Jan Kara was testing filesystem freeze, and was consistently locking
up, although my tests of 068 were passing. He pointed out that
he was running in read/write mode, and it was atime updates causing
the trouble. Sure enough, dropping "-w" from fsstress locked me up
too. Change this so we get better (and more realistic) coverage.
Signed-off-by: Eric Sandeen <sandeen@xxxxxxxxxx>
cc: Jan Kara <jack@xxxxxxx>
---
diff --git a/068 b/068
index 6f08f18..b15409e 100755
--- a/068
+++ b/068
@@ -79,8 +79,9 @@ touch $tmp.running
nops=200
while [ -f "$tmp.running" ]
do
- # -w ensures that the only ops are ones which cause write I/O
- $FSSTRESS_PROG -d $STRESS_DIR -w -p $procs -n $nops $FSSTRESS_AVOID \
+ # We do both read & write IO - not only is this more realistic,
+ # but it also potentially tests atime updates
+ $FSSTRESS_PROG -d $STRESS_DIR -p $procs -n $nops $FSSTRESS_AVOID \
> /dev/null 2>&1
done
_______________________________________________
xfs mailing list
xfs@xxxxxxxxxxx
http://oss.sgi.com/mailman/listinfo/xfs
