From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([140.186.70.92]:42695)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <jcmvbkbc@gmail.com>) id 1QJWI8-00072Z-Tt
	for qemu-devel@nongnu.org; Mon, 09 May 2011 15:38:53 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <jcmvbkbc@gmail.com>) id 1QJWI8-0004dE-2N
	for qemu-devel@nongnu.org; Mon, 09 May 2011 15:38:52 -0400
Received: from mail-ey0-f173.google.com ([209.85.215.173]:54205)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <jcmvbkbc@gmail.com>) id 1QJWI7-0004d2-Te
	for qemu-devel@nongnu.org; Mon, 09 May 2011 15:38:52 -0400
Received: by eyb6 with SMTP id 6so1988718eyb.4
	for <qemu-devel@nongnu.org>; Mon, 09 May 2011 12:38:50 -0700 (PDT)
From: Max Filippov <jcmvbkbc@gmail.com>
Date: Mon, 9 May 2011 23:38:45 +0400
References: <1304470768-16924-1-git-send-email-jcmvbkbc@gmail.com>
	<4DC17FCF.2090706@twiddle.net> <4DC18641.2080904@twiddle.net>
In-Reply-To: <4DC18641.2080904@twiddle.net>
MIME-Version: 1.0
Content-Type: Text/Plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Message-Id: <201105092338.46339.jcmvbkbc@gmail.com>
Subject: Re: [Qemu-devel] [RFC 18/28] target-xtensa: implement exceptions
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: Richard Henderson <rth@twiddle.net>
Cc: qemu-devel@nongnu.org

> > This is a case where you almost certainly want to check this 
> > condition inside QEMU and translate the opcode differently.
> > 
> > See cpu_get_tb_cpu_state, which sets bits in *flags.  These
> > flags can then be checked in tb->flags while translating.
> > At which point you'd avoid all the conditionalization on
> > the value in PS here in check_privilege and merely issue
> > the exception_cause.
> > 
> > The ARM port is a good example for testing these sorts of bits.
> 
> Actually, while the tb flags are useful, privileged instructions
> are usually checked by testing mmu_index, since we already have
> to generate different code for the TB based on which TLB entry we
> need to access.

Richard,
do I understand it right that if I use tb->flags or mem_index to make decisions during translation, then I must issue exit_tb on instructions that can change state they reflect?

Thanks.
-- Max