Re: [PATCH] gpio: virtio: remove timeout

On Mon, Dec 20, 2021 at 03:54:56PM +0100, Bartosz Golaszewski wrote:

On Mon, Dec 20, 2021 at 2:07 PM Vincent Whitchurch
[off-list ref] wrote:

quoted

The driver imposes an arbitrary one second timeout on virtio requests,
but the specification doesn't prevent the virtio device from taking
longer to process requests, so remove this timeout to support all
systems and device implementations.

Fixes: 3a29355a22c0275fe86 ("gpio: Add virtio-gpio driver")
Signed-off-by: Vincent Whitchurch <redacted>
---
 drivers/gpio/gpio-virtio.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/drivers/gpio/gpio-virtio.c b/drivers/gpio/gpio-virtio.c
index 84f96b78f32a..9f4941bc5760 100644
--- a/drivers/gpio/gpio-virtio.c
+++ b/drivers/gpio/gpio-virtio.c

@@ -100,11 +100,7 @@ static int _virtio_gpio_req(struct virtio_gpio *vgpio, u16 type, u16 gpio,
        virtqueue_kick(vgpio->request_vq);
        mutex_unlock(&vgpio->lock);

-       if (!wait_for_completion_timeout(&line->completion, HZ)) {
-               dev_err(dev, "GPIO operation timed out\n");
-               ret = -ETIMEDOUT;
-               goto out;
-       }
+       wait_for_completion(&line->completion);

        if (unlikely(res->status != VIRTIO_GPIO_STATUS_OK)) {
                dev_err(dev, "GPIO request failed: %d\n", gpio);
--

2.33.1

My knowledge of virtio is limited, I hope this is not a stupid question.

Does this mean the operation can get stuck indefinitely?

Bart

Only if the device is broken. which given it's part of the
hypervisor, is par for the course.

-- 
MST

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization