git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH 2/2] remote: fix trivial memory leak

From
Jeff King <peff@peff.net>
Date
Sep 24, 2013, 05:19 UTC
Message-ID
<20130924051908.GG2766@sigill.intra.peff.net>
In-Reply-To
<1379772563-11000-3-git-send-email-felipe.contreras@gmail.com>
On Sat, Sep 21, 2013 at 09:09:23AM -0500, Felipe Contreras wrote:
Show 19 quoted lines
> diff --git a/remote.c b/remote.c
> index efcba93..654e7f5 100644
> --- a/remote.c
> +++ b/remote.c
> @@ -480,7 +480,6 @@ static void read_config(void)
>  	int flag;
>  	if (default_remote_name) /* did this already */
>  		return;
> -	default_remote_name = xstrdup("origin");
>  	current_branch = NULL;
>  	head_ref = resolve_ref_unsafe("HEAD", sha1, 0, &flag);
>  	if (head_ref && (flag & REF_ISSYMREF) &&
> @@ -489,6 +488,8 @@ static void read_config(void)
>  			make_branch(head_ref + strlen("refs/heads/"), 0);
>  	}
>  	git_config(handle_config, NULL);
> +	if (!default_remote_name)
> +		default_remote_name = xstrdup("origin");
>  	alias_all_urls();

I wondered if we might also leak when seeing duplicate config options (i.e., leaking the old one when replacing it with the new). But we don't actually strdup() the configured remote names, but instead just point into the "struct branch", which owns the data.

So I think an even better fix would be:
-- >8 --
Subject: remote: do not copy "origin" string literal

Our default_remote_name starts at "origin", but may be overridden by the config file. In the former case, we allocate a new string, but in the latter case, we point to the remote name in an existing "struct branch".

This gives the variable inconsistent free() semantics (we are sometimes responsible for freeing the string and sometimes pointing to somebody else's storage), and causes a small leak when the allocated string is overridden by config.

We can fix both by simply dropping the extra copy and pointing to the string literal.

Noticed-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Jeff King <peff@peff.net>
---
 remote.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/remote.c b/remote.c
index e9fedfa..9f1a8aa 100644
--- a/remote.c
+++ b/remote.c
@@ -483,7 +483,7 @@ static void read_config(void)
 	int flag;
 	if (default_remote_name) /* did this already */
 		return;
-	default_remote_name = xstrdup("origin");
+	default_remote_name = "origin";
 	current_branch = NULL;
 	head_ref = resolve_ref_unsafe("HEAD", sha1, 0, &flag);
 	if (head_ref && (flag & REF_ISSYMREF) &&
-- 
1.8.4.rc3.19.g9da5bf6
Previous: Felipe ContrerasNext: Junio C Hamano
Message 14 of 15 in “fetch: trivial fixes”
  1. 0/2 fetch: trivial fixesFelipe Contreras, Sep 21, 2013
  2. 1/2 fetch: add missing documentationFelipe Contreras, Sep 21, 2013
  3. Jeff KingSep 24, 2013
  4. Felipe ContrerasSep 24, 2013
  5. Jeff KingSep 24, 2013
  6. Felipe ContrerasSep 24, 2013
  7. Jeff KingSep 24, 2013
  8. Felipe ContrerasSep 24, 2013
  9. Jeff KingSep 24, 2013
  10. Felipe ContrerasSep 24, 2013
  11. Jeff KingSep 24, 2013
  12. Felipe ContrerasSep 24, 2013
  13. 2/2 remote: fix trivial memory leakFelipe Contreras, Sep 21, 2013
  14. Jeff KingSep 24, 2013
  15. Junio C HamanoOct 15, 2013

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.