[PATCH v8 0/3] Speedup mremap on ppc64

[PATCH v8 0/3] Speedup mremap on ppc64

[Aneesh Kumar K.V]

This patchset enables MOVE_PMD/MOVE_PUD support on power. This requires
the platform to support updating higher-level page tables without
updating page table entries. This also needs to invalidate the Page Walk
Cache on architecture supporting the same.

This patchset is dependent on
https://lore.kernel.org/linux-mm/20210616045239.370802-1-aneesh.kumar@linux.ibm.com/

Changes from v7:
* Split mremap fixes to a separate series

Changes from v6:
* Update ppc64 flush_tlb_range to invalidate page walk cache.
* Add patches to fix race between mremap and page out
* Add patch to fix build error with page table levels 2

Changes from v5:
* Drop patch mm/mremap: Move TLB flush outside page table lock
* Add fixes for race between optimized mremap and page out

Changes from v4:
* Change function name and arguments based on review feedback.

Changes from v3:
* Fix build error reported by kernel test robot
* Address review feedback.

Changes from v2:
* switch from using mmu_gather to flush_pte_tlb_pwc_range() 

Changes from v1:
* Rebase to recent upstream
* Fix build issues with tlb_gather_mmu changes



Aneesh Kumar K.V (3):
  mm/mremap: Allow arch runtime override
  powerpc/book3s64/mm: Update flush_tlb_range to flush page walk cache
  powerpc/mm: Enable HAVE_MOVE_PMD support

 .../include/asm/book3s/64/tlbflush-radix.h    |  2 +
 arch/powerpc/include/asm/tlb.h                |  6 +++
 arch/powerpc/mm/book3s64/radix_hugetlbpage.c  |  8 +++-
 arch/powerpc/mm/book3s64/radix_tlb.c          | 44 ++++++++++++-------
 arch/powerpc/platforms/Kconfig.cputype        |  2 +
 mm/mremap.c                                   | 15 ++++++-
 6 files changed, 58 insertions(+), 19 deletions(-)

-- 
2.31.1

[PATCH v8 0/3] Speedup mremap on ppc64

[Aneesh Kumar K.V]

This patchset enables MOVE_PMD/MOVE_PUD support on power. This requires
the platform to support updating higher-level page tables without
updating page table entries. This also needs to invalidate the Page Walk
Cache on architecture supporting the same.

This patchset is dependent on
https://lore.kernel.org/linux-mm/20210616045239.370802-1-aneesh.kumar@linux.ibm.com/

Changes from v7:
* Split mremap fixes to a separate series

Changes from v6:
* Update ppc64 flush_tlb_range to invalidate page walk cache.
* Add patches to fix race between mremap and page out
* Add patch to fix build error with page table levels 2

Changes from v5:
* Drop patch mm/mremap: Move TLB flush outside page table lock
* Add fixes for race between optimized mremap and page out

Changes from v4:
* Change function name and arguments based on review feedback.

Changes from v3:
* Fix build error reported by kernel test robot
* Address review feedback.

Changes from v2:
* switch from using mmu_gather to flush_pte_tlb_pwc_range() 

Changes from v1:
* Rebase to recent upstream
* Fix build issues with tlb_gather_mmu changes



Aneesh Kumar K.V (3):
  mm/mremap: Allow arch runtime override
  powerpc/book3s64/mm: Update flush_tlb_range to flush page walk cache
  powerpc/mm: Enable HAVE_MOVE_PMD support

 .../include/asm/book3s/64/tlbflush-radix.h    |  2 +
 arch/powerpc/include/asm/tlb.h                |  6 +++
 arch/powerpc/mm/book3s64/radix_hugetlbpage.c  |  8 +++-
 arch/powerpc/mm/book3s64/radix_tlb.c          | 44 ++++++++++++-------
 arch/powerpc/platforms/Kconfig.cputype        |  2 +
 mm/mremap.c                                   | 15 ++++++-
 6 files changed, 58 insertions(+), 19 deletions(-)

-- 
2.31.1

[PATCH v8 1/3] mm/mremap: Allow arch runtime override

[Aneesh Kumar K.V]

Architectures like ppc64 support faster mremap only with radix
translation. Hence allow a runtime check w.r.t support for fast mremap.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/tlb.h |  6 ++++++
 mm/mremap.c                    | 15 ++++++++++++++-
 2 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/include/asm/tlb.h b/arch/powerpc/include/asm/tlb.h
index 160422a439aa..09a9ae5f3656 100644
--- a/arch/powerpc/include/asm/tlb.h
+++ b/arch/powerpc/include/asm/tlb.h
@@ -83,5 +83,11 @@ static inline int mm_is_thread_local(struct mm_struct *mm)
 }
 #endif
 
+#define arch_supports_page_table_move arch_supports_page_table_move
+static inline bool arch_supports_page_table_move(void)
+{
+	return radix_enabled();
+}
+
 #endif /* __KERNEL__ */
 #endif /* __ASM_POWERPC_TLB_H */
diff --git a/mm/mremap.c b/mm/mremap.c
index c3cad539a7aa..ca9d345f22e8 100644
--- a/mm/mremap.c
+++ b/mm/mremap.c
@@ -25,7 +25,7 @@
 #include <linux/userfaultfd_k.h>
 
 #include <asm/cacheflush.h>
-#include <asm/tlbflush.h>
+#include <asm/tlb.h>
 #include <asm/pgalloc.h>
 
 #include "internal.h"
@@ -210,6 +210,15 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd,
 		drop_rmap_locks(vma);
 }
 
+#ifndef arch_supports_page_table_move
+#define arch_supports_page_table_move arch_supports_page_table_move
+static inline bool arch_supports_page_table_move(void)
+{
+	return IS_ENABLED(CONFIG_HAVE_MOVE_PMD) ||
+		IS_ENABLED(CONFIG_HAVE_MOVE_PUD);
+}
+#endif
+
 #ifdef CONFIG_HAVE_MOVE_PMD
 static bool move_normal_pmd(struct vm_area_struct *vma, unsigned long old_addr,
 		  unsigned long new_addr, pmd_t *old_pmd, pmd_t *new_pmd)
@@ -218,6 +227,8 @@ static bool move_normal_pmd(struct vm_area_struct *vma, unsigned long old_addr,
 	struct mm_struct *mm = vma->vm_mm;
 	pmd_t pmd;
 
+	if (!arch_supports_page_table_move())
+		return false;
 	/*
 	 * The destination pmd shouldn't be established, free_pgtables()
 	 * should have released it.
@@ -284,6 +295,8 @@ static bool move_normal_pud(struct vm_area_struct *vma, unsigned long old_addr,
 	struct mm_struct *mm = vma->vm_mm;
 	pud_t pud;
 
+	if (!arch_supports_page_table_move())
+		return false;
 	/*
 	 * The destination pud shouldn't be established, free_pgtables()
 	 * should have released it.
-- 
2.31.1

[PATCH v8 1/3] mm/mremap: Allow arch runtime override

[Aneesh Kumar K.V]

Architectures like ppc64 support faster mremap only with radix
translation. Hence allow a runtime check w.r.t support for fast mremap.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/tlb.h |  6 ++++++
 mm/mremap.c                    | 15 ++++++++++++++-
 2 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/include/asm/tlb.h b/arch/powerpc/include/asm/tlb.h
index 160422a439aa..09a9ae5f3656 100644
--- a/arch/powerpc/include/asm/tlb.h
+++ b/arch/powerpc/include/asm/tlb.h
@@ -83,5 +83,11 @@ static inline int mm_is_thread_local(struct mm_struct *mm)
 }
 #endif
 
+#define arch_supports_page_table_move arch_supports_page_table_move
+static inline bool arch_supports_page_table_move(void)
+{
+	return radix_enabled();
+}
+
 #endif /* __KERNEL__ */
 #endif /* __ASM_POWERPC_TLB_H */
diff --git a/mm/mremap.c b/mm/mremap.c
index c3cad539a7aa..ca9d345f22e8 100644
--- a/mm/mremap.c
+++ b/mm/mremap.c
@@ -25,7 +25,7 @@
 #include <linux/userfaultfd_k.h>
 
 #include <asm/cacheflush.h>
-#include <asm/tlbflush.h>
+#include <asm/tlb.h>
 #include <asm/pgalloc.h>
 
 #include "internal.h"
@@ -210,6 +210,15 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd,
 		drop_rmap_locks(vma);
 }
 
+#ifndef arch_supports_page_table_move
+#define arch_supports_page_table_move arch_supports_page_table_move
+static inline bool arch_supports_page_table_move(void)
+{
+	return IS_ENABLED(CONFIG_HAVE_MOVE_PMD) ||
+		IS_ENABLED(CONFIG_HAVE_MOVE_PUD);
+}
+#endif
+
 #ifdef CONFIG_HAVE_MOVE_PMD
 static bool move_normal_pmd(struct vm_area_struct *vma, unsigned long old_addr,
 		  unsigned long new_addr, pmd_t *old_pmd, pmd_t *new_pmd)
@@ -218,6 +227,8 @@ static bool move_normal_pmd(struct vm_area_struct *vma, unsigned long old_addr,
 	struct mm_struct *mm = vma->vm_mm;
 	pmd_t pmd;
 
+	if (!arch_supports_page_table_move())
+		return false;
 	/*
 	 * The destination pmd shouldn't be established, free_pgtables()
 	 * should have released it.
@@ -284,6 +295,8 @@ static bool move_normal_pud(struct vm_area_struct *vma, unsigned long old_addr,
 	struct mm_struct *mm = vma->vm_mm;
 	pud_t pud;
 
+	if (!arch_supports_page_table_move())
+		return false;
 	/*
 	 * The destination pud shouldn't be established, free_pgtables()
 	 * should have released it.
-- 
2.31.1

[PATCH v8 2/3] powerpc/book3s64/mm: Update flush_tlb_range to flush page walk cache

[Aneesh Kumar K.V]

flush_tlb_range is special in that we don't specify the page size used
for the translation. Hence when flushing TLB we flush the translation cache
for all possible page sizes. The kernel also uses the same interface when
moving page tables around. Such a move requires us to flush the page walk cache.

Instead of adding another interface to force page walk cache flush,
update flush_tlb_range to flush page walk cache if the range flushed
is more than the PMD range. A page table move will always involve an
invalidate range more than PMD_SIZE.

Running microbenchmark with mprotect and parallel memory access
didn't show any observable performance impact.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 .../include/asm/book3s/64/tlbflush-radix.h    |  2 +
 arch/powerpc/mm/book3s64/radix_hugetlbpage.c  |  8 +++-
 arch/powerpc/mm/book3s64/radix_tlb.c          | 44 ++++++++++++-------
 3 files changed, 36 insertions(+), 18 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h b/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h
index 8b33601cdb9d..ab9d5e535000 100644
--- a/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h
+++ b/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h
@@ -60,6 +60,8 @@ extern void radix__flush_hugetlb_tlb_range(struct vm_area_struct *vma,
 					   unsigned long start, unsigned long end);
 extern void radix__flush_tlb_range_psize(struct mm_struct *mm, unsigned long start,
 					 unsigned long end, int psize);
+void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
+				      unsigned long end, int psize);
 extern void radix__flush_pmd_tlb_range(struct vm_area_struct *vma,
 				       unsigned long start, unsigned long end);
 extern void radix__flush_tlb_range(struct vm_area_struct *vma, unsigned long start,
diff --git a/arch/powerpc/mm/book3s64/radix_hugetlbpage.c b/arch/powerpc/mm/book3s64/radix_hugetlbpage.c
index cb91071eef52..23d3e08911d3 100644
--- a/arch/powerpc/mm/book3s64/radix_hugetlbpage.c
+++ b/arch/powerpc/mm/book3s64/radix_hugetlbpage.c
@@ -32,7 +32,13 @@ void radix__flush_hugetlb_tlb_range(struct vm_area_struct *vma, unsigned long st
 	struct hstate *hstate = hstate_file(vma->vm_file);
 
 	psize = hstate_get_psize(hstate);
-	radix__flush_tlb_range_psize(vma->vm_mm, start, end, psize);
+	/*
+	 * Flush PWC even if we get PUD_SIZE hugetlb invalidate to keep this simpler.
+	 */
+	if (end - start >= PUD_SIZE)
+		radix__flush_tlb_pwc_range_psize(vma->vm_mm, start, end, psize);
+	else
+		radix__flush_tlb_range_psize(vma->vm_mm, start, end, psize);
 }
 
