From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stefan Bader Subject: Re: [Bug #15096] Resume lock up -- bisected, commit 3a1151e3f124fd1a2c54b8153f510f1a7c715369 Date: Mon, 08 Feb 2010 20:55:33 +0100 Message-ID: <4B706C35.2040100@canonical.com> References: <2uzMDuY-7_O.A.M1G.7X1bLB@chimera> <4B7004F4.3030804@canonical.com> <201002082022.43515.rjw@sisk.pl> <4B7066BA.7040806@canonical.com> Mime-Version: 1.0 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: Sender: kernel-testers-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: Content-Type: text/plain; charset="utf-8" To: =?UTF-8?B?UmFmYcWCIE1pxYJlY2tp?= Cc: "Rafael J. Wysocki" , Linux Kernel Mailing List , Kernel Testers List , Maciej Rutecki , Len Brown , Zhang Rui Rafa=C5=82 Mi=C5=82ecki wrote: > 2010/2/8 Stefan Bader : >> Rafael J. Wysocki wrote: >>> On Monday 08 February 2010, Rafa=C5=82 Mi=C5=82ecki wrote: >>>> 2010/2/8 Stefan Bader : >>>>> Rafael J. Wysocki wrote: >>>>>> On Monday 08 February 2010, Rafa=C5=82 Mi=C5=82ecki wrote: >>>>>>> 2010/2/8 Rafael J. Wysocki : >>>>>>>> This message has been generated automatically as a part of a r= eport >>>>>>>> of regressions introduced between 2.6.31 and 2.6.32. >>>>>>>> >>>>>>>> The following bug entry is on the current list of known regres= sions >>>>>>>> introduced between 2.6.31 and 2.6.32. Please verify if it sti= ll should >>>>>>>> be listed and let the tracking team know (either way). >>>>>>>> >>>>>>>> >>>>>>>> Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=3D= 15096 >>>>>>>> Subject : Resume lock up -- bisected, commit 3a1151e3f= 124fd1a2c54b8153f510f1a7c715369 >>>>>>>> Submitter : Rafa=C5=82 Mi=C5=82ecki >>>>>>>> Date : 2010-01-20 23:15 (19 days old) >>>>>>>> First-Bad-Commit: http://git.kernel.org/git/linus/3a1151e3f124= fd1a2c54b8153f510f1a7c715369 >>>>>> It looks like we should revert this commit, seems broken. >>>>> Hm, otoh we had definitely people with some Asus laptops who wher= e only be able >>>>> to use backlight control with it. So I would rather prefer some w= ork around to >>>>> the suspend/resume problem (_DOS if I understood correctly). >>>> It's same case with my Sony VAIO. Without this I can not control >>>> backlight. Eventually I can hack on GPU registers directly, but I >>>> don't think it can be called solution. >>> Well, do I understand correctly that without commit >>> 3a1151e3f124fd1a2c54b8153f510f1a7c715369 you can't control the back= light, >>> but with the commit applied there's a resume problem on your box? >>> >>> Rafael >> Rafa=C5=82 can correct me. I would phrase it that way that without t= he bisected >> patch, there is no backlight control. And because there is with it, = we hit the >> next BIOS bug because the backlight driver now tries to correctly br= ing it back >> on resume. >=20 > Do you suspect some conflict between video module and BIOS? Any more = ideas? >=20 That would be my guess based on the (admitted not to be very deep readi= ng across some of your and Rui's comments. He probably has a better Idea on that.= But what the patch in question does is to allow video devices that have not all = the methods defined (before _DOD *and* _DOS, after _DOD *or* _DOS) to be ha= ndled by the generic acpi video driver. And this gets you backlight control (by = some additional methods). But now, as the driver is in use, it is in the suspend/resume path and this causes problems. =46rom cross reading bugmail I had the feeling Rui was on a track while= I did not follow close enough to take in any details. I would suspect that code assumes details from those two (_DOD and _DOS= which is not there if the BIOS provides them both and is not handled as the driv= er used to be only used in that case). I probably should get back and read the = bug report with all the info in detail somewhen tomorrow. Stefan