7193 and driver

[Johnny Teveßen]

Quoting Juergen Gaertner (gaertner at mbox.si.uni-hannover.de):

> > I've got a WD 7193RTL PCI card and Linux-2.2.9. I downloaded the
> what's that?     ---^^^

I don't know. That's what the cardbox says:

"MODEL: 7193RTL" on the front, "Model Number: 7193RTL" on the side.

On the front: "PCI-SCSI ADAPTER KIT  WESTERN DIGITAL"
"western digital guaranteed compatibility", "designed for
windows95", "Free inside Corel SCSI 2", and so on.

On the back, under "Specifications", it says: "Controller:
WD7193", and I wasn't able to find any other references to
the "RTL" postfix.

> Perhaps you can send me the patch? What do you changed in the driver?

I'll checkout the kernel patch and send it to the list. For the
driver patch, I'll first have to remove some helpful hacks I did.

> > My problem now is that I get a "WCS checksum error" because the
> > card returns a 0x00 data status ("no interrupt pending") instead
> > of a 0x01 ("ok") after the ENABLE_ADVANCE_MODE command. And the
> > Linux interrupt counter in /proc also says that IRQ 9 has never
> > been triggered (maybe I'm confusing something with internal
> > Spider interrupts here?).
> 
> that's bad. Think it's the new SMP-Interrupt sheme. 

Quite possible. I should add that I'm running a UP kernel.

> > Jun  8 18:09:04 trib0026 kernel: wd719x: init: RISC-Type: 1; fw_release: 3e!
> 
> did you tried FW 3a? (you can change this in the source code

Not yet, since the driver code suggests that 3a firmware is only
necessary if a 7296 is detected. I'll have to change that and
then try the 3a code.

> strange... the jiffies are not increasing... normaly I have some delay when
> loading the risc. This may be a timming problem. Oh no, not the good old f***ing
> problem again...

I noticed that. And since I never before had to fiddle with jiffies,
I was puzzled about how long a jiffie is on 2.2.x and x86. Did a
web search and asked in 5 Linux channels. Nobody knew or wanted to
say, so I did some timing and found out that a jiffie is 1/100 sec.

I think this is the changed interrupt behaviour (no sei/cli anymore),
and that jiffies is not updated while interrupts are turned off.

> The newest code (as of 4th April 1998) can be found in
>        
> http://schering.si2.uni-hannover.de/stuff/gaertner/WD719X/Linux/4.4.1998/
> or
> 	http://sun1.rrzn.uni-hannover.de/gaertner.juergen/wd719x/Linux/4.4.1998/
> 
> this was my last version.

That's what I started with. Somebody should update the timestamps in
the files.

> Maybe you like to see the docu about the RISC at
> 
> http://sun1.rrzn.uni-hannover.de/gaertner.juergen/wd719x/Linux/Docu/Spider/

A friend of a friend already downloaded that. ;-)
It's great but a bunch of pages to read.
-- 
J. Teveßen, Germany
PGPkey: http://pgp.lsi.upc.es:11371/pks/lookup?op=get&search=0x5C77B04D