On Sat, Dec 6, 2025 at 6:03 PM Pratyush Yadav <[email protected]> wrote:
>
> KHO currently restricts the maximum order of a restored page to the
> maximum order supported by the buddy allocator. While this works fine
> for much of the data passed across kexec, it is possible to have pages
> larger than MAX_PAGE_ORDER.
>
> For one, it is possible to get a larger order when using
> kho_preserve_pages() if the number of pages is large enough, since it
> tries to combine multiple aligned 0-order preservations into one higher
> order preservation.
>
> For another, upcoming support for hugepages can have gigantic hugepages
> being preserved over KHO.
>
> There is no real reason for this limit. The KHO preservation machinery
> can handle any page order. Remove this artificial restriction on max
> page order.
Before removing this restriction, please also add a patch that makes
sure that the calculated order does not intermix memory from different
topologies.
I suspect we need something like this in kho_preserve_pages():
order = min(count_trailing_zeros(pfn), ilog2(end_pfn - pfn));
/* Reduce order if the block spans across NUMA nodes. */
while (order > 0) {
size_pfn = 1UL << order;
if (pfn_to_nid(pfn) == pfn_to_nid(pfn + size_pfn - 1))
break;
order--;
}
Pasha
