From: Nicolas Pitre <nico@fluxnic.net>
To: Junio C Hamano <gitster@pobox.com>
Cc: "Andrzej K. Haczewski" <ahaczewski@gmail.com>,
git@vger.kernel.org, Johannes Sixt <j.sixt@viscovery.net>
Subject: [PATCH] pack-objects: move thread autodetection closer to relevant code
Date: Wed, 04 Nov 2009 16:32:46 -0500 (EST) [thread overview]
Message-ID: <alpine.LFD.2.00.0911041623570.10340@xanadu.home> (raw)
In-Reply-To: <16cee31f0911041316n20fc9f12s6595dadc813d8f46@mail.gmail.com>
[-- Attachment #1: Type: TEXT/PLAIN, Size: 1920 bytes --]
Let's keep thread stuff close together if possible. And in this case,
this even reduces the #ifdef noise, and allows for skipping the
autodetection altogether if delta search is not needed (like with a pure
clone).
Signed-off-by: Nicolas Pitre <nico@fluxnic.net>
---
> 2009/11/4 Nicolas Pitre <nico@fluxnic.net>:
> >> @@ -2327,6 +2354,8 @@ int cmd_pack_objects(int argc, const char **argv, const char *prefix)
> >> #ifdef THREADED_DELTA_SEARCH
> >> if (!delta_search_threads) /* --threads=0 means autodetect */
> >> delta_search_threads = online_cpus();
> >> +
> >> + init_threaded_delta_search();
> >
> > What about doing this at the beginning of ll_find_deltas() instead?
> > And similarly for cleanup_threaded_delta_search(): call it right before
> > leaving ll_find_deltas(). This way thread issues would remain more
> > localized. In fact I'd move the whole thing above in ll_find_deltas()
> > as well (separately from this patch though).
So here it is.
diff --git a/builtin-pack-objects.c b/builtin-pack-objects.c
index 02f9246..4c91e94 100644
--- a/builtin-pack-objects.c
+++ b/builtin-pack-objects.c
@@ -1629,6 +1629,8 @@ static void ll_find_deltas(struct object_entry **list, unsigned list_size,
struct thread_params *p;
int i, ret, active_threads = 0;
+ if (!delta_search_threads) /* --threads=0 means autodetect */
+ delta_search_threads = online_cpus();
if (delta_search_threads <= 1) {
find_deltas(list, &list_size, window, depth, processed);
return;
@@ -2324,11 +2326,6 @@ int cmd_pack_objects(int argc, const char **argv, const char *prefix)
if (keep_unreachable && unpack_unreachable)
die("--keep-unreachable and --unpack-unreachable are incompatible.");
-#ifdef THREADED_DELTA_SEARCH
- if (!delta_search_threads) /* --threads=0 means autodetect */
- delta_search_threads = online_cpus();
-#endif
-
prepare_packed_git();
if (progress)
next prev parent reply other threads:[~2009-11-04 21:34 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-11-03 21:30 [PATCH 0/1] Port of pthreads to Windows API threads Andrzej K. Haczewski
2009-11-03 21:30 ` [PATCH 1/1] MSVC: port pthread code to native Windows threads Andrzej K. Haczewski
2009-11-03 23:38 ` Johannes Schindelin
2009-11-04 2:34 ` Joshua Jensen
2009-11-04 7:44 ` Andrzej K. Haczewski
2009-11-04 8:24 ` Johannes Sixt
2009-11-04 11:02 ` Johannes Schindelin
2009-11-04 8:17 ` Andrzej K. Haczewski
2009-11-04 8:15 ` Johannes Sixt
2009-11-04 8:48 ` Michael Wookey
2009-11-04 10:53 ` Andrzej K. Haczewski
2009-11-04 10:37 ` [PATCH] " Andrzej K. Haczewski
2009-11-04 10:50 ` Erik Faye-Lund
2009-11-04 10:56 ` Andrzej K. Haczewski
2009-11-04 11:14 ` Paolo Bonzini
2009-11-04 11:23 ` Paolo Bonzini
2009-11-04 12:39 ` Johannes Sixt
2009-11-04 13:47 ` Andrzej K. Haczewski
2009-11-04 14:34 ` Johannes Sixt
2009-11-04 14:50 ` Andrzej K. Haczewski
2009-11-04 20:43 ` Daniel Barkalow
2009-11-04 21:17 ` Nicolas Pitre
2009-11-04 22:22 ` Daniel Barkalow
2009-11-05 0:27 ` Nicolas Pitre
2009-11-05 13:48 ` Dmitry Potapov
2009-11-04 14:14 ` Andrzej K. Haczewski
2009-11-04 14:19 ` Erik Faye-Lund
2009-11-04 14:04 ` Johannes Schindelin
2009-11-04 15:55 ` [PATCH] MSVC: Windows-native implementation for subset of Pthreads API Andrzej K. Haczewski
2009-11-04 18:10 ` Nicolas Pitre
2009-11-04 21:16 ` Andrzej K. Haczewski
2009-11-04 21:32 ` Nicolas Pitre [this message]
2009-11-06 7:20 ` [PATCH] pack-objects: move thread autodetection closer to relevant code Junio C Hamano
2009-11-04 21:41 ` [PATCH] MSVC: Windows-native implementation for subset of Pthreads API Erik Faye-Lund
2009-11-04 22:50 ` Andrzej K. Haczewski
2009-11-05 2:47 ` Nicolas Pitre
2009-11-05 9:00 ` Andrzej K. Haczewski
2009-11-05 9:41 ` Erik Faye-Lund
2009-11-05 10:18 ` Andrzej K. Haczewski
2009-11-05 12:27 ` Johannes Sixt
2009-11-05 12:53 ` Andrzej K. Haczewski
2009-11-05 19:25 ` Nicolas Pitre
2009-11-05 20:38 ` Andrzej K. Haczewski
2009-11-05 22:15 ` Nicolas Pitre
2009-11-04 21:52 ` Nicolas Pitre
2009-11-04 23:47 ` Andrzej K. Haczewski
2009-11-04 23:57 ` Andrzej K. Haczewski
2009-11-05 0:22 ` Nicolas Pitre
2009-11-05 8:51 ` Andrzej K. Haczewski
2009-11-05 19:22 ` Nicolas Pitre
2009-11-05 2:10 ` Nicolas Pitre
2009-11-05 8:45 ` Andrzej K. Haczewski
2009-11-05 19:17 ` Nicolas Pitre
2009-11-05 7:33 ` Johannes Sixt
2009-11-04 23:58 ` Junio C Hamano
2009-11-05 16:45 ` Andrzej K. Haczewski
2009-11-05 17:31 ` Johannes Sixt
2009-11-05 19:39 ` Nicolas Pitre
2009-11-05 20:09 ` Andrzej K. Haczewski
2009-11-05 20:36 ` Nicolas Pitre
2009-11-06 8:10 ` Andrzej K. Haczewski
2009-11-06 8:25 ` Johannes Sixt
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.LFD.2.00.0911041623570.10340@xanadu.home \
--to=nico@fluxnic.net \
--cc=ahaczewski@gmail.com \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=j.sixt@viscovery.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.
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).