search for: control_intr

...,6 +1484,10 @@ static void control_work_handler(struct work_struct *work) } } spin_unlock(&portdev->cvq_lock); + if (unlikely(!virtqueue_enable_cb(vq))) { + virtqueue_disable_cb(vq); + goto start; + } } static void out_intr(struct virtqueue *vq) @@ -1533,6 +1538,7 @@ static void control_intr(struct virtqueue *vq) { struct ports_device *portdev; + virtqueue_disable_cb(vq); portdev = vq->vdev->priv; schedule_work(&portdev->control_work); } -- 1.7.7.4 Amit

...,6 +1484,10 @@ static void control_work_handler(struct work_struct *work) } } spin_unlock(&portdev->cvq_lock); + if (unlikely(!virtqueue_enable_cb(vq))) { + virtqueue_disable_cb(vq); + goto start; + } } static void out_intr(struct virtqueue *vq) @@ -1533,6 +1538,7 @@ static void control_intr(struct virtqueue *vq) { struct ports_device *portdev; + virtqueue_disable_cb(vq); portdev = vq->vdev->priv; schedule_work(&portdev->control_work); } -- 1.7.7.4 Amit

Hello, The first patch here fixes the race seen by Miche. He hasn't yet reported back if this fixes the races he saw, but Joy Pu from Red Hat tested this patch with hot-plugging/unplugging ports in a loop. Before this patch, he saw some freezes as well as sysfs warnings. After applying the patch, all was well. The second patch can be folded into the series fixing S4 for virtio-console. It

Hello, The first patch here fixes the race seen by Miche. He hasn't yet reported back if this fixes the races he saw, but Joy Pu from Red Hat tested this patch with hot-plugging/unplugging ports in a loop. Before this patch, he saw some freezes as well as sysfs warnings. After applying the patch, all was well. The second patch can be folded into the series fixing S4 for virtio-console. It

Hey Rusty, These updated patches in the series return -EFAULT on copy_xx_user errors and also move the copy_from_user into fops_write() instead of it being in send_buf. This enables send_buf to just read from kernel buffers, making it simpler. This also allows write()s to write more to the host in one go, removingthe 4k limitation. I do limit the writes to 32k at once to not put too much

Hey Rusty, These updated patches in the series return -EFAULT on copy_xx_user errors and also move the copy_from_user into fops_write() instead of it being in send_buf. This enables send_buf to just read from kernel buffers, making it simpler. This also allows write()s to write more to the host in one go, removingthe 4k limitation. I do limit the writes to 32k at once to not put too much