On Tue, Feb 15, 2022 at 02:15:03PM -0800, Yang Shi wrote: > commit 24d7275ce2791829953ed4e72f68277ceb2571c6 upstream > > The syzbot reported the below BUG: > > kernel BUG at include/linux/page-flags.h:785! > invalid opcode: 0000 [#1] PREEMPT SMP KASAN > CPU: 1 PID: 4392 Comm: syz-executor560 Not tainted 5.16.0-rc6-syzkaller #0 > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 > RIP: 0010:PageDoubleMap include/linux/page-flags.h:785 [inline] > RIP: 0010:__page_mapcount+0x2d2/0x350 mm/util.c:744 > Call Trace: > page_mapcount include/linux/mm.h:837 [inline] > smaps_account+0x470/0xb10 fs/proc/task_mmu.c:466 > smaps_pte_entry fs/proc/task_mmu.c:538 [inline] > smaps_pte_range+0x611/0x1250 fs/proc/task_mmu.c:601 > walk_pmd_range mm/pagewalk.c:128 [inline] > walk_pud_range mm/pagewalk.c:205 [inline] > walk_p4d_range mm/pagewalk.c:240 [inline] > walk_pgd_range mm/pagewalk.c:277 [inline] > __walk_page_range+0xe23/0x1ea0 mm/pagewalk.c:379 > walk_page_vma+0x277/0x350 mm/pagewalk.c:530 > smap_gather_stats.part.0+0x148/0x260 fs/proc/task_mmu.c:768 > smap_gather_stats fs/proc/task_mmu.c:741 [inline] > show_smap+0xc6/0x440 fs/proc/task_mmu.c:822 > seq_read_iter+0xbb0/0x1240 fs/seq_file.c:272 > seq_read+0x3e0/0x5b0 fs/seq_file.c:162 > vfs_read+0x1b5/0x600 fs/read_write.c:479 > ksys_read+0x12d/0x250 fs/read_write.c:619 > do_syscall_x64 arch/x86/entry/common.c:50 [inline] > do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80 > entry_SYSCALL_64_after_hwframe+0x44/0xae > > The reproducer was trying to read /proc/$PID/smaps when calling > MADV_FREE at the mean time. MADV_FREE may split THPs if it is called > for partial THP. It may trigger the below race: > > CPU A CPU B > ----- ----- > smaps walk: MADV_FREE: > page_mapcount() > PageCompound() > split_huge_page() > page = compound_head(page) > PageDoubleMap(page) > > When calling PageDoubleMap() this page is not a tail page of THP anymore > so the BUG is triggered. > > This could be fixed by elevated refcount of the page before calling > mapcount, but that would prevent it from counting migration entries, and > it seems overkilling because the race just could happen when PMD is > split so all PTE entries of tail pages are actually migration entries, > and smaps_account() does treat migration entries as mapcount == 1 as > Kirill pointed out. > > Add a new parameter for smaps_account() to tell this entry is migration > entry then skip calling page_mapcount(). Don't skip getting mapcount > for device private entries since they do track references with mapcount. > > Pagemap also has the similar issue although it was not reported. Fixed > it as well. > > [shy828301@xxxxxxxxx: v4] > Link: https://lkml.kernel.org/r/20220203182641.824731-1-shy828301@xxxxxxxxx > [nathan@xxxxxxxxxx: avoid unused variable warning in pagemap_pmd_range()] > Link: https://lkml.kernel.org/r/20220207171049.1102239-1-nathan@xxxxxxxxxx > Link: https://lkml.kernel.org/r/20220120202805.3369-1-shy828301@xxxxxxxxx > Fixes: e9b61f19858a ("thp: reintroduce split_huge_page()") > Signed-off-by: Yang Shi <shy828301@xxxxxxxxx> > Signed-off-by: Nathan Chancellor <nathan@xxxxxxxxxx> > Reported-by: syzbot+1f52b3a18d5633fa7f82@xxxxxxxxxxxxxxxxxxxxxxxxx > Acked-by: David Hildenbrand <david@xxxxxxxxxx> > Cc: "Kirill A. Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx> > Cc: Jann Horn <jannh@xxxxxxxxxx> > Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx> > Cc: Alexey Dobriyan <adobriyan@xxxxxxxxx> > Cc: <stable@xxxxxxxxxxxxxxx> > Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> > Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> > --- > fs/proc/task_mmu.c | 40 +++++++++++++++++++++++++++++++--------- > 1 file changed, 31 insertions(+), 9 deletions(-) Both backports applied, thanks. greg k-h
