[U-Boot] [RFC PATCH v4 0/23] Add Driver Model support to network stack

Simon Glass sjg at chromium.org
Wed Mar 4 00:40:46 CET 2015


Hi Joe,

On 3 March 2015 at 16:32, Joe Hershberger <joe.hershberger at gmail.com> wrote:
> Hi Simon,
>
>
> On Tue, Mar 3, 2015 at 5:14 PM, Simon Glass <sjg at chromium.org> wrote:
>>
>> Hi Joe,
>>
>> On 3 March 2015 at 15:29, Joe Hershberger <joe.hershberger at gmail.com>
>> wrote:
>> > On Sun, Mar 1, 2015 at 10:13 AM, Simon Glass <sjg at chromium.org> wrote:
>> >
>> > [snip]
>> >
>> >> - I think the recv() method should change before long. The
>> >> NetReceive() call should be made from the uclass since it is common to
>> >> all drivers. Then the recv() method can return a packet if it finds
>> >> one, but not submit it for processing
>> >
>> > I looked into doing this and I think it may be more of a step backward.
>> > Currently devices can directly hand their DMA buffers to the network
>> > stack.
>> > With this change, most drivers would be forced to memcpy each packet
>> > into
>> > the supplied buffer.
>>
>> I don't see why. The uclass can request the address and length of the
>> buffer.
>>
>> int (*recv)(char **buffp, int **lenp);
>
> That works... I think I'll return the length though. I don't see much value
> in having a pointer to the length.
>
> int (*recv)(uchar **packetp);

LGTM.

Regards,
Simon


More information about the U-Boot mailing list