[Bug tapsets/7030] New: signal tapset may be referring to inline functions.

[Bug tapsets/7030] New: signal tapset may be referring to inline functions.

[srikar at linux dot vnet dot ibm dot com]

__group_send_sig_info and specific_send_sig_info seem to be inlined in 2.6.27
kernels. 
__group_send_sig_info and specific_send_sig_info are now just wrapper functions
for send_signal.
Though this issue has been talked in
http://sources.redhat.com/bugzilla/show_bug.cgi?id=1155, we could fix this by
make changes in signal.stp to refer to send_signal rather than
__group_send_sig_info and specific_send_sig_info. 

Currently the failure message when running  sig-by-pid.stp on 2.6.26+ kernel:
stap sig-by-pid.stp 
semantic error: failed to retrieve location attribute for local 'sig'
(dieoffset: 0x4e01ef): identifier '$sig' at
/usr/share/systemtap/tapset/signal.stp:92:11
semantic error: failed to retrieve location attribute for local 't' (dieoffset:
0x4e01e5): identifier '$t' at :93:12
semantic error: failed to retrieve location attribute for local 'sig'
(dieoffset: 0x4e031b): identifier '$sig' at :92:11
semantic error: failed to retrieve location attribute for local 't' (dieoffset:
0x4e0311): identifier '$t' at :93:12
semantic error: failed to retrieve location attribute for local 'sig'
(dieoffset: 0x4e17ef): identifier '$sig' at :92:11
semantic error: failed to retrieve location attribute for local 't' (dieoffset:
0x4e17e5): identifier '$t' at :93:12
semantic error: failed to retrieve location attribute for local 'sig'
(dieoffset: 0x4e1fa3): identifier '$sig' at :92:11
semantic error: failed to retrieve location attribute for local 't' (dieoffset:
0x4e1f99): identifier '$t' at :93:12
semantic error: failed to retrieve location attribute for local 'sig'
(dieoffset: 0x4e01ef): identifier '$sig' at :92:11
Pass 2: analysis failed.  Try again with more '-v' (verbose) options.

sig-by-pid.stp is a pretty popular script with customers and It got a mention in
a developerworks article.

-- 
           Summary: signal tapset may be referring to inline functions.
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: tapsets
        AssignedTo: systemtap at sources dot redhat dot com
        ReportedBy: srikar at linux dot vnet dot ibm dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=7030

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug tapsets/7030] signal tapset may be referring to inline functions.

[mjw at redhat dot com]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mjw at redhat dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=7030

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

[Bug tapsets/7030] signal tapset may be referring to inline functions.

[prasad at linux dot vnet dot ibm dot com]

------- Additional Comments From prasad at linux dot vnet dot ibm dot com  2008-11-18 09:51 -------
The behaviour described below can be attributed to the fact that probe 
_signal.send.part1 and probe _signal.send.part4 which respectively point to the 
kernel functions __group_send_sig_info() and specific_send_sig_info() have been 
inlined.

The sudden inlining of these funtions could be the result of changes made to 
these functions through commit 4cd4b6d4e0372075f846feb85aea016cbdbfec4c that 
made them a mere wrapper over send_signal(). Following this change, and due to 
the fact that the wrapper functions are being inlined by the compiler, we shall 
point these probe aliases directly to send_signal() with the appropriate names 
and function parameters. The probe aliases are restricted to kernel versions 
newer than 2.6.25.

On application of the attached patch, the script sig_by_pid.stp is found to 
work fine on the Fedora kernel 2.6.27.5-101.fc10.i686. It is pending testing on 
a kernel older than 2.6.25 though. Requesting comments, if any.


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=7030

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.