The p4 program is finicky about making sure the recorded client Root
matches the current working directory. The way it discovers the latter
seems to be to inspect shell variable $PWD. This could involve symlinks,
that while leading to the same place as the client Root, look different,
and cause p4 to fail.
Resolve all client paths using "test-path-utils real_path $path". This
removes ".." and resolves all symlinks.
Discovered while running with --root=/dev/shm, which is a link to
/run/shm.
Signed-off-by: Pete Wyckoff <pw@padd.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
export P4EDITOR=:
db="$TRASH_DIRECTORY/db"
-cli="$TRASH_DIRECTORY/cli"
+cli=$(test-path-utils real_path "$TRASH_DIRECTORY/cli")
git="$TRASH_DIRECTORY/git"
pidfile="$TRASH_DIRECTORY/p4d.pid"
exec >/dev/null &&
test_must_fail git p4 clone --dest="$git" --use-client-spec
) &&
- cli2="$TRASH_DIRECTORY/cli2" &&
+ cli2=$(test-path-utils real_path "$TRASH_DIRECTORY/cli2") &&
mkdir -p "$cli2" &&
test_when_finished "rmdir \"$cli2\"" &&
(
cd "$git" &&
git config git-p4.skipSubmitEdit true &&
git config git-p4.attemptRCSCleanup true &&
- (cd ../cli && p4_append_to_file kwfile1.c) &&
+ (cd "$cli" && p4_append_to_file kwfile1.c) &&
old_lines=$(wc -l <kwfile1.c) &&
perl -n -i -e "print unless m/Revision:/" kwfile1.c &&
new_lines=$(wc -l <kwfile1.c) &&