Minutes of SystemTap Conference Call, Nov 3, 2005

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Minutes of SystemTap Conference Call, Nov 3, 2005

Kevin Stafford
Minutes of SystemTap Conference Call, Nov 3, 2005

Participants:
IBM:
        Ananth Mavinakayanahalli ([hidden email] -- Boston)
        Jim Keniston ([hidden email] -- Beaverton)
        Kevin Stafford ([hidden email] -- Beaverton)
        Hien Nguyen ([hidden email] -- Beaverton)
Red Hat:
        Elena Zannoni ([hidden email] -- Boston)
        Martin Hunt ([hidden email] -- Seattle)
        Frank Eigler ([hidden email] -- Toronto)
        Will Cohen ([hidden email] -- Raleigh)
Hitachi:
        Satoshi Oshima ([hidden email])

Actions Required (ARs):
 >From 5/19/05:

23. Jim, Martin: Ensure that the runtime's stack-trace feature works
    appropriately when there are return probes in the call stack.
    [Update: Martin is working on this.  Jim will prototype an improvement
    whereby the saved IP is updated before the handler is called.]

 >From 6/9/05:
26. Rusty/Brad: Forward email address of Intel test person to Jay
    Turner.  [Update 10/6/05: Will be done now that Intel has Jay's email
    address: [hidden email]]

 >From 6/23/05:
30. Rusty: Review Will's rendition of Rusty's test that exercises
    probes on a wide variety of instruction types.
    [Update: As of 7/7/05, Will still needs to hear back from Rusty.]

 >From 7/28/05:
41. Anil: Post a summary of known djprobes implementation issues.
    [Update 10/6/05: Defer at least until djprobes discussion settles
    down a bit.]

 >From 8/11/05:
44. Tom: Automate relayfs tests and post them to mailing list.
    [Update: In progress]

 >From 8/18/05:

53. Anil: Test RCU-based SMP scalability feature on ia64.
    [Update: continue testing on 32 way machine and post to ML]

 >From 8/25/05:

55. Elena: Investigate approaches to making the full SystemTap bug
    list available to SystemTap developers outside Red Hat.
    [Update: generally this is found at http://bugzilla.redhat.com/bugzilla/
     This AR will be removed as of next week.]

 >From 9/8/05:

57. Martin: Runtime build issues fix or unlink documentation from
Runtime webpage.

 >From 9/22/05:

73. Prasanna: post user space probes paper to ML
    [Update: Will post to ML after 10/11-14/05 conference]

76. Hien: post dumpstack routine that solves kallsyms_lookup PPC issue.

77. Frank: post prototype to do symbol lookup in the translator.
(alternative
    to #76)
    [Done: as part of bz #1332.]

 >From 9/29/05:

83. Kevin: Post SystemTap documentation, mirror examples on website.
    [Update: stap manpages mirrored on site, package requirement list
    to come (see #86)]

 >From 10/6/05

86. Kevin: Compile list of known SystemTap package requirements to post to
    website.
    [Update: internal review of list in progress]

 >From 10/20/05

89. Hien: Investigate new PPC64 stap symbol resolution mechanism.

90. Prasanna: Complete user-space probes (preferably RCU enhancement
    compliant) design/implementation in time for face-2-face.

91. Elena: Investigate solutions to elfutils license situation.

 From 11/03/05

96. Hien: Run the testsuite for PPC64 stap build and post results to ML.
    [Done: posted 11/04/05]

97. Jim: Open a BZ for a feature request that would give the probe language
    the ability to specify maxactive return probe instances.
    [DONE 11/3/05: #1802]

Martin has replaced the Runtime Map API with an improved design. Graydon has
altered the translator accordingly to take advantage of the new API.

Masami has implemented an updated i386 Djprobes patch set, and is working on
an x86_64 port. He awaits review at the face-2-face before submitting to
LKML.

Hien has submitted ppc64 runtime patches to the mailing list. The patch
relies
on kallsyms, requests that kallsyms gets exported in RHEL kernels. He is
also
exploring other avenues for symbol lookup.

U3 updates were discussed.

Frank is concerned about kretprobe instabilities on particular functions
that
are known to be dangerous probe points (exec,etc). He requests some fresh
thinking on the matter, before a blacklist solution is accepted.

F2F meeting agenda items reviewed, topic presenters chosen.

--
Kevin Stafford
DES 2 | MS 2M3
Beaverton - OR
Linux Technology Center
IBM Systems & Technology
Phone: 1-503-578-3039
Email: [hidden email]