From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:41783)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <fred.konrad@greensocs.com>) id 1YKlAZ-0006c3-V2
	for qemu-devel@nongnu.org; Mon, 09 Feb 2015 05:02:23 -0500
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <fred.konrad@greensocs.com>) id 1YKlAW-0008OS-O1
	for qemu-devel@nongnu.org; Mon, 09 Feb 2015 05:02:19 -0500
Received: from greensocs.com ([193.104.36.180]:33202)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <fred.konrad@greensocs.com>) id 1YKlAW-0008OG-Ep
	for qemu-devel@nongnu.org; Mon, 09 Feb 2015 05:02:16 -0500
Message-ID: <54D885A0.3080609@greensocs.com>
Date: Mon, 09 Feb 2015 11:02:08 +0100
From: Frederic Konrad <fred.konrad@greensocs.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Subject: [Qemu-devel] CPU TLB flush with multithread TCG.
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: qemu-devel <qemu-devel@nongnu.org>, mttcg@listserver.greensocs.com
Cc: Peter Maydell <peter.maydell@linaro.org>, Jan Kiszka <jan.kiszka@siemens.com>, Alexander Graf <agraf@suse.de>, Paolo Bonzini <pbonzini@redhat.com>

Hi everybody,

In multithread tlb_flush is broken as CPUA can flush an other CPUB and 
CPUB can be
executing code, and fixing this can be quite hard:
   * We need to exit the CPU which is flushed.
   * Makes sure the CPU is stopped.
   * Then we can flush tlb.
The big issues are:
   * Two threads can be doing a flush at the same time.
   * Something can restart the CPU during the flush.

A better idea I think is that instead of flushing tlb we can put a flag 
in CPUState such
as flush_request and ask the cpu to exit.
Then later once the CPU is exited we can flush tlbs if flush_request is set.
It will ensure that the CPU won't execute code as it's associated thread 
will be
flushing.

Can this work?

Thanks,
Fred