gmane.linux.ltp

[PATCH] mq_send/5-1.c: Fix race conditions in the test

Brandon Philips — 2008-12-03T11:56:43

This test is racy. It relies on signals interrupting sleeps to do synchronization between processes. Reported failure: https://bugzilla.novell.com/show_bug.cgi?id=375897 Instead use a synchronization pipe to have the child thread send messages to the parent thread (with reasonable timeouts since this is testing blocking calls). # Apply to root of ltp-full-20081031 tree $ cd testcases/open_posix_testsuite/ $ make conformance/interfaces/mq_send/5-1.test $ conformance/interfaces/mq_send/5-1.test Signed-off-by: Brandon Philips public.gmane.org> --- testcases/open_posix_testsuite/conformance/interfaces/mq_send/5-1.c | 53 +++--- testcases/open_posix_testsuite/include/posixtest.h | 81 ++++++++++ 2 files changed, 106 insertions(+), 28 deletions(-) Index: ltp-full-20081031/testcases/open_posix_testsuite/conformance/interfaces/mq_send/5-1.c =================================================================== --- ltp-full-20081031.orig/testcases/open_posix_testsuite/c

Re: [PATCH] mmapstress03: constants should be of type 'long'

Mike Frysinger — 2008-12-02T20:12:01

------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/_______________________________________________ Ltp-list mailing list Ltp-list-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f< at >public.gmane.org https://lists.sourceforge.net/lists/listinfo/ltp-list

Re: mmapstress03: errno = 12: couldn't brk back over holes

Dmitry Guryanov — 2008-12-02T09:32:07

I've fixed your issue with patch http://sourceforge.net/mailarchive/message.php?msg_name=200812021159.10177.dguryanov%40parallels.com

[PATCH] mmapstress03: constants should be of type 'long'

Dmitry Guryanov — 2008-12-02T08:59:09

sbrk has arguments of type intptr_t, which is long in on x86_64. There is following line in mmapstress03: if (sbrk(-NUM_SEGS * pagesize) == NEG1) { here the type of argument is int, because of constant NUM_SEGS, and this test fails here on x86_64. So let constants will be long. Index: testcases/kernel/mem/mmapstress/mmapstress03.c =================================================================== RCS file: /cvsroot/ltp/ltp/testcases/kernel/mem/mmapstress/mmapstress03.c,v retrieving revision 1.3 diff -u -p -r1.3 mmapstress03.c --- testcases/kernel/mem/mmapstress/mmapstress03.c15 Feb 2006 06:34:46 -0000 1.3 +++ testcases/kernel/mem/mmapstress/mmapstress03.c2 Dec 2008 08:54:07 -0000 < at >< at > -58,11 +58,11 < at >< at > int anyfail(); void ok_exit(); /***** ** ** *****/ -#define AS_SVSM_VSEG_MAX48U -#define AS_SVSM_MMAP_MAX16U +#define AS_SVSM_VSEG_MAX48UL +#define AS_SVSM_MMAP_MAX16UL -#define EXTRA_VSEGS2 +#define EXTRA_VSEGS2L #define NUM_SEGS(AS_SVSM_VSEG_MAX + EXTRA_VSEGS) #define ERRO

Re: [PATCH 0/4][RT][RFC][Take#5] XML dumps for realtime tests

Gilles.Carry — 2008-12-02T07:29:02

I forgot to mention the changes from the previous patchset (take#3 to #4) API changes: - generic tag printing function + macros on top of it: more flexible - no longer need to specify tag name when closing: use of a dynamic stack to remember names - more features to handle carriage returns, indentation, self-closing tags and attributes Change of XML output grammar: - records data provided as attributes instead of nodes: easier to read and more compact Doc updates Minor fixes Gilles. Gilles Carry a écrit :

[PATCH] Perform the test in a private directory

Roy Lee — 2008-12-02T01:12:56

This test case requires write permission for the dummy program. It would fail for those who put LTP on an read-only environment. So this patch copies the dummy test program to and performs the test in a private directory. p.s. this patch copy the one Renaud Lottiaux sent for execve02.c. --- testcases/kernel/syscalls/execve/execve05.c | 47 +++++++++++++++++++++++++++ 1 files changed, 47 insertions(+), 0 deletions(-) diff --git a/testcases/kernel/syscalls/execve/execve05.c b/testcases/kernel/syscalls/execve/execve05.c index 8952d6f..ec23a43 100644 --- a/testcases/kernel/syscalls/execve/execve05.c +++ b/testcases/kernel/syscalls/execve/execve05.c < at >< at > -55,6 +55,7 < at >< at > #include #include #include +#include #include "test.h" #include "usctest.h" #include "libtestsuite.h" < at >< at > -212,9 +213,52 < at >< at > help() void setup() { +char *cmd, *dirc, *basec, *bname, *dname, *path, *pwd = NULL; +int res; + /* capture signals */ tst_sig(FORK, DEF_HANDLER, cleanup); +/

