Re: BZ 108848: Enable SPARSE memory config on ia64

[Dave Jones <davej redhat com>]

On Wed, Jul 26, 2006 at 07:14:55AM -0400, Prarit Bhargava wrote:
 > Aron Griffis wrote:
 > >Dave Jones wrote:  [Tue Jul 25 2006, 03:37:28PM EDT]
 > >  
 > >>This diff is against the generated files.  Can you create the same
 > >>result but based on the templates in the configs/ dir in CVS ?
 > >>    
 > >
 > >This seems to be sufficient:
 > >
 > >diff -r 07fae80a782c devel/configs/config-ia64-generic
 > >--- a/devel/configs/config-ia64-generic	Tue Jul 25 20:07:47 2006 
 > >-0400
 > >+++ b/devel/configs/config-ia64-generic	Tue Jul 25 22:28:05 2006 
 > >-0400
 > >@@ -27,9 +27,9 @@ CONFIG_IA64_PAGE_SIZE_16KB=y
 > > CONFIG_IA64_PAGE_SIZE_16KB=y
 > > # CONFIG_IA64_PAGE_SIZE_64KB is not set
 > > CONFIG_IA64_L1_CACHE_SHIFT=7
 > >-CONFIG_VIRTUAL_MEM_MAP=y
 > > CONFIG_NUMA=y
 > >-CONFIG_DISCONTIGMEM=y
 > >+CONFIG_SPARSEMEM_MANUAL=y
 > >+CONFIG_SPARSEMEM=y
 > > CONFIG_IA64_MCA_RECOVERY=m
 > > CONFIG_IA64_CYCLONE=y
 > > # CONFIG_MMTIMER is not set
 > >  
 > 
 > I'll ack this -- this does seem to sufficient...

I'm glad you did, as Aron's mail is doing that funky "turns up already
marked as read" thing, so I missed his reply.

After applying the above diff, it complained about unset symbols.
CONFIG_MEMORY_HOTPLUG
CONFIG_ACPI_HOTPLUG_MEMORY

which I set.

Then it complained about

CONFIG_VIRTUAL_MEM_MAP

which the above diff removes. I added it back.
The resulting config looks sane to me, but you may want
to double check things are ok in builds 2456 and above.

		Dave

-- 
http://www.codemonkey.org.uk