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 2F777386EC6E; Thu, 18 Jan 2024 13:01:36 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 2F777386EC6E
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1705582896;
	bh=4vEQlLKmaWkKezhOHcz2TpmY7j7TzfQQ9X/DAe1/d6c=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=hij1wSLPUVRZSppLkTEl4u7KMpX6hYWILv+85P1Bk42/Ez6ly7zXlY2JvASrtAbtP
	 az1OCFrZcfULXvMPUec17t6ECJbOlf/NA31kJC0PzoMUKH+UbXfnZzve86oNOQa14W
	 eDgNdkE72jPpS2J1ERCkOPJmFhnvYSgnEUXWNjHs=
From: "hjl.tools at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/113312] Add __attribute__((no_callee_saved_registers))
 for Intel FRED
Date: Thu, 18 Jan 2024 13:01:34 +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: 14.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: hjl.tools at gmail dot com
X-Bugzilla-Status: NEW
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-113312-4-aLTtxkklqD@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-113312-4@http.gcc.gnu.org/bugzilla/>
References: <bug-113312-4@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=3D113312

--- Comment #23 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Xin Li from comment #22)
> Per Peter's suggestion, I added __attribute__((no_callee_saved_registers))
> to a linux source tree containing FRED patches:
> https://github.com/xinli-intel/linux-fred-public/commit/
> 12c38143a5c33e89f2b3d8906629dd4f23f8d79c.  And I compiled the linux code
> with a gcc built from
> https://gitlab.com/x86-gcc/gcc/-/tree/users/hjl/pr113312/gcc-13.
>=20
> Following are my observations:
> 1) the generated kernel boots fine on both FRED Simics model and bare met=
al.
> 2) the asm code generated for fred_entry_from_{user,kernel}() are the sam=
e,
> i.e., __attribute__((no_callee_saved_registers)) makes no difference (Pet=
er
> said the FRED dispatch points simply do not have significant register
> pressure =E2=80=93 intentionally).
> 3) other functions with __attribute__((no_callee_saved_registers)) do get
> rid of pushing/popping clobbered registers, and cause no issues because t=
hey
> are top-level functions, only invoked by tail call, meaning the following
> case won't happen:
>=20
> <func>:
> ...
> mov (%rbx), %r13
> call foo
> mov %rax, (%r13)
> ...
>=20
> otherwise foo() is NOT a top-level function.

Do these mean that __attribute__((no_callee_saved_registers)) works for you?
BTW, my patches also avoid saving callee-saved registers in
__attribute__((noreturn))
functions.=