Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1415935 > unrolled thread

Re: x86: bad pte in pageattr_test

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: x86: bad pte in pageattr_test Dmitry Vyukov <dvyukov@google.com> - 2016-06-07 11:40 +0200
    Re: x86: bad pte in pageattr_test Thomas Gleixner <tglx@linutronix.de> - 2016-06-09 23:40 +0200
      Re: x86: bad pte in pageattr_test Dmitry Vyukov <dvyukov@google.com> - 2016-06-10 12:20 +0200
        Re: x86: bad pte in pageattr_test Thomas Gleixner <tglx@linutronix.de> - 2016-06-10 15:00 +0200
          Re: x86: bad pte in pageattr_test Dmitry Vyukov <dvyukov@google.com> - 2016-06-10 15:10 +0200

#1415935 — Re: x86: bad pte in pageattr_test

On Tue, Jun 7, 2016 at 11:34 AM, Dmitry Vyukov <dvyukov@google.com> wrote:
> On Mon, Apr 11, 2016 at 10:28 AM, Dmitry Vyukov <dvyukov@google.com> wrote:
>> Hello,
>>
>> I've got the following WARNING while running syzkaller fuzzer:
>>
>> CPA ffff880054118000: bad pte after revert 8000000054118363
>> ------------[ cut here ]------------
>> WARNING: CPU: 2 PID: 1503 at arch/x86/mm/pageattr-test.c:226
>> pageattr_test+0xa6c/0xd10
>> NOT PASSED. Please report.
>> Modules linked in:
>> CPU: 2 PID: 1503 Comm: pageattr-test Not tainted 4.6.0-rc2+ #346
>> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>>  ffffffff87eb25c0 ffff88003b627a70 ffffffff82c8b17f ffffffff81490b58
>>  fffffbfff0fd64b8 ffff88003b627ae8 0000000000000000 ffffffff86a77e00
>>  ffffffff8129487c 0000000000000009 ffff88003b627ab8 ffffffff8136639f
>> Call Trace:
>>  [<     inline     >] __dump_stack lib/dump_stack.c:15
>>  [<ffffffff82c8b17f>] dump_stack+0x12e/0x18f lib/dump_stack.c:51
>>  [<ffffffff8136639f>] __warn+0x19f/0x1e0 kernel/panic.c:512
>>  [<ffffffff8136648c>] warn_slowpath_fmt+0xac/0xd0 kernel/panic.c:527
>>  [<ffffffff8129487c>] pageattr_test+0xa6c/0xd10 arch/x86/mm/pageattr-test.c:226
>>  [<ffffffff81294b3b>] do_pageattr_test+0x1b/0x60 arch/x86/mm/pageattr-test.c:240
>>  [<ffffffff813cde7f>] kthread+0x23f/0x2d0 drivers/block/aoe/aoecmd.c:1303
>>  [<ffffffff867b6252>] ret_from_fork+0x22/0x50 arch/x86/entry/entry_64.S:392
>> ---[ end trace e669a8d69b836be8 ]---
>>
>> Unfortunately it is not reproducible, tried bumping checking frequency to 1.
>>
>> On commit 541d8f4d59d79f5d37c8c726f723d42ff307db57 (Apr 5).
>>
>> For the repcord, full syzkaller log:
>> https://gist.githubusercontent.com/dvyukov/323ff7275c5ac38156cb40caeacac057/raw/0836f8dd81024e441f81caebdcb73ca1221aef97/gistfile1.txt
>
>
> Ping.
>
> Hit it again on af8c34ce6ae32addda3788d54a7e340cad22516b (4.7-rc2)

Should we delete this test if it is not important?


> CPA ffff880059990000: bad pte 8000000059990060
> ------------[ cut here ]------------
> WARNING: CPU: 1 PID: 1602 at arch/x86/mm/pageattr-test.c:226[<
> none      >] pageattr_test+0xa6f/0xd10 arch/x86/mm/pageattr-test.c:226
> NOT PASSED. Please report.
> Modules linked in:
> CPU: 1 PID: 1602 Comm: pageattr-test Not tainted 4.7.0-rc2+ #2
> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>  ffffffff880b59e0 ffff88003a41fa70 ffffffff82cc5e2f ffffffff81495ed8
>  fffffbfff1016b3c ffff88003a41fae8 0000000000000000 ffffffff86c7a060
>  ffffffff8129ac8f 0000000000000009 ffff88003a41fab8 ffffffff8136d23f
> Call Trace:
>  [<     inline     >] __dump_stack lib/dump_stack.c:15
>  [<ffffffff82cc5e2f>] dump_stack+0x12e/0x18f lib/dump_stack.c:51
>  [<ffffffff8136d23f>] __warn+0x19f/0x1e0 kernel/panic.c:516
>  [<ffffffff8136d32c>] warn_slowpath_fmt+0xac/0xd0 kernel/panic.c:531
>  [<ffffffff8129ac8f>] pageattr_test+0xa6f/0xd10 arch/x86/mm/pageattr-test.c:226
>  [<ffffffff8129af4b>] do_pageattr_test+0x1b/0x60 arch/x86/mm/pageattr-test.c:240
>  [<ffffffff813d58df>] kthread+0x23f/0x2d0 drivers/block/aoe/aoecmd.c:1303
>  [<ffffffff86a8da0f>] ret_from_fork+0x1f/0x40 arch/x86/entry/entry_64.S:389
> ---[ end trace 1e5de26a2555fc00 ]---

