This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH v1.1] Improve unaligned memcpy and memmove.
- From: OndÅej BÃlka <neleai at seznam dot cz>
- To: Liubov Dmitrieva <liubov dot dmitrieva at gmail dot com>, r at domone dot podge
- Cc: GNU C Library <libc-alpha at sourceware dot org>
- Date: Fri, 27 Sep 2013 12:58:30 +0200
- Subject: Re: [PATCH v1.1] Improve unaligned memcpy and memmove.
- Authentication-results: sourceware.org; auth=none
- References: <20130819085220 dot GB19541 at domone> <20130829153829 dot GA6105 at domone dot kolej dot mff dot cuni dot cz> <20130923194237 dot GE11240 at domone dot podge> <CAHjhQ93v+GC2u7eFw7--MV7OZQ6nQcMGpWUD7kLBX3NFzUrmkA at mail dot gmail dot com> <20130926174239 dot GA8548 at domone dot podge> <CAHjhQ91R8cf=GRmkByn838b4m+H78NZ8eX8bhaVTx7SmaYCqmw at mail dot gmail dot com>
On Fri, Sep 27, 2013 at 02:08:41PM +0400, Liubov Dmitrieva wrote:
> It seems it doesn't work even when I removed the files. The script
> goes to infinite loop now on Silvermont.
> Can you please check that it works or upload fixed version.
>
Here I decided to gather more data to get more accurate graphs so it
probably it is running too long
A probable step there gathers data by repeately running make which
could be checked by looking at results/result.html file. A graph should
like following, and when you refresh page after one call iteration
number after Total calls: should increase.
http://kam.mff.cuni.cz/~ondra/benchmark_string/i7_ivy_bridge/memcpy_profile_loop/results_gcc/result.html
On benchmarks file you could modify time by adding TIMES=number line
before tests, default is ten times so adding
TIMES=1 # before
test results_gcc make bench_gcc
should speed this up.