linux-kernel - Re: [RFC] fuse: do not generate interrupt requests for fatal signals

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

Date: Thu, 20 Jun 2024 23:40:41 +0200
From: Bernd Schubert <bernd.schubert@...tmail.fm>
To: Haifeng Xu <haifeng.xu@...pee.com>, Christian Brauner
 <brauner@...nel.org>, Miklos Szeredi <miklos@...redi.hu>
Cc: linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [RFC] fuse: do not generate interrupt requests for fatal signals



On 6/20/24 08:43, Haifeng Xu wrote:
> 
> 
> On 2024/6/17 15:25, Christian Brauner wrote:
>> On Fri, Jun 14, 2024 at 12:01:39PM GMT, Miklos Szeredi wrote:
>>> On Thu, 13 Jun 2024 at 12:44, Haifeng Xu <haifeng.xu@...pee.com> wrote:
>>>
>>>> So why the client doesn't get woken up?
>>>
>>> Need to find out what the server (lxcfs) is doing.  Can you do a
>>> strace of lxcfs to see the communication on the fuse device?
>>
>> Fwiw, I'm one of the orignal authors and maintainers of LXCFS so if you
>> have specific questions, I may be able to help.
> 
> Thanks. All server threads of lcxfs wokrs fine now.
> 
> So can we add another interface to abort those dead request?
> If the client thread got killed and wait for relpy, but the fuse sever didn't 
> send reply for some unknown reason，we can use this interface to wakeup the client thread.

Isn't that a manual workaround? I.e. an admin or a script needs to trigger it?

There is a discussion in this thread to add request timeouts
https://lore.kernel.org/linux-kernel/20240605153552.GB21567@localhost.localdomain/T/
I guess for interrupted requests that would be definitely a case where timeouts could be
applied?


Thanks,
Bernd