Re: general protection fault in transparent_hugepage_enabled

From: Dmitry Vyukov
Date: Wed Jan 02 2019 - 05:03:33 EST


On Wed, Jan 2, 2019 at 10:47 AM Vlastimil Babka <vbabka@xxxxxxx> wrote:
>
> On 1/2/19 10:42 AM, Dmitry Vyukov wrote:
> > On Wed, Jan 2, 2019 at 8:42 AM Vlastimil Babka <vbabka@xxxxxxx> wrote:
> >>
> >> On 12/24/18 4:48 PM, syzbot wrote:
> >>> Hello,
> >>>
> >>> syzbot found the following crash on:
> >>>
> >>> HEAD commit: 6a1d293238c1 Add linux-next specific files for 20181224
> >>> git tree: linux-next
> >>> console output: https://syzkaller.appspot.com/x/log.txt?x=149a2add400000
> >>> kernel config: https://syzkaller.appspot.com/x/.config?x=c190b602a5d2d731
> >>> dashboard link: https://syzkaller.appspot.com/bug?extid=a5fea9200aefd1cf4818
> >>> compiler: gcc (GCC) 8.0.1 20180413 (experimental)
> >>> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=1798bfb7400000
> >>> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=17f4dc57400000
> >>>
> >>> IMPORTANT: if you fix the bug, please add the following tag to the commit:
> >>> Reported-by: syzbot+a5fea9200aefd1cf4818@xxxxxxxxxxxxxxxxxxxxxxxxx
> >>>
> >>> sshd (6016) used greatest stack depth: 15720 bytes left
> >>> kasan: CONFIG_KASAN_INLINE enabled
> >>> kasan: GPF could be caused by NULL-ptr deref or user memory access
> >>> general protection fault: 0000 [#1] PREEMPT SMP KASAN
> >>> CPU: 1 PID: 6032 Comm: syz-executor045 Not tainted 4.20.0-rc7-next-20181224
> >>> #187
> >>> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
> >>> Google 01/01/2011
> >>> RIP: 0010:transparent_hugepage_enabled+0x8c/0x5e0 mm/huge_memory.c:69
> >>
> >> FTR, it was most certainly the same thing as here:
> >> https://marc.info/?l=linux-mm&m=154563777207477&w=2
> >
> > Thanks for the update. I see the fix is still not in linux-next, which has:
> >
> > +bool transparent_hugepage_enabled(struct vm_area_struct *vma)
> > +{
> > + if (vma_is_anonymous(vma))
> > + return __transparent_hugepage_enabled(vma);
> > + if (vma_is_shmem(vma) && shmem_huge_enabled(vma))
> > + return __transparent_hugepage_enabled(vma);
> > +
> > + return false;
> > +}
> >
> > Let's wait until the patch is updated and then tell syzbot that "mm,
> > thp, proc: report THP eligibility for each vma" fixes this.
>
> Actually the fix was folded into the patch that caused the bug, and was
> already sent to and merged by Linus, commit
> 7635d9cbe8327e131a1d3d8517dc186c2796ce2e


But the email thread you referenced says that we need:

@@ -66,6 +66,8 @@ bool transparent_hugepage_enabled(struct vm_area_struct *vma)
{
if (vma_is_anonymous(vma))
return __transparent_hugepage_enabled(vma);
+ if (!vma->vm_file || !vma->vm_file->f_mapping)
+ return false;
if (shmem_mapping(vma->vm_file->f_mapping) && shmem_huge_enabled(vma))
return __transparent_hugepage_enabled(vma);

and 7635d9cbe8327e131a1d3d8517dc186c2796ce2e contains:

+bool transparent_hugepage_enabled(struct vm_area_struct *vma)
+{
+ if (vma_is_anonymous(vma))
+ return __transparent_hugepage_enabled(vma);
+ if (vma_is_shmem(vma) && shmem_huge_enabled(vma))
+ return __transparent_hugepage_enabled(vma);
+
+ return false;
+}

What am I missing?



