Skip to content

K
Kernel

openeuler / Kernel
1 年多 前同步成功

通知 8
Star 0
Fork 0
K
Kernel
切换分支/标签
  1. 19 10月, 2021 5 次提交
    • H
      mm/migrate: fix CPUHP state to update node demotion order · a6a0251c
      由 Huang Ying 提交于 
      The node demotion order needs to be updated during CPU hotplug.  Because
whether a NUMA node has CPU may influence the demotion order.  The
update function should be called during CPU online/offline after the
node_states[N_CPU] has been updated.  That is done in
CPUHP_AP_ONLINE_DYN during CPU online and in CPUHP_MM_VMSTAT_DEAD during
CPU offline.  But in commit 884a6e5d ("mm/migrate: update node
demotion order on hotplug events"), the function to update node demotion
order is called in CPUHP_AP_ONLINE_DYN during CPU online/offline.  This
doesn't satisfy the order requirement.

For example, there are 4 CPUs (P0, P1, P2, P3) in 2 sockets (P0, P1 in S0
and P2, P3 in S1), the demotion order is

 - S0 -> NUMA_NO_NODE
 - S1 -> NUMA_NO_NODE

After P2 and P3 is offlined, because S1 has no CPU now, the demotion
order should have been changed to

 - S0 -> S1
 - S1 -> NO_NODE

but it isn't changed, because the order updating callback for CPU
hotplug doesn't see the new nodemask.  After that, if P1 is offlined,
the demotion order is changed to the expected order as above.

So in this patch, we added CPUHP_AP_MM_DEMOTION_ONLINE and
CPUHP_MM_DEMOTION_DEAD to be called after CPUHP_AP_ONLINE_DYN and
CPUHP_MM_VMSTAT_DEAD during CPU online and offline, and register the
update function on them.

Link: https://lkml.kernel.org/r/20210929060351.7293-1-ying.huang@intel.com
Fixes: 884a6e5d ("mm/migrate: update node demotion order on hotplug events")
Signed-off-by: N"Huang, Ying" <ying.huang@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Wei Xu <weixugc@google.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Keith Busch <kbusch@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      a6a0251c
    • D
      mm/migrate: add CPU hotplug to demotion #ifdef · 76af6a05
      由 Dave Hansen 提交于 
      Once upon a time, the node demotion updates were driven solely by memory
hotplug events.  But now, there are handlers for both CPU and memory
hotplug.

However, the #ifdef around the code checks only memory hotplug.  A
system that has HOTPLUG_CPU=y but MEMORY_HOTPLUG=n would miss CPU
hotplug events.

Update the #ifdef around the common code.  Add memory and CPU-specific
#ifdefs for their handlers.  These memory/CPU #ifdefs avoid unused
function warnings when their Kconfig option is off.

[arnd@arndb.de: rework hotplug_memory_notifier() stub]
  Link: https://lkml.kernel.org/r/20211013144029.2154629-1-arnd@kernel.org

Link: https://lkml.kernel.org/r/20210924161255.E5FE8F7E@davehans-spike.ostc.intel.com
Fixes: 884a6e5d ("mm/migrate: update node demotion order on hotplug events")
Signed-off-by: NDave Hansen <dave.hansen@linux.intel.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Cc: "Huang, Ying" <ying.huang@intel.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Wei Xu <weixugc@google.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Yang Shi <yang.shi@linux.alibaba.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      76af6a05
    • D
      mm/migrate: optimize hotplug-time demotion order updates · 295be91f
      由 Dave Hansen 提交于 
      Patch series "mm/migrate: 5.15 fixes for automatic demotion", v2.

This contains two fixes for the "automatic demotion" code which was
merged into 5.15:

 * Fix memory hotplug performance regression by watching
   suppressing any real action on irrelevant hotplug events.

 * Ensure CPU hotplug handler is registered when memory hotplug
   is disabled.

This patch (of 2):

== tl;dr ==

Automatic demotion opted for a simple, lazy approach to handling hotplug
events.  This noticeably slows down memory hotplug[1].  Optimize away
updates to the demotion order when memory hotplug events should have no
effect.

This has no effect on CPU hotplug.  There is no known problem on the CPU
side and any work there will be in a separate series.

== Background ==

Automatic demotion is a memory migration strategy to ensure that new
allocations have room in faster memory tiers on tiered memory systems.
The kernel maintains an array (node_demotion[]) to drive these
migrations.

The node_demotion[] path is calculated by starting at nodes with CPUs
and then "walking" to nodes with memory.  Only hotplug events which
online or offline a node with memory (N_ONLINE) or CPUs (N_CPU) will
actually affect the migration order.

== Problem ==

However, the current code is lazy.  It completely regenerates the
migration order on *any* CPU or memory hotplug event.  The logic was
that these events are extremely rare and that the overhead from
indiscriminate order regeneration is minimal.

Part of the update logic involves a synchronize_rcu(), which is a pretty
big hammer.  Its overhead was large enough to be detected by some 0day
tests that watch memory hotplug performance[1].

== Solution ==

Add a new helper (node_demotion_topo_changed()) which can differentiate
between superfluous and impactful hotplug events.  Skip the expensive
update operation for superfluous events.

== Aside: Locking ==

It took me a few moments to declare the locking to be safe enough for
node_demotion_topo_changed() to work.  It all hinges on the memory
hotplug lock:

During memory hotplug events, 'mem_hotplug_lock' is held for write.
This ensures that two memory hotplug events can not be called
simultaneously.