 /*
diff --git a/arch/powerpc/mm/book3s64/radix_tlb.c b/arch/powerpc/mm/book3s64/radix_tlb.c
index 409e61210789..9f1a177f6bb6 100644
--- a/arch/powerpc/mm/book3s64/radix_tlb.c
+++ b/arch/powerpc/mm/book3s64/radix_tlb.c
@@ -989,14 +989,13 @@ static unsigned long tlb_local_single_page_flush_ceiling __read_mostly = POWER9_
 
 static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 					    unsigned long start, unsigned long end)
-
 {
 	unsigned long pid;
 	unsigned int page_shift = mmu_psize_defs[mmu_virtual_psize].shift;
 	unsigned long page_size = 1UL << page_shift;
 	unsigned long nr_pages = (end - start) >> page_shift;
 	bool fullmm = (end == TLB_FLUSH_ALL);
-	bool flush_pid;
+	bool flush_pid, flush_pwc = false;
 	enum tlb_flush_type type;
 
 	pid = mm->context.id;
@@ -1015,8 +1014,16 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 		flush_pid = nr_pages > tlb_single_page_flush_ceiling;
 	else
 		flush_pid = nr_pages > tlb_local_single_page_flush_ceiling;
+	/*
+	 * full pid flush already does the PWC flush. if it is not full pid
+	 * flush check the range is more than PMD and force a pwc flush
+	 * mremap() depends on this behaviour.
+	 */
+	if (!flush_pid && (end - start) >= PMD_SIZE)
+		flush_pwc = true;
 
 	if (!mmu_has_feature(MMU_FTR_GTSE) && type == FLUSH_TYPE_GLOBAL) {
+		unsigned long type = H_RPTI_TYPE_TLB;
 		unsigned long tgt = H_RPTI_TARGET_CMMU;
 		unsigned long pg_sizes = psize_to_rpti_pgsize(mmu_virtual_psize);
 
@@ -1024,19 +1031,20 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 			pg_sizes |= psize_to_rpti_pgsize(MMU_PAGE_2M);
 		if (atomic_read(&mm->context.copros) > 0)
 			tgt |= H_RPTI_TARGET_NMMU;
-		pseries_rpt_invalidate(pid, tgt, H_RPTI_TYPE_TLB, pg_sizes,
-				       start, end);
+		if (flush_pwc)
+			type |= H_RPTI_TYPE_PWC;
+		pseries_rpt_invalidate(pid, tgt, type, pg_sizes, start, end);
 	} else if (flush_pid) {
+		/*
+		 * We are now flushing a range larger than PMD size force a RIC_FLUSH_ALL
+		 */
 		if (type == FLUSH_TYPE_LOCAL) {
-			_tlbiel_pid(pid, RIC_FLUSH_TLB);
+			_tlbiel_pid(pid, RIC_FLUSH_ALL);
 		} else {
 			if (cputlb_use_tlbie()) {
-				if (mm_needs_flush_escalation(mm))
-					_tlbie_pid(pid, RIC_FLUSH_ALL);
-				else
-					_tlbie_pid(pid, RIC_FLUSH_TLB);
+				_tlbie_pid(pid, RIC_FLUSH_ALL);
 			} else {
-				_tlbiel_pid_multicast(mm, pid, RIC_FLUSH_TLB);
+				_tlbiel_pid_multicast(mm, pid, RIC_FLUSH_ALL);
 			}
 		}
 	} else {
@@ -1052,6 +1060,9 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 
 		if (type == FLUSH_TYPE_LOCAL) {
 			asm volatile("ptesync": : :"memory");
+			if (flush_pwc)
+				/* For PWC, only one flush is needed */
+				__tlbiel_pid(pid, 0, RIC_FLUSH_PWC);
 			__tlbiel_va_range(start, end, pid, page_size, mmu_virtual_psize);
 			if (hflush)
 				__tlbiel_va_range(hstart, hend, pid,
@@ -1059,6 +1070,8 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 			ppc_after_tlbiel_barrier();
 		} else if (cputlb_use_tlbie()) {
 			asm volatile("ptesync": : :"memory");
+			if (flush_pwc)
+				__tlbie_pid(pid, RIC_FLUSH_PWC);
 			__tlbie_va_range(start, end, pid, page_size, mmu_virtual_psize);
 			if (hflush)
 				__tlbie_va_range(hstart, hend, pid,
@@ -1066,10 +1079,10 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 			asm volatile("eieio; tlbsync; ptesync": : :"memory");
 		} else {
 			_tlbiel_va_range_multicast(mm,
-					start, end, pid, page_size, mmu_virtual_psize, false);
+					start, end, pid, page_size, mmu_virtual_psize, flush_pwc);
 			if (hflush)
 				_tlbiel_va_range_multicast(mm,
-					hstart, hend, pid, PMD_SIZE, MMU_PAGE_2M, false);
+					hstart, hend, pid, PMD_SIZE, MMU_PAGE_2M, flush_pwc);
 		}
 	}
 out:
@@ -1143,9 +1156,6 @@ void radix__flush_all_lpid_guest(unsigned int lpid)
 	_tlbie_lpid_guest(lpid, RIC_FLUSH_ALL);
 }
 
