For implicit feedback endpoints, the number of bytes for each packet is matched by the corresponding synchronizing endpoint. The size is calculated by taking the actual size and dividing it by the stride - currently by the endpoint's stride, but we should use the synchronization source's stride. This is evident when the number of channels differ between the synchronization source and the implicitly fed-back endpoint, as with M-Audio Fast Track C400 - the synchronization source (capture) has 4 channels, while the implicit feedback mode endpoint has 6.
Signed-off-by: Eldad Zack eldad@fogrefinery.com --- sound/usb/endpoint.c | 9 ++++++--- 1 files changed, 6 insertions(+), 3 deletions(-)
diff --git a/sound/usb/endpoint.c b/sound/usb/endpoint.c index 34de6f2..72e711c 100644 --- a/sound/usb/endpoint.c +++ b/sound/usb/endpoint.c @@ -1034,15 +1034,18 @@ void snd_usb_handle_sync_urb(struct snd_usb_endpoint *ep, /* * Iterate through the inbound packet and prepare the lengths * for the output packet. The OUT packet we are about to send - * will have the same amount of payload bytes than the IN - * packet we just received. + * will have the same amount of payload bytes per stride as the + * IN packet we just received. Since the actual size is scaled + * by the stride, use the sender stride to calculate the length + * in case the number of channels differ between the implicitly + * fed-back endpoint and the synchronizing endpoint. */
out_packet->packets = in_ctx->packets; for (i = 0; i < in_ctx->packets; i++) { if (urb->iso_frame_desc[i].status == 0) out_packet->packet_size[i] = - urb->iso_frame_desc[i].actual_length / ep->stride; + urb->iso_frame_desc[i].actual_length / sender->stride; else out_packet->packet_size[i] = 0; }