Skip to contentSkip to conversation

powerpc: minor fix of numa message

PATCH In review Geoff Levand 2 participants 3 messages Sep 14, 2006 – Sep 16, 2006 PowerPC

lore
Share & tools
lore

Geoff Levand

author
#0 lore raw

PATCH 1/1 powerpc: minor fix of numa message

This is a very minor fix that effects no current system. The NUMA startup message values for 'Top of RAM' and 'Memory hole size' were reported incorrectly if more than one lmb region was added.

A comment in the lmb code states that lmb_analyze() must be called before lmb_phys_mem_size().

Signed-off-by Geoff Levand <[email protected]>

---

arch/powerpc/mm/numa.c +6 -2
Index: cell--common--4/arch/powerpc/mm/numa.c
===================================================================
@@ -483,10 +483,14 @@
 
 static void __init setup_nonnuma(void)
 {
-	unsigned long top_of_ram = lmb_end_of_DRAM();
-	unsigned long total_ram = lmb_phys_mem_size();
+	unsigned long top_of_ram;
+	unsigned long total_ram;
 	unsigned int i;
 
+	lmb_analyze();
+	top_of_ram = lmb_end_of_DRAM();
+	total_ram = lmb_phys_mem_size();
+
 	printk(KERN_DEBUG "Top of RAM: 0x%lx, Total RAM: 0x%lx\n",
 	       top_of_ram, total_ram);
 	printk(KERN_DEBUG "Memory hole size: %ldMB\n",
#1 lore raw
Quoted 6 lines · Geoff Levand

This is a very minor fix that effects no current system. The NUMA startup message values for 'Top of RAM' and 'Memory hole size' were reported incorrectly if more than one lmb region was added.

A comment in the lmb code states that lmb_analyze() must be called before lmb_phys_mem_size().

But we've already called it in early_init_devtree()?

cheers

Signature · 8 lines
Michael Ellerman
OzLabs, IBM Australia Development Lab

wwweb: http://michael.ellerman.id.au
phone: +61 2 6212 1183 (tie line 70 21183)

We do not inherit the earth from our ancestors,
we borrow it from our children. - S.M.A.R.T Person
author
#2 lore raw
Quoted 9 lines · Michael Ellerman
Quoted 6 lines · Geoff Levand

This is a very minor fix that effects no current system. The NUMA startup message values for 'Top of RAM' and 'Memory hole size' were reported incorrectly if more than one lmb region was added.

A comment in the lmb code states that lmb_analyze() must be called before lmb_phys_mem_size().

But we've already called it in early_init_devtree()?

Yes, of course you are correct. Updated patch follows.

-Geoff

This is a very minor comment change to the powerpc lbm code reflect current usage.

Signed-off-by Geoff Levand <[email protected]>

---

arch/powerpc/mm/lmb.c +3 -1
Index: cell--common--4/arch/powerpc/mm/lmb.c
===================================================================
@@ -285,7 +285,9 @@
 	return base;
 }
 
-/* You must call lmb_analyze() before this. */
+/* You must call lmb_analyze() sometime before this, typcally just after
+ * adding regions.
+ */
 unsigned long __init lmb_phys_mem_size(void)
 {
 	return lmb.memory.size;