[Xen-devel] [PATCH] linux/x86_64: correctly handle mem= reducing memory amount

["Jan Beulich" <JBeulich@xxxxxxxxxx>]

When mem= is being used to specify a value below the amount a domain
got passed from Xen, init_memory_mapping() got called with the higher
original value (end_pfn_map), triggering the BUG()s in maddr.h
checking PFNs against end_pfn.

Written and tested on 2.6.16.60 and made apply to the 2.6.18 tree
without further testing.

Signed-off-by: Jan Beulich <jbeulich@xxxxxxxxxx>

--- sle10sp3-2009-11-27.orig/arch/x86_64/kernel/e820-xen.c      2009-10-01 
14:16:23.000000000 +0200
+++ sle10sp3-2009-11-27/arch/x86_64/kernel/e820-xen.c   2009-12-01 
11:31:11.000000000 +0100
@@ -30,12 +30,14 @@
 unsigned long end_pfn; 
 EXPORT_SYMBOL(end_pfn);
 
+#ifndef CONFIG_XEN
 /* 
  * end_pfn only includes RAM, while end_pfn_map includes all e820 entries.
  * The direct mapping extends to end_pfn_map, so that we can directly access
  * apertures, ACPI and other tables without having to play with fixmaps.
  */ 
 unsigned long end_pfn_map; 
+#endif
 
 /* 
  * Last pfn which the user wants to use.
--- sle10sp3-2009-11-27.orig/include/asm-x86_64/proto.h 2009-08-10 
10:13:16.000000000 +0200
+++ sle10sp3-2009-11-27/include/asm-x86_64/proto.h      2009-12-01 
11:30:58.000000000 +0100
@@ -75,7 +75,11 @@ extern void load_gs_index(unsigned gs);
 extern void stop_timer_interrupt(void);
 extern void main_timer_handler(struct pt_regs *regs);
 
+#ifndef CONFIG_XEN
 extern unsigned long end_pfn_map; 
+#else
+#define end_pfn_map end_pfn
+#endif
 
 extern void show_trace(struct task_struct *, struct pt_regs *, unsigned long * 
rsp);
 extern void show_registers(struct pt_regs *regs);

Attachment: xenlinux-x86_64-mem-reduction.patch
Description: Text document

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel