fe42481d6f
Add allow_oversubscribe kernel argument
...
It's not allowed in the default setting.
Execute mcreboot.sh with -O option to allow it.
refs #1072
2018-03-10 13:08:38 +09:00
65667709a8
Fix thread status race-condition caused by hold_thread() in do_kill() and terminate()
...
Conflicts:
arch/x86_64/kernel/syscall.c
kernel/syscall.c
2018-03-09 17:53:17 +09:00
c260b5c6f3
xpmem: support for fork()
...
refs #925
2018-02-22 09:37:48 +09:00
4bb65494e9
signal: When the process receives a termination signal, it first terminates mcexec.
...
refs #863
refs #870
2018-01-23 14:40:38 +09:00
2f2b3cdc6f
signal: interrupt_syscall is called by the core executing the thread that recieved the signal.
...
refs #999
2018-01-23 14:31:04 +09:00
589504dc33
mcreboot: -h to indicate halting CPU in idle threads (e.g., in futex_wait())
2017-12-18 11:22:15 +09:00
d649d6fc2d
Include mbind support (this is a rebase commit to merge into development)
2017-11-27 11:16:53 +09:00
12840601e1
support PERF_TYPE_{HARDWARE|HW_CACHE} in perf_event_open
...
refs #829
2017-10-20 23:10:20 +09:00
d5629606c5
mcexec: -m: interpret as numactl -m (i.e., MPOL_BIND)
...
Conflicts:
executer/include/uprotocol.h
executer/user/mcexec.c
kernel/include/syscall.h
2017-10-18 16:54:34 +09:00
4d4279121b
process/vm; replace vm_range list by a rbtree
...
This replaces the chained list used to keep track of all memory ranges
of a process by a standard rbtree (no need of interval tree here
because there is no overlap)
Accesses that were done directly through vm_range_list before were
replaced by lookup_process_memory_range, even full list scan (e.g.
coredump).
The full scans will thus be less efficient because calls to rb_next()
will not be inlined, but these are rarer calls that can probably afford
this compared to code simplicity.
The only reference to the actual backing structure left outside of
process.c is a call to rb_erase in xpmem_free_process_memory_range.
v2: fix lookup_process_memory_range with small start address
v3: make vm_range_insert error out properly
Panic does not lead to easy debug, all error paths
are handled to just return someting on error
v4: fix lookup_process_memory_range (again)
That optimistically going left was a more serious bug than just
last iteration, we could just pass by a match and continue down
the tree if the match was not a leaf.
v5: some users actually needed leftmost match, so restore behavior
without the breakage (hopefully)
2017-10-13 10:00:27 +09:00
99da5b6484
ptrace: unify flags PT_TRACE_SYSCALL_ENTER and PT_TRACE_SYSCALL_EXIT to PT_TRACE_SYSCALL
...
refs #961
2017-10-11 15:43:57 +09:00
a2fbe99b60
madvise: support MADV_DONTDUMP/DODUMP
...
refs #661
2017-09-26 14:21:40 +09:00
9989f41fd3
add arm64 support
...
- add arm64 dependent codes with GICv3 and SVE support
- fix bugs based on architecture separation requests
2017-09-05 15:06:27 +09:00
236a072311
Add qlmpi and swap to mckernel (This is rebase commit for merging to development)
2017-08-29 15:04:58 +09:00
992a292c08
profile: better time breakdown and exclusion of idle cycles
2017-07-20 17:36:34 +09:00
07efb3ab9a
support to utility thread offloading
2017-06-27 13:27:09 +09:00
e901d42fb6
mcexec: --extend-heap-by: argument to specify heap extension size
2017-05-23 19:58:49 +09:00
0dd2fad33b
brk(): more forceful heap extension
2017-05-23 19:57:08 +09:00
e3e0f6a174
mcexec: introduction of --profile
2017-05-23 02:42:06 +09:00
9c25d47d9b
mcexec: transfer job information to LWK
2017-05-23 02:42:06 +09:00
e2f424846c
profile: rewrite syscall tracker for generic profiling code
2017-05-23 02:42:05 +09:00
610463ff39
sched_setaffinity(): respect process cpu_set
2017-05-23 02:39:42 +09:00
26b9484bae
mcexec: --mpol-threshold to control MPOL_BIND/MPOL_PREFERRED
2017-05-23 02:39:42 +09:00
182202523e
mcexec/mm: user memory policy control for heap, stack, etc.
