Corrupt packets in Jnet (Was: Re: rtentry and rtrequest)
julian at elischer.org
Wed Apr 25 03:18:15 UTC 2007
Alan Garfield wrote:
> Peter Jeremy wrote:
>> Given that we are effectivly dealing with a shared memory block, how
>> does the SP now when the server has finished writing and vice versa?
>> Is jnet's handling of multiple mbufs making the SP think there are
>> multiple packets?
> D'oh! /me slaps forehead
> I wondereded what the NAK response I saw I was getting after each TX. RX
> gets an interrupt, TX gets a NAK.
> If I block sending the next packet until I receive a NAK or I timeout
> that should fix it. Silly silly boy!
I'd say you need to wait for an ACK not a NAK
>>> Your jnet_start() routine fills the tail of the buffer w/zeros
>>> already, doesn't it?
>> I would also suggest padding to 256 bytes with zeroes.
> Already does that as Yar correctly pointed out. The ADDR port is reset
> to zero, a bus_space_write_multi1 dumps into the DATA port the packet
> till there is no packet left, and a for loop fills what's left.
> freebsd-net at freebsd.org mailing list
> To unsubscribe, send any mail to "freebsd-net-unsubscribe at freebsd.org"
More information about the freebsd-net