This is the mail archive of the systemtap@sources.redhat.com mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Minutes of 5/19/05 meeting


Minutes of SystemTAP Conference Call, May 19, 2005

Participants:
IBM:
	Larry Kessler (lkessler@us.ibm.com -- Beaverton, OR)
	Vara Prasad (varap@us.ibm.com -- Beaverton)
	Jim Keniston (jkenisto@us.ibm.com -- Beaverton)
	Ananth Mavinakayanahalli (amavin@redhat.com, ananth@in.ibm.com --
Boston)
	Prasanna Panchamukhi (prasanna@in.ibm.com -- Bangalore)
	Maneesh Soni (maneesh@in.ibm.com -- Bangalore)
Red Hat:
	Will Cohen (wcohen@redhat.com -- Raleigh)
	Martin Hunt (hunt@redhat.com -- Seattle)
	Andrew Cagney (cagney@redhat.com -- Toronto)
Intel:
	Brad Chen (brad.chen@intel.com -- Santa Clara)
	Rusty Lynch (rusty.lynch@intel.com -- Hillsboro, OR)
	Anil KeshavaMurthy (anil.s.keshavamurthy@intel.com -- Hillsboro)
	Charles Spirakis (charles.spirakis@intel.com)
	Chris Elford 

Actions Required (ARs):
>From 5/12/05:
17. Vara: Publish new call-in number via private email.  [Update:
The current number should work at least through the 5/26/05 call.]

18. All: Verify that your email address, as published in these
minutes, is correct.

>From 5/19/05:
19. Vara: Collect sample kprobes modules and post them to the
SystemTap web site.

20. Rusty: Create a kprobes test for ia64, per his 5/12/05 design,
that tests probes on a large subset of the ia64 instruction set.
Work with Prasanna to create such tests for other architectures.

21. Rusty, Will: Look into adding kprobes tests to LDP and the
Red Hat suite, respectively.

22. Rusty: Look into x86_64 limitation re: temporary disarming of
reentrant probes.

23. Jim, Martin: Ensure that the runtime's stack-trace feature works
appropriately when there are return probes in the call stack.


The ia64 port of kprobes is looking good.  It is undergoing testing
and internal code review.  It includes support for jprobes, but not
return probes yet.  Rusty's hoping to post this to LKML in the next
couple of days.

Any guess as to when Linus will release v2.6.12?  Will he be less
likely to accept patches after that?  Probably not.

There is a perceived need for:
- a kprobes test suite -- e.g., a suite we could add to LTP (Linux
Test Project)
- kprobes documentation

For documentation, we typically point people at Will's article in
Red Hat Magazine
(http://www.redhat.com/magazine/005mar05/features/kprobes/)
and Prasanna's article in developerWorks
(http://www-106.ibm.com/developerworks/library/l-kprobes.html?ca=dgr-lnxw42Kprobe).
It was suggested that we post a how-to article for inclusion in the
kernel Documentation directory.  Vara acknowledged that he had an AR to
collect sample kprobe modules and post them on the SystemTap web page.

Regarding a test suite, Rusty has experience adding tests to
LTP, and is willing to lead an effort to add a kprobes suite
to LTP.  Rusty reviewed his 5/12/05 suggestion for exhaustive,
architecture-specific kprobes tests ("Added kprobe functional test
module to CVS").  He is creating such a test for ia64, and will
work with Prasanna to create similar tests for other architectures.
Will will look into incorporating these tests into Red Hat's suite.

Prasanna has implemented temporary disarming of a reentrant probe,
and has posted patches for i386, x86_64, ppc64, and sparc64 to LKML.
There's a limitation re: post_handlers on x86_64; Rusty will get
an expert at Intel to look into this.

Ananth has a fix that allows a jprobe and multiple kprobes to
coexist at the same address.  Prasanna has suggested an improvement.
They hope to post the fix to LKML by early next week.

Ananth has return probes working for ppc64, and will post to LKML
and other appropriate lists soon.

Ananth expects to defer his SMP scalability patch(es) until after
RHEL4 U2.  He has a working prototype, but it needs serious review
and it needs to be updated to reflect recently added features such
as return probes.  Prasanna has offered to consult.

There was some discussion of the mechanics of completing the OLS
paper.  We seem to be in good shape, thanks in part to John
Lockhart's indulgence regarding the submission deadline.

Martin's runtime work is checked into CVS, although the documentation
is out of date.  He has added a stack-trace feature, among others. Jim
and Martin need to ensure that this feature will work appropriately
even when a return probe is in the call stack.

The x86_64 return-probes port is taking a back seat the the ia64
kprobes/jprobes port.

Vara has been working on the OLS paper all week.  Take 2 of his
proposal for C tapsets should be posted early next week.

Brad has been very interested in testing.  He will check a proposal
into CVS soon.  Charles has been working with Victoria Gromova on the
event-based tapset.  The Red Hat Summit (?) in New Orleans June 1-3
was mentioned.

Who will be at OLS?  Will and Vara will present the SystemTap paper.
Andrew will present a debugger-related paper.  Brad, Anil, and Rusty
will also be there.

The Beaverton IBM people (Larry, Vara, Hien, Jim) will probably miss
the 5/26 call.



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]