Subject: Re: mol-0.9.40 on 603 machine
From: Ed Jaeger (ed.jaeger@bgcorp.com)
Date: Thu Feb 10 2000 - 09:30:20 MST
I have also had no success on my PowerBase 200 (603ev).
The latest version (0.9.40) patches the 2.2.14 source tree cleanly, and
the kernel compiles with no problems. The MOL "hooks" show up in
/proc/ksyms and in System.map. I'm no programmer, but I did trace the
code (I recompiled MOL from the source RPM just to be sure) and the
"device or resource busy" message appears to result from mol.o thinking
the kernel is not patched even though it is. From init.c:
int
init_module( void )
{
int i;
char *ptr;
/* if we have a patched kernel, then mol_interface is defined... */
molif = (molif_t*)get_module_symbol( "", "mol_interface" );
/* otherwise look for a runtime patched kernel... */
if( !molif ){
rtif = phys_to_virt( RUNTIME_IF_ADDR );
if( rtif->magic != RUNTIME_MAGIC || rtif->version != RUNTIME_VERSION ) {
printk("This kernel lacks MOL support!\n");
return 1;
}
molif = &rtif->molif;
}
--rest of init.c snipped--
}
The line
molif = (molif_t*)get_module_symbol( "", "mol_interface" );
should work, but doesn't seem to. I'm stuck, since my programming
abilities are now exhausted.
Keith Clayton wrote:
>
> Hi,
>
> Has anyone succeed in getting mol-0.9.40 running on a 603 based machine?
> I'm running 2.2.15pre6 rsync'd from linuxcare.au.com
>
> Patch that was included with mol-0.9.40 for 2.2.15pre4 applied cleanly to
> my kernel source and in menu config I set mac-on-linux support to Y. Just
> to be sure I wasn't a spaz, I double checked my .config just now and
> CONGIG_MOL=Y
>
-- Ed Jaeger, CFO, Bohlender Graebener Corporation ed.jaeger@bgcorp.com http://www.bgcorp.com --- "A man never discloses his own character so clearly as when he describes another's."-- Jean Paul Richter (1763-1825)
This archive was generated by hypermail 2a24 : Wed Mar 01 2000 - 00:06:20 MST