Re: [PATCH 1/4] cdc-acm: reassemble fragmented notifications

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Am Dienstag, den 14.03.2017, 21:14 +0100 schrieb Tobias Herzog:
> USB devices may have very limitited endpoint packet sizes, so that
> notifications can not be transferred within one single usb packet.
> Reassembling of multiple packages may be necessary.

Hi,

thank you for the patch. Unfortunately it has some issue.
Please see the comments inside.

	Regards
		Oliver

> 
> Signed-off-by: Tobias Herzog <t-herzog@xxxxxx>
> ---
>  drivers/usb/class/cdc-acm.c | 102 +++++++++++++++++++++++++++++++-------------
>  drivers/usb/class/cdc-acm.h |   2 +
>  2 files changed, 75 insertions(+), 29 deletions(-)
> 
> diff --git a/drivers/usb/class/cdc-acm.c b/drivers/usb/class/cdc-acm.c
> index e35b150..40714fe 100644
> --- a/drivers/usb/class/cdc-acm.c
> +++ b/drivers/usb/class/cdc-acm.c
> @@ -282,39 +282,13 @@ static DEVICE_ATTR(iCountryCodeRelDate, S_IRUGO, show_country_rel_date, NULL);
>   * Interrupt handlers for various ACM device responses
>   */
>  
> -/* control interface reports status changes with "interrupt" transfers */
> -static void acm_ctrl_irq(struct urb *urb)
> +static void acm_process_notification(struct acm *acm, unsigned char *buf)
>  {
> -	struct acm *acm = urb->context;
> -	struct usb_cdc_notification *dr = urb->transfer_buffer;
> -	unsigned char *data;
>  	int newctrl;
>  	int difference;
> -	int retval;
> -	int status = urb->status;
> +	struct usb_cdc_notification *dr = (struct usb_cdc_notification *)buf;
> +	unsigned char *data = (unsigned char *)(dr + 1);
>  
> -	switch (status) {
> -	case 0:
> -		/* success */
> -		break;
> -	case -ECONNRESET:
> -	case -ENOENT:
> -	case -ESHUTDOWN:
> -		/* this urb is terminated, clean up */
> -		dev_dbg(&acm->control->dev,
> -			"%s - urb shutting down with status: %d\n",
> -			__func__, status);
> -		return;
> -	default:
> -		dev_dbg(&acm->control->dev,
> -			"%s - nonzero urb status received: %d\n",
> -			__func__, status);
> -		goto exit;
> -	}
> -
> -	usb_mark_last_busy(acm->dev);
> -
> -	data = (unsigned char *)(dr + 1);
>  	switch (dr->bNotificationType) {
>  	case USB_CDC_NOTIFY_NETWORK_CONNECTION:
>  		dev_dbg(&acm->control->dev,
> @@ -363,8 +337,74 @@ static void acm_ctrl_irq(struct urb *urb)
>  			__func__,
>  			dr->bNotificationType, dr->wIndex,
>  			dr->wLength, data[0], data[1]);
> +	}
> +}
> +
> +/* control interface reports status changes with "interrupt" transfers */
> +static void acm_ctrl_irq(struct urb *urb)
> +{
> +	struct acm *acm = urb->context;
> +	struct usb_cdc_notification *dr = urb->transfer_buffer;
> +	unsigned int current_size = urb->actual_length;
> +	unsigned int expected_size, copy_size;
> +	int retval;
> +	int status = urb->status;
> +
> +	switch (status) {
> +	case 0:
> +		/* success */
>  		break;
> +	case -ECONNRESET:
> +	case -ENOENT:
> +	case -ESHUTDOWN:
> +		/* this urb is terminated, clean up */
> +		kfree(acm->notification_buffer);
> +		acm->notification_buffer = NULL;

Why? Disconnect() will free it anyway. It should be enough
to discard the content.

> +		dev_dbg(&acm->control->dev,
> +			"%s - urb shutting down with status: %d\n",
> +			__func__, status);
> +		return;
> +	default:
> +		dev_dbg(&acm->control->dev,
> +			"%s - nonzero urb status received: %d\n",
> +			__func__, status);
> +		goto exit;
>  	}
> +
> +	usb_mark_last_busy(acm->dev);
> +
> +	if (acm->notification_buffer)
> +		dr = (struct usb_cdc_notification *)acm->notification_buffer;
> +
> +	/* assume the first package contains at least two bytes */
> +	expected_size = dr->wLength + 8;

You need the explain where you got the 8 from. In fact a define would
be best.

> +
> +	if (current_size < expected_size) {
> +		/* notification is transmitted framented, reassemble */

Please fix the typo.

> +		if (!acm->notification_buffer) {
> +			acm->notification_buffer =
> +					kmalloc(expected_size, GFP_ATOMIC);

This can fail. You _must_ check for that.

> +			acm->nb_index = 0;
> +		}
> +
> +		copy_size = min(current_size,
> +				expected_size - acm->nb_index);
> +
> +		memcpy(&acm->notification_buffer[acm->nb_index],
> +		       urb->transfer_buffer, copy_size);
> +		acm->nb_index += copy_size;
> +		current_size = acm->nb_index;
> +	}
> +
> +	if (current_size < expected_size)
> +		goto exit;

This is an unneeded goto.

> +	/* notification complete */
> +	acm_process_notification(acm, (unsigned char *)dr);
> +
> +	kfree(acm->notification_buffer);

Why? If one message was fragmented, the next one will also likely be
fragmented. Why release the buffer before you know whether it can be
reused?

> +	acm->notification_buffer = NULL;
> +
>  exit:
>  	retval = usb_submit_urb(urb, GFP_ATOMIC);
>  	if (retval && retval != -EPERM)
> @@ -1488,6 +1528,8 @@ static int acm_probe(struct usb_interface *intf,
>  			 epctrl->bInterval ? epctrl->bInterval : 16);
>  	acm->ctrlurb->transfer_flags |= URB_NO_TRANSFER_DMA_MAP;
>  	acm->ctrlurb->transfer_dma = acm->ctrl_dma;
> +	acm->notification_buffer = NULL;
> +	acm->nb_index = 0;
>  
>  	dev_info(&intf->dev, "ttyACM%d: USB ACM device\n", minor);
>  
> @@ -1580,6 +1622,8 @@ static void acm_disconnect(struct usb_interface *intf)
>  	usb_free_coherent(acm->dev, acm->ctrlsize, acm->ctrl_buffer, acm->ctrl_dma);
>  	acm_read_buffers_free(acm);
>  
> +	kfree(acm->notification_buffer);
> +
>  	if (!acm->combined_interfaces)
>  		usb_driver_release_interface(&acm_driver, intf == acm->control ?
>  					acm->data : acm->control);
> diff --git a/drivers/usb/class/cdc-acm.h b/drivers/usb/class/cdc-acm.h
> index c980f11..bc07fb2 100644
> --- a/drivers/usb/class/cdc-acm.h
> +++ b/drivers/usb/class/cdc-acm.h
> @@ -98,6 +98,8 @@ struct acm {
>  	struct acm_wb *putbuffer;			/* for acm_tty_put_char() */
>  	int rx_buflimit;
>  	spinlock_t read_lock;
> +	u8 *notification_buffer;			/* to reassemble fragmented notifications */
> +	unsigned int nb_index;
>  	int write_used;					/* number of non-empty write buffers */
>  	int transmitting;
>  	spinlock_t write_lock;

--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html



[Index of Archives]     [Linux Media]     [Linux Input]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Old Linux USB Devel Archive]

  Powered by Linux