| Message ID | 2bd4ac94-f7c3-41d6-27a7-352f3319bda7@ivitera.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | usb: gadget: f_uac2: EP OUT adaptive instead of async | expand |
On Fri, Feb 7, 2020 at 6:55 PM Pavel Hofman <pavel.hofman@ivitera.com> wrote: > > The existing UAC2 implementation presents its EP OUT as > USB_ENDPOINT_SYNC_ASYNC. > > However: > 1) f_uac2 does not define any feedback endpoint > > 2) IMO in reality it is adaptive - the USB host is the one which sets > the pace of data. > > Changing USB_ENDPOINT_SYNC_ASYNC to USB_ENDPOINT_SYNC_ADAPTIVE for the FS > and HS output endpoints corrects the config to reflect real functionality. That's a good idea but ADAPTIVE endpoint still requires feedback endpoint for source (USB IN) case so the host can synchronize with such endpoint (see 3.16.2.2 of UAC2 spec "For adaptive audio source endpoints and asynchronous audio sink endpoints, an explicit synchronization mechanism is needed to maintain synchronization during transfers"). > > Also, the change makes the UAC2 gadget recognized and working > in MS Windows. Does it recognizes well with both IN and OUT (e.g. capture+playback enabled) adaptive endpoints? Thanks, Ruslan > > Signed-off-by: Pavel Hofman <pavel.hofman@ivitera.com> > --- > drivers/usb/gadget/function/f_uac2.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/usb/gadget/function/f_uac2.c > b/drivers/usb/gadget/function/f_uac2.c > index db2d498..e8c9dd1 100644 > --- a/drivers/usb/gadget/function/f_uac2.c > +++ b/drivers/usb/gadget/function/f_uac2.c > @@ -273,7 +273,7 @@ enum { > .bDescriptorType = USB_DT_ENDPOINT, > > .bEndpointAddress = USB_DIR_OUT, > - .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, > + .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ADAPTIVE, > .wMaxPacketSize = cpu_to_le16(1023), > .bInterval = 1, > }; > @@ -282,7 +282,7 @@ enum { > .bLength = USB_DT_ENDPOINT_SIZE, > .bDescriptorType = USB_DT_ENDPOINT, > > - .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, > + .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ADAPTIVE, > .wMaxPacketSize = cpu_to_le16(1024), > .bInterval = 4, > }; > -- > 1.9.1
Dne 11. 02. 20 v 17:16 Ruslan Bilovol napsal(a): > On Fri, Feb 7, 2020 at 6:55 PM Pavel Hofman <pavel.hofman@ivitera.com> wrote: >> >> The existing UAC2 implementation presents its EP OUT as >> USB_ENDPOINT_SYNC_ASYNC. >> >> However: >> 1) f_uac2 does not define any feedback endpoint >> >> 2) IMO in reality it is adaptive - the USB host is the one which sets >> the pace of data. >> >> Changing USB_ENDPOINT_SYNC_ASYNC to USB_ENDPOINT_SYNC_ADAPTIVE for the FS >> and HS output endpoints corrects the config to reflect real functionality. > > That's a good idea but ADAPTIVE endpoint still requires feedback endpoint for > source (USB IN) case so the host can synchronize with such endpoint > (see 3.16.2.2 of > UAC2 spec "For adaptive audio source endpoints and asynchronous audio sink > endpoints, an explicit synchronization mechanism is needed to maintain > synchronization > during transfers"). I apologize for missing this message. Please can we resume the discussion? The tested combination is (not-changed) async IN and (changed) adaptive OUT, which is what most USB-adaptive soundcards use. Such combination does not require any feedback endpoint. > >> >> Also, the change makes the UAC2 gadget recognized and working >> in MS Windows. > > Does it recognizes well with both IN and OUT (e.g. capture+playback enabled) > adaptive endpoints? Only OUT is adaptive, IN stays async. Thanks a lot, Pavel. > > Thanks, > Ruslan > >> >> Signed-off-by: Pavel Hofman <pavel.hofman@ivitera.com> >> --- >> drivers/usb/gadget/function/f_uac2.c | 4 ++-- >> 1 file changed, 2 insertions(+), 2 deletions(-) >> >> diff --git a/drivers/usb/gadget/function/f_uac2.c >> b/drivers/usb/gadget/function/f_uac2.c >> index db2d498..e8c9dd1 100644 >> --- a/drivers/usb/gadget/function/f_uac2.c >> +++ b/drivers/usb/gadget/function/f_uac2.c >> @@ -273,7 +273,7 @@ enum { >> .bDescriptorType = USB_DT_ENDPOINT, >> >> .bEndpointAddress = USB_DIR_OUT, >> - .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, >> + .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ADAPTIVE, >> .wMaxPacketSize = cpu_to_le16(1023), >> .bInterval = 1, >> }; >> @@ -282,7 +282,7 @@ enum { >> .bLength = USB_DT_ENDPOINT_SIZE, >> .bDescriptorType = USB_DT_ENDPOINT, >> >> - .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, >> + .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ADAPTIVE, >> .wMaxPacketSize = cpu_to_le16(1024), >> .bInterval = 4, >> }; >> -- >> 1.9.1 >
diff --git a/drivers/usb/gadget/function/f_uac2.c b/drivers/usb/gadget/function/f_uac2.c index db2d498..e8c9dd1 100644 --- a/drivers/usb/gadget/function/f_uac2.c +++ b/drivers/usb/gadget/function/f_uac2.c @@ -273,7 +273,7 @@ enum { .bDescriptorType = USB_DT_ENDPOINT, .bEndpointAddress = USB_DIR_OUT, - .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, + .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ADAPTIVE, .wMaxPacketSize = cpu_to_le16(1023), .bInterval = 1, }; @@ -282,7 +282,7 @@ enum { .bLength = USB_DT_ENDPOINT_SIZE, .bDescriptorType = USB_DT_ENDPOINT, - .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, + .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ADAPTIVE, .wMaxPacketSize = cpu_to_le16(1024), .bInterval = 4, };
The existing UAC2 implementation presents its EP OUT as USB_ENDPOINT_SYNC_ASYNC. However: 1) f_uac2 does not define any feedback endpoint 2) IMO in reality it is adaptive - the USB host is the one which sets the pace of data. Changing USB_ENDPOINT_SYNC_ASYNC to USB_ENDPOINT_SYNC_ADAPTIVE for the FS and HS output endpoints corrects the config to reflect real functionality. Also, the change makes the UAC2 gadget recognized and working in MS Windows. Signed-off-by: Pavel Hofman <pavel.hofman@ivitera.com> --- drivers/usb/gadget/function/f_uac2.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)