Re: [PATCH 3/3][RFC/RFT] PM / sleep: Do not delay the synchronization of MTRR during resume

From: Lukas Wunner
Date: Tue Feb 06 2018 - 09:04:32 EST


On Thu, Dec 14, 2017 at 12:02:42AM +0800, Yu Chen wrote:
> On Wed, Dec 13, 2017 at 01:31:50AM +0100, Rafael J. Wysocki wrote:
> > On Tuesday, October 31, 2017 10:58:50 AM CET Yu Chen wrote:
[snip]
> > > In one word, with the patch applied, the cpu online time during resume
> > > has decreased by about 6 seconds on a bogus MTRR platform, and decreased
> > > by about 600ms on a 88 cpus Xeon platform after resumed.
> > >
> > > Cc: Len Brown <len.brown@xxxxxxxxx>
> > > Cc: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> > > Cc: Rui Zhang <rui.zhang@xxxxxxxxx>
> > > Signed-off-by: Chen Yu <yu.c.chen@xxxxxxxxx>
> >
> > It will be better to combine this with patch [2/3] IMO, because that makes
> > it clear why the changes in that patch are needed.
> >
> > Also you can define the new flag in mtrr/main.c, set it in
> > arch_enable_nonboot_cpus_begin() and clear it in
> > arch_enable_nonboot_cpus_end(). It is better to put it into the
> > arch-specific code as the flag itself is arch-specific.
> >
> > Then, of course, you don't need patch [1/3] and all can be done in one
> > patch.
>
> Ok, will rewrite the patch, thanks!

Just for the record, this series cuts down resume time from system sleep
by 4-5 seconds on my MacBookPro9,1. Great work, looking forward to this
being respun and merged.

Tested-by: Lukas Wunner <lukas@xxxxxxxxx>

Thanks,

Lukas