Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To:     linux-kernel@vger.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        stable@vger.kernel.org,
        =?UTF-8?q?Ga=C3=ABl=20PORTAY?= <gael.portay@collabora.com>,
        Steve Longerbeam <slongerbeam@gmail.com>,
        Hans Verkuil <hverkuil-cisco@xs4all.nl>,
        Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Subject: [PATCH 4.14 171/183] media: imx: prpencvf: Stop upstream before disabling IDMA channel
Date:   Fri, 22 Mar 2019 12:16:39 +0100
Message-Id: <20190322111254.289598788@linuxfoundation.org>
In-Reply-To: <20190322111241.819468003@linuxfoundation.org>
References: <20190322111241.819468003@linuxfoundation.org>
User-Agent: quilt/0.65
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

4.14-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Steve Longerbeam <slongerbeam@gmail.com>

commit a19c22677377b87e4354f7306f46ad99bc982a9f upstream.

Upstream must be stopped immediately after receiving the last EOF and
before disabling the IDMA channel. This can be accomplished by moving
upstream stream off to just after receiving the last EOF completion in
prp_stop(). For symmetry also move upstream stream on to end of
prp_start().

This fixes a complete system hard lockup on the SabreAuto when streaming
from the ADV7180, by repeatedly sending a stream off immediately followed
by stream on:

while true; do v4l2-ctl  -d1 --stream-mmap --stream-count=3; done

Eventually this either causes the system lockup or EOF timeouts at all
subsequent stream on, until a system reset.

The lockup occurs when disabling the IDMA channel at stream off. Stopping
the video data stream entering the IDMA channel before disabling the
channel itself appears to be a reliable fix for the hard lockup.

Fixes: f0d9c8924e2c3 ("[media] media: imx: Add IC subdev drivers")

Reported-by: Gaël PORTAY <gael.portay@collabora.com>
Tested-by: Gaël PORTAY <gael.portay@collabora.com>
Signed-off-by: Steve Longerbeam <slongerbeam@gmail.com>
Cc: stable@vger.kernel.org	# for 4.13 and up
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 drivers/staging/media/imx/imx-ic-prpencvf.c |   26 +++++++++++++++++---------
 1 file changed, 17 insertions(+), 9 deletions(-)

--- a/drivers/staging/media/imx/imx-ic-prpencvf.c
+++ b/drivers/staging/media/imx/imx-ic-prpencvf.c
@@ -676,12 +676,23 @@ static int prp_start(struct prp_priv *pr
 		goto out_free_nfb4eof_irq;
 	}
 
+	/* start upstream */
+	ret = v4l2_subdev_call(priv->src_sd, video, s_stream, 1);
+	ret = (ret && ret != -ENOIOCTLCMD) ? ret : 0;
+	if (ret) {
+		v4l2_err(&ic_priv->sd,
+			 "upstream stream on failed: %d\n", ret);
+		goto out_free_eof_irq;
+	}
+
 	/* start the EOF timeout timer */
 	mod_timer(&priv->eof_timeout_timer,
 		  jiffies + msecs_to_jiffies(IMX_MEDIA_EOF_TIMEOUT));
 
 	return 0;
 
+out_free_eof_irq:
+	devm_free_irq(ic_priv->dev, priv->eof_irq, priv);
 out_free_nfb4eof_irq:
 	devm_free_irq(ic_priv->dev, priv->nfb4eof_irq, priv);
 out_unsetup:
@@ -713,6 +724,12 @@ static void prp_stop(struct prp_priv *pr
 	if (ret == 0)
 		v4l2_warn(&ic_priv->sd, "wait last EOF timeout\n");
 
+	/* stop upstream */
+	ret = v4l2_subdev_call(priv->src_sd, video, s_stream, 0);
+	if (ret && ret != -ENOIOCTLCMD)
+		v4l2_warn(&ic_priv->sd,
+			  "upstream stream off failed: %d\n", ret);
+
 	devm_free_irq(ic_priv->dev, priv->eof_irq, priv);
 	devm_free_irq(ic_priv->dev, priv->nfb4eof_irq, priv);
 
@@ -1144,15 +1161,6 @@ static int prp_s_stream(struct v4l2_subd
 	if (ret)
 		goto out;
 
-	/* start/stop upstream */
-	ret = v4l2_subdev_call(priv->src_sd, video, s_stream, enable);
-	ret = (ret && ret != -ENOIOCTLCMD) ? ret : 0;
-	if (ret) {
-		if (enable)
-			prp_stop(priv);
-		goto out;
-	}
-
 update_count:
 	priv->stream_count += enable ? 1 : -1;
 	if (priv->stream_count < 0)