Madhavan Srinivasan [off-list ref] writes:

On Tuesday 19 July 2016 10:18 AM, Michael