[Bug 925961] New: acerhdf and coretemp interfering on Acer Aspire One 150
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 Bug ID: 925961 Summary: acerhdf and coretemp interfering on Acer Aspire One 150 Classification: openSUSE Product: openSUSE Distribution Version: 13.2 Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kernel Assignee: kernel-maintainers@forge.provo.novell.com Reporter: dieter.jurzitza@t-online.de QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- After searching for quite a while I found out why the fan control on my Acer Aspire One did not work any more after upgrading from 13.1: 13.2 loads both the modules "coretemp" and "acerhdf". They interfere with each other, each one tries to control the fan accordingly, however, "coretemp" does not do the right thing for this netbook. To efficiently adjust fan settings "acerhdf" is the right choice, so coretemp needs to be blacklisted on the above mentioned netbook. Please do not say to report this upstream as my email address (not my personal one, it is t-online.de) is blacklisted on kernel org - if anyone considers this as an upstream topic, whoever may report it - please report. Who wants to set thing as required, please put: ################################################ blacklist coretemp options acerhdf verbose=0 fanon=70000 fanoff=63000 interval=15 kernelmode=1 ################################################ into /etc/modprobe.d/99-local.conf. This corresponds to 70°C for fan - on and 63°C for fan - off. Thank you for looking into this, take care Dieter Jurzitza -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 Dieter Jurzitza <dieter.jurzitza@t-online.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|CONFIRMED |RESOLVED Resolution|--- |WORKSFORME Flags|needinfo?(dieter.jurzitza@t | |-online.de) | --- Comment #20 from Dieter Jurzitza <dieter.jurzitza@t-online.de> --- Your assumptions were totally right - I removed the three lines (took my 5 minutes plus 2 days compile time :-)) and acerhdf is functional again. Could not observe any side effects so far, acerhdf does now what it is and was expected to do. So I add two patches, one for the three lines you readily mentioned and one for series.conf - I've been putting the patch into patches.suse.tar.bz2, you'll probably find a better place for this. Nevertheless, many thanks for your support - and if there is a kernel update currently pending it would be great if this could find it's way into ... Have a good start into the new week, take care Dieter Jurzitza -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 --- Comment #21 from Dieter Jurzitza <dieter.jurzitza@t-online.de> --- Created attachment 631534 --> http://bugzilla.opensuse.org/attachment.cgi?id=631534&action=edit Make sure that the acer - patch actually gets into the kernel ... -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 --- Comment #22 from Dieter Jurzitza <dieter.jurzitza@t-online.de> --- Created attachment 631535 --> http://bugzilla.opensuse.org/attachment.cgi?id=631535&action=edit This is needed to make acerhdf behave again as used to, before. This modification as suggested by Jeadn Delvare reverts the behaviour of the kernel so acerhdf does what it is excpected to, again. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 --- Comment #23 from Dieter Jurzitza <dieter.jurzitza@t-online.de> --- Last side-note: in comment #16 I talked about 3.11 not working, however I should have said 3.16, sorry for this, just discovered my error. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 Jean Delvare <jdelvare@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Resolution|WORKSFORME |FIXED --- Comment #24 from Jean Delvare <jdelvare@suse.com> --- Fix committed to the openSUSE 13.2 kernel branch, so the next maintenance update will include it. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 Jean Delvare <jdelvare@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |bpetkov@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 --- Comment #25 from Rui Zhang <rui.zhang@intel.com> --- hmmm, I think I'm trying to fix the bug in another way, please refer to patches at https://patchwork.kernel.org/patch/6166681/ https://patchwork.kernel.org/patch/6166691/ https://patchwork.kernel.org/patch/6166701/ with these three patches, it should work even with step_wise governor, and I'd like to push these patches to 4.1 and 3.18 - 4.0 stable kernel. about commit 178c2490b99f898efc06d1ad75cadc84f13021a6, are you going to revert it in suse kernel, or upstream stable kernel? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 --- Comment #26 from Jean Delvare <jdelvare@suse.com> --- Rui, thanks for looking into this. I am not familiar enough with the thermal subsystem to review your patches. Also I don't own an Acer laptop myself so I can't test if they would solve the problem that was reported here. But I have to say I have some doubts that they would actually help. My understanding of the problem is that the acerhdf wanted to be called back even when the step_wise governor considered that no change in cooling was needed, and the optimization broke that mechanism/assumption. If you believe that the problem would actually be solved by your patches, please explain how, as it isn't obvious to me. This is mostly theoretical though, as I don't think that your 3 patches qualify for stable kernels. They are large and somewhat intrusive, so they could come with regressions or side effects. I wouldn't take the risk of applying them to the openSUSE 13.2 kernel. Reverting the optimization as I did feels much safer. I did send a patch reverting 178c2490b99f898efc06d1ad75cadc84f13021a6 to the stable kernel mailing list, with suggestion to apply it to kernel branches v3.12 to v3.17. You were Cc'd. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 --- Comment #27 from Rui Zhang <rui.zhang@intel.com> --- (In reply to Jean Delvare from comment #26)
Rui, thanks for looking into this.
I am not familiar enough with the thermal subsystem to review your patches. Also I don't own an Acer laptop myself so I can't test if they would solve the problem that was reported here. But I have to say I have some doubts that they would actually help. My understanding of the problem is that the acerhdf wanted to be called back even when the step_wise governor considered that no change in cooling was needed, and the optimization broke that mechanism/assumption.
If you believe that the problem would actually be solved by your patches, please explain how, as it isn't obvious to me. This is mostly theoretical though, as I don't think that your 3 patches qualify for stable kernels. They are large and somewhat intrusive, so they could come with regressions or side effects. I wouldn't take the risk of applying them to the openSUSE 13.2 kernel. Reverting the optimization as I did feels much safer.
I did send a patch reverting 178c2490b99f898efc06d1ad75cadc84f13021a6 to the stable kernel mailing list, with suggestion to apply it to kernel branches v3.12 to v3.17. You were Cc'd.
Yes, IMO, the patch should be applied to all stable kernel as it is actually a bug in step_wise governor. and I will rework the three patches above on top of the fix then. can you please resend with cc: all stable kernel? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.opensuse.org/show_bug.cgi?id=925961 --- Comment #28 from Jean Delvare <jdelvare@suse.com> --- Sorry Rui, I don't really understand what you expect from me. I already sent the revert patch to stable: http://article.gmane.org/gmane.linux.kernel.stable/131678 I see no point in sending it again. I'll resend a copy to you in private if you lost it. I did not get any feedback from the stable folks, maybe a ack from you will help get the revert patch in. I also can't see how this is relevant to you, as your 3 patches are not stable material anyway, they should go in kernel v4.1 or v4.2, they will not apply on top of my revert patch which is only for older kernels. And I disagree that this was a bug in the step_wise governor. The optimization was correct, the problem was that acerhdf was abusing the step_wise governor. But this is fixed since kernel v3.19. -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠novell.com