On 12.06.25 04:17, Alistair Popple wrote:
On Wed, Jun 11, 2025 at 02:06:53PM +0200, David Hildenbrand wrote:
Marking PMDs that map a "normal" refcounted folios as special is
against our rules documented for vm_normal_page().
Fortunately, there are not that many pmd_special() check that can be
mislead, and most vm_normal_page_pmd()/vm_normal_folio_pmd() users that
would get this wrong right now are rather harmless: e.g., none so far
bases decisions whether to grab a folio reference on that decision.
Well, and GUP-fast will fallback to GUP-slow. All in all, so far no big
implications as it seems.
Getting this right will get more important as we use
folio_normal_page_pmd() in more places.
Fix it by teaching insert_pfn_pmd() to properly handle folios and
pfns -- moving refcount/mapcount/etc handling in there, renaming it to
insert_pmd(), and distinguishing between both cases using a new simple
"struct folio_or_pfn" structure.
Use folio_mk_pmd() to create a pmd for a folio cleanly.
Fixes: 6c88f72691f8 ("mm/huge_memory: add vmf_insert_folio_pmd()")
Signed-off-by: David Hildenbrand <da...@redhat.com>
---
mm/huge_memory.c | 58 ++++++++++++++++++++++++++++++++----------------
1 file changed, 39 insertions(+), 19 deletions(-)
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 49b98082c5401..7e3e9028873e5 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1372,9 +1372,17 @@ vm_fault_t do_huge_pmd_anonymous_page(struct vm_fault
*vmf)
return __do_huge_pmd_anonymous_page(vmf);
}
-static int insert_pfn_pmd(struct vm_area_struct *vma, unsigned long addr,
- pmd_t *pmd, pfn_t pfn, pgprot_t prot, bool write,
- pgtable_t pgtable)
+struct folio_or_pfn {
+ union {
+ struct folio *folio;
+ pfn_t pfn;
+ };
+ bool is_folio;
+};
I know it's simple, but I'm still not a fan particularly as these types of
patterns tend to proliferate once introduced. See below for a suggestion.
It's much better than abusing pfn_t for folios -- and I don't
particularly see a problem with this pattern here as long as it stays in
this file.
+static int insert_pmd(struct vm_area_struct *vma, unsigned long addr,
+ pmd_t *pmd, struct folio_or_pfn fop, pgprot_t prot,
+ bool write, pgtable_t pgtable)
{
struct mm_struct *mm = vma->vm_mm;
pmd_t entry;
@@ -1382,8 +1390,11 @@ static int insert_pfn_pmd(struct vm_area_struct *vma,
unsigned long addr,
lockdep_assert_held(pmd_lockptr(mm, pmd));
if (!pmd_none(*pmd)) {
+ const unsigned long pfn = fop.is_folio ? folio_pfn(fop.folio) :
+ pfn_t_to_pfn(fop.pfn);
+
if (write) {
- if (pmd_pfn(*pmd) != pfn_t_to_pfn(pfn)) {
+ if (pmd_pfn(*pmd) != pfn) {
WARN_ON_ONCE(!is_huge_zero_pmd(*pmd));
return -EEXIST;
}
@@ -1396,11 +1407,19 @@ static int insert_pfn_pmd(struct vm_area_struct *vma,
unsigned long addr,
return -EEXIST;
}
- entry = pmd_mkhuge(pfn_t_pmd(pfn, prot));
- if (pfn_t_devmap(pfn))
- entry = pmd_mkdevmap(entry);
- else
- entry = pmd_mkspecial(entry);
+ if (fop.is_folio) {
+ entry = folio_mk_pmd(fop.folio, vma->vm_page_prot);
+
+ folio_get(fop.folio);
+ folio_add_file_rmap_pmd(fop.folio, &fop.folio->page, vma);
+ add_mm_counter(mm, mm_counter_file(fop.folio), HPAGE_PMD_NR);
+ } else {
+ entry = pmd_mkhuge(pfn_t_pmd(fop.pfn, prot));
+ if (pfn_t_devmap(fop.pfn))
+ entry = pmd_mkdevmap(entry);
+ else
+ entry = pmd_mkspecial(entry);
+ }
Could we change insert_pfn_pmd() to insert_pmd_entry() and have callers call
something like pfn_to_pmd_entry() or folio_to_pmd_entry() to create the pmd_t
entry as appropriate, which is then passed to insert_pmd_entry() to do the bits
common to both?
Yeah, I had that idea as well but discarded it, because the
refcounting+mapcounting handling is better placed where we are actually
inserting the pmd (not possibly only upgrading permissions of an
existing mapping). Avoid 4-line comments as the one we are removing in
patch #3 ...
--
Cheers,
David / dhildenb
