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

Re: remote-curl: segfault parsing remote.<name>.fetch outside a repository

From
Junio C Hamano <gitster@pobox.com>
Date
Mar 22, 2026, 01:20 UTC
Message-ID
<xmqqa4w0acdr.fsf@gitster.g>
In-Reply-To
<20260321210602.GA736981@coredump.intra.peff.net>
Jeff King <peff@peff.net> writes:
Show 8 quoted lines
> It's a curious case, though. The crashing code is parse_refspec() does
> this:
>
>   if (llen == the_hash_algo->hexsz && !get_oid_hex(item->src, &unused))
>         item->exact_sha1 = 1; /* ok */
>
> But what is the correct hash algo to use here when we are outside a
> repository?

Hmph, who is calling into the transport outside a repository in the first place? Even "git clone" should create the receiving repository before it calls into the transport, no? Is this "git ls-remote" or something?

Show 7 quoted lines
> In this particular case, the origin refspecs are not even going to be
> used, but you can construct a similar one where they are:
>
>   git -C / \
>       -c remote.foo.url=https://github.com/git/git \
>       -c remote.foo.fetch=whatever \
>       ls-remote foo
OK.
Show 21 quoted lines
> We could do this:
>
> diff --git a/refspec.c b/refspec.c
> index 0775358d96..e6c29b7dd0 100644
> --- a/refspec.c
> +++ b/refspec.c
> @@ -101,7 +101,7 @@ static int parse_refspec(struct refspec_item *item, const char *refspec, int fet
>  		/* LHS */
>  		if (!*item->src)
>  			; /* empty is ok; it means "HEAD" */
> -		else if (llen == the_hash_algo->hexsz && !get_oid_hex(item->src, &unused))
> +		else if (the_hash_algo && llen == the_hash_algo->hexsz && !get_oid_hex(item->src, &unused))
>  			item->exact_sha1 = 1; /* ok */
>  		else if (!check_refname_format(item->src, flags))
>  			; /* valid looking ref is ok */
>
> to make the segfault go away, but it is mostly papering over the
> problem. I'm not sure if the exact_sha1 flag would matter when we are
> not actually fetching (and we cannot fetch when we are not in a local
> repo). Grepping around, it looks like it does influence the ref prefixes
> we send to the other side (yet another chicken-and-egg!).

Yup, I agree with your assessment that exact_sha1 should mostly be garbage if we do not have a repository in the first place.

Previous: Jeff KingNext: Jeff King
Message 16 of 17 in “remote-curl: segfault parsing remote.<name>.fetch outside a repository”
  1. Jo LissMar 21, 2026
  2. remote-curl: set fallback hash algorithm outside repoK Jayatheerth, Mar 21, 2026
  3. brian m. carlsonMar 21, 2026
  4. refspec: safely parse refspecs outside a repositoryK Jayatheerth, Mar 22, 2026
  5. Junio C HamanoMar 22, 2026
  6. Jeff KingMar 22, 2026
  7. 1/2 refspec: safely parse refspecs outside a repositoryK Jayatheerth, Mar 22, 2026
  8. 2/2 refspec: fix typo in commentK Jayatheerth, Mar 22, 2026
  9. Junio C HamanoMar 23, 2026
  10. Jeff KingMar 23, 2026
  11. Junio C HamanoMar 23, 2026
  12. 1/2 remote-curl: fall back to default hash outside repoK Jayatheerth, Mar 24, 2026
  13. 2/2 refspec: fix typo in commentK Jayatheerth, Mar 24, 2026
  14. Junio C HamanoMar 24, 2026
  15. Jeff KingMar 21, 2026
  16. Junio C HamanoMar 22, 2026
  17. Jeff KingMar 22, 2026

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.