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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox