Re: [syzbot] WARNING in split_huge_page_to_list (2)

[HORIGUCHI NAOYA(堀口 直也)]

On Tue, Aug 23, 2022 at 06:36:24PM -0700, Yang Shi wrote:
> On Tue, Aug 23, 2022 at 3:08 PM syzbot
> <syzbot+07a218429c8d19b1fb25@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
> >
> > Hello,
> >
> > syzbot found the following issue on:
> >
> > HEAD commit:    680fb5b009e8 Merge tag 'arm64-upstream' into for-kernelci
> > git tree:       git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git for-kernelci
> > console output: https://syzkaller.appspot.com/x/log.txt?x=10ef7dc3080000
> > kernel config:  https://syzkaller.appspot.com/x/.config?x=4557ad2600fc45f4
> > dashboard link: https://syzkaller.appspot.com/bug?extid=07a218429c8d19b1fb25
> > compiler:       Debian clang version 13.0.1-++20220126092033+75e33f71c2da-1~exp1~20220126212112.63, GNU ld (GNU Binutils for Debian) 2.35.2
> > userspace arch: arm64
> > syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=17993aa5080000
> > C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=16521a3d080000
> >
> > IMPORTANT: if you fix the issue, please add the following tag to the commit:
> > Reported-by: syzbot+07a218429c8d19b1fb25@xxxxxxxxxxxxxxxxxxxxxxxxx
> >
> > raw: 05ffc00000010801 0000000000000000 dead000000000122 0000000000000000
> > raw: 0000000000000000 0000000000000000 00000002ffffffff 0000000000000000
> > page dumped because: VM_WARN_ON_ONCE_PAGE(is_hzp)
> > ------------[ cut here ]------------
> > WARNING: CPU: 1 PID: 3025 at mm/huge_memory.c:2555 split_huge_page_to_list+0x310/0xe2c mm/huge_memory.c:2555
> > Modules linked in:
> > CPU: 1 PID: 3025 Comm: syz-executor201 Not tainted 5.19.0-rc8-syzkaller-01618-g680fb5b009e8 #0
> > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/20/2022
> > pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
> > pc : split_huge_page_to_list+0x310/0xe2c mm/huge_memory.c:2555
> > lr : split_huge_page_to_list+0x308/0xe2c mm/huge_memory.c:2555
> > sp : ffff80001321bba0
> > x29: ffff80001321bbf0 x28: 0000000000400000 x27: 0000000000000001
> > x26: ffff80000ceda000 x25: 05ffc00000000000 x24: ffff80000d8c1000
> > x23: fffffc0003300000 x22: ffff0001ff012d40 x21: fffffc0003304400
> > x20: 0000000000000000 x19: fffffc0003300000 x18: 00000000000000c0
> > x17: ffff80000dcfd698 x16: ffff80000db3b658 x15: ffff0000c6708000
> > x14: 0000000000000000 x13: 00000000ffffffff x12: ffff0000c6708000
> > x11: ff808000081bf324 x10: 0000000000000000 x9 : 461f0eca467c0e00
> > x8 : 0000000000000001 x7 : ffff8000081937c0 x6 : 0000000000000000
> > x5 : 0000000000000080 x4 : 0000000000000001 x3 : 0000000000000000
> > x2 : 0000000000000000 x1 : 0000000100000000 x0 : 0000000000000031
> > Call trace:
> >  split_huge_page_to_list+0x310/0xe2c mm/huge_memory.c:2555
> >  split_huge_page include/linux/huge_mm.h:206 [inline]
> >  try_to_split_thp_page mm/memory-failure.c:1488 [inline]
> >  memory_failure+0x510/0x14cc mm/memory-failure.c:1870
> >  madvise_inject_error+0x260/0x334 mm/madvise.c:1115
> 
> MADV_HWPOISON is called on a huge zero page. The warning was raised
> for this case intentionally (it was a BUG before), see commit
> 478d134e9506c7e9bfe2830ed03dd85e97966313 ("mm/huge_memory: do not
> overkill when splitting huge_zero_page").

Sorry for my delayed response, I just now posted a patch to avoid this warning.
https://lore.kernel.org/linux-mm/20230406082004.2185420-1-naoya.horiguchi@xxxxxxxxx/T/#u

Thanks,
Naoya Horiguchi