> >>> Code: 80 3c 02 00 0f 85 ae 04 00 00 4c 8b a3 a0 00 00 00 48 b8 00 00 00 00
> >>> 00 fc ff df 49 8d bc 24 b8 01 00 00 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f
> >>> 85 91 04 00 00 49 8b bc 24 b8 01 00 00 e8 2d 70 e6
> >>> RSP: 0018:ffff8881c2237138 EFLAGS: 00010202
> >>> RAX: dffffc0000000000 RBX: ffff8881c2bdbc60 RCX: 0000000000000000
> >>> RDX: 0000000000000037 RSI: ffffffff81c8fa1a RDI: 00000000000001b8
> >>> RBP: ffff8881c2237160 R08: ffffed10383b25ed R09: ffffed10383b25ec
> >>> R10: ffffed10383b25ec R11: ffff8881c1d92f63 R12: 0000000000000000
> >>> R13: ffff8881c2bdbd00 R14: dffffc0000000000 R15: 0000000000000f5e
> >>> FS: 0000000001a48880(0000) GS:ffff8881dad00000(0000) knlGS:0000000000000000
> >>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> >>> CR2: 0000000020b58000 CR3: 00000001c2210000 CR4: 00000000001406e0
> >>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> >>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> >>> Call Trace:
> >>> show_smap+0x167/0x580 fs/proc/task_mmu.c:805
> >>> traverse+0x344/0x7b0 fs/seq_file.c:113
> >>> seq_read+0xc76/0x1150 fs/seq_file.c:188
> >>> do_loop_readv_writev fs/read_write.c:700 [inline]
> >>> do_iter_read+0x4bc/0x670 fs/read_write.c:924
> >>> vfs_readv+0x175/0x1c0 fs/read_write.c:986
> >>> kernel_readv fs/splice.c:362 [inline]
> >>> default_file_splice_read+0x539/0xb20 fs/splice.c:417
> >>> do_splice_to+0x12e/0x190 fs/splice.c:880
> >>> splice_direct_to_actor+0x31c/0x9d0 fs/splice.c:957
> >>> do_splice_direct+0x2d4/0x420 fs/splice.c:1066
> >>> do_sendfile+0x62a/0xe50 fs/read_write.c:1439
> >>> __do_sys_sendfile64 fs/read_write.c:1494 [inline]
> >>> __se_sys_sendfile64 fs/read_write.c:1486 [inline]
> >>> __x64_sys_sendfile64+0x15d/0x250 fs/read_write.c:1486
> >>> do_syscall_64+0x1b9/0x820 arch/x86/entry/common.c:290
> >>> entry_SYSCALL_64_after_hwframe+0x49/0xbe
> >>> RIP: 0033:0x440089
> >>> Code: 18 89 d0 c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 00 48 89 f8 48 89 f7
> >>> 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff
> >>> ff 0f 83 5b 14 fc ff c3 66 2e 0f 1f 84 00 00 00 00
> >>> RSP: 002b:00007fff3d710a18 EFLAGS: 00000213 ORIG_RAX: 0000000000000028
> >>> RAX: ffffffffffffffda RBX: 00007fff3d710a20 RCX: 0000000000440089
> >>> RDX: 0000000020b58000 RSI: 0000000000000003 RDI: 0000000000000003
> >>> RBP: 00000000006ca018 R08: 0000000000000010 R09: 65732f636f72702f
> >>> R10: 000000000000ffff R11: 0000000000000213 R12: 0000000000401970
> >>> R13: 0000000000401a00 R14: 0000000000000000 R15: 0000000000000000
> >>> Modules linked in:
> >>> ---[ end trace faf026efd8795e93 ]---
> >>> RIP: 0010:transparent_hugepage_enabled+0x8c/0x5e0 mm/huge_memory.c:69
> >>> Code: 80 3c 02 00 0f 85 ae 04 00 00 4c 8b a3 a0 00 00 00 48 b8 00 00 00 00
> >>> 00 fc ff df 49 8d bc 24 b8 01 00 00 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f
> >>> 85 91 04 00 00 49 8b bc 24 b8 01 00 00 e8 2d 70 e6
> >>> RSP: 0018:ffff8881c2237138 EFLAGS: 00010202
> >>> RAX: dffffc0000000000 RBX: ffff8881c2bdbc60 RCX: 0000000000000000
> >>> RDX: 0000000000000037 RSI: ffffffff81c8fa1a RDI: 00000000000001b8
> >>> RBP: ffff8881c2237160 R08: ffffed10383b25ed R09: ffffed10383b25ec
> >>> R10: ffffed10383b25ec R11: ffff8881c1d92f63 R12: 0000000000000000
> >>> R13: ffff8881c2bdbd00 R14: dffffc0000000000 R15: 0000000000000f5e
> >>> FS: 0000000001a48880(0000) GS:ffff8881dad00000(0000) knlGS:0000000000000000
> >>> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> >>> CR2: 0000000020b58000 CR3: 00000001c2210000 CR4: 00000000001406e0
> >>> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> >>> DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> >>>
> >>>
> >>> ---
> >>> This bug is generated by a bot. It may contain errors.
> >>> See https://goo.gl/tpsmEJ for more information about syzbot.
> >>> syzbot engineers can be reached at syzkaller@xxxxxxxxxxxxxxxxx
> >>>
> >>> syzbot will keep track of this bug report. See:
> >>> https://goo.gl/tpsmEJ#bug-status-tracking for how to communicate with
> >>> syzbot.
> >>> syzbot can test patches for this bug, for details see:
> >>> https://goo.gl/tpsmEJ#testing-patches
>