mm/memblock: expose only miminal interface to add/walk physmem

"physmem" in the memblock allocator is somewhat weird: it's not actually
used for allocation, it's simply information collected during boot, which
describes the unmodified physical memory map at boot time, without any
standby/hotplugged memory. It's only used on s390 and is currently the
only reason s390 keeps using CONFIG_ARCH_KEEP_MEMBLOCK.

Physmem isn't numa aware and current users don't specify any flags. Let's
hide it from the user, exposing only for_each_physmem(), and simplify. The
interface for physmem is now really minimalistic:
- memblock_physmem_add() to add ranges
- for_each_physmem() / __next_physmem_range() to walk physmem ranges

Don't place it into an __init section and don't discard it without
CONFIG_ARCH_KEEP_MEMBLOCK. As we're reusing __next_mem_range(), remove
the __meminit notifier to avoid section mismatch warnings once
CONFIG_ARCH_KEEP_MEMBLOCK is no longer used with
CONFIG_HAVE_MEMBLOCK_PHYS_MAP.

While fixing up the documentation, sneak in some related cleanups. We can
stop setting CONFIG_ARCH_KEEP_MEMBLOCK for s390 next.

Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Mike Rapoport <rppt@linux.ibm.com>
Message-Id: <20200701141830.18749-2-david@redhat.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>

arch/s390/kernel/crash_dump.c

@@ -549,8 +549,7 @@ static int get_mem_chunk_cnt(void)
 	int cnt = 0;
 	u64 idx;
 
-	for_each_mem_range(idx, &memblock.physmem, &oldmem_type, NUMA_NO_NODE,
+	for_each_physmem_range(idx, &oldmem_type, NULL, NULL)
-			   MEMBLOCK_NONE, NULL, NULL, NULL)
 		cnt++;
 	return cnt;
 }
@@ -563,8 +562,7 @@ static void loads_init(Elf64_Phdr *phdr, u64 loads_offset)
 	phys_addr_t start, end;
 	u64 idx;
 
-	for_each_mem_range(idx, &memblock.physmem, &oldmem_type, NUMA_NO_NODE,
+	for_each_physmem_range(idx, &oldmem_type, &start, &end) {
-			   MEMBLOCK_NONE, &start, &end, NULL) {
 		phdr->p_filesz = end - start;
 		phdr->p_type = PT_LOAD;
 		phdr->p_offset = start;

include/linux/memblock.h

@@ -77,16 +77,12 @@ struct memblock_type {
  * @current_limit: physical address of the current allocation limit
  * @memory: usable memory regions
  * @reserved: reserved memory regions
- * @physmem: all physical memory
  */
 struct memblock {
 	bool bottom_up;  /* is bottom up direction? */
 	phys_addr_t current_limit;
 	struct memblock_type memory;
 	struct memblock_type reserved;
-#ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP
-	struct memblock_type physmem;
-#endif
 };
 
 extern struct memblock memblock;
@@ -145,6 +141,30 @@ void __next_reserved_mem_region(u64 *idx, phys_addr_t *out_start,
 
 void __memblock_free_late(phys_addr_t base, phys_addr_t size);
 
+#ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP
+static inline void __next_physmem_range(u64 *idx, struct memblock_type *type,
+					phys_addr_t *out_start,
+					phys_addr_t *out_end)
+{
+	extern struct memblock_type physmem;
+
+	__next_mem_range(idx, NUMA_NO_NODE, MEMBLOCK_NONE, &physmem, type,
+			 out_start, out_end, NULL);
+}
+
+/**
+ * for_each_physmem_range - iterate through physmem areas not included in type.
+ * @i: u64 used as loop variable
+ * @type: ptr to memblock_type which excludes from the iteration, can be %NULL
+ * @p_start: ptr to phys_addr_t for start address of the range, can be %NULL
+ * @p_end: ptr to phys_addr_t for end address of the range, can be %NULL
+ */
+#define for_each_physmem_range(i, type, p_start, p_end)			\
+	for (i = 0, __next_physmem_range(&i, type, p_start, p_end);	\
+	     i != (u64)ULLONG_MAX;					\
+	     __next_physmem_range(&i, type, p_start, p_end))
+#endif /* CONFIG_HAVE_MEMBLOCK_PHYS_MAP */
+
 /**
  * for_each_mem_range - iterate through memblock areas from type_a and not
  * included in type_b. Or just type_a if type_b is NULL.

