problem with x86-prepare-kprobes-code-for-unification.patch

problem with x86-prepare-kprobes-code-for-unification.patch

[Jeremy Fitzhardinge]

Hi,

This code in kprobes_64.c from
x86-prepare-kprobes-code-for-unification.patch is problematic:

+static __always_inline int is_IF_modifier(kprobe_opcode_t *insn)
+{
+	switch (*insn) {
+	case 0xfa:		/* cli */
+	case 0xfb:		/* sti */
+	case 0xcf:		/* iret/iretd */
+	case 0x9d:		/* popf/popfd */
+		return 1;
+	}
+
+	/*
+	 * on 64 bit x86, 0x40-0x4f are prefixes so we need to look
+	 * at the next byte instead.. but of course not recurse infinitely
+	 */
+	if (*insn  >= 0x40 && *insn <= 0x4f && *insn > 0x4f)
+		return is_IF_modifier(++insn);
+	return 0;
+}


Firstly, the conditional in the if() doesn't seem to make much sense. 
It appears to be miscopied from:

-	if (*insn  >= 0x40 && *insn <= 0x4f && *++insn == 0xcf)
-		return 1;


Secondly, the recursive inlining is making 4.0.2 complain:

/home/jeremy/hg/xen/paravirt/linux/arch/x86/kernel/kprobes_64.c: In function 'set_current_kprobe':
/home/jeremy/hg/xen/paravirt/linux/arch/x86/kernel/kprobes_64.c:152: sorry, unimplemented: inlining failed in call to 'is_IF_modifier': recursive inlining
/home/jeremy/hg/xen/paravirt/linux/arch/x86/kernel/kprobes_64.c:166: sorry, unimplemented: called from here

    J