From: Philipp Hahn <phahn-oss@avm.de>
To: git@vger.kernel.org
Cc: Philipp Hahn <phahn-oss@avm.de>
Subject: BUG: commit-graph + objects/alternate -> SIGSEGV
Date: Fri, 17 Oct 2025 11:47:11 +0200 [thread overview]
Message-ID: <20251017094717.626300-1-phahn-oss@avm.de> (raw)
Thank you for filling out a Git bug report!
Please answer the following questions to help us understand your issue.
What did you do before the bug happened? (Steps to reproduce your issue)
git config --global fetch.writeCommitGraph true
SOME_REPO='git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git'
git clone "$SOME_REPO" 1st
git clone --reference base/.git/objects/ "$SOME_REPO" 2nd
sleep 14d # lots of changes upstream happen
cat 1st/.git/objects/info/commit-graphs/commit-graph-chain
git -C 1st fetch origin
cat 1st/.git/objects/info/commit-graphs/commit-graph-chain
cat 2nd/.git/objects/info/commit-graphs/commit-graph-chain
git -C 2nd fetch origin
cat 2nd/.git/objects/info/commit-graphs/commit-graph-chain
What did you expect to happen? (Expected behavior)
should work
What happened instead? (Actual behavior)
`git fetch` crashes with a SIGSEGV respective complains about "missing commit-graphs":
> Warnung: konnte nicht alle Commit-Graph-Dateien finden
(gdb) bt
#0 __memcpy_avx_unaligned_erms () at ../sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S:342
#1 0x000063f176070f97 in hashwrite (f=0x63f1ad38c3b0, buf=0x0, count=20) at ./csum-file.c:126
#2 0x000063f17604b43d in write_graph_chunk_data (f=0x63f1ad38c3b0, data=0x63f1ad7e7f20) at ./commit-graph.c:1240
#3 0x000063f176235dc2 in write_chunkfile (cf=0x63f1ad38c070, data=0x63f1ad7e7f20) at ./chunk-format.c:85
#4 0x000063f17604df55 in write_commit_graph_file (ctx=0x63f1ad7e7f20) at ./commit-graph.c:2132
#5 0x000063f17604f9eb in write_commit_graph (odb=0x63f1ad05d130, pack_indexes=0x0, commits=0x7ffd0f657160, flags=(COMMIT_GRAPH_WRITE_PROGRESS | COMMIT_GRAPH_WRITE_SPLIT), opts=0x0)
at ./commit-graph.c:2630
#6 0x000063f17604d0d3 in write_commit_graph_reachable (odb=0x63f1ad05d130, flags=(COMMIT_GRAPH_WRITE_PROGRESS | COMMIT_GRAPH_WRITE_SPLIT), opts=0x0) at ./commit-graph.c:1874
#7 0x000063f175f58cdd in cmd_fetch (argc=1, argv=0x7ffd0f658638, prefix=0x0) at builtin/fetch.c:2471
#8 0x000063f175efadda in run_builtin (p=0x63f176349b98 <commands+984>, argc=4, argv=0x7ffd0f658630) at ./git.c:469
#9 0x000063f175efb240 in handle_builtin (argc=4, argv=0x7ffd0f658630) at ./git.c:723
#10 0x000063f175efb4c2 in run_argv (argcp=0x7ffd0f65847c, argv=0x7ffd0f658470) at ./git.c:787
#11 0x000063f175efba9e in cmd_main (argc=4, argv=0x7ffd0f658630) at ./git.c:922
#12 0x000063f176007a64 in main (argc=5, argv=0x7ffd0f658628) at ./common-main.c:62
(gdb) up
#2 0x000063f17604b43d in write_graph_chunk_data (f=0x63f1ad38c3b0, data=0x63f1ad7e7f20) at ./commit-graph.c:1240
1240 hashwrite(f, tree->hash, the_hash_algo->rawsz);
(gdb) p tree
$7 = (struct object_id *) 0x0
(gdb) l
1235
1236 if (repo_parse_commit_no_graph(ctx->r, *list))
1237 die(_("unable to parse commit %s"),
1238 oid_to_hex(&(*list)->object.oid));
1239 tree = get_commit_tree_oid(*list);
1240 hashwrite(f, tree->hash, the_hash_algo->rawsz);
1241
1242 parent = (*list)->parents;
1243
1244 if (!parent)
What's different between what you expected and what actually happened?
`git` should not crash but detect, that the referenced commit-graphs from the
alternate repository are no longer there, ignore them and build a new complete
graph on its own.
Anything else you want to add:
May `git` should detect, that "alternate repositories" don't work well with
"commit graphs" and should warn about that.
Or recomment to disabble `fetch.writeCommitGraph` in repositories, which are used as alternates.
2nd/.git/objects/info/commit-graphs/:
commit-graph-chain
graph-eaf006965bd96cbfaea3127fe6faded17e337d81.graph
2nd/.git/objects/info/commit-graphs/commit-graph-chain:
947ca043c9d5c054195bff874e37bfed738607b3
087b52e4f21a976a00a96d80fc10329066c87267
c28bc2ddb938daffda137d8847dff685dc361ebf
eaf006965bd96cbfaea3127fe6faded17e337d81
1st/.git/objects/info/commit-graphs/:
commit-graph-chain
graph-00ab433b37dfb9c3a3b2930a68b55005c39fdd1a.graph
graph-7c2471bbb8d6980cbe716b8ffdc0a171c8d12ea9.graph
graph-947ca043c9d5c054195bff874e37bfed738607b3.graph
Please review the rest of the bug report below.
You can delete any lines you don't wish to share.
[System Info]
git version:
git version 2.43.0
cpu: x86_64
no commit associated with this build
sizeof-long: 8
sizeof-size_t: 8
shell-path: /bin/sh
uname: Linux 6.8.0-85-generic #85-Ubuntu SMP PREEMPT_DYNAMIC Thu Sep 18 15:26:59 UTC 2025 x86_64
compiler info: gnuc: 12.2
libc info: glibc: 2.39
$SHELL (typically, interactive shell): /bin/bash
[Enabled Hooks]
next reply other threads:[~2025-10-17 9:48 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-10-17 9:47 Philipp Hahn [this message]
2025-10-17 9:54 ` BUG: commit-graph + objects/alternate -> SIGSEGV Kristoffer Haugsbakk
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=20251017094717.626300-1-phahn-oss@avm.de \
--to=phahn-oss@avm.de \
--cc=git@vger.kernel.org \
/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).