[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20191024061311.GA4541@gmail.com>
Date: Wed, 23 Oct 2019 23:13:11 -0700
From: Andrei Vagin <avagin@...il.com>
To: Vincenzo Frascino <vincenzo.frascino@....com>
Cc: Dmitry Safonov <dima@...sta.com>, linux-kernel@...r.kernel.org,
Dmitry Safonov <0x7f454c46@...il.com>,
Adrian Reber <adrian@...as.de>,
Andrei Vagin <avagin@...nvz.org>,
Andy Lutomirski <luto@...nel.org>,
Arnd Bergmann <arnd@...db.de>,
Christian Brauner <christian.brauner@...ntu.com>,
Cyrill Gorcunov <gorcunov@...nvz.org>,
"Eric W. Biederman" <ebiederm@...ssion.com>,
"H. Peter Anvin" <hpa@...or.com>, Ingo Molnar <mingo@...hat.com>,
Jann Horn <jannh@...gle.com>, Jeff Dike <jdike@...toit.com>,
Oleg Nesterov <oleg@...hat.com>,
Pavel Emelyanov <xemul@...tuozzo.com>,
Shuah Khan <shuah@...nel.org>,
Thomas Gleixner <tglx@...utronix.de>,
containers@...ts.linux-foundation.org, criu@...nvz.org,
linux-api@...r.kernel.org, x86@...nel.org
Subject: Re: [PATCHv7 18/33] lib/vdso: Add unlikely() hint into
vdso_read_begin()
On Wed, Oct 16, 2019 at 12:24:14PM +0100, Vincenzo Frascino wrote:
> On 10/11/19 2:23 AM, Dmitry Safonov wrote:
> > From: Andrei Vagin <avagin@...il.com>
> >
> > Place the branch with no concurrent write before contended case.
> >
> > Performance numbers for Intel(R) Core(TM) i5-6300U CPU @ 2.40GHz
> > (more clock_gettime() cycles - the better):
> > | before | after
> > -----------------------------------
> > | 150252214 | 153242367
> > | 150301112 | 153324800
> > | 150392773 | 153125401
> > | 150373957 | 153399355
> > | 150303157 | 153489417
> > | 150365237 | 153494270
> > -----------------------------------
> > avg | 150331408 | 153345935
> > diff % | 2 | 0
> > -----------------------------------
> > stdev % | 0.3 | 0.1
> >
> > Signed-off-by: Andrei Vagin <avagin@...il.com>
> > Co-developed-by: Dmitry Safonov <dima@...sta.com>
> > Signed-off-by: Dmitry Safonov <dima@...sta.com>
>
> Reviewed-by: Vincenzo Frascino <vincenzo.frascino@....com>
> Tested-by: Vincenzo Frascino <vincenzo.frascino@....com>
Hello Vincenzo,
Could you test the attached patch on aarch64? On x86, it gives about 9%
performance improvement for CLOCK_MONOTONIC and CLOCK_BOOTTIME.
Here is my test:
https://github.com/avagin/vdso-perf
It is calling clock_gettime() in a loop for three seconds and then
reports a number of iterations.
Thanks,
Andrei
View attachment "0001-lib-vdso-make-do_hres-and-do_coarse-as-__always_inli.patch" of type "text/plain" (2209 bytes)
Powered by blists - more mailing lists