So, this patch started out with me trying to keep from passing
contiguous, node-specific mem_map into free_area_init_node() and
cousins. Instead, I relied on some calls to pfn_to_page().
This works fine and dandy when all you need is the pgdat->node_mem_map
to do pfn_to_page(). However, the non-NUMA/DISCONTIG architectures use
the real, global mem_map[] instead of a node_mem_map in the
pfn_to_page() calculation. So, I ended up effectively trying to
initialize mem_map from itself, when it was NULL. That was bad, and
caused some very pretty colors on someone's screen when he tested it.
So, I had to make sure to initialize the global mem_map[] before calling
into free_area_init_node(). Then, I realized how many architectures do
this on their own, and have comments like this:
/* XXX: MRB-remove - this doesn't seem sane, should this be done somewhere else ?*/
mem_map = NODE_DATA(0)->node_mem_map;
Also, ppc64 has some interesting hacks^Wcode in this area to make up for
any empty-memory NUMA nodes, which I think can go away now.
The following patch does what my first one did (don't pass mem_map into
the init functions), incorporates Jesse Barnes' ia64 fixes on top of
that, and gets rid of all but one of the global mem_map initializations
(parisc is weird). It also magically removes more code than it adds.
It could be smaller, but I shamelessly added some comments.
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
---
taka-dave/arch/arm/mm/init.c | 4 ----
taka-dave/arch/arm26/mm/init.c | 2 --
taka-dave/arch/cris/arch-v10/mm/init.c | 1 -
taka-dave/arch/i386/mm/discontig.c | 9 +++------
taka-dave/arch/ia64/mm/contig.c | 5 ++---
taka-dave/arch/ia64/mm/init.c | 2 +-
taka-dave/arch/ppc64/mm/init.c | 1 -
taka-dave/arch/sh/mm/init.c | 2 --
taka-dave/arch/sh64/mm/init.c | 3 ---
taka-dave/arch/sparc/mm/srmmu.c | 1 -
taka-dave/arch/sparc/mm/sun4c.c | 1 -
taka-dave/arch/sparc64/mm/init.c | 1 -
taka-dave/arch/v850/kernel/setup.c | 1 -
taka-dave/mm/page_alloc.c | 18 ++++++++++++++----
14 files changed, 20 insertions(+), 31 deletions(-)
diff -puN arch/arm/mm/init.c~A6-no_arch_mem_map_init arch/arm/mm/init.c
--- taka/arch/arm/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/arm/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -499,10 +499,6 @@ void __init paging_init(struct meminfo *
bdata->node_boot_start >> PAGE_SHIFT, zhole_size);
}
-#ifndef CONFIG_DISCONTIGMEM
- mem_map = contig_page_data.node_mem_map;
-#endif
-
/*
* finish off the bad pages once
* the mem_map is initialised
diff -puN arch/arm26/mm/init.c~A6-no_arch_mem_map_init arch/arm26/mm/init.c
--- taka/arch/arm26/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/arm26/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -309,8 +309,6 @@ void __init paging_init(struct meminfo *
free_area_init_node(0, pgdat, zone_size,
bdata->node_boot_start >> PAGE_SHIFT, zhole_size);
- mem_map = NODE_DATA(0)->node_mem_map;
-
/*
* finish off the bad pages once
* the mem_map is initialised
diff -puN arch/cris/arch-v10/mm/init.c~A6-no_arch_mem_map_init arch/cris/arch-v10/mm/init.c
--- taka/arch/cris/arch-v10/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/cris/arch-v10/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -184,7 +184,6 @@ paging_init(void)
*/
free_area_init_node(0, &contig_page_data, zones_size, PAGE_OFFSET >> PAGE_SHIFT, 0);
- mem_map = contig_page_data.node_mem_map;
}
/* Initialize remaps of some I/O-ports. It is important that this
