From: Don Zickus <dzickus@redhat.com>
To: Jiri Olsa <jolsa@redhat.com>
Cc: acme@ghostprotocols.net, LKML <linux-kernel@vger.kernel.org>,
jmario@redhat.com, fowles@inreach.com, eranian@google.com
Subject: Re: [PATCH 10/21] perf, c2c: Add stats to track data source bits and cpu to node maps
Date: Tue, 18 Feb 2014 21:51:57 -0500 [thread overview]
Message-ID: <20140219025157.GN25953@redhat.com> (raw)
In-Reply-To: <20140218130530.GJ4343@krava.brq.redhat.com>
On Tue, Feb 18, 2014 at 02:05:31PM +0100, Jiri Olsa wrote:
> On Mon, Feb 10, 2014 at 12:29:05PM -0500, Don Zickus wrote:
> > This patch adds a bunch of stats that will be used later in post-processing
> > to determine where and with what frequency the HITMs are coming from.
> >
> > Most of the stats are decoded from the data source response. Another
> > piece of the stats is tracking which cpu the record came in on.
> >
> > In order to properly build a cpu map to map where interesting events are coming
> > from, I shamelessly copy-n-pasted the cpu->NUMA node code from builtin-kmem.c.
> >
> > As HITMs are most expensive when going across NUMA nodes, it only made sense
> > to create a quick cpu->NUMA lookup for when processing the records.
> >
> > Credit to Dick Fowles for determining which bits are important and how to
> > properly track them. Ported to perf by me.
> >
> > Original-by: Dick Fowles <rfowles@redhat.com>
> > Signed-off-by: Don Zickus <dzickus@redhat.com>
> > ---
>
> SNIP
>
> > +
> > +static int setup_cpunode_map(void)
> > +{
> > + struct dirent *dent1, *dent2;
> > + DIR *dir1, *dir2;
> > + unsigned int cpu, mem;
> > + char buf[PATH_MAX];
> > +
> > + /* initialize globals */
> > + if (init_cpunode_map())
> > + return -1;
> > +
> > + dir1 = opendir(PATH_SYS_NODE);
> > + if (!dir1)
> > + return 0;
> > +
> > + /* walk tree and setup map */
> > + while ((dent1 = readdir(dir1)) != NULL) {
> > + if (dent1->d_type != DT_DIR ||
> > + sscanf(dent1->d_name, "node%u", &mem) < 1)
> > + continue;
> > +
> > + snprintf(buf, PATH_MAX, "%s/%s", PATH_SYS_NODE, dent1->d_name);
> > + dir2 = opendir(buf);
> > + if (!dir2)
> > + continue;
> > + while ((dent2 = readdir(dir2)) != NULL) {
> > + if (dent2->d_type != DT_LNK ||
> > + sscanf(dent2->d_name, "cpu%u", &cpu) < 1)
> > + continue;
> > + cpunode_map[cpu] = mem;
> > + }
> > + closedir(dir2);
> > + }
> > + closedir(dir1);
> > + return 0;
> > +}
>
>
> There's already setup_cpunode_map interface in builtin-kmem.c
> Please make it global (maybe place in separate object?)
> and use this one.
Heh, where do you think I got this from? :-) Though I did tweak it for my
needs, namely I used 'possible' cpus as opposed to 'online' cpus to deal
with hotplug.
I also ran into a bug here, where this code populating an array based on
what is on the running system, not the system where the data was
collected. Is it possible to have perf-archive add this info?
I try to make this function global on the next version.
Cheers,
Don
>
> jirka
next prev parent reply other threads:[~2014-02-19 2:52 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-02-10 17:28 [PATCH 00/21] perf, c2c: Add new tool to analyze cacheline contention on NUMA systems Don Zickus
2014-02-10 17:28 ` [PATCH 03/21] Revert "perf: Disable PERF_RECORD_MMAP2 support" Don Zickus
2014-02-10 17:28 ` [PATCH 04/21] perf, machine: Use map as success in ip__resolve_ams Don Zickus
2014-02-10 17:29 ` [PATCH 05/21] perf, session: Change header.misc dump from decimal to hex Don Zickus
2014-02-18 12:56 ` Jiri Olsa
2014-02-19 2:40 ` Don Zickus
2014-02-10 17:29 ` [PATCH 06/21] perf, stat: FIXME Stddev calculation is incorrect Don Zickus
2014-02-10 17:29 ` [PATCH 07/21] perf, callchain: Add generic callchain print handler for stdio Don Zickus
2014-02-10 17:29 ` [PATCH 08/21] perf, c2c: Rework setup code to prepare for features Don Zickus
