From: Vince Weaver <vweaver1@eecs.utk.edu>
To: Ingo Molnar <mingo@elte.hu>
Cc: Avi Kivity <avi@redhat.com>,
Robert Richter <robert.richter@amd.com>,
Benjamin Block <bebl@mageta.org>,
Hans Rosenfeld <hans.rosenfeld@amd.com>, <hpa@zytor.com>,
<tglx@linutronix.de>, <suresh.b.siddha@intel.com>,
<eranian@google.com>, <brgerst@gmail.com>,
<Andreas.Herrmann3@amd.com>, <x86@kernel.org>,
<linux-kernel@vger.kernel.org>,
Benjamin Block <benjamin.block@amd.com>
Subject: Re: [RFC 4/5] x86, perf: implements lwp-perf-integration (rc1)
Date: Tue, 20 Dec 2011 10:48:26 -0500 [thread overview]
Message-ID: <alpine.DEB.2.00.1112201040190.11369@cl320.eecs.utk.edu> (raw)
In-Reply-To: <20111220091511.GB3091@elte.hu>
On Tue, 20 Dec 2011, Ingo Molnar wrote:
> Granted, LWP was mis-designed to quite a degree, those AMD chip
> engineers should have talked to people who understand how modern
> PMU abstractions are added to the OS kernel properly.
You do realize that LWP was probably in design 5+ years ago, at a time
when most Linux kernel developers wanted nothing to do with perf counters,
and thus anyone they did contact for help would have been from the
since-rejected perfctr or perfmon2 camp.
Also, I'm sure Linux isn't the only Operating System that they had in mind
when designing this functionality.
Running LWP through the kernel is a foolish idea. Does anyone have any
numbers on what that would do to overhead?
perf_events creates huge overhead when doing self monitoring. For simple
self-monintoring counter reads it is an *order of magnitude* worse than
doing the same thing with perfctr.
(see numbers here if you don't believe me:
http://web.eecs.utk.edu/~vweaver1/projects/perf-events/benchmarks/rdtsc_overhead/ )
Vince
next prev parent reply other threads:[~2011-12-20 15:48 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-11-29 12:41 [PATCH 0/9] rework of extended state handling, LWP support Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 1/9] x86, xsave: warn on #NM exceptions caused by the kernel Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 2/9] x86, xsave: cleanup fpu/xsave support Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 3/9] x86, xsave: cleanup fpu/xsave signal frame setup Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 4/9] x86, xsave: rework fpu/xsave support Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 5/9] x86, xsave: remove unused code Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 6/9] x86, xsave: more cleanups Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 7/9] x86, xsave: remove lazy allocation of xstate area Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 8/9] x86, xsave: add support for non-lazy xstates Hans Rosenfeld
2011-11-29 12:41 ` [PATCH 9/9] x86, xsave: add kernel support for AMDs Lightweight Profiling (LWP) Hans Rosenfeld
2011-11-29 21:31 ` [PATCH 0/9] rework of extended state handling, LWP support Andi Kleen
2011-11-30 17:37 ` Hans Rosenfeld
2011-11-30 21:52 ` Andi Kleen
2011-12-01 20:36 ` Hans Rosenfeld
2011-12-02 2:01 ` H. Peter Anvin
2011-12-02 11:20 ` Hans Rosenfeld
2011-12-07 19:57 ` Hans Rosenfeld
2011-12-07 20:00 ` [PATCH 7/8] x86, xsave: add support for non-lazy xstates Hans Rosenfeld
2011-12-07 20:00 ` [PATCH 8/8] x86, xsave: add kernel support for AMDs Lightweight Profiling (LWP) Hans Rosenfeld
2011-12-05 10:22 ` [PATCH 0/9] rework of extended state handling, LWP support Ingo Molnar
2011-12-16 16:07 ` Hans Rosenfeld
2011-12-16 16:12 ` [RFC 1/5] x86, perf: Implement software-activation of lwp Hans Rosenfeld
2011-12-16 16:12 ` [RFC 2/5] perf: adds prototype for a new perf-context-type Hans Rosenfeld
2011-12-16 16:12 ` [RFC 3/5] perf: adds a new pmu-initialization-call Hans Rosenfeld
2011-12-16 16:12 ` [RFC 4/5] x86, perf: implements lwp-perf-integration (rc1) Hans Rosenfeld
2011-12-18 8:04 ` Ingo Molnar
2011-12-18 15:22 ` Benjamin Block
2011-12-18 23:43 ` Ingo Molnar
2011-12-19 9:09 ` Robert Richter
2011-12-19 10:54 ` Ingo Molnar
2011-12-19 11:12 ` Avi Kivity
2011-12-19 11:40 ` Ingo Molnar
2011-12-19 11:58 ` Avi Kivity
2011-12-19 18:13 ` Benjamin
2011-12-20 8:56 ` Ingo Molnar
2011-12-20 9:15 ` Ingo Molnar
2011-12-20 9:47 ` Avi Kivity
2011-12-20 10:09 ` Ingo Molnar
2011-12-20 15:27 ` Joerg Roedel
2011-12-20 18:40 ` Ingo Molnar
2011-12-21 0:07 ` Joerg Roedel
2011-12-21 12:34 ` Ingo Molnar
2011-12-21 12:44 ` Avi Kivity
2011-12-21 13:22 ` Ingo Molnar
2011-12-21 22:49 ` Joerg Roedel
2011-12-23 10:53 ` Ingo Molnar
2011-12-21 11:46 ` Gleb Natapov
2011-12-23 10:56 ` Ingo Molnar
2011-12-20 15:48 ` Vince Weaver [this message]
2011-12-20 18:27 ` Ingo Molnar
2011-12-20 22:47 ` Vince Weaver
2011-12-21 12:00 ` Ingo Molnar
2011-12-21 13:55 ` Vince Weaver
2011-12-16 16:12 ` [RFC 5/5] x86, perf: adds support for the LWP threshold-int Hans Rosenfeld
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=alpine.DEB.2.00.1112201040190.11369@cl320.eecs.utk.edu \
--to=vweaver1@eecs.utk.edu \
--cc=Andreas.Herrmann3@amd.com \
--cc=avi@redhat.com \
--cc=bebl@mageta.org \
--cc=benjamin.block@amd.com \
--cc=brgerst@gmail.com \
--cc=eranian@google.com \
--cc=hans.rosenfeld@amd.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=robert.richter@amd.com \
--cc=suresh.b.siddha@intel.com \
--cc=tglx@linutronix.de \
--cc=x86@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.