Content-type: multipart/alternative; Boundary="1__=08BBFEA7DFFBF6138f9e8a93df938690918c08BBFEA7DFFBF613" --1__=08BBFEA7DFFBF6138f9e8a93df938690918c08BBFEA7DFFBF613 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: quoted-printable Ashok: I just checked the 2.6.11 kernel tree, arch/ppc/oprofile and do not see= any call graph kernel support. When you get to the 2.6.17 (not sure exactl= y when it got added) or so time frame, there is a file in the arch/ppc/oprofile directory called backtrace.c which provides the call graph support. I only see oprofile support in the 2.6.11 kernel for th= e freescale architecture. I assume that is what you are running on. Note, the call graph support that is in the 2.6.17 and newer kernels ha= s an issue for leaf function. The leaf function is a function that does not= call another function. These leaf functions do not appear in the call graph correctly. We are looking to see if we can fix this issue in a reasonable way. Carl Love = Ashok V = = To Sent by: oprofile-list@lists.sourceforge.= net oprofile-list-bou = cc nces@lists.source = forge.net Subj= ect Opreport Crash = = 04/23/2008 07:56 = AM = = = = Hi, Thanking you in advance. I am profiling kernel 2.6.11 in powerpc 32= bit architechture and I am using oprofile-o.9.3.tar.gz. I am able to run ./oprofile without any issues but while I am trying to generate call gr= aph using the cmd ./opreport -c, I am getting the following crash: (gdb) bt 5 #0 op_bfd (this=3D0x7faf3910, archive=3D, fname=3D, symbol_filter=3D@0x1013ca9c, ok=3D@0x7faf388c) at op_bfd.cpp:138 #1 0x10048358 in populate_for_image (archive_path=3D@0x7faf3950, samples=3D@0x7faf39f0, ip=3D@0x10150d88, symbol_filter=3D@0x1013ca9c, has_debug_info=3D0x1013ca00) at populate.cpp:68 #2 0x100285ac in callgraph_container::populate (this=3D0x7faf3af0, archive_path=3D@0x1013bb98, iprofiles=3D@0x7faf3a74, extra=3D@0x1013cc54, debug_info=3Dfalse, threshold=3D0, merge_lib=3Dfalse, sym_filter=3D@0x1013ca9c) at callgraph_container.cpp:408 #3 0x1000782c in (anonymous namespace)::opreport (spec=3D) at opreport.cpp:562 #4 0x10018df4 in run_pp_tool (argc=3D, argv=3D0x7faf3e04, fct=3D0x10007440 <(anonymous namespace)::opreport(options::spec const&)>) at common_option.cpp:207 Please help me in solving this issue. -Ashok Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try= it now. -----------------------------------------------------------------------= -- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100.= Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/= javaone _______________________________________________ oprofile-list mailing list oprofile-list@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/oprofile-list = --1__=08BBFEA7DFFBF6138f9e8a93df938690918c08BBFEA7DFFBF613 Content-type: text/html; charset=US-ASCII Content-Disposition: inline Content-transfer-encoding: quoted-printable

Ashok:

I just checked the 2.6.11 kernel tree, arch/ppc/oprofile and do not see= any call graph kernel support. When you get to the 2.6.17 (not sure e= xactly when it got added) or so time frame, there is a file in the arch= /ppc/oprofile directory called backtrace.c which provides the call grap= h support. I only see oprofile support in the 2.6.11 kernel for the fr= eescale architecture. I assume that is what you are running on.

Note, the call graph support that is in the 2.6.17 and newer kernels ha= s an issue for leaf function. The leaf function is a function that doe= s not call another function. These leaf functions do not appear in the= call graph correctly. We are looking to see if we can fix this issue = in a reasonable way.

Carl Love

3D"InactiveAshok V <ashok_july14@yahoo.com&g= t;


=
          Ashok V <ashok_july14@yahoo.com>
          Sent by: oprofile-list-bounces@lists.sourceforge.net

          04/23/2008 07:56 AM

=
3D=
To
3D""
oprofile-list@lists.sourceforge.net
3D=
cc
3D""
3D=
Subject
3D""
Opreport Crash
3D=3D""

Hi,

Thanking you in advance. I am profiling kernel 2.6.11 in powerpc 32= bit architechture and I am using oprofile-o.9.3.tar.gz. I am able to r= un ./oprofile without any issues but while I am trying to generate call= graph using the cmd ./opreport -c, I am getting the
following crash:

(gdb) bt 5
#0 op_bfd (this=3D0x7faf3910, archive=3D<value optimized out>, fname=3D<value optimized out>, symbol_filter=3D@0x1013ca9c,
ok=3D@0x7faf388c) at op_bfd.cpp:138


#1 0x10048358 in populate_for_image
(archive_path=3D@0x7faf3950, samples=3D@0x7faf39f0,
ip=3D@0x10150d88, symbol_filter=3D@0x1013ca9c,
has_debug_info=3D0x1013ca00) at populate.cpp:68


#2 0x100285ac in callgraph_container::populate
(this=3D0x7faf3af0, archive_path=3D@0x1013bb98,
iprofiles=3D@0x7faf3a74, extra=3D@0x1013cc54, debug_info=3Dfalse,
threshold=3D0,
merge_lib=3Dfalse, sym_filter=3D@0x1013ca9c) at
callgraph_container.cpp:408


#3 0x1000782c in (anonymous namespace)::opreport
(spec=3D<value optimized out>) at opreport.cpp:562
#4 0x10018df4 in run_pp_tool (argc=3D<value optimized out>,
argv=3D0x7faf3e04, fct=3D0x10007440 <(anonymous
namespace)::opreport(options::spec const&)>) at
common_option.cpp:207

Please help me in solving this issue.

-Ashok


Be a better friend, newshound, and know-it-all with = Yahoo! Mobile. Try it now.--------------= -----------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.=
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13= 503038;p?http://java.sun.com/javaone______________________= _________________________
oprofile-list mailing list
oprofile-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/oprofile-list=

= --1__=08BBFEA7DFFBF6138f9e8a93df938690918c08BBFEA7DFFBF613--