* [Patch]patch for some systemtap test cases failure
@ 2006-03-06 8:57 bibo,mao
2006-03-06 13:48 ` Frank Ch. Eigler
0 siblings, 1 reply; 6+ messages in thread
From: bibo,mao @ 2006-03-06 8:57 UTC (permalink / raw)
To: systemtap
Hi,
This is patch for some indent.stp and args.exp test case failure. And I
tested in my IA64 box.
--- src/testsuite/buildok/indent.stp 2006-03-04 01:44:23.000000000 +0800
+++ /root/stap_testing_200603051706/src/testsuite/buildok/indent.stp
2006-03-06 08:42:02.000000000 +0800
@@ -1,3 +1,5 @@
+#! stap -p4
+
probe begin {
print (thread_indent (1)) print ("yo\n")
print (thread_indent (-1)) print ("ta\n")
--- tests/testsuite/systemtap.samples/args.exp 2006-02-26
05:09:05.000000000 +0800
+++
/root/stap_testing_200603051706/tests/testsuite/systemtap.samples/args.exp
2006-03-06 08:35:03.000000000 +0800
@@ -1,13 +1,18 @@
set test "args"
set stappath [exec which stap]
-set stpdpath [exec dirname $stappath]/../libexec/systemtap/stpd
+set stpdpath [exec dirname $stappath]/stpd
if [file exists $stpdpath] {
pass "$test search for stpd ($stpdpath)"
} else {
- fail "$test search for stpd"
- return
+ set stpdpath [exec dirname $stappath]/../libexec/systemtap/stpd
+ if [file exists $stpdpath] {
+ pass "$test search for stpd ($stpdpath)"
+ } else {
+ fail "$test search for stpd ($stpdpath)"
+ return
+ }
}
set modname "args_[pid]"
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Patch]patch for some systemtap test cases failure
2006-03-06 8:57 [Patch]patch for some systemtap test cases failure bibo,mao
@ 2006-03-06 13:48 ` Frank Ch. Eigler
2006-03-07 1:02 ` bibo mao
2006-03-08 15:11 ` William Cohen
0 siblings, 2 replies; 6+ messages in thread
From: Frank Ch. Eigler @ 2006-03-06 13:48 UTC (permalink / raw)
To: bibo,mao; +Cc: systemtap
"bibo,mao" <bibo.mao@intel.com> writes:
> [...]
> set stappath [exec which stap]
> -set stpdpath [exec dirname $stappath]/../libexec/systemtap/stpd
> +set stpdpath [exec dirname $stappath]/stpd
> [...]
The pass-5 tests are meant to run against an *installed* copy of
systemtap (i.e., the result of a "make install"), not the parts as
they happen to be arranged in the build tree. In the former, the
stpd binary will not be alongside stap.
- FChE
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Patch]patch for some systemtap test cases failure
2006-03-06 13:48 ` Frank Ch. Eigler
@ 2006-03-07 1:02 ` bibo mao
2006-03-07 15:00 ` William Cohen
2006-03-08 15:11 ` William Cohen
1 sibling, 1 reply; 6+ messages in thread
From: bibo mao @ 2006-03-07 1:02 UTC (permalink / raw)
To: Frank Ch. Eigler; +Cc: bibo,mao, systemtap
Frank Ch. Eigler wrote:
> "bibo,mao" <bibo.mao@intel.com> writes:
>
>> [...]
>> set stappath [exec which stap]
>> -set stpdpath [exec dirname $stappath]/../libexec/systemtap/stpd
>> +set stpdpath [exec dirname $stappath]/stpd
>> [...]
>
> The pass-5 tests are meant to run against an *installed* copy of
> systemtap (i.e., the result of a "make install"), not the parts as
> they happen to be arranged in the build tree. In the former, the
> stpd binary will not be alongside stap.
>
> - FChE
>
yes, it is. But in the nightly test script stap_testing, it defines
export PATH=${STAP_OBJ}:$PATH, and it will first search STAP_OBJ
directory, where stap is in the same directory with stpd.