-static void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
-				  unsigned long end, int psize);
-
 void radix__tlb_flush(struct mmu_gather *tlb)
 {
 	int psize = 0;
@@ -1252,8 +1262,8 @@ void radix__flush_tlb_range_psize(struct mm_struct *mm, unsigned long start,
 	return __radix__flush_tlb_range_psize(mm, start, end, psize, false);
 }
 
-static void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
-				  unsigned long end, int psize)
+void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
+				      unsigned long end, int psize)
 {
 	__radix__flush_tlb_range_psize(mm, start, end, psize, true);
 }
-- 
2.31.1

[PATCH v8 2/3] powerpc/book3s64/mm: Update flush_tlb_range to flush page walk cache

[Aneesh Kumar K.V]

flush_tlb_range is special in that we don't specify the page size used
for the translation. Hence when flushing TLB we flush the translation cache
for all possible page sizes. The kernel also uses the same interface when
moving page tables around. Such a move requires us to flush the page walk cache.

Instead of adding another interface to force page walk cache flush,
update flush_tlb_range to flush page walk cache if the range flushed
is more than the PMD range. A page table move will always involve an
invalidate range more than PMD_SIZE.

Running microbenchmark with mprotect and parallel memory access
didn't show any observable performance impact.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 .../include/asm/book3s/64/tlbflush-radix.h    |  2 +
 arch/powerpc/mm/book3s64/radix_hugetlbpage.c  |  8 +++-
 arch/powerpc/mm/book3s64/radix_tlb.c          | 44 ++++++++++++-------
 3 files changed, 36 insertions(+), 18 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h b/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h
index 8b33601cdb9d..ab9d5e535000 100644
--- a/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h
+++ b/arch/powerpc/include/asm/book3s/64/tlbflush-radix.h
@@ -60,6 +60,8 @@ extern void radix__flush_hugetlb_tlb_range(struct vm_area_struct *vma,
 					   unsigned long start, unsigned long end);
 extern void radix__flush_tlb_range_psize(struct mm_struct *mm, unsigned long start,
 					 unsigned long end, int psize);
+void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
+				      unsigned long end, int psize);
 extern void radix__flush_pmd_tlb_range(struct vm_area_struct *vma,
 				       unsigned long start, unsigned long end);
 extern void radix__flush_tlb_range(struct vm_area_struct *vma, unsigned long start,
diff --git a/arch/powerpc/mm/book3s64/radix_hugetlbpage.c b/arch/powerpc/mm/book3s64/radix_hugetlbpage.c
index cb91071eef52..23d3e08911d3 100644
--- a/arch/powerpc/mm/book3s64/radix_hugetlbpage.c
+++ b/arch/powerpc/mm/book3s64/radix_hugetlbpage.c
@@ -32,7 +32,13 @@ void radix__flush_hugetlb_tlb_range(struct vm_area_struct *vma, unsigned long st
 	struct hstate *hstate = hstate_file(vma->vm_file);
 
 	psize = hstate_get_psize(hstate);
-	radix__flush_tlb_range_psize(vma->vm_mm, start, end, psize);
+	/*
+	 * Flush PWC even if we get PUD_SIZE hugetlb invalidate to keep this simpler.
+	 */
+	if (end - start >= PUD_SIZE)
+		radix__flush_tlb_pwc_range_psize(vma->vm_mm, start, end, psize);
+	else
+		radix__flush_tlb_range_psize(vma->vm_mm, start, end, psize);
 }
 
