[PATCH 0/3] Small fixups for `OBJECT_INFO` flags
- From
Patrick Steinhardt <ps@pks.im>
- Date
- Jan 26, 2026, 12:17 UTC
- Message-ID
- <20260126-b4-pks-read-object-info-flags-v1-0-e682a003b17c@pks.im>
Hi,
I was kind of curious why there were gaps in the `OBJECT_INFO_*` flags, but eventually found out that these gaps are of historic nature: there used to be more flags, but their respective values got removed at one point in time. So naturally, I wanted to clean this up a bit so that the next reader wouldn't have the same question.
Surprisingly though I found out that this breaks tests, which of course puzzled me. As it turns out though, we were incorrectly using a couple of these flags for `odb_has_object()`, and the changed definitions had overlap with the existing meaning of other `HAS_OBJECT_*` flags. There isn't really any bug here as far as I can see, but this is only really by chance.
In any case, the first two commits fix calls to `odb_has_object()` that used invalid flags. The last commit then removes the gaps and converts the flags to use an enum instead.
Thanks!
Patrick
---
Patrick Steinhardt (3):
builtin/backfill: fix flags passed to `odb_has_object()`
builtin/fsck: fix flags passed to `odb_has_object()`
odb: drop gaps in object info flag valuesbuiltin/backfill.c | 3 +-- builtin/fsck.c | 3 ++- odb.h | 38 ++++++++++++++++++++++---------------- 3 files changed, 25 insertions(+), 19 deletions(-)
--- base-commit: ea24e2c55433012a0a6c4ae947a87bc66404e484 change-id: 20260126-b4-pks-read-object-info-flags-236c4437cfc5