[LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[Alexey Kodanev]

One of the possible ways to test RCU is to use rcutorture kernel module.
The test requires that kernel configured with CONFIG_RCU_TORTURE_TEST.
It runs rcutorture module using particular options and then inspects
dmesg output for module's test results.
For more information, please read Linux Documentation: RCU/torture.txt

Signed-off-by: Alexey Kodanev <alexey.kodanev@oracle.com>
---
 runtest/kernel_misc                                |    1 +
 testcases/kernel/device-drivers/Makefile           |    1 +
 testcases/kernel/device-drivers/rcu/Makefile       |   22 ++++
 testcases/kernel/device-drivers/rcu/rcu_torture.sh |  120 ++++++++++++++++++++
 4 files changed, 144 insertions(+), 0 deletions(-)
 create mode 100644 testcases/kernel/device-drivers/rcu/Makefile
 create mode 100755 testcases/kernel/device-drivers/rcu/rcu_torture.sh

diff --git a/runtest/kernel_misc b/runtest/kernel_misc
index c1826fb..8f37cc6 100644
--- a/runtest/kernel_misc
+++ b/runtest/kernel_misc
@@ -7,3 +7,4 @@ tbio tbio
 ltp_acpi ltp_acpi
 ltp_acpi_cpufreq ltp_acpi_cpufreq
 uaccess uaccess
+rcu_torture.sh rcu_torture.sh
diff --git a/testcases/kernel/device-drivers/Makefile b/testcases/kernel/device-drivers/Makefile
index d207736..37b6c64 100644
--- a/testcases/kernel/device-drivers/Makefile
+++ b/testcases/kernel/device-drivers/Makefile
@@ -21,6 +21,7 @@ include $(top_srcdir)/include/mk/env_pre.mk
 SUBDIRS		:= acpi \
 		   block \
 		   pci \
+		   rcu \
 		   rtc \
 		   tbio \
 		   uaccess
diff --git a/testcases/kernel/device-drivers/rcu/Makefile b/testcases/kernel/device-drivers/rcu/Makefile
new file mode 100644
index 0000000..96c6b5b
--- /dev/null
+++ b/testcases/kernel/device-drivers/rcu/Makefile
@@ -0,0 +1,22 @@
+# Copyright (c) 2014 Oracle and/or its affiliates. All Rights Reserved.
+#
+# 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 2 of
+# the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it would 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, write the Free Software Foundation,
+# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+
+top_srcdir	?= ../../../..
+include $(top_srcdir)/include/mk/testcases.mk
+
+INSTALL_TARGETS		:= rcu_torture.sh
+
+include $(top_srcdir)/include/mk/generic_leaf_target.mk
diff --git a/testcases/kernel/device-drivers/rcu/rcu_torture.sh b/testcases/kernel/device-drivers/rcu/rcu_torture.sh
new file mode 100755
index 0000000..c3db889
--- /dev/null
+++ b/testcases/kernel/device-drivers/rcu/rcu_torture.sh
@@ -0,0 +1,120 @@
+#!/bin/sh
+# Copyright (c) 2014 Oracle and/or its affiliates. All Rights Reserved.
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation; either version 2 of
+# the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it would be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write the Free Software Foundation,
+# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+#
+# Author: Alexey Kodanev <alexey.kodanev@oracle.com>
+#
+# One of the possible ways to test RCU is to use rcutorture kernel module.
+# The test requiers that kernel configured with CONFIG_RCU_TORTURE_TEST.
+# It runs rcutorture module using particular options and then inspects
+# dmesg output for module's test results.
+# For more information, please read Linux Documentation: RCU/torture.txt
+
+export RC=0
+export TST_TOTAL=14
+export TCID="rcu_torture"
+export TST_COUNT=0
+# default options
+test_time=60
+num_readers=10
+num_writers=5
+
+while getopts :ht:r:w: opt; do
+	case "$opt" in
+	h)
+		echo "Usage:"
+		echo "h        help"
+		echo "t x      time in seconds for each test-case"
+		echo "r x      number of readers"
+		echo "w x      number of writers"
+		exit 0
+	;;
+	t) test_time=$OPTARG ;;
+	r) num_readers=$OPTARG ;;
+	w) num_writers=$OPTARG ;;
+	*)
+		tst_brkm TBROK NULL "unknown option: $opt"
+		exit 2
+	;;
+	esac
+done
+
+cleanup()
+{
+	rmmod rcutorture > /dev/null 2>&1
+}
+
+if [ "`id -u`" -ne 0 ]; then
+	tst_brkm TCONF NULL "Test must be run as root"
+	exit 0
+fi
+
+# check if module is present
+modprobe rcutorture > /dev/null 2>&1
+if [ $? -ne 0 ]; then
+	tst_brkm TCONF NULL "Test requiers rcutorture module"
+	exit 0
+fi
+cleanup
+
+trap cleanup EXIT
+
+rcu_type="rcu rcu_sync rcu_expedited rcu_bh rcu_bh_sync rcu_bh_expedited \
+          srcu srcu_sync srcu_expedited srcu_raw srcu_raw_sync sched \
+          sched_sync sched_expedited"
+
+est_time=`echo "scale=2; $test_time * $TST_TOTAL / 60 " | bc`
+tst_resm TINFO "estimate time $est_time min"
+
+for type in $rcu_type; do
+
+	tst_resm TINFO "tc$TST_COUNT: $type running ($test_time sec)..."
+
+	modprobe rcutorture nreaders=$num_readers nfakewriters=$num_writers \
+	         stat_interval=60 test_no_idle_hz=1 shuffle_interval=3 \
+	         stutter=5 irqreader=1 fqs_duration=0 fqs_holdoff=0 \
+	         fqs_stutter=3 test_boost=1 test_boost_interval=7 \
+	         test_boost_duration=4 shutdown_secs=0 \
+	         stall_cpu=0 stall_cpu_holdoff=10 n_barrier_cbs=0 \
+	         onoff_interval=0 onoff_holdoff=0 torture_type=$type \
+	         > /dev/null 2>&1
+
+	if [ $? -ne 0 ]; then
+		tst_brkm TBROK NULL "failed to load module"
+		exit 2
+	fi
+
+	sleep $test_time
+
+	rmmod rcutorture > /dev/null 2>&1
+	if [ $? -ne 0 ]; then
+		tst_brkm TBROK NULL "failed to unload module"
+		exit 2
+	fi
+
+	# check module status in dmesg
+	result_str=`dmesg | tail -n 1`
+	echo "$result_str" | grep SUCCESS > /dev/null 2>&1
+	if [ $? -eq 0 ]; then
+		tst_resm TPASS "tc$TST_COUNT: completed"
+	else
+		tst_resm TFAIL "tc$TST_COUNT: failed, please check dmesg"
+		RC=1
+	fi
+	TST_COUNT=$(( $TST_COUNT + 1 ))
+done
+
+exit $RC
-- 
1.7.1


------------------------------------------------------------------------------
Managing the Performance of Cloud-Based Applications
Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
Read the Whitepaper.
http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/ostg.clktrk
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: [LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[chrubis]

Hi!
> One of the possible ways to test RCU is to use rcutorture kernel module.
> The test requires that kernel configured with CONFIG_RCU_TORTURE_TEST.
> It runs rcutorture module using particular options and then inspects
> dmesg output for module's test results.
> For more information, please read Linux Documentation: RCU/torture.txt
> 
> Signed-off-by: Alexey Kodanev <alexey.kodanev@oracle.com>
> ---
>  runtest/kernel_misc                                |    1 +
>  testcases/kernel/device-drivers/Makefile           |    1 +
>  testcases/kernel/device-drivers/rcu/Makefile       |   22 ++++
>  testcases/kernel/device-drivers/rcu/rcu_torture.sh |  120 ++++++++++++++++++++
>  4 files changed, 144 insertions(+), 0 deletions(-)
>  create mode 100644 testcases/kernel/device-drivers/rcu/Makefile
>  create mode 100755 testcases/kernel/device-drivers/rcu/rcu_torture.sh
> 
> diff --git a/runtest/kernel_misc b/runtest/kernel_misc
> index c1826fb..8f37cc6 100644
> --- a/runtest/kernel_misc
> +++ b/runtest/kernel_misc
> @@ -7,3 +7,4 @@ tbio tbio
>  ltp_acpi ltp_acpi
>  ltp_acpi_cpufreq ltp_acpi_cpufreq
>  uaccess uaccess
> +rcu_torture.sh rcu_torture.sh
> diff --git a/testcases/kernel/device-drivers/Makefile b/testcases/kernel/device-drivers/Makefile
> index d207736..37b6c64 100644
> --- a/testcases/kernel/device-drivers/Makefile
> +++ b/testcases/kernel/device-drivers/Makefile
> @@ -21,6 +21,7 @@ include $(top_srcdir)/include/mk/env_pre.mk
>  SUBDIRS		:= acpi \
>  		   block \
>  		   pci \
> +		   rcu \
>  		   rtc \
>  		   tbio \
>  		   uaccess
> diff --git a/testcases/kernel/device-drivers/rcu/Makefile b/testcases/kernel/device-drivers/rcu/Makefile
> new file mode 100644
> index 0000000..96c6b5b
> --- /dev/null
> +++ b/testcases/kernel/device-drivers/rcu/Makefile
> @@ -0,0 +1,22 @@
> +# Copyright (c) 2014 Oracle and/or its affiliates. All Rights Reserved.
> +#
> +# 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 2 of
> +# the License, or (at your option) any later version.
> +#
> +# This program is distributed in the hope that it would 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, write the Free Software Foundation,
> +# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
> +
> +top_srcdir	?= ../../../..
> +include $(top_srcdir)/include/mk/testcases.mk
> +
> +INSTALL_TARGETS		:= rcu_torture.sh
> +
> +include $(top_srcdir)/include/mk/generic_leaf_target.mk
> diff --git a/testcases/kernel/device-drivers/rcu/rcu_torture.sh b/testcases/kernel/device-drivers/rcu/rcu_torture.sh
> new file mode 100755
> index 0000000..c3db889
> --- /dev/null
> +++ b/testcases/kernel/device-drivers/rcu/rcu_torture.sh
> @@ -0,0 +1,120 @@
> +#!/bin/sh
> +# Copyright (c) 2014 Oracle and/or its affiliates. All Rights Reserved.
> +#
> +# This program is free software; you can redistribute it and/or
