Re: [PATCHv3] tracing - avoid soft lockup in trace_pipe

[Steven Rostedt]

On Fri, 2011-03-25 at 12:05 +0100, Jiri Olsa wrote:

> right, I'm now able to trigger the issue when running setting the tracing
> as said in the patch and plus running this in few instacies:
> 
> (while [ 1 ]; do find / > /dev/null ; done) &

That didn't work for me either. But if I run hackbench 50 on the box,
then the trace does hang while it is running.

> 
> It's strange I did not need this last time.. not sure why :)
> 
> 
> anyway, the original reproducer I picked this issue from is ltp suite:
> http://sourceforge.net/projects/ltp/files/LTP%20Source/ltp-20100131/ltp-full-20100131.tgz/download
> 
> and running following test:
> ltp-full-20101031/testcases/kernel/tracing/ftrace_stress_test/ftrace_stress_test.sh 
> 
> reproduces the issue all the time for me, and the patch fixies it.
> Attaching patch rebased to the current master with changed comment.

Thanks, I'll queue it up for 2.6.40, as I do not think this is that
critical of a bug.

-- Steve


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/