Re: [RFC][PATCH 1/2] i386: consolidate discontig functions into normalones

[Mika Penttilä]

Dave Hansen wrote:

> There are quite a few functions in i386's discontig.c which are
> actually NUMA-specific, not discontigmem.  They are also very
> similar to the generic, flat functions found in setup.c.
>
> This patch takes the versions in setup.c and makes them work
> for both NUMA and non-NUMA cases.  In the process, quite a
> few nasty #ifdef and externs can be removed.
>
> One of the main mechanisms to do this is that highstart_pfn
> and highend_pfn are now gone, replaced by node_start/end_pfn[].
> However, this has no real impact on storage space, because
> those arrays are declared with a length of MAX_NUMNODES, which
> is 1 when NUMA is off.
>
>
>  
I think you allocate remap pages for nothing in the flatmem case for 
node0...those aren't used for the mem map in !NUMA.

--Mika

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/