linux-kernel - RE: [PATCH 2/2] ufs: don't use the fair tag sharings

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: <DM6PR04MB6575F344EF2D962103888A56FC769@DM6PR04MB6575.namprd04.prod.outlook.com>
Date:   Tue, 9 May 2023 16:19:52 +0000
From:   Avri Altman <Avri.Altman@....com>
To:     Bart Van Assche <bvanassche@....org>,
        Ed Tsai <ed.tsai@...iatek.com>,
        "axboe@...nel.dk" <axboe@...nel.dk>
CC:     "linux-block@...r.kernel.org" <linux-block@...r.kernel.org>,
        "linux-scsi@...r.kernel.org" <linux-scsi@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "martin.petersen@...cle.com" <martin.petersen@...cle.com>,
        "stanley.chu@...iatek.com" <stanley.chu@...iatek.com>,
        "peter.wang@...iatek.com" <peter.wang@...iatek.com>,
        "chun-hung.wu@...iatek.com" <chun-hung.wu@...iatek.com>,
        "alice.chao@...iatek.com" <alice.chao@...iatek.com>,
        "powen.kao@...iatek.com" <powen.kao@...iatek.com>,
        "naomi.chu@...iatek.com" <naomi.chu@...iatek.com>,
        "wsd_upstream@...iatek.com" <wsd_upstream@...iatek.com>
Subject: RE: [PATCH 2/2] ufs: don't use the fair tag sharings

> On 5/9/23 01:03, Avri Altman wrote:
> > However, I think the decision of that should be of the platform owner,
> > And not in the core driver.
> 
> Hi Avri,
> 
> I don't see any use case in which performance of a UFS device would be
> improved
> by leaving QUEUE_FLAG_FAIR_TAG_SHARING enabled. Are you perhaps aware
> of such a
> use case?
Following your argument, then why fair allocation exists in the first place?

When running benchmarks I am hacking the scheduler's "fair" tag allocation as well.
That's why I acked this change.

Since this change may affect the IO profile as a whole,
I think the platform owners should have the flexibility not to use it,
Should they choose to.

Thanks,
Avri
> 
> Thanks,
> 
> Bart.