From: nils.carlson@ludd.ltu.se (Nils Carlson)
Subject: [ltt-dev] PATCH [0/1]: Replace gettimeofday with clock_gettime
Date: Tue, 25 May 2010 21:38:20 +0200 [thread overview]
Message-ID: <B7D438A9-94F6-480E-AA6C-3FD42C2B7024@ludd.ltu.se> (raw)
In-Reply-To: <4BFBDF38.7090200@polymtl.ca>
Comments below,
On May 25, 2010, at 4:31 PM, Pierre-Marc Fournier wrote:
> Thanks, comments below.
>
>
> On 05/24/2010 10:54 AM, Nils Carlson wrote:
>> This patch replaces gettimeofday with clock_gettime(CLOCK_MONOTONIC,)
>>
>> Signed-off-by: Nils Carlson<nils.carlson at ludd.ltu.se>
>>
>> diff --git a/include/ust/clock.h b/include/ust/clock.h
>> index d4b6a9d..bc26d26 100644
>> --- a/include/ust/clock.h
>> +++ b/include/ust/clock.h
>> @@ -18,6 +18,7 @@
>> #ifndef UST_CLOCK_H
>> #define UST_CLOCK_H
>>
>> +#include<time.h>
>> #include<sys/time.h>
>> #include<ust/kcompat/kcompat.h>
>>
>> @@ -25,7 +26,7 @@
>>
>> /* There are two types of clocks that can be used.
>> - TSC based clock
>> - - gettimeofday() clock
>> + - gettimeofday() clock - to guarantee monotonicity use
>> clock_gettime
>>
>> Microbenchmarks on Linux 2.6.30 on Core2 Duo 3GHz (functions are
>> inlined):
>> Calls (100000000) to tsc(): 4004035641 cycles or 40 cycles/call
>> @@ -41,13 +42,13 @@
>>
>> static __inline__ u64 trace_clock_read64(void)
>> {
>> - struct timeval tv;
>> + struct timespec ts;
>> u64 retval;
>>
>> - gettimeofday(&tv, NULL);
>> - retval = tv.tv_sec;
>> + clock_gettime(CLOCK_MONOTONIC,&ts);
>> + retval = ts.tv_sec;
>> retval *= 1000000;
>> - retval += tv.tv_usec;
>> + retval += ts.tv_nsec / 1000;
>
> Any specific reason for truncating the nanoseconds here?
Only to provide the same fidelity as the previous code. I don't know
enough
about how you use the timer, so I didn't take it upon myself to change
the
resolution.
>
>>
>> return retval;
>> }
>> diff --git a/libust/Makefile.am b/libust/Makefile.am
>> index 2de5857..e357ad1 100644
>> --- a/libust/Makefile.am
>> +++ b/libust/Makefile.am
>> @@ -27,6 +27,7 @@ libust_la_LDFLAGS = -no-undefined -version-info
>> 0:0:0
>>
>> libust_la_LIBADD = \
>> -lpthread \
>> + -lrt \
>
> I guess this isn't supposed to be part of the patch?
It is. clock_gettime requires the realtime clock library.
>
>> $(top_builddir)/snprintf/libustsnprintf.la \
>> $(top_builddir)/libustcomm/libustcomm.la
>>
>>
>>
>>
>>
>> _______________________________________________
>> ltt-dev mailing list
>> ltt-dev at lists.casi.polymtl.ca
>> http://lists.casi.polymtl.ca/cgi-bin/mailman/listinfo/ltt-dev
next prev parent reply other threads:[~2010-05-25 19:38 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-24 14:54 Nils Carlson
2010-05-25 14:31 ` Pierre-Marc Fournier
2010-05-25 19:38 ` Nils Carlson [this message]
2010-05-26 16:33 ` Pierre-Marc Fournier
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=B7D438A9-94F6-480E-AA6C-3FD42C2B7024@ludd.ltu.se \
--to=nils.carlson@ludd.ltu.se \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox