[ltt-dev] A problem in net instrumentation code of lttng-0.64

[Neil Horman]

On Tue, Dec 09, 2008 at 11:07:30AM -0500, Mathieu Desnoyers wrote:
> * Sunzen Wang (sunzen at gmail.com) wrote:
> > Hi Mathieu,
> > 
> > I found a strange problem in net instrumentation code while scanning
> > lttng instrumentation code via gitweb, and would like to inform you
> > with the attached patch.
> > 
> > Is it ok?  If you have noticed it, just ignore it. If it is a not
> > problem, please teach me.
> > 
> > Thanks,
> 
> Hi Suzen,
> 
> Yes, the problem you identified must be fixed. I'll apply your patch
> into the next LTTng release.
> 
> Mathieu
> 
> (your patch pasted here)
> 
Apologies.  That was a finger check on my part.  I must have posted an unfixed
version of my patch rather than extracting from my git tree right before
posting.  I'll double check that in the future.
Neil

> 
> diff -Naru linux-2.6.27.8-lttng-0.64/net/core/dev.c linux-2.6.27.8-lttng-new/net/core/dev.c
> --- linux-2.6.27.8-lttng-0.64/net/core/dev.c	2008-12-09 14:51:16.972757754 +0800
> +++ linux-2.6.27.8-lttng-new/net/core/dev.c	2008-12-09 14:54:00.636758086 +0800
> @@ -2336,7 +2336,7 @@
>  {
>  	unsigned long flags;
>  
> -	trace_net_napi_poll(n);
> +	trace_net_napi_schedule(n);
>  
>  	local_irq_save(flags);
>  	list_add_tail(&n->poll_list, &__get_cpu_var(softnet_data).poll_list);
> -- 
> Mathieu Desnoyers
> OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
> 

-- 
/****************************************************
 * Neil Horman <nhorman at tuxdriver.com>
 * Software Engineer, Red Hat
 ****************************************************/