X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=mm%2Fbootmem.c;h=4af15d0340adaa7bc9cded10d36302679d86f90a;hb=ada662f3eb6231ab27f5e6366d4e5c395d25edd3;hp=94ea612deccff5dba2e0f69ca7e57a3b5633221a;hpb=5f2809e69c7128f86316048221cf45146f69a4a0;p=linux-2.6

diff --git a/mm/bootmem.c b/mm/bootmem.c
index 94ea612dec..4af15d0340 100644
--- a/mm/bootmem.c
+++ b/mm/bootmem.c
@@ -80,7 +80,7 @@ static void __init link_bootmem(bootmem_data_t *bdata)
 		bootmem_data_t *ent;
 
 		ent = list_entry(iter, bootmem_data_t, list);
-		if (bdata->node_boot_start < ent->node_boot_start)
+		if (bdata->node_min_pfn < ent->node_min_pfn)
 			break;
 	}
 	list_add_tail(&bdata->list, iter);
@@ -96,7 +96,7 @@ static unsigned long __init init_bootmem_core(bootmem_data_t *bdata,
 
 	mminit_validate_memmodel_limits(&start, &end);
 	bdata->node_bootmem_map = phys_to_virt(PFN_PHYS(mapstart));
-	bdata->node_boot_start = PFN_PHYS(start);
+	bdata->node_min_pfn = start;
 	bdata->node_low_pfn = end;
 	link_bootmem(bdata);
 
@@ -151,7 +151,7 @@ static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata)
 	if (!bdata->node_bootmem_map)
 		return 0;
 
-	start = PFN_DOWN(bdata->node_boot_start);
+	start = bdata->node_min_pfn;
 	end = bdata->node_low_pfn;
 
 	/*
@@ -167,7 +167,7 @@ static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata)
 		unsigned long *map, idx, vec;
 
 		map = bdata->node_bootmem_map;
-		idx = start - PFN_DOWN(bdata->node_boot_start);
+		idx = start - bdata->node_min_pfn;
 		vec = ~map[idx / BITS_PER_LONG];
 
 		if (aligned && vec == ~0UL && start + BITS_PER_LONG < end) {
@@ -192,7 +192,7 @@ static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata)
 	}
 
 	page = virt_to_page(bdata->node_bootmem_map);
-	pages = bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start);
+	pages = bdata->node_low_pfn - bdata->node_min_pfn;
 	pages = bootmem_bootmap_pages(pages);
 	count += pages;
 	while (pages--)
@@ -225,47 +225,99 @@ unsigned long __init free_all_bootmem(void)
 	return free_all_bootmem_core(NODE_DATA(0)->bdata);
 }
 
-static void __init free_bootmem_core(bootmem_data_t *bdata, unsigned long addr,
-				     unsigned long size)
+static void __init __free(bootmem_data_t *bdata,
+			unsigned long sidx, unsigned long eidx)
+{
+	unsigned long idx;
+
+	bdebug("nid=%td start=%lx end=%lx\n", bdata - bootmem_node_data,
+		sidx + bdata->node_min_pfn,
+		eidx + bdata->node_min_pfn);
+
+	if (bdata->hint_idx > sidx)
+		bdata->hint_idx = sidx;
+
+	for (idx = sidx; idx < eidx; idx++)
+		if (!test_and_clear_bit(idx, bdata->node_bootmem_map))
+			BUG();
+}
+
+static int __init __reserve(bootmem_data_t *bdata, unsigned long sidx,
+			unsigned long eidx, int flags)
+{
+	unsigned long idx;
+	int exclusive = flags & BOOTMEM_EXCLUSIVE;
+
+	bdebug("nid=%td start=%lx end=%lx flags=%x\n",
+		bdata - bootmem_node_data,
+		sidx + bdata->node_min_pfn,
+		eidx + bdata->node_min_pfn,
+		flags);
+
+	for (idx = sidx; idx < eidx; idx++)
+		if (test_and_set_bit(idx, bdata->node_bootmem_map)) {
+			if (exclusive) {
+				__free(bdata, sidx, idx);
+				return -EBUSY;
+			}
+			bdebug("silent double reserve of PFN %lx\n",
+				idx + bdata->node_min_pfn);
+		}
+	return 0;
+}
+
+static int __init mark_bootmem_node(bootmem_data_t *bdata,
+				unsigned long start, unsigned long end,
+				int reserve, int flags)
 {
 	unsigned long sidx, eidx;
-	unsigned long i;
 