CPU hotplug has a similar lock (cpuhp_state_mutex) which also provides
mutual exclusion between CPU hotplug events.  In addition, the demotion
code acquire and hold the mem_hotplug_lock for read during its CPU
hotplug handlers.  This provides mutual exclusion between the demotion
memory hotplug callbacks and the CPU hotplug callbacks.

This effectively allows treating the migration target generation code to
act as if it is single-threaded.

1. https://lore.kernel.org/all/20210905135932.GE15026@xsang-OptiPlex-9020/

Link: https://lkml.kernel.org/r/20210924161251.093CCD06@davehans-spike.ostc.intel.com
Link: https://lkml.kernel.org/r/20210924161253.D7673E31@davehans-spike.ostc.intel.com
Fixes: 884a6e5d ("mm/migrate: update node demotion order on hotplug events")
Signed-off-by: NDave Hansen <dave.hansen@linux.intel.com>
Reported-by: Nkernel test robot <oliver.sang@intel.com>
Reviewed-by: NDavid Hildenbrand <david@redhat.com>
Cc: "Huang, Ying" <ying.huang@intel.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Wei Xu <weixugc@google.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Yang Shi <yang.shi@linux.alibaba.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      295be91f
    • N
      userfaultfd: fix a race between writeprotect and exit_mmap() · cb185d5f
      由 Nadav Amit 提交于 
      A race is possible when a process exits, its VMAs are removed by
exit_mmap() and at the same time userfaultfd_writeprotect() is called.

The race was detected by KASAN on a development kernel, but it appears
to be possible on vanilla kernels as well.

Use mmget_not_zero() to prevent the race as done in other userfaultfd
operations.

Link: https://lkml.kernel.org/r/20210921200247.25749-1-namit@vmware.com
Fixes: 63b2d417 ("userfaultfd: wp: add the writeprotect API to userfaultfd ioctl")
Signed-off-by: NNadav Amit <namit@vmware.com>
Tested-by: NLi  Wang <liwang@redhat.com>
Reviewed-by: NPeter Xu <peterx@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      cb185d5f
    • P
      mm/userfaultfd: selftests: fix memory corruption with thp enabled · 8913970c
      由 Peter Xu 提交于 
      In RHEL's gating selftests we've encountered memory corruption in the
uffd event test even with upstream kernel:

        # ./userfaultfd anon 128 4
        nr_pages: 32768, nr_pages_per_cpu: 32768
        bounces: 3, mode: rnd racing read, userfaults: 6240 missing (6240) 14729 wp (14729)
        bounces: 2, mode: racing read, userfaults: 1444 missing (1444) 28877 wp (28877)
        bounces: 1, mode: rnd read, userfaults: 6055 missing (6055) 14699 wp (14699)
        bounces: 0, mode: read, userfaults: 82 missing (82) 25196 wp (25196)
        testing uffd-wp with pagemap (pgsize=4096): done
        testing uffd-wp with pagemap (pgsize=2097152): done
        testing events (fork, remap, remove): ERROR: nr 32427 memory corruption 0 1 (errno=0, line=963)
        ERROR: faulting process failed (errno=0, line=1117)

It can be easily reproduced when global thp enabled, which is the
default for RHEL.

It's also known as a side effect of commit 0db282ba ("selftest: use
mmap instead of posix_memalign to allocate memory", 2021-07-23), which
is imho right itself on using mmap() to make sure the addresses will be
untagged even on arm.

The problem is, for each test we allocate buffers using two
allocate_area() calls.  We assumed these two buffers won't affect each
other, however they could, because mmap() could have found that the two
buffers are near each other and having the same VMA flags, so they got
merged into one VMA.

It won't be a big problem if thp is not enabled, but when thp is
agressively enabled it means when initializing the src buffer it could
accidentally setup part of the dest buffer too when there's a shared THP
that overlaps the two regions.  Then some of the dest buffer won't be
able to be trapped by userfaultfd missing mode, then it'll cause memory
corruption as described.

To fix it, do release_pages() after initializing the src buffer.

Since the previous two release_pages() calls are after
uffd_test_ctx_clear() which will unmap all the buffers anyway (which is
stronger than release pages; as unmap() also tear town pgtables), drop
them as they shouldn't really be anything useful.

We can mark the Fixes tag upon 0db282ba as it's reported to only
happen there, however the real "Fixes" IMHO should be 8ba6e864, as
before that commit we'll always do explicit release_pages() before
registration of uffd, and 8ba6e864 changed that logic by adding
extra unmap/map and we didn't release the pages at the right place.
Meanwhile I don't have a solid glue anyway on whether posix_memalign()
could always avoid triggering this bug, hence it's safer to attach this
fix to commit 8ba6e864.

Link: https://lkml.kernel.org/r/20210923232512.210092-1-peterx@redhat.com
Fixes: 8ba6e864 ("userfaultfd/selftests: reinitialize test context in each test")
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1994931Signed-off-by: NPeter Xu <peterx@redhat.com>
Reported-by: NLi Wang <liwan@redhat.com>
Tested-by: NLi Wang <liwang@redhat.com>
Reviewed-by: NAxel Rasmussen <axelrasmussen@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      8913970c
  3. 18 10月, 2021 17 次提交
  5. 17 10月, 2021 13 次提交
  7. 16 10月, 2021 5 次提交