Re: [PATCH 0/7] Backport uclamp vs margin fixes into 5.15.y

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 03/16/23 08:43, Greg KH wrote:
> On Wed, Mar 15, 2023 at 12:53:04PM +0000, Qais Yousef wrote:
> > On 03/15/23 09:53, Greg KH wrote:
> > > On Wed, Mar 15, 2023 at 09:03:45AM +0100, Greg KH wrote:
> > > > On Wed, Mar 08, 2023 at 04:22:00PM +0000, Qais Yousef wrote:
> > > > > Portion of the fixes were ported in 5.15 but missed some.
> > > > > 
> > > > > This ports the remainder of the fixes.
> > > > > 
> > > > > Based on 5.15.98.
> > > > > 
> > > > > Build tested on x86 with and without uclamp config enabled.
> > > > 
> > > > Now queued up, thanks.
> > > 
> > > Wait, should we also queue up a2e90611b9f4 ("sched/fair: Remove capacity
> > > inversion detection") to 5.10 and 5.15 now as well?
> > 
> > It has a dependency on e5ed0550c04c ("sched/fair: unlink misfit task from cpu
> > overutilized") which is nice to have but not strictly required. It improves the
> > search for best CPU under adverse thermal pressure to try harder. And the new
> > search effectively replaces the capacity inversion detection, so it is removed
> > afterwards.
> > 
> > I'd like to have it but find_energy_efficient_cpu() looks a bit different for
> > a straightforward backport and opted to avoid the risk.
> > 
> > Happy to reconsider though.
> 
> Ok, just wanting to verify as that commit looked like it was needed
> after these.
> 
> But, as I've just now dropped both 5.10.y and 5.15.y sets of patches,
> this is going to need to be redone anyway, so perhaps put that note in
> the 00/XX email when you resubmit so that I don't get confused again.

Will do. There were no failure for the 5.15 set though. The reports were
against 5.10 only.

I retested 5.15 series against:

	1. default ubuntu config which has uclamp + smp
	2. default ubuntu config without uclamp + smp
	3. default ubunto config without smp (which automatically disabled
	   uclamp)
	4. reported riscv-allnoconfig, mips-randconfig, x86_64-randocnfigs

And they all compiled without an error. Happy to retake them or prefer a resend
anyway though nothing has changed?

Just finishing testing 5.10 and will send v2 with the updated cover-letter.


Thanks!

--
Qais Yousef



[Index of Archives]     [Linux Kernel]     [Kernel Development Newbies]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite Hiking]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux