Re: what will get backported?

[Janne Blomqvist <jblomqvi at cc dot hut dot fi>]

On Tue, Mar 14, 2006 at 12:47:19AM -0800, Brooks Moses wrote:
> Janne Blomqvist wrote:
> >One thing that was deemed significant enough to backport to 4.1 was
> >Paul Thomas patch inlining dot_product. It's the one responsible for
> >the huge improvement in induct which you can see at
> >http://www.suse.de/~rguenther/polyhedron/ .
> 
> Speaking of which, what happened to channel on March 5th?  That was a 
> pretty sizable jump, and I don't think I've seen any mention of it here....

See the thread starting at 

http://gcc.gnu.org/ml/fortran/2006-03/msg00039.html

Incidentally, you posted a couple of messages to the thread above
yourself. :)

Roger committed two patches, but the second one was reverted since it
caused a regression. Apparently the patches were supposed to work
together, since removing the second patch caused the channel
regression. Paul Thomas suggested to allow Roger some time to fix the
second patch, and I agree with that since the approach described in
the paper referenced by Roger seems quite clever.


-- 
Janne Blomqvist

Attachment: pgp00000.pgp
Description: PGP signature