Re: [PATCH V9 1/4] perf/core: Add PERF_SAMPLE_DATA_PAGE_SIZE

[Peter Zijlstra]

On Tue, Oct 13, 2020 at 05:46:15PM +0200, Peter Zijlstra wrote:
> Nah, that generic function, should work for 90% of all archs, it's just
> a few oddballs that need something else.
> 
> Also, if we add that hugetlb exception, we'll even get the usermap for
> those oddballs right.
> 
> I'll take this version after the merge window, I'll add __weak for the
> oddballs and also add the hugetlb userspace thing on top.

Like so.. 

---
Subject: perf,mm: Handle non-page-table-aligned hugetlbfs
From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Date: Fri, 9 Oct 2020 11:09:27 +0200

A limited nunmber of architectures support hugetlbfs sizes that do not
align with the page-tables (ARM64, Power, Sparc64). Add support for
this to the generic perf_get_page_size() implementation, and also
allow an architecture to override this implementation.

This latter is only needed when it uses non-page-table aligned huge
pages in its kernel map.

Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
---
 kernel/events/core.c |   39 +++++++++++++++++++++++++++++++++------
 1 file changed, 33 insertions(+), 6 deletions(-)

--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -6996,10 +6996,18 @@ static u64 perf_virt_to_phys(u64 virt)
 #ifdef CONFIG_MMU
 
 /*
- * Return the MMU page size of a given virtual address
+ * Return the MMU page size of a given virtual address.
+ *
+ * This generic implementation handles page-table aligned huge pages, as well
+ * as non-page-table aligned hugetlbfs compound pages.
+ *
+ * If an architecture supports and uses non-page-table aligned pages in their
+ * kernel mapping it will need to provide it's own implementation of this
+ * function.
  */
-static u64 __perf_get_page_size(struct mm_struct *mm, unsigned long addr)
+__weak u64 arch_perf_get_page_size(struct mm_struct *mm, unsigned long addr)
 {
+	struct page *page;
 	pgd_t *pgd;
 	p4d_t *p4d;
 	pud_t *pud;
@@ -7021,15 +7029,27 @@ static u64 __perf_get_page_size(struct m
 	if (!pud_present(*pud))
 		return 0;
 
-	if (pud_leaf(*pud))
+	if (pud_leaf(*pud)) {
+#ifdef pud_page
+		page = pud_page(*pud);
+		if (PageHuge(page))
+			return page_size(compound_head(page));
+#endif
 		return 1ULL << PUD_SHIFT;
+	}
 
 	pmd = pmd_offset(pud, addr);
 	if (!pmd_present(*pmd))
 		return 0;
 
-	if (pmd_leaf(*pmd))
+	if (pmd_leaf(*pmd)) {
+#ifdef pmd_page
+		page = pmd_page(*pmd);
+		if (PageHuge(page))
+			return page_size(compound_head(page));
+#endif
 		return 1ULL << PMD_SHIFT;
+	}
 
 	pte = pte_offset_map(pmd, addr);
 	if (!pte_present(*pte)) {
@@ -7037,13 +7057,20 @@ static u64 __perf_get_page_size(struct m
 		return 0;
 	}
 
+	page = pte_page(*pte);
+	if (PageHuge(page)) {
+		u64 size = page_size(compound_head(page));
+		pte_unmap(pte);
+		return size;
+	}
+
 	pte_unmap(pte);
 	return PAGE_SIZE;
 }
 
 #else
 
-static u64 __perf_get_page_size(struct mm_struct *mm, unsigned long addr)
+static u64 arch_perf_get_page_size(struct mm_struct *mm, unsigned long addr)
 {
 	return 0;
 }
@@ -7074,7 +7101,7 @@ static u64 perf_get_page_size(unsigned l
 		mm = &init_mm;
 	}
 
-	size = __perf_get_page_size(mm, addr);
+	size = arch_perf_get_page_size(mm, addr);
 
 	local_irq_restore(flags);