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: <3836423e-8f0f-1278-38e5-d35fbd7298e4@kernel.org>
Date:   Fri, 28 Jul 2023 13:09:06 +0900
From:   Damien Le Moal <dlemoal@...nel.org>
To:     TW <dalzot@...il.com>,
        Thorsten Leemhuis <regressions@...mhuis.info>
Cc:     regressions@...ts.linux.dev,
        Mario Limonciello <mario.limonciello@....com>,
        Bart Van Assche <bvanassche@....org>,
        LKML <linux-kernel@...r.kernel.org>, stable@...r.kernel.org
Subject: Re: Scsi_bus_resume+0x0/0x90 returns -5 when resuming from s3 sleep

On 7/28/23 11:49, TW wrote:
> Comparing to the 5.15 kernel which had almost no delay. They are HDDs 
> though, it was working flawlessly earlier but didn't want to top post 
> again. Tried a systemd suspend instead of from the Xfce4 logout menu and 
> everything works as intended. So I'd say that it's fixed now.

I think that asynchronous scsi suspend/resume using the PM infrastructure,
introduced in 5.16, is creating this additional delay. The reason is that now,
ata and scsi drivers do their resume operations from within the PM resume
process, resulting in some delays to reach "PM: suspend exit".

I am looking into improving this, but it may take some more time.

I will post officially the patch you tried and CC you. A "tested-by" tag from
you would be appreciated.

Thanks.

> 
> On 7/27/23 20:33, Damien Le Moal wrote:
>> "Slow coming back" -> Compared to which version of the kernel ? Do you 
>> have numbers ? If the devices are HDDs, resume will wait for these to 
>> spin up. That takes a while (about 10s normally).

-- 
Damien Le Moal
Western Digital Research

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