diff -puN arch/i386/mm/discontig.c~A6-no_arch_mem_map_init arch/i386/mm/discontig.c
--- taka/arch/i386/mm/discontig.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/i386/mm/discontig.c 2004-09-27 13:49:13.000000000 -0700
@@ -417,18 +417,15 @@ void __init zone_sizes_init(void)
* normal bootmem allocator, but other nodes come from the
* remapped KVA area - mbligh
*/
- if (!nid)
- free_area_init_node(nid, NODE_DATA(nid),
- zones_size, start, zholes_size);
- else {
+ if (nid) {
unsigned long lmem_map;
lmem_map = (unsigned long)node_remap_start_vaddr[nid];
lmem_map += sizeof(pg_data_t) + PAGE_SIZE - 1;
lmem_map &= PAGE_MASK;
NODE_DATA(nid)->node_mem_map = (struct page *)lmem_map;
- free_area_init_node(nid, NODE_DATA(nid), zones_size,
- start, zholes_size);
}
+ free_area_init_node(nid, NODE_DATA(nid), zones_size,
+ start, zholes_size);
}
return;
}
diff -puN arch/ia64/mm/contig.c~A6-no_arch_mem_map_init arch/ia64/mm/contig.c
--- taka/arch/ia64/mm/contig.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/ia64/mm/contig.c 2004-09-27 13:49:13.000000000 -0700
@@ -269,7 +269,7 @@ paging_init (void)
vmem_map = (struct page *) 0;
free_area_init_node(0, &contig_page_data, zones_size, 0,
zholes_size);
- vmem_map = mem_map = contig_page_data.node_mem_map;
+ vmem_map = contig_page_data.node_mem_map;
} else {
unsigned long map_size;
@@ -280,11 +280,10 @@ paging_init (void)
vmem_map = (struct page *) vmalloc_end;
efi_memmap_walk(create_mem_map_page_table, 0);
- contig_page_data.node_mem_map = vmem_map;
+ NODE_DATA(0)->node_mem_map = vmem_map;
free_area_init_node(0, &contig_page_data, zones_size,
0, zholes_size);
- mem_map = contig_page_data.node_mem_map;
printk("Virtual mem_map starts at 0x%p\n", mem_map);
}
#else /* !CONFIG_VIRTUAL_MEM_MAP */
diff -puN arch/ia64/mm/discontig.c~A6-no_arch_mem_map_init arch/ia64/mm/discontig.c
diff -puN arch/ia64/mm/init.c~A6-no_arch_mem_map_init arch/ia64/mm/init.c
--- taka/arch/ia64/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/ia64/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -429,7 +429,7 @@ virtual_memmap_init (u64 start, u64 end,
/ sizeof(struct page));
if (map_start < map_end)
- memmap_init_zone((unsigned long)(map_end - map_start),
+ memmap_init_zone(map_start, (unsigned long) (map_end - map_start),
args->nid, args->zone, page_to_pfn(map_start));
return 0;
}
diff -puN arch/ppc64/mm/init.c~A6-no_arch_mem_map_init arch/ppc64/mm/init.c
--- taka/arch/ppc64/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/ppc64/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -662,7 +662,6 @@ void __init paging_init(void)
free_area_init_node(0, &contig_page_data, zones_size,
__pa(PAGE_OFFSET) >> PAGE_SHIFT, zholes_size);
- mem_map = contig_page_data.node_mem_map;
}
#endif /* CONFIG_DISCONTIGMEM */
diff -puN arch/sh/mm/init.c~A6-no_arch_mem_map_init arch/sh/mm/init.c
--- taka/arch/sh/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/sh/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -216,8 +216,6 @@ void __init paging_init(void)
#endif
NODE_DATA(0)->node_mem_map = NULL;
free_area_init_node(0, NODE_DATA(0), zones_size, __MEMORY_START >> PAGE_SHIFT, 0);
- /* XXX: MRB-remove - this doesn't seem sane, should this be done somewhere else ?*/
- mem_map = NODE_DATA(0)->node_mem_map;
#ifdef CONFIG_DISCONTIGMEM
/*
diff -puN arch/sh64/mm/init.c~A6-no_arch_mem_map_init arch/sh64/mm/init.c
