Re: [PATCH] Add failing test for fetching from multiple packs over dumb httpd

[Charles Bailey <charles@hashpling.org>]

On Tue, Jan 27, 2015 at 01:12:21PM -0500, Jeff King wrote:
> On Tue, Jan 27, 2015 at 03:20:41PM +0000, Charles Bailey wrote:
> 
> > From: Charles Bailey <cbailey32@bloomberg.net>
> > 
> > When objects are spread across multiple packs, if an initial fetch does
> > require all pack files, a subsequent fetch for objects in packs not
> > retrieved in the initial fetch will fail.
> 
> s/does/does not/, I think?

Yes, that's definitely what I meant to write.

[...]
> It looks like the culprit is 7b64469 (Allow parse_pack_index on
> temporary files, 2010-04-19). It added a new "idx_path" parameter to
> parse_pack_index, which we pass as NULL.  That causes its call to
> check_packed_git_idx to fail (because it has no idea what file we are
> talking about!).

That change looks like it went into 1.7.1.1. I cannot confirm this
working before then but we've definitely seen the bug in 1.7.12.3 and
more recent versions.

> This seems to fix it:
> 
> diff --git a/sha1_file.c b/sha1_file.c
> index 30995e6..eda4d90 100644
> --- a/sha1_file.c
> +++ b/sha1_file.c
> @@ -1149,6 +1149,9 @@ struct packed_git *parse_pack_index(unsigned char *sha1, const char *idx_path)
>  	const char *path = sha1_pack_name(sha1);
>  	struct packed_git *p = alloc_packed_git(strlen(path) + 1);
>  
> +	if (!idx_path)
> +		idx_path = sha1_pack_index_name(sha1);
> +
>  	strcpy(p->pack_name, path);
>  	hashcpy(p->sha1, sha1);
>  	if (check_packed_git_idx(idx_path, p)) {

It certainly fixes my test script and I can give this patch a test in
the 'real' world.