Skip to content

K
Kernel

openeuler / Kernel
接近 2 年 前同步成功

通知 8
Star 0
Fork 0
K
Kernel
切换分支/标签
  1. 02 7月, 2021 1 次提交
    • A
      mm: remove special swap entry functions · af5cdaf8
      由 Alistair Popple 提交于 
      Patch series "Add support for SVM atomics in Nouveau", v11.

Introduction
============

Some devices have features such as atomic PTE bits that can be used to
implement atomic access to system memory.  To support atomic operations to
a shared virtual memory page such a device needs access to that page which
is exclusive of the CPU.  This series introduces a mechanism to
temporarily unmap pages granting exclusive access to a device.

These changes are required to support OpenCL atomic operations in Nouveau
to shared virtual memory (SVM) regions allocated with the
CL_MEM_SVM_ATOMICS clSVMAlloc flag.  A more complete description of the
OpenCL SVM feature is available at
https://www.khronos.org/registry/OpenCL/specs/3.0-unified/html/
OpenCL_API.html#_shared_virtual_memory .

Implementation
==============

Exclusive device access is implemented by adding a new swap entry type
(SWAP_DEVICE_EXCLUSIVE) which is similar to a migration entry.  The main
difference is that on fault the original entry is immediately restored by
the fault handler instead of waiting.

Restoring the entry triggers calls to MMU notifers which allows a device
driver to revoke the atomic access permission from the GPU prior to the
CPU finalising the entry.

Patches
=======

Patches 1 & 2 refactor existing migration and device private entry
functions.

Patches 3 & 4 rework try_to_unmap_one() by splitting out unrelated
functionality into separate functions - try_to_migrate_one() and
try_to_munlock_one().

Patch 5 renames some existing code but does not introduce functionality.

Patch 6 is a small clean-up to swap entry handling in copy_pte_range().

Patch 7 contains the bulk of the implementation for device exclusive
memory.

Patch 8 contains some additions to the HMM selftests to ensure everything
works as expected.

Patch 9 is a cleanup for the Nouveau SVM implementation.

Patch 10 contains the implementation of atomic access for the Nouveau
driver.

Testing
=======

This has been tested with upstream Mesa 21.1.0 and a simple OpenCL program
which checks that GPU atomic accesses to system memory are atomic.
Without this series the test fails as there is no way of write-protecting
the page mapping which results in the device clobbering CPU writes.  For
reference the test is available at
https://ozlabs.org/~apopple/opencl_svm_atomics/

Further testing has been performed by adding support for testing exclusive
access to the hmm-tests kselftests.

This patch (of 10):

Remove multiple similar inline functions for dealing with different types
of special swap entries.

Both migration and device private swap entries use the swap offset to
store a pfn.  Instead of multiple inline functions to obtain a struct page
for each swap entry type use a common function pfn_swap_entry_to_page().
Also open-code the various entry_to_pfn() functions as this results is
shorter code that is easier to understand.

Link: https://lkml.kernel.org/r/20210616105937.23201-1-apopple@nvidia.com
Link: https://lkml.kernel.org/r/20210616105937.23201-2-apopple@nvidia.comSigned-off-by: NAlistair Popple <apopple@nvidia.com>
Reviewed-by: NRalph Campbell <rcampbell@nvidia.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Ben Skeggs <bskeggs@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      af5cdaf8
  3. 01 7月, 2021 4 次提交
    • Y
      mm: memory: make numa_migrate_prep() non-static · f4c0d836
      由 Yang Shi 提交于 
      The numa_migrate_prep() will be used by huge NUMA fault as well in the
following patch, make it non-static.

Link: https://lkml.kernel.org/r/20210518200801.7413-3-shy828301@gmail.comSigned-off-by: NYang Shi <shy828301@gmail.com>
Acked-by: NMel Gorman <mgorman@suse.de>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      f4c0d836
    • Y
      mm: memory: add orig_pmd to struct vm_fault · 5db4f15c
      由 Yang Shi 提交于 
      Pach series "mm: thp: use generic THP migration for NUMA hinting fault", v3.

