[PATCH v3 0/3] Add update_mmu_tlb_range() to simplify code

[PATCH v3 0/3] Add update_mmu_tlb_range() to simplify code

[Bang Li]

This series of commits mainly adds the update_mmu_tlb_range() to
batch update tlb in an address range and refactor the update_mmu_tlb().

After the commit 19eaf44954df ("mm: thp: support allocation of anonymous
multi-size THP"), We may need to batch update tlb of a certain address
range by calling update_mmu_tlb() in a loop. Using the
update_mmu_tlb_range(), we can simplify the code and possibly reduce the
execution of some unnecessary code in some architectures.

Thanks,
Bang

Changes since v2 [2]:
 - Use generic architecture to implement update_mmu_tlb() (per Ryan)

Changes since v1 [1]:
 - Add __HAVE_ARCH_UPDATE_MMU_TLB_RANGE macro (per Lance Yang)

[1] https://lore.kernel.org/linux-mm/20240429103346.59115-1-libang.li@antgroup.com/
[2] https://lore.kernel.org/linux-mm/20240506155120.83105-1-libang.li@antgroup.com/

Bang Li (3):
  mm: Add update_mmu_tlb_range()
  mm: Refactor update_mmu_tlb()
  mm: Use update_mmu_tlb_range() to simplify code

 arch/loongarch/include/asm/pgtable.h |  4 ++--
 arch/mips/include/asm/pgtable.h      |  4 ++--
 arch/riscv/include/asm/pgtable.h     |  4 ++--
 arch/xtensa/include/asm/pgtable.h    |  6 +++---
 arch/xtensa/mm/tlb.c                 |  6 +++---
 include/linux/pgtable.h              | 11 ++++++++---
 mm/memory.c                          |  4 +---
 7 files changed, 21 insertions(+), 18 deletions(-)

-- 
2.19.1.6.gb485710b

[PATCH v3 1/3] mm: Add update_mmu_tlb_range()

[Bang Li]

Added update_mmu_tlb_range(), we can batch update tlb of an
address range.

Signed-off-by: Bang Li <libang.li@antgroup.com>
---
 arch/loongarch/include/asm/pgtable.h | 2 ++
 arch/mips/include/asm/pgtable.h      | 2 ++
 arch/riscv/include/asm/pgtable.h     | 2 ++
 arch/xtensa/include/asm/pgtable.h    | 3 +++
 arch/xtensa/mm/tlb.c                 | 6 ++++++
 5 files changed, 15 insertions(+)

diff --git a/arch/loongarch/include/asm/pgtable.h b/arch/loongarch/include/asm/pgtable.h
index af3acdf3481a..5ccc2a3a6f7a 100644
--- a/arch/loongarch/include/asm/pgtable.h
+++ b/arch/loongarch/include/asm/pgtable.h
@@ -469,6 +469,8 @@ static inline void update_mmu_cache_range(struct vm_fault *vmf,
 
 #define __HAVE_ARCH_UPDATE_MMU_TLB
 #define update_mmu_tlb	update_mmu_cache
+#define update_mmu_tlb_range(vma, addr, ptep, nr) \
+	update_mmu_cache_range(NULL, vma, addr, ptep, nr)
 
 static inline void update_mmu_cache_pmd(struct vm_area_struct *vma,
 			unsigned long address, pmd_t *pmdp)
diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h
index e27a4c83c548..0891ad7d43b6 100644
--- a/arch/mips/include/asm/pgtable.h
+++ b/arch/mips/include/asm/pgtable.h
@@ -596,6 +596,8 @@ static inline void update_mmu_cache_range(struct vm_fault *vmf,
 
 #define	__HAVE_ARCH_UPDATE_MMU_TLB
 #define update_mmu_tlb	update_mmu_cache
+#define update_mmu_tlb_range(vma, address, ptep, nr) \
+	update_mmu_cache_range(NULL, vma, address, ptep, nr)
 
 static inline void update_mmu_cache_pmd(struct vm_area_struct *vma,
 	unsigned long address, pmd_t *pmdp)
diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h
index 661b2b4fe758..fc07b829ac4a 100644
--- a/arch/riscv/include/asm/pgtable.h
+++ b/arch/riscv/include/asm/pgtable.h
@@ -488,6 +488,8 @@ static inline void update_mmu_cache_range(struct vm_fault *vmf,
 