 /*
diff --git a/arch/powerpc/mm/book3s64/radix_tlb.c b/arch/powerpc/mm/book3s64/radix_tlb.c
index 409e61210789..9f1a177f6bb6 100644
--- a/arch/powerpc/mm/book3s64/radix_tlb.c
+++ b/arch/powerpc/mm/book3s64/radix_tlb.c
@@ -989,14 +989,13 @@ static unsigned long tlb_local_single_page_flush_ceiling __read_mostly = POWER9_
 
 static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 					    unsigned long start, unsigned long end)
-
 {
 	unsigned long pid;
 	unsigned int page_shift = mmu_psize_defs[mmu_virtual_psize].shift;
 	unsigned long page_size = 1UL << page_shift;
 	unsigned long nr_pages = (end - start) >> page_shift;
 	bool fullmm = (end == TLB_FLUSH_ALL);
-	bool flush_pid;
+	bool flush_pid, flush_pwc = false;
 	enum tlb_flush_type type;
 
 	pid = mm->context.id;
@@ -1015,8 +1014,16 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 		flush_pid = nr_pages > tlb_single_page_flush_ceiling;
 	else
 		flush_pid = nr_pages > tlb_local_single_page_flush_ceiling;
+	/*
+	 * full pid flush already does the PWC flush. if it is not full pid
+	 * flush check the range is more than PMD and force a pwc flush
+	 * mremap() depends on this behaviour.
+	 */
+	if (!flush_pid && (end - start) >= PMD_SIZE)
+		flush_pwc = true;
 
 	if (!mmu_has_feature(MMU_FTR_GTSE) && type == FLUSH_TYPE_GLOBAL) {
+		unsigned long type = H_RPTI_TYPE_TLB;
 		unsigned long tgt = H_RPTI_TARGET_CMMU;
 		unsigned long pg_sizes = psize_to_rpti_pgsize(mmu_virtual_psize);
 
@@ -1024,19 +1031,20 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 			pg_sizes |= psize_to_rpti_pgsize(MMU_PAGE_2M);
 		if (atomic_read(&mm->context.copros) > 0)
 			tgt |= H_RPTI_TARGET_NMMU;
-		pseries_rpt_invalidate(pid, tgt, H_RPTI_TYPE_TLB, pg_sizes,
-				       start, end);
+		if (flush_pwc)
+			type |= H_RPTI_TYPE_PWC;
+		pseries_rpt_invalidate(pid, tgt, type, pg_sizes, start, end);
 	} else if (flush_pid) {
+		/*
+		 * We are now flushing a range larger than PMD size force a RIC_FLUSH_ALL
+		 */
 		if (type == FLUSH_TYPE_LOCAL) {
-			_tlbiel_pid(pid, RIC_FLUSH_TLB);
+			_tlbiel_pid(pid, RIC_FLUSH_ALL);
 		} else {
 			if (cputlb_use_tlbie()) {
-				if (mm_needs_flush_escalation(mm))
-					_tlbie_pid(pid, RIC_FLUSH_ALL);
-				else
-					_tlbie_pid(pid, RIC_FLUSH_TLB);
+				_tlbie_pid(pid, RIC_FLUSH_ALL);
 			} else {
-				_tlbiel_pid_multicast(mm, pid, RIC_FLUSH_TLB);
+				_tlbiel_pid_multicast(mm, pid, RIC_FLUSH_ALL);
 			}
 		}
 	} else {
@@ -1052,6 +1060,9 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 
 		if (type == FLUSH_TYPE_LOCAL) {
 			asm volatile("ptesync": : :"memory");
+			if (flush_pwc)
+				/* For PWC, only one flush is needed */
+				__tlbiel_pid(pid, 0, RIC_FLUSH_PWC);
 			__tlbiel_va_range(start, end, pid, page_size, mmu_virtual_psize);
 			if (hflush)
 				__tlbiel_va_range(hstart, hend, pid,
@@ -1059,6 +1070,8 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 			ppc_after_tlbiel_barrier();
 		} else if (cputlb_use_tlbie()) {
 			asm volatile("ptesync": : :"memory");
+			if (flush_pwc)
+				__tlbie_pid(pid, RIC_FLUSH_PWC);
 			__tlbie_va_range(start, end, pid, page_size, mmu_virtual_psize);
 			if (hflush)
 				__tlbie_va_range(hstart, hend, pid,
@@ -1066,10 +1079,10 @@ static inline void __radix__flush_tlb_range(struct mm_struct *mm,
 			asm volatile("eieio; tlbsync; ptesync": : :"memory");
 		} else {
 			_tlbiel_va_range_multicast(mm,
-					start, end, pid, page_size, mmu_virtual_psize, false);
+					start, end, pid, page_size, mmu_virtual_psize, flush_pwc);
 			if (hflush)
 				_tlbiel_va_range_multicast(mm,
-					hstart, hend, pid, PMD_SIZE, MMU_PAGE_2M, false);
+					hstart, hend, pid, PMD_SIZE, MMU_PAGE_2M, flush_pwc);
 		}
 	}
 out:
@@ -1143,9 +1156,6 @@ void radix__flush_all_lpid_guest(unsigned int lpid)
 	_tlbie_lpid_guest(lpid, RIC_FLUSH_ALL);
 }
 
-static void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
-				  unsigned long end, int psize);
-
 void radix__tlb_flush(struct mmu_gather *tlb)
 {
 	int psize = 0;
@@ -1252,8 +1262,8 @@ void radix__flush_tlb_range_psize(struct mm_struct *mm, unsigned long start,
 	return __radix__flush_tlb_range_psize(mm, start, end, psize, false);
 }
 
-static void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
-				  unsigned long end, int psize)
+void radix__flush_tlb_pwc_range_psize(struct mm_struct *mm, unsigned long start,
+				      unsigned long end, int psize)
 {
 	__radix__flush_tlb_range_psize(mm, start, end, psize, true);
 }
-- 
2.31.1

[PATCH v8 3/3] powerpc/mm: Enable HAVE_MOVE_PMD support

[Aneesh Kumar K.V]

mremap HAVE_MOVE_PMD/PUD optimization time comparison for 1GB region:
1GB mremap - Source PTE-aligned, Destination PTE-aligned
  mremap time:      2292772ns
1GB mremap - Source PMD-aligned, Destination PMD-aligned
  mremap time:      1158928ns
1GB mremap - Source PUD-aligned, Destination PUD-aligned
  mremap time:        63886ns

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/platforms/Kconfig.cputype | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/arch/powerpc/platforms/Kconfig.cputype b/arch/powerpc/platforms/Kconfig.cputype
index f998e655b570..be8ceb5bece4 100644
--- a/arch/powerpc/platforms/Kconfig.cputype
+++ b/arch/powerpc/platforms/Kconfig.cputype
@@ -101,6 +101,8 @@ config PPC_BOOK3S_64
 	select ARCH_ENABLE_THP_MIGRATION if TRANSPARENT_HUGEPAGE
 	select ARCH_SUPPORTS_HUGETLBFS
 	select ARCH_SUPPORTS_NUMA_BALANCING
+	select HAVE_MOVE_PMD
+	select HAVE_MOVE_PUD
 	select IRQ_WORK
 	select PPC_MM_SLICES
 	select PPC_HAVE_KUEP
-- 
2.31.1

[PATCH v8 3/3] powerpc/mm: Enable HAVE_MOVE_PMD support

[Aneesh Kumar K.V]

mremap HAVE_MOVE_PMD/PUD optimization time comparison for 1GB region:
1GB mremap - Source PTE-aligned, Destination PTE-aligned
  mremap time:      2292772ns
1GB mremap - Source PMD-aligned, Destination PMD-aligned
  mremap time:      1158928ns
1GB mremap - Source PUD-aligned, Destination PUD-aligned
  mremap time:        63886ns

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/platforms/Kconfig.cputype | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/arch/powerpc/platforms/Kconfig.cputype b/arch/powerpc/platforms/Kconfig.cputype
index f998e655b570..be8ceb5bece4 100644
--- a/arch/powerpc/platforms/Kconfig.cputype
+++ b/arch/powerpc/platforms/Kconfig.cputype
@@ -101,6 +101,8 @@ config PPC_BOOK3S_64
 	select ARCH_ENABLE_THP_MIGRATION if TRANSPARENT_HUGEPAGE
 	select ARCH_SUPPORTS_HUGETLBFS
 	select ARCH_SUPPORTS_NUMA_BALANCING
+	select HAVE_MOVE_PMD
+	select HAVE_MOVE_PUD
 	select IRQ_WORK
 	select PPC_MM_SLICES
 	select PPC_HAVE_KUEP
-- 
2.31.1