* [Bug uprobes/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
@ 2012-10-27 16:11 ` tschan+sourceware at devzone dot ch
2012-10-27 16:22 ` [Bug translator/14774] " fche at redhat dot com
` (6 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: tschan+sourceware at devzone dot ch @ 2012-10-27 16:11 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
--- Comment #1 from Daniel Tschan <tschan+sourceware at devzone dot ch> 2012-10-27 16:10:51 UTC ---
Created attachment 6704
--> http://sourceware.org/bugzilla/attachment.cgi?id=6704
Proposed bug fix
Tracked down the bug to the method iterate_over_srcfile_lines in dwflpp.cxx by
using systemtap on itself:
if (line_type == RELATIVE)
{
// body omitted
}
else if (line_type == WILDCARD)
function_line (&lineno);
else if (line_type == RANGE) { /* correct lineno */
// body ommited
}
for (int l = lineno; ; l = l + 1)
{
// body ommited
}
If both function name and line number are wildcards the above code matches a
"random" function in the source file an sets lineno to its start. The following
code then only looks a this matched function an all functions defined later in
the source file.
The attached patch seems to fix this problem.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug translator/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
2012-10-27 16:11 ` [Bug uprobes/14774] " tschan+sourceware at devzone dot ch
@ 2012-10-27 16:22 ` fche at redhat dot com
2012-10-27 22:38 ` tschan+sourceware at devzone dot ch
` (5 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: fche at redhat dot com @ 2012-10-27 16:22 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
Frank Ch. Eigler <fche at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |fche at redhat dot com
Component|uprobes |translator
--- Comment #2 from Frank Ch. Eigler <fche at redhat dot com> 2012-10-27 16:22:31 UTC ---
That looks correct to me.
Could you run the testsuite before & after your fix, to see if there are any
differences? Your fix might change the results. If not, would you like to try
adding a new systemtap.base/* test case?
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug translator/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
2012-10-27 16:11 ` [Bug uprobes/14774] " tschan+sourceware at devzone dot ch
2012-10-27 16:22 ` [Bug translator/14774] " fche at redhat dot com
@ 2012-10-27 22:38 ` tschan+sourceware at devzone dot ch
2012-10-27 23:38 ` fche at redhat dot com
` (4 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: tschan+sourceware at devzone dot ch @ 2012-10-27 22:38 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
--- Comment #3 from Daniel Tschan <tschan+sourceware at devzone dot ch> 2012-10-27 22:38:35 UTC ---
I tried running the test suite. But apparently 8 GB RAM are not enough to run
all the tests and I have no machine available that has more RAM. As far as the
test suite did run (it stopped at semko/nodwf03.stp) there were no differences
with and without the fix.
I'll try to write a test case though.
Meanwhile I noticed that lineno is zero-based. So it should be set to 0, not 1
if the function name is a wildcard. I'll update the patch together with the
test case.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug translator/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
` (2 preceding siblings ...)
2012-10-27 22:38 ` tschan+sourceware at devzone dot ch
@ 2012-10-27 23:38 ` fche at redhat dot com
2012-10-29 20:01 ` fche at redhat dot com
` (3 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: fche at redhat dot com @ 2012-10-27 23:38 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
--- Comment #4 from Frank Ch. Eigler <fche at redhat dot com> 2012-10-27 23:37:33 UTC ---
(FWIW, we run the testsuite on 1-2 GB machines regularly.)
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug translator/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
` (3 preceding siblings ...)
2012-10-27 23:38 ` fche at redhat dot com
@ 2012-10-29 20:01 ` fche at redhat dot com
2012-10-30 11:52 ` tschan+sourceware at devzone dot ch
` (2 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: fche at redhat dot com @ 2012-10-29 20:01 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
Frank Ch. Eigler <fche at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
--- Comment #5 from Frank Ch. Eigler <fche at redhat dot com> 2012-10-29 20:01:24 UTC ---
commit 78e2a671 fixes this.
A test case would still be nice.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug translator/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
` (4 preceding siblings ...)
2012-10-29 20:01 ` fche at redhat dot com
@ 2012-10-30 11:52 ` tschan+sourceware at devzone dot ch
2012-10-30 15:08 ` fche at redhat dot com
2012-11-03 19:19 ` tschan+sourceware at devzone dot ch
7 siblings, 0 replies; 9+ messages in thread
From: tschan+sourceware at devzone dot ch @ 2012-10-30 11:52 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
--- Comment #6 from Daniel Tschan <tschan+sourceware at devzone dot ch> 2012-10-30 11:51:54 UTC ---
Created attachment 6709
--> http://sourceware.org/bugzilla/attachment.cgi?id=6709
Differences of test suite runs before and after patch.
Thank you for committing the patch. Test case to follow.
Turns out that I have to run the test suite as root, otherwise the semko.exp
test case runs indefinitely, or rather till the system runs out of memory. Now
after the complete test suite ran through I see some differences before and
after my fix. There are 4 fewer unexpected failures, see attached diff.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug translator/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
` (5 preceding siblings ...)
2012-10-30 11:52 ` tschan+sourceware at devzone dot ch
@ 2012-10-30 15:08 ` fche at redhat dot com
2012-11-03 19:19 ` tschan+sourceware at devzone dot ch
7 siblings, 0 replies; 9+ messages in thread
From: fche at redhat dot com @ 2012-10-30 15:08 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
--- Comment #7 from Frank Ch. Eigler <fche at redhat dot com> 2012-10-30 15:07:39 UTC ---
> Turns out that I have to run the test suite as root, otherwise the semko.exp
> test case runs indefinitely, or rather till the system runs out of memory.
That's not supposed to happen; will try to reproduce that here.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug translator/14774] process.statement("*@file:*") probes don't match all functions
2012-10-27 15:14 [Bug uprobes/14774] New: process.statement("*@file:*") probes don't match all functions tschan+sourceware at devzone dot ch
` (6 preceding siblings ...)
2012-10-30 15:08 ` fche at redhat dot com
@ 2012-11-03 19:19 ` tschan+sourceware at devzone dot ch
7 siblings, 0 replies; 9+ messages in thread
From: tschan+sourceware at devzone dot ch @ 2012-11-03 19:19 UTC (permalink / raw)
To: systemtap
http://sourceware.org/bugzilla/show_bug.cgi?id=14774
--- Comment #8 from Daniel Tschan <tschan+sourceware at devzone dot ch> 2012-11-03 19:19:01 UTC ---
Created attachment 6713
--> http://sourceware.org/bugzilla/attachment.cgi?id=6713
Proposed test case
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread