[OpenBIOS] [RFC v2] WIP fix for /memory reg property

[Andreas Färber]

Am 04.11.2010 um 01:25 schrieb Andreas Färber:

> v2:
> * Use my-#acells for address size.
> * Introduce my-#scells for size size.
> * Confine special-handling of "available" to /memory and MMU.
> ---
> Hi,
>
> How can we determine the cell size within Forth?
>
> 64bit? wouldn't work as criteria for ppc64.
> Is an explicit [IFDEF] CONFIG_{SPARC64,PPC64} the only way here?
>
> Thanks,
> Andreas
>
> arch/ppc/qemu/init.c     |   14 +++++++++-----
> forth/admin/devices.fs   |   40 ++++++++++++++++++++++++++++++++ 
> +-------
> forth/device/property.fs |   16 +++++++++++++++-
> 3 files changed, 57 insertions(+), 13 deletions(-)

Sorry, too tired... I merged the wrong patches. Above question still  
applies to the v2 "Pretty-print reg property" though!

Below is my obviously-to-be-cleaned-up fix for the ppc /memory reg  
property, which my other patch pretty-prints to properly debug such  
issues. ;)

Andreas

> diff --git a/arch/ppc/qemu/init.c b/arch/ppc/qemu/init.c
> index 08eb59c..d7ebc15 100644
> --- a/arch/ppc/qemu/init.c
> +++ b/arch/ppc/qemu/init.c
> @@ -774,15 +774,19 @@ arch_of_init( void )
>
> 	/* all memory */
>
> -	PUSH(ram_size >> 32);
> -	fword("encode-int");
> -	PUSH(ram_size & 0xffffffff);
> -	fword("encode-int");
> +	PUSH(0);
> +	fword("encode-phys");
> +#if 0
> 	fword("encode+");
> 	PUSH(0);
> 	fword("encode-int");
> 	fword("encode+");
> -	PUSH(0);
> +#endif
> +#if 0
> +	PUSH(ram_size >> 32);
> +	fword("encode-int");
> +#endif
> +	PUSH(ram_size & 0xffffffff);
> 	fword("encode-int");
> 	fword("encode+");
> 	push_str("reg");