Bug 15728

Summary: 2.6.34-rc3 doesn't suspend to disk (-rc2 did)
Product: Power Management Reporter: Maciej Rutecki (maciej.rutecki)
Component: Hibernation/SuspendAssignee: power-management_other
Status: CLOSED CODE_FIX    
Severity: normal CC: jirislaby, maciej.rutecki, nigel, rjw
Priority: P1    
Hardware: All   
OS: Linux   
Kernel Version: 2.6.34-rc3 Subsystem:
Regression: Yes Bisected commit-id:
Bug Depends on:    
Bug Blocks: 7216, 15310    
Attachments: patch for the issue

Description Maciej Rutecki 2010-04-08 18:42:41 UTC 
Subject    : 2.6.34-rc3 doesn't suspend to disk (-rc2 did)
Submitter  : Jiri Kosina <jkosina@suse.cz>
Date       : 2010-04-02 0:19
Message-ID : alpine.LNX.2.00.1004020215100.8345@pobox.suse.cz
References : http://marc.info/?l=linux-kernel&m=127016758131762&w=2

This entry is being used for tracking a regression from 2.6.33.  Please don't
close it until the problem is fixed in the mainline.
Comment 1 Jiri Slaby 2010-04-08 20:48:06 UTC 
I hit that with opensuse factory today. echo disk > /sys/power/state works fine there. The kernel is -rc2-git3 (plus other patches of course). I'll look into that later if noone beats me to it.
Comment 2 Jiri Slaby 2010-04-08 21:42:43 UTC 
Created attachment 25923 [details]
patch for the issue

I'm not sure whether the second paragraph in the changelog holds. For me it works.
Comment 3 Rafael J. Wysocki 2010-04-08 22:01:20 UTC 
OK, I assume you have tested it both with and without CONFIG_DEBUG_BLOCK_EXT_DEVT.

I guess it should be pushed to Linus soon.

I wonder which change broke it, though.

Handled-By : Jiri Slaby <jslaby@suse.cz>
Patch : https://bugzilla.kernel.org/attachment.cgi?id=25923
Comment 4 Jiri Slaby 2010-04-09 07:36:09 UTC 
(In reply to comment #3)
> OK, I assume you have tested it both with and without
> CONFIG_DEBUG_BLOCK_EXT_DEVT.

Yes.

> I wonder which change broke it, though.

I will bisect it later and make a note here.
Comment 5 Jiri Slaby 2010-04-10 08:40:04 UTC 
(In reply to comment #4)
> (In reply to comment #3)
> > I wonder which change broke it, though.
> 
> I will bisect it later and make a note here.

I'm confused. I can reproduce it in 2.6.33 with DEBUG_BLOCK_EXT_DEVT set too. But suse with 33 and that option set worked OK. Waiting for Jiri's input.