Hi, Here's a regression fix that I noticed this morning while deploying 2.28 out to GitHub. The gist is that we accidentally call 'string_list_clear(..., 1)' when freeing the list of allowed/banned object filters, but the ->util pointer either contains '0', or '1'. So, the free() of the util pointer is bogus, and we end up crashing. The fix in the second patch is straightforward, but it doesn't work without the first patch, which teaches 'git clone' to match 'git fetch' and not ignore an error from 'transport_fetch_refs()'. Unfortunately, this regression has been in Git since 2.28. Fortunately, nobody has seemed to notice, so I doubt that it's bitten anybody out in the wild. But, it's still worth fixing, since it is so obviously broken. Thanks, Taylor Taylor Blau (2): builtin/clone.c: don't ignore transport_fetch_refs() errors upload-pack.c: don't free allowed_filters util pointers builtin/clone.c | 15 +++++++++++---- t/t5616-partial-clone.sh | 10 +++++++++- upload-pack.c | 2 +- 3 files changed, 21 insertions(+), 6 deletions(-) -- 2.29.2.533.g07db1f5344
