git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Jeff King <peff@peff.net>
To: "Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
Cc: git@vger.kernel.org, "Junio C Hamano" <gitster@pobox.com>,
	"Nguyễn Thái Ngọc Duy" <pclouds@gmail.com>,
	"Michael Haggerty" <mhagger@alum.mit.edu>,
	"Stefan Beller" <stefanbeller@gmail.com>,
	"Jonathan Nieder" <jrnieder@gmail.com>
Subject: Re: [PATCH 4/5] gc: don't run "reflog expire" when keeping reflogs
Date: Wed, 13 Mar 2019 20:40:06 -0400	[thread overview]
Message-ID: <20190314004006.GI31968@sigill.intra.peff.net> (raw)
In-Reply-To: <20190313235439.30439-5-avarab@gmail.com>

On Thu, Mar 14, 2019 at 12:54:38AM +0100, Ævar Arnfjörð Bjarmason wrote:

> Don't redundantly run "git reflog expire --all" when gc.reflogExpire
> and gc.reflogExpireUnreachable are set to "never".
> 
> I'm being careful to not repeat the issue fixed in
> 8ab5aa4bd8 ("parseopt: handle malformed --expire arguments more
> nicely", 2018-04-21). We'll die early if the config variables are set
> to invalid values.

I think the general sentiment here makes sense, that reflog expiration
should be a noop if we're set to "never" anyway.

It does feel a little funny for "gc" to be peeking into the internals of
how "reflog" will make its decision about how to run, though. I doubt
those rules are likely to change, but if they did, there'd be very
little to remind somebody working on reflog that they need to change the
matching rules here.

Is our problem running "git reflog" at all, or is it that "git reflog"
does too much work even when it's been told "never"? If the latter,
could we just have it exit early as soon as it's figured out the prune
expiry dates it's using?

I.e., something like:

diff --git a/builtin/reflog.c b/builtin/reflog.c
index 4d3430900d..aab221f315 100644
--- a/builtin/reflog.c
+++ b/builtin/reflog.c
@@ -594,6 +594,15 @@ static int cmd_reflog_expire(int argc, const char **argv, const char *prefix)
 			break;
 	}
 
+	/*
+	 * If we're not expiring anything and not dropping stale entries,
+	 * there's no point in even opening the reflogs, since we're guaranteed
+	 * to do nothing.
+	 */
+	if (!cb.cmd.expire_total && !cb.cmd.expire_unreachable &&
+	    !cb.cmd.stalefix)
+		return 0;
+
 	/*
 	 * We can trust the commits and objects reachable from refs
 	 * even in older repository.  We cannot trust what's reachable

Seeing "--stale-fix" again reminded me: that may be the "oops, we can
spend tons of CPU walking history" bit that I mentioned in the other
part of the thread. But I don't think git-gc would ever run with that
option.

-Peff

  reply	other threads:[~2019-03-14  0:41 UTC|newest]

