Andrea Arcangeli
1c2fb7a4c2
ksm: fix deadlock with munlock in exit_mmap ...
Rawhide users have reported hang at startup when cryptsetup is run: the
same problem can be simply reproduced by running a program int main() {
mlockall(MCL_CURRENT | MCL_FUTURE); return 0; }
The problem is that exit_mmap() applies munlock_vma_pages_all() to
clean up VM_LOCKED areas, and its current implementation (stupidly)
tries to fault in absent pages, for example where PROT_NONE prevented
them being faulted in when mlocking. Whereas the "ksm: fix oom
deadlock" patch, knowing there's a race by which KSM might try to fault
in pages after exit_mmap() had finally zapped the range, backs out of
such faults doing nothing when its ksm_test_exit() notices mm_users 0.
So revert that part of "ksm: fix oom deadlock" which moved the
ksm_exit() call from before exit_mmap() to the middle of exit_mmap();
and remove those ksm_test_exit() checks from the page fault paths, so
allowing the munlocking to proceed without interference.
ksm_exit, if there are rmap_items still chained on this mm slot, takes
mmap_sem write side: so preventing KSM from working on an mm while
exit_mmap runs. And KSM will bail out as soon as it notices that
mm_users is already zero, thanks to its internal ksm_test_exit checks.
So that when a task is killed by OOM killer or the user, KSM will not
indefinitely prevent it from running exit_mmap to release its memory.
This does break a part of what "ksm: fix oom deadlock" was trying to
achieve. When unmerging KSM (echo 2 >/sys/kernel/mm/ksm), and even
when ksmd itself has to cancel a KSM page, it is possible that the
first OOM-kill victim would be the KSM process being faulted: then its
memory won't be freed until a second victim has been selected (freeing
memory for the unmerging fault to complete).
But the OOM killer is already liable to kill a second victim once the
intended victim's p->mm goes to NULL: so there's not much point in
rejecting this KSM patch before fixing that OOM behaviour. It is very
much more important to allow KSM users to boot up, than to haggle over
an unlikely and poorly supported OOM case.
We also intend to fix munlocking to not fault pages: at which point
this patch _could_ be reverted; though that would be controversial, so
we hope to find a better solution.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Justin M. Forbes <jforbes@redhat.com>
Acked-for-now-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Cc: Izik Eidus <ieidus@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-09-22 07:17:32 -07:00
powerpc: Enable GCOV
2009-08-20 10:29:28 +10:00
Merge branch 'irq-threaded-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-11 13:21:31 -07:00
vt: remove power stuff from kernel/power
2009-09-19 13:13:25 -07:00
Merge branch 'timers-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-18 09:15:24 -07:00
Merge branch 'perfcounters-rename-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-21 09:15:07 -07:00
bsdacct: switch credentials for writing to the accounting file
2009-08-24 11:33:40 +10:00
async: Fix lack of boot-time console due to insufficient synchronization
2009-06-08 12:31:53 -07:00
Fix rule eviction order for AUDIT_DIR
2009-06-24 00:02:38 -04:00
Fix rule eviction order for AUDIT_DIR
2009-06-24 00:02:38 -04:00
Fix rule eviction order for AUDIT_DIR
2009-06-24 00:02:38 -04:00
Audit: clean up all op= output to include string quoting
2009-06-24 00:00:52 -04:00
Audit: clean up all op= output to include string quoting
2009-06-24 00:00:52 -04:00
Fix rule eviction order for AUDIT_DIR
2009-06-24 00:02:38 -04:00
const: mark remaining inode_operations as const
2009-09-22 07:17:24 -07:00
debug cgroup: remove unneeded cgroup_lock
2009-04-02 19:04:54 -07:00
signals: implement sys_rt_tgsigqueueinfo
2009-04-30 19:24:24 +02:00
Merge branch 'x86-pat-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-15 09:19:38 -07:00
cpuset,mm: update tasks' mems_allowed in time
2009-06-16 19:47:31 -07:00
CRED: Allow put_cred() to cope with a NULL groups list
2009-09-15 09:10:57 +10:00
headers: taskstats_kern.h trim
2009-09-18 09:48:52 -07:00
Get rid of indirect include of fs_struct.h
2009-03-31 23:00:27 -04:00
perf: Do the big rename: Performance Counters -> Performance Events
2009-09-21 14:28:04 +02:00
Merge branch 'tracing-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-04-05 11:04:19 -07:00
ksm: fix deadlock with munlock in exit_mmap
2009-09-22 07:17:32 -07:00
sched: fix nr_uninterruptible accounting of frozen tasks really
2009-07-18 14:19:53 +02:00
Merge branch 'core-futexes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-11 13:16:22 -07:00
futex: Fix compat_futex to be same as futex for REQUEUE_PI
2009-08-10 15:41:12 +02:00
groups: move code to kernel/groups.c
2009-06-16 19:47:48 -07:00
Merge branch 'timers-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-18 09:15:24 -07:00
kernel/kallsyms.c: replace deprecated __initcall with device_initcall and fix whitespace
2009-06-09 22:37:52 +02:00
kexec: fix omitting offset in extended crashkernel syntax
2009-07-29 19:10:34 -07:00
kfifo: Use "const" definitions
2009-09-19 13:13:17 -07:00
sysrq, intel_fb: fix sysrq g collision
2009-05-15 07:56:24 -05:00
Merge branch 'tracing-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-11 13:24:03 -07:00
Merge branch 'linus' into tracing/core
2009-08-11 14:19:09 +02:00
sched: Keep kthreads at default priority
2009-09-09 17:30:06 +02:00
lockdep: Remove recursion stattistics
2009-08-29 13:22:36 +02:00
lockdep: BFS cleanup
2009-07-24 10:53:29 +02:00
Merge commit 'v2.6.31-rc8' into core/locking
2009-08-31 18:05:25 +02:00
Merge branch 'perfcounters-rename-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-21 09:15:07 -07:00
tracing: Remove markers
2009-09-18 21:22:08 +02:00
Merge branch 'linus' into perfcounters/core
2009-06-11 17:55:42 +02:00
cgroups: relax ns_can_attach checks to allow attaching to grandchild cgroups
2009-04-02 19:04:53 -07:00
nsproxy: extract create_nsproxy()
2009-06-18 13:03:56 -07:00
trace: stop tracer in oops_enter()
2009-07-24 15:30:45 -04:00
module_param: allow 'bool' module_params to be bool, not just int.
