git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Junio C Hamano <gitster@pobox.com>
To: Jeff King <peff@peff.net>
Cc: Eric Sunshine <sunshine@sunshineco.com>,
	Stefan Beller <sbeller@google.com>,
	Git List <git@vger.kernel.org>
Subject: Re: [PATCH 4/6] abbrev_sha1_in_line: don't leak memory
Date: Wed, 30 Mar 2016 10:06:41 -0700	[thread overview]
Message-ID: <xmqqzitfuc7y.fsf@gitster.mtv.corp.google.com> (raw)
In-Reply-To: <20160330013127.GA6680@sigill.intra.peff.net> (Jeff King's message of "Tue, 29 Mar 2016 21:31:28 -0400")

Jeff King <peff@peff.net> writes:

> On Tue, Mar 29, 2016 at 09:30:38PM -0400, Eric Sunshine wrote:
>
>> The implementation of strbuf_list_free() is this:
>> 
>>     struct strbuf **s = sbs;
>>     while (*s) {
>>         strbuf_release(*s);
>>         free(*s++);
>>     }
>>     free(sbs);
>> 
>> which means that wt-status.c is leaking not only 'split', but also
>> each element of split[], right?
>
> Yeah, I didn't notice that, but I think you're right.

Correct.

I suspect that we would be better off in the longer term if
we made conscious effort to deprecate and eradicate the use
of strbuf_split* API functions.  They are easy to write
crappy code with, inefficient, and error prone.  You would
rarely need to have N resulting pieces as strbufs to be
easily manipulatable [*1*].

The function can be written by not using the "split and then
rebuild" pattern, perhaps like so, and the result is even
easier to read and understand, I would say.  A sample rewrite
is attached at the end.


[Footnote]

*1* I wouldn't be this harsh if the function were to fill an
    array of pointers or offsets into the original string.

 wt-status.c | 43 +++++++++++++++++++------------------------
 1 file changed, 19 insertions(+), 24 deletions(-)

diff --git a/wt-status.c b/wt-status.c
index ef74864..4886c35 100644
--- a/wt-status.c
+++ b/wt-status.c
@@ -1037,35 +1037,30 @@ static int split_commit_in_progress(struct wt_status *s)
  */
 static void abbrev_sha1_in_line(struct strbuf *line)
 {
-	struct strbuf **split;
-	int i;
+	const char *cp, *ep, *abbrev;
+	unsigned char sha1[20];
 
-	if (starts_with(line->buf, "exec ") ||
-	    starts_with(line->buf, "x "))
+	/* the oddball "exec" does not have 40-hex as the second token */
+	if (starts_with(line->buf, "exec ") || starts_with(line->buf, "x "))
 		return;
 
-	split = strbuf_split_max(line, ' ', 3);
-	if (split[0] && split[1]) {
-		unsigned char sha1[20];
-		const char *abbrev;
+	/* find the second token on the line */
+	cp = strchr(line->buf, ' ');
+	if (!cp)
+		return;
+	cp++;
+	ep = strchr(cp, ' ');
+	if (!ep)
+		return;
 
-		/*
-		 * strbuf_split_max left a space. Trim it and re-add
-		 * it after abbreviation.
-		 */
-		strbuf_trim(split[1]);
-		if (!get_sha1(split[1]->buf, sha1)) {
-			abbrev = find_unique_abbrev(sha1, DEFAULT_ABBREV);
-			strbuf_reset(split[1]);
-			strbuf_addf(split[1], "%s ", abbrev);
-			strbuf_reset(line);
-			for (i = 0; split[i]; i++)
-				strbuf_addf(line, "%s", split[i]->buf);
-		}
-	}
-	for (i = 0; split[i]; i++)
-		strbuf_release(split[i]);
+	/* is it 40-hex? */
+	if (ep - cp != GIT_SHA1_HEXSZ || get_sha1_hex(cp, sha1))
+		return;
 
+	/* replace it with its abbreviation */
+	abbrev = find_unique_abbrev(sha1, DEFAULT_ABBREV);
+	strbuf_splice(line, cp - line->buf, GIT_SHA1_HEXSZ,
+		      abbrev, strlen(abbrev));
 }
 
 static void read_rebase_todolist(const char *fname, struct string_list *lines)

  reply	other threads:[~2016-03-30 17:06 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-30  0:38 [PATCH 0/6] Some cleanups Stefan Beller
2016-03-30  0:38 ` [PATCH 1/6] path.c: allocate enough memory for string Stefan Beller
2016-03-30  0:56   ` Junio C Hamano
2016-03-30  0:57   ` Eric Sunshine
2016-03-30 16:41     ` Stefan Beller
2016-03-30 17:16       ` Junio C Hamano
2016-03-30  0:38 ` [PATCH 2/6] imap-send.c, cram: allocate enough memory for null terminated string Stefan Beller
2016-03-30  1:02   ` Eric Sunshine
2016-03-30  1:07   ` Jeff King
2016-03-30  0:38 ` [PATCH 3/6] notes: don't leak memory in git_config_get_notes_strategy Stefan Beller
2016-03-30  1:11   ` Eric Sunshine
2016-03-30  1:13   ` Jeff King
2016-03-30 17:17     ` Junio C Hamano
2016-03-30  0:38 ` [PATCH 4/6] abbrev_sha1_in_line: don't leak memory Stefan Beller
2016-03-30  1:11   ` Jeff King
2016-03-30  1:30     ` Eric Sunshine
2016-03-30  1:31       ` Jeff King
2016-03-30 17:06         ` Junio C Hamano [this message]
2016-03-30 17:21           ` Jeff King
2016-03-30  0:38 ` [PATCH 5/6] bundle: don't leak an fd in case of early return Stefan Beller
2016-03-30  1:17   ` Jeff King
2016-03-30 17:19     ` Junio C Hamano
2016-03-30  0:38 ` [PATCH 6/6] credential-cache, send_request: close fd when done Stefan Beller
2016-03-30  1:20   ` 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=xmqqzitfuc7y.fsf@gitster.mtv.corp.google.com \
    --to=gitster@pobox.com \
    --cc=git@vger.kernel.org \
    --cc=peff@peff.net \
    --cc=sbeller@google.com \
    --cc=sunshine@sunshineco.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).