 #define __HAVE_ARCH_UPDATE_MMU_TLB
 #define update_mmu_tlb update_mmu_cache
+#define update_mmu_tlb_range(vma, addr, ptep, nr) \
+	update_mmu_cache_range(NULL, vma, addr, ptep, nr)
 
 static inline void update_mmu_cache_pmd(struct vm_area_struct *vma,
 		unsigned long address, pmd_t *pmdp)
diff --git a/arch/xtensa/include/asm/pgtable.h b/arch/xtensa/include/asm/pgtable.h
index 9a7e5e57ee9a..436158bd9030 100644
--- a/arch/xtensa/include/asm/pgtable.h
+++ b/arch/xtensa/include/asm/pgtable.h
@@ -413,6 +413,9 @@ typedef pte_t *pte_addr_t;
 void update_mmu_tlb(struct vm_area_struct *vma,
 		    unsigned long address, pte_t *ptep);
 #define __HAVE_ARCH_UPDATE_MMU_TLB
+void update_mmu_tlb_range(struct vm_area_struct *vma,
+		unsigned long address, pte_t *ptep, unsigned int nr);
+#define update_mmu_tlb_range update_mmu_tlb_range
 
 #endif /* !defined (__ASSEMBLY__) */
 
diff --git a/arch/xtensa/mm/tlb.c b/arch/xtensa/mm/tlb.c
index d8b60d6e50a8..05efba86b870 100644
--- a/arch/xtensa/mm/tlb.c
+++ b/arch/xtensa/mm/tlb.c
@@ -169,6 +169,12 @@ void update_mmu_tlb(struct vm_area_struct *vma,
 	local_flush_tlb_page(vma, address);
 }
 
+void update_mmu_tlb_range(struct vm_area_struct *vma,
+			unsigned long address, pte_t *ptep, unsigned int nr)
+{
+	local_flush_tlb_range(vma, address, address + PAGE_SIZE * nr);
+}
+
 #ifdef CONFIG_DEBUG_TLB_SANITY
 
 static unsigned get_pte_for_vaddr(unsigned vaddr)
-- 
2.19.1.6.gb485710b

Re: [PATCH v3 1/3] mm: Add update_mmu_tlb_range()

[Bang Li]

Hi David,

Thanks for you review!

On 2024/5/21 17:32, David Hildenbrand wrote:
> On 18.05.24 09:49, Bang Li wrote:
>> Added update_mmu_tlb_range(), we can batch update tlb of an
>> address range.
>>
>> Signed-off-by: Bang Li <libang.li@antgroup.com>
>> ---
> 
> The generic variant that is a NOP from patch #2 should go into this 
> patch as well.

Good point. Thanks for pointing this out.

Thanks,
Bang

> 
> Apart from that LGTM
> 

[PATCH v3 2/3] mm: Refactor update_mmu_tlb()

[Bang Li]

Remove update_mmu_tlb() from those architectures and define
generically via update_mmu_tlb_range(), removing the ability
for arches to override it.

Signed-off-by: Bang Li <libang.li@antgroup.com>
---
 arch/loongarch/include/asm/pgtable.h |  2 --
 arch/mips/include/asm/pgtable.h      |  2 --
 arch/riscv/include/asm/pgtable.h     |  2 --
 arch/xtensa/include/asm/pgtable.h    |  3 ---
 arch/xtensa/mm/tlb.c                 |  6 ------
 include/linux/pgtable.h              | 11 ++++++++---
 6 files changed, 8 insertions(+), 18 deletions(-)

diff --git a/arch/loongarch/include/asm/pgtable.h b/arch/loongarch/include/asm/pgtable.h
index 5ccc2a3a6f7a..161dd6e10479 100644
--- a/arch/loongarch/include/asm/pgtable.h
+++ b/arch/loongarch/include/asm/pgtable.h
@@ -467,8 +467,6 @@ static inline void update_mmu_cache_range(struct vm_fault *vmf,
 #define update_mmu_cache(vma, addr, ptep) \
 	update_mmu_cache_range(NULL, vma, addr, ptep, 1)
 
-#define __HAVE_ARCH_UPDATE_MMU_TLB
-#define update_mmu_tlb	update_mmu_cache
 #define update_mmu_tlb_range(vma, addr, ptep, nr) \
 	update_mmu_cache_range(NULL, vma, addr, ptep, nr)
 
diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h
index 0891ad7d43b6..c29a551eb0ca 100644
--- a/arch/mips/include/asm/pgtable.h
+++ b/arch/mips/include/asm/pgtable.h
@@ -594,8 +594,6 @@ static inline void update_mmu_cache_range(struct vm_fault *vmf,
 #define update_mmu_cache(vma, address, ptep) \
 	update_mmu_cache_range(NULL, vma, address, ptep, 1)
 
-#define	__HAVE_ARCH_UPDATE_MMU_TLB
-#define update_mmu_tlb	update_mmu_cache
 #define update_mmu_tlb_range(vma, address, ptep, nr) \
 	update_mmu_cache_range(NULL, vma, address, ptep, nr)
 
diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h
index fc07b829ac4a..158170442d2f 100644
--- a/arch/riscv/include/asm/pgtable.h
+++ b/arch/riscv/include/asm/pgtable.h
@@ -486,8 +486,6 @@ static inline void update_mmu_cache_range(struct vm_fault *vmf,
 #define update_mmu_cache(vma, addr, ptep) \
 	update_mmu_cache_range(NULL, vma, addr, ptep, 1)
 
