On Thu, Apr 04, 2019 at 07:25:46PM -0400, Jeff King wrote:
> When we have a .midx that covers many packfiles, we try to avoid opening
> the .idx for those packfiles. However, there are a few problems with the
> filename comparison we use:
>
> - we ask midx_contains_pack() about the .pack name, not the .idx name.
> But it compares to the latter.
>
> - we compute the basename of the pack using strrchr() to find the
> final slash. But that leaves an extra "/" at the start of our
> string; we need to advance past it.
>
> That also raises the question of what to do when the name does not
> have a slash at all. This should generally not happen (we always
> find files in "pack/"), but it doesn't hurt to be defensive here.
>
> The tests don't notice because there's nothing about opening those .idx
> files that would cause us to give incorrect output. It's just a little
> slower. The new test checks this case by corrupting the covered .idx,
> and then making sure we don't complain about it.
When the test suite is run with GIT_TEST_MULTI_PACK_INDEX=1, the
following tests fail with this patch:
t1006-cat-file.sh (Wstat: 256 Tests: 111 Failed: 1)
Failed test: 87
Non-zero exit status: 1
t5320-delta-islands.sh (Wstat: 256 Tests: 15 Failed: 10)
Failed tests: 2-4, 6-10, 12-13
Non-zero exit status: 1
t5310-pack-bitmaps.sh (Wstat: 256 Tests: 46 Failed: 1)
Failed test: 44
Non-zero exit status: 1
t5570-git-daemon.sh (Wstat: 256 Tests: 21 Failed: 1)
Failed test: 10
Non-zero exit status: 1
> Signed-off-by: Jeff King <peff@xxxxxxxx>
> ---
> packfile.c | 17 ++++++++++++++---
> t/t5319-multi-pack-index.sh | 14 ++++++++++++++
> 2 files changed, 28 insertions(+), 3 deletions(-)
>
> diff --git a/packfile.c b/packfile.c
> index 054269ae5d..e7ca135ed5 100644
> --- a/packfile.c
> +++ b/packfile.c
> @@ -472,6 +472,16 @@ static unsigned int get_max_fd_limit(void)
> #endif
> }
>
> +static const char *pack_basename(struct packed_git *p)
> +{
> + const char *ret = strrchr(p->pack_name, '/');
> + if (ret)
> + ret = ret + 1; /* skip past slash */
> + else
> + ret = p->pack_name; /* we only have a base */
> + return ret;
> +}
> +
> /*
> * Do not call this directly as this leaks p->pack_fd on error return;
> * call open_packed_git() instead.
> @@ -486,15 +496,16 @@ static int open_packed_git_1(struct packed_git *p)
> ssize_t read_result;
> const unsigned hashsz = the_hash_algo->rawsz;
>
> - if (!p->index_data) {
> + if (!p->index_data && the_repository->objects->multi_pack_index) {
> struct multi_pack_index *m;
> - const char *pack_name = strrchr(p->pack_name, '/');
> + char *idx_name = pack_name_to_idx(pack_basename(p));
>
> for (m = the_repository->objects->multi_pack_index;
> m; m = m->next) {
> - if (midx_contains_pack(m, pack_name))
> + if (midx_contains_pack(m, idx_name))
> break;
> }
> + free(idx_name);
>
> if (!m && open_pack_index(p))
> return error("packfile %s index unavailable", p->pack_name);
> diff --git a/t/t5319-multi-pack-index.sh b/t/t5319-multi-pack-index.sh
> index 8c4d2bd849..1ebf19ec3c 100755
> --- a/t/t5319-multi-pack-index.sh
> +++ b/t/t5319-multi-pack-index.sh
> @@ -117,6 +117,20 @@ test_expect_success 'write midx with one v2 pack' '
>
> compare_results_with_midx "one v2 pack"
>
> +test_expect_success 'corrupt idx not opened' '
> + idx=$(test-tool read-midx $objdir | grep "\.idx\$") &&
> + mv $objdir/pack/$idx backup-$idx &&
> + test_when_finished "mv backup-\$idx \$objdir/pack/\$idx" &&
> +
> + # This is the minimum size for a sha-1 based .idx; this lets
> + # us pass perfunctory tests, but anything that actually opens and reads
> + # the idx file will complain.
> + test_copy_bytes 1064 <backup-$idx >$objdir/pack/$idx &&
> +
> + git -c core.multiPackIndex=true rev-list --objects --all 2>err &&
> + test_must_be_empty err
> +'
> +
> test_expect_success 'add more objects' '
> for i in $(test_seq 6 10)
> do
> --
> 2.21.0.714.gd1be1d035b
>
