perf record: Add record.build-id config option

Post processing at 'perf record' takes a long time on big machines.

What it does is to find the build-id of binaries found in the event
stream, so that it can make sure, at 'report' time, that the symtabs (be
it ELF, kallsyms, etc) being used to resolve symbols are the ones
matching the binaries found at 'record' time.

Sometimes we just want to skip this processing of events at the end of
the session to get quicker results, making sure the binaries haven't
changed from 'record' to 'report' time.

Add a new config option to control this behavior.

The record.build-id config variable can have one of the following
values:

 - cache: post-process data and save/update the binaries into the
          build-id cache (in ~/.debug).  This is the default.
 - no-cache: post-process the data but not update the build-id cache.
             Same effect as using the -N option.
 - skip: skip post-processing and do not update the cache.
         Same effect as using the -B option.

Reported-and-Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Taeung Song <treeze.taeung@gmail.com>
Link: http://lkml.kernel.org/r/1450144196-22957-1-git-send-email-namhyung@kernel.org
[ Added some more text to the documentation ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This commit is contained in:
Namhyung Kim 2015-12-15 10:49:56 +09:00 committed by Arnaldo Carvalho de Melo
parent 7efe0e034c
commit 7a29c087ff
2 changed files with 26 additions and 1 deletions

View File

@ -207,11 +207,23 @@ comma-separated list with no space: 0,1. Ranges of CPUs are specified with -: 0-
In per-thread mode with inheritance mode on (default), samples are captured only when
the thread executes on the designated CPUs. Default is to monitor all CPUs.
-B::
--no-buildid::
Do not save the build ids of binaries in the perf.data files. This skips
post processing after recording, which sometimes makes the final step in
the recording process to take a long time, as it needs to process all
events looking for mmap records. The downside is that it can misresolve
symbols if the workload binaries used when recording get locally rebuilt
or upgraded, because the only key available in this case is the
pathname. You can also set the "record.build-id" config variable to
'skip to have this behaviour permanently.
-N::
--no-buildid-cache::
Do not update the buildid cache. This saves some overhead in situations
where the information in the perf.data file (which includes buildids)
is sufficient.
is sufficient. You can also set the "record.build-id" config variable to
'no-cache' to have the same effect.
-G name,...::
--cgroup name,...::

View File

@ -837,6 +837,19 @@ int record_callchain_opt(const struct option *opt,
static int perf_record_config(const char *var, const char *value, void *cb)
{
struct record *rec = cb;
if (!strcmp(var, "record.build-id")) {
if (!strcmp(value, "cache"))
rec->no_buildid_cache = false;
else if (!strcmp(value, "no-cache"))
rec->no_buildid_cache = true;
else if (!strcmp(value, "skip"))
rec->no_buildid = true;
else
return -1;
return 0;
}
if (!strcmp(var, "record.call-graph"))
var = "call-graph.record-mode"; /* fall-through */