Re: regression -next -- scsi: sd: Rely on the driver core for asynchronous probing was Re: next-20190408..0418: Suspend/resume problems on Thinkpad X60

[Bart Van Assche]

On 4/25/19 12:33 AM, Pavel Machek wrote:
> On Wed 2019-04-24 13:56:01, Bart Van Assche wrote:
>> On Wed, 2019-04-24 at 22:51 +0200, Pavel Machek wrote:
>>> Unfortunately, that one does not revert cleanly on top of -next.
>>
>> Can you try the following:
>>
>> git revert d16ece577bf2cee7f94bab75a0d967bcb89dd2a7 &&
>>   git revert 21e6ba3f0e0257cce1a226c1f15e0a8ba4338ca3
>>
>> I will see whether I can come up with a better way to analyze what is
>> going on. I had not expected that these patches would cause any suspend/
>> resume problems.
> 
> Not even d16ece reverts:
> 
> pavel@duo:/data/l/linux-next-32$ git show  | head -3
> commit 76c938fcaa4b4a5d8f05fa907925d5043834964e
> Author: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>
> Date:   Tue Apr 23 20:24:59 2019 +1000
> pavel@duo:/data/l/linux-next-32$ git revert
> d16ece577bf2cee7f94bab75a0d967bcb89dd2a7
> error: could not revert d16ece5... scsi: sd: Inline sd_probe_part2()
> hint: after resolving the conflicts, mark the corrected paths
> hint: with 'git add <paths>' or 'git rm <paths>'
> hint: and commit the result with 'git commit'

There has been a non-trivial merge between the block and scsi trees in
linux-next. That's probably what prevents these patches to revert
cleanly. How about performing the following tests:
* Build, boot and test Martin's latest for-5.2 branch
(git://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git; branch
5.2/scsi-queue).
* If suspend/resume does not work reliably with that branch, revert the
two patches above, rebuild, reboot and retest.

Thanks,

Bart.