When the THP NUMA fault support was added THP migration was not supported
yet.  So the ad hoc THP migration was implemented in NUMA fault handling.
Since v4.14 THP migration has been supported so it doesn't make too much
sense to still keep another THP migration implementation rather than using
the generic migration code.  It is definitely a maintenance burden to keep
two THP migration implementation for different code paths and it is more
error prone.  Using the generic THP migration implementation allows us
remove the duplicate code and some hacks needed by the old ad hoc
implementation.

A quick grep shows x86_64, PowerPC (book3s), ARM64 ans S390 support both
THP and NUMA balancing.  The most of them support THP migration except for
S390.  Zi Yan tried to add THP migration support for S390 before but it
was not accepted due to the design of S390 PMD.  For the discussion,
please see: https://lkml.org/lkml/2018/4/27/953.

Per the discussion with Gerald Schaefer in v1 it is acceptible to skip
huge PMD for S390 for now.

I saw there were some hacks about gup from git history, but I didn't
figure out if they have been removed or not since I just found FOLL_NUMA
code in the current gup implementation and they seems useful.

Patch #1 ~ #2 are preparation patches.
Patch #3 is the real meat.
Patch #4 ~ #6 keep consistent counters and behaviors with before.
Patch #7 skips change huge PMD to prot_none if thp migration is not supported.

Test
----
Did some tests to measure the latency of do_huge_pmd_numa_page.  The test
VM has 80 vcpus and 64G memory.  The test would create 2 processes to
consume 128G memory together which would incur memory pressure to cause
THP splits.  And it also creates 80 processes to hog cpu, and the memory
consumer processes are bound to different nodes periodically in order to
increase NUMA faults.

The below test script is used:

echo 3 > /proc/sys/vm/drop_caches

# Run stress-ng for 24 hours
./stress-ng/stress-ng --vm 2 --vm-bytes 64G --timeout 24h &
PID=$!

./stress-ng/stress-ng --cpu $NR_CPUS --timeout 24h &

# Wait for vm stressors forked
sleep 5

PID_1=`pgrep -P $PID | awk 'NR == 1'`
PID_2=`pgrep -P $PID | awk 'NR == 2'`

JOB1=`pgrep -P $PID_1`
JOB2=`pgrep -P $PID_2`

# Bind load jobs to different nodes periodically to force generate
# cross node memory access
while [ -d "/proc/$PID" ]
do
        taskset -apc 8 $JOB1
        taskset -apc 8 $JOB2
        sleep 300
        taskset -apc 58 $JOB1
        taskset -apc 58 $JOB2
        sleep 300
done

With the above test the histogram of latency of do_huge_pmd_numa_page is
as shown below.  Since the number of do_huge_pmd_numa_page varies
drastically for each run (should be due to scheduler), so I converted the
raw number to percentage.

                             patched               base
@us[stress-ng]:
[0]                          3.57%                 0.16%
[1]                          55.68%                18.36%
[2, 4)                       10.46%                40.44%
[4, 8)                       7.26%                 17.82%
[8, 16)                      21.12%                13.41%
[16, 32)                     1.06%                 4.27%
[32, 64)                     0.56%                 4.07%
[64, 128)                    0.16%                 0.35%
[128, 256)                   < 0.1%                < 0.1%
[256, 512)                   < 0.1%                < 0.1%
[512, 1K)                    < 0.1%                < 0.1%
[1K, 2K)                     < 0.1%                < 0.1%
[2K, 4K)                     < 0.1%                < 0.1%
[4K, 8K)                     < 0.1%                < 0.1%
[8K, 16K)                    < 0.1%                < 0.1%
[16K, 32K)                   < 0.1%                < 0.1%
[32K, 64K)                   < 0.1%                < 0.1%

Per the result, patched kernel is even slightly better than the base
kernel.  I think this is because the lock contention against THP split is
less than base kernel due to the refactor.