2017-05-23 02:39:42 +09:00
fdbdcbd0ee
VR_AP_USER: memory range flag to respect user mempolicy (e.g., in PF handler)
2017-05-23 02:39:41 +09:00
e6c4d7731d
Merge remote-tracking branch 'origin/rusage'
...
Conflicts:
configure
kernel/process.c
2017-04-27 15:10:38 +09:00
c38d536aaa
xpmem: porting xpmem v2.6.3
...
implement xpmem_get, xpmem_release, xpmem_attach, xpmem_detach
2017-03-29 18:20:53 +09:00
3fe2257929
create rusage branch.
2017-03-15 23:22:51 +09:00
addbe91e59
do_migrate(): signal migrated thread before releasing runq lock
2017-01-30 07:24:09 +09:00
5cc420a6c3
syscall/offload tracker: clean-up and support process-wise aggregation
2017-01-30 07:24:09 +09:00
bd47b909bf
futex(): spin wait when CPU not oversubscribed and fix lost wake-up bug
2017-01-13 08:43:25 +09:00
ec75095073
add_process_memory_range(): optionally return range object
2016-12-30 15:51:17 +09:00
f81927b85b
Revert "brk(): larger allocation units internally"
...
This reverts commit c58ab0f648
2016-12-20 11:11:09 +09:00
c58ab0f648
brk(): larger allocation units internally
2016-12-18 21:12:37 +09:00
fdcf766337
prepare_process(): pass cpu_set in program_load_desc
2016-12-09 16:32:20 +09:00
7d13bfb14e
set_mempolicy(): limit maxnode to PROCESS_NUMA_MASK_BITS
2016-12-08 21:05:10 +09:00
a8c4ab221b
use MCS locks in signal handling code
2016-12-05 18:10:20 +09:00
f78d031e64
syscall and offload tracking (disabled by default)
2016-12-05 18:10:20 +09:00
1130cafe41
ptrace: fixed for threads.
2016-11-28 11:19:30 +09:00
64607152ee
VM: introduction of range lookup cache
2016-11-23 08:48:44 +09:00
0f826290d0
NUMA: get_mempolicy(), set_mempolicy() and mbind() implementation
2016-11-05 13:32:02 +09:00
9705a80c82
get/set_mempolicy(): support for query/set process level policy
2016-10-16 14:01:14 +09:00
99a02e2941
get_mempolicy(): store policy in per-process VM structure
2016-10-16 09:10:36 +09:00
faa929e717
NUMA: add NUMA mask to process VM structure
2016-10-14 21:34:31 +09:00
e03f377326
interrupt_syscall: interrupt valid thread
2016-10-03 00:49:56 +09:00
02536b7724
Merge remote-tracking branch 'remotes/origin/ikc2'
...
Conflicts:
executer/kernel/mcctrl/syscall.c
It is resolved.
2016-09-27 11:48:12 +09:00
7c562d0539
support madvise(MADV_DONTFORK)
2016-09-01 11:22:53 +09:00
f584e2ec25
increase kernel stack size and eliminate unused waitq declaration in do_syscall()
2016-08-16 09:20:55 +09:00
3aa06444f4
do_syscall(): allow descheduling threads in offloaded syscalls if CPU core oversubscribed
2016-08-16 08:58:22 +09:00
6aae35cb3d
process: transfer TIDs in bulk and reuse them locally
2016-08-02 16:59:04 +09:00