Thread overview: 60+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-12 23:28 BUG: Race condition due to reflog expiry in "gc" Ævar Arnfjörð Bjarmason
2019-03-13  1:43 ` Junio C Hamano
2019-03-13 10:28   ` Ævar Arnfjörð Bjarmason
2019-03-13 16:02     ` Jeff King
2019-03-13 16:22       ` Ævar Arnfjörð Bjarmason
2019-03-13 23:54         ` [PATCH 0/5] gc: minor code cleanup + contention fixes Ævar Arnfjörð Bjarmason
2019-03-14 12:34           ` [PATCH v2 0/7] " Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 0/8] " Ævar Arnfjörð Bjarmason
2019-03-18  6:13               ` Junio C Hamano
2019-03-28 16:14               ` [PATCH v4 0/7] gc: tests and handle reflog expire config Ævar Arnfjörð Bjarmason
2019-03-28 16:14               ` [PATCH v4 1/7] gc: remove redundant check for gc_auto_threshold Ævar Arnfjörð Bjarmason
2019-03-28 16:14               ` [PATCH v4 2/7] gc: convert to using the_hash_algo Ævar Arnfjörð Bjarmason
2019-03-28 16:14               ` [PATCH v4 3/7] gc: refactor a "call me once" pattern Ævar Arnfjörð Bjarmason
2019-03-28 16:14               ` [PATCH v4 4/7] reflog tests: make use of "test_config" idiom Ævar Arnfjörð Bjarmason
2019-03-28 16:14               ` [PATCH v4 5/7] reflog tests: test for the "points nowhere" warning Ævar Arnfjörð Bjarmason
2019-03-28 16:14               ` [PATCH v4 6/7] reflog tests: assert lack of early exit with expiry="never" Ævar Arnfjörð Bjarmason
2019-03-28 16:14               ` [PATCH v4 7/7] gc: handle & check gc.reflogExpire config Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 1/8] gc: remove redundant check for gc_auto_threshold Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 2/8] gc: convert to using the_hash_algo Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 3/8] gc: refactor a "call me once" pattern Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 4/8] reflog tests: make use of "test_config" idiom Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 5/8] reflog tests: test for the "points nowhere" warning Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 6/8] reflog tests: assert lack of early exit with expiry="never" Ævar Arnfjörð Bjarmason
2019-03-19  6:20               ` Jeff King
2019-03-15 15:59             ` [PATCH v3 7/8] gc: handle & check gc.reflogExpire config Ævar Arnfjörð Bjarmason
2019-03-15 15:59             ` [PATCH v3 8/8] reflog expire: don't assert the OID when locking refs Ævar Arnfjörð Bjarmason
     [not found]               ` <b870a17d-2103-41b8-3cbc-7389d5fff33a@alum.mit.edu>
2019-03-21 14:10                 ` Ævar Arnfjörð Bjarmason
2019-03-19  6:27             ` [PATCH v2 0/7] gc: minor code cleanup + contention fixes Jeff King
2019-03-14 12:34           ` [PATCH v2 1/7] gc: remove redundant check for gc_auto_threshold Ævar Arnfjörð Bjarmason
2019-03-14 12:34           ` [PATCH v2 2/7] gc: convert to using the_hash_algo Ævar Arnfjörð Bjarmason
2019-03-15  9:51             ` Duy Nguyen
2019-03-15 10:42               ` Ævar Arnfjörð Bjarmason
2019-03-15 15:49                 ` Johannes Schindelin
2019-03-14 12:34           ` [PATCH v2 3/7] gc: refactor a "call me once" pattern Ævar Arnfjörð Bjarmason
2019-03-14 12:34           ` [PATCH v2 4/7] reflog tests: make use of "test_config" idiom Ævar Arnfjörð Bjarmason
2019-03-14 12:34           ` [PATCH v2 5/7] reflog: exit early if there's no work to do Ævar Arnfjörð Bjarmason
2019-03-15 10:01             ` Duy Nguyen
2019-03-15 15:51               ` Ævar Arnfjörð Bjarmason
2019-03-14 12:34           ` [PATCH v2 6/7] gc: don't run "reflog expire" when keeping reflogs Ævar Arnfjörð Bjarmason
2019-03-15 10:05             ` Duy Nguyen
2019-03-15 10:24               ` Ævar Arnfjörð Bjarmason
2019-03-15 10:32                 ` Duy Nguyen
2019-03-15 15:51                 ` Johannes Schindelin
2019-03-18  6:04                 ` Junio C Hamano
2019-03-14 12:34           ` [PATCH v2 7/7] reflog expire: don't assert the OID when locking refs Ævar Arnfjörð Bjarmason
2019-03-15 11:10             ` Duy Nguyen
2019-03-15 15:52               ` Ævar Arnfjörð Bjarmason
2019-03-13 23:54         ` [PATCH 1/5] gc: remove redundant check for gc_auto_threshold Ævar Arnfjörð Bjarmason
2019-03-14  0:25           ` Jeff King
2019-03-13 23:54         ` [PATCH 2/5] gc: convert to using the_hash_algo Ævar Arnfjörð Bjarmason
2019-03-14  0:26           ` Jeff King
2019-03-13 23:54         ` [PATCH 3/5] gc: refactor a "call me once" pattern Ævar Arnfjörð Bjarmason
2019-03-14  0:30           ` Jeff King
2019-03-13 23:54         ` [PATCH 4/5] gc: don't run "reflog expire" when keeping reflogs Ævar Arnfjörð Bjarmason
2019-03-14  0:40           ` Jeff King [this message]
2019-03-14  4:51             ` Junio C Hamano
2019-03-14 19:26               ` Jeff King
2019-03-13 23:54         ` [PATCH 5/5] reflog expire: don't assert the OID when locking refs Ævar Arnfjörð Bjarmason
2019-03-14  0:44           ` Jeff King
2019-03-14  0:25         ` BUG: Race condition due to reflog expiry in "gc" Jeff King

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=20190314004006.GI31968@sigill.intra.peff.net \
    --to=peff@peff.net \
    --cc=avarab@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=jrnieder@gmail.com \
    --cc=mhagger@alum.mit.edu \
    --cc=pclouds@gmail.com \
    --cc=stefanbeller@gmail.com \
    /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).