To exclude the affect from THP split, I also did test w/o memory pressure.
No obvious regression is spotted.  The below is the test result *w/o*
memory pressure.

                           patched                  base
@us[stress-ng]:
[0]                        7.97%                   18.4%
[1]                        69.63%                  58.24%
[2, 4)                     4.18%                   2.63%
[4, 8)                     0.22%                   0.17%
[8, 16)                    1.03%                   0.92%
[16, 32)                   0.14%                   < 0.1%
[32, 64)                   < 0.1%                  < 0.1%
[64, 128)                  < 0.1%                  < 0.1%
[128, 256)                 < 0.1%                  < 0.1%
[256, 512)                 0.45%                   1.19%
[512, 1K)                  15.45%                  17.27%
[1K, 2K)                   < 0.1%                  < 0.1%
[2K, 4K)                   < 0.1%                  < 0.1%
[4K, 8K)                   < 0.1%                  < 0.1%
[8K, 16K)                  0.86%                   0.88%
[16K, 32K)                 < 0.1%                  0.15%
[32K, 64K)                 < 0.1%                  < 0.1%
[64K, 128K)                < 0.1%                  < 0.1%
[128K, 256K)               < 0.1%                  < 0.1%

The series also survived a series of tests that exercise NUMA balancing
migrations by Mel.

This patch (of 7):

Add orig_pmd to struct vm_fault so the "orig_pmd" parameter used by huge
page fault could be removed, just like its PTE counterpart does.

Link: https://lkml.kernel.org/r/20210518200801.7413-1-shy828301@gmail.com
Link: https://lkml.kernel.org/r/20210518200801.7413-2-shy828301@gmail.comSigned-off-by: NYang Shi <shy828301@gmail.com>
Acked-by: NMel Gorman <mgorman@suse.de>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      5db4f15c
    • A
      userfaultfd/shmem: support minor fault registration for shmem · c949b097
      由 Axel Rasmussen 提交于 
      This patch allows shmem-backed VMAs to be registered for minor faults.
Minor faults are appropriately relayed to userspace in the fault path, for
VMAs with the relevant flag.

This commit doesn't hook up the UFFDIO_CONTINUE ioctl for shmem-backed
minor faults, though, so userspace doesn't yet have a way to resolve such
faults.

Because of this, we also don't yet advertise this as a supported feature.
That will be done in a separate commit when the feature is fully
implemented.

Link: https://lkml.kernel.org/r/20210503180737.2487560-4-axelrasmussen@google.comSigned-off-by: NAxel Rasmussen <axelrasmussen@google.com>
Acked-by: NPeter Xu <peterx@redhat.com>
Acked-by: NHugh Dickins <hughd@google.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Brian Geffon <bgeffon@google.com>
Cc: "Dr . David Alan Gilbert" <dgilbert@redhat.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: Joe Perches <joe@perches.com>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Lokesh Gidra <lokeshgidra@google.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.vnet.ibm.com>
Cc: Mina Almasry <almasrymina@google.com>
Cc: Oliver Upton <oupton@google.com>
Cc: Shaohua Li <shli@fb.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Wang Qing <wangqing@vivo.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      c949b097
    • P
      mm/userfaultfd: fix uffd-wp special cases for fork() · 8f34f1ea
      由 Peter Xu 提交于 
      We tried to do something similar in b569a176 ("userfaultfd: wp: drop
_PAGE_UFFD_WP properly when fork") previously, but it's not doing it all
right..  A few fixes around the code path:

1. We were referencing VM_UFFD_WP vm_flags on the _old_ vma rather
   than the new vma.  That's overlooked in b569a176, so it won't work
   as expected.  Thanks to the recent rework on fork code
   (7a4830c3), we can easily get the new vma now, so switch the
   checks to that.

2. Dropping the uffd-wp bit in copy_huge_pmd() could be wrong if the
   huge pmd is a migration huge pmd.  When it happens, instead of using
   pmd_uffd_wp(), we should use pmd_swp_uffd_wp().  The fix is simply to
   handle them separately.