--- taka/arch/sh64/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/sh64/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -124,9 +124,6 @@ void __init paging_init(void)
zones_size[ZONE_DMA] = MAX_LOW_PFN - START_PFN;
NODE_DATA(0)->node_mem_map = NULL;
free_area_init_node(0, NODE_DATA(0), zones_size, __MEMORY_START >> PAGE_SHIFT, 0);
-
- /* XXX: MRB-remove - this doesn't seem sane, should this be done somewhere else ?*/
- mem_map = NODE_DATA(0)->node_mem_map;
}
void __init mem_init(void)
diff -puN arch/sparc/mm/srmmu.c~A6-no_arch_mem_map_init arch/sparc/mm/srmmu.c
--- taka/arch/sparc/mm/srmmu.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/sparc/mm/srmmu.c 2004-09-27 13:49:13.000000000 -0700
@@ -1343,7 +1343,6 @@ void __init srmmu_paging_init(void)
free_area_init_node(0, &contig_page_data, zones_size,
pfn_base, zholes_size);
- mem_map = contig_page_data.node_mem_map;
}
}
diff -puN arch/sparc/mm/sun4c.c~A6-no_arch_mem_map_init arch/sparc/mm/sun4c.c
--- taka/arch/sparc/mm/sun4c.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/sparc/mm/sun4c.c 2004-09-27 13:49:13.000000000 -0700
@@ -2116,7 +2116,6 @@ void __init sun4c_paging_init(void)
free_area_init_node(0, &contig_page_data, zones_size,
pfn_base, zholes_size);
- mem_map = contig_page_data.node_mem_map;
}
cnt = 0;
diff -puN arch/sparc64/mm/init.c~A6-no_arch_mem_map_init arch/sparc64/mm/init.c
--- taka/arch/sparc64/mm/init.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/sparc64/mm/init.c 2004-09-27 13:49:13.000000000 -0700
@@ -1511,7 +1511,6 @@ void __init paging_init(void)
free_area_init_node(0, &contig_page_data, zones_size,
phys_base >> PAGE_SHIFT, zholes_size);
- mem_map = contig_page_data.node_mem_map;
}
device_scan();
diff -puN arch/v850/kernel/setup.c~A6-no_arch_mem_map_init arch/v850/kernel/setup.c
--- taka/arch/v850/kernel/setup.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/arch/v850/kernel/setup.c 2004-09-27 13:49:13.000000000 -0700
@@ -283,5 +283,4 @@ init_mem_alloc (unsigned long ram_start,
NODE_DATA(0)->node_mem_map = NULL;
free_area_init_node (0, NODE_DATA(0), zones_size,
ADDR_TO_PAGE (PAGE_OFFSET), 0);
- mem_map = NODE_DATA(0)->node_mem_map;
}
diff -puN mm/page_alloc.c~A6-no_arch_mem_map_init mm/page_alloc.c
--- taka/mm/page_alloc.c~A6-no_arch_mem_map_init 2004-09-27 13:49:13.000000000 -0700
+++ taka-dave/mm/page_alloc.c 2004-09-27 13:49:13.000000000 -0700
@@ -1691,14 +1691,25 @@ static void __init free_area_init_core(s
}
}
-void __init node_alloc_mem_map(struct pglist_data *pgdat)
+static void __init alloc_node_mem_map(struct pglist_data *pgdat)
{
unsigned long size;
+ /*
+ * Make sure that the architecture hasn't already allocated
+ * a node_mem_map, and that the node contains memory.
+ */
+ if (pgdat->node_mem_map || !pgdat->node_spanned_pages)
+ return;
+
size = (pgdat->node_spanned_pages + 1) * sizeof(struct page);
pgdat->node_mem_map = alloc_bootmem_node(pgdat, size);
#ifndef CONFIG_DISCONTIGMEM
- mem_map = contig_page_data.node_mem_map;
+ /*
+ * With no DISCONTIG, the global mem_map is just set as node 0's
+ */
+ if (pgdat == NODE_DATA(0))
+ mem_map = NODE_DATA(0)->node_mem_map;
#endif
}
@@ -1710,8 +1721,7 @@ void __init free_area_init_node(int nid,
pgdat->node_start_pfn = node_start_pfn;
calculate_zone_totalpages(pgdat, zones_size, zholes_size);
- if (!pfn_to_page(node_start_pfn))
- node_alloc_mem_map(pgdat);
+ alloc_node_mem_map(pgdat);
free_area_init_core(pgdat, zones_size, zholes_size);
}
_