-#define __HAVE_ARCH_UPDATE_MMU_TLB
-#define update_mmu_tlb update_mmu_cache
 #define update_mmu_tlb_range(vma, addr, ptep, nr) \
 	update_mmu_cache_range(NULL, vma, addr, ptep, nr)
 
diff --git a/arch/xtensa/include/asm/pgtable.h b/arch/xtensa/include/asm/pgtable.h
index 436158bd9030..1647a7cc3fbf 100644
--- a/arch/xtensa/include/asm/pgtable.h
+++ b/arch/xtensa/include/asm/pgtable.h
@@ -410,9 +410,6 @@ void update_mmu_cache_range(struct vm_fault *vmf, struct vm_area_struct *vma,
 
 typedef pte_t *pte_addr_t;
 
-void update_mmu_tlb(struct vm_area_struct *vma,
-		    unsigned long address, pte_t *ptep);
-#define __HAVE_ARCH_UPDATE_MMU_TLB
 void update_mmu_tlb_range(struct vm_area_struct *vma,
 		unsigned long address, pte_t *ptep, unsigned int nr);
 #define update_mmu_tlb_range update_mmu_tlb_range
diff --git a/arch/xtensa/mm/tlb.c b/arch/xtensa/mm/tlb.c
index 05efba86b870..0a1a815dc796 100644
--- a/arch/xtensa/mm/tlb.c
+++ b/arch/xtensa/mm/tlb.c
@@ -163,12 +163,6 @@ void local_flush_tlb_kernel_range(unsigned long start, unsigned long end)
 	}
 }
 
-void update_mmu_tlb(struct vm_area_struct *vma,
-		    unsigned long address, pte_t *ptep)
-{
-	local_flush_tlb_page(vma, address);
-}
-
 void update_mmu_tlb_range(struct vm_area_struct *vma,
 			unsigned long address, pte_t *ptep, unsigned int nr)
 {
diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
index 18019f037bae..117b807e3f89 100644
--- a/include/linux/pgtable.h
+++ b/include/linux/pgtable.h
@@ -729,13 +729,18 @@ static inline void clear_full_ptes(struct mm_struct *mm, unsigned long addr,
  * fault. This function updates TLB only, do nothing with cache or others.
  * It is the difference with function update_mmu_cache.
  */
-#ifndef __HAVE_ARCH_UPDATE_MMU_TLB
+#ifndef update_mmu_tlb_range
+static inline void update_mmu_tlb_range(struct vm_area_struct *vma,
+				unsigned long address, pte_t *ptep, unsigned int nr)
+{
+}
+#endif
+
 static inline void update_mmu_tlb(struct vm_area_struct *vma,
 				unsigned long address, pte_t *ptep)
 {
+	update_mmu_tlb_range(vma, address, ptep, 1);
 }
-#define __HAVE_ARCH_UPDATE_MMU_TLB
-#endif
 
 /*
  * Some architectures may be able to avoid expensive synchronization
-- 
2.19.1.6.gb485710b

Re: [PATCH v3 2/3] mm: Refactor update_mmu_tlb()

[David Hildenbrand]

On 18.05.24 09:49, Bang Li wrote:
> Remove update_mmu_tlb() from those architectures and define
> generically via update_mmu_tlb_range(), removing the ability
> for arches to override it.

I'd suggest something like

"mm: implement update_mmu_tlb() using update_mmu_tlb_range()

Let's make update_mmu_tlb() simply a generic wrapper around 
update_mmu_tlb_range(). Only the latter can now be overridden by the 
architecture. We can now remove __HAVE_ARCH_UPDATE_MMU_TLB as well.
"

[...]

> +#ifndef update_mmu_tlb_range
> +static inline void update_mmu_tlb_range(struct vm_area_struct *vma,
> +				unsigned long address, pte_t *ptep, unsigned int nr)
> +{
> +}
> +#endif

With that in patch #1

Acked-by: David Hildenbrand <david@redhat.com>

-- 
Cheers,

David / dhildenb

Re: [PATCH v3 2/3] mm: Refactor update_mmu_tlb()

[Bang Li]

Hi David,

Thanks for you review!

