On 12/21/2019 2:49 PM, brian m. carlson wrote:
> This test corrupts various locations in a multi-pack index to test
> various error responses. However, these offsets differ between SHA-1
> indexes and SHA-256 indexes due to differences in object length. Use
> test_oid to look up the correct offsets based on the algorithm.
> ---
> t/t5319-multi-pack-index.sh | 22 ++++++++++++++++++----
> 1 file changed, 18 insertions(+), 4 deletions(-)
>
> diff --git a/t/t5319-multi-pack-index.sh b/t/t5319-multi-pack-index.sh
> index 464bb68e89..43a7a66c9d 100755
> --- a/t/t5319-multi-pack-index.sh
> +++ b/t/t5319-multi-pack-index.sh
> @@ -28,6 +28,20 @@ midx_read_expect () {
> test_cmp expect actual
> }
>
> +test_expect_success 'setup' '
> + test_oid_init &&
> + test_oid_cache <<-EOF
> + idxoff sha1:2999
> + idxoff sha256:3739
> +
> + packnameoff sha1:652
> + packnameoff sha256:940
> +
> + fanoutoff sha1:1
> + fanoutoff sha256:3
> + EOF
> +'
> +
> test_expect_success 'write midx with no packs' '
> test_when_finished rm -f pack/multi-pack-index &&
> git multi-pack-index --object-dir=. write &&
> @@ -225,7 +239,7 @@ test_expect_success 'verify bad signature' '
> "multi-pack-index signature"
> '
>
> -HASH_LEN=20
> +HASH_LEN=$(test_oid rawsz)
I'm glad this was an easy update.
> +MIDX_OFFSET_OID_FANOUT=$(($MIDX_OFFSET_PACKNAMES + $(test_oid packnameoff)))
> +MIDX_BYTE_OID_FANOUT_ORDER=$((MIDX_OFFSET_OID_FANOUT + 250 * $MIDX_OID_FANOUT_WIDTH + $(test_oid fanoutoff)))
I see why these are necessary: we need to find an exact position that
causes an order check to fail. In that sense, the offsets need to match
the early bytes of a packfile name or an OID fanout value.
> @@ -387,7 +401,7 @@ test_expect_success 'force some 64-bit offsets with pack-objects' '
> pack64=$(git pack-objects --index-version=2,0x40 objects64/pack/test-64 <obj-list) &&
> idx64=objects64/pack/test-64-$pack64.idx &&
> chmod u+w $idx64 &&
> - corrupt_data $idx64 2999 "\02" &&
> + corrupt_data $idx64 $(test_oid idxoff) "\02" &&
Sorry that this was not a better-calculated value, but
your approach works well here.
Thanks,
-Stolee
