Re: [ANNOUNCE] xfs; for-next branch updated to 31bb236

[Dave Chinner <david@xxxxxxxxxxxxx>]

On Wed, Mar 02, 2016 at 01:45:59AM -0800, Christoph Hellwig wrote:
> > Mateusz Guzik (1):
> >       [2e83b79] xfs: fix two memory leaks in xfs_attr_list.c error paths
> 
> This didn't seem to address the comment about using a goto for
> error handling?

I looked at it, and there isn't a case for a goto. The allocated
memory is in a variable local to the if branch where the leak was
found. Using a jump-based error handling stack means making
all those variables global, and then having to make the freeing
of it conditional. Simply didn't seem worth it to me.

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx