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 C8319384D163; Thu,  6 Oct 2022 06:00:40 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org C8319384D163
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1665036040;
	bh=I9EgJeqPMvV0pwapWVYqY6l78GjEV8tQ1pOBXkIQXQs=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=CSdR4wh4fYNlG6/lnvTE4T5kyp5drZ2dztSm8s/+nxbGif/OFAFZxaxcAfKHwTRvO
	 JDO6g/tfU3Vu0HusMugU5zR/VNIaF7m8i4V35ca98X0rOKDAq8/EEUVJhFsHLiwsn2
	 A3wncC86ewo4CvAH5A3kPDf10VOGeAAvAbJINhTg=
From: "unlvsur at live dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug rtl-optimization/107167] It looks like GCC wastes registers on
 trivial computations when result can be cached
Date: Thu, 06 Oct 2022 06:00:40 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: rtl-optimization
X-Bugzilla-Version: 13.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: enhancement
X-Bugzilla-Who: unlvsur at live dot com
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Resolution: 
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-107167-4-oiGctYq2MQ@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107167-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107167-4@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=3D107167

--- Comment #2 from cqwrteur <unlvsur at live dot com> ---
(In reply to Andrew Pinski from comment #1)
> This is a reassociation, scheduling issue and register allocation issue.
>=20
> Plus your example might be slower due to dependencies.
>=20
> Without a full example of where gcc ra goes wrong, gcc actually produces
> much better code for this example due to register renaming in hw.
> Note many x86_64 also does register renaming for the stack too

The problem I do things like sha512_round:
=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20
sha512_round(x[0]=3Dbig_endian(W[0]),a,b,d,e,f,g,h,bpc,0x428a2f98d728ae22);
sha512_round(x[1]=3Dbig_endian(W[1]),h,a,c,d,e,f,g,bpc,0x7137449123ef65cd);
sha512_round(x[2]=3Dbig_endian(W[2]),g,h,b,c,d,e,f,bpc,0xb5c0fbcfec4d3b2f);
sha512_round(x[3]=3Dbig_endian(W[3]),f,g,a,b,c,d,e,bpc,0xe9b5dba58189dbbc);
sha512_round(x[4]=3Dbig_endian(W[4]),e,f,h,a,b,c,d,bpc,0x3956c25bf348b538);

They use tons of registers. If GCC wastes registers, tons of time would was=
te
on stack push/load.

My implementation by GCC on x86_64 is slower than openssl's asm version
particularly due to this reason. GCC just pushes/stores too many values on =
the
stack.

https://github.com/openssl/openssl/blob/master/crypto/sha/asm/sha512-x86_64=
.pl#L192

OpenSSL does exactly what I do here.=