2014-02-18 13:02 ` Jiri Olsa
2014-02-19 2:45 ` Don Zickus
2014-02-10 17:29 ` [PATCH 09/21] perf, c2c: Add rbtree sorted on mmap2 data Don Zickus
2014-02-18 13:04 ` Jiri Olsa
2014-02-19 2:48 ` Don Zickus
2014-02-21 2:45 ` Don Zickus
2014-02-21 16:59 ` Jiri Olsa
2014-02-26 3:12 ` Don Zickus
2014-02-10 17:29 ` [PATCH 10/21] perf, c2c: Add stats to track data source bits and cpu to node maps Don Zickus
2014-02-18 13:05 ` Jiri Olsa
2014-02-19 2:51 ` Don Zickus [this message]
2014-02-10 17:29 ` [PATCH 11/21] perf, c2c: Sort based on hottest cache line Don Zickus
2014-02-10 17:29 ` [PATCH 12/21] perf, c2c: Display cacheline HITM analysis to stdout Don Zickus
2014-02-10 17:29 ` [PATCH 13/21] perf, c2c: Add callchain support Don Zickus
2014-02-18 13:07 ` Jiri Olsa
2014-02-19 2:54 ` Don Zickus
2014-02-10 17:29 ` [PATCH 14/21] perf, c2c: Output summary stats Don Zickus
2014-02-10 17:29 ` [PATCH 15/21] perf, c2c: Dump rbtree for debugging Don Zickus
2014-02-10 17:29 ` [PATCH 16/21] perf, c2c: Fixup tid because of perf map is broken Don Zickus
2014-02-10 17:29 ` [PATCH 17/21] perf, c2c: Add symbol count table Don Zickus
2014-02-18 13:09 ` Jiri Olsa
2014-02-19 2:56 ` Don Zickus
2014-02-10 17:29 ` [PATCH 18/21] perf, c2c: Add shared cachline summary table Don Zickus
2014-02-10 17:29 ` [PATCH 19/21] perf, c2c: Add framework to analyze latency and display summary stats Don Zickus
2014-02-10 17:29 ` [PATCH 20/21] perf, c2c: Add selected extreme latencies to output cacheline stats table Don Zickus
2014-02-10 17:29 ` [PATCH 21/21] perf, c2c: Add summary latency table for various parts of caches Don Zickus
2014-02-10 18:59 ` [PATCH 00/21] perf, c2c: Add new tool to analyze cacheline contention on NUMA systems Davidlohr Bueso
2014-02-10 19:17 ` Don Zickus
2014-02-10 19:18 ` [PATCH 01/21] perf c2c: Shared data analyser Don Zickus
2014-02-10 22:10 ` Davidlohr Bueso
2014-02-11 11:24 ` Jiri Olsa
2014-02-11 11:31 ` Arnaldo Carvalho de Melo
2014-02-11 13:54 ` Don Zickus
2014-02-11 14:36 ` Don Zickus
2014-02-11 15:41 ` Arnaldo Carvalho de Melo
2014-02-10 19:18 ` [PATCH 02/21] perf c2c: Dump raw records, decode data_src bits Don Zickus
2014-02-10 21:18 ` [PATCH 00/21] perf, c2c: Add new tool to analyze cacheline contention on NUMA systems Peter Zijlstra
2014-02-10 22:11 ` Don Zickus
2014-02-10 21:29 ` Peter Zijlstra
2014-02-10 22:20 ` Don Zickus
2014-02-10 22:21 ` Stephane Eranian
2014-02-11 7:14 ` Peter Zijlstra
2014-02-11 10:35 ` Stephane Eranian
2014-02-11 10:52 ` Peter Zijlstra
2014-02-11 10:58 ` Stephane Eranian
2014-02-11 11:02 ` Peter Zijlstra
2014-02-11 11:04 ` Stephane Eranian
2014-02-11 11:08 ` Peter Zijlstra
2014-02-11 11:08 ` Stephane Eranian
2014-02-11 11:14 ` Peter Zijlstra
2014-02-11 11:28 ` Stephane Eranian
2014-02-11 11:31 ` Peter Zijlstra
2014-02-11 11:51 ` Peter Zijlstra
2014-02-11 11:50 ` Arnaldo Carvalho de Melo
2014-02-11 12:09 ` Peter Zijlstra
2014-02-13 13:02 ` Jiri Olsa
2014-02-13 13:10 ` Stephane Eranian
[not found] ` <1392053356-23024-2-git-send-email-dzickus@redhat.com>
2014-02-18 12:52 ` [PATCH 01/21] perf c2c: Shared data analyser Jiri Olsa
2014-02-18 12:56 ` Arnaldo Carvalho de Melo
2014-02-19 2:42 ` Don Zickus
[not found] ` <1392053356-23024-3-git-send-email-dzickus@redhat.com>
2014-02-18 12:53 ` [PATCH 02/21] perf c2c: Dump raw records, decode data_src bits Jiri Olsa
2014-02-18 13:49 ` Arnaldo Carvalho de Melo
2014-02-19 3:04 ` Don Zickus
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=20140219025157.GN25953@redhat.com \
--to=dzickus@redhat.com \
--cc=acme@ghostprotocols.net \
--cc=eranian@google.com \
--cc=fowles@inreach.com \
--cc=jmario@redhat.com \
--cc=jolsa@redhat.com \
--cc=linux-kernel@vger.kernel.org \
/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;
as well as URLs for NNTP newsgroup(s).