linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Adrian Hunter <adrian.hunter@intel.com>
To: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Jiri Olsa <jolsa@redhat.com>, linux-kernel@vger.kernel.org
Subject: [PATCH 12/22] perf intel-pt: Document IPC usage
Date: Mon, 20 May 2019 14:37:18 +0300	[thread overview]
Message-ID: <20190520113728.14389-13-adrian.hunter@intel.com> (raw)
In-Reply-To: <20190520113728.14389-1-adrian.hunter@intel.com>

Add brief documentation about instructions-per-cycle (IPC) information
derived from Intel PT.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/Documentation/intel-pt.txt | 30 +++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

diff --git a/tools/perf/Documentation/intel-pt.txt b/tools/perf/Documentation/intel-pt.txt
index 60d99e5e7921..50c5b60101bd 100644
--- a/tools/perf/Documentation/intel-pt.txt
+++ b/tools/perf/Documentation/intel-pt.txt
@@ -103,6 +103,36 @@ The flags are "bcrosyiABEx" which stand for branch, call, return, conditional,
 system, asynchronous, interrupt, transaction abort, trace begin, trace end, and
 in transaction, respectively.
 
+Another interesting field that is not printed by default is 'ipc' which can be
+displayed as follows:
+
+	perf script --itrace=be -F+ipc
+
+There are two ways that instructions-per-cycle (IPC) can be calculated depending
+on the recording.
+
+If the 'cyc' config term (see config terms section below) was used, then IPC is
+calculated using the cycle count from CYC packets, otherwise MTC packets are
+used - refer to the 'mtc' config term.  When MTC is used, however, the values
+are less accurate because the timing is less accurate.
+
+Because Intel PT does not update the cycle count on every branch or instruction,
+the values will often be zero.  When there are values, they will be the number
+of instructions and number of cycles since the last update, and thus represent
+the average IPC since the last IPC for that event type.  Note IPC for "branches"
+events is calculated separately from IPC for "instructions" events.
+
+Also note that the IPC instruction count may or may not include the current
+instruction.  If the cycle count is associated with an asynchronous branch
+(e.g. page fault or interrupt), then the instruction count does not include the
+current instruction, otherwise it does.  That is consistent with whether or not
+that instruction has retired when the cycle count is updated.
+
+Another note, in the case of "branches" events, non-taken branches are not
+presently sampled, so IPC values for them do not appear e.g. a CYC packet with a
+TNT packet that starts with a non-taken branch.  To see every possible IPC
+value, "instructions" events can be used e.g. --itrace=i0ns
+
 While it is possible to create scripts to analyze the data, an alternative
 approach is available to export the data to a sqlite or postgresql database.
 Refer to script export-to-sqlite.py or export-to-postgresql.py for more details,
-- 
2.17.1


  parent reply	other threads:[~2019-05-20 11:38 UTC|newest]

Thread overview: 52+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-05-20 11:37 [PATCH 00/22] perf intel-pt: Add support for instructions-per-cycle (IPC) Adrian Hunter
2019-05-20 11:37 ` [PATCH 01/22] perf intel-pt: Fix itrace defaults for perf script Adrian Hunter
2019-05-20 13:58   ` Sasha Levin
2019-05-20 14:45   ` Arnaldo Carvalho de Melo
2019-05-31 16:45     ` Arnaldo Carvalho de Melo
2019-06-04 11:32       ` Adrian Hunter
2019-06-04 13:20         ` Arnaldo Carvalho de Melo
2019-05-30  7:53   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 02/22] perf auxtrace: " Adrian Hunter
2019-05-30  7:54   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 03/22] perf intel-pt: Fix itrace defaults for perf script intel-pt documentation Adrian Hunter
2019-05-30  7:54   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 04/22] perf intel-pt: Factor out intel_pt_update_sample_time Adrian Hunter
2019-06-17 18:59   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 05/22] perf intel-pt: Accumulate cycle count from CYC packets Adrian Hunter
2019-06-17 18:59   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 06/22] perf tools: Add IPC information to perf_sample Adrian Hunter
2019-06-17 19:00   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 07/22] perf intel-pt: Add support for samples to contain IPC ratio Adrian Hunter
2019-06-17 19:01   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 08/22] perf script: Add output of " Adrian Hunter
2019-06-17 19:01   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 09/22] perf intel-pt: Record when decoding PSB+ packets Adrian Hunter
2019-06-17 19:02   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 10/22] perf intel-pt: Re-factor TIP cases in intel_pt_walk_to_ip Adrian Hunter
2019-06-17 19:03   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 11/22] perf intel-pt: Accumulate cycle count from TSC/TMA/MTC packets Adrian Hunter
2019-06-17 19:04   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` Adrian Hunter [this message]
2019-06-17 19:04   ` [tip:perf/core] perf intel-pt: Document IPC usage tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 13/22] perf thread-stack: Accumulate IPC information Adrian Hunter
2019-06-17 19:05   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 14/22] perf db-export: Add brief documentation Adrian Hunter
2019-06-17 19:06   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 15/22] perf db-export: Export IPC information Adrian Hunter
2019-06-17 19:07   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 16/22] perf scripts python: export-to-sqlite.py: " Adrian Hunter
2019-06-17 19:07   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 17/22] perf scripts python: export-to-postgresql.py: " Adrian Hunter
2019-06-17 19:08   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 18/22] perf scripts python: exported-sql-viewer.py: Add IPC information to the Branch reports Adrian Hunter
2019-05-31 16:44   ` Arnaldo Carvalho de Melo
2019-06-03  5:50     ` Adrian Hunter
2019-06-17 19:09   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 19/22] perf scripts python: exported-sql-viewer.py: Add CallGraphModelParams Adrian Hunter
2019-06-17 19:09   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 20/22] perf scripts python: exported-sql-viewer.py: Add IPC information to Call Graph Graph Adrian Hunter
2019-06-17 19:10   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 21/22] perf scripts python: exported-sql-viewer.py: Add IPC information to Call Tree Adrian Hunter
2019-06-17 19:11   ` [tip:perf/core] " tip-bot for Adrian Hunter
2019-05-20 11:37 ` [PATCH 22/22] perf scripts python: exported-sql-viewer.py: Select find text when find bar is activated Adrian Hunter
2019-06-17 19:11   ` [tip:perf/core] " tip-bot for Adrian Hunter

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=20190520113728.14389-13-adrian.hunter@intel.com \
    --to=adrian.hunter@intel.com \
    --cc=acme@kernel.org \
    --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).