UPDATE for V2: We now know the full repro, and a test is added. Thanks Szeder and Peff for your insights! UPDATE in V3: Cleaned up the commit messages and some test details. While dogfooding, Johannes found a bug in the fetch.writeCommitGraph config behavior. While his example initially happened during a clone with --recurse-submodules, (UPDATE) and the submodule is important, but --recurse-submodules is not: $ git clone <url> test $ cd test $ git -c fetch.writeCommitGraph=true fetch origin Computing commit graph generation numbers: 100% (12/12), done. BUG: commit-graph.c:886: missing parent <hash1> for commit <hash2> Aborted (core dumped) In the repo I had cloned, there were really 60 commits to scan, but only 12 were in the list to write when calling compute_generation_numbers(). A commit in the list expects to see a parent, but that parent is not in the list. The simple example I used for my testing was https://github.com/derrickstolee/numbers. Thie repo HAS A SUBMODULE, I just forgot. Sorry for derailing the investigation somewhat. The details above are the start of the commit message for [PATCH 1/2], including a test that fails when fetching after cloning a repo with a submodule. In [PATCH 2/2], I actually have the fix. I tried to include as much detail as I could for how I investigated the problem and why I think this is the right solution. I added details that have come from the on-list discussion, including what the submodule code is doing and why REACHABLE is no longer used in commit-reach.c. Thanks, -Stolee Derrick Stolee (2): t5510-fetch.sh: demonstrate fetch.writeCommitGraph bug commit-graph: fix writing first commit-graph during fetch commit-graph.c | 11 +++++++---- commit-reach.c | 1 - object.h | 3 ++- t/t5510-fetch.sh | 16 ++++++++++++++++ 4 files changed, 25 insertions(+), 6 deletions(-) base-commit: d966095db01190a2196e31195ea6fa0c722aa732 Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-415%2Fderrickstolee%2Ffetch-first-write-fail-v3 Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-415/derrickstolee/fetch-first-write-fail-v3 Pull-Request: https://github.com/gitgitgadget/git/pull/415 Range-diff vs v2: 1: 6ac0a05746 ! 1: ce53b5a7bf t5510-fetch.sh: demonstrate fetch.writeCommitGraph bug @@ -22,10 +22,6 @@ A follow-up will fix the bug, but first we create a test that demonstrates the problem. - I used "test_expect_failure" for the entire test instead of - "test_must_fail" only on the command that I expect to fail. This is - because the BUG() returns an exit code so test_must_fail complains. - Helped-by: Jeff King <peff@xxxxxxxx> Helped-by: Johannes Schindelin <johannes.schindelin@xxxxxx> Helped-by: Szeder Gábor <szeder.dev@xxxxxxxxx> @@ -39,16 +35,15 @@ ' +test_expect_failure 'fetch.writeCommitGraph with submodules' ' -+ pwd="$(pwd)" && + git clone dups super && + ( + cd super && -+ git submodule add "file://$pwd/three" && ++ git submodule add "file://$TRASH_DIRECTORY/three" && + git commit -m "add submodule" + ) && -+ git clone "super" writeError && ++ git clone "super" super-clone && + ( -+ cd writeError && ++ cd super-clone && + test_path_is_missing .git/objects/info/commit-graphs/commit-graph-chain && + git -c fetch.writeCommitGraph=true fetch origin && + test_path_is_file .git/objects/info/commit-graphs/commit-graph-chain 2: ca59b118f1 ! 2: edacfff490 commit-graph: fix writing first commit-graph during fetch @@ -6,7 +6,7 @@ fetch.writeCommitGraph and fetching in a repo with a submodule. Here, we fix that bug and set the test to "test_expect_success". - The prolem arises with this set of commands when the remote repo at + The problem arises with this set of commands when the remote repo at <url> has a submodule. Note that --recurse-submodules is not needed to demonstrate the bug. @@ -44,16 +44,6 @@ negotiation is comparing the remote refs with the local refs and marking some commits as UNINTERESTING. - You may ask: did this feature ever work at all? Yes, it did, as long as - you had a commit-graph covering all of your local refs. My testing was - unfortunately limited to this scenario. The UNINTERESTING commits are - always part of the "old" commit-graph, and when we add new commits to a - top layer of the commit-graph chain those are not needed. If we happen - to merge layers of the chain, then the commits are added as a list, not - using a commit walk. Further, the test added for this config option in - t5510-fetch.sh uses local filesystem clones, which somehow avoids this - logic. - I tested running clear_commit_marks_many() to clear the UNINTERESTING flag inside close_reachable(), but the tips did not have the flag, so that did nothing. @@ -62,7 +52,7 @@ fault. Thanks, Peff, for pointing out this detail! More specifically, for each submodule, the collect_changed_submodules() runs a revision walk to essentially do file-history on the list of submodules. That - revision walk marks commits UNININTERESTING if they are simiplified away + revision walk marks commits UNININTERESTING if they are simplified away by not changing the submodule. Instead, I finally arrived on the conclusion that I should use a flag @@ -163,6 +153,6 @@ -test_expect_failure 'fetch.writeCommitGraph with submodules' ' +test_expect_success 'fetch.writeCommitGraph with submodules' ' - pwd="$(pwd)" && git clone dups super && ( + cd super && -- gitgitgadget
