git/t/t7415-submodule-names.sh
Jeff King 368b4e5906 index-pack: handle --strict checks of non-repo packs
Commit 73c3f0f704 (index-pack: check .gitmodules files with
--strict, 2018-05-04) added a call to add_packed_git(), with
the intent that the newly-indexed objects would be available
to the process when we run fsck_finish().  But that's not
what add_packed_git() does. It only allocates the struct,
and you must install_packed_git() on the result. So that
call was effectively doing nothing (except leaking a
struct).

But wait, we passed all of the tests! Does that mean we
don't need the call at all?

For normal cases, no. When we run "index-pack --stdin"
inside a repository, we write the new pack into the object
directory. If fsck_finish() needs to access one of the new
objects, then our initial lookup will fail to find it, but
we'll follow up by running reprepare_packed_git() and
looking again. That logic was meant to handle somebody else
repacking simultaneously, but it ends up working for us
here.

But there is a case that does need this, that we were not
testing. You can run "git index-pack foo.pack" on any file,
even when it is not inside the object directory. Or you may
not even be in a repository at all! This case fails without
doing the proper install_packed_git() call.

We can make this work by adding the install call.

Note that we should be prepared to handle add_packed_git()
failing. We can just silently ignore this case, though. If
fsck_finish() later needs the objects and they're not
available, it will complain itself. And if it doesn't
(because we were able to resolve the whole fsck in the first
pass), then it actually isn't an interesting error at all.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2018-06-01 11:48:56 +09:00

165 lines
4.6 KiB
Bash
Executable File

#!/bin/sh
test_description='check handling of .. in submodule names
Exercise the name-checking function on a variety of names, and then give a
real-world setup that confirms we catch this in practice.
'
. ./test-lib.sh
. "$TEST_DIRECTORY"/lib-pack.sh
test_expect_success 'check names' '
cat >expect <<-\EOF &&
valid
valid/with/paths
EOF
git submodule--helper check-name >actual <<-\EOF &&
valid
valid/with/paths
../foo
/../foo
..\foo
\..\foo
foo/..
foo/../
foo\..
foo\..\
foo/../bar
EOF
test_cmp expect actual
'
test_expect_success 'create innocent subrepo' '
git init innocent &&
git -C innocent commit --allow-empty -m foo
'
test_expect_success 'submodule add refuses invalid names' '
test_must_fail \
git submodule add --name ../../modules/evil "$PWD/innocent" evil
'
test_expect_success 'add evil submodule' '
git submodule add "$PWD/innocent" evil &&
mkdir modules &&
cp -r .git/modules/evil modules &&
write_script modules/evil/hooks/post-checkout <<-\EOF &&
echo >&2 "RUNNING POST CHECKOUT"
EOF
git config -f .gitmodules submodule.evil.update checkout &&
git config -f .gitmodules --rename-section \
submodule.evil submodule.../../modules/evil &&
git add modules &&
git commit -am evil
'
# This step seems like it shouldn't be necessary, since the payload is
# contained entirely in the evil submodule. But due to the vagaries of the
# submodule code, checking out the evil module will fail unless ".git/modules"
# exists. Adding another submodule (with a name that sorts before "evil") is an
# easy way to make sure this is the case in the victim clone.
test_expect_success 'add other submodule' '
git submodule add "$PWD/innocent" another-module &&
git add another-module &&
git commit -am another
'
test_expect_success 'clone evil superproject' '
git clone --recurse-submodules . victim >output 2>&1 &&
! grep "RUNNING POST CHECKOUT" output
'
test_expect_success 'fsck detects evil superproject' '
test_must_fail git fsck
'
test_expect_success 'transfer.fsckObjects detects evil superproject (unpack)' '
rm -rf dst.git &&
git init --bare dst.git &&
git -C dst.git config transfer.fsckObjects true &&
test_must_fail git push dst.git HEAD
'
test_expect_success 'transfer.fsckObjects detects evil superproject (index)' '
rm -rf dst.git &&
git init --bare dst.git &&
git -C dst.git config transfer.fsckObjects true &&
git -C dst.git config transfer.unpackLimit 1 &&
test_must_fail git push dst.git HEAD
'
# Normally our packs contain commits followed by trees followed by blobs. This
# reverses the order, which requires backtracking to find the context of a
# blob. We'll start with a fresh gitmodules-only tree to make it simpler.
test_expect_success 'create oddly ordered pack' '
git checkout --orphan odd &&
git rm -rf --cached . &&
git add .gitmodules &&
git commit -m odd &&
{
pack_header 3 &&
pack_obj $(git rev-parse HEAD:.gitmodules) &&
pack_obj $(git rev-parse HEAD^{tree}) &&
pack_obj $(git rev-parse HEAD)
} >odd.pack &&
pack_trailer odd.pack
'
test_expect_success 'transfer.fsckObjects handles odd pack (unpack)' '
rm -rf dst.git &&
git init --bare dst.git &&
test_must_fail git -C dst.git unpack-objects --strict <odd.pack
'
test_expect_success 'transfer.fsckObjects handles odd pack (index)' '
rm -rf dst.git &&
git init --bare dst.git &&
test_must_fail git -C dst.git index-pack --strict --stdin <odd.pack
'
test_expect_success 'index-pack --strict works for non-repo pack' '
rm -rf dst.git &&
git init --bare dst.git &&
cp odd.pack dst.git &&
test_must_fail git -C dst.git index-pack --strict odd.pack 2>output &&
# Make sure we fail due to bad gitmodules content, not because we
# could not read the blob in the first place.
grep gitmodulesName output
'
test_expect_success 'fsck detects symlinked .gitmodules file' '
git init symlink &&
(
cd symlink &&
# Make the tree directly to avoid index restrictions.
#
# Because symlinks store the target as a blob, choose
# a pathname that could be parsed as a .gitmodules file
# to trick naive non-symlink-aware checking.
tricky="[foo]bar=true" &&
content=$(git hash-object -w ../.gitmodules) &&
target=$(printf "$tricky" | git hash-object -w --stdin) &&
tree=$(
{
printf "100644 blob $content\t$tricky\n" &&
printf "120000 blob $target\t.gitmodules\n"
} | git mktree
) &&
commit=$(git commit-tree $tree) &&
# Check not only that we fail, but that it is due to the
# symlink detector; this grep string comes from the config
# variable name and will not be translated.
test_must_fail git fsck 2>output &&
grep gitmodulesSymlink output
)
'
test_done