On Sun, May 27, 2007 at 10:32:58 +0200, Anders Gavare wrote:

> However, I ran in to a problem which I think is in NetBSD, not in the
> emulator. From mmeye/machdep.c, uvm_page_physload() is called. Here is a
> function call trace from the emulator:
> 
>  <uvm_page_physload(0x4c209,0xd000,0x4c209,0xd000,&_end ...
>    <panic("uvm_page_physload: start >= end" ...
> 
> The source code in mmeye/machdep.c, just before the uvm_page_physload()
> call, looks something like this:
> 
>        kernend = atop(round_page(SH3_P1SEG_TO_PHYS(end)));
>   #if NKSYMS || defined(DDB) || defined(LKM)
>        /* XXX Currently symbol table size is not passed to the kernel. */
>        kernend += 0x40000;                                     /* XXX */
>   #endif
> 
> If I manually change the first and third argument to, say, 0xc209 + 0x40,
> then the kernel gets further (and prints some boot messages).
> 
> So, should the source code perhaps read:
> 
>        kernend += atop(0x40000);                               /* XXX */
> 
> instead?

Looks like it.

I don't have any mmEye hardware and Brains Corp. hasn't been very
active in maintaining the port (e.g. this code has been there for 5
years, so obviously i was not tested).

SY, Uwe
-- 
uwe@stderr.spb.ru                       |       Zu Grunde kommen
http://snark.ptc.spbu.ru/~uwe/          |       Ist zu Grunde gehen