| Message ID | 20221107191136.18048-2-pvorel@suse.cz |
|---|---|
| State | New |
| Headers | show |
| Series | Possible bug in zram on ppc64le on vfat | expand |
Hi Li, > Hi Petr, > On Tue, Nov 8, 2022 at 3:12 AM Petr Vorel <pvorel@suse.cz> wrote: > > Repeatedly read /sys/block/zram*/mm_stat for 1 sec. This should fix bug > > on ppc64le on stable kernels, where mem_used_total is often 0. > > Signed-off-by: Petr Vorel <pvorel@suse.cz> > > --- > > .../kernel/device-drivers/zram/zram01.sh | 27 +++++++++++++++++-- > > 1 file changed, 25 insertions(+), 2 deletions(-) > > diff --git a/testcases/kernel/device-drivers/zram/zram01.sh > > b/testcases/kernel/device-drivers/zram/zram01.sh > > index 58d233f91..76a8ccab4 100755 > > --- a/testcases/kernel/device-drivers/zram/zram01.sh > > +++ b/testcases/kernel/device-drivers/zram/zram01.sh > > @@ -105,6 +105,26 @@ zram_mount() > > tst_res TPASS "mount of zram device(s) succeeded" > > } > > +read_mem_used_total() > > +{ > > + echo $(awk '{print $3}' $1) > > +} > > + > > +# Reads /sys/block/zram*/mm_stat until mem_used_total is not 0. > > +loop_read_mem_used_total() > This is not a looping function to check if mem_used_total is equal to zero, > the loop part is by means of the TST_RETRY_FUNC macro. Thanks for your review! > So, I'd suggest renaming it to check_read_mem_used_total(). Agree. Unfortunately even this didn't help on ppc64le system where I was able to reproduce it, thus probably not worth to merge. Unfortunately later I was not able to reproduce the problem any more, I'll try it more this week. Kind regards, Petr > Reviewed-by: Li Wang <liwang@redhat.com> > > +{ > > + local file="$1" > > + local mem_used_total > > + > > + tst_res TINFO "$file" > > + cat $file >&2 > > + > > + mem_used_total=$(read_mem_used_total $file) > > + [ "$mem_used_total" -eq 0 ] && return 1 > > + > > + return 0 > > +} > > + > > zram_fill_fs() > > { > > local mem_used_total > > @@ -133,9 +153,12 @@ zram_fill_fs() > > continue > > fi > > - mem_used_total=`awk '{print $3}' > > "/sys/block/zram$i/mm_stat"` > > + TST_RETRY_FUNC "loop_read_mem_used_total > > /sys/block/zram$i/mm_stat" 0 > > + mem_used_total=$(read_mem_used_total > > /sys/block/zram$i/mm_stat) > > + tst_res TINFO "mem_used_total: $mem_used_total" > > + > > v=$((100 * 1024 * $b / $mem_used_total)) > > - r=`echo "scale=2; $v / 100 " | bc` > > + r=$(echo "scale=2; $v / 100 " | bc) > > if [ "$v" -lt 100 ]; then > > tst_res TFAIL "compression ratio: $r:1" > > -- > > 2.38.0 > > -- > > Mailing list info: https://lists.linux.it/listinfo/ltp
diff --git a/testcases/kernel/device-drivers/zram/zram01.sh b/testcases/kernel/device-drivers/zram/zram01.sh index 58d233f91..76a8ccab4 100755 --- a/testcases/kernel/device-drivers/zram/zram01.sh +++ b/testcases/kernel/device-drivers/zram/zram01.sh @@ -105,6 +105,26 @@ zram_mount() tst_res TPASS "mount of zram device(s) succeeded" } +read_mem_used_total() +{ + echo $(awk '{print $3}' $1) +} + +# Reads /sys/block/zram*/mm_stat until mem_used_total is not 0. +loop_read_mem_used_total() +{ + local file="$1" + local mem_used_total + + tst_res TINFO "$file" + cat $file >&2 + + mem_used_total=$(read_mem_used_total $file) + [ "$mem_used_total" -eq 0 ] && return 1 + + return 0 +} + zram_fill_fs() { local mem_used_total @@ -133,9 +153,12 @@ zram_fill_fs() continue fi - mem_used_total=`awk '{print $3}' "/sys/block/zram$i/mm_stat"` + TST_RETRY_FUNC "loop_read_mem_used_total /sys/block/zram$i/mm_stat" 0 + mem_used_total=$(read_mem_used_total /sys/block/zram$i/mm_stat) + tst_res TINFO "mem_used_total: $mem_used_total" + v=$((100 * 1024 * $b / $mem_used_total)) - r=`echo "scale=2; $v / 100 " | bc` + r=$(echo "scale=2; $v / 100 " | bc) if [ "$v" -lt 100 ]; then tst_res TFAIL "compression ratio: $r:1"
Repeatedly read /sys/block/zram*/mm_stat for 1 sec. This should fix bug on ppc64le on stable kernels, where mem_used_total is often 0. Signed-off-by: Petr Vorel <pvorel@suse.cz> --- .../kernel/device-drivers/zram/zram01.sh | 27 +++++++++++++++++-- 1 file changed, 25 insertions(+), 2 deletions(-)