lkml.org 
[lkml]   [2016]   [Nov]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [tip:x86/core] x86: Enable Intel Turbo Boost Max Technology 3.0
From
Date
On Mon, 2016-11-28 at 09:51 +0100, Ingo Molnar wrote:
> * Thomas Gleixner <tglx@linutronix.de> wrote:
>
> > 
> >
> > >
> > > >
> > > > +#include <linux/sched.h>
> > > > +#include <linux/cpumask.h>
> > > > +#include <linux/cpuset.h>
> > > > +#include <asm/mutex.h>
> > > > +#include <linux/sched.h>
> > > > +#include <linux/sysctl.h>
> > > > +#include <linux/nodemask.h>
> > > arch/x86/kernel/itmt.c:26:23: fatal error: asm/mutex.h: No such file or directory
> > >
> > > >
> > > > +config SCHED_ITMT
> > > > + bool "Intel Turbo Boost Max Technology (ITMT) scheduler support"
> > > > + depends on SCHED_MC && CPU_SUP_INTEL && X86_INTEL_PSTATE
> > > > + ---help---
> > > > +   ITMT enabled scheduler support improves the CPU scheduler's decision
> > > > +   to move tasks to cpu core that can be boosted to a higher frequency
> > > > +   than others. It will have better performance at a cost of slightly
> > > > +   increased overhead in task migrations. If unsure say N here.
> > > Argh, so the 'itmt' name really sucks as well - could we please make it something 
> > > more obvious - like SCHED_INTEL_TURBO or so - and similarly rename the file as 
> > > well?
> > >
> > > The sched_intel_turbo.c file could thus host all things related to scheduler 
> > > support of turbo frequencies - it shouldn't be named after the Intel acronym of 
> > > the day...
> > It would be nice to come up with such nitpicks during review. This thing went 
> > through 8 iterations, but nothing came up and I didn't mind the itmt naming.
> Yeah, so I had to NAK an early iteration and didn't get around to doing a really 
> detailed review yet - and after (falsely) thinking it had a build failure I got 
> overly worked up about the bad naming: my bad and apologies!
>
> So the code looks good to me but the naming still sucks a bit - I'm fine with 
> having the commits re-merged as-is and renaming the Kconfig variable to something 
> more expressive: I've done this in tip:sched/core and have fixed the asm/mutex.h 
> thing as well.
>
> Wrt. improving the naming:
>
> Firstly, popular tech news has coined the 'Turbo Boost Max' technology 'TBM' (TBM2 
> and TBM3) as the natural acronym of the Intel feature - not 'ITMT'. So to anyone 
> except people well aware of Intel acronyms the term 'ITMT' will be pretty 
> meaningless.
>
> Does something more generic like SCHED_MC_PRIO (as an extension to SCHED_MC) work 
> with everyone? Intel Turbo Max 3.0 is the current (only) implementation of it, but 
> I don't think the technology will stop at that stage as dies are getting larger 
> but thinner.
>
> I also think the Kconfig text is somewhat misleading and the default-disabled 
> status is counterproductive:
>
> +config SCHED_ITMT
> +       bool "Intel Turbo Boost Max Technology (ITMT) scheduler support"
> +       depends on SCHED_MC && CPU_SUP_INTEL && X86_INTEL_PSTATE
> +       ---help---
> +         ITMT enabled scheduler support improves the CPU scheduler's decision
> +         to move tasks to cpu core that can be boosted to a higher frequency
> +         than others. It will have better performance at a cost of slightly
> +         increased overhead in task migrations. If unsure say N here.
>
> ... the extra cost of smarter CPU selection is IMHO overwhelmed by the negative 
> effects of not knowing about core frequency ordering, on most workloads.
>
> A better default would be default-y I believe (that is what we do for CPU hardware 
> enablement typically), and a better description would be something like:
>
> +config SCHED_MC_PRIO
> +       bool "CPU core priorities scheduler support"
> +       depends on SCHED_MC && CPU_SUP_INTEL && X86_INTEL_PSTATE
> + default y
> +       ---help---
> +       Intel Turbo Boost Max 3.0 enabled CPUs have a core ordering determined at 
> + manufacturing time, which allows certain cores to reach higher turbo
> + frequencies (when running single threaded workloads) than others.
> +
> + Enabling this kernel feature teaches the scheduler about the TBM3 priority
> + order of the CPU cores and adjusts the scheduler's CPU selection logic 
> + accordingly, so that higher overall system performance can be achieved.
> +
> + This feature will have no effect on CPUs without this feature.
> +
> + If unsure say Y here.
>
> If/when other architectures make use of this the Kconfig entry can be moved into 
> the scheduler Kconfig - but for the time being it can stay in arch/x86/.
>
> Another variant would be to eliminate the Kconfig option altogether and make it a 
> natural feature of SCHED_MC (like it is in the core scheduler).
>

I am fine with renaming SCHED_ITMT to SCHED_MC_PRIO.  Patch 7 and 8 that
Rafael merged into his tree also have SCHED_ITMT so they will need to
be updated if we renamed it.

Thanks.

Tim

\
 
 \ /
  Last update: 2016-11-28 18:37    [W:0.092 / U:0.040 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site