When a zero-length packet has been requested and another packet is
written into the fifo, the MX1 tends to send the first byte of the
previous packet instead of the first byte of the current packet.
The CRC is adjusted accordingly so that this packet is _not_
discarded by the host.
Waiting for the ZLPS bit to clear avoids these bad packets.
Signed-off-by: Daniel Glöckner <dg@xxxxxxxxx>
---
drivers/usb/gadget/imx_udc.c | 7 +++++++
1 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/drivers/usb/gadget/imx_udc.c b/drivers/usb/gadget/imx_udc.c
index ff7658e..642cceb 100644
--- a/drivers/usb/gadget/imx_udc.c
+++ b/drivers/usb/gadget/imx_udc.c
@@ -415,6 +415,13 @@ static int write_packet(struct imx_ep_struct *imx_ep, struct imx_request *req)
u8 *buf;
int length, count, temp;
+ if (unlikely(__raw_readl(imx_ep->imx_usb->base +
+ USB_EP_STAT(EP_NO(imx_ep))) & EPSTAT_ZLPS)) {
+ D_TRX(imx_ep->imx_usb->dev, "<%s> zlp still queued in EP %s\n",
+ __func__, imx_ep->ep.name);
+ return -1;
+ }
+
buf = req->req.buf + req->req.actual;
prefetch(buf);
--
1.6.1.3
--
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
