From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:56654)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <alex.bennee@linaro.org>) id 1c1yJW-0005oi-S6
	for qemu-devel@nongnu.org; Wed, 02 Nov 2016 12:22:59 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <alex.bennee@linaro.org>) id 1c1yJT-00052f-NZ
	for qemu-devel@nongnu.org; Wed, 02 Nov 2016 12:22:58 -0400
Received: from mail-wm0-x230.google.com ([2a00:1450:400c:c09::230]:33011)
	by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16)
	(Exim 4.71) (envelope-from <alex.bennee@linaro.org>)
	id 1c1yJT-00052H-Gq
	for qemu-devel@nongnu.org; Wed, 02 Nov 2016 12:22:55 -0400
Received: by mail-wm0-x230.google.com with SMTP id 79so47817339wmy.0
	for <qemu-devel@nongnu.org>; Wed, 02 Nov 2016 09:22:55 -0700 (PDT)
References: <20161102142559.23395-1-bobby.prani@gmail.com>
From: Alex =?utf-8?Q?Benn=C3=A9e?= <alex.bennee@linaro.org>
In-reply-to: <20161102142559.23395-1-bobby.prani@gmail.com>
Date: Wed, 02 Nov 2016 16:22:52 +0000
Message-ID: <8737j9vokj.fsf@linaro.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
Subject: Re: [Qemu-devel] [PATCH] mttcg: Handle EXCP_ATOMIC exception
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: Pranith Kumar <bobby.prani@gmail.com>
Cc: qemu-devel@nongnu.org, pbonzini@redhat.com


Pranith Kumar <bobby.prani@gmail.com> writes:

> The patch enables handling atomic code in the guest. This should be
> preferably done in cpu_handle_exception(), but the current assumptions
> regarding when we can execute atomic sections cause a deadlock.
>
> Signed-off-by: Pranith Kumar <bobby.prani@gmail.com>
> ---
>  cpus.c | 7 +++++++
>  1 file changed, 7 insertions(+)
>
> diff --git a/cpus.c b/cpus.c
> index 8f98060..c4ba7d8 100644
> --- a/cpus.c
> +++ b/cpus.c
> @@ -1315,6 +1315,9 @@ static void *qemu_tcg_rr_cpu_thread_fn(void *arg)
>                  if (r == EXCP_DEBUG) {
>                      cpu_handle_guest_debug(cpu);
>                      break;
> +                } else if (r == EXCP_ATOMIC) {
> +                    cpu_exec_step_atomic(cpu);
> +                    break;

Hmm don't we need to unlock the iothread here as well? I suspect you
never see a deadlock because the rr thread can't by definition race with
itself but the locking practice should be the same for both cases.

>                  }
>              } else if (cpu->stop) {
>                  if (cpu->unplug) {
> @@ -1385,6 +1388,10 @@ static void *qemu_tcg_cpu_thread_fn(void *arg)
>                   */
>                  g_assert(cpu->halted);
>                  break;
> +            case EXCP_ATOMIC:
> +                qemu_mutex_unlock_iothread();
> +                cpu_exec_step_atomic(cpu);
> +                qemu_mutex_lock_iothread();
>              default:
>                  /* Ignore everything else? */
>                  break;


--
Alex Bennée