x86, mm: introduce vmem_altmap to augment vmemmap_populate()

In support of providing struct page for large persistent memory
capacities, use struct vmem_altmap to change the default policy for
allocating memory for the memmap array.  The default vmemmap_populate()
allocates page table storage area from the page allocator.  Given
persistent memory capacities relative to DRAM it may not be feasible to
store the memmap in 'System Memory'.  Instead vmem_altmap represents
pre-allocated "device pages" to satisfy vmemmap_alloc_block_buf()
requests.

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Reported-by: kbuild test robot <lkp@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

arch/x86/mm/init_64.c

@@ -30,6 +30,7 @@
 #include <linux/module.h>
 #include <linux/memory.h>
 #include <linux/memory_hotplug.h>
+#include <linux/memremap.h>
 #include <linux/nmi.h>
 #include <linux/gfp.h>
 #include <linux/kcore.h>
@@ -714,6 +715,12 @@ static void __meminit free_pagetable(struct page *page, int order)
 {
 	unsigned long magic;
 	unsigned int nr_pages = 1 << order;
+	struct vmem_altmap *altmap = to_vmem_altmap((unsigned long) page);
+
+	if (altmap) {
+		vmem_altmap_free(altmap, nr_pages);
+		return;
+	}
 
 	/* bootmem page has reserved flag */
 	if (PageReserved(page)) {
@@ -1017,13 +1024,19 @@ int __ref arch_remove_memory(u64 start, u64 size)
 {
 	unsigned long start_pfn = start >> PAGE_SHIFT;
 	unsigned long nr_pages = size >> PAGE_SHIFT;
+	struct page *page = pfn_to_page(start_pfn);
+	struct vmem_altmap *altmap;
 	struct zone *zone;
 	int ret;
 
-	zone = page_zone(pfn_to_page(start_pfn));
-	kernel_physical_mapping_remove(start, start + size);
+	/* With altmap the first mapped page is offset from @start */
+	altmap = to_vmem_altmap((unsigned long) page);
+	if (altmap)
+		page += vmem_altmap_offset(altmap);
+	zone = page_zone(page);
 	ret = __remove_pages(zone, start_pfn, nr_pages);
 	WARN_ON_ONCE(ret);
+	kernel_physical_mapping_remove(start, start + size);
 
 	return ret;
 }
@@ -1235,7 +1248,7 @@ static void __meminitdata *p_start, *p_end;
 static int __meminitdata node_start;
 
 static int __meminit vmemmap_populate_hugepages(unsigned long start,
-						unsigned long end, int node)
+		unsigned long end, int node, struct vmem_altmap *altmap)
 {
 	unsigned long addr;
 	unsigned long next;
@@ -1258,7 +1271,7 @@ static int __meminit vmemmap_populate_hugepages(unsigned long start,
 		if (pmd_none(*pmd)) {
 			void *p;
 
-			p = vmemmap_alloc_block_buf(PMD_SIZE, node);
+			p = __vmemmap_alloc_block_buf(PMD_SIZE, node, altmap);
 			if (p) {
 				pte_t entry;
 
@@ -1279,7 +1292,8 @@ static int __meminit vmemmap_populate_hugepages(unsigned long start,
 				addr_end = addr + PMD_SIZE;
 				p_end = p + PMD_SIZE;
 				continue;
-			}
+			} else if (altmap)
+				return -ENOMEM; /* no fallback */
 		} else if (pmd_large(*pmd)) {
 			vmemmap_verify((pte_t *)pmd, node, addr, next);
 			continue;
@@ -1293,11 +1307,16 @@ static int __meminit vmemmap_populate_hugepages(unsigned long start,
 
 int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node)
 {
+	struct vmem_altmap *altmap = to_vmem_altmap(start);
 	int err;
 
 	if (cpu_has_pse)
-		err = vmemmap_populate_hugepages(start, end, node);
-	else
+		err = vmemmap_populate_hugepages(start, end, node, altmap);
+	else if (altmap) {
+		pr_err_once("%s: no cpu support for altmap allocations\n",
+				__func__);
+		err = -ENOMEM;
+	} else
 		err = vmemmap_populate_basepages(start, end, node);
 	if (!err)
 		sync_global_pgds(start, end - 1, 0);

drivers/nvdimm/pmem.c

@@ -178,7 +178,8 @@ static struct pmem_device *pmem_alloc(struct device *dev,
 
 	pmem->pfn_flags = PFN_DEV;
 	if (pmem_should_map_pages(dev)) {
-		pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, res);
+		pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, res,
+				NULL);
 		pmem->pfn_flags |= PFN_MAP;
 	} else
 		pmem->virt_addr = (void __pmem *) devm_memremap(dev,
@@ -387,7 +388,8 @@ static int nvdimm_namespace_attach_pfn(struct nd_namespace_common *ndns)
 	/* establish pfn range for lookup, and switch to direct map */
 	pmem = dev_get_drvdata(dev);
 	devm_memunmap(dev, (void __force *) pmem->virt_addr);
-	pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, &nsio->res);
+	pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, &nsio->res,
+			NULL);
 	pmem->pfn_flags |= PFN_MAP;
 	if (IS_ERR(pmem->virt_addr)) {
 		rc = PTR_ERR(pmem->virt_addr);