3. Forget to carry over uffd-wp bit for a write migration huge pmd
   entry.  This also happens in copy_huge_pmd(), where we converted a
   write huge migration entry into a read one.

4. In copy_nonpresent_pte(), drop uffd-wp if necessary for swap ptes.

5. In copy_present_page() when COW is enforced when fork(), we also
   need to pass over the uffd-wp bit if VM_UFFD_WP is armed on the new
   vma, and when the pte to be copied has uffd-wp bit set.

Remove the comment in copy_present_pte() about this.  It won't help a huge
lot to only comment there, but comment everywhere would be an overkill.
Let's assume the commit messages would help.

[peterx@redhat.com: fix a few thp pmd missing uffd-wp bit]
  Link: https://lkml.kernel.org/r/20210428225030.9708-4-peterx@redhat.com

Link: https://lkml.kernel.org/r/20210428225030.9708-3-peterx@redhat.com
Fixes: b569a176 ("userfaultfd: wp: drop _PAGE_UFFD_WP properly when fork")
Signed-off-by: NPeter Xu <peterx@redhat.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: Mike Rapoport <rppt@linux.vnet.ibm.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Axel Rasmussen <axelrasmussen@google.com>
Cc: Brian Geffon <bgeffon@google.com>
Cc: "Dr . David Alan Gilbert" <dgilbert@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Joe Perches <joe@perches.com>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Lokesh Gidra <lokeshgidra@google.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mina Almasry <almasrymina@google.com>
Cc: Oliver Upton <oupton@google.com>
Cc: Shaohua Li <shli@fb.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Wang Qing <wangqing@vivo.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      8f34f1ea
  5. 30 6月, 2021 5 次提交
  7. 17 6月, 2021 1 次提交
    • H
      mm/thp: unmap_mapping_page() to fix THP truncate_cleanup_page() · 22061a1f
      由 Hugh Dickins 提交于 
      There is a race between THP unmapping and truncation, when truncate sees
pmd_none() and skips the entry, after munmap's zap_huge_pmd() cleared
it, but before its page_remove_rmap() gets to decrement
compound_mapcount: generating false "BUG: Bad page cache" reports that
the page is still mapped when deleted.  This commit fixes that, but not
in the way I hoped.

The first attempt used try_to_unmap(page, TTU_SYNC|TTU_IGNORE_MLOCK)
instead of unmap_mapping_range() in truncate_cleanup_page(): it has
often been an annoyance that we usually call unmap_mapping_range() with
no pages locked, but there apply it to a single locked page.
try_to_unmap() looks more suitable for a single locked page.

However, try_to_unmap_one() contains a VM_BUG_ON_PAGE(!pvmw.pte,page):
it is used to insert THP migration entries, but not used to unmap THPs.
Copy zap_huge_pmd() and add THP handling now? Perhaps, but their TLB
needs are different, I'm too ignorant of the DAX cases, and couldn't
decide how far to go for anon+swap.  Set that aside.

The second attempt took a different tack: make no change in truncate.c,
but modify zap_huge_pmd() to insert an invalidated huge pmd instead of
clearing it initially, then pmd_clear() between page_remove_rmap() and
unlocking at the end.  Nice.  But powerpc blows that approach out of the
water, with its serialize_against_pte_lookup(), and interesting pgtable
usage.  It would need serious help to get working on powerpc (with a
minor optimization issue on s390 too).  Set that aside.

Just add an "if (page_mapped(page)) synchronize_rcu();" or other such
delay, after unmapping in truncate_cleanup_page()? Perhaps, but though
that's likely to reduce or eliminate the number of incidents, it would
give less assurance of whether we had identified the problem correctly.

This successful iteration introduces "unmap_mapping_page(page)" instead
of try_to_unmap(), and goes the usual unmap_mapping_range_tree() route,
with an addition to details.  Then zap_pmd_range() watches for this
case, and does spin_unlock(pmd_lock) if so - just like
page_vma_mapped_walk() now does in the PVMW_SYNC case.  Not pretty, but
safe.

Note that unmap_mapping_page() is doing a VM_BUG_ON(!PageLocked) to
assert its interface; but currently that's only used to make sure that
page->mapping is stable, and zap_pmd_range() doesn't care if the page is
locked or not.  Along these lines, in invalidate_inode_pages2_range()
move the initial unmap_mapping_range() out from under page lock, before
then calling unmap_mapping_page() under page lock if still mapped.

Link: https://lkml.kernel.org/r/a2a4a148-cdd8-942c-4ef8-51b77f643dbe@google.com
Fixes: fc127da0 ("truncate: handle file thp")
Signed-off-by: NHugh Dickins <hughd@google.com>
Acked-by: NKirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: NYang Shi <shy828301@gmail.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jue Wang <juew@google.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Wang Yugui <wangyugui@e16-tech.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      22061a1f
  9. 05 6月, 2021 1 次提交
  11. 07 5月, 2021 2 次提交
  13. 01 5月, 2021 5 次提交
  15. 31 3月, 2021 1 次提交
    • I
      mm: fix race by making init_zero_pfn() early_initcall · e720e7d0
      由 Ilya Lipnitskiy 提交于 
      There are code paths that rely on zero_pfn to be fully initialized
before core_initcall.  For example, wq_sysfs_init() is a core_initcall
function that eventually results in a call to kernel_execve, which
causes a page fault with a subsequent mmput.  If zero_pfn is not
initialized by then it may not get cleaned up properly and result in an
error:

  BUG: Bad rss-counter state mm:(ptrval) type:MM_ANONPAGES val:1

Here is an analysis of the race as seen on a MIPS device. On this
particular MT7621 device (Ubiquiti ER-X), zero_pfn is PFN 0 until
initialized, at which point it becomes PFN 5120:

  1. wq_sysfs_init calls into kobject_uevent_env at core_initcall:
       kobject_uevent_env+0x7e4/0x7ec
       kset_register+0x68/0x88
       bus_register+0xdc/0x34c
       subsys_virtual_register+0x34/0x78
       wq_sysfs_init+0x1c/0x4c
       do_one_initcall+0x50/0x1a8
       kernel_init_freeable+0x230/0x2c8
       kernel_init+0x10/0x100
       ret_from_kernel_thread+0x14/0x1c

  2. kobject_uevent_env() calls call_usermodehelper_exec() which executes
     kernel_execve asynchronously.

  3. Memory allocations in kernel_execve cause a page fault, bumping the
     MM reference counter:
       add_mm_counter_fast+0xb4/0xc0
       handle_mm_fault+0x6e4/0xea0
       __get_user_pages.part.78+0x190/0x37c
       __get_user_pages_remote+0x128/0x360
       get_arg_page+0x34/0xa0
       copy_string_kernel+0x194/0x2a4
       kernel_execve+0x11c/0x298
       call_usermodehelper_exec_async+0x114/0x194

  4. In case zero_pfn has not been initialized yet, zap_pte_range does
     not decrement the MM_ANONPAGES RSS counter and the BUG message is
     triggered shortly afterwards when __mmdrop checks the ref counters:
       __mmdrop+0x98/0x1d0
       free_bprm+0x44/0x118
       kernel_execve+0x160/0x1d8
       call_usermodehelper_exec_async+0x114/0x194
       ret_from_kernel_thread+0x14/0x1c

To avoid races such as described above, initialize init_zero_pfn at
early_initcall level.  Depending on the architecture, ZERO_PAGE is
either constant or gets initialized even earlier, at paging_init, so
there is no issue with initializing zero_pfn earlier.

