From: Cyril Hrubis <chrubis@suse.cz>
To: ltp@lists.linux.it
Subject: [LTP] [PATCH RFC 9/9] ftrace_stress: add two new tests
Date: Thu, 17 Mar 2016 19:08:33 +0100 [thread overview]
Message-ID: <20160317180833.GJ31815@rei.lan> (raw)
In-Reply-To: <1457079898-9449-10-git-send-email-liwang@redhat.com>
Hi!
> +###########################################################################
> +## ##
> +## Copyright (c) 2015, Red Hat Inc. ##
> +## ##
> +## This program is free software: you can redistribute it and/or modify ##
> +## it under the terms of the GNU General Public License as published by ##
> +## the Free Software Foundation, either version 3 of the License, or ##
> +## (at your option) any later version. ##
> +## ##
> +## This program is distributed in the hope that it will be useful, ##
> +## but WITHOUT ANY WARRANTY; without even the implied warranty of ##
> +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ##
> +## GNU General Public License for more details. ##
> +## ##
> +## You should have received a copy of the GNU General Public License ##
> +## along with this program. If not, see <http://www.gnu.org/licenses/>. ##
> +## ##
> +## Author: Chunyu Hu <chuhu@redhat.com> ##
> +## ##
> +###########################################################################
> +
> +triggers=( traceon traceoff enable_event disable_event snapshot\
> + dump cpudump stacktrace module function )
> +n_triggers=${#triggers}
Bashism again.
> +module_pick()
> +{
> + nr_module=$(lsmod | wc -l)
> + pick_one=$(( (RANDOM % nr_module) +1 ))
> + picked_module=$(lsmod | sed -n ''$pick_one'p')
Unfortunately there is no portable way to get random numbers in shell.
So the best we can do is to write a small helper tst_random to take a
range and produce random number in that range. I would have just added a
simple C program to testcases/lib/ the same way we did for portable
sleep shorted than 1s in tst_sleep.c.
> +}
> +
> +nr_functions=$(cat $TRACING_PATH/available_filter_functions|wc -l)
You can do just 'wc -l $TRACING_PATH/available_filter_functions'
> +function_pick()
> +{
local AVAILABLE_FUNCTIONS="$TRACING_PATH/available_filter_functions"
> + if [ -f $TRACING_PATH/available_filter_functions ]; then
> + local pick_one=$(( (RANDOM % nr_functions) + 1 ))
> + picked_function=$(cat $TRACING_PATH/available_filter_functions |\
> + sed -n ''$pick_one'p'|awk '{print $1}')
> + echo $picked_function
You can just do:
awk "{if (NR==$pick_one) {print \$1}}" $AVAILABLE_FUNCTIONS
instead of cat, sed and echo.
> + else
> + echo "\*sched\*"
> + fi
> +}
> +
> +event_pick()
> +{
> + if [ -f $TRACING_PATH/available_events ]; then
> + nr_events=$(cat $TRACING_PATH/available_events | wc -l)
> + local pick_one=$(( (RANDOM % nr_events) + 1 ))
> + picked_event=$(cat $TRACING_PATH/available_events | sed -n ''$pick_one'p')
> +
> + echo "$picked_event"
Here as well no need to use echo and cat:
sed -n ''$pick_one'p' $AVAILABLE_EVENTS
> + else
> + echo "sched:sched_switch"
> + fi
> +}
> +
> +filter_formatter()
> +{
> + local count=$(( (RANDOM % 5) + 1 ))
> + function_name=$(function_pick)
> + case $1 in
> + traceon|traceoff|snapshot|dump|cpudump|stacktrace|module|function)
> + format=$1
> + ;;
> + enable_event|disable_event)
> + event_sys_name=$(event_pick)
> + format=$1:$event_sys_name
> + ;;
> + module)
> + module_pick
> + echo "$picked_module:\*"
> + return
> + ;;
> + function)
> + echo "$picked_function"
> + return
> + ;;
> + *)
> + format=$1
> + ;;
> + esac
> + ((count)) && format=$format:$count
Bahsism. Also the count is non-zero since it's RANDOM % 5 + 1.
> + echo $function_name:$format
> +}
> +
> +for ((; ;))
> +{
Bash style loop. Use while true; do ... done
> + cat $TRACING_PATH/set_ftrace_filter > /dev/null
Why do we cat the file to /dev/null?
> + trigger_index=$((RANDOM % n_triggers))
> + trigger_name=${triggers[$trigger_index]}
> + filter_format=$(filter_formatter $trigger_name)
> +
> + echo "$filter_format" > $TRACING_PATH/set_ftrace_filter
> + [ $? -ne 0 ] && echo "setup filter <$filter_format> failed"
Shouldn't we report a failure here?
> + sleep 1
> +
> + echo "!$filter_format" > $TRACING_PATH/set_ftrace_filter
> + [ $? -ne 0 ] && echo "remove filter <$filter_format> failed"
Here as well.
> +}
> diff --git a/testcases/kernel/tracing/ftrace_test/ftrace_stress/ftrace_tracing_cpumask.sh b/testcases/kernel/tracing/ftrace_test/ftrace_stress/ftrace_tracing_cpumask.sh
> new file mode 100755
> index 0000000..28bbc27
> --- /dev/null
> +++ b/testcases/kernel/tracing/ftrace_test/ftrace_stress/ftrace_tracing_cpumask.sh
> @@ -0,0 +1,51 @@
> +#! /bin/sh
> +
> +###########################################################################
> +## ##
> +## Copyright (c) 2015, Red Hat Inc. ##
> +## ##
> +## This program is free software: you can redistribute it and/or modify ##
> +## it under the terms of the GNU General Public License as published by ##
> +## the Free Software Foundation, either version 3 of the License, or ##
> +## (at your option) any later version. ##
> +## ##
> +## This program is distributed in the hope that it will be useful, ##
> +## but WITHOUT ANY WARRANTY; without even the implied warranty of ##
> +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ##
> +## GNU General Public License for more details. ##
> +## ##
> +## You should have received a copy of the GNU General Public License ##
> +## along with this program. If not, see <http://www.gnu.org/licenses/>. ##
> +## ##
> +## Author: Chunyu Hu <chuhu@redhat.com> ##
> +## ##
> +###########################################################################
> +
> +nr_cpus=`tst_ncpus`
> +
> +get_random_value()
> +{
> + local max=$1
> + local min=${2:-0}
> + local random=${RANDOM:-$(date +%N)}
> + echo $(( random % max ))
> +}
Well looks like you have implemented portable random function after all
although the range does not work at all.
it should really be:
echo $((min + (random % (max - min + 1))))
Which would cover range [min, max] including both min and max.
And the function should be in a library used by all the testcases.
> +get_test_cpumask()
> +{
> + local random=${RANDOM:-$(date +%N)}
Unused variable.
> + local set_cnt=$(get_random_value $nr_cpus);
> + mask=0
> + for ((c=0; c<set_cnt; c++));do
Bashism.
> + local cpuid=$(get_random_value $nr_cpus)
> + mask=$(( mask| (1<<$cpuid) ))
> + done
> + mask=`echo $mask | awk '{printf "%x",$0}'`
Why can't we echo the mask here and do:
get_test_cpumask > $TRACING_PATH/tracing_cpumask
in the main loop?
> +}
> +
> +for ((; ;))
> +{
Bashism.
> + get_test_cpumask
> + echo $mask > $TRACING_PATH/tracing_cpumask
> + sleep 5
Why sleep 5?
> +}
> diff --git a/testcases/kernel/tracing/ftrace_test/ftrace_stress_test.sh b/testcases/kernel/tracing/ftrace_test/ftrace_stress_test.sh
> index d1be49a..49cae86 100755
> --- a/testcases/kernel/tracing/ftrace_test/ftrace_stress_test.sh
> +++ b/testcases/kernel/tracing/ftrace_test/ftrace_stress_test.sh
> @@ -31,7 +31,8 @@ test_targets=" \
> trace_pipe current_tracer ftrace_enabled function_profile_enabled \
> set_event set_ftrace_pid stack_max_size stack_trace trace trace_clock \
> trace_options trace_stat tracing_enabled tracing_max_latency \
> -tracing_on function_profile_enabled buffer_size_kb"
> +tracing_on function_profile_enabled buffer_size_kb tracing_cpumask \
> +set_ftrace_filter"
>
> get_skip_targets()
> {
> --
> 1.8.3.1
>
>
> --
> Mailing list info: http://lists.linux.it/listinfo/ltp
--
Cyril Hrubis
chrubis@suse.cz
next prev parent reply other threads:[~2016-03-17 18:08 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-03-04 8:24 [LTP] [PATCH RFC 0/9] tracing: make ftrace tests to be extended Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 1/9] ftrace_stress: remove the useless file ftrace_get_page_size.c Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 2/9] tracing[1]: reorganize ftrace-stress tests to general tests Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 3/9] tracing[2]: reorganize ftrace stress " Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 4/9] ftrace_regression: add new case to ftrace_regression/ dir Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 5/9] ftrace_regression: add a new testcase Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 6/9] ftrace_stress: skip unsupported tests Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 7/9] ftrace_stress: keep the name of testscipt in sync with tracing file Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 8/9] ftrace_stress: update the trace_options test Li Wang
2016-03-04 8:24 ` [LTP] [PATCH RFC 9/9] ftrace_stress: add two new tests Li Wang
2016-03-17 18:08 ` Cyril Hrubis [this message]
2016-03-18 14:08 ` Chunyu Hu
2016-03-21 12:09 ` Cyril Hrubis
2016-03-17 17:19 ` [LTP] [PATCH RFC 8/9] ftrace_stress: update the trace_options test Cyril Hrubis
2016-03-17 17:28 ` Cyril Hrubis
2016-03-18 13:24 ` Chunyu Hu
2016-03-18 13:21 ` Chunyu Hu
2016-03-17 17:07 ` [LTP] [PATCH RFC 6/9] ftrace_stress: skip unsupported tests Cyril Hrubis
2016-03-18 13:08 ` Chunyu Hu
2016-03-17 16:50 ` [LTP] [PATCH RFC 5/9] ftrace_regression: add a new testcase Cyril Hrubis
2016-03-18 7:57 ` Li Wang
2016-03-17 16:49 ` [LTP] [PATCH RFC 4/9] ftrace_regression: add new case to ftrace_regression/ dir Cyril Hrubis
2016-03-17 16:39 ` [LTP] [PATCH RFC 3/9] tracing[2]: reorganize ftrace stress tests to general tests Cyril Hrubis
2016-03-18 10:42 ` Li Wang
2016-03-21 12:21 ` Cyril Hrubis
2016-03-22 8:27 ` Li Wang
2016-03-17 16:29 ` [LTP] [PATCH RFC 1/9] ftrace_stress: remove the useless file ftrace_get_page_size.c Cyril Hrubis
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160317180833.GJ31815@rei.lan \
--to=chrubis@suse.cz \
--cc=ltp@lists.linux.it \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.