linux-kernel - Re: INFO: task hung in cancel_delayed_work

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Wed, 18 Sep 2019 22:50:16 -0700
From:   Eric Biggers <ebiggers@...nel.org>
To:     Steffen Klassert <steffen.klassert@...unet.com>
Cc:     syzbot <syzbot+f39ab8494f6015e62360@...kaller.appspotmail.com>,
        ast@...nel.org, aviadye@...lanox.com, borisp@...lanox.com,
        bpf@...r.kernel.org, daniel@...earbox.net, davejwatson@...com,
        davem@...emloft.net, ilyal@...lanox.com,
        jakub.kicinski@...ronome.com, john.fastabend@...il.com,
        kafai@...com, linux-crypto@...r.kernel.org,
        linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
        songliubraving@...com, syzkaller-bugs@...glegroups.com, yhs@...com
Subject: Re: INFO: task hung in cancel_delayed_work_sync

On Thu, Sep 19, 2019 at 07:36:20AM +0200, Steffen Klassert wrote:
> On Wed, Sep 18, 2019 at 10:15:45PM -0700, Eric Biggers wrote:
> > 
> > Reproducer involves pcrypt, so probably the pcrypt deadlock again...
> > https://lkml.kernel.org/linux-crypto/20190817054743.GE8209@sol.localdomain/
> 
> I'll submit the patch I proposed here in case noone has a better idea
> how to fix this now:
> 
> https://lkml.kernel.org/linux-crypto/20190821063704.GM2879@gauss3.secunet.de/
> 
> The original patch is from you, I did some modifications to forbid
> pcrypt if an underlying algorithm needs a fallback.
> 
> May I leave your 'Signed off' on this patch, or just
> quote that the initial version is from you?
> 

Keeping my Signed-off-by is fine, but please leave a note about what you
changed, like:

Signed-off-by: Eric Biggers <ebiggers@...gle.com>
[SK: also require that the underlying algorithm doesn't need a fallback]
Signed-off-by: Steffen Klassert <steffen.klassert@...unet.com>

Also, a nit: in the commit message,

> Fix this by making pcrypt forbid instantiation if pcrypt appears in the
> underlying ->cra_driver_name and if an underlying algorithm needs a
> fallback.

... the word "and" should be "or".

- Eric