Link: https://lkml.kernel.org/r/CALCv0x2YqOXEAy2Q=hafjhHCtTHVodChv1qpM=niAXOpqEbt7w@mail.gmail.comSigned-off-by: NIlya Lipnitskiy <ilya.lipnitskiy@gmail.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: stable@vger.kernel.org
Tested-by: N周琰杰 (Zhou Yanjie) <zhouyanjie@wanyeetech.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      e720e7d0
  17. 14 3月, 2021 2 次提交
    • N
      mm/userfaultfd: fix memory corruption due to writeprotect · 6ce64428
      由 Nadav Amit 提交于 
      Userfaultfd self-test fails occasionally, indicating a memory corruption.

Analyzing this problem indicates that there is a real bug since mmap_lock
is only taken for read in mwriteprotect_range() and defers flushes, and
since there is insufficient consideration of concurrent deferred TLB
flushes in wp_page_copy().  Although the PTE is flushed from the TLBs in
wp_page_copy(), this flush takes place after the copy has already been
performed, and therefore changes of the page are possible between the time
of the copy and the time in which the PTE is flushed.

To make matters worse, memory-unprotection using userfaultfd also poses a
problem.  Although memory unprotection is logically a promotion of PTE
permissions, and therefore should not require a TLB flush, the current
userrfaultfd code might actually cause a demotion of the architectural PTE
permission: when userfaultfd_writeprotect() unprotects memory region, it
unintentionally *clears* the RW-bit if it was already set.  Note that this
unprotecting a PTE that is not write-protected is a valid use-case: the
userfaultfd monitor might ask to unprotect a region that holds both
write-protected and write-unprotected PTEs.

The scenario that happens in selftests/vm/userfaultfd is as follows:

cpu0				cpu1			cpu2
----				----			----
							[ Writable PTE
							  cached in TLB ]
userfaultfd_writeprotect()
[ write-*unprotect* ]
mwriteprotect_range()
mmap_read_lock()
change_protection()

change_protection_range()
...
change_pte_range()
[ *clear* “write”-bit ]
[ defer TLB flushes ]
				[ page-fault ]
				...
				wp_page_copy()
				 cow_user_page()
				  [ copy page ]
							[ write to old
							  page ]
				...
				 set_pte_at_notify()

A similar scenario can happen:

cpu0		cpu1		cpu2		cpu3
----		----		----		----
						[ Writable PTE
				  		  cached in TLB ]
userfaultfd_writeprotect()
[ write-protect ]
[ deferred TLB flush ]
		userfaultfd_writeprotect()
		[ write-unprotect ]
		[ deferred TLB flush]
				[ page-fault ]
				wp_page_copy()
				 cow_user_page()
				 [ copy page ]
				 ...		[ write to page ]
				set_pte_at_notify()

This race exists since commit 292924b2 ("userfaultfd: wp: apply
_PAGE_UFFD_WP bit").  Yet, as Yu Zhao pointed, these races became apparent
since commit 09854ba9 ("mm: do_wp_page() simplification") which made
wp_page_copy() more likely to take place, specifically if page_count(page)
> 1.

To resolve the aforementioned races, check whether there are pending
flushes on uffd-write-protected VMAs, and if there are, perform a flush
before doing the COW.

Further optimizations will follow to avoid during uffd-write-unprotect
unnecassary PTE write-protection and TLB flushes.

Link: https://lkml.kernel.org/r/20210304095423.3825684-1-namit@vmware.com
Fixes: 09854ba9 ("mm: do_wp_page() simplification")
Signed-off-by: NNadav Amit <namit@vmware.com>
Suggested-by: NYu Zhao <yuzhao@google.com>
Reviewed-by: NPeter Xu <peterx@redhat.com>
Tested-by: NPeter Xu <peterx@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.vnet.ibm.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Will Deacon <will@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: <stable@vger.kernel.org>	[5.9+]
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      6ce64428
    • P
      mm: introduce page_needs_cow_for_dma() for deciding whether cow · 97a7e473
      由 Peter Xu 提交于 
      We've got quite a few places (pte, pmd, pud) that explicitly checked
against whether we should break the cow right now during fork().  It's
easier to provide a helper, especially before we work the same thing on
hugetlbfs.