[toc] | [next] | [standalone]

#1418766

On Tue, 7 Jun 2016, Dmitry Vyukov wrote:
> >> I've got the following WARNING while running syzkaller fuzzer:
> >>
> >> CPA ffff880054118000: bad pte after revert 8000000054118363
> 
> > CPA ffff880059990000: bad pte 8000000059990060

In both cases the PTE bit which the test modifies is in the wrong state.

> Should we delete this test if it is not important?

No. There is something badly wrong.

PAGE_BIT_CPA_TEST is the same as PAGE_BIT_SPECIAL. And the latter is used by
the mm code to mark user space mappings. The test code only modifies the
direct mapping, i.e. the kernel side one.

So something sets PAGE_BIT_SPECIAL on a kernel PTE. And that's definitely a
bug.

These are the last entries from your syzkaller log file of the first incident:

r0 = perf_event_open(&(0x7f000000f000-0x78)={0x2, 0x78, 0x11, 0x7, 0xd537, 0x6, 0x0, 0xc1, 0xffff, 0x5, 0x0, 0x40, 0x4, 0x9, 0x5369, 0x8, 0x7, 0x8508, 0x3, 0x80, 0x0}, 0x0, 0xffffffff, 0xffffffffffffffff, 0x0)
mmap(&(0x7f0000cbb000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
r1 = syz_open_dev$mouse(&(0x7f0000cbb000)="2f6465762f696e7075742f6d6f7573652300", 0x100, 0xa00)
mmap(&(0x7f0000cbc000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
setsockopt$BT_SNDMTU(r1, 0x112, 0xc, &(0x7f0000cbc000)=0x5, 0x2)
mmap(&(0x7f0000cbb000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
ioctl$EVIOCGEFFECTS(r1, 0x80044584, &(0x7f0000cbc000-0x942)=nil)
r2 = fcntl$dupfd(r0, 0x406, r0)
mmap(&(0x7f0000cbc000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
mmap(&(0x7f00002bf000)=nil, (0x1000), 0x3, 0x8010, 0xffffffffffffffff, 0x0)
mmap(&(0x7f0000000000)=nil, (0x0), 0x3, 0x32, 0xffffffffffffffff, 0x0)
pwritev(r2, &(0x7f00007e9000)=[{&(0x7f0000cbc000)=....

Do you have log of the second one available as well?

CC'ing mm and perf folks.

Thanks,

	tglx

[toc] | [prev] | [next] | [standalone]

