 0;
 	// call memory_stat_rss_add() here because pgshift is resolved here
-	if (rusage_memory_stat_add(args->range, phys, PTL1_SIZE, PTL1_SIZE)) {
-		dkprintf("%lx+,%s: calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
-			 phys, __func__, base, phys, PTL1_SIZE, PTL1_SIZE);
-	} else {
-		dkprintf("%s: !calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
-			 __func__, base, phys, PTL1_SIZE, PTL1_SIZE);
+	if (!(args->attr[0] & PTE_CONT)) {
+		if (rusage_memory_stat_add(args->range, phys,
+					   PTL1_SIZE, PTL1_SIZE)) {
+			dkprintf("%lx+,%s: calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
+				phys, __func__, base, phys,
+				PTL1_SIZE, PTL1_SIZE);
+		}
 	}
 
 out:
@@ -2761,7 +2772,9 @@ retry:
 
 				phys = args->phys + (base - start);
 
-				//check head pte
+				/* Check if we can begin / end a series of
+				 * contiguous PTEs
+				 */
 				if (__page_offset(base, tbl.cont_pgsize) == 0) {
 					uintptr_t next_addr = base +
 						tbl.cont_pgsize;
@@ -2770,11 +2783,24 @@ retry:
 						next_addr = end;
 					}
 
-					// set contiguous bit until the
-					// next head pte if phys is aligned
-					// and range does not end early.
+					/* Begin the series if physical address
+					 * is also aligned and the range covers
+					 * the series. Don't start or end it if
+					 * physical address is not aligned or
+					 * the range ends early.
+					 */
 					if (__page_offset(phys | next_addr, tbl.cont_pgsize) == 0) {
 						args->attr[level-1] |= PTE_CONT;
+						if (rusage_memory_stat_add(args->range,
+									   phys,
+									   tbl.cont_pgsize,
+									   tbl.cont_pgsize)) {
+							dkprintf("%lx+,%s: calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
+								 phys, __func__,
+								 base, phys,
+								 tbl.cont_pgsize,
+								 tbl.cont_pgsize);
+						}
 					} else {
 						args->attr[level-1] &= ~PTE_CONT;
 					}
@@ -2782,21 +2808,23 @@ retry:
 
 				ptl_set(ptep, phys | args->attr[level-1],
 					level);
+
 				error = 0;
 				dkprintf("set_range_middle(%lx,%lx,%lx,%d):"
 					 "large page. %d %lx\n",
 					 base, start, end, level, error, *ptep);
 				// Call memory_stat_rss_add() here because pgshift is resolved here
-				if (rusage_memory_stat_add(args->range, phys,
-							   tbl.pgsize,
-							   tbl.pgsize)) {
-					dkprintf("%lx+,%s: calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
-						 phys, __func__, base, phys,
-						 tbl.pgsize, tbl.pgsize);
-				} else {
-					dkprintf("%s: !calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
-						 __func__, base, phys,
-						 tbl.pgsize, tbl.pgsize);
+				if (!(args->attr[level-1] & PTE_CONT)) {
+					if (rusage_memory_stat_add(args->range,
+								   phys,
+								   tbl.pgsize,
+								   tbl.pgsize)) {
+						dkprintf("%lx+,%s: calling memory_stat_rss_add(),base=%lx,phys=%lx,size=%ld,pgsize=%ld\n",
+							 phys, __func__, base,
+							 phys,
+							 tbl.pgsize,
+							 tbl.pgsize);
+					}
 				}
 				goto out;
 			}
@@ -2848,7 +2876,7 @@ retry:
 	error = 0;
 out:
 	if (tt_pa) {
-		ihk_mc_free_pages(tt_pa, 1);
+		ihk_mc_free_pages(phys_to_virt((unsigned long)tt_pa), 1);
 	}
 	dkprintf("set_range_middle(%lx,%lx,%lx,%d): %d %lx\n",
 		 base, start, end, level, error, *ptep);
@@ -3200,6 +3228,7 @@ void load_page_table(struct page_table *pt)
 {
 	if (pt == NULL) {
 		// load page table for idle(EL1) process.
+		switch_mm(init_pt);
 		return;
 	}
 	// load page table for user(EL0) thread.
@@ -3259,7 +3288,7 @@ void *map_fixed_area(unsigned long phys, unsigned long size, int uncachable)
 		attr |= PTATTR_UNCACHABLE;
 	}
 
-	kprintf("map_fixed: phys: 0x%lx => 0x%lx (%d pages)\n",
+	dkprintf("map_fixed: phys: 0x%lx => 0x%lx (%d pages)\n",
 			paligned, v, npages);
 
 	pt = get_init_page_table();
@@ -3335,15 +3364,15 @@ unsigned long virt_to_phys(void *v)
 {
 	unsigned long va = (unsigned long)v;
 
-	if (MAP_KERNEL_START <= va) {
-		return va - MAP_KERNEL_START + arm64_kernel_phys_base;
+	if (va >= MAP_ST_START) {
+		return va - MAP_ST_START + arm64_st_phys_base;
 	}
-	return va - MAP_ST_START;
+	return va - MAP_KERNEL_START + arm64_kernel_phys_base;
 }
 
 void *phys_to_virt(unsigned long p)
 {
-	return (void *)(p | MAP_ST_START);
+	return (void *)((p - arm64_st_phys_base) | MAP_ST_START);
 }
 
 int copy_from_user(void *dst, const void *src, size_t siz)
@@ -3716,44 +3745,6 @@ translation_table_t* get_translation_table_as_paddr(const struct page_table *pt)
 	return pt->tt_pa;
 }
 
-#ifdef POSTK_DEBUG_ARCH_DEP_8
-void remote_flush_tlb_cpumask(struct process_vm *vm, 
-		unsigned long addr, int cpu_id)
-{
-	unsigned long cpu;
-	cpu_set_t _cpu_set;
-	int flush_ind;
-
-	if (addr) {
-		flush_ind = (addr >> PAGE_SHIFT) % IHK_TLB_FLUSH_IRQ_VECTOR_SIZE;
-	}
-	/* Zero address denotes full TLB flush */
-	else {	
-		/* Random.. */
-		flush_ind = (rdtsc()) % IHK_TLB_FLUSH_IRQ_VECTOR_SIZE;
-	}
-
-	/* Take a copy of the cpu set so that we don't hold the lock
-	 * all the way while interrupting other cores */
-	ihk_mc_spinlock_lock_noirq(&vm->address_space->cpu_set_lock);
-	memcpy(&_cpu_set, &vm->address_space->cpu_set, sizeof(cpu_set_t));
-	ihk_mc_spinlock_unlock_noirq(&vm->address_space->cpu_set_lock);
-
-	/* Loop through CPUs in this address space and interrupt them for
-	 * TLB flush on the specified address */
-	for_each_set_bit(cpu, (const unsigned long*)&_cpu_set.__bits, CPU_SETSIZE) {
-		if (ihk_mc_get_processor_id() == cpu) 
-			continue;
-
-		dkprintf("remote_flush_tlb_cpumask: flush_ind: %d, addr: 0x%lX, interrupting cpu: %d\n",
-		        flush_ind, addr, cpu);
-
-		ihk_mc_interrupt_cpu(cpu,
-				     ihk_mc_get_vector(flush_ind + IHK_TLB_FLUSH_IRQ_VECTOR_START));
-	}
-}
-#endif /* POSTK_DEBUG_ARCH_DEP_8 */
-
 void arch_adjust_allocate_page_size(struct page_table *pt,
 				    uintptr_t fault_addr,
 				    pte_t *ptep,

arch/arm64/kernel/mikc.c

@@ -19,7 +19,7 @@ int ihk_mc_ikc_init_first_local(struct ihk_ikc_channel_desc *channel,
 
 	memset(channel, 0, sizeof(struct ihk_ikc_channel_desc));
 
-	mikc_queue_pages = ((4 * num_processors * MASTER_IKCQ_PKTSIZE)
+	mikc_queue_pages = ((8 * num_processors * MASTER_IKCQ_PKTSIZE)
 			+ (PAGE_SIZE - 1)) / PAGE_SIZE;
 
 	/* Place both sides in this side */

arch/arm64/kernel/perfctr.c

@@ -8,6 +8,7 @@
 #include <string.h>
 #include <ihk/mm.h>
 #include <irq.h>
+#include <process.h>
 
 /*
  * @ref.impl arch/arm64/kernel/perf_event.c
@@ -85,25 +86,17 @@ void arm64_disable_user_access_pmu_regs(void)
 	cpu_pmu.disable_user_access_pmu_regs();
 }
 
-extern unsigned int *arm64_march_perfmap;
-
 static int __ihk_mc_perfctr_init(int counter, uint32_t type, uint64_t config, int mode)
 {
 	int ret = -1;
 	unsigned long config_base = 0;
-	int mapping;
 
-	mapping = cpu_pmu.map_event(type, config);
-	if (mapping < 0) {
-		return mapping;
-	}
-
-	ret = cpu_pmu.disable_counter(counter);
+	ret = cpu_pmu.disable_counter(1UL << counter);
 	if (ret < 0) {
 		return ret;
 	}
 
-	ret = cpu_pmu.enable_intens(counter);
+	ret = cpu_pmu.enable_intens(1UL << counter);
 	if (ret < 0) {
 		return ret;
 	}
@@ -112,7 +105,7 @@ static int __ihk_mc_perfctr_init(int counter, uint32_t type, uint64_t config, in
 	if (ret) {
 		return ret;
 	}
-	config_base |= (unsigned long)mapping;
+	config_base |= config;
 	cpu_pmu.write_evtype(counter, config_base);
 	return ret;
 }
@@ -124,68 +117,24 @@ int ihk_mc_perfctr_init_raw(int counter, uint64_t config, int mode)
 	return ret;
 }
 
-int ihk_mc_perfctr_init(int counter, uint64_t config, int mode)
-{
-	int ret;
-	ret = __ihk_mc_perfctr_init(counter, PERF_TYPE_RAW, config, mode);
-	return ret;
-}
-
 int ihk_mc_perfctr_start(unsigned long counter_mask)
 {
-	int ret = 0, i;
-
-	for (i = 0; i < sizeof(counter_mask) * BITS_PER_BYTE; i++) {
-		if (counter_mask & (1UL << i)) {
-			ret = cpu_pmu.enable_counter(i);
-			if (ret < 0) {
-				kprintf("%s: enable failed(idx=%d)\n",
-					__func__, i);
-				break;
-			}
-		}
-	}
-	return ret;
+	return cpu_pmu.enable_counter(counter_mask);
 }
 
 int ihk_mc_perfctr_stop(unsigned long counter_mask, int flags)
 {
-	int i = 0;
-
-	for (i = 0; i < sizeof(counter_mask) * BITS_PER_BYTE; i++) {
-		if (!(counter_mask & (1UL << i)))
-			continue;
-
-		int ret = 0;
-
-		ret = cpu_pmu.disable_counter(i);
-		if (ret < 0) {
-			continue;
-		}
-
-		if (flags & IHK_MC_PERFCTR_DISABLE_INTERRUPT) {
-			// when ihk_mc_perfctr_start is called,
-			// ihk_mc_perfctr_init is also called so disable
-			// interrupt
-			ret = cpu_pmu.disable_intens(i);
-			if (ret < 0) {
-				continue;
-			}
-		}
-	}
-	return 0;
+	return cpu_pmu.disable_counter(counter_mask);
 }
 
 int ihk_mc_perfctr_reset(int counter)
 {
-	// TODO[PMU]: ihk_mc_perfctr_setと同様にサンプリングレートの共通部実装の扱いを見てから本実装。
 	cpu_pmu.write_counter(counter, 0);
 	return 0;
 }
 
 int ihk_mc_perfctr_set(int counter, long val)
 {
-	// TODO[PMU]: 共通部でサンプリングレートの計算をして、設定するカウンタ値をvalに渡してくるようになると想定。サンプリングレートの扱いを見てから本実装。
 	uint32_t v = val;
 	cpu_pmu.write_counter(counter, v);
 	return 0;
@@ -198,6 +147,15 @@ int ihk_mc_perfctr_read_mask(unsigned long counter_mask, unsigned long *value)
 	return 0;
 }
 
+int ihk_mc_perfctr_alloc(struct thread *thread, struct mc_perf_event *event)
+{
+	const int counters = ihk_mc_perf_get_num_counters();
+
+	return cpu_pmu.get_event_idx(counters,
+				     thread->pmc_alloc_map,
+				     event->hw_config);
+}
+
 unsigned long ihk_mc_perfctr_read(int counter)
 {
 	unsigned long count;
@@ -205,6 +163,14 @@ unsigned long ihk_mc_perfctr_read(int counter)
 	return count;
 }
 
+unsigned long ihk_mc_perfctr_value(int counter, unsigned long correction)
+{
+	unsigned long count = ihk_mc_perfctr_read(counter) + correction;
+
+	count &= ((1UL << 32) - 1);
+	return count;
+}
+
 int ihk_mc_perfctr_alloc_counter(unsigned int *type, unsigned long *config,
 				 unsigned long pmc_status)
 {
@@ -234,12 +200,14 @@ int ihk_mc_perfctr_alloc_counter(unsigned int *type, unsigned long *config,
 
 int ihk_mc_perf_counter_mask_check(unsigned long counter_mask)
 {
-	return 1;
+	return cpu_pmu.counter_mask_valid(counter_mask);
 }
 
 int ihk_mc_perf_get_num_counters(void)
 {
-	return cpu_pmu.per_cpu[ihk_mc_get_processor_id()].num_events;
+	const struct per_cpu_arm_pmu *per_cpu_arm_pmu = get_per_cpu_pmu();
+
+	return per_cpu_arm_pmu->num_events;
 }
 
 int ihk_mc_perfctr_set_extra(struct mc_perf_event *event)
@@ -247,3 +215,83 @@ int ihk_mc_perfctr_set_extra(struct mc_perf_event *event)
 	/* Nothing to do. */
 	return 0;
 }
+
+static inline uint64_t arm_pmu_event_max_period(struct mc_perf_event *event)
+{
+	return 0xFFFFFFFF;
+}
+
+int hw_perf_event_init(struct mc_perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+
+	if (!is_sampling_event(event)) {
+		hwc->sample_period  = arm_pmu_event_max_period(event) >> 1;
+		hwc->last_period    = hwc->sample_period;
+		ihk_atomic64_set(&hwc->period_left, hwc->sample_period);
+	}
+	return 0;
+}
+
+int ihk_mc_event_set_period(struct mc_perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int64_t left = ihk_atomic64_read(&hwc->period_left);
+	int64_t period = hwc->sample_period;
+	uint64_t max_period;
+	int ret = 0;
+
+	max_period = arm_pmu_event_max_period(event);
+	if (unlikely(left <= -period)) {
+		left = period;
+		ihk_atomic64_set(&hwc->period_left, left);
+		hwc->last_period = period;
+		ret = 1;
+	}
+
+	if (unlikely(left <= 0)) {
+		left += period;
+		ihk_atomic64_set(&hwc->period_left, left);
+		hwc->last_period = period;
+		ret = 1;
+	}
+
+	/*
+	 * Limit the maximum period to prevent the counter value
+	 * from overtaking the one we are about to program. In
+	 * effect we are reducing max_period to account for
+	 * interrupt latency (and we are being very conservative).
+	 */
+	if (left > (max_period >> 1))
+		left = (max_period >> 1);
+
+	ihk_atomic64_set(&hwc->prev_count, (uint64_t)-left);
+
+	cpu_pmu.write_counter(event->counter_id,
+			      (uint64_t)(-left) & max_period);
+
+	return ret;
+}
+
+uint64_t ihk_mc_event_update(struct mc_perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int64_t delta;
+	uint64_t prev_raw_count, new_raw_count;
+	uint64_t max_period = arm_pmu_event_max_period(event);
+
+again:
+	prev_raw_count = ihk_atomic64_read(&hwc->prev_count);
+	new_raw_count = cpu_pmu.read_counter(event->counter_id);
+
+	if (ihk_atomic64_cmpxchg(&hwc->prev_count, prev_raw_count,
+			     new_raw_count) != prev_raw_count)
+		goto again;
+
+	delta = (new_raw_count - prev_raw_count) & max_period;
+
+	ihk_atomic64_add(delta, &event->count);
+	ihk_atomic64_add(-delta, &hwc->period_left);
+
+	return new_raw_count;
+}

arch/arm64/kernel/perfctr_armv8pmu.c

@@ -4,7 +4,6 @@
 #include <ihk/perfctr.h>
 #include <errno.h>
 #include <ihk/debug.h>
-#include <debug.h>
 #include <sysreg.h>
 #include <virt.h>
 #include <bitops.h>
@@ -21,29 +20,174 @@
 #define DDEBUG_DEFAULT DDEBUG_PRINT
 #endif
 
+/*
+ * read pmevcntr<n>_el0 functions
+ */
+#define read_pmevcntrN_el0(N) \
+static uint32_t read_pmevcntr##N##_el0(void) \
+{ \
+	return read_sysreg(pmevcntr##N##_el0); \
+}
+
+read_pmevcntrN_el0(0)
+read_pmevcntrN_el0(1)
+read_pmevcntrN_el0(2)
+read_pmevcntrN_el0(3)
+read_pmevcntrN_el0(4)
+read_pmevcntrN_el0(5)
+read_pmevcntrN_el0(6)
+read_pmevcntrN_el0(7)
+read_pmevcntrN_el0(8)
+read_pmevcntrN_el0(9)
+read_pmevcntrN_el0(10)
+read_pmevcntrN_el0(11)
+read_pmevcntrN_el0(12)
+read_pmevcntrN_el0(13)
+read_pmevcntrN_el0(14)
+read_pmevcntrN_el0(15)
+read_pmevcntrN_el0(16)
+read_pmevcntrN_el0(17)
+read_pmevcntrN_el0(18)
+read_pmevcntrN_el0(19)
+read_pmevcntrN_el0(20)
+read_pmevcntrN_el0(21)
+read_pmevcntrN_el0(22)
+read_pmevcntrN_el0(23)
+read_pmevcntrN_el0(24)
+read_pmevcntrN_el0(25)
+read_pmevcntrN_el0(26)
+read_pmevcntrN_el0(27)
+read_pmevcntrN_el0(28)
+read_pmevcntrN_el0(29)
+read_pmevcntrN_el0(30)
+
+static uint32_t (* const read_pmevcntr_el0[])(void) = {
+	read_pmevcntr0_el0, read_pmevcntr1_el0, read_pmevcntr2_el0,
+	read_pmevcntr3_el0, read_pmevcntr4_el0, read_pmevcntr5_el0,
+	read_pmevcntr6_el0, read_pmevcntr7_el0, read_pmevcntr8_el0,
+	read_pmevcntr9_el0, read_pmevcntr10_el0, read_pmevcntr11_el0,
+	read_pmevcntr12_el0, read_pmevcntr13_el0, read_pmevcntr14_el0,
+	read_pmevcntr15_el0, read_pmevcntr16_el0, read_pmevcntr17_el0,
+	read_pmevcntr18_el0, read_pmevcntr19_el0, read_pmevcntr20_el0,
+	read_pmevcntr21_el0, read_pmevcntr22_el0, read_pmevcntr23_el0,
+	read_pmevcntr24_el0, read_pmevcntr25_el0, read_pmevcntr26_el0,
+	read_pmevcntr27_el0, read_pmevcntr28_el0, read_pmevcntr29_el0,
+	read_pmevcntr30_el0,
+};
+
 
 /*
- * @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c
- * Perf Events' indices
+ * write pmevcntr<n>_el0 functions
  */
-#define	ARMV8_IDX_CYCLE_COUNTER	0
-#define	ARMV8_IDX_COUNTER0	1
-#define	ARMV8_IDX_COUNTER_LAST	(ARMV8_IDX_CYCLE_COUNTER + get_per_cpu_pmu()->num_events - 1)
+#define write_pmevcntrN_el0(N) \
+static void write_pmevcntr##N##_el0(uint32_t v) \
+{ \
+	write_sysreg(v, pmevcntr##N##_el0); \
+}
 
-/* @ref.impl linux-v4.15-rc3 arch/arm64/include/asm/perf_event.h */
-#define	ARMV8_PMU_MAX_COUNTERS	32
-#define	ARMV8_PMU_COUNTER_MASK	(ARMV8_PMU_MAX_COUNTERS - 1)
+write_pmevcntrN_el0(0)
+write_pmevcntrN_el0(1)
+write_pmevcntrN_el0(2)
+write_pmevcntrN_el0(3)
+write_pmevcntrN_el0(4)
+write_pmevcntrN_el0(5)
+write_pmevcntrN_el0(6)
+write_pmevcntrN_el0(7)
+write_pmevcntrN_el0(8)
+write_pmevcntrN_el0(9)
+write_pmevcntrN_el0(10)
+write_pmevcntrN_el0(11)
+write_pmevcntrN_el0(12)
+write_pmevcntrN_el0(13)
+write_pmevcntrN_el0(14)
+write_pmevcntrN_el0(15)
+write_pmevcntrN_el0(16)
+write_pmevcntrN_el0(17)
+write_pmevcntrN_el0(18)
+write_pmevcntrN_el0(19)
+write_pmevcntrN_el0(20)
+write_pmevcntrN_el0(21)
+write_pmevcntrN_el0(22)
+write_pmevcntrN_el0(23)
+write_pmevcntrN_el0(24)
+write_pmevcntrN_el0(25)
+write_pmevcntrN_el0(26)
+write_pmevcntrN_el0(27)
+write_pmevcntrN_el0(28)
+write_pmevcntrN_el0(29)
+write_pmevcntrN_el0(30)
+
+static void (* const write_pmevcntr_el0[])(uint32_t) = {
+	write_pmevcntr0_el0, write_pmevcntr1_el0, write_pmevcntr2_el0,
+	write_pmevcntr3_el0, write_pmevcntr4_el0, write_pmevcntr5_el0,
+	write_pmevcntr6_el0, write_pmevcntr7_el0, write_pmevcntr8_el0,
+	write_pmevcntr9_el0, write_pmevcntr10_el0, write_pmevcntr11_el0,
+	write_pmevcntr12_el0, write_pmevcntr13_el0, write_pmevcntr14_el0,
+	write_pmevcntr15_el0, write_pmevcntr16_el0, write_pmevcntr17_el0,
+	write_pmevcntr18_el0, write_pmevcntr19_el0, write_pmevcntr20_el0,
+	write_pmevcntr21_el0, write_pmevcntr22_el0, write_pmevcntr23_el0,
+	write_pmevcntr24_el0, write_pmevcntr25_el0, write_pmevcntr26_el0,
+	write_pmevcntr27_el0, write_pmevcntr28_el0, write_pmevcntr29_el0,
+	write_pmevcntr30_el0,
+};
 
 /*
- * ARMv8 low level PMU access
+ * write pmevtyper<n>_el0 functions
  */
+#define write_pmevtyperN_el0(N) \
+static void write_pmevtyper##N##_el0(uint32_t v) \
+{ \
+	write_sysreg(v, pmevtyper##N##_el0); \
+}
 
-/*
- * @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c
- * Perf Event to low level counters mapping
- */
-#define	ARMV8_IDX_TO_COUNTER(x)	\
-	(((x) - ARMV8_IDX_COUNTER0) & ARMV8_PMU_COUNTER_MASK)
+write_pmevtyperN_el0(0)
+write_pmevtyperN_el0(1)
+write_pmevtyperN_el0(2)
+write_pmevtyperN_el0(3)
+write_pmevtyperN_el0(4)
+write_pmevtyperN_el0(5)
+write_pmevtyperN_el0(6)
+write_pmevtyperN_el0(7)
+write_pmevtyperN_el0(8)
+write_pmevtyperN_el0(9)
+write_pmevtyperN_el0(10)
+write_pmevtyperN_el0(11)
+write_pmevtyperN_el0(12)
+write_pmevtyperN_el0(13)
+write_pmevtyperN_el0(14)
+write_pmevtyperN_el0(15)
+write_pmevtyperN_el0(16)
+write_pmevtyperN_el0(17)
+write_pmevtyperN_el0(18)
+write_pmevtyperN_el0(19)
+write_pmevtyperN_el0(20)
+write_pmevtyperN_el0(21)
+write_pmevtyperN_el0(22)
+write_pmevtyperN_el0(23)
+write_pmevtyperN_el0(24)
+write_pmevtyperN_el0(25)
+write_pmevtyperN_el0(26)
+write_pmevtyperN_el0(27)
+write_pmevtyperN_el0(28)
+write_pmevtyperN_el0(29)
+write_pmevtyperN_el0(30)
+
+static void (* const write_pmevtyper_el0[])(uint32_t) = {
+	write_pmevtyper0_el0, write_pmevtyper1_el0, write_pmevtyper2_el0,
+	write_pmevtyper3_el0, write_pmevtyper4_el0, write_pmevtyper5_el0,
+	write_pmevtyper6_el0, write_pmevtyper7_el0, write_pmevtyper8_el0,
+	write_pmevtyper9_el0, write_pmevtyper10_el0, write_pmevtyper11_el0,
+	write_pmevtyper12_el0, write_pmevtyper13_el0, write_pmevtyper14_el0,
+	write_pmevtyper15_el0, write_pmevtyper16_el0, write_pmevtyper17_el0,
+	write_pmevtyper18_el0, write_pmevtyper19_el0, write_pmevtyper20_el0,
+	write_pmevtyper21_el0, write_pmevtyper22_el0, write_pmevtyper23_el0,
+	write_pmevtyper24_el0, write_pmevtyper25_el0, write_pmevtyper26_el0,
+	write_pmevtyper27_el0, write_pmevtyper28_el0, write_pmevtyper29_el0,
+	write_pmevtyper30_el0,
+};
+
+#define	ARMV8_IDX_CYCLE_COUNTER	31
+#define	ARMV8_IDX_COUNTER0	0
 
 /*
  * @ref.impl linux-v4.15-rc3 arch/arm64/include/asm/perf_event.h
@@ -175,6 +319,10 @@
 
 /* PMUv3 HW events mapping. */
 
+/* disable -Woverride-init for the following initializations */
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Woverride-init"
+
 /*
  * @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c
  * ARMv8 Architectural defined events, not all of these may
@@ -220,6 +368,9 @@ static const unsigned armv8_pmuv3_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
 	[C(BPU)][C(OP_WRITE)][C(RESULT_MISS)]	= ARMV8_PMUV3_PERFCTR_BR_MIS_PRED,
 };
 
+/* restore warnings */
+#pragma GCC diagnostic pop
+
 /* @ref.impl linux-v4.15-rc3 drivers/perf/arm_pmu.c */
 static int
 armpmu_map_cache_event(const unsigned (*cache_map)
@@ -298,11 +449,25 @@ armpmu_map_event(uint32_t type, uint64_t config,
 	return -ENOENT;
 }
 
-/* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
+static inline int armv8pmu_counter_mask_valid(unsigned long counter_mask)
+{
+	int num;
+	unsigned long event;
+	unsigned long cycle;
+	unsigned long invalid_mask;
+
+	num = get_per_cpu_pmu()->num_events;
+	num--; /* Sub the CPU cycles counter */
+	event = ((1UL << num) - 1) << ARMV8_IDX_COUNTER0;
+	cycle = 1UL << ARMV8_IDX_CYCLE_COUNTER;
+	invalid_mask = ~(event | cycle);
+
+	return !(counter_mask & invalid_mask);
+}
+
 static inline int armv8pmu_counter_valid(int idx)
 {
-	return idx >= ARMV8_IDX_CYCLE_COUNTER &&
-		idx <= ARMV8_IDX_COUNTER_LAST;
+	return armv8pmu_counter_mask_valid(1UL << idx);
 }
 
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
@@ -326,6 +491,11 @@ static inline int armv8pmu_has_overflowed(uint32_t pmovsr)
 	return pmovsr & ARMV8_PMU_OVERFLOWED_MASK;
 }
 
+static inline int armv8pmu_counter_has_overflowed(uint32_t pmnc, int idx)
+{
+	return pmnc & BIT(idx);
+}
+
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
 static int __armv8_pmuv3_map_event(uint32_t type, uint64_t config,
 				   const unsigned int (*extra_event_map)
@@ -357,6 +527,23 @@ static int armv8_pmuv3_map_event(uint32_t type, uint64_t config)
 	return __armv8_pmuv3_map_event(type, config, NULL, NULL);
 }
 
+
+static int armv8_pmuv3_map_hw_event(uint64_t config)
+{
+	return __armv8_pmuv3_map_event(PERF_TYPE_HARDWARE, config, NULL, NULL);
+}
+
+
+static int armv8_pmuv3_map_cache_event(uint64_t config)
+{
+	return __armv8_pmuv3_map_event(PERF_TYPE_HW_CACHE, config, NULL, NULL);
+}
+
+static int armv8_pmuv3_map_raw_event(uint64_t config)
+{
+	return __armv8_pmuv3_map_event(PERF_TYPE_RAW, config, NULL, NULL);
+}
+
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
 static inline uint32_t armv8pmu_pmcr_read(void)
 {
@@ -371,24 +558,6 @@ static inline void armv8pmu_pmcr_write(uint32_t val)
 	write_sysreg(val, pmcr_el0);
 }
 
-/* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
-static inline int armv8pmu_select_counter(int idx)
-{
-	uint32_t counter;
-
-	if (!armv8pmu_counter_valid(idx)) {
-		ekprintf("%s: The count_register#%d is not implemented.\n",
-			__func__, idx);
-		return -EINVAL;
-	}
-
-	counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(counter, pmselr_el0);
-	isb();
-
-	return idx;
-}
-
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
 static inline uint32_t armv8pmu_read_counter(int idx)
 {
@@ -401,8 +570,8 @@ static inline uint32_t armv8pmu_read_counter(int idx)
 	else if (idx == ARMV8_IDX_CYCLE_COUNTER) {
 		value = read_sysreg(pmccntr_el0);
 	}
-	else if (armv8pmu_select_counter(idx) == idx) {
-		value = read_sysreg(pmxevcntr_el0);
+	else {
+		value = read_pmevcntr_el0[idx]();
 	}
 
 	return value;
@@ -421,43 +590,42 @@ static inline void armv8pmu_write_counter(int idx, uint32_t value)
 		 * count using the lower 32bits and we want an interrupt when
 		 * it overflows.
 		 */
-		uint64_t value64 = 0xffffffff00000000ULL | value;
+		uint64_t value64 = (int32_t)value;
 
 		write_sysreg(value64, pmccntr_el0);
 	}
-	else if (armv8pmu_select_counter(idx) == idx) {
-		write_sysreg(value, pmxevcntr_el0);
+	else {
+		write_pmevcntr_el0[idx](value);
 	}
 }
 
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
-static inline int armv8pmu_enable_intens(int idx)
+static inline int armv8pmu_enable_intens(unsigned long counter_mask)
 {
-	uint32_t counter;
-
-	if (!armv8pmu_counter_valid(idx)) {
-		ekprintf("%s: The count_register#%d is not implemented.\n",
-			__func__, idx);
+	if (!armv8pmu_counter_mask_valid(counter_mask)) {
+		ekprintf("%s: invalid counter mask(%#lx)\n",
+			__func__, counter_mask);
 		return -EINVAL;
 	}
 
-	counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmintenset_el1);
-	return idx;
+	write_sysreg(counter_mask, pmintenset_el1);
+	return 0;
 }
 
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
-static inline int armv8pmu_disable_intens(int idx)
+static inline int armv8pmu_disable_intens(unsigned long counter_mask)
 {
-	uint32_t counter = ARMV8_IDX_TO_COUNTER(idx);
-
-	write_sysreg(BIT(counter), pmintenclr_el1);
+	if (!armv8pmu_counter_mask_valid(counter_mask)) {
+		ekprintf("%s: invalid counter mask(%#lx)\n",
+			__func__, counter_mask);
+		return -EINVAL;
+	}
+	write_sysreg(counter_mask, pmintenclr_el1);
 	isb();
 	/* Clear the overflow flag in case an interrupt is pending. */
-	write_sysreg(BIT(counter), pmovsclr_el0);
+	write_sysreg(counter_mask, pmovsclr_el0);
 	isb();
-
-	return idx;
+	return 0;
 }
 
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
@@ -492,42 +660,37 @@ static int armv8pmu_set_event_filter(unsigned long *config_base, int mode)
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
 static inline void armv8pmu_write_evtype(int idx, uint32_t val)
 {
-	if (armv8pmu_select_counter(idx) == idx) {
-		val &= ARMV8_PMU_EVTYPE_MASK;
-		write_sysreg(val, pmxevtyper_el0);
+	if (!armv8pmu_counter_valid(idx)) {
+		ekprintf("%s: The count_register#%d is not implemented.\n",
+			 __func__, idx);
+		return;
+	} else if (idx != ARMV8_IDX_CYCLE_COUNTER) {
+		write_pmevtyper_el0[idx](val);
 	}
 }
 
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
-static inline int armv8pmu_enable_counter(int idx)
+static inline int armv8pmu_enable_counter(unsigned long counter_mask)
 {
-	uint32_t counter;
-
-	if (!armv8pmu_counter_valid(idx)) {
-		ekprintf("%s: The count_register#%d is not implemented.\n",
-			__func__, idx);
+	if (!armv8pmu_counter_mask_valid(counter_mask)) {
+		ekprintf("%s: invalid counter mask 0x%lx.\n",
+			 __func__, counter_mask);
 		return -EINVAL;
 	}
-
-	counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmcntenset_el0);
-	return idx;
+	write_sysreg(counter_mask, pmcntenset_el0);
+	return 0;
 }
 
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
-static inline int armv8pmu_disable_counter(int idx)
+static inline int armv8pmu_disable_counter(unsigned long counter_mask)
 {
-	uint32_t counter;
-
-	if (!armv8pmu_counter_valid(idx)) {
-		ekprintf("%s: The count_register#%d is not implemented.\n",
-			__func__, idx);
+	if (!armv8pmu_counter_mask_valid(counter_mask)) {
+		ekprintf("%s: invalid counter mask 0x%lx.\n",
+			 __func__, counter_mask);
 		return -EINVAL;
 	}
-
-	counter = ARMV8_IDX_TO_COUNTER(idx);
-	write_sysreg(BIT(counter), pmcntenclr_el0);
-	return idx;
+	write_sysreg(counter_mask, pmcntenclr_el0);
+	return 0;
 }
 
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
@@ -555,41 +718,20 @@ static void armv8pmu_stop(void)
 	ihk_mc_spinlock_unlock(&pmu_lock, flags);
 }
 
-/* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
-static void armv8pmu_disable_event(int idx)
-{
-	unsigned long flags;
-
-	/*
-	 * Disable counter and interrupt
-	 */
-	flags = ihk_mc_spinlock_lock(&pmu_lock);
-
-	/*
-	 * Disable counter
-	 */
-	armv8pmu_disable_counter(idx);
-
-	/*
-	 * Disable interrupt for this counter
-	 */
-	armv8pmu_disable_intens(idx);
-
-	ihk_mc_spinlock_unlock(&pmu_lock, flags);
-}
-
 /* @ref.impl linux-v4.15-rc3 arch/arm64/kernel/perf_event.c */
 static void armv8pmu_reset(void *info)
 {
 	struct arm_pmu *cpu_pmu = (struct arm_pmu *)info;
-	uint32_t idx, nb_cnt =
+	uint32_t nb_cnt =
 		cpu_pmu->per_cpu[ihk_mc_get_processor_id()].num_events;
+	nb_cnt--; /* Sub the CPU cycles counter */
+	unsigned long event = ((1UL << nb_cnt) - 1) << ARMV8_IDX_COUNTER0;
+	unsigned long cycle = 1UL << ARMV8_IDX_CYCLE_COUNTER;
+	unsigned long valid_mask = event | cycle;
 
 	/* The counter and interrupt enable registers are unknown at reset. */
-	for (idx = ARMV8_IDX_CYCLE_COUNTER; idx < nb_cnt; ++idx) {
-		armv8pmu_disable_counter(idx);
-		armv8pmu_disable_intens(idx);
-	}
+	armv8pmu_disable_counter(valid_mask);
+	armv8pmu_disable_intens(valid_mask);
 
 	/*
 	 * Initialize & Reset PMNC. Request overflow interrupt for
@@ -603,7 +745,7 @@ static void armv8pmu_reset(void *info)
 static int armv8pmu_get_event_idx(int num_events, unsigned long used_mask,
 				  unsigned long config)
 {
-	int idx;
+	int idx, end;
 	unsigned long evtype = config & ARMV8_PMU_EVTYPE_EVENT;
 
 	/* Always prefer to place a cycle counter into the cycle counter. */
@@ -615,7 +757,9 @@ static int armv8pmu_get_event_idx(int num_events, unsigned long used_mask,
 	/*
 	 * Otherwise use events counters
 	 */
-	for (idx = ARMV8_IDX_COUNTER0; idx < num_events; ++idx) {
+	end = ARMV8_IDX_COUNTER0 + num_events;
+	end--; /* Sub the CPU cycles counter */
+	for (idx = ARMV8_IDX_COUNTER0; idx < end; ++idx) {
 		if (!(used_mask & (1UL << idx)))
 			return idx;
 	}
@@ -642,13 +786,11 @@ static uint32_t armv8pmu_read_num_pmnc_events(void)
 
 static void armv8pmu_handle_irq(void *priv)
 {
-	struct siginfo info;
 	uint32_t pmovsr;
 	struct thread *thread = cpu_local_var(current);
 	struct process *proc = thread->proc;
-	long irqstate;
-	struct mckfd *fdp;
-	struct pt_regs *regs = (struct pt_regs *)priv;
+	const struct per_cpu_arm_pmu *cpu_pmu = get_per_cpu_pmu();
+	int idx;
 
 	/*
 	 * Get and reset the IRQ flags
@@ -661,27 +803,40 @@ static void armv8pmu_handle_irq(void *priv)
 	if (!armv8pmu_has_overflowed(pmovsr))
 		return;
 
+	if (!proc->monitoring_event) {
+		return;
+	}
 	/*
 	 * Handle the counter(s) overflow(s)
 	 */
-	/* same as x86_64 mckernel */
-	irqstate = ihk_mc_spinlock_lock(&proc->mckfd_lock);
-	for (fdp = proc->mckfd; fdp; fdp = fdp->next) {
-		if (fdp->sig_no > 0)
-			break;
-	}
-	ihk_mc_spinlock_unlock(&proc->mckfd_lock, irqstate);
+	for (idx = 0; idx < cpu_pmu->num_events; idx++) {
+		struct mc_perf_event *event = NULL;
+		struct mc_perf_event *sub;
 
-	if (fdp) {
-		memset(&info, '\0', sizeof(info));
-		info.si_signo = fdp->sig_no;
-		info._sifields._sigfault.si_addr = (void *)regs->pc;
-		info._sifields._sigpoll.si_fd = fdp->fd;
-		set_signal(fdp->sig_no, regs, &info);
-	}
-	else {
-		set_signal(SIGIO, regs, NULL);
+		if (!armv8pmu_counter_has_overflowed(pmovsr, idx)) {
+			continue;
+		}
+
+		if (proc->monitoring_event->counter_id == idx) {
+			event = proc->monitoring_event;
+		} else {
+			list_for_each_entry(sub,
+					&proc->monitoring_event->sibling_list,
+					group_entry) {
+				if (sub->counter_id == idx) {
+					event = sub;
+					break;
+				}
+			}
+		}
+
+		if (!event) {
+			continue;
+		}
+		ihk_mc_event_update(event);
+		ihk_mc_event_set_period(event);
 	}
+	return;
 }
 
 static void armv8pmu_enable_user_access_pmu_regs(void)
@@ -735,11 +890,15 @@ int armv8pmu_init(struct arm_pmu* cpu_pmu)
 	cpu_pmu->write_evtype = armv8pmu_write_evtype;
 	cpu_pmu->get_event_idx = armv8pmu_get_event_idx;
 	cpu_pmu->map_event = armv8_pmuv3_map_event;
+	cpu_pmu->map_hw_event = armv8_pmuv3_map_hw_event;
+	cpu_pmu->map_cache_event = armv8_pmuv3_map_cache_event;
+	cpu_pmu->map_raw_event = armv8_pmuv3_map_raw_event;
 	cpu_pmu->enable_user_access_pmu_regs =
 		armv8pmu_enable_user_access_pmu_regs;
 	cpu_pmu->disable_user_access_pmu_regs =
 		armv8pmu_disable_user_access_pmu_regs;
 	cpu_pmu->handler = &armv8pmu_handler;
+	cpu_pmu->counter_mask_valid = &armv8pmu_counter_mask_valid;
 	return 0;
 }
 

arch/arm64/kernel/psci.c

@@ -18,10 +18,9 @@
 #include <psci.h>
 #include <errno.h>
 #include <ihk/types.h>
-#include <ihk/debug.h>
 #include <compiler.h>
 #include <lwk/compiler.h>
-#include <debug.h>
+#include <ihk/debug.h>
 
 //#define DEBUG_PRINT_PSCI
 

arch/arm64/kernel/ptrace.c

@@ -1,4 +1,4 @@
-/* ptrace.c COPYRIGHT FUJITSU LIMITED 2016-2018 */
+/* ptrace.c COPYRIGHT FUJITSU LIMITED 2016-2019 */
 #include <errno.h>
 #include <debug-monitors.h>
 #include <hw_breakpoint.h>
@@ -11,7 +11,8 @@
 #include <hwcap.h>
 #include <string.h>
 #include <thread_info.h>
-#include <debug.h>
+#include <ptrace.h>
+#include <ihk/debug.h>
 
 //#define DEBUG_PRINT_SC
 
@@ -25,37 +26,6 @@
 extern void save_debugreg(unsigned long *debugreg);
 extern int interrupt_from_user(void *);
 
-enum aarch64_regset {
-	REGSET_GPR,
-	REGSET_FPR,
-	REGSET_TLS,
-	REGSET_HW_BREAK,
-	REGSET_HW_WATCH,
-	REGSET_SYSTEM_CALL,
-#ifdef CONFIG_ARM64_SVE
-	REGSET_SVE,
-#endif /* CONFIG_ARM64_SVE */
-};
-
-struct user_regset;
-typedef long user_regset_get_fn(struct thread *target,
-				const struct user_regset *regset,
-				unsigned int pos, unsigned int count,
-				void *kbuf, void __user *ubuf);
-
-typedef long user_regset_set_fn(struct thread *target,
-				const struct user_regset *regset,
-				unsigned int pos, unsigned int count,
-				const void *kbuf, const void __user *ubuf);
-
-struct user_regset {
-	user_regset_get_fn *get;
-	user_regset_set_fn *set;
-	unsigned int n;
-	unsigned int size;
-	unsigned int core_note_type;
-};
-
 long ptrace_read_user(struct thread *thread, long addr, unsigned long *value)
 {
 	return -EIO;
@@ -273,6 +243,17 @@ static inline long copy_regset_from_user(struct thread *target,
 	return regset->set(target, regset, offset, size, NULL, data);
 }
 
+unsigned int regset_size(struct thread *target,
+			 const struct user_regset *regset)
+{
+	if (!regset->get_size) {
+		return regset->n * regset->size;
+	}
+	else {
+		return regset->get_size(target, regset);
+	}
+}
+
 /*
  * Bits which are always architecturally RES0 per ARM DDI 0487A.h
  * Userspace cannot use these until they have an architectural meaning.
@@ -624,6 +605,48 @@ out:
 
 #ifdef CONFIG_ARM64_SVE
 
+static void sve_init_header_from_thread(struct user_sve_header *header,
+					struct thread *target)
+{
+	unsigned int vq;
+
+	memset(header, 0, sizeof(*header));
+
+	/* McKernel processes always enable SVE. */
+	header->flags = SVE_PT_REGS_SVE;
+
+	if (target->ctx.thread->sve_flags & SVE_PT_VL_INHERIT) {
+		header->flags |= SVE_PT_VL_INHERIT;
+	}
+
+	header->vl = target->ctx.thread->sve_vl;
+	vq = sve_vq_from_vl(header->vl);
+
+	header->max_vl = sve_max_vl;
+	header->size = SVE_PT_SIZE(vq, header->flags);
+	header->max_size = SVE_PT_SIZE(sve_vq_from_vl(header->max_vl),
+				       SVE_PT_REGS_SVE);
+}
+
+static unsigned int sve_size_from_header(struct user_sve_header const *header)
+{
+	return ALIGN(header->size, SVE_VQ_BYTES);
+}
+
+static unsigned int sve_get_size(struct thread *target,
+				 const struct user_regset *regset)
+{
+	struct user_sve_header header;
+
+	/* Instead of system_supports_sve() */
+	if (unlikely(!(elf_hwcap & HWCAP_SVE))) {
+		return 0;
+	}
+
+	sve_init_header_from_thread(&header, target);
+	return sve_size_from_header(&header);
+}
+
 /* read NT_ARM_SVE */
 static long sve_get(struct thread *target,
 		    const struct user_regset *regset,
@@ -646,23 +669,9 @@ static long sve_get(struct thread *target,
 	}
 
 	/* Header */
-	memset(&header, 0, sizeof(header));
-
-	header.vl = target->ctx.thread->sve_vl;
-
-	BUG_ON(!sve_vl_valid(header.vl));
+	sve_init_header_from_thread(&header, target);
 	vq = sve_vq_from_vl(header.vl);
 
-	BUG_ON(!sve_vl_valid(sve_max_vl));
-	header.max_vl = sve_max_vl;
-
-	/* McKernel processes always enable SVE. */
-	header.flags = SVE_PT_REGS_SVE;
-
-	header.size = SVE_PT_SIZE(vq, header.flags);
-	header.max_size = SVE_PT_SIZE(sve_vq_from_vl(header.max_vl),
-				      SVE_PT_REGS_SVE);
-
 	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf, &header,
 				  0, sizeof(header));
 	if (ret) {
@@ -676,11 +685,9 @@ static long sve_get(struct thread *target,
 	 */
 
 	/* Otherwise: full SVE case */
+
 	start = SVE_PT_SVE_OFFSET;
 	end = SVE_PT_SVE_FFR_OFFSET(vq) + SVE_PT_SVE_FFR_SIZE(vq);
-
-	BUG_ON(end < start);
-	BUG_ON(end - start > sve_state_size(target));
 	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf,
 				  target->ctx.thread->sve_state,
 				  start, end);
@@ -690,24 +697,18 @@ static long sve_get(struct thread *target,
 
 	start = end;
 	end = SVE_PT_SVE_FPSR_OFFSET(vq);
-
-	BUG_ON(end < start);
 	ret = user_regset_copyout_zero(&pos, &count, &kbuf, &ubuf,
 				       start, end);
 	if (ret) {
 		goto out;
 	}
 
+	/*
+	 * Copy fpsr, and fpcr which must follow contiguously in
+	 * struct fpsimd_state:
+	 */
 	start = end;
 	end = SVE_PT_SVE_FPCR_OFFSET(vq) + SVE_PT_SVE_FPCR_SIZE;
-
-	BUG_ON((char *)(&target->fp_regs->fpcr + 1) <
-	       (char *)&target->fp_regs->fpsr);
-	BUG_ON(end < start);
-	BUG_ON((char *)(&target->fp_regs->fpcr + 1) -
-	       (char *)&target->fp_regs->fpsr !=
-	        end - start);
-
 	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf,
 				  &target->fp_regs->fpsr,
 				  start, end);
@@ -716,9 +717,7 @@ static long sve_get(struct thread *target,
 	}
 
 	start = end;
-	end = (SVE_PT_SIZE(SVE_VQ_MAX, SVE_PT_REGS_SVE) + 15) / 16 * 16;
-
-	BUG_ON(end < start);
+	end = sve_size_from_header(&header);
 	ret = user_regset_copyout_zero(&pos, &count, &kbuf, &ubuf,
 				       start, end);
 out:
@@ -762,13 +761,12 @@ static long sve_set(struct thread *target,
 	 * sve_set_vector_length(), which will also validate them for us:
 	 */
 	ret = sve_set_vector_length(target, header.vl,
-				    header.flags & ~SVE_PT_REGS_MASK);
+		((unsigned long)header.flags & ~SVE_PT_REGS_MASK) << 16);
 	if (ret) {
 		goto out;
 	}
 
 	/* Actual VL set may be less than the user asked for: */
-	BUG_ON(!sve_vl_valid(target->ctx.thread->sve_vl));
 	vq = sve_vq_from_vl(target->ctx.thread->sve_vl);
 
 	/* Registers: FPSIMD-only case */
@@ -779,11 +777,19 @@ static long sve_set(struct thread *target,
 	}
 
 	/* Otherwise: full SVE case */
+
+	/*
+	 * If setting a different VL from the requested VL and there is
+	 * register data, the data layout will be wrong: don't even
+	 * try to set the registers in this case.
+	 */
+	if (count && vq != sve_vq_from_vl(header.vl)) {
+		ret = -EIO;
+		goto out;
+	}
+
 	start = SVE_PT_SVE_OFFSET;
 	end = SVE_PT_SVE_FFR_OFFSET(vq) + SVE_PT_SVE_FFR_SIZE(vq);
-
-	BUG_ON(end < start);
-	BUG_ON(end - start > sve_state_size(target));
 	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
 				 target->ctx.thread->sve_state,
 				 start, end);
@@ -793,27 +799,21 @@ static long sve_set(struct thread *target,
 
 	start = end;
 	end = SVE_PT_SVE_FPSR_OFFSET(vq);
-
-	BUG_ON(end < start);
 	ret = user_regset_copyin_ignore(&pos, &count, &kbuf, &ubuf,
 					start, end);
 	if (ret) {
 		goto out;
 	}
 
+	/*
+	 * Copy fpsr, and fpcr which must follow contiguously in
+	 * struct fpsimd_state:
+	 */
 	start = end;
 	end = SVE_PT_SVE_FPCR_OFFSET(vq) + SVE_PT_SVE_FPCR_SIZE;
-
-	BUG_ON((char *)(&target->fp_regs->fpcr + 1) <
-		(char *)&target->fp_regs->fpsr);
-	BUG_ON(end < start);
-	BUG_ON((char *)(&target->fp_regs->fpcr + 1) -
-		(char *)&target->fp_regs->fpsr !=
-		 end - start);
-
-	user_regset_copyin(&pos, &count, &kbuf, &ubuf,
-			   &target->fp_regs->fpsr,
-			   start, end);
+	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
+				 &target->fp_regs->fpsr,
+				 start, end);
 out:
 	return ret;
 }
@@ -825,8 +825,9 @@ static const struct user_regset aarch64_regsets[] = {
 		.core_note_type = NT_PRSTATUS,
 		.n = sizeof(struct user_pt_regs) / sizeof(uint64_t),
 		.size = sizeof(uint64_t),
+		.align = sizeof(uint64_t),
 		.get = gpr_get,
-		.set = gpr_set
+		.set = gpr_set,
 	},
 	[REGSET_FPR] = {
 		.core_note_type = NT_PRFPREG,
@@ -836,56 +837,75 @@ static const struct user_regset aarch64_regsets[] = {
 		 * fpcr are 32-bits wide.
 		 */
 		.size = sizeof(uint32_t),
+		.align = sizeof(uint32_t),
 		.get = fpr_get,
-		.set = fpr_set
+		.set = fpr_set,
 	},
 	[REGSET_TLS] = {
 		.core_note_type = NT_ARM_TLS,
 		.n = 1,
 		.size = sizeof(void *),
+		.align = sizeof(void *),
 		.get = tls_get,
-		.set = tls_set
+		.set = tls_set,
 	},
 	[REGSET_HW_BREAK] = {
 		.core_note_type = NT_ARM_HW_BREAK,
 		.n = sizeof(struct user_hwdebug_state) / sizeof(uint32_t),
 		.size = sizeof(uint32_t),
+		.align = sizeof(uint32_t),
 		.get = hw_break_get,
-		.set = hw_break_set
+		.set = hw_break_set,
 	},
 	[REGSET_HW_WATCH] = {
 		.core_note_type = NT_ARM_HW_WATCH,
 		.n = sizeof(struct user_hwdebug_state) / sizeof(uint32_t),
 		.size = sizeof(uint32_t),
+		.align = sizeof(uint32_t),
 		.get = hw_break_get,
-		.set = hw_break_set
+		.set = hw_break_set,
 	},
 	[REGSET_SYSTEM_CALL] = {
 		.core_note_type = NT_ARM_SYSTEM_CALL,
 		.n = 1,
 		.size = sizeof(int),
+		.align = sizeof(int),
 		.get = system_call_get,
-		.set = system_call_set
+		.set = system_call_set,
 	},
 #ifdef CONFIG_ARM64_SVE
 	[REGSET_SVE] = { /* Scalable Vector Extension */
 		.core_note_type = NT_ARM_SVE,
-		.n = (SVE_PT_SIZE(SVE_VQ_MAX, SVE_PT_REGS_SVE) + 15) / 16,
-		.size = 16,
+		.n = (SVE_PT_SIZE(SVE_VQ_MAX, SVE_PT_REGS_SVE) +
+			(SVE_VQ_BYTES - 1)) / SVE_VQ_BYTES,
+		.size = SVE_VQ_BYTES,
+		.align = SVE_VQ_BYTES,
 		.get = sve_get,
-		.set = sve_set
+		.set = sve_set,
+		.get_size = sve_get_size,
 	},
 #endif /* CONFIG_ARM64_SVE */
 };
 
-static const struct user_regset *
-find_regset(const struct user_regset *regset, unsigned int type, int n)
+static const struct user_regset_view user_aarch64_view = {
+	.name = "aarch64", .e_machine = EM_AARCH64,
+	.regsets = aarch64_regsets,
+	.n = sizeof(aarch64_regsets) / sizeof(aarch64_regsets[0])
+};
+
+const struct user_regset_view *current_user_regset_view(void)
+{
+	return &user_aarch64_view;
+}
+
+const struct user_regset *find_regset(const struct user_regset_view *view,
+				      unsigned int type)
 {
 	int i = 0;
 
-	for (i = 0; i < n; i++) {
-		if (regset[i].core_note_type == type) {
-			return &regset[i];
+	for (i = 0; i < view->n; i++) {
+		if (view->regsets[i].core_note_type == type) {
+			return &view->regsets[i];
 		}
 	}
 	return NULL;
@@ -894,8 +914,8 @@ find_regset(const struct user_regset *regset, unsigned int type, int n)
 static long ptrace_regset(struct thread *thread, int req, long type, struct iovec *iov)
 {
 	long rc = -EINVAL;
-	const struct user_regset *regset = find_regset(aarch64_regsets, type,
-					sizeof(aarch64_regsets) / sizeof(aarch64_regsets[0]));
+	const struct user_regset *regset =
+		find_regset(&user_aarch64_view, type);
 
 	if (!regset) {
 		kprintf("%s: not supported type 0x%x\n", __FUNCTION__, type);
@@ -944,6 +964,7 @@ void ptrace_report_signal(struct thread *thread, int sig)
 	/* save thread_info, if called by ptrace_report_exec() */
 	if (sig == ((SIGTRAP | (PTRACE_EVENT_EXEC << 8)))) {
 		memcpy(&tinfo, thread->ctx.thread, sizeof(struct thread_info));
+		thread->uctx->user_regs.regs[0] = 0;
 	}
 
 	mcs_rwlock_writer_lock(&proc->update_lock, &lock);
@@ -956,6 +977,13 @@ void ptrace_report_signal(struct thread *thread, int sig)
 	thread->exit_status = sig;
 	thread->status = PS_TRACED;
 	thread->ptrace &= ~PT_TRACE_SYSCALL;
+	if (sig == ((SIGTRAP | (PTRACE_EVENT_EXEC << 8))) &&
+				thread->ptrace & PTRACE_O_TRACEEXEC) {
+		/* PTRACE_O_TRACEEXEC: since Linux 3.0, the former
+		 * thread ID can be retrieved with PTRACE_GETEVENTMSG.
+		 * Report no change. */
+		thread->ptrace_eventmsg = thread->tid;
+	}
 	save_debugreg(thread->ptrace_debugreg);
 	if (sig == SIGSTOP || sig == SIGTSTP ||
 	    sig == SIGTTIN || sig == SIGTTOU) {
@@ -991,6 +1019,7 @@ void ptrace_report_signal(struct thread *thread, int sig)
 	if (sig == ((SIGTRAP | (PTRACE_EVENT_EXEC << 8)))) {
 		memcpy(thread->ctx.thread, &tinfo, sizeof(struct thread_info));
 	}
+	arch_flush_icache_all();
 }
 
 long

arch/arm64/kernel/syscall.c

@@ -1,4 +1,4 @@
-/* syscall.c COPYRIGHT FUJITSU LIMITED 2015-2018 */
+/* syscall.c COPYRIGHT FUJITSU LIMITED 2015-2019 */
 #include <cpulocal.h>
 #include <string.h>
 #include <kmalloc.h>
@@ -15,7 +15,8 @@
 #include <limits.h>
 #include <uio.h>
 #include <syscall.h>
-#include <debug.h>
+#include <rusage_private.h>
+#include <ihk/debug.h>
 
 void terminate_mcexec(int, int);
 extern void ptrace_report_signal(struct thread *thread, int sig);
@@ -42,7 +43,7 @@ uintptr_t debug_constants[] = {
 	offsetof(struct cpu_local_var, runq),
 	offsetof(struct cpu_local_var, status),
 	offsetof(struct cpu_local_var, idle),
-	offsetof(struct thread, ctx) + offsetof(struct thread_info, cpu_context),
+	offsetof(struct thread, ctx),
 	offsetof(struct thread, sched_list),
 	offsetof(struct thread, proc),
 	offsetof(struct thread, status),
@@ -56,13 +57,34 @@ extern int num_processors;
 int obtain_clone_cpuid(cpu_set_t *cpu_set, int use_last)
 {
 	int min_queue_len = -1;
-	int cpu, min_cpu = -1, uti_cpu = -1;
-	unsigned long irqstate;
+	int cpu, min_cpu = -1;
+#if 0
+	int uti_cpu = -1;
+#endif
+	unsigned long irqstate = 0;
 
-	irqstate = ihk_mc_spinlock_lock(&runq_reservation_lock);
+	int start, end, step;
+
+	if (use_last) {
+		start = num_processors - 1;
+		end = -1;
+		step = -1;
+	}
+	else {
+		start = 0;
+		end = num_processors;
+		step = 1;
+	}
+
+	if (!cpu_local_var(current)->proc->nr_processes) {
+		irqstate = ihk_mc_spinlock_lock(&runq_reservation_lock);
+	}
+	else {
+		irqstate = cpu_disable_interrupt_save();
+	}
 
 	/* Find the first allowed core with the shortest run queue */
-	for (cpu = 0; cpu < num_processors; ++cpu) {
+	for (cpu = start; cpu != end; cpu += step) {
 		struct cpu_local_var *v;
 
 		if (!CPU_ISSET(cpu, cpu_set))
@@ -73,11 +95,14 @@ int obtain_clone_cpuid(cpu_set_t *cpu_set, int use_last)
 		dkprintf("%s: cpu=%d,runq_len=%d,runq_reserved=%d\n",
 			 __func__, cpu, v->runq_len, v->runq_reserved);
 		if (min_queue_len == -1 ||
-		    v->runq_len + v->runq_reserved < min_queue_len) {
-			min_queue_len = v->runq_len + v->runq_reserved;
+		    //v->runq_len + v->runq_reserved < min_queue_len) {
+		    v->runq_len < min_queue_len) {
+			//min_queue_len = v->runq_len + v->runq_reserved;
+			min_queue_len = v->runq_len;
 			min_cpu = cpu;
 		}
 
+#if 0
 		/* Record the last tie CPU */
 		if (min_cpu != cpu &&
 		    v->runq_len + v->runq_reserved == min_queue_len) {
@@ -86,14 +111,15 @@ int obtain_clone_cpuid(cpu_set_t *cpu_set, int use_last)
 		dkprintf("%s: cpu=%d,runq_len=%d,runq_reserved=%d,min_cpu=%d,uti_cpu=%d\n",
 			 __func__, cpu, v->runq_len, v->runq_reserved,
 			 min_cpu, uti_cpu);
+#else
 
 		ihk_mc_spinlock_unlock_noirq(&v->runq_lock);
-#if 0
 		if (min_queue_len == 0)
 			break;
 #endif
 	}
 
+#if 0
 	min_cpu = use_last ? uti_cpu : min_cpu;
 	if (min_cpu != -1) {
 		if (get_cpu_local_var(min_cpu)->status != CPU_STATUS_RESERVED)
@@ -102,22 +128,20 @@ int obtain_clone_cpuid(cpu_set_t *cpu_set, int use_last)
 		__sync_fetch_and_add(&get_cpu_local_var(min_cpu)->runq_reserved,
 				     1);
 	}
-	ihk_mc_spinlock_unlock(&runq_reservation_lock, irqstate);
+#else
+	__sync_fetch_and_add(&get_cpu_local_var(min_cpu)->runq_reserved, 1);
+#endif
+
+	if (!cpu_local_var(current)->proc->nr_processes) {
+		ihk_mc_spinlock_unlock(&runq_reservation_lock, irqstate);
+	}
+	else {
+		cpu_restore_interrupt(irqstate);
+	}
 
 	return min_cpu;
 }
 
-int
-arch_clear_host_user_space()
-{
-	struct thread *th = cpu_local_var(current);
-
-	/* XXX: might be unnecessary */
-	clear_host_pte(th->vm->region.user_start,
-			(th->vm->region.user_end - th->vm->region.user_start));
-	return 0;
-}
-
 /* archtecture-depended syscall handlers */
 extern unsigned long do_fork(int clone_flags, unsigned long newsp,
 			     unsigned long parent_tidptr, unsigned long child_tidptr,
@@ -126,10 +150,18 @@ extern unsigned long do_fork(int clone_flags, unsigned long newsp,
 
 SYSCALL_DECLARE(clone)
 {
+	struct process *proc = cpu_local_var(current)->proc;
+	struct mcs_rwlock_node_irqsave lock_dump;
+	unsigned long ret;
+
+	/* mutex coredump */
+	mcs_rwlock_reader_lock(&proc->coredump_lock, &lock_dump);
+
 	if ((int)ihk_mc_syscall_arg0(ctx) & CLONE_VFORK) {
-		return do_fork(CLONE_VFORK|SIGCHLD, 0, 0, 0, 0, ihk_mc_syscall_pc(ctx), ihk_mc_syscall_sp(ctx));
+		ret = do_fork(CLONE_VFORK|SIGCHLD, 0, 0, 0, 0,
+				ihk_mc_syscall_pc(ctx), ihk_mc_syscall_sp(ctx));
 	} else {
-		return do_fork((int)ihk_mc_syscall_arg0(ctx),	/* clone_flags */
+		ret = do_fork((int)ihk_mc_syscall_arg0(ctx), /* clone_flags */
 			       ihk_mc_syscall_arg1(ctx),	/* newsp */
 			       ihk_mc_syscall_arg2(ctx),	/* parent_tidptr */
 			       ihk_mc_syscall_arg4(ctx),	/* child_tidptr (swap arg3) */
@@ -137,33 +169,9 @@ SYSCALL_DECLARE(clone)
 			       ihk_mc_syscall_pc(ctx),		/* curpc */
 			       ihk_mc_syscall_sp(ctx));		/* cursp */
 	}
-}
+	mcs_rwlock_reader_unlock(&proc->coredump_lock, &lock_dump);
 
-SYSCALL_DECLARE(rt_sigaction)
-{
-	int sig = ihk_mc_syscall_arg0(ctx);
-	const struct sigaction *act = (const struct sigaction *)ihk_mc_syscall_arg1(ctx);
-	struct sigaction *oact = (struct sigaction *)ihk_mc_syscall_arg2(ctx);
-	size_t sigsetsize = ihk_mc_syscall_arg3(ctx);
-	struct k_sigaction new_sa, old_sa;
-	int rc;
-
-	if (sigsetsize != sizeof(sigset_t))
-		return -EINVAL;
-
-	if(act)
-		if(copy_from_user(&new_sa.sa, act, sizeof new_sa.sa)){
-			goto fault;
-		}
-	rc = do_sigaction(sig, act? &new_sa: NULL, oact? &old_sa: NULL);
-	if(rc == 0 && oact)
-		if(copy_to_user(oact, &old_sa.sa, sizeof old_sa.sa)){
-			goto fault;
-		}
-
-	return rc;
-fault:
-	return -EFAULT;
+	return ret;
 }
 
 SYSCALL_DECLARE(prctl)
@@ -178,11 +186,10 @@ SYSCALL_DECLARE(prctl)
 
 	switch (option) {
 	case PR_SVE_SET_VL:
-		error = SVE_SET_VL(cpu_local_var(current),
-				ihk_mc_syscall_arg1(ctx), ihk_mc_syscall_arg2(ctx));
+		error = SVE_SET_VL(ihk_mc_syscall_arg1(ctx));
 		break;
 	case PR_SVE_GET_VL:
-		error = SVE_GET_VL(cpu_local_var(current));
+		error = SVE_GET_VL();
 		break;
 	case PR_SET_THP_DISABLE:
 		if (arg3 || arg4 || arg5) {
@@ -657,7 +664,7 @@ void set_single_step(struct thread *thread)
 	set_regs_spsr_ss(thread->uctx);
 }
 
-extern void coredump(struct thread *thread, void *regs);
+extern int coredump(struct thread *thread, void *regs, int sig);
 
 static int
 isrestart(int syscallno, unsigned long rc, int sig, int restart)
@@ -1096,6 +1103,7 @@ do_signal(unsigned long rc, void *regs0, struct thread *thread, struct sig_pendi
 	struct mcs_rwlock_node_irqsave lock;
 	struct mcs_rwlock_node_irqsave mcs_rw_node;
 	int restart = 0;
+	int ret;
 
 	for(w = pending->sigmask.__val[0], sig = 0; w; sig++, w >>= 1);
 	dkprintf("do_signal(): tid=%d, pid=%d, sig=%d\n", thread->tid, proc->pid, sig);
@@ -1270,15 +1278,6 @@ do_signal(unsigned long rc, void *regs0, struct thread *thread, struct sig_pendi
 			dkprintf("SIGTRAP(): woken up\n");
 			break;
 		case SIGCONT:
-			memset(&info, '\0', sizeof info);
-			info.si_signo = SIGCHLD;
-			info.si_code = CLD_CONTINUED;
-			info._sifields._sigchld.si_pid = proc->pid;
-			info._sifields._sigchld.si_status = 0x0000ffff;
-			do_kill(cpu_local_var(current), proc->parent->pid, -1, SIGCHLD, &info, 0);
-			proc->main_thread->signal_flags = SIGNAL_STOP_CONTINUED;
-			proc->status = PS_RUNNING;
-			dkprintf("do_signal,SIGCONT,do nothing\n");
 			break;
 		case SIGQUIT:
 		case SIGILL:
@@ -1290,9 +1289,31 @@ do_signal(unsigned long rc, void *regs0, struct thread *thread, struct sig_pendi
 		case SIGXCPU:
 		case SIGXFSZ:
 		core:
-			dkprintf("do_signal,default,core,sig=%d\n", sig);
-			coredump(thread, regs);
-			coredumped = 0x80;
+			thread->coredump_regs =
+				kmalloc(sizeof(struct pt_regs),
+					IHK_MC_AP_NOWAIT);
+			if (!thread->coredump_regs) {
+				kprintf("%s: Out of memory\n", __func__);
+				goto skip;
+			}
+			memcpy(thread->coredump_regs, regs,
+			       sizeof(struct pt_regs));
+
+			ret = coredump(thread, regs, sig);
+			switch (ret) {
+			case -EBUSY:
+				kprintf("%s: INFO: coredump not performed, try ulimit -c <non-zero>\n",
+					__func__);
+				break;
+			case 0:
+				coredumped = 0x80;
+				break;
+			default:
+				kprintf("%s: ERROR: coredump failed (%d)\n",
+					__func__, ret);
+				break;
+			}
+skip:
 			terminate(0, sig | coredumped);
 			break;
 		case SIGCHLD:
@@ -1309,70 +1330,6 @@ out:
 	return restart;
 }
 
-static struct sig_pending *
-getsigpending(struct thread *thread, int delflag){
-	struct list_head *head;
-	mcs_rwlock_lock_t *lock;
-	struct mcs_rwlock_node_irqsave mcs_rw_node;
-	struct sig_pending *next;
-	struct sig_pending *pending;
-	__sigset_t w;
-
-	w = thread->sigmask.__val[0];
-
-	lock = &thread->sigcommon->lock;
-	head = &thread->sigcommon->sigpending;
-	for(;;) {
-		if (delflag) {
-			mcs_rwlock_writer_lock(lock, &mcs_rw_node);
-		}
-		else {
-			mcs_rwlock_reader_lock(lock, &mcs_rw_node);
-		}
-
-		list_for_each_entry_safe(pending, next, head, list){
-			if(!(pending->sigmask.__val[0] & w)){
-				if(delflag) 
-					list_del(&pending->list);
-
-				if (delflag) {
-					mcs_rwlock_writer_unlock(lock, &mcs_rw_node);
-				}
-				else {
-					mcs_rwlock_reader_unlock(lock, &mcs_rw_node);
-				}
-				return pending;
-			}
-		}
-
-		if (delflag) {
-			mcs_rwlock_writer_unlock(lock, &mcs_rw_node);
-		}
-		else {
-			mcs_rwlock_reader_unlock(lock, &mcs_rw_node);
-		}
-
-		if(lock == &thread->sigpendinglock)
-			return NULL;
-
-		lock = &thread->sigpendinglock;
-		head = &thread->sigpending;
-	}
-
-	return NULL;
-}
-
-struct sig_pending *
-hassigpending(struct thread *thread)
-{
-	if (list_empty(&thread->sigpending) &&
-	    list_empty(&thread->sigcommon->sigpending)) {
-		return NULL;
-	}
-
-	return getsigpending(thread, 0);
-}
-
 int
 interrupt_from_user(void *regs0)
 {
@@ -1396,185 +1353,6 @@ void save_syscall_return_value(int num, unsigned long rc)
 	}
 }
 
