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 21B153858C2C; Sat, 26 Aug 2023 16:43:26 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 21B153858C2C
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1693068206;
	bh=sjY4vOZBnX9VA9mKZv2AtqZhrxzRKsuQzZJSL1urGHc=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=edc8rddSXy4rFQ6ARtNt3ZjdYoiM7x6aiVBuz1lOAJ1qWdYG9s0gDloWvy7OYSOz7
	 DgrRmy9dxjdjO9BPK3y17YEBxdZF3LlBoLZ5BgPuecuWyYUlKs9CYYf1C9FRwW0OGh
	 wOmVdACh9xOKJVtKv3VEDLzhbQ/yh38lGOAXSkVg=
From: "eggert at cs dot ucla.edu" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug rtl-optimization/111143] [missed optimization] unlikely code
 slows down diffutils x86-64 ASCII processing
Date: Sat, 26 Aug 2023 16:43:24 +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: 13.1.1
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: eggert at cs dot ucla.edu
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-111143-4-Cu5HTyFMyQ@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-111143-4@http.gcc.gnu.org/bugzilla/>
References: <bug-111143-4@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=3D111143

--- Comment #7 from Paul Eggert <eggert at cs dot ucla.edu> ---
(In reply to Alexander Monakov from comment #6)

> Are you binding the benchmark to some core in particular?

I did the benchmark on performance cores, which was my original use case. On
efficiency cores, adding the (unnecessary) 'mov eax, 1' doesn't change timi=
ng
much (0.9% speedup on one test).

> it is better to have 'add rbx, 1' instead of 'add rbx, rax' in this loop =
on any CPU

Somewhat counterintuitively, that doesn't seem to be the case for the
efficiency cores on this platform, as the "38% faster" code is 7% slower on
E-cores. However, the use cases I'm concerned about are typically run on
performance cores.=