linux-toolchains.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Re: [PATCH] sched: Optimize __calc_delta.
       [not found]     ` <CABk29Ntt88BfqGz=Rjd5Mwyj+N0YLv0Z9QsZp8DgB-n+FV1Zgw@mail.gmail.com>
@ 2021-03-03 10:02       ` Peter Zijlstra
  2021-03-03 22:00         ` Josh Don
  0 siblings, 1 reply; 2+ messages in thread
From: Peter Zijlstra @ 2021-03-03 10:02 UTC (permalink / raw)
  To: Josh Don
  Cc: Ingo Molnar, Juri Lelli, Vincent Guittot, Dietmar Eggemann,
	Steven Rostedt, Ben Segall, Mel Gorman,
	Daniel Bristot de Oliveira, linux-kernel, Clement Courbet,
	Oleg Rombakh, Nick Desaulniers, linux-toolchains

On Tue, Mar 02, 2021 at 12:57:37PM -0800, Josh Don wrote:
> On gcc, the asm versions of `fls` are about the same speed as the
> builtin. On clang, the versions that use fls (fls,fls64) are more than
> twice as slow as the builtin. This is because the way the `fls` function
> is written, clang puts the value in memory:
> https://godbolt.org/z/EfMbYe. This can be fixed in a separate patch.

Is this because clang gets the asm constraints wrong? ISTR that
happening before, surely the right thing is to fix clang?

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH] sched: Optimize __calc_delta.
  2021-03-03 10:02       ` [PATCH] sched: Optimize __calc_delta Peter Zijlstra
@ 2021-03-03 22:00         ` Josh Don
  0 siblings, 0 replies; 2+ messages in thread
From: Josh Don @ 2021-03-03 22:00 UTC (permalink / raw)
  To: Peter Zijlstra
  Cc: Ingo Molnar, Juri Lelli, Vincent Guittot, Dietmar Eggemann,
	Steven Rostedt, Ben Segall, Mel Gorman,
	Daniel Bristot de Oliveira, linux-kernel, Clement Courbet,
	Oleg Rombakh, Nick Desaulniers, linux-toolchains

On Wed, Mar 3, 2021 at 2:02 AM Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Tue, Mar 02, 2021 at 12:57:37PM -0800, Josh Don wrote:
> > On gcc, the asm versions of `fls` are about the same speed as the
> > builtin. On clang, the versions that use fls (fls,fls64) are more than
> > twice as slow as the builtin. This is because the way the `fls` function
> > is written, clang puts the value in memory:
> > https://godbolt.org/z/EfMbYe. This can be fixed in a separate patch.
>
> Is this because clang gets the asm constraints wrong? ISTR that
> happening before, surely the right thing is to fix clang?

https://bugs.llvm.org/show_bug.cgi?id=49406 filed

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2021-03-04  0:49 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <20210226195239.3905966-1-joshdon@google.com>
     [not found] ` <YDliCOPy9s1RdLwd@hirez.programming.kicks-ass.net>
     [not found]   ` <CABk29Nv7iJEcDg3rgSvfTkXEM69ZeLByJAsZYuA5qpdj645nZw@mail.gmail.com>
     [not found]     ` <CABk29Ntt88BfqGz=Rjd5Mwyj+N0YLv0Z9QsZp8DgB-n+FV1Zgw@mail.gmail.com>
2021-03-03 10:02       ` [PATCH] sched: Optimize __calc_delta Peter Zijlstra
2021-03-03 22:00         ` Josh Don

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).