Subject: Re: Working 1.6?
To: None <hamajima@nagoya.ydc.co.jp>
From: Michael L. Hitch <mhitch@msu.oscs.montana.edu>
List: port-amiga
Date: 06/04/2002 21:45:56
HAMAJIMA Katsuomi <hamajima@nagoya.ydc.co.jp> writes: 
> --- sys/arch/amiga/amiga/amiga_init.c.orig      Fri May 31 06:00:23 2002
> +++ sys/arch/amiga/amiga/amiga_init.c   Wed Jun  5 01:20:29 2002
> @@ -249,7 +249,6 @@
>  	RELOC(boot_cphysize, u_long) = cphysize;
>  
>  	RELOC(machineid, int) = id;
> -	RELOC(chipmem_start, vaddr_t) = cphysize;
>  	RELOC(chipmem_end, vaddr_t) = cphysize;
>  	RELOC(esym, char *) = esym_addr;
>  	RELOC(boot_flags, u_long) = flags;

  Where did that come frome (Aymeric?).  Setting chipmem_start and
chipmem_end to the same address makes no sense whatsoever.

  I'm still puzzled as to why my A2000/Zeus boots and runs.  I would think
the chip memory allocation would be messed up, but my display works normally.
From looking at things with DDB, it looks like the chip memory allocator
has allocated memory from memory mapping the Zorro II address space.  I do
note that if there isn't any Zorro II memory in the machine, it would be
using memory maps the CIA chips - random accesses to that could do bad things.

Michael
---
Michael L. Hitch                        mhitch@montana.edu
Operations Consulting,  Information Technology Center
Montana State University, Bozeman, MT     USA