Junio C Hamano [off-list ref] writes:
Junio C Hamano [off-list ref] writes:
quoted
Bad news is that this does not seem to have much to do with the
incore-index series. v1.5.4-rc4 shows the same breakage.
Please disregard that bad news -- that was my stupidity while
bisecting. v1.5.4-rc4 seems to be Ok.
And the bisection points at the first one in the series "Updated
in-memory index cleanup". Which is kind of odd and sad, as it
is not supposed to change the semantics at all, and I did not
see anything obviously wrong with it.
$ cd t
$ while :
do
sh -x ./t7501-commit.sh -i -v >,,log 2>&1 || break
done