-void
-check_signal(unsigned long rc, void *regs0, int num)
-{
-	__check_signal(rc, regs0, num, 0);
-}
-
-void
-check_signal_irq_disabled(unsigned long rc, void *regs0, int num)
-{
-	__check_signal(rc, regs0, num, 1);
-}
-
-static void
-__check_signal(unsigned long rc, void *regs0, int num, int irq_disabled)
-{
-	ihk_mc_user_context_t *regs = regs0;
-	struct thread *thread;
-	struct sig_pending *pending;
-	int	irqstate;
-
-	if(clv == NULL)
-		return;
-	thread = cpu_local_var(current);
-
-	if(thread == NULL || thread->proc->pid == 0){
-		struct thread *t;
-		irqstate = ihk_mc_spinlock_lock(&(cpu_local_var(runq_lock)));
-		list_for_each_entry(t, &(cpu_local_var(runq)), sched_list){
-			if(t->proc->pid <= 0)
-				continue;
-			if(t->status == PS_INTERRUPTIBLE &&
-			   hassigpending(t)){
-				t->status = PS_RUNNING;
-				break;
-			}
-		}
-		ihk_mc_spinlock_unlock(&(cpu_local_var(runq_lock)), irqstate);
-		goto out;
-	}
-
-	if(regs != NULL && !interrupt_from_user(regs)) {
-		goto out;
-	}
-
-	if (list_empty(&thread->sigpending) &&
-	    list_empty(&thread->sigcommon->sigpending)) {
-		goto out;
-	}
-
-	for(;;){
-		/* When this function called from check_signal_irq_disabled,
-		 * return with interrupt invalid.
-		 * This is to eliminate signal loss.
-		 */
-		if (irq_disabled == 1) {
-			irqstate = cpu_disable_interrupt_save();
-		}
-		pending = getsigpending(thread, 1);
-		if(!pending) {
-			dkprintf("check_signal,queue is empty\n");
-			goto out;
-		}
-		if (irq_disabled == 1) {
-			cpu_restore_interrupt(irqstate);
-		}
-		if (do_signal(rc, regs, thread, pending, num)) {
-			num = -1;
-		}
-	}
-
-out:
-	return;
-}
-
-static int
-check_sig_pending_thread(struct thread *thread)
-{
-	int found = 0;
-	struct list_head *head;
-	mcs_rwlock_lock_t *lock;
-	struct mcs_rwlock_node_irqsave mcs_rw_node;
-	struct sig_pending *next;
-	struct sig_pending *pending;
-	__sigset_t w;
-	__sigset_t x;
-	int sig = 0;
-	struct k_sigaction *k;
-	struct cpu_local_var *v;
-
-	v = get_this_cpu_local_var();
-	w = thread->sigmask.__val[0];
-
-	lock = &thread->sigcommon->lock;
-	head = &thread->sigcommon->sigpending;
-	for (;;) {
-		mcs_rwlock_reader_lock(lock, &mcs_rw_node);
-
-		list_for_each_entry_safe(pending, next, head, list) {
-			for (x = pending->sigmask.__val[0], sig = 0; x;
-			     sig++, x >>= 1)
-				;
-			k = thread->sigcommon->action + sig - 1;
-			if ((sig != SIGCHLD && sig != SIGURG) ||
-			    (k->sa.sa_handler != SIG_IGN &&
-			     k->sa.sa_handler != NULL)) {
-				if (!(pending->sigmask.__val[0] & w)) {
-					if (pending->interrupted == 0) {
-						pending->interrupted = 1;
-						found = 1;
-						if (sig != SIGCHLD &&
-						    sig != SIGURG &&
-						    !k->sa.sa_handler) {
-							found = 2;
-							break;
-						}
-					}
-				}
-			}
-		}
-
-		mcs_rwlock_reader_unlock(lock, &mcs_rw_node);
-
-		if (found == 2) {
-			break;
-		}
-
-		if (lock == &thread->sigpendinglock) {
-			break;
-		}
-
-		lock = &thread->sigpendinglock;
-		head = &thread->sigpending;
-	}
-
-	if (found == 2) {
-		ihk_mc_spinlock_unlock(&v->runq_lock, v->runq_irqstate);
-		terminate_mcexec(0, sig);
-		return 1;
-	}
-	else if (found == 1) {
-		ihk_mc_spinlock_unlock(&v->runq_lock, v->runq_irqstate);
-		interrupt_syscall(thread, 0);
-		return 1;
-	}
-	return 0;
-}
-
-void
-check_sig_pending(void)
-{
-	struct thread *thread;
-	struct cpu_local_var *v;
-
-	if (clv == NULL)
-		return;
-
-	v = get_this_cpu_local_var();
-repeat:
-	v->runq_irqstate = ihk_mc_spinlock_lock(&v->runq_lock);
-	list_for_each_entry(thread, &(v->runq), sched_list) {
-
-		if (thread == NULL || thread == &cpu_local_var(idle)) {
-			continue;
-		}
-
-		if (thread->in_syscall_offload == 0) {
-			continue;
-		}
-
-		if (thread->proc->group_exit_status & 0x0000000100000000L) {
-			continue;
-		}
-
-		if (check_sig_pending_thread(thread))
-			goto repeat;
-	}
-	ihk_mc_spinlock_unlock(&v->runq_lock, v->runq_irqstate);
-}
-
 unsigned long
 do_kill(struct thread * thread, int pid, int tid, int sig, siginfo_t *info, int ptracecont)
 {
@@ -1590,7 +1368,6 @@ do_kill(struct thread * thread, int pid, int tid, int sig, siginfo_t *info, int
 	struct list_head *head = NULL;
 	int rc;
 	unsigned long irqstate = 0;
-	struct k_sigaction *k;
 	int doint;
 	int found = 0;
 	siginfo_t info0;
@@ -1600,6 +1377,7 @@ do_kill(struct thread * thread, int pid, int tid, int sig, siginfo_t *info, int
 	struct process_hash *phash = rset->process_hash;
 	struct mcs_rwlock_node lock;
 	struct mcs_rwlock_node updatelock;
+	struct sig_pending *pending = NULL;
 
 	if(sig > SIGRTMAX || sig < 0)
 		return -EINVAL;
@@ -1786,47 +1564,61 @@ done:
 
 	mcs_rwlock_writer_lock_noirq(savelock, &mcs_rw_node);
 
-	/* Put signal event even when handler is SIG_IGN or SIG_DFL
-	   because target ptraced thread must call ptrace_report_signal 
-	   in check_signal */
 	rc = 0;
-	k = tthread->sigcommon->action + sig - 1;
-	if ((sig != SIGKILL && (tthread->ptrace & PT_TRACED)) ||
-	    (k->sa.sa_handler != SIG_IGN &&
-	     (k->sa.sa_handler != NULL ||
-	      (sig != SIGCHLD && sig != SIGURG)))) {
-		struct sig_pending *pending = NULL;
-		if (sig < SIGRTMIN) { // SIGRTMIN - SIGRTMAX
-			list_for_each_entry(pending, head, list){
-				if(pending->sigmask.__val[0] == mask &&
-				   pending->ptracecont == ptracecont)
-					break;
-			}
-			if(&pending->list == head)
-				pending = NULL;
+
+	if (sig < SIGRTMIN) { // SIGRTMIN - SIGRTMAX
+		list_for_each_entry(pending, head, list) {
+			if (pending->sigmask.__val[0] == mask &&
+			    pending->ptracecont == ptracecont)
+				break;
 		}
-		if(pending == NULL){
-			doint = 1;
-			pending = kmalloc(sizeof(struct sig_pending), IHK_MC_AP_NOWAIT);
-			if(!pending){
-				rc = -ENOMEM;
-			}
-			else{
-				memset(pending, 0, sizeof(struct sig_pending));
-				pending->sigmask.__val[0] = mask;
-				memcpy(&pending->info, info, sizeof(siginfo_t));
-				pending->ptracecont = ptracecont;
-				if(sig == SIGKILL || sig == SIGSTOP)
-					list_add(&pending->list, head);
-				else
-					list_add_tail(&pending->list, head);
-				tthread->sigevent = 1;
-			}
+		if (&pending->list == head)
+			pending = NULL;
+	}
+	if (pending == NULL) {
+		doint = 1;
+		pending = kmalloc(sizeof(struct sig_pending), IHK_MC_AP_NOWAIT);
+		if (!pending) {
+			rc = -ENOMEM;
+		}
+		else {
+			memset(pending, 0, sizeof(struct sig_pending));
+			pending->sigmask.__val[0] = mask;
+			memcpy(&pending->info, info, sizeof(siginfo_t));
+			pending->ptracecont = ptracecont;
+			if (sig == SIGKILL || sig == SIGSTOP)
+				list_add(&pending->list, head);
+			else
+				list_add_tail(&pending->list, head);
+			tthread->sigevent = 1;
 		}
 	}
+
 	mcs_rwlock_writer_unlock_noirq(savelock, &mcs_rw_node);
 	cpu_restore_interrupt(irqstate);
 
+	if (sig == SIGCONT || ptracecont == 1) {
+		/* Wake up the target only when stopped by SIGSTOP */
+		if (sched_wakeup_thread(tthread, PS_STOPPED) == 0) {
+			struct siginfo info;
+
+			tthread->proc->main_thread->signal_flags =
+							SIGNAL_STOP_CONTINUED;
+			tthread->proc->status = PS_RUNNING;
+			memset(&info, '\0', sizeof(info));
+			info.si_signo = SIGCHLD;
+			info.si_code = CLD_CONTINUED;
+			info._sifields._sigchld.si_pid = tthread->proc->pid;
+			info._sifields._sigchld.si_status = 0x0000ffff;
+			do_kill(tthread, tthread->proc->parent->pid, -1,
+							SIGCHLD, &info, 0);
+			if (thread != tthread) {
+				ihk_mc_interrupt_cpu(tthread->cpu_id,
+						ihk_mc_get_vector(IHK_GV_IKC));
+			}
+			doint = 0;
+		}
+	}
 	if (doint && !(mask & tthread->sigmask.__val[0])) {
 		int status = tthread->status;
 
@@ -1841,11 +1633,6 @@ done:
 				/* Wake up the target only when stopped by ptrace-reporting */
 				sched_wakeup_thread(tthread, PS_TRACED | PS_STOPPED | PS_INTERRUPTIBLE);
 			}
-			else if(sig == SIGCONT || ptracecont == 1){
-				/* Wake up the target only when stopped by SIGSTOP */
-				sched_wakeup_thread(tthread, PS_STOPPED);
-				tthread->proc->status = PS_RUNNING;
-			}
 			else {
 				sched_wakeup_thread(tthread, PS_INTERRUPTIBLE);
 			}
@@ -1870,7 +1657,7 @@ set_signal(int sig, void *regs0, siginfo_t *info)
 	}
 
 	if ((__sigmask(sig) & thread->sigmask.__val[0])) {
-		coredump(thread, regs0);
+		coredump(thread, regs0, sig);
 		terminate(0, sig | 0x80);
 	}
 	do_kill(thread, thread->proc->pid, thread->tid, sig, info, 0);
@@ -1900,7 +1687,7 @@ SYSCALL_DECLARE(mmap)
 		;
 
 	const uintptr_t addr0 = ihk_mc_syscall_arg0(ctx);
-	const size_t len0 = ihk_mc_syscall_arg1(ctx);
+	size_t len0 = ihk_mc_syscall_arg1(ctx);
 	const int prot = ihk_mc_syscall_arg2(ctx);
 	const int flags0 = ihk_mc_syscall_arg3(ctx);
 	const int fd = ihk_mc_syscall_arg4(ctx);
@@ -1941,7 +1728,8 @@ SYSCALL_DECLARE(mmap)
 
 		if (hugeshift == 0) {
 			/* default hugepage size */
-			flags |= MAP_HUGE_SECOND_BLOCK;
+			flags |= ihk_mc_get_linux_default_huge_page_shift() <<
+				MAP_HUGE_SHIFT;
 		} else if ((first_level_block_support &&
 				hugeshift == MAP_HUGE_FIRST_BLOCK) ||
 			   (first_level_block_support &&
@@ -1958,6 +1746,14 @@ SYSCALL_DECLARE(mmap)
 			goto out;
 		}
 		pgsize = (size_t)1 << ((flags >> MAP_HUGE_SHIFT) & 0x3F);
+		/* Round-up map length by pagesize */
+		len0 = ALIGN(len0, pgsize);
+
+		if (rusage_check_overmap(len0,
+				(flags >> MAP_HUGE_SHIFT) & 0x3F)) {
+			error = -ENOMEM;
+			goto out;
+		}
 	}
 
 #define	VALID_DUMMY_ADDR	((region->user_start + PTL3_SIZE - 1) & ~(PTL3_SIZE - 1))
@@ -2018,7 +1814,8 @@ SYSCALL_DECLARE(shmget)
 
 		if (hugeshift == 0) {
 			/* default hugepage size */
-			shmflg |= SHM_HUGE_SECOND_BLOCK;
+			shmflg |= ihk_mc_get_linux_default_huge_page_shift() <<
+				MAP_HUGE_SHIFT;
 		} else if ((first_level_block_support &&
 				hugeshift == SHM_HUGE_FIRST_BLOCK) ||
 			   (first_level_block_support &&
@@ -2082,11 +1879,13 @@ int do_process_vm_read_writev(int pid,
 	struct process *rproc;
 	struct process *lproc = lthread->proc;
 	struct process_vm *rvm = NULL;
-	unsigned long rphys;
-	unsigned long rpage_left;
-	unsigned long psize;
-	void *rva;
+	unsigned long lphys, rphys;
+	unsigned long lpage_left, rpage_left;
+	unsigned long lpsize, rpsize;
+	void *rva, *lva;
+#if 0
 	struct vm_range *range;
+#endif
 	struct mcs_rwlock_node_irqsave lock;
 	struct mcs_rwlock_node update_lock;
 
@@ -2099,8 +1898,9 @@ int do_process_vm_read_writev(int pid,
 		return -EINVAL;
 	}
 
+#if 0
 	/* Check if parameters are okay */
-	ihk_mc_spinlock_lock_noirq(&lthread->vm->memory_range_lock);
+	ihk_rwspinlock_read_lock_noirq(&lthread->vm->memory_range_lock);
 
 	range = lookup_process_memory_range(lthread->vm, 
 			(uintptr_t)local_iov, 
@@ -2122,11 +1922,12 @@ int do_process_vm_read_writev(int pid,
 
 	ret = 0;
 arg_out:
-	ihk_mc_spinlock_unlock_noirq(&lthread->vm->memory_range_lock);
+	ihk_rwspinlock_read_unlock_noirq(&lthread->vm->memory_range_lock);
 
 	if (ret != 0) {
 		goto out;
 	}
+#endif
 
 	for (li = 0; li < liovcnt; ++li) {
 		llen += local_iov[li].iov_len;
@@ -2191,7 +1992,7 @@ arg_out:
 		if (pli != li) {
 			struct vm_range *range;
 
-			ihk_mc_spinlock_lock_noirq(&lthread->vm->memory_range_lock);
+			ihk_rwspinlock_read_lock_noirq(&lthread->vm->memory_range_lock);
 
 			/* Is base valid? */
 			range = lookup_process_memory_range(lthread->vm,
@@ -2221,7 +2022,7 @@ arg_out:
 
 			ret = 0;
 pli_out:
-			ihk_mc_spinlock_unlock_noirq(&lthread->vm->memory_range_lock);
+			ihk_rwspinlock_read_unlock_noirq(&lthread->vm->memory_range_lock);
 
 			if (ret != 0) {
 				goto out;
@@ -2234,7 +2035,7 @@ pli_out:
 		if (pri != ri) {
 			struct vm_range *range;
 
-			ihk_mc_spinlock_lock_noirq(&rvm->memory_range_lock);
+			ihk_rwspinlock_read_lock_noirq(&rvm->memory_range_lock);
 
 			/* Is base valid? */
 			range = lookup_process_memory_range(rvm,
@@ -2264,7 +2065,7 @@ pli_out:
 
 			ret = 0;
 pri_out:
-			ihk_mc_spinlock_unlock_noirq(&rvm->memory_range_lock);
+			ihk_rwspinlock_read_unlock_noirq(&rvm->memory_range_lock);
 
 			if (ret != 0) {
 				goto out;
@@ -2279,10 +2080,53 @@ pri_out:
 			to_copy = remote_iov[ri].iov_len - roff;
 		}
 
-retry_lookup:
+retry_llookup:
+		/* Figure out local physical */
+		/* TODO: remember page and do this only if necessary */
+		ret = ihk_mc_pt_virt_to_phys_size(lthread->vm->address_space->page_table,
+				local_iov[li].iov_base + loff, &lphys, &lpsize);
+
+		if (ret) {
+			uint64_t reason = PF_POPULATE | PF_WRITE | PF_USER;
+			void *addr;
+
+			if (faulted) {
+				ret = -EFAULT;
+				goto out;
+			}
+
+			/* Fault in pages */
+			for (addr = (void *)
+					(((unsigned long)local_iov[li].iov_base + loff)
+					& PAGE_MASK);
+					addr < (local_iov[li].iov_base + loff + to_copy);
+					addr += PAGE_SIZE) {
+
+				ret = page_fault_process_vm(lthread->vm, addr, reason);
+				if (ret) {
+					ret = -EFAULT;
+					goto out;
+				}
+			}
+
+			faulted = 1;
+			goto retry_llookup;
+		}
+
+		lpage_left = ((((unsigned long)local_iov[li].iov_base + loff +
+			lpsize) & ~(lpsize - 1)) -
+			((unsigned long)local_iov[li].iov_base + loff));
+		if (lpage_left < to_copy) {
+			to_copy = lpage_left;
+		}
+
+		lva = phys_to_virt(lphys);
+
+retry_rlookup:
+		/* Figure out remote physical */
 		/* TODO: remember page and do this only if necessary */
 		ret = ihk_mc_pt_virt_to_phys_size(rvm->address_space->page_table,
-				remote_iov[ri].iov_base + roff, &rphys, &psize);
+				remote_iov[ri].iov_base + roff, &rphys, &rpsize);
 
 		if (ret) {
 			uint64_t reason = PF_POPULATE | PF_WRITE | PF_USER;
@@ -2308,11 +2152,11 @@ retry_lookup:
 			}
 
 			faulted = 1;
-			goto retry_lookup;
+			goto retry_rlookup;
 		}
 
 		rpage_left = ((((unsigned long)remote_iov[ri].iov_base + roff +
-			psize) & ~(psize - 1)) -
+			rpsize) & ~(rpsize - 1)) -
 			((unsigned long)remote_iov[ri].iov_base + roff));
 		if (rpage_left < to_copy) {
 			to_copy = rpage_left;
@@ -2321,16 +2165,16 @@ retry_lookup:
 		rva = phys_to_virt(rphys);
 
 		fast_memcpy(
-				(op == PROCESS_VM_READ) ? local_iov[li].iov_base + loff : rva,
-				(op == PROCESS_VM_READ) ? rva : local_iov[li].iov_base + loff,
+				(op == PROCESS_VM_READ) ? lva : rva,
+				(op == PROCESS_VM_READ) ? rva : lva,
 				to_copy);
 
 		copied += to_copy;
-		dkprintf("local_iov[%d]: 0x%lx %s remote_iov[%d]: 0x%lx, %lu copied, psize: %lu, rpage_left: %lu\n",
+		dkprintf("local_iov[%d]: 0x%lx %s remote_iov[%d]: 0x%lx, %lu copied, rpsize: %lu, rpage_left: %lu\n",
 			li, local_iov[li].iov_base + loff, 
 			(op == PROCESS_VM_READ) ? "<-" : "->", 
 			ri, remote_iov[ri].iov_base + roff, to_copy,
-			psize, rpage_left);
+			rpsize, rpage_left);
 
 		loff += to_copy;
 		roff += to_copy;
@@ -2700,4 +2544,48 @@ SYSCALL_DECLARE(time)
 	return time();
 }
 
+void calculate_time_from_tsc(struct timespec *ts)
+{
+	long ver;
+	unsigned long current_tsc;
+	time_t sec_delta;
+	long ns_delta;
+
+	for (;;) {
+		while ((ver = ihk_atomic64_read(&tod_data.version)) & 1) {
+			/* settimeofday() is in progress */
+			cpu_pause();
+		}
+		rmb(); /* fetch version before time */
+		*ts = tod_data.origin;
+		rmb(); /* fetch time before checking version */
+		if (ver == ihk_atomic64_read(&tod_data.version)) {
+			break;
+		}
+
+		/* settimeofday() has intervened */
+		cpu_pause();
+	}
+
+	current_tsc = rdtsc();
+	sec_delta = current_tsc / tod_data.clocks_per_sec;
+	ns_delta = NS_PER_SEC * (current_tsc % tod_data.clocks_per_sec)
+		/ tod_data.clocks_per_sec;
+	/* calc. of ns_delta overflows if clocks_per_sec exceeds 18.44 GHz */
+
+	ts->tv_sec += sec_delta;
+	ts->tv_nsec += ns_delta;
+	if (ts->tv_nsec >= NS_PER_SEC) {
+		ts->tv_nsec -= NS_PER_SEC;
+		++ts->tv_sec;
+	}
+}
+
+extern void ptrace_syscall_event(struct thread *thread);
+long arch_ptrace_syscall_event(struct thread *thread,
+			       ihk_mc_user_context_t *ctx, long setret)
+{
+	ptrace_syscall_event(thread);
+	return setret;
+}
 /*** End of File ***/

arch/arm64/kernel/timer.c

@@ -8,7 +8,7 @@
 #include <cputype.h>
 #include <irq.h>
 #include <arch-timer.h>
-#include <debug.h>
+#include <ihk/debug.h>
 
 //#define DEBUG_PRINT_TIMER
 
@@ -111,6 +111,8 @@ static void timer_handler(void *priv)
 		/* set timer re-enable for periodic */
 		arch_timer_reg_write(ARCH_TIMER_REG_TVAL, clocks);
 		arch_timer_reg_write(ARCH_TIMER_REG_CTRL, ctrl);
+
+		do_backlog();
 	}
 }
 

arch/arm64/kernel/vdso.c

@@ -11,10 +11,9 @@
 #include <process.h>
 #include <string.h>
 #include <syscall.h>
-#include <ihk/debug.h>
 #include <ikc/queue.h>
 #include <vdso.h>
-#include <debug.h>
+#include <ihk/debug.h>
 
 //#define DEBUG_PRINT_VDSO
 
@@ -23,7 +22,6 @@
 #define DDEBUG_DEFAULT DDEBUG_PRINT
 #endif
 
-#ifdef POSTK_DEBUG_ARCH_DEP_52
 #define VDSO_MAXPAGES 1
 struct vdso {
 	long busy;
@@ -34,7 +32,6 @@ struct vdso {
 	long lbase;
 	long offset_sigtramp;
 };
-#endif /*POSTK_DEBUG_ARCH_DEP_52*/
 
 extern char vdso_start, vdso_end;
 static struct vdso vdso;
@@ -90,6 +87,7 @@ int arch_setup_vdso(void)
 	}
 
 	panic("Only support host mapping vDSO");
+	return -1;
 }
 
 static int get_free_area(struct process_vm *vm, size_t len, intptr_t hint,

arch/x86_64/kboot/main.c

@@ -18,7 +18,7 @@ extern char data_start[], data_end[];
 #define LARGE_PAGE_MASK  (~((unsigned long)LARGE_PAGE_SIZE - 1))
 
 #define MAP_ST_START       0xffff800000000000UL
-#define MAP_KERNEL_START   0xffffffff80000000UL
+/* MAP_KERNEL_START is defined by cmake */
 
 #define PTL4_SHIFT         39
 #define PTL3_SHIFT         30

arch/x86_64/kernel/coredump.c

@@ -1,8 +1,9 @@
-/* coredump.c COPYRIGHT FUJITSU LIMITED 2018 */
+/* coredump.c COPYRIGHT FUJITSU LIMITED 2018-2019 */
 #include <process.h>
 #include <elfcore.h>
 
-void arch_fill_prstatus(struct elf_prstatus64 *prstatus, struct thread *thread, void *regs0)
+void arch_fill_prstatus(struct elf_prstatus64 *prstatus,
+	struct thread *thread, void *regs0, int sig)
 {
 	struct x86_user_context *uctx = regs0;
 	struct x86_basic_regs *regs = &uctx->gpr;
@@ -18,8 +19,6 @@ void arch_fill_prstatus(struct elf_prstatus64 *prstatus, struct thread *thread,
 	short int pr_cursig;
 	a8_uint64_t pr_sigpend;
 	a8_uint64_t pr_sighold;
-	pid_t pr_pid;
-	pid_t pr_ppid;
 	pid_t pr_pgrp;
 	pid_t pr_sid;
 	struct prstatus64_timeval pr_utime;
@@ -28,6 +27,14 @@ void arch_fill_prstatus(struct elf_prstatus64 *prstatus, struct thread *thread,
 	struct prstatus64_timeval pr_cstime;
  */
 
+	prstatus->pr_pid = thread->tid;
+	if (thread->proc->parent) {
+		prstatus->pr_ppid = thread->proc->parent->pid;
+	}
+
+	prstatus->pr_info.si_signo = sig;
+	prstatus->pr_cursig = sig;
+
 	prstatus->pr_reg[0] = _r15;
 	prstatus->pr_reg[1] = _r14;
 	prstatus->pr_reg[2] = _r13;
@@ -55,3 +62,13 @@ void arch_fill_prstatus(struct elf_prstatus64 *prstatus, struct thread *thread,
 
 	prstatus->pr_fpvalid = 0;	/* We assume no fp */
 }
+
+void arch_fill_thread_core_info(struct note *head,
+				struct thread *thread, void *regs)
+{
+}
+
+int arch_get_thread_core_info_size(void)
+{
+	return 0;
+}

arch/x86_64/kernel/cpu.c

@@ -1,4 +1,4 @@
-/* cpu.c COPYRIGHT FUJITSU LIMITED 2018 */
+/* cpu.c COPYRIGHT FUJITSU LIMITED 2018-2019 */
 /**
  * \file cpu.c
  *  License details are found in the file LICENSE.
@@ -16,7 +16,6 @@
  */
 
 #include <ihk/cpu.h>
-#include <ihk/debug.h>
 #include <ihk/mm.h>
 #include <types.h>
 #include <errno.h>
@@ -32,7 +31,7 @@
 #include <prctl.h>
 #include <page.h>
 #include <kmalloc.h>
-#include <debug.h>
+#include <ihk/debug.h>
 
 #define LAPIC_ID            0x020
 #define LAPIC_TIMER         0x320
@@ -45,11 +44,9 @@
 #define LAPIC_ICR0          0x300
 #define LAPIC_ICR2          0x310
 #define LAPIC_ESR           0x280
-#ifdef POSTK_DEBUG_ARCH_DEP_75 /* x86 depend hide */
 #define LOCAL_TIMER_VECTOR  0xef
 #define LOCAL_PERF_VECTOR   0xf0
 #define LOCAL_SMP_FUNC_CALL_VECTOR   0xf1
-#endif /* POSTK_DEBUG_ARCH_DEP_75 */
 
 #define APIC_INT_LEVELTRIG      0x08000
 #define APIC_INT_ASSERT         0x04000
@@ -148,7 +145,7 @@ void reload_idt(void)
 }
 
 static struct list_head handlers[256 - 32];
-extern char nmi[];
+extern char nmi_handler[];
 extern char page_fault[], general_protection_exception[];
 extern char debug_exception[], int3_exception[];
 
@@ -175,7 +172,7 @@ static void init_idt(void)
 		set_idt_entry(i, generic_common_handlers[i]);
 	}
 
-	set_idt_entry(2, (uintptr_t)nmi);
+	set_idt_entry(2, (uintptr_t)nmi_handler);
 	set_idt_entry(13, (unsigned long)general_protection_exception);
 	set_idt_entry(14, (unsigned long)page_fault);
 
@@ -955,6 +952,8 @@ void handle_interrupt(int vector, struct x86_user_context *regs)
 		v->flags |= CPU_FLAG_NEED_RESCHED;
 		ihk_mc_spinlock_unlock(&v->runq_lock, irqstate);
 		dkprintf("timer[%lu]: CPU_FLAG_NEED_RESCHED \n", rdtsc());
+
+		do_backlog();
 	}
 	else if (vector == LOCAL_PERF_VECTOR) {
 		struct siginfo info;
@@ -1206,6 +1205,15 @@ unsigned long cpu_disable_interrupt_save(void)
 	return flags;
 }
 
+unsigned long cpu_enable_interrupt_save(void)
+{
+	unsigned long flags;
+
+	asm volatile("pushf; pop %0; sti" : "=r"(flags) : : "memory", "cc");
+
+	return flags;
+}
+
 /*@
   @ behavior valid_vector:
   @   assumes 32 <= vector <= 255;
@@ -1602,14 +1610,18 @@ int ihk_mc_arch_get_special_register(enum ihk_asr_type type,
 }
 
 /*@
-  @ requires \valid_apicid(cpu);	// valid APIC ID or not
+  @ requires \valid_cpuid(cpu);     // valid CPU logical ID
   @ ensures \result == 0
   @*/
 int ihk_mc_interrupt_cpu(int cpu, int vector)
 {
+	if (cpu < 0 || cpu >= num_processors) {
+		kprintf("%s: invalid CPU id: %d\n", __func__, cpu);
+		return -1;
+	}
 	dkprintf("[%d] ihk_mc_interrupt_cpu: %d\n", ihk_mc_get_processor_id(), cpu);
 
-	x86_issue_ipi(cpu, vector);
+	x86_issue_ipi(get_x86_cpu_local_variable(cpu)->apic_id, vector);
 	return 0;
 }
 
@@ -1624,6 +1636,7 @@ struct thread *arch_switch_context(struct thread *prev, struct thread *next)
 	/* Set up new TLS.. */
 	ihk_mc_init_user_tlsbase(next->uctx, next->tlsblock_base);
 
+#ifdef ENABLE_PERF
 	/* Performance monitoring inherit */
 	if(next->proc->monitoring_event) {
 		if(next->proc->perf_status == PP_RESET)
@@ -1633,6 +1646,7 @@ struct thread *arch_switch_context(struct thread *prev, struct thread *next)
 			perf_start(next->proc->monitoring_event);
 		}
 	}
+#endif
 
 #ifdef PROFILE_ENABLE
 	if (prev && prev->profile && prev->profile_start_ts != 0) {
@@ -1708,7 +1722,7 @@ check_and_allocate_fp_regs(struct thread *thread)
 
 		if (!thread->fp_regs) {
 			kprintf("error: allocating fp_regs pages\n");
-			result = 1;
+			result = -ENOMEM;
 			goto out;
 		}
 
@@ -1721,12 +1735,14 @@ out:
 /*@
   @ requires \valid(thread);
   @*/
-void
+int
 save_fp_regs(struct thread *thread)
 {
-	if (check_and_allocate_fp_regs(thread) != 0) {
-		// alloc error
-		return;
+	int ret = 0;
+
+	ret = check_and_allocate_fp_regs(thread);
+	if (ret) {
+		goto out;
 	}
 
 	if (xsave_available) {
@@ -1741,13 +1757,23 @@ save_fp_regs(struct thread *thread)
 
 		dkprintf("fp_regs for TID %d saved\n", thread->tid);
 	}
+out:
+	return ret;
 }
 
-void copy_fp_regs(struct thread *from, struct thread *to)
+int copy_fp_regs(struct thread *from, struct thread *to)
 {
-	if ((from->fp_regs != NULL) && (check_and_allocate_fp_regs(to) == 0)) {
-		memcpy(to->fp_regs, from->fp_regs, sizeof(fp_regs_struct));
+	int ret = 0;
+
+	if (from->fp_regs != NULL) {
+		ret = check_and_allocate_fp_regs(to);
+		if (!ret) {
+			memcpy(to->fp_regs,
+				from->fp_regs,
+				sizeof(fp_regs_struct));
+		}
 	}
+	return ret;
 }
 
 /*@
@@ -1820,6 +1846,10 @@ ihk_mc_init_user_tlsbase(ihk_mc_user_context_t *ctx,
 	do_arch_prctl(ARCH_SET_FS, tls_base_addr);
 }
 
+void arch_flush_icache_all(void)
+{
+	return;
+}
 
 /*@
   @ assigns \nothing;
@@ -1973,6 +2003,92 @@ mod_nmi_ctx(void *nmi_ctx, void (*func)())
 	l[i++] = 0x28;				// KERNEL DS
 }
 
+void arch_save_panic_regs(void *irq_regs)
+{
+	struct thread *current = cpu_local_var(current);
+	struct x86_user_context *regs =
+		(struct x86_user_context *)irq_regs;
+	struct x86_cpu_local_variables *x86v =
+		get_x86_cpu_local_variable(ihk_mc_get_processor_id());
+	struct segment_regs {
+		uint32_t rflags;
+		uint32_t cs;
+		uint32_t ss;
+		uint32_t ds;
+		uint32_t es;
+		uint32_t fs;
+		uint32_t gs;
+	} *sregs;
+
+	/* Kernel space? */
+	if (regs->gpr.rip > USER_END) {
+		x86v->panic_regs[0] = regs->gpr.rax;
+		x86v->panic_regs[1] = regs->gpr.rbx;
+		x86v->panic_regs[2] = regs->gpr.rcx;
+		x86v->panic_regs[3] = regs->gpr.rdx;
+		x86v->panic_regs[4] = regs->gpr.rsi;
+		x86v->panic_regs[5] = regs->gpr.rdi;
+		x86v->panic_regs[6] = regs->gpr.rbp;
+		x86v->panic_regs[7] = regs->gpr.rsp;
+		x86v->panic_regs[8] = regs->gpr.r8;
+		x86v->panic_regs[9] = regs->gpr.r9;
+		x86v->panic_regs[10] = regs->gpr.r10;
+		x86v->panic_regs[11] = regs->gpr.r11;
+		x86v->panic_regs[12] = regs->gpr.r12;
+		x86v->panic_regs[13] = regs->gpr.r13;
+		x86v->panic_regs[14] = regs->gpr.r14;
+		x86v->panic_regs[15] = regs->gpr.r15;
+		x86v->panic_regs[16] = regs->gpr.rip;
+		sregs = (struct segment_regs *)&x86v->panic_regs[17];
+		sregs->rflags = regs->gpr.rflags;
+		sregs->cs = regs->gpr.cs;
+		sregs->ss = regs->gpr.ss;
+		sregs->ds = regs->sr.ds;
+		sregs->es = regs->sr.es;
+		sregs->fs = regs->sr.fs;
+		sregs->gs = regs->sr.gs;
+	}
+	/* User-space, show kernel context */
+	else {
+		kprintf("%s: in user-space: %p\n", __func__, regs->gpr.rip);
+		x86v->panic_regs[0] = 0;
+		x86v->panic_regs[1] = current->ctx.rbx;
+		x86v->panic_regs[2] = 0;
+		x86v->panic_regs[3] = 0;
+		x86v->panic_regs[4] = current->ctx.rsi;
+		x86v->panic_regs[5] = current->ctx.rdi;
+		x86v->panic_regs[6] = current->ctx.rbp;
+		x86v->panic_regs[7] = current->ctx.rsp;
+		x86v->panic_regs[8] = 0;
+		x86v->panic_regs[9] = 0;
+		x86v->panic_regs[10] = 0;
+		x86v->panic_regs[11] = 0;
+		x86v->panic_regs[12] = regs->gpr.r12;
+		x86v->panic_regs[13] = regs->gpr.r13;
+		x86v->panic_regs[14] = regs->gpr.r14;
+		x86v->panic_regs[15] = regs->gpr.r15;
+		x86v->panic_regs[16] = (unsigned long)enter_user_mode;
+		sregs = (struct segment_regs *)&x86v->panic_regs[17];
+		sregs->rflags = regs->gpr.rflags;
+		sregs->cs = regs->gpr.cs;
+		sregs->ss = regs->gpr.ss;
+		sregs->ds = regs->sr.ds;
+		sregs->es = regs->sr.es;
+		sregs->fs = regs->sr.fs;
+		sregs->gs = regs->sr.gs;
+	}
+
+	x86v->paniced = 1;
+}
+
+void arch_clear_panic(void)
+{
+	struct x86_cpu_local_variables *x86v =
+		get_x86_cpu_local_variable(ihk_mc_get_processor_id());
+
+	x86v->paniced = 0;
+}
+
 int arch_cpu_read_write_register(
 		struct ihk_os_cpu_register *desc,
 		enum mcctrl_os_cpu_operation op)
@@ -2096,9 +2212,7 @@ int smp_call_func(cpu_set_t *__cpu_set, smp_func_t __func, void *__arg)
 		ihk_mc_spinlock_unlock(&get_cpu_local_var(cpu)->smp_func_req_lock,
 			irq_flags);
 
-		ihk_mc_interrupt_cpu(
-				get_x86_cpu_local_variable(cpu)->apic_id,
-				LOCAL_SMP_FUNC_CALL_VECTOR);
+		ihk_mc_interrupt_cpu(cpu, LOCAL_SMP_FUNC_CALL_VECTOR);
 
 		++cpu_index;
 	}
@@ -2130,4 +2244,48 @@ free_out:
 	return ret;
 }
 
+extern int nmi_mode;
+extern long freeze_thaw(void *nmi_ctx);
+
+void multi_nm_interrupt_handler(void *irq_regs)
+{
+	dkprintf("%s: ...\n", __func__);
+	switch (nmi_mode) {
+	case 1:
+	case 2:
+		/* mode == 1 or 2, for FREEZER NMI */
+		dkprintf("%s: freeze mode NMI catch. (nmi_mode=%d)\n",
+			 __func__, nmi_mode);
+		freeze_thaw(NULL);
+		break;
+
+	case 0:
+		/* mode == 0, for MEMDUMP NMI */
+		arch_save_panic_regs(irq_regs);
+		ihk_mc_query_mem_areas();
+		/* memdump-nmi is halted McKernel, break is unnecessary. */
+		/* fall through */
+	case 3:
+		/* mode == 3, for SHUTDOWN-WAIT NMI */
+		kprintf("%s: STOP\n", __func__);
+		while (nmi_mode != 4)
+			cpu_halt();
+		break;
+
+	case 4:
+		/* mode == 4, continue NMI */
+		arch_clear_panic();
+		if (!ihk_mc_get_processor_id()) {
+			ihk_mc_clear_dump_page_completion();
+		}
+		kprintf("%s: RESUME, nmi_mode: %d\n", __func__, nmi_mode);
+		break;
+
+	default:
+		ekprintf("%s: Unknown nmi-mode(%d) detected.\n",
+			 __func__, nmi_mode);
+		break;
+	}
+}
+
 /*** end of file ***/

arch/x86_64/kernel/include/arch-lock.h

@@ -33,7 +33,7 @@ extern void preempt_disable(void);
 
 #define IHK_STATIC_SPINLOCK_FUNCS
 
-static void ihk_mc_spinlock_init(ihk_spinlock_t *lock)
+static inline void ihk_mc_spinlock_init(ihk_spinlock_t *lock)
 {
 	lock->head_tail = 0;
 }
@@ -50,10 +50,13 @@ rc = __ihk_mc_spinlock_trylock_noirq(l); \
 #define ihk_mc_spinlock_trylock_noirq __ihk_mc_spinlock_trylock_noirq
 #endif
 
-static int __ihk_mc_spinlock_trylock_noirq(ihk_spinlock_t *lock)
+static inline int __ihk_mc_spinlock_trylock_noirq(ihk_spinlock_t *lock)
 {
 	ihk_spinlock_t cur = { .head_tail = lock->head_tail };
-	ihk_spinlock_t next = { .tickets.head = cur.tickets.head, .tickets.tail = cur.tickets.tail + 2 };
+	ihk_spinlock_t next = { .tickets = {
+		.head = cur.tickets.head,
+		.tail = cur.tickets.tail + 2
+	} };
 	int success;
 
 	if (cur.tickets.head != cur.tickets.tail) {
@@ -80,7 +83,8 @@ __kprintf("[%d] ret ihk_mc_spinlock_trylock\n", ihk_mc_get_processor_id()); rc;\
 #else
 #define ihk_mc_spinlock_trylock __ihk_mc_spinlock_trylock
 #endif
-static unsigned long __ihk_mc_spinlock_trylock(ihk_spinlock_t *lock, int *result)
+static inline unsigned long __ihk_mc_spinlock_trylock(ihk_spinlock_t *lock,
+						      int *result)
 {
 	unsigned long flags;
 	
@@ -101,7 +105,7 @@ __kprintf("[%d] ret ihk_mc_spinlock_lock_noirq\n", ihk_mc_get_processor_id()); \
 #define ihk_mc_spinlock_lock_noirq __ihk_mc_spinlock_lock_noirq
 #endif
 
-static void __ihk_mc_spinlock_lock_noirq(ihk_spinlock_t *lock)
+static inline void __ihk_mc_spinlock_lock_noirq(ihk_spinlock_t *lock)
 {
 	register struct __raw_tickets inc = { .tail = 0x0002 };
 
@@ -132,7 +136,7 @@ __kprintf("[%d] ret ihk_mc_spinlock_lock\n", ihk_mc_get_processor_id()); rc;\
 #else
 #define ihk_mc_spinlock_lock __ihk_mc_spinlock_lock
 #endif
-static unsigned long __ihk_mc_spinlock_lock(ihk_spinlock_t *lock)
+static inline unsigned long __ihk_mc_spinlock_lock(ihk_spinlock_t *lock)
 {
 	unsigned long flags;
 	
@@ -152,7 +156,7 @@ __kprintf("[%d] ret ihk_mc_spinlock_unlock_noirq\n", ihk_mc_get_processor_id());
 #else
 #define ihk_mc_spinlock_unlock_noirq __ihk_mc_spinlock_unlock_noirq
 #endif
-static void __ihk_mc_spinlock_unlock_noirq(ihk_spinlock_t *lock)
+static inline void __ihk_mc_spinlock_unlock_noirq(ihk_spinlock_t *lock)
 {
 	__ticket_t inc = 0x0002;
 
@@ -171,100 +175,14 @@ __kprintf("[%d] ret ihk_mc_spinlock_unlock\n", ihk_mc_get_processor_id()); \
 #else
 #define ihk_mc_spinlock_unlock __ihk_mc_spinlock_unlock
 #endif
-static void __ihk_mc_spinlock_unlock(ihk_spinlock_t *lock, unsigned long flags)
+static inline void __ihk_mc_spinlock_unlock(ihk_spinlock_t *lock,
+					    unsigned long flags)
 {
 	__ihk_mc_spinlock_unlock_noirq(lock);
 
 	cpu_restore_interrupt(flags);
 }
 
-/* An implementation of the Mellor-Crummey Scott (MCS) lock */
-typedef struct mcs_lock_node {
-	unsigned long locked;
-	struct mcs_lock_node *next;
-	unsigned long irqsave;
-#ifndef ENABLE_UBSAN
-} __aligned(64) mcs_lock_node_t;
-#else
-} mcs_lock_node_t;
-#endif
-
-typedef mcs_lock_node_t mcs_lock_t;
-
-static void mcs_lock_init(struct mcs_lock_node *node)
-{
-	node->locked = 0;
-	node->next = NULL;
-}
-
-static void __mcs_lock_lock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	struct mcs_lock_node *pred;
-
-	node->next = NULL;
-	node->locked = 0;
-	pred = (struct mcs_lock_node *)xchg8((unsigned long *)&lock->next,
-			(unsigned long)node);
-
-	if (pred) {
-		node->locked = 1;
-		pred->next = node;
-		while (node->locked != 0) {
-			cpu_pause();
-		}
-	}
-}
-
-static void __mcs_lock_unlock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	if (node->next == NULL) {
-		struct mcs_lock_node *old = (struct mcs_lock_node *)
-			atomic_cmpxchg8((unsigned long *)&lock->next,
-					(unsigned long)node, (unsigned long)0);
-
-		if (old == node) {
-			return;
-		}
-
-		while (node->next == NULL) {
-			cpu_pause();
-		}
-	}
-
-	node->next->locked = 0;
-}
-
-static void mcs_lock_lock_noirq(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	preempt_disable();
-	__mcs_lock_lock(lock, node);
-}
-
-static void mcs_lock_unlock_noirq(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	__mcs_lock_unlock(lock, node);
-	preempt_enable();
-}
-
-static void mcs_lock_lock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	node->irqsave = cpu_disable_interrupt_save();
-	mcs_lock_lock_noirq(lock, node);
-}
-
-static void mcs_lock_unlock(struct mcs_lock_node *lock,
-		struct mcs_lock_node *node)
-{
-	mcs_lock_unlock_noirq(lock, node);
-	cpu_restore_interrupt(node->irqsave);
-}
-
-
 #define SPINLOCK_IN_MCS_RWLOCK
 
 // reader/writer lock
@@ -310,7 +228,7 @@ typedef struct mcs_rwlock_lock {
 } mcs_rwlock_lock_t;
 #endif
 
-static void
+static inline void
 mcs_rwlock_init(struct mcs_rwlock_lock *lock)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -331,7 +249,7 @@ __kprintf("[%d] ret mcs_rwlock_writer_lock_noirq\n", ihk_mc_get_processor_id());
 #else
 #define mcs_rwlock_writer_lock_noirq __mcs_rwlock_writer_lock_noirq
 #endif
-static void
+static inline void
 __mcs_rwlock_writer_lock_noirq(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -358,7 +276,7 @@ __mcs_rwlock_writer_lock_noirq(struct mcs_rwlock_lock *lock, struct mcs_rwlock_n
 }
 
 #ifndef SPINLOCK_IN_MCS_RWLOCK
-static void
+static inline void
 mcs_rwlock_unlock_readers(struct mcs_rwlock_lock *lock)
 {
 	struct mcs_rwlock_node *p;
@@ -425,7 +343,7 @@ __kprintf("[%d] ret mcs_rwlock_writer_unlock_noirq\n", ihk_mc_get_processor_id()
 #else
 #define mcs_rwlock_writer_unlock_noirq __mcs_rwlock_writer_unlock_noirq
 #endif
-static void
+static inline void
 __mcs_rwlock_writer_unlock_noirq(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -485,7 +403,7 @@ atomic_inc_ifnot0(ihk_atomic_t *v)
 	return old;
 }
 
-static void
+static inline void
 __mcs_rwlock_reader_lock_noirq(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -551,7 +469,7 @@ __kprintf("[%d] ret mcs_rwlock_reader_unlock_noirq\n", ihk_mc_get_processor_id()
 #else
 #define mcs_rwlock_reader_unlock_noirq __mcs_rwlock_reader_unlock_noirq
 #endif
-static void
+static inline void
 __mcs_rwlock_reader_unlock_noirq(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -598,7 +516,7 @@ __kprintf("[%d] ret mcs_rwlock_writer_lock\n", ihk_mc_get_processor_id()); \
 #else
 #define mcs_rwlock_writer_lock __mcs_rwlock_writer_lock
 #endif
-static void
+static inline void
 __mcs_rwlock_writer_lock(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node_irqsave *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -618,7 +536,7 @@ __kprintf("[%d] ret mcs_rwlock_writer_unlock\n", ihk_mc_get_processor_id()); \
 #else
 #define mcs_rwlock_writer_unlock __mcs_rwlock_writer_unlock
 #endif
-static void
+static inline void
 __mcs_rwlock_writer_unlock(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node_irqsave *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -638,7 +556,7 @@ __kprintf("[%d] ret mcs_rwlock_reader_lock\n", ihk_mc_get_processor_id()); \
 #else
 #define mcs_rwlock_reader_lock __mcs_rwlock_reader_lock
 #endif
-static void
+static inline void
 __mcs_rwlock_reader_lock(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node_irqsave *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -658,7 +576,7 @@ __kprintf("[%d] ret mcs_rwlock_reader_unlock\n", ihk_mc_get_processor_id()); \
 #else
 #define mcs_rwlock_reader_unlock __mcs_rwlock_reader_unlock
 #endif
-static void
+static inline void
 __mcs_rwlock_reader_unlock(struct mcs_rwlock_lock *lock, struct mcs_rwlock_node_irqsave *node)
 {
 #ifdef SPINLOCK_IN_MCS_RWLOCK
@@ -674,4 +592,90 @@ static inline int irqflags_can_interrupt(unsigned long flags)
 	return !!(flags & 0x200);
 }
 
+struct ihk_rwlock {
+	union {
+		long lock;
+		struct {
+			unsigned int read;
+			int write;
+		};
+	} lock;
+};
+
+static inline void ihk_mc_rwlock_init(struct ihk_rwlock *rw)
+{
+	rw->lock.read = 0;
+	rw->lock.write = 1;
+}
+
+static inline void ihk_mc_read_lock(struct ihk_rwlock *rw)
+{
+	asm volatile("1:\t"
+		     "lock; decq %0\n\t"
+		     "jns 3f\n\t"
+		     "lock incq %0\n\t"
+		     "2:\t"
+		     "pause\n\t"
+		     "cmpq $0x1, %0\n\t"
+		     "jns 1b\n\t"
+		     "jmp 2b\n\t"
+		     "3:"
+		     : "+m" (rw->lock.lock) : : "memory");
+}
+
+static inline void ihk_mc_write_lock(struct ihk_rwlock *rw)
+{
+	asm volatile("1:\t"
+		     "lock; decl %0\n\t"
+		     "je 3f\n\t"
+		     "lock; incl %0\n\t"
+		     "2:\t"
+		     "pause\n\t"
+		     "cmpl $0x1,%0\n\t"
+		     "je 1b\n\t"
+		     "jmp 2b\n\t"
+		     "3:"
+		     : "+m" (rw->lock.write) : "i" (((1L) << 32)) : "memory");
+}
+
+static inline int ihk_mc_read_trylock(struct ihk_rwlock *rw)
+{
+	ihk_atomic64_t *count = (ihk_atomic64_t *)rw;
+
+	if (ihk_atomic64_sub_return(1, count) >= 0)
+		return 1;
+	ihk_atomic64_inc(count);
+	return 0;
+}
+
+static inline int ihk_mc_write_trylock(struct ihk_rwlock *rw)
+{
+	ihk_atomic_t *count = (ihk_atomic_t *)&rw->lock.write;
+
+	if (ihk_atomic_dec_and_test(count))
+		return 1;
+	ihk_atomic_inc(count);
+	return 0;
+}
+
+static inline void ihk_mc_read_unlock(struct ihk_rwlock *rw)
+{
+	asm volatile("lock; incq %0" : "+m" (rw->lock.lock) : : "memory");
+}
+
+static inline void ihk_mc_write_unlock(struct ihk_rwlock *rw)
+{
+	asm volatile("lock; incl %0"
+		     : "+m" (rw->lock.write) : "i" (((1L) << 32)) : "memory");
+}
+
+static inline int ihk_mc_write_can_lock(struct ihk_rwlock *rw)
+{
+	return rw->lock.write == 1;
+}
+
+static inline int ihk_mc_read_can_lock(struct ihk_rwlock *rw)
+{
+	return rw->lock.lock > 0;
+}
 #endif

arch/x86_64/kernel/include/arch-memory.h

@@ -17,6 +17,7 @@
 #define __HEADER_X86_COMMON_ARCH_MEMORY_H
 
 #include <ihk/types.h>
+#include <errno.h>
 
 #define KERNEL_CS_ENTRY    4
 #define KERNEL_DS_ENTRY    5
@@ -66,8 +67,8 @@
  * Placing the LWK image in the virtual address space at the end of
  * the Linux modules section enables us to map the LWK TEXT in Linux
  * as well, so that Linux can also call into LWK text.
+ * It's defined by cmake.
  */
-#define MAP_KERNEL_START   0xFFFFFFFFFE800000UL
 #define STACK_TOP(region)  ((region)->user_end)
 
 #define MAP_VMAP_SIZE      0x0000000100000000UL
@@ -183,12 +184,10 @@ enum ihk_mc_pt_attribute {
 
 enum ihk_mc_pt_attribute attr_mask;
 
-#ifdef POSTK_DEBUG_ARCH_DEP_12
 static inline int pfn_is_write_combined(uintptr_t pfn)
 {
 	return ((pfn & PFL1_PWT) && !(pfn & PFL1_PCD));
 }
-#endif /* #ifdef POSTK_DEBUG_ARCH_DEP_12 */
 
 static inline int pte_is_null(pte_t *ptep)
 {
@@ -365,6 +364,17 @@ static inline int pgsize_to_tbllv(size_t pgsize)
 	return 0;
 }
 
+static inline int pgsize_to_pgshift(size_t pgsize)
+{
+	switch (pgsize) {
+	case PTL1_SIZE:	return PTL1_SHIFT;
+	case PTL2_SIZE:	return PTL2_SHIFT;
+	case PTL3_SIZE:	return PTL3_SHIFT;
+	case PTL4_SIZE:	return PTL4_SHIFT;
+	default: return -EINVAL;
+	}
+}
+
 static inline size_t tbllv_to_pgsize(int level)
 {
 	switch (level) {

arch/x86_64/kernel/include/arch/cpu.h

@@ -13,19 +13,17 @@
 #ifndef ARCH_CPU_H
 #define ARCH_CPU_H
 
+#define mb()    asm volatile("mfence":::"memory")
+#define rmb()   asm volatile("lfence":::"memory")
+#define wmb()   asm volatile("sfence" ::: "memory")
+
+#define smp_mb()    mb()
+#define smp_rmb()   rmb()
+#define smp_wmb()	barrier()
+
 #define arch_barrier()	asm volatile("" : : : "memory")
 
-static inline void rmb(void)
-{
-	arch_barrier();
-}
-
-static inline void wmb(void)
-{
-	arch_barrier();
-}
-
-static unsigned long read_tsc(void)
+static inline unsigned long read_tsc(void)
 {
 	unsigned int low, high;
 
@@ -34,4 +32,21 @@ static unsigned long read_tsc(void)
 	return (low | ((unsigned long)high << 32));
 }
 
+#define smp_load_acquire(p)						\
+({									\
+	typeof(*p) ___p1 = ACCESS_ONCE(*p);				\
+	compiletime_assert_atomic_type(*p);				\
+	barrier();							\
+	___p1;								\
+})
+
+#define smp_store_release(p, v)			\
+({							\
+	compiletime_assert_atomic_type(*p);	\
+	barrier();							\
+	WRITE_ONCE(*p, v);					\
+})
+
+void arch_flush_icache_all(void);
+
 #endif /* ARCH_CPU_H */

arch/x86_64/kernel/include/arch_rusage.h

@@ -1,32 +0,0 @@
-#ifndef ARCH_RUSAGE_H_INCLUDED
-#define ARCH_RUSAGE_H_INCLUDED
-
-#define DEBUG_RUSAGE
-
-#define IHK_OS_PGSIZE_4KB 0
-#define IHK_OS_PGSIZE_2MB 1
-#define IHK_OS_PGSIZE_1GB 2
-
-extern struct rusage_global rusage;
-
-static inline int rusage_pgsize_to_pgtype(size_t pgsize)
-{
-	int ret = IHK_OS_PGSIZE_4KB;
-	switch (pgsize) {
-	case PTL1_SIZE:
-		ret = IHK_OS_PGSIZE_4KB;
-		break;
-	case PTL2_SIZE:
-		ret = IHK_OS_PGSIZE_2MB;
-		break;
-	case PTL3_SIZE:
-		ret = IHK_OS_PGSIZE_1GB;
-		break;
-	default:
-		kprintf("%s: Error: Unknown pgsize=%ld\n", __FUNCTION__, pgsize);
-		break;
-	}
-	return ret;
-}
-
-#endif /* !defined(ARCH_RUSAGE_H_INCLUDED) */

arch/x86_64/kernel/include/ihk/atomic.h

@@ -13,6 +13,8 @@
 #ifndef HEADER_X86_COMMON_IHK_ATOMIC_H
 #define HEADER_X86_COMMON_IHK_ATOMIC_H
  
+#include <lwk/compiler.h>
+
 /***********************************************************************
  * ihk_atomic_t
  */
@@ -114,7 +116,7 @@ static inline long ihk_atomic64_read(const ihk_atomic64_t *v)
 	return *(volatile long *)&(v)->counter64;
 }
 
-static inline void ihk_atomic64_set(ihk_atomic64_t *v, int i)
+static inline void ihk_atomic64_set(ihk_atomic64_t *v, long i)
 {
 	v->counter64 = i;
 }
@@ -124,6 +126,22 @@ static inline void ihk_atomic64_inc(ihk_atomic64_t *v)
 	asm volatile ("lock incq %0" : "+m"(v->counter64));
 }
 
+static inline long ihk_atomic64_add_return(long i, ihk_atomic64_t *v)
+{
+	long __i;
+
+	__i = i;
+	asm volatile("lock xaddq %0, %1"
+		     : "+r" (i), "+m" (v->counter64)
+		     : : "memory");
+	return i + __i;
+}
+
+static inline long ihk_atomic64_sub_return(long i, ihk_atomic64_t *v)
+{
+	return ihk_atomic64_add_return(-i, v);
+}
+
 /***********************************************************************
  * others
  */
@@ -156,43 +174,55 @@ static inline unsigned long xchg8(unsigned long *ptr, unsigned long x)
 	return __x;
 }
 
-#define __xchg(x, ptr, size)						\
-({									\
-	__typeof(*(ptr)) __x = (x);					\
-	switch (size) {							\
-	case 1:								\
-		asm volatile("xchgb %b0,%1"				\
-			     : "=q" (__x)				\
-			     : "m" (*__xg(ptr)), "0" (__x)		\
-			     : "memory");				\
-		break;							\
-	case 2:								\
-		asm volatile("xchgw %w0,%1"				\
-			     : "=r" (__x)				\
-			     : "m" (*__xg(ptr)), "0" (__x)		\
-			     : "memory");				\
-		break;							\
-	case 4:								\
-		asm volatile("xchgl %k0,%1"				\
-			     : "=r" (__x)				\
-			     : "m" (*__xg(ptr)), "0" (__x)		\
-			     : "memory");				\
-		break;							\
-	case 8:								\
-		asm volatile("xchgq %0,%1"				\
-			     : "=r" (__x)				\
-			     : "m" (*__xg(ptr)), "0" (__x)		\
-			     : "memory");				\
-		break;							\
-	default:							\
-		panic("xchg for wrong size");					\
-	}								\
-	__x;								\
-})
+#define __X86_CASE_B	1
+#define __X86_CASE_W	2
+#define __X86_CASE_L	4
+#define __X86_CASE_Q	8
 
+extern void __xchg_wrong_size(void)
+	__compiletime_error("Bad argument size for xchg");
 
-#define xchg(ptr, v)							\
-	__xchg((v), (ptr), sizeof(*ptr))
+/*
+ * An exchange-type operation, which takes a value and a pointer, and
+ * returns the old value.
+ */
+#define __xchg_op(ptr, arg, op, lock)					\
+	({								\
+		__typeof__(*(ptr)) __ret = (arg);			\
+		switch (sizeof(*(ptr))) {				\
+		case __X86_CASE_B:					\
+			asm volatile (lock #op "b %b0, %1\n"		\
+				      : "+q" (__ret), "+m" (*(ptr))	\
+				      : : "memory", "cc");		\
+			break;						\
+		case __X86_CASE_W:					\
+			asm volatile (lock #op "w %w0, %1\n"		\
+				      : "+r" (__ret), "+m" (*(ptr))	\
+				      : : "memory", "cc");		\
+			break;						\
+		case __X86_CASE_L:					\
+			asm volatile (lock #op "l %0, %1\n"		\
+				      : "+r" (__ret), "+m" (*(ptr))	\
+				      : : "memory", "cc");		\
+			break;						\
+		case __X86_CASE_Q:					\
+			asm volatile (lock #op "q %q0, %1\n"		\
+				      : "+r" (__ret), "+m" (*(ptr))	\
+				      : : "memory", "cc");		\
+			break;						\
+		default:						\
+			__xchg_wrong_size();			\
+		}							\
+		__ret;							\
+	})
+
+/*
+ * Note: no "lock" prefix even on SMP: xchg always implies lock anyway.
+ * Since this is generally used to protect other memory information, we
+ * use "asm volatile" and "memory" clobbers to prevent gcc from moving
+ * information around.
+ */
+#define xchg(ptr, v)	__xchg_op((ptr), (v), xchg, "")
 
 static inline unsigned long atomic_cmpxchg8(unsigned long *addr,
 		unsigned long oldval,
@@ -241,4 +271,66 @@ static inline unsigned long ihk_atomic_add_long_return(long i, long *v) {
         return i + __i;
 }
 
+extern void __cmpxchg_wrong_size(void)
+	__compiletime_error("Bad argument size for cmpxchg");
+
+/*
+ * Atomic compare and exchange.  Compare OLD with MEM, if identical,
+ * store NEW in MEM.  Return the initial value in MEM.  Success is
+ * indicated by comparing RETURN with OLD.
+ */
+#define __raw_cmpxchg(ptr, old, new, size, lock)		\
+({									\
+	__typeof__(*(ptr)) __ret;					\
+	__typeof__(*(ptr)) __old = (old);				\
+	__typeof__(*(ptr)) __new = (new);				\
+	switch (size) {							\
+	case __X86_CASE_B:						\
+	{								\
+		volatile uint8_t *__ptr = (volatile uint8_t *)(ptr);\
+		asm volatile(lock "cmpxchgb %2,%1"			\
+			     : "=a" (__ret), "+m" (*__ptr)		\
+			     : "q" (__new), "0" (__old)			\
+			     : "memory");				\
+		break;							\
+	}								\
+	case __X86_CASE_W:						\
+	{								\
+		volatile uint16_t *__ptr = (volatile uint16_t *)(ptr);\
+		asm volatile(lock "cmpxchgw %2,%1"			\
+			     : "=a" (__ret), "+m" (*__ptr)		\
+			     : "r" (__new), "0" (__old)			\
+			     : "memory");				\
+		break;							\
+	}								\
+	case __X86_CASE_L:						\
+	{								\
+		volatile uint32_t *__ptr = (volatile uint32_t *)(ptr);\
+		asm volatile(lock "cmpxchgl %2,%1"			\
+			     : "=a" (__ret), "+m" (*__ptr)		\
+			     : "r" (__new), "0" (__old)			\
+			     : "memory");				\
+		break;							\
+	}								\
+	case __X86_CASE_Q:						\
+	{								\
+		volatile uint64_t *__ptr = (volatile uint64_t *)(ptr);\
+		asm volatile(lock "cmpxchgq %2,%1"			\
+			     : "=a" (__ret), "+m" (*__ptr)		\
+			     : "r" (__new), "0" (__old)			\
+			     : "memory");				\
+		break;							\
+	}								\
+	default:							\
+		__cmpxchg_wrong_size();		\
+	}								\
+	__ret;								\
+})
+
+#define __cmpxchg(ptr, old, new, size)					\
+	__raw_cmpxchg((ptr), (old), (new), (size), "lock; ")
+
+#define cmpxchg(ptr, old, new)						\
+	__cmpxchg(ptr, old, new, sizeof(*(ptr)))
+
 #endif

arch/x86_64/kernel/include/registers.h

@@ -71,7 +71,7 @@
 #define MSR_PERF_CTL_0 0xc0010000
 #define MSR_PERF_CTR_0 0xc0010004
 
-static unsigned long xgetbv(unsigned int index)
+static inline unsigned long xgetbv(unsigned int index)
 {
 	unsigned int low, high;
 
@@ -80,7 +80,7 @@ static unsigned long xgetbv(unsigned int index)
 	return low | ((unsigned long)high << 32);
 }
 
-static void xsetbv(unsigned int index, unsigned long val)
+static inline void xsetbv(unsigned int index, unsigned long val)
 {
 	unsigned int low, high;
 
@@ -90,7 +90,8 @@ static void xsetbv(unsigned int index, unsigned long val)
 	asm volatile("xsetbv" : : "a" (low), "d" (high), "c" (index));
 }
 
-static void wrmsr(unsigned int idx, unsigned long value){
+static inline void wrmsr(unsigned int idx, unsigned long value)
+{
 	unsigned int high, low;
 
 	high = value >> 32;
@@ -99,7 +100,7 @@ static void wrmsr(unsigned int idx, unsigned long value){
 	asm volatile("wrmsr" : : "c" (idx), "a" (low), "d" (high) : "memory");
 }
 
-static unsigned long rdpmc(unsigned int counter)
+static inline unsigned long rdpmc(unsigned int counter)
 {
 	unsigned int high, low;
 
@@ -108,7 +109,7 @@ static unsigned long rdpmc(unsigned int counter)
 	return (unsigned long)high << 32 | low;
 }
 
-static unsigned long rdmsr(unsigned int index)
+static inline unsigned long rdmsr(unsigned int index)
 {
 	unsigned int high, low;
 
@@ -117,7 +118,7 @@ static unsigned long rdmsr(unsigned int index)
 	return (unsigned long)high << 32 | low;
 }
 
-static unsigned long rdtsc(void)
+static inline unsigned long rdtsc(void)
 {
 	unsigned int high, low;
 
@@ -126,7 +127,7 @@ static unsigned long rdtsc(void)
 	return (unsigned long)high << 32 | low;
 }
 
-static void set_perfctl(int counter, int event, int mask)
+static inline void set_perfctl(int counter, int event, int mask)
 {
 	unsigned long value;
 
@@ -137,7 +138,7 @@ static void set_perfctl(int counter, int event, int mask)
 	wrmsr(MSR_PERF_CTL_0 + counter, value);
 }
 
-static void start_perfctr(int counter)
+static inline void start_perfctr(int counter)
 {
 	unsigned long value;
 
@@ -145,7 +146,7 @@ static void start_perfctr(int counter)
 	value |= (1 << 22);
 	wrmsr(MSR_PERF_CTL_0 + counter, value);
 }
-static void stop_perfctr(int counter)
+static inline void stop_perfctr(int counter)
 {
 	unsigned long value;
 
@@ -154,17 +155,17 @@ static void stop_perfctr(int counter)
 	wrmsr(MSR_PERF_CTL_0 + counter, value);
 }
 
-static void clear_perfctl(int counter)
+static inline void clear_perfctl(int counter)
 {
 	wrmsr(MSR_PERF_CTL_0 + counter, 0);
 }
 
-static void set_perfctr(int counter, unsigned long value)
+static inline void set_perfctr(int counter, unsigned long value)
 {
 	wrmsr(MSR_PERF_CTR_0 + counter, value);
 }
 
-static unsigned long read_perfctr(int counter)
+static inline unsigned long read_perfctr(int counter)
 {
 	return rdpmc(counter);
 }

arch/x86_64/kernel/include/rlimit.h

@@ -84,7 +84,11 @@ enum __rlimit_resource
   __RLIMIT_RTPRIO = 14,
 #define RLIMIT_RTPRIO __RLIMIT_RTPRIO
 
-  __RLIMIT_NLIMITS = 15,
+  /* timeout for RT tasks in us */
+  __RLIMIT_RTTIME = 15,
+#define RLIMIT_RTTIME __RLIMIT_RTTIME
+
+  __RLIMIT_NLIMITS = 16,
   __RLIM_NLIMITS = __RLIMIT_NLIMITS
 #define RLIMIT_NLIMITS __RLIMIT_NLIMITS
 #define RLIM_NLIMITS __RLIM_NLIMITS

arch/x86_64/kernel/include/syscall_list.h

@@ -74,6 +74,7 @@ SYSCALL_DELEGATED(89, readlink)
 SYSCALL_HANDLED(96, gettimeofday)
 SYSCALL_HANDLED(97, getrlimit)
 SYSCALL_HANDLED(98, getrusage)
+SYSCALL_HANDLED(99, sysinfo)
 SYSCALL_HANDLED(100, times)
 SYSCALL_HANDLED(101, ptrace)
 SYSCALL_HANDLED(102, getuid)
@@ -147,24 +148,24 @@ SYSCALL_DELEGATED(266, symlinkat)
 SYSCALL_DELEGATED(267, readlinkat)
 SYSCALL_DELEGATED(268, fchmodat)
 SYSCALL_DELEGATED(269, faccessat)
-SYSCALL_DELEGATED(270, pselect6)
-SYSCALL_DELEGATED(271, ppoll)
+SYSCALL_HANDLED(270, pselect6)
+SYSCALL_HANDLED(271, ppoll)
 SYSCALL_HANDLED(273, set_robust_list)
 SYSCALL_HANDLED(279, move_pages)
-SYSCALL_DELEGATED(281, epoll_pwait)
+SYSCALL_HANDLED(281, epoll_pwait)
 SYSCALL_HANDLED(282, signalfd)
 SYSCALL_HANDLED(289, signalfd4)
+#ifdef ENABLE_PERF
 SYSCALL_HANDLED(298, perf_event_open)
+#endif
+SYSCALL_HANDLED(302, prlimit64)
 #ifdef DCFA_KMOD
 SYSCALL_HANDLED(303, mod_call)
 #endif
 SYSCALL_HANDLED(309, getcpu)
 SYSCALL_HANDLED(310, process_vm_readv)
 SYSCALL_HANDLED(311, process_vm_writev)
-SYSCALL_HANDLED(601, pmc_init)
-SYSCALL_HANDLED(602, pmc_start)
-SYSCALL_HANDLED(603, pmc_stop)
-SYSCALL_HANDLED(604, pmc_reset)
+SYSCALL_HANDLED(322, execveat)
 SYSCALL_HANDLED(700, get_cpu_id)
 #ifdef PROFILE_ENABLE
 SYSCALL_HANDLED(__NR_profile, profile)
@@ -180,4 +181,8 @@ SYSCALL_HANDLED(802, linux_mlock)
 SYSCALL_HANDLED(803, suspend_threads)
 SYSCALL_HANDLED(804, resume_threads)
 SYSCALL_HANDLED(811, linux_spawn)
-/**** End of File ****/
+
+/* Do not edit the lines including this comment and
+ * EOF just after it because those are used as a
+ * robust marker for the autotest patch.
+ */

arch/x86_64/kernel/interrupt.S

@@ -1,3 +1,4 @@
+/* interrupt.S COPYRIGHT FUJITSU LIMITED 2019 */
 /**
  * \file interrupt.S
  *  License details are found in the file LICENSE.
@@ -91,6 +92,9 @@ vector=vector+1
 .endr
 
 common_interrupt:
+#define MULT_INTR_VECTOR 242
+		cmp $(MULT_INTR_VECTOR),%rdi
+		je  1f
 		PUSH_ALL_REGS
 		movq    ERROR_OFFSET(%rsp), %rdi
 		movq	%rsp, %rsi
@@ -99,6 +103,19 @@ common_interrupt:
 		addq	$8, %rsp
 		iretq
 
+
+.globl  nmi_handler
+nmi_handler:
+		cld
+		pushq   $0 /* error field of x86_basic_regs */
+		PUSH_ALL_REGS
+		movq	%rsp, %rdi
+		call 	multi_nm_interrupt_handler /* Enter C code */
+		POP_ALL_REGS
+		addq	$8, %rsp
+		iretq
+
+
 .globl __page_fault_handler_address
 __page_fault_handler_address:
 		.quad	0
@@ -137,74 +154,6 @@ __freeze:
 		POP_ALL_REGS
 		iretq
 
