Re: [PATCH] #2 VIA Rhine stalls: TxAbort handling

[Ivan G.]

> [1] "aborted due to excessive collisions" according to the doc, but it also
>     mentions that for "excessive collisions", bit 13 would have to be set.
>     It isn't.
>     (I have seen it on my VT6102, though, with interrupt status of 0x2008
>     for instance)

bit 13 = IntrTxAborted
I don't see it below.

/* Enable interrupts by setting the interrupt mask. */
writew(IntrRxDone | IntrRxErr | IntrRxEmpty| IntrRxOverflow| IntrRxDropped|
   IntrTxDone | IntrTxAbort | IntrTxUnderrun | IntrPCIErr | IntrStatsMax | 
IntrLinkChange | IntrMIIChange, ioaddr + IntrEnable);

Interrupts referenced in the driver and not listed here are: IntrRxNoBuf,
IntrRxWakeUp, IntrTxAborted

Interrupts included here but not used in the driver are:
IntrRxOverflow, IntrRxDropped.

I've known about this for a while but I wasn't sure how to fix everything...
I wasn't sure if every interrupt was handled correctly.
For example what exactly is the difference between IntrTxAbort and 
IntrTxAborted. 

I was also puzzled as to why the docs say:
Transmit Descriptor Underflow for IntrMIIChange

I am talking about the newest VT86C100A docs.
I believe Urban Widmark had a patch that redefined that interrupt but it was 
never included.