Vince,
On Mon, Oct 30, 2006 at 01:06:25PM -0500, Vince Weaver wrote:
> On Mon, 30 Oct 2006, Stephane Eranian wrote:
>
> > Did that problem affect arch/i386/kernel/process.c?
> >
> > Please verify the following:
>
> I checked, and the patched version of the code looks exactly as you say it
> should.
>
Ok, now I remember, there was a bogus assertion in __pfm_ctxswin_thread().
I thought it would be a illegal case, but it is not. So if line 99 in
your perfmon_ctxsw.c, has the BUG_ON() just like below, remove it and
this should fix your problem.
static void __pfm_ctxswin_thread(struct task_struct *task,
struct pfm_context *ctx, u64 now)
{
u64 cur_act;
struct pfm_event_set *set;
int reload_pmcs, reload_pmds;
BUG_ON(!task->mm);
> There was a message during patching:
>
> patching file arch/i386/kernel/process.c
> Hunk #2 succeeded at 380 (offset -32 lines).
> Hunk #3 succeeded at 433 (offset -32 lines).
> Hunk #4 succeeded at 539 (offset -32 lines).
> Hunk #5 succeeded at 557 (offset -32 lines).
> Hunk #6 succeeded at 700 (offset -32 lines).
>
That is fine.
--
-Stephane
_______________________________________________
perfmon mailing list
[email protected]
http://www.hpl.hp.com/hosted/linux/mail-archives/perfmon/
