From: Jing Leng <[email protected]>
EP0 transfer is bi-directional, but in the cdnsp gadget, the
transmission direction of EP0 is not changed after it is
initialized to IN, so the OUT data from EP0 received by the host
is invalid.
The value of ep0_expect_in will change according to the value of
bRequestType in the SETUP transaction of control transfer, so we
can use it as the transmission direction of EP0.
Signed-off-by: Jing Leng <[email protected]>
---
drivers/usb/cdns3/cdnsp-gadget.c | 11 +++++++----
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/drivers/usb/cdns3/cdnsp-gadget.c b/drivers/usb/cdns3/cdnsp-gadget.c
index c67715f6f756..526f80651d35 100644
--- a/drivers/usb/cdns3/cdnsp-gadget.c
+++ b/drivers/usb/cdns3/cdnsp-gadget.c
@@ -345,6 +345,7 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
{
struct cdnsp_device *pdev = pep->pdev;
struct usb_request *request;
+ u8 direction;
int ret;
if (preq->epnum == 0 && !list_empty(&pep->pending_list)) {
@@ -355,11 +356,14 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
request = &preq->request;
request->actual = 0;
request->status = -EINPROGRESS;
- preq->direction = pep->direction;
+
+ direction = usb_endpoint_xfer_control(pep->endpoint.desc) ?
+ pdev->ep0_expect_in : pep->direction;
+ preq->direction = direction;
preq->epnum = pep->number;
preq->td.drbl = 0;
- ret = usb_gadget_map_request_by_dev(pdev->dev, request, pep->direction);
+ ret = usb_gadget_map_request_by_dev(pdev->dev, request, direction);
if (ret) {
trace_cdnsp_request_enqueue_error(preq);
return ret;
@@ -387,8 +391,7 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
return 0;
unmap:
- usb_gadget_unmap_request_by_dev(pdev->dev, &preq->request,
- pep->direction);
+ usb_gadget_unmap_request_by_dev(pdev->dev, &preq->request, direction);
list_del(&preq->list);
trace_cdnsp_request_enqueue_error(preq);
--
2.34.1
On Mon, Sep 26, 2022 at 03:59:02PM +0800, [email protected] wrote:
> From: Jing Leng <[email protected]>
>
> EP0 transfer is bi-directional, but in the cdnsp gadget, the
> transmission direction of EP0 is not changed after it is
> initialized to IN, so the OUT data from EP0 received by the host
> is invalid.
>
> The value of ep0_expect_in will change according to the value of
> bRequestType in the SETUP transaction of control transfer, so we
> can use it as the transmission direction of EP0.
>
> Signed-off-by: Jing Leng <[email protected]>
Your email does not match your From: line and the signature of it shows
that it is invalid and does not pass authentication :(
Please work with your university to get an email address that works
properly.
thanks,
greg k-h
EP0 transfer is bi-directional, but in the cdnsp gadget, the
transmission direction of EP0 is not changed after it is
initialized to IN, so the OUT data from EP0 received by the host
is invalid.
The value of ep0_expect_in will change according to the value of
bRequestType in the SETUP transaction of control transfer, so we
can use it as the transmission direction of EP0.
Signed-off-by: Jing Leng <[email protected]>
Signed-off-by: Jing Leng <[email protected]>
---
ChangeLog v1->v2:
- Rebase the patch.
- Make email addresses ('From' and 'Signed-off-by') consistent.
---
drivers/usb/cdns3/cdnsp-gadget.c | 11 +++++++----
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/drivers/usb/cdns3/cdnsp-gadget.c b/drivers/usb/cdns3/cdnsp-gadget.c
index c67715f6f756..526f80651d35 100644
--- a/drivers/usb/cdns3/cdnsp-gadget.c
+++ b/drivers/usb/cdns3/cdnsp-gadget.c
@@ -345,6 +345,7 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
{
struct cdnsp_device *pdev = pep->pdev;
struct usb_request *request;
+ u8 direction;
int ret;
if (preq->epnum == 0 && !list_empty(&pep->pending_list)) {
@@ -355,11 +356,14 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
request = &preq->request;
request->actual = 0;
request->status = -EINPROGRESS;
- preq->direction = pep->direction;
+
+ direction = usb_endpoint_xfer_control(pep->endpoint.desc) ?
+ pdev->ep0_expect_in : pep->direction;
+ preq->direction = direction;
preq->epnum = pep->number;
preq->td.drbl = 0;
- ret = usb_gadget_map_request_by_dev(pdev->dev, request, pep->direction);
+ ret = usb_gadget_map_request_by_dev(pdev->dev, request, direction);
if (ret) {
trace_cdnsp_request_enqueue_error(preq);
return ret;
@@ -387,8 +391,7 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
return 0;
unmap:
- usb_gadget_unmap_request_by_dev(pdev->dev, &preq->request,
- pep->direction);
+ usb_gadget_unmap_request_by_dev(pdev->dev, &preq->request, direction);
list_del(&preq->list);
trace_cdnsp_request_enqueue_error(preq);
--
2.17.1
EP0 transfer is bi-directional, but in the cdnsp gadget, the
transmission direction of EP0 is not changed after it is
initialized to IN, so the OUT data from EP0 received by the host
is invalid.
The value of ep0_expect_in will change according to the value of
bRequestType in the SETUP transaction of control transfer, so we
can use it as the transmission direction of EP0.
