2020-06-29 18:51:27

by Jörg Otte

Subject: segmentation faults with kernels >v5.8-rc2

I frequently get segmentation faults in newer kernel >v5.8-rc2 e.g.
Chrome_ChildIOT[1298]: segfault at 40d048 ip 000056111a41970d sp
00007fdced6931b0 error 6 in chrome[561115f64000+785b000]

Bisection gave me the following
first bad commit:

e9c15badbb7b20ccdbadf5da14e0a68fbad51015 is the first bad commit
commit e9c15badbb7b20ccdbadf5da14e0a68fbad51015
Author: Mel Gorman <[email protected]>
Date: Mon Jun 15 13:13:58 2020 +0100

fs: Do not check if there is a fsnotify watcher on pseudo inodes
...

I double checked by reverting this commit on top of v5.8-rc3 and the
segmentation faults are gone.


Jörg

2020-06-29 21:26:26

by Jan Kara

Subject: Re: segmentation faults with kernels >v5.8-rc2

Hello!

On Mon 29-06-20 15:58:50, J?rg Otte wrote:
> I frequently get segmentation faults in newer kernel >v5.8-rc2 e.g.
> Chrome_ChildIOT[1298]: segfault at 40d048 ip 000056111a41970d sp
> 00007fdced6931b0 error 6 in chrome[561115f64000+785b000]
>
> Bisection gave me the following
> first bad commit:
>
> e9c15badbb7b20ccdbadf5da14e0a68fbad51015 is the first bad commit
> commit e9c15badbb7b20ccdbadf5da14e0a68fbad51015
> Author: Mel Gorman <[email protected]>
> Date: Mon Jun 15 13:13:58 2020 +0100
>
> fs: Do not check if there is a fsnotify watcher on pseudo inodes
> ...
>
> I double checked by reverting this commit on top of v5.8-rc3 and the
> segmentation faults are gone.

We've already reverted that commit today... Thanks for report!

Honza
--
Jan Kara <[email protected]>
SUSE Labs, CR