mm: mmu_gather rework

Rework the existing mmu_gather infrastructure.

The direct purpose of these patches was to allow preemptible mmu_gather,
but even without that I think these patches provide an improvement to the
status quo.

The first 9 patches rework the mmu_gather infrastructure.  For review
purpose I've split them into generic and per-arch patches with the last of
those a generic cleanup.

The next patch provides generic RCU page-table freeing, and the followup
is a patch converting s390 to use this.  I've also got 4 patches from
DaveM lined up (not included in this series) that uses this to implement
gup_fast() for sparc64.

Then there is one patch that extends the generic mmu_gather batching.

After that follow the mm preemptibility patches, these make part of the mm
a lot more preemptible.  It converts i_mmap_lock and anon_vma->lock to
mutexes which together with the mmu_gather rework makes mmu_gather
preemptible as well.

Making i_mmap_lock a mutex also enables a clean-up of the truncate code.

This also allows for preemptible mmu_notifiers, something that XPMEM I
think wants.

Furthermore, it removes the new and universially detested unmap_mutex.

This patch:

Remove the first obstacle towards a fully preemptible mmu_gather.

The current scheme assumes mmu_gather is always done with preemption
disabled and uses per-cpu storage for the page batches.  Change this to
try and allocate a page for batching and in case of failure, use a small
on-stack array to make some progress.

Preemptible mmu_gather is desired in general and usable once i_mmap_lock
becomes a mutex.  Doing it before the mutex conversion saves us from
having to rework the code by moving the mmu_gather bits inside the
pte_lock.

Also avoid flushing the tlb batches from under the pte lock, this is
useful even without the i_mmap_lock conversion as it significantly reduces
pte lock hold times.

[akpm@linux-foundation.org: fix comment tpyo]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Miller <davem@davemloft.net>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Richard Weinberger <richard@nod.at>
Cc: Tony Luck <tony.luck@intel.com>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Hugh Dickins <hughd@google.com>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Nick Piggin <npiggin@kernel.dk>
Cc: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/exec.c

@@ -600,7 +600,7 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift)
 	unsigned long length = old_end - old_start;
 	unsigned long new_start = old_start - shift;
 	unsigned long new_end = old_end - shift;
-	struct mmu_gather *tlb;
+	struct mmu_gather tlb;
 
 	BUG_ON(new_start > new_end);
 
@@ -626,12 +626,12 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift)
 		return -ENOMEM;
 
 	lru_add_drain();
-	tlb = tlb_gather_mmu(mm, 0);
+	tlb_gather_mmu(&tlb, mm, 0);
 	if (new_end > old_start) {
 		/*
 		 * when the old and new regions overlap clear from new_end.
 		 */
-		free_pgd_range(tlb, new_end, old_end, new_end,
+		free_pgd_range(&tlb, new_end, old_end, new_end,
 			vma->vm_next ? vma->vm_next->vm_start : 0);
 	} else {
 		/*
@@ -640,10 +640,10 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift)
 		 * have constraints on va-space that make this illegal (IA64) -
 		 * for the others its just a little faster.
 		 */
-		free_pgd_range(tlb, old_start, old_end, new_end,
+		free_pgd_range(&tlb, old_start, old_end, new_end,
 			vma->vm_next ? vma->vm_next->vm_start : 0);
 	}
-	tlb_finish_mmu(tlb, new_end, old_end);
+	tlb_finish_mmu(&tlb, new_end, old_end);
 
 	/*
 	 * Shrink the vma to just the new range.  Always succeeds.

include/asm-generic/tlb.h

@@ -5,6 +5,8 @@
  * Copyright 2001 Red Hat, Inc.
  * Based on code from mm/memory.c Copyright Linus Torvalds and others.
  *
+ * Copyright 2011 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com>
+ *
  * This program is free software; you can redistribute it and/or
  * modify it under the terms of the GNU General Public License
  * as published by the Free Software Foundation; either version
@@ -22,51 +24,71 @@
  * and page free order so much..
  */
 #ifdef CONFIG_SMP
-  #ifdef ARCH_FREE_PTR_NR
-    #define FREE_PTR_NR   ARCH_FREE_PTR_NR
-  #else
-    #define FREE_PTE_NR	506
-  #endif
   #define tlb_fast_mode(tlb) ((tlb)->nr == ~0U)
 #else
-  #define FREE_PTE_NR	1
   #define tlb_fast_mode(tlb) 1
 #endif
 
