[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Slink bootdisks on an MCA machine (PS/2 9595)




On 18 Apr 1999, Adam Di Carlo wrote:

> Jor-el <jorel@ibm.net> writes:
> 
> > 	I just tried the Slink bootdisks (the standard and the safe ones)
> > on a PS/2 model 9595 machine that I have access to. This specific model
> > has the known problem that the FPU coupling test at boot time disables the
> > halt interrupt timer, and so the boot will only work if one uses the
> > 'no-hlt' boot parameter to the kernel. This disables the check for the
> > 'hlt' instruction and lets the boot proceed. (*)
> 
> > (*)  Obviously needs to be documented somewhere in the Debian install
> > guide.
> 
> I'd love to document it.  Obviously, it would go somewhere around
> <URL:http://www.debian.org/releases/slink/i386/ch-rescue-boot.en.html>.
> 
> What is the symptom of this happening?  A hang when after the kernel says
>    Checking 'hlt' instruction...
> ?

Exactly.

> 
> > 	Ok, so I did that, but the boot hang at the point where kswapd is
> > being started. Or atleast, that is the last message that appears on the
> > console before the machine goes comatose. Has anyone gotten past this
> > point?
> 
> Of course -- I've installed the system dozens of times, personally, on
> two different architectures.  Do you really think the Debian
> boot-floppies would be that broken?
> 
	I never meant to imply that. I myself have used the Debian
bootdisks on several machines successfully. However, this is an MCA
machine and a troublesome model to boot (no pun intended). My question was
actually stating 'Has anyone gotten past this point on a model 9595?'.

	Someone who saw my post responded to me saying that if I used the
<Ctrl>-<ScrollLock> combination, I would be able to see what was the
actual process that was hanging. I tried this, and a lot of other key
combinations that I could think of, with nary a response. Whereupon, the
correspondant suggested that the kernel itself is hanging.

	What do you suggest I do? I would have filed a bug ordinarily, but
I'm pretty sure that you guys who are the ones actively building boot
disks dont have access to such a machine, and hence, might not be able to
debug it. 


Regards,
Jor-el


Reply to: