Status

[Jürgen Gärtner]

Kernel > 2.0.36: NOT WORKING :-(

Finaly I got a 2.2.7 system working and installed the sources for the WD.
Unfortunly there are many changes in the kernel code, so nothing seems to work.
1. There is a bug in wd719x.h that impeaches compiling (parse error before
"qeuecommand") (a "," missing in the scsi_host definition). This is because I
never tested kenel-versions > 2.0.36. Maybe there are other quirks like this.
2. PCI stuff has completly changed. I did a fix, but didn't release it yet.
3. Init doesn't work anymore. I have no idea why. Is somebody any further than
me? I suspect the SMP stuff and changes to uwait() or such alike. Maybe it's
also due to egcs's "optimization"? Ihave to completely change the wait-stuff.
This is verry bad, because we can't go further without this working.
4. SMP-Interrupts ???
5. Patch: J. Teveßen did a patch against 2.2.9. I looked inside and it seems to
be ok. IMHO we don't need to patch "pci.h", as the card detection is done "by
hand" (see below). Also I think the Makefile should only configure, if
experimetal-stuff was checked in code mutualy.

Things working:
Card type detection. Yes, all WD's are reported as WD7197. This is normal
behaviour. The real detection is done with some obscure "fizzle" code, that is
still working.

CVS:
I changed to CVS. If somebody can explains how to set-up anonymous-cvs, I'm
willing to implement this on my HP-Server. 
Perhaps the most simple way is, if I put all the stuff changed so far into a new
directory called "CVS"?

Mailing-list:
The new seems to work (Thanx to Shane). Shane, where is the mailing-list
archive? If you like, I can send you my pesonal wd719x-folder to populize the
list with some messages?

Interesting:
After 2 years loosing contact to WD there are now people from WD asking for the
Linux driver!!!

Rgs. Gärti