Re: [PATCH 3/3] perfcounter: Align ftrace events raw samples to 8 bytes

[Frederic Weisbecker <fweisbec@gmail.com>]

On Mon, Aug 10, 2009 at 09:13:57AM +0200, Peter Zijlstra wrote:
> On Sat, 2009-08-08 at 04:26 +0200, Frederic Weisbecker wrote:
> > Align the ftrace events raw samples to 8 bytes so that they meet the
> > perf output event alignements requirements.
> > 
> > Reported-by: Peter Zijlstra <peterz@infradead.org>
> > Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
> > Cc: Peter Zijlstra <peterz@infradead.org>
> > Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
> > Cc: Mike Galbraith <efault@gmx.de>
> > Cc: Paul Mackerras <paulus@samba.org>
> > ---
> >  include/trace/ftrace.h |    2 +-
> >  1 files changed, 1 insertions(+), 1 deletions(-)
> > 
> > diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h
> > index 80e5f6c..7b0834c 100644
> > --- a/include/trace/ftrace.h
> > +++ b/include/trace/ftrace.h
> > @@ -687,7 +687,7 @@ static void ftrace_profile_##call(proto)				\
> >  	pc = preempt_count();						\
> >  									\
> >  	__data_size = ftrace_get_offsets_##call(&__data_offsets, args); \
> > -	__entry_size = __data_size + sizeof(*entry);			\
> > +	__entry_size = ALIGN(__data_size + sizeof(*entry), sizeof(u64));\
> >  									\
> >  	do {								\
> >  		char raw_data[__entry_size];				\
> 
> Make sure to clear the trailing few bytes, otherwise we've got a stack
> data leak here.


Oh, right.


 
> Also, tracepoints using __string() could still easily overflow the stack
> here and cause great havoc, esp since the tracepoint could already be
> deep into a call-chain.


I know yeah, I'll think about a way to store the string parts as pointers
and copy their content on perf_output only.