Re: [PATCH 1/2] mmiotrace: fix buffer overrun detection

[Pekka Paalanen <pq@iki.fi>]

On Tue, 25 Nov 2008 12:59:22 +0100
"Frédéric Weisbecker" <fweisbec@gmail.com> wrote:

> 2008/11/25 Frédéric Weisbecker <fweisbec@gmail.com>:
> > Perhaps that:
> >
> > int trace_vprintk(unsigned long ip, const char *fmt, va_list args)
> > {
> >     ......
> >     if (tracing_disabled)
> >                return 0;
> 
> Since this is the return value for tracing_mark_write(), it may better
> return an error to avoid a spin when tracing
> is disabled. -ENODEV I guess, or -EBUSY
> 
> The strange thing is that tracing_open_generic() shoud have return
> -ENODEV in this case during the opening of trace_mark.
> 

This sounds worth investigating, thanks for the tip.
Hmm, maybe if I changed the current tracer while trace_enabled was
zero...

Anyway, this isn't too high on my priority list.

-- 
Pekka Paalanen
http://www.iki.fi/pq/