linux-kernel - Re: [PATCH 00/53] Let create*_workqueue() accept a printf-style format string

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]

Message-ID: <63917f94-d9ed-4d43-b0a8-50ae4afcf1ee@acm.org>
Date: Mon, 1 Jul 2024 13:13:57 -0700
From: Bart Van Assche <bvanassche@....org>
To: Tejun Heo <tj@...nel.org>
Cc: Lai Jiangshan <jiangshanlai@...il.com>,
 "Martin K . Petersen" <martin.petersen@...cle.com>,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH 00/53] Let create*_workqueue() accept a printf-style
 format string

On 7/1/24 12:39 PM, Tejun Heo wrote:
> On Sun, Jun 30, 2024 at 03:26:18PM -0700, Bart Van Assche wrote:
>> 1. Introduce the create*_workqueue2() macros that accept a printf-style
>>     format string and argument list.
> 
> I don't think we want to update create_workqueue() interface at all. Please
> convert them to use alloc_workqueue() instead. Once we convert all
> create_workqueue() users then we can just drop the old interface.
> 
> The challenge with converging to alloc_workqueue() interface is that the
> conversion needs to explicitly decide whether WQ_MEM_RECLAIM is needed or
> not but that shouldn't be *too* difficult.

The create*_workqueue() macros are used about 270 times in the Linux
kernel (the macro names also occur in some comments):

$ git grep -nHEw 
'create_workqueue|create_freezable_workqueue|create_singlethread_workqueue' 
| grep -Ev 'Documentation/|include/linux/workqueue' | wc -l
     271

That's too much for a single patch series. I can start with converting
the SCSI workqueue creation code.

Thanks,

Bart.