Re: Please revert commit 4287509b4d21e34dc492 from 5.16.y

Re: Please revert commit 4287509b4d21e34dc492 from 5.16.y

[Rafael J. Wysocki]

I messed up the stable list address, sorry.

On Thu, Mar 10, 2022 at 1:26 PM Rafael J. Wysocki <rafael@kernel.org> wrote:
>
> Hi Greg & Sasha,
>
> Commit 4287509b4d21e34dc492 that went into 5.16.y as a backport of
> mainline commit dc0075ba7f38 ("ACPI: PM: s2idle: Cancel wakeup before
> dispatching EC GPE") is causing trouble in 5.16.y, but 5.17-rc7
> including the original commit is fine.
>
> This is most likely due to some other changes that commit dc0075ba7f38
> turns out to depend on which have not been backported, but because it
> is not an essential fix (and it was backported, because it carried a
> Fixes tag and not because it was marked for backporting), IMV it is
> better to revert it from 5.16.y than to try to pull all of the
> dependencies in (and risk missing any of them), so please do that.
>
> Please see this thread:
>
> https://lore.kernel.org/linux-pm/31b9d1cd-6a67-218b-4ada-12f72e6f00dc@redhat.com/
>
> for reference.
>
> Cheers!

Re: Please revert commit 4287509b4d21e34dc492 from 5.16.y

[Greg Kroah-Hartman]

On Thu, Mar 10, 2022 at 01:26:16PM +0100, Rafael J. Wysocki wrote:
> Hi Greg & Sasha,
> 
> Commit 4287509b4d21e34dc492 that went into 5.16.y as a backport of
> mainline commit dc0075ba7f38 ("ACPI: PM: s2idle: Cancel wakeup before
> dispatching EC GPE") is causing trouble in 5.16.y, but 5.17-rc7
> including the original commit is fine.
> 
> This is most likely due to some other changes that commit dc0075ba7f38
> turns out to depend on which have not been backported, but because it
> is not an essential fix (and it was backported, because it carried a
> Fixes tag and not because it was marked for backporting), IMV it is
> better to revert it from 5.16.y than to try to pull all of the
> dependencies in (and risk missing any of them), so please do that.
> 
> Please see this thread:
> 
> https://lore.kernel.org/linux-pm/31b9d1cd-6a67-218b-4ada-12f72e6f00dc@redhat.com/

Odd that this is only showing up in 5.16 as this commit also is in 5.4
and 5.10 and 5.15.  Should I drop it from there as well?

thanks,

greg k-h

Re: Please revert commit 4287509b4d21e34dc492 from 5.16.y

[Hans de Goede]

Hi,

On 3/10/22 13:45, Greg Kroah-Hartman wrote:
> On Thu, Mar 10, 2022 at 01:26:16PM +0100, Rafael J. Wysocki wrote:
>> Hi Greg & Sasha,
>>
>> Commit 4287509b4d21e34dc492 that went into 5.16.y as a backport of
>> mainline commit dc0075ba7f38 ("ACPI: PM: s2idle: Cancel wakeup before
>> dispatching EC GPE") is causing trouble in 5.16.y, but 5.17-rc7
>> including the original commit is fine.
>>
>> This is most likely due to some other changes that commit dc0075ba7f38
>> turns out to depend on which have not been backported, but because it
>> is not an essential fix (and it was backported, because it carried a
>> Fixes tag and not because it was marked for backporting), IMV it is
>> better to revert it from 5.16.y than to try to pull all of the
>> dependencies in (and risk missing any of them), so please do that.
>>
>> Please see this thread:
>>
>> https://lore.kernel.org/linux-pm/31b9d1cd-6a67-218b-4ada-12f72e6f00dc@redhat.com/
> 
> Odd that this is only showing up in 5.16 as this commit also is in 5.4
> and 5.10 and 5.15. 

This has sofar mostly(only?) been reported by Fedora and Arch users which
both have 5.16.y kernels in their update repos ATM.

Regards,

Hans

Re: Please revert commit 4287509b4d21e34dc492 from 5.16.y

[Rafael J. Wysocki]

On Thu, Mar 10, 2022 at 1:45 PM Greg Kroah-Hartman
<gregkh@linuxfoundation.org> wrote:
>
> On Thu, Mar 10, 2022 at 01:26:16PM +0100, Rafael J. Wysocki wrote:
> > Hi Greg & Sasha,
> >
> > Commit 4287509b4d21e34dc492 that went into 5.16.y as a backport of
> > mainline commit dc0075ba7f38 ("ACPI: PM: s2idle: Cancel wakeup before
> > dispatching EC GPE") is causing trouble in 5.16.y, but 5.17-rc7
> > including the original commit is fine.
> >
> > This is most likely due to some other changes that commit dc0075ba7f38
> > turns out to depend on which have not been backported, but because it
> > is not an essential fix (and it was backported, because it carried a
> > Fixes tag and not because it was marked for backporting), IMV it is
> > better to revert it from 5.16.y than to try to pull all of the
> > dependencies in (and risk missing any of them), so please do that.
> >
> > Please see this thread:
> >
> > https://lore.kernel.org/linux-pm/31b9d1cd-6a67-218b-4ada-12f72e6f00dc@redhat.com/
>
> Odd that this is only showing up in 5.16 as this commit also is in 5.4
> and 5.10 and 5.15.  Should I drop it from there as well?

It's better to do so, because these series are also likely to be
missing the mainline commits it depends on.

Re: Please revert commit 4287509b4d21e34dc492 from 5.16.y

[Greg Kroah-Hartman]

On Thu, Mar 10, 2022 at 02:01:04PM +0100, Rafael J. Wysocki wrote:
> On Thu, Mar 10, 2022 at 1:45 PM Greg Kroah-Hartman
> <gregkh@linuxfoundation.org> wrote:
> >
> > On Thu, Mar 10, 2022 at 01:26:16PM +0100, Rafael J. Wysocki wrote:
> > > Hi Greg & Sasha,
> > >
> > > Commit 4287509b4d21e34dc492 that went into 5.16.y as a backport of
> > > mainline commit dc0075ba7f38 ("ACPI: PM: s2idle: Cancel wakeup before
> > > dispatching EC GPE") is causing trouble in 5.16.y, but 5.17-rc7
> > > including the original commit is fine.
> > >
> > > This is most likely due to some other changes that commit dc0075ba7f38
> > > turns out to depend on which have not been backported, but because it
> > > is not an essential fix (and it was backported, because it carried a
> > > Fixes tag and not because it was marked for backporting), IMV it is
> > > better to revert it from 5.16.y than to try to pull all of the
> > > dependencies in (and risk missing any of them), so please do that.
> > >
> > > Please see this thread:
> > >
> > > https://lore.kernel.org/linux-pm/31b9d1cd-6a67-218b-4ada-12f72e6f00dc@redhat.com/
> >
> > Odd that this is only showing up in 5.16 as this commit also is in 5.4
> > and 5.10 and 5.15.  Should I drop it from there as well?
> 
> It's better to do so, because these series are also likely to be
> missing the mainline commits it depends on.

Ok, will go do that now, thanks!

greg k-h