Re: [RFC] Testing Support for FILECAPS improvements in 2.6.27

Serge E. Hallyn — 2008-12-01T18:04:57

Quoting Subrata Modak (subrata-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8< at >public.gmane.org): That probably ranks lower priority than the securebits tests. That should be tested by existing tests. It's a code refactoring, not a new feature. thanks, -serge ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/

fails in syscalls/stat04_64 and syscalls/lstat01A_64 testcases

Dmitry Guryanov — 2008-12-01T17:55:54

Hi, stat04 and lstat01A are testcases of symlink01 test: from runtest/syscalls: lstat01A symlink01 -T lstat01 lstat01A_64 symlink01 -T lstat01_64 ... stat04 symlink01 -T stat04 stat04_64 symlink01 -T stat04_64 symlink01 test has no testcases with name lstat01_64 and stat04_64, so this testcases fail now. What can we do with this ?

Re: problems with signalfd testcase on different systems

Dmitry Guryanov — 2008-12-01T16:44:03

thanks

[PATCH 1/4] [RT][RFC][Take#5] New library: libxml

Gilles Carry — 2008-12-01T15:10:35

This library provides a set of function to help the saving of data as XML formatted files. Note: this is a lightweight library as it does not check if the document is well formed or grammaticallt correct. It must be considered as a commodity that helps tag indentation and avoids user to toss with '<' and '>'. To embbed it into executables, run: XML_LIB=1 make Also, it adds heading tags with data such as timestamp, system information, test conditions... This to facilitate post processing. (eg. further comparisons of different testruns, formatting for plotting...) This patch does not alter the LTP/RT traditional stats dump. A new global command line option is used: -x Compilation is conditional to LIB_XML. Signed-off-by: Gilles Carry public.gmane.org> --- testcases/realtime/config.mk | 5 + testcases/realtime/include/libxml.h | 134 ++++++++++++ testcases/realtime/lib/librttest.c | 23 ++- testcases/realtime/lib/libxml.c | 385 +++++++++++++++++++++++++++++

[PATCH 2/4] [RT][RFC][Take#5] Integration of xml into stats.

Gilles Carry — 2008-12-01T15:10:36

This patch adds a new function into libstat: xml_stats_container_save which handles an xml_stream_t instead of a file. It dumps the same data stats_container_save with XML format. Also added two fields to stats: timestamp and cpuid. Compilation is conditional to LIB_XML. --- testcases/realtime/include/libstats.h | 16 ++++++++++++ testcases/realtime/lib/libstats.c | 42 +++++++++++++++++++++++++++++++++ 2 files changed, 58 insertions(+), 0 deletions(-) diff --git a/testcases/realtime/include/libstats.h b/testcases/realtime/include/libstats.h index 05c26d8..fefad7b 100644 --- a/testcases/realtime/include/libstats.h +++ b/testcases/realtime/include/libstats.h < at >< at > -45,6 +45,10 < at >< at > #include #include #include +#include +#ifdef LIB_XML +#include +#endif #define MIN(A,B) ((A)<(B)?(A):(B)) #define MAX(A,B) ((A)>(B)?(A):(B)) < at >< at > -52,6 +56,8 < at >< at > typedef struct stats_record { long x; long y; +nsec_t timestamp; +int cpuid; } stats_record_t;

[PATCH 3/4] [RT][RFC][Take#5] matrix_mult.c: add xml dump ofstatistics and results.

Gilles Carry — 2008-12-01T15:10:37

This simply adds xml function calls to allow the dumping of results and samples into an xml file. Compilation is conditional to LIB_XML. Signed-off-by: Gilles Carry public.gmane.org> --- testcases/realtime/func/matrix_mult/matrix_mult.c | 68 +++++++++++++++++++- 1 files changed, 64 insertions(+), 4 deletions(-) diff --git a/testcases/realtime/func/matrix_mult/matrix_mult.c b/testcases/realtime/func/matrix_mult/matrix_mult.c index e27f47d..f336ac8 100644 --- a/testcases/realtime/func/matrix_mult/matrix_mult.c +++ b/testcases/realtime/func/matrix_mult/matrix_mult.c < at >< at > -43,6 +43,9 < at >< at > #include #include #include +#ifdef LIB_XML +#include +#endif #define MAX_CPUS8192 #define PRIO43 < at >< at > -137,7 +140,7 < at >< at > void matrix_mult(int m_size) } } -void matrix_mult_record(int m_size, int index) +void matrix_mult_record(int m_size, int index, int cpuid) { nsec_t start, end, delta; int i; < at >< at > -149,6 +152,8 < at >< at > void matrix_mult_r

[PATCH 4/4] [RT][RFC][Take#5] Documentation and tools for XMLlibrary

Gilles Carry — 2008-12-01T15:10:38

Documentation: just the basics on how to - run - code - compile Tooles: - xml_collect.sh Also an example of XML dump. Signed-off-by: Gilles Carry public.gmane.org> --- testcases/realtime/doc/XML | 261 ++++++++++++++++++++++++++++ testcases/realtime/scripts/xml_collect.xml | 15 ++ 2 files changed, 276 insertions(+), 0 deletions(-) create mode 100644 testcases/realtime/doc/XML create mode 100755 testcases/realtime/scripts/xml_collect.xml diff --git a/testcases/realtime/doc/XML b/testcases/realtime/doc/XML new file mode 100644 index 0000000..cd886e3 --- /dev/null +++ b/testcases/realtime/doc/XML < at >< at > -0,0 +1,261 < at >< at > +XML library +=========== + +1. Why dump results and stats with XML? +--------------------------------------- + +RT tests already generate stats and plot files. What XML can give more? + +Post processing: +XML provides an easy way to transform the data without the need to change and +recompile the test. Just create or modify an Xquery to get the req

[PATCH 0/4][RT][RFC][Take#5] XML dumps for realtime tests

Gilles Carry — 2008-12-01T15:10:34

Hello, To overcome libstats limitations, I have implemented this library and integrated it into libstats matrix_mult.c I've used this because I had to compare/correlate stats of differents tests runs on several kernels and architectures. This proved very useful as once the tests have run, many extracts can be done from original results, without having to modify/recompile/rerun the test itself. This really made things easier. With a single test run output, I could plot 2-D or 3-D graphs just by changing the xquery that generates the gnuplot data file. Basic understanding of XML is necessary to use this. I think this work is worth sharing. I wish people review it and comment it. Though it's been implemented on top of realtime tests, it should be easily portable to the reset of LTP, should some be interested... This series consists of four patches: first is the library itself, second is the integration with libstat, third is a first use with matrix_mult, forth is the documentation and a tool. ---

[ANNOUNCE] The Linux Test Project has been Released forNOVEMBER2008

Subrata Modak — 2008-12-01T09:30:22

Dear All, The Linux Test Project test suite has been released for the month of NOV 2008. The latest version of the test-suite contains 3000+ tests for the Linux OS and can be found at http://ltp.sourceforge.net/, Latest happenings in LTP can also be found at: http://ltp.sourceforge.net/wiki/, http://ltp.sourceforge.net/wikiArchives.php, and, IRC: irc.freenode.org #ltp. ======================== NOVEMBER 2008 Highlights: ======================== * Addition of 23 new _16 & _64 bit versions of syscall tests, * Addition of 1 PIDNS test, * Introduction of AUTOCONF, * Integration of various other tests to default run, * Release of LCOV-1.7, * Major updates to RT, TI-RPC & CONTROLLER tests, ------------------------------ ============================== NOVEMBER 2008 LTP Contributors: ============================== * Gilles Carry, * Darren Hart, * Mike Frysinger, * Elder Costa, * Henry Yei, * Naresh kamboju, * Alan Cox, * Sukadev, * Nageswara R Sastry, * Peter Oberparleiter, * Veerendra C, * Stephen Smalley, * Sudh

Re: [PATCH] [TEST_ADD] Interface to query tun/tap features

Rusty Russell — 2008-12-01T04:25:02

Thanks, of course! Rusty. ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/

Re: Update for the Sun-RPC /TIRPC LTP Test Suite

Subrata Modak — 2008-11-30T15:25:25

I think this messed up a little bit. Your last patch was integrated and hence this will fail to apply. I am going to do the fresh release soon. Send across all the changes w.r.t this new release. Regards-- Subrata ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/

Re: problems with signalfd testcase on different systems

Mike Frysinger — 2008-11-28T20:35:16

problems with signalfd testcase on different systems

Dmitry Guryanov — 2008-11-28T18:23:50

Hi, signalfd syscall is absent on a lot of systems. As I understand from discussion about using autoconf for this test, it's supposed to use default config on new systems and run autoconf on old ones. But required autoconf version - 2.61, rhel-5.2 for example has autoconf-2.59. I need to run ltp on different distributions and installing non-native packages is undesirable. Can we do something with it ?

Update for the Sun-RPC /TIRPC LTP Test Suite

Le Rouzic — 2008-11-28T14:34:05

Subrata Modak a écrit : Hi, If you don't mind I prefer an integrated patch I join at this mail. It contents: - contents of the first already sent patch - fix about issues you mentionned on it (ltpstress.sh, configure~) - new extensions about the build/install scripts - the adaptation of some rpc scripts to take account if rpcbind is used rather than portmap Bets Regards

Re: [PATCH] ltp iptables testcase failed

Subrata Modak — 2008-11-28T11:52:11

Merged. Regards-- Subrata On Wed, 2008-11-26 at 17:53 +0530, Subrata Modak wrote: ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/