From: git@jeffhostetler.com
To: git@vger.kernel.org
Cc: gitster@pobox.com, peff@peff.net,
Jeff Hostetler <jeffhost@microsoft.com>
Subject: [PATCH 0/6] thread lazy_init_name_hash
Date: Wed, 22 Mar 2017 17:14:19 +0000 [thread overview]
Message-ID: <1490202865-31325-1-git-send-email-git@jeffhostetler.com> (raw)
From: Jeff Hostetler <jeffhost@microsoft.com>
This patch series is a performance optimization for
lazy_init_name_hash() in name-hash.c on very large
repositories.
This change allows lazy_init_name_hash() to optionally
use multiple threads when building the the_index.dir_hash
and the_index.name_hash hashmaps. The original code path
has been preserved and is used when the repo is small or
the system does not have sufficient CPUs.
A helper command (t/helper/test-lazy-init-name-hash) was
created to demonstrate performance differences and validate
output. For example, use the '-p' option to compare both
code paths on a large repo.
During our testing on the Windows source tree (3.1M
files, 500K folders, 450MB index), this change reduced
the runtime of lazy_init_name_hash() from 1.4 to 0.27
seconds.
This patch series replaces my earlier
* jh/memihash-opt (2017-02-17) 5 commits
patch series.
Jeff Hostetler (6):
name-hash: specify initial size for istate.dir_hash table
hashmap: allow memihash computation to be continued
hashmap: Add disallow_rehash setting
name-hash: perf improvement for lazy_init_name_hash
name-hash: add test-lazy-init-name-hash
name-hash: add perf test for lazy_init_name_hash
Makefile | 1 +
cache.h | 1 +
hashmap.c | 29 ++-
hashmap.h | 25 ++
name-hash.c | 490 +++++++++++++++++++++++++++++++++++-
t/helper/test-lazy-init-name-hash.c | 264 +++++++++++++++++++
t/perf/p0004-lazy-init-name-hash.sh | 19 ++
7 files changed, 820 insertions(+), 9 deletions(-)
create mode 100644 t/helper/test-lazy-init-name-hash.c
create mode 100644 t/perf/p0004-lazy-init-name-hash.sh
--
2.7.4
next reply other threads:[~2017-03-22 17:16 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-03-22 17:14 git [this message]
2017-03-22 17:14 ` [PATCH 1/6] name-hash: specify initial size for istate.dir_hash table git
2017-03-22 17:14 ` [PATCH 2/6] hashmap: allow memihash computation to be continued git
2017-03-22 17:14 ` [PATCH 3/6] hashmap: Add disallow_rehash setting git
2017-03-22 17:14 ` [PATCH 4/6] name-hash: perf improvement for lazy_init_name_hash git
2017-03-22 17:14 ` [PATCH 5/6] name-hash: add test-lazy-init-name-hash git
2017-03-22 17:14 ` [PATCH 6/6] name-hash: add perf test for lazy_init_name_hash git
2017-03-22 18:02 ` [PATCH 0/6] thread lazy_init_name_hash Stefan Beller
2017-03-22 19:22 ` Jeff Hostetler
2017-03-22 19:38 ` Junio C Hamano
2017-03-22 20:54 ` Junio C Hamano
2017-03-22 21:04 ` Jeff Hostetler
2017-03-22 21:29 ` Junio C Hamano
2017-03-22 20:39 ` Junio C Hamano
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
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1490202865-31325-1-git-send-email-git@jeffhostetler.com \
--to=git@jeffhostetler.com \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=jeffhost@microsoft.com \
--cc=peff@peff.net \
/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.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
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 read-only IMAP folder(s) and NNTP newsgroup(s).