[PATCH] sparsemem extreme implementation

With cleanups from Dave Hansen <haveblue@us.ibm.com>

SPARSEMEM_EXTREME makes mem_section a one dimensional array of pointers to
mem_sections.  This two level layout scheme is able to achieve smaller
memory requirements for SPARSEMEM with the tradeoff of an additional shift
and load when fetching the memory section.  The current SPARSEMEM
implementation is a one dimensional array of mem_sections which is the
default SPARSEMEM configuration.  The patch attempts isolates the
implementation details of the physical layout of the sparsemem section
array.

SPARSEMEM_EXTREME requires bootmem to be functioning at the time of
memory_present() calls.  This is not always feasible, so architectures
which do not need it may allocate everything statically by using
SPARSEMEM_STATIC.

Signed-off-by: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Bob Picco <bob.picco@hp.com>
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

4 files changed, 49 insertions, 37 deletions

diff --git a/arch/i386/Kconfig b/arch/i386/Kconfig
index 619d843ba231..dcb0ad098c60 100644
--- a/arch/i386/Kconfig
+++ b/arch/i386/Kconfig
@@ -754,6 +754,7 @@ config NUMA
         depends on SMP && HIGHMEM64G && (X86_NUMAQ || X86_GENERICARCH || (X86_SUMMIT && ACPI))
         default n if X86_PC
         default y if (X86_NUMAQ || X86_SUMMIT)
+        select SPARSEMEM_STATIC
 
 # Need comments to help the hapless user trying to turn on NUMA support
 comment "NUMA (NUMA-Q) requires SMP, 64GB highmem support"
diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index b97054bbc394..79cf578e21b9 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -487,39 +487,29 @@ struct mem_section {
         unsigned long section_mem_map;
 };
 
-#ifdef CONFIG_ARCH_SPARSEMEM_EXTREME
-/*
- * Should we ever require GCC 4 or later then the flat array scheme
- * can be eliminated and a uniform solution for EXTREME and !EXTREME can
- * be arrived at.
- */
-#define SECTION_ROOT_SHIFT      (PAGE_SHIFT-3)
-#define SECTION_ROOT_MASK       ((1UL<<SECTION_ROOT_SHIFT) - 1)
-#define SECTION_TO_ROOT(_sec)   ((_sec) >> SECTION_ROOT_SHIFT)
-#define NR_SECTION_ROOTS        (NR_MEM_SECTIONS >>  SECTION_ROOT_SHIFT)
+#ifdef CONFIG_SPARSEMEM_EXTREME
+#define SECTIONS_PER_ROOT       (PAGE_SIZE / sizeof (struct mem_section))
+#else
+#define SECTIONS_PER_ROOT       1
+#endif
 
-extern struct mem_section *mem_section[NR_SECTION_ROOTS];
-
-static inline struct mem_section *__nr_to_section(unsigned long nr)
-{
-        if (!mem_section[SECTION_TO_ROOT(nr)])
-                return NULL;
-        return &mem_section[SECTION_TO_ROOT(nr)][nr & SECTION_ROOT_MASK];
-}
+#define SECTION_NR_TO_ROOT(sec) ((sec) / SECTIONS_PER_ROOT)
+#define NR_SECTION_ROOTS        (NR_MEM_SECTIONS / SECTIONS_PER_ROOT)
+#define SECTION_ROOT_MASK       (SECTIONS_PER_ROOT - 1)
 
+#ifdef CONFIG_SPARSEMEM_EXTREME
+extern struct mem_section *mem_section[NR_SECTION_ROOTS];
 #else
-
-extern struct mem_section mem_section[NR_MEM_SECTIONS];
+extern struct mem_section mem_section[NR_SECTION_ROOTS][SECTIONS_PER_ROOT];
+#endif
 
 static inline struct mem_section *__nr_to_section(unsigned long nr)
 {
-        return &mem_section[nr];
+        if (!mem_section[SECTION_NR_TO_ROOT(nr)])
+                return NULL;
+        return &mem_section[SECTION_NR_TO_ROOT(nr)][nr & SECTION_ROOT_MASK];
 }
 
-#define sparse_index_init(_sec, _nid)  do {} while (0)
-
-#endif
-
 /*
  * We use the lower bits of the mem_map pointer to store
  * a little bit of information.  There should be at least
diff --git a/mm/Kconfig b/mm/Kconfig
index fc644c5c065d..4e9937ac3529 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -91,10 +91,23 @@ config HAVE_MEMORY_PRESENT
         depends on ARCH_HAVE_MEMORY_PRESENT || SPARSEMEM
 
 #
+# SPARSEMEM_EXTREME (which is the default) does some bootmem
+# allocations when memory_present() is called.  If this can not
+# be done on your architecture, select this option.  However,
+# statically allocating the mem_section[] array can potentially
+# consume vast quantities of .bss, so be careful.
+#
+# This option will also potentially produce smaller runtime code
+# with gcc 3.4 and later.
+#
+config SPARSEMEM_STATIC
+        def_bool n
+
+#
 # Architectecture platforms which require a two level mem_section in SPARSEMEM
 # must select this option. This is usually for architecture platforms with
 # an extremely sparse physical address space.
 #
-config ARCH_SPARSEMEM_EXTREME
-        def_bool n
-        depends on SPARSEMEM && 64BIT
+config SPARSEMEM_EXTREME
+        def_bool y
+        depends on SPARSEMEM && !SPARSEMEM_STATIC
diff --git a/mm/sparse.c b/mm/sparse.c
index b2b456bf0a5d..fa01292157a9 100644
--- a/mm/sparse.c
+++ b/mm/sparse.c
@@ -13,28 +13,36 @@
  *
  * 1) mem_section       - memory sections, mem_map's for valid memory
  */
-#ifdef CONFIG_ARCH_SPARSEMEM_EXTREME
+#ifdef CONFIG_SPARSEMEM_EXTREME
 struct mem_section *mem_section[NR_SECTION_ROOTS]
         ____cacheline_maxaligned_in_smp;
+#else
+struct mem_section mem_section[NR_SECTION_ROOTS][SECTIONS_PER_ROOT]
+        ____cacheline_maxaligned_in_smp;
+#endif
+EXPORT_SYMBOL(mem_section);
+
+static void sparse_alloc_root(unsigned long root, int nid)
+{
+#ifdef CONFIG_SPARSEMEM_EXTREME
+        mem_section[root] = alloc_bootmem_node(NODE_DATA(nid), PAGE_SIZE);
+#endif
+}
 
 static void sparse_index_init(unsigned long section, int nid)
 {
-        unsigned long root = SECTION_TO_ROOT(section);
+        unsigned long root = SECTION_NR_TO_ROOT(section);
 
         if (mem_section[root])
                 return;
-        mem_section[root] = alloc_bootmem_node(NODE_DATA(nid), PAGE_SIZE);
+
+        sparse_alloc_root(root, nid);
+
         if (mem_section[root])
                 memset(mem_section[root], 0, PAGE_SIZE);
         else
                 panic("memory_present: NO MEMORY\n");
 }
-#else
-struct mem_section mem_section[NR_MEM_SECTIONS]
-        ____cacheline_maxaligned_in_smp;
-#endif
-EXPORT_SYMBOL(mem_section);
-
 /* Record a memory area against a node. */
 void memory_present(int nid, unsigned long start, unsigned long end)
 {