> +# modify it under the terms of the GNU General Public License as
> +# published by the Free Software Foundation; either version 2 of
> +# the License, or (at your option) any later version.
> +#
> +# This program is distributed in the hope that it would be useful,
> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> +# GNU General Public License for more details.
> +#
> +# You should have received a copy of the GNU General Public License
> +# along with this program; if not, write the Free Software Foundation,
> +# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
> +#
> +# Author: Alexey Kodanev <alexey.kodanev@oracle.com>
> +#
> +# One of the possible ways to test RCU is to use rcutorture kernel module.
> +# The test requiers that kernel configured with CONFIG_RCU_TORTURE_TEST.
> +# It runs rcutorture module using particular options and then inspects
> +# dmesg output for module's test results.
> +# For more information, please read Linux Documentation: RCU/torture.txt
> +
> +export RC=0
> +export TST_TOTAL=14
> +export TCID="rcu_torture"
> +export TST_COUNT=0
> +# default options
> +test_time=60
> +num_readers=10
> +num_writers=5
> +
> +while getopts :ht:r:w: opt; do
> +	case "$opt" in
> +	h)
> +		echo "Usage:"
> +		echo "h        help"
> +		echo "t x      time in seconds for each test-case"
> +		echo "r x      number of readers"
> +		echo "w x      number of writers"
> +		exit 0
> +	;;
> +	t) test_time=$OPTARG ;;
> +	r) num_readers=$OPTARG ;;
> +	w) num_writers=$OPTARG ;;
> +	*)
> +		tst_brkm TBROK NULL "unknown option: $opt"
> +		exit 2
> +	;;
> +	esac
> +done
> +
> +cleanup()
> +{
> +	rmmod rcutorture > /dev/null 2>&1
> +}
> +
> +if [ "`id -u`" -ne 0 ]; then
> +	tst_brkm TCONF NULL "Test must be run as root"
> +	exit 0
> +fi
> +
> +# check if module is present
> +modprobe rcutorture > /dev/null 2>&1
> +if [ $? -ne 0 ]; then
> +	tst_brkm TCONF NULL "Test requiers rcutorture module"
> +	exit 0
> +fi
> +cleanup
> +
> +trap cleanup EXIT
> +
> +rcu_type="rcu rcu_sync rcu_expedited rcu_bh rcu_bh_sync rcu_bh_expedited \
> +          srcu srcu_sync srcu_expedited srcu_raw srcu_raw_sync sched \
> +          sched_sync sched_expedited"
> +
> +est_time=`echo "scale=2; $test_time * $TST_TOTAL / 60 " | bc`
> +tst_resm TINFO "estimate time $est_time min"
> +
> +for type in $rcu_type; do
> +
> +	tst_resm TINFO "tc$TST_COUNT: $type running ($test_time sec)..."
> +
> +	modprobe rcutorture nreaders=$num_readers nfakewriters=$num_writers \
> +	         stat_interval=60 test_no_idle_hz=1 shuffle_interval=3 \
> +	         stutter=5 irqreader=1 fqs_duration=0 fqs_holdoff=0 \
> +	         fqs_stutter=3 test_boost=1 test_boost_interval=7 \
> +	         test_boost_duration=4 shutdown_secs=0 \
> +	         stall_cpu=0 stall_cpu_holdoff=10 n_barrier_cbs=0 \
> +	         onoff_interval=0 onoff_holdoff=0 torture_type=$type \
> +	         > /dev/null 2>&1
> +
> +	if [ $? -ne 0 ]; then
> +		tst_brkm TBROK NULL "failed to load module"
> +		exit 2
> +	fi
> +
> +	sleep $test_time
> +
> +	rmmod rcutorture > /dev/null 2>&1
> +	if [ $? -ne 0 ]; then
> +		tst_brkm TBROK NULL "failed to unload module"
> +		exit 2
> +	fi
> +
> +	# check module status in dmesg
> +	result_str=`dmesg | tail -n 1`

This is quite fragile, if anything has added a line into dmesg while the
test was running it will fail. Better approach would be to save the
dmesg output into a file and grep it for the rcutorture resutl.

> +	echo "$result_str" | grep SUCCESS > /dev/null 2>&1
> +	if [ $? -eq 0 ]; then
> +		tst_resm TPASS "tc$TST_COUNT: completed"
> +	else
> +		tst_resm TFAIL "tc$TST_COUNT: failed, please check dmesg"
> +		RC=1

What about using the test.sh library I've send as a RFC into the mailing
list instead? Have you seen the code? It's modeled more closely to the C
interface and stores the exit value internally, manages the TST_COUNT,
etc...

(http://sourceforge.net/p/ltp/mailman/message/31941597/)

> +	fi
> +	TST_COUNT=$(( $TST_COUNT + 1 ))
> +done
> +
> +exit $RC
> -- 
> 1.7.1
> 
> 
> ------------------------------------------------------------------------------
> Managing the Performance of Cloud-Based Applications
> Take advantage of what the Cloud has to offer - Avoid Common Pitfalls.
> Read the Whitepaper.
> http://pubads.g.doubleclick.net/gampad/clk?id=121051231&iu=/4140/ostg.clktrk
> _______________________________________________
> Ltp-list mailing list
> Ltp-list@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ltp-list

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: [LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[chrubis]

Hi!
> > +	rmmod rcutorture > /dev/null 2>&1
> > +	if [ $? -ne 0 ]; then
> > +		tst_brkm TBROK NULL "failed to unload module"
> > +		exit 2
> > +	fi
> > +
> > +	# check module status in dmesg
> > +	result_str=`dmesg | tail -n 1`
> 
> This is quite fragile, if anything has added a line into dmesg while the
> test was running it will fail. Better approach would be to save the
> dmesg output into a file and grep it for the rcutorture resutl.

And the best possible solution would be to add a more convinient
interface to the module itself, but that would need to be comunicated
with kernel upstream.

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: [LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[chrubis]

Hi!
> >>> +	rmmod rcutorture > /dev/null 2>&1
> >>> +	if [ $? -ne 0 ]; then
> >>> +		tst_brkm TBROK NULL "failed to unload module"
> >>> +		exit 2
> >>> +	fi
> >>> +
> >>> +	# check module status in dmesg
> >>> +	result_str=`dmesg | tail -n 1`
> >> This is quite fragile, if anything has added a line into dmesg while the
> >> test was running it will fail. Better approach would be to save the
> >> dmesg output into a file and grep it for the rcutorture resutl.
> > And the best possible solution would be to add a more convinient
> > interface to the module itself, but that would need to be comunicated
> > with kernel upstream.
> Messages quite informative and could be parsed easily. Do you mean 
> adding sysfs parameters to control the module?

Yes. I would rather see the test writing into a sysfs file and be
blocked in the write till the test is finished which is more robust than
the sleep it does currently.