-.globl nmi
-nmi:
-#define PANICED 232
-#define PANIC_REGS 240
-		movq	%rax,%gs:PANIC_REGS+0x00
-		movq	%rsp,%gs:PANIC_REGS+0x08
-
-		movl	nmi_mode(%rip),%eax
-		cmp $3,%rax
-		je  4f
-		cmp	$1,%rax
-		je	1f
-		cmp	$2,%rax
-		jne	3f
-1:
-		cld
-		movq	%gs:PANIC_REGS+0x00,%rax
-		PUSH_ALL_REGS
-		subq	$40, %rsp
-		movq	%rsp,%gs:PANIC_REGS+0x10
-		movq	%rsp, %rdi
-		call	freeze_thaw
-		cmpq	$0, %rax
-		jnz	2f
-		addq	$40, %rsp
-2:
-		POP_ALL_REGS
-		iretq
-3:
-		movq	%rbx,%gs:PANIC_REGS+0x08
-		movq	%rcx,%gs:PANIC_REGS+0x10
-		movq	%rdx,%gs:PANIC_REGS+0x18
-		movq	%rsi,%gs:PANIC_REGS+0x20
-		movq	%rdi,%gs:PANIC_REGS+0x28
-		movq	%rbp,%gs:PANIC_REGS+0x30
-		movq	0x18(%rsp),%rax		/* rsp */
-		movq	%rax,%gs:PANIC_REGS+0x38
-		movq	%r8, %gs:PANIC_REGS+0x40
-		movq	%r9, %gs:PANIC_REGS+0x48
-		movq	%r10,%gs:PANIC_REGS+0x50
-		movq	%r11,%gs:PANIC_REGS+0x58
-		movq	%r12,%gs:PANIC_REGS+0x60
-		movq	%r13,%gs:PANIC_REGS+0x68
-		movq	%r14,%gs:PANIC_REGS+0x70
-		movq	%r15,%gs:PANIC_REGS+0x78
-		movq	0x00(%rsp),%rax		/* rip */
-		movq	%rax,%gs:PANIC_REGS+0x80
-		movq	0x10(%rsp),%rax		/* rflags */
-		movl	%eax,%gs:PANIC_REGS+0x88
-		movq	0x08(%rsp),%rax		/* cs */
-		movl	%eax,%gs:PANIC_REGS+0x8C
-		movq	0x20(%rsp),%rax		/* ss */
-		movl	%eax,%gs:PANIC_REGS+0x90
-		xorq	%rax,%rax
-		movw	%ds,%ax
-		movl	%eax,%gs:PANIC_REGS+0x94
-		movw	%es,%ax
-		movl	%eax,%gs:PANIC_REGS+0x98
-		movw	%fs,%ax
-		movl	%eax,%gs:PANIC_REGS+0x9C
-		movw	%gs,%ax
-		movl	%eax,%gs:PANIC_REGS+0xA0
-		movq	$1,%gs:PANICED
-		call	ihk_mc_query_mem_areas
-4:
-		hlt
-		jmp	4b
-
 .globl x86_syscall
 x86_syscall:
 		cld

arch/x86_64/kernel/memory.c

@@ -14,7 +14,6 @@
  */
 
 #include <ihk/cpu.h>
-#include <ihk/debug.h>
 #include <ihk/mm.h>
 #include <types.h>
 #include <memory.h>
@@ -26,7 +25,7 @@
 #include <cls.h>
 #include <kmalloc.h>
 #include <rusage_private.h>
-#include <debug.h>
+#include <ihk/debug.h>
 
 //#define DEBUG
 
@@ -38,6 +37,7 @@
 static char *last_page;
 extern char _head[], _end[];
 
+extern unsigned long linux_page_offset_base;
 extern unsigned long x86_kernel_phys_base;
 
 /* Arch specific early allocation routine */
@@ -1355,109 +1355,6 @@ struct clear_range_args {
 	int max_nr_addr;
 };
 
-#ifdef POSTK_DEBUG_ARCH_DEP_8
-void remote_flush_tlb_cpumask(struct process_vm *vm,
-		unsigned long addr, int cpu_id)
-{
-	unsigned long __addr = addr;
-	return remote_flush_tlb_array_cpumask(vm, &__addr, 1, cpu_id);
-}
-
-void remote_flush_tlb_array_cpumask(struct process_vm *vm,
-		unsigned long *addr,
-		int nr_addr,
-		int cpu_id)
-{
-	unsigned long cpu;
-	int flush_ind;
-	struct tlb_flush_entry *flush_entry;
-	cpu_set_t _cpu_set;
-
-	if (addr[0]) {
-		flush_ind = (addr[0] >> PAGE_SHIFT) % IHK_TLB_FLUSH_IRQ_VECTOR_SIZE;
-	}
-	/* Zero address denotes full TLB flush */
-	else {
-		/* Random.. */
-		flush_ind = (rdtsc()) % IHK_TLB_FLUSH_IRQ_VECTOR_SIZE;
-	}
-
-	flush_entry = &tlb_flush_vector[flush_ind];
-
-	/* Take a copy of the cpu set so that we don't hold the lock
-	 * all the way while interrupting other cores */
-	ihk_mc_spinlock_lock_noirq(&vm->address_space->cpu_set_lock);
-	memcpy(&_cpu_set, &vm->address_space->cpu_set, sizeof(cpu_set_t));
-	ihk_mc_spinlock_unlock_noirq(&vm->address_space->cpu_set_lock);
-
-	dkprintf("trying to aquire flush_entry->lock flush_ind: %d\n", flush_ind);
-
-	ihk_mc_spinlock_lock_noirq(&flush_entry->lock);
-
-	flush_entry->vm = vm;
-	flush_entry->addr = addr;
-	flush_entry->nr_addr = nr_addr;
-	ihk_atomic_set(&flush_entry->pending, 0);
-
-	dkprintf("lock aquired, iterating cpu mask.. flush_ind: %d\n", flush_ind);
-
-	/* Loop through CPUs in this address space and interrupt them for
-	 * TLB flush on the specified address */
-	for_each_set_bit(cpu, (const unsigned long*)&_cpu_set.__bits, CPU_SETSIZE) {
-
-		if (ihk_mc_get_processor_id() == cpu)
-			continue;
-
-		ihk_atomic_inc(&flush_entry->pending);
-		dkprintf("remote_flush_tlb_cpumask: flush_ind: %d, addr: 0x%lX, interrupting cpu: %d\n",
-		        flush_ind, addr, cpu);
-
-#ifdef POSTK_DEBUG_ARCH_DEP_8 /* arch depend hide */
-		/* TODO(pka_idke) Interim support */
-		ihk_mc_interrupt_cpu(cpu, 
-				     ihk_mc_get_vector(flush_ind + IHK_TLB_FLUSH_IRQ_VECTOR_START));
-#else /* POSTK_DEBUG_ARCH_DEP_8 */
-		ihk_mc_interrupt_cpu(get_x86_cpu_local_variable(cpu)->apic_id,
-		                     flush_ind + IHK_TLB_FLUSH_IRQ_VECTOR_START);
-#endif /* POSTK_DEBUG_ARCH_DEP_8 */
-	}
-
-#ifdef DEBUG_IC_TLB
-	{
-		unsigned long tsc;
-		tsc = rdtsc() + 12884901888;  /* 1.2GHz =>10 sec */
-#endif
-		if (flush_entry->addr[0]) {
-			int i;
-
-			for (i = 0; i < flush_entry->nr_addr; ++i) {
-				flush_tlb_single(flush_entry->addr[i] & PAGE_MASK);
-			}
-		}
-		/* Zero address denotes full TLB flush */
-		else {
-			flush_tlb();
-		}
-
-		/* Wait for all cores */
-		while (ihk_atomic_read(&flush_entry->pending) != 0) {
-			cpu_pause();
-
-#ifdef DEBUG_IC_TLB
-			if (rdtsc() > tsc) {
-				kprintf("waited 10 secs for remote TLB!! -> panic_all()\n");
-				panic_all_cores("waited 10 secs for remote TLB!!\n");
-			}
-#endif
-		}
-#ifdef DEBUG_IC_TLB
-	}
-#endif
-
-	ihk_mc_spinlock_unlock_noirq(&flush_entry->lock);
-}
-#endif /* POSTK_DEBUG_ARCH_DEP_8 */
-
 static void remote_flush_tlb_add_addr(struct clear_range_args *args,
 		unsigned long addr)
 {
@@ -1622,7 +1519,7 @@ static int clear_range_l3(void *args0, pte_t *ptep, uint64_t base,
 {
 	struct clear_range_args *args = args0;
 	int error;
-	uint64_t phys;
+	uint64_t phys = 0;
 	pte_t old;
 	struct page *page;
 	struct page_table *pt;
@@ -2572,10 +2469,10 @@ static void init_linux_kernel_mapping(struct page_table *pt)
 		map_start = 0;
 		map_end = 0x20000000000;
 
-		virt = (void *)LINUX_PAGE_OFFSET;
+		virt = (void *)linux_page_offset_base;
 
 		kprintf("Linux kernel virtual: 0x%lx - 0x%lx -> 0x%lx - 0x%lx\n",
-			LINUX_PAGE_OFFSET, LINUX_PAGE_OFFSET + map_end, 0, map_end);
+			virt, virt + map_end, 0, map_end);
 
 		for (phys = map_start; phys < map_end; phys += LARGE_PAGE_SIZE) {
 			if (set_pt_large_page(pt, virt, phys, PTATTR_WRITABLE) != 0) {
@@ -2599,9 +2496,11 @@ static void init_linux_kernel_mapping(struct page_table *pt)
 			}
 
 			dkprintf("Linux kernel virtual: 0x%lx - 0x%lx -> 0x%lx - 0x%lx\n",
-					 LINUX_PAGE_OFFSET + map_start, LINUX_PAGE_OFFSET + map_end, map_start, map_end);
+				 linux_page_offset_base + map_start,
+				 linux_page_offset_base + map_end,
+				 map_start, map_end);
 
-			virt = (void *)(LINUX_PAGE_OFFSET + map_start);
+			virt = (void *)(linux_page_offset_base + map_start);
 			for (phys = map_start; phys < map_end; phys += LARGE_PAGE_SIZE, virt += LARGE_PAGE_SIZE) {
 				if (set_pt_large_page(pt, virt, phys, PTATTR_WRITABLE) != 0) {
 					kprintf("%s: set_pt_large_page() failed for 0x%lx\n", __FUNCTION__, virt);
@@ -2652,7 +2551,7 @@ void *map_fixed_area(unsigned long phys, unsigned long size, int uncachable)
 		attr |= PTATTR_UNCACHABLE;
 	}
 
-	kprintf("map_fixed: phys: 0x%lx => 0x%lx (%d pages)\n",
+	dkprintf("map_fixed: phys: 0x%lx => 0x%lx (%d pages)\n",
 			paligned, v, npages);
 
 	for (i = 0; i < npages; i++) {
@@ -2745,12 +2644,12 @@ unsigned long virt_to_phys(void *v)
 	unsigned long va = (unsigned long)v;
 
 	if (va >= MAP_KERNEL_START) {
-		dkprintf("%s: MAP_KERNEL_START <= 0x%lx <= LINUX_PAGE_OFFSET\n",
+		dkprintf("%s: MAP_KERNEL_START <= 0x%lx <= linux_page_offset_base\n",
 				__FUNCTION__, va);
 		return va - MAP_KERNEL_START + x86_kernel_phys_base;
 	}
-	else if (va >= LINUX_PAGE_OFFSET) {
-		return va - LINUX_PAGE_OFFSET;
+	else if (va >= linux_page_offset_base) {
+		return va - linux_page_offset_base;
 	}
 	else if (va >= MAP_FIXED_START) {
 		return va - MAP_FIXED_START;
@@ -2769,7 +2668,7 @@ void *phys_to_virt(unsigned long p)
 		return (void *)(p + MAP_ST_START);
 	}
 
-	return (void *)(p + LINUX_PAGE_OFFSET);
+	return (void *)(p + linux_page_offset_base);
 }
 
 int copy_from_user(void *dst, const void *src, size_t siz)

arch/x86_64/kernel/perfctr.c

@@ -12,12 +12,12 @@
 #include <march.h>
 #include <errno.h>
 #include <cls.h>
-#include <ihk/debug.h>
 #include <ihk/cpu.h>
 #include <registers.h>
 #include <mc_perf_event.h>
 #include <config.h>
-#include <debug.h>
+#include <ihk/debug.h>
+#include <process.h>
 
 extern unsigned int *x86_march_perfmap;
 extern int running_on_kvm(void);
@@ -223,41 +223,6 @@ int ihk_mc_perfctr_init_raw(int counter, unsigned int code, int mode)
 #endif /*POSTK_DEBUG_TEMP_FIX_29*/
 }
 
-#ifdef POSTK_DEBUG_TEMP_FIX_29
-int ihk_mc_perfctr_init(int counter, uint64_t config, int mode)
-#else
-int ihk_mc_perfctr_init(int counter, enum ihk_perfctr_type type, int mode)
-#endif /*POSTK_DEBUG_TEMP_FIX_29*/
-{
-#ifdef POSTK_DEBUG_TEMP_FIX_29
-	enum ihk_perfctr_type type;
-
-	switch (config) {
-	case PERF_COUNT_HW_CPU_CYCLES :
-		type = APT_TYPE_CYCLE;
-		break;
-	case PERF_COUNT_HW_INSTRUCTIONS :
-		type = APT_TYPE_INSTRUCTIONS;
-		break;
-	default :
-		// Not supported config.
-		type = PERFCTR_MAX_TYPE;
-	}
-#endif /*POSTK_DEBUG_TEMP_FIX_29*/
-
-	if (counter < 0 || counter >= NUM_PERF_COUNTERS) {
-		return -EINVAL;
-	}
-	if (type < 0 || type >= PERFCTR_MAX_TYPE) {
-		return -EINVAL;
-	}
-	if (!x86_march_perfmap[type]) {
-		return -EINVAL;
-	}
-
-	return set_perfctr_x86_direct(counter, mode, x86_march_perfmap[type]);
-}
-
 int ihk_mc_perfctr_set_extra(struct mc_perf_event *event)
 {
 	struct thread *thread = cpu_local_var(current);
@@ -412,6 +377,23 @@ int ihk_mc_perfctr_read_mask(unsigned long counter_mask, unsigned long *value)
 	return 0;
 }
 
+int ihk_mc_perfctr_alloc(struct thread *thread, struct mc_perf_event *event)
+{
+	int ret = -EINVAL;
+	int i = 0;
+	const int counters = ihk_mc_perf_get_num_counters();
+
+	// find avail generic counter
+	for (i = 0; i < counters; i++) {
+		if (!(thread->pmc_alloc_map & (1 << i))) {
+			ret = i;
+			break;
+		}
+	}
+
+	return ret;
+}
+
 unsigned long ihk_mc_perfctr_read(int counter)
 {
 	unsigned long retval = 0;
@@ -439,6 +421,14 @@ unsigned long ihk_mc_perfctr_read(int counter)
 	return retval;
 }
 
+unsigned long ihk_mc_perfctr_value(int counter, unsigned long correction)
+{
+	unsigned long count = ihk_mc_perfctr_read(counter) + correction;
+
+	count &= 0x000000ffffffffffL;
+	return count;
+}
+
 // read by rdmsr
 unsigned long ihk_mc_perfctr_read_msr(int counter)
 {
@@ -513,3 +503,18 @@ int ihk_mc_perf_get_num_counters(void)
 {
 	return NUM_PERF_COUNTERS;
 }
+
+int hw_perf_event_init(struct mc_perf_event *event)
+{
+	return 0;
+}
+
+int ihk_mc_event_set_period(struct mc_perf_event *event)
+{
+	return 0;
+}
+
+uint64_t ihk_mc_event_update(struct mc_perf_event *event)
+{
+	return 0;
+}

arch/x86_64/kernel/syscall.c

@@ -16,7 +16,6 @@
  */
 
 #include <ihk/cpu.h>
-#include <ihk/debug.h>
 #include <cls.h>
 #include <cpulocal.h>
 #include <syscall.h>
@@ -32,7 +31,8 @@
 #include <page.h>
 #include <limits.h>
 #include <syscall.h>
-#include <debug.h>
+#include <rusage_private.h>
+#include <ihk/debug.h>
 
 void terminate_mcexec(int, int);
 extern long do_sigaction(int sig, struct k_sigaction *act, struct k_sigaction *oact);
@@ -64,7 +64,6 @@ uintptr_t debug_constants[] = {
 	-1,
 };
 
-#ifdef POSTK_DEBUG_ARCH_DEP_52
 #define VDSO_MAXPAGES 2
 struct vdso {
 	long busy;
@@ -80,8 +79,24 @@ struct vdso {
 	long hpet_phys;
 	void *pvti_virt;
 	long pvti_phys;
+	void *vgtod_virt;
+};
+
+struct vsyscall_gtod_data {
+	int seq;
+
+	struct {
+		int vclock_mode;
+		unsigned long cycle_last;
+		unsigned long mask;
+		unsigned int mult;
+		unsigned int shift;
+	} clock;
+
+	/* open coded 'struct timespec' */
+	time_t wall_time_sec;
+	unsigned long wall_time_snsec;
 };
-#endif /*POSTK_DEBUG_ARCH_DEP_52*/
 
 static struct vdso vdso;
 static size_t container_size = 0;
@@ -132,44 +147,6 @@ int obtain_clone_cpuid(cpu_set_t *cpu_set, int use_last) {
     return min_cpu;
 }
 
-int
-arch_clear_host_user_space()
-{
-	struct thread *th = cpu_local_var(current);
-
-	/* XXX: might be unnecessary */
-	clear_host_pte(th->vm->region.user_start,
-			(th->vm->region.user_end - th->vm->region.user_start));
-	return 0;
-}
-
-SYSCALL_DECLARE(rt_sigaction)
-{
-	int sig = ihk_mc_syscall_arg0(ctx);
-	const struct sigaction *act = (const struct sigaction *)ihk_mc_syscall_arg1(ctx);
-	struct sigaction *oact = (struct sigaction *)ihk_mc_syscall_arg2(ctx);
-	size_t sigsetsize = ihk_mc_syscall_arg3(ctx);
-	struct k_sigaction new_sa, old_sa;
-	int rc;
-
-	if (sigsetsize != sizeof(sigset_t))
-		return -EINVAL;
-
-	if(act)
-		if(copy_from_user(&new_sa.sa, act, sizeof new_sa.sa)){
-			goto fault;
-		}
-	rc = do_sigaction(sig, act? &new_sa: NULL, oact? &old_sa: NULL);
-	if(rc == 0 && oact)
-		if(copy_to_user(oact, &old_sa.sa, sizeof old_sa.sa)){
-			goto fault;
-		}
-
-	return rc;
-fault:
-	return -EFAULT;
-}
-
 SYSCALL_DECLARE(prctl)
 {
 	struct process *proc = cpu_local_var(current)->proc;
@@ -558,7 +535,7 @@ long ptrace_write_regset(struct thread *thread, long type, struct iovec *iov)
 	return rc;
 }
 
-extern void coredump(struct thread *thread, void *regs);
+extern int coredump(struct thread *thread, void *regs, int sig);
 
 void ptrace_report_signal(struct thread *thread, int sig)
 {
@@ -726,6 +703,7 @@ do_signal(unsigned long rc, void *regs0, struct thread *thread, struct sig_pendi
 	struct mcs_rwlock_node_irqsave lock;
 	struct mcs_rwlock_node_irqsave mcs_rw_node;
 	int restart = 0;
+	int ret;
 
 	for(w = pending->sigmask.__val[0], sig = 0; w; sig++, w >>= 1);
 	dkprintf("do_signal(): tid=%d, pid=%d, sig=%d\n", thread->tid, proc->pid, sig);
@@ -971,15 +949,6 @@ do_signal(unsigned long rc, void *regs0, struct thread *thread, struct sig_pendi
 			dkprintf("SIGTRAP(): woken up\n");
 			break;
 		case SIGCONT:
-			memset(&info, '\0', sizeof info);
-			info.si_signo = SIGCHLD;
-			info.si_code = CLD_CONTINUED;
-			info._sifields._sigchld.si_pid = proc->pid;
-			info._sifields._sigchld.si_status = 0x0000ffff;
-			do_kill(cpu_local_var(current), proc->parent->pid, -1, SIGCHLD, &info, 0);
-			proc->main_thread->signal_flags = SIGNAL_STOP_CONTINUED;
-			proc->status = PS_RUNNING;
-			dkprintf("do_signal,SIGCONT,do nothing\n");
 			break;
 		case SIGQUIT:
 		case SIGILL:
@@ -991,9 +960,31 @@ do_signal(unsigned long rc, void *regs0, struct thread *thread, struct sig_pendi
 		case SIGXCPU:
 		case SIGXFSZ:
 		core:
-			dkprintf("do_signal,default,core,sig=%d\n", sig);
-			coredump(thread, regs);
-			coredumped = 0x80;
+			thread->coredump_regs =
+				kmalloc(sizeof(struct x86_user_context),
+					IHK_MC_AP_NOWAIT);
+			if (!thread->coredump_regs) {
+				kprintf("%s: Out of memory\n", __func__);
+				goto skip;
+			}
+			memcpy(thread->coredump_regs, regs,
+			       sizeof(struct x86_user_context));
+
+			ret = coredump(thread, regs, sig);
+			switch (ret) {
+			case -EBUSY:
+				kprintf("%s: INFO: coredump not performed, try ulimit -c <non-zero>\n",
+					__func__);
+				break;
+			case 0:
+				coredumped = 0x80;
+				break;
+			default:
+				kprintf("%s: ERROR: coredump failed (%d)\n",
+					__func__, ret);
+				break;
+			}
+skip:
 			terminate(0, sig | coredumped);
 			break;
 		case SIGCHLD:
@@ -1010,80 +1001,6 @@ out:
 	return restart;
 }
 
-static struct sig_pending *
-getsigpending(struct thread *thread, int delflag){
-	struct list_head *head;
-	mcs_rwlock_lock_t *lock;
-	struct mcs_rwlock_node_irqsave mcs_rw_node;
-	struct sig_pending *next;
-	struct sig_pending *pending;
-	__sigset_t w;
-	__sigset_t x;
-	int sig;
-	struct k_sigaction *k;
-
-	w = thread->sigmask.__val[0];
-
-	lock = &thread->sigcommon->lock;
-	head = &thread->sigcommon->sigpending;
-	for(;;) {
-		if (delflag) {
-			mcs_rwlock_writer_lock(lock, &mcs_rw_node);
-		}
-		else {
-			mcs_rwlock_reader_lock(lock, &mcs_rw_node);
-		}
-
-		list_for_each_entry_safe(pending, next, head, list){
-			for(x = pending->sigmask.__val[0], sig = 0; x; sig++, x >>= 1);
-			k = thread->sigcommon->action + sig - 1;
-			if(delflag ||
-			   (sig != SIGCHLD && sig != SIGURG) ||
-			   (k->sa.sa_handler != (void *)1 &&
-			    k->sa.sa_handler != NULL)){
-				if(!(pending->sigmask.__val[0] & w)){
-					if(delflag) 
-						list_del(&pending->list);
-
-					if (delflag) {
-						mcs_rwlock_writer_unlock(lock, &mcs_rw_node);
-					}
-					else {
-						mcs_rwlock_reader_unlock(lock, &mcs_rw_node);
-					}
-					return pending;
-				}
-			}
-		}
-
-		if (delflag) {
-			mcs_rwlock_writer_unlock(lock, &mcs_rw_node);
-		}
-		else {
-			mcs_rwlock_reader_unlock(lock, &mcs_rw_node);
-		}
-
-		if(lock == &thread->sigpendinglock)
-			return NULL;
-
-		lock = &thread->sigpendinglock;
-		head = &thread->sigpending;
-	}
-
-	return NULL;
-}
-
-struct sig_pending *
-hassigpending(struct thread *thread)
-{
-	if (list_empty(&thread->sigpending) &&
-	    list_empty(&thread->sigcommon->sigpending)) {
-		return NULL;
-	}
-
-	return getsigpending(thread, 0);
-}
-
 int
 interrupt_from_user(void *regs0)
 {
@@ -1098,170 +1015,6 @@ void save_syscall_return_value(int num, unsigned long rc)
 	return;
 }
 
-/** \brief check arrived signals and processing
- *
- * @param rc    return value of syscall
- * @param regs0 context
- * @param num   syscall number (-1: Not called on exiting system call)
- */
-void
-check_signal(unsigned long rc, void *regs0, int num)
-{
-	struct x86_user_context *regs = regs0;
-	struct thread *thread;
-	struct sig_pending *pending;
-	int	irqstate;
-
-	if(clv == NULL)
-		return;
-	thread = cpu_local_var(current);
-
-	if(thread == NULL || thread == &cpu_local_var(idle)){
-		struct thread *t;
-
-		irqstate = ihk_mc_spinlock_lock(&(cpu_local_var(runq_lock)));
-		list_for_each_entry(t, &(cpu_local_var(runq)), sched_list){
-			if(t == &cpu_local_var(idle))
-				continue;
-			if(t->status == PS_INTERRUPTIBLE &&
-			   hassigpending(t)){
-				t->status = PS_RUNNING;
-				break;
-			}
-		}
-		ihk_mc_spinlock_unlock(&(cpu_local_var(runq_lock)), irqstate);
-		goto out;
-	}
-
-	if(regs != NULL && !interrupt_from_user(regs)) {
-		goto out;
-	}
-
-	if (list_empty(&thread->sigpending) &&
-	    list_empty(&thread->sigcommon->sigpending)) {
-		goto out;
-	}
-
-	for(;;){
-		pending = getsigpending(thread, 1);
-		if(!pending) {
-			dkprintf("check_signal,queue is empty\n");
-			goto out;
-		}
-
-		if (do_signal(rc, regs, thread, pending, num)) {
-			num = -1;
-		}
-	}
-
-out:
-	return;
-}
-
-static int
-check_sig_pending_thread(struct thread *thread)
-{
-	int found = 0;
-	struct list_head *head;
-	mcs_rwlock_lock_t *lock;
-	struct mcs_rwlock_node_irqsave mcs_rw_node;
-	struct sig_pending *next;
-	struct sig_pending *pending;
-	__sigset_t w;
-	__sigset_t x;
-	int sig = 0;
-	struct k_sigaction *k;
-	struct cpu_local_var *v;
-
-	v = get_this_cpu_local_var();
-	w = thread->sigmask.__val[0];
-
-	lock = &thread->sigcommon->lock;
-	head = &thread->sigcommon->sigpending;
-	for (;;) {
-		mcs_rwlock_reader_lock(lock, &mcs_rw_node);
-
-		list_for_each_entry_safe(pending, next, head, list){
-			for (x = pending->sigmask.__val[0], sig = 0; x;
-			     sig++, x >>= 1);
-			k = thread->sigcommon->action + sig - 1;
-			if ((sig != SIGCHLD && sig != SIGURG) ||
-			    (k->sa.sa_handler != (void *)1 &&
-			     k->sa.sa_handler != NULL)) {
-				if (!(pending->sigmask.__val[0] & w)) {
-					if (pending->interrupted == 0) {
-						pending->interrupted = 1;
-						found = 1;
-						if (sig != SIGCHLD &&
-						    sig != SIGURG &&
-						    !k->sa.sa_handler) {
-							found = 2;
-							break;
-						}
-					}
-				}
-			}
-		}
-
-		mcs_rwlock_reader_unlock(lock, &mcs_rw_node);
-
-		if (found == 2) {
-			break;
-		}
-
-		if (lock == &thread->sigpendinglock) {
-			break;
-		}
-
-		lock = &thread->sigpendinglock;
-		head = &thread->sigpending;
-	}
-
-	if (found == 2) {
-		ihk_mc_spinlock_unlock(&v->runq_lock, v->runq_irqstate);
-		terminate_mcexec(0, sig);
-		return 1;
-	}
-	else if (found == 1) {
-		ihk_mc_spinlock_unlock(&v->runq_lock, v->runq_irqstate);
-		interrupt_syscall(thread, 0);
-		return 1;
-	}
-	return 0;
-}
-
-void
-check_sig_pending(void)
-{
-	struct thread *thread;
-	struct cpu_local_var *v;
-
-	if (clv == NULL)
-		return;
-
-	v = get_this_cpu_local_var();
-repeat:
-	v->runq_irqstate = ihk_mc_spinlock_lock(&v->runq_lock);
-	list_for_each_entry(thread, &(v->runq), sched_list) {
-
-		if (thread == NULL || thread == &cpu_local_var(idle)) {
-			continue;
-		}
-
-		if (thread->in_syscall_offload == 0) {
-			continue;
-		}
-
-		if (thread->proc->group_exit_status & 0x0000000100000000L) {
-			continue;
-		}
-
-		if (check_sig_pending_thread(thread))
-			goto repeat;
-	}
-	ihk_mc_spinlock_unlock(&v->runq_lock, v->runq_irqstate);
-}
-
 unsigned long
 do_kill(struct thread *thread, int pid, int tid, int sig, siginfo_t *info,
         int ptracecont)
@@ -1278,7 +1031,6 @@ do_kill(struct thread *thread, int pid, int tid, int sig, siginfo_t *info,
 	struct list_head *head = NULL;
 	int rc;
 	unsigned long irqstate = 0;
-	struct k_sigaction *k;
 	int doint;
 	int found = 0;
 	siginfo_t info0;
@@ -1288,6 +1040,7 @@ do_kill(struct thread *thread, int pid, int tid, int sig, siginfo_t *info,
 	struct process_hash *phash = rset->process_hash;
 	struct mcs_rwlock_node lock;
 	struct mcs_rwlock_node updatelock;
+	struct sig_pending *pending = NULL;
 
 	if(sig > 64 || sig < 0)
 		return -EINVAL;
@@ -1509,54 +1262,70 @@ done:
 
 	mcs_rwlock_writer_lock_noirq(savelock, &mcs_rw_node);
 
-	/* Put signal event even when handler is SIG_IGN or SIG_DFL
-	   because target ptraced thread must call ptrace_report_signal 
-	   in check_signal */
 	rc = 0;
-	k = tthread->sigcommon->action + sig - 1;
-	if ((sig != SIGKILL && (tthread->ptrace & PT_TRACED)) ||
-	    (k->sa.sa_handler != (void *)1 &&
-	     (k->sa.sa_handler != NULL ||
-	      (sig != SIGCHLD && sig != SIGURG)))) {
-		struct sig_pending *pending = NULL;
-		if (sig < 33) { // SIGRTMIN - SIGRTMAX
-			list_for_each_entry(pending, head, list){
-				if(pending->sigmask.__val[0] == mask &&
-				   pending->ptracecont == ptracecont)
-					break;
-			}
-			if(&pending->list == head)
-				pending = NULL;
+
+	if (sig < 33) { // SIGRTMIN - SIGRTMAX
+		list_for_each_entry(pending, head, list) {
+			if (pending->sigmask.__val[0] == mask &&
+			    pending->ptracecont == ptracecont)
+				break;
 		}
-		if(pending == NULL){
-			doint = 1;
-			pending = kmalloc(sizeof(struct sig_pending), IHK_MC_AP_NOWAIT);
-			if(!pending){
-				rc = -ENOMEM;
-			}
-			else{
-				memset(pending, 0, sizeof(struct sig_pending));
-				pending->sigmask.__val[0] = mask;
-				memcpy(&pending->info, info, sizeof(siginfo_t));
-				pending->ptracecont = ptracecont;
-				if(sig == SIGKILL || sig == SIGSTOP)
-					list_add(&pending->list, head);
-				else
-					list_add_tail(&pending->list, head);
-				tthread->sigevent = 1;
-			}
+		if (&pending->list == head)
+			pending = NULL;
+	}
+	if (pending == NULL) {
+		doint = 1;
+		pending = kmalloc(sizeof(struct sig_pending), IHK_MC_AP_NOWAIT);
+		if (!pending) {
+			rc = -ENOMEM;
+		}
+		else {
+			memset(pending, 0, sizeof(struct sig_pending));
+			pending->sigmask.__val[0] = mask;
+			memcpy(&pending->info, info, sizeof(siginfo_t));
+			pending->ptracecont = ptracecont;
+			if (sig == SIGKILL || sig == SIGSTOP)
+				list_add(&pending->list, head);
+			else
+				list_add_tail(&pending->list, head);
+			tthread->sigevent = 1;
 		}
 	}
+
 	mcs_rwlock_writer_unlock_noirq(savelock, &mcs_rw_node);
 	cpu_restore_interrupt(irqstate);
 
+	if (sig == SIGCONT || ptracecont == 1) {
+		/* Wake up the target only when stopped by SIGSTOP */
+		if (sched_wakeup_thread(tthread, PS_STOPPED) == 0) {
+			struct siginfo info;
+
+			tthread->proc->main_thread->signal_flags =
+							SIGNAL_STOP_CONTINUED;
+			tthread->proc->status = PS_RUNNING;
+			memset(&info, '\0', sizeof(info));
+			info.si_signo = SIGCHLD;
+			info.si_code = CLD_CONTINUED;
+			info._sifields._sigchld.si_pid = tthread->proc->pid;
+			info._sifields._sigchld.si_status = 0x0000ffff;
+			do_kill(tthread, tthread->proc->parent->pid, -1,
+							SIGCHLD, &info, 0);
+			tthread->proc->status = PS_RUNNING;
+			if (thread != tthread) {
+				ihk_mc_interrupt_cpu(tthread->cpu_id,
+						ihk_mc_get_vector(IHK_GV_IKC));
+			}
+			doint = 0;
+		}
+	}
 	if (doint && !(mask & tthread->sigmask.__val[0])) {
 		int status = tthread->status;
 
 		if (thread != tthread) {
 			dkprintf("do_kill,ipi,pid=%d,cpu_id=%d\n",
 				 tproc->pid, tthread->cpu_id);
-			ihk_mc_interrupt_cpu(get_x86_cpu_local_variable(tthread->cpu_id)->apic_id, 0xd0);
+			ihk_mc_interrupt_cpu(tthread->cpu_id,
+					ihk_mc_get_vector(IHK_GV_IKC));
 		}
 
 		if (status != PS_RUNNING) {
@@ -1564,11 +1333,6 @@ done:
 				/* Wake up the target only when stopped by ptrace-reporting */
 				sched_wakeup_thread(tthread, PS_TRACED | PS_STOPPED | PS_INTERRUPTIBLE);
 			}
-			else if(sig == SIGCONT || ptracecont == 1){
-				/* Wake up the target only when stopped by SIGSTOP */
-				sched_wakeup_thread(tthread, PS_STOPPED);
-				tthread->proc->status = PS_RUNNING;
-			}
 			else {
 				sched_wakeup_thread(tthread, PS_INTERRUPTIBLE);
 			}
@@ -1593,7 +1357,7 @@ set_signal(int sig, void *regs0, siginfo_t *info)
         }
 
 	if ((__sigmask(sig) & thread->sigmask.__val[0])) {
-		coredump(thread, regs0);
+		coredump(thread, regs0, sig);
 		terminate(0, sig | 0x80);
 	}
 	do_kill(thread, thread->proc->pid, thread->tid, sig, info, 0);
@@ -1629,7 +1393,7 @@ SYSCALL_DECLARE(mmap)
 		;
 
 	const uintptr_t addr0 = ihk_mc_syscall_arg0(ctx);
-	const size_t len0 = ihk_mc_syscall_arg1(ctx);
+	size_t len0 = ihk_mc_syscall_arg1(ctx);
 	const int prot = ihk_mc_syscall_arg2(ctx);
 	const int flags0 = ihk_mc_syscall_arg3(ctx);
 	const int fd = ihk_mc_syscall_arg4(ctx);
@@ -1668,7 +1432,9 @@ SYSCALL_DECLARE(mmap)
 	if (flags & MAP_HUGETLB) {
 		switch (flags & (0x3F << MAP_HUGE_SHIFT)) {
 		case 0:
-			flags |= MAP_HUGE_2MB;	/* default hugepage size */
+			/* default hugepage size */
+			flags |= ihk_mc_get_linux_default_huge_page_shift() <<
+				MAP_HUGE_SHIFT;
 			break;
 
 		case MAP_HUGE_2MB:
@@ -1684,16 +1450,29 @@ SYSCALL_DECLARE(mmap)
 		}
 
 		pgsize = (size_t)1 << ((flags >> MAP_HUGE_SHIFT) & 0x3F);
+		/* Round-up map length by pagesize */
+		len0 = ALIGN(len0, pgsize);
+
+		if (rusage_check_overmap(len0,
+				(flags >> MAP_HUGE_SHIFT) & 0x3F)) {
+			error = -ENOMEM;
+			goto out;
+		}
 	}
 
 #define	VALID_DUMMY_ADDR	((region->user_start + PTL3_SIZE - 1) & ~(PTL3_SIZE - 1))
-	addr = (flags & MAP_FIXED)? addr0: VALID_DUMMY_ADDR;
+	addr = addr0;
 	len = (len0 + pgsize - 1) & ~(pgsize - 1);
+recheck:
 	if ((addr & (pgsize - 1))
 			|| (len == 0)
 			|| !(flags & (MAP_SHARED | MAP_PRIVATE))
 			|| ((flags & MAP_SHARED) && (flags & MAP_PRIVATE))
 			|| (off0 & (pgsize - 1))) {
+		if (!(flags & MAP_FIXED) && addr != VALID_DUMMY_ADDR) {
+			addr = VALID_DUMMY_ADDR;
+			goto recheck;
+		}
 		ekprintf("sys_mmap(%lx,%lx,%x,%x,%x,%lx):EINVAL\n",
 				addr0, len0, prot, flags0, fd, off0);
 		error = -EINVAL;
@@ -1703,6 +1482,10 @@ SYSCALL_DECLARE(mmap)
 	if (addr < region->user_start
 			|| region->user_end <= addr
 			|| len > (region->user_end - region->user_start)) {
+		if (!(flags & MAP_FIXED) && addr != VALID_DUMMY_ADDR) {
+			addr = VALID_DUMMY_ADDR;
+			goto recheck;
+		}
 		ekprintf("sys_mmap(%lx,%lx,%x,%x,%x,%lx):ENOMEM\n",
 				addr0, len0, prot, flags0, fd, off0);
 		error = -ENOMEM;
@@ -1730,10 +1513,20 @@ out:
 
 SYSCALL_DECLARE(clone)
 {
-    return do_fork((int)ihk_mc_syscall_arg0(ctx), ihk_mc_syscall_arg1(ctx),
+	struct process *proc = cpu_local_var(current)->proc;
+	struct mcs_rwlock_node_irqsave lock_dump;
+	unsigned long ret;
+
+	/* mutex coredump */
+	mcs_rwlock_reader_lock(&proc->coredump_lock, &lock_dump);
+
+	ret =  do_fork((int)ihk_mc_syscall_arg0(ctx), ihk_mc_syscall_arg1(ctx),
                    ihk_mc_syscall_arg2(ctx), ihk_mc_syscall_arg3(ctx),
                    ihk_mc_syscall_arg4(ctx), ihk_mc_syscall_pc(ctx),
                    ihk_mc_syscall_sp(ctx));
+
+	mcs_rwlock_reader_unlock(&proc->coredump_lock, &lock_dump);
+	return ret;
 }
 
 SYSCALL_DECLARE(fork)
@@ -1761,7 +1554,9 @@ SYSCALL_DECLARE(shmget)
 		int hugeshift = shmflg & (0x3F << SHM_HUGE_SHIFT);
 
 		if (hugeshift == 0) {
-			shmflg |= SHM_HUGE_2MB;	/* default hugepage size */
+			/* default hugepage size */
+			shmflg |= ihk_mc_get_linux_default_huge_page_shift() <<
+				MAP_HUGE_SHIFT;
 		} else if (hugeshift == SHM_HUGE_2MB ||
 			   hugeshift == SHM_HUGE_1GB) {
 			/*nop*/
@@ -2210,7 +2005,7 @@ int do_process_vm_read_writev(int pid,
 	}
 
 	/* Check if parameters are okay */
-	ihk_mc_spinlock_lock_noirq(&lthread->vm->memory_range_lock);
+	ihk_rwspinlock_read_lock_noirq(&lthread->vm->memory_range_lock);
 
 	range = lookup_process_memory_range(lthread->vm, 
 			(uintptr_t)local_iov, 
@@ -2232,7 +2027,7 @@ int do_process_vm_read_writev(int pid,
 
 	ret = 0;
 arg_out:
-	ihk_mc_spinlock_unlock_noirq(&lthread->vm->memory_range_lock);
+	ihk_rwspinlock_read_unlock_noirq(&lthread->vm->memory_range_lock);
 
 	if (ret != 0) {
 		goto out;
@@ -2301,7 +2096,7 @@ arg_out:
 		if (pli != li) {
 			struct vm_range *range;
 
-			ihk_mc_spinlock_lock_noirq(&lthread->vm->memory_range_lock);
+			ihk_rwspinlock_read_lock_noirq(&lthread->vm->memory_range_lock);
 
 			/* Is base valid? */
 			range = lookup_process_memory_range(lthread->vm,
@@ -2331,7 +2126,7 @@ arg_out:
 
 			ret = 0;
 pli_out:
-			ihk_mc_spinlock_unlock_noirq(&lthread->vm->memory_range_lock);
+			ihk_rwspinlock_read_unlock_noirq(&lthread->vm->memory_range_lock);
 
 			if (ret != 0) {
 				goto out;
@@ -2344,7 +2139,7 @@ pli_out:
 		if (pri != ri) {
 			struct vm_range *range;
 
-			ihk_mc_spinlock_lock_noirq(&rvm->memory_range_lock);
+			ihk_rwspinlock_read_lock_noirq(&rvm->memory_range_lock);
 
 			/* Is base valid? */
 			range = lookup_process_memory_range(rvm,
@@ -2374,7 +2169,7 @@ pli_out:
 
 			ret = 0;
 pri_out:
-			ihk_mc_spinlock_unlock_noirq(&rvm->memory_range_lock);
+			ihk_rwspinlock_read_unlock_noirq(&rvm->memory_range_lock);
 
 			if (ret != 0) {
 				goto out;
@@ -2811,4 +2606,46 @@ time_t time(void) {
 	return ret;
 }
 
+void calculate_time_from_tsc(struct timespec *ts)
+{
+	unsigned long seq;
+	unsigned long seq2;
+	unsigned long ns;
+	unsigned long delta;
+	struct vsyscall_gtod_data *gtod = vdso.vgtod_virt;
+
+	do {
+		for (;;) {
+			seq = ACCESS_ONCE(gtod->seq);
+			if (unlikely(seq & 1)) {
+				cpu_pause();
+				continue;
+			}
+			break;
+		}
+		rmb(); /* fetch sequence before time */
+		ts->tv_sec = gtod->wall_time_sec;
+		ns = gtod->wall_time_snsec;
+		delta = rdtsc() - gtod->clock.cycle_last;
+		ns += delta * gtod->clock.mult;
+		ns >>= gtod->clock.shift;
+		seq2 = ACCESS_ONCE(gtod->seq);
+		rmb(); /* fetch time before checking sequence */
+	} while (seq != seq2);
+	ts->tv_nsec = ns;
+
+	if (ts->tv_nsec >= NS_PER_SEC) {
+		ts->tv_nsec -= NS_PER_SEC;
+		++ts->tv_sec;
+	}
+}
+
+extern void ptrace_syscall_event(struct thread *thread);
+long arch_ptrace_syscall_event(struct thread *thread,
+				ihk_mc_user_context_t *ctx, long setret)
+{
+	ihk_mc_syscall_ret(ctx) = setret;
+	ptrace_syscall_event(thread);
+	return ihk_mc_syscall_ret(ctx);
+}
 /*** End of File ***/

arch/x86_64/tools/irqbalance_mck.service

@@ -1,10 +0,0 @@
-[Unit]
-Description=irqbalance daemon
-After=syslog.target
-
-[Service]
-EnvironmentFile=/tmp/irqbalance_mck
-ExecStart=/usr/sbin/irqbalance --foreground $IRQBALANCE_ARGS
-
-[Install]
-WantedBy=multi-user.target

arch/x86_64/tools/mcoverlay-create-smp-x86.sh.in

@@ -1,150 +0,0 @@
-# mcoverlay-create-smp-x86.sh.in COPYRIGHT FUJITSU LIMITED 2018
-# Overlay /proc, /sys with McKernel specific contents
-
-#
-# Revert any state that has been initialized before the error occured.
-#
-if [ -z "$(declare -f error_exit)" ]; then
-error_exit() {
-	local status=$1
-
-	case $status in
-	mcos_sys_mounted)
-		if [ "$enable_mcoverlay" == "yes" ]; then
-			umount /tmp/mcos/mcos0_sys
-		fi
-		;&
-	mcos_proc_mounted)
-		if [ "$enable_mcoverlay" == "yes" ]; then
-			umount /tmp/mcos/mcos0_proc
-		fi
-		;&
-	mcoverlayfs_loaded)
-		if [ "$enable_mcoverlay" == "yes" ]; then
-			rmmod mcoverlay 2>/dev/null
-		fi
-		;&
-	linux_proc_bind_mounted)
-		if [ "$enable_mcoverlay" == "yes" ]; then
-			umount /tmp/mcos/linux_proc
-		fi
-		;&
-	tmp_mcos_mounted)
-		if [ "$enable_mcoverlay" == "yes" ]; then
-			umount /tmp/mcos
-		fi
-		;&
-	tmp_mcos_created)
-		if [ "$enable_mcoverlay" == "yes" ]; then
-			rm -rf /tmp/mcos
-		fi
-		;&
-	initial)
-		# Nothing more to revert
-		;;
-	esac
-
-	# Retun -EINVAL
-	exit -22
-}
-fi
-
-if [ ! -e /tmp/mcos ]; then 
-    mkdir -p /tmp/mcos; 
-fi
-if ! mount -t tmpfs tmpfs /tmp/mcos; then
-    echo "error: mount /tmp/mcos" >&2
-    error_exit "tmp_mcos_created"
-fi
-if [ ! -e /tmp/mcos/linux_proc ]; then 
-    mkdir -p /tmp/mcos/linux_proc; 
-fi
-if ! mount --bind /proc /tmp/mcos/linux_proc; then
-    echo "error: mount /tmp/mcos/linux_proc" >&2
-    error_exit "tmp_mcos_mounted"
-fi
-if ! taskset -c 0 insmod @KMODDIR@/mcoverlay.ko 2>/dev/null; then
-    echo "error: inserting mcoverlay.ko" >&2
-    error_exit "linux_proc_bind_mounted"
-fi
-while [ ! -e /proc/mcos0 ]
-do
-    sleep 0.1
-done
-if [ ! -e /tmp/mcos/mcos0_proc ]; then 
-    mkdir -p /tmp/mcos/mcos0_proc; 
-fi
-if [ ! -e /tmp/mcos/mcos0_proc_upper ]; then 
-    mkdir -p /tmp/mcos/mcos0_proc_upper; 
-fi
-if [ ! -e /tmp/mcos/mcos0_proc_work ]; then 
-    mkdir -p /tmp/mcos/mcos0_proc_work; 
-fi
-if ! mount -t mcoverlay mcoverlay -o lowerdir=/proc/mcos0:/proc,upperdir=/tmp/mcos/mcos0_proc_upper,workdir=/tmp/mcos/mcos0_proc_work,nocopyupw,nofscheck /tmp/mcos/mcos0_proc; then
-    echo "error: mounting /tmp/mcos/mcos0_proc" >&2
-    error_exit "mcoverlayfs_loaded"
-fi
-# TODO: How de we revert this in case of failure??
-mount --make-rprivate /proc
-
-while [ ! -e /sys/devices/virtual/mcos/mcos0/sys/setup_complete ]
-do
-    sleep 0.1
-done
-if [ ! -e /tmp/mcos/mcos0_sys ]; then 
-    mkdir -p /tmp/mcos/mcos0_sys; 
-fi
-if [ ! -e /tmp/mcos/mcos0_sys_upper ]; then 
-    mkdir -p /tmp/mcos/mcos0_sys_upper; 
-fi
-if [ ! -e /tmp/mcos/mcos0_sys_work ]; then 
-    mkdir -p /tmp/mcos/mcos0_sys_work; 
-fi
-if ! mount -t mcoverlay mcoverlay -o lowerdir=/sys/devices/virtual/mcos/mcos0/sys:/sys,upperdir=/tmp/mcos/mcos0_sys_upper,workdir=/tmp/mcos/mcos0_sys_work,nocopyupw,nofscheck /tmp/mcos/mcos0_sys; then
-    echo "error: mount /tmp/mcos/mcos0_sys" >&2
-    error_exit "mcos_proc_mounted"
-fi
-# TODO: How de we revert this in case of failure??
-mount --make-rprivate /sys
-
-touch /tmp/mcos/mcos0_proc/mckernel
-
-rm -rf /tmp/mcos/mcos0_sys/setup_complete
-
-# Hide NUMA related files which are outside the LWK partition
-for cpuid in `find /sys/devices/system/cpu/* -maxdepth 0 -name "cpu[0123456789]*" -printf "%f "`; do
-    if [ ! -e "/sys/devices/virtual/mcos/mcos0/sys/devices/system/cpu/$cpuid" ]; then
-	rm -rf /tmp/mcos/mcos0_sys/devices/system/cpu/$cpuid
-	rm -rf /tmp/mcos/mcos0_sys/bus/cpu/devices/$cpuid
-	rm -rf /tmp/mcos/mcos0_sys/bus/cpu/drivers/processor/$cpuid
-    else
-	for nodeid in `find /sys/devices/system/cpu/$cpuid/* -maxdepth 0 -name "node[0123456789]*" -printf "%f "`; do
-	    if [ ! -e "/sys/devices/virtual/mcos/mcos0/sys/devices/system/cpu/$cpuid/$nodeid" ]; then
-		rm -f /tmp/mcos/mcos0_sys/devices/system/cpu/$cpuid/$nodeid
-	    fi
-	done
-    fi
-done
-for nodeid in `find /sys/devices/system/node/* -maxdepth 0 -name "node[0123456789]*" -printf "%f "`; do
-    if [ ! -e "/sys/devices/virtual/mcos/mcos0/sys/devices/system/node/$nodeid" ]; then
-	rm -rf /tmp/mcos/mcos0_sys/devices/system/node/$nodeid/*
-	rm -rf /tmp/mcos/mcos0_sys/bus/node/devices/$nodeid
-    else
-	# Delete non-existent symlinks
-	for cpuid in `find /sys/devices/system/node/$nodeid/* -maxdepth 0 -name "cpu[0123456789]*" -printf "%f "`; do
-	    if [ ! -e "/sys/devices/virtual/mcos/mcos0/sys/devices/system/node/$nodeid/$cpuid" ]; then
-		rm -f /tmp/mcos/mcos0_sys/devices/system/node/$nodeid/$cpuid
-	    fi
-	done
-
-	rm -f /tmp/mcos/mcos0_sys/devices/system/node/$nodeid/memory*
-    fi
-done
-rm -f /tmp/mcos/mcos0_sys/devices/system/node/has_*
-for cpuid in `find /sys/bus/cpu/devices/* -maxdepth 0 -name "cpu[0123456789]*" -printf "%f "`; do
-    if [ ! -e "/sys/devices/virtual/mcos/mcos0/sys/bus/cpu/devices/$cpuid" ]; then
-	rm -rf /tmp/mcos/mcos0_sys/bus/cpu/devices/$cpuid
-    fi
-done
-
-exit 0

arch/x86_64/tools/mcoverlay-destroy-smp-x86.sh.in

@@ -1,16 +0,0 @@
-# Remove mcoverlay if loaded
-
-if grep mcoverlay /proc/modules &>/dev/null; then
-    if [ "`cat /proc/mounts | grep /tmp/mcos/mcos0_sys`" != "" ]; then umount -l /tmp/mcos/mcos0_sys; fi
-    if [ "`cat /proc/mounts | grep /tmp/mcos/mcos0_proc`" != "" ]; then umount -l /tmp/mcos/mcos0_proc; fi
-    if [ "`cat /proc/mounts | grep /tmp/mcos/linux_proc`" != "" ]; then umount -l /tmp/mcos/linux_proc; fi
-    if [ "`cat /proc/mounts | grep /tmp/mcos`" != "" ]; then umount -l /tmp/mcos; fi
-    if [ -e /tmp/mcos ]; then rm -rf /tmp/mcos; fi
-    if ! rmmod mcoverlay 2>/dev/null; then
-	echo "error: removing mcoverlay" >&2
-	# Return -EINVAL
-	exit -22
-    fi
-fi
-
-exit 0

cmake/modules/AutoconfHelper.cmake

@@ -0,0 +1,383 @@
+# Helper functions for translating autoconf projects. Several functions
+# are lifted from the Mono sources
+
+include (CheckCSourceCompiles)
+include (CheckIncludeFile)
+include (TestBigEndian)
+include (CheckFunctionExists)
+include (CheckTypeSize)
+include (CheckCSourceRuns)
+
+
+# Function to get the version information from the configure.ac file in the
+# current directory. Its argument is the name of the library as passed to
+# AC_INIT. It will set the variables ${LIBNAME}_VERSION and ${LIBNAME}_SOVERSION
+function (ac_get_version libname)
+  string(TOUPPER "${libname}" libname_upper)
+
+  # Read the relevant content from configure.ac
+  file (STRINGS configure.ac tmp_configure_ac
+    REGEX "${libname_upper}_[_A-Z]+=[ \\t]*[0-9]+")
+
+  # Product version
+  string (REGEX REPLACE ".+MAJOR[_A-Z]+=([0-9]+).+MINOR[_A-Z]+=([0-9]+).+MICRO[_A-Z]+=([0-9]+).*"
+    "\\1.\\2.\\3" ${libname_upper}_VERSION "${tmp_configure_ac}")
+
+  # Library version for libtool
+  string (REGEX REPLACE ".+CURRENT=([0-9]+).+REVISION=([0-9]+).+AGE=([0-9]+).*"
+    "\\1.\\2.\\3" ${libname_upper}_SOVERSION "${tmp_configure_ac}")
+
+  # Checks if the string needs to be displayed
+  set (${libname_upper}_DISPLAYSTR_AUX
+    "Found ${libname} version ${${libname_upper}_VERSION}, soversion ${${libname_upper}_SOVERSION} from configure.ac"
+  )
+  if ((NOT ${libname_upper}_DISPLAYSTR) OR (NOT ${libname_upper}_DISPLAYSTR STREQUAL ${libname_upper}_DISPLAYSTR_AUX))
+    set (${libname_upper}_DISPLAYSTR ${${libname_upper}_DISPLAYSTR_AUX}
+      CACHE INTERNAL "Version string from ${libname}" FORCE)
+    message (STATUS ${${libname_upper}_DISPLAYSTR})
+  endif ()
+
+  # Export the result to the caller
+  set(${libname_upper}_VERSION "${${libname_upper}_VERSION}" PARENT_SCOPE)
+  set(${libname_upper}_SOVERSION "${${libname_upper}_SOVERSION}" PARENT_SCOPE)
+endfunction()
+
+
+# Also from mono's source code
+# Implementation of AC_CHECK_HEADERS
+# In addition, it also records the list of variables in the variable
+# 'autoheader_vars', and for each variable, a documentation string in the
+# variable ${var}_doc
+function(ac_check_headers)
+  foreach (header ${ARGV})
+	string(TOUPPER ${header} header_var)
+	string(REPLACE "." "_" header_var ${header_var})
+	string(REPLACE "/" "_" header_var ${header_var})
+	set(header_var "HAVE_${header_var}")
+	check_include_file (${header} ${header_var})
+	set("${header_var}_doc" "Define to 1 if you have the <${header}> header file." PARENT_SCOPE)
+	if (${header_var})
+	  set("${header_var}_defined" "1" PARENT_SCOPE)
+	endif()
+	set("${header_var}_val" "1" PARENT_SCOPE)
+	set (autoheader_vars ${autoheader_vars} ${header_var})
+  endforeach()
+  set (autoheader_vars ${autoheader_vars} PARENT_SCOPE)
+endfunction()
+
+# Function taken from mono's source code
+function (ac_check_funcs)
+  foreach (func ${ARGV})
+    string(TOUPPER ${func} var)
+    set(var "HAVE_${var}")
+    set(${var})
+    check_function_exists (${func} ${var})
+    set("${var}_doc" "Define to 1 if you have the '${func}' function." PARENT_SCOPE)
+    if (${var})
+      set("${var}_defined" "1" PARENT_SCOPE)
+      set(${var} yes PARENT_SCOPE)
+    endif()
+    set("${var}_val" "1" PARENT_SCOPE)
+    set (autoheader_vars ${autoheader_vars} ${var})
+  endforeach()
+  set (autoheader_vars ${autoheader_vars} PARENT_SCOPE)
+endfunction()
+
+
+# Specifically, this macro checks for stdlib.h', stdarg.h',
+# string.h', and float.h'; if the system has those, it probably
+# has the rest of the ANSI C header files.  This macro also checks
+# whether string.h' declares memchr' (and thus presumably the
+# other mem' functions), whether stdlib.h' declare free' (and
+# thus presumably malloc' and other related functions), and whether
+# the ctype.h' macros work on characters with the high bit set, as
+# ANSI C requires.
+function (ac_header_stdc)
+  if (STDC_HEADERS)
+    return()
+  endif()
+  message(STATUS "Looking for ANSI-C headers")
+  set(code "
+#include <stdlib.h>
+#include <stdarg.h>
+#include <string.h>
+#include <float.h>
+
+int main(int argc, char **argv)
+{
+  void *ptr;
+  free((void*)1);
+  ptr = memchr((void*)1, 0, 0);
+
+  return (int)ptr;
+}
+")
+  # FIXME Check the ctype.h high bit
+  CHECK_C_SOURCE_COMPILES("${code}" STDC_HEADERS)
+  if (STDC_HEADERS)
+    set(STDC_HEADERS 1 PARENT_SCOPE)
+	message(STATUS "Looking for ANSI-C headers - found")
+  else()
+    message(STATUS "Looking for ANSI-C headers - not found")
+  endif()
+endfunction()
+
+
+# Also from the mono sources, kind of implements AC_SYS_LARGEFILE
+function (ac_sys_largefile)
+  CHECK_C_SOURCE_RUNS("
+#include <sys/types.h>
+#define BIG_OFF_T (((off_t)1<<62)-1+((off_t)1<<62))
+int main (int argc, char **argv) {
+    int big_off_t=((BIG_OFF_T%2147483629==721) &&
+                   (BIG_OFF_T%2147483647==1));
+    return big_off ? 0 : 1;
+}
+" HAVE_LARGE_FILE_SUPPORT)
+
+# Check if it makes sense to define _LARGE_FILES or _FILE_OFFSET_BITS
+  if (HAVE_LARGE_FILE_SUPPORT)
+    return()
+  endif()
+
+  set (_LARGE_FILE_EXTRA_SRC "
+#include <sys/types.h>
+int main (int argc, char **argv) {
+  return sizeof(off_t) == 8 ? 0 : 1;
+}
+")
+  CHECK_C_SOURCE_RUNS ("#define _LARGE_FILES\n${_LARGE_FILE_EXTRA_SRC}"
+    HAVE_USEFUL_D_LARGE_FILES)
+  if (NOT HAVE_USEFUL_D_LARGE_FILES)
+    if (NOT DEFINED HAVE_USEFUL_D_FILE_OFFSET_BITS)
+      set (SHOW_LARGE_FILE_WARNING TRUE)
+    endif ()
+    CHECK_C_SOURCE_RUNS ("#define _FILE_OFFSET_BITS 64\n${_LARGE_FILE_EXTRA_SRC}"
+      HAVE_USEFUL_D_FILE_OFFSET_BITS)
+	  if (HAVE_USEFUL_D_FILE_OFFSET_BITS)
+      set (_FILE_OFFSET_BITS 64 PARENT_SCOPE)
+	  elseif (SHOW_LARGE_FILE_WARNING)
+	    message (WARNING "No 64 bit file support through off_t available.")
+	  endif ()
+  else ()
+    set (_LARGE_FILES 1 PARENT_SCOPE)
+  endif ()
+endfunction ()
+
+
+# Quick way to set some basic variables
+# FIXME add support for variable number of arguments: only package and version are mandatory
+# arguments are package version bug_report tarname url
+function (ac_init)
+    set(package ${ARGV0})
+    set(version ${ARGV1})
+    set(bug_report ${ARGV2})
+    set(tarname ${ARGV3})
+    set(url ${ARGV4})
+    set(PACKAGE_NAME "\"${package}\"" PARENT_SCOPE)
+    set(PACKAGE_VERSION "\"${version}\"" PARENT_SCOPE)
+    set(VERSION "\"${version}\"" PARENT_SCOPE)
+    if(version)
+        set(PACKAGE_STRING "\"${package} ${version}\"" PARENT_SCOPE)
+    else()
+        set(PACKAGE_STRING "\"${package}\"" PARENT_SCOPE)
+    endif()
+
+    set(PACKAGE_BUGREPORT "\"${bug_report}\"" PARENT_SCOPE)
+
+    if(NOT tarname)
+        string(REGEX REPLACE "[^a-zA-Z0-9_]" "-" tarname "${package}")
+    endif()
+    set(PACKAGE_TARNAME "\"${tarname}\"" PARENT_SCOPE)
+
+    set(PACKAGE_URL "\"${url}\"" PARENT_SCOPE)
+endfunction()
+
+
+# Checks for the const keyword, defining "HAS_CONST_SUPPORT"
+# If it does not have support, defines "const" to 0 in the parent scope
+function (ac_c_const)
+  CHECK_C_SOURCE_COMPILES(
+    "int main(int argc, char **argv){const int r = 0;return r;}"
+    HAS_CONST_SUPPORT)
+  if (NOT HAS_CONST_SUPPORT)
+    set(const 0 PARENT_SCOPE)
+  endif()
+endfunction()
+
+
+# Inline keyword support. Defines "inline" in the parent scope to the
+# compiler internal keyword for inline in C
+# TODO write a better test!
+function (ac_c_inline)
+  if (MSVC)
+    set (inline __inline)
+  elseif(CMAKE_COMPILER_IS_GNUC)
+    set (inline __inline__)
+  endif()
+  set(inline "${inline}" PARENT_SCOPE)
+endfunction()
+
+
+# Test if you can safely include both <sys/time.h> and <time.h>
+function (ac_header_time)
+  CHECK_C_SOURCE_COMPILES(
+    "#include <sys/time.h>\n#include <time.h>\nint main(int argc, char **argv) { return 0; }"
+    TIME_WITH_SYS_TIME)
+  set(TIME_WITH_SYS_TIME ${TIME_WITH_SYS_TIME} PARENT_SCOPE)
+endfunction()
+
+
+# Native cpu byte order: 1 if big-endian (Motorola) or 0 if little-endian
+# (Intel), setting "WORDS_BIGENDIAN" to 1 if big endian
+function (ac_c_bigendian)
+  TEST_BIG_ENDIAN(HOST_BIGENDIAN)
+  if (HOST_BIGENDIAN)
+    set(WORDS_BIGENDIAN 1 PARENT_SCOPE)
+  endif()
+endfunction()
+
+
+# Check for off_t, setting "off_t" in the parent scope
+function(ac_type_off_t)
+  CHECK_TYPE_SIZE("off_t" SIZEOF_OFF_T)
+  if (NOT SIZEOF_OFF_T)
+    set(off_t "long int")
+  endif()
+  set(off_t ${off_t} PARENT_SCOPE)
+endfunction()
+
+
+# Check for size_t, setting "size_t" in the parent scope
+function(ac_type_size_t)
+  CHECK_TYPE_SIZE("size_t" SIZEOF_SIZE_T)
+  if (NOT SIZEOF_SIZE_T)
+    set(size_t "unsigned int")
+  endif()
+  set(size_t ${size_t} PARENT_SCOPE)
+endfunction()
+
+
+# Define "TM_IN_SYS_TIME" to 1 if <sys/time.h> declares "struct tm"
+function(ac_struct_tm)
+  CHECK_C_SOURCE_COMPILES(
+    "#include <sys/time.h>\nint main(int argc, char **argv) { struct tm x; return 0; }"
+    TM_IN_SYS_TIME
+  )
+  if (TM_IN_SYS_TIME)
+    set (TM_IN_SYS_TIME 1 PARENT_SCOPE)
+  endif()
+endfunction()
+
+
+# Obtain size of an 'type' and define as SIZEOF_TYPE
+function (ac_check_sizeof typename)
+  string(TOUPPER "SIZEOF_${typename}" varname)
+	string(REPLACE " " "_" varname "${varname}")
+  string(REPLACE "*" "p" varname "${varname}")
+  CHECK_TYPE_SIZE("${typename}" ${varname} BUILTIN_TYPES_ONLY)
+  if(NOT ${varname})
+    set(${varname} 0 PARENT_SCOPE)
+  endif()
+endfunction()
+
+
+# Check if the type exists, defines HAVE_<type>
+function (ac_check_type typename)
+  string(TOUPPER "${typename}" varname)
+	string(REPLACE " " "_" varname "${varname}")
+  string(REPLACE "*" "p" varname "${varname}")
+  CHECK_TYPE_SIZE("${typename}" ${varname})
+  if (NOT "${varname}" STREQUAL "")
+    set("HAVE_${varname}" 1 PARENT_SCOPE)
+    set("${varname}" "${typename}" PARENT_SCOPE)
+  else()
+    set("${varname}" "unknown" PARENT_SCOPE)
+  endif()
+endfunction()
+
+
+# Verifies if each type on the list exists, using the given prelude
+function (ac_check_types type_list prelude)
+  foreach(typename ${type_list})
+    string(TOUPPER "HAVE_${typename}" varname)
+    string(REPLACE " " "_" varname "${varname}")
+    string(REPLACE "*" "p" varname "${varname}")
+    CHECK_C_SOURCE_COMPILES("${prelude}\n ${typename} foo;" ${varname})
+  endforeach()
+endfunction()
+
+function(ac_path_prog variable prog_to_check_for value_if_not_found env_var)
+    find_program(${variable} NAMES ${prog_to_check_for} PATHS ENV ${env_var} NO_DEFAULT_PATH)
+    if(NOT ${variable})
+        message(STATUS "Looking for ${prog_to_check_for} - not found")
+        set(${variable} ${value_if_not_fount} PARENT_SCOPE)
+    else()
+        message(STATUS "Looking for ${prog_to_check_for} - ${variable}")
+        set(${variable} ${${variable}} PARENT_SCOPE)
+    endif()
+endfunction()
+
+# check if function func exists in library lib
+function(ac_check_lib lib func)
+    string(TOUPPER "HAVE_${func}" varname)
+    set(CMAKE_REQUIRED_LIBRARIES ${lib})
+    check_function_exists(${func} ${varname})
+    set(CMAKE_REQUIRED_LIBRARIES)
+endfunction()
+
+# check if source compiles without linking
+function(ac_try_compile SOURCE VAR)
+    set(CMAKE_TMP_DIR ${CMAKE_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/CMakeTmp)
+    if(NOT DEFINED "${VAR}")
+        file(WRITE
+            "${CMAKE_TMP_DIR}/src.c"
+            "${SOURCE}\n"
+        )
+
+        if(NOT CMAKE_REQUIRED_QUIET)
+            message(STATUS "Performing Test ${VAR}")
+        endif()
+        # Set up CMakeLists.txt for static library:
+        file(WRITE
+            ${CMAKE_TMP_DIR}/CMakeLists.txt
+            "add_library(compile STATIC src.c)"
+        )
+
+        # Configure:
+        execute_process(
+            COMMAND ${CMAKE_COMMAND} -G "${CMAKE_GENERATOR}" .
+            WORKING_DIRECTORY ${CMAKE_TMP_DIR}
+        )
+
+        # Build:
+        execute_process(
+            COMMAND ${CMAKE_COMMAND} --build ${CMAKE_TMP_DIR}
+            RESULT_VARIABLE RESVAR
+            OUTPUT_VARIABLE OUTPUT
+            ERROR_VARIABLE OUTPUT
+        )
+
+        # Set up result:
+        if(RESVAR EQUAL 0)
+            set(${VAR} 1 CACHE INTERNAL "Test ${VAR}")
+            if(NOT CMAKE_REQUIRED_QUIET)
+                message(STATUS "Performing Test ${VAR} - Success")
+            endif()
+
+            file(APPEND ${CMAKE_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/CMakeOutput.log
+                "Performing C SOURCE FILE Test ${VAR} succeded with the following output:\n"
+                "${OUTPUT}\n"
+                "Source file was:\n${SOURCE}\n")
+        else()
+            if(NOT CMAKE_REQUIRED_QUIET)
+                message(STATUS "Performing Test ${VAR} - Failed")
+            endif()
+            set(${VAR} "" CACHE INTERNAL "Test ${VAR}")
+            file(APPEND ${CMAKE_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/CMakeError.log
+            "Performing C SOURCE FILE Test ${VAR} failed with the following output:\n"
+            "${OUTPUT}\n"
+            "Source file was:\n${SOURCE}\n")
+        endif()
+    endif()
+endfunction()

cmake/modules/FindLibElf.cmake

@@ -0,0 +1,64 @@
+# - Try to find libelf
+# Once done this will define
+#
+#  LIBELF_FOUND - system has libelf
+#  LIBELF_INCLUDE_DIRS - the libelf include directory
+#  LIBELF_LIBRARIES - Link these to use libelf
+#  LIBELF_DEFINITIONS - Compiler switches required for using libelf
+#
+# This module reads hints about search locations from variables:
+#
+#  LIBELF_ROOT - Preferred installation prefix
+#
+#  Copyright (c) 2008 Bernhard Walle <bernhard.walle@gmx.de>
+#
+#  Redistribution and use is allowed according to the terms of the New
+#  BSD license.
+#  For details see the accompanying COPYING-CMAKE-SCRIPTS file.
+#
+
+
+if (LIBELF_LIBRARIES AND LIBELF_INCLUDE_DIRS)
+  set (LibElf_FIND_QUIETLY TRUE)
+endif (LIBELF_LIBRARIES AND LIBELF_INCLUDE_DIRS)
+
+find_path (LIBELF_INCLUDE_DIRS
+  NAMES
+    libelf/libelf.h libelf.h
+  HINTS
+    ${LIBELF_ROOT}
+  PATH_SUFFIXES
+    include
+	libelf/include
+)
+
+find_library (LIBELF_LIBRARIES
+  NAMES
+    elf libelf
+  HINTS
+    ${LIBELF_ROOT}
+  PATH_SUFFIXES
+    lib
+	libelf/lib
+)
+
+include (FindPackageHandleStandardArgs)
+
+
+# handle the QUIETLY and REQUIRED arguments and set LIBELF_FOUND to TRUE if all listed variables are TRUE
+FIND_PACKAGE_HANDLE_STANDARD_ARGS(LibElf DEFAULT_MSG
+  LIBELF_LIBRARIES
+  LIBELF_INCLUDE_DIRS)
+
+set(CMAKE_REQUIRED_LIBRARIES elf)
+include(CheckCXXSourceCompiles)
+check_cxx_source_compiles("#include <libelf.h>
+int main() {
+  Elf *e = (Elf*)0;
+  size_t sz;
+  elf_getshdrstrndx(e, &sz);
+  return 0;
+}" ELF_GETSHDRSTRNDX)
+unset(CMAKE_REQUIRED_LIBRARIES)
+
+mark_as_advanced(LIBELF_INCLUDE_DIRS LIBELF_LIBRARIES ELF_GETSHDRSTRNDX)

cmake/modules/Kbuild.cmake

@@ -14,6 +14,28 @@ mark_as_advanced(
 	KBUILD_MAKE_FLAGS
 )
 
+if (${CMAKE_GENERATOR} STREQUAL Ninja)
+	set(MAKE "make")
+	list(APPEND KBUILD_MAKE_FLAGS "-j")
+else ()
+	set(MAKE "$(MAKE)")
+endif ()
+
+# Convert McKernel "arm64" into Linux "aarch64"
+if ("${ARCH}" STREQUAL "arm64")
+	set(LINUX_ARCH "aarch64")
+else ()
+	set(LINUX_ARCH "${ARCH}")
+endif ()
+
+if (NOT "${LINUX_ARCH}" STREQUAL "${CMAKE_HOST_SYSTEM_PROCESSOR}")
+	string(REGEX REPLACE "ld$" "" CROSS_COMPILE "${CMAKE_LINKER}")
+	list(APPEND KBUILD_MAKE_FLAGS "ARCH=${ARCH}")
+	list(APPEND KBUILD_MAKE_FLAGS "CROSS_COMPILE=${CROSS_COMPILE}")
+endif()
+
+string(REPLACE ";" " " KBUILD_MAKE_FLAGS_STR "${KBUILD_MAKE_FLAGS}")
+
 function(kmod MODULE_NAME)
 	cmake_parse_arguments(KMOD "" "INSTALL_DEST" "C_FLAGS;SOURCES;EXTRA_SYMBOLS;DEPENDS" ${ARGN})
 
