2023-07-27 10:40:40

by Damien Le Moal

[permalink] [raw]
Subject: Re: Scsi_bus_resume+0x0/0x90 returns -5 when resuming from s3 sleep

On 7/27/23 19:06, TW wrote:
> I retried on 6.5 rc3 without the Nvidia drivers and still received the same
> error and going to try for the patch next but got a malformed patch error on
> line 6 for the first patch for libata-scsi.c. The other two seem to go through
> just fine however.

Which other two patches are you talking about ?

> Also the bugzilla link is similar to what I have but the disk doesn't
> disappear, comes back but just takes awhile to come back out of sleep mode.

The switch to async resume has revealed many issues with the way ata devices
are resumed with regard to their scsi representation. The issues manifest in
different form. The drive "gone" problem was fixed recently. The start-stop
command seem to cause most of the time a delay in resume that several users
noticed. In your case though, it is an outright failure.

Hence the *single* patch I asked you to test (another user with a delay issue
is testing the same as well). Not sure what other patches you are talking about.

--
Damien Le Moal
Western Digital Research