Since we'll reference is_cow_mapping() in mm.h, move it there too.
Actually it suites mm.h more since internal.h is mm/ only, but mm.h is
exported to the whole kernel.  With that we should expect another patch to
use is_cow_mapping() whenever we can across the kernel since we do use it
quite a lot but it's always done with raw code against VM_* flags.

Link: https://lkml.kernel.org/r/20210217233547.93892-4-peterx@redhat.comSigned-off-by: NPeter Xu <peterx@redhat.com>
Reviewed-by: NJason Gunthorpe <jgg@ziepe.ca>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Airlie <airlied@linux.ie>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: Gal Pressman <galpress@amazon.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jann Horn <jannh@google.com>
Cc: Kirill Shutemov <kirill@shutemov.name>
Cc: Kirill Tkhai <ktkhai@virtuozzo.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Mike Rapoport <rppt@linux.vnet.ibm.com>
Cc: Roland Scheidegger <sroland@vmware.com>
Cc: VMware Graphics <linux-graphics-maintainer@vmware.com>
Cc: Wei Zhang <wzam@amazon.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      97a7e473
  19. 27 2月, 2021 1 次提交
  21. 25 2月, 2021 3 次提交
  23. 09 2月, 2021 1 次提交
    • P
      mm: provide a saner PTE walking API for modules · 9fd6dad1
      由 Paolo Bonzini 提交于 
      Currently, the follow_pfn function is exported for modules but
follow_pte is not.  However, follow_pfn is very easy to misuse,
because it does not provide protections (so most of its callers
assume the page is writable!) and because it returns after having
already unlocked the page table lock.

Provide instead a simplified version of follow_pte that does
not have the pmdpp and range arguments.  The older version
survives as follow_invalidate_pte() for use by fs/dax.c.
Reviewed-by: NJason Gunthorpe <jgg@nvidia.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
      9fd6dad1
  25. 30 1月, 2021 2 次提交
  27. 21 1月, 2021 1 次提交
    • W
      mm: Pass 'address' to map to do_set_pte() and drop FAULT_FLAG_PREFAULT · 9d3af4b4
      由 Will Deacon 提交于 
      Rather than modifying the 'address' field of the 'struct vm_fault'
passed to do_set_pte(), leave that to identify the real faulting address
and pass in the virtual address to be mapped by the new pte as a
separate argument.

This makes FAULT_FLAG_PREFAULT redundant, as a prefault entry can be
identified simply by comparing the new address parameter with the
faulting address, so remove the redundant flag at the same time.

Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NWill Deacon <will@kernel.org>
      9d3af4b4
  29. 20 1月, 2021 2 次提交
    • W
      mm: Allow architectures to request 'old' entries when prefaulting · 46bdb427
      由 Will Deacon 提交于 
      Commit 5c0a85fa ("mm: make faultaround produce old ptes") changed
the "faultaround" behaviour to initialise prefaulted PTEs as 'old',
since this avoids vmscan wrongly assuming that they are hot, despite
having never been explicitly accessed by userspace. The change has been
shown to benefit numerous arm64 micro-architectures (with hardware
access flag) running Android, where both application launch latency and
direct reclaim time are significantly reduced (by 10%+ and ~80%
respectively).

Unfortunately, commit 315d09bf ("Revert "mm: make faultaround
produce old ptes"") reverted the change due to it being identified as
the cause of a ~6% regression in unixbench on x86. Experiments on a
variety of recent arm64 micro-architectures indicate that unixbench is
not affected by the original commit, which appears to yield a 0-1%
performance improvement.

Since one size does not fit all for the initial state of prefaulted
PTEs, introduce arch_wants_old_prefaulted_pte(), which allows an
architecture to opt-in to 'old' prefaulted PTEs at runtime based on
whatever criteria it may have.

