From patchwork Thu Dec 18 14:49:46 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ben Hutchings X-Patchwork-Id: 27570 X-Patchwork-Delegate: g.liakhovetski@gmx.de Received: from mail.tu-berlin.de ([130.149.7.33]) by www.linuxtv.org with esmtp (Exim 4.72) (envelope-from ) id 1Y1cOp-0003j3-2T; Thu, 18 Dec 2014 15:49:55 +0100 X-tubIT-Incoming-IP: 209.132.180.67 Received: from vger.kernel.org ([209.132.180.67]) by mail.tu-berlin.de (exim-4.72/mailfrontend-8) with esmtp id 1Y1cOm-0005LD-kk; Thu, 18 Dec 2014 15:49:54 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751141AbaLROtu (ORCPT + 1 other); Thu, 18 Dec 2014 09:49:50 -0500 Received: from ducie-dc1.codethink.co.uk ([185.25.241.215]:52201 "EHLO ducie-dc1.codethink.co.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750899AbaLROtt (ORCPT ); Thu, 18 Dec 2014 09:49:49 -0500 Received: from localhost (localhost [127.0.0.1]) by ducie-dc1.codethink.co.uk (Postfix) with ESMTP id 7B9AE4605EB; Thu, 18 Dec 2014 14:49:48 +0000 (GMT) X-Virus-Scanned: Debian amavisd-new at ducie-dc1.codethink.co.uk Received: from ducie-dc1.codethink.co.uk ([127.0.0.1]) by localhost (ducie-dc1.codethink.co.uk [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id oF7RJQdwHV-K; Thu, 18 Dec 2014 14:49:46 +0000 (GMT) Received: from [10.24.1.188] (xylophone.dyn.ducie.codethink.co.uk [10.24.1.188]) by ducie-dc1.codethink.co.uk (Postfix) with ESMTPSA id 43CEC46011C; Thu, 18 Dec 2014 14:49:46 +0000 (GMT) Message-ID: <1418914186.22813.16.camel@xylophone.i.decadent.org.uk> Subject: [RFC PATCH 3/5] media: rcar_vin: Fix race condition terminating stream From: Ben Hutchings To: linux-media@vger.kernel.org Cc: Guennadi Liakhovetski , linux-kernel@codethink.co.uk, William Towle , Sergei Shtylyov , Hans Verkuil Date: Thu, 18 Dec 2014 14:49:46 +0000 In-Reply-To: <1418914070.22813.13.camel@xylophone.i.decadent.org.uk> References: <1418914070.22813.13.camel@xylophone.i.decadent.org.uk> Organization: Codethink Ltd. X-Mailer: Evolution 3.4.4-3 Mime-Version: 1.0 Sender: linux-media-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-media@vger.kernel.org X-PMX-Version: 6.0.0.2142326, Antispam-Engine: 2.7.2.2107409, Antispam-Data: 2014.12.18.144220 X-PMX-Spam: Gauge=IIIIIIIII, Probability=9%, Report=' MULTIPLE_RCPTS 0.1, HTML_00_01 0.05, HTML_00_10 0.05, MIME_LOWER_CASE 0.05, MSGID_ADDED_BY_MTA 0.05, BODY_SIZE_3000_3999 0, BODY_SIZE_5000_LESS 0, BODY_SIZE_7000_LESS 0, CT_TEXT_PLAIN_UTF8_CAPS 0, REFERENCES 0, URI_ENDS_IN_HTML 0, __ANY_URI 0, __CP_MEDIA_BODY 0, __CP_URI_IN_BODY 0, __CT 0, __CTE 0, __CT_TEXT_PLAIN 0, __HAS_FROM 0, __HAS_MSGID 0, __HAS_X_MAILER 0, __HAS_X_MAILING_LIST 0, __IN_REP_TO 0, __MIME_TEXT_ONLY 0, __MIME_VERSION 0, __MULTIPLE_RCPTS_CC_X2 0, __REFERENCES 0, __SANE_MSGID 0, __SUBJ_ALPHA_END 0, __TO_MALFORMED_2 0, __TO_NO_NAME 0, __URI_NO_WWW 0, __URI_NS ' From: Ian Molton This patch fixes a race condition whereby a frame being captured may generate an interrupt between requesting capture to halt and freeing buffers. This condition is exposed by the earlier patch that explicitly calls vb2_buffer_done() during stop streaming. The solution is to wait for capture to finish prior to finalising these buffers. Signed-off-by: Ian Molton Signed-off-by: William Towle --- drivers/media/platform/soc_camera/rcar_vin.c | 43 +++++++++++++++++--------- 1 file changed, 28 insertions(+), 15 deletions(-) diff --git a/drivers/media/platform/soc_camera/rcar_vin.c b/drivers/media/platform/soc_camera/rcar_vin.c index 7069176..b234e57 100644 --- a/drivers/media/platform/soc_camera/rcar_vin.c +++ b/drivers/media/platform/soc_camera/rcar_vin.c @@ -458,6 +458,29 @@ error: vb2_buffer_done(vb, VB2_BUF_STATE_ERROR); } +/* + * Wait for capture to stop and all in-flight buffers to be finished with by + * the video hardware. This must be called under &priv->lock + * + */ +static void rcar_vin_wait_stop_streaming(struct rcar_vin_priv *priv) +{ + while (priv->state != STOPPED) { + + /* issue stop if running */ + if (priv->state == RUNNING) + rcar_vin_request_capture_stop(priv); + + /* wait until capturing has been stopped */ + if (priv->state == STOPPING) { + priv->request_to_stop = true; + spin_unlock_irq(&priv->lock); + wait_for_completion(&priv->capture_stop); + spin_lock_irq(&priv->lock); + } + } +} + static void rcar_vin_videobuf_release(struct vb2_buffer *vb) { struct soc_camera_device *icd = soc_camera_from_vb2q(vb->vb2_queue); @@ -465,7 +488,6 @@ static void rcar_vin_videobuf_release(struct vb2_buffer *vb) struct rcar_vin_priv *priv = ici->priv; unsigned int i; int buf_in_use = 0; - spin_lock_irq(&priv->lock); /* Is the buffer in use by the VIN hardware? */ @@ -477,20 +499,8 @@ static void rcar_vin_videobuf_release(struct vb2_buffer *vb) } if (buf_in_use) { - while (priv->state != STOPPED) { - - /* issue stop if running */ - if (priv->state == RUNNING) - rcar_vin_request_capture_stop(priv); - - /* wait until capturing has been stopped */ - if (priv->state == STOPPING) { - priv->request_to_stop = true; - spin_unlock_irq(&priv->lock); - wait_for_completion(&priv->capture_stop); - spin_lock_irq(&priv->lock); - } - } + rcar_vin_wait_stop_streaming(priv); + /* * Capturing has now stopped. The buffer we have been asked * to release could be any of the current buffers in use, so @@ -520,12 +530,15 @@ static void rcar_vin_stop_streaming(struct vb2_queue *vq) spin_lock_irq(&priv->lock); + rcar_vin_wait_stop_streaming(priv); + for (i = 0; i < vq->num_buffers; ++i) if (vq->bufs[i]->state == VB2_BUF_STATE_ACTIVE) vb2_buffer_done(vq->bufs[i], VB2_BUF_STATE_ERROR); list_for_each_safe(buf_head, tmp, &priv->capture) list_del_init(buf_head); + spin_unlock_irq(&priv->lock); }