diff options
author | Andrea Arcangeli <aarcange@redhat.com> | 2012-03-21 16:33:42 -0700 |
---|---|---|
committer | Greg Kroah-Hartman <gregkh@linuxfoundation.org> | 2012-04-02 09:27:10 -0700 |
commit | 5a3e1f550cfc86a68729770bcfa28f36b238b34d (patch) | |
tree | 68104c35c656a8b90ea720351c090c64fa3f1a04 /include/asm-generic | |
parent | c2ec63edaf48c90c3495eeb0b75bb05102fbf71a (diff) |
mm: thp: fix pmd_bad() triggering in code paths holding mmap_sem read mode
commit 1a5a9906d4e8d1976b701f889d8f35d54b928f25 upstream.
In some cases it may happen that pmd_none_or_clear_bad() is called with
the mmap_sem hold in read mode. In those cases the huge page faults can
allocate hugepmds under pmd_none_or_clear_bad() and that can trigger a
false positive from pmd_bad() that will not like to see a pmd
materializing as trans huge.
It's not khugepaged causing the problem, khugepaged holds the mmap_sem
in write mode (and all those sites must hold the mmap_sem in read mode
to prevent pagetables to go away from under them, during code review it
seems vm86 mode on 32bit kernels requires that too unless it's
restricted to 1 thread per process or UP builds). The race is only with
the huge pagefaults that can convert a pmd_none() into a
pmd_trans_huge().
Effectively all these pmd_none_or_clear_bad() sites running with
mmap_sem in read mode are somewhat speculative with the page faults, and
the result is always undefined when they run simultaneously. This is
probably why it wasn't common to run into this. For example if the
madvise(MADV_DONTNEED) runs zap_page_range() shortly before the page
fault, the hugepage will not be zapped, if the page fault runs first it
will be zapped.
Altering pmd_bad() not to error out if it finds hugepmds won't be enough
to fix this, because zap_pmd_range would then proceed to call
zap_pte_range (which would be incorrect if the pmd become a
pmd_trans_huge()).
The simplest way to fix this is to read the pmd in the local stack
(regardless of what we read, no need of actual CPU barriers, only
compiler barrier needed), and be sure it is not changing under the code
that computes its value. Even if the real pmd is changing under the
value we hold on the stack, we don't care. If we actually end up in
zap_pte_range it means the pmd was not none already and it was not huge,
and it can't become huge from under us (khugepaged locking explained
above).
All we need is to enforce that there is no way anymore that in a code
path like below, pmd_trans_huge can be false, but pmd_none_or_clear_bad
can run into a hugepmd. The overhead of a barrier() is just a compiler
tweak and should not be measurable (I only added it for THP builds). I
don't exclude different compiler versions may have prevented the race
too by caching the value of *pmd on the stack (that hasn't been
verified, but it wouldn't be impossible considering
pmd_none_or_clear_bad, pmd_bad, pmd_trans_huge, pmd_none are all inlines
and there's no external function called in between pmd_trans_huge and
pmd_none_or_clear_bad).
if (pmd_trans_huge(*pmd)) {
if (next-addr != HPAGE_PMD_SIZE) {
VM_BUG_ON(!rwsem_is_locked(&tlb->mm->mmap_sem));
split_huge_page_pmd(vma->vm_mm, pmd);
} else if (zap_huge_pmd(tlb, vma, pmd, addr))
continue;
/* fall through */
}
if (pmd_none_or_clear_bad(pmd))
Because this race condition could be exercised without special
privileges this was reported in CVE-2012-1179.
The race was identified and fully explained by Ulrich who debugged it.
I'm quoting his accurate explanation below, for reference.
====== start quote =======
mapcount 0 page_mapcount 1
kernel BUG at mm/huge_memory.c:1384!
At some point prior to the panic, a "bad pmd ..." message similar to the
following is logged on the console:
mm/memory.c:145: bad pmd ffff8800376e1f98(80000000314000e7).
The "bad pmd ..." message is logged by pmd_clear_bad() before it clears
the page's PMD table entry.
143 void pmd_clear_bad(pmd_t *pmd)
144 {
-> 145 pmd_ERROR(*pmd);
146 pmd_clear(pmd);
147 }
After the PMD table entry has been cleared, there is an inconsistency
between the actual number of PMD table entries that are mapping the page
and the page's map count (_mapcount field in struct page). When the page
is subsequently reclaimed, __split_huge_page() detects this inconsistency.
1381 if (mapcount != page_mapcount(page))
1382 printk(KERN_ERR "mapcount %d page_mapcount %d\n",
1383 mapcount, page_mapcount(page));
-> 1384 BUG_ON(mapcount != page_mapcount(page));
The root cause of the problem is a race of two threads in a multithreaded
process. Thread B incurs a page fault on a virtual address that has never
been accessed (PMD entry is zero) while Thread A is executing an madvise()
system call on a virtual address within the same 2 MB (huge page) range.
virtual address space
.---------------------.
| |
| |
.-|---------------------|
| | |
| | |<-- B(fault)
| | |
2 MB | |/////////////////////|-.
huge < |/////////////////////| > A(range)
page | |/////////////////////|-'
| | |
| | |
'-|---------------------|
| |
| |
'---------------------'
- Thread A is executing an madvise(..., MADV_DONTNEED) system call
on the virtual address range "A(range)" shown in the picture.
sys_madvise
// Acquire the semaphore in shared mode.
down_read(¤t->mm->mmap_sem)
...
madvise_vma
switch (behavior)
case MADV_DONTNEED:
madvise_dontneed
zap_page_range
unmap_vmas
unmap_page_range
zap_pud_range
zap_pmd_range
//
// Assume that this huge page has never been accessed.
// I.e. content of the PMD entry is zero (not mapped).
//
if (pmd_trans_huge(*pmd)) {
// We don't get here due to the above assumption.
}
//
// Assume that Thread B incurred a page fault and
.---------> // sneaks in here as shown below.
| //
| if (pmd_none_or_clear_bad(pmd))
| {
| if (unlikely(pmd_bad(*pmd)))
| pmd_clear_bad
| {
| pmd_ERROR
| // Log "bad pmd ..." message here.
| pmd_clear
| // Clear the page's PMD entry.
| // Thread B incremented the map count
| // in page_add_new_anon_rmap(), but
| // now the page is no longer mapped
| // by a PMD entry (-> inconsistency).
| }
| }
|
v
- Thread B is handling a page fault on virtual address "B(fault)" shown
in the picture.
...
do_page_fault
__do_page_fault
// Acquire the semaphore in shared mode.
down_read_trylock(&mm->mmap_sem)
...
handle_mm_fault
if (pmd_none(*pmd) && transparent_hugepage_enabled(vma))
// We get here due to the above assumption (PMD entry is zero).
do_huge_pmd_anonymous_page
alloc_hugepage_vma
// Allocate a new transparent huge page here.
...
__do_huge_pmd_anonymous_page
...
spin_lock(&mm->page_table_lock)
...
page_add_new_anon_rmap
// Here we increment the page's map count (starts at -1).
atomic_set(&page->_mapcount, 0)
set_pmd_at
// Here we set the page's PMD entry which will be cleared
// when Thread A calls pmd_clear_bad().
...
spin_unlock(&mm->page_table_lock)
The mmap_sem does not prevent the race because both threads are acquiring
it in shared mode (down_read). Thread B holds the page_table_lock while
the page's map count and PMD table entry are updated. However, Thread A
does not synchronize on that lock.
====== end quote =======
[akpm@linux-foundation.org: checkpatch fixes]
Reported-by: Ulrich Obergfell <uobergfe@redhat.com>
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: Dave Jones <davej@redhat.com>
Acked-by: Larry Woodman <lwoodman@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Mark Salter <msalter@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Diffstat (limited to 'include/asm-generic')
-rw-r--r-- | include/asm-generic/pgtable.h | 61 |
1 files changed, 61 insertions, 0 deletions
diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h index 76bff2bff15..a03c098b0cc 100644 --- a/include/asm-generic/pgtable.h +++ b/include/asm-generic/pgtable.h @@ -425,6 +425,8 @@ extern void untrack_pfn_vma(struct vm_area_struct *vma, unsigned long pfn, unsigned long size); #endif +#ifdef CONFIG_MMU + #ifndef CONFIG_TRANSPARENT_HUGEPAGE static inline int pmd_trans_huge(pmd_t pmd) { @@ -441,7 +443,66 @@ static inline int pmd_write(pmd_t pmd) return 0; } #endif /* __HAVE_ARCH_PMD_WRITE */ +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ + +/* + * This function is meant to be used by sites walking pagetables with + * the mmap_sem hold in read mode to protect against MADV_DONTNEED and + * transhuge page faults. MADV_DONTNEED can convert a transhuge pmd + * into a null pmd and the transhuge page fault can convert a null pmd + * into an hugepmd or into a regular pmd (if the hugepage allocation + * fails). While holding the mmap_sem in read mode the pmd becomes + * stable and stops changing under us only if it's not null and not a + * transhuge pmd. When those races occurs and this function makes a + * difference vs the standard pmd_none_or_clear_bad, the result is + * undefined so behaving like if the pmd was none is safe (because it + * can return none anyway). The compiler level barrier() is critically + * important to compute the two checks atomically on the same pmdval. + */ +static inline int pmd_none_or_trans_huge_or_clear_bad(pmd_t *pmd) +{ + /* depend on compiler for an atomic pmd read */ + pmd_t pmdval = *pmd; + /* + * The barrier will stabilize the pmdval in a register or on + * the stack so that it will stop changing under the code. + */ +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + barrier(); +#endif + if (pmd_none(pmdval)) + return 1; + if (unlikely(pmd_bad(pmdval))) { + if (!pmd_trans_huge(pmdval)) + pmd_clear_bad(pmd); + return 1; + } + return 0; +} + +/* + * This is a noop if Transparent Hugepage Support is not built into + * the kernel. Otherwise it is equivalent to + * pmd_none_or_trans_huge_or_clear_bad(), and shall only be called in + * places that already verified the pmd is not none and they want to + * walk ptes while holding the mmap sem in read mode (write mode don't + * need this). If THP is not enabled, the pmd can't go away under the + * code even if MADV_DONTNEED runs, but if THP is enabled we need to + * run a pmd_trans_unstable before walking the ptes after + * split_huge_page_pmd returns (because it may have run when the pmd + * become null, but then a page fault can map in a THP and not a + * regular page). + */ +static inline int pmd_trans_unstable(pmd_t *pmd) +{ +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + return pmd_none_or_trans_huge_or_clear_bad(pmd); +#else + return 0; #endif +} + +#endif /* CONFIG_MMU */ #endif /* !__ASSEMBLY__ */ |