-	BUG_ON(!size);
+	bdebug("nid=%td start=%lx end=%lx reserve=%d flags=%x\n",
+		bdata - bootmem_node_data, start, end, reserve, flags);
 
-	/* out range */
-	if (addr + size < bdata->node_boot_start ||
-		PFN_DOWN(addr) > bdata->node_low_pfn)
-		return;
-	/*
-	 * round down end of usable mem, partially free pages are
-	 * considered reserved.
-	 */
+	BUG_ON(start < bdata->node_min_pfn);
+	BUG_ON(end > bdata->node_low_pfn);
 
-	if (addr >= bdata->node_boot_start &&
-			PFN_DOWN(addr - bdata->node_boot_start) < bdata->hint_idx)
-		bdata->hint_idx = PFN_DOWN(addr - bdata->node_boot_start);
+	sidx = start - bdata->node_min_pfn;
+	eidx = end - bdata->node_min_pfn;
 
-	/*
-	 * Round up to index to the range.
-	 */
-	if (PFN_UP(addr) > PFN_DOWN(bdata->node_boot_start))
-		sidx = PFN_UP(addr) - PFN_DOWN(bdata->node_boot_start);
+	if (reserve)
+		return __reserve(bdata, sidx, eidx, flags);
 	else
-		sidx = 0;
+		__free(bdata, sidx, eidx);
+	return 0;
+}
 
-	eidx = PFN_DOWN(addr + size - bdata->node_boot_start);
-	if (eidx > bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start))
-		eidx = bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start);
+static int __init mark_bootmem(unsigned long start, unsigned long end,
+				int reserve, int flags)
+{
+	unsigned long pos;
+	bootmem_data_t *bdata;
 
-	bdebug("nid=%td start=%lx end=%lx\n", bdata - bootmem_node_data,
-		sidx + PFN_DOWN(bdata->node_boot_start),
-		eidx + PFN_DOWN(bdata->node_boot_start));
+	pos = start;
+	list_for_each_entry(bdata, &bdata_list, list) {
+		int err;
+		unsigned long max;
 
-	for (i = sidx; i < eidx; i++) {
-		if (unlikely(!test_and_clear_bit(i, bdata->node_bootmem_map)))
-			BUG();
+		if (pos < bdata->node_min_pfn ||
+		    pos >= bdata->node_low_pfn) {
+			BUG_ON(pos != start);
+			continue;
+		}
+
+		max = min(bdata->node_low_pfn, end);
+
+		err = mark_bootmem_node(bdata, pos, max, reserve, flags);
+		if (reserve && err) {
+			mark_bootmem(start, pos, 0, 0);
+			return err;
+		}
+
+		if (max == end)
+			return 0;
+		pos = bdata->node_low_pfn;
 	}
+	BUG();
 }
 
 /**
@@ -276,12 +328,17 @@ static void __init free_bootmem_core(bootmem_data_t *bdata, unsigned long addr,
  *
  * Partial pages will be considered reserved and left as they are.
  *
- * Only physical pages that actually reside on @pgdat are marked.
+ * The range must reside completely on the specified node.
  */
 void __init free_bootmem_node(pg_data_t *pgdat, unsigned long physaddr,
 			      unsigned long size)
 {
-	free_bootmem_core(pgdat->bdata, physaddr, size);
+	unsigned long start, end;
+
+	start = PFN_UP(physaddr);
+	end = PFN_DOWN(physaddr + size);
+
+	mark_bootmem_node(pgdat->bdata, start, end, 0, 0);
 }
 
 /**
@@ -291,92 +348,16 @@ void __init free_bootmem_node(pg_data_t *pgdat, unsigned long physaddr,
  *
  * Partial pages will be considered reserved and left as they are.
  *
- * All physical pages within the range are marked, no matter what
- * node they reside on.
+ * The range must be contiguous but may span node boundaries.
  */
 void __init free_bootmem(unsigned long addr, unsigned long size)
 {
-	bootmem_data_t *bdata;
-	list_for_each_entry(bdata, &bdata_list, list)
-		free_bootmem_core(bdata, addr, size);
-}
-
-/*
- * Marks a particular physical memory range as unallocatable. Usable RAM
- * might be used for boot-time allocations - or it might get added
- * to the free page pool later on.
- */
-static int __init can_reserve_bootmem_core(bootmem_data_t *bdata,
-			unsigned long addr, unsigned long size, int flags)
-{
-	unsigned long sidx, eidx;
-	unsigned long i;
-
-	BUG_ON(!size);
-
-	/* out of range, don't hold other */
-	if (addr + size < bdata->node_boot_start ||
-		PFN_DOWN(addr) > bdata->node_low_pfn)
-		return 0;
-
-	/*
-	 * Round up to index to the range.
-	 */
-	if (addr > bdata->node_boot_start)
-		sidx= PFN_DOWN(addr - bdata->node_boot_start);
-	else
-		sidx = 0;
+	unsigned long start, end;
 
