linux-kernel - Re: [PATCH v2 1/2] Revert "locking/rwsem: Remove reader optimistic spinning"

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:   Mon, 4 Sep 2023 15:56:56 -0400
From:   Waiman Long <longman@...hat.com>
To:     Peter Zijlstra <peterz@...radead.org>,
        Bongkyu Kim <bongkyu7.kim@...sung.com>
Cc:     mingo@...hat.com, will@...nel.org, boqun.feng@...il.com,
        linux-kernel@...r.kernel.org, gregkh@...uxfoundation.org
Subject: Re: [PATCH v2 1/2] Revert "locking/rwsem: Remove reader optimistic
 spinning"

On 9/4/23 11:10, Peter Zijlstra wrote:
> On Fri, Sep 01, 2023 at 10:07:03AM +0900, Bongkyu Kim wrote:
>> This reverts commit 617f3ef95177840c77f59c2aec1029d27d5547d6.
>>
>> In mobile environment, reader optimistic spinning is still useful
>> because there're not many readers. In my test result at android device,
>> it improves application startup time about 3.8%
>> App startup time is most important factor for android user expriences.
>> So, re-enable reader optimistic spinning by this commit. And,
>> the later patch will make it optional feature by cmdline.
> I'm not seeing any mention on how this interacts with all the rwsem work
> that has been done since that commit, like the handoff rework.
>
> Why is a straight revert a sane thing at this point?

I also agree that a revert is not the best way to reintroduce the 
feature. It should document the reason why reader optimistic spinning is 
not the default as discussed in commit 617f3ef9517 ("locking/rwsem: 
Remove reader optimistic spinning") and under what condition should 
reader optimistic spinning can be turned back on.

Besides, I now think we may not really need 2 separate nonspinnable 
bits. We can go with one that is set by writer timing out when spinning 
on reader.

Cheers,
Longman