@@ -33,17 +55,6 @@ endif(ENABLE_WERROR)
 	configure_file(${CMAKE_SOURCE_DIR}/cmake/modules/Kbuild.in
 		${CMAKE_CURRENT_BINARY_DIR}/Kbuild)
 
-	if (${CMAKE_GENERATOR} STREQUAL Ninja)
-		set(MAKE "make")
-		list(APPEND KBUILD_MAKE_FLAGS "-j")
-	else ()
-		set(MAKE "$(MAKE)")
-	endif ()
-	if (NOT "${ARCH}" STREQUAL "${CMAKE_HOST_SYSTEM_PROCESSOR}")
-		string(REGEX REPLACE "ld$" "" CROSS_COMPILE "${CMAKE_LINKER}")
-		list(APPEND KBUILD_MAKE_FLAGS "ARCH=${ARCH};CROSS_COMPILE=${CROSS_COMPILE}")
-	endif()
-
 	string(REGEX REPLACE "\\.c(;|$)" ".o.cmd\\1" KMOD_O_CMD "${KMOD_SOURCES}")
 	string(REGEX REPLACE "[^/;]+(;|$)" ".\\0" KMOD_O_CMD "${KMOD_O_CMD}")
 
@@ -78,6 +89,10 @@ endif(ENABLE_WERROR)
 	# the native build system do these checks, if possible at all...
 	add_custom_command(OUTPUT kmod_always_rebuild COMMAND touch kmod_always_rebuild)
 
+	if (NOT EXISTS "${KERNEL_DIR}/Makefile")
+		message(FATAL_ERROR "${KERNEL_DIR} does not contain a Makefile and is probably missing. install kernel development package or set the KERNEL_DIR variable")
+	endif()
+
 	add_custom_command(
 		OUTPUT "${MODULE_NAME}.ko"
 			"Module.symvers"

config.h.in

@@ -6,8 +6,9 @@
 /* version number */
 #define MCKERNEL_VERSION "${MCKERNEL_VERSION}"
 
-/* whether mcoverlayfs is enabled */
-#cmakedefine ENABLE_MCOVERLAYFS 1
+/* enable the required code for mcexec to be able to use bind mount
+ * there is no config option as its use is discouraged */
+// #define MCEXEC_BIND_MOUNT 1
 
 /* whether memdump feature is enabled */
 #cmakedefine ENABLE_MEMDUMP 1
@@ -27,18 +28,12 @@
 /* whether undefined behaviour sanitizer is enabled */
 #cmakedefine ENABLE_UBSAN 1
 
+/* whether per-CPU allocator cache (ThunderX2 workaround) is enabled */
+#cmakedefine ENABLE_PER_CPU_ALLOC_CACHE 1
+
 /* Path of bind-mount source directory */
 #cmakedefine ROOTFSDIR "${ROOTFSDIR}"
 
-/* Path of install directory for libraries */
-#cmakedefine MCKERNEL_LIBDIR "${MCKERNEL_LIBDIR}"
-
-/* Path of install directory for binary */
-#cmakedefine BINDIR "${BINDIR}"
-
-/* Path of install directory for system binary */
-#cmakedefine SBINDIR "${SBINDIR}"
-
 /* for non-RHEL kernels */
 #ifndef RHEL_RELEASE_VERSION
 #define RHEL_RELEASE_VERSION(a,b) (((a) << 8) + (b))

executer/include/ihklib_rusage.h

@@ -1,27 +0,0 @@
-#ifndef IHKLIB_RUSAGE_H_INCLUDED
-#define IHKLIB_RUSAGE_H_INCLUDED
-
-#define IHK_MAX_NUM_PGSIZES 4
-#define IHK_MAX_NUM_NUMA_NODES 1024
-#define IHK_MAX_NUM_CPUS 1024
-
-#define IHK_OS_PGSIZE_4KB 0
-#define IHK_OS_PGSIZE_2MB 1
-#define IHK_OS_PGSIZE_1GB 2
-
-struct mckernel_rusage {
-	unsigned long memory_stat_rss[IHK_MAX_NUM_PGSIZES];
-	unsigned long memory_stat_mapped_file[IHK_MAX_NUM_PGSIZES];
-	unsigned long memory_max_usage;
-	unsigned long memory_kmem_usage;
-	unsigned long memory_kmem_max_usage;
-	unsigned long memory_numa_stat[IHK_MAX_NUM_NUMA_NODES];
-	unsigned long cpuacct_stat_system;
-	unsigned long cpuacct_stat_user;
-	unsigned long cpuacct_usage;
-	unsigned long cpuacct_usage_percpu[IHK_MAX_NUM_CPUS];
-	int num_threads;
-	int max_num_threads;
-};
-
-#endif /* !defined(IHKLIB_RUSAGE_H_INCLUDED) */

executer/include/uprotocol.h

@@ -55,7 +55,7 @@
 #define MCEXEC_UP_SYS_UNSHARE    0x30a02916
 
 #define MCEXEC_UP_UTI_GET_CTX    0x30a02920
-#define MCEXEC_UP_UTI_SAVE_FS    0x30a02921
+#define MCEXEC_UP_UTI_SWITCH_CTX 0x30a02921
 #define MCEXEC_UP_SIG_THREAD     0x30a02922
 #define MCEXEC_UP_SYSCALL_THREAD 0x30a02924
 #define MCEXEC_UP_TERMINATE_THREAD 0x30a02925
@@ -91,7 +91,10 @@ struct program_image_section {
 
 struct get_cpu_set_arg {
 	int nr_processes;
+	char *req_cpu_list;   // Requested by user-space
+	int req_cpu_list_len; // Lenght of request string
 	int *process_rank;
+	pid_t ppid;
 	void *cpu_set;
 	size_t cpu_set_size;	// Size in bytes
 	int *target_core;
@@ -193,7 +196,6 @@ struct syscall_response {
 	unsigned long req_thread_status;
 	long ret;
 	unsigned long fault_address;
-	unsigned long fault_reason;
 };
 
 struct syscall_ret_desc {
@@ -359,7 +361,7 @@ struct uti_get_ctx_desc {
 	unsigned long key; /* OUT: struct task_struct* of mcexec thread, used to search struct host_thread */
 };
 
-struct uti_save_fs_desc {
+struct uti_switch_ctx_desc {
 	void *rctx; /* Remote context */
 	void *lctx; /* Local context */
 };

executer/kernel/mcctrl/CMakeLists.txt

@@ -4,6 +4,8 @@ if(ARCH STREQUAL "x86_64")
 	set(ARCH_C_FLAGS "-mno-red-zone -mcmodel=kernel")
 endif()
 
+set(MCEXEC_PATH "${CMAKE_INSTALL_FULL_BINDIR}/mcexec" CACHE STRING "mcexec path for binfmt")
+
 kmod(mcctrl
 	C_FLAGS
 		-I${IHK_FULL_SOURCE_DIR}/linux/include
@@ -16,7 +18,7 @@ kmod(mcctrl
 		-I${CMAKE_CURRENT_SOURCE_DIR}/arch/${ARCH}/include
 		-I${PROJECT_BINARY_DIR}
 		-I${PROJECT_SOURCE_DIR}/kernel/include
-		-DMCEXEC_PATH=\\"${CMAKE_INSTALL_FULL_BINDIR}/mcexec\\"
+		-DMCEXEC_PATH=\\"${MCEXEC_PATH}\\"
 		${ARCH_C_FLAGS}
 	SOURCES
 		driver.c control.c ikc.c syscall.c procfs.c binfmt_mcexec.c

executer/kernel/mcctrl/arch/arm64/archdeps.c

@@ -1,7 +1,12 @@
-/* archdeps.c COPYRIGHT FUJITSU LIMITED 2016 */
+/* archdeps.c COPYRIGHT FUJITSU LIMITED 2016-2019 */
 #include <linux/version.h>
 #include <linux/mm_types.h>
 #include <linux/kallsyms.h>
+#if KERNEL_VERSION(4, 11, 0) <= LINUX_VERSION_CODE
+#include <linux/sched/task_stack.h>
+#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4, 11, 0) */
+#include <linux/ptrace.h>
+#include <linux/uaccess.h>
 #include <asm/vdso.h>
 #include "config.h"
 #include "../../mcctrl.h"
@@ -42,7 +47,6 @@ int arch_symbols_init(void)
 }
 
 
-#ifdef POSTK_DEBUG_ARCH_DEP_52
 #define VDSO_MAXPAGES 1
 struct vdso {
 	long busy;
@@ -53,7 +57,6 @@ struct vdso {
 	long lbase;
 	long offset_sigtramp;
 };
-#endif /*POSTK_DEBUG_ARCH_DEP_52*/
 
 unsigned long
 reserve_user_space_common(struct mcctrl_usrdata *usrdata, unsigned long start, unsigned long end);
@@ -95,6 +98,74 @@ reserve_user_space(struct mcctrl_usrdata *usrdata, unsigned long *startp, unsign
 	return 0;
 }
 
+#if KERNEL_VERSION(4, 0, 0) <= LINUX_VERSION_CODE
+static long elf_search_vdso_sigtramp(void)
+{
+	int i = 0;
+	long ans = -1;
+	char *shstr = NULL, *dynstr = NULL;
+	Elf64_Ehdr *eh = NULL;
+	Elf64_Shdr *tmp_sh = NULL, *sym_sh = NULL;
+	Elf64_Sym *sym = NULL;
+
+	/* ELF header */
+	eh = (Elf64_Ehdr *)vdso_start;
+	if (eh == NULL) {
+		D("vdso_start is NULL.\n");
+		goto out;
+	}
+
+	/* ELF magic check */
+	if (eh->e_ident[EI_MAG0] != ELFMAG0 &&
+	    eh->e_ident[EI_MAG1] != ELFMAG1 &&
+	    eh->e_ident[EI_MAG2] != ELFMAG2 &&
+	    eh->e_ident[EI_MAG3] != ELFMAG3) {
+		D("vdso_start ELF MAGIC Mismatch.\n"
+		  "e_ident[EI_MAG0 - EI_MAG3]: %02x %02x %02x %02x\n",
+		  eh->e_ident[EI_MAG0], eh->e_ident[EI_MAG1],
+		  eh->e_ident[EI_MAG2], eh->e_ident[EI_MAG3]);
+		goto out;
+	}
+
+	/* Search dynsym-table and dynstr-table offset
+	 * from section header table
+	 */
+	tmp_sh = (Elf64_Shdr *)(vdso_start + eh->e_shoff);
+	shstr = vdso_start + (tmp_sh + eh->e_shstrndx)->sh_offset;
+	for (i = 0; i < eh->e_shnum; i++, tmp_sh++) {
+		if (tmp_sh->sh_type == SHT_DYNSYM) {
+			sym_sh = tmp_sh;
+		}
+
+		if (tmp_sh->sh_type == SHT_STRTAB &&
+		    !strcmp(&shstr[tmp_sh->sh_name], ".dynstr")) {
+			dynstr = vdso_start + tmp_sh->sh_offset;
+		}
+	}
+
+	if (sym_sh == NULL) {
+		D("dynsym-table not found.\n");
+		goto out;
+	}
+
+	if (dynstr == 0) {
+		D("dynstr-table not found.\n");
+		goto out;
+	}
+
+	/* Search __kernel_rt_sigreturn offset from dynsym-table */
+	sym = (Elf64_Sym *)(vdso_start + sym_sh->sh_offset);
+	for (i = 0; (i * sym_sh->sh_entsize) < sym_sh->sh_size; i++, sym++) {
+		if (!strcmp(dynstr + sym->st_name, "__kernel_rt_sigreturn")) {
+			ans = sym->st_value;
+		}
+	}
+
+out:
+	return ans;
+}
+#endif /*LINUX_VERSION_CODE >= KERNEL_VERSION(4,0,0)*/
+
 void get_vdso_info(ihk_os_t os, long vdso_rpa)
 {
 	ihk_device_t dev = ihk_os_to_dev(os);
@@ -128,7 +199,12 @@ void get_vdso_info(ihk_os_t os, long vdso_rpa)
 
 	/* offsets */
 	vdso->lbase = VDSO_LBASE;
-	vdso->offset_sigtramp = vdso_offset_sigtramp;
+	vdso->offset_sigtramp = elf_search_vdso_sigtramp();
+
+	if (unlikely(vdso->offset_sigtramp == -1)) {
+		D("Use vdso_offset_sigtramp in header-file.\n");
+		vdso->offset_sigtramp = vdso_offset_sigtramp;
+	}
 #endif /*LINUX_VERSION_CODE >= KERNEL_VERSION(4,0,0)*/
 out:
 	wmb();
@@ -142,59 +218,61 @@ out:
 void *
 get_user_sp(void)
 {
-	/* TODO; skeleton for UTI */
-	return NULL;
+	return (void *)current_pt_regs()->sp;
 }
 
 void
 set_user_sp(void *usp)
 {
-	/* TODO; skeleton for UTI */
+	current_pt_regs()->sp = (unsigned long)usp;
 }
 
-/* TODO; skeleton for UTI */
 struct trans_uctx {
 	volatile int cond;
 	int fregsize;
-
-	unsigned long rax;
-	unsigned long rbx;
-	unsigned long rcx;
-	unsigned long rdx;
-	unsigned long rsi;
-	unsigned long rdi;
-	unsigned long rbp;
-	unsigned long r8;
-	unsigned long r9;
-	unsigned long r10;
-	unsigned long r11;
-	unsigned long r12;
-	unsigned long r13;
-	unsigned long r14;
-	unsigned long r15;
-	unsigned long rflags;
-	unsigned long rip;
-	unsigned long rsp;
-	unsigned long fs;
+	struct user_pt_regs regs;
+	unsigned long tls_baseaddr;
 };
 
 void
-restore_fs(unsigned long fs)
+restore_tls(unsigned long addr)
 {
-	/* TODO; skeleton for UTI */
+	const unsigned long tpidrro = 0;
+
+	asm volatile(
+	"	msr	tpidr_el0, %0\n"
+	"	msr	tpidrro_el0, %1"
+	: : "r" (addr), "r" (tpidrro));
 }
 
 void
-save_fs_ctx(void *ctx)
+save_tls_ctx(void __user *ctx)
 {
-	/* TODO; skeleton for UTI */
+	struct trans_uctx __user *tctx = ctx;
+	unsigned long baseaddr;
+
+	asm volatile(
+	"	mrs	%0, tpidr_el0"
+	: "=r" (baseaddr));
+
+	if (copy_to_user(&tctx->tls_baseaddr, &baseaddr,
+			 sizeof(tctx->tls_baseaddr))) {
+		pr_err("%s: copy_to_user failed.\n", __func__);
+		return;
+	}
 }
 
 unsigned long
-get_fs_ctx(void *ctx)
+get_tls_ctx(void __user *ctx)
 {
-	/* TODO; skeleton for UTI */
-	return 0;
+	struct trans_uctx __user *tctx = ctx;
+	struct trans_uctx kctx;
+
+	if (copy_from_user(&kctx, tctx, sizeof(struct trans_uctx))) {
+		pr_err("%s: copy_from_user failed.\n", __func__);
+		return 0;
+	}
+	return kctx.tls_baseaddr;
 }
 
 #if LINUX_VERSION_CODE < KERNEL_VERSION(4,2,0)
@@ -304,3 +382,38 @@ out:
 			error, rva, rpa, pgsize);
 	return error;
 }
+
+/*
+ * Assembler switch_ctx executes only ioctl.
+ * Context register save/load is done on Linux (get from current_pt_regs).
+ * Do TLS save/load and register host_thread with ioctl.
+ */
+long arch_switch_ctx(struct uti_switch_ctx_desc *desc)
+{
+	int rc = 0;
+	struct trans_uctx *__user rctx = NULL;
+	struct trans_uctx *__user lctx = NULL;
+	struct trans_uctx klctx = {
+		.regs = current_pt_regs()->user_regs,
+	};
+
+	rctx = desc->rctx;
+	lctx = desc->lctx;
+
+	if (copy_to_user(lctx, &klctx, sizeof(klctx))) {
+		pr_err("%s: Error: copy_to_user failed\n", __func__);
+		rc = -EFAULT;
+		goto out;
+	}
+
+	if (copy_from_user(&current_pt_regs()->user_regs,
+			   &rctx->regs, sizeof(rctx->regs))) {
+		pr_err("%s: Error: copy_from_user failed\n", __func__);
+		rc = -EFAULT;
+		goto out;
+	}
+	restore_tls(get_tls_ctx(rctx));
+
+out:
+	return rc;
+}

executer/kernel/mcctrl/arch/arm64/include/archdeps.h

@@ -9,8 +9,6 @@
 extern int translate_rva_to_rpa(ihk_os_t os, unsigned long rpt, unsigned long rva,
 				unsigned long *rpap, unsigned long *pgsizep);
 
-#ifdef POSTK_DEBUG_ARCH_DEP_12
-
 #define PFN_WRITE_COMBINED PTE_ATTRINDX(MT_NORMAL_NC)
 static inline bool pte_is_write_combined(pte_t pte)
 {
@@ -31,9 +29,8 @@ static inline bool pte_is_write_combined(pte_t pte)
 #endif
 	return ((pte_val(pte) & PTE_ATTRINDX_MASK) == PFN_WRITE_COMBINED);
 }
-#endif /* POSTK_DEBUG_ARCH_DEP_12 */
 
-#define ARMV8_IDX_COUNTER0	1
+#define ARMV8_IDX_COUNTER0	0
 #define ARCH_PERF_COUNTER_START	ARMV8_IDX_COUNTER0
 
 #if LINUX_VERSION_CODE < KERNEL_VERSION(3,7,0)

executer/kernel/mcctrl/arch/x86_64/archdeps.c

@@ -2,9 +2,13 @@
 #include <linux/version.h>
 #include <linux/kallsyms.h>
 #include <linux/uaccess.h>
+#include <asm/vsyscall.h>
+#include <asm/vgtod.h>
 #include "config.h"
 #include "../../mcctrl.h"
 
+#define gtod (&VVAR(vsyscall_gtod_data))
+
 //#define SC_DEBUG
 
 #ifdef SC_DEBUG
@@ -54,7 +58,6 @@ int arch_symbols_init(void)
 }
 
 
-#ifdef POSTK_DEBUG_ARCH_DEP_52
 #define VDSO_MAXPAGES 2
 struct vdso {
 	long busy;
@@ -70,8 +73,8 @@ struct vdso {
 	long hpet_phys;
 	void *pvti_virt;
 	long pvti_phys;
+	void *vgtod_virt;
 };
-#endif /*POSTK_DEBUG_ARCH_DEP_52*/
 
 unsigned long
 reserve_user_space_common(struct mcctrl_usrdata *usrdata, unsigned long start, unsigned long end);
@@ -207,6 +210,7 @@ void get_vdso_info(ihk_os_t os, long vdso_rpa)
 #endif
 	}
 
+	vdso->vgtod_virt = (void *)gtod;
 out:
 	wmb();
 	vdso->busy = 0;
@@ -257,25 +261,35 @@ struct trans_uctx {
 };
 
 void
-restore_fs(unsigned long fs)
+restore_tls(unsigned long addr)
 {
-	wrmsrl(MSR_FS_BASE, fs);
+	wrmsrl(MSR_FS_BASE, addr);
 }
 
 void
-save_fs_ctx(void *ctx)
+save_tls_ctx(void __user *ctx)
 {
-	struct trans_uctx *tctx = ctx;
+	struct trans_uctx __user *tctx = ctx;
+	struct trans_uctx kctx;
 
-	rdmsrl(MSR_FS_BASE, tctx->fs);
+	if (copy_from_user(&kctx, tctx, sizeof(struct trans_uctx))) {
+		pr_err("%s: copy_from_user failed.\n", __func__);
+		return;
+	}
+	rdmsrl(MSR_FS_BASE, kctx.fs);
 }
 
 unsigned long
-get_fs_ctx(void *ctx)
+get_tls_ctx(void __user *ctx)
 {
-	struct trans_uctx *tctx = ctx;
+	struct trans_uctx __user *tctx = ctx;
+	struct trans_uctx kctx;
 
-	return tctx->fs;
+	if (copy_from_user(&kctx, tctx, sizeof(struct trans_uctx))) {
+		pr_err("%s: copy_from_user failed.\n", __func__);
+		return 0;
+	}
+	return kctx.fs;
 }
 
 unsigned long
@@ -356,11 +370,17 @@ out:
 	return error;
 }
 
-#ifdef POSTK_DEBUG_ARCH_DEP_12
 #define PFN_WRITE_COMBINED _PAGE_PWT
 static inline bool pte_is_write_combined(pte_t pte)
 {
 	return ((pte_flags(pte) & _PAGE_PWT) && !(pte_flags(pte) & _PAGE_PCD));
 }
-#endif /* POSTK_DEBUG_ARCH_DEP_12 */
 
+/*
+ * The assembler switch_ctx is save/load registers in the context.
+ * Do TLS save/load and register host_thread with ioctl.
+ */
+long arch_switch_ctx(struct uti_switch_ctx_desc *desc)
+{
+	return 0;
+}

executer/kernel/mcctrl/arch/x86_64/include/archdeps.h

@@ -9,14 +9,12 @@
 extern int translate_rva_to_rpa(ihk_os_t os, unsigned long rpt, unsigned long rva,
 				unsigned long *rpap, unsigned long *pgsizep);
 
-#ifdef POSTK_DEBUG_ARCH_DEP_12
 #define PFN_WRITE_COMBINED _PAGE_PWT
 
 static inline bool pte_is_write_combined(pte_t pte)
 {
 	return ((pte_flags(pte) & _PAGE_PWT) && !(pte_flags(pte) & _PAGE_PCD));
 }
-#endif /* POSTK_DEBUG_ARCH_DEP_12 */
 
 #define ARCH_PERF_COUNTER_START	0
 

executer/kernel/mcctrl/control.c

@@ -44,7 +44,6 @@
 #include <config.h>
 #include "mcctrl.h"
 #include <ihk/ihk_host_user.h>
-#include <ihklib_rusage.h>
 #include <rusage.h>
 #if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 11, 0)
 #include <uapi/linux/sched/types.h>
@@ -87,8 +86,19 @@ int syscall_backward(struct mcctrl_usrdata *, int, unsigned long, unsigned long,
                      unsigned long, unsigned long, unsigned long,
                      unsigned long, unsigned long *);
 
+struct mcos_handler_info {
+	int pid;
+	int cpu;
+	struct mcctrl_usrdata *ud;
+	struct file *file;
+	unsigned long user_start;
+	unsigned long user_end;
+	unsigned long prepare_thread;
+};
+
 static long mcexec_prepare_image(ihk_os_t os,
-                                 struct program_load_desc * __user udesc)
+				struct program_load_desc * __user udesc,
+				struct file *file)
 {
 	struct program_load_desc *desc = NULL;
 	struct program_load_desc *pdesc = NULL;
@@ -100,6 +110,7 @@ static long mcexec_prepare_image(ihk_os_t os,
 	struct mcctrl_per_proc_data *ppd = NULL;
 	int num_sections;
 	int free_ikc_pointers = 1;
+	struct mcos_handler_info *info;
 
 	if (!usrdata) {
 		pr_err("%s: error: mcctrl_usrdata not found\n", __func__);
@@ -122,6 +133,14 @@ static long mcexec_prepare_image(ihk_os_t os,
 		goto free_out;
 	}
 
+	info = ihk_os_get_mcos_private_data(file);
+	if (!info) {
+		ret = -EFAULT;
+		goto free_out;
+	}
+	/* To serialize SCD_MSG_SCHEDULE_PROCESS and SCD_MSG_CLEANUP_PROCESS */
+	info->cpu = desc->cpu;
+
 	ppd = mcctrl_get_per_proc_data(usrdata, desc->pid);
 	if (!ppd) {
 		printk("%s: ERROR: no per process data for PID %d\n",
@@ -193,6 +212,11 @@ static long mcexec_prepare_image(ihk_os_t os,
 		/* either send or remote prepare_process failed */
 		goto put_and_free_out;
 	}
+	/*
+	 * Used as SCD_MSG_CLEANUP_PROCESS target which isn't scheduled
+	 * with SCD_MSG_SCHEDULE_PROCESS
+	 */
+	info->prepare_thread = pdesc->rprocess;
 
 	/* Update rpgtable */
 	ppd->rpgtable = pdesc->rpgtable;
@@ -307,30 +331,10 @@ int mcexec_transfer_image(ihk_os_t os, struct remote_transfer *__user upt)
 #endif
 }
 
-struct mcos_handler_info {
-	int pid;
-	int cpu;
-	struct mcctrl_usrdata *ud;
-	struct file *file;
-	unsigned long user_start;
-	unsigned long user_end;
-};
-
 struct mcos_handler_info;
 static LIST_HEAD(host_threads); /* Used for FS switch */
 DEFINE_RWLOCK(host_thread_lock);
 
-/* Info of Linux counterpart of migrated-to-Linux thread */
-struct host_thread {
-	struct list_head list;
-	struct mcos_handler_info *handler;
-	int     pid;
-	int     tid;
-	unsigned long usp;
-	unsigned long lfs;
-	unsigned long rfs;
-};
-
 struct mcos_handler_info *new_mcos_handler_info(ihk_os_t os, struct file *file)
 {
 	struct mcos_handler_info *info;
@@ -391,6 +395,7 @@ static void release_handler(ihk_os_t os, void *param)
 	memset(&isp, '\0', sizeof isp);
 	isp.msg = SCD_MSG_CLEANUP_PROCESS;
 	isp.pid = info->pid;
+	isp.arg = info->prepare_thread;
 
 	dprintk("%s: SCD_MSG_CLEANUP_PROCESS, info: %p, cpu: %d\n",
 			__FUNCTION__, info, info->cpu);
@@ -426,6 +431,7 @@ static long mcexec_start_image(ihk_os_t os,
 	struct mcctrl_channel *c;
 	struct mcctrl_usrdata *usrdata = ihk_host_os_get_usrdata(os);
 	struct mcos_handler_info *info;
+	struct mcos_handler_info *prev_info;
 	int ret = 0;
 
 	if (!usrdata) {
@@ -446,6 +452,7 @@ static long mcexec_start_image(ihk_os_t os,
 		goto out;
 	}
 
+	prev_info = ihk_os_get_mcos_private_data(file);
 	info = new_mcos_handler_info(os, file);
 	if (info == NULL) {
 		ret = -ENOMEM;
@@ -456,6 +463,7 @@ static long mcexec_start_image(ihk_os_t os,
 	info->cpu = desc->cpu;
 	info->user_start = desc->user_start;
 	info->user_end = desc->user_end;
+	info->prepare_thread = prev_info->prepare_thread;
 	ihk_os_register_release_handler(file, release_handler, info);
 	ihk_os_set_mcos_private_data(file, info);
 
@@ -472,8 +480,10 @@ static long mcexec_start_image(ihk_os_t os,
 	ret = mcctrl_ikc_send(os, desc->cpu, &isp);
 	if (ret < 0) {
 		printk("%s: error: sending IKC msg\n", __FUNCTION__);
+		goto out;
 	}
-
+	/* clear prepared thread struct */
+	info->prepare_thread = 0;
 out:
 	kfree(desc);
 	return ret;
@@ -577,17 +587,14 @@ extern int mckernel_cpu_2_linux_cpu(struct mcctrl_usrdata *udp, int cpu_id);
 static long mcexec_get_cpuset(ihk_os_t os, unsigned long arg)
 {
 	struct mcctrl_usrdata *udp = ihk_host_os_get_usrdata(os);
-	struct mcctrl_part_exec *pe;
+	struct mcctrl_part_exec *pe = NULL, *pe_itr;
 	struct get_cpu_set_arg req;
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 	struct mcctrl_cpu_topology *cpu_top, *cpu_top_i;
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-	struct cpu_topology *cpu_top, *cpu_top_i;
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 	struct cache_topology *cache_top;
 	int cpu, cpus_assigned, cpus_to_assign, cpu_prev;
 	int ret = 0;
 	int mcexec_linux_numa;
+	int pe_list_len = 0;
 	cpumask_t *mcexec_cpu_set = NULL;
 	cpumask_t *cpus_used = NULL;
 	cpumask_t *cpus_to_use = NULL;
@@ -607,24 +614,126 @@ static long mcexec_get_cpuset(ihk_os_t os, unsigned long arg)
 		return -EINVAL;
 	}
 
-	pe = &udp->part_exec;
-
-	mutex_lock(&pe->lock);
-
 	if (copy_from_user(&req, (void *)arg, sizeof(req))) {
-		printk("%s: error copying user request\n", __FUNCTION__);
+		pr_err("%s: error copying user request\n", __func__);
 		ret = -EINVAL;
-		goto put_and_unlock_out;
+		goto put_out;
 	}
 
-	/* First process to enter CPU partitioning */
-	if (pe->nr_processes == -1) {
+	/* User requested CPU mask? */
+	if (req.req_cpu_list && req.req_cpu_list_len) {
+		char *cpu_list = NULL;
+
+		cpu_list = kmalloc(req.req_cpu_list_len, GFP_KERNEL);
+		if (!cpu_list) {
+			printk("%s: error: allocating CPU list\n", __FUNCTION__);
+			ret = -ENOMEM;
+			goto put_out;
+		}
+
+		if (copy_from_user(cpu_list,
+					req.req_cpu_list, req.req_cpu_list_len)) {
+			printk("%s: error copying CPU list request\n", __FUNCTION__);
+			kfree(cpu_list);
+			ret = -EINVAL;
+			goto put_out;
+		}
+
+		cpus_used = kmalloc(sizeof(cpumask_t), GFP_KERNEL);
+		cpus_to_use = kmalloc(sizeof(cpumask_t), GFP_KERNEL);
+		if (!cpus_to_use || !cpus_used) {
+			printk("%s: error: allocating CPU mask\n", __FUNCTION__);
+			ret = -ENOMEM;
+			kfree(cpu_list);
+			goto put_out;
+		}
+		memset(cpus_used, 0, sizeof(cpumask_t));
+		memset(cpus_to_use, 0, sizeof(cpumask_t));
+
+		/* Parse CPU list */
+		if (cpulist_parse(cpu_list, cpus_to_use) < 0) {
+			printk("%s: invalid CPUs requested: %s\n",
+				__FUNCTION__, cpu_list);
+			ret = -EINVAL;
+			kfree(cpu_list);
+			goto put_out;
+		}
+
+		memcpy(cpus_used, cpus_to_use, sizeof(cpumask_t));
+
+		/* Copy mask to user-space */
+		if (copy_to_user(req.cpu_set, cpus_used,
+					(req.cpu_set_size < sizeof(cpumask_t) ?
+					 req.cpu_set_size : sizeof(cpumask_t)))) {
+			printk("%s: error copying mask to user\n", __FUNCTION__);
+			ret = -EINVAL;
+			kfree(cpu_list);
+			goto put_out;
+		}
+
+		/* Copy IKC target core */
+		cpu = cpumask_next(-1, cpus_used);
+		if (copy_to_user(req.target_core, &cpu, sizeof(cpu))) {
+			printk("%s: error copying target core to user\n",
+					__FUNCTION__);
+			ret = -EINVAL;
+			kfree(cpu_list);
+			goto put_out;
+		}
+
+		/* Save in per-process structure */
+		memcpy(&ppd->cpu_set, cpus_used, sizeof(cpumask_t));
+		ppd->ikc_target_cpu = cpu;
+		printk("%s: %s -> target McKernel CPU: %d\n",
+			__func__, cpu_list, cpu);
+
+		ret = 0;
+		kfree(cpu_list);
+		goto put_out;
+	}
+
+	mutex_lock(&udp->part_exec_lock);
+	/* Find part_exec having same node_proxy */
+	list_for_each_entry_reverse(pe_itr, &udp->part_exec_list, chain) {
+		pe_list_len++;
+		if (pe_itr->node_proxy_pid == req.ppid) {
+			pe = pe_itr;
+			break;
+		}
+	}
+
+	if (!pe) {
+		/* First process to enter CPU partitioning */
+		pr_debug("%s: pe_list_len:%d\n", __func__, pe_list_len);
+		if (pe_list_len >= PE_LIST_MAXLEN) {
+			/* delete head entry of pe_list */
+			pe_itr = list_first_entry(&udp->part_exec_list,
+					struct mcctrl_part_exec, chain);
+			list_del(&pe_itr->chain);
+			kfree(pe_itr);
+		}
+
+		pe = kzalloc(sizeof(struct mcctrl_part_exec), GFP_KERNEL);
+		if (!pe) {
+			mutex_unlock(&udp->part_exec_lock);
+			ret = -ENOMEM;
+			goto put_out;
+		}
+		/* Init part_exec */
+		mutex_init(&pe->lock);
+		INIT_LIST_HEAD(&pe->pli_list);
 		pe->nr_processes = req.nr_processes;
 		pe->nr_processes_left = req.nr_processes;
+		pe->nr_processes_joined = 0;
+		pe->node_proxy_pid = req.ppid;
+
+		list_add_tail(&pe->chain, &udp->part_exec_list);
 		dprintk("%s: nr_processes: %d (partitioned exec starts)\n",
-				__FUNCTION__,
-				pe->nr_processes);
+				__func__, pe->nr_processes);
 	}
+	mutex_unlock(&udp->part_exec_lock);
+
+	mutex_lock(&pe->lock);
 
 	if (pe->nr_processes != req.nr_processes) {
 		printk("%s: error: requested number of processes"
@@ -634,7 +743,15 @@ static long mcexec_get_cpuset(ihk_os_t os, unsigned long arg)
 		goto put_and_unlock_out;
 	}
 
+	if (pe->nr_processes_joined >= pe->nr_processes) {
+		printk("%s: too many processes have joined to the group of %d\n",
+				__func__, req.ppid);
+		ret = -EINVAL;
+		goto put_and_unlock_out;
+	}
+
 	--pe->nr_processes_left;
+	++pe->nr_processes_joined;
 	dprintk("%s: nr_processes: %d, nr_processes_left: %d\n",
 			__FUNCTION__,
 			pe->nr_processes,
@@ -720,8 +837,6 @@ static long mcexec_get_cpuset(ihk_os_t os, unsigned long arg)
 				wake_up_interruptible(&pli_next->pli_wq);
 			}
 
-			/* Reset process counter to start state */
-			pe->nr_processes = -1;
 			ret = -ETIMEDOUT;
 			goto put_and_unlock_out;
 		}
@@ -969,16 +1084,8 @@ next_cpu:
 	/* Commit used cores to OS structure */
 	memcpy(&pe->cpus_used, cpus_used, sizeof(*cpus_used));
 
-	/* Reset if last process */
-	if (pe->nr_processes_left == 0) {
-		dprintk("%s: nr_processes: %d (partitioned exec ends)\n",
-				__FUNCTION__,
-				pe->nr_processes);
-		pe->nr_processes = -1;
-		memset(&pe->cpus_used, 0, sizeof(pe->cpus_used));
-	}
-	/* Otherwise wake up next process in list */
-	else {
+	/* If not last process, wake up next process in list */
+	if (pe->nr_processes_left != 0) {
 		++pe->process_rank;
 		pli_next = list_first_entry(&pe->pli_list,
 			struct process_list_item, list);
@@ -991,11 +1098,14 @@ next_cpu:
 	ret = 0;
 
 put_and_unlock_out:
+	mutex_unlock(&pe->lock);
+
+put_out:
+	mcctrl_put_per_proc_data(ppd);
+
 	kfree(cpus_to_use);
 	kfree(cpus_used);
 	kfree(mcexec_cpu_set);
-	mcctrl_put_per_proc_data(ppd);
-	mutex_unlock(&pe->lock);
 
 	return ret;
 }
@@ -1199,7 +1309,7 @@ int mcexec_syscall(struct mcctrl_usrdata *ud, struct ikc_scd_packet *packet)
 	ppd = mcctrl_get_per_proc_data(ud, pid);
 
 	if (unlikely(!ppd)) {
-		kprintf("%s: ERROR: no per-process structure for PID %d, "
+		dprintk("%s: ERROR: no per-process structure for PID %d, "
 				"syscall nr: %lu\n",
 				__FUNCTION__, pid, packet->req.number);
 
@@ -1414,7 +1524,7 @@ retry_alloc:
 			__FUNCTION__, task_pid_vnr(current), packet->ref);
 
 	mb();
-	if (!packet->req.valid) {
+	if (!smp_load_acquire(&packet->req.valid)) {
 		printk("%s: ERROR: stray wakeup pid: %d, tid: %d: SC %lu\n",
 				__FUNCTION__,
 				task_tgid_vnr(current),
@@ -1424,7 +1534,7 @@ retry_alloc:
 		goto retry;
 	}
 
-	packet->req.valid = 0; /* ack */
+	smp_store_release(&packet->req.valid,  0); /* ack */
 	dprintk("%s: system call: %d, args[0]: %lu, args[1]: %lu, args[2]: %lu, "
 			"args[3]: %lu, args[4]: %lu, args[5]: %lu\n",
 			__FUNCTION__,
@@ -2376,7 +2486,7 @@ long mcctrl_getrusage(ihk_os_t ihk_os, struct mcctrl_ioctl_getrusage_desc *__use
 {
 	struct mcctrl_ioctl_getrusage_desc desc;
 	struct rusage_global *rusage_global = ihk_os_get_rusage(ihk_os);
-	struct mckernel_rusage *rusage = NULL;
+	struct ihk_os_rusage *rusage = NULL;
 	int ret = 0;
 	int i;
 	unsigned long ut;
@@ -2388,13 +2498,13 @@ long mcctrl_getrusage(ihk_os_t ihk_os, struct mcctrl_ioctl_getrusage_desc *__use
 		goto out;
 	}
 
-	rusage = kmalloc(sizeof(struct mckernel_rusage), GFP_KERNEL);
+	rusage = kmalloc(sizeof(struct ihk_os_rusage), GFP_KERNEL);
 	if (!rusage) {
 		printk("%s: kmalloc failed\n", __FUNCTION__);
 		ret = -ENOMEM;
 		goto out;
 	}
-	memset(rusage, 0, sizeof(struct mckernel_rusage));
+	memset(rusage, 0, sizeof(struct ihk_os_rusage));
 
 	/* Compile statistics */
 	for (i = 0; i < IHK_MAX_NUM_PGSIZES; i++) {
@@ -2415,15 +2525,17 @@ long mcctrl_getrusage(ihk_os_t ihk_os, struct mcctrl_ioctl_getrusage_desc *__use
 		st += rusage_global->cpu[i].system_tsc * rusage_global->ns_per_tsc / 1000;
 		rusage->cpuacct_usage_percpu[i] = wt;
 	}
-	rusage->cpuacct_stat_system = st / 10000000;
-	rusage->cpuacct_stat_user = ut / 10000000;
+	rusage->cpuacct_stat_system = (st + 10000000 - 1) / 10000000;
+	rusage->cpuacct_stat_user = (ut + 10000000 - 1) / 10000000;
 	rusage->cpuacct_usage = ut;
 
 	rusage->num_threads = rusage_global->num_threads;
 	rusage->max_num_threads = rusage_global->max_num_threads;
 
-	if (desc.size_rusage > sizeof(struct mckernel_rusage)) {
-		printk("%s: desc.size_rusage=%ld > sizeof(struct mckernel_rusage)=%ld\n", __FUNCTION__, desc.size_rusage, sizeof(struct mckernel_rusage));
+	if (desc.size_rusage > sizeof(struct ihk_os_rusage)) {
+		printk("%s: desc.size_rusage=%ld > sizeof(struct mckernel_rusage)=%ld\n",
+		       __func__, desc.size_rusage,
+		       sizeof(struct ihk_os_rusage));
 		ret = -EINVAL;
 		goto out;
 	}
@@ -2444,10 +2556,10 @@ long mcctrl_getrusage(ihk_os_t ihk_os, struct mcctrl_ioctl_getrusage_desc *__use
 
 extern void *get_user_sp(void);
 extern void set_user_sp(unsigned long);
-extern void restore_fs(unsigned long fs);
-extern void save_fs_ctx(void *);
-extern unsigned long get_fs_ctx(void *);
-extern unsigned long get_rsp_ctx(void *);
+extern void restore_tls(unsigned long addr);
+extern void save_tls_ctx(void __user *ctx);
+extern unsigned long get_tls_ctx(void __user *ctx);
+extern unsigned long get_rsp_ctx(void *ctx);
 
 long mcexec_uti_get_ctx(ihk_os_t os, struct uti_get_ctx_desc __user *udesc)
 {
@@ -2491,14 +2603,15 @@ long mcexec_uti_get_ctx(ihk_os_t os, struct uti_get_ctx_desc __user *udesc)
 	return rc;
 }
 
-long mcexec_uti_save_fs(ihk_os_t os, struct uti_save_fs_desc __user *udesc, struct file *file)
+long mcctrl_switch_ctx(ihk_os_t os, struct uti_switch_ctx_desc __user *udesc,
+		       struct file *file)
 {
 	int rc = 0;
 	void *usp = get_user_sp();
 	struct mcos_handler_info *info;
 	struct host_thread *thread;
 	unsigned long flags;
-	struct uti_save_fs_desc desc;
+	struct uti_switch_ctx_desc desc;
 	struct mcctrl_usrdata *usrdata = ihk_host_os_get_usrdata(os);
 	struct mcctrl_per_proc_data *ppd;
 
@@ -2508,21 +2621,26 @@ long mcexec_uti_save_fs(ihk_os_t os, struct uti_save_fs_desc __user *udesc, stru
 		goto out;
 	}
 
-	if(copy_from_user(&desc, udesc, sizeof(struct uti_save_fs_desc))) {
+	if (copy_from_user(&desc, udesc, sizeof(struct uti_switch_ctx_desc))) {
 		printk("%s: Error: copy_from_user failed\n", __FUNCTION__);
 		rc = -EFAULT;
 		goto out;
 	}
 
-	save_fs_ctx(desc.lctx);
+	rc = arch_switch_ctx(&desc);
+	if (rc < 0) {
+		goto out;
+	}
+
+	save_tls_ctx(desc.lctx);
 	info = ihk_os_get_mcos_private_data(file);
 	thread = kmalloc(sizeof(struct host_thread), GFP_KERNEL);
 	memset(thread, '\0', sizeof(struct host_thread));
 	thread->pid = task_tgid_vnr(current);
 	thread->tid = task_pid_vnr(current);
 	thread->usp = (unsigned long)usp;
-	thread->lfs = get_fs_ctx(desc.lctx);
-	thread->rfs = get_fs_ctx(desc.rctx);
+	thread->ltls = get_tls_ctx(desc.lctx);
+	thread->rtls = get_tls_ctx(desc.rctx);
 	thread->handler = info;
 
 	write_lock_irqsave(&host_thread_lock, flags);
@@ -2568,9 +2686,9 @@ mcexec_sig_thread(ihk_os_t os, unsigned long arg, struct file *file)
 	read_unlock_irqrestore(&host_thread_lock, flags);
 	if (thread) {
 		if (arg)
-			restore_fs(thread->lfs);
+			restore_tls(thread->ltls);
 		else
-			restore_fs(thread->rfs);
+			restore_tls(thread->rtls);
 		goto out;
 	}
 	ret = -EINVAL;
@@ -2774,8 +2892,7 @@ long mcexec_syscall_thread(ihk_os_t os, unsigned long arg, struct file *file)
 		return -EFAULT;
 	}
 
-	/* debug */
-	if (0 && param.number == __NR_futex) {
+	if (param.number == __NR_futex) {
 		struct uti_futex_resp resp = {
 			.done = 0
 		};
@@ -2971,13 +3088,8 @@ mcexec_uti_attr(ihk_os_t os, struct uti_attr_desc __user *_desc)
 	cpumask_t *cpuset = NULL, *env_cpuset = NULL;
 	struct mcctrl_usrdata *ud = ihk_host_os_get_usrdata(os);
 	ihk_device_t dev = ihk_os_to_dev(os);
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 	struct mcctrl_cpu_topology *cpu_topo;
 	struct mcctrl_cpu_topology *target_cpu = NULL;
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-	struct cpu_topology *cpu_topo;
-	struct cpu_topology *target_cpu = NULL;
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 	struct node_topology *node_topo;
 	struct ihk_cache_topology *lcache_topo;
 	struct ihk_node_topology *lnode_topo;
@@ -3200,13 +3312,51 @@ out:
 	return rc;
 }
 
+static int __mcctrl_control_perm(unsigned int request)
+{
+	int ret = 0;
+	kuid_t euid;
+
+	/* black list */
+	switch (request) {
+	case IHK_OS_AUX_PERF_NUM:
+	case IHK_OS_AUX_PERF_SET:
+	case IHK_OS_AUX_PERF_GET:
+	case IHK_OS_AUX_PERF_ENABLE:
+	case IHK_OS_AUX_PERF_DISABLE:
+	case IHK_OS_AUX_PERF_DESTROY:
+		euid = current_euid();
+		pr_debug("%s: request=0x%x, euid=%u\n",
+			 __func__, request, euid.val);
+		if (euid.val) {
+			ret = -EPERM;
+		}
+		break;
+	default:
+		break;
+	}
+	pr_debug("%s: request=0x%x, ret=%d\n", __func__, request, ret);
+
+	return ret;
+}
+
 long __mcctrl_control(ihk_os_t os, unsigned int req, unsigned long arg,
                       struct file *file)
 {
+	int ret;
+
+	ret = __mcctrl_control_perm(req);
+	if (ret) {
+		pr_err("%s: error: permission denied, req: %x\n",
+		       __func__, req);
+		return ret;
+	}
+
 	switch (req) {
 	case MCEXEC_UP_PREPARE_IMAGE:
 		return mcexec_prepare_image(os,
-		                            (struct program_load_desc *)arg);
+					(struct program_load_desc *)arg,
+					file);
 	case MCEXEC_UP_TRANSFER:
 		return mcexec_transfer_image(os, (struct remote_transfer *)arg);
 
@@ -3272,8 +3422,9 @@ long __mcctrl_control(ihk_os_t os, unsigned int req, unsigned long arg,
 	case MCEXEC_UP_UTI_GET_CTX:
 		return mcexec_uti_get_ctx(os, (struct uti_get_ctx_desc *)arg);
 
-	case MCEXEC_UP_UTI_SAVE_FS:
-		return mcexec_uti_save_fs(os, (struct uti_save_fs_desc *)arg, file);
+	case MCEXEC_UP_UTI_SWITCH_CTX:
+		return mcctrl_switch_ctx(os, (struct uti_switch_ctx_desc *)arg,
+					 file);
 
 	case MCEXEC_UP_SIG_THREAD:
 		return mcexec_sig_thread(os, arg, file);
@@ -3320,14 +3471,6 @@ long __mcctrl_control(ihk_os_t os, unsigned int req, unsigned long arg,
 	return -EINVAL;
 }
 
-/* Per-CPU register manipulation functions */
-struct mcctrl_os_cpu_response {
-	int done;
-	unsigned long val;
-	int err;
-	wait_queue_head_t wq;
-};
-
 int mcctrl_get_request_os_cpu(ihk_os_t os, int *ret_cpu)
 {
 	struct mcctrl_usrdata *usrdata;
@@ -3379,7 +3522,8 @@ int mcctrl_get_request_os_cpu(ihk_os_t os, int *ret_cpu)
 	*ret_cpu = ch->send.queue->read_cpu;
 	ret = 0;
 
-	printk("%s: OS: %p, CPU: %d\n", __FUNCTION__, os, *ret_cpu);
+	pr_info("%s: OS: %lx, CPU: %d\n",
+		__func__, (unsigned long)os, *ret_cpu);
 
 out_put_ppd:
 	mcctrl_put_per_thread_data(ptd);
@@ -3390,73 +3534,67 @@ out_put_ppd:
 	return ret;
 }
 
-void mcctrl_os_read_write_cpu_response(ihk_os_t os,
-		struct ikc_scd_packet *pisp)
-{
-	struct mcctrl_os_cpu_response *resp;
-
-	/* XXX: What if caller thread is unblocked by a signal
-	 * before this message arrives? */
-	resp = pisp->resp;
-	if (!resp) {
-		return;
-	}
-
-	resp->val = pisp->desc.val;
-	resp->done = 1;
-	resp->err = pisp->err;
-	wake_up_interruptible(&resp->wq);
-}
-
 int __mcctrl_os_read_write_cpu_register(ihk_os_t os, int cpu,
 		struct ihk_os_cpu_register *desc,
 		enum mcctrl_os_cpu_operation op)
 {
+	struct mcctrl_usrdata *udp = ihk_host_os_get_usrdata(os);
 	struct ikc_scd_packet isp;
-	struct mcctrl_os_cpu_response resp;
+	struct ihk_os_cpu_register *ldesc = NULL;
+	int do_free = 0;
 	int ret = -EINVAL;
 
+	if (!udp) {
+		pr_err("%s: error: mcctrl_usrdata not found\n", __func__);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (cpu < 0 || cpu >= udp->cpu_info->n_cpus) {
+		pr_err("%s: error: cpu (%d) is out of range\n",
+		       __func__, cpu);
+		ret = -EINVAL;
+		goto out;
+
+	}
+
+	/* Keep a dynamic structure around that can
+	 * survive an early return due to a signal */
+	ldesc = kmalloc(sizeof(*ldesc), GFP_KERNEL);
+	if (!ldesc) {
+		printk("%s: ERROR: allocating cpu register desc\n", __FUNCTION__);
+		return -ENOMEM;
+	}
+	*ldesc = *desc;
+
 	memset(&isp, '\0', sizeof(struct ikc_scd_packet));
 	isp.msg = SCD_MSG_CPU_RW_REG;
 	isp.op = op;
-	isp.desc = *desc;
-	isp.resp = &resp;
+	isp.pdesc = virt_to_phys(ldesc);
 
-	resp.done = 0;
-	resp.err = 0;
-	init_waitqueue_head(&resp.wq);
-
-	mb();
-	ret = mcctrl_ikc_send(os, cpu, &isp);
-	if (ret < 0) {
+	ret = mcctrl_ikc_send_wait(os, cpu, &isp, 0, NULL, &do_free, 1, ldesc);
+	if (ret != 0) {
 		printk("%s: ERROR sending IKC msg: %d\n", __FUNCTION__, ret);
 		goto out;
 	}
 
-	/* Wait for response */
-	ret = wait_event_interruptible(resp.wq, resp.done);
-	if (ret < 0) {
-		printk("%s: ERROR after wait: %d\n", __FUNCTION__, ret);
-		goto out;
-	}
-
-	ret = resp.err;
-	if (ret != 0) {
-		printk("%s: ERROR receive: %d\n", __FUNCTION__, resp.err);
-		goto out;
-	}
-
 	/* Update if read */
-	if (ret == 0 && op == MCCTRL_OS_CPU_READ_REGISTER) {
-		desc->val = resp.val;
+	if (op == MCCTRL_OS_CPU_READ_REGISTER) {
+		desc->val = ldesc->val;
 	}
 
-	dprintk("%s: MCCTRL_OS_CPU_%s_REGISTER: reg: 0x%lx, val: 0x%lx\n",
+	/* Notify caller (for future async implementation) */
+	atomic_set(&desc->sync, 1);
+
+	dprintk("%s: MCCTRL_OS_CPU_%s_REGISTER: CPU: %d, addr_ext: 0x%lx, val: 0x%lx\n",
 		__FUNCTION__,
-		(op == MCCTRL_OS_CPU_READ_REGISTER ? "READ" : "WRITE"),
-		desc->addr, desc->val);
+		(op == MCCTRL_OS_CPU_READ_REGISTER ? "READ" : "WRITE"), cpu,
+		desc->addr_ext, desc->val);
 
 out:
+	if (do_free) {
+		kfree(ldesc);
+	}
 	return ret;
 }
 

executer/kernel/mcctrl/driver.c

@@ -21,8 +21,10 @@
  *  2013/08/19 shirasawa mcexec forward signal to MIC process
  */
 
-#include <linux/sched.h>
+#include <linux/init.h>
 #include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
 #include <linux/fs.h>
 #include <linux/miscdevice.h>
 #include <linux/slab.h>
@@ -80,11 +82,13 @@ static struct ihk_os_user_call_handler mcctrl_uchs[] = {
 	{ .request = MCEXEC_UP_CLOSE_EXEC, .func = mcctrl_ioctl },
 	{ .request = MCEXEC_UP_GET_CRED, .func = mcctrl_ioctl },
 	{ .request = MCEXEC_UP_GET_CREDV, .func = mcctrl_ioctl },
+#ifdef MCEXEC_BIND_MOUNT
 	{ .request = MCEXEC_UP_SYS_MOUNT, .func = mcctrl_ioctl },
 	{ .request = MCEXEC_UP_SYS_UMOUNT, .func = mcctrl_ioctl },
 	{ .request = MCEXEC_UP_SYS_UNSHARE, .func = mcctrl_ioctl },
+#endif // MCEXEC_BIND_MOUNT
 	{ .request = MCEXEC_UP_UTI_GET_CTX, .func = mcctrl_ioctl },
-	{ .request = MCEXEC_UP_UTI_SAVE_FS, .func = mcctrl_ioctl },
+	{ .request = MCEXEC_UP_UTI_SWITCH_CTX, .func = mcctrl_ioctl },
 	{ .request = MCEXEC_UP_SIG_THREAD, .func = mcctrl_ioctl },
 	{ .request = MCEXEC_UP_SYSCALL_THREAD, .func = mcctrl_ioctl },
 	{ .request = MCEXEC_UP_TERMINATE_THREAD, .func = mcctrl_ioctl },
@@ -170,14 +174,6 @@ error_cleanup_channels:
 int mcctrl_os_shutdown_notifier(int os_index)
 {
 	if (os[os_index]) {
-		/* Wait for os running */
-		if (ihk_os_wait_for_status(os[os_index], IHK_OS_STATUS_RUNNING, 0, 200) != 0) {
-			printk("IHK: OS does not become RUNNING in shutdown. Force shutdown.\n");
-			/* send nmi to force shutdown */
-			ihk_os_send_nmi(os[os_index], 3);
-			mdelay(200);
-		}
-
 		pager_cleanup();
 		sysfsm_cleanup(os[os_index]);
 		free_topology_info(os[os_index]);

executer/kernel/mcctrl/ikc.c

@@ -208,6 +208,8 @@ static int syscall_packet_handler(struct ihk_ikc_channel_desc *c,
 	case SCD_MSG_PERF_ACK:
 	case SCD_MSG_SEND_SIGNAL_ACK:
 	case SCD_MSG_PROCFS_ANSWER:
+	case SCD_MSG_REMOTE_PAGE_FAULT_ANSWER:
+	case SCD_MSG_CPU_RW_REG_RESP:
 		mcctrl_wakeup_cb(__os, pisp);
 		break;
 
@@ -238,10 +240,6 @@ static int syscall_packet_handler(struct ihk_ikc_channel_desc *c,
 		get_vdso_info(__os, pisp->arg);
 		break;
 
-	case SCD_MSG_CPU_RW_REG_RESP:
-		mcctrl_os_read_write_cpu_response(__os, pisp);
-		break;
-
 	case SCD_MSG_EVENTFD:
 		dkprintf("%s: SCD_MSG_EVENTFD,pisp->eventfd_type=%d\n", __FUNCTION__, pisp->eventfd_type);
 		mcctrl_eventfd(__os, pisp);
@@ -464,7 +462,7 @@ int prepare_ikc_channels(ihk_os_t os)
 	int i;
 	int ret = 0;
 
-	usrdata = kzalloc(sizeof(struct mcctrl_usrdata), GFP_KERNEL);
+	usrdata = kzalloc(sizeof(struct mcctrl_usrdata), GFP_ATOMIC);
 	if (!usrdata) {
 		printk("%s: error: allocating mcctrl_usrdata\n", __FUNCTION__);
 		ret = -ENOMEM;
@@ -490,7 +488,7 @@ int prepare_ikc_channels(ihk_os_t os)
 	usrdata->num_channels = usrdata->cpu_info->n_cpus;
 	usrdata->channels = kzalloc(sizeof(struct mcctrl_channel) *
 			usrdata->num_channels,
-			GFP_KERNEL);
+			GFP_ATOMIC);
 
 	if (!usrdata->channels) {
 		printk("Error: cannot allocate channels.\n");
@@ -499,7 +497,7 @@ int prepare_ikc_channels(ihk_os_t os)
 	}
 
 	usrdata->ikc2linux = kzalloc(sizeof(struct ihk_ikc_channel_desc *) *
-			nr_cpu_ids, GFP_KERNEL);
+			nr_cpu_ids, GFP_ATOMIC);
 
 	if (!usrdata->ikc2linux) {
 		printk("Error: cannot allocate ikc2linux channels.\n");
@@ -515,6 +513,7 @@ int prepare_ikc_channels(ihk_os_t os)
 
 	init_waitqueue_head(&usrdata->wq_procfs);
 	mutex_init(&usrdata->reserve_lock);
+	mutex_init(&usrdata->part_exec_lock);
 
 	for (i = 0; i < MCCTRL_PER_PROC_DATA_HASH_SIZE; ++i) {
 		INIT_LIST_HEAD(&usrdata->per_proc_data_hash[i]);
@@ -523,10 +522,8 @@ int prepare_ikc_channels(ihk_os_t os)
 
 	INIT_LIST_HEAD(&usrdata->cpu_topology_list);
 	INIT_LIST_HEAD(&usrdata->node_topology_list);
+	INIT_LIST_HEAD(&usrdata->part_exec_list);
 
-	mutex_init(&usrdata->part_exec.lock);
-	INIT_LIST_HEAD(&usrdata->part_exec.pli_list);
-	usrdata->part_exec.nr_processes = -1;
 	INIT_LIST_HEAD(&usrdata->wakeup_descs_list);
 	spin_lock_init(&usrdata->wakeup_descs_lock);
 
@@ -582,6 +579,18 @@ void destroy_ikc_channels(ihk_os_t os)
 
 	kfree(usrdata->channels);
 	kfree(usrdata->ikc2linux);
+
+	mutex_lock(&usrdata->part_exec_lock);
+	while (!list_empty(&usrdata->part_exec_list)) {
+		struct mcctrl_part_exec *pe;
+
+		pe = list_first_entry(&usrdata->part_exec_list,
+				struct mcctrl_part_exec, chain);
+		list_del(&pe->chain);
+		kfree(pe);
+	}
+	mutex_unlock(&usrdata->part_exec_lock);
+
 	kfree(usrdata);
 }
 

executer/kernel/mcctrl/mcctrl.h

@@ -69,6 +69,9 @@
 #define	SCD_MSG_PROCFS_ANSWER		0x13
 #define	SCD_MSG_PROCFS_RELEASE		0x15
 
+#define SCD_MSG_REMOTE_PAGE_FAULT	0x18
+#define SCD_MSG_REMOTE_PAGE_FAULT_ANSWER 0x19
+
 #define SCD_MSG_DEBUG_LOG		0x20
 
 #define SCD_MSG_SYSFS_REQ_CREATE        0x30
@@ -121,12 +124,6 @@ enum mcctrl_os_cpu_operation {
 	MCCTRL_OS_CPU_MAX_OP
 };
 
-/* Used to wake-up a Linux thread futex_wait()-ing */
-struct uti_futex_resp {
-	int done;
-	wait_queue_head_t wq;
-};
-
 struct ikc_scd_packet {
 	struct ihk_ikc_packet_header header;
 	int msg;
@@ -157,7 +154,7 @@ struct ikc_scd_packet {
 
 		/* SCD_MSG_CPU_RW_REG */
 		struct {
-			struct ihk_os_cpu_register desc;
+			unsigned long pdesc; /* Physical addr of the descriptor */
 			enum mcctrl_os_cpu_operation op;
 			void *resp;
 		};
@@ -172,6 +169,14 @@ struct ikc_scd_packet {
 			void *resp;
 			int *spin_sleep; /* 1: waiting in linux_wait_event() 0: woken up by someone else */
 		} futex;
+
+		/* SCD_MSG_REMOTE_PAGE_FAULT */
+		struct {
+			int target_cpu;
+			int fault_tid;
+			unsigned long fault_address;
+			unsigned long fault_reason;
+		};
 	};
 	/* char padding[8]; */ /* We want the size to be 128 bytes */
 };
@@ -289,11 +294,7 @@ struct cache_topology {
 	struct list_head chain;
 };
 
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 struct mcctrl_cpu_topology {
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-struct cpu_topology {
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 	//struct mcctrl_usrdata *udp;
 	struct ihk_cpu_topology *saved;
 	int mckernel_cpu_id;
@@ -323,13 +324,20 @@ struct process_list_item {
 	wait_queue_head_t pli_wq;
 };
 
+#define PE_LIST_MAXLEN 5
+
 struct mcctrl_part_exec {
 	struct mutex lock;	
 	int nr_processes;
+	/* number of processes to let in / out the synchronization point */
 	int nr_processes_left;
+	/* number of processes which have joined the partition */
+	int nr_processes_joined;
 	int process_rank;
+	pid_t node_proxy_pid;
 	cpumask_t cpus_used;
 	struct list_head pli_list;
+	struct list_head chain;
 };
 
 #define CPU_LONGS (((NR_CPUS) + (BITS_PER_LONG) - 1) / (BITS_PER_LONG))
@@ -352,6 +360,7 @@ struct mcctrl_usrdata {
 	int	job_pos;
 	int	mcctrl_dma_abort;
 	struct mutex reserve_lock;
+	struct mutex part_exec_lock;
 	unsigned long	last_thread_exec;
 	wait_queue_head_t wq_procfs;
 	struct list_head per_proc_data_hash[MCCTRL_PER_PROC_DATA_HASH_SIZE];
@@ -367,7 +376,7 @@ struct mcctrl_usrdata {
 	nodemask_t numa_online;
 	struct list_head cpu_topology_list;
 	struct list_head node_topology_list;
-	struct mcctrl_part_exec part_exec;
+	struct list_head part_exec_list;
 	int perf_event_num;
 };
 
@@ -448,40 +457,8 @@ void mcctrl_put_per_proc_data(struct mcctrl_per_proc_data *ppd);
 int mcctrl_add_per_thread_data(struct mcctrl_per_proc_data *ppd, void *data);
 void mcctrl_put_per_thread_data_unsafe(struct mcctrl_per_thread_data *ptd);
 void mcctrl_put_per_thread_data(struct mcctrl_per_thread_data* ptd);
-#ifdef POSTK_DEBUG_ARCH_DEP_56 /* Strange how to use inline declaration fix. */
-static inline struct mcctrl_per_thread_data *mcctrl_get_per_thread_data(struct mcctrl_per_proc_data *ppd,
-									struct task_struct *task)
-{
-	struct mcctrl_per_thread_data *ptd_iter, *ptd = NULL;
-	int hash = (((uint64_t)task >> 4) & MCCTRL_PER_THREAD_DATA_HASH_MASK);
-	unsigned long flags;
-
-	/* Check if data for this thread exists */
-	write_lock_irqsave(&ppd->per_thread_data_hash_lock[hash], flags);
-
-	list_for_each_entry(ptd_iter, &ppd->per_thread_data_hash[hash], hash) {
-		if (ptd_iter->task == task) {
-			ptd = ptd_iter;
-			break;
-		}
-	}
-
-	if (ptd) {
-		if (atomic_read(&ptd->refcount) <= 0) {
-			printk("%s: ERROR: use-after-free detected (%d)", __FUNCTION__, atomic_read(&ptd->refcount));
-			ptd = NULL;
-			goto out;
-		}
-		atomic_inc(&ptd->refcount);
-	}
-
- out:
-	write_unlock_irqrestore(&ppd->per_thread_data_hash_lock[hash], flags);
-	return ptd;
-}
-#else /* POSTK_DEBUG_ARCH_DEP_56 */
-inline struct mcctrl_per_thread_data *mcctrl_get_per_thread_data(struct mcctrl_per_proc_data *ppd, struct task_struct *task);
-#endif /* POSTK_DEBUG_ARCH_DEP_56 */
+struct mcctrl_per_thread_data *mcctrl_get_per_thread_data(struct mcctrl_per_proc_data *ppd,
+							  struct task_struct *task);
 int mcctrl_clear_pte_range(uintptr_t start, uintptr_t len);
 
 void __return_syscall(ihk_os_t os, struct ikc_scd_packet *packet, 
@@ -526,24 +503,6 @@ void reply_get_cpu_mapping(long req_pa);
 void free_topology_info(ihk_os_t os);
 
 /* archdep.c */
-#ifndef POSTK_DEBUG_ARCH_DEP_52
-#define VDSO_MAXPAGES 2
-struct vdso {
-	long busy;
-	int vdso_npages;
-	char vvar_is_global;
-	char hpet_is_global;
-	char pvti_is_global;
-	char padding;
-	long vdso_physlist[VDSO_MAXPAGES];
-	void *vvar_virt;
-	long vvar_phys;
-	void *hpet_virt;
-	long hpet_phys;
-	void *pvti_virt;
-	long pvti_phys;
-};
-#endif /*POSTK_DEBUG_ARCH_DEP_52*/
 
 int reserve_user_space(struct mcctrl_usrdata *usrdata, unsigned long *startp,
 		unsigned long *endp);
@@ -573,8 +532,28 @@ struct ihk_perf_event_attr{
 };
 
 struct mcctrl_ioctl_getrusage_desc {
-	void* rusage;
+	struct ihk_os_rusage *rusage;
 	size_t size_rusage;
 };
 
+/* uti */
+long mcctrl_switch_ctx(ihk_os_t os, struct uti_switch_ctx_desc __user *desc,
+			struct file *file);
+long arch_switch_ctx(struct uti_switch_ctx_desc *desc);
+
+struct host_thread {
+	struct list_head list;
+	struct mcos_handler_info *handler;
+	int     pid;
+	int     tid;
+	unsigned long usp;
+	unsigned long ltls;
+	unsigned long rtls;
+};
+
+/* Used to wake-up a Linux thread futex_wait()-ing */
+struct uti_futex_resp {
+	int done;
+	wait_queue_head_t wq;
+};
 #endif

executer/kernel/mcctrl/procfs.c

@@ -1110,10 +1110,10 @@ static const struct procfs_entry pid_entry_stuff[] = {
 //	PROC_LNK("exe",        mckernel_readlink),
 //	PROC_REG("limits",     S_IRUSR|S_IWUSR, NULL),
 	PROC_REG("maps",       0444, &mckernel_buff_io),
-	PROC_REG("mem",        0400, NULL),
+	PROC_REG("mem",        0600, NULL),
 	PROC_REG("pagemap",    0444, NULL),
 //	PROC_REG("smaps",      S_IRUGO, NULL),
-//	PROC_REG("stat",       0444, &mckernel_buff_io),
+	PROC_REG("stat",       0444, &mckernel_buff_io),
 //	PROC_REG("statm",      S_IRUGO, NULL),
 	PROC_REG("status",     0444, &mckernel_buff_io),
 //	PROC_REG("syscall",    S_IRUGO, NULL),

executer/kernel/mcctrl/syscall.c

@@ -43,6 +43,8 @@
 #include <linux/semaphore.h>
 #include <linux/spinlock.h>
 #include <linux/mount.h>
+#include <linux/kdev_t.h>
+#include <linux/hugetlb.h>
 #include <asm/uaccess.h>
 #include <asm/delay.h>
 #include <asm/io.h>
@@ -178,8 +180,8 @@ int mcctrl_add_per_thread_data(struct mcctrl_per_proc_data *ppd, void *data)
 	return ret;
 }
 
-#ifndef POSTK_DEBUG_ARCH_DEP_56 /* Strange how to use inline declaration fix. */
-struct mcctrl_per_thread_data *mcctrl_get_per_thread_data(struct mcctrl_per_proc_data *ppd, struct task_struct *task)
+struct mcctrl_per_thread_data *mcctrl_get_per_thread_data(struct mcctrl_per_proc_data *ppd,
+							  struct task_struct *task)
 {
 	struct mcctrl_per_thread_data *ptd_iter, *ptd = NULL;
 	int hash = (((uint64_t)task >> 4) & MCCTRL_PER_THREAD_DATA_HASH_MASK);
@@ -208,7 +210,6 @@ struct mcctrl_per_thread_data *mcctrl_get_per_thread_data(struct mcctrl_per_proc
 	read_unlock_irqrestore(&ppd->per_thread_data_hash_lock[hash], flags);
 	return ptd;
 }
-#endif /* !POSTK_DEBUG_ARCH_DEP_56 */
 
 static int __notify_syscall_requester(ihk_os_t os, struct ikc_scd_packet *packet,
 		struct syscall_response *res)
@@ -227,19 +228,19 @@ static int __notify_syscall_requester(ihk_os_t os, struct ikc_scd_packet *packet
 	c = (usrdata->channels + packet->ref)->c;
 
 	/* If spinning, no need for IKC message */
-	if (__sync_bool_compare_and_swap(&res->req_thread_status,
+	if (cmpxchg(&res->req_thread_status,
 				IHK_SCD_REQ_THREAD_SPINNING,
-				IHK_SCD_REQ_THREAD_TO_BE_WOKEN)) {
+				IHK_SCD_REQ_THREAD_TO_BE_WOKEN) ==
+			IHK_SCD_REQ_THREAD_SPINNING) {
 		dprintk("%s: no need to send IKC message for PID %d\n",
-			   __FUNCTION__, packet->pid);
+				__FUNCTION__, packet->pid);
 		return ret;
 	}
 
 	/* Wait until the status goes back to IHK_SCD_REQ_THREAD_SPINNING or
 	   IHK_SCD_REQ_THREAD_DESCHEDULED because two wake-up attempts are competing.
-	   Note that mcexec_terminate_thread() and remote page fault and
-	   returning EINTR would compete. */
-	if (res->req_thread_status == IHK_SCD_REQ_THREAD_TO_BE_WOKEN) {
+	   Note that mcexec_terminate_thread() and returning EINTR would compete. */
+	if (smp_load_acquire(&res->req_thread_status) == IHK_SCD_REQ_THREAD_TO_BE_WOKEN) {
 		printk("%s: INFO: someone else is waking up the McKernel thread, "
 				"pid: %d, req status: %lu, syscall nr: %lu\n",
 				__FUNCTION__, packet->pid,
@@ -247,9 +248,10 @@ static int __notify_syscall_requester(ihk_os_t os, struct ikc_scd_packet *packet
 	}
 
 	/* The thread is not spinning any more, make sure it's descheduled */
-	if (!__sync_bool_compare_and_swap(&res->req_thread_status,
+	if (cmpxchg(&res->req_thread_status,
 				IHK_SCD_REQ_THREAD_DESCHEDULED,
-				IHK_SCD_REQ_THREAD_TO_BE_WOKEN)) {
+				IHK_SCD_REQ_THREAD_TO_BE_WOKEN) !=
+			IHK_SCD_REQ_THREAD_DESCHEDULED) {
 		printk("%s: WARNING: inconsistent requester status, "
 				"pid: %d, req status: %lu, syscall nr: %lu\n",
 				__FUNCTION__, packet->pid,
@@ -273,7 +275,7 @@ long syscall_backward(struct mcctrl_usrdata *usrdata, int num,
                       unsigned long *ret)
 {
 	struct ikc_scd_packet *packet;
-    struct ikc_scd_packet *free_packet = NULL;
+	struct ikc_scd_packet *free_packet = NULL;
 	struct syscall_request *req;
 	struct syscall_response *resp;
 	unsigned long syscall_ret;
@@ -282,15 +284,16 @@ long syscall_backward(struct mcctrl_usrdata *usrdata, int num,
 	struct mcctrl_per_proc_data *ppd;
 	struct mcctrl_per_thread_data *ptd;
 	unsigned long phys;
-	struct syscall_request _request[2];
-	struct syscall_request *request;
+	struct syscall_request *request = NULL;
 	int retry;
 
