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 0D0D33846403; Tue,  9 Apr 2024 08:05:10 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 0D0D33846403
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1712649911;
	bh=aF5eRGUoURgPYbF1us2eSgAuUrcKAHoDgf/FLQBsZCs=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=gjbQVnvaTLTmjT/I33IxGJHjLGSznp83wipr7qKAnmSPZB7pWzPFePm05dcsyb6+k
	 4C1T2/3Fbr/6dhTY5yFoUTIDCOmzuaXDP2Cuv2CnqJrTH8kUT+RH2ArWc6piO2bs/t
	 HRkLTeaxd+W1uwLq202H3wmDXlomR2maNGXGez4g=
From: "ubizjak at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/114639] [riscv] ICE in create_pre_exit, at
 mode-switching.cc:451
Date: Tue, 09 Apr 2024 08:05:10 +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: ice-on-valid-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: ubizjak 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-114639-4-vZxv6PyqAd@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-114639-4@http.gcc.gnu.org/bugzilla/>
References: <bug-114639-4@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=3D114639

--- Comment #9 from Uro=C5=A1 Bizjak <ubizjak at gmail dot com> ---
(In reply to Andrew Pinski from comment #2)
>             /* If we didn't see a full return value copy, verify that the=
re
>                is a plausible reason for this.  If some, but not all of t=
he
>                return register is likely spilled, we can expect that there
>                is a copy for the likely spilled part.  */

This part of the mode-switching pass is a real PITA. The trick here is with=
 the
calculation of forced_late_switch (but please see N.b. comment at the begin=
ning
of the function where some failed assumptions are described).=