From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
	id 5712B385841A; Tue, 26 Mar 2024 23:52:05 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 5712B385841A
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1711497125;
	bh=CeFnfS4oVkEWtV5PvNtkPGA8SxLnR3v1zarUB1M4bTI=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=awo9kpvd7qQct7flPBzI5SVFs6nYviE1JzSdVuL5KjkgSCA6qpnlZ0DnR9h7kExxf
	 GaIyDuJ4M8QvHCGwaPR+5hfHUrKw5PxagA85sQh8txK87ihGAMfLMnbXMAorqbDcJ+
	 kfe1SAba1YAzpM61bdhkn3YKWCgzSlLONyqDYsaU=
From: "Explorer09 at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/114490] Optimization: x86 "shl" condition codes never
 reused
Date: Tue, 26 Mar 2024 23:52:04 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 13.2.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: Explorer09 at gmail dot com
X-Bugzilla-Status: RESOLVED
X-Bugzilla-Resolution: INVALID
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-114490-4-vSCrzkdqLo@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-114490-4@http.gcc.gnu.org/bugzilla/>
References: <bug-114490-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
List-Id: <gcc-bugs.sourceware.org>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D114490

--- Comment #3 from Kang-Che Sung <Explorer09 at gmail dot com> ---
(In reply to Andrew Pinski from comment #2)
>=20
> Basically if you are shifting by 0, then the CF bit never changes from the
> previous and this can't be done.
>=20
> So the comment in the backend is correct and there is no way to remove the
> test (unless you know ch can't be 0).

What if I had the CF cleared to 0 already before the shift?

Even when the flags don't change when the value in the CL register is zero,=
 if
I had the flags in a known state, the optimization can still be done.

And my suggested 'my_isxdigit_2' code worked before I know the caveat of the
shift instructions... ("xorl  %eax, %eax" -> CF =3D SF =3D 0)=