From: Ingo Molnar <mingo@kernel.org>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Jiri Olsa <jolsa@redhat.com>,
linux-kernel@vger.kernel.org,
Corey Ashford <cjashfor@linux.vnet.ibm.com>,
Frederic Weisbecker <fweisbec@gmail.com>,
Ingo Molnar <mingo@elte.hu>, Namhyung Kim <namhyung@kernel.org>,
Paul Mackerras <paulus@samba.org>,
Arnaldo Carvalho de Melo <acme@redhat.com>,
Andi Kleen <ak@linux.intel.com>, David Ahern <dsahern@gmail.com>,
Stephane Eranian <eranian@google.com>
Subject: Re: [PATCH 0/9] perf: Adding better precise_ip field handling
Date: Fri, 10 May 2013 12:55:36 +0200 [thread overview]
Message-ID: <20130510105536.GA18805@gmail.com> (raw)
In-Reply-To: <20130510103436.GC31235@dyad.programming.kicks-ass.net>
* Peter Zijlstra <peterz@infradead.org> wrote:
> On Fri, May 10, 2013 at 12:31:12PM +0200, Ingo Molnar wrote:
> > PERF_COUNT_HW_CPU_CYCLES_PRECISE is not a vague request at all: it means
> > 'get me the most precise cycles profiling available on this system'.
>
> And how will you interpret the results? Do you know to manually adjust
> for skid or will you assume the results 'correct'?
Look at the tools/perf/ patches, they don't actually need or use that
information to adjust for skid!
If user-space wants _that_ level of control because it wants to correct
for skid (if there's skid), or if it wants to display to the user how
precise the profiling is, then they can do the (much) more complex probing
dance.
What is absolutely indefensible is to not give a good shortcut for the
most common case of 'give me the most precise cycles event you got'...
> I see such a feature only causing confusion; I told it to be precise,
> therefore this register op after the memory load really is the more
> expensive thing.
You are creating confusion where there's none: "give me the best profiling
you've got" is a pretty reasonable thing to ask.
The thing is, there's variations in the quality of profiling between CPUs,
sometimes even between CPU models. 99.999% of the people don't care about
that, because 99.9% of the time the profile is unambiguous: functions are
typically big enough, with the overhead somewhere in the middle, so skid
just doesn't matter.
So to 'solve' this corner case information you worry about to extract
(which cannot be solved - there's more profiling artifacts and imprecision
than skid) you sacrifice the thing that _DOES_ matter: for the kernel to
offer our best profiling feature as easily as possible...
What explanation do you have for that failure?
> People generally don't volunteer to think, you have to force them to --
> even if that makes them complain.
It is a mistake to 'force' people to consider stuff _they don't care about
in 99% of the cases_.
Thanks,
Ingo
next prev parent reply other threads:[~2013-05-10 10:55 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-05-09 13:32 [PATCH 0/9] perf: Adding better precise_ip field handling Jiri Olsa
2013-05-09 13:32 ` [PATCH 1/9] perf x86: Add precise sysfs cpu pmu attribute Jiri Olsa
2013-05-09 13:32 ` [PATCH 2/9] perf tools: Add precise object to interface sysfs precise Jiri Olsa
2013-05-10 1:34 ` Namhyung Kim
2013-05-10 9:06 ` Jiri Olsa
2013-05-09 13:32 ` [PATCH 3/9] perf tests: Add precise event automated test Jiri Olsa
2013-05-09 13:32 ` [PATCH 4/9] perf tools: Add a precise event qualifier Jiri Olsa
2013-05-10 1:43 ` Namhyung Kim
2013-05-10 9:10 ` Jiri Olsa
2013-05-09 13:32 ` [PATCH 5/9] perf tools: Set maximum precise value for event 'p' modifier Jiri Olsa
2013-05-09 19:43 ` David Ahern
2013-05-10 9:12 ` Jiri Olsa
2013-05-10 1:53 ` Namhyung Kim
2013-05-10 9:16 ` Jiri Olsa
2013-05-09 13:32 ` [PATCH 6/9] perf tools: Set maximum precise value for 'precise' term Jiri Olsa
2013-05-09 13:32 ` [PATCH 7/9] perf tests: Add automated precise term test Jiri Olsa
2013-05-09 13:32 ` [PATCH 8/9] perf: Document the ABI for 'precise' sysfs attribute Jiri Olsa
2013-05-10 1:57 ` Namhyung Kim
2013-05-09 13:32 ` [PATCH 9/9] perf: Document the ABI for 'rdpmc' " Jiri Olsa
2013-05-09 15:07 ` [PATCH 0/9] perf: Adding better precise_ip field handling Peter Zijlstra
2013-05-09 15:20 ` Jiri Olsa
2013-05-10 9:27 ` Peter Zijlstra
2013-05-10 9:40 ` Jiri Olsa
2013-05-10 9:53 ` Peter Zijlstra
2013-05-10 10:18 ` Ingo Molnar
2013-05-10 10:22 ` Peter Zijlstra
2013-05-10 10:31 ` Ingo Molnar
2013-05-10 10:34 ` Peter Zijlstra
2013-05-10 10:55 ` Ingo Molnar [this message]
2013-05-10 11:27 ` Peter Zijlstra
2013-05-11 7:50 ` Ingo Molnar
2013-05-13 9:36 ` Peter Zijlstra
2013-05-13 19:43 ` Ingo Molnar
2013-05-14 7:22 ` Peter Zijlstra
2013-05-14 8:37 ` Ingo Molnar
2013-05-14 8:36 ` Gleb Natapov
2013-05-15 13:27 ` Stephane Eranian
2013-05-28 9:54 ` Ingo Molnar
2013-05-10 9:29 ` Peter Zijlstra
2013-05-10 9:43 ` Jiri Olsa
-- strict thread matches above, loose matches on Subject: below --
2013-01-26 17:27 Jiri Olsa
2013-01-26 18:53 ` Jiri Olsa
2013-01-27 12:41 ` Ingo Molnar
2013-01-27 12:57 ` Jiri Olsa
2013-01-27 13:02 ` Ingo Molnar
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=20130510105536.GA18805@gmail.com \
--to=mingo@kernel.org \
--cc=acme@redhat.com \
--cc=ak@linux.intel.com \
--cc=cjashfor@linux.vnet.ibm.com \
--cc=dsahern@gmail.com \
--cc=eranian@google.com \
--cc=fweisbec@gmail.com \
--cc=jolsa@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=namhyung@kernel.org \
--cc=paulus@samba.org \
--cc=peterz@infradead.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).