Commit Graph

21 Commits

Author SHA1 Message Date
3949ab65a8 TO RESET: Add kernel argument to toggle on-demand paging for hugetlbfs map
Change-Id: Id748e0a2afc4ea59142fedb652a15b4007c5dee4
2020-12-29 16:31:33 +09:00
d7cf39883f Revert "shmobj: Support large page"
This reverts commit 9a60997ea0.

Change-Id: Id60959b4e03451987239faa0bbc2e780b72fafaa
2020-07-19 12:53:45 +00:00
9a60997ea0 shmobj: Support large page
Mixing page sizes is allowed by shmobj.

Change-Id: Ic48b71da2db6ce3f68fa3dbc8ad5ae96347d6018
Refs: #1381
Refs: #1458
2020-07-15 03:50:56 +00:00
e774e1b984 Revert "xpmem: fix mapping of attachment and segment"
This reverts commit a5fcc91656.

Change-Id: If29415369d724391b291939ecce76482138e82f5
2020-06-16 11:28:02 +09:00
a5fcc91656 xpmem: fix mapping of attachment and segment
* Mapping attached part of segment is done at attach time instead of
  make time to work with runtimes (e.g. OpenMPI) xpmem_make-ing the
  entire user-space
* Mapping attached part of segment at attach time can be turned off by
  specifying xpmem_remote_on_demand in kernel argument
* Mapping attachment chooses appropriate page-sizes, i.e., largest
  allowed by memory range and segment page boundary

Fixes: a8696d8 "xpmem: Support large page attachment"
Change-Id: I44663865204036520e5f62fe22b9134ee4629f9b
2020-06-15 10:11:29 +09:00
74f15783d2 ihk_os_getrusage(): Add per-page-size memory usage accounting 2017-08-17 12:49:34 +09:00
8eb3bf3559 physical page management: eliminate static page frame array and
maintain page structures dynamically covering only file mappings.
use hash table for address <-> page structure conversion.
2016-10-14 21:34:29 +09:00
9b82f1a52c use ihk_mc_alloc/free_pages() and eliminate direct calls to low level routines 2016-10-14 21:34:29 +09:00
e1a01803d0 disable demand paging on ANONYMOUS mappings unless anon_on_demand kernel argument is passed 2015-09-14 17:26:37 +09:00
8097b057da add page_is_in_memobj() and page_is_multi_mapped()
These are page test functions.
2014-05-07 21:24:45 +09:00
ce7af5839d keep page.count valid through its life cycle 2014-05-07 21:24:45 +09:00
29c7d40005 use atomic operations for manipulating page.count 2014-05-07 21:24:45 +09:00
506caa7e7f page.h, process.h, rlimit.h, syscall.h, time.h, timer.h copyright 2013-11-14 17:34:30 +09:00
27172ad413 support private mapped file 2013-10-15 12:30:41 +09:00
78d9d3fcd2 Revert "trial implementation of private file mapping"
This reverts commit abe57218c4.
2013-07-26 16:44:39 +09:00
abe57218c4 trial implementation of private file mapping
for review only.  will soon be reverted.
2013-07-26 14:40:12 +09:00
deb9cd4e75 implement sys_munmap() 2013-07-18 09:25:58 +09:00
4693789608 change function names 2012-12-17 16:15:05 +09:00
a1490da112 Revert "make munmap free physical pages, and one correction for aligned mmap in syscall.c (extend_process_region is called with memory-region start-address which is aligned (map_end_aligned), instead of tail-address of current-region (region->map_end)) for aligned mmap (in syscall.c)"
This reverts commit e7317cca98.
2012-11-26 13:38:07 +09:00
e7317cca98 make munmap free physical pages, and one correction for aligned mmap in syscall.c (extend_process_region is called with memory-region start-address which is aligned (map_end_aligned), instead of tail-address of current-region (region->map_end)) for aligned mmap (in syscall.c) 2012-10-30 21:32:49 +09:00
83a17650b9 AP, kmalloc 2011-11-06 19:27:09 +09:00