From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jeffrey A Law <law@cygnus.com>
To: Joern Rennecke <amylaar@cygnus.co.uk>
Cc: m.hayes@elec.canterbury.ac.nz, gcc@gcc.gnu.org
Subject: Re: Loop patch update (Was: Re: Autoincrement examples) 
Date: Fri, 31 Dec 1999 23:54:00 -0000
Message-ID: <27712.945144991@upchuck>
References: <199912132235.WAA28605@phal.cygnus.co.uk>
X-SW-Source: 1999-12n/msg00223.html
Message-ID: <19991231235400.Xepwts1N_Z_I19ZRMMccQB4dS4RpZsC3HJxw2aC5BOc@z>

  In message < 199912132235.WAA28605@phal.cygnus.co.uk >you write:
[ ... ]

  > The first problem - fixed by the first hunk - was an outright bug: the
  > sign of the calculated lifetime was wrong.
  > 
  > The second one is rather a missing feature: we didn't take into account tha
  > t
  > reducing a DEST_REG giv that can be combined with a giv is really pretty
  > cheap.
  > 
  > Mon Dec 13 22:30:36 1999  J"orn Rennecke <amylaar@cygnus.co.uk>
  > 
  > 	* loop.c (strength_reduce): Fix sign of giv lifetime calculation
  > 	for givs made from biv increments.
  > 	Don't reject giv for its cost if can likely be combined with the biv.
Assuming this doesn't depend on the other loop changes in the queue, you can
install it.

Thanks,
jeff