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 E48DE3858C60; Mon, 23 Oct 2023 13:54:49 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org E48DE3858C60
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1698069289;
	bh=vfSLg6w2c1ZWMsGW0ElmW7Y5vUwRGCueJmq2Q3ZlkAY=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=gstCITap9EUVno6PbvtbBkRqIlft/mZIDSMy3ATmY+4BX7YY6fQjdmxQkQn0ginnU
	 mvqJmkKMNf5C0Ex+XrJG3oPdvRCEz4V9REcsCGoXbqVziQL1K2jYuwmYGVU52GW9rx
	 KYVgDnr3WnBzFESF/S9ZlRixhc3k9/VXXUhJQhnQ=
From: "bettio.davide at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug middle-end/111933] memcpy on Xtensa not optimized when n ==
 sizeof(uint32_t) or sizeof(uint64_t)
Date: Mon, 23 Oct 2023 13:54:49 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: middle-end
X-Bugzilla-Version: 11.2.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: bettio.davide at gmail 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-111933-4-19eePuVWIC@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-111933-4@http.gcc.gnu.org/bugzilla/>
References: <bug-111933-4@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=3D111933

--- Comment #2 from Davide Bettio <bettio.davide at gmail dot com> ---
(In reply to Richard Biener from comment #1)
> uint32_t from_unaligned_u32(uint32_t *unaligned)
> {
>     uint32_t tmp;
>     tmp =3D *unaligned;
>=20
> note this isn't an unaligned access since you dereference a uint32_t poin=
ter

Let me explain better my purpose: I was writing a function for reading uint=
32_t
and uint64_t values at any address, including odd ones such as 0x00002345.

So as far as I know the only generic way to do this is either casting them =
to
uint8_t and reading them as 4 or 8 bytes, or using memcpy.
My expectation is that memcpy is always replaced with a load instruction wh=
en
fast unaligned memory access is available.

Let me know if I am making any wrong assumptiom.=