Cc: Jan Kara <jack@suse.cz>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Reported-by: NVinayak Menon <vinmenon@codeaurora.org>
Signed-off-by: NWill Deacon <will@kernel.org>
      46bdb427
    • K
      mm: Cleanup faultaround and finish_fault() codepaths · f9ce0be7
      由 Kirill A. Shutemov 提交于 
      alloc_set_pte() has two users with different requirements: in the
faultaround code, it called from an atomic context and PTE page table
has to be preallocated. finish_fault() can sleep and allocate page table
as needed.

PTL locking rules are also strange, hard to follow and overkill for
finish_fault().

Let's untangle the mess. alloc_set_pte() has gone now. All locking is
explicit.

The price is some code duplication to handle huge pages in faultaround
path, but it should be fine, having overall improvement in readability.

Link: https://lore.kernel.org/r/20201229132819.najtavneutnf7ajp@boxSigned-off-by: NKirill A. Shutemov <kirill.shutemov@linux.intel.com>
[will: s/from from/from/ in comment; spotted by willy]
Signed-off-by: NWill Deacon <will@kernel.org>
      f9ce0be7
  31. 12 1月, 2021 1 次提交
    • D
      mm: Close race in generic_access_phys · 96667f8a
      由 Daniel Vetter 提交于 
      Way back it was a reasonable assumptions that iomem mappings never
change the pfn range they point at. But this has changed:

- gpu drivers dynamically manage their memory nowadays, invalidating
  ptes with unmap_mapping_range when buffers get moved

- contiguous dma allocations have moved from dedicated carvetouts to
  cma regions. This means if we miss the unmap the pfn might contain
  pagecache or anon memory (well anything allocated with GFP_MOVEABLE)

- even /dev/mem now invalidates mappings when the kernel requests that
  iomem region when CONFIG_IO_STRICT_DEVMEM is set, see 3234ac66
  ("/dev/mem: Revoke mappings when a driver claims the region")

Accessing pfns obtained from ptes without holding all the locks is
therefore no longer a good idea. Fix this.

Since ioremap might need to manipulate pagetables too we need to drop
the pt lock and have a retry loop if we raced.

While at it, also add kerneldoc and improve the comment for the
vma_ops->access function. It's for accessing, not for moving the
memory from iomem to system memory, as the old comment seemed to
suggest.

References: 28b2ee20 ("access_process_vm device memory infrastructure")
Signed-off-by: NDaniel Vetter <daniel.vetter@intel.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Benjamin Herrensmidt <benh@kernel.crashing.org>
Cc: Dave Airlie <airlied@linux.ie>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: linux-mm@kvack.org
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-samsung-soc@vger.kernel.org
Cc: linux-media@vger.kernel.org
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20201127164131.2244124-8-daniel.vetter@ffwll.ch
      96667f8a
  33. 30 12月, 2020 1 次提交
    • N
      mm: generalise COW SMC TLB flushing race comment · 111fe718
      由 Nicholas Piggin 提交于 
      I'm not sure if I'm completely missing something here, but AFAIKS the
reference to the mysterious "COW SMC race" confuses the issue.  The
original changelog and mailing list thread didn't help me either.

This SMC race is where the problem was detected, but isn't the general
problem bigger and more obvious: that the new PTE could be picked up at
any time by any TLB while entries for the old PTE exist in other TLBs
before the TLB flush takes effect?

The case where the iTLB and dTLB of a CPU are pointing at different pages
is an interesting one but follows from the general problem.

The other (minor) thing with the comment I think it makes it a bit clearer
to say what the old code was doing (i.e., it avoids the race as opposed to
what?).

References: 4ce072f1 ("mm: fix a race condition under SMC + COW")
Link: https://lkml.kernel.org/r/20201215121119.351650-1-npiggin@gmail.comReviewed-by: NMatthew Wilcox (Oracle) <willy@infradead.org>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Suresh Siddha <sbsiddha@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      111fe718
  35. 16 12月, 2020 4 次提交
  37. 19 10月, 2020 1 次提交
  39. 17 10月, 2020 1 次提交