> > > That's right. usb_ep_alloc_buffer doesn't really use the endpoint
> > > argument, ....
> >
> > That's not true. Controller drivers are certainly allowed to use
> > that argument ... it's provided so that they _can_ use it. Some do.
>
> I thought that might provoke you into commenting... :-)
>
> ISTR we discussed this very issue back when g_file_storage was being
> developed. I can't recall the actual issue or what we decided. It may
> had something to do with a desire to avoid allocating separate buffers for
> bulk-in and bulk-out.
>
> Do you think the code in file_storage.c needs to be changed?
It deserves a comment that the current code is nonportable.
The only hardware with a current UDC driver that would seem to be
problematic is "goku_udc", since that's got a broken DMA design;
only one direction (TX/IN) really has usable DMA support(*).
And that chip doesn't seem to all that widely used.
- Dave
(*) For some reason, peripheral RX DMA -- "OUT" -- seems to be
something that silicon designers have a really hard time
with. Even on peripheral controllers that do have usable
RXDMA support, it almost always has major limitations
compared to TX/IN transfers.
-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