mm/memblock.c

@@ -44,19 +44,20 @@
  *   in the system, for instance when the memory is restricted with
  *   ``mem=`` command line parameter
  * * ``reserved`` - describes the regions that were allocated
- * * ``physmap`` - describes the actual physical memory regardless of
+ * * ``physmem`` - describes the actual physical memory available during
- *   the possible restrictions; the ``physmap`` type is only available
+ *   boot regardless of the possible restrictions and memory hot(un)plug;
- *   on some architectures.
+ *   the ``physmem`` type is only available on some architectures.
  *
  * Each region is represented by :c:type:`struct memblock_region` that
  * defines the region extents, its attributes and NUMA node id on NUMA
  * systems. Every memory type is described by the :c:type:`struct
  * memblock_type` which contains an array of memory regions along with
- * the allocator metadata. The memory types are nicely wrapped with
+ * the allocator metadata. The "memory" and "reserved" types are nicely
- * :c:type:`struct memblock`. This structure is statically initialzed
+ * wrapped with :c:type:`struct memblock`. This structure is statically
- * at build time. The region arrays for the "memory" and "reserved"
+ * initialized at build time. The region arrays are initially sized to
- * types are initially sized to %INIT_MEMBLOCK_REGIONS and for the
+ * %INIT_MEMBLOCK_REGIONS for "memory" and %INIT_MEMBLOCK_RESERVED_REGIONS
- * "physmap" type to %INIT_PHYSMEM_REGIONS.
+ * for "reserved". The region array for "physmem" is initially sized to
+ * %INIT_PHYSMEM_REGIONS.
  * The memblock_allow_resize() enables automatic resizing of the region
  * arrays during addition of new regions. This feature should be used
  * with care so that memory allocated for the region array will not
@@ -87,8 +88,8 @@
  * function frees all the memory to the buddy page allocator.
  *
  * Unless an architecture enables %CONFIG_ARCH_KEEP_MEMBLOCK, the
- * memblock data structures will be discarded after the system
+ * memblock data structures (except "physmem") will be discarded after the
- * initialization completes.
+ * system initialization completes.
  */
 
 #ifndef CONFIG_NEED_MULTIPLE_NODES
@@ -104,7 +105,7 @@ unsigned long long max_possible_pfn;
 static struct memblock_region memblock_memory_init_regions[INIT_MEMBLOCK_REGIONS] __initdata_memblock;
 static struct memblock_region memblock_reserved_init_regions[INIT_MEMBLOCK_RESERVED_REGIONS] __initdata_memblock;
 #ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP
-static struct memblock_region memblock_physmem_init_regions[INIT_PHYSMEM_REGIONS] __initdata_memblock;
+static struct memblock_region memblock_physmem_init_regions[INIT_PHYSMEM_REGIONS];
 #endif
 
 struct memblock memblock __initdata_memblock = {
@@ -118,17 +119,19 @@ struct memblock memblock __initdata_memblock = {
 	.reserved.max		= INIT_MEMBLOCK_RESERVED_REGIONS,
 	.reserved.name		= "reserved",
 
-#ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP
-	.physmem.regions	= memblock_physmem_init_regions,
-	.physmem.cnt		= 1,	/* empty dummy entry */
-	.physmem.max		= INIT_PHYSMEM_REGIONS,
-	.physmem.name		= "physmem",
-#endif
-
 	.bottom_up		= false,
 	.current_limit		= MEMBLOCK_ALLOC_ANYWHERE,
 };
 
+#ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP
+struct memblock_type physmem = {
+	.regions		= memblock_physmem_init_regions,
+	.cnt			= 1,	/* empty dummy entry */
+	.max			= INIT_PHYSMEM_REGIONS,
+	.name			= "physmem",
+};
+#endif
+
 int memblock_debug __initdata_memblock;
 static bool system_has_some_mirror __initdata_memblock = false;
 static int memblock_can_resize __initdata_memblock;
@@ -838,7 +841,7 @@ int __init_memblock memblock_physmem_add(phys_addr_t base, phys_addr_t size)
 	memblock_dbg("%s: [%pa-%pa] %pS\n", __func__,
 		     &base, &end, (void *)_RET_IP_);
 
-	return memblock_add_range(&memblock.physmem, base, size, MAX_NUMNODES, 0);
+	return memblock_add_range(&physmem, base, size, MAX_NUMNODES, 0);
 }
 #endif
 
@@ -1019,11 +1022,9 @@ static bool should_skip_region(struct memblock_region *m, int nid, int flags)
  * As both region arrays are sorted, the function advances the two indices
  * in lockstep and returns each intersection.
  */
-void __init_memblock __next_mem_range(u64 *idx, int nid,
+void __next_mem_range(u64 *idx, int nid, enum memblock_flags flags,
-				      enum memblock_flags flags,
 		      struct memblock_type *type_a,
-				      struct memblock_type *type_b,
+		      struct memblock_type *type_b, phys_addr_t *out_start,
-				      phys_addr_t *out_start,
 		      phys_addr_t *out_end, int *out_nid)
 {
 	int idx_a = *idx & 0xffffffff;
@@ -1924,7 +1925,7 @@ void __init_memblock __memblock_dump_all(void)
 	memblock_dump(&memblock.memory);
 	memblock_dump(&memblock.reserved);
 #ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP
-	memblock_dump(&memblock.physmem);
+	memblock_dump(&physmem);
 #endif
 }
 
@@ -2064,8 +2065,8 @@ static int __init memblock_init_debugfs(void)
 	debugfs_create_file("reserved", 0444, root,
 			    &memblock.reserved, &memblock_debug_fops);
 #ifdef CONFIG_HAVE_MEMBLOCK_PHYS_MAP
-	debugfs_create_file("physmem", 0444, root,
+	debugfs_create_file("physmem", 0444, root, &physmem,
-			    &memblock.physmem, &memblock_debug_fops);
+			    &memblock_debug_fops);
 #endif
 
 	return 0;