sparse-checkout: fix segfault on malformed patterns

This series fixes some issues with parsing sparse-checkout patterns when core.sparseCheckoutCone is enabled but the sparse-checkout file itself contains patterns that don't match the cone mode format.

The first patch fixes a segfault first reported in [1]. The other two patches are from an earlier submission [2] that never got picked up and I lost track of. There was another patch involving 'git sparse-checkout init --cone' that isn't necessary, especially with Elijah doing some work in that space right now.

[1] https://github.com/git-for-windows/git/issues/3498 [2]
https://lore.kernel.org/git/[email protected]

Thanks, -Stolee

Update in v4
============

* For added precaution, this kind of unexpected duplicate pattern will disable cone mode matching.
* Tests are updated to verify this new behavior.

Updates in v2 and v3
====================

* I intended to fix a typo in a patch, but accidentally sent the amend! commit in v2
* v3 has the typo fix properly squashed in.
* Added Elijah's review.

Derrick Stolee (3):
sparse-checkout: fix segfault on malformed patterns
sparse-checkout: fix OOM error with mixed patterns
sparse-checkout: refuse to add to bad patterns

 builtin/sparse-checkout.c          |  5 +++-
 dir.c                              |  6 ++---
 t/t1091-sparse-checkout-builtin.sh | 37 +++++++++++++++++++++++++++++-
 3 files changed, 42 insertions(+), 6 deletions(-)

base-commit: abe6bb3905392d5eb6b01fa6e54d7e784e0522aa
Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-1069%2Fderrickstolee%2Fsparse-checkout%2Finput-bug-v4
Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-1069/derrickstolee/sparse-checkout/input-bug-v4
Pull-Request: https://github.com/gitgitgadget/git/pull/1069

Range-diff vs v3:

1: 1744a26845f ! 1: 5353c541d9f sparse-checkout: fix segfault on malformed patterns
@@ Commit message
list' command because it iterates over the contents of the hashset, which is now invalid.

- The fix here is to stop trying to remove from the hashset. Better to leave
- bad data in the sparse-checkout matching logic (with a warning) than to
- segfault. If we are in this state, then we are already traversing into
- undefined behavior, so this change to keep the entry in the hashset is no
- worse than removing it.
+ The fix here is to stop trying to remove from the hashset. In addition,
+ we disable cone mode sparse-checkout because of the malformed data. This
+ results in the pattern-matching working with a possibly-slower
+ algorithm, but using the patterns as they are in the sparse-checkout
+ file.
+
+ This also changes the behavior of commands such as 'git sparse-checkout
+ list' because the output patterns will be the contents of the
+ sparse-checkout file instead of the list of directories. This is an
+ existing behavior for other types of bad patterns.

Add a test that triggers the segfault without the code change.

@@ dir.c: static void add_pattern_to_hashsets(struct pattern_list *pl, struct path_
- hashmap_remove(&pl->parent_hashmap, &translated->ent, &data);
- free(data);
- free(translated);
++ goto clear_hashmaps;
}

return;
@@ t/t1091-sparse-checkout-builtin.sh: test_expect_success 'cone mode clears ignore
+ !/foo/*/
+ /foo/\*/
+ EOF
-+ cat repo/.git/info/sparse-checkout &&
-+ git -C repo sparse-checkout list
++
++ # Listing the patterns will notice the duplicate pattern and
++ # emit a warning. It will list the patterns directly instead
++ # of using the cone-mode translation to a set of directories.
++ git -C repo sparse-checkout list >actual 2>err &&
++ test_cmp repo/.git/info/sparse-checkout actual &&
++ grep "warning: your sparse-checkout file may have issues: pattern .* is repeated" err &&
++ grep "warning: disabling cone pattern matching" err
+'
+
test_done

 2:  a2fe867222e = 2:  3fd625290a3 sparse-checkout: fix OOM error with mixed patterns
 3:  a0e5a942ae0 = 3:  f5f7b8b8e04 sparse-checkout: refuse to add to bad patterns

gitgitgadget