Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Yao Qi <yao@codesourcery.com>
To: <gdb-patches@sourceware.org>
Subject: [PATCH 0/3 V3] Cache code access for disassemble
Date: Thu, 21 Nov 2013 01:34:00 -0000	[thread overview]
Message-ID: <1384996594-20865-1-git-send-email-yao@codesourcery.com> (raw)

Hi,
This is the V3 of this patch series.  Most of the patches were
committed because they are cleanups, and V3 only has three patches.

Patch 1/3 is to rename some functions and variables related to
"stack-cache", to address review comments in V2 that we find they are
improper.  Patch 2/3 is to add commands "set/show code-cache", and we
choose a way that "both stack cache and code cache use the single global
dcache, and transitions of either options will invalidate the dcache".
Patch 3/3 is unchanged.

The performance improvement is measured by gdb.perf/disassemble.exp,
posted https://sourceware.org/ml/gdb-patches/2013-11/msg00574.html
(note that gdb.perf/disassemble.py is modified to flush code cache after
each run)

			      Original Patched
disassemble cpu_time	0	0.49	0.16
disassemble cpu_time	1	0.93	0.21
disassemble cpu_time	2	1.35	0.33
disassemble wall_time	0	0.722337007523	0.167083024979
disassemble wall_time	1	1.37996888161	0.215560913086
disassemble wall_time	2	2.07086896896	0.323045969009
disassemble vmsize	0	151000	150880
disassemble vmsize	1	151692	151740
disassemble vmsize	2	151692	151740

Regression tested x86_64-linux.

*** BLURB HERE ***

Yao Qi (3):
  Renaming in target-dcache.c
  set/show code-cache
  Use target_read_code in disassemble.

 gdb/NEWS            |    6 ++++
 gdb/disasm.c        |    2 +-
 gdb/doc/gdb.texinfo |   16 +++++++++-
 gdb/target-dcache.c |   81 +++++++++++++++++++++++++++++++++++++++++---------
 gdb/target-dcache.h |    4 ++-
 gdb/target.c        |   23 ++++++++++++--
 gdb/target.h        |    5 +++
 7 files changed, 115 insertions(+), 22 deletions(-)

-- 
1.7.7.6


             reply	other threads:[~2013-11-21  1:18 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-11-21  1:34 Yao Qi [this message]
2013-11-21  1:18 ` [PATCH 3/3] Use target_read_code in disassemble Yao Qi
2013-11-21  2:56   ` Doug Evans
2013-11-21  1:18 ` [PATCH 1/3] Renaming in target-dcache.c Yao Qi
2013-11-21  1:47   ` Doug Evans
2013-11-21  3:05     ` Yao Qi
2013-11-21  1:18 ` [PATCH 2/3] set/show code-cache Yao Qi
2013-11-21  3:33   ` Yao Qi
2013-11-21  5:06     ` Eli Zaretskii
2013-11-21 12:21     ` Pedro Alves
2013-11-21 14:03       ` Yao Qi
2013-11-21 16:13         ` Eli Zaretskii
2013-11-22  1:51           ` Yao Qi
2013-11-21 20:50       ` Doug Evans
2013-11-21 19:38     ` Tom Tromey
2013-11-21 21:06       ` Doug Evans
2013-11-21 21:17         ` Tom Tromey
2013-11-22  0:38           ` Doug Evans
2013-11-22  2:19           ` Yao Qi
2013-11-22  7:43             ` Eli Zaretskii
2013-11-24  9:02 ` [PATCH 0/3 V3] Cache code access for disassemble Yao Qi

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=1384996594-20865-1-git-send-email-yao@codesourcery.com \
    --to=yao@codesourcery.com \
    --cc=gdb-patches@sourceware.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