-	eidx = PFN_UP(addr + size - bdata->node_boot_start);
-	if (eidx > bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start))
-		eidx = bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start);
-
-	for (i = sidx; i < eidx; i++) {
-		if (test_bit(i, bdata->node_bootmem_map)) {
-			if (flags & BOOTMEM_EXCLUSIVE)
-				return -EBUSY;
-		}
-	}
-
-	return 0;
+	start = PFN_UP(addr);
+	end = PFN_DOWN(addr + size);
 
-}
-
-static void __init reserve_bootmem_core(bootmem_data_t *bdata,
-			unsigned long addr, unsigned long size, int flags)
-{
-	unsigned long sidx, eidx;
-	unsigned long i;
-
-	BUG_ON(!size);
-
-	/* out of range */
-	if (addr + size < bdata->node_boot_start ||
-		PFN_DOWN(addr) > bdata->node_low_pfn)
-		return;
-
-	/*
-	 * Round up to index to the range.
-	 */
-	if (addr > bdata->node_boot_start)
-		sidx= PFN_DOWN(addr - bdata->node_boot_start);
-	else
-		sidx = 0;
-
-	eidx = PFN_UP(addr + size - bdata->node_boot_start);
-	if (eidx > bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start))
-		eidx = bdata->node_low_pfn - PFN_DOWN(bdata->node_boot_start);
-
-	bdebug("nid=%td start=%lx end=%lx flags=%x\n",
-		bdata - bootmem_node_data,
-		sidx + PFN_DOWN(bdata->node_boot_start),
-		eidx + PFN_DOWN(bdata->node_boot_start),
-		flags);
-
-	for (i = sidx; i < eidx; i++)
-		if (test_and_set_bit(i, bdata->node_bootmem_map))
-			bdebug("hm, page %lx reserved twice.\n",
-				PFN_DOWN(bdata->node_boot_start) + i);
+	mark_bootmem(start, end, 0, 0);
 }
 
 /**
@@ -388,18 +369,17 @@ static void __init reserve_bootmem_core(bootmem_data_t *bdata,
  *
  * Partial pages will be reserved.
  *
- * Only physical pages that actually reside on @pgdat are marked.
+ * The range must reside completely on the specified node.
  */
 int __init reserve_bootmem_node(pg_data_t *pgdat, unsigned long physaddr,
 				 unsigned long size, int flags)
 {
-	int ret;
+	unsigned long start, end;
 
-	ret = can_reserve_bootmem_core(pgdat->bdata, physaddr, size, flags);
-	if (ret < 0)
-		return -ENOMEM;
-	reserve_bootmem_core(pgdat->bdata, physaddr, size, flags);
-	return 0;
+	start = PFN_DOWN(physaddr);
+	end = PFN_UP(physaddr + size);
+
+	return mark_bootmem_node(pgdat->bdata, start, end, 1, flags);
 }
 
 #ifndef CONFIG_HAVE_ARCH_BOOTMEM_NODE
@@ -411,24 +391,17 @@ int __init reserve_bootmem_node(pg_data_t *pgdat, unsigned long physaddr,
  *
  * Partial pages will be reserved.
  *
- * All physical pages within the range are marked, no matter what
- * node they reside on.
+ * The range must be contiguous but may span node boundaries.
  */
 int __init reserve_bootmem(unsigned long addr, unsigned long size,
 			    int flags)
 {
-	bootmem_data_t *bdata;
-	int ret;
+	unsigned long start, end;
 
-	list_for_each_entry(bdata, &bdata_list, list) {
-		ret = can_reserve_bootmem_core(bdata, addr, size, flags);
-		if (ret < 0)
-			return ret;
-	}
-	list_for_each_entry(bdata, &bdata_list, list)
-		reserve_bootmem_core(bdata, addr, size, flags);
+	start = PFN_DOWN(addr);
+	end = PFN_UP(addr + size);
 
-	return 0;
+	return mark_bootmem(start, end, 1, flags);
 }
 #endif /* !CONFIG_HAVE_ARCH_BOOTMEM_NODE */
 
@@ -436,6 +409,7 @@ static void * __init alloc_bootmem_core(struct bootmem_data *bdata,
 				unsigned long size, unsigned long align,
 				unsigned long goal, unsigned long limit)
 {
+	unsigned long fallback = 0;
 	unsigned long min, max, start, sidx, midx, step;
 
