realtek docu, ASUS 486 MB

[Donald Becker]

On Fri, 24 Sep 1999, Hermann Lauer wrote:

> I have bought two rtl8139 Ethernetkontrollers,
> the one in my Pentium works fine, but the one in the
> 486 hang's the machine.

This isn't uncommon for newer PCI cards.

> I have read some Comments about broken 486/PCI implementations
> (I have a Intel Saturn chipset (rev.2), Motherboard ASUS SP3I/PCI or

I have a SP3G ('486 w/PCI) that I test with.  There are a number of cards
that Just Won't Work in the machine.  The Tulip driver used to have a
special work-around to drop the burst size conditional on a '486, but I
eventually removed it.

> something like that) and I'm hacking on the rtl8129.c at the moment
> to at least not crash the machine - at the moment at least TX don't
> crash the machine any more (always hard crashes, IMHO pci deadlocks).

Hmmm, it's usually corrupted data.  That's bad but not fatal for IP with
Ethernet, but can really suck for a disk.

> http://www.realtek.com.tw
> is gone (maybe due to the earthquake ?).

It's the earthquake.  There isn't much damage in the industrial park where
Realtek is located, but the only power is from generators.

Where do I get this info from?   I had set up meetings this week for David
Lesher with various Taiwanese companies (ADMtek, Lite-On, Realtek, SiS, VIA,
Winbond) to discuss Linux driver issues.  The earthquake hit just after he
landed :-<, so none of the meetings happened.  I just got email from him --
he left Taiwan early and is now in Australia, 

> Can all the transfer defines be safely set to the lowest possible value
> (i.e. 0) (RX_FIFO_THRESH RX_DMA_BURST TX_DMA_BURST and 
> so on) and the driver will still work ?

Yes, I believe so.  The *DMA_BURST numbers are important ones to try.  It's
trivial to make them module paramters so that you can try them more easily.

> Is there any docu available on the net about programming the MB Chipset ?

I don't know.

Donald Becker					  becker@cesdis.gsfc.nasa.gov
USRA-CESDIS, Center of Excellence in Space Data and Information Sciences.
Code 930.5, Goddard Space Flight Center,  Greenbelt, MD.  20771
301-286-0882	     http://cesdis.gsfc.nasa.gov/people/becker/whoiam.html

 | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the
 |  body of the mail, include only the text:
 |   unsubscribe this-list-name youraddress@wherever.org
 | You will be unsubscribed as speedily as possible.