c/7344: performance regression on huge case statements

[Jan Hubicka]

The following reply was made to PR c/7344; it has been noted by GNATS.

From: Jan Hubicka <jh@suse.cz>
To: Nathanael Nerode <neroden@twcny.rr.com>
Cc: gcc-gnats@gcc.gnu.org, jh@suse.cz, gcc-bugs@gcc.gnu.org
Subject: Re: c/7344: performance regression on huge case statements
Date: Thu, 17 Oct 2002 15:32:09 +0200

 > Jan, your patch of October 14 had the ChangeLog entries for predict.c 
 > (can_predict_insn_p etc) and toplev.c (rest_of_compilation), but somehow 
 > the actual files failed to get committed.  Either commit the files 
 > (preferred) or remove the ChangeLog entry, please. :-)
 
 Oops, I will commit that one.
 
 > 
 > http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=7344
 > 
 > Incidentally, with only the patch of October 11th (edge cache) the test 
 > case still seems to take forever.  We'll see what happens once the patch 
 > of October 14th (eliminating unnecessary queries) is in place; I expect 
 > that to be more important based on your earlier numbers.
 
 There is still bit too much time spent in crossjumping and branch
 prediction.  I didn't have time to take a look at that yet, I will try
 to find some later.
 
 Honza
 > 
 > --Nathanael