Greeting,
FYI, we noticed the following commit (built with gcc-9):
commit: 2418c34937c42a30ef4bccd923ad664a89e1fbd4 ("iov_iter: optimize iov_iter_advance() for iovec and kvec")
https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git untested.iov_iter
in testcase: boot
on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 16G
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
+--------------------------+------------+------------+
| | c5f070c68e | 2418c34937 |
+--------------------------+------------+------------+
| boot_failures | 0 | 11 |
| Initiating_system_reboot | 0 | 11 |
+--------------------------+------------+------------+
If you fix the issue, kindly add following tag
Reported-by: kernel test robot <[email protected]>
[ 25.942407] prompt_ramdisk=0
[ 25.943148] vga=normal
[ 25.943836] watchdog_thresh=240
[ 26.029535] init: Console is alive
[ 26.030789] init: - watchdog -
[ 56.035061] softdog: Initiating system reboot
Kboot worker: lkp-worker13
Elapsed time: 60
kvm=(
as a contrast, this was not observed in parent, e.g.
[ 13.467468] prompt_ramdisk=0
[ 13.467787] vga=normal
[ 13.468059] watchdog_thresh=240
[ 13.519759] init: Console is alive
[ 13.520462] init: - watchdog -
[ 13.526430] kmodloader: loading kernel modules from /etc/modules-boot.d/*
[ 13.527899] kmodloader: done loading kernel modules from /etc/modules-boot.d/*
[ 13.529071] kmodloader (124) used greatest stack depth: 6708 bytes left
[ 13.539875] init: - preinit -
... <keep going on after then>
To reproduce:
# build kernel
cd linux
cp config-5.12.0-rc2-00003-g2418c34937c4 .config
make HOSTCC=gcc-9 CC=gcc-9 ARCH=i386 olddefconfig prepare modules_prepare bzImage
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp qemu -k <bzImage> job-script # job-script is attached in this email
---
0DAY/LKP+ Test Infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/[email protected] Intel Corporation
Thanks,
Oliver Sang
On Wed, Apr 28, 2021 at 10:37:47AM +0800, kernel test robot wrote:
>
>
> Greeting,
>
> FYI, we noticed the following commit (built with gcc-9):
>
> commit: 2418c34937c42a30ef4bccd923ad664a89e1fbd4 ("iov_iter: optimize iov_iter_advance() for iovec and kvec")
> https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git untested.iov_iter
>
>
> in testcase: boot
>
> on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 16G
>
> caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
>
>
> +--------------------------+------------+------------+
> | | c5f070c68e | 2418c34937 |
> +--------------------------+------------+------------+
> | boot_failures | 0 | 11 |
> | Initiating_system_reboot | 0 | 11 |
> +--------------------------+------------+------------+
>
>
> If you fix the issue, kindly add following tag
> Reported-by: kernel test robot <[email protected]>
Could you run it with soft_panic=1 in command line? Alternatively, some
information about how to reproduce that without running hell knows what
as root on host would be very welcome; I can't imagine a single reason
for needing root to run qemu, to be honest...
On Wed, Apr 28, 2021 at 10:37:47AM +0800, kernel test robot wrote:
>
>
> Greeting,
>
> FYI, we noticed the following commit (built with gcc-9):
>
> commit: 2418c34937c42a30ef4bccd923ad664a89e1fbd4 ("iov_iter: optimize iov_iter_advance() for iovec and kvec")
> https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git untested.iov_iter
>
>
> in testcase: boot
>
> on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 16G
>
> caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
Could you try https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git for-lkp
on the same test and see what (if anything) gets spewed into dmesg?
hi, Al Viro,
On Wed, Apr 28, 2021 at 04:32:44AM +0000, Al Viro wrote:
> On Wed, Apr 28, 2021 at 10:37:47AM +0800, kernel test robot wrote:
> >
> >
> > Greeting,
> >
> > FYI, we noticed the following commit (built with gcc-9):
> >
> > commit: 2418c34937c42a30ef4bccd923ad664a89e1fbd4 ("iov_iter: optimize iov_iter_advance() for iovec and kvec")
> > https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git untested.iov_iter
> >
> >
> > in testcase: boot
> >
> > on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 16G
> >
> > caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
>
> Could you try https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git for-lkp
> on the same test and see what (if anything) gets spewed into dmesg?
please check attached dmesg.xz for detail
[ 18.549991] ------------[ cut here ]------------
[ 18.550798] WARNING: CPU: 0 PID: 124 at lib/iov_iter.c:1117 iov_iter_advance.cold+0x13/0x7b
[ 18.552115] Modules linked in:
[ 18.552693] CPU: 0 PID: 124 Comm: kmodloader Not tainted 5.12.0-rc2-00004-g039686f7e8a0 #1
[ 18.554221] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
[ 18.555733] EIP: iov_iter_advance.cold+0x13/0x7b
[ 18.556616] Code: 00 00 b8 90 78 2f b2 31 c9 e8 cd 9e 79 ff 58 31 c0 5a e9 cc 4b a8 ff 6a 01 31 c9 ba 01 00 00 00 b8 c8 6d 2f b2 e8 b1 9e 79 ff <0f> 0
b 6a 01 31 c9 ba 01 00 00 00 b8 b0 6d 2f b2 31 db e8 9a 9e 79
[ 18.559751] EAX: 00000000 EBX: 00000000 ECX: 00000000 EDX: 00000001
[ 18.560818] ESI: 00000000 EDI: ddf89ee4 EBP: ddf89eac ESP: ddf89e5c
[ 18.561894] DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068 EFLAGS: 00010206
[ 18.563020] CR0: 80050033 CR2: a7f30844 CR3: 2deee000 CR4: 000406d0
[ 18.564011] Call Trace:
[ 18.564483] do_iter_read+0x138/0x1c0
[ 18.565159] vfs_readv+0x47/0x80
[ 18.565771] ? lock_acquire+0xa8/0x2e0
[ 18.566463] ? __fget_light+0x4d/0x90
[ 18.567152] do_readv+0x5f/0xe0
[ 18.567723] __ia32_sys_readv+0x12/0x20
[ 18.568451] __do_fast_syscall_32+0x74/0xb0
[ 18.569223] do_fast_syscall_32+0x28/0x60
[ 18.569977] do_SYSENTER_32+0x15/0x20
[ 18.570676] entry_SYSENTER_32+0x9f/0xf1
[ 18.571419] EIP: 0xa7f44549
[ 18.571962] Code: 03 74 c0 01 10 05 03 74 b8 01 10 06 03 74 b4 01 10 07 03 74 b0 01 10 08 03 74 d8 01 00 00 00 00 00 51 52 55 89 e5 0f 34 cd 80 <5d> 5
a 59 c3 90 90 90 90 8d b4 26 00 00 00 00 8d b4 26 00 00 00 00
[ 18.575391] EAX: ffffffda EBX: 00000004 ECX: af810710 EDX: 00000002
[ 18.576611] ESI: 00000001 EDI: 00003061 EBP: af810738 ESP: af8106e4
[ 18.577748] DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 007b EFLAGS: 00000282
[ 18.578970] irq event stamp: 2455
[ 18.579587] hardirqs last enabled at (2463): [<b10bd9b5>] console_unlock+0x455/0x610
[ 18.580951] hardirqs last disabled at (2470): [<b10bd9a5>] console_unlock+0x445/0x610
[ 18.582269] softirqs last enabled at (2322): [<b19d6aad>] __do_softirq+0x29d/0x434
[ 18.583483] softirqs last disabled at (2313): [<b1024220>] call_on_stack+0x40/0x50
[ 18.585165] ---[ end trace 6e5f78d431f11767 ]---
hi, Al Viro,
On Wed, Apr 28, 2021 at 03:18:18AM +0000, Al Viro wrote:
> On Wed, Apr 28, 2021 at 10:37:47AM +0800, kernel test robot wrote:
> >
> >
> > Greeting,
> >
> > FYI, we noticed the following commit (built with gcc-9):
> >
> > commit: 2418c34937c42a30ef4bccd923ad664a89e1fbd4 ("iov_iter: optimize iov_iter_advance() for iovec and kvec")
> > https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git untested.iov_iter
> >
> >
> > in testcase: boot
> >
> > on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 16G
> >
> > caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
> >
> >
> > +--------------------------+------------+------------+
> > | | c5f070c68e | 2418c34937 |
> > +--------------------------+------------+------------+
> > | boot_failures | 0 | 11 |
> > | Initiating_system_reboot | 0 | 11 |
> > +--------------------------+------------+------------+
> >
> >
> > If you fix the issue, kindly add following tag
> > Reported-by: kernel test robot <[email protected]>
>
> Could you run it with soft_panic=1 in command line? Alternatively, some
> information about how to reproduce that without running hell knows what
> as root on host would be very welcome; I can't imagine a single reason
> for needing root to run qemu, to be honest...
not sure if this soft_panic=1 test is still useful, but from test, it seems
not generate more useful information (attached one for example).
and if you still want this test, could you help check our command line and
maybe supply further information how to enable it properly?
On Fri, Apr 30, 2021 at 11:54:31PM +0800, Oliver Sang wrote:
> not sure if this soft_panic=1 test is still useful, but from test, it seems
> not generate more useful information (attached one for example).
>
> and if you still want this test, could you help check our command line and
> maybe supply further information how to enable it properly?
Should be fixed in the current branch.