include/linux/memory_hotplug.h

@@ -275,7 +275,8 @@ extern int offline_pages(unsigned long start_pfn, unsigned long nr_pages);
 extern bool is_memblock_offlined(struct memory_block *mem);
 extern void remove_memory(int nid, u64 start, u64 size);
 extern int sparse_add_one_section(struct zone *zone, unsigned long start_pfn);
-extern void sparse_remove_one_section(struct zone *zone, struct mem_section *ms);
+extern void sparse_remove_one_section(struct zone *zone, struct mem_section *ms,
+		unsigned long map_offset);
 extern struct page *sparse_decode_mem_map(unsigned long coded_mem_map,
 					  unsigned long pnum);
 

include/linux/memremap.h

@@ -4,21 +4,53 @@
 
 struct resource;
 struct device;
+
+/**
+ * struct vmem_altmap - pre-allocated storage for vmemmap_populate
+ * @base_pfn: base of the entire dev_pagemap mapping
+ * @reserve: pages mapped, but reserved for driver use (relative to @base)
+ * @free: free pages set aside in the mapping for memmap storage
+ * @align: pages reserved to meet allocation alignments
+ * @alloc: track pages consumed, private to vmemmap_populate()
+ */
+struct vmem_altmap {
+	const unsigned long base_pfn;
+	const unsigned long reserve;
+	unsigned long free;
+	unsigned long align;
+	unsigned long alloc;
+};
+
+unsigned long vmem_altmap_offset(struct vmem_altmap *altmap);
+void vmem_altmap_free(struct vmem_altmap *altmap, unsigned long nr_pfns);
+
+#if defined(CONFIG_SPARSEMEM_VMEMMAP) && defined(CONFIG_ZONE_DEVICE)
+struct vmem_altmap *to_vmem_altmap(unsigned long memmap_start);
+#else
+static inline struct vmem_altmap *to_vmem_altmap(unsigned long memmap_start)
+{
+	return NULL;
+}
+#endif
+
 /**
  * struct dev_pagemap - metadata for ZONE_DEVICE mappings
+ * @altmap: pre-allocated/reserved memory for vmemmap allocations
  * @dev: host device of the mapping for debug
  */
 struct dev_pagemap {
-	/* TODO: vmem_altmap and percpu_ref count */
+	struct vmem_altmap *altmap;
+	const struct resource *res;
 	struct device *dev;
 };
 
 #ifdef CONFIG_ZONE_DEVICE
-void *devm_memremap_pages(struct device *dev, struct resource *res);
+void *devm_memremap_pages(struct device *dev, struct resource *res,
+		struct vmem_altmap *altmap);
 struct dev_pagemap *find_dev_pagemap(resource_size_t phys);
 #else
 static inline void *devm_memremap_pages(struct device *dev,
-		struct resource *res)
+		struct resource *res, struct vmem_altmap *altmap)
 {
 	/*
 	 * Fail attempts to call devm_memremap_pages() without
@@ -34,5 +66,4 @@ static inline struct dev_pagemap *find_dev_pagemap(resource_size_t phys)
 	return NULL;
 }
 #endif
-
 #endif /* _LINUX_MEMREMAP_H_ */

include/linux/mm.h

@@ -2217,7 +2217,14 @@ pud_t *vmemmap_pud_populate(pgd_t *pgd, unsigned long addr, int node);
 pmd_t *vmemmap_pmd_populate(pud_t *pud, unsigned long addr, int node);
 pte_t *vmemmap_pte_populate(pmd_t *pmd, unsigned long addr, int node);
 void *vmemmap_alloc_block(unsigned long size, int node);
-void *vmemmap_alloc_block_buf(unsigned long size, int node);
+struct vmem_altmap;
+void *__vmemmap_alloc_block_buf(unsigned long size, int node,
+		struct vmem_altmap *altmap);
+static inline void *vmemmap_alloc_block_buf(unsigned long size, int node)
+{
+	return __vmemmap_alloc_block_buf(size, node, NULL);
+}
+
 void vmemmap_verify(pte_t *, int, unsigned long, unsigned long);
 int vmemmap_populate_basepages(unsigned long start, unsigned long end,
 			       int node);

kernel/memremap.c

@@ -166,6 +166,7 @@ struct page_map {
 	struct resource res;
 	struct percpu_ref *ref;
 	struct dev_pagemap pgmap;
+	struct vmem_altmap altmap;
 };
 
 static void pgmap_radix_release(struct resource *res)
@@ -183,13 +184,16 @@ static void devm_memremap_pages_release(struct device *dev, void *data)
 	struct page_map *page_map = data;
 	struct resource *res = &page_map->res;
 	resource_size_t align_start, align_size;
+	struct dev_pagemap *pgmap = &page_map->pgmap;
 
 	pgmap_radix_release(res);
 
 	/* pages are dead and unused, undo the arch mapping */
 	align_start = res->start & ~(SECTION_SIZE - 1);
 	align_size = ALIGN(resource_size(res), SECTION_SIZE);
 	arch_remove_memory(align_start, align_size);
+	dev_WARN_ONCE(dev, pgmap->altmap && pgmap->altmap->alloc,
+			"%s: failed to free all reserved pages\n", __func__);
 }
 
 /* assumes rcu_read_lock() held at entry */
@@ -203,11 +207,23 @@ struct dev_pagemap *find_dev_pagemap(resource_size_t phys)
 	return page_map ? &page_map->pgmap : NULL;
 }
 
-void *devm_memremap_pages(struct device *dev, struct resource *res)
+/**
+ * devm_memremap_pages - remap and provide memmap backing for the given resource
+ * @dev: hosting device for @res
+ * @res: "host memory" address range
+ * @altmap: optional descriptor for allocating the memmap from @res
+ *
+ * Note, the expectation is that @res is a host memory range that could
+ * feasibly be treated as a "System RAM" range, i.e. not a device mmio
+ * range, but this is not enforced.
+ */
+void *devm_memremap_pages(struct device *dev, struct resource *res,
+		struct vmem_altmap *altmap)
 {
 	int is_ram = region_intersects(res->start, resource_size(res),
 			"System RAM");
 	resource_size_t key, align_start, align_size;
+	struct dev_pagemap *pgmap;
 	struct page_map *page_map;
 	int error, nid;
 
@@ -220,14 +236,27 @@ void *devm_memremap_pages(struct device *dev, struct resource *res)
 	if (is_ram == REGION_INTERSECTS)
 		return __va(res->start);
 
+	if (altmap && !IS_ENABLED(CONFIG_SPARSEMEM_VMEMMAP)) {
+		dev_err(dev, "%s: altmap requires CONFIG_SPARSEMEM_VMEMMAP=y\n",
+				__func__);
+		return ERR_PTR(-ENXIO);
+	}
+
 	page_map = devres_alloc_node(devm_memremap_pages_release,
 			sizeof(*page_map), GFP_KERNEL, dev_to_node(dev));
 	if (!page_map)
 		return ERR_PTR(-ENOMEM);
+	pgmap = &page_map->pgmap;
 
 	memcpy(&page_map->res, res, sizeof(*res));
 
-	page_map->pgmap.dev = dev;
+	pgmap->dev = dev;
+	if (altmap) {
+		memcpy(&page_map->altmap, altmap, sizeof(*altmap));
+		pgmap->altmap = &page_map->altmap;
+	}
+	pgmap->res = &page_map->res;
+
 	mutex_lock(&pgmap_lock);
 	error = 0;
 	for (key = res->start; key <= res->end; key += SECTION_SIZE) {
@@ -273,4 +302,43 @@ void *devm_memremap_pages(struct device *dev, struct resource *res)
 	return ERR_PTR(error);
 }
 EXPORT_SYMBOL(devm_memremap_pages);
+
+unsigned long vmem_altmap_offset(struct vmem_altmap *altmap)
+{
+	/* number of pfns from base where pfn_to_page() is valid */
+	return altmap->reserve + altmap->free;
+}
+
+void vmem_altmap_free(struct vmem_altmap *altmap, unsigned long nr_pfns)
+{
+	altmap->alloc -= nr_pfns;
+}
+
+#ifdef CONFIG_SPARSEMEM_VMEMMAP
+struct vmem_altmap *to_vmem_altmap(unsigned long memmap_start)
+{
+	/*
+	 * 'memmap_start' is the virtual address for the first "struct
+	 * page" in this range of the vmemmap array.  In the case of
+	 * CONFIG_SPARSE_VMEMMAP a page_to_pfn conversion is simple
+	 * pointer arithmetic, so we can perform this to_vmem_altmap()
+	 * conversion without concern for the initialization state of
+	 * the struct page fields.
+	 */
+	struct page *page = (struct page *) memmap_start;
+	struct dev_pagemap *pgmap;
+
+	/*
+	 * Uncoditionally retrieve a dev_pagemap associated with the
+	 * given physical address, this is only for use in the
+	 * arch_{add|remove}_memory() for setting up and tearing down
+	 * the memmap.
+	 */
+	rcu_read_lock();
+	pgmap = find_dev_pagemap(__pfn_to_phys(page_to_pfn(page)));
+	rcu_read_unlock();
+
+	return pgmap ? pgmap->altmap : NULL;
+}
+#endif /* CONFIG_SPARSEMEM_VMEMMAP */
 #endif /* CONFIG_ZONE_DEVICE */