On Fri, 9 Jun 2000, Alan Cox wrote:
> Just a random pondering - the only other way I can see short frames occuring
> is if the card is managing to receive a frame into a buffer that is still
> being set up again after use ?
I don't see how ... :(
The descriptors are protected by a ownership flag, when 1 the driver owns
it, when 0 the driver owns it. If you don't own it, you don't touch it. I
could be wrong, my english isn't good enough to decipher this:
"Owner : This bit control by driver, 1 to identify this descriptor own
by VT86C100A controller, 0 means this descriptor be a free
descriptor; Driver must set this bit be zero when initialed."
The driver only sets it to 1 after setting it up (allocating a fresh skb
and such). The card (I assume) sets it to 0 after having received into it
and then raises the interrupt. The driver only looks at the ones with
There is someone else that reported problems with IP masquerading
(possibly with packet filter?) and the via-rhine (different error message
and not related to sniffing). Here it was a packet-filtering firewall
(with ip-masq?) ...
IP masq is working fine here, I have flood pings, ftp downloads, tcpdump
(on the internal via interface).
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to firstname.lastname@example.org
This archive was generated by hypermail 2b29 : Thu Jun 15 2000 - 21:00:40 EST