From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 6267 invoked by alias); 21 Nov 2001 21:46:23 -0000 Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org Received: (qmail 6172 invoked from network); 21 Nov 2001 21:46:12 -0000 Received: from unknown (HELO gash2.peakpeak.com) (207.174.178.17) by sourceware.cygnus.com with SMTP; 21 Nov 2001 21:46:12 -0000 Received: from creche.cygnus.com (ta0200.peakpeak.com [204.144.244.200]) by gash2.peakpeak.com (8.9.3/8.9.3) with ESMTP id KAA24017; Wed, 21 Nov 2001 10:45:41 -0700 Received: (from tromey@localhost) by creche.cygnus.com (8.9.3/8.9.3) id KAA17479; Wed, 21 Nov 2001 10:45:52 -0700 To: Per Bothner Cc: Gcc Patch List , Java Patch List , Alexandre Petit-Bianco Subject: Re: Patch: don't generate dead bytecode References: <87elocq19k.fsf@creche.redhat.com> <3BC39195.6040803@bothner.com> <3BC3990B.3000000@bothner.com> <3BC62CED.1080903@bothner.com> <87u1x5lejy.fsf@creche.redhat.com> <3BC78628.7080809@bothner.com> <87bshz2pg0.fsf@creche.redhat.com> <3BF8B29D.5040907@bothner.com> X-Zippy: You mean now I can SHOOT YOU in the back and further BLUR th' distinction between FANTASY and REALITY? X-Attribution: Tom Reply-To: tromey@redhat.com From: Tom Tromey Date: Tue, 13 Nov 2001 15:03:00 -0000 In-Reply-To: Per Bothner's message of "Sun, 18 Nov 2001 23:19:57 -0800" Message-ID: <87n11gj9v3.fsf@creche.redhat.com> X-Mailer: Gnus v5.7/Emacs 20.5 X-SW-Source: 2001-11/txt/msg00751.txt.bz2 >>>>> "Per" == Per Bothner writes: Per> Perhaps rather than writing our own fold it would be better to Per> add some flag to force fold to respect evaluation order and Per> otherwise not do anything that might break Java semantics. The other problem we face is that a Java fold() must handle string operations as well. Whether we should write our own or modify the existing one, I don't know. fold() certainly seems complicated, so rewriting it might be a mistake. Anyway, last time I looked this area was one of our leading causes of failure for Jacks tests, so on a correctness metric it is a big win to fix. Tom