It appears that the goal of this case was to test balance and replace concurrency. But these two operations aren't meant to run concurrently and the replace failing errors are captured in the seqres.full output. Which are expected errors. To avoid further confusion, this patch adds comments. The reason to keep this test case is at the Link. Link: https://patchwork.kernel.org/patch/11806307/ Reported-by: Josef Bacik <josef@xxxxxxxxxxxxxx> Signed-off-by: Anand Jain <anand.jain@xxxxxxxxxx> --- v2: Instead of removing the test case, add comments to clarify. Title is changed old: fstests: delete btrfs/064 it makes no sens tests/btrfs/064 | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/tests/btrfs/064 b/tests/btrfs/064 index 683a69f113bf..ce73acf7872a 100755 --- a/tests/btrfs/064 +++ b/tests/btrfs/064 @@ -4,9 +4,11 @@ # # FSQA Test No. btrfs/064 # -# Run btrfs balance and replace operations simultaneously with fsstress -# running in background. -# +# Run btrfs balance and replace operations simultaneously with fsstress running +# in the background, check with the scrub if all the blocks are ok. +# Balance and replace operations are mutually exclusive operations they can't +# run simultaneously. One of them is expected to fail when the other is running. + seq=`basename $0` seqres=$RESULT_DIR/$seq echo "QA output created by $seq" @@ -62,6 +64,8 @@ run_test() $FSSTRESS_PROG $args >/dev/null 2>&1 & fsstress_pid=$! + # Start both balance and replace in the background. + # Either balance or replace shall run, the other fails. echo -n "Start balance worker: " >>$seqres.full _btrfs_stress_balance $SCRATCH_MNT >/dev/null 2>&1 & balance_pid=$! -- 2.25.1
