[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-changelog] [xen-3.2-testing] x86: Clarify early_page_fault() handler intent, and improve debug



# HG changeset patch
# User Keir Fraser <keir.fraser@xxxxxxxxxx>
# Date 1201516571 0
# Node ID 36a8e6763f3af8938f3c5c60e67cedd8a7ef0120
# Parent  c8c91e2e552e3b614d9c3f33bdc3d89c465d09c8
x86: Clarify early_page_fault() handler intent, and improve debug
tracing.
Signed-off-by: Keir Fraser <keir.fraser@xxxxxxxxxx>
xen-unstable changeset:   16887:923f2f7365079861e0e2341de22839376183208e
xen-unstable date:        Fri Jan 25 13:42:36 2008 +0000
---
 xen/arch/x86/traps.c |   22 ++++++++++++++--------
 1 files changed, 14 insertions(+), 8 deletions(-)

diff -r c8c91e2e552e -r 36a8e6763f3a xen/arch/x86/traps.c
--- a/xen/arch/x86/traps.c      Mon Jan 28 10:35:47 2008 +0000
+++ b/xen/arch/x86/traps.c      Mon Jan 28 10:36:11 2008 +0000
@@ -1116,12 +1116,11 @@ asmlinkage void do_page_fault(struct cpu
 }
 
 /*
- * Early handler to deal with spurious page faults. For example, consider a 
- * routine that uses a mapping immediately after installing it (making it 
- * present). The CPU may speculatively execute the memory access before 
- * executing the PTE write. The instruction will then be marked to cause a 
- * page fault when it is retired, despite the fact that the PTE is present and 
- * correct at that point in time.
+ * Early #PF handler to print CR2, error code, and stack.
+ * 
+ * We also deal with spurious faults here, even though they should never happen
+ * during early boot (an issue was seen once, but was most likely a hardware 
+ * problem).
  */
 asmlinkage void do_early_page_fault(struct cpu_user_regs *regs)
 {
@@ -1140,8 +1139,15 @@ asmlinkage void do_early_page_fault(stru
     }
 
     if ( stuck++ == 1000 )
-        panic("Early fatal page fault at %04x:%p (cr2=%p, ec=%04x)\n", 
-              regs->cs, _p(regs->eip), _p(cr2), regs->error_code);
+    {
+        unsigned long *stk = (unsigned long *)regs;
+        printk("Early fatal page fault at %04x:%p (cr2=%p, ec=%04x)\n", 
+               regs->cs, _p(regs->eip), _p(cr2), regs->error_code);
+        printk("Stack dump: ");
+        while ( ((long)stk & ((PAGE_SIZE - 1) & ~(BYTES_PER_LONG - 1))) != 0 )
+            printk("%p ", _p(*stk++));
+        for ( ; ; ) ;
+    }
 }
 
 long do_fpu_taskswitch(int set)

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


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.