That looks good too me.
I'm certain that'll compile on ppc.
OK, great. The change affects the audio pre-buffer (ringbuffer) indexing, so with this change on ppc the audio should _still_ always be in sync.
Now that I think of this, there is another type of ringbuffer that doesn't use atomic setting/getting of the indexes.... Might change it to use the introduced functions as well to be very sure it works reliable on ppc as well.... (Used for memory/cpu load detection, not that important..)
ui/ExternalProcessing.ui is missing in cvs right now, though ;) so
I'll have to wait and test later.
Strange, I remember seen the commit log in my mailbox, which means the file was committed. It's in a new directory, you might have to do a cvs -d update ?
(iirc, the -d checks for new dirs, and downloads the new content if there is any.