From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:41058)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <rth7680@gmail.com>) id 1ZPP8b-0000TX-Qp
	for qemu-devel@nongnu.org; Wed, 12 Aug 2015 02:03:46 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <rth7680@gmail.com>) id 1ZPP8W-0005Rt-RY
	for qemu-devel@nongnu.org; Wed, 12 Aug 2015 02:03:45 -0400
Received: from mail-qg0-x231.google.com ([2607:f8b0:400d:c04::231]:35109)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <rth7680@gmail.com>) id 1ZPP8W-0005Rj-No
	for qemu-devel@nongnu.org; Wed, 12 Aug 2015 02:03:40 -0400
Received: by qgj62 with SMTP id 62so4658766qgj.2
	for <qemu-devel@nongnu.org>; Tue, 11 Aug 2015 23:03:40 -0700 (PDT)
Sender: Richard Henderson <rth7680@gmail.com>
References: <1439151229-27747-1-git-send-email-laurent@vivier.eu>
	<1439151229-27747-13-git-send-email-laurent@vivier.eu>
From: Richard Henderson <rth@twiddle.net>
Message-ID: <55CAE1B7.90104@twiddle.net>
Date: Tue, 11 Aug 2015 23:03:35 -0700
MIME-Version: 1.0
In-Reply-To: <1439151229-27747-13-git-send-email-laurent@vivier.eu>
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH for-2.5 12/30] m68k: Manage divw overflow
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Laurent Vivier <laurent@vivier.eu>, qemu-devel@nongnu.org
Cc: peter.maydell@linaro.org, peter.crosthwaite@xilinx.com, Andreas Schwab <schwab@linux-m68k.org>, gerg@uclinux.org

On 08/09/2015 01:13 PM, Laurent Vivier wrote:
> "Overflow may be detected and set before the instruction completes.
> If the instruction detects an overflow, it sets the overflow condition
> code, and the operands are unaffected."

May also implies may not.  I presume this is important for matching hardware? 
Is there some program you know of that depends on this?

> +    /* dest.l / src.w */
> +
> +    dest = DREG(insn, 9);
> +    tcg_gen_mov_i32(QREG_DIV1, dest);
> +
>       SRC_EA(env, src, OS_WORD, sign, NULL);
>       tcg_gen_mov_i32(QREG_DIV2, src);
> +
> +    /* div1 / div2 */
> +
>       if (sign) {
>           gen_helper_divs(cpu_env, tcg_const_i32(1));
>       } else {
>           gen_helper_divu(cpu_env, tcg_const_i32(1));
>       }
>
> +    set_cc_op(s, CC_OP_FLAGS);
> +
> +    l1 = gen_new_label();
> +    gen_jmpcc(s, 9 /* V */, l1);
>       tmp = tcg_temp_new();
>       src = tcg_temp_new();
>       tcg_gen_ext16u_i32(tmp, QREG_DIV1);
>       tcg_gen_shli_i32(src, QREG_DIV2, 16);
> -    tcg_gen_or_i32(reg, tmp, src);
> -    set_cc_op(s, CC_OP_FLAGS);
> +    tcg_gen_or_i32(dest, tmp, src);
> +    gen_set_label(l1);

All that said, it's possible to implement this branch inside the helper via 
exception.  Or simply return the inputs to effect no change.


r~