mmu_gather: Remove per arch tlb_{start,end}_vma()

Scattered across the archs are 3 basic forms of tlb_{start,end}_vma().
Provide two new MMU_GATHER_knobs to enumerate them and remove the per
arch tlb_{start,end}_vma() implementations.

- MMU_GATHER_NO_FLUSH_CACHE indicates the arch has flush_cache_range()
but does *NOT* want to call it for each VMA.

- MMU_GATHER_MERGE_VMAS indicates the arch wants to merge the
invalidate across multiple VMAs if possible.

With these it is possible to capture the three forms:

1) empty stubs;
select MMU_GATHER_NO_FLUSH_CACHE and MMU_GATHER_MERGE_VMAS

2) start: flush_cache_range(), end: empty;
select MMU_GATHER_MERGE_VMAS

3) start: flush_cache_range(), end: flush_tlb_range();
default

Obviously, if the architecture does not have flush_cache_range() then
it also doesn't need to select MMU_GATHER_NO_FLUSH_CACHE.

Signed-off-by: Peter Zijlstra (Intel) <[e-mail]>
Acked-by: Will Deacon <[e-mail]>
Cc: David Miller <[e-mail]>
Signed-off-by: Linus Torvalds <[e-mail]>

arch/Kconfig

···
 
 config MMU_GATHER_NO_RANGE
 	bool
+	select MMU_GATHER_MERGE_VMAS
+
+config MMU_GATHER_NO_FLUSH_CACHE
+	bool
+
+config MMU_GATHER_MERGE_VMAS
+	bool
 
 config MMU_GATHER_NO_GATHER
 	bool

arch/csky/include/asm/tlb.h

···
 #define __ASM_CSKY_TLB_H
 
 #include <asm/cacheflush.h>
-
-#define tlb_start_vma(tlb, vma) \
-	do { \
-		if (!(tlb)->fullmm) \
-			flush_cache_range(vma, (vma)->vm_start, (vma)->vm_end); \
-	}  while (0)
-
-#define tlb_end_vma(tlb, vma) \
-	do { \
-		if (!(tlb)->fullmm) \
-			flush_tlb_range(vma, (vma)->vm_start, (vma)->vm_end); \
-	}  while (0)
-
 #define tlb_flush(tlb) flush_tlb_mm((tlb)->mm)
 
 #include <asm-generic/tlb.h>

arch/loongarch/Kconfig

···
 	select TRACE_IRQFLAGS_SUPPORT
 	select USE_PERCPU_NUMA_NODE_ID
 	select ZONE_DMA32
+	select MMU_GATHER_MERGE_VMAS if MMU
 
 config 32BIT
 	bool

arch/loongarch/include/asm/tlb.h

···
 		);
 }
 
-/*
- * LoongArch doesn't need any special per-pte or per-vma handling, except
- * we need to flush cache for area to be unmapped.
- */
-#define tlb_start_vma(tlb, vma)					\
-	do {							\
-		if (!(tlb)->fullmm)				\
-			flush_cache_range(vma, vma->vm_start, vma->vm_end); \
-	}  while (0)
-#define tlb_end_vma(tlb, vma) do { } while (0)
 #define __tlb_remove_tlb_entry(tlb, ptep, address) do { } while (0)
 
 static void tlb_flush(struct mmu_gather *tlb);

arch/powerpc/Kconfig

···
 	select IRQ_FORCED_THREADING
 	select MMU_GATHER_PAGE_SIZE
 	select MMU_GATHER_RCU_TABLE_FREE
+	select MMU_GATHER_MERGE_VMAS
 	select MODULES_USE_ELF_RELA
 	select NEED_DMA_MAP_STATE		if PPC64 || NOT_COHERENT_CACHE
 	select NEED_PER_CPU_EMBED_FIRST_CHUNK	if PPC64

arch/powerpc/include/asm/tlb.h

···
 
 #include <linux/pagemap.h>
 
-#define tlb_start_vma(tlb, vma)	do { } while (0)
-#define tlb_end_vma(tlb, vma)	do { } while (0)
 #define __tlb_remove_tlb_entry	__tlb_remove_tlb_entry
 
