Re: Testers needed for LTTng 0.6.0pre8

At 2006/9/28 14:46-0400  Mathieu Desnoyers writes:
> Hi,
> 
> 0.6.0pre9 fixes the powerpc optimization. There is a weird difference between
> x86 and powerpc for the asm operands.
> 
> x86 is weird, not powerpc. :)

:-)

Thanks! I can now build and boot it on a power 5 machine. For
anyone else interested you do need to apply a small patch:

--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -922,8 +922,6 @@ #endif
        DBG(" <- early_init_devtree()\n");
 }

-#undef printk
-
 int
 prom_n_addr_cells(struct device_node* np)
 {

but this is a mainline bug.

I did do quick test run. Maybe I'm driving it wrong, but I'm not
getting much output in /tmp/trace

cyeoh@Diego2:~$ ls -la /tmp/trace/
total 12
drwxr-xr-x 3 root root 4096 Sep 29 14:50 .
drwxrwxrwt 3 root root 4096 Sep 29 15:05 ..
drwxr-xr-x 2 root root 4096 Sep 29 14:50 eventdefs

I do get dmesg output indicating it is doing something:

ltt-control ltt_control_input
ltt_control : trace trace
Creating trace trace
ltt-control ltt_control_input
ltt_control : trace trace
Start tracing trace
Dumping facility core
Dumping facility socket
Dumping facility kernel
Dumping facility fs_data
Dumping facility kernel_arch
Dumping facility network
Dumping facility ipc
Dumping facility statedump
Dumping facility timer
Dumping facility memory
Dumping facility process
Dumping facility fs
Dumping facility network_ip_interface
ltt_statedump_start
do_ltt_statedump
trace_statedump_statedump_end
ltt-control ltt_control_input
ltt_control : trace trace
Stop tracing trace
ltt-control ltt_control_input
ltt_control : trace trace
Destroying trace trace

but nothing else appears in /tmp/trace

Chris
-- 
[email protected]
IBM OzLabs Linux Development Group, ADL
Canberra, Australia