+/*
+ * If we can't allocate a page to make a big batch of page pointers
+ * to work on, then just handle a few from the on-stack structure.
+ */
+#define MMU_GATHER_BUNDLE	8
+
 /* struct mmu_gather is an opaque type used by the mm code for passing around
  * any data needed by arch specific code for tlb_remove_page.
  */
 struct mmu_gather {
 	struct mm_struct	*mm;
 	unsigned int		nr;	/* set to ~0U means fast mode */
+	unsigned int		max;	/* nr < max */
 	unsigned int		need_flush;/* Really unmapped some ptes? */
 	unsigned int		fullmm; /* non-zero means full mm flush */
-	struct page *		pages[FREE_PTE_NR];
+#ifdef HAVE_ARCH_MMU_GATHER
+	struct arch_mmu_gather	arch;
+#endif
+	struct page		**pages;
+	struct page		*local[MMU_GATHER_BUNDLE];
 };
 
-/* Users of the generic TLB shootdown code must declare this storage space. */
-DECLARE_PER_CPU(struct mmu_gather, mmu_gathers);
+static inline void __tlb_alloc_page(struct mmu_gather *tlb)
+{
+	unsigned long addr = __get_free_pages(GFP_NOWAIT | __GFP_NOWARN, 0);
+
+	if (addr) {
+		tlb->pages = (void *)addr;
+		tlb->max = PAGE_SIZE / sizeof(struct page *);
+	}
+}
 
 /* tlb_gather_mmu
- *	Return a pointer to an initialized struct mmu_gather.
+ *	Called to initialize an (on-stack) mmu_gather structure for page-table
+ *	tear-down from @mm. The @fullmm argument is used when @mm is without
+ *	users and we're going to destroy the full address space (exit/execve).
  */
-static inline struct mmu_gather *
-tlb_gather_mmu(struct mm_struct *mm, unsigned int full_mm_flush)
+static inline void
+tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, bool fullmm)
 {
-	struct mmu_gather *tlb = &get_cpu_var(mmu_gathers);
-
 	tlb->mm = mm;
 
-	/* Use fast mode if only one CPU is online */
-	tlb->nr = num_online_cpus() > 1 ? 0U : ~0U;
+	tlb->max = ARRAY_SIZE(tlb->local);
+	tlb->pages = tlb->local;
 
-	tlb->fullmm = full_mm_flush;
+	if (num_online_cpus() > 1) {
+		tlb->nr = 0;
+		__tlb_alloc_page(tlb);
+	} else /* Use fast mode if only one CPU is online */
+		tlb->nr = ~0U;
 
-	return tlb;
+	tlb->fullmm = fullmm;
+
+#ifdef HAVE_ARCH_MMU_GATHER
+	tlb->arch = ARCH_MMU_GATHER_INIT;
+#endif
 }
 
 static inline void
-tlb_flush_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
+tlb_flush_mmu(struct mmu_gather *tlb)
 {
 	if (!tlb->need_flush)
 		return;
@@ -75,6 +97,13 @@ tlb_flush_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
 	if (!tlb_fast_mode(tlb)) {
 		free_pages_and_swap_cache(tlb->pages, tlb->nr);
 		tlb->nr = 0;
+		/*
+		 * If we are using the local on-stack array of pages for MMU
+		 * gather, try allocating an off-stack array again as we have
+		 * recently freed pages.
+		 */
+		if (tlb->pages == tlb->local)
+			__tlb_alloc_page(tlb);
 	}
 }
 
@@ -85,29 +114,42 @@ tlb_flush_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
 static inline void
 tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end)
 {
-	tlb_flush_mmu(tlb, start, end);
+	tlb_flush_mmu(tlb);
 
 	/* keep the page table cache within bounds */
 	check_pgt_cache();
 
-	put_cpu_var(mmu_gathers);
+	if (tlb->pages != tlb->local)
+		free_pages((unsigned long)tlb->pages, 0);
 }
 
-/* tlb_remove_page
+/* __tlb_remove_page
  *	Must perform the equivalent to __free_pte(pte_get_and_clear(ptep)), while
  *	handling the additional races in SMP caused by other CPUs caching valid
- *	mappings in their TLBs.
+ *	mappings in their TLBs. Returns the number of free page slots left.
+ *	When out of page slots we must call tlb_flush_mmu().
  */
-static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
+static inline int __tlb_remove_page(struct mmu_gather *tlb, struct page *page)
 {
 	tlb->need_flush = 1;
 	if (tlb_fast_mode(tlb)) {
 		free_page_and_swap_cache(page);
-		return;
+		return 1; /* avoid calling tlb_flush_mmu() */
 	}
 	tlb->pages[tlb->nr++] = page;
-	if (tlb->nr >= FREE_PTE_NR)
-		tlb_flush_mmu(tlb, 0, 0);
+	VM_BUG_ON(tlb->nr > tlb->max);
+
+	return tlb->max - tlb->nr;
+}
+
+/* tlb_remove_page
+ *	Similar to __tlb_remove_page but will call tlb_flush_mmu() itself when
+ *	required.
+ */
+static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page)
+{
+	if (!__tlb_remove_page(tlb, page))
+		tlb_flush_mmu(tlb);
 }
 
 /**

include/linux/mm.h

@@ -906,7 +906,7 @@ int zap_vma_ptes(struct vm_area_struct *vma, unsigned long address,
 		unsigned long size);
 unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address,
 		unsigned long size, struct zap_details *);
-unsigned long unmap_vmas(struct mmu_gather **tlb,
+unsigned long unmap_vmas(struct mmu_gather *tlb,
 		struct vm_area_struct *start_vma, unsigned long start_addr,
 		unsigned long end_addr, unsigned long *nr_accounted,
 		struct zap_details *);

mm/memory.c

@@ -912,12 +912,13 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				long *zap_work, struct zap_details *details)
 {
 	struct mm_struct *mm = tlb->mm;
+	int force_flush = 0;
 	pte_t *pte;
 	spinlock_t *ptl;
 	int rss[NR_MM_COUNTERS];
 
 	init_rss_vec(rss);
-
+again:
 	pte = pte_offset_map_lock(mm, pmd, addr, &ptl);
 	arch_enter_lazy_mmu_mode();
 	do {
@@ -974,7 +975,9 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			page_remove_rmap(page);
 			if (unlikely(page_mapcount(page) < 0))
 				print_bad_pte(vma, addr, ptent, page);
-			tlb_remove_page(tlb, page);
+			force_flush = !__tlb_remove_page(tlb, page);
+			if (force_flush)
+				break;
 			continue;
 		}
 		/*
@@ -1001,6 +1004,18 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 	arch_leave_lazy_mmu_mode();
 	pte_unmap_unlock(pte - 1, ptl);
 
+	/*
+	 * mmu_gather ran out of room to batch pages, we break out of
+	 * the PTE lock to avoid doing the potential expensive TLB invalidate
+	 * and page-free while holding it.
+	 */
+	if (force_flush) {
+		force_flush = 0;
+		tlb_flush_mmu(tlb);
+		if (addr != end)
+			goto again;
+	}
+
 	return addr;
 }
 
@@ -1121,17 +1136,14 @@ static unsigned long unmap_page_range(struct mmu_gather *tlb,
  * ensure that any thus-far unmapped pages are flushed before unmap_vmas()
  * drops the lock and schedules.
  */
-unsigned long unmap_vmas(struct mmu_gather **tlbp,
+unsigned long unmap_vmas(struct mmu_gather *tlb,
 		struct vm_area_struct *vma, unsigned long start_addr,
 		unsigned long end_addr, unsigned long *nr_accounted,
 		struct zap_details *details)
 {
 	long zap_work = ZAP_BLOCK_SIZE;
-	unsigned long tlb_start = 0;	/* For tlb_finish_mmu */
-	int tlb_start_valid = 0;
 	unsigned long start = start_addr;
 	spinlock_t *i_mmap_lock = details? details->i_mmap_lock: NULL;
-	int fullmm = (*tlbp)->fullmm;
 	struct mm_struct *mm = vma->vm_mm;
 
 	mmu_notifier_invalidate_range_start(mm, start_addr, end_addr);
@@ -1152,11 +1164,6 @@ unsigned long unmap_vmas(struct mmu_gather **tlbp,
 			untrack_pfn_vma(vma, 0, 0);
 
 		while (start != end) {
-			if (!tlb_start_valid) {
-				tlb_start = start;
-				tlb_start_valid = 1;
-			}
-
 			if (unlikely(is_vm_hugetlb_page(vma))) {
 				/*
 				 * It is undesirable to test vma->vm_file as it
@@ -1177,7 +1184,7 @@ unsigned long unmap_vmas(struct mmu_gather **tlbp,
 
 				start = end;
 			} else
-				start = unmap_page_range(*tlbp, vma,
+				start = unmap_page_range(tlb, vma,
 						start, end, &zap_work, details);
 
 			if (zap_work > 0) {
@@ -1185,19 +1192,13 @@ unsigned long unmap_vmas(struct mmu_gather **tlbp,
 				break;
 			}
 
