Re: Bisected regression: 4287509b4d21 causes HP Spectre 14t-ea100 to overheat while suspended

Re: Bisected regression: 4287509b4d21 causes HP Spectre 14t-ea100 to overheat while suspended

[Rafael J. Wysocki]

On 3/9/2022 4:50 AM, Philippe Troin wrote:
> Linux 5.16.9 was working fine, but starting with 5.16.10 my main laptop
> has been overheating while suspended.
> I've also tried 5.16.11, 5.16.12 and 5.16.13 and they also show the
> same issue (overheating while suspended).
>
> I could not spot any difference between the dmesg messages issued
> during suspend between a working version (5.16.9) and the the broken
> ones (5.16.1[0-3]).
>
> I've bisected the regression down to commit 4287509b4d21
> https://lore.kernel.org/lkml/20220214092510.074083242@linuxfoundation.org/
>
> I have also tried reverting that change on top of 5.16.12, and the
> overheating behavior is gone.

OK, so it is not clear if this is a regression in -stable only or in the 
mainline.

Would it be possible to try 5.17-rc7 and see if the issue is still there?

Re: Bisected regression: 4287509b4d21 causes HP Spectre 14t-ea100 to overheat while suspended

[Philippe Troin]

On Wed, 2022-03-09 at 16:21 +0100, Rafael J. Wysocki wrote:
> On 3/9/2022 4:50 AM, Philippe Troin wrote:
> > Linux 5.16.9 was working fine, but starting with 5.16.10 my main laptop
> > has been overheating while suspended.
> > I've also tried 5.16.11, 5.16.12 and 5.16.13 and they also show the
> > same issue (overheating while suspended).
> > 
> > I could not spot any difference between the dmesg messages issued
> > during suspend between a working version (5.16.9) and the the broken
> > ones (5.16.1[0-3]).
> > 
> > I've bisected the regression down to commit 4287509b4d21
> > https://lore.kernel.org/lkml/20220214092510.074083242@linuxfoundation.org/
> > 
> > I have also tried reverting that change on top of 5.16.12, and the
> > overheating behavior is gone.
> 
> OK, so it is not clear if this is a regression in -stable only or in the 
> mainline.
> 
> Would it be possible to try 5.17-rc7 and see if the issue is still there?

Tried 5.17-rc7, and the issue is not appearing when running it (the
laptop does not overheat when suspended).

So it looks like the problem only appears on the 5.16 branch, starting
on 4287509b4d21.

Phil.

Re: Bisected regression: 4287509b4d21 causes HP Spectre 14t-ea100 to overheat while suspended

[Rafael J. Wysocki]

On Wed, Mar 9, 2022 at 7:02 PM Philippe Troin <phil@fifi.org> wrote:
>
> On Wed, 2022-03-09 at 16:21 +0100, Rafael J. Wysocki wrote:
> > On 3/9/2022 4:50 AM, Philippe Troin wrote:
> > > Linux 5.16.9 was working fine, but starting with 5.16.10 my main laptop
> > > has been overheating while suspended.
> > > I've also tried 5.16.11, 5.16.12 and 5.16.13 and they also show the
> > > same issue (overheating while suspended).
> > >
> > > I could not spot any difference between the dmesg messages issued
> > > during suspend between a working version (5.16.9) and the the broken
> > > ones (5.16.1[0-3]).
> > >
> > > I've bisected the regression down to commit 4287509b4d21
> > > https://lore.kernel.org/lkml/20220214092510.074083242@linuxfoundation.org/
> > >
> > > I have also tried reverting that change on top of 5.16.12, and the
> > > overheating behavior is gone.
> >
> > OK, so it is not clear if this is a regression in -stable only or in the
> > mainline.
> >
> > Would it be possible to try 5.17-rc7 and see if the issue is still there?
>
> Tried 5.17-rc7, and the issue is not appearing when running it (the
> laptop does not overheat when suspended).
>
> So it looks like the problem only appears on the 5.16 branch, starting
> on 4287509b4d21.

Thanks for the confirmation!