Ben Peart <peart...@gmail.com> writes: > From: Ben Peart <benpe...@microsoft.com> > > When doing a 'checkout -b' do a full checkout including updating the working > tree when doing the initial checkout. This fixes the regression in behavior > caused by fa655d8411 checkout: optimize "git checkout -b <new_branch>" > > Signed-off-by: Ben Peart <benpe...@microsoft.com> > --- > builtin/checkout.c | 6 ++++++ > t/t2018-checkout-branch.sh | 2 +- > 2 files changed, 7 insertions(+), 1 deletion(-) > > diff --git a/builtin/checkout.c b/builtin/checkout.c > index 6fadf412e8..af6b5c8336 100644 > --- a/builtin/checkout.c > +++ b/builtin/checkout.c > @@ -517,6 +517,12 @@ static int skip_merge_working_tree(const struct > checkout_opts *opts, > if (core_apply_sparse_checkout && !checkout_optimize_new_branch) > return 0; > > + /* > + * We must do the merge if this is the initial checkout > + */ > + if (is_cache_unborn()) > + return 0; > +
Yup, that's a trivial fix ;-) > /* > * We must do the merge if we are actually moving to a new commit. > */ > diff --git a/t/t2018-checkout-branch.sh b/t/t2018-checkout-branch.sh > index 35999b3adb..c438889b0c 100755 > --- a/t/t2018-checkout-branch.sh > +++ b/t/t2018-checkout-branch.sh > @@ -206,7 +206,7 @@ test_expect_success 'checkout -b after clone > --no-checkout does a checkout of HE > rev="$(git -C src rev-parse HEAD)" && > git clone --no-checkout src dest && > git -C dest checkout "$rev" -b branch && > - test_must_fail test -f dest/a > + test -f dest/a > ' This is flipping the wrong thing. Rather, introduce the whole test as test_expect_failure that wants to make sure dest/a exists, and with this 2/2 patch flip test_expect_failure into test_expect_success. Thanks.
