[PATCH] powerpc/64s/radix: Don't prefetch DAR in update_mmu_cache

[PATCH] powerpc/64s/radix: Don't prefetch DAR in update_mmu_cache

[Nicholas Piggin]

The idea behind this prefetch was to kick off a page table walk before
returning from the fault, getting some pipelining advantage.

But this never showed up any noticable performance advantage, and in
fact with KUAP the prefetches are actually blocked and cause some
kind of micro-architectural fault. Removing this improves page fault
microbenchmark performance by about 9%.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/book3s/64/pgtable.h |  7 +++++--
 arch/powerpc/mm/book3s64/hash_utils.c        |  5 -----
 arch/powerpc/mm/book3s64/pgtable.c           | 13 -------------
 3 files changed, 5 insertions(+), 20 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/pgtable.h b/arch/powerpc/include/asm/book3s/64/pgtable.h
index 368b136517e0..59ed15e43e89 100644
--- a/arch/powerpc/include/asm/book3s/64/pgtable.h
+++ b/arch/powerpc/include/asm/book3s/64/pgtable.h
@@ -1139,8 +1139,11 @@ extern pmd_t mk_pmd(struct page *page, pgprot_t pgprot);
 extern pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot);
 extern void set_pmd_at(struct mm_struct *mm, unsigned long addr,
 		       pmd_t *pmdp, pmd_t pmd);
-extern void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long addr,
-				 pmd_t *pmd);
+static inline void update_mmu_cache_pmd(struct vm_area_struct *vma,
+					unsigned long addr, pmd_t *pmd)
+{
+}
+
 extern int hash__has_transparent_hugepage(void);
 static inline int has_transparent_hugepage(void)
 {
diff --git a/arch/powerpc/mm/book3s64/hash_utils.c b/arch/powerpc/mm/book3s64/hash_utils.c
index 845da1e8ca4f..2458615805ee 100644
--- a/arch/powerpc/mm/book3s64/hash_utils.c
+++ b/arch/powerpc/mm/book3s64/hash_utils.c
@@ -1672,11 +1672,6 @@ void update_mmu_cache(struct vm_area_struct *vma, unsigned long address,
 	unsigned long trap;
 	bool is_exec;
 
-	if (radix_enabled()) {
-		prefetch((void *)address);
-		return;
-	}
-
 	/* We only want HPTEs for linux PTEs that have _PAGE_ACCESSED set */
 	if (!pte_young(*ptep) || address >= TASK_SIZE)
 		return;
diff --git a/arch/powerpc/mm/book3s64/pgtable.c b/arch/powerpc/mm/book3s64/pgtable.c
index e0bb69c616e4..821b483a5ac3 100644
--- a/arch/powerpc/mm/book3s64/pgtable.c
+++ b/arch/powerpc/mm/book3s64/pgtable.c
@@ -146,19 +146,6 @@ pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot)
 	pmdv &= _HPAGE_CHG_MASK;
 	return pmd_set_protbits(__pmd(pmdv), newprot);
 }
-
-/*
- * This is called at the end of handling a user page fault, when the
- * fault has been handled by updating a HUGE PMD entry in the linux page tables.
- * We use it to preload an HPTE into the hash table corresponding to
- * the updated linux HUGE PMD entry.
- */
-void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long addr,
-			  pmd_t *pmd)
-{
-	if (radix_enabled())
-		prefetch((void *)addr);
-}
 #endif /* CONFIG_TRANSPARENT_HUGEPAGE */
 
 /* For use by kexec */
-- 
2.23.0

Re: [PATCH] powerpc/64s/radix: Don't prefetch DAR in update_mmu_cache

[Michael Ellerman]

On Mon, 4 May 2020 22:29:07 +1000, Nicholas Piggin wrote:
> The idea behind this prefetch was to kick off a page table walk before
> returning from the fault, getting some pipelining advantage.
> 
> But this never showed up any noticable performance advantage, and in
> fact with KUAP the prefetches are actually blocked and cause some
> kind of micro-architectural fault. Removing this improves page fault
> microbenchmark performance by about 9%.

Applied to powerpc/next.

[1/1] powerpc/64s/radix: Don't prefetch DAR in update_mmu_cache
      https://git.kernel.org/powerpc/c/18594f9b8c45484bd527ebc6b08383b95f58ba73

cheers