-			tlb_finish_mmu(*tlbp, tlb_start, start);
-
 			if (need_resched() ||
 				(i_mmap_lock && spin_needbreak(i_mmap_lock))) {
-				if (i_mmap_lock) {
-					*tlbp = NULL;
+				if (i_mmap_lock)
 					goto out;
-				}
 				cond_resched();
 			}
 
-			*tlbp = tlb_gather_mmu(vma->vm_mm, fullmm);
-			tlb_start_valid = 0;
 			zap_work = ZAP_BLOCK_SIZE;
 		}
 	}
@@ -1217,16 +1218,15 @@ unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address,
 		unsigned long size, struct zap_details *details)
 {
 	struct mm_struct *mm = vma->vm_mm;
-	struct mmu_gather *tlb;
+	struct mmu_gather tlb;
 	unsigned long end = address + size;
 	unsigned long nr_accounted = 0;
 
 	lru_add_drain();
-	tlb = tlb_gather_mmu(mm, 0);
+	tlb_gather_mmu(&tlb, mm, 0);
 	update_hiwater_rss(mm);
 	end = unmap_vmas(&tlb, vma, address, end, &nr_accounted, details);
-	if (tlb)
-		tlb_finish_mmu(tlb, address, end);
+	tlb_finish_mmu(&tlb, address, end);
 	return end;
 }
 

mm/mmap.c

@@ -1903,17 +1903,17 @@ static void unmap_region(struct mm_struct *mm,
 		unsigned long start, unsigned long end)
 {
 	struct vm_area_struct *next = prev? prev->vm_next: mm->mmap;
-	struct mmu_gather *tlb;
+	struct mmu_gather tlb;
 	unsigned long nr_accounted = 0;
 
 	lru_add_drain();
-	tlb = tlb_gather_mmu(mm, 0);
+	tlb_gather_mmu(&tlb, mm, 0);
 	update_hiwater_rss(mm);
 	unmap_vmas(&tlb, vma, start, end, &nr_accounted, NULL);
 	vm_unacct_memory(nr_accounted);
-	free_pgtables(tlb, vma, prev? prev->vm_end: FIRST_USER_ADDRESS,
-				 next? next->vm_start: 0);
-	tlb_finish_mmu(tlb, start, end);
+	free_pgtables(&tlb, vma, prev ? prev->vm_end : FIRST_USER_ADDRESS,
+				 next ? next->vm_start : 0);
+	tlb_finish_mmu(&tlb, start, end);
 }
 
 /*
@@ -2255,7 +2255,7 @@ EXPORT_SYMBOL(do_brk);
 /* Release all mmaps. */
 void exit_mmap(struct mm_struct *mm)
 {
-	struct mmu_gather *tlb;
+	struct mmu_gather tlb;
 	struct vm_area_struct *vma;
 	unsigned long nr_accounted = 0;
 	unsigned long end;
@@ -2280,14 +2280,14 @@ void exit_mmap(struct mm_struct *mm)
 
 	lru_add_drain();
 	flush_cache_mm(mm);
-	tlb = tlb_gather_mmu(mm, 1);
+	tlb_gather_mmu(&tlb, mm, 1);
 	/* update_hiwater_rss(mm) here? but nobody should be looking */
 	/* Use -1 here to ensure all VMAs in the mm are unmapped */
 	end = unmap_vmas(&tlb, vma, 0, -1, &nr_accounted, NULL);
 	vm_unacct_memory(nr_accounted);
 
-	free_pgtables(tlb, vma, FIRST_USER_ADDRESS, 0);
-	tlb_finish_mmu(tlb, 0, end);
+	free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, 0);
+	tlb_finish_mmu(&tlb, 0, end);
 
 	/*
 	 * Walk the list again, actually closing and freeing it,