[PATCH] OProfile: fixed x86_64 incorrect kernel call graphs
Fix the problem in kernel 2.6.15.1 (and early versions) that OProfile on
x86_64 does not correctly collect the stack traces for kernel functions.
The original code in valid_kernel_stack() in arch/i386/oprofile/backtrace.c
assumes that the frame pointer (headaddr) should be greater than stack
(i.e., regs).
This assumption is wrong for x86_64 because NMIs in x86_64 use a seperate
stack different from the kernel stack. Therefore, the variable stack now
points to some location on the NMI stack, which turns out to be at a higher
address than the frame pointer (headaddr) on the kernel stack. The correct
comparison here should be between headaddr and regs->rsp for x86_64.
Signed-off-by: Tong Li <tong.n.li@intel.com>
Cc: John Levon <levon@movementarian.org>
Cc: Philippe Elie <phil.el@wanadoo.fr>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
diff --git a/arch/i386/oprofile/backtrace.c b/arch/i386/oprofile/backtrace.c
index 21654be..acc1813 100644
--- a/arch/i386/oprofile/backtrace.c
+++ b/arch/i386/oprofile/backtrace.c
@@ -49,7 +49,9 @@
* | stack |
* --------------- saved regs->ebp value if valid (frame_head address)
* . .
- * --------------- struct pt_regs stored on stack (struct pt_regs *)
+ * --------------- saved regs->rsp value if x86_64
+ * | |
+ * --------------- struct pt_regs * stored on stack if 32-bit
* | |
* . .
* | |
@@ -57,13 +59,26 @@
* | |
* | | \/ Lower addresses
*
- * Thus, &pt_regs <-> stack base restricts the valid(ish) ebp values
+ * Thus, regs (or regs->rsp for x86_64) <-> stack base restricts the
+ * valid(ish) ebp values. Note: (1) for x86_64, NMI and several other
+ * exceptions use special stacks, maintained by the interrupt stack table
+ * (IST). These stacks are set up in trap_init() in
+ * arch/x86_64/kernel/traps.c. Thus, for x86_64, regs now does not point
+ * to the kernel stack; instead, it points to some location on the NMI
+ * stack. On the other hand, regs->rsp is the stack pointer saved when the
+ * NMI occurred. (2) For 32-bit, regs->esp is not valid because the
+ * processor does not save %esp on the kernel stack when interrupts occur
+ * in the kernel mode.
*/
#ifdef CONFIG_FRAME_POINTER
static int valid_kernel_stack(struct frame_head * head, struct pt_regs * regs)
{
unsigned long headaddr = (unsigned long)head;
+#ifdef CONFIG_X86_64
+ unsigned long stack = (unsigned long)regs->rsp;
+#else
unsigned long stack = (unsigned long)regs;
+#endif
unsigned long stack_base = (stack & ~(THREAD_SIZE - 1)) + THREAD_SIZE;
return headaddr > stack && headaddr < stack_base;