#1419199

On Thu, Jun 9, 2016 at 11:34 PM, Thomas Gleixner <tglx@linutronix.de> wrote:
> On Tue, 7 Jun 2016, Dmitry Vyukov wrote:
>> >> I've got the following WARNING while running syzkaller fuzzer:
>> >>
>> >> CPA ffff880054118000: bad pte after revert 8000000054118363
>>
>> > CPA ffff880059990000: bad pte 8000000059990060
>
> In both cases the PTE bit which the test modifies is in the wrong state.
>
>> Should we delete this test if it is not important?
>
> No. There is something badly wrong.
>
> PAGE_BIT_CPA_TEST is the same as PAGE_BIT_SPECIAL. And the latter is used by
> the mm code to mark user space mappings. The test code only modifies the
> direct mapping, i.e. the kernel side one.
>
> So something sets PAGE_BIT_SPECIAL on a kernel PTE. And that's definitely a
> bug.
>
> These are the last entries from your syzkaller log file of the first incident:
>
> r0 = perf_event_open(&(0x7f000000f000-0x78)={0x2, 0x78, 0x11, 0x7, 0xd537, 0x6, 0x0, 0xc1, 0xffff, 0x5, 0x0, 0x40, 0x4, 0x9, 0x5369, 0x8, 0x7, 0x8508, 0x3, 0x80, 0x0}, 0x0, 0xffffffff, 0xffffffffffffffff, 0x0)
> mmap(&(0x7f0000cbb000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
> r1 = syz_open_dev$mouse(&(0x7f0000cbb000)="2f6465762f696e7075742f6d6f7573652300", 0x100, 0xa00)
> mmap(&(0x7f0000cbc000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
> setsockopt$BT_SNDMTU(r1, 0x112, 0xc, &(0x7f0000cbc000)=0x5, 0x2)
> mmap(&(0x7f0000cbb000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
> ioctl$EVIOCGEFFECTS(r1, 0x80044584, &(0x7f0000cbc000-0x942)=nil)
> r2 = fcntl$dupfd(r0, 0x406, r0)
> mmap(&(0x7f0000cbc000)=nil, (0x1000), 0x3, 0x32, 0xffffffffffffffff, 0x0)
> mmap(&(0x7f00002bf000)=nil, (0x1000), 0x3, 0x8010, 0xffffffffffffffff, 0x0)
> mmap(&(0x7f0000000000)=nil, (0x0), 0x3, 0x32, 0xffffffffffffffff, 0x0)
> pwritev(r2, &(0x7f00007e9000)=[{&(0x7f0000cbc000)=....
>
> Do you have log of the second one available as well?
>
> CC'ing mm and perf folks.


Here is the second log:
https://gist.githubusercontent.com/dvyukov/dd7970a5daaa7a30f6d37fa5592b56de/raw/f29182024538e604c95d989f7b398816c3c595dc/gistfile1.txt

I've hit only twice. The first time I tried hard to reproduce it, with
no success. So unfortunately that's all we have.

Re logs: my setup executes up to 16 programs in parallel. So for
normal BUGs any of the preceding 16 programs can be guilty. But since
this check is asynchronous, it can be just any preceding program in
the log.

I would expect that it is triggered by some rarely-executing poorly
tested code. Maybe mmap of some device?

[toc] | [prev] | [next] | [standalone]

#1419364

On Fri, 10 Jun 2016, Dmitry Vyukov wrote:
> Here is the second log:
> https://gist.githubusercontent.com/dvyukov/dd7970a5daaa7a30f6d37fa5592b56de/raw/f29182024538e604c95d989f7b398816c3c595dc/gistfile1.txt
> 
> I've hit only twice. The first time I tried hard to reproduce it, with
> no success. So unfortunately that's all we have.
> 
> Re logs: my setup executes up to 16 programs in parallel. So for
> normal BUGs any of the preceding 16 programs can be guilty. But since
> this check is asynchronous, it can be just any preceding program in
> the log.

Ok.
 
> I would expect that it is triggered by some rarely-executing poorly
> tested code. Maybe mmap of some device?

That's the mmap(dev) list which is common between the two log files:

vcsn
ircomm
rfkill
userio
dspn
mice
midi
sndpcmc
hidraw0
vga_arbiter
lightnvm
sr

Dunno, if that's the right direction, but exposing these a bit more might be
worth to try.

Thanks,

	tglx

[toc] | [prev] | [next] | [standalone]

#1419387

On Fri, Jun 10, 2016 at 2:54 PM, Thomas Gleixner <tglx@linutronix.de> wrote:
> On Fri, 10 Jun 2016, Dmitry Vyukov wrote:
>> Here is the second log:
>> https://gist.githubusercontent.com/dvyukov/dd7970a5daaa7a30f6d37fa5592b56de/raw/f29182024538e604c95d989f7b398816c3c595dc/gistfile1.txt
>>
>> I've hit only twice. The first time I tried hard to reproduce it, with
>> no success. So unfortunately that's all we have.
>>
>> Re logs: my setup executes up to 16 programs in parallel. So for
>> normal BUGs any of the preceding 16 programs can be guilty. But since
>> this check is asynchronous, it can be just any preceding program in
>> the log.
>
> Ok.
>
>> I would expect that it is triggered by some rarely-executing poorly
>> tested code. Maybe mmap of some device?
>
> That's the mmap(dev) list which is common between the two log files:
>
> vcsn
> ircomm
> rfkill
> userio
> dspn
> mice
> midi
> sndpcmc
> hidraw0
> vga_arbiter
> lightnvm
> sr
>
> Dunno, if that's the right direction, but exposing these a bit more might be
> worth to try.


I am now running both of these logs for several hours (2.5M
executions). No failures so far...

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web