-	if (((unsigned long)_request ^ (unsigned long)(_request + 1)) &
-	    ~(PAGE_SIZE -1))
-		request = _request + 1;
-	else
-		request = _request;
+	request = kmalloc(sizeof(struct syscall_request), GFP_ATOMIC);
+	if (!request) {
+		printk("%s: ERROR: allocating request\n", __func__);
+		syscall_ret = -ENOMEM;
+		goto no_ppd;
+	}
+
 	request->number = num;
 	request->args[0] = arg1;
 	request->args[1] = arg2;
@@ -305,8 +308,9 @@ long syscall_backward(struct mcctrl_usrdata *usrdata, int num,
 
 	if (!ppd) {
 		kprintf("%s: ERROR: no per-process structure for PID %d??\n", 
-				__FUNCTION__, task_tgid_vnr(current));
-		return -EINVAL;
+			__func__, task_tgid_vnr(current));
+		syscall_ret = -EINVAL;
+		goto no_ppd;
 	}
 
 	ptd = mcctrl_get_per_thread_data(ppd, current);
@@ -454,11 +458,13 @@ out:
 out_put_ppd:
 	mcctrl_put_per_thread_data(ptd);
 	pr_ptd("put", task_pid_vnr(current), ptd);
- no_ptd:
+no_ptd:
 	dprintk("%s: tid: %d, syscall: %d, syscall_ret: %lx\n",
 		__FUNCTION__, task_pid_vnr(current), num, syscall_ret);
 
 	mcctrl_put_per_proc_data(ppd);
+no_ppd:
+	kfree(request);
 	return syscall_ret;
 }
 
@@ -479,214 +485,43 @@ extern struct host_thread *host_threads;
 extern rwlock_t host_thread_lock;
 #endif
 
-int remote_page_fault(struct mcctrl_usrdata *usrdata, void *fault_addr, uint64_t reason)
+int remote_page_fault(struct mcctrl_usrdata *usrdata, void *fault_addr,
+		      uint64_t reason, struct mcctrl_per_proc_data *ppd,
+		      struct ikc_scd_packet *packet)
 {
-	struct ikc_scd_packet *packet;
-	struct ikc_scd_packet *free_packet = NULL;
-	struct syscall_request *req;
-	struct syscall_response *resp;
 	int error;
-	struct wait_queue_head_list_node *wqhln;
-	unsigned long irqflags;
-	struct mcctrl_per_proc_data *ppd;
-	struct mcctrl_per_thread_data *ptd;
-	unsigned long phys;
-	int retry;
+	struct mcctrl_wakeup_desc *desc;
+	int do_frees = 1;
 	
 	dprintk("%s: tid: %d, fault_addr: %p, reason: %lu\n",
 			__FUNCTION__, task_pid_vnr(current), fault_addr, (unsigned long)reason);
-	
-	/* Look up per-process structure */
-	ppd = mcctrl_get_per_proc_data(usrdata, task_tgid_vnr(current));
-
-	if (!ppd) {
-		kprintf("%s: ERROR: no per-process structure for PID %d??\n", 
-				__FUNCTION__, task_tgid_vnr(current));
-		return -EINVAL;
-	}
-
-	ptd = mcctrl_get_per_thread_data(ppd, current);
-	if (!ptd) {
-		printk("%s: ERROR: mcctrl_get_per_thread_data failed\n", __FUNCTION__);		
-		error = -ENOENT;
-		goto no_ptd;
-	}
-	pr_ptd("get", task_pid_vnr(current), ptd);
-	packet = (struct ikc_scd_packet *)ptd->data;
-	if (!packet) {
-		printk("%s: no packet registered for TID %d\n",
-				__FUNCTION__, task_pid_vnr(current));
-		error = -ENOENT;
-		goto out_put_ppd;
-	}
-
-	req = &packet->req;
-
-	/* Map response structure */
-	phys = ihk_device_map_memory(ihk_os_to_dev(usrdata->os), 
-			packet->resp_pa, sizeof(*resp));
-	resp = ihk_device_map_virtual(ihk_os_to_dev(usrdata->os), 
-			phys, sizeof(*resp), NULL, 0);
-	if (!resp) {
-		printk("%s: ERROR: invalid response structure address\n",
-			__FUNCTION__);
-		error = -EINVAL;
-		goto out;
-	}
-
-retry_alloc:
-	wqhln = kmalloc(sizeof(*wqhln), GFP_ATOMIC);
-	if (!wqhln) {
-		printk("WARNING: coudln't alloc wait queue head, retrying..\n");
-		goto retry_alloc;
-	}
-	memset(wqhln, 0, sizeof(struct wait_queue_head_list_node));
-
-	/* Prepare per-thread wait queue head */
-	wqhln->task = current;
-	/* Save the TID explicitly, because mcexec_syscall(), where the request
-	 * will be matched, is in IRQ context and can't call task_pid_vnr() */
-	wqhln->rtid = task_pid_vnr(current);
-	wqhln->req = 0;
-	init_waitqueue_head(&wqhln->wq_syscall);
-
-	irqflags = ihk_ikc_spinlock_lock(&ppd->wq_list_lock);
-	/* Add to exact list */
-	list_add_tail(&wqhln->list, &ppd->wq_list_exact);
-	ihk_ikc_spinlock_unlock(&ppd->wq_list_lock, irqflags);
 
 	/* Request page fault */
-	resp->ret = -EFAULT;
-	resp->fault_address = (unsigned long)fault_addr;
-	resp->fault_reason = reason;
-	resp->stid = task_pid_vnr(current);
+	packet->msg = SCD_MSG_REMOTE_PAGE_FAULT;
+	packet->fault_address = (unsigned long)fault_addr;
+	packet->fault_reason = reason;
 
-#define STATUS_PAGER_COMPLETED	1
-#define	STATUS_PAGE_FAULT	3
-	req->valid = 0;
-
-	if (__notify_syscall_requester(usrdata->os, packet, resp) < 0) {
-		printk("%s: WARNING: failed to notify PID %d\n",
-			__FUNCTION__, packet->pid);
+	/* we need to alloc desc ourselves because GFP_ATOMIC */
+retry_alloc:
+	desc = kmalloc(sizeof(*desc), GFP_ATOMIC);
+	if (!desc) {
+		pr_warn("WARNING: coudln't alloc remote page fault wait desc, retrying..\n");
+		goto retry_alloc;
 	}
 
-	mb();
-	resp->status = STATUS_PAGE_FAULT;
-
-	retry = 0;
-	for (;;) {
-		dprintk("%s: tid: %d, fault_addr: %p SLEEPING\n", 
-				__FUNCTION__, task_pid_vnr(current), fault_addr);
-		/* wait for response */
-		error = wait_event_interruptible(wqhln->wq_syscall, wqhln->req);
-
-		/* Delay signal handling */
-		if (error == -ERESTARTSYS) {
-			printk("%s: INFO: interrupted by signal\n", __FUNCTION__);
-			retry++;
-			if (retry < 5) { /* mcexec is alive */
-				printk("%s: INFO: retry=%d\n", __FUNCTION__, retry);
-				continue;
-			}
-		}
-
-		/* Remove per-thread wait queue head */
-		irqflags = ihk_ikc_spinlock_lock(&ppd->wq_list_lock);
-		list_del(&wqhln->list);
-		ihk_ikc_spinlock_unlock(&ppd->wq_list_lock, irqflags);
-
-		dprintk("%s: tid: %d, fault_addr: %p WOKEN UP\n", 
-				__FUNCTION__, task_pid_vnr(current), fault_addr);
-
-		if (retry >= 5) {
-			kfree(wqhln);
-			kprintf("%s: INFO: mcexec is gone or retry count exceeded,pid=%d,retry=%d\n", __FUNCTION__, task_tgid_vnr(current), retry);
-			error = -EINVAL;
-			goto out;
-		}
-
-		if (error) {
-			kfree(wqhln);
-			printk("remote_page_fault:interrupted. %d\n", error);
-			goto out;
-		}
-		else {
-			/* Update packet reference */
-			packet = wqhln->packet;
-			free_packet = packet;
-			req = &packet->req;
-			{
-				unsigned long phys2;
-				struct syscall_response *resp2;
-				phys2 = ihk_device_map_memory(ihk_os_to_dev(usrdata->os), 
-						packet->resp_pa, sizeof(*resp));
-				resp2 = ihk_device_map_virtual(ihk_os_to_dev(usrdata->os), 
-						phys2, sizeof(*resp), NULL, 0);
-
-				if (resp != resp2) {
-					resp = resp2;
-					phys = phys2;
-					printk("%s: updated new remote PA for resp\n", __FUNCTION__);
-				}
-			}
-		}
-
-		if (!req->valid) {
-			printk("remote_page_fault:not valid\n");
-		}
-		req->valid = 0;
-
-		/* check result */
-		if (req->number != __NR_mmap) {
-			printk("remote_page_fault:unexpected response. %lx %lx\n",
-					req->number, req->args[0]);
-			error = -EIO;
-			goto out;
-		}
-#define	PAGER_REQ_RESUME	0x0101
-		else if (req->args[0] != PAGER_REQ_RESUME) {
-			resp->ret = pager_call(usrdata->os, (void *)req);
-
-			if (__notify_syscall_requester(usrdata->os, packet, resp) < 0) {
-				printk("%s: WARNING: failed to notify PID %d\n",
-						__FUNCTION__, packet->pid);
-			}
-
-			mb();
-			resp->status = STATUS_PAGER_COMPLETED;
-			break;
-			//continue;
-		}
-		else {
-			error = req->args[1];
-			if (error) {
-				printk("remote_page_fault:response %d\n", error);
-				kfree(wqhln);
-				goto out;
-			}
-		}
-		break;
+	/* packet->target_cpu was set in rus_vm_fault if a thread was found */
+	error = mcctrl_ikc_send_wait(usrdata->os, packet->target_cpu, packet,
+				     0, desc, &do_frees, 0);
+	if (do_frees)
+		kfree(desc);
+	if (error < 0) {
+		pr_warn("%s: WARNING: failed to request remote page fault PID %d: %d\n",
+			__func__, packet->pid, error);
 	}
 
-	kfree(wqhln);
-	error = 0;
-out:
-	/* Release remote page-fault response packet */
-	if (free_packet) {
-		ihk_ikc_release_packet((struct ihk_ikc_free_packet *)free_packet);
-	}
-
-	ihk_device_unmap_virtual(ihk_os_to_dev(usrdata->os), resp, sizeof(*resp));
-	ihk_device_unmap_memory(ihk_os_to_dev(usrdata->os), phys, sizeof(*resp));
-
-out_put_ppd:
-	mcctrl_put_per_thread_data(ptd);
-	pr_ptd("put", task_pid_vnr(current), ptd);
- no_ptd:
 	dprintk("%s: tid: %d, fault_addr: %p, reason: %lu, error: %d\n",
-			__FUNCTION__, task_pid_vnr(current), fault_addr, (unsigned long)reason, error);
-
-	mcctrl_put_per_proc_data(ppd);
+		__func__, task_pid_vnr(current), fault_addr,
+		(unsigned long)reason, error);
 	return error;
 }
 
@@ -704,7 +539,11 @@ out_put_ppd:
 #define	USE_VM_INSERT_PFN	1
 
 #if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 11, 0)
+#if defined(RHEL_RELEASE_CODE) && RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(8, 2)
+static vm_fault_t rus_vm_fault(struct vm_fault *vmf)
+#else
 static int rus_vm_fault(struct vm_fault *vmf)
+#endif
 {
 	struct vm_area_struct *vma = vmf->vma;
 #else
@@ -726,70 +565,70 @@ static int rus_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 #endif
 	struct mcctrl_per_proc_data *ppd;
 	struct mcctrl_per_thread_data *ptd;
-	struct ikc_scd_packet *packet;
+	struct task_struct *task = current;
+	struct ikc_scd_packet packet = { };
+	unsigned long rsysnum = 0;
 	int ret = 0;
-
 #if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
-	dprintk("mcctrl:page fault:flags %#x pgoff %#lx va %#lx page %p\n",
-			vmf->flags, vmf->pgoff, vmf->address, vmf->page);
+	unsigned long addr = vmf->address;
 #else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-	dprintk("mcctrl:page fault:flags %#x pgoff %#lx va %p page %p\n",
-			vmf->flags, vmf->pgoff, vmf->virtual_address, vmf->page);
+	void __user *addr = vmf->virtual_address;
 #endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
 
 	/* Look up per-process structure */
-	ppd = mcctrl_get_per_proc_data(usrdata, task_tgid_vnr(current));
+	ppd = mcctrl_get_per_proc_data(usrdata, task_tgid_vnr(task));
 	if (!ppd) {
-		kprintf("%s: INFO: no per-process structure for pid %d (tid %d), try to use pid %d\n", 
-				__FUNCTION__, task_tgid_vnr(current), task_pid_vnr(current), vma->vm_mm->owner->pid);
-		ppd = mcctrl_get_per_proc_data(usrdata, vma->vm_mm->owner->pid);
+		pr_err("%s: INFO: no per-process structure for "
+				"pid %d (tid %d), trying to use pid %d\n",
+				__func__,
+				task_tgid_vnr(task), task_pid_vnr(task),
+				vma->vm_mm->owner->pid);
+		task = vma->vm_mm->owner;
+		ppd = mcctrl_get_per_proc_data(usrdata, task_tgid_vnr(task));
 	}
 
 	if (!ppd) {
-		kprintf("%s: ERROR: no per-process structure for PID %d??\n", 
-				__FUNCTION__, task_tgid_vnr(current));
+		pr_err("%s: ERROR: no per-process structure for PID %d??\n",
+				__func__, task_tgid_vnr(task));
 		ret = VM_FAULT_SIGBUS;
 		goto no_ppd;
 	}
+	packet.fault_tid = ppd->pid;
 
-	ptd = mcctrl_get_per_thread_data(ppd, current);
-	if (!ptd) {
-		printk("%s: ERROR: mcctrl_get_per_thread_data failed\n", __FUNCTION__);
-		ret = VM_FAULT_SIGBUS;
-		goto no_ptd;
+	ptd = mcctrl_get_per_thread_data(ppd, task);
+	if (ptd) {
+		struct ikc_scd_packet *ptd_packet;
+
+		pr_ptd("get", task_pid_vnr(task), ptd);
+		ptd_packet = (struct ikc_scd_packet *)ptd->data;
+		if (ptd_packet) {
+			packet.target_cpu = ptd_packet->ref;
+			packet.fault_tid = ptd_packet->req.rtid;
+			rsysnum = ptd_packet->req.number;
+		}
+		mcctrl_put_per_thread_data(ptd);
+		pr_ptd("put", task_pid_vnr(task), ptd);
 	}
-	pr_ptd("get", task_pid_vnr(current), ptd);
-	packet = (struct ikc_scd_packet *)ptd->data;
-	if (!packet) {
+
+	/* Don't even bother looking up NULL */
+	if (!addr) {
+		pr_warn("%s: WARNING: attempted NULL pointer access\n",
+				__func__);
 		ret = VM_FAULT_SIGBUS;
-		printk("%s: no packet registered for TID %d\n",
-				__FUNCTION__, task_pid_vnr(current));
 		goto put_and_out;
 	}
 
 	for (try = 1; ; ++try) {
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
 		error = translate_rva_to_rpa(usrdata->os, ppd->rpgtable,
-				vmf->address, &rpa, &pgsize);
-#else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-		error = translate_rva_to_rpa(usrdata->os, ppd->rpgtable,
-				(unsigned long)vmf->virtual_address,
-				&rpa, &pgsize);
-#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
+				(unsigned long)addr, &rpa, &pgsize);
 #define	NTRIES 2
 		if (!error || (try >= NTRIES)) {
 			if (error) {
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
-				printk("%s: error translating 0x%#lx "
-						"(req: TID: %u, syscall: %lu)\n",
-						__FUNCTION__, vmf->address,
-						packet->req.rtid, packet->req.number);
-#else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-				printk("%s: error translating 0x%p "
-						"(req: TID: %u, syscall: %lu)\n",
-						__FUNCTION__, vmf->virtual_address,
-						packet->req.rtid, packet->req.number);
-#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
+				pr_err("%s: error translating 0x%#lx "
+					"(req: TID: %u, syscall: %lu)\n",
+					__func__,
+					(unsigned long)addr,
+					packet.fault_tid, rsysnum);
 			}
 
 			break;
@@ -800,23 +639,14 @@ static int rus_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 #define	PF_WRITE	0x02
 			reason |= PF_WRITE;
 		}
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
-		error = remote_page_fault(usrdata, (void *)vmf->address, reason);
-#else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-		error = remote_page_fault(usrdata, vmf->virtual_address, reason);
-#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
+		error = remote_page_fault(usrdata, (void *)addr,
+					  reason, ppd, &packet);
 		if (error) {
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
-				printk("%s: error forwarding PF for 0x%#lx "
-						"(req: TID: %d, syscall: %lu)\n",
-						__FUNCTION__, vmf->address,
-						packet->req.rtid, packet->req.number);
-#else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-				printk("%s: error forwarding PF for 0x%p "
-						"(req: TID: %d, syscall: %lu)\n",
-						__FUNCTION__, vmf->virtual_address,
-						packet->req.rtid, packet->req.number);
-#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
+			pr_err("%s: error forwarding PF for 0x%#lx "
+					"(req: TID: %d, syscall: %lu)\n",
+					__func__,
+					(unsigned long)addr,
+					packet.fault_tid, rsysnum);
 			break;
 		}
 	}
@@ -825,11 +655,7 @@ static int rus_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 		goto put_and_out;
 	}
 
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
-	rva = vmf->address & ~(pgsize - 1);
-#else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-	rva = (unsigned long)vmf->virtual_address & ~(pgsize - 1);
-#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
+	rva = (unsigned long)addr & ~(pgsize - 1);
 	rpa = rpa & ~(pgsize - 1);
 
 	phys = ihk_device_map_memory(dev, rpa, pgsize);
@@ -849,21 +675,13 @@ static int rus_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 
 			error = vm_insert_page(vma, rva+(pix*PAGE_SIZE), page);
 			if (error) {
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
-				printk("%s: error inserting mapping for 0x%#lx "
-						"(req: TID: %d, syscall: %lu) error: %d, " 
-						"vm_start: 0x%lx, vm_end: 0x%lx\n",
-						__FUNCTION__, vmf->address,
-						packet->req.rtid, packet->req.number, error,
-						vma->vm_start, vma->vm_end);
-#else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-				printk("%s: error inserting mapping for 0x%p "
-						"(req: TID: %d, syscall: %lu) error: %d, " 
-						"vm_start: 0x%lx, vm_end: 0x%lx\n",
-						__FUNCTION__, vmf->virtual_address,
-						packet->req.rtid, packet->req.number, error,
-						vma->vm_start, vma->vm_end);
-#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
+				pr_err("%s: error inserting mapping for 0x%#lx "
+					"(req: TID: %d, syscall: %lu) error: %d,"
+					" vm_start: 0x%lx, vm_end: 0x%lx\n",
+					__func__,
+					(unsigned long)addr, packet.fault_tid,
+					rsysnum, error,
+					vma->vm_start, vma->vm_end);
 			}
 		}
 		else
@@ -875,16 +693,13 @@ static int rus_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 					      pfn+pix);
 #endif
 		if (error) {
-#if 1 /* POSTK_DEBUG_TEMP_FIX_11 */ /* rus_vm_fault() multi-thread fix */
-			printk("%s: vm_insert_pfn returned %d\n", __FUNCTION__, error);
+			pr_err("%s: vm_insert_pfn returned %d\n",
+			       __func__, error);
 			if (error == -EBUSY) {
 				error = 0;
 			} else {
 				break;
 			}
-#else /* POSTK_DEBUG_TEMP_FIX_11 */
-			break;
-#endif /* POSTK_DEBUG_TEMP_FIX_11 */
 		}
 	}
 #else
@@ -892,17 +707,11 @@ static int rus_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 #endif
 	ihk_device_unmap_memory(dev, phys, pgsize);
 	if (error) {
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0)
-		printk("%s: remote PF failed for 0x%#lx, pgoff: %lu "
-				"(req: TID: %d, syscall: %lu)\n",
-				__FUNCTION__, vmf->address, vmf->pgoff,
-				packet->req.rtid, packet->req.number);
-#else /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
-		printk("%s: remote PF failed for 0x%p, pgoff: %lu "
-				"(req: TID: %d, syscall: %lu)\n",
-				__FUNCTION__, vmf->virtual_address, vmf->pgoff,
-				packet->req.rtid, packet->req.number);
-#endif /* LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) */
+		pr_err("%s: remote PF failed for 0x%#lx, pgoff: %lu"
+				" (req: TID: %d, syscall: %lu)\n",
+				__func__,
+				(unsigned long)addr, vmf->pgoff,
+				packet.fault_tid, rsysnum);
 		ret = VM_FAULT_SIGBUS;
 		goto put_and_out;
 	}
@@ -910,9 +719,6 @@ static int rus_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	ret = VM_FAULT_NOPAGE;
 
 put_and_out:
-	mcctrl_put_per_thread_data(ptd);
-	pr_ptd("put", task_pid_vnr(current), ptd);
- no_ptd:
 	mcctrl_put_per_proc_data(ppd);
  no_ppd:
 	return ret;
@@ -1075,6 +881,7 @@ struct pager_create_result {
 	int		maxprot;
 	uint32_t flags;
 	size_t size;
+	int pgshift;
 	char path[PATH_MAX];
 };
 
@@ -1136,6 +943,7 @@ static int pager_req_create(ihk_os_t os, int fd, uintptr_t result_pa)
 	struct kstat st;
 	int mf_flags = 0;
 	unsigned long irqflags;
+	int pgshift = 0;
 
 	dprintk("pager_req_create(%d,%lx)\n", fd, (long)result_pa);
 
@@ -1144,8 +952,16 @@ static int pager_req_create(ihk_os_t os, int fd, uintptr_t result_pa)
 		printk("pager_req_create(%d,%lx):vfs_stat failed. %d\n", fd, (long)result_pa, error);
 		goto out;
 	}
-	if (S_ISCHR(st.mode) && (MAJOR(st.rdev) == 1)) {
-		/* treat memory devices as regular files */
+	if (S_ISCHR(st.mode) && (MAJOR(st.rdev) == 1) &&
+			(MINOR(st.rdev) == 1 ||   // /dev/mem
+			 MINOR(st.rdev) == 5)) {  // /dev/zero
+		/* treat memory devices and zero devices as regular files */
+	}
+	else if (S_ISCHR(st.mode) && (MAJOR(st.rdev) == 1)) {
+		error = -ENODEV;
+		dprintk("%s(%d,%lx):unmappable device %x\n",
+				__func__, fd, (long)result_pa, st.mode);
+		goto out;
 	}
 	else if (!S_ISREG(st.mode)) {
 		error = -ESRCH;
@@ -1160,6 +976,30 @@ static int pager_req_create(ihk_os_t os, int fd, uintptr_t result_pa)
 		goto out;
 	}
 
+	/* Shared memory hack */
+	{
+		char *pathbuf, *fullpath;
+		pathbuf = kmalloc(PATH_MAX, GFP_ATOMIC);
+		if (pathbuf) {
+			fullpath = d_path(&file->f_path, pathbuf, PATH_MAX);
+			if (!IS_ERR(fullpath)) {
+				if (!strncmp("/tmp/ompi.", fullpath, 10) ||
+						!strncmp("/dev/shm/", fullpath, 9) ||
+						(!strncmp("/var/opt/FJSVtcs/ple/daemonif/",
+							fullpath, 30) && !strstr(fullpath, "dstore_sm.lock"))) {
+					printk("%s: treating %s as a device file..\n",
+						__func__, fullpath);
+					kfree(pathbuf);
+
+					error = -ESRCH;
+					goto out;
+				}
+
+				kfree(pathbuf);
+			}
+		}
+	}
+
 	inode = file->f_path.dentry->d_inode;
 	if (!inode) {
 		error = -EBADF;
@@ -1167,6 +1007,10 @@ static int pager_req_create(ihk_os_t os, int fd, uintptr_t result_pa)
 		goto out;
 	}
 
+	if (!strcmp(inode->i_sb->s_type->name, "tmpfs")) {
+		mf_flags = MF_IS_REMOVABLE;
+	}
+
 	if (!strcmp(inode->i_sb->s_type->name, "proc")) {
 		error = -ESRCH;
 		goto out;
@@ -1188,13 +1032,14 @@ static int pager_req_create(ihk_os_t os, int fd, uintptr_t result_pa)
 	}
 
 	if (inode->i_op == mcctrl_hugetlbfs_inode_operations) {
+		struct hstate *h = hstate_file(file);
+
+		pgshift = PAGE_SHIFT + huge_page_order(h);
 		mf_flags = MF_HUGETLBFS;
 		/* pager is used as handle id on mckernel side, use inode */
 		pager = (void *)st.ino;
-		/* retrofit blksize in resp as well through st.size field;
-		 * the actual file size is not used
-		 */
-		st.size = st.blksize;
+		/* file size is not used */
+		st.size = 0;
 		goto out_reply;
 	}
 
@@ -1214,7 +1059,7 @@ static int pager_req_create(ihk_os_t os, int fd, uintptr_t result_pa)
 			pager = newpager;
 			newpager = NULL;
 
-			/* Intel MPI library and shared memory "prefetch" */
+			/* Shared libraries prefetch */
 			{
 				char *pathbuf, *fullpath;
 
@@ -1222,15 +1067,7 @@ static int pager_req_create(ihk_os_t os, int fd, uintptr_t result_pa)
 				if (pathbuf) {
 					fullpath = d_path(&file->f_path, pathbuf, PATH_MAX);
 					if (!IS_ERR(fullpath)) {
-						if (!strncmp("/dev/shm/Intel_MPI", fullpath, 18)) {
-							mf_flags = (MF_PREMAP | MF_ZEROFILL);
-							dprintk("%s: filename: %s, premap & zerofill\n",
-									__FUNCTION__, fullpath);
-						}
-						else if (strstr(fullpath, "libmpi") ||
-								strstr(fullpath, "libiomp") ||
-								strstr(fullpath, "libpthread") ||
-								strstr(fullpath, "libc.so")) {
+						if (strstr(fullpath, ".so")) {
 							mf_flags = MF_PREFETCH;
 							dprintk("%s: filename: %s, prefetch\n",
 									__FUNCTION__, fullpath);
@@ -1281,6 +1118,7 @@ out_reply:
 	resp->maxprot = maxprot;
 	resp->flags = mf_flags;
 	resp->size = st.size;
+	resp->pgshift = pgshift;
 
 	error = pager_get_path(file, resp->path);
 
@@ -1355,8 +1193,9 @@ static int pager_req_read(ihk_os_t os, uintptr_t handle, off_t off, size_t size,
 	uintptr_t phys = -1;
 	ihk_device_t dev = ihk_os_to_dev(os);
 	void *buf = NULL;
-	loff_t pos;
+	loff_t pos, fsize;
 	unsigned long flags;
+	unsigned int major, minor;
 
 	dprintk("pager_req_read(%lx,%lx,%lx,%lx)\n", handle, off, size, rpa);
 
@@ -1378,6 +1217,21 @@ static int pager_req_read(ihk_os_t os, uintptr_t handle, off_t off, size_t size,
 		goto out;
 	}
 
+	major = MAJOR(file->f_mapping->host->i_rdev);
+	minor = MINOR(file->f_mapping->host->i_rdev);
+	if ((major == 1 && minor == 1) || // /dev/mem
+		(major == 1 && minor == 5)) { // /dev/zero
+		/* Nothing to check */
+	}
+	else {
+		/* Check if the target page fits in the file */
+		fsize = i_size_read(file->f_mapping->host);
+		if (off > fsize) {
+			ss = 0;
+			goto out;
+		}
+	}
+
 	phys = ihk_device_map_memory(dev, rpa, size);
 	buf = ihk_device_map_virtual(dev, phys, size, NULL, 0);
 	if (!buf) {
@@ -1585,6 +1439,26 @@ static int pager_req_map(ihk_os_t os, int fd, size_t len, off_t off,
 #define	ANY_WHERE 0
 	if (prot_and_flags & MAP_LOCKED) prot_and_flags |= MAP_POPULATE;
 
+	/* Shared memory hack */
+	{
+		char *pathbuf, *fullpath;
+		pathbuf = kmalloc(PATH_MAX, GFP_ATOMIC);
+		if (pathbuf) {
+			fullpath = d_path(&file->f_path, pathbuf, PATH_MAX);
+			if (!IS_ERR(fullpath)) {
+				if (!strncmp("/tmp/ompi.", fullpath, 10) ||
+						!strncmp("/dev/shm/", fullpath, 9) ||
+						!strncmp("/var/opt/FJSVtcs/ple/daemonif/",
+							fullpath, 30)) {
+					dprintk("%s: pre-populating %s..\n",
+						__func__, fullpath);
+					prot_and_flags |= MAP_POPULATE;
+				}
+				kfree(pathbuf);
+			}
+		}
+	}
+
 #if LINUX_VERSION_CODE < KERNEL_VERSION(3,5,0)
 	down_write(&current->mm->mmap_sem);
 
@@ -1598,7 +1472,12 @@ static int pager_req_map(ihk_os_t os, int fd, size_t len, off_t off,
 #endif
 
 	if (IS_ERR_VALUE(va)) {
-		printk("pager_req_map(%p,%d,%lx,%lx,%lx):do_mmap_pgoff failed. %d\n", os, fd, len, off, result_rpa, (int)va);
+		if ((int)va != -ENOTSUPP) {
+			pr_err("%s(%p,%d,%lx,%lx,%lx): "
+			       "do_mmap_pgoff failed. %d\n",
+			       __func__, os, fd, len, off, result_rpa,
+			       (int)va);
+		}
 		error = va;
 		goto out;
 	}
@@ -1712,16 +1591,9 @@ retry:
 						pfn |= PFN_VALID | PFN_PRESENT;
 
 						/* Check if mapping is write-combined */
-#ifdef POSTK_DEBUG_ARCH_DEP_12
 						if (pte_is_write_combined(*pte)) {
 							pfn |= PFN_WRITE_COMBINED;
 						}
-#else /* POSTK_DEBUG_ARCH_DEP_12 */
-						if ((pte_flags(*pte) & _PAGE_PWT) &&
-						    !(pte_flags(*pte) & _PAGE_PCD)) {
-							pfn |= _PAGE_PWT;
-						}
-#endif /* POSTK_DEBUG_ARCH_DEP_12 */
 					}
 					pte_unmap(pte);
 				}
@@ -1754,10 +1626,27 @@ retry:
 #else
 		fault = handle_mm_fault(current->mm, vma, va, flags);
 #endif
+#ifdef SC_DEBUG
 		if (fault != 0) {
-			printk("%s: error: faulting %lx at off: %lu\n", 
-					__FUNCTION__, va, off);
+			char *pathbuf = NULL;
+			char *fullpath;
+
+			if (vma->vm_file) {
+				pathbuf = kmalloc(PATH_MAX, GFP_ATOMIC);
+				if (pathbuf) {
+					fullpath = d_path(&vma->vm_file->f_path,
+							pathbuf, PATH_MAX);
+					if (!IS_ERR(fullpath)) {
+						printk("%s: WARNING: couldn't fault 0x%lx"
+								" at off: %lu in %s\n",
+								__FUNCTION__, va, off, fullpath);
+					}
+
+					kfree(pathbuf);
+				}
+			}
 		}
+#endif
 
 		page_fault_attempted = 1;
 		goto retry;
@@ -2052,6 +1941,8 @@ int mcctrl_clear_pte_range(uintptr_t start, uintptr_t len)
 		}
 		if (addr < end) {
 #if LINUX_VERSION_CODE < KERNEL_VERSION(4, 18, 0)
+			/* Revert permission */
+			vma->vm_flags |= VM_READ | VM_WRITE | VM_EXEC;
 			error = zap_vma_ptes(vma, addr, end-addr);
 			if (error) {
 				mcctrl_zap_page_range(vma, addr, end-addr,
@@ -2069,6 +1960,8 @@ int mcctrl_clear_pte_range(uintptr_t start, uintptr_t len)
 						      NULL);
 			}
 			else {
+				/* Revert permission */
+				vma->vm_flags |= VM_READ | VM_WRITE | VM_EXEC;
 				zap_vma_ptes(vma, addr, end-addr);
 			}
 #endif /* LINUX_VERSION_CODE < KERNEL_VERSION(4, 18, 0) */
@@ -2124,7 +2017,10 @@ int release_user_space(uintptr_t start, uintptr_t len)
  * \param chunks The number of chunks which make a core file image in the whole.
  */
 
-static int writecore(ihk_os_t os, unsigned long rcoretable, int chunks) {
+static int writecore(ihk_os_t os, unsigned long rcoretable, int chunks,
+		     unsigned long cmdline_rphys, unsigned long cmdline_len)
+{
+	char *fn = NULL;
 	struct file *file;
 	struct coretable *coretable;
 	int i, tablesize, error = 0;
@@ -2133,22 +2029,43 @@ static int writecore(ihk_os_t os, unsigned long rcoretable, int chunks) {
 	unsigned long phys, tablephys, rphys;
 	ihk_device_t dev = ihk_os_to_dev(os);
 	char *pt;
+	unsigned long cmdline_phys;
+	char *cmdline;
 
 	dprintk("coredump called as a pseudo syscall\n");
 
+	fn = kmalloc(PATH_MAX, GFP_ATOMIC);
+	if (!fn) {
+		dprintk("%s: ERROR: allocating file name\n", __func__);
+		error = -ENOMEM;
+		goto fail;
+	}
+
 	if (chunks <= 0) {
 		dprintk("no core data found!(%d)\n", chunks);
 		error = -EINVAL;
 		goto fail;
 	}
 
+	cmdline_phys = ihk_device_map_memory(dev, cmdline_rphys, cmdline_len);
+	cmdline = ihk_device_map_virtual(dev, cmdline_phys, cmdline_len, NULL,
+					 0);
+	sprintf(fn, "mccore-%s.%d",
+		strrchr(cmdline, '/') ?
+		strrchr(cmdline, '/') + 1 : cmdline,
+		task_tgid_vnr(current));
+	pr_info("%s: fn=%s\n", __func__, fn);
+
+	ihk_device_unmap_virtual(dev, cmdline, cmdline_len);
+	ihk_device_unmap_memory(dev, cmdline_phys, cmdline_len);
+
 	/* Every Linux documentation insists we should not 
 	 * open a file in the kernel module, but our karma 
 	 * leads us here. Precisely, Here we emulate the core 
 	 * dump routine of the Linux kernel in linux/fs/exec.c. 
 	 * So we have a legitimate reason to do this.
 	 */
-	file = filp_open("core", O_CREAT | O_RDWR | O_LARGEFILE | O_TRUNC, 0600);
+	file = filp_open(fn, O_CREAT | O_RDWR | O_LARGEFILE | O_TRUNC, 0600);
 	if (IS_ERR(file) || !file->f_op) {
 		dprintk("cannot open core file\n");
 		error = PTR_ERR(file);
@@ -2218,6 +2135,7 @@ fail:
 		/* make sure we do not travel to user land */
 		error = -EINVAL;
 	}
+	kfree(fn);
 	return error;
 }
 
@@ -2273,7 +2191,8 @@ int __do_in_kernel_syscall(ihk_os_t os, struct ikc_scd_packet *packet)
 		}
 
 	case __NR_coredump:
-		ret = writecore(os, sc->args[1], sc->args[0]);
+		ret = writecore(os, sc->args[1], sc->args[0], sc->args[2],
+				sc->args[3]);
 		break;
 	
 	case __NR_sched_setparam: {

executer/kernel/mcctrl/sysfs_files.c

@@ -157,13 +157,8 @@ static void free_node_topology(struct mcctrl_usrdata *udp)
 	return;
 } /* free_node_topology() */
 
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 static void free_cpu_topology_one(struct mcctrl_usrdata *udp,
 		struct mcctrl_cpu_topology *cpu)
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-static void free_cpu_topology_one(struct mcctrl_usrdata *udp,
-		struct cpu_topology *cpu)
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 {
 	struct cache_topology *cache;
 	struct cache_topology *next;
@@ -179,13 +174,8 @@ static void free_cpu_topology_one(struct mcctrl_usrdata *udp,
 
 static void free_cpu_topology(struct mcctrl_usrdata *udp)
 {
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 	struct mcctrl_cpu_topology *cpu;
 	struct mcctrl_cpu_topology *next;
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-	struct cpu_topology *cpu;
-	struct cpu_topology *next;
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 
 	list_for_each_entry_safe(cpu, next, &udp->cpu_topology_list, chain) {
 		list_del(&cpu->chain);
@@ -315,13 +305,8 @@ static int translate_cpumap(struct mcctrl_usrdata *udp,
 	return error;
 } /* translate_cpumap() */
 
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 static struct cache_topology *get_cache_topology(struct mcctrl_usrdata *udp,
 		struct mcctrl_cpu_topology *cpu_topo, struct ihk_cache_topology *saved)
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-static struct cache_topology *get_cache_topology(struct mcctrl_usrdata *udp,
-		struct cpu_topology *cpu_topo, struct ihk_cache_topology *saved)
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 {
 	int error;
 	struct cache_topology *topo = NULL;
@@ -355,21 +340,12 @@ out:
 	return (error)? ERR_PTR(error): topo;
 } /* get_cache_topology() */
 
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 static struct mcctrl_cpu_topology *get_one_cpu_topology(struct mcctrl_usrdata *udp,
 		int index)
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-static struct cpu_topology *get_one_cpu_topology(struct mcctrl_usrdata *udp,
-		int index)
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 {
 	int error;
 	ihk_device_t dev = ihk_os_to_dev(udp->os);
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 	struct mcctrl_cpu_topology *topology = NULL;
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-	struct cpu_topology *topology = NULL;
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 	struct cache_topology *cache;
 	struct ihk_cache_topology *saved_cache;
 
@@ -387,12 +363,8 @@ static struct cpu_topology *get_one_cpu_topology(struct mcctrl_usrdata *udp,
 	topology->saved = ihk_device_get_cpu_topology(dev, 
 			mckernel_cpu_2_hw_id(udp, index));
 
-#ifdef POSTK_DEBUG_TEMP_FIX_21 /* IS_ERR() through return NULL */
 	if (!topology->saved) {
-#else /* POSTK_DEBUG_TEMP_FIX_21 */
-	if (IS_ERR(topology->saved)) {
-#endif /* POSTK_DEBUG_TEMP_FIX_21 */
-		error = PTR_ERR(topology->saved);
+		error = -ENOENT;
 		eprintk("mcctrl:get_one_cpu_topology:"
 				"ihk_device_get_cpu_topology failed. %d\n",
 				error);
@@ -428,6 +400,9 @@ static struct cpu_topology *get_one_cpu_topology(struct mcctrl_usrdata *udp,
 					"get_cache_topology failed. %d\n",
 					error);
 			goto out;
+		} else if (!cache) {
+			error = -ENOENT;
+			goto out;
 		}
 
 		list_add(&cache->chain, &topology->cache_list);
@@ -447,11 +422,7 @@ static int get_cpu_topology(struct mcctrl_usrdata *udp)
 {
 	int error;
 	int index;
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 	struct mcctrl_cpu_topology *topology;
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-	struct cpu_topology *topology;
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 
 	dprintk("get_cpu_topology(%p)\n", udp);
 	for (index = 0; index < udp->cpu_info->n_cpus; ++index) {
@@ -473,13 +444,8 @@ out:
 	return error;
 } /* get_cpu_topology() */
 
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 static void setup_cpu_sysfs_cache_files(struct mcctrl_usrdata *udp,
 		struct mcctrl_cpu_topology *cpu, struct cache_topology *cache)
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-static void setup_cpu_sysfs_cache_files(struct mcctrl_usrdata *udp,
-		struct cpu_topology *cpu, struct cache_topology *cache)
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 {
 	char *prefix = "/sys/devices/system/cpu";
 	int cpu_number = cpu->mckernel_cpu_id;
@@ -531,13 +497,8 @@ static void setup_cpu_sysfs_cache_files(struct mcctrl_usrdata *udp,
 	return;
 } /* setup_cpu_sysfs_cache_files() */
 
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 static void setup_cpu_sysfs_files(struct mcctrl_usrdata *udp,
 		struct mcctrl_cpu_topology *cpu)
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-static void setup_cpu_sysfs_files(struct mcctrl_usrdata *udp,
-		struct cpu_topology *cpu)
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 {
 	char *prefix = "/sys/devices/system/cpu";
 	int cpu_number = cpu->mckernel_cpu_id;
@@ -586,11 +547,7 @@ static void setup_cpu_sysfs_files(struct mcctrl_usrdata *udp,
 static void setup_cpus_sysfs_files(struct mcctrl_usrdata *udp)
 {
 	int error;
-#ifdef POSTK_DEBUG_ARCH_DEP_40 /* cpu_topology name change */
 	struct mcctrl_cpu_topology *cpu;
-#else /* POSTK_DEBUG_ARCH_DEP_40 */
-	struct cpu_topology *cpu;
-#endif /* POSTK_DEBUG_ARCH_DEP_40 */
 
 	error = get_cpu_topology(udp);
 	if (error) {

executer/kernel/mcoverlayfs/CMakeLists.txt

@@ -1,14 +0,0 @@
-# LESS/GREATER_EQUAL appears somewhere in 3.7... meh compat until we stop caring about 2.x
-# ...apparently can't define macros ot use inside if, so unfold manually
-
-if(NOT (LINUX_VERSION_CODE LESS 262144) AND NOT (LINUX_VERSION_CODE GREATER 262400))
-	add_subdirectory("linux-4.0.9")
-elseif(NOT (LINUX_VERSION_CODE LESS 263680) AND NOT (LINUX_VERSION_CODE GREATER 263936))
-	add_subdirectory("linux-4.6.7")
-elseif(LINUX_VERSION_CODE EQUAL 199168)
-	add_subdirectory("linux-3.10.0-327.36.1.el7")
-else()
-	#add_subdirectory("linux-3.10.0-327.36.1.el7")
-	add_subdirectory("linux-4.18.14")
-	#message(FATAL_ERROR "mcoverlayfs enabled but kernel version not compatible")
-endif()

executer/kernel/mcoverlayfs/linux-3.10.0-327.36.1.el7/CMakeLists.txt

@@ -1,7 +0,0 @@
-kmod(mcoverlay
-	SOURCES
-		copy_up.c dir.c inode.c readdir.c super.c
-	INSTALL_DEST
-		${KMODDIR}
-)
-

executer/kernel/mcoverlayfs/linux-3.10.0-327.36.1.el7/copy_up.c

@@ -1,461 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/module.h>
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/file.h>
-#include <linux/splice.h>
-#include <linux/xattr.h>
-#include <linux/security.h>
-#include <linux/uaccess.h>
-#include <linux/sched.h>
-#include <linux/namei.h>
-#include <linux/fdtable.h>
-#include <linux/ratelimit.h>
-#include "overlayfs.h"
-
-#define OVL_COPY_UP_CHUNK_SIZE (1 << 20)
-
-static unsigned ovl_check_copy_up = 1;
-module_param_named(check_copy_up, ovl_check_copy_up, uint,
-		   S_IWUSR | S_IRUGO);
-MODULE_PARM_DESC(ovl_check_copy_up,
-		 "Warn on copy-up when causing process also has a R/O fd open");
-
-static int ovl_check_fd(const void *data, struct file *f, unsigned fd)
-{
-	const struct dentry *dentry = data;
-
-	if (f->f_path.dentry == dentry)
-		pr_warn_ratelimited("overlayfs: Warning: Copying up %pD, but open R/O on fd %u which will cease to be coherent [pid=%d %s]\n",
-				    f, fd, current->pid, current->comm);
-	return 0;
-}
-
-/*
- * Check the fds open by this process and warn if something like the following
- * scenario is about to occur:
- *
- *	fd1 = open("foo", O_RDONLY);
- *	fd2 = open("foo", O_RDWR);
- */
-static void ovl_do_check_copy_up(struct dentry *dentry)
-{
-	if (ovl_check_copy_up)
-		iterate_fd(current->files, 0, ovl_check_fd, dentry);
-}
-
-int ovl_copy_xattr(struct dentry *old, struct dentry *new)
-{
-	ssize_t list_size, size, value_size = 0;
-	char *buf, *name, *value = NULL;
-	int uninitialized_var(error);
-
-	if (!old->d_inode->i_op->getxattr ||
-	    !new->d_inode->i_op->getxattr)
-		return 0;
-
-	list_size = vfs_listxattr(old, NULL, 0);
-	if (list_size <= 0) {
-		if (list_size == -EOPNOTSUPP)
-			return 0;
-		return list_size;
-	}
-
-	buf = kzalloc(list_size, GFP_KERNEL);
-	if (!buf)
-		return -ENOMEM;
-
-	list_size = vfs_listxattr(old, buf, list_size);
-	if (list_size <= 0) {
-		error = list_size;
-		goto out;
-	}
-
-	for (name = buf; name < (buf + list_size); name += strlen(name) + 1) {
-retry:
-		size = vfs_getxattr(old, name, value, value_size);
-		if (size == -ERANGE)
-			size = vfs_getxattr(old, name, NULL, 0);
-
-		if (size < 0) {
-			error = size;
-			break;
-		}
-
-		if (size > value_size) {
-			void *new;
-
-			new = krealloc(value, size, GFP_KERNEL);
-			if (!new) {
-				error = -ENOMEM;
-				break;
-			}
-			value = new;
-			value_size = size;
-			goto retry;
-		}
-
-		error = vfs_setxattr(new, name, value, size, 0);
-		if (error)
-			break;
-	}
-	kfree(value);
-out:
-	kfree(buf);
-	return error;
-}
-
-static int ovl_copy_up_data(struct path *old, struct path *new, loff_t len)
-{
-	struct file *old_file;
-	struct file *new_file;
-	loff_t old_pos = 0;
-	loff_t new_pos = 0;
-	int error = 0;
-
-	if (len == 0)
-		return 0;
-
-	old_file = ovl_path_open(old, O_RDONLY);
-	if (IS_ERR(old_file))
-		return PTR_ERR(old_file);
-
-	new_file = ovl_path_open(new, O_WRONLY);
-	if (IS_ERR(new_file)) {
-		error = PTR_ERR(new_file);
-		goto out_fput;
-	}
-
-	/* FIXME: copy up sparse files efficiently */
-	while (len) {
-		size_t this_len = OVL_COPY_UP_CHUNK_SIZE;
-		long bytes;
-
-		if (len < this_len)
-			this_len = len;
-
-		if (signal_pending_state(TASK_KILLABLE, current)) {
-			error = -EINTR;
-			break;
-		}
-
-		bytes = do_splice_direct(old_file, &old_pos,
-					 new_file, &new_pos,
-					 this_len, SPLICE_F_MOVE);
-		if (bytes <= 0) {
-			error = bytes;
-			break;
-		}
-		WARN_ON(old_pos != new_pos);
-
-		len -= bytes;
-	}
-
-	fput(new_file);
-out_fput:
-	fput(old_file);
-	return error;
-}
-
-static char *ovl_read_symlink(struct dentry *realdentry)
-{
-	int res;
-	char *buf;
-	struct inode *inode = realdentry->d_inode;
-	mm_segment_t old_fs;
-
-	res = -EINVAL;
-	if (!inode->i_op->readlink)
-		goto err;
-
-	res = -ENOMEM;
-	buf = (char *) __get_free_page(GFP_KERNEL);
-	if (!buf)
-		goto err;
-
-	old_fs = get_fs();
-	set_fs(get_ds());
-	/* The cast to a user pointer is valid due to the set_fs() */
-	res = inode->i_op->readlink(realdentry,
-				    (char __user *)buf, PAGE_SIZE - 1);
-	set_fs(old_fs);
-	if (res < 0) {
-		free_page((unsigned long) buf);
-		goto err;
-	}
-	buf[res] = '\0';
-
-	return buf;
-
-err:
-	return ERR_PTR(res);
-}
-
-static int ovl_set_timestamps(struct dentry *upperdentry, struct kstat *stat)
-{
-	struct iattr attr = {
-		.ia_valid =
-		     ATTR_ATIME | ATTR_MTIME | ATTR_ATIME_SET | ATTR_MTIME_SET,
-		.ia_atime = stat->atime,
-		.ia_mtime = stat->mtime,
-	};
-
-	return notify_change(upperdentry, &attr, NULL);
-}
-
-int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat)
-{
-	int err = 0;
-
-	if (!S_ISLNK(stat->mode)) {
-		struct iattr attr = {
-			.ia_valid = ATTR_MODE,
-			.ia_mode = stat->mode,
-		};
-		err = notify_change(upperdentry, &attr, NULL);
-	}
-	if (!err) {
-		struct iattr attr = {
-			.ia_valid = ATTR_UID | ATTR_GID,
-			.ia_uid = stat->uid,
-			.ia_gid = stat->gid,
-		};
-		err = notify_change(upperdentry, &attr, NULL);
-	}
-	if (!err)
-		ovl_set_timestamps(upperdentry, stat);
-
-	return err;
-}
-
-static int ovl_copy_up_locked(struct dentry *workdir, struct dentry *upperdir,
-			      struct dentry *dentry, struct path *lowerpath,
-			      struct kstat *stat, struct iattr *attr,
-			      const char *link)
-{
-	struct inode *wdir = workdir->d_inode;
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *newdentry = NULL;
-	struct dentry *upper = NULL;
-	umode_t mode = stat->mode;
-	int err;
-
-	newdentry = ovl_lookup_temp(workdir, dentry);
-	err = PTR_ERR(newdentry);
-	if (IS_ERR(newdentry))
-		goto out;
-
-	upper = lookup_one_len(dentry->d_name.name, upperdir,
-			       dentry->d_name.len);
-	err = PTR_ERR(upper);
-	if (IS_ERR(upper))
-		goto out1;
-
-	/* Can't properly set mode on creation because of the umask */
-	stat->mode &= S_IFMT;
-	err = ovl_create_real(wdir, newdentry, stat, link, NULL, true);
-	stat->mode = mode;
-	if (err)
-		goto out2;
-
-	if (S_ISREG(stat->mode)) {
-		struct path upperpath;
-		ovl_path_upper(dentry, &upperpath);
-		BUG_ON(upperpath.dentry != NULL);
-		upperpath.dentry = newdentry;
-
-		err = ovl_copy_up_data(lowerpath, &upperpath, stat->size);
-		if (err)
-			goto out_cleanup;
-	}
-
-	err = ovl_copy_xattr(lowerpath->dentry, newdentry);
-	if (err)
-		goto out_cleanup;
-
-	mutex_lock(&newdentry->d_inode->i_mutex);
-	err = ovl_set_attr(newdentry, stat);
-	if (!err && attr)
-		err = notify_change(newdentry, attr, NULL);
-	mutex_unlock(&newdentry->d_inode->i_mutex);
-	if (err)
-		goto out_cleanup;
-
-	err = ovl_do_rename(wdir, newdentry, udir, upper, 0);
-	if (err)
-		goto out_cleanup;
-
-	ovl_dentry_update(dentry, newdentry);
-	newdentry = NULL;
-
-	/*
-	 * Non-directores become opaque when copied up.
-	 */
-	if (!S_ISDIR(stat->mode))
-		ovl_dentry_set_opaque(dentry, true);
-out2:
-	dput(upper);
-out1:
-	dput(newdentry);
-out:
-	return err;
-
-out_cleanup:
-	ovl_cleanup(wdir, newdentry);
-	goto out;
-}
-
-/*
- * Copy up a single dentry
- *
- * Directory renames only allowed on "pure upper" (already created on
- * upper filesystem, never copied up).  Directories which are on lower or
- * are merged may not be renamed.  For these -EXDEV is returned and
- * userspace has to deal with it.  This means, when copying up a
- * directory we can rely on it and ancestors being stable.
- *
- * Non-directory renames start with copy up of source if necessary.  The
- * actual rename will only proceed once the copy up was successful.  Copy
- * up uses upper parent i_mutex for exclusion.  Since rename can change
- * d_parent it is possible that the copy up will lock the old parent.  At
- * that point the file will have already been copied up anyway.
- */
-int ovl_copy_up_one(struct dentry *parent, struct dentry *dentry,
-		    struct path *lowerpath, struct kstat *stat,
-		    struct iattr *attr)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	int err;
-	struct kstat pstat;
-	struct path parentpath;
-	struct dentry *upperdir;
-	struct dentry *upperdentry;
-	const struct cred *old_cred;
-	struct cred *override_cred;
-	char *link = NULL;
-
-	if (WARN_ON(!workdir))
-		return -EROFS;
-
-	ovl_do_check_copy_up(lowerpath->dentry);
-
-	ovl_path_upper(parent, &parentpath);
-	upperdir = parentpath.dentry;
-
-	err = vfs_getattr(&parentpath, &pstat);
-	if (err)
-		return err;
-
-	if (S_ISLNK(stat->mode)) {
-		link = ovl_read_symlink(lowerpath->dentry);
-		if (IS_ERR(link))
-			return PTR_ERR(link);
-	}
-
-	err = -ENOMEM;
-	override_cred = prepare_creds();
-	if (!override_cred)
-		goto out_free_link;
-
-	override_cred->fsuid = stat->uid;
-	override_cred->fsgid = stat->gid;
-	/*
-	 * CAP_SYS_ADMIN for copying up extended attributes
-	 * CAP_DAC_OVERRIDE for create
-	 * CAP_FOWNER for chmod, timestamp update
-	 * CAP_FSETID for chmod
-	 * CAP_CHOWN for chown
-	 * CAP_MKNOD for mknod
-	 */
-	cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-	cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-	cap_raise(override_cred->cap_effective, CAP_FOWNER);
-	cap_raise(override_cred->cap_effective, CAP_FSETID);
-	cap_raise(override_cred->cap_effective, CAP_CHOWN);
-	cap_raise(override_cred->cap_effective, CAP_MKNOD);
-	old_cred = override_creds(override_cred);
-
-	err = -EIO;
-	if (lock_rename(workdir, upperdir) != NULL) {
-		pr_err("overlayfs: failed to lock workdir+upperdir\n");
-		goto out_unlock;
-	}
-	upperdentry = ovl_dentry_upper(dentry);
-	if (upperdentry) {
-		unlock_rename(workdir, upperdir);
-		err = 0;
-		/* Raced with another copy-up?  Do the setattr here */
-		if (attr) {
-			mutex_lock(&upperdentry->d_inode->i_mutex);
-			err = notify_change(upperdentry, attr, NULL);
-			mutex_unlock(&upperdentry->d_inode->i_mutex);
-		}
-		goto out_put_cred;
-	}
-
-	err = ovl_copy_up_locked(workdir, upperdir, dentry, lowerpath,
-				 stat, attr, link);
-	if (!err) {
-		/* Restore timestamps on parent (best effort) */
-		ovl_set_timestamps(upperdir, &pstat);
-	}
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out_put_cred:
-	revert_creds(old_cred);
-	put_cred(override_cred);
-
-out_free_link:
-	if (link)
-		free_page((unsigned long) link);
-
-	return err;
-}
-
-int ovl_copy_up(struct dentry *dentry)
-{
-	int err;
-
-	err = 0;
-	while (!err) {
-		struct dentry *next;
-		struct dentry *parent;
-		struct path lowerpath;
-		struct kstat stat;
-		enum ovl_path_type type = ovl_path_type(dentry);
-
-		if (OVL_TYPE_UPPER(type))
-			break;
-
-		next = dget(dentry);
-		/* find the topmost dentry not yet copied up */
-		for (;;) {
-			parent = dget_parent(next);
-
-			type = ovl_path_type(parent);
-			if (OVL_TYPE_UPPER(type))
-				break;
-
-			dput(next);
-			next = parent;
-		}
-
-		ovl_path_lower(next, &lowerpath);
-		err = vfs_getattr(&lowerpath, &stat);
-		if (!err)
-			err = ovl_copy_up_one(parent, next, &lowerpath, &stat, NULL);
-
-		dput(parent);
-		dput(next);
-	}
-
-	return err;
-}

