Linux Virtualization - Mar 2013 - [PATCH] Revert "virtio_console: Initialize guest_connected=true for rproc_serial"

From: Sjur Br?ndeland <sjur.brandeland at stericsson.com>

This reverts commit 8078db789a92b10ff6e2d713231b5367e014c53b.

The reverted patch caused opening of ports to fail for rproc_serial.
In probe guest_connected was set to true, but port_fops_open()
fails with -EMFILE if guest_connected already is true.

Signed-off-by: Sjur Br?ndeland <sjur.brandeland at stericsson.com>
---
Hi Rusty,

Here is a fix intended for 3.9.
Sorry for the churn here :-(

Regards,
Sjur

 drivers/char/virtio_console.c |    7 +++++--
 1 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c
index e905d5f..031be0b 100644
--- a/drivers/char/virtio_console.c
+++ b/drivers/char/virtio_console.c
@@ -1436,7 +1436,7 @@ static int add_port(struct ports_device *portdev, u32 id)
 		 * rproc_serial does not want the console port, only
 		 * the generic port implementation.
 		 */
-		port->host_connected = port->guest_connected = true;
+		port->host_connected = true;
 	else if (!use_multiport(port->portdev)) {
 		/*
 		 * If we're not using multiport support,
@@ -1757,8 +1757,11 @@ static void in_intr(struct virtqueue *vq)
 	 * tty is spawned) and the host sends out data to console
 	 * ports.  For generic serial ports, the host won't
 	 * (shouldn't) send data till the guest is connected.
+	 * However a remote device might send data before the port is
+	 * connected. So don't remove data from a rproc_serial device.
 	 */
-	if (!port->guest_connected)
+
+	if (!port->guest_connected &&
!is_rproc_serial(port->portdev->vdev))
 		discard_port_data(port);
 
 	spin_unlock_irqrestore(&port->inbuf_lock, flags);
-- 
1.7.5.4

sjur.brandeland at stericsson.com writes:
> From: Sjur Br?ndeland <sjur.brandeland at stericsson.com>
>
> This reverts commit 8078db789a92b10ff6e2d713231b5367e014c53b.
>
> The reverted patch caused opening of ports to fail for rproc_serial.
> In probe guest_connected was set to true, but port_fops_open()
> fails with -EMFILE if guest_connected already is true.
>
> Signed-off-by: Sjur Br?ndeland <sjur.brandeland at stericsson.com>
> ---
> Hi Rusty,
>
> Here is a fix intended for 3.9.
> Sorry for the churn here :-(
Applied,
Rusty.

On (Mon) 11 Mar 2013 [16:15:00], sjur.brandeland at stericsson.com
wrote:
> From: Sjur Br?ndeland <sjur.brandeland at stericsson.com>
> 
> This reverts commit 8078db789a92b10ff6e2d713231b5367e014c53b.
> 
> The reverted patch caused opening of ports to fail for rproc_serial.
> In probe guest_connected was set to true, but port_fops_open()
> fails with -EMFILE if guest_connected already is true.
OK, I missed that.  Can you add a comment near the 2nd hunk mentioning
this?


		Amit

Amit Shah <amit.shah at redhat.com> writes:
> On (Wed) 13 Mar 2013 [12:10:48], sjur.brandeland at stericsson.com wrote:
>> From: Sjur Br??ndeland <sjur.brandeland at stericsson.com>
>> 
>> 
>> This reverts commit 8078db789a92b10ff6e2d713231b5367e014c53b, and
>> adds a lengthy comment explaining the problem area.
>> 
>> The reverted patch caused opening of ports to fail for rproc_serial.
>> In probe guest_connected was set to true, but port_fops_open()
>> fails with -EMFILE if guest_connected already is true.
>> 
>> Signed-off-by: Sjur Br?ndeland <sjur.brandeland at
stericsson.com>
>
> Acked-by: Amit Shah <amit.shah at redhat.com>
>
> 		Amit
Applied to my fixes branch.

Thanks,
Rusty.