On Thu, Feb 20, 2014 at 04:48:30PM -0500, Will Woods wrote:
On Tue, 2014-02-18 at 10:20 -0800, Brian C. Lane wrote:
> On Tue, Feb 18, 2014 at 11:33:45AM -0500, Chris Lumens wrote:
> > > > I think this, the coverage patch, and the existing instperf code need
to
> > > > get together and decide on a common command line argument and
location
> > > > to dump their data files.
> > >
> > > Good idea. How about /tmp/debug/ ?
> >
> > That'd work for me. Should we also consider moving the files over to
> > /root/wherever post-installation?
>
> I looked at doing that, but couldn't find a clean way to get a reference
> to the anaconda object from inside doConfiguration so that I could tell
> it to dump the main thread's stats. It would be nice, but I don't think
> it really needs to.
Could do that pretty easily in data/post-scripts/99-copy-logs.ks, right?
Right, if you can get it to write the file first. That would have to be
done in doConfiguration right before we run the post scripts, but since
I put the profiler into the anaconda object you can't get there from
here.
I don't really want to make it global, so suggestions are welcome :)
--
Brian C. Lane | Anaconda Team | IRC: bcl #anaconda | Port Orchard, WA (PST8PDT)