usb: renesas_usbhs: Don't disable the pipe if Control write status stage

This patch fixes an issue that sometimes this controller is not able
to complete the Control write status stage.

This driver should enable DCPCTR.CCPL and PID_BUF to complete the status
stage. However, if this driver detects the ctrl_stage interruption first
before the control write data is received, this driver will clear the
PID_BUF wrongly in the usbhsf_pio_try_pop(). To avoid this issue, this
patch doesn't clear the PID_BUF in the usbhsf_pio_try_pop().
(Since also the privious code doesn't disable the PID_BUF after a control
 transfer was finished, this patch doesn't have any side efforts.)

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/drivers/usb/renesas_usbhs/fifo.c b/drivers/usb/renesas_usbhs/fifo.c
index 50f4e3a120727524bed9bbd1797ccd6b49542110..fa126a2b0c33c18ae174e57ce087bd1fe9f4cdff 100644 (file)
--- a/drivers/usb/renesas_usbhs/fifo.c
+++ b/drivers/usb/renesas_usbhs/fifo.c
@@ -678,7 +678,14 @@ static int usbhsf_pio_try_pop(struct usbhs_pkt *pkt, int *is_done)
                *is_done = 1;
                usbhsf_rx_irq_ctrl(pipe, 0);
                usbhs_pipe_running(pipe, 0);
-               usbhs_pipe_disable(pipe);       /* disable pipe first */
+               /*
+                * If function mode, since this controller is possible to enter
+                * Control Write status stage at this timing, this driver
+                * should not disable the pipe. If such a case happens, this
+                * controller is not able to complete the status stage.
+                */
+               if (!usbhs_mod_is_host(priv) && !usbhs_pipe_is_dcp(pipe))
+                       usbhs_pipe_disable(pipe);       /* disable pipe first */
        }
 
        /*