So for the time being I would go with the test as it is and try to
start a discussion about making the module more friendly to automated
testing on LKML. Does that sound reasonably?

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: [LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[chrubis]

Hi!
> > So for the time being I would go with the test as it is and try to
> > start a discussion about making the module more friendly to automated
> > testing on LKML. Does that sound reasonably?
> Yes it does.
> 
> One more thing, e.g we have kernel which is not configured with 
> rcutorture but we have kernel-devel. What if we copy rcutorture source 
> file to LTP and build it when the above condition is true?

I would like to avoid adding the module source into the LTP git tree,
because that would bring another piece of code depending on unstable
kernel API which will fail to compile sooner or later.

And trying to install kernel source package and locating the module
source is IMHO overcomplicated.

Or do you have a better ide how to do this?

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: [LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[chrubis]

Hi!
> >> +
> >> +	rmmod rcutorture > /dev/null 2>&1
> >> +	if [ $? -ne 0 ]; then
> >> +		tst_brkm TBROK NULL "failed to unload module"
> >> +		exit 2
> >> +	fi
> >> +
> >> +	# check module status in dmesg
> >> +	result_str=`dmesg | tail -n 1`
> > This is quite fragile, if anything has added a line into dmesg while the
> > test was running it will fail. Better approach would be to save the
> > dmesg output into a file and grep it for the rcutorture resutl.
> Why copy it to a file? I'm thinking about changing to this command, it 
> will return the last test result.
>
> result_str=`dmesg | sed -nE '$s/.*End of test: ([A-Z]+):.*/\1/p'`

I was just thinking how to make the race window as small as possible. In
order to do that I would save the output right after the test has
finished etc.

But it's quite unlikely that the whole kernel log buffer will be filled
with messages meanwhile.

> Don't need the last grep as well, will be as follows:
> if [ "$result_str" == "SUCCESS" ]; then
> ...
> >> +	echo "$result_str" | grep SUCCESS > /dev/null 2>&1
> >> +	if [ $? -eq 0 ]; then
> >> +		tst_resm TPASS "tc$TST_COUNT: completed"
> >> +	else
> >> +		tst_resm TFAIL "tc$TST_COUNT: failed, please check dmesg"
> >> +		RC=1
> > What about using the test.sh library I've send as a RFC into the mailing
> > list instead? Have you seen the code? It's modeled more closely to the C
> > interface and stores the exit value internally, manages the TST_COUNT,
> > etc...
> >
> > (http://sourceforge.net/p/ltp/mailman/message/31941597/)
> No, I missed the attachment :(

Sorry, I forgot to change the attachement to be inlined.

> Well done! I've tried it today with the rcutorture test,  have a few 
> questions:
> - tst_require_root returns with TBROK, I thought it should be TCONF 
> instead, shouldn't it?

Right, thanks for noticing.

> - LTP_TST_COUNT incremented on each call to tst_resm(), would it be 
> better to count only PASS/FAIL calls as it is in ltplib?

I've noticed this too a few days ago.

I will fix these problems and add push it into LTP git and update
test-writing-guidelines.

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: [LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[chrubis]

Hi!
> >>> This is quite fragile, if anything has added a line into dmesg while the
> >>> test was running it will fail. Better approach would be to save the
> >>> dmesg output into a file and grep it for the rcutorture resutl.
> >> Why copy it to a file? I'm thinking about changing to this command, it
> >> will return the last test result.
> >>
> >> result_str=`dmesg | sed -nE '$s/.*End of test: ([A-Z]+):.*/\1/p'`
> > I was just thinking how to make the race window as small as possible. In
> > order to do that I would save the output right after the test has
> > finished etc.
> >
> > But it's quite unlikely that the whole kernel log buffer will be filled
> > with messages meanwhile.
> >
> Got it, so I will save it to a file, then parse the file with sed, OK?

Giving it second though piping it to sed directly is also acceptable.
The kernel log buffer is not that small and the code will be a bit less
complicated.

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: [LTP] [PATCH] device-drivers/rcu: add wrapper script for rcutorture test

[chrubis]

Hi!
> I've noticed this too a few days ago.
> 
> I will fix these problems and add push it into LTP git and update
> test-writing-guidelines.

I've fixed one more thing (masked out the TINFO, TCONF and TRETR out of
the test return value as the C library does) and updated the
test-writing-guidelines.

-- 
Cyril Hrubis
chrubis@suse.cz

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list