2009-06-12 21:46:58 +09:30
perf: Tidy up after the big rename
2009-09-21 14:34:11 +02:00
kmemleak: Remove alloc_bootmem annotations introduced in the past
2009-07-09 17:07:02 +01:00
pidns: rewrite copy_pid_ns()
2009-06-18 13:03:55 -07:00
posix_cpu_timers_exit_group(): Do not use thread_group_cputimer()
2009-08-08 18:30:25 +02:00
time: Introduce CLOCK_REALTIME_COARSE
2009-08-21 21:43:46 +02:00
cleanup console_print()
2009-09-14 17:41:42 -07:00
kernel/profile.c: Switch /proc/irq/prof_cpu_mask to seq_file
2009-09-20 20:15:40 +02:00
Merge branch 'master' into next
2009-07-14 00:30:40 +10:00
rcu: Fix whitespace inconsistencies
2009-09-19 08:53:22 +02:00
rcu: Fix whitespace inconsistencies
2009-09-19 08:53:22 +02:00
rcu: Fix whitespace inconsistencies
2009-09-19 08:53:22 +02:00
rcu: Fix whitespace inconsistencies
2009-09-19 08:53:22 +02:00
rcu: Fix whitespace inconsistencies
2009-09-19 08:53:22 +02:00
rcu: Fix whitespace inconsistencies
2009-09-19 08:53:22 +02:00
Merge branch 'tracing-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-04-05 11:04:19 -07:00
memcg: add interface to reset limits
2009-06-18 13:03:48 -07:00
kernel/resource.c: fix sign extension in reserve_setup()
2009-06-30 18:56:00 -07:00
rtmutex: Avoid deadlock in rt_mutex_start_proxy_lock()
2009-08-06 05:50:21 +02:00
rt_mutex: add proxy lock routines
2009-04-06 11:14:02 +02:00
Merge branch 'perfcounters-rename-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-21 09:15:07 -07:00
sched_clock: Make it NMI safe
2009-09-18 20:47:30 +02:00
sched: Add new prio to cpupri before removing old prio
2009-08-02 14:26:09 +02:00
sched: Add new wakeup preemption mode: WAKEUP_RUNNING
2009-09-17 10:17:25 +02:00
Merge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-21 09:06:17 -07:00
sched: Add new wakeup preemption mode: WAKEUP_RUNNING
2009-09-17 10:17:25 +02:00
sched: Simplify sys_sched_rr_get_interval() system call
2009-09-21 09:53:55 +02:00
sched: Simplify sys_sched_rr_get_interval() system call
2009-09-21 09:53:55 +02:00
do_sigaltstack: small cleanups
2009-08-01 11:18:56 -07:00
slow-work: use round_jiffies() for thread pool's cull and OOM timers
2009-06-16 19:47:49 -07:00
Merge branch 'x86/urgent' into x86/pat
2009-08-26 17:24:28 -07:00
softirq: add BLOCK_IOPOLL to softirq_to_name
2009-09-17 15:53:44 -04:00
locking: Allow arch-inlined spinlocks
2009-08-31 18:08:50 +02:00
perf: Do the big rename: Performance Counters -> Performance Events
2009-09-21 14:28:04 +02:00
perf: Do the big rename: Performance Counters -> Performance Events
2009-09-21 14:28:04 +02:00
perf: Do the big rename: Performance Counters -> Performance Events
2009-09-21 14:28:04 +02:00
genetlink: make netns aware
2009-07-12 14:03:27 -07:00
time: Prevent 32 bit overflow with set_normalized_timespec()
2009-09-15 10:17:30 +02:00
perf: Do the big rename: Performance Counters -> Performance Events
2009-09-21 14:28:04 +02:00
tracing: Don't trace kernel thread syscalls
2009-08-26 21:29:52 +02:00
sched: delayed cleanup of user_struct
2009-06-15 21:30:23 -07:00
utsns: extract creeate_uts_ns()
2009-06-18 13:03:55 -07:00
proc_sysctl: use CONFIG_PROC_SYSCTL around ipc and utsname proc_handlers
2009-04-02 19:05:01 -07:00
locking, sched: Give waitqueue spinlocks their own lockdep classes
2009-08-10 14:43:09 +02:00
Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
2009-09-11 13:23:18 -07:00
1c2fb7a4c2