usb: chipidea: udc: workaround for endpoint conflict issue
Subsystems (2)
PATCH v2 1/1 usb: chipidea: udc: workaround for endpoint conflict issue
An endpoint conflict occurs when the USB is working in device mode during an isochronous communication. When the endpointA IN direction is an isochronous IN endpoint, and the host sends an IN token to endpointA on another device, then the OUT transaction may be missed regardless the OUT endpoint number. Generally, this occurs when the device is connected to the host through a hub and other devices are connected to the same hub.
The affected OUT endpoint can be either control, bulk, isochronous, or an interrupt endpoint. After the OUT endpoint is primed, if an IN token to the same endpoint number on another device is received, then the OUT endpoint may be unprimed (cannot be detected by software), which causes this endpoint to no longer respond to the host OUT token, and thus, no corresponding interrupt occurs.
There is no good workaround for this issue, the only thing the software could do is numbering isochronous IN from the highest endpoint since we have observed most of device number endpoint from the lowest.
---
Changes for v2:
- Improve the code sytle
drivers/usb/chipidea/udc.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+)
drivers/usb/chipidea/udc.c +23 -0
@@ -1622,6 +1622,28 @@ static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on)static int ci_udc_start (struct usb_gadget *gadget,struct usb_gadget_driver *driver);static int ci_udc_stop (struct usb_gadget *gadget); + + +/* Match ISOC IN from the highest endpoint */ +static struct usb_ep *ci_udc_match_ep (struct usb_gadget *gadget, +struct usb_endpoint_descriptor *desc, +struct usb_ss_ep_comp_descriptor *comp_desc) +{ + struct ci_hdrc *ci = container_of(gadget, struct ci_hdrc, gadget); +struct usb_ep *ep; +u8 type = desc->bmAttributes &USB_ENDPOINT_XFERTYPE_MASK ; + +if ((type ==USB_ENDPOINT_XFER_ISOC ) && + (desc->bEndpointAddress &USB_DIR_IN )) { +list_for_each_entry_reverse (ep, &ci->gadget.ep_list, ep_list) { +if (ep->caps.dir_in && !ep->claimed) +return ep; + } + } + +return NULL ; +} +/** * Device operations part of the API to the USB controller hardware, * which don't involve endpoints (or i/o)
@@ -1635,6 +1657,7 @@ static const struct usb_gadget_ops usb_gadget_ops = { .vbus_draw = ci_udc_vbus_draw, .udc_start = ci_udc_start, .udc_stop = ci_udc_stop, + .match_ep = ci_udc_match_ep, };static int init_eps (struct ci_hdrc *ci)
Signature · 1 line
2.14.1
Greg KH
maintainerQuoted 39 lines · Peter Chen
An endpoint conflict occurs when the USB is working in device mode during an isochronous communication. When the endpointA IN direction is an isochronous IN endpoint, and the host sends an IN token to endpointA on another device, then the OUT transaction may be missed regardless the OUT endpoint number. Generally, this occurs when the device is connected to the host through a hub and other devices are connected to the same hub.
The affected OUT endpoint can be either control, bulk, isochronous, or an interrupt endpoint. After the OUT endpoint is primed, if an IN token to the same endpoint number on another device is received, then the OUT endpoint may be unprimed (cannot be detected by software), which causes this endpoint to no longer respond to the host OUT token, and thus, no corresponding interrupt occurs.
There is no good workaround for this issue, the only thing the software could do is numbering isochronous IN from the highest endpoint since we have observed most of device number endpoint from the lowest.
Cc Sergei Shtylyov <[email protected]>Cc Jun Li <[email protected]>Signed-off-by Peter Chen <[email protected]>---
Changes for v2:
- Improve the code sytledrivers/usb/chipidea/udc.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+)drivers/usb/chipidea/udc.c +2
-0@@ -1622,6 +1622,28 @@ static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on)static int ci_udc_start (struct usb_gadget *gadget,struct usb_gadget_driver *driver);static int ci_udc_stop (struct usb_gadget *gadget); + +
drivers/usb/chipidea/udc.c static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on) · line 1636
Quoted 10 lines
drivers/usb/chipidea/udc.c +10
-0+/* Match ISOC IN from the highest endpoint */ +static struct usb_ep *ci_udc_match_ep (struct usb_gadget *gadget, +struct usb_endpoint_descriptor *desc, +struct usb_ss_ep_comp_descriptor *comp_desc) +{ + struct ci_hdrc *ci = container_of(gadget, struct ci_hdrc, gadget); +struct usb_ep *ep; +u8 type = desc->bmAttributes &USB_ENDPOINT_XFERTYPE_MASK ; + + if ((type == USB_ENDPOINT_XFER_ISOC) &&
usb_endpoint_xfer_isoc()?
drivers/usb/chipidea/udc.c static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on) · line 1637
Quoted 1 line
drivers/usb/chipidea/udc.c +1
-0+ (desc->bEndpointAddress & USB_DIR_IN)) {
usb_endpoint_dir_in()?
thanks,
greg k-h
Hi Peter,
Quoted 16 lines · Peter Chen
An endpoint conflict occurs when the USB is working in device mode during an isochronous communication. When the endpointA IN direction is an isochronous IN endpoint, and the host sends an IN token to endpointA on another device, then the OUT transaction may be missed regardless the OUT endpoint number. Generally, this occurs when the device is connected to the host through a hub and other devices are connected to the same hub.
The affected OUT endpoint can be either control, bulk, isochronous, or an interrupt endpoint. After the OUT endpoint is primed, if an IN token to the same endpoint number on another device is received, then the OUT endpoint may be unprimed (cannot be detected by software), which causes this endpoint to no longer respond to the host OUT token, and thus, no corresponding interrupt occurs.
Quoted 7 lines
There is no good workaround for this issue, the only thing the software could do is numbering isochronous IN from the highest endpoint since we have observed most of device number endpoint from the lowest.
Cc Sergei Shtylyov <[email protected]>Cc Jun Li <[email protected]>Signed-off-by Peter Chen <[email protected]>
Should this patch have Cc stable so that it gets applied for older kernels?
drivers/usb/chipidea/udc.c static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on) · line 1626
Quoted 10 lines
drivers/usb/chipidea/udc.c +2
-0@@ -1622,6 +1622,28 @@ static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on)static int ci_udc_start (struct usb_gadget *gadget,struct usb_gadget_driver *driver);static int ci_udc_stop (struct usb_gadget *gadget); + +
One blank line is enough.
Quoted 30 lines · Fabio Estevam
Hi Peter,
Quoted 16 lines · Peter Chen
An endpoint conflict occurs when the USB is working in device mode during an isochronous communication. When the endpointA IN direction is an isochronous IN endpoint, and the host sends an IN token to endpointA on another device, then the OUT transaction may be missed regardless the OUT endpoint number. Generally, this occurs when the device is connected to the host through a hub and other devices are connected to the same hub.
The affected OUT endpoint can be either control, bulk, isochronous, or an interrupt endpoint. After the OUT endpoint is primed, if an IN token to the same endpoint number on another device is received, then the OUT endpoint may be unprimed (cannot be detected by software), which causes this endpoint to no longer respond to the host OUT token, and thus, no corresponding interrupt occurs.
Quoted 7 lines
There is no good workaround for this issue, the only thing the software could do is numbering isochronous IN from the highest endpoint since we have observed most of device number endpoint from the lowest.
Cc Sergei Shtylyov <[email protected]>Cc Jun Li <[email protected]>Signed-off-by Peter Chen <[email protected]>Should this patch have Cc stable so that it gets applied for older kernels?
Quoted 12 lines
drivers/usb/chipidea/udc.c static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on) · line 1626
Quoted 10 lines
drivers/usb/chipidea/udc.c +2
-0@@ -1622,6 +1622,28 @@ static int ci_udc_pullup(struct usb_gadget *_gadget, int is_on)static int ci_udc_start (struct usb_gadget *gadget,struct usb_gadget_driver *driver);static int ci_udc_stop (struct usb_gadget *gadget); + +One blank line is enough.
Thanks, both.
Will be addressed at next version.
Peter