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 A06993852C79; Fri, 25 Nov 2022 05:16:26 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org A06993852C79
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1669353386;
	bh=+t3SCZknKKEJblrSP5uR890ixLL6z2EuiLAjWuh6k0M=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=Pn0oqYbRJniJYdQkF9aJkv03yVoolhOz44v6Bt8SNPXH0UxgWJDJPxmuGoP2oLwBR
	 a8Zfy5qxBcdcSuprLAMgk6Fyf1Zume7aop7FKxer8lyiXVho1Q/6mDXfkiYoBI7khw
	 lkRyc2ztTkiXWu2rIOXD54qIdUukLPN8vRiUiOyY=
From: "crazylht at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/107863] [10/11/12/13 Regression] ICE with unrecognizable
 insn when using -funsigned-char with some SSE/AVX builtins
Date: Fri, 25 Nov 2022 05:16:25 +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: 12.2.1
X-Bugzilla-Keywords: FIXME, ice-on-valid-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: crazylht 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: 10.5
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-107863-4-qVy9WNQktz@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107863-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107863-4@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=3D107863

--- Comment #6 from Hongtao.liu <crazylht at gmail dot com> ---
For pattern
(set (reg:QI 607)
    (const_int 255 [0xff]))

general_operand return false for op const_int 255 QImode since
trunc_int_for_mode (INTVAL (op), mode) return -1, INVAL (op) is 255.

---cut from general_operand (rtx, machine_mode)----------
  if (CONST_INT_P (op)
      && mode !=3D VOIDmode
      && trunc_int_for_mode (INTVAL (op), mode) !=3D INTVAL (op))
    return false;
---cut end-----------------


and in trunc_int_for_mode, it does signed extend, not unsigned_extend for
!flag_signed_char.

----cut from trunc_int_for_mode------------
  /* Sign-extend for the requested mode.  */

  if (width < HOST_BITS_PER_WIDE_INT)
    {
      HOST_WIDE_INT sign =3D 1;
      sign <<=3D width - 1;
      c &=3D (sign << 1) - 1;
      c ^=3D sign;
      c -=3D sign;
    }

  return c;
--------------cut end--------------


Should we do something like=20


modified   gcc/explow.cc
@@ -64,7 +64,8 @@ trunc_int_for_mode (HOST_WIDE_INT c, machine_mode mode)

   /* Sign-extend for the requested mode.  */

-  if (width < HOST_BITS_PER_WIDE_INT)
+  if (width < HOST_BITS_PER_WIDE_INT
+      && (mode !=3D QImode || !flag_signed_char))
     {
       HOST_WIDE_INT sign =3D 1;
       sign <<=3D width - 1;=