From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugs-return-469089-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 32644 invoked by alias); 1 Dec 2014 14:15:24 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Received: (qmail 32586 invoked by uid 48); 1 Dec 2014 14:15:19 -0000
From: "rguenth at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/14541] [tree-ssa] built-in math functions are not fully optimized at tree level
Date: Mon, 01 Dec 2014 14:15:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: tree-optimization
X-Bugzilla-Version: tree-ssa
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: enhancement
X-Bugzilla-Who: rguenth at gcc dot gnu.org
X-Bugzilla-Status: ASSIGNED
X-Bugzilla-Priority: P2
X-Bugzilla-Assigned-To: pinskia at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-14541-4-x92LuPfnlC@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-14541-4@http.gcc.gnu.org/bugzilla/>
References: <bug-14541-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2014-12/txt/msg00096.txt.bz2

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=14541

--- Comment #16 from Richard Biener <rguenth at gcc dot gnu.org> ---
(for logfn (BUILT_IN_LOGF BUILT_IN_LOG BUILT_IN_LOGL)
     expfn (BUILT_IN_EXPF BUILT_IN_EXP BUILT_IN_EXPL)
 (simplify
  (logfn (expfn @0))
  (if (flag_unsafe_math_optimizations)
   @0)))

with the still somewhat awkward syntax, patch for proposed better one
is not yet ready but would look like

(define_operator_list log (BUILT_IN_LOGF BUILT_IN_LOG BUILT_IN_LOGL))
(define_operator_list exp (BUILT_IN_EXPF BUILT_IN_EXP BUILT_IN_EXPL))

(simplify
 (log (exp @0))
 (if (flag_unsafe_math_optimizations)
  @0)))

thus the 'for' would be implicit.  Note that fold_builtin_logarithm does handle
some more cases.  The match-and-simplify branch already covers quite some
patterns in match-builtins.def (but not the above).

I'll see if I can take the syntax patch (I was waiting for that so not to
merge the "ugly" syntax).