executer/kernel/mcoverlayfs/linux-3.10.0-327.36.1.el7/dir.c

@@ -1,972 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/namei.h>
-#include <linux/xattr.h>
-#include <linux/security.h>
-#include <linux/cred.h>
-#include "overlayfs.h"
-
-void ovl_cleanup(struct inode *wdir, struct dentry *wdentry)
-{
-	int err;
-
-	dget(wdentry);
-	if (S_ISDIR(wdentry->d_inode->i_mode))
-		err = ovl_do_rmdir(wdir, wdentry);
-	else
-		err = ovl_do_unlink(wdir, wdentry);
-	dput(wdentry);
-
-	if (err) {
-		pr_err("overlayfs: cleanup of '%pd2' failed (%i)\n",
-		       wdentry, err);
-	}
-}
-
-struct dentry *ovl_lookup_temp(struct dentry *workdir, struct dentry *dentry)
-{
-	struct dentry *temp;
-	char name[20];
-
-	snprintf(name, sizeof(name), "#%lx", (unsigned long) dentry);
-
-	temp = lookup_one_len(name, workdir, strlen(name));
-	if (!IS_ERR(temp) && temp->d_inode) {
-		pr_err("overlayfs: workdir/%s already exists\n", name);
-		dput(temp);
-		temp = ERR_PTR(-EIO);
-	}
-
-	return temp;
-}
-
-/* caller holds i_mutex on workdir */
-static struct dentry *ovl_whiteout(struct dentry *workdir,
-				   struct dentry *dentry)
-{
-	int err;
-	struct dentry *whiteout;
-	struct inode *wdir = workdir->d_inode;
-
-	whiteout = ovl_lookup_temp(workdir, dentry);
-	if (IS_ERR(whiteout))
-		return whiteout;
-
-	err = ovl_do_whiteout(wdir, whiteout);
-	if (err) {
-		dput(whiteout);
-		whiteout = ERR_PTR(err);
-	}
-
-	return whiteout;
-}
-
-int ovl_create_real(struct inode *dir, struct dentry *newdentry,
-		    struct kstat *stat, const char *link,
-		    struct dentry *hardlink, bool debug)
-{
-	int err;
-
-	if (newdentry->d_inode)
-		return -ESTALE;
-
-	if (hardlink) {
-		err = ovl_do_link(hardlink, dir, newdentry, debug);
-	} else {
-		switch (stat->mode & S_IFMT) {
-		case S_IFREG:
-			err = ovl_do_create(dir, newdentry, stat->mode, debug);
-			break;
-
-		case S_IFDIR:
-			err = ovl_do_mkdir(dir, newdentry, stat->mode, debug);
-			break;
-
-		case S_IFCHR:
-		case S_IFBLK:
-		case S_IFIFO:
-		case S_IFSOCK:
-			err = ovl_do_mknod(dir, newdentry,
-					   stat->mode, stat->rdev, debug);
-			break;
-
-		case S_IFLNK:
-			err = ovl_do_symlink(dir, newdentry, link, debug);
-			break;
-
-		default:
-			err = -EPERM;
-		}
-	}
-	if (!err && WARN_ON(!newdentry->d_inode)) {
-		/*
-		 * Not quite sure if non-instantiated dentry is legal or not.
-		 * VFS doesn't seem to care so check and warn here.
-		 */
-		err = -ENOENT;
-	}
-	return err;
-}
-
-static int ovl_set_opaque(struct dentry *upperdentry)
-{
-	return ovl_do_setxattr(upperdentry, OVL_XATTR_OPAQUE, "y", 1, 0);
-}
-
-static void ovl_remove_opaque(struct dentry *upperdentry)
-{
-	int err;
-
-	err = ovl_do_removexattr(upperdentry, OVL_XATTR_OPAQUE);
-	if (err) {
-		pr_warn("overlayfs: failed to remove opaque from '%s' (%i)\n",
-			upperdentry->d_name.name, err);
-	}
-}
-
-static int ovl_dir_getattr(struct vfsmount *mnt, struct dentry *dentry,
-			 struct kstat *stat)
-{
-	int err;
-	enum ovl_path_type type;
-	struct path realpath;
-
-	type = ovl_path_real(dentry, &realpath);
-	err = vfs_getattr(&realpath, stat);
-	if (err)
-		return err;
-
-	stat->dev = dentry->d_sb->s_dev;
-	stat->ino = dentry->d_inode->i_ino;
-
-	/*
-	 * It's probably not worth it to count subdirs to get the
-	 * correct link count.  nlink=1 seems to pacify 'find' and
-	 * other utilities.
-	 */
-	if (OVL_TYPE_MERGE(type))
-		stat->nlink = 1;
-
-	return 0;
-}
-
-static int ovl_create_upper(struct dentry *dentry, struct inode *inode,
-			    struct kstat *stat, const char *link,
-			    struct dentry *hardlink)
-{
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *newdentry;
-	int err;
-
-	mutex_lock_nested(&udir->i_mutex, I_MUTEX_PARENT);
-	newdentry = lookup_one_len(dentry->d_name.name, upperdir,
-				   dentry->d_name.len);
-	err = PTR_ERR(newdentry);
-	if (IS_ERR(newdentry))
-		goto out_unlock;
-	err = ovl_create_real(udir, newdentry, stat, link, hardlink, false);
-	if (err)
-		goto out_dput;
-
-	ovl_dentry_version_inc(dentry->d_parent);
-	ovl_dentry_update(dentry, newdentry);
-	ovl_copyattr(newdentry->d_inode, inode);
-	d_instantiate(dentry, inode);
-	newdentry = NULL;
-out_dput:
-	dput(newdentry);
-out_unlock:
-	mutex_unlock(&udir->i_mutex);
-	return err;
-}
-
-static int ovl_lock_rename_workdir(struct dentry *workdir,
-				   struct dentry *upperdir)
-{
-	/* Workdir should not be the same as upperdir */
-	if (workdir == upperdir)
-		goto err;
-
-	/* Workdir should not be subdir of upperdir and vice versa */
-	if (lock_rename(workdir, upperdir) != NULL)
-		goto err_unlock;
-
-	return 0;
-
-err_unlock:
-	unlock_rename(workdir, upperdir);
-err:
-	pr_err("overlayfs: failed to lock workdir+upperdir\n");
-	return -EIO;
-}
-
-static struct dentry *ovl_clear_empty(struct dentry *dentry,
-				      struct list_head *list)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	struct inode *wdir = workdir->d_inode;
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct path upperpath;
-	struct dentry *upper;
-	struct dentry *opaquedir;
-	struct kstat stat;
-	int err;
-
-	if (WARN_ON(!workdir))
-		return ERR_PTR(-EROFS);
-
-	err = ovl_lock_rename_workdir(workdir, upperdir);
-	if (err)
-		goto out;
-
-	ovl_path_upper(dentry, &upperpath);
-	err = vfs_getattr(&upperpath, &stat);
-	if (err)
-		goto out_unlock;
-
-	err = -ESTALE;
-	if (!S_ISDIR(stat.mode))
-		goto out_unlock;
-	upper = upperpath.dentry;
-	if (upper->d_parent->d_inode != udir)
-		goto out_unlock;
-
-	opaquedir = ovl_lookup_temp(workdir, dentry);
-	err = PTR_ERR(opaquedir);
-	if (IS_ERR(opaquedir))
-		goto out_unlock;
-
-	err = ovl_create_real(wdir, opaquedir, &stat, NULL, NULL, true);
-	if (err)
-		goto out_dput;
-
-	err = ovl_copy_xattr(upper, opaquedir);
-	if (err)
-		goto out_cleanup;
-
-	err = ovl_set_opaque(opaquedir);
-	if (err)
-		goto out_cleanup;
-
-	mutex_lock(&opaquedir->d_inode->i_mutex);
-	err = ovl_set_attr(opaquedir, &stat);
-	mutex_unlock(&opaquedir->d_inode->i_mutex);
-	if (err)
-		goto out_cleanup;
-
-	err = ovl_do_rename(wdir, opaquedir, udir, upper, RENAME_EXCHANGE);
-	if (err)
-		goto out_cleanup;
-
-	ovl_cleanup_whiteouts(upper, list);
-	ovl_cleanup(wdir, upper);
-	unlock_rename(workdir, upperdir);
-
-	/* dentry's upper doesn't match now, get rid of it */
-	d_drop(dentry);
-
-	return opaquedir;
-
-out_cleanup:
-	ovl_cleanup(wdir, opaquedir);
-out_dput:
-	dput(opaquedir);
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out:
-	return ERR_PTR(err);
-}
-
-static struct dentry *ovl_check_empty_and_clear(struct dentry *dentry)
-{
-	int err;
-	struct dentry *ret = NULL;
-	LIST_HEAD(list);
-
-	err = ovl_check_empty_dir(dentry, &list);
-	if (err)
-		ret = ERR_PTR(err);
-	else {
-		/*
-		 * If no upperdentry then skip clearing whiteouts.
-		 *
-		 * Can race with copy-up, since we don't hold the upperdir
-		 * mutex.  Doesn't matter, since copy-up can't create a
-		 * non-empty directory from an empty one.
-		 */
-		if (ovl_dentry_upper(dentry))
-			ret = ovl_clear_empty(dentry, &list);
-	}
-
-	ovl_cache_free(&list);
-
-	return ret;
-}
-
-static int ovl_create_over_whiteout(struct dentry *dentry, struct inode *inode,
-				    struct kstat *stat, const char *link,
-				    struct dentry *hardlink)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	struct inode *wdir = workdir->d_inode;
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *upper;
-	struct dentry *newdentry;
-	int err;
-
-	if (WARN_ON(!workdir))
-		return -EROFS;
-
-	err = ovl_lock_rename_workdir(workdir, upperdir);
-	if (err)
-		goto out;
-
-	newdentry = ovl_lookup_temp(workdir, dentry);
-	err = PTR_ERR(newdentry);
-	if (IS_ERR(newdentry))
-		goto out_unlock;
-
-	upper = lookup_one_len(dentry->d_name.name, upperdir,
-			       dentry->d_name.len);
-	err = PTR_ERR(upper);
-	if (IS_ERR(upper))
-		goto out_dput;
-
-	err = ovl_create_real(wdir, newdentry, stat, link, hardlink, true);
-	if (err)
-		goto out_dput2;
-
-	if (S_ISDIR(stat->mode)) {
-		err = ovl_set_opaque(newdentry);
-		if (err)
-			goto out_cleanup;
-
-		err = ovl_do_rename(wdir, newdentry, udir, upper,
-				    RENAME_EXCHANGE);
-		if (err)
-			goto out_cleanup;
-
-		ovl_cleanup(wdir, upper);
-	} else {
-		err = ovl_do_rename(wdir, newdentry, udir, upper, 0);
-		if (err)
-			goto out_cleanup;
-	}
-	ovl_dentry_version_inc(dentry->d_parent);
-	ovl_dentry_update(dentry, newdentry);
-	ovl_copyattr(newdentry->d_inode, inode);
-	d_instantiate(dentry, inode);
-	newdentry = NULL;
-out_dput2:
-	dput(upper);
-out_dput:
-	dput(newdentry);
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out:
-	return err;
-
-out_cleanup:
-	ovl_cleanup(wdir, newdentry);
-	goto out_dput2;
-}
-
-static int ovl_create_or_link(struct dentry *dentry, int mode, dev_t rdev,
-			      const char *link, struct dentry *hardlink)
-{
-	int err;
-	struct inode *inode;
-	struct kstat stat = {
-		.mode = mode,
-		.rdev = rdev,
-	};
-
-	err = -ENOMEM;
-	inode = ovl_new_inode(dentry->d_sb, mode, dentry->d_fsdata);
-	if (!inode)
-		goto out;
-
-	err = ovl_copy_up(dentry->d_parent);
-	if (err)
-		goto out_iput;
-
-	if (!ovl_dentry_is_opaque(dentry)) {
-		err = ovl_create_upper(dentry, inode, &stat, link, hardlink);
-	} else {
-		const struct cred *old_cred;
-		struct cred *override_cred;
-
-		err = -ENOMEM;
-		override_cred = prepare_creds();
-		if (!override_cred)
-			goto out_iput;
-
-		/*
-		 * CAP_SYS_ADMIN for setting opaque xattr
-		 * CAP_DAC_OVERRIDE for create in workdir, rename
-		 * CAP_FOWNER for removing whiteout from sticky dir
-		 */
-		cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		cap_raise(override_cred->cap_effective, CAP_FOWNER);
-		old_cred = override_creds(override_cred);
-
-		err = ovl_create_over_whiteout(dentry, inode, &stat, link,
-					       hardlink);
-
-		revert_creds(old_cred);
-		put_cred(override_cred);
-	}
-
-	if (!err)
-		inode = NULL;
-out_iput:
-	iput(inode);
-out:
-	return err;
-}
-
-static int ovl_create_object(struct dentry *dentry, int mode, dev_t rdev,
-			     const char *link)
-{
-	int err;
-
-	err = ovl_want_write(dentry);
-	if (!err) {
-		err = ovl_create_or_link(dentry, mode, rdev, link, NULL);
-		ovl_drop_write(dentry);
-	}
-
-	return err;
-}
-
-static int ovl_create(struct inode *dir, struct dentry *dentry, umode_t mode,
-		      bool excl)
-{
-	return ovl_create_object(dentry, (mode & 07777) | S_IFREG, 0, NULL);
-}
-
-static int ovl_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
-{
-	return ovl_create_object(dentry, (mode & 07777) | S_IFDIR, 0, NULL);
-}
-
-static int ovl_mknod(struct inode *dir, struct dentry *dentry, umode_t mode,
-		     dev_t rdev)
-{
-	/* Don't allow creation of "whiteout" on overlay */
-	if (S_ISCHR(mode) && rdev == WHITEOUT_DEV)
-		return -EPERM;
-
-	return ovl_create_object(dentry, mode, rdev, NULL);
-}
-
-static int ovl_symlink(struct inode *dir, struct dentry *dentry,
-		       const char *link)
-{
-	return ovl_create_object(dentry, S_IFLNK, 0, link);
-}
-
-static int ovl_link(struct dentry *old, struct inode *newdir,
-		    struct dentry *new)
-{
-	int err;
-	struct dentry *upper;
-
-	err = ovl_want_write(old);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(old);
-	if (err)
-		goto out_drop_write;
-
-	upper = ovl_dentry_upper(old);
-	err = ovl_create_or_link(new, upper->d_inode->i_mode, 0, NULL, upper);
-
-out_drop_write:
-	ovl_drop_write(old);
-out:
-	return err;
-}
-
-static int ovl_remove_and_whiteout(struct dentry *dentry, bool is_dir)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	struct inode *wdir = workdir->d_inode;
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *whiteout;
-	struct dentry *upper;
-	struct dentry *opaquedir = NULL;
-	int err;
-	int flags = 0;
-
-	if (WARN_ON(!workdir))
-		return -EROFS;
-
-	if (is_dir) {
-		if (OVL_TYPE_MERGE_OR_LOWER(ovl_path_type(dentry))) {
-			opaquedir = ovl_check_empty_and_clear(dentry);
-			err = PTR_ERR(opaquedir);
-			if (IS_ERR(opaquedir))
-				goto out;
-		} else {
-			LIST_HEAD(list);
-
-			/*
-			 * When removing an empty opaque directory, then it
-			 * makes no sense to replace it with an exact replica of
-			 * itself.  But emptiness still needs to be checked.
-			 */
-			err = ovl_check_empty_dir(dentry, &list);
-			ovl_cache_free(&list);
-			if (err)
-				goto out;
-		}
-	}
-
-	err = ovl_lock_rename_workdir(workdir, upperdir);
-	if (err)
-		goto out_dput;
-
-	upper = lookup_one_len(dentry->d_name.name, upperdir,
-			       dentry->d_name.len);
-	err = PTR_ERR(upper);
-	if (IS_ERR(upper))
-		goto out_unlock;
-
-	err = -ESTALE;
-	if ((opaquedir && upper != opaquedir) ||
-	    (!opaquedir && ovl_dentry_upper(dentry) &&
-	     upper != ovl_dentry_upper(dentry))) {
-		goto out_dput_upper;
-	}
-
-	whiteout = ovl_whiteout(workdir, dentry);
-	err = PTR_ERR(whiteout);
-	if (IS_ERR(whiteout))
-		goto out_dput_upper;
-
-	if (d_is_dir(upper))
-		flags = RENAME_EXCHANGE;
-
-	err = ovl_do_rename(wdir, whiteout, udir, upper, flags);
-	if (err)
-		goto kill_whiteout;
-	if (flags)
-		ovl_cleanup(wdir, upper);
-
-	ovl_dentry_version_inc(dentry->d_parent);
-out_d_drop:
-	d_drop(dentry);
-	dput(whiteout);
-out_dput_upper:
-	dput(upper);
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out_dput:
-	dput(opaquedir);
-out:
-	return err;
-
-kill_whiteout:
-	ovl_cleanup(wdir, whiteout);
-	goto out_d_drop;
-}
-
-static int ovl_remove_upper(struct dentry *dentry, bool is_dir)
-{
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *dir = upperdir->d_inode;
-	struct dentry *upper;
-	int err;
-
-	mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
-	upper = lookup_one_len(dentry->d_name.name, upperdir,
-			       dentry->d_name.len);
-	err = PTR_ERR(upper);
-	if (IS_ERR(upper))
-		goto out_unlock;
-
-	err = -ESTALE;
-	if (upper == ovl_dentry_upper(dentry)) {
-		if (is_dir)
-			err = vfs_rmdir(dir, upper);
-		else
-			err = vfs_unlink(dir, upper, NULL);
-		ovl_dentry_version_inc(dentry->d_parent);
-	}
-	dput(upper);
-
-	/*
-	 * Keeping this dentry hashed would mean having to release
-	 * upperpath/lowerpath, which could only be done if we are the
-	 * sole user of this dentry.  Too tricky...  Just unhash for
-	 * now.
-	 */
-	if (!err)
-		d_drop(dentry);
-out_unlock:
-	mutex_unlock(&dir->i_mutex);
-
-	return err;
-}
-
-static inline int ovl_check_sticky(struct dentry *dentry)
-{
-	struct inode *dir = ovl_dentry_real(dentry->d_parent)->d_inode;
-	struct inode *inode = ovl_dentry_real(dentry)->d_inode;
-
-	if (check_sticky(dir, inode))
-		return -EPERM;
-
-	return 0;
-}
-
-static int ovl_do_remove(struct dentry *dentry, bool is_dir)
-{
-	enum ovl_path_type type;
-	int err;
-
-	err = ovl_check_sticky(dentry);
-	if (err)
-		goto out;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(dentry->d_parent);
-	if (err)
-		goto out_drop_write;
-
-	type = ovl_path_type(dentry);
-	if (OVL_TYPE_PURE_UPPER(type)) {
-		err = ovl_remove_upper(dentry, is_dir);
-	} else {
-		const struct cred *old_cred;
-		struct cred *override_cred;
-
-		err = -ENOMEM;
-		override_cred = prepare_creds();
-		if (!override_cred)
-			goto out_drop_write;
-
-		/*
-		 * CAP_SYS_ADMIN for setting xattr on whiteout, opaque dir
-		 * CAP_DAC_OVERRIDE for create in workdir, rename
-		 * CAP_FOWNER for removing whiteout from sticky dir
-		 * CAP_FSETID for chmod of opaque dir
-		 * CAP_CHOWN for chown of opaque dir
-		 */
-		cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		cap_raise(override_cred->cap_effective, CAP_FOWNER);
-		cap_raise(override_cred->cap_effective, CAP_FSETID);
-		cap_raise(override_cred->cap_effective, CAP_CHOWN);
-		old_cred = override_creds(override_cred);
-
-		err = ovl_remove_and_whiteout(dentry, is_dir);
-
-		revert_creds(old_cred);
-		put_cred(override_cred);
-	}
-out_drop_write:
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static int ovl_unlink(struct inode *dir, struct dentry *dentry)
-{
-	return ovl_do_remove(dentry, false);
-}
-
-static int ovl_rmdir(struct inode *dir, struct dentry *dentry)
-{
-	return ovl_do_remove(dentry, true);
-}
-
-static int ovl_rename2(struct inode *olddir, struct dentry *old,
-		       struct inode *newdir, struct dentry *new,
-		       unsigned int flags)
-{
-	int err;
-	enum ovl_path_type old_type;
-	enum ovl_path_type new_type;
-	struct dentry *old_upperdir;
-	struct dentry *new_upperdir;
-	struct dentry *olddentry;
-	struct dentry *newdentry;
-	struct dentry *trap;
-	bool old_opaque;
-	bool new_opaque;
-	bool new_create = false;
-	bool cleanup_whiteout = false;
-	bool overwrite = !(flags & RENAME_EXCHANGE);
-	bool is_dir = S_ISDIR(old->d_inode->i_mode);
-	bool new_is_dir = false;
-	struct dentry *opaquedir = NULL;
-	const struct cred *old_cred = NULL;
-	struct cred *override_cred = NULL;
-
-	err = -EINVAL;
-	if (flags & ~(RENAME_EXCHANGE | RENAME_NOREPLACE))
-		goto out;
-
-	flags &= ~RENAME_NOREPLACE;
-
-	err = ovl_check_sticky(old);
-	if (err)
-		goto out;
-
-	/* Don't copy up directory trees */
-	old_type = ovl_path_type(old);
-	err = -EXDEV;
-	if (OVL_TYPE_MERGE_OR_LOWER(old_type) && is_dir)
-		goto out;
-
-	if (new->d_inode) {
-		err = ovl_check_sticky(new);
-		if (err)
-			goto out;
-
-		if (S_ISDIR(new->d_inode->i_mode))
-			new_is_dir = true;
-
-		new_type = ovl_path_type(new);
-		err = -EXDEV;
-		if (!overwrite && OVL_TYPE_MERGE_OR_LOWER(new_type) && new_is_dir)
-			goto out;
-
-		err = 0;
-		if (!OVL_TYPE_UPPER(new_type) && !OVL_TYPE_UPPER(old_type)) {
-			if (ovl_dentry_lower(old)->d_inode ==
-			    ovl_dentry_lower(new)->d_inode)
-				goto out;
-		}
-		if (OVL_TYPE_UPPER(new_type) && OVL_TYPE_UPPER(old_type)) {
-			if (ovl_dentry_upper(old)->d_inode ==
-			    ovl_dentry_upper(new)->d_inode)
-				goto out;
-		}
-	} else {
-		if (ovl_dentry_is_opaque(new))
-			new_type = __OVL_PATH_UPPER;
-		else
-			new_type = __OVL_PATH_UPPER | __OVL_PATH_PURE;
-	}
-
-	err = ovl_want_write(old);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(old);
-	if (err)
-		goto out_drop_write;
-
-	err = ovl_copy_up(new->d_parent);
-	if (err)
-		goto out_drop_write;
-	if (!overwrite) {
-		err = ovl_copy_up(new);
-		if (err)
-			goto out_drop_write;
-	}
-
-	old_opaque = !OVL_TYPE_PURE_UPPER(old_type);
-	new_opaque = !OVL_TYPE_PURE_UPPER(new_type);
-
-	if (old_opaque || new_opaque) {
-		err = -ENOMEM;
-		override_cred = prepare_creds();
-		if (!override_cred)
-			goto out_drop_write;
-
-		/*
-		 * CAP_SYS_ADMIN for setting xattr on whiteout, opaque dir
-		 * CAP_DAC_OVERRIDE for create in workdir
-		 * CAP_FOWNER for removing whiteout from sticky dir
-		 * CAP_FSETID for chmod of opaque dir
-		 * CAP_CHOWN for chown of opaque dir
-		 */
-		cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		cap_raise(override_cred->cap_effective, CAP_FOWNER);
-		cap_raise(override_cred->cap_effective, CAP_FSETID);
-		cap_raise(override_cred->cap_effective, CAP_CHOWN);
-		old_cred = override_creds(override_cred);
-	}
-
-	if (overwrite && OVL_TYPE_MERGE_OR_LOWER(new_type) && new_is_dir) {
-		opaquedir = ovl_check_empty_and_clear(new);
-		err = PTR_ERR(opaquedir);
-		if (IS_ERR(opaquedir)) {
-			opaquedir = NULL;
-			goto out_revert_creds;
-		}
-	}
-
-	if (overwrite) {
-		if (old_opaque) {
-			if (new->d_inode || !new_opaque) {
-				/* Whiteout source */
-				flags |= RENAME_WHITEOUT;
-			} else {
-				/* Switch whiteouts */
-				flags |= RENAME_EXCHANGE;
-			}
-		} else if (is_dir && !new->d_inode && new_opaque) {
-			flags |= RENAME_EXCHANGE;
-			cleanup_whiteout = true;
-		}
-	}
-
-	old_upperdir = ovl_dentry_upper(old->d_parent);
-	new_upperdir = ovl_dentry_upper(new->d_parent);
-
-	trap = lock_rename(new_upperdir, old_upperdir);
-
-
-	olddentry = lookup_one_len(old->d_name.name, old_upperdir,
-				   old->d_name.len);
-	err = PTR_ERR(olddentry);
-	if (IS_ERR(olddentry))
-		goto out_unlock;
-
-	err = -ESTALE;
-	if (olddentry != ovl_dentry_upper(old))
-		goto out_dput_old;
-
-	newdentry = lookup_one_len(new->d_name.name, new_upperdir,
-				   new->d_name.len);
-	err = PTR_ERR(newdentry);
-	if (IS_ERR(newdentry))
-		goto out_dput_old;
-
-	err = -ESTALE;
-	if (ovl_dentry_upper(new)) {
-		if (opaquedir) {
-			if (newdentry != opaquedir)
-				goto out_dput;
-		} else {
-			if (newdentry != ovl_dentry_upper(new))
-				goto out_dput;
-		}
-	} else {
-		new_create = true;
-		if (!d_is_negative(newdentry) &&
-		    (!new_opaque || !ovl_is_whiteout(newdentry)))
-			goto out_dput;
-	}
-
-	if (olddentry == trap)
-		goto out_dput;
-	if (newdentry == trap)
-		goto out_dput;
-
-	if (is_dir && !old_opaque && new_opaque) {
-		err = ovl_set_opaque(olddentry);
-		if (err)
-			goto out_dput;
-	}
-	if (!overwrite && new_is_dir && old_opaque && !new_opaque) {
-		err = ovl_set_opaque(newdentry);
-		if (err)
-			goto out_dput;
-	}
-
-	if (old_opaque || new_opaque) {
-		err = ovl_do_rename(old_upperdir->d_inode, olddentry,
-				    new_upperdir->d_inode, newdentry,
-				    flags);
-	} else {
-		/* No debug for the plain case */
-		BUG_ON(flags & ~RENAME_EXCHANGE);
-		err = vfs_rename(old_upperdir->d_inode, olddentry,
-				 new_upperdir->d_inode, newdentry,
-				 NULL, flags);
-	}
-
-	if (err) {
-		if (is_dir && !old_opaque && new_opaque)
-			ovl_remove_opaque(olddentry);
-		if (!overwrite && new_is_dir && old_opaque && !new_opaque)
-			ovl_remove_opaque(newdentry);
-		goto out_dput;
-	}
-
-	if (is_dir && old_opaque && !new_opaque)
-		ovl_remove_opaque(olddentry);
-	if (!overwrite && new_is_dir && !old_opaque && new_opaque)
-		ovl_remove_opaque(newdentry);
-
-	if (old_opaque != new_opaque) {
-		ovl_dentry_set_opaque(old, new_opaque);
-		if (!overwrite)
-			ovl_dentry_set_opaque(new, old_opaque);
-	}
-
-	if (cleanup_whiteout)
-		ovl_cleanup(old_upperdir->d_inode, newdentry);
-
-	ovl_dentry_version_inc(old->d_parent);
-	ovl_dentry_version_inc(new->d_parent);
-
-out_dput:
-	dput(newdentry);
-out_dput_old:
-	dput(olddentry);
-out_unlock:
-	unlock_rename(new_upperdir, old_upperdir);
-out_revert_creds:
-	if (old_opaque || new_opaque) {
-		revert_creds(old_cred);
-		put_cred(override_cred);
-	}
-out_drop_write:
-	ovl_drop_write(old);
-out:
-	dput(opaquedir);
-	return err;
-}
-
-static int ovl_rename(struct inode *olddir, struct dentry *old,
-		      struct inode *newdir, struct dentry *new)
-{
-	return ovl_rename2(olddir, old, newdir, new, 0);
-}
-
-const struct inode_operations_wrapper ovl_dir_inode_operations = {
-	.ops = {
-	.lookup		= ovl_lookup,
-	.mkdir		= ovl_mkdir,
-	.symlink	= ovl_symlink,
-	.unlink		= ovl_unlink,
-	.rmdir		= ovl_rmdir,
-	.rename		= ovl_rename,
-	.link		= ovl_link,
-	.setattr	= ovl_setattr,
-	.create		= ovl_create,
-	.mknod		= ovl_mknod,
-	.permission	= ovl_permission,
-	.getattr	= ovl_dir_getattr,
-	.setxattr	= ovl_setxattr,
-	.getxattr	= ovl_getxattr,
-	.listxattr	= ovl_listxattr,
-	.removexattr	= ovl_removexattr,
-	},
-	.rename2	= ovl_rename2,
-};

executer/kernel/mcoverlayfs/linux-3.10.0-327.36.1.el7/inode.c

@@ -1,442 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/xattr.h>
-#include "overlayfs.h"
-
-static int ovl_copy_up_last(struct dentry *dentry, struct iattr *attr,
-			    bool no_data)
-{
-	int err;
-	struct dentry *parent;
-	struct kstat stat;
-	struct path lowerpath;
-
-	parent = dget_parent(dentry);
-	err = ovl_copy_up(parent);
-	if (err)
-		goto out_dput_parent;
-
-	ovl_path_lower(dentry, &lowerpath);
-	err = vfs_getattr(&lowerpath, &stat);
-	if (err)
-		goto out_dput_parent;
-
-	if (no_data)
-		stat.size = 0;
-
-	err = ovl_copy_up_one(parent, dentry, &lowerpath, &stat, attr);
-
-out_dput_parent:
-	dput(parent);
-	return err;
-}
-
-int ovl_setattr(struct dentry *dentry, struct iattr *attr)
-{
-	int err;
-	struct dentry *upperdentry;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(dentry);
-	if (!err) {
-		upperdentry = ovl_dentry_upper(dentry);
-
-		mutex_lock(&upperdentry->d_inode->i_mutex);
-		err = notify_change(upperdentry, attr, NULL);
-		mutex_unlock(&upperdentry->d_inode->i_mutex);
-	}
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static int ovl_getattr(struct vfsmount *mnt, struct dentry *dentry,
-			 struct kstat *stat)
-{
-	struct path realpath;
-
-	ovl_path_real(dentry, &realpath);
-	return vfs_getattr(&realpath, stat);
-}
-
-int ovl_permission(struct inode *inode, int mask)
-{
-	struct ovl_entry *oe;
-	struct dentry *alias = NULL;
-	struct inode *realinode;
-	struct dentry *realdentry;
-	bool is_upper;
-	int err;
-
-	if (S_ISDIR(inode->i_mode)) {
-		oe = inode->i_private;
-	} else if (mask & MAY_NOT_BLOCK) {
-		return -ECHILD;
-	} else {
-		/*
-		 * For non-directories find an alias and get the info
-		 * from there.
-		 */
-		alias = d_find_any_alias(inode);
-		if (WARN_ON(!alias))
-			return -ENOENT;
-
-		oe = alias->d_fsdata;
-	}
-
-	realdentry = ovl_entry_real(oe, &is_upper);
-
-	/* Careful in RCU walk mode */
-	realinode = ACCESS_ONCE(realdentry->d_inode);
-	if (!realinode) {
-		WARN_ON(!(mask & MAY_NOT_BLOCK));
-		err = -ENOENT;
-		goto out_dput;
-	}
-
-	if (mask & MAY_WRITE) {
-		umode_t mode = realinode->i_mode;
-
-		/*
-		 * Writes will always be redirected to upper layer, so
-		 * ignore lower layer being read-only.
-		 *
-		 * If the overlay itself is read-only then proceed
-		 * with the permission check, don't return EROFS.
-		 * This will only happen if this is the lower layer of
-		 * another overlayfs.
-		 *
-		 * If upper fs becomes read-only after the overlay was
-		 * constructed return EROFS to prevent modification of
-		 * upper layer.
-		 */
-		err = -EROFS;
-		if (is_upper && !IS_RDONLY(inode) && IS_RDONLY(realinode) &&
-		    (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)))
-			goto out_dput;
-	}
-
-	err = __inode_permission(realinode, mask);
-out_dput:
-	dput(alias);
-	return err;
-}
-
-
-struct ovl_link_data {
-	struct dentry *realdentry;
-	void *cookie;
-};
-
-static void *ovl_follow_link(struct dentry *dentry, struct nameidata *nd)
-{
-	void *ret;
-	struct dentry *realdentry;
-	struct inode *realinode;
-	struct ovl_link_data *data = NULL;
-
-	realdentry = ovl_dentry_real(dentry);
-	realinode = realdentry->d_inode;
-
-	if (WARN_ON(!realinode->i_op->follow_link))
-		return ERR_PTR(-EPERM);
-
-	if (realinode->i_op->put_link) {
-		data = kmalloc(sizeof(struct ovl_link_data), GFP_KERNEL);
-		if (!data)
-			return ERR_PTR(-ENOMEM);
-		data->realdentry = realdentry;
-	}
-
-	ret = realinode->i_op->follow_link(realdentry, nd);
-	if (IS_ERR(ret)) {
-		kfree(data);
-		return ret;
-	}
-
-	if (data)
-		data->cookie = ret;
-
-	return data;
-}
-
-static void ovl_put_link(struct dentry *dentry, struct nameidata *nd, void *c)
-{
-	struct inode *realinode;
-	struct ovl_link_data *data = c;
-
-	if (!data)
-		return;
-
-	realinode = data->realdentry->d_inode;
-	realinode->i_op->put_link(data->realdentry, nd, data->cookie);
-	kfree(data);
-}
-
-static int ovl_readlink(struct dentry *dentry, char __user *buf, int bufsiz)
-{
-	struct path realpath;
-	struct inode *realinode;
-
-	ovl_path_real(dentry, &realpath);
-	realinode = realpath.dentry->d_inode;
-
-	if (!realinode->i_op->readlink)
-		return -EINVAL;
-
-	touch_atime(&realpath);
-
-	return realinode->i_op->readlink(realpath.dentry, buf, bufsiz);
-}
-
-
-static bool ovl_is_private_xattr(const char *name)
-{
-	return strncmp(name, OVL_XATTR_PRE_NAME, OVL_XATTR_PRE_LEN) == 0;
-}
-
-int ovl_setxattr(struct dentry *dentry, const char *name,
-		 const void *value, size_t size, int flags)
-{
-	int err;
-	struct dentry *upperdentry;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = -EPERM;
-	if (ovl_is_private_xattr(name))
-		goto out_drop_write;
-
-	err = ovl_copy_up(dentry);
-	if (err)
-		goto out_drop_write;
-
-	upperdentry = ovl_dentry_upper(dentry);
-	err = vfs_setxattr(upperdentry, name, value, size, flags);
-
-out_drop_write:
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static bool ovl_need_xattr_filter(struct dentry *dentry,
-				  enum ovl_path_type type)
-{
-	if ((type & (__OVL_PATH_PURE | __OVL_PATH_UPPER)) == __OVL_PATH_UPPER)
-		return S_ISDIR(dentry->d_inode->i_mode);
-	else
-		return false;
-}
-
-ssize_t ovl_getxattr(struct dentry *dentry, const char *name,
-		     void *value, size_t size)
-{
-	struct path realpath;
-	enum ovl_path_type type = ovl_path_real(dentry, &realpath);
-
-	if (ovl_need_xattr_filter(dentry, type) && ovl_is_private_xattr(name))
-		return -ENODATA;
-
-	return vfs_getxattr(realpath.dentry, name, value, size);
-}
-
-ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size)
-{
-	struct path realpath;
-	enum ovl_path_type type = ovl_path_real(dentry, &realpath);
-	ssize_t res;
-	int off;
-
-	res = vfs_listxattr(realpath.dentry, list, size);
-	if (res <= 0 || size == 0)
-		return res;
-
-	if (!ovl_need_xattr_filter(dentry, type))
-		return res;
-
-	/* filter out private xattrs */
-	for (off = 0; off < res;) {
-		char *s = list + off;
-		size_t slen = strlen(s) + 1;
-
-		BUG_ON(off + slen > res);
-
-		if (ovl_is_private_xattr(s)) {
-			res -= slen;
-			memmove(s, s + slen, res - off);
-		} else {
-			off += slen;
-		}
-	}
-
-	return res;
-}
-
-int ovl_removexattr(struct dentry *dentry, const char *name)
-{
-	int err;
-	struct path realpath;
-	enum ovl_path_type type = ovl_path_real(dentry, &realpath);
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = -ENODATA;
-	if (ovl_need_xattr_filter(dentry, type) && ovl_is_private_xattr(name))
-		goto out_drop_write;
-
-	if (!OVL_TYPE_UPPER(type)) {
-		err = vfs_getxattr(realpath.dentry, name, NULL, 0);
-		if (err < 0)
-			goto out_drop_write;
-
-		err = ovl_copy_up(dentry);
-		if (err)
-			goto out_drop_write;
-
-		ovl_path_upper(dentry, &realpath);
-	}
-
-	err = vfs_removexattr(realpath.dentry, name);
-out_drop_write:
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static bool ovl_open_need_copy_up(int flags, enum ovl_path_type type,
-				  struct dentry *realdentry)
-{
-	if (OVL_TYPE_UPPER(type))
-		return false;
-
-	if (special_file(realdentry->d_inode->i_mode))
-		return false;
-
-	if (!(OPEN_FMODE(flags) & FMODE_WRITE) && !(flags & O_TRUNC))
-		return false;
-
-	return true;
-}
-
-static int ovl_dentry_open(struct dentry *dentry, struct file *file,
-		    const struct cred *cred)
-{
-	int err;
-	struct path realpath;
-	enum ovl_path_type type;
-	bool want_write = false;
-
-	type = ovl_path_real(dentry, &realpath);
-	if (!ovl_is_nocopyupw(dentry)) {
-		if (ovl_open_need_copy_up(file->f_flags, type,
-		    realpath.dentry)) {
-			want_write = true;
-			err = ovl_want_write(dentry);
-			if (err)
-				goto out;
-
-			if (file->f_flags & O_TRUNC)
-				err = ovl_copy_up_last(dentry, NULL, true);
-			else
-				err = ovl_copy_up(dentry);
-			if (err)
-				goto out_drop_write;
-
-			ovl_path_upper(dentry, &realpath);
-		}
-	}
-
-	err = vfs_open(&realpath, file, cred);
-out_drop_write:
-	if (want_write)
-		ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static const struct inode_operations_wrapper ovl_file_inode_operations = {
-	.ops = {
-	.setattr	= ovl_setattr,
-	.permission	= ovl_permission,
-	.getattr	= ovl_getattr,
-	.setxattr	= ovl_setxattr,
-	.getxattr	= ovl_getxattr,
-	.listxattr	= ovl_listxattr,
-	.removexattr	= ovl_removexattr,
-	},
-	.dentry_open	= ovl_dentry_open,
-};
-
-static const struct inode_operations ovl_symlink_inode_operations = {
-	.setattr	= ovl_setattr,
-	.follow_link	= ovl_follow_link,
-	.put_link	= ovl_put_link,
-	.readlink	= ovl_readlink,
-	.getattr	= ovl_getattr,
-	.setxattr	= ovl_setxattr,
-	.getxattr	= ovl_getxattr,
-	.listxattr	= ovl_listxattr,
-	.removexattr	= ovl_removexattr,
-};
-
-struct inode *ovl_new_inode(struct super_block *sb, umode_t mode,
-			    struct ovl_entry *oe)
-{
-	struct inode *inode;
-
-	inode = new_inode(sb);
-	if (!inode)
-		return NULL;
-
-	mode &= S_IFMT;
-
-	inode->i_ino = get_next_ino();
-	inode->i_mode = mode;
-	inode->i_flags |= S_NOATIME | S_NOCMTIME;
-
-	switch (mode) {
-	case S_IFDIR:
-		inode->i_private = oe;
-		inode->i_op = &ovl_dir_inode_operations.ops;
-		inode->i_fop = &ovl_dir_operations;
-		inode->i_flags |= S_IOPS_WRAPPER;
-		break;
-
-	case S_IFLNK:
-		inode->i_op = &ovl_symlink_inode_operations;
-		break;
-
-	case S_IFREG:
-	case S_IFSOCK:
-	case S_IFBLK:
-	case S_IFCHR:
-	case S_IFIFO:
-		inode->i_op = &ovl_file_inode_operations.ops;
-		inode->i_flags |= S_IOPS_WRAPPER;
-		break;
-
-	default:
-		WARN(1, "illegal file type: %i\n", mode);
-		iput(inode);
-		inode = NULL;
-	}
-
-	return inode;
-}

executer/kernel/mcoverlayfs/linux-3.10.0-327.36.1.el7/overlayfs.h

@@ -1,200 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/kernel.h>
-
-struct ovl_entry;
-
-enum ovl_path_type {
-	__OVL_PATH_PURE		= (1 << 0),
-	__OVL_PATH_UPPER	= (1 << 1),
-	__OVL_PATH_MERGE	= (1 << 2),
-};
-
-#define OVL_TYPE_UPPER(type)	((type) & __OVL_PATH_UPPER)
-#define OVL_TYPE_MERGE(type)	((type) & __OVL_PATH_MERGE)
-#define OVL_TYPE_PURE_UPPER(type) ((type) & __OVL_PATH_PURE)
-#define OVL_TYPE_MERGE_OR_LOWER(type) \
-	(OVL_TYPE_MERGE(type) || !OVL_TYPE_UPPER(type))
-
-#define OVL_XATTR_PRE_NAME "trusted.overlay."
-#define OVL_XATTR_PRE_LEN  16
-#define OVL_XATTR_OPAQUE   OVL_XATTR_PRE_NAME"opaque"
-
-static inline int ovl_do_rmdir(struct inode *dir, struct dentry *dentry)
-{
-	int err = vfs_rmdir(dir, dentry);
-	pr_debug("rmdir(%pd2) = %i\n", dentry, err);
-	return err;
-}
-
-static inline int ovl_do_unlink(struct inode *dir, struct dentry *dentry)
-{
-	int err = vfs_unlink(dir, dentry, NULL);
-	pr_debug("unlink(%pd2) = %i\n", dentry, err);
-	return err;
-}
-
-static inline int ovl_do_link(struct dentry *old_dentry, struct inode *dir,
-			      struct dentry *new_dentry, bool debug)
-{
-	int err = vfs_link(old_dentry, dir, new_dentry, NULL);
-	if (debug) {
-		pr_debug("link(%pd2, %pd2) = %i\n",
-			 old_dentry, new_dentry, err);
-	}
-	return err;
-}
-
-static inline int ovl_do_create(struct inode *dir, struct dentry *dentry,
-			     umode_t mode, bool debug)
-{
-	int err = vfs_create(dir, dentry, mode, true);
-	if (debug)
-		pr_debug("create(%pd2, 0%o) = %i\n", dentry, mode, err);
-	return err;
-}
-
-static inline int ovl_do_mkdir(struct inode *dir, struct dentry *dentry,
-			       umode_t mode, bool debug)
-{
-	int err = vfs_mkdir(dir, dentry, mode);
-	if (debug)
-		pr_debug("mkdir(%pd2, 0%o) = %i\n", dentry, mode, err);
-	return err;
-}
-
-static inline int ovl_do_mknod(struct inode *dir, struct dentry *dentry,
-			       umode_t mode, dev_t dev, bool debug)
-{
-	int err = vfs_mknod(dir, dentry, mode, dev);
-	if (debug) {
-		pr_debug("mknod(%pd2, 0%o, 0%o) = %i\n",
-			 dentry, mode, dev, err);
-	}
-	return err;
-}
-
-static inline int ovl_do_symlink(struct inode *dir, struct dentry *dentry,
-				 const char *oldname, bool debug)
-{
-	int err = vfs_symlink(dir, dentry, oldname);
-	if (debug)
-		pr_debug("symlink(\"%s\", %pd2) = %i\n", oldname, dentry, err);
-	return err;
-}
-
-static inline int ovl_do_setxattr(struct dentry *dentry, const char *name,
-				  const void *value, size_t size, int flags)
-{
-	int err = vfs_setxattr(dentry, name, value, size, flags);
-	pr_debug("setxattr(%pd2, \"%s\", \"%*s\", 0x%x) = %i\n",
-		 dentry, name, (int) size, (char *) value, flags, err);
-	return err;
-}
-
-static inline int ovl_do_removexattr(struct dentry *dentry, const char *name)
-{
-	int err = vfs_removexattr(dentry, name);
-	pr_debug("removexattr(%pd2, \"%s\") = %i\n", dentry, name, err);
-	return err;
-}
-
-static inline int ovl_do_rename(struct inode *olddir, struct dentry *olddentry,
-				struct inode *newdir, struct dentry *newdentry,
-				unsigned int flags)
-{
-	int err;
-
-	pr_debug("rename2(%pd2, %pd2, 0x%x)\n",
-		 olddentry, newdentry, flags);
-
-	err = vfs_rename(olddir, olddentry, newdir, newdentry, NULL, flags);
-
-	if (err) {
-		pr_debug("...rename2(%pd2, %pd2, ...) = %i\n",
-			 olddentry, newdentry, err);
-	}
-	return err;
-}
-
-static inline int ovl_do_whiteout(struct inode *dir, struct dentry *dentry)
-{
-	int err = vfs_whiteout(dir, dentry);
-	pr_debug("whiteout(%pd2) = %i\n", dentry, err);
-	return err;
-}
-
-bool ovl_is_nocopyupw(struct dentry *dentry);
-enum ovl_path_type ovl_path_type(struct dentry *dentry);
-u64 ovl_dentry_version_get(struct dentry *dentry);
-void ovl_dentry_version_inc(struct dentry *dentry);
-void ovl_path_upper(struct dentry *dentry, struct path *path);
-void ovl_path_lower(struct dentry *dentry, struct path *path);
-enum ovl_path_type ovl_path_real(struct dentry *dentry, struct path *path);
-int ovl_path_next(int idx, struct dentry *dentry, struct path *path);
-struct dentry *ovl_dentry_upper(struct dentry *dentry);
-struct dentry *ovl_dentry_lower(struct dentry *dentry);
-struct dentry *ovl_dentry_real(struct dentry *dentry);
-struct dentry *ovl_entry_real(struct ovl_entry *oe, bool *is_upper);
-struct ovl_dir_cache *ovl_dir_cache(struct dentry *dentry);
-void ovl_set_dir_cache(struct dentry *dentry, struct ovl_dir_cache *cache);
-struct dentry *ovl_workdir(struct dentry *dentry);
-int ovl_want_write(struct dentry *dentry);
-void ovl_drop_write(struct dentry *dentry);
-bool ovl_dentry_is_opaque(struct dentry *dentry);
-void ovl_dentry_set_opaque(struct dentry *dentry, bool opaque);
-bool ovl_is_whiteout(struct dentry *dentry);
-void ovl_dentry_update(struct dentry *dentry, struct dentry *upperdentry);
-struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry,
-			  unsigned int flags);
-struct file *ovl_path_open(struct path *path, int flags);
-
-struct dentry *ovl_upper_create(struct dentry *upperdir, struct dentry *dentry,
-				struct kstat *stat, const char *link);
-
-/* readdir.c */
-extern const struct file_operations ovl_dir_operations;
-int ovl_check_empty_dir(struct dentry *dentry, struct list_head *list);
-void ovl_cleanup_whiteouts(struct dentry *upper, struct list_head *list);
-void ovl_cache_free(struct list_head *list);
-
-/* inode.c */
-int ovl_setattr(struct dentry *dentry, struct iattr *attr);
-int ovl_permission(struct inode *inode, int mask);
-int ovl_setxattr(struct dentry *dentry, const char *name,
-		 const void *value, size_t size, int flags);
-ssize_t ovl_getxattr(struct dentry *dentry, const char *name,
-		     void *value, size_t size);
-ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size);
-int ovl_removexattr(struct dentry *dentry, const char *name);
-
-struct inode *ovl_new_inode(struct super_block *sb, umode_t mode,
-			    struct ovl_entry *oe);
-static inline void ovl_copyattr(struct inode *from, struct inode *to)
-{
-	to->i_uid = from->i_uid;
-	to->i_gid = from->i_gid;
-}
-
-/* dir.c */
-extern const struct inode_operations_wrapper ovl_dir_inode_operations;
-struct dentry *ovl_lookup_temp(struct dentry *workdir, struct dentry *dentry);
-int ovl_create_real(struct inode *dir, struct dentry *newdentry,
-		    struct kstat *stat, const char *link,
-		    struct dentry *hardlink, bool debug);
-void ovl_cleanup(struct inode *dir, struct dentry *dentry);
-
-/* copy_up.c */
-int ovl_copy_up(struct dentry *dentry);
-int ovl_copy_up_one(struct dentry *parent, struct dentry *dentry,
-		    struct path *lowerpath, struct kstat *stat,
-		    struct iattr *attr);
-int ovl_copy_xattr(struct dentry *old, struct dentry *new);
-int ovl_set_attr(struct dentry *upper, struct kstat *stat);

executer/kernel/mcoverlayfs/linux-3.10.0-327.36.1.el7/readdir.c

@@ -1,626 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/namei.h>
-#include <linux/file.h>
-#include <linux/xattr.h>
-#include <linux/rbtree.h>
-#include <linux/security.h>
-#include <linux/cred.h>
-#include <linux/version.h>
-#include "overlayfs.h"
-
-struct ovl_cache_entry {
-	unsigned int len;
-	unsigned int type;
-	u64 ino;
-	struct list_head l_node;
-	struct rb_node node;
-	struct ovl_cache_entry *next_maybe_whiteout;
-	bool is_whiteout;
-	char name[];
-};
-
-struct ovl_dir_cache {
-	long refcount;
-	u64 version;
-	struct list_head entries;
-};
-
-/* vfs_readdir vs. iterate_dir compat */
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(3, 11, 0) || \
-	(defined(RHEL_RELEASE_CODE) && RHEL_RELEASE_CODE >= RHEL_RELEASE_VERSION(7, 5))
-#define USE_ITERATE_DIR 1
-#endif
-
-#ifndef USE_ITERATE_DIR
-struct dir_context {
-	const filldir_t actor;
-	//loff_t pos;
-};
-#endif
-
-struct ovl_readdir_data {
-	struct dir_context ctx;
-	bool is_merge;
-	struct rb_root root;
-	struct list_head *list;
-	struct list_head middle;
-	struct ovl_cache_entry *first_maybe_whiteout;
-	int count;
-	int err;
-};
-
-struct ovl_dir_file {
-	bool is_real;
-	bool is_upper;
-	struct ovl_dir_cache *cache;
-	struct list_head *cursor;
-	struct file *realfile;
-	struct file *upperfile;
-};
-
-static struct ovl_cache_entry *ovl_cache_entry_from_node(struct rb_node *n)
-{
-	return container_of(n, struct ovl_cache_entry, node);
-}
-
-static struct ovl_cache_entry *ovl_cache_entry_find(struct rb_root *root,
-						    const char *name, int len)
-{
-	struct rb_node *node = root->rb_node;
-	int cmp;
-
-	while (node) {
-		struct ovl_cache_entry *p = ovl_cache_entry_from_node(node);
-
-		cmp = strncmp(name, p->name, len);
-		if (cmp > 0)
-			node = p->node.rb_right;
-		else if (cmp < 0 || len < p->len)
-			node = p->node.rb_left;
-		else
-			return p;
-	}
-
-	return NULL;
-}
-
-static struct ovl_cache_entry *ovl_cache_entry_new(struct ovl_readdir_data *rdd,
-						   const char *name, int len,
-						   u64 ino, unsigned int d_type)
-{
-	struct ovl_cache_entry *p;
-	size_t size = offsetof(struct ovl_cache_entry, name[len + 1]);
-
-	p = kmalloc(size, GFP_KERNEL);
-	if (!p)
-		return NULL;
-
-	memcpy(p->name, name, len);
-	p->name[len] = '\0';
-	p->len = len;
-	p->type = d_type;
-	p->ino = ino;
-	p->is_whiteout = false;
-
-	if (d_type == DT_CHR) {
-		p->next_maybe_whiteout = rdd->first_maybe_whiteout;
-		rdd->first_maybe_whiteout = p;
-	}
-	return p;
-}
-
-static int ovl_cache_entry_add_rb(struct ovl_readdir_data *rdd,
-				  const char *name, int len, u64 ino,
-				  unsigned int d_type)
-{
-	struct rb_node **newp = &rdd->root.rb_node;
-	struct rb_node *parent = NULL;
-	struct ovl_cache_entry *p;
-
-	while (*newp) {
-		int cmp;
-		struct ovl_cache_entry *tmp;
-
-		parent = *newp;
-		tmp = ovl_cache_entry_from_node(*newp);
-		cmp = strncmp(name, tmp->name, len);
-		if (cmp > 0)
-			newp = &tmp->node.rb_right;
-		else if (cmp < 0 || len < tmp->len)
-			newp = &tmp->node.rb_left;
-		else
-			return 0;
-	}
-
-	p = ovl_cache_entry_new(rdd, name, len, ino, d_type);
-	if (p == NULL)
-		return -ENOMEM;
-
-	list_add_tail(&p->l_node, rdd->list);
-	rb_link_node(&p->node, parent, newp);
-	rb_insert_color(&p->node, &rdd->root);
-
-	return 0;
-}
-
-static int ovl_fill_lower(struct ovl_readdir_data *rdd,
-			  const char *name, int namelen,
-			  loff_t offset, u64 ino, unsigned int d_type)
-{
-	struct ovl_cache_entry *p;
-
-	p = ovl_cache_entry_find(&rdd->root, name, namelen);
-	if (p) {
-		list_move_tail(&p->l_node, &rdd->middle);
-	} else {
-		p = ovl_cache_entry_new(rdd, name, namelen, ino, d_type);
-		if (p == NULL)
-			rdd->err = -ENOMEM;
-		else
-			list_add_tail(&p->l_node, &rdd->middle);
-	}
-
-	return rdd->err;
-}
-
-void ovl_cache_free(struct list_head *list)
-{
-	struct ovl_cache_entry *p;
-	struct ovl_cache_entry *n;
-
-	list_for_each_entry_safe(p, n, list, l_node)
-		kfree(p);
-
-	INIT_LIST_HEAD(list);
-}
-
-static void ovl_cache_put(struct ovl_dir_file *od, struct dentry *dentry)
-{
-	struct ovl_dir_cache *cache = od->cache;
-
-	WARN_ON(cache->refcount <= 0);
-	cache->refcount--;
-	if (!cache->refcount) {
-		if (ovl_dir_cache(dentry) == cache)
-			ovl_set_dir_cache(dentry, NULL);
-
-		ovl_cache_free(&cache->entries);
-		kfree(cache);
-	}
-}
-
-static int ovl_fill_merge(void *buf, const char *name, int namelen,
-			  loff_t offset, u64 ino, unsigned int d_type)
-{
-	struct dir_context *ctx = buf;
-	struct ovl_readdir_data *rdd =
-		container_of(ctx, struct ovl_readdir_data, ctx);
-
-	rdd->count++;
-	if (!rdd->is_merge)
-		return ovl_cache_entry_add_rb(rdd, name, namelen, ino, d_type);
-	else
-		return ovl_fill_lower(rdd, name, namelen, offset, ino, d_type);
-}
-
-static int ovl_check_whiteouts(struct dentry *dir, struct ovl_readdir_data *rdd)
-{
-	int err;
-	struct ovl_cache_entry *p;
-	struct dentry *dentry;
-	const struct cred *old_cred;
-	struct cred *override_cred;
-
-	override_cred = prepare_creds();
-	if (!override_cred)
-		return -ENOMEM;
-
-	/*
-	 * CAP_DAC_OVERRIDE for lookup
-	 */
-	cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-	old_cred = override_creds(override_cred);
-
-	err = mutex_lock_killable(&dir->d_inode->i_mutex);
-	if (!err) {
-		while (rdd->first_maybe_whiteout) {
-			p = rdd->first_maybe_whiteout;
-			rdd->first_maybe_whiteout = p->next_maybe_whiteout;
-			dentry = lookup_one_len(p->name, dir, p->len);
-			if (!IS_ERR(dentry)) {
-				p->is_whiteout = ovl_is_whiteout(dentry);
-				dput(dentry);
-			}
-		}
-		mutex_unlock(&dir->d_inode->i_mutex);
-	}
-	revert_creds(old_cred);
-	put_cred(override_cred);
-
-	return err;
-}
-
-static inline int ovl_dir_read(struct path *realpath,
-			       struct ovl_readdir_data *rdd)
-{
-	struct file *realfile;
-	int err;
-
-	realfile = ovl_path_open(realpath, O_RDONLY | O_DIRECTORY);
-	if (IS_ERR(realfile))
-		return PTR_ERR(realfile);
-
-	rdd->first_maybe_whiteout = NULL;
-	//rdd->ctx.pos = 0;
-	do {
-		rdd->count = 0;
-		rdd->err = 0;
-#ifdef USE_ITERATE_DIR
-		err = iterate_dir(realfile, &rdd->ctx);
-#else
-		err = vfs_readdir(realfile, rdd->ctx.actor, rdd);
-#endif
-		if (err >= 0)
-			err = rdd->err;
-	} while (!err && rdd->count);
-
-	if (!err && rdd->first_maybe_whiteout)
-		err = ovl_check_whiteouts(realpath->dentry, rdd);
-
-	fput(realfile);
-
-	return err;
-}
-
-static void ovl_dir_reset(struct file *file)
-{
-	struct ovl_dir_file *od = file->private_data;
-	struct ovl_dir_cache *cache = od->cache;
-	struct dentry *dentry = file->f_path.dentry;
-	enum ovl_path_type type = ovl_path_type(dentry);
-
-	if (cache && ovl_dentry_version_get(dentry) != cache->version) {
-		ovl_cache_put(od, dentry);
-		od->cache = NULL;
-		od->cursor = NULL;
-	}
-	WARN_ON(!od->is_real && !OVL_TYPE_MERGE(type));
-	if (od->is_real && OVL_TYPE_MERGE(type))
-		od->is_real = false;
-}
-
-static int ovl_dir_read_merged(struct dentry *dentry, struct list_head *list)
-{
-	int err;
-	struct path realpath;
-	struct ovl_readdir_data rdd = {
-		.ctx.actor = ovl_fill_merge,
-		.list = list,
-		.root = RB_ROOT,
-		.is_merge = false,
-	};
-	int idx, next;
-
-	for (idx = 0; idx != -1; idx = next) {
-		next = ovl_path_next(idx, dentry, &realpath);
-
-		if (next != -1) {
-			err = ovl_dir_read(&realpath, &rdd);
-			if (err)
-				break;
-		} else {
-			/*
-			 * Insert lowest layer entries before upper ones, this
-			 * allows offsets to be reasonably constant
-			 */
-			list_add(&rdd.middle, rdd.list);
-			rdd.is_merge = true;
-			err = ovl_dir_read(&realpath, &rdd);
-			list_del(&rdd.middle);
-		}
-	}
-	return err;
-}
-
-static void ovl_seek_cursor(struct ovl_dir_file *od, loff_t pos)
-{
-	struct list_head *p;
-	loff_t off = 0;
-
-	list_for_each(p, &od->cache->entries) {
-		if (off >= pos)
-			break;
-		off++;
-	}
-	/* Cursor is safe since the cache is stable */
-	od->cursor = p;
-}
-
-static struct ovl_dir_cache *ovl_cache_get(struct dentry *dentry)
-{
-	int res;
-	struct ovl_dir_cache *cache;
-
-	cache = ovl_dir_cache(dentry);
-	if (cache && ovl_dentry_version_get(dentry) == cache->version) {
-		cache->refcount++;
-		return cache;
-	}
-	ovl_set_dir_cache(dentry, NULL);
-
-	cache = kzalloc(sizeof(struct ovl_dir_cache), GFP_KERNEL);
-	if (!cache)
-		return ERR_PTR(-ENOMEM);
-
-	cache->refcount = 1;
-	INIT_LIST_HEAD(&cache->entries);
-
-	res = ovl_dir_read_merged(dentry, &cache->entries);
-	if (res) {
-		ovl_cache_free(&cache->entries);
-		kfree(cache);
-		return ERR_PTR(res);
-	}
-
-	cache->version = ovl_dentry_version_get(dentry);
-	ovl_set_dir_cache(dentry, cache);
-
-	return cache;
-}
-
-#ifdef USE_ITERATE_DIR
-struct iterate_wrapper {
-	struct dir_context ctx;
-	filldir_t actor;
-	void *buf;
-};
-
-static int ovl_wrap_readdir(void *ctx, const char *name, int namelen,
-			    loff_t offset, u64 ino, unsigned int d_type)
-{
-	struct iterate_wrapper *w = ctx;
-
-	return w->actor(w->buf, name, namelen, offset, ino, d_type);
-}
-#endif
-
-static int ovl_readdir(struct file *file, void *buf, filldir_t filler)
-{
-	struct ovl_dir_file *od = file->private_data;
-	struct dentry *dentry = file->f_path.dentry;
-	struct ovl_cache_entry *p;
-	int res;
-
-	if (!file->f_pos)
-		ovl_dir_reset(file);
-
-	if (od->is_real) {
-#ifdef USE_ITERATE_DIR
-		struct iterate_wrapper w = {
-			.ctx.actor = ovl_wrap_readdir,
-			.actor = filler,
-			.buf = buf,
-		};
-		res = iterate_dir(od->realfile, &w.ctx);
-#else
-		res = vfs_readdir(od->realfile, filler, buf);
-#endif
-		file->f_pos = od->realfile->f_pos;
-
-		return res;
-	}
-
-	if (!od->cache) {
-		struct ovl_dir_cache *cache;
-
-		cache = ovl_cache_get(dentry);
-		if (IS_ERR(cache))
-			return PTR_ERR(cache);
-
-		od->cache = cache;
-		ovl_seek_cursor(od, file->f_pos);
-	}
-
-	while (od->cursor != &od->cache->entries) {
-		p = list_entry(od->cursor, struct ovl_cache_entry, l_node);
-		if (!p->is_whiteout)
-			if (filler(buf, p->name, p->len, file->f_pos, p->ino, p->type))
-				break;
-		od->cursor = p->l_node.next;
-		file->f_pos++;
-	}
-	return 0;
-}
-
-static loff_t ovl_dir_llseek(struct file *file, loff_t offset, int origin)
-{
-	loff_t res;
-	struct ovl_dir_file *od = file->private_data;
-
-	mutex_lock(&file_inode(file)->i_mutex);
-	if (!file->f_pos)
-		ovl_dir_reset(file);
-
-	if (od->is_real) {
-		res = vfs_llseek(od->realfile, offset, origin);
-		file->f_pos = od->realfile->f_pos;
-	} else {
-		res = -EINVAL;
-
-		switch (origin) {
-		case SEEK_CUR:
-			offset += file->f_pos;
-			break;
-		case SEEK_SET:
-			break;
-		default:
-			goto out_unlock;
-		}
-		if (offset < 0)
-			goto out_unlock;
-
-		if (offset != file->f_pos) {
-			file->f_pos = offset;
-			if (od->cache)
-				ovl_seek_cursor(od, offset);
-		}
-		res = offset;
-	}
-out_unlock:
-	mutex_unlock(&file_inode(file)->i_mutex);
-
-	return res;
-}
-
-static int ovl_dir_fsync(struct file *file, loff_t start, loff_t end,
-			 int datasync)
-{
-	struct ovl_dir_file *od = file->private_data;
-	struct dentry *dentry = file->f_path.dentry;
-	struct file *realfile = od->realfile;
-
-	/*
-	 * Need to check if we started out being a lower dir, but got copied up
-	 */
-	if (!od->is_upper && OVL_TYPE_UPPER(ovl_path_type(dentry))) {
-		struct inode *inode = file_inode(file);
-
-		realfile = lockless_dereference(od->upperfile);
-		if (!realfile) {
-			struct path upperpath;
-
-			ovl_path_upper(dentry, &upperpath);
-			realfile = ovl_path_open(&upperpath, O_RDONLY);
-			smp_mb__before_spinlock();
-			mutex_lock(&inode->i_mutex);
-			if (!od->upperfile) {
-				if (IS_ERR(realfile)) {
-					mutex_unlock(&inode->i_mutex);
-					return PTR_ERR(realfile);
-				}
-				od->upperfile = realfile;
-			} else {
-				/* somebody has beaten us to it */
-				if (!IS_ERR(realfile))
-					fput(realfile);
-				realfile = od->upperfile;
-			}
-			mutex_unlock(&inode->i_mutex);
-		}
-	}
-
-	return vfs_fsync_range(realfile, start, end, datasync);
-}
-
-static int ovl_dir_release(struct inode *inode, struct file *file)
-{
-	struct ovl_dir_file *od = file->private_data;
-
-	if (od->cache) {
-		mutex_lock(&inode->i_mutex);
-		ovl_cache_put(od, file->f_path.dentry);
-		mutex_unlock(&inode->i_mutex);
-	}
-	fput(od->realfile);
-	if (od->upperfile)
-		fput(od->upperfile);
-	kfree(od);
-
-	return 0;
-}
-
-static int ovl_dir_open(struct inode *inode, struct file *file)
-{
-	struct path realpath;
-	struct file *realfile;
-	struct ovl_dir_file *od;
-	enum ovl_path_type type;
-
-	od = kzalloc(sizeof(struct ovl_dir_file), GFP_KERNEL);
-	if (!od)
-		return -ENOMEM;
-
-	type = ovl_path_real(file->f_path.dentry, &realpath);
-	realfile = ovl_path_open(&realpath, file->f_flags);
-	if (IS_ERR(realfile)) {
-		kfree(od);
-		return PTR_ERR(realfile);
-	}
-	od->realfile = realfile;
-	od->is_real = !OVL_TYPE_MERGE(type);
-	od->is_upper = OVL_TYPE_UPPER(type);
-	file->private_data = od;
-
-	return 0;
-}
-
-const struct file_operations ovl_dir_operations = {
-	.read		= generic_read_dir,
-	.open		= ovl_dir_open,
-	.readdir	= ovl_readdir,
-	.llseek		= ovl_dir_llseek,
-	.fsync		= ovl_dir_fsync,
-	.release	= ovl_dir_release,
-};
-
-int ovl_check_empty_dir(struct dentry *dentry, struct list_head *list)
-{
-	int err;
-	struct ovl_cache_entry *p;
-
-	err = ovl_dir_read_merged(dentry, list);
-	if (err)
-		return err;
-
-	err = 0;
-
-	list_for_each_entry(p, list, l_node) {
-		if (p->is_whiteout)
-			continue;
-
-		if (p->name[0] == '.') {
-			if (p->len == 1)
-				continue;
-			if (p->len == 2 && p->name[1] == '.')
-				continue;
-		}
-		err = -ENOTEMPTY;
-		break;
-	}
-
-	return err;
-}
-
-void ovl_cleanup_whiteouts(struct dentry *upper, struct list_head *list)
-{
-	struct ovl_cache_entry *p;
-
-	mutex_lock_nested(&upper->d_inode->i_mutex, I_MUTEX_CHILD);
-	list_for_each_entry(p, list, l_node) {
-		struct dentry *dentry;
-
-		if (!p->is_whiteout)
-			continue;
-
-		dentry = lookup_one_len(p->name, upper, p->len);
-		if (IS_ERR(dentry)) {
-			pr_err("overlayfs: lookup '%s/%.*s' failed (%i)\n",
-			       upper->d_name.name, p->len, p->name,
-			       (int) PTR_ERR(dentry));
-			continue;
-		}
-		ovl_cleanup(upper->d_inode, dentry);
-		dput(dentry);
-	}
-	mutex_unlock(&upper->d_inode->i_mutex);
-}

executer/kernel/mcoverlayfs/linux-4.0.9/copy_up.c

@@ -1,416 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/file.h>
-#include <linux/splice.h>
-#include <linux/xattr.h>
-#include <linux/security.h>
-#include <linux/uaccess.h>
-#include <linux/sched.h>
-#include <linux/namei.h>
-#include "overlayfs.h"
-
-#define OVL_COPY_UP_CHUNK_SIZE (1 << 20)
-
-int ovl_copy_xattr(struct dentry *old, struct dentry *new)
-{
-	ssize_t list_size, size;
-	char *buf, *name, *value;
-	int error;
-
-	if (!old->d_inode->i_op->getxattr ||
-	    !new->d_inode->i_op->getxattr)
-		return 0;
-
-	list_size = vfs_listxattr(old, NULL, 0);
-	if (list_size <= 0) {
-		if (list_size == -EOPNOTSUPP)
-			return 0;
-		return list_size;
-	}
-
-	buf = kzalloc(list_size, GFP_KERNEL);
-	if (!buf)
-		return -ENOMEM;
-
-	error = -ENOMEM;
-	value = kmalloc(XATTR_SIZE_MAX, GFP_KERNEL);
-	if (!value)
-		goto out;
-
-	list_size = vfs_listxattr(old, buf, list_size);
-	if (list_size <= 0) {
-		error = list_size;
-		goto out_free_value;
-	}
-
-	for (name = buf; name < (buf + list_size); name += strlen(name) + 1) {
-		size = vfs_getxattr(old, name, value, XATTR_SIZE_MAX);
-		if (size <= 0) {
-			error = size;
-			goto out_free_value;
-		}
-		error = vfs_setxattr(new, name, value, size, 0);
-		if (error)
-			goto out_free_value;
-	}
-
-out_free_value:
-	kfree(value);
-out:
-	kfree(buf);
-	return error;
-}
-
-static int ovl_copy_up_data(struct path *old, struct path *new, loff_t len)
-{
-	struct file *old_file;
-	struct file *new_file;
-	loff_t old_pos = 0;
-	loff_t new_pos = 0;
-	int error = 0;
-
-	if (len == 0)
-		return 0;
-
-	old_file = ovl_path_open(old, O_RDONLY);
-	if (IS_ERR(old_file))
-		return PTR_ERR(old_file);
-
-	new_file = ovl_path_open(new, O_WRONLY);
-	if (IS_ERR(new_file)) {
-		error = PTR_ERR(new_file);
-		goto out_fput;
-	}
-
-	/* FIXME: copy up sparse files efficiently */
-	while (len) {
-		size_t this_len = OVL_COPY_UP_CHUNK_SIZE;
-		long bytes;
-
-		if (len < this_len)
-			this_len = len;
-
-		if (signal_pending_state(TASK_KILLABLE, current)) {
-			error = -EINTR;
-			break;
-		}
-
-		bytes = do_splice_direct(old_file, &old_pos,
-					 new_file, &new_pos,
-					 this_len, SPLICE_F_MOVE);
-		if (bytes <= 0) {
-			error = bytes;
-			break;
-		}
-		WARN_ON(old_pos != new_pos);
-
-		len -= bytes;
-	}
-
-	fput(new_file);
-out_fput:
-	fput(old_file);
-	return error;
-}
-
-static char *ovl_read_symlink(struct dentry *realdentry)
-{
-	int res;
-	char *buf;
-	struct inode *inode = realdentry->d_inode;
-	mm_segment_t old_fs;
-
-	res = -EINVAL;
-	if (!inode->i_op->readlink)
-		goto err;
-
-	res = -ENOMEM;
-	buf = (char *) __get_free_page(GFP_KERNEL);
-	if (!buf)
-		goto err;
-
-	old_fs = get_fs();
-	set_fs(get_ds());
-	/* The cast to a user pointer is valid due to the set_fs() */
-	res = inode->i_op->readlink(realdentry,
-				    (char __user *)buf, PAGE_SIZE - 1);
-	set_fs(old_fs);
-	if (res < 0) {
-		free_page((unsigned long) buf);
-		goto err;
-	}
-	buf[res] = '\0';
-
-	return buf;
-
-err:
-	return ERR_PTR(res);
-}
-
-static int ovl_set_timestamps(struct dentry *upperdentry, struct kstat *stat)
-{
-	struct iattr attr = {
-		.ia_valid =
-		     ATTR_ATIME | ATTR_MTIME | ATTR_ATIME_SET | ATTR_MTIME_SET,
-		.ia_atime = stat->atime,
-		.ia_mtime = stat->mtime,
-	};
-
-	return notify_change(upperdentry, &attr, NULL);
-}
-
-int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat)
-{
-	int err = 0;
-
-	if (!S_ISLNK(stat->mode)) {
-		struct iattr attr = {
-			.ia_valid = ATTR_MODE,
-			.ia_mode = stat->mode,
-		};
-		err = notify_change(upperdentry, &attr, NULL);
-	}
-	if (!err) {
-		struct iattr attr = {
-			.ia_valid = ATTR_UID | ATTR_GID,
-			.ia_uid = stat->uid,
-			.ia_gid = stat->gid,
-		};
-		err = notify_change(upperdentry, &attr, NULL);
-	}
-	if (!err)
-		ovl_set_timestamps(upperdentry, stat);
-
-	return err;
-}
-
-static int ovl_copy_up_locked(struct dentry *workdir, struct dentry *upperdir,
-			      struct dentry *dentry, struct path *lowerpath,
-			      struct kstat *stat, struct iattr *attr,
-			      const char *link)
-{
-	struct inode *wdir = workdir->d_inode;
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *newdentry = NULL;
-	struct dentry *upper = NULL;
-	umode_t mode = stat->mode;
-	int err;
-
-	newdentry = ovl_lookup_temp(workdir, dentry);
-	err = PTR_ERR(newdentry);
-	if (IS_ERR(newdentry))
-		goto out;
-
-	upper = lookup_one_len(dentry->d_name.name, upperdir,
-			       dentry->d_name.len);
-	err = PTR_ERR(upper);
-	if (IS_ERR(upper))
-		goto out1;
-
-	/* Can't properly set mode on creation because of the umask */
-	stat->mode &= S_IFMT;
-	err = ovl_create_real(wdir, newdentry, stat, link, NULL, true);
-	stat->mode = mode;
-	if (err)
-		goto out2;
-
-	if (S_ISREG(stat->mode)) {
-		struct path upperpath;
-		ovl_path_upper(dentry, &upperpath);
-		BUG_ON(upperpath.dentry != NULL);
-		upperpath.dentry = newdentry;
-
-		err = ovl_copy_up_data(lowerpath, &upperpath, stat->size);
-		if (err)
-			goto out_cleanup;
-	}
-
-	err = ovl_copy_xattr(lowerpath->dentry, newdentry);
-	if (err)
-		goto out_cleanup;
-
-	mutex_lock(&newdentry->d_inode->i_mutex);
-	err = ovl_set_attr(newdentry, stat);
-	if (!err && attr)
-		err = notify_change(newdentry, attr, NULL);
-	mutex_unlock(&newdentry->d_inode->i_mutex);
-	if (err)
-		goto out_cleanup;
-
-	err = ovl_do_rename(wdir, newdentry, udir, upper, 0);
-	if (err)
-		goto out_cleanup;
-
-	ovl_dentry_update(dentry, newdentry);
-	newdentry = NULL;
-
-	/*
-	 * Non-directores become opaque when copied up.
-	 */
-	if (!S_ISDIR(stat->mode))
-		ovl_dentry_set_opaque(dentry, true);
-out2:
-	dput(upper);
-out1:
-	dput(newdentry);
-out:
-	return err;
-
-out_cleanup:
-	ovl_cleanup(wdir, newdentry);
-	goto out;
-}
-
-/*
- * Copy up a single dentry
- *
- * Directory renames only allowed on "pure upper" (already created on
- * upper filesystem, never copied up).  Directories which are on lower or
- * are merged may not be renamed.  For these -EXDEV is returned and
- * userspace has to deal with it.  This means, when copying up a
- * directory we can rely on it and ancestors being stable.
- *
- * Non-directory renames start with copy up of source if necessary.  The
- * actual rename will only proceed once the copy up was successful.  Copy
- * up uses upper parent i_mutex for exclusion.  Since rename can change
- * d_parent it is possible that the copy up will lock the old parent.  At
- * that point the file will have already been copied up anyway.
- */
-int ovl_copy_up_one(struct dentry *parent, struct dentry *dentry,
-		    struct path *lowerpath, struct kstat *stat,
-		    struct iattr *attr)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	int err;
-	struct kstat pstat;
-	struct path parentpath;
-	struct dentry *upperdir;
-	struct dentry *upperdentry;
-	const struct cred *old_cred;
-	struct cred *override_cred;
-	char *link = NULL;
-
-	if (WARN_ON(!workdir))
-		return -EROFS;
-
-	ovl_path_upper(parent, &parentpath);
-	upperdir = parentpath.dentry;
-
-	err = vfs_getattr(&parentpath, &pstat);
-	if (err)
-		return err;
-
-	if (S_ISLNK(stat->mode)) {
-		link = ovl_read_symlink(lowerpath->dentry);
-		if (IS_ERR(link))
-			return PTR_ERR(link);
-	}
-
-	err = -ENOMEM;
-	override_cred = prepare_creds();
-	if (!override_cred)
-		goto out_free_link;
-
-	override_cred->fsuid = stat->uid;
-	override_cred->fsgid = stat->gid;
-	/*
-	 * CAP_SYS_ADMIN for copying up extended attributes
-	 * CAP_DAC_OVERRIDE for create
-	 * CAP_FOWNER for chmod, timestamp update
-	 * CAP_FSETID for chmod
-	 * CAP_CHOWN for chown
-	 * CAP_MKNOD for mknod
-	 */
-	cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-	cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-	cap_raise(override_cred->cap_effective, CAP_FOWNER);
-	cap_raise(override_cred->cap_effective, CAP_FSETID);
-	cap_raise(override_cred->cap_effective, CAP_CHOWN);
-	cap_raise(override_cred->cap_effective, CAP_MKNOD);
-	old_cred = override_creds(override_cred);
-
-	err = -EIO;
-	if (lock_rename(workdir, upperdir) != NULL) {
-		pr_err("overlayfs: failed to lock workdir+upperdir\n");
-		goto out_unlock;
-	}
-	upperdentry = ovl_dentry_upper(dentry);
-	if (upperdentry) {
-		unlock_rename(workdir, upperdir);
-		err = 0;
-		/* Raced with another copy-up?  Do the setattr here */
-		if (attr) {
-			mutex_lock(&upperdentry->d_inode->i_mutex);
-			err = notify_change(upperdentry, attr, NULL);
-			mutex_unlock(&upperdentry->d_inode->i_mutex);
-		}
-		goto out_put_cred;
-	}
-
-	err = ovl_copy_up_locked(workdir, upperdir, dentry, lowerpath,
-				 stat, attr, link);
-	if (!err) {
-		/* Restore timestamps on parent (best effort) */
-		ovl_set_timestamps(upperdir, &pstat);
-	}
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out_put_cred:
-	revert_creds(old_cred);
-	put_cred(override_cred);
-
-out_free_link:
-	if (link)
-		free_page((unsigned long) link);
-
-	return err;
-}
-
-int ovl_copy_up(struct dentry *dentry)
-{
-	int err;
-
-	err = 0;
-	while (!err) {
-		struct dentry *next;
-		struct dentry *parent;
-		struct path lowerpath;
-		struct kstat stat;
-		enum ovl_path_type type = ovl_path_type(dentry);
-
-		if (OVL_TYPE_UPPER(type))
-			break;
-
-		next = dget(dentry);
-		/* find the topmost dentry not yet copied up */
-		for (;;) {
-			parent = dget_parent(next);
-
-			type = ovl_path_type(parent);
-			if (OVL_TYPE_UPPER(type))
-				break;
-
-			dput(next);
-			next = parent;
-		}
-
-		ovl_path_lower(next, &lowerpath);
-		err = vfs_getattr(&lowerpath, &stat);
-		if (!err)
-			err = ovl_copy_up_one(parent, next, &lowerpath, &stat, NULL);
-
-		dput(parent);
-		dput(next);
-	}
-
-	return err;
-}

