From: Filipe Manana <fdmanana@xxxxxxxx> The test is ignoring failures completely: 1) If mounting the scratch device fails it redirects both stdout and stderr to /dev/null, so it gets unnoticed and the test runs against a different fs than expected (in my test environment $SCRATCH_MNT points to a directory in an ext4 fs for example but I want to test btrfs); 2) We are redirecting the stdout and stderr of fiemap-tester to the $seqres.full file, so if it fails it gets completely unnoticed and the test succeeds. For the first issue fix this by not even using the scratch filesystem and use instead the test filesystem, since the test creates a 2M file which is small enough. For the second issue simply don't redirect the stdout and stderr, so that if the test program fails it causes a mismatch with the golden output, making the test fail. Signed-off-by: Filipe Manana <fdmanana@xxxxxxxx> --- tests/generic/094 | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/tests/generic/094 b/tests/generic/094 index c82efced..6ae417c3 100755 --- a/tests/generic/094 +++ b/tests/generic/094 @@ -9,18 +9,22 @@ . ./common/preamble _begin_fstest auto quick prealloc fiemap +_cleanup() +{ + cd / + rm -fr $tmp.* + rm -f $fiemapfile +} + # Import common functions. . ./common/filter -_require_scratch +_require_test _require_odirect _require_xfs_io_command "fiemap" _require_xfs_io_command "falloc" -_scratch_mkfs > /dev/null 2>&1 -_scratch_mount > /dev/null 2>&1 - -fiemapfile=$SCRATCH_MNT/$seq.fiemap +fiemapfile=$TEST_DIR/$seq.fiemap _require_test_program "fiemap-tester" @@ -29,12 +33,10 @@ seed=`date +%s` echo "using seed $seed" >> $seqres.full echo "fiemap run with sync" -$here/src/fiemap-tester -q -S -s $seed -r 200 $fiemapfile 2>&1 | \ - tee -a $seqres.full +$here/src/fiemap-tester -q -S -s $seed -r 200 $fiemapfile echo "fiemap run without sync" -$here/src/fiemap-tester -q -s $seed -r 200 $fiemapfile 2>&1 | \ - tee -a $seqres.full +$here/src/fiemap-tester -q -s $seed -r 200 $fiemapfile status=0 exit -- 2.47.2