 	BUG_ON(!size);
@@ -449,7 +423,7 @@ static void * __init alloc_bootmem_core(struct bootmem_data *bdata,
 		bdata - bootmem_node_data, size, PAGE_ALIGN(size) >> PAGE_SHIFT,
 		align, goal, limit);
 
-	min = PFN_DOWN(bdata->node_boot_start);
+	min = bdata->node_min_pfn;
 	max = bdata->node_low_pfn;
 
 	goal >>= PAGE_SHIFT;
@@ -467,12 +441,15 @@ static void * __init alloc_bootmem_core(struct bootmem_data *bdata,
 	else
 		start = ALIGN(min, step);
 
-	sidx = start - PFN_DOWN(bdata->node_boot_start);
-	midx = max - PFN_DOWN(bdata->node_boot_start);
+	sidx = start - bdata->node_min_pfn;;
+	midx = max - bdata->node_min_pfn;
 
 	if (bdata->hint_idx > sidx) {
-		/* Make sure we retry on failure */
-		goal = 1;
+		/*
+		 * Handle the valid case of sidx being zero and still
+		 * catch the fallback below.
+		 */
+		fallback = sidx + 1;
 		sidx = ALIGN(bdata->hint_idx, step);
 	}
 
@@ -511,20 +488,49 @@ find_block:
 		/*
 		 * Reserve the area now:
 		 */
-		for (i = PFN_DOWN(start_off) + merge;
-				i < PFN_UP(end_off); i++)
-			if (test_and_set_bit(i, bdata->node_bootmem_map))
-				BUG();
+		if (__reserve(bdata, PFN_DOWN(start_off) + merge,
+				PFN_UP(end_off), BOOTMEM_EXCLUSIVE))
+			BUG();
 
-		region = phys_to_virt(bdata->node_boot_start + start_off);
+		region = phys_to_virt(PFN_PHYS(bdata->node_min_pfn) +
+				start_off);
 		memset(region, 0, size);
 		return region;
 	}
 
+	if (fallback) {
+		sidx = ALIGN(fallback - 1, step);
+		fallback = 0;
+		goto find_block;
+	}
+
+	return NULL;
+}
+
+static void * __init ___alloc_bootmem_nopanic(unsigned long size,
+					unsigned long align,
+					unsigned long goal,
+					unsigned long limit)
+{
+	bootmem_data_t *bdata;
+
+restart:
+	list_for_each_entry(bdata, &bdata_list, list) {
+		void *region;
+
+		if (goal && bdata->node_low_pfn <= PFN_DOWN(goal))
+			continue;
+		if (limit && bdata->node_min_pfn >= PFN_DOWN(limit))
+			break;
+
+		region = alloc_bootmem_core(bdata, size, align, goal, limit);
+		if (region)
+			return region;
+	}
+
 	if (goal) {
 		goal = 0;
-		sidx = 0;
-		goto find_block;
+		goto restart;
 	}
 
 	return NULL;
@@ -544,16 +550,23 @@ find_block:
  * Returns NULL on failure.
  */
 void * __init __alloc_bootmem_nopanic(unsigned long size, unsigned long align,
-				      unsigned long goal)
+					unsigned long goal)
 {
-	bootmem_data_t *bdata;
-	void *ptr;
+	return ___alloc_bootmem_nopanic(size, align, goal, 0);
+}
 
-	list_for_each_entry(bdata, &bdata_list, list) {
-		ptr = alloc_bootmem_core(bdata, size, align, goal, 0);
-		if (ptr)
-			return ptr;
-	}
+static void * __init ___alloc_bootmem(unsigned long size, unsigned long align,
+					unsigned long goal, unsigned long limit)
+{
+	void *mem = ___alloc_bootmem_nopanic(size, align, goal, limit);
+
+	if (mem)
+		return mem;
+	/*
+	 * Whoops, we cannot satisfy the allocation request.
+	 */
+	printk(KERN_ALERT "bootmem alloc of %lu bytes failed!\n", size);
+	panic("Out of memory");
 	return NULL;
 }
 
@@ -573,16 +586,20 @@ void * __init __alloc_bootmem_nopanic(unsigned long size, unsigned long align,
 void * __init __alloc_bootmem(unsigned long size, unsigned long align,
 			      unsigned long goal)
 {
-	void *mem = __alloc_bootmem_nopanic(size,align,goal);
+	return ___alloc_bootmem(size, align, goal, 0);
+}
 