On 2024/5/21 17:36, David Hildenbrand wrote:
> On 18.05.24 09:49, Bang Li wrote:
>> Remove update_mmu_tlb() from those architectures and define
>> generically via update_mmu_tlb_range(), removing the ability
>> for arches to override it.
> 
> I'd suggest something like
> 
> "mm: implement update_mmu_tlb() using update_mmu_tlb_range()
> 
> Let's make update_mmu_tlb() simply a generic wrapper around 
> update_mmu_tlb_range(). Only the latter can now be overridden by the 
> architecture. We can now remove __HAVE_ARCH_UPDATE_MMU_TLB as well.
> "

Agree! Thank you for your suggestion, I will modify it in the next version

> 
> [...]
> 
>> +#ifndef update_mmu_tlb_range
>> +static inline void update_mmu_tlb_range(struct vm_area_struct *vma,
>> +                unsigned long address, pte_t *ptep, unsigned int nr)
>> +{
>> +}
>> +#endif
> 
> With that in patch #1

Thanks again.
Bang

> 
> Acked-by: David Hildenbrand <david@redhat.com>
> 

[PATCH v3 3/3] mm: Use update_mmu_tlb_range() to simplify code

[Bang Li]

Let us simplify the code by update_mmu_tlb_range().

Signed-off-by: Bang Li <libang.li@antgroup.com>
---
 mm/memory.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index eea6e4984eae..2d53e29cf76e 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -4421,7 +4421,6 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf)
 	vm_fault_t ret = 0;
 	int nr_pages = 1;
 	pte_t entry;
-	int i;
 
 	/* File mapping without ->vm_ops ? */
 	if (vma->vm_flags & VM_SHARED)
@@ -4491,8 +4490,7 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf)
 		update_mmu_tlb(vma, addr, vmf->pte);
 		goto release;
 	} else if (nr_pages > 1 && !pte_range_none(vmf->pte, nr_pages)) {
-		for (i = 0; i < nr_pages; i++)
-			update_mmu_tlb(vma, addr + PAGE_SIZE * i, vmf->pte + i);
+		update_mmu_tlb_range(vma, addr, vmf->pte, nr_pages);
 		goto release;
 	}
 
-- 
2.19.1.6.gb485710b

Re: [PATCH v3 3/3] mm: Use update_mmu_tlb_range() to simplify code

[Lance Yang]

On Sat, May 18, 2024 at 3:51 PM Bang Li <libang.li@antgroup.com> wrote:
>
> Let us simplify the code by update_mmu_tlb_range().
>

LGTM.
Reviewed-by: Lance Yang <ioworker0@gmail.com>

> Signed-off-by: Bang Li <libang.li@antgroup.com>
> ---
>  mm/memory.c | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/mm/memory.c b/mm/memory.c
> index eea6e4984eae..2d53e29cf76e 100644
> --- a/mm/memory.c
> +++ b/mm/memory.c
> @@ -4421,7 +4421,6 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf)
>         vm_fault_t ret = 0;
>         int nr_pages = 1;
>         pte_t entry;
> -       int i;
>
>         /* File mapping without ->vm_ops ? */
>         if (vma->vm_flags & VM_SHARED)
> @@ -4491,8 +4490,7 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf)
>                 update_mmu_tlb(vma, addr, vmf->pte);
>                 goto release;
>         } else if (nr_pages > 1 && !pte_range_none(vmf->pte, nr_pages)) {
> -               for (i = 0; i < nr_pages; i++)
> -                       update_mmu_tlb(vma, addr + PAGE_SIZE * i, vmf->pte + i);
> +               update_mmu_tlb_range(vma, addr, vmf->pte, nr_pages);
>                 goto release;
>         }
>
> --
> 2.19.1.6.gb485710b
>

Re: [PATCH v3 3/3] mm: Use update_mmu_tlb_range() to simplify code

[David Hildenbrand]

On 18.05.24 09:49, Bang Li wrote:
> Let us simplify the code by update_mmu_tlb_range().
> 
> Signed-off-by: Bang Li <libang.li@antgroup.com>
> ---
>   mm/memory.c | 4 +---
>   1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/mm/memory.c b/mm/memory.c
> index eea6e4984eae..2d53e29cf76e 100644
> --- a/mm/memory.c
> +++ b/mm/memory.c
> @@ -4421,7 +4421,6 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf)
>   	vm_fault_t ret = 0;
>   	int nr_pages = 1;
>   	pte_t entry;
> -	int i;
>   
>   	/* File mapping without ->vm_ops ? */
>   	if (vma->vm_flags & VM_SHARED)
> @@ -4491,8 +4490,7 @@ static vm_fault_t do_anonymous_page(struct vm_fault *vmf)
>   		update_mmu_tlb(vma, addr, vmf->pte);
>   		goto release;
>   	} else if (nr_pages > 1 && !pte_range_none(vmf->pte, nr_pages)) {
> -		for (i = 0; i < nr_pages; i++)
> -			update_mmu_tlb(vma, addr + PAGE_SIZE * i, vmf->pte + i);
> +		update_mmu_tlb_range(vma, addr, vmf->pte, nr_pages);
>   		goto release;
>   	}
>   

Acked-by: David Hildenbrand <david@redhat.com>

-- 
Cheers,

David / dhildenb