 #define tlb_flush tlb_flush

arch/s390/Kconfig

···
 	select IOMMU_SUPPORT		if PCI
 	select MMU_GATHER_NO_GATHER
 	select MMU_GATHER_RCU_TABLE_FREE
+	select MMU_GATHER_MERGE_VMAS
 	select MODULES_USE_ELF_RELA
 	select NEED_DMA_MAP_STATE	if PCI
 	select NEED_SG_DMA_LENGTH	if PCI

arch/s390/include/asm/tlb.h

···
 static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
 					  struct page *page, int page_size);
 
-#define tlb_start_vma(tlb, vma)			do { } while (0)
-#define tlb_end_vma(tlb, vma)			do { } while (0)
-
 #define tlb_flush tlb_flush
 #define pte_free_tlb pte_free_tlb
 #define pmd_free_tlb pmd_free_tlb

arch/sparc/Kconfig

···
 	select HAVE_KRETPROBES
 	select HAVE_KPROBES
 	select MMU_GATHER_RCU_TABLE_FREE if SMP
+	select MMU_GATHER_MERGE_VMAS
+	select MMU_GATHER_NO_FLUSH_CACHE
 	select HAVE_ARCH_TRANSPARENT_HUGEPAGE
 	select HAVE_DYNAMIC_FTRACE
 	select HAVE_FTRACE_MCOUNT_RECORD

arch/sparc/include/asm/tlb_64.h

···
 void __flush_tlb_pending(unsigned long, unsigned long, unsigned long *);
 void flush_tlb_pending(void);
 
-#define tlb_start_vma(tlb, vma) do { } while (0)
-#define tlb_end_vma(tlb, vma)	do { } while (0)
 #define tlb_flush(tlb)	flush_tlb_pending()
 
 /*

arch/x86/Kconfig

···
 	select HAVE_PERF_REGS
 	select HAVE_PERF_USER_STACK_DUMP
 	select MMU_GATHER_RCU_TABLE_FREE	if PARAVIRT
+	select MMU_GATHER_MERGE_VMAS
 	select HAVE_POSIX_CPU_TIMERS_TASK_WORK
 	select HAVE_REGS_AND_STACK_ACCESS_API
 	select HAVE_RELIABLE_STACKTRACE		if UNWINDER_ORC || STACK_VALIDATION

arch/x86/include/asm/tlb.h

···
 #ifndef _ASM_X86_TLB_H
 #define _ASM_X86_TLB_H
 
-#define tlb_start_vma(tlb, vma) do { } while (0)
-#define tlb_end_vma(tlb, vma) do { } while (0)
-
 #define tlb_flush tlb_flush
 static inline void tlb_flush(struct mmu_gather *tlb);
 

include/asm-generic/tlb.h

···
  *  Useful if your architecture doesn't use IPIs for remote TLB invalidates
  *  and therefore doesn't naturally serialize with software page-table walkers.
  *
+ *  MMU_GATHER_NO_FLUSH_CACHE
+ *
+ *  Indicates the architecture has flush_cache_range() but it needs *NOT* be called
+ *  before unmapping a VMA.
+ *
+ *  NOTE: strictly speaking we shouldn't have this knob and instead rely on
+ *	  flush_cache_range() being a NOP, except Sparc64 seems to be
+ *	  different here.
+ *
+ *  MMU_GATHER_MERGE_VMAS
+ *
+ *  Indicates the architecture wants to merge ranges over VMAs; typical when
+ *  multiple range invalidates are more expensive than a full invalidate.
+ *
  *  MMU_GATHER_NO_RANGE
  *
- *  Use this if your architecture lacks an efficient flush_tlb_range().
+ *  Use this if your architecture lacks an efficient flush_tlb_range(). This
+ *  option implies MMU_GATHER_MERGE_VMAS above.
  *
  *  MMU_GATHER_NO_GATHER
  *
···
 		return;
 
 	tlb_update_vma_flags(tlb, vma);
+#ifndef CONFIG_MMU_GATHER_NO_FLUSH_CACHE
 	flush_cache_range(vma, vma->vm_start, vma->vm_end);
+#endif
 }
 #endif
 
 #ifndef tlb_end_vma
 static inline void tlb_end_vma(struct mmu_gather *tlb, struct vm_area_struct *vma)
 {
-	if (tlb->fullmm)
+	if (tlb->fullmm || IS_ENABLED(CONFIG_MMU_GATHER_MERGE_VMAS))
 		return;
 
 	/*