Re: [PATCH: 002/017]Memory hotplug for new nodes v.4.(change name old add_memory() to arch_add_memory())

From: KAMEZAWA Hiroyuki <kamezawa.hiroyu_at_jp.fujitsu.com>
Date: 2006-03-22 12:38:39
On Tue, 21 Mar 2006 17:08:18 -0800
Dave Hansen <haveblue@us.ibm.com> wrote:
> If I missed it before, please refresh my memory.  But, if we're
> providing arch_nid_probe(addr), then why don't we just call it inside of
> add_memory() on the start address, instead of in the generic code?
> 
I think just *probe* needs it. The firmware which supports memory-hotplug, ACPI, 
i386/x86_64/ia64, can tell node number as pxm.(proximity domain)

add_memory() can pass address of paddr as args, but can't pass *pxm*. 

We already maintain pxm <-> nid map. But paddr <-> nid map isn't now.
If add_memory() doesn't has nid as args, we have to maintain
(1) pxm <-> nid map.
(2) paddr <-> nid map.
Becasue pfn_to_nid() is maintained by SPARSEMEM itself now, *new* paddr<->nid map
is redundant, I think. And I think the firmware already has the map before calling
add_memory(). 

> I'm also getting a bit confused in your patches whether add_memory() is
> the _original_ add_memory(), or the new one.  It tends to get lost in 17
> patches. :(
> 
maybe a big patch :(, We (I and Goto-san) are discussing to split them to
a bit small chunks. 

> I don't really like the arch_nid_probe() name.  We need to make it very
> apparent that it is to be used _only_ for memory hotplug operations.  It
> has no meaning for anything else.
> 
> 	hotplug_physaddr_to_nid()?
> 
> Maybe with a "memory_" in front.  Maybe even
> memory_add_physaddr_to_nid()?
> 
Okay, we'll rename it.

> It was probably to keep from changing as little code as possible, but
> please convert the u64 values to pfns as soon as possible.  I noticed
> that hotadd_new_pgdat() still deals with them, and does the shift as
> well.  Is that really necessary.
> 
> The u64s should not be kept for more than one level of calls.  That
> level of calls should be the firmware.  So, let the firmware call into
> the VM code with u64s, then have all of the plain VM code deal in pfns.
> 
Okay.

-- Kame

-
To unsubscribe from this list: send the line "unsubscribe linux-ia64" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Received on Wed Mar 22 12:39:11 2006

This archive was generated by hypermail 2.1.8 : 2006-03-22 12:39:20 EST