diff options
| author | Kirill A. Shutemov <kirill.shutemov@linux.intel.com> | 2016-03-17 14:20:13 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2016-03-17 15:09:34 -0700 |
| commit | 5f7377147ce0b5d521c0ee2aab2ec3ead51297db (patch) | |
| tree | 4525af234d84b4ff3fe3338ce6faee08630b11f2 | |
| parent | fec89c109f3a7737fe3a7bf0f40d1fb7709d353b (diff) | |
| download | linux-5f7377147ce0b5d521c0ee2aab2ec3ead51297db.tar.bz2 | |
thp: fix deadlock in split_huge_pmd()
split_huge_pmd() tries to munlock page with munlock_vma_page(). That
requires the page to locked.
If the is locked by caller, we would get a deadlock:
Unable to find swap-space signature
INFO: task trinity-c85:1907 blocked for more than 120 seconds.
Not tainted 4.4.0-00032-gf19d0bdced41-dirty #1606
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
trinity-c85 D ffff88084d997608 0 1907 309 0x00000000
Call Trace:
schedule+0x9f/0x1c0
schedule_timeout+0x48e/0x600
io_schedule_timeout+0x1c3/0x390
bit_wait_io+0x29/0xd0
__wait_on_bit_lock+0x94/0x140
__lock_page+0x1d4/0x280
__split_huge_pmd+0x5a8/0x10f0
split_huge_pmd_address+0x1d9/0x230
try_to_unmap_one+0x540/0xc70
rmap_walk_anon+0x284/0x810
rmap_walk_locked+0x11e/0x190
try_to_unmap+0x1b1/0x4b0
split_huge_page_to_list+0x49d/0x18a0
follow_page_mask+0xa36/0xea0
SyS_move_pages+0xaf3/0x1570
entry_SYSCALL_64_fastpath+0x12/0x6b
2 locks held by trinity-c85/1907:
#0: (&mm->mmap_sem){++++++}, at: SyS_move_pages+0x933/0x1570
#1: (&anon_vma->rwsem){++++..}, at: split_huge_page_to_list+0x402/0x18a0
I don't think the deadlock is triggerable without split_huge_page()
simplifilcation patchset.
But munlock_vma_page() here is wrong: we want to munlock the page
unconditionally, no need in rmap lookup, that munlock_vma_page() does.
Let's use clear_page_mlock() instead. It can be called under ptl.
Fixes: e90309c9f772 ("thp: allow mlocked THP again")
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
| -rw-r--r-- | mm/huge_memory.c | 13 |
1 files changed, 2 insertions, 11 deletions
diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 4a58fa19afac..021db1781872 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -2981,29 +2981,20 @@ void __split_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd, { spinlock_t *ptl; struct mm_struct *mm = vma->vm_mm; - struct page *page = NULL; unsigned long haddr = address & HPAGE_PMD_MASK; mmu_notifier_invalidate_range_start(mm, haddr, haddr + HPAGE_PMD_SIZE); ptl = pmd_lock(mm, pmd); if (pmd_trans_huge(*pmd)) { - page = pmd_page(*pmd); + struct page *page = pmd_page(*pmd); if (PageMlocked(page)) - get_page(page); - else - page = NULL; + clear_page_mlock(page); } else if (!pmd_devmap(*pmd)) goto out; __split_huge_pmd_locked(vma, pmd, haddr, freeze); out: spin_unlock(ptl); mmu_notifier_invalidate_range_end(mm, haddr, haddr + HPAGE_PMD_SIZE); - if (page) { - lock_page(page); - munlock_vma_page(page); - unlock_page(page); - put_page(page); - } } void split_huge_pmd_address(struct vm_area_struct *vma, unsigned long address, |