-	if (mem)
-		return mem;
-	/*
-	 * Whoops, we cannot satisfy the allocation request.
-	 */
-	printk(KERN_ALERT "bootmem alloc of %lu bytes failed!\n", size);
-	panic("Out of memory");
-	return NULL;
+static void * __init ___alloc_bootmem_node(bootmem_data_t *bdata,
+				unsigned long size, unsigned long align,
+				unsigned long goal, unsigned long limit)
+{
+	void *ptr;
+
+	ptr = alloc_bootmem_core(bdata, size, align, goal, limit);
+	if (ptr)
+		return ptr;
+
+	return ___alloc_bootmem(size, align, goal, limit);
 }
 
 /**
@@ -603,13 +620,7 @@ void * __init __alloc_bootmem(unsigned long size, unsigned long align,
 void * __init __alloc_bootmem_node(pg_data_t *pgdat, unsigned long size,
 				   unsigned long align, unsigned long goal)
 {
-	void *ptr;
-
-	ptr = alloc_bootmem_core(pgdat->bdata, size, align, goal, 0);
-	if (ptr)
-		return ptr;
-
-	return __alloc_bootmem(size, align, goal);
+	return ___alloc_bootmem_node(pgdat->bdata, size, align, goal, 0);
 }
 
 #ifdef CONFIG_SPARSEMEM
@@ -623,30 +634,15 @@ void * __init __alloc_bootmem_node(pg_data_t *pgdat, unsigned long size,
 void * __init alloc_bootmem_section(unsigned long size,
 				    unsigned long section_nr)
 {
-	void *ptr;
-	unsigned long limit, goal, start_nr, end_nr, pfn;
-	struct pglist_data *pgdat;
+	bootmem_data_t *bdata;
+	unsigned long pfn, goal, limit;
 
 	pfn = section_nr_to_pfn(section_nr);
-	goal = PFN_PHYS(pfn);
-	limit = PFN_PHYS(section_nr_to_pfn(section_nr + 1)) - 1;
-	pgdat = NODE_DATA(early_pfn_to_nid(pfn));
-	ptr = alloc_bootmem_core(pgdat->bdata, size, SMP_CACHE_BYTES, goal,
-				limit);
+	goal = pfn << PAGE_SHIFT;
+	limit = section_nr_to_pfn(section_nr + 1) << PAGE_SHIFT;
+	bdata = &bootmem_node_data[early_pfn_to_nid(pfn)];
 
-	if (!ptr)
-		return NULL;
-
-	start_nr = pfn_to_section_nr(PFN_DOWN(__pa(ptr)));
-	end_nr = pfn_to_section_nr(PFN_DOWN(__pa(ptr) + size));
-	if (start_nr != section_nr || end_nr != section_nr) {
-		printk(KERN_WARNING "alloc_bootmem failed on section %ld.\n",
-		       section_nr);
-		free_bootmem_core(pgdat->bdata, __pa(ptr), size);
-		ptr = NULL;
-	}
-
-	return ptr;
+	return alloc_bootmem_core(bdata, size, SMP_CACHE_BYTES, goal, limit);
 }
 #endif
 
@@ -682,22 +678,7 @@ void * __init __alloc_bootmem_node_nopanic(pg_data_t *pgdat, unsigned long size,
 void * __init __alloc_bootmem_low(unsigned long size, unsigned long align,
 				  unsigned long goal)
 {
-	bootmem_data_t *bdata;
-	void *ptr;
-
-	list_for_each_entry(bdata, &bdata_list, list) {
-		ptr = alloc_bootmem_core(bdata, size, align, goal,
-					ARCH_LOW_ADDRESS_LIMIT);
-		if (ptr)
-			return ptr;
-	}
-
-	/*
-	 * Whoops, we cannot satisfy the allocation request.
-	 */
-	printk(KERN_ALERT "low bootmem alloc of %lu bytes failed!\n", size);
-	panic("Out of low memory");
-	return NULL;
+	return ___alloc_bootmem(size, align, goal, ARCH_LOW_ADDRESS_LIMIT);
 }
 
 /**
@@ -718,6 +699,6 @@ void * __init __alloc_bootmem_low(unsigned long size, unsigned long align,
 void * __init __alloc_bootmem_low_node(pg_data_t *pgdat, unsigned long size,
 				       unsigned long align, unsigned long goal)
 {
-	return alloc_bootmem_core(pgdat->bdata, size, align, goal,
-				ARCH_LOW_ADDRESS_LIMIT);
+	return ___alloc_bootmem_node(pgdat->bdata, size, align,
+				goal, ARCH_LOW_ADDRESS_LIMIT);
 }