git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jonathan Nieder <jrnieder@gmail.com>
To: Junio C Hamano <gitster@pobox.com>
Cc: git@vger.kernel.org, Andreas Krey <a.krey@gmx.de>
Subject: [PATCH] cherry-pick: do not dump core when iconv fails
Date: Sat, 8 May 2010 18:17:29 -0500	[thread overview]
Message-ID: <20100508231729.GA6718@progeny.tock> (raw)
In-Reply-To: <20100430133228.GA1620@progeny.tock>

When cherry-picking, usually the new and old commit encodings are both
UTF-8.  Most old iconv implementations do not support this trivial
conversion, so on old platforms, out->message remains NULL, and later
attempts to read it segfault.

Fix this by noticing the input and output encodings match and skipping
the iconv step, like the other reencode_string() call sites already do.
Also stop segfaulting on other iconv failures: if iconv fails for some
other reason, the best we can do is to pass the old message through.

This fixes a regression introduced in v1.7.1-rc0~15^2~2 (revert:
clarify label on conflict hunks, 2010-03-20).

Reported-by: Andreas Krey <a.krey@gmx.de>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
---
Andreas reported this cherry-pick regression about a week ago, and a
patch similar to this one seemed to fix it.  The only question that
made me stall was what to do if iconv just doesn’t understand an
encoding:

 - from a certain point of view it might make sense to pass through
   the message and note the old encoding

 - but on the other hand, that would not respect the user’s preference
   as expressed in i18n.commitencoding, and it would deny the caller
   the chance to fix the encoding problem in a text editor before
   commiting.

That second point was driven home when I tried to implement this: it
required teaching ‘git commit’ to respect a new GIT_COMMIT_ENCODING
variable, but this was a pain (there is a sizeable amount of code
paying attention to i18n.commitencoding) and the result would have
been cherry-pick ignoring the $GIT_COMMIT_ENCODING preference,
creating encoding mismatches between ident and commit message to boot.

Everyone should be using utf8 anyway. :)

Anyway, this minimal fix should be safe (it just restores the old
behavior, except we do not use iconv for trivial conversions any
more).  Patch is against v1.7.1-rc0~15^2~2.

Thanks to Andreas for the help.

 builtin/revert.c |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/builtin/revert.c b/builtin/revert.c
index 5a5b721..5adaf27 100644
--- a/builtin/revert.c
+++ b/builtin/revert.c
@@ -97,8 +97,13 @@ static int get_message(const char *raw_message, struct commit_message *out)
 		encoding = "UTF-8";
 	if (!git_commit_encoding)
 		git_commit_encoding = "UTF-8";
-	if ((out->reencoded_message = reencode_string(raw_message,
-					git_commit_encoding, encoding)))
+
+	out->reencoded_message = NULL;
+	out->message = raw_message;
+	if (strcmp(encoding, git_commit_encoding))
+		out->reencoded_message = reencode_string(raw_message,
+					git_commit_encoding, encoding);
+	if (out->reencoded_message)
 		out->message = out->reencoded_message;
 
 	abbrev = find_unique_abbrev(commit->object.sha1, DEFAULT_ABBREV);
-- 
1.7.1

  reply	other threads:[~2010-05-08 23:17 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-04-28 19:38 Proper way to abort incorrect cherry-picking? Eugene Sajine
2010-04-28 19:49 ` David Borowitz
2010-04-28 19:59   ` Eugene Sajine
2010-04-28 22:39     ` Jon Seymour
2010-04-28 23:37       ` Jonathan Nieder
2010-04-29  0:07         ` Jon Seymour
2010-04-29 19:11         ` git cherry(pick) dumps core Andreas Krey
2010-04-29 19:49           ` Jonathan Nieder
2010-04-29 20:21             ` Andreas Krey
2010-04-30 13:32               ` Jonathan Nieder
2010-05-08 23:17                 ` Jonathan Nieder [this message]
2010-05-08 23:55                   ` [PATCH] cherry-pick: do not dump core when iconv fails Junio C Hamano
2010-04-28 19:50 ` Proper way to abort incorrect cherry-picking? Jonathan Nieder
2010-04-28 20:05   ` Eugene Sajine

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=20100508231729.GA6718@progeny.tock \
    --to=jrnieder@gmail.com \
    --cc=a.krey@gmx.de \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.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.
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).