Christophe Leroy [off-list ref] writes:

Le 10/07/2023 à 18:08, Aneesh Kumar K.V a écrit :

quoted

Architectures like powerpc will like to use different page table allocators
and mapping mechanisms to implement vmemmap optimization. Similar to
vmemmap_populate allow architectures to implement
vmemap_populate_compound_pages

Signed-off-by: Aneesh Kumar K.V <redacted>
---
  mm/sparse-vmemmap.c | 3 +++
  1 file changed, 3 insertions(+)

diff --git a/mm/sparse-vmemmap.c b/mm/sparse-vmemmap.c
index a044a130405b..541b3f69a481 100644
--- a/mm/sparse-vmemmap.c
+++ b/mm/sparse-vmemmap.c

@@ -141,6 +141,7 @@ void __meminit vmemmap_verify(pte_t *pte, int node,
  			start, end - 1);
  }
  
+#ifndef vmemmap_populate_compound_pages
  pte_t * __meminit vmemmap_pte_populate(pmd_t *pmd, unsigned long addr, int node,
  				       struct vmem_altmap *altmap,
  				       struct page *reuse)

Should vmemmap_pte_populate() be static ?

It looks odd to exclude a non-static function based on a non related macro.

There are several such function in the block being excluded here. Can 
you explain why it is correct to do that ?

Those functions can actually be made static. But I will do that as a
part of different patch. I will update this patch and make sure the
 #ifdef will only override the vmemmap_populate_compound_pages.

modified   mm/sparse-vmemmap.c

@@ -141,7 +141,6 @@ void __meminit vmemmap_verify(pte_t *pte, int node,

			start, end - 1);
 }

-#ifndef vmemmap_populate_compound_pages
 pte_t * __meminit vmemmap_pte_populate(pmd_t *pmd, unsigned long addr, int node,
				       struct vmem_altmap *altmap,
				       struct page *reuse)

@@ -359,6 +358,7 @@ int __meminit vmemmap_populate_hugepages(unsigned long start, unsigned long end,

	return 0;
 }

+#ifndef vmemmap_populate_compound_pages
 /*
  * For compound pages bigger than section size (e.g. x86 1G compound
  * pages with 2M subsection size) fill the rest of sections as tail

-aneesh