executer/kernel/mcoverlayfs/linux-4.0.9/dir.c

@@ -1,951 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/namei.h>
-#include <linux/xattr.h>
-#include <linux/security.h>
-#include <linux/cred.h>
-#include "overlayfs.h"
-
-void ovl_cleanup(struct inode *wdir, struct dentry *wdentry)
-{
-	int err;
-
-	dget(wdentry);
-	if (d_is_dir(wdentry))
-		err = ovl_do_rmdir(wdir, wdentry);
-	else
-		err = ovl_do_unlink(wdir, wdentry);
-	dput(wdentry);
-
-	if (err) {
-		pr_err("overlayfs: cleanup of '%pd2' failed (%i)\n",
-		       wdentry, err);
-	}
-}
-
-struct dentry *ovl_lookup_temp(struct dentry *workdir, struct dentry *dentry)
-{
-	struct dentry *temp;
-	char name[20];
-
-	snprintf(name, sizeof(name), "#%lx", (unsigned long) dentry);
-
-	temp = lookup_one_len(name, workdir, strlen(name));
-	if (!IS_ERR(temp) && temp->d_inode) {
-		pr_err("overlayfs: workdir/%s already exists\n", name);
-		dput(temp);
-		temp = ERR_PTR(-EIO);
-	}
-
-	return temp;
-}
-
-/* caller holds i_mutex on workdir */
-static struct dentry *ovl_whiteout(struct dentry *workdir,
-				   struct dentry *dentry)
-{
-	int err;
-	struct dentry *whiteout;
-	struct inode *wdir = workdir->d_inode;
-
-	whiteout = ovl_lookup_temp(workdir, dentry);
-	if (IS_ERR(whiteout))
-		return whiteout;
-
-	err = ovl_do_whiteout(wdir, whiteout);
-	if (err) {
-		dput(whiteout);
-		whiteout = ERR_PTR(err);
-	}
-
-	return whiteout;
-}
-
-int ovl_create_real(struct inode *dir, struct dentry *newdentry,
-		    struct kstat *stat, const char *link,
-		    struct dentry *hardlink, bool debug)
-{
-	int err;
-
-	if (newdentry->d_inode)
-		return -ESTALE;
-
-	if (hardlink) {
-		err = ovl_do_link(hardlink, dir, newdentry, debug);
-	} else {
-		switch (stat->mode & S_IFMT) {
-		case S_IFREG:
-			err = ovl_do_create(dir, newdentry, stat->mode, debug);
-			break;
-
-		case S_IFDIR:
-			err = ovl_do_mkdir(dir, newdentry, stat->mode, debug);
-			break;
-
-		case S_IFCHR:
-		case S_IFBLK:
-		case S_IFIFO:
-		case S_IFSOCK:
-			err = ovl_do_mknod(dir, newdentry,
-					   stat->mode, stat->rdev, debug);
-			break;
-
-		case S_IFLNK:
-			err = ovl_do_symlink(dir, newdentry, link, debug);
-			break;
-
-		default:
-			err = -EPERM;
-		}
-	}
-	if (!err && WARN_ON(!newdentry->d_inode)) {
-		/*
-		 * Not quite sure if non-instantiated dentry is legal or not.
-		 * VFS doesn't seem to care so check and warn here.
-		 */
-		err = -ENOENT;
-	}
-	return err;
-}
-
-static int ovl_set_opaque(struct dentry *upperdentry)
-{
-	return ovl_do_setxattr(upperdentry, OVL_XATTR_OPAQUE, "y", 1, 0);
-}
-
-static void ovl_remove_opaque(struct dentry *upperdentry)
-{
-	int err;
-
-	err = ovl_do_removexattr(upperdentry, OVL_XATTR_OPAQUE);
-	if (err) {
-		pr_warn("overlayfs: failed to remove opaque from '%s' (%i)\n",
-			upperdentry->d_name.name, err);
-	}
-}
-
-static int ovl_dir_getattr(struct vfsmount *mnt, struct dentry *dentry,
-			 struct kstat *stat)
-{
-	int err;
-	enum ovl_path_type type;
-	struct path realpath;
-
-	type = ovl_path_real(dentry, &realpath);
-	err = vfs_getattr(&realpath, stat);
-	if (err)
-		return err;
-
-	stat->dev = dentry->d_sb->s_dev;
-	stat->ino = dentry->d_inode->i_ino;
-
-	/*
-	 * It's probably not worth it to count subdirs to get the
-	 * correct link count.  nlink=1 seems to pacify 'find' and
-	 * other utilities.
-	 */
-	if (OVL_TYPE_MERGE(type))
-		stat->nlink = 1;
-
-	return 0;
-}
-
-static int ovl_create_upper(struct dentry *dentry, struct inode *inode,
-			    struct kstat *stat, const char *link,
-			    struct dentry *hardlink)
-{
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *newdentry;
-	int err;
-
-	mutex_lock_nested(&udir->i_mutex, I_MUTEX_PARENT);
-	newdentry = lookup_one_len(dentry->d_name.name, upperdir,
-				   dentry->d_name.len);
-	err = PTR_ERR(newdentry);
-	if (IS_ERR(newdentry))
-		goto out_unlock;
-	err = ovl_create_real(udir, newdentry, stat, link, hardlink, false);
-	if (err)
-		goto out_dput;
-
-	ovl_dentry_version_inc(dentry->d_parent);
-	ovl_dentry_update(dentry, newdentry);
-	ovl_copyattr(newdentry->d_inode, inode);
-	d_instantiate(dentry, inode);
-	newdentry = NULL;
-out_dput:
-	dput(newdentry);
-out_unlock:
-	mutex_unlock(&udir->i_mutex);
-	return err;
-}
-
-static int ovl_lock_rename_workdir(struct dentry *workdir,
-				   struct dentry *upperdir)
-{
-	/* Workdir should not be the same as upperdir */
-	if (workdir == upperdir)
-		goto err;
-
-	/* Workdir should not be subdir of upperdir and vice versa */
-	if (lock_rename(workdir, upperdir) != NULL)
-		goto err_unlock;
-
-	return 0;
-
-err_unlock:
-	unlock_rename(workdir, upperdir);
-err:
-	pr_err("overlayfs: failed to lock workdir+upperdir\n");
-	return -EIO;
-}
-
-static struct dentry *ovl_clear_empty(struct dentry *dentry,
-				      struct list_head *list)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	struct inode *wdir = workdir->d_inode;
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct path upperpath;
-	struct dentry *upper;
-	struct dentry *opaquedir;
-	struct kstat stat;
-	int err;
-
-	if (WARN_ON(!workdir))
-		return ERR_PTR(-EROFS);
-
-	err = ovl_lock_rename_workdir(workdir, upperdir);
-	if (err)
-		goto out;
-
-	ovl_path_upper(dentry, &upperpath);
-	err = vfs_getattr(&upperpath, &stat);
-	if (err)
-		goto out_unlock;
-
-	err = -ESTALE;
-	if (!S_ISDIR(stat.mode))
-		goto out_unlock;
-	upper = upperpath.dentry;
-	if (upper->d_parent->d_inode != udir)
-		goto out_unlock;
-
-	opaquedir = ovl_lookup_temp(workdir, dentry);
-	err = PTR_ERR(opaquedir);
-	if (IS_ERR(opaquedir))
-		goto out_unlock;
-
-	err = ovl_create_real(wdir, opaquedir, &stat, NULL, NULL, true);
-	if (err)
-		goto out_dput;
-
-	err = ovl_copy_xattr(upper, opaquedir);
-	if (err)
-		goto out_cleanup;
-
-	err = ovl_set_opaque(opaquedir);
-	if (err)
-		goto out_cleanup;
-
-	mutex_lock(&opaquedir->d_inode->i_mutex);
-	err = ovl_set_attr(opaquedir, &stat);
-	mutex_unlock(&opaquedir->d_inode->i_mutex);
-	if (err)
-		goto out_cleanup;
-
-	err = ovl_do_rename(wdir, opaquedir, udir, upper, RENAME_EXCHANGE);
-	if (err)
-		goto out_cleanup;
-
-	ovl_cleanup_whiteouts(upper, list);
-	ovl_cleanup(wdir, upper);
-	unlock_rename(workdir, upperdir);
-
-	/* dentry's upper doesn't match now, get rid of it */
-	d_drop(dentry);
-
-	return opaquedir;
-
-out_cleanup:
-	ovl_cleanup(wdir, opaquedir);
-out_dput:
-	dput(opaquedir);
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out:
-	return ERR_PTR(err);
-}
-
-static struct dentry *ovl_check_empty_and_clear(struct dentry *dentry)
-{
-	int err;
-	struct dentry *ret = NULL;
-	LIST_HEAD(list);
-
-	err = ovl_check_empty_dir(dentry, &list);
-	if (err)
-		ret = ERR_PTR(err);
-	else {
-		/*
-		 * If no upperdentry then skip clearing whiteouts.
-		 *
-		 * Can race with copy-up, since we don't hold the upperdir
-		 * mutex.  Doesn't matter, since copy-up can't create a
-		 * non-empty directory from an empty one.
-		 */
-		if (ovl_dentry_upper(dentry))
-			ret = ovl_clear_empty(dentry, &list);
-	}
-
-	ovl_cache_free(&list);
-
-	return ret;
-}
-
-static int ovl_create_over_whiteout(struct dentry *dentry, struct inode *inode,
-				    struct kstat *stat, const char *link,
-				    struct dentry *hardlink)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	struct inode *wdir = workdir->d_inode;
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *upper;
-	struct dentry *newdentry;
-	int err;
-
-	if (WARN_ON(!workdir))
-		return -EROFS;
-
-	err = ovl_lock_rename_workdir(workdir, upperdir);
-	if (err)
-		goto out;
-
-	newdentry = ovl_lookup_temp(workdir, dentry);
-	err = PTR_ERR(newdentry);
-	if (IS_ERR(newdentry))
-		goto out_unlock;
-
-	upper = lookup_one_len(dentry->d_name.name, upperdir,
-			       dentry->d_name.len);
-	err = PTR_ERR(upper);
-	if (IS_ERR(upper))
-		goto out_dput;
-
-	err = ovl_create_real(wdir, newdentry, stat, link, hardlink, true);
-	if (err)
-		goto out_dput2;
-
-	if (S_ISDIR(stat->mode)) {
-		err = ovl_set_opaque(newdentry);
-		if (err)
-			goto out_cleanup;
-
-		err = ovl_do_rename(wdir, newdentry, udir, upper,
-				    RENAME_EXCHANGE);
-		if (err)
-			goto out_cleanup;
-
-		ovl_cleanup(wdir, upper);
-	} else {
-		err = ovl_do_rename(wdir, newdentry, udir, upper, 0);
-		if (err)
-			goto out_cleanup;
-	}
-	ovl_dentry_version_inc(dentry->d_parent);
-	ovl_dentry_update(dentry, newdentry);
-	ovl_copyattr(newdentry->d_inode, inode);
-	d_instantiate(dentry, inode);
-	newdentry = NULL;
-out_dput2:
-	dput(upper);
-out_dput:
-	dput(newdentry);
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out:
-	return err;
-
-out_cleanup:
-	ovl_cleanup(wdir, newdentry);
-	goto out_dput2;
-}
-
-static int ovl_create_or_link(struct dentry *dentry, int mode, dev_t rdev,
-			      const char *link, struct dentry *hardlink)
-{
-	int err;
-	struct inode *inode;
-	struct kstat stat = {
-		.mode = mode,
-		.rdev = rdev,
-	};
-
-	err = -ENOMEM;
-	inode = ovl_new_inode(dentry->d_sb, mode, dentry->d_fsdata);
-	if (!inode)
-		goto out;
-
-	err = ovl_copy_up(dentry->d_parent);
-	if (err)
-		goto out_iput;
-
-	if (!ovl_dentry_is_opaque(dentry)) {
-		err = ovl_create_upper(dentry, inode, &stat, link, hardlink);
-	} else {
-		const struct cred *old_cred;
-		struct cred *override_cred;
-
-		err = -ENOMEM;
-		override_cred = prepare_creds();
-		if (!override_cred)
-			goto out_iput;
-
-		/*
-		 * CAP_SYS_ADMIN for setting opaque xattr
-		 * CAP_DAC_OVERRIDE for create in workdir, rename
-		 * CAP_FOWNER for removing whiteout from sticky dir
-		 */
-		cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		cap_raise(override_cred->cap_effective, CAP_FOWNER);
-		old_cred = override_creds(override_cred);
-
-		err = ovl_create_over_whiteout(dentry, inode, &stat, link,
-					       hardlink);
-
-		revert_creds(old_cred);
-		put_cred(override_cred);
-	}
-
-	if (!err)
-		inode = NULL;
-out_iput:
-	iput(inode);
-out:
-	return err;
-}
-
-static int ovl_create_object(struct dentry *dentry, int mode, dev_t rdev,
-			     const char *link)
-{
-	int err;
-
-	err = ovl_want_write(dentry);
-	if (!err) {
-		err = ovl_create_or_link(dentry, mode, rdev, link, NULL);
-		ovl_drop_write(dentry);
-	}
-
-	return err;
-}
-
-static int ovl_create(struct inode *dir, struct dentry *dentry, umode_t mode,
-		      bool excl)
-{
-	return ovl_create_object(dentry, (mode & 07777) | S_IFREG, 0, NULL);
-}
-
-static int ovl_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
-{
-	return ovl_create_object(dentry, (mode & 07777) | S_IFDIR, 0, NULL);
-}
-
-static int ovl_mknod(struct inode *dir, struct dentry *dentry, umode_t mode,
-		     dev_t rdev)
-{
-	/* Don't allow creation of "whiteout" on overlay */
-	if (S_ISCHR(mode) && rdev == WHITEOUT_DEV)
-		return -EPERM;
-
-	return ovl_create_object(dentry, mode, rdev, NULL);
-}
-
-static int ovl_symlink(struct inode *dir, struct dentry *dentry,
-		       const char *link)
-{
-	return ovl_create_object(dentry, S_IFLNK, 0, link);
-}
-
-static int ovl_link(struct dentry *old, struct inode *newdir,
-		    struct dentry *new)
-{
-	int err;
-	struct dentry *upper;
-
-	err = ovl_want_write(old);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(old);
-	if (err)
-		goto out_drop_write;
-
-	upper = ovl_dentry_upper(old);
-	err = ovl_create_or_link(new, upper->d_inode->i_mode, 0, NULL, upper);
-
-out_drop_write:
-	ovl_drop_write(old);
-out:
-	return err;
-}
-
-static int ovl_remove_and_whiteout(struct dentry *dentry, bool is_dir)
-{
-	struct dentry *workdir = ovl_workdir(dentry);
-	struct inode *wdir = workdir->d_inode;
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *udir = upperdir->d_inode;
-	struct dentry *whiteout;
-	struct dentry *upper;
-	struct dentry *opaquedir = NULL;
-	int err;
-
-	if (WARN_ON(!workdir))
-		return -EROFS;
-
-	if (is_dir) {
-		if (OVL_TYPE_MERGE_OR_LOWER(ovl_path_type(dentry))) {
-			opaquedir = ovl_check_empty_and_clear(dentry);
-			err = PTR_ERR(opaquedir);
-			if (IS_ERR(opaquedir))
-				goto out;
-		} else {
-			LIST_HEAD(list);
-
-			/*
-			 * When removing an empty opaque directory, then it
-			 * makes no sense to replace it with an exact replica of
-			 * itself.  But emptiness still needs to be checked.
-			 */
-			err = ovl_check_empty_dir(dentry, &list);
-			ovl_cache_free(&list);
-			if (err)
-				goto out;
-		}
-	}
-
-	err = ovl_lock_rename_workdir(workdir, upperdir);
-	if (err)
-		goto out_dput;
-
-	whiteout = ovl_whiteout(workdir, dentry);
-	err = PTR_ERR(whiteout);
-	if (IS_ERR(whiteout))
-		goto out_unlock;
-
-	upper = ovl_dentry_upper(dentry);
-	if (!upper) {
-		upper = lookup_one_len(dentry->d_name.name, upperdir,
-				       dentry->d_name.len);
-		err = PTR_ERR(upper);
-		if (IS_ERR(upper))
-			goto kill_whiteout;
-
-		err = ovl_do_rename(wdir, whiteout, udir, upper, 0);
-		dput(upper);
-		if (err)
-			goto kill_whiteout;
-	} else {
-		int flags = 0;
-
-		if (opaquedir)
-			upper = opaquedir;
-		err = -ESTALE;
-		if (upper->d_parent != upperdir)
-			goto kill_whiteout;
-
-		if (is_dir)
-			flags |= RENAME_EXCHANGE;
-
-		err = ovl_do_rename(wdir, whiteout, udir, upper, flags);
-		if (err)
-			goto kill_whiteout;
-
-		if (is_dir)
-			ovl_cleanup(wdir, upper);
-	}
-	ovl_dentry_version_inc(dentry->d_parent);
-out_d_drop:
-	d_drop(dentry);
-	dput(whiteout);
-out_unlock:
-	unlock_rename(workdir, upperdir);
-out_dput:
-	dput(opaquedir);
-out:
-	return err;
-
-kill_whiteout:
-	ovl_cleanup(wdir, whiteout);
-	goto out_d_drop;
-}
-
-static int ovl_remove_upper(struct dentry *dentry, bool is_dir)
-{
-	struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent);
-	struct inode *dir = upperdir->d_inode;
-	struct dentry *upper = ovl_dentry_upper(dentry);
-	int err;
-
-	mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
-	err = -ESTALE;
-	if (upper->d_parent == upperdir) {
-		/* Don't let d_delete() think it can reset d_inode */
-		dget(upper);
-		if (is_dir)
-			err = vfs_rmdir(dir, upper);
-		else
-			err = vfs_unlink(dir, upper, NULL);
-		dput(upper);
-		ovl_dentry_version_inc(dentry->d_parent);
-	}
-
-	/*
-	 * Keeping this dentry hashed would mean having to release
-	 * upperpath/lowerpath, which could only be done if we are the
-	 * sole user of this dentry.  Too tricky...  Just unhash for
-	 * now.
-	 */
-	d_drop(dentry);
-	mutex_unlock(&dir->i_mutex);
-
-	return err;
-}
-
-static inline int ovl_check_sticky(struct dentry *dentry)
-{
-	struct inode *dir = ovl_dentry_real(dentry->d_parent)->d_inode;
-	struct inode *inode = ovl_dentry_real(dentry)->d_inode;
-
-	if (check_sticky(dir, inode))
-		return -EPERM;
-
-	return 0;
-}
-
-static int ovl_do_remove(struct dentry *dentry, bool is_dir)
-{
-	enum ovl_path_type type;
-	int err;
-
-	err = ovl_check_sticky(dentry);
-	if (err)
-		goto out;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(dentry->d_parent);
-	if (err)
-		goto out_drop_write;
-
-	type = ovl_path_type(dentry);
-	if (OVL_TYPE_PURE_UPPER(type)) {
-		err = ovl_remove_upper(dentry, is_dir);
-	} else {
-		const struct cred *old_cred;
-		struct cred *override_cred;
-
-		err = -ENOMEM;
-		override_cred = prepare_creds();
-		if (!override_cred)
-			goto out_drop_write;
-
-		/*
-		 * CAP_SYS_ADMIN for setting xattr on whiteout, opaque dir
-		 * CAP_DAC_OVERRIDE for create in workdir, rename
-		 * CAP_FOWNER for removing whiteout from sticky dir
-		 * CAP_FSETID for chmod of opaque dir
-		 * CAP_CHOWN for chown of opaque dir
-		 */
-		cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		cap_raise(override_cred->cap_effective, CAP_FOWNER);
-		cap_raise(override_cred->cap_effective, CAP_FSETID);
-		cap_raise(override_cred->cap_effective, CAP_CHOWN);
-		old_cred = override_creds(override_cred);
-
-		err = ovl_remove_and_whiteout(dentry, is_dir);
-
-		revert_creds(old_cred);
-		put_cred(override_cred);
-	}
-out_drop_write:
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static int ovl_unlink(struct inode *dir, struct dentry *dentry)
-{
-	return ovl_do_remove(dentry, false);
-}
-
-static int ovl_rmdir(struct inode *dir, struct dentry *dentry)
-{
-	return ovl_do_remove(dentry, true);
-}
-
-static int ovl_rename2(struct inode *olddir, struct dentry *old,
-		       struct inode *newdir, struct dentry *new,
-		       unsigned int flags)
-{
-	int err;
-	enum ovl_path_type old_type;
-	enum ovl_path_type new_type;
-	struct dentry *old_upperdir;
-	struct dentry *new_upperdir;
-	struct dentry *olddentry;
-	struct dentry *newdentry;
-	struct dentry *trap;
-	bool old_opaque;
-	bool new_opaque;
-	bool new_create = false;
-	bool cleanup_whiteout = false;
-	bool overwrite = !(flags & RENAME_EXCHANGE);
-	bool is_dir = d_is_dir(old);
-	bool new_is_dir = false;
-	struct dentry *opaquedir = NULL;
-	const struct cred *old_cred = NULL;
-	struct cred *override_cred = NULL;
-
-	err = -EINVAL;
-	if (flags & ~(RENAME_EXCHANGE | RENAME_NOREPLACE))
-		goto out;
-
-	flags &= ~RENAME_NOREPLACE;
-
-	err = ovl_check_sticky(old);
-	if (err)
-		goto out;
-
-	/* Don't copy up directory trees */
-	old_type = ovl_path_type(old);
-	err = -EXDEV;
-	if (OVL_TYPE_MERGE_OR_LOWER(old_type) && is_dir)
-		goto out;
-
-	if (new->d_inode) {
-		err = ovl_check_sticky(new);
-		if (err)
-			goto out;
-
-		if (d_is_dir(new))
-			new_is_dir = true;
-
-		new_type = ovl_path_type(new);
-		err = -EXDEV;
-		if (!overwrite && OVL_TYPE_MERGE_OR_LOWER(new_type) && new_is_dir)
-			goto out;
-
-		err = 0;
-		if (!OVL_TYPE_UPPER(new_type) && !OVL_TYPE_UPPER(old_type)) {
-			if (ovl_dentry_lower(old)->d_inode ==
-			    ovl_dentry_lower(new)->d_inode)
-				goto out;
-		}
-		if (OVL_TYPE_UPPER(new_type) && OVL_TYPE_UPPER(old_type)) {
-			if (ovl_dentry_upper(old)->d_inode ==
-			    ovl_dentry_upper(new)->d_inode)
-				goto out;
-		}
-	} else {
-		if (ovl_dentry_is_opaque(new))
-			new_type = __OVL_PATH_UPPER;
-		else
-			new_type = __OVL_PATH_UPPER | __OVL_PATH_PURE;
-	}
-
-	err = ovl_want_write(old);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(old);
-	if (err)
-		goto out_drop_write;
-
-	err = ovl_copy_up(new->d_parent);
-	if (err)
-		goto out_drop_write;
-	if (!overwrite) {
-		err = ovl_copy_up(new);
-		if (err)
-			goto out_drop_write;
-	}
-
-	old_opaque = !OVL_TYPE_PURE_UPPER(old_type);
-	new_opaque = !OVL_TYPE_PURE_UPPER(new_type);
-
-	if (old_opaque || new_opaque) {
-		err = -ENOMEM;
-		override_cred = prepare_creds();
-		if (!override_cred)
-			goto out_drop_write;
-
-		/*
-		 * CAP_SYS_ADMIN for setting xattr on whiteout, opaque dir
-		 * CAP_DAC_OVERRIDE for create in workdir
-		 * CAP_FOWNER for removing whiteout from sticky dir
-		 * CAP_FSETID for chmod of opaque dir
-		 * CAP_CHOWN for chown of opaque dir
-		 */
-		cap_raise(override_cred->cap_effective, CAP_SYS_ADMIN);
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		cap_raise(override_cred->cap_effective, CAP_FOWNER);
-		cap_raise(override_cred->cap_effective, CAP_FSETID);
-		cap_raise(override_cred->cap_effective, CAP_CHOWN);
-		old_cred = override_creds(override_cred);
-	}
-
-	if (overwrite && OVL_TYPE_MERGE_OR_LOWER(new_type) && new_is_dir) {
-		opaquedir = ovl_check_empty_and_clear(new);
-		err = PTR_ERR(opaquedir);
-		if (IS_ERR(opaquedir)) {
-			opaquedir = NULL;
-			goto out_revert_creds;
-		}
-	}
-
-	if (overwrite) {
-		if (old_opaque) {
-			if (new->d_inode || !new_opaque) {
-				/* Whiteout source */
-				flags |= RENAME_WHITEOUT;
-			} else {
-				/* Switch whiteouts */
-				flags |= RENAME_EXCHANGE;
-			}
-		} else if (is_dir && !new->d_inode && new_opaque) {
-			flags |= RENAME_EXCHANGE;
-			cleanup_whiteout = true;
-		}
-	}
-
-	old_upperdir = ovl_dentry_upper(old->d_parent);
-	new_upperdir = ovl_dentry_upper(new->d_parent);
-
-	trap = lock_rename(new_upperdir, old_upperdir);
-
-	olddentry = ovl_dentry_upper(old);
-	newdentry = ovl_dentry_upper(new);
-	if (newdentry) {
-		if (opaquedir) {
-			newdentry = opaquedir;
-			opaquedir = NULL;
-		} else {
-			dget(newdentry);
-		}
-	} else {
-		new_create = true;
-		newdentry = lookup_one_len(new->d_name.name, new_upperdir,
-					   new->d_name.len);
-		err = PTR_ERR(newdentry);
-		if (IS_ERR(newdentry))
-			goto out_unlock;
-	}
-
-	err = -ESTALE;
-	if (olddentry->d_parent != old_upperdir)
-		goto out_dput;
-	if (newdentry->d_parent != new_upperdir)
-		goto out_dput;
-	if (olddentry == trap)
-		goto out_dput;
-	if (newdentry == trap)
-		goto out_dput;
-
-	if (is_dir && !old_opaque && new_opaque) {
-		err = ovl_set_opaque(olddentry);
-		if (err)
-			goto out_dput;
-	}
-	if (!overwrite && new_is_dir && old_opaque && !new_opaque) {
-		err = ovl_set_opaque(newdentry);
-		if (err)
-			goto out_dput;
-	}
-
-	if (old_opaque || new_opaque) {
-		err = ovl_do_rename(old_upperdir->d_inode, olddentry,
-				    new_upperdir->d_inode, newdentry,
-				    flags);
-	} else {
-		/* No debug for the plain case */
-		BUG_ON(flags & ~RENAME_EXCHANGE);
-		err = vfs_rename(old_upperdir->d_inode, olddentry,
-				 new_upperdir->d_inode, newdentry,
-				 NULL, flags);
-	}
-
-	if (err) {
-		if (is_dir && !old_opaque && new_opaque)
-			ovl_remove_opaque(olddentry);
-		if (!overwrite && new_is_dir && old_opaque && !new_opaque)
-			ovl_remove_opaque(newdentry);
-		goto out_dput;
-	}
-
-	if (is_dir && old_opaque && !new_opaque)
-		ovl_remove_opaque(olddentry);
-	if (!overwrite && new_is_dir && !old_opaque && new_opaque)
-		ovl_remove_opaque(newdentry);
-
-	if (old_opaque != new_opaque) {
-		ovl_dentry_set_opaque(old, new_opaque);
-		if (!overwrite)
-			ovl_dentry_set_opaque(new, old_opaque);
-	}
-
-	if (cleanup_whiteout)
-		ovl_cleanup(old_upperdir->d_inode, newdentry);
-
-	ovl_dentry_version_inc(old->d_parent);
-	ovl_dentry_version_inc(new->d_parent);
-
-out_dput:
-	dput(newdentry);
-out_unlock:
-	unlock_rename(new_upperdir, old_upperdir);
-out_revert_creds:
-	if (old_opaque || new_opaque) {
-		revert_creds(old_cred);
-		put_cred(override_cred);
-	}
-out_drop_write:
-	ovl_drop_write(old);
-out:
-	dput(opaquedir);
-	return err;
-}
-
-const struct inode_operations ovl_dir_inode_operations = {
-	.lookup		= ovl_lookup,
-	.mkdir		= ovl_mkdir,
-	.symlink	= ovl_symlink,
-	.unlink		= ovl_unlink,
-	.rmdir		= ovl_rmdir,
-	.rename2	= ovl_rename2,
-	.link		= ovl_link,
-	.setattr	= ovl_setattr,
-	.create		= ovl_create,
-	.mknod		= ovl_mknod,
-	.permission	= ovl_permission,
-	.getattr	= ovl_dir_getattr,
-	.setxattr	= ovl_setxattr,
-	.getxattr	= ovl_getxattr,
-	.listxattr	= ovl_listxattr,
-	.removexattr	= ovl_removexattr,
-};

executer/kernel/mcoverlayfs/linux-4.0.9/inode.c

@@ -1,438 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/xattr.h>
-#include "overlayfs.h"
-
-static int ovl_copy_up_last(struct dentry *dentry, struct iattr *attr,
-			    bool no_data)
-{
-	int err;
-	struct dentry *parent;
-	struct kstat stat;
-	struct path lowerpath;
-
-	parent = dget_parent(dentry);
-	err = ovl_copy_up(parent);
-	if (err)
-		goto out_dput_parent;
-
-	ovl_path_lower(dentry, &lowerpath);
-	err = vfs_getattr(&lowerpath, &stat);
-	if (err)
-		goto out_dput_parent;
-
-	if (no_data)
-		stat.size = 0;
-
-	err = ovl_copy_up_one(parent, dentry, &lowerpath, &stat, attr);
-
-out_dput_parent:
-	dput(parent);
-	return err;
-}
-
-int ovl_setattr(struct dentry *dentry, struct iattr *attr)
-{
-	int err;
-	struct dentry *upperdentry;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	upperdentry = ovl_dentry_upper(dentry);
-	if (upperdentry) {
-		mutex_lock(&upperdentry->d_inode->i_mutex);
-		err = notify_change(upperdentry, attr, NULL);
-		mutex_unlock(&upperdentry->d_inode->i_mutex);
-	} else {
-		err = ovl_copy_up_last(dentry, attr, false);
-	}
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static int ovl_getattr(struct vfsmount *mnt, struct dentry *dentry,
-			 struct kstat *stat)
-{
-	struct path realpath;
-
-	ovl_path_real(dentry, &realpath);
-	return vfs_getattr(&realpath, stat);
-}
-
-int ovl_permission(struct inode *inode, int mask)
-{
-	struct ovl_entry *oe;
-	struct dentry *alias = NULL;
-	struct inode *realinode;
-	struct dentry *realdentry;
-	bool is_upper;
-	int err;
-
-	if (S_ISDIR(inode->i_mode)) {
-		oe = inode->i_private;
-	} else if (mask & MAY_NOT_BLOCK) {
-		return -ECHILD;
-	} else {
-		/*
-		 * For non-directories find an alias and get the info
-		 * from there.
-		 */
-		alias = d_find_any_alias(inode);
-		if (WARN_ON(!alias))
-			return -ENOENT;
-
-		oe = alias->d_fsdata;
-	}
-
-	realdentry = ovl_entry_real(oe, &is_upper);
-
-	/* Careful in RCU walk mode */
-	realinode = ACCESS_ONCE(realdentry->d_inode);
-	if (!realinode) {
-		WARN_ON(!(mask & MAY_NOT_BLOCK));
-		err = -ENOENT;
-		goto out_dput;
-	}
-
-	if (mask & MAY_WRITE) {
-		umode_t mode = realinode->i_mode;
-
-		/*
-		 * Writes will always be redirected to upper layer, so
-		 * ignore lower layer being read-only.
-		 *
-		 * If the overlay itself is read-only then proceed
-		 * with the permission check, don't return EROFS.
-		 * This will only happen if this is the lower layer of
-		 * another overlayfs.
-		 *
-		 * If upper fs becomes read-only after the overlay was
-		 * constructed return EROFS to prevent modification of
-		 * upper layer.
-		 */
-		err = -EROFS;
-		if (is_upper && !IS_RDONLY(inode) && IS_RDONLY(realinode) &&
-		    (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)))
-			goto out_dput;
-	}
-
-	err = __inode_permission(realinode, mask);
-out_dput:
-	dput(alias);
-	return err;
-}
-
-
-struct ovl_link_data {
-	struct dentry *realdentry;
-	void *cookie;
-};
-
-static void *ovl_follow_link(struct dentry *dentry, struct nameidata *nd)
-{
-	void *ret;
-	struct dentry *realdentry;
-	struct inode *realinode;
-
-	realdentry = ovl_dentry_real(dentry);
-	realinode = realdentry->d_inode;
-
-	if (WARN_ON(!realinode->i_op->follow_link))
-		return ERR_PTR(-EPERM);
-
-	ret = realinode->i_op->follow_link(realdentry, nd);
-	if (IS_ERR(ret))
-		return ret;
-
-	if (realinode->i_op->put_link) {
-		struct ovl_link_data *data;
-
-		data = kmalloc(sizeof(struct ovl_link_data), GFP_KERNEL);
-		if (!data) {
-			realinode->i_op->put_link(realdentry, nd, ret);
-			return ERR_PTR(-ENOMEM);
-		}
-		data->realdentry = realdentry;
-		data->cookie = ret;
-
-		return data;
-	} else {
-		return NULL;
-	}
-}
-
-static void ovl_put_link(struct dentry *dentry, struct nameidata *nd, void *c)
-{
-	struct inode *realinode;
-	struct ovl_link_data *data = c;
-
-	if (!data)
-		return;
-
-	realinode = data->realdentry->d_inode;
-	realinode->i_op->put_link(data->realdentry, nd, data->cookie);
-	kfree(data);
-}
-
-static int ovl_readlink(struct dentry *dentry, char __user *buf, int bufsiz)
-{
-	struct path realpath;
-	struct inode *realinode;
-
-	ovl_path_real(dentry, &realpath);
-	realinode = realpath.dentry->d_inode;
-
-	if (!realinode->i_op->readlink)
-		return -EINVAL;
-
-	touch_atime(&realpath);
-
-	return realinode->i_op->readlink(realpath.dentry, buf, bufsiz);
-}
-
-
-static bool ovl_is_private_xattr(const char *name)
-{
-	return strncmp(name, OVL_XATTR_PRE_NAME, OVL_XATTR_PRE_LEN) == 0;
-}
-
-int ovl_setxattr(struct dentry *dentry, const char *name,
-		 const void *value, size_t size, int flags)
-{
-	int err;
-	struct dentry *upperdentry;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = -EPERM;
-	if (ovl_is_private_xattr(name))
-		goto out_drop_write;
-
-	err = ovl_copy_up(dentry);
-	if (err)
-		goto out_drop_write;
-
-	upperdentry = ovl_dentry_upper(dentry);
-	err = vfs_setxattr(upperdentry, name, value, size, flags);
-
-out_drop_write:
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static bool ovl_need_xattr_filter(struct dentry *dentry,
-				  enum ovl_path_type type)
-{
-	if ((type & (__OVL_PATH_PURE | __OVL_PATH_UPPER)) == __OVL_PATH_UPPER)
-		return S_ISDIR(dentry->d_inode->i_mode);
-	else
-		return false;
-}
-
-ssize_t ovl_getxattr(struct dentry *dentry, const char *name,
-		     void *value, size_t size)
-{
-	struct path realpath;
-	enum ovl_path_type type = ovl_path_real(dentry, &realpath);
-
-	if (ovl_need_xattr_filter(dentry, type) && ovl_is_private_xattr(name))
-		return -ENODATA;
-
-	return vfs_getxattr(realpath.dentry, name, value, size);
-}
-
-ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size)
-{
-	struct path realpath;
-	enum ovl_path_type type = ovl_path_real(dentry, &realpath);
-	ssize_t res;
-	int off;
-
-	res = vfs_listxattr(realpath.dentry, list, size);
-	if (res <= 0 || size == 0)
-		return res;
-
-	if (!ovl_need_xattr_filter(dentry, type))
-		return res;
-
-	/* filter out private xattrs */
-	for (off = 0; off < res;) {
-		char *s = list + off;
-		size_t slen = strlen(s) + 1;
-
-		BUG_ON(off + slen > res);
-
-		if (ovl_is_private_xattr(s)) {
-			res -= slen;
-			memmove(s, s + slen, res - off);
-		} else {
-			off += slen;
-		}
-	}
-
-	return res;
-}
-
-int ovl_removexattr(struct dentry *dentry, const char *name)
-{
-	int err;
-	struct path realpath;
-	enum ovl_path_type type = ovl_path_real(dentry, &realpath);
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = -ENODATA;
-	if (ovl_need_xattr_filter(dentry, type) && ovl_is_private_xattr(name))
-		goto out_drop_write;
-
-	if (!OVL_TYPE_UPPER(type)) {
-		err = vfs_getxattr(realpath.dentry, name, NULL, 0);
-		if (err < 0)
-			goto out_drop_write;
-
-		err = ovl_copy_up(dentry);
-		if (err)
-			goto out_drop_write;
-
-		ovl_path_upper(dentry, &realpath);
-	}
-
-	err = vfs_removexattr(realpath.dentry, name);
-out_drop_write:
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static bool ovl_open_need_copy_up(int flags, enum ovl_path_type type,
-				  struct dentry *realdentry)
-{
-	if (OVL_TYPE_UPPER(type))
-		return false;
-
-	if (special_file(realdentry->d_inode->i_mode))
-		return false;
-
-	if (!(OPEN_FMODE(flags) & FMODE_WRITE) && !(flags & O_TRUNC))
-		return false;
-
-	return true;
-}
-
-static int ovl_dentry_open(struct dentry *dentry, struct file *file,
-		    const struct cred *cred)
-{
-	int err;
-	struct path realpath;
-	enum ovl_path_type type;
-	bool want_write = false;
-
-	type = ovl_path_real(dentry, &realpath);
-	if (!ovl_is_nocopyupw(dentry)) {
-		if (ovl_open_need_copy_up(file->f_flags, type, realpath.dentry)) {
-			want_write = true;
-			err = ovl_want_write(dentry);
-			if (err)
-				goto out;
-
-			if (file->f_flags & O_TRUNC)
-				err = ovl_copy_up_last(dentry, NULL, true);
-			else
-				err = ovl_copy_up(dentry);
-			if (err)
-				goto out_drop_write;
-
-			ovl_path_upper(dentry, &realpath);
-		}
-	}
-
-	err = vfs_open(&realpath, file, cred);
-out_drop_write:
-	if (want_write)
-		ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static const struct inode_operations ovl_file_inode_operations = {
-	.setattr	= ovl_setattr,
-	.permission	= ovl_permission,
-	.getattr	= ovl_getattr,
-	.setxattr	= ovl_setxattr,
-	.getxattr	= ovl_getxattr,
-	.listxattr	= ovl_listxattr,
-	.removexattr	= ovl_removexattr,
-	.dentry_open	= ovl_dentry_open,
-};
-
-static const struct inode_operations ovl_symlink_inode_operations = {
-	.setattr	= ovl_setattr,
-	.follow_link	= ovl_follow_link,
-	.put_link	= ovl_put_link,
-	.readlink	= ovl_readlink,
-	.getattr	= ovl_getattr,
-	.setxattr	= ovl_setxattr,
-	.getxattr	= ovl_getxattr,
-	.listxattr	= ovl_listxattr,
-	.removexattr	= ovl_removexattr,
-};
-
-struct inode *ovl_new_inode(struct super_block *sb, umode_t mode,
-			    struct ovl_entry *oe)
-{
-	struct inode *inode;
-
-	inode = new_inode(sb);
-	if (!inode)
-		return NULL;
-
-	mode &= S_IFMT;
-
-	inode->i_ino = get_next_ino();
-	inode->i_mode = mode;
-	inode->i_flags |= S_NOATIME | S_NOCMTIME;
-
-	switch (mode) {
-	case S_IFDIR:
-		inode->i_private = oe;
-		inode->i_op = &ovl_dir_inode_operations;
-		inode->i_fop = &ovl_dir_operations;
-		break;
-
-	case S_IFLNK:
-		inode->i_op = &ovl_symlink_inode_operations;
-		break;
-
-	case S_IFREG:
-	case S_IFSOCK:
-	case S_IFBLK:
-	case S_IFCHR:
-	case S_IFIFO:
-		inode->i_op = &ovl_file_inode_operations;
-		break;
-
-	default:
-		WARN(1, "illegal file type: %i\n", mode);
-		iput(inode);
-		inode = NULL;
-	}
-
-	return inode;
-}

executer/kernel/mcoverlayfs/linux-4.0.9/overlayfs.h

@@ -1,200 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/kernel.h>
-
-struct ovl_entry;
-
-enum ovl_path_type {
-	__OVL_PATH_PURE		= (1 << 0),
-	__OVL_PATH_UPPER	= (1 << 1),
-	__OVL_PATH_MERGE	= (1 << 2),
-};
-
-#define OVL_TYPE_UPPER(type)	((type) & __OVL_PATH_UPPER)
-#define OVL_TYPE_MERGE(type)	((type) & __OVL_PATH_MERGE)
-#define OVL_TYPE_PURE_UPPER(type) ((type) & __OVL_PATH_PURE)
-#define OVL_TYPE_MERGE_OR_LOWER(type) \
-	(OVL_TYPE_MERGE(type) || !OVL_TYPE_UPPER(type))
-
-#define OVL_XATTR_PRE_NAME "trusted.overlay."
-#define OVL_XATTR_PRE_LEN  16
-#define OVL_XATTR_OPAQUE   OVL_XATTR_PRE_NAME"opaque"
-
-static inline int ovl_do_rmdir(struct inode *dir, struct dentry *dentry)
-{
-	int err = vfs_rmdir(dir, dentry);
-	pr_debug("rmdir(%pd2) = %i\n", dentry, err);
-	return err;
-}
-
-static inline int ovl_do_unlink(struct inode *dir, struct dentry *dentry)
-{
-	int err = vfs_unlink(dir, dentry, NULL);
-	pr_debug("unlink(%pd2) = %i\n", dentry, err);
-	return err;
-}
-
-static inline int ovl_do_link(struct dentry *old_dentry, struct inode *dir,
-			      struct dentry *new_dentry, bool debug)
-{
-	int err = vfs_link(old_dentry, dir, new_dentry, NULL);
-	if (debug) {
-		pr_debug("link(%pd2, %pd2) = %i\n",
-			 old_dentry, new_dentry, err);
-	}
-	return err;
-}
-
-static inline int ovl_do_create(struct inode *dir, struct dentry *dentry,
-			     umode_t mode, bool debug)
-{
-	int err = vfs_create(dir, dentry, mode, true);
-	if (debug)
-		pr_debug("create(%pd2, 0%o) = %i\n", dentry, mode, err);
-	return err;
-}
-
-static inline int ovl_do_mkdir(struct inode *dir, struct dentry *dentry,
-			       umode_t mode, bool debug)
-{
-	int err = vfs_mkdir(dir, dentry, mode);
-	if (debug)
-		pr_debug("mkdir(%pd2, 0%o) = %i\n", dentry, mode, err);
-	return err;
-}
-
-static inline int ovl_do_mknod(struct inode *dir, struct dentry *dentry,
-			       umode_t mode, dev_t dev, bool debug)
-{
-	int err = vfs_mknod(dir, dentry, mode, dev);
-	if (debug) {
-		pr_debug("mknod(%pd2, 0%o, 0%o) = %i\n",
-			 dentry, mode, dev, err);
-	}
-	return err;
-}
-
-static inline int ovl_do_symlink(struct inode *dir, struct dentry *dentry,
-				 const char *oldname, bool debug)
-{
-	int err = vfs_symlink(dir, dentry, oldname);
-	if (debug)
-		pr_debug("symlink(\"%s\", %pd2) = %i\n", oldname, dentry, err);
-	return err;
-}
-
-static inline int ovl_do_setxattr(struct dentry *dentry, const char *name,
-				  const void *value, size_t size, int flags)
-{
-	int err = vfs_setxattr(dentry, name, value, size, flags);
-	pr_debug("setxattr(%pd2, \"%s\", \"%*s\", 0x%x) = %i\n",
-		 dentry, name, (int) size, (char *) value, flags, err);
-	return err;
-}
-
-static inline int ovl_do_removexattr(struct dentry *dentry, const char *name)
-{
-	int err = vfs_removexattr(dentry, name);
-	pr_debug("removexattr(%pd2, \"%s\") = %i\n", dentry, name, err);
-	return err;
-}
-
-static inline int ovl_do_rename(struct inode *olddir, struct dentry *olddentry,
-				struct inode *newdir, struct dentry *newdentry,
-				unsigned int flags)
-{
-	int err;
-
-	pr_debug("rename2(%pd2, %pd2, 0x%x)\n",
-		 olddentry, newdentry, flags);
-
-	err = vfs_rename(olddir, olddentry, newdir, newdentry, NULL, flags);
-
-	if (err) {
-		pr_debug("...rename2(%pd2, %pd2, ...) = %i\n",
-			 olddentry, newdentry, err);
-	}
-	return err;
-}
-
-static inline int ovl_do_whiteout(struct inode *dir, struct dentry *dentry)
-{
-	int err = vfs_whiteout(dir, dentry);
-	pr_debug("whiteout(%pd2) = %i\n", dentry, err);
-	return err;
-}
-
-bool ovl_is_nocopyupw(struct dentry *dentry);
-enum ovl_path_type ovl_path_type(struct dentry *dentry);
-u64 ovl_dentry_version_get(struct dentry *dentry);
-void ovl_dentry_version_inc(struct dentry *dentry);
-void ovl_path_upper(struct dentry *dentry, struct path *path);
-void ovl_path_lower(struct dentry *dentry, struct path *path);
-enum ovl_path_type ovl_path_real(struct dentry *dentry, struct path *path);
-int ovl_path_next(int idx, struct dentry *dentry, struct path *path);
-struct dentry *ovl_dentry_upper(struct dentry *dentry);
-struct dentry *ovl_dentry_lower(struct dentry *dentry);
-struct dentry *ovl_dentry_real(struct dentry *dentry);
-struct dentry *ovl_entry_real(struct ovl_entry *oe, bool *is_upper);
-struct ovl_dir_cache *ovl_dir_cache(struct dentry *dentry);
-void ovl_set_dir_cache(struct dentry *dentry, struct ovl_dir_cache *cache);
-struct dentry *ovl_workdir(struct dentry *dentry);
-int ovl_want_write(struct dentry *dentry);
-void ovl_drop_write(struct dentry *dentry);
-bool ovl_dentry_is_opaque(struct dentry *dentry);
-void ovl_dentry_set_opaque(struct dentry *dentry, bool opaque);
-bool ovl_is_whiteout(struct dentry *dentry);
-void ovl_dentry_update(struct dentry *dentry, struct dentry *upperdentry);
-struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry,
-			  unsigned int flags);
-struct file *ovl_path_open(struct path *path, int flags);
-
-struct dentry *ovl_upper_create(struct dentry *upperdir, struct dentry *dentry,
-				struct kstat *stat, const char *link);
-
-/* readdir.c */
-extern const struct file_operations ovl_dir_operations;
-int ovl_check_empty_dir(struct dentry *dentry, struct list_head *list);
-void ovl_cleanup_whiteouts(struct dentry *upper, struct list_head *list);
-void ovl_cache_free(struct list_head *list);
-
-/* inode.c */
-int ovl_setattr(struct dentry *dentry, struct iattr *attr);
-int ovl_permission(struct inode *inode, int mask);
-int ovl_setxattr(struct dentry *dentry, const char *name,
-		 const void *value, size_t size, int flags);
-ssize_t ovl_getxattr(struct dentry *dentry, const char *name,
-		     void *value, size_t size);
-ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size);
-int ovl_removexattr(struct dentry *dentry, const char *name);
-
-struct inode *ovl_new_inode(struct super_block *sb, umode_t mode,
-			    struct ovl_entry *oe);
-static inline void ovl_copyattr(struct inode *from, struct inode *to)
-{
-	to->i_uid = from->i_uid;
-	to->i_gid = from->i_gid;
-}
-
-/* dir.c */
-extern const struct inode_operations ovl_dir_inode_operations;
-struct dentry *ovl_lookup_temp(struct dentry *workdir, struct dentry *dentry);
-int ovl_create_real(struct inode *dir, struct dentry *newdentry,
-		    struct kstat *stat, const char *link,
-		    struct dentry *hardlink, bool debug);
-void ovl_cleanup(struct inode *dir, struct dentry *dentry);
-
-/* copy_up.c */
-int ovl_copy_up(struct dentry *dentry);
-int ovl_copy_up_one(struct dentry *parent, struct dentry *dentry,
-		    struct path *lowerpath, struct kstat *stat,
-		    struct iattr *attr);
-int ovl_copy_xattr(struct dentry *old, struct dentry *new);
-int ovl_set_attr(struct dentry *upper, struct kstat *stat);

executer/kernel/mcoverlayfs/linux-4.0.9/readdir.c

@@ -1,557 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/namei.h>
-#include <linux/file.h>
-#include <linux/xattr.h>
-#include <linux/rbtree.h>
-#include <linux/security.h>
-#include <linux/cred.h>
-#include "overlayfs.h"
-
-struct ovl_cache_entry {
-	unsigned int len;
-	unsigned int type;
-	u64 ino;
-	struct list_head l_node;
-	struct rb_node node;
-	bool is_whiteout;
-	char name[];
-};
-
-struct ovl_dir_cache {
-	long refcount;
-	u64 version;
-	struct list_head entries;
-};
-
-struct ovl_readdir_data {
-	struct dir_context ctx;
-	bool is_merge;
-	struct rb_root root;
-	struct list_head *list;
-	struct list_head middle;
-	struct dentry *dir;
-	int count;
-	int err;
-};
-
-struct ovl_dir_file {
-	bool is_real;
-	bool is_upper;
-	struct ovl_dir_cache *cache;
-	struct list_head *cursor;
-	struct file *realfile;
-	struct file *upperfile;
-};
-
-static struct ovl_cache_entry *ovl_cache_entry_from_node(struct rb_node *n)
-{
-	return container_of(n, struct ovl_cache_entry, node);
-}
-
-static struct ovl_cache_entry *ovl_cache_entry_find(struct rb_root *root,
-						    const char *name, int len)
-{
-	struct rb_node *node = root->rb_node;
-	int cmp;
-
-	while (node) {
-		struct ovl_cache_entry *p = ovl_cache_entry_from_node(node);
-
-		cmp = strncmp(name, p->name, len);
-		if (cmp > 0)
-			node = p->node.rb_right;
-		else if (cmp < 0 || len < p->len)
-			node = p->node.rb_left;
-		else
-			return p;
-	}
-
-	return NULL;
-}
-
-static struct ovl_cache_entry *ovl_cache_entry_new(struct dentry *dir,
-						   const char *name, int len,
-						   u64 ino, unsigned int d_type)
-{
-	struct ovl_cache_entry *p;
-	size_t size = offsetof(struct ovl_cache_entry, name[len + 1]);
-
-	p = kmalloc(size, GFP_KERNEL);
-	if (!p)
-		return NULL;
-
-	memcpy(p->name, name, len);
-	p->name[len] = '\0';
-	p->len = len;
-	p->type = d_type;
-	p->ino = ino;
-	p->is_whiteout = false;
-
-	if (d_type == DT_CHR) {
-		struct dentry *dentry;
-		const struct cred *old_cred;
-		struct cred *override_cred;
-
-		override_cred = prepare_creds();
-		if (!override_cred) {
-			kfree(p);
-			return NULL;
-		}
-
-		/*
-		 * CAP_DAC_OVERRIDE for lookup
-		 */
-		cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE);
-		old_cred = override_creds(override_cred);
-
-		dentry = lookup_one_len(name, dir, len);
-		if (!IS_ERR(dentry)) {
-			p->is_whiteout = ovl_is_whiteout(dentry);
-			dput(dentry);
-		}
-		revert_creds(old_cred);
-		put_cred(override_cred);
-	}
-	return p;
-}
-
-static int ovl_cache_entry_add_rb(struct ovl_readdir_data *rdd,
-				  const char *name, int len, u64 ino,
-				  unsigned int d_type)
-{
-	struct rb_node **newp = &rdd->root.rb_node;
-	struct rb_node *parent = NULL;
-	struct ovl_cache_entry *p;
-
-	while (*newp) {
-		int cmp;
-		struct ovl_cache_entry *tmp;
-
-		parent = *newp;
-		tmp = ovl_cache_entry_from_node(*newp);
-		cmp = strncmp(name, tmp->name, len);
-		if (cmp > 0)
-			newp = &tmp->node.rb_right;
-		else if (cmp < 0 || len < tmp->len)
-			newp = &tmp->node.rb_left;
-		else
-			return 0;
-	}
-
-	p = ovl_cache_entry_new(rdd->dir, name, len, ino, d_type);
-	if (p == NULL)
-		return -ENOMEM;
-
-	list_add_tail(&p->l_node, rdd->list);
-	rb_link_node(&p->node, parent, newp);
-	rb_insert_color(&p->node, &rdd->root);
-
-	return 0;
-}
-
-static int ovl_fill_lower(struct ovl_readdir_data *rdd,
-			  const char *name, int namelen,
-			  loff_t offset, u64 ino, unsigned int d_type)
-{
-	struct ovl_cache_entry *p;
-
-	p = ovl_cache_entry_find(&rdd->root, name, namelen);
-	if (p) {
-		list_move_tail(&p->l_node, &rdd->middle);
-	} else {
-		p = ovl_cache_entry_new(rdd->dir, name, namelen, ino, d_type);
-		if (p == NULL)
-			rdd->err = -ENOMEM;
-		else
-			list_add_tail(&p->l_node, &rdd->middle);
-	}
-
-	return rdd->err;
-}
-
-void ovl_cache_free(struct list_head *list)
-{
-	struct ovl_cache_entry *p;
-	struct ovl_cache_entry *n;
-
-	list_for_each_entry_safe(p, n, list, l_node)
-		kfree(p);
-
-	INIT_LIST_HEAD(list);
-}
-
-static void ovl_cache_put(struct ovl_dir_file *od, struct dentry *dentry)
-{
-	struct ovl_dir_cache *cache = od->cache;
-
-	WARN_ON(cache->refcount <= 0);
-	cache->refcount--;
-	if (!cache->refcount) {
-		if (ovl_dir_cache(dentry) == cache)
-			ovl_set_dir_cache(dentry, NULL);
-
-		ovl_cache_free(&cache->entries);
-		kfree(cache);
-	}
-}
-
-static int ovl_fill_merge(struct dir_context *ctx, const char *name,
-			  int namelen, loff_t offset, u64 ino,
-			  unsigned int d_type)
-{
-	struct ovl_readdir_data *rdd =
-		container_of(ctx, struct ovl_readdir_data, ctx);
-
-	rdd->count++;
-	if (!rdd->is_merge)
-		return ovl_cache_entry_add_rb(rdd, name, namelen, ino, d_type);
-	else
-		return ovl_fill_lower(rdd, name, namelen, offset, ino, d_type);
-}
-
-static inline int ovl_dir_read(struct path *realpath,
-			       struct ovl_readdir_data *rdd)
-{
-	struct file *realfile;
-	int err;
-
-	realfile = ovl_path_open(realpath, O_RDONLY | O_DIRECTORY);
-	if (IS_ERR(realfile))
-		return PTR_ERR(realfile);
-
-	rdd->dir = realpath->dentry;
-	rdd->ctx.pos = 0;
-	do {
-		rdd->count = 0;
-		rdd->err = 0;
-		err = iterate_dir(realfile, &rdd->ctx);
-		if (err >= 0)
-			err = rdd->err;
-	} while (!err && rdd->count);
-	fput(realfile);
-
-	return err;
-}
-
-static void ovl_dir_reset(struct file *file)
-{
-	struct ovl_dir_file *od = file->private_data;
-	struct ovl_dir_cache *cache = od->cache;
-	struct dentry *dentry = file->f_path.dentry;
-	enum ovl_path_type type = ovl_path_type(dentry);
-
-	if (cache && ovl_dentry_version_get(dentry) != cache->version) {
-		ovl_cache_put(od, dentry);
-		od->cache = NULL;
-		od->cursor = NULL;
-	}
-	WARN_ON(!od->is_real && !OVL_TYPE_MERGE(type));
-	if (od->is_real && OVL_TYPE_MERGE(type))
-		od->is_real = false;
-}
-
-static int ovl_dir_read_merged(struct dentry *dentry, struct list_head *list)
-{
-	int err;
-	struct path realpath;
-	struct ovl_readdir_data rdd = {
-		.ctx.actor = ovl_fill_merge,
-		.list = list,
-		.root = RB_ROOT,
-		.is_merge = false,
-	};
-	int idx, next;
-
-	for (idx = 0; idx != -1; idx = next) {
-		next = ovl_path_next(idx, dentry, &realpath);
-
-		if (next != -1) {
-			err = ovl_dir_read(&realpath, &rdd);
-			if (err)
-				break;
-		} else {
-			/*
-			 * Insert lowest layer entries before upper ones, this
-			 * allows offsets to be reasonably constant
-			 */
-			list_add(&rdd.middle, rdd.list);
-			rdd.is_merge = true;
-			err = ovl_dir_read(&realpath, &rdd);
-			list_del(&rdd.middle);
-		}
-	}
-	return err;
-}
-
-static void ovl_seek_cursor(struct ovl_dir_file *od, loff_t pos)
-{
-	struct list_head *p;
-	loff_t off = 0;
-
-	list_for_each(p, &od->cache->entries) {
-		if (off >= pos)
-			break;
-		off++;
-	}
-	/* Cursor is safe since the cache is stable */
-	od->cursor = p;
-}
-
-static struct ovl_dir_cache *ovl_cache_get(struct dentry *dentry)
-{
-	int res;
-	struct ovl_dir_cache *cache;
-
-	cache = ovl_dir_cache(dentry);
-	if (cache && ovl_dentry_version_get(dentry) == cache->version) {
-		cache->refcount++;
-		return cache;
-	}
-	ovl_set_dir_cache(dentry, NULL);
-
-	cache = kzalloc(sizeof(struct ovl_dir_cache), GFP_KERNEL);
-	if (!cache)
-		return ERR_PTR(-ENOMEM);
-
-	cache->refcount = 1;
-	INIT_LIST_HEAD(&cache->entries);
-
-	res = ovl_dir_read_merged(dentry, &cache->entries);
-	if (res) {
-		ovl_cache_free(&cache->entries);
-		kfree(cache);
-		return ERR_PTR(res);
-	}
-
-	cache->version = ovl_dentry_version_get(dentry);
-	ovl_set_dir_cache(dentry, cache);
-
-	return cache;
-}
-
-static int ovl_iterate(struct file *file, struct dir_context *ctx)
-{
-	struct ovl_dir_file *od = file->private_data;
-	struct dentry *dentry = file->f_path.dentry;
-	struct ovl_cache_entry *p;
-
-	if (!ctx->pos)
-		ovl_dir_reset(file);
-
-	if (od->is_real)
-		return iterate_dir(od->realfile, ctx);
-
-	if (!od->cache) {
-		struct ovl_dir_cache *cache;
-
-		cache = ovl_cache_get(dentry);
-		if (IS_ERR(cache))
-			return PTR_ERR(cache);
-
-		od->cache = cache;
-		ovl_seek_cursor(od, ctx->pos);
-	}
-
-	while (od->cursor != &od->cache->entries) {
-		p = list_entry(od->cursor, struct ovl_cache_entry, l_node);
-		if (!p->is_whiteout)
-			if (!dir_emit(ctx, p->name, p->len, p->ino, p->type))
-				break;
-		od->cursor = p->l_node.next;
-		ctx->pos++;
-	}
-	return 0;
-}
-
-static loff_t ovl_dir_llseek(struct file *file, loff_t offset, int origin)
-{
-	loff_t res;
-	struct ovl_dir_file *od = file->private_data;
-
-	mutex_lock(&file_inode(file)->i_mutex);
-	if (!file->f_pos)
-		ovl_dir_reset(file);
-
-	if (od->is_real) {
-		res = vfs_llseek(od->realfile, offset, origin);
-		file->f_pos = od->realfile->f_pos;
-	} else {
-		res = -EINVAL;
-
-		switch (origin) {
-		case SEEK_CUR:
-			offset += file->f_pos;
-			break;
-		case SEEK_SET:
-			break;
-		default:
-			goto out_unlock;
-		}
-		if (offset < 0)
-			goto out_unlock;
-
-		if (offset != file->f_pos) {
-			file->f_pos = offset;
-			if (od->cache)
-				ovl_seek_cursor(od, offset);
-		}
-		res = offset;
-	}
-out_unlock:
-	mutex_unlock(&file_inode(file)->i_mutex);
-
-	return res;
-}
-
-static int ovl_dir_fsync(struct file *file, loff_t start, loff_t end,
-			 int datasync)
-{
-	struct ovl_dir_file *od = file->private_data;
-	struct dentry *dentry = file->f_path.dentry;
-	struct file *realfile = od->realfile;
-
-	/*
-	 * Need to check if we started out being a lower dir, but got copied up
-	 */
-	if (!od->is_upper && OVL_TYPE_UPPER(ovl_path_type(dentry))) {
-		struct inode *inode = file_inode(file);
-
-		realfile = lockless_dereference(od->upperfile);
-		if (!realfile) {
-			struct path upperpath;
-
-			ovl_path_upper(dentry, &upperpath);
-			realfile = ovl_path_open(&upperpath, O_RDONLY);
-			smp_mb__before_spinlock();
-			mutex_lock(&inode->i_mutex);
-			if (!od->upperfile) {
-				if (IS_ERR(realfile)) {
-					mutex_unlock(&inode->i_mutex);
-					return PTR_ERR(realfile);
-				}
-				od->upperfile = realfile;
-			} else {
-				/* somebody has beaten us to it */
-				if (!IS_ERR(realfile))
-					fput(realfile);
-				realfile = od->upperfile;
-			}
-			mutex_unlock(&inode->i_mutex);
-		}
-	}
-
-	return vfs_fsync_range(realfile, start, end, datasync);
-}
-
-static int ovl_dir_release(struct inode *inode, struct file *file)
-{
-	struct ovl_dir_file *od = file->private_data;
-
-	if (od->cache) {
-		mutex_lock(&inode->i_mutex);
-		ovl_cache_put(od, file->f_path.dentry);
-		mutex_unlock(&inode->i_mutex);
-	}
-	fput(od->realfile);
-	if (od->upperfile)
-		fput(od->upperfile);
-	kfree(od);
-
-	return 0;
-}
-
-static int ovl_dir_open(struct inode *inode, struct file *file)
-{
-	struct path realpath;
-	struct file *realfile;
-	struct ovl_dir_file *od;
-	enum ovl_path_type type;
-
-	od = kzalloc(sizeof(struct ovl_dir_file), GFP_KERNEL);
-	if (!od)
-		return -ENOMEM;
-
-	type = ovl_path_real(file->f_path.dentry, &realpath);
-	realfile = ovl_path_open(&realpath, file->f_flags);
-	if (IS_ERR(realfile)) {
-		kfree(od);
-		return PTR_ERR(realfile);
-	}
-	od->realfile = realfile;
-	od->is_real = !OVL_TYPE_MERGE(type);
-	od->is_upper = OVL_TYPE_UPPER(type);
-	file->private_data = od;
-
-	return 0;
-}
-
-const struct file_operations ovl_dir_operations = {
-	.read		= generic_read_dir,
-	.open		= ovl_dir_open,
-	.iterate	= ovl_iterate,
-	.llseek		= ovl_dir_llseek,
-	.fsync		= ovl_dir_fsync,
-	.release	= ovl_dir_release,
-};
-
-int ovl_check_empty_dir(struct dentry *dentry, struct list_head *list)
-{
-	int err;
-	struct ovl_cache_entry *p;
-
-	err = ovl_dir_read_merged(dentry, list);
-	if (err)
-		return err;
-
-	err = 0;
-
-	list_for_each_entry(p, list, l_node) {
-		if (p->is_whiteout)
-			continue;
-
-		if (p->name[0] == '.') {
-			if (p->len == 1)
-				continue;
-			if (p->len == 2 && p->name[1] == '.')
-				continue;
-		}
-		err = -ENOTEMPTY;
-		break;
-	}
-
-	return err;
-}
-
-void ovl_cleanup_whiteouts(struct dentry *upper, struct list_head *list)
-{
-	struct ovl_cache_entry *p;
-
-	mutex_lock_nested(&upper->d_inode->i_mutex, I_MUTEX_CHILD);
-	list_for_each_entry(p, list, l_node) {
-		struct dentry *dentry;
-
-		if (!p->is_whiteout)
-			continue;
-
-		dentry = lookup_one_len(p->name, upper, p->len);
-		if (IS_ERR(dentry)) {
-			pr_err("overlayfs: lookup '%s/%.*s' failed (%i)\n",
-			       upper->d_name.name, p->len, p->name,
-			       (int) PTR_ERR(dentry));
-			continue;
-		}
-		ovl_cleanup(upper->d_inode, dentry);
-		dput(dentry);
-	}
-	mutex_unlock(&upper->d_inode->i_mutex);
-}

executer/kernel/mcoverlayfs/linux-4.18.14/CMakeLists.txt

@@ -1,7 +0,0 @@
-kmod(mcoverlay
-	SOURCES
-		copy_up.c dir.c inode.c readdir.c super.c namei.c util.c export.c
-	INSTALL_DEST
-		${KMODDIR}
-)
-

