On Tuesday 13 April 2010 15:24:17 Michael Ellerman wrote:

On Tue, 2010-04-13 at 14:16 +1000, Mark Nelson wrote:

quoted

We need to keep track of the backing pages that get allocated by
vmemmap_populate() so that when we use kdump, the dump-capture kernel knows
where these pages are.

We use a linked list of structures that contain the physical address of the
backing page and corresponding virtual address to track the backing pages.
We can use an hlist to save space, because we never remove nodes.

What if we remove_memory() ? Or does that not do what I think it does?

That's a good question, and one that I probably should have added to the
commit message.
But, following through, it looks like we end up calling into
__remove_section() from mm/memory_hotplug.c and if
CONFIG_SPARSEMEM_VMEMMAP is enabled we just return EBUSY as freeing
memmap with vmemmap isn't implemented yet.

So for the moment, I'm not sure we have to worry about it.

Thanks!
Mark.