[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-mac68k
Subject:    [linux-mac68k] Re: linux-mac68k (fwd)
From:       Michael Schmitz <SCHMITZ () lcbvax ! cchem ! berkeley ! edu>
Date:       1998-04-12 2:23:44
[Download RAW message or body]

Hi,

>> I want to install linux on my Performa 475, which has an "upgraded" full
>> 68040 CPU. I read about the SCSI-controller not being supported yet. Does
>> that mean, I will not be able to boot from the harddrive? And if so, what
>> does "actively been worked on" mean? Is there hope for the next weeks?
>
>	"Actively be[ing] worked on" means that there are (in this case) two
>individuals scratching their heads over how to write a driver taht will
>force this scsi device into submission.  I'm one of those folks and it's
>tough going.  I'm not nearly as driver literate as I would have figured
>from my general O/S background (which is terribly frustrating!), but I'm
>also software savvy, so I'm making the best of a bad situation.  David
>Weis is the other individual.  He's bright guy and has been published in
>"Linux Journal."  However, I cannot comment on his current state (with
>respect to his progress).

To answer the 'hope for the next weeks' question first: I'd be surprised if 
it worked that fast. Even the IDE driver took a couple of weeks to figure out
completely (and there's still weirdness in the Powerbook case).

As encouragement to the driver hackers: Linux driver structure sure needs some 
getting used to - and lots of source reading and head scratching. 
The first thing I'll do for the new SCSI driver is rewriting the interrupt code 
to return something meaningful on irq_pending(), and the next thing should be 
to figure out where the interrupt from the 53C9x gets stuck. That part took 
Alan quite a while, and it was the stupid VIA1 PCR setup - which is done only 
for Macs with the 'old' SCSI driver currently, so extending this to the 53C96 
Macs might help some; look at via_init() ...

If it helps, I can patch the current 53C96 code I got from David a while back
into 2.1.90 (what kernel version did you work on, David??) and add the 
via_scsi_poll to check on the SCSI status on each timer tick. 

But maybe all of this has been tried in the meantime, and I should wait for a
new diff??

	Michael


--

[ Unsubscription requests to 'linux-mac68k-request@wave.lm.com', body of 'unsubscribe' ]

[prev in list] [next in list] [prev in thread] [next in thread]