Maybe the test script need modification :)
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Patch]patch for some systemtap test cases failure
2006-03-07 1:02 ` bibo mao
@ 2006-03-07 15:00 ` William Cohen
0 siblings, 0 replies; 6+ messages in thread
From: William Cohen @ 2006-03-07 15:00 UTC (permalink / raw)
To: bibo mao; +Cc: Frank Ch. Eigler, bibo,mao, systemtap
bibo mao wrote:
> Frank Ch. Eigler wrote:
>
>> "bibo,mao" <bibo.mao@intel.com> writes:
>>
>>> [...]
>>> set stappath [exec which stap]
>>> -set stpdpath [exec dirname $stappath]/../libexec/systemtap/stpd
>>> +set stpdpath [exec dirname $stappath]/stpd
>>> [...]
>>
>>
>> The pass-5 tests are meant to run against an *installed* copy of
>> systemtap (i.e., the result of a "make install"), not the parts as
>> they happen to be arranged in the build tree. In the former, the
>> stpd binary will not be alongside stap.
>>
>> - FChE
>>
> yes, it is. But in the nightly test script stap_testing, it defines
> export PATH=${STAP_OBJ}:$PATH, and it will first search STAP_OBJ
> directory, where stap is in the same directory with stpd.
> Maybe the test script need modification :)
Hi,
I made some modifications in the script Monday to do an install and use
the installed stap. You should check out the new version in tests/tools.
-Will
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Patch]patch for some systemtap test cases failure
2006-03-06 13:48 ` Frank Ch. Eigler
2006-03-07 1:02 ` bibo mao
@ 2006-03-08 15:11 ` William Cohen
2006-03-09 2:45 ` Frank Ch. Eigler
1 sibling, 1 reply; 6+ messages in thread
From: William Cohen @ 2006-03-08 15:11 UTC (permalink / raw)
To: Frank Ch. Eigler; +Cc: bibo,mao, systemtap
Frank Ch. Eigler wrote:
> "bibo,mao" <bibo.mao@intel.com> writes:
>
>
>>[...]
>> set stappath [exec which stap]
>>-set stpdpath [exec dirname $stappath]/../libexec/systemtap/stpd
>>+set stpdpath [exec dirname $stappath]/stpd
>>[...]
>
>
> The pass-5 tests are meant to run against an *installed* copy of
> systemtap (i.e., the result of a "make install"), not the parts as
> they happen to be arranged in the build tree. In the former, the
> stpd binary will not be alongside stap.
>
> - FChE
What is the logic of making the pass-5 tests being run again *installed*
copy of systemtap which the pass-[1234] tests are only run against the
build tree? It seems logical to make it possible to be able to run that
pass-[1234] tests against installed systemtap also.
-Will
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [Patch]patch for some systemtap test cases failure
2006-03-08 15:11 ` William Cohen
@ 2006-03-09 2:45 ` Frank Ch. Eigler
0 siblings, 0 replies; 6+ messages in thread
From: Frank Ch. Eigler @ 2006-03-09 2:45 UTC (permalink / raw)
To: William Cohen; +Cc: systemtap
Hi -
wcohen wrote:
> What is the logic of making the pass-5 tests being run [against]
> *installed* copy of systemtap which the pass-[1234] tests are only
> run against the build tree?
The pass-5 tests best simulate real usage. The pass-[1-4] tests are
tantamount to unit tests, which are easily (and I hope frequently) run
within unprivileged build sessions.
> It seems logical to make it possible to be able to run that
> pass-[1234] tests against installed systemtap also.
I suppose one could, but it's redundant. Chances are that any
systemtap binary that was installed will already have had the
build-time test suite run against it.
- FChE
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2006-03-09 2:45 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2006-03-06 8:57 [Patch]patch for some systemtap test cases failure bibo,mao
2006-03-06 13:48 ` Frank Ch. Eigler
2006-03-07 1:02 ` bibo mao
2006-03-07 15:00 ` William Cohen
2006-03-08 15:11 ` William Cohen
2006-03-09 2:45 ` Frank Ch. Eigler
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).