Root-cause:
There is an issue like endpoint is not recognized as primed, when bus
have more pressure and the add dTD tripwire semaphore (ATDTW bit in
USBCMD register) that can cause the controller to ignore a dTD that is
added to a primed endpoint.
This issue observed with the Windows10 host machine.

Workaround:
The software must implement a periodic cycle, and check for each dTD,
if the endpoint is primed. It can do this by reading the corresponding
bits in the ENDPTPRIME and ENDPTSTAT registers. If these bits are read
at 0, the software needs to re-prime the endpoint by writing 1 to the
corresponding bit in the ENDPTPRIME register.

Added conditional revision check of 2.20[CI_REVISION_22].

Signed-off-by: Piyush Mehta <redacted>
---
Changes for V2:
- Addressed Peter review comments - Remove unrelated new-line
- Updated commit message

Link: https://lore.kernel.org/linux-usb/SJ0PR02MB8644CBBA848A0F68323F1AA5D4D99@SJ0PR02MB8644.namprd02.prod.outlook.com/ (local)
---
 drivers/usb/chipidea/udc.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/usb/chipidea/udc.c b/drivers/usb/chipidea/udc.c
index 8834ca6..f9ca501 100644
--- a/drivers/usb/chipidea/udc.c
+++ b/drivers/usb/chipidea/udc.c

@@ -49,6 +49,8 @@ ctrl_endpt_in_desc = {
 	.wMaxPacketSize  = cpu_to_le16(CTRL_PAYLOAD_MAX),
 };
 
+static int reprime_dtd(struct ci_hdrc *ci, struct ci_hw_ep *hwep,
+		       struct td_node *node);
 /**
  * hw_ep_bit: calculates the bit number
  * @num: endpoint number

@@ -599,6 +601,12 @@ static int _hardware_enqueue(struct ci_hw_ep *hwep, struct ci_hw_req *hwreq)
 
 		prevlastnode->ptr->next = cpu_to_le32(next);
 		wmb();
+
+		if (ci->rev == CI_REVISION_22) {
+			if (!hw_read(ci, OP_ENDPTSTAT, BIT(n)))
+				reprime_dtd(ci, hwep, prevlastnode);
+		}
+
 		if (hw_read(ci, OP_ENDPTPRIME, BIT(n)))
 			goto done;
 		do {

-- 
2.7.4