This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Interesting observations wrt FDO and tramp3d-v4
- From: Steven Bosscher <stevenb at suse dot de>
- To: gcc at gcc dot gnu dot org
- Date: Thu, 8 Dec 2005 16:49:20 +0100
- Subject: Re: Interesting observations wrt FDO and tramp3d-v4
- References: <Pine.LNX.4.63.0512081302140.6048@t148.fhfr.qr>
On Thursday 08 December 2005 13:11, Richard Guenther wrote:
> look at those CSE numbers!
Why do they surprise you?
> The profile generating numbers suggest we're either doing something
> stupid, or that we want some heuristics applied to not instrument
> every edge, but only interesting ones.
We already only instrument the interesting edges ;-) We only add
instrumentation on edges that are not on a spanning tree of the CFG.
Selecting the interesting ones from that subset of all edges sounds
a bit strange, because the whole point of profiling is to _find_
those interesting edges :-P
Gr.
Steven