[PATCH] for_each_online_pgdat: for_each_bootmem

Add a list_head to bootmem_data_t and make bootmems use it.  bootmem list is
sorted by node_boot_start.

Only nodes against which init_bootmem() is called are linked to the list.
(i386 allocates bootmem only from one node(0) not from all online nodes.)

A summary:
 1. for_each_online_pgdat() traverses all *online* nodes.
 2. alloc_bootmem() allocates memory only from initialized-for-bootmem nodes.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

2 files changed, 30 insertions, 10 deletions

diff --git a/include/linux/bootmem.h b/include/linux/bootmem.h
index 7155452fb4a8..de3eb8d8ae26 100644
--- a/include/linux/bootmem.h
+++ b/include/linux/bootmem.h
@@ -38,6 +38,7 @@ typedef struct bootmem_data {
         unsigned long last_pos;
         unsigned long last_success;     /* Previous allocation point.  To speed
                                          * up searching */
+        struct list_head list;
 } bootmem_data_t;
 
 extern unsigned long __init bootmem_bootmap_pages (unsigned long);
diff --git a/mm/bootmem.c b/mm/bootmem.c
index b55bd39fc5dd..d3e3bd2ffcea 100644
--- a/mm/bootmem.c
+++ b/mm/bootmem.c
@@ -33,6 +33,7 @@ EXPORT_SYMBOL(max_pfn);		/* This is exported so
                                  * dma_get_required_mask(), which uses
                                  * it, can be an inline function */
 
+static LIST_HEAD(bdata_list);
 #ifdef CONFIG_CRASH_DUMP
 /*
  * If we have booted due to a crash, max_pfn will be a very low value. We need
@@ -52,6 +53,27 @@ unsigned long __init bootmem_bootmap_pages (unsigned long pages)
 
         return mapsize;
 }
+/*
+ * link bdata in order
+ */
+static void link_bootmem(bootmem_data_t *bdata)
+{
+        bootmem_data_t *ent;
+        if (list_empty(&bdata_list)) {
+                list_add(&bdata->list, &bdata_list);
+                return;
+        }
+        /* insert in order */
+        list_for_each_entry(ent, &bdata_list, list) {
+                if (bdata->node_boot_start < ent->node_boot_start) {
+                        list_add_tail(&bdata->list, &ent->list);
+                        return;
+                }
+        }
+        list_add_tail(&bdata->list, &bdata_list);
+        return;
+}
+
 
 /*
  * Called once to set up the allocator itself.
@@ -62,13 +84,11 @@ static unsigned long __init init_bootmem_core (pg_data_t *pgdat,
         bootmem_data_t *bdata = pgdat->bdata;
         unsigned long mapsize = ((end - start)+7)/8;
 
-        pgdat->pgdat_next = pgdat_list;
-        pgdat_list = pgdat;
-
         mapsize = ALIGN(mapsize, sizeof(long));
         bdata->node_bootmem_map = phys_to_virt(mapstart << PAGE_SHIFT);
         bdata->node_boot_start = (start << PAGE_SHIFT);
         bdata->node_low_pfn = end;
+        link_bootmem(bdata);
 
         /*
          * Initially all pages are reserved - setup_arch() has to
@@ -383,12 +403,11 @@ unsigned long __init free_all_bootmem (void)
 
 void * __init __alloc_bootmem(unsigned long size, unsigned long align, unsigned long goal)
 {
-        pg_data_t *pgdat = pgdat_list;
+        bootmem_data_t *bdata;
         void *ptr;
 
-        for_each_pgdat(pgdat)
-                if ((ptr = __alloc_bootmem_core(pgdat->bdata, size,
-                                                 align, goal, 0)))
+        list_for_each_entry(bdata, &bdata_list, list)
+                if ((ptr = __alloc_bootmem_core(bdata, size, align, goal, 0)))
                         return(ptr);
 
         /*
@@ -416,11 +435,11 @@ void * __init __alloc_bootmem_node(pg_data_t *pgdat, unsigned long size, unsigne
 
 void * __init __alloc_bootmem_low(unsigned long size, unsigned long align, unsigned long goal)
 {
-        pg_data_t *pgdat = pgdat_list;
+        bootmem_data_t *bdata;
         void *ptr;
 
-        for_each_pgdat(pgdat)
-                if ((ptr = __alloc_bootmem_core(pgdat->bdata, size,
+        list_for_each_entry(bdata, &bdata_list, list)
+                if ((ptr = __alloc_bootmem_core(bdata, size,
                                                  align, goal, LOW32LIMIT)))
                         return(ptr);