2010/9/16 Sergei Shtylyov <sshtylyov@xxxxxxxxxx>:
> Hello.
>
> On 16-09-2010 8:46, Ming Lei wrote:
>
>>>> If so, once the dma interrupt comes, will request->actual be same
>>>> with request->length in musb_g_tx? And if it is true, could we remove
>>>> the
>>>> check for 'is_dma'?
>
>>> see that is_dma is set to true by just checking if dma in enabled in
>>> txcsr, it might be that dma didn't complete everything and you need to
>>> write txpktrdy by hand to send last short packet. So to remove that you
>>> would need to re-work a bit more code.
>
>>> You need to know when this is a dma IRQ or an endpoint IRQ.
>
>> How about the fix below(against the previous patch set) for the
>> ZLP problem pointed by Sergei?
>
>> diff --git a/drivers/usb/musb/musb_gadget.c
>> b/drivers/usb/musb/musb_gadget.c
>> index 46cf94a..a15a83c 100644
>> --- a/drivers/usb/musb/musb_gadget.c
>> +++ b/drivers/usb/musb/musb_gadget.c
>> @@ -477,40 +477,37 @@ void musb_g_tx(struct musb *musb, u8 epnum)
>> epnum, csr, musb_ep->dma->actual_len,
>> request);
>> }
>>
>> - if (is_dma || request->actual == request->length) {
>> - /*
>> - * First, maybe a terminating short packet. Some
>> DMA
>> - * engines might handle this by themselves.
>> - */
>> - if ((request->zero && request->length
>> - && request->length % musb_ep->packet_sz
>> == 0)
>> + /*
>> + * First, maybe a terminating short packet. Some DMA
>> + * engines might handle this by themselves.
>> + */
>> + if ((request->zero && (request->actual ==
>> request->length))
>
> No. You should only send ZLP only if (request->length % musb_ep->packet_sz
> == 0)
Suppose the condition of "request->zero && (request->actual ==
request->length)" is true now:
- if request->length % musb_ep->packet_sz == 0, the patch can send zlp out
without any problem
-if request->length % musb_ep->packet_sz != 0, in the dma case we may set
txpktrdy to send out short packet since the 2nd condition is triggered, and in
pio case txstate has set it already if request->actual == request->length;
So from the analysis, seems the patch can process correctly either
request->length % musb_ep->packet_sz is 0 or not.
Please help to check the above further.
thanks,
--
Lei Ming
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