executer/kernel/mcoverlayfs/linux-4.18.14/copy_up.c

@@ -1,804 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/module.h>
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/file.h>
-#include <linux/splice.h>
-#include <linux/xattr.h>
-#include <linux/security.h>
-#include <linux/uaccess.h>
-#include <linux/sched/signal.h>
-#include <linux/cred.h>
-#include <linux/namei.h>
-#include <linux/fdtable.h>
-#include <linux/ratelimit.h>
-#include <linux/exportfs.h>
-#include "overlayfs.h"
-
-#define OVL_COPY_UP_CHUNK_SIZE (1 << 20)
-
-static bool __read_mostly ovl_check_copy_up;
-module_param_named(check_copy_up, ovl_check_copy_up, bool,
-		   S_IWUSR | S_IRUGO);
-MODULE_PARM_DESC(ovl_check_copy_up,
-		 "Warn on copy-up when causing process also has a R/O fd open");
-
-static int ovl_check_fd(const void *data, struct file *f, unsigned int fd)
-{
-	const struct dentry *dentry = data;
-
-	if (file_inode(f) == d_inode(dentry))
-		pr_warn_ratelimited("overlayfs: Warning: Copying up %pD, but open R/O on fd %u which will cease to be coherent [pid=%d %s]\n",
-				    f, fd, current->pid, current->comm);
-	return 0;
-}
-
-/*
- * Check the fds open by this process and warn if something like the following
- * scenario is about to occur:
- *
- *	fd1 = open("foo", O_RDONLY);
- *	fd2 = open("foo", O_RDWR);
- */
-static void ovl_do_check_copy_up(struct dentry *dentry)
-{
-	if (ovl_check_copy_up)
-		iterate_fd(current->files, 0, ovl_check_fd, dentry);
-}
-
-int ovl_copy_xattr(struct dentry *old, struct dentry *new)
-{
-	ssize_t list_size, size, value_size = 0;
-	char *buf, *name, *value = NULL;
-	int uninitialized_var(error);
-	size_t slen;
-
-	if (!(old->d_inode->i_opflags & IOP_XATTR) ||
-	    !(new->d_inode->i_opflags & IOP_XATTR))
-		return 0;
-
-	list_size = vfs_listxattr(old, NULL, 0);
-	if (list_size <= 0) {
-		if (list_size == -EOPNOTSUPP)
-			return 0;
-		return list_size;
-	}
-
-	buf = kzalloc(list_size, GFP_KERNEL);
-	if (!buf)
-		return -ENOMEM;
-
-	list_size = vfs_listxattr(old, buf, list_size);
-	if (list_size <= 0) {
-		error = list_size;
-		goto out;
-	}
-
-	for (name = buf; list_size; name += slen) {
-		slen = strnlen(name, list_size) + 1;
-
-		/* underlying fs providing us with an broken xattr list? */
-		if (WARN_ON(slen > list_size)) {
-			error = -EIO;
-			break;
-		}
-		list_size -= slen;
-
-		if (ovl_is_private_xattr(name))
-			continue;
-retry:
-		size = vfs_getxattr(old, name, value, value_size);
-		if (size == -ERANGE)
-			size = vfs_getxattr(old, name, NULL, 0);
-
-		if (size < 0) {
-			/* NOFSCHECK */
-			continue;
-
-			error = size;
-			break;
-		}
-
-		if (size > value_size) {
-			void *new;
-
-			new = krealloc(value, size, GFP_KERNEL);
-			if (!new) {
-				error = -ENOMEM;
-				break;
-			}
-			value = new;
-			value_size = size;
-			goto retry;
-		}
-
-		error = security_inode_copy_up_xattr(name);
-		if (error < 0 && error != -EOPNOTSUPP)
-			break;
-		if (error == 1) {
-			error = 0;
-			continue; /* Discard */
-		}
-		error = vfs_setxattr(new, name, value, size, 0);
-		if (error)
-			break;
-	}
-	kfree(value);
-out:
-	kfree(buf);
-	return error;
-}
-
-static int ovl_copy_up_data(struct path *old, struct path *new, loff_t len)
-{
-	struct file *old_file;
-	struct file *new_file;
-	loff_t old_pos = 0;
-	loff_t new_pos = 0;
-	int error = 0;
-
-	if (len == 0)
-		return 0;
-
-	old_file = ovl_path_open(old, O_LARGEFILE | O_RDONLY);
-	if (IS_ERR(old_file))
-		return PTR_ERR(old_file);
-
-	new_file = ovl_path_open(new, O_LARGEFILE | O_WRONLY);
-	if (IS_ERR(new_file)) {
-		error = PTR_ERR(new_file);
-		goto out_fput;
-	}
-
-	/* Try to use clone_file_range to clone up within the same fs */
-	error = vfs_clone_file_range(old_file, 0, new_file, 0, len);
-	if (!error)
-		goto out;
-	/* Couldn't clone, so now we try to copy the data */
-	error = 0;
-
-	/* FIXME: copy up sparse files efficiently */
-	while (len) {
-		size_t this_len = OVL_COPY_UP_CHUNK_SIZE;
-		long bytes;
-
-		if (len < this_len)
-			this_len = len;
-
-		if (signal_pending_state(TASK_KILLABLE, current)) {
-			error = -EINTR;
-			break;
-		}
-
-		bytes = do_splice_direct(old_file, &old_pos,
-					 new_file, &new_pos,
-					 this_len, SPLICE_F_MOVE);
-		if (bytes <= 0) {
-			error = bytes;
-			break;
-		}
-		WARN_ON(old_pos != new_pos);
-
-		len -= bytes;
-	}
-out:
-	if (!error)
-		error = vfs_fsync(new_file, 0);
-	fput(new_file);
-out_fput:
-	fput(old_file);
-	return error;
-}
-
-static int ovl_set_timestamps(struct dentry *upperdentry, struct kstat *stat)
-{
-	struct iattr attr = {
-		.ia_valid =
-		     ATTR_ATIME | ATTR_MTIME | ATTR_ATIME_SET | ATTR_MTIME_SET,
-		.ia_atime = stat->atime,
-		.ia_mtime = stat->mtime,
-	};
-
-	return notify_change(upperdentry, &attr, NULL);
-}
-
-int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat)
-{
-	int err = 0;
-
-	if (!S_ISLNK(stat->mode)) {
-		struct iattr attr = {
-			.ia_valid = ATTR_MODE,
-			.ia_mode = stat->mode,
-		};
-		err = notify_change(upperdentry, &attr, NULL);
-	}
-	if (!err) {
-		struct iattr attr = {
-			.ia_valid = ATTR_UID | ATTR_GID,
-			.ia_uid = stat->uid,
-			.ia_gid = stat->gid,
-		};
-		err = notify_change(upperdentry, &attr, NULL);
-	}
-	if (!err)
-		ovl_set_timestamps(upperdentry, stat);
-
-	return err;
-}
-
-struct ovl_fh *ovl_encode_real_fh(struct dentry *real, bool is_upper)
-{
-	struct ovl_fh *fh;
-	int fh_type, fh_len, dwords;
-	void *buf;
-	int buflen = MAX_HANDLE_SZ;
-	uuid_t *uuid = &real->d_sb->s_uuid;
-
-	buf = kmalloc(buflen, GFP_KERNEL);
-	if (!buf)
-		return ERR_PTR(-ENOMEM);
-
-	/*
-	 * We encode a non-connectable file handle for non-dir, because we
-	 * only need to find the lower inode number and we don't want to pay
-	 * the price or reconnecting the dentry.
-	 */
-	dwords = buflen >> 2;
-	fh_type = exportfs_encode_fh(real, buf, &dwords, 0);
-	buflen = (dwords << 2);
-
-	fh = ERR_PTR(-EIO);
-	if (WARN_ON(fh_type < 0) ||
-	    WARN_ON(buflen > MAX_HANDLE_SZ) ||
-	    WARN_ON(fh_type == FILEID_INVALID))
-		goto out;
-
-	BUILD_BUG_ON(MAX_HANDLE_SZ + offsetof(struct ovl_fh, fid) > 255);
-	fh_len = offsetof(struct ovl_fh, fid) + buflen;
-	fh = kmalloc(fh_len, GFP_KERNEL);
-	if (!fh) {
-		fh = ERR_PTR(-ENOMEM);
-		goto out;
-	}
-
-	fh->version = OVL_FH_VERSION;
-	fh->magic = OVL_FH_MAGIC;
-	fh->type = fh_type;
-	fh->flags = OVL_FH_FLAG_CPU_ENDIAN;
-	/*
-	 * When we will want to decode an overlay dentry from this handle
-	 * and all layers are on the same fs, if we get a disconncted real
-	 * dentry when we decode fid, the only way to tell if we should assign
-	 * it to upperdentry or to lowerstack is by checking this flag.
-	 */
-	if (is_upper)
-		fh->flags |= OVL_FH_FLAG_PATH_UPPER;
-	fh->len = fh_len;
-	fh->uuid = *uuid;
-	memcpy(fh->fid, buf, buflen);
-
-out:
-	kfree(buf);
-	return fh;
-}
-
-int ovl_set_origin(struct dentry *dentry, struct dentry *lower,
-		   struct dentry *upper)
-{
-	const struct ovl_fh *fh = NULL;
-	int err;
-
-	/*
-	 * When lower layer doesn't support export operations store a 'null' fh,
-	 * so we can use the overlay.origin xattr to distignuish between a copy
-	 * up and a pure upper inode.
-	 */
-	if (ovl_can_decode_fh(lower->d_sb)) {
-		fh = ovl_encode_real_fh(lower, false);
-		if (IS_ERR(fh))
-			return PTR_ERR(fh);
-	}
-
-	/*
-	 * Do not fail when upper doesn't support xattrs.
-	 */
-	err = ovl_check_setxattr(dentry, upper, OVL_XATTR_ORIGIN, fh,
-				 fh ? fh->len : 0, 0);
-	kfree(fh);
-
-	return err;
-}
-
-/* Store file handle of @upper dir in @index dir entry */
-static int ovl_set_upper_fh(struct dentry *upper, struct dentry *index)
-{
-	const struct ovl_fh *fh;
-	int err;
-
-	fh = ovl_encode_real_fh(upper, true);
-	if (IS_ERR(fh))
-		return PTR_ERR(fh);
-
-	err = ovl_do_setxattr(index, OVL_XATTR_UPPER, fh, fh->len, 0);
-
-	kfree(fh);
-	return err;
-}
-
-/*
- * Create and install index entry.
- *
- * Caller must hold i_mutex on indexdir.
- */
-static int ovl_create_index(struct dentry *dentry, struct dentry *origin,
-			    struct dentry *upper)
-{
-	struct dentry *indexdir = ovl_indexdir(dentry->d_sb);
-	struct inode *dir = d_inode(indexdir);
-	struct dentry *index = NULL;
-	struct dentry *temp = NULL;
-	struct qstr name = { };
-	int err;
-
-	/*
-	 * For now this is only used for creating index entry for directories,
-	 * because non-dir are copied up directly to index and then hardlinked
-	 * to upper dir.
-	 *
-	 * TODO: implement create index for non-dir, so we can call it when
-	 * encoding file handle for non-dir in case index does not exist.
-	 */
-	if (WARN_ON(!d_is_dir(dentry)))
-		return -EIO;
-
-	/* Directory not expected to be indexed before copy up */
-	if (WARN_ON(ovl_test_flag(OVL_INDEX, d_inode(dentry))))
-		return -EIO;
-
-	err = ovl_get_index_name(origin, &name);
-	if (err)
-		return err;
-
-	temp = ovl_create_temp(indexdir, OVL_CATTR(S_IFDIR | 0));
-	err = PTR_ERR(temp);
-	if (IS_ERR(temp))
-		goto free_name;
-
-	err = ovl_set_upper_fh(upper, temp);
-	if (err)
-		goto out;
-
-	index = lookup_one_len(name.name, indexdir, name.len);
-	if (IS_ERR(index)) {
-		err = PTR_ERR(index);
-	} else {
-		err = ovl_do_rename(dir, temp, dir, index, 0);
-		dput(index);
-	}
-out:
-	if (err)
-		ovl_cleanup(dir, temp);
-	dput(temp);
-free_name:
-	kfree(name.name);
-	return err;
-}
-
-struct ovl_copy_up_ctx {
-	struct dentry *parent;
-	struct dentry *dentry;
-	struct path lowerpath;
-	struct kstat stat;
-	struct kstat pstat;
-	const char *link;
-	struct dentry *destdir;
-	struct qstr destname;
-	struct dentry *workdir;
-	bool tmpfile;
-	bool origin;
-	bool indexed;
-};
-
-static int ovl_link_up(struct ovl_copy_up_ctx *c)
-{
-	int err;
-	struct dentry *upper;
-	struct dentry *upperdir = ovl_dentry_upper(c->parent);
-	struct inode *udir = d_inode(upperdir);
-
-	/* Mark parent "impure" because it may now contain non-pure upper */
-	err = ovl_set_impure(c->parent, upperdir);
-	if (err)
-		return err;
-
-	err = ovl_set_nlink_lower(c->dentry);
-	if (err)
-		return err;
-
-	inode_lock_nested(udir, I_MUTEX_PARENT);
-	upper = lookup_one_len(c->dentry->d_name.name, upperdir,
-			       c->dentry->d_name.len);
-	err = PTR_ERR(upper);
-	if (!IS_ERR(upper)) {
-		err = ovl_do_link(ovl_dentry_upper(c->dentry), udir, upper);
-		dput(upper);
-
-		if (!err) {
-			/* Restore timestamps on parent (best effort) */
-			ovl_set_timestamps(upperdir, &c->pstat);
-			ovl_dentry_set_upper_alias(c->dentry);
-		}
-	}
-	inode_unlock(udir);
-	if (err)
-		return err;
-
-	err = ovl_set_nlink_upper(c->dentry);
-
-	return err;
-}
-
-static int ovl_install_temp(struct ovl_copy_up_ctx *c, struct dentry *temp,
-			    struct dentry **newdentry)
-{
-	int err;
-	struct dentry *upper;
-	struct inode *udir = d_inode(c->destdir);
-
-	upper = lookup_one_len(c->destname.name, c->destdir, c->destname.len);
-	if (IS_ERR(upper))
-		return PTR_ERR(upper);
-
-	if (c->tmpfile)
-		err = ovl_do_link(temp, udir, upper);
-	else
-		err = ovl_do_rename(d_inode(c->workdir), temp, udir, upper, 0);
-
-	if (!err)
-		*newdentry = dget(c->tmpfile ? upper : temp);
-	dput(upper);
-
-	return err;
-}
-
-static struct dentry *ovl_get_tmpfile(struct ovl_copy_up_ctx *c)
-{
-	int err;
-	struct dentry *temp;
-	const struct cred *old_creds = NULL;
-	struct cred *new_creds = NULL;
-	struct ovl_cattr cattr = {
-		/* Can't properly set mode on creation because of the umask */
-		.mode = c->stat.mode & S_IFMT,
-		.rdev = c->stat.rdev,
-		.link = c->link
-	};
-
-	err = security_inode_copy_up(c->dentry, &new_creds);
-	temp = ERR_PTR(err);
-	if (err < 0)
-		goto out;
-
-	if (new_creds)
-		old_creds = override_creds(new_creds);
-
-	if (c->tmpfile)
-		temp = ovl_do_tmpfile(c->workdir, c->stat.mode);
-	else
-		temp = ovl_create_temp(c->workdir, &cattr);
-out:
-	if (new_creds) {
-		revert_creds(old_creds);
-		put_cred(new_creds);
-	}
-
-	return temp;
-}
-
-static int ovl_copy_up_inode(struct ovl_copy_up_ctx *c, struct dentry *temp)
-{
-	int err;
-
-	if (S_ISREG(c->stat.mode)) {
-		struct path upperpath;
-
-		ovl_path_upper(c->dentry, &upperpath);
-		BUG_ON(upperpath.dentry != NULL);
-		upperpath.dentry = temp;
-
-		err = ovl_copy_up_data(&c->lowerpath, &upperpath, c->stat.size);
-		if (err)
-			return err;
-	}
-
-	err = ovl_copy_xattr(c->lowerpath.dentry, temp);
-	if (err)
-		return err;
-
-	inode_lock(temp->d_inode);
-	err = ovl_set_attr(temp, &c->stat);
-	inode_unlock(temp->d_inode);
-	if (err)
-		return err;
-
-	/*
-	 * Store identifier of lower inode in upper inode xattr to
-	 * allow lookup of the copy up origin inode.
-	 *
-	 * Don't set origin when we are breaking the association with a lower
-	 * hard link.
-	 */
-	if (c->origin) {
-		err = ovl_set_origin(c->dentry, c->lowerpath.dentry, temp);
-		if (err)
-			return err;
-	}
-
-	return 0;
-}
-
-static int ovl_copy_up_locked(struct ovl_copy_up_ctx *c)
-{
-	struct inode *udir = c->destdir->d_inode;
-	struct inode *inode;
-	struct dentry *newdentry = NULL;
-	struct dentry *temp;
-	int err;
-
-	temp = ovl_get_tmpfile(c);
-	if (IS_ERR(temp))
-		return PTR_ERR(temp);
-
-	err = ovl_copy_up_inode(c, temp);
-	if (err)
-		goto out;
-
-	if (S_ISDIR(c->stat.mode) && c->indexed) {
-		err = ovl_create_index(c->dentry, c->lowerpath.dentry, temp);
-		if (err)
-			goto out;
-	}
-
-	if (c->tmpfile) {
-		inode_lock_nested(udir, I_MUTEX_PARENT);
-		err = ovl_install_temp(c, temp, &newdentry);
-		inode_unlock(udir);
-	} else {
-		err = ovl_install_temp(c, temp, &newdentry);
-	}
-	if (err)
-		goto out;
-
-	inode = d_inode(c->dentry);
-	ovl_inode_update(inode, newdentry);
-	if (S_ISDIR(inode->i_mode))
-		ovl_set_flag(OVL_WHITEOUTS, inode);
-
-out:
-	if (err && !c->tmpfile)
-		ovl_cleanup(d_inode(c->workdir), temp);
-	dput(temp);
-	return err;
-
-}
-
-/*
- * Copy up a single dentry
- *
- * All renames start with copy up of source if necessary.  The actual
- * rename will only proceed once the copy up was successful.  Copy up uses
- * upper parent i_mutex for exclusion.  Since rename can change d_parent it
- * is possible that the copy up will lock the old parent.  At that point
- * the file will have already been copied up anyway.
- */
-static int ovl_do_copy_up(struct ovl_copy_up_ctx *c)
-{
-	int err;
-	struct ovl_fs *ofs = c->dentry->d_sb->s_fs_info;
-	bool to_index = false;
-
-	/*
-	 * Indexed non-dir is copied up directly to the index entry and then
-	 * hardlinked to upper dir. Indexed dir is copied up to indexdir,
-	 * then index entry is created and then copied up dir installed.
-	 * Copying dir up to indexdir instead of workdir simplifies locking.
-	 */
-	if (ovl_need_index(c->dentry)) {
-		c->indexed = true;
-		if (S_ISDIR(c->stat.mode))
-			c->workdir = ovl_indexdir(c->dentry->d_sb);
-		else
-			to_index = true;
-	}
-
-	if (S_ISDIR(c->stat.mode) || c->stat.nlink == 1 || to_index)
-		c->origin = true;
-
-	if (to_index) {
-		c->destdir = ovl_indexdir(c->dentry->d_sb);
-		err = ovl_get_index_name(c->lowerpath.dentry, &c->destname);
-		if (err)
-			return err;
-	} else if (WARN_ON(!c->parent)) {
-		/* Disconnected dentry must be copied up to index dir */
-		return -EIO;
-	} else {
-		/*
-		 * Mark parent "impure" because it may now contain non-pure
-		 * upper
-		 */
-		err = ovl_set_impure(c->parent, c->destdir);
-		if (err)
-			return err;
-	}
-
-	/* Should we copyup with O_TMPFILE or with workdir? */
-	if (S_ISREG(c->stat.mode) && ofs->tmpfile) {
-		c->tmpfile = true;
-		err = ovl_copy_up_locked(c);
-	} else {
-		err = ovl_lock_rename_workdir(c->workdir, c->destdir);
-		if (!err) {
-			err = ovl_copy_up_locked(c);
-			unlock_rename(c->workdir, c->destdir);
-		}
-	}
-
-
-	if (err)
-		goto out;
-
-	if (c->indexed)
-		ovl_set_flag(OVL_INDEX, d_inode(c->dentry));
-
-	if (to_index) {
-		/* Initialize nlink for copy up of disconnected dentry */
-		err = ovl_set_nlink_upper(c->dentry);
-	} else {
-		struct inode *udir = d_inode(c->destdir);
-
-		/* Restore timestamps on parent (best effort) */
-		inode_lock(udir);
-		ovl_set_timestamps(c->destdir, &c->pstat);
-		inode_unlock(udir);
-
-		ovl_dentry_set_upper_alias(c->dentry);
-	}
-
-out:
-	if (to_index)
-		kfree(c->destname.name);
-	return err;
-}
-
-static int ovl_copy_up_one(struct dentry *parent, struct dentry *dentry,
-			   int flags)
-{
-	int err;
-	DEFINE_DELAYED_CALL(done);
-	struct path parentpath;
-	struct ovl_copy_up_ctx ctx = {
-		.parent = parent,
-		.dentry = dentry,
-		.workdir = ovl_workdir(dentry),
-	};
-
-	if (WARN_ON(!ctx.workdir))
-		return -EROFS;
-
-	ovl_path_lower(dentry, &ctx.lowerpath);
-	err = vfs_getattr(&ctx.lowerpath, &ctx.stat,
-			  STATX_BASIC_STATS, AT_STATX_SYNC_AS_STAT);
-	if (err)
-		return err;
-
-	if (parent) {
-		ovl_path_upper(parent, &parentpath);
-		ctx.destdir = parentpath.dentry;
-		ctx.destname = dentry->d_name;
-
-		err = vfs_getattr(&parentpath, &ctx.pstat,
-				  STATX_ATIME | STATX_MTIME,
-				  AT_STATX_SYNC_AS_STAT);
-		if (err)
-			return err;
-	}
-
-	/* maybe truncate regular file. this has no effect on dirs */
-	if (flags & O_TRUNC)
-		ctx.stat.size = 0;
-
-	if (S_ISLNK(ctx.stat.mode)) {
-		ctx.link = vfs_get_link(ctx.lowerpath.dentry, &done);
-		if (IS_ERR(ctx.link))
-			return PTR_ERR(ctx.link);
-	}
-	ovl_do_check_copy_up(ctx.lowerpath.dentry);
-
-	err = ovl_copy_up_start(dentry);
-	/* err < 0: interrupted, err > 0: raced with another copy-up */
-	if (unlikely(err)) {
-		if (err > 0)
-			err = 0;
-	} else {
-		if (!ovl_dentry_upper(dentry))
-			err = ovl_do_copy_up(&ctx);
-		if (!err && parent && !ovl_dentry_has_upper_alias(dentry))
-			err = ovl_link_up(&ctx);
-		ovl_copy_up_end(dentry);
-	}
-	do_delayed_call(&done);
-
-	return err;
-}
-
-int ovl_copy_up_flags(struct dentry *dentry, int flags)
-{
-	int err = 0;
-	const struct cred *old_cred = ovl_override_creds(dentry->d_sb);
-	bool disconnected = (dentry->d_flags & DCACHE_DISCONNECTED);
-
-	/*
-	 * With NFS export, copy up can get called for a disconnected non-dir.
-	 * In this case, we will copy up lower inode to index dir without
-	 * linking it to upper dir.
-	 */
-	if (WARN_ON(disconnected && d_is_dir(dentry)))
-		return -EIO;
-
-	while (!err) {
-		struct dentry *next;
-		struct dentry *parent = NULL;
-
-		/*
-		 * Check if copy-up has happened as well as for upper alias (in
-		 * case of hard links) is there.
-		 *
-		 * Both checks are lockless:
-		 *  - false negatives: will recheck under oi->lock
-		 *  - false positives:
-		 *    + ovl_dentry_upper() uses memory barriers to ensure the
-		 *      upper dentry is up-to-date
-		 *    + ovl_dentry_has_upper_alias() relies on locking of
-		 *      upper parent i_rwsem to prevent reordering copy-up
-		 *      with rename.
-		 */
-		if (ovl_dentry_upper(dentry) &&
-		    (ovl_dentry_has_upper_alias(dentry) || disconnected))
-			break;
-
-		next = dget(dentry);
-		/* find the topmost dentry not yet copied up */
-		for (; !disconnected;) {
-			parent = dget_parent(next);
-
-			if (ovl_dentry_upper(parent))
-				break;
-
-			dput(next);
-			next = parent;
-		}
-
-		err = ovl_copy_up_one(parent, next, flags);
-
-		dput(parent);
-		dput(next);
-	}
-	revert_creds(old_cred);
-
-	return err;
-}
-
-int ovl_copy_up(struct dentry *dentry)
-{
-	return ovl_copy_up_flags(dentry, 0);
-}

executer/kernel/mcoverlayfs/linux-4.18.14/export.c

@@ -1,853 +0,0 @@
-/*
- * Overlayfs NFS export support.
- *
- * Amir Goldstein <amir73il@gmail.com>
- *
- * Copyright (C) 2017-2018 CTERA Networks. All Rights Reserved.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/cred.h>
-#include <linux/mount.h>
-#include <linux/namei.h>
-#include <linux/xattr.h>
-#include <linux/exportfs.h>
-#include <linux/ratelimit.h>
-#include <linux/version.h>
-#include "overlayfs.h"
-
-static int ovl_encode_maybe_copy_up(struct dentry *dentry)
-{
-	int err;
-
-	if (ovl_dentry_upper(dentry))
-		return 0;
-
-	err = ovl_want_write(dentry);
-	if (!err) {
-		err = ovl_copy_up(dentry);
-		ovl_drop_write(dentry);
-	}
-
-	if (err) {
-		pr_warn_ratelimited("overlayfs: failed to copy up on encode (%pd2, err=%i)\n",
-				    dentry, err);
-	}
-
-	return err;
-}
-
-/*
- * Before encoding a non-upper directory file handle from real layer N, we need
- * to check if it will be possible to reconnect an overlay dentry from the real
- * lower decoded dentry. This is done by following the overlay ancestry up to a
- * "layer N connected" ancestor and verifying that all parents along the way are
- * "layer N connectable". If an ancestor that is NOT "layer N connectable" is
- * found, we need to copy up an ancestor, which is "layer N connectable", thus
- * making that ancestor "layer N connected". For example:
- *
- * layer 1: /a
- * layer 2: /a/b/c
- *
- * The overlay dentry /a is NOT "layer 2 connectable", because if dir /a is
- * copied up and renamed, upper dir /a will be indexed by lower dir /a from
- * layer 1. The dir /a from layer 2 will never be indexed, so the algorithm (*)
- * in ovl_lookup_real_ancestor() will not be able to lookup a connected overlay
- * dentry from the connected lower dentry /a/b/c.
- *
- * To avoid this problem on decode time, we need to copy up an ancestor of
- * /a/b/c, which is "layer 2 connectable", on encode time. That ancestor is
- * /a/b. After copy up (and index) of /a/b, it will become "layer 2 connected"
- * and when the time comes to decode the file handle from lower dentry /a/b/c,
- * ovl_lookup_real_ancestor() will find the indexed ancestor /a/b and decoding
- * a connected overlay dentry will be accomplished.
- *
- * (*) the algorithm in ovl_lookup_real_ancestor() can be improved to lookup an
- * entry /a in the lower layers above layer N and find the indexed dir /a from
- * layer 1. If that improvement is made, then the check for "layer N connected"
- * will need to verify there are no redirects in lower layers above N. In the
- * example above, /a will be "layer 2 connectable". However, if layer 2 dir /a
- * is a target of a layer 1 redirect, then /a will NOT be "layer 2 connectable":
- *
- * layer 1: /A (redirect = /a)
- * layer 2: /a/b/c
- */
-
-/* Return the lowest layer for encoding a connectable file handle */
-static int ovl_connectable_layer(struct dentry *dentry)
-{
-	struct ovl_entry *oe = OVL_E(dentry);
-
-	/* We can get overlay root from root of any layer */
-	if (dentry == dentry->d_sb->s_root)
-		return oe->numlower;
-
-	/*
-	 * If it's an unindexed merge dir, then it's not connectable with any
-	 * lower layer
-	 */
-	if (ovl_dentry_upper(dentry) &&
-	    !ovl_test_flag(OVL_INDEX, d_inode(dentry)))
-		return 0;
-
-	/* We can get upper/overlay path from indexed/lower dentry */
-	return oe->lowerstack[0].layer->idx;
-}
-
-/*
- * @dentry is "connected" if all ancestors up to root or a "connected" ancestor
- * have the same uppermost lower layer as the origin's layer. We may need to
- * copy up a "connectable" ancestor to make it "connected". A "connected" dentry
- * cannot become non "connected", so cache positive result in dentry flags.
- *
- * Return the connected origin layer or < 0 on error.
- */
-static int ovl_connect_layer(struct dentry *dentry)
-{
-	struct dentry *next, *parent = NULL;
-	int origin_layer;
-	int err = 0;
-
-	if (WARN_ON(dentry == dentry->d_sb->s_root) ||
-	    WARN_ON(!ovl_dentry_lower(dentry)))
-		return -EIO;
-
-	origin_layer = OVL_E(dentry)->lowerstack[0].layer->idx;
-	if (ovl_dentry_test_flag(OVL_E_CONNECTED, dentry))
-		return origin_layer;
-
-	/* Find the topmost origin layer connectable ancestor of @dentry */
-	next = dget(dentry);
-	for (;;) {
-		parent = dget_parent(next);
-		if (WARN_ON(parent == next)) {
-			err = -EIO;
-			break;
-		}
-
-		/*
-		 * If @parent is not origin layer connectable, then copy up
-		 * @next which is origin layer connectable and we are done.
-		 */
-		if (ovl_connectable_layer(parent) < origin_layer) {
-			err = ovl_encode_maybe_copy_up(next);
-			break;
-		}
-
-		/* If @parent is connected or indexed we are done */
-		if (ovl_dentry_test_flag(OVL_E_CONNECTED, parent) ||
-		    ovl_test_flag(OVL_INDEX, d_inode(parent)))
-			break;
-
-		dput(next);
-		next = parent;
-	}
-
-	dput(parent);
-	dput(next);
-
-	if (!err)
-		ovl_dentry_set_flag(OVL_E_CONNECTED, dentry);
-
-	return err ?: origin_layer;
-}
-
-/*
- * We only need to encode origin if there is a chance that the same object was
- * encoded pre copy up and then we need to stay consistent with the same
- * encoding also after copy up. If non-pure upper is not indexed, then it was
- * copied up before NFS export was enabled. In that case we don't need to worry
- * about staying consistent with pre copy up encoding and we encode an upper
- * file handle. Overlay root dentry is a private case of non-indexed upper.
- *
- * The following table summarizes the different file handle encodings used for
- * different overlay object types:
- *
- *  Object type		| Encoding
- * --------------------------------
- *  Pure upper		| U
- *  Non-indexed upper	| U
- *  Indexed upper	| L (*)
- *  Non-upper		| L (*)
- *
- * U = upper file handle
- * L = lower file handle
- *
- * (*) Connecting an overlay dir from real lower dentry is not always
- * possible when there are redirects in lower layers and non-indexed merge dirs.
- * To mitigate those case, we may copy up the lower dir ancestor before encode
- * a lower dir file handle.
- *
- * Return 0 for upper file handle, > 0 for lower file handle or < 0 on error.
- */
-static int ovl_check_encode_origin(struct dentry *dentry)
-{
-	struct ovl_fs *ofs = dentry->d_sb->s_fs_info;
-
-	/* Upper file handle for pure upper */
-	if (!ovl_dentry_lower(dentry))
-		return 0;
-
-	/*
-	 * Upper file handle for non-indexed upper.
-	 *
-	 * Root is never indexed, so if there's an upper layer, encode upper for
-	 * root.
-	 */
-	if (ovl_dentry_upper(dentry) &&
-	    !ovl_test_flag(OVL_INDEX, d_inode(dentry)))
-		return 0;
-
-	/*
-	 * Decoding a merge dir, whose origin's ancestor is under a redirected
-	 * lower dir or under a non-indexed upper is not always possible.
-	 * ovl_connect_layer() will try to make origin's layer "connected" by
-	 * copying up a "connectable" ancestor.
-	 */
-	if (d_is_dir(dentry) && ofs->upper_mnt)
-		return ovl_connect_layer(dentry);
-
-	/* Lower file handle for indexed and non-upper dir/non-dir */
-	return 1;
-}
-
-static int ovl_d_to_fh(struct dentry *dentry, char *buf, int buflen)
-{
-	struct ovl_fh *fh = NULL;
-	int err, enc_lower;
-
-	/*
-	 * Check if we should encode a lower or upper file handle and maybe
-	 * copy up an ancestor to make lower file handle connectable.
-	 */
-	err = enc_lower = ovl_check_encode_origin(dentry);
-	if (enc_lower < 0)
-		goto fail;
-
-	/* Encode an upper or lower file handle */
-	fh = ovl_encode_real_fh(enc_lower ? ovl_dentry_lower(dentry) :
-				ovl_dentry_upper(dentry), !enc_lower);
-	err = PTR_ERR(fh);
-	if (IS_ERR(fh))
-		goto fail;
-
-	err = -EOVERFLOW;
-	if (fh->len > buflen)
-		goto fail;
-
-	memcpy(buf, (char *)fh, fh->len);
-	err = fh->len;
-
-out:
-	kfree(fh);
-	return err;
-
-fail:
-	pr_warn_ratelimited("overlayfs: failed to encode file handle (%pd2, err=%i, buflen=%d, len=%d, type=%d)\n",
-			    dentry, err, buflen, fh ? (int)fh->len : 0,
-			    fh ? fh->type : 0);
-	goto out;
-}
-
-static int ovl_dentry_to_fh(struct dentry *dentry, u32 *fid, int *max_len)
-{
-	int res, len = *max_len << 2;
-
-	res = ovl_d_to_fh(dentry, (char *)fid, len);
-	if (res <= 0)
-		return FILEID_INVALID;
-
-	len = res;
-
-	/* Round up to dwords */
-	*max_len = (len + 3) >> 2;
-	return OVL_FILEID;
-}
-
-static int ovl_encode_fh(struct inode *inode, u32 *fid, int *max_len,
-			 struct inode *parent)
-{
-	struct dentry *dentry;
-	int type;
-
-	/* TODO: encode connectable file handles */
-	if (parent)
-		return FILEID_INVALID;
-
-	dentry = d_find_any_alias(inode);
-	if (WARN_ON(!dentry))
-		return FILEID_INVALID;
-
-	type = ovl_dentry_to_fh(dentry, fid, max_len);
-
-	dput(dentry);
-	return type;
-}
-
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 16, 0)
-/*
- * Find or instantiate an overlay dentry from real dentries and index.
- */
-static struct dentry *ovl_obtain_alias(struct super_block *sb,
-				       struct dentry *upper_alias,
-				       struct ovl_path *lowerpath,
-				       struct dentry *index)
-{
-	struct dentry *lower = lowerpath ? lowerpath->dentry : NULL;
-	struct dentry *upper = upper_alias ?: index;
-	struct dentry *dentry;
-	struct inode *inode;
-	struct ovl_entry *oe;
-	struct ovl_inode_params oip = {
-		.lowerpath = lowerpath,
-		.index = index,
-		.numlower = !!lower
-	};
-
-	/* We get overlay directory dentries with ovl_lookup_real() */
-	if (d_is_dir(upper ?: lower))
-		return ERR_PTR(-EIO);
-
-	oip.upperdentry = dget(upper);
-	inode = ovl_get_inode(sb, &oip);
-	if (IS_ERR(inode)) {
-		dput(upper);
-		return ERR_CAST(inode);
-	}
-
-	dentry = d_find_any_alias(inode);
-	if (!dentry) {
-		dentry = d_alloc_anon(inode->i_sb);
-		if (!dentry)
-			goto nomem;
-		oe = ovl_alloc_entry(lower ? 1 : 0);
-		if (!oe)
-			goto nomem;
-
-		if (lower) {
-			oe->lowerstack->dentry = dget(lower);
-			oe->lowerstack->layer = lowerpath->layer;
-		}
-		dentry->d_fsdata = oe;
-		if (upper_alias)
-			ovl_dentry_set_upper_alias(dentry);
-	}
-
-	return d_instantiate_anon(dentry, inode);
-
-nomem:
-	iput(inode);
-	dput(dentry);
-	return ERR_PTR(-ENOMEM);
-}
-
-/* Get the upper or lower dentry in stach whose on layer @idx */
-static struct dentry *ovl_dentry_real_at(struct dentry *dentry, int idx)
-{
-	struct ovl_entry *oe = dentry->d_fsdata;
-	int i;
-
-	if (!idx)
-		return ovl_dentry_upper(dentry);
-
-	for (i = 0; i < oe->numlower; i++) {
-		if (oe->lowerstack[i].layer->idx == idx)
-			return oe->lowerstack[i].dentry;
-	}
-
-	return NULL;
-}
-
-/*
- * Lookup a child overlay dentry to get a connected overlay dentry whose real
- * dentry is @real. If @real is on upper layer, we lookup a child overlay
- * dentry with the same name as the real dentry. Otherwise, we need to consult
- * index for lookup.
- */
-static struct dentry *ovl_lookup_real_one(struct dentry *connected,
-					  struct dentry *real,
-					  struct ovl_layer *layer)
-{
-	struct inode *dir = d_inode(connected);
-	struct dentry *this, *parent = NULL;
-	struct name_snapshot name;
-	int err;
-
-	/*
-	 * Lookup child overlay dentry by real name. The dir mutex protects us
-	 * from racing with overlay rename. If the overlay dentry that is above
-	 * real has already been moved to a parent that is not under the
-	 * connected overlay dir, we return -ECHILD and restart the lookup of
-	 * connected real path from the top.
-	 */
-	inode_lock_nested(dir, I_MUTEX_PARENT);
-	err = -ECHILD;
-	parent = dget_parent(real);
-	if (ovl_dentry_real_at(connected, layer->idx) != parent)
-		goto fail;
-
-	/*
-	 * We also need to take a snapshot of real dentry name to protect us
-	 * from racing with underlying layer rename. In this case, we don't
-	 * care about returning ESTALE, only from dereferencing a free name
-	 * pointer because we hold no lock on the real dentry.
-	 */
-	take_dentry_name_snapshot(&name, real);
-	this = lookup_one_len(name.name, connected, strlen(name.name));
-	err = PTR_ERR(this);
-	if (IS_ERR(this)) {
-		goto fail;
-	} else if (!this || !this->d_inode) {
-		dput(this);
-		err = -ENOENT;
-		goto fail;
-	} else if (ovl_dentry_real_at(this, layer->idx) != real) {
-		dput(this);
-		err = -ESTALE;
-		goto fail;
-	}
-
-out:
-	release_dentry_name_snapshot(&name);
-	dput(parent);
-	inode_unlock(dir);
-	return this;
-
-fail:
-	pr_warn_ratelimited("overlayfs: failed to lookup one by real (%pd2, layer=%d, connected=%pd2, err=%i)\n",
-			    real, layer->idx, connected, err);
-	this = ERR_PTR(err);
-	goto out;
-}
-
-static struct dentry *ovl_lookup_real(struct super_block *sb,
-				      struct dentry *real,
-				      struct ovl_layer *layer);
-
-/*
- * Lookup an indexed or hashed overlay dentry by real inode.
- */
-static struct dentry *ovl_lookup_real_inode(struct super_block *sb,
-					    struct dentry *real,
-					    struct ovl_layer *layer)
-{
-	struct ovl_fs *ofs = sb->s_fs_info;
-	struct ovl_layer upper_layer = { .mnt = ofs->upper_mnt };
-	struct dentry *index = NULL;
-	struct dentry *this = NULL;
-	struct inode *inode;
-
-	/*
-	 * Decoding upper dir from index is expensive, so first try to lookup
-	 * overlay dentry in inode/dcache.
-	 */
-	inode = ovl_lookup_inode(sb, real, !layer->idx);
-	if (IS_ERR(inode))
-		return ERR_CAST(inode);
-	if (inode) {
-		this = d_find_any_alias(inode);
-		iput(inode);
-	}
-
-	/*
-	 * For decoded lower dir file handle, lookup index by origin to check
-	 * if lower dir was copied up and and/or removed.
-	 */
-	if (!this && layer->idx && ofs->indexdir && !WARN_ON(!d_is_dir(real))) {
-		index = ovl_lookup_index(ofs, NULL, real, false);
-		if (IS_ERR(index))
-			return index;
-	}
-
-	/* Get connected upper overlay dir from index */
-	if (index) {
-		struct dentry *upper = ovl_index_upper(ofs, index);
-
-		dput(index);
-		if (IS_ERR_OR_NULL(upper))
-			return upper;
-
-		/*
-		 * ovl_lookup_real() in lower layer may call recursively once to
-		 * ovl_lookup_real() in upper layer. The first level call walks
-		 * back lower parents to the topmost indexed parent. The second
-		 * recursive call walks back from indexed upper to the topmost
-		 * connected/hashed upper parent (or up to root).
-		 */
-		this = ovl_lookup_real(sb, upper, &upper_layer);
-		dput(upper);
-	}
-
-	if (IS_ERR_OR_NULL(this))
-		return this;
-
-	if (WARN_ON(ovl_dentry_real_at(this, layer->idx) != real)) {
-		dput(this);
-		this = ERR_PTR(-EIO);
-	}
-
-	return this;
-}
-
-/*
- * Lookup an indexed or hashed overlay dentry, whose real dentry is an
- * ancestor of @real.
- */
-static struct dentry *ovl_lookup_real_ancestor(struct super_block *sb,
-					       struct dentry *real,
-					       struct ovl_layer *layer)
-{
-	struct dentry *next, *parent = NULL;
-	struct dentry *ancestor = ERR_PTR(-EIO);
-
-	if (real == layer->mnt->mnt_root)
-		return dget(sb->s_root);
-
-	/* Find the topmost indexed or hashed ancestor */
-	next = dget(real);
-	for (;;) {
-		parent = dget_parent(next);
-
-		/*
-		 * Lookup a matching overlay dentry in inode/dentry
-		 * cache or in index by real inode.
-		 */
-		ancestor = ovl_lookup_real_inode(sb, next, layer);
-		if (ancestor)
-			break;
-
-		if (parent == layer->mnt->mnt_root) {
-			ancestor = dget(sb->s_root);
-			break;
-		}
-
-		/*
-		 * If @real has been moved out of the layer root directory,
-		 * we will eventully hit the real fs root. This cannot happen
-		 * by legit overlay rename, so we return error in that case.
-		 */
-		if (parent == next) {
-			ancestor = ERR_PTR(-EXDEV);
-			break;
-		}
-
-		dput(next);
-		next = parent;
-	}
-
-	dput(parent);
-	dput(next);
-
-	return ancestor;
-}
-
-/*
- * Lookup a connected overlay dentry whose real dentry is @real.
- * If @real is on upper layer, we lookup a child overlay dentry with the same
- * path the real dentry. Otherwise, we need to consult index for lookup.
- */
-static struct dentry *ovl_lookup_real(struct super_block *sb,
-				      struct dentry *real,
-				      struct ovl_layer *layer)
-{
-	struct dentry *connected;
-	int err = 0;
-
-	connected = ovl_lookup_real_ancestor(sb, real, layer);
-	if (IS_ERR(connected))
-		return connected;
-
-	while (!err) {
-		struct dentry *next, *this;
-		struct dentry *parent = NULL;
-		struct dentry *real_connected = ovl_dentry_real_at(connected,
-								   layer->idx);
-
-		if (real_connected == real)
-			break;
-
-		/* Find the topmost dentry not yet connected */
-		next = dget(real);
-		for (;;) {
-			parent = dget_parent(next);
-
-			if (parent == real_connected)
-				break;
-
-			/*
-			 * If real has been moved out of 'real_connected',
-			 * we will not find 'real_connected' and hit the layer
-			 * root. In that case, we need to restart connecting.
-			 * This game can go on forever in the worst case. We
-			 * may want to consider taking s_vfs_rename_mutex if
-			 * this happens more than once.
-			 */
-			if (parent == layer->mnt->mnt_root) {
-				dput(connected);
-				connected = dget(sb->s_root);
-				break;
-			}
-
-			/*
-			 * If real file has been moved out of the layer root
-			 * directory, we will eventully hit the real fs root.
-			 * This cannot happen by legit overlay rename, so we
-			 * return error in that case.
-			 */
-			if (parent == next) {
-				err = -EXDEV;
-				break;
-			}
-
-			dput(next);
-			next = parent;
-		}
-
-		if (!err) {
-			this = ovl_lookup_real_one(connected, next, layer);
-			if (IS_ERR(this))
-				err = PTR_ERR(this);
-
-			/*
-			 * Lookup of child in overlay can fail when racing with
-			 * overlay rename of child away from 'connected' parent.
-			 * In this case, we need to restart the lookup from the
-			 * top, because we cannot trust that 'real_connected' is
-			 * still an ancestor of 'real'. There is a good chance
-			 * that the renamed overlay ancestor is now in cache, so
-			 * ovl_lookup_real_ancestor() will find it and we can
-			 * continue to connect exactly from where lookup failed.
-			 */
-			if (err == -ECHILD) {
-				this = ovl_lookup_real_ancestor(sb, real,
-								layer);
-				err = PTR_ERR_OR_ZERO(this);
-			}
-			if (!err) {
-				dput(connected);
-				connected = this;
-			}
-		}
-
-		dput(parent);
-		dput(next);
-	}
-
-	if (err)
-		goto fail;
-
-	return connected;
-
-fail:
-	pr_warn_ratelimited("overlayfs: failed to lookup by real (%pd2, layer=%d, connected=%pd2, err=%i)\n",
-			    real, layer->idx, connected, err);
-	dput(connected);
-	return ERR_PTR(err);
-}
-
-/*
- * Get an overlay dentry from upper/lower real dentries and index.
- */
-static struct dentry *ovl_get_dentry(struct super_block *sb,
-				     struct dentry *upper,
-				     struct ovl_path *lowerpath,
-				     struct dentry *index)
-{
-	struct ovl_fs *ofs = sb->s_fs_info;
-	struct ovl_layer upper_layer = { .mnt = ofs->upper_mnt };
-	struct ovl_layer *layer = upper ? &upper_layer : lowerpath->layer;
-	struct dentry *real = upper ?: (index ?: lowerpath->dentry);
-
-	/*
-	 * Obtain a disconnected overlay dentry from a non-dir real dentry
-	 * and index.
-	 */
-	if (!d_is_dir(real))
-		return ovl_obtain_alias(sb, upper, lowerpath, index);
-
-	/* Removed empty directory? */
-	if ((real->d_flags & DCACHE_DISCONNECTED) || d_unhashed(real))
-		return ERR_PTR(-ENOENT);
-
-	/*
-	 * If real dentry is connected and hashed, get a connected overlay
-	 * dentry whose real dentry is @real.
-	 */
-	return ovl_lookup_real(sb, real, layer);
-}
-
-static struct dentry *ovl_upper_fh_to_d(struct super_block *sb,
-					struct ovl_fh *fh)
-{
-	struct ovl_fs *ofs = sb->s_fs_info;
-	struct dentry *dentry;
-	struct dentry *upper;
-
-	if (!ofs->upper_mnt)
-		return ERR_PTR(-EACCES);
-
-	upper = ovl_decode_real_fh(fh, ofs->upper_mnt, true);
-	if (IS_ERR_OR_NULL(upper))
-		return upper;
-
-	dentry = ovl_get_dentry(sb, upper, NULL, NULL);
-	dput(upper);
-
-	return dentry;
-}
-
-static struct dentry *ovl_lower_fh_to_d(struct super_block *sb,
-					struct ovl_fh *fh)
-{
-	struct ovl_fs *ofs = sb->s_fs_info;
-	struct ovl_path origin = { };
-	struct ovl_path *stack = &origin;
-	struct dentry *dentry = NULL;
-	struct dentry *index = NULL;
-	struct inode *inode;
-	int err;
-
-	/* First lookup overlay inode in inode cache by origin fh */
-	err = ovl_check_origin_fh(ofs, fh, false, NULL, &stack);
-	if (err)
-		return ERR_PTR(err);
-
-	if (!d_is_dir(origin.dentry) ||
-	    !(origin.dentry->d_flags & DCACHE_DISCONNECTED)) {
-		inode = ovl_lookup_inode(sb, origin.dentry, false);
-		err = PTR_ERR(inode);
-		if (IS_ERR(inode))
-			goto out_err;
-		if (inode) {
-			dentry = d_find_any_alias(inode);
-			iput(inode);
-			if (dentry)
-				goto out;
-		}
-	}
-
-	/* Then lookup indexed upper/whiteout by origin fh */
-	if (ofs->indexdir) {
-		index = ovl_get_index_fh(ofs, fh);
-		err = PTR_ERR(index);
-		if (IS_ERR(index)) {
-			index = NULL;
-			goto out_err;
-		}
-	}
-
-	/* Then try to get a connected upper dir by index */
-	if (index && d_is_dir(index)) {
-		struct dentry *upper = ovl_index_upper(ofs, index);
-
-		err = PTR_ERR(upper);
-		if (IS_ERR_OR_NULL(upper))
-			goto out_err;
-
-		dentry = ovl_get_dentry(sb, upper, NULL, NULL);
-		dput(upper);
-		goto out;
-	}
-
-	/* Otherwise, get a connected non-upper dir or disconnected non-dir */
-	if (d_is_dir(origin.dentry) &&
-	    (origin.dentry->d_flags & DCACHE_DISCONNECTED)) {
-		dput(origin.dentry);
-		origin.dentry = NULL;
-		err = ovl_check_origin_fh(ofs, fh, true, NULL, &stack);
-		if (err)
-			goto out_err;
-	}
-	if (index) {
-		err = ovl_verify_origin(index, origin.dentry, false);
-		if (err)
-			goto out_err;
-	}
-
-	dentry = ovl_get_dentry(sb, NULL, &origin, index);
-
-out:
-	dput(origin.dentry);
-	dput(index);
-	return dentry;
-
-out_err:
-	dentry = ERR_PTR(err);
-	goto out;
-}
-
-static struct dentry *ovl_fh_to_dentry(struct super_block *sb, struct fid *fid,
-				       int fh_len, int fh_type)
-{
-	struct dentry *dentry = NULL;
-	struct ovl_fh *fh = (struct ovl_fh *) fid;
-	int len = fh_len << 2;
-	unsigned int flags = 0;
-	int err;
-
-	err = -EINVAL;
-	if (fh_type != OVL_FILEID)
-		goto out_err;
-
-	err = ovl_check_fh_len(fh, len);
-	if (err)
-		goto out_err;
-
-	flags = fh->flags;
-	dentry = (flags & OVL_FH_FLAG_PATH_UPPER) ?
-		 ovl_upper_fh_to_d(sb, fh) :
-		 ovl_lower_fh_to_d(sb, fh);
-	err = PTR_ERR(dentry);
-	if (IS_ERR(dentry) && err != -ESTALE)
-		goto out_err;
-
-	return dentry;
-
-out_err:
-	pr_warn_ratelimited("overlayfs: failed to decode file handle (len=%d, type=%d, flags=%x, err=%i)\n",
-			    len, fh_type, flags, err);
-	return ERR_PTR(err);
-}
-
-static struct dentry *ovl_fh_to_parent(struct super_block *sb, struct fid *fid,
-				       int fh_len, int fh_type)
-{
-	pr_warn_ratelimited("overlayfs: connectable file handles not supported; use 'no_subtree_check' exportfs option.\n");
-	return ERR_PTR(-EACCES);
-}
-
-static int ovl_get_name(struct dentry *parent, char *name,
-			struct dentry *child)
-{
-	/*
-	 * ovl_fh_to_dentry() returns connected dir overlay dentries and
-	 * ovl_fh_to_parent() is not implemented, so we should not get here.
-	 */
-	WARN_ON_ONCE(1);
-	return -EIO;
-}
-
-static struct dentry *ovl_get_parent(struct dentry *dentry)
-{
-	/*
-	 * ovl_fh_to_dentry() returns connected dir overlay dentries, so we
-	 * should not get here.
-	 */
-	WARN_ON_ONCE(1);
-	return ERR_PTR(-EIO);
-}
-#endif
-
-const struct export_operations ovl_export_operations = {
-	.encode_fh	= ovl_encode_fh,
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 16, 0)
-	.fh_to_dentry	= ovl_fh_to_dentry,
-	.fh_to_parent	= ovl_fh_to_parent,
-	.get_name	= ovl_get_name,
-	.get_parent	= ovl_get_parent,
-#endif
-};

executer/kernel/mcoverlayfs/linux-4.18.14/inode.c

@@ -1,874 +0,0 @@
-/*
- *
- * Copyright (C) 2011 Novell Inc.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- */
-
-#include <linux/fs.h>
-#include <linux/slab.h>
-#include <linux/cred.h>
-#include <linux/xattr.h>
-#include <linux/posix_acl.h>
-#include <linux/ratelimit.h>
-#include <linux/version.h>
-#include "overlayfs.h"
-
-
-int ovl_setattr(struct dentry *dentry, struct iattr *attr)
-{
-	int err;
-	struct dentry *upperdentry;
-	const struct cred *old_cred;
-
-	/* NOCOPYUPW */
-	return 0;
-
-	/*
-	 * Check for permissions before trying to copy-up.  This is redundant
-	 * since it will be rechecked later by ->setattr() on upper dentry.  But
-	 * without this, copy-up can be triggered by just about anybody.
-	 *
-	 * We don't initialize inode->size, which just means that
-	 * inode_newsize_ok() will always check against MAX_LFS_FILESIZE and not
-	 * check for a swapfile (which this won't be anyway).
-	 */
-	err = setattr_prepare(dentry, attr);
-	if (err)
-		return err;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	err = ovl_copy_up(dentry);
-	if (!err) {
-		upperdentry = ovl_dentry_upper(dentry);
-
-		if (attr->ia_valid & (ATTR_KILL_SUID|ATTR_KILL_SGID))
-			attr->ia_valid &= ~ATTR_MODE;
-
-		inode_lock(upperdentry->d_inode);
-		old_cred = ovl_override_creds(dentry->d_sb);
-		err = notify_change(upperdentry, attr, NULL);
-		revert_creds(old_cred);
-		if (!err)
-			ovl_copyattr(upperdentry->d_inode, dentry->d_inode);
-		inode_unlock(upperdentry->d_inode);
-	}
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-static int ovl_map_dev_ino(struct dentry *dentry, struct kstat *stat,
-			   struct ovl_layer *lower_layer)
-{
-	bool samefs = ovl_same_sb(dentry->d_sb);
-	unsigned int xinobits = ovl_xino_bits(dentry->d_sb);
-
-	if (samefs) {
-		/*
-		 * When all layers are on the same fs, all real inode
-		 * number are unique, so we use the overlay st_dev,
-		 * which is friendly to du -x.
-		 */
-		stat->dev = dentry->d_sb->s_dev;
-		return 0;
-	} else if (xinobits) {
-		unsigned int shift = 64 - xinobits;
-		/*
-		 * All inode numbers of underlying fs should not be using the
-		 * high xinobits, so we use high xinobits to partition the
-		 * overlay st_ino address space. The high bits holds the fsid
-		 * (upper fsid is 0). This way overlay inode numbers are unique
-		 * and all inodes use overlay st_dev. Inode numbers are also
-		 * persistent for a given layer configuration.
-		 */
-		if (stat->ino >> shift) {
-			pr_warn_ratelimited("overlayfs: inode number too big (%pd2, ino=%llu, xinobits=%d)\n",
-					    dentry, stat->ino, xinobits);
-		} else {
-			if (lower_layer)
-				stat->ino |= ((u64)lower_layer->fsid) << shift;
-
-			stat->dev = dentry->d_sb->s_dev;
-			return 0;
-		}
-	}
-
-	/* The inode could not be mapped to a unified st_ino address space */
-	if (S_ISDIR(dentry->d_inode->i_mode)) {
-		/*
-		 * Always use the overlay st_dev for directories, so 'find
-		 * -xdev' will scan the entire overlay mount and won't cross the
-		 * overlay mount boundaries.
-		 *
-		 * If not all layers are on the same fs the pair {real st_ino;
-		 * overlay st_dev} is not unique, so use the non persistent
-		 * overlay st_ino for directories.
-		 */
-		stat->dev = dentry->d_sb->s_dev;
-		stat->ino = dentry->d_inode->i_ino;
-	} else if (lower_layer && lower_layer->fsid) {
-		/*
-		 * For non-samefs setup, if we cannot map all layers st_ino
-		 * to a unified address space, we need to make sure that st_dev
-		 * is unique per lower fs. Upper layer uses real st_dev and
-		 * lower layers use the unique anonymous bdev assigned to the
-		 * lower fs.
-		 */
-		stat->dev = lower_layer->fs->pseudo_dev;
-	}
-
-	return 0;
-}
-
-int ovl_getattr(const struct path *path, struct kstat *stat,
-		u32 request_mask, unsigned int flags)
-{
-	struct dentry *dentry = path->dentry;
-	enum ovl_path_type type;
-	struct path realpath;
-	const struct cred *old_cred;
-	bool is_dir = S_ISDIR(dentry->d_inode->i_mode);
-	bool samefs = ovl_same_sb(dentry->d_sb);
-	struct ovl_layer *lower_layer = NULL;
-	int err;
-
-	type = ovl_path_real(dentry, &realpath);
-	old_cred = ovl_override_creds(dentry->d_sb);
-	err = vfs_getattr(&realpath, stat, request_mask, flags);
-	if (err)
-		goto out;
-
-	/*
-	 * For non-dir or same fs, we use st_ino of the copy up origin.
-	 * This guaranties constant st_dev/st_ino across copy up.
-	 * With xino feature and non-samefs, we use st_ino of the copy up
-	 * origin masked with high bits that represent the layer id.
-	 *
-	 * If lower filesystem supports NFS file handles, this also guaranties
-	 * persistent st_ino across mount cycle.
-	 */
-	if (!is_dir || samefs || ovl_xino_bits(dentry->d_sb)) {
-		if (!OVL_TYPE_UPPER(type)) {
-			lower_layer = ovl_layer_lower(dentry);
-		} else if (OVL_TYPE_ORIGIN(type)) {
-			struct kstat lowerstat;
-			u32 lowermask = STATX_INO | (!is_dir ? STATX_NLINK : 0);
-
-			ovl_path_lower(dentry, &realpath);
-			err = vfs_getattr(&realpath, &lowerstat,
-					  lowermask, flags);
-			if (err)
-				goto out;
-
-			/*
-			 * Lower hardlinks may be broken on copy up to different
-			 * upper files, so we cannot use the lower origin st_ino
-			 * for those different files, even for the same fs case.
-			 *
-			 * Similarly, several redirected dirs can point to the
-			 * same dir on a lower layer. With the "verify_lower"
-			 * feature, we do not use the lower origin st_ino, if
-			 * we haven't verified that this redirect is unique.
-			 *
-			 * With inodes index enabled, it is safe to use st_ino
-			 * of an indexed origin. The index validates that the
-			 * upper hardlink is not broken and that a redirected
-			 * dir is the only redirect to that origin.
-			 */
-			if (ovl_test_flag(OVL_INDEX, d_inode(dentry)) ||
-			    (!ovl_verify_lower(dentry->d_sb) &&
-			     (is_dir || lowerstat.nlink == 1))) {
-				stat->ino = lowerstat.ino;
-				lower_layer = ovl_layer_lower(dentry);
-			}
-		}
-	}
-
-	err = ovl_map_dev_ino(dentry, stat, lower_layer);
-	if (err)
-		goto out;
-
-	/*
-	 * It's probably not worth it to count subdirs to get the
-	 * correct link count.  nlink=1 seems to pacify 'find' and
-	 * other utilities.
-	 */
-	if (is_dir && OVL_TYPE_MERGE(type))
-		stat->nlink = 1;
-
-	/*
-	 * Return the overlay inode nlinks for indexed upper inodes.
-	 * Overlay inode nlink counts the union of the upper hardlinks
-	 * and non-covered lower hardlinks. It does not include the upper
-	 * index hardlink.
-	 */
-	if (!is_dir && ovl_test_flag(OVL_INDEX, d_inode(dentry)))
-		stat->nlink = dentry->d_inode->i_nlink;
-
-out:
-	revert_creds(old_cred);
-
-	return err;
-}
-
-int ovl_permission(struct inode *inode, int mask)
-{
-	struct inode *upperinode = ovl_inode_upper(inode);
-	struct inode *realinode = upperinode ?: ovl_inode_lower(inode);
-	const struct cred *old_cred;
-	int err;
-
-	/* Careful in RCU walk mode */
-	if (!realinode) {
-		WARN_ON(!(mask & MAY_NOT_BLOCK));
-		return -ECHILD;
-	}
-
-	/*
-	 * Check overlay inode with the creds of task and underlying inode
-	 * with creds of mounter
-	 */
-	err = generic_permission(inode, mask);
-	if (err)
-		return err;
-
-	old_cred = ovl_override_creds(inode->i_sb);
-	if (!upperinode &&
-	    !special_file(realinode->i_mode) && mask & MAY_WRITE) {
-		mask &= ~(MAY_WRITE | MAY_APPEND);
-		/* Make sure mounter can read file for copy up later */
-		mask |= MAY_READ;
-	}
-	err = inode_permission(realinode, mask);
-	revert_creds(old_cred);
-
-	return err;
-}
-
-static int ovl_readlink(struct dentry *dentry, char __user *buf, int bufsiz)
-{
-	struct path realpath;
-	struct inode *realinode;
-
-	ovl_path_real(dentry, &realpath);
-	realinode = realpath.dentry->d_inode;
-
-	if (!realinode->i_op->readlink)
-		return -EINVAL;
-
-	touch_atime(&realpath);
-
-	return realinode->i_op->readlink(realpath.dentry, buf, bufsiz);
-}
-
-static const char *ovl_get_link(struct dentry *dentry,
-				struct inode *inode,
-				struct delayed_call *done)
-{
-	const struct cred *old_cred;
-	const char *p;
-
-	if (!dentry)
-		return ERR_PTR(-ECHILD);
-
-	old_cred = ovl_override_creds(dentry->d_sb);
-	p = vfs_get_link(ovl_dentry_real(dentry), done);
-	revert_creds(old_cred);
-	return p;
-}
-
-bool ovl_is_private_xattr(const char *name)
-{
-	return strncmp(name, OVL_XATTR_PREFIX,
-		       sizeof(OVL_XATTR_PREFIX) - 1) == 0;
-}
-
-int ovl_xattr_set(struct dentry *dentry, struct inode *inode, const char *name,
-		  const void *value, size_t size, int flags)
-{
-	int err;
-	struct dentry *upperdentry = ovl_i_dentry_upper(inode);
-	struct dentry *realdentry = upperdentry ?: ovl_dentry_lower(dentry);
-	const struct cred *old_cred;
-
-	/* NOCOPYUPW */
-	return 0;
-
-	err = ovl_want_write(dentry);
-	if (err)
-		goto out;
-
-	if (!value && !upperdentry) {
-		err = vfs_getxattr(realdentry, name, NULL, 0);
-		if (err < 0)
-			goto out_drop_write;
-	}
-
-	if (!upperdentry) {
-		err = ovl_copy_up(dentry);
-		if (err)
-			goto out_drop_write;
-
-		realdentry = ovl_dentry_upper(dentry);
-	}
-
-	old_cred = ovl_override_creds(dentry->d_sb);
-	if (value)
-		err = vfs_setxattr(realdentry, name, value, size, flags);
-	else {
-		WARN_ON(flags != XATTR_REPLACE);
-		err = vfs_removexattr(realdentry, name);
-	}
-	revert_creds(old_cred);
-
-out_drop_write:
-	ovl_drop_write(dentry);
-out:
-	return err;
-}
-
-int ovl_xattr_get(struct dentry *dentry, struct inode *inode, const char *name,
-		  void *value, size_t size)
-{
-	ssize_t res;
-	const struct cred *old_cred;
-	struct dentry *realdentry =
-		ovl_i_dentry_upper(inode) ?: ovl_dentry_lower(dentry);
-
-	old_cred = ovl_override_creds(dentry->d_sb);
-	res = vfs_getxattr(realdentry, name, value, size);
-	revert_creds(old_cred);
-	return res;
-}
-
-static bool ovl_can_list(const char *s)
-{
-	/* List all non-trusted xatts */
-	if (strncmp(s, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN) != 0)
-		return true;
-
-	/* Never list trusted.overlay, list other trusted for superuser only */
-	return !ovl_is_private_xattr(s) && capable(CAP_SYS_ADMIN);
-}
-
-ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size)
-{
-	struct dentry *realdentry = ovl_dentry_real(dentry);
-	ssize_t res;
-	size_t len;
-	char *s;
-	const struct cred *old_cred;
-
-	old_cred = ovl_override_creds(dentry->d_sb);
-	res = vfs_listxattr(realdentry, list, size);
-	revert_creds(old_cred);
-	if (res <= 0 || size == 0)
-		return res;
-
-	/* filter out private xattrs */
-	for (s = list, len = res; len;) {
-		size_t slen = strnlen(s, len) + 1;
-
-		/* underlying fs providing us with an broken xattr list? */
-		if (WARN_ON(slen > len))
-			return -EIO;
-
-		len -= slen;
-		if (!ovl_can_list(s)) {
-			res -= slen;
-			memmove(s, s + slen, len);
-		} else {
-			s += slen;
-		}
-	}
-
-	return res;
-}
-
-struct posix_acl *ovl_get_acl(struct inode *inode, int type)
-{
-	struct inode *realinode = ovl_inode_real(inode);
-	const struct cred *old_cred;
-	struct posix_acl *acl;
-
-	if (!IS_ENABLED(CONFIG_FS_POSIX_ACL) || !IS_POSIXACL(realinode))
-		return NULL;
-
-	old_cred = ovl_override_creds(inode->i_sb);
-	acl = get_acl(realinode, type);
-	revert_creds(old_cred);
-
-	return acl;
-}
-
-static bool ovl_open_need_copy_up(struct dentry *dentry, int flags)
-{
-	/* Copy up of disconnected dentry does not set upper alias */
-	if (ovl_dentry_upper(dentry) &&
-	    (ovl_dentry_has_upper_alias(dentry) ||
-	     (dentry->d_flags & DCACHE_DISCONNECTED)))
-		return false;
-
-	if (special_file(d_inode(dentry)->i_mode))
-		return false;
-
-	if (!(OPEN_FMODE(flags) & FMODE_WRITE) && !(flags & O_TRUNC))
-		return false;
-
-	return true;
-}
-
-int ovl_open_maybe_copy_up(struct dentry *dentry, unsigned int file_flags)
-{
-	int err = 0;
-
-	/* NOCOPYUPW */
-	return err;
-
-	if (ovl_open_need_copy_up(dentry, file_flags)) {
-		err = ovl_want_write(dentry);
-		if (!err) {
-			err = ovl_copy_up_flags(dentry, file_flags);
-			ovl_drop_write(dentry);
-		}
-	}
-
-	return err;
-}
-
-int ovl_update_time(struct inode *inode, struct timespec64 *ts, int flags)
-{
-	if (flags & S_ATIME) {
-		struct ovl_fs *ofs = inode->i_sb->s_fs_info;
-		struct path upperpath = {
-			.mnt = ofs->upper_mnt,
-			.dentry = ovl_upperdentry_dereference(OVL_I(inode)),
-		};
-
-		if (upperpath.dentry) {
-			touch_atime(&upperpath);
-			inode->i_atime = d_inode(upperpath.dentry)->i_atime;
-		}
-	}
-	return 0;
-}
-
-static const struct inode_operations ovl_file_inode_operations = {
-	.setattr	= ovl_setattr,
-	.permission	= ovl_permission,
-	.getattr	= ovl_getattr,
-	.listxattr	= ovl_listxattr,
-	.get_acl	= ovl_get_acl,
-	.update_time	= ovl_update_time,
-};
-
-static const struct inode_operations ovl_symlink_inode_operations = {
-	.setattr	= ovl_setattr,
-	.get_link	= ovl_get_link,
-	.readlink	= ovl_readlink,
-	.getattr	= ovl_getattr,
-	.listxattr	= ovl_listxattr,
-	.update_time	= ovl_update_time,
-};
-
-/*
- * It is possible to stack overlayfs instance on top of another
- * overlayfs instance as lower layer. We need to annonate the
- * stackable i_mutex locks according to stack level of the super
- * block instance. An overlayfs instance can never be in stack
- * depth 0 (there is always a real fs below it).  An overlayfs
- * inode lock will use the lockdep annotaion ovl_i_mutex_key[depth].
- *
- * For example, here is a snip from /proc/lockdep_chains after
- * dir_iterate of nested overlayfs:
- *
- * [...] &ovl_i_mutex_dir_key[depth]   (stack_depth=2)
- * [...] &ovl_i_mutex_dir_key[depth]#2 (stack_depth=1)
- * [...] &type->i_mutex_dir_key        (stack_depth=0)
- */
-#define OVL_MAX_NESTING FILESYSTEM_MAX_STACK_DEPTH
-
-static inline void ovl_lockdep_annotate_inode_mutex_key(struct inode *inode)
-{
-#ifdef CONFIG_LOCKDEP
-	static struct lock_class_key ovl_i_mutex_key[OVL_MAX_NESTING];
-	static struct lock_class_key ovl_i_mutex_dir_key[OVL_MAX_NESTING];
-	static struct lock_class_key ovl_i_lock_key[OVL_MAX_NESTING];
-
-	int depth = inode->i_sb->s_stack_depth - 1;
-
-	if (WARN_ON_ONCE(depth < 0 || depth >= OVL_MAX_NESTING))
-		depth = 0;
-
-	if (S_ISDIR(inode->i_mode))
-		lockdep_set_class(&inode->i_rwsem, &ovl_i_mutex_dir_key[depth]);
-	else
-		lockdep_set_class(&inode->i_rwsem, &ovl_i_mutex_key[depth]);
-
-	lockdep_set_class(&OVL_I(inode)->lock, &ovl_i_lock_key[depth]);
-#endif
-}
-
-static void ovl_fill_inode(struct inode *inode, umode_t mode, dev_t rdev,
-			   unsigned long ino, int fsid)
-{
-	int xinobits = ovl_xino_bits(inode->i_sb);
-
-	/*
-	 * When NFS export is enabled and d_ino is consistent with st_ino
-	 * (samefs or i_ino has enough bits to encode layer), set the same
-	 * value used for d_ino to i_ino, because nfsd readdirplus compares
-	 * d_ino values to i_ino values of child entries. When called from
-	 * ovl_new_inode(), ino arg is 0, so i_ino will be updated to real
-	 * upper inode i_ino on ovl_inode_init() or ovl_inode_update().
-	 */
-	if (inode->i_sb->s_export_op &&
-	    (ovl_same_sb(inode->i_sb) || xinobits)) {
-		inode->i_ino = ino;
-		if (xinobits && fsid && !(ino >> (64 - xinobits)))
-			inode->i_ino |= (unsigned long)fsid << (64 - xinobits);
-	} else {
-		inode->i_ino = get_next_ino();
-	}
-	inode->i_mode = mode;
-	inode->i_flags |= S_NOCMTIME;
-#ifdef CONFIG_FS_POSIX_ACL
-	inode->i_acl = inode->i_default_acl = ACL_DONT_CACHE;
-#endif
-
-	ovl_lockdep_annotate_inode_mutex_key(inode);
-
-	switch (mode & S_IFMT) {
-	case S_IFREG:
-		inode->i_op = &ovl_file_inode_operations;
-		break;
-
-	case S_IFDIR:
-		inode->i_op = &ovl_dir_inode_operations;
-		inode->i_fop = &ovl_dir_operations;
-		break;
-
-	case S_IFLNK:
-		inode->i_op = &ovl_symlink_inode_operations;
-		break;
-
-	default:
-		inode->i_op = &ovl_file_inode_operations;
-		init_special_inode(inode, mode, rdev);
-		break;
-	}
-}
-
-/*
- * With inodes index enabled, an overlay inode nlink counts the union of upper
- * hardlinks and non-covered lower hardlinks. During the lifetime of a non-pure
- * upper inode, the following nlink modifying operations can happen:
- *
- * 1. Lower hardlink copy up
- * 2. Upper hardlink created, unlinked or renamed over
- * 3. Lower hardlink whiteout or renamed over
- *
- * For the first, copy up case, the union nlink does not change, whether the
- * operation succeeds or fails, but the upper inode nlink may change.
- * Therefore, before copy up, we store the union nlink value relative to the
- * lower inode nlink in the index inode xattr trusted.overlay.nlink.
- *
- * For the second, upper hardlink case, the union nlink should be incremented
- * or decremented IFF the operation succeeds, aligned with nlink change of the
- * upper inode. Therefore, before link/unlink/rename, we store the union nlink
- * value relative to the upper inode nlink in the index inode.
- *
- * For the last, lower cover up case, we simplify things by preceding the
- * whiteout or cover up with copy up. This makes sure that there is an index
- * upper inode where the nlink xattr can be stored before the copied up upper
- * entry is unlink.
- */
-#define OVL_NLINK_ADD_UPPER	(1 << 0)
-
-/*
- * On-disk format for indexed nlink:
- *
- * nlink relative to the upper inode - "U[+-]NUM"
- * nlink relative to the lower inode - "L[+-]NUM"
- */
-
-static int ovl_set_nlink_common(struct dentry *dentry,
-				struct dentry *realdentry, const char *format)
-{
-	struct inode *inode = d_inode(dentry);
-	struct inode *realinode = d_inode(realdentry);
-	char buf[13];
-	int len;
-
-	len = snprintf(buf, sizeof(buf), format,
-		       (int) (inode->i_nlink - realinode->i_nlink));
-
-	if (WARN_ON(len >= sizeof(buf)))
-		return -EIO;
-
-	return ovl_do_setxattr(ovl_dentry_upper(dentry),
-			       OVL_XATTR_NLINK, buf, len, 0);
-}
-
-int ovl_set_nlink_upper(struct dentry *dentry)
-{
-	return ovl_set_nlink_common(dentry, ovl_dentry_upper(dentry), "U%+i");
-}
-
-int ovl_set_nlink_lower(struct dentry *dentry)
-{
-	return ovl_set_nlink_common(dentry, ovl_dentry_lower(dentry), "L%+i");
-}
-
-unsigned int ovl_get_nlink(struct dentry *lowerdentry,
-			   struct dentry *upperdentry,
-			   unsigned int fallback)
-{
-	int nlink_diff;
-	int nlink;
-	char buf[13];
-	int err;
-
-	if (!lowerdentry || !upperdentry || d_inode(lowerdentry)->i_nlink == 1)
-		return fallback;
-
-	err = vfs_getxattr(upperdentry, OVL_XATTR_NLINK, &buf, sizeof(buf) - 1);
-	if (err < 0)
-		goto fail;
-
-	buf[err] = '\0';
-	if ((buf[0] != 'L' && buf[0] != 'U') ||
-	    (buf[1] != '+' && buf[1] != '-'))
-		goto fail;
-
-	err = kstrtoint(buf + 1, 10, &nlink_diff);
-	if (err < 0)
-		goto fail;
-
-	nlink = d_inode(buf[0] == 'L' ? lowerdentry : upperdentry)->i_nlink;
-	nlink += nlink_diff;
-
-	if (nlink <= 0)
-		goto fail;
-
-	return nlink;
-
-fail:
-	pr_warn_ratelimited("overlayfs: failed to get index nlink (%pd2, err=%i)\n",
-			    upperdentry, err);
-	return fallback;
-}
-
-struct inode *ovl_new_inode(struct super_block *sb, umode_t mode, dev_t rdev)
-{
-	struct inode *inode;
-
-	inode = new_inode(sb);
-	if (inode)
-		ovl_fill_inode(inode, mode, rdev, 0, 0);
-
-	return inode;
-}
-
-static int ovl_inode_test(struct inode *inode, void *data)
-{
-	return inode->i_private == data;
-}
-
-static int ovl_inode_set(struct inode *inode, void *data)
-{
-	inode->i_private = data;
-	return 0;
-}
-
-static bool ovl_verify_inode(struct inode *inode, struct dentry *lowerdentry,
-			     struct dentry *upperdentry, bool strict)
-{
-	/*
-	 * For directories, @strict verify from lookup path performs consistency
-	 * checks, so NULL lower/upper in dentry must match NULL lower/upper in
-	 * inode. Non @strict verify from NFS handle decode path passes NULL for
-	 * 'unknown' lower/upper.
-	 */
-	if (S_ISDIR(inode->i_mode) && strict) {
-		/* Real lower dir moved to upper layer under us? */
-		if (!lowerdentry && ovl_inode_lower(inode))
-			return false;
-
-		/* Lookup of an uncovered redirect origin? */
-		if (!upperdentry && ovl_inode_upper(inode))
-			return false;
-	}
-
-	/*
-	 * Allow non-NULL lower inode in ovl_inode even if lowerdentry is NULL.
-	 * This happens when finding a copied up overlay inode for a renamed
-	 * or hardlinked overlay dentry and lower dentry cannot be followed
-	 * by origin because lower fs does not support file handles.
-	 */
-	if (lowerdentry && ovl_inode_lower(inode) != d_inode(lowerdentry))
-		return false;
-
-	/*
-	 * Allow non-NULL __upperdentry in inode even if upperdentry is NULL.
-	 * This happens when finding a lower alias for a copied up hard link.
-	 */
-	if (upperdentry && ovl_inode_upper(inode) != d_inode(upperdentry))
-		return false;
-
-	return true;
-}
-
-struct inode *ovl_lookup_inode(struct super_block *sb, struct dentry *real,
-			       bool is_upper)
-{
-	struct inode *inode, *key = d_inode(real);
-
-	inode = ilookup5(sb, (unsigned long) key, ovl_inode_test, key);
-	if (!inode)
-		return NULL;
-
-	if (!ovl_verify_inode(inode, is_upper ? NULL : real,
-			      is_upper ? real : NULL, false)) {
-		iput(inode);
-		return ERR_PTR(-ESTALE);
-	}
-
-	return inode;
-}
-
-/*
- * Does overlay inode need to be hashed by lower inode?
- */
-static bool ovl_hash_bylower(struct super_block *sb, struct dentry *upper,
-			     struct dentry *lower, struct dentry *index)
-{
-	struct ovl_fs *ofs = sb->s_fs_info;
-
-	/* No, if pure upper */
-	if (!lower)
-		return false;
-
-	/* Yes, if already indexed */
-	if (index)
-		return true;
-
-	/* Yes, if won't be copied up */
-	if (!ofs->upper_mnt)
-		return true;
-
-	/* No, if lower hardlink is or will be broken on copy up */
-	if ((upper || !ovl_indexdir(sb)) &&
-	    !d_is_dir(lower) && d_inode(lower)->i_nlink > 1)
-		return false;
-
-	/* No, if non-indexed upper with NFS export */
-	if (sb->s_export_op && upper)
-		return false;
-
-	/* Otherwise, hash by lower inode for fsnotify */
-	return true;
-}
-
-static struct inode *ovl_iget5(struct super_block *sb, struct inode *newinode,
-			       struct inode *key)
-{
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 18, 0)
-	return newinode ? inode_insert5(newinode, (unsigned long) key,
-					 ovl_inode_test, ovl_inode_set, key) :
-#else
-	return
-#endif
-			  iget5_locked(sb, (unsigned long) key,
-				       ovl_inode_test, ovl_inode_set, key);
-}
-
-struct inode *ovl_get_inode(struct super_block *sb,
-			    struct ovl_inode_params *oip)
-{
-	struct dentry *upperdentry = oip->upperdentry;
-	struct ovl_path *lowerpath = oip->lowerpath;
-	struct inode *realinode = upperdentry ? d_inode(upperdentry) : NULL;
-	struct inode *inode;
-	struct dentry *lowerdentry = lowerpath ? lowerpath->dentry : NULL;
-	bool bylower = ovl_hash_bylower(sb, upperdentry, lowerdentry,
-					oip->index);
-	int fsid = bylower ? oip->lowerpath->layer->fsid : 0;
-	bool is_dir;
-	unsigned long ino = 0;
-
-	if (!realinode)
-		realinode = d_inode(lowerdentry);
-
-	/*
-	 * Copy up origin (lower) may exist for non-indexed upper, but we must
-	 * not use lower as hash key if this is a broken hardlink.
-	 */
-	is_dir = S_ISDIR(realinode->i_mode);
-	if (upperdentry || bylower) {
-		struct inode *key = d_inode(bylower ? lowerdentry :
-						      upperdentry);
-		unsigned int nlink = is_dir ? 1 : realinode->i_nlink;
-
-		inode = ovl_iget5(sb, oip->newinode, key);
-		if (!inode)
-			goto out_nomem;
-		if (!(inode->i_state & I_NEW)) {
-			/*
-			 * Verify that the underlying files stored in the inode
-			 * match those in the dentry.
-			 */
-			if (!ovl_verify_inode(inode, lowerdentry, upperdentry,
-					      true)) {
-				iput(inode);
-				inode = ERR_PTR(-ESTALE);
-				goto out;
-			}
-
-			dput(upperdentry);
-			goto out;
-		}
-
-		/* Recalculate nlink for non-dir due to indexing */
-		if (!is_dir)
-			nlink = ovl_get_nlink(lowerdentry, upperdentry, nlink);
-		set_nlink(inode, nlink);
-		ino = key->i_ino;
-	} else {
-		/* Lower hardlink that will be broken on copy up */
-		inode = new_inode(sb);
-		if (!inode)
-			goto out_nomem;
-	}
-	ovl_fill_inode(inode, realinode->i_mode, realinode->i_rdev, ino, fsid);
-	ovl_inode_init(inode, upperdentry, lowerdentry);
-
-	if (upperdentry && ovl_is_impuredir(upperdentry))
-		ovl_set_flag(OVL_IMPURE, inode);
-
-	if (oip->index)
-		ovl_set_flag(OVL_INDEX, inode);
-
-	/* Check for non-merge dir that may have whiteouts */
-	if (is_dir) {
-		if (((upperdentry && lowerdentry) || oip->numlower > 1) ||
-		    ovl_check_origin_xattr(upperdentry ?: lowerdentry)) {
-			ovl_set_flag(OVL_WHITEOUTS, inode);
-		}
-	}
-
-	if (inode->i_state & I_NEW)
-		unlock_new_inode(inode);
-out:
-	return inode;
-
-out_nomem:
-	inode = ERR_PTR(-ENOMEM);
-	goto out;
-}