Skip to contentSkip to conversation

i386: nid_zone_sizes_init() update

PATCH In review Magnus Damm 4 participants 5 messages Oct 5, 2005 – Oct 6, 2005 Memory management

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 i386: nid_zone_sizes_init() update

Broken out nid_zone_sizes_init() change from i386 NUMA emulation code.

---

Applies on top of linux-2.6.14-rc2-git8-mhp1

arch/i386/kernel/setup.c +9 -16
@@ -1215,31 +1215,24 @@ static inline unsigned long max_hardware
 {
 	return virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT;
 }
-static inline unsigned long  nid_size_pages(int nid)
-{
-	return node_end_pfn[nid] - node_start_pfn[nid];
-}
-static inline int nid_starts_in_highmem(int nid)
-{
-	return node_start_pfn[nid] >= max_low_pfn;
-}
 
 void __init nid_zone_sizes_init(int nid)
 {
 	unsigned long zones_size[MAX_NR_ZONES] = {0, 0, 0};
-	unsigned long max_dma;
+	unsigned long max_dma = min(max_hardware_dma_pfn(), max_low_pfn);
 	unsigned long start = node_start_pfn[nid];
 	unsigned long end = node_end_pfn[nid];
 
 	if (node_has_online_mem(nid)){
-		if (nid_starts_in_highmem(nid)) {
-			zones_size[ZONE_HIGHMEM] = nid_size_pages(nid);
-		} else {
-			max_dma = min(max_hardware_dma_pfn(), max_low_pfn);
-			zones_size[ZONE_DMA] = max_dma;
-			zones_size[ZONE_NORMAL] = max_low_pfn - max_dma;
-			zones_size[ZONE_HIGHMEM] = end - max_low_pfn;
+		if (start < max_dma) {
+			zones_size[ZONE_DMA] = min(end, max_dma) - start;
+		}
+		if (start < max_low_pfn && max_dma < end) {
+			zones_size[ZONE_NORMAL] = min(end, max_low_pfn) - max(start, max_dma);
 		}
+		if (max_low_pfn <= end) {
+			zones_size[ZONE_HIGHMEM] = end - max(start, max_low_pfn);
+               }
 	}
 
 	free_area_init_node(nid, NODE_DATA(nid), zones_size, start,

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

#1 lore raw
Quoted 1 line

Broken out nid_zone_sizes_init() change from i386 NUMA emulation code.

Mmmm. what's the purpose of this change? Not sure I understand what you're trying to acheive here ... looks like you're just removing some abstractions? To me, they made the code a bit easier to read.

Quoted 50 lines

---

Applies on top of linux-2.6.14-rc2-git8-mhp1

arch/i386/kernel/setup.c +9 -17
@@ -1215,31 +1215,24 @@ static inline unsigned long max_hardware
 {
 	return virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT;
 }
-static inline unsigned long  nid_size_pages(int nid)
-{
-	return node_end_pfn[nid] - node_start_pfn[nid];
-}
-static inline int nid_starts_in_highmem(int nid)
-{
-	return node_start_pfn[nid] >= max_low_pfn;
-}
 
 void __init nid_zone_sizes_init(int nid)
 {
 	unsigned long zones_size[MAX_NR_ZONES] = {0, 0, 0};
-	unsigned long max_dma;
+	unsigned long max_dma = min(max_hardware_dma_pfn(), max_low_pfn);
 	unsigned long start = node_start_pfn[nid];
 	unsigned long end = node_end_pfn[nid];
 
 	if (node_has_online_mem(nid)){
-		if (nid_starts_in_highmem(nid)) {
-			zones_size[ZONE_HIGHMEM] = nid_size_pages(nid);
-		} else {
-			max_dma = min(max_hardware_dma_pfn(), max_low_pfn);
-			zones_size[ZONE_DMA] = max_dma;
-			zones_size[ZONE_NORMAL] = max_low_pfn - max_dma;
-			zones_size[ZONE_HIGHMEM] = end - max_low_pfn;
+		if (start < max_dma) {
+			zones_size[ZONE_DMA] = min(end, max_dma) - start;
+		}
+		if (start < max_low_pfn && max_dma < end) {
+			zones_size[ZONE_NORMAL] = min(end, max_low_pfn) - max(start, max_dma);
 		}
+		if (max_low_pfn <= end) {
+			zones_size[ZONE_HIGHMEM] = end - max(start, max_low_pfn);
+               }
 	}
 
 	free_area_init_node(nid, NODE_DATA(nid), zones_size, start,
 
--
Quoted 6 lines

To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

#2 lore raw
Quoted 5 lines · Martin J. Bligh
Quoted 1 line

Broken out nid_zone_sizes_init() change from i386 NUMA emulation code.

Mmmm. what's the purpose of this change? Not sure I understand what you're trying to acheive here ... looks like you're just removing some abstractions? To me, they made the code a bit easier to read.

Thanks for the compliment. Perhaps we should merge this patch upstream:

http://www.sr71.net/patches/2.6.14/2.6.14-rc2-git8-mhp1/broken-out/B2.1-i386-discontig-consolidation.patch

-- Dave

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

#3 lore raw
In reply to Magnus Damm
Quoted 1 line · Magnus Damm

Broken out nid_zone_sizes_init() change from i386 NUMA emulation code.

...

arch/i386/kernel/setup.c static inline unsigned long max_hardware · line 1225

Quoted 8 lines
arch/i386/kernel/setup.c +0 -8
-static inline unsigned long  nid_size_pages(int nid)
-{
-	return node_end_pfn[nid] - node_start_pfn[nid];
-}
-static inline int nid_starts_in_highmem(int nid)
-{
-	return node_start_pfn[nid] >= max_low_pfn;
-}

Hey, I liked those helpers!

When I suggested that you make your patches apply on top of the existing -mhp stuff, I didn't just mean that they should _apply_, they should probably mesh a little bit better. For instance, it would be very helpful to use those 'static inlines', or make a couple new ones if you need them.

-- Dave

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>

#4 lore raw
Quoted 13 lines · Dave Hansen
Quoted 1 line · Magnus Damm

Broken out nid_zone_sizes_init() change from i386 NUMA emulation code.

...

arch/i386/kernel/setup.c static inline unsigned long max_hardware · line 1225

Quoted 8 lines
arch/i386/kernel/setup.c +0 -8
-static inline unsigned long  nid_size_pages(int nid)
-{
-     return node_end_pfn[nid] - node_start_pfn[nid];
-}
-static inline int nid_starts_in_highmem(int nid)
-{
-     return node_start_pfn[nid] >= max_low_pfn;
-}

Hey, I liked those helpers!

Well, too bad for you! ;)

Quoted 5 lines

When I suggested that you make your patches apply on top of the existing -mhp stuff, I didn't just mean that they should _apply_, they should probably mesh a little bit better. For instance, it would be very helpful to use those 'static inlines', or make a couple new ones if you need them.

Okey, I think I understand. And I agree that more effort could be put into this to make the code look better. I thought this change was required for the NUMA emulation code, but it turns out it wasn't. I guess keeping the code as is and ignoring the patch is probably the best solution.

Thanks!

/ magnus

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>