Signed-off-by: Jing Leng <[email protected]>
---
ChangeLog v2->v3:
- Repair my email address.
ChangeLog v1->v2:
- Rebase the patch.
- Make email addresses ('From' and 'Signed-off-by') consistent.
---
drivers/usb/cdns3/cdnsp-gadget.c | 11 +++++++----
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/drivers/usb/cdns3/cdnsp-gadget.c b/drivers/usb/cdns3/cdnsp-gadget.c
index c67715f6f756..526f80651d35 100644
--- a/drivers/usb/cdns3/cdnsp-gadget.c
+++ b/drivers/usb/cdns3/cdnsp-gadget.c
@@ -345,6 +345,7 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
{
struct cdnsp_device *pdev = pep->pdev;
struct usb_request *request;
+ u8 direction;
int ret;
if (preq->epnum == 0 && !list_empty(&pep->pending_list)) {
@@ -355,11 +356,14 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
request = &preq->request;
request->actual = 0;
request->status = -EINPROGRESS;
- preq->direction = pep->direction;
+
+ direction = usb_endpoint_xfer_control(pep->endpoint.desc) ?
+ pdev->ep0_expect_in : pep->direction;
+ preq->direction = direction;
preq->epnum = pep->number;
preq->td.drbl = 0;
- ret = usb_gadget_map_request_by_dev(pdev->dev, request, pep->direction);
+ ret = usb_gadget_map_request_by_dev(pdev->dev, request, direction);
if (ret) {
trace_cdnsp_request_enqueue_error(preq);
return ret;
@@ -387,8 +391,7 @@ int cdnsp_ep_enqueue(struct cdnsp_ep *pep, struct cdnsp_request *preq)
return 0;
unmap:
- usb_gadget_unmap_request_by_dev(pdev->dev, &preq->request,
- pep->direction);
+ usb_gadget_unmap_request_by_dev(pdev->dev, &preq->request, direction);
list_del(&preq->list);
trace_cdnsp_request_enqueue_error(preq);
--
2.17.1
**********************************************************************
This email and attachments contain Ambarella Proprietary and/or Confidential Information and is intended solely for the use of the individual(s) to whom it is addressed. Any unauthorized review, use, disclosure, distribute, copy, or print is prohibited. If you are not an intended recipient, please contact the sender by reply email and destroy all copies of the original message. Thank you.
On Tue, Nov 01, 2022 at 02:17:30PM +0800, Jing Leng wrote:
> EP0 transfer is bi-directional, but in the cdnsp gadget, the
> transmission direction of EP0 is not changed after it is
> initialized to IN, so the OUT data from EP0 received by the host
> is invalid.
>
> The value of ep0_expect_in will change according to the value of
> bRequestType in the SETUP transaction of control transfer, so we
> can use it as the transmission direction of EP0.
>
> Signed-off-by: Jing Leng <[email protected]>
> ---
> ChangeLog v2->v3:
> - Repair my email address.
Yes, it works, and it's validated!
Nice job, thanks. I'll let the cdns3 maintainer review it first, but
just wanted to say thanks for fixing this up, it makes my life a lot
easier when accepting patches.
greg k-h
>
>On Tue, Nov 01, 2022 at 02:17:30PM +0800, Jing Leng wrote:
>> EP0 transfer is bi-directional, but in the cdnsp gadget, the
>> transmission direction of EP0 is not changed after it is initialized
>> to IN, so the OUT data from EP0 received by the host is invalid.
>>
>> The value of ep0_expect_in will change according to the value of
>> bRequestType in the SETUP transaction of control transfer, so we can
>> use it as the transmission direction of EP0.
>>
>> Signed-off-by: Jing Leng <[email protected]>
Acked-by: Pawel Laszczak <[email protected]>
Thanks,
Pawel Laaszczak
>> ---
>> ChangeLog v2->v3:
>> - Repair my email address.
>
>Yes, it works, and it's validated!
>
>Nice job, thanks. I'll let the cdns3 maintainer review it first, but just wanted to
>say thanks for fixing this up, it makes my life a lot easier when accepting
>patches.
>
>greg k-h
On Tue, Nov 01, 2022 at 10:43:21AM +0100, Greg KH wrote:
> On Tue, Nov 01, 2022 at 02:17:30PM +0800, Jing Leng wrote:
> > EP0 transfer is bi-directional, but in the cdnsp gadget, the
> > transmission direction of EP0 is not changed after it is
> > initialized to IN, so the OUT data from EP0 received by the host
> > is invalid.
> >
> > The value of ep0_expect_in will change according to the value of
> > bRequestType in the SETUP transaction of control transfer, so we
> > can use it as the transmission direction of EP0.
> >
> > Signed-off-by: Jing Leng <[email protected]>
> > ---
> > ChangeLog v2->v3:
> > - Repair my email address.
>
> Yes, it works, and it's validated!
>
> Nice job, thanks. I'll let the cdns3 maintainer review it first, but
> just wanted to say thanks for fixing this up, it makes my life a lot
> easier when accepting patches.
Oops, I missed that email footer, sorry. That is NOT compatible with
Linux kernel development, and as per the recommendation from my legal
people, I have to go revert that change as we can't take it as-is
because it might have been sent out incorrectly as it was stated to be
"Proprietary and/or Confidential Information" which is not allowed in
the Linux kernel.
thanks,
greg k-h