Re: [PATCH 5.10.y] drm/cirrus: fix a NULL vs IS_ERR() checks

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Dec 28, 2021 at 10:19:30PM +0800, Shile Zhang wrote:
> 
> 
> On 2021/12/28 22:05, Greg Kroah-Hartman wrote:
> > On Tue, Dec 28, 2021 at 09:56:25PM +0800, Shile Zhang wrote:
> > > 
> > > 
> > > On 2021/12/28 21:51, Greg Kroah-Hartman wrote:
> > > > On Tue, Dec 28, 2021 at 09:25:56PM +0800, Shile Zhang wrote:
> > > > > The function drm_gem_shmem_vmap can returns error pointers as well,
> > > > > which could cause following kernel crash:
> > > > > 
> > > > > BUG: unable to handle page fault for address: fffffffffffffffc
> > > > > PGD 1426a12067 P4D 1426a12067 PUD 1426a14067 PMD 0
> > > > > Oops: 0000 [#1] SMP NOPTI
> > > > > CPU: 12 PID: 3598532 Comm: stress-ng Kdump: loaded Not tainted 5.10.50.x86_64 #1
> > > > > ...
> > > > > RIP: 0010:memcpy_toio+0x23/0x50
> > > > > Code: 00 00 00 00 0f 1f 00 0f 1f 44 00 00 48 85 d2 74 28 40 f6 c7 01 75 2b 48 83 fa 01 76 06 40 f6 c7 02 75 17 48 89 d1 48 c1 e9 02 <f3> a5 f6 c2 02 74 02 66 a5 f6 c2 01 74 01 a4 c3 66 a5 48 83 ea 02
> > > > > RSP: 0018:ffffafbf8a203c68 EFLAGS: 00010216
> > > > > RAX: 0000000000000000 RBX: fffffffffffffffc RCX: 0000000000000200
> > > > > RDX: 0000000000000800 RSI: fffffffffffffffc RDI: ffffafbf82000000
> > > > > RBP: ffffafbf82000000 R08: 0000000000000002 R09: 0000000000000000
> > > > > R10: 00000000000002b5 R11: 0000000000000000 R12: 0000000000000800
> > > > > R13: ffff8a6801099300 R14: 0000000000000001 R15: 0000000000000300
> > > > > FS:  00007f4a6bc5f740(0000) GS:ffff8a8641900000(0000) knlGS:0000000000000000
> > > > > CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > > > > CR2: fffffffffffffffc CR3: 00000016d3874001 CR4: 00000000003606e0
> > > > > Call Trace:
> > > > >    drm_fb_memcpy_dstclip+0x5e/0x80 [drm_kms_helper]
> > > > >    cirrus_fb_blit_rect.isra.0+0xb7/0xe0 [cirrus]
> > > > >    cirrus_pipe_update+0x9f/0xa8 [cirrus]
> > > > >    drm_atomic_helper_commit_planes+0xb8/0x220 [drm_kms_helper]
> > > > >    drm_atomic_helper_commit_tail+0x42/0x80 [drm_kms_helper]
> > > > >    commit_tail+0xce/0x130 [drm_kms_helper]
> > > > >    drm_atomic_helper_commit+0x113/0x140 [drm_kms_helper]
> > > > >    drm_client_modeset_commit_atomic+0x1c4/0x200 [drm]
> > > > >    drm_client_modeset_commit_locked+0x53/0x80 [drm]
> > > > >    drm_client_modeset_commit+0x24/0x40 [drm]
> > > > >    drm_fbdev_client_restore+0x48/0x85 [drm_kms_helper]
> > > > >    drm_client_dev_restore+0x64/0xb0 [drm]
> > > > >    drm_release+0xf2/0x110 [drm]
> > > > >    __fput+0x96/0x240
> > > > >    task_work_run+0x5c/0x90
> > > > >    exit_to_user_mode_loop+0xce/0xd0
> > > > >    exit_to_user_mode_prepare+0x6a/0x70
> > > > >    syscall_exit_to_user_mode+0x12/0x40
> > > > >    entry_SYSCALL_64_after_hwframe+0x44/0xa9
> > > > > RIP: 0033:0x7f4a6bd82c2b
> > > > > 
> > > > > Fixes: ab3e023b1b4c9 ("drm/cirrus: rewrite and modernize driver.")
> > > > > 
> > > > > CC: stable@xxxxxxxxxxxxxxx
> > > > > Reported-by: Wen Kang <kw01107137@xxxxxxxxxxxxxxx>
> > > > > Signed-off-by: Shile Zhang <shile.zhang@xxxxxxxxxxxxxxxxx>
> > > > > ---
> > > > >    drivers/gpu/drm/tiny/cirrus.c | 2 +-
> > > > >    1 file changed, 1 insertion(+), 1 deletion(-)
> > > > 
> > > > What is the git commit id of this patch in Linus's tree?
> > > 
> > > Sorry, I checked that this issue seems fixed by the improvement in following
> > > series:
> > > https://patchwork.freedesktop.org/series/82217/
> > 
> > I do not understand, that is a huge patch series.  What individual
> > commit in Linus's tree resolves this?
> 
> Sorry,
> 1. This crash only happened in 5.10.y tree now, which fixed in Linus's tree
> by refactoring in above huge series.

Which specific patch resolved the issue?

> 2. It's hard to get the individual commit to fix this issue from that
> series. So I try to send this simple fix help to fix only for 5.10.y, which
> is needless to Linus's tree.

'git bisect' should be able to help you out.

> 3. If this patch is not OK for stable tree, Could you please help to
> backport the correct fix from Linus's tree in next version of 5.10.y?

If you can provide the commit id of the fix, sure.

thanks,

greg k-h



[Index of Archives]     [Linux Kernel]     [Kernel Development Newbies]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite Hiking]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux