Am Mittwoch, 30 M=E4rz 2016, 20:09:36 schrieb Michael Ellerman:

On Wed, 2016-03-30 at 13:34 +0530, Naveen N. Rao wrote:

quoted

On 2016/03/29 08:35PM, Thiago Jung Bauermann wrote:

quoted

Am Dienstag, 29 M=E4rz 2016, 14:31:34 schrieb Michael Ellerman:

quoted

On Mon, 2016-03-28 at 17:06 -0300, Thiago Jung Bauermann wrote:=

quoted

quoted

quoted

quoted

With this patch, all vmlinux symbols match /proc/kallsyms and=

 the

quoted

quoted

quoted

quoted

testcase passes.

=20
Have you tested this on an LE system?

=20
No, I was focusing on ppc64 BE.

=20
Which kernel did you use? I don't see this with the latest 4.6.0-rc=

1.

=20
Yeah I can't reproduce the perf test failure either.

I can still reproduce on 4.6.0-rc1.

Not sure what's going on?

Try disabling CONFIG_FUNCTION_TRACER. The test failure happens when .kr=
etprobe_trampoline_holder and
kretprobe_trampoline have the same address. If that option is enabled, =
the symbols have different addresses:

$ eu-readelf -s /boot/vmlinux-4.6.0-rc1-dirty | grep kretprobe_tramp   =
                                                =20
 2479: c000000000ef1d68     24 NOTYPE  LOCAL  DEFAULT       37 kretprob=
e_trampoline_holder
 2480: c000000000043ab0     40 FUNC    LOCAL  DEFAULT        1 .kretpro=
be_trampoline_holder
86903: c000000000043ac4      0 NOTYPE  GLOBAL DEFAULT        1 kretprob=
e_trampoline
$ grep kretprobe_tramp /proc/kallsyms
c000000000043ab0 t .kretprobe_trampoline_holder
c000000000043ac4 T kretprobe_trampoline
c000000000ef1d68 d kretprobe_trampoline_holder

--=20
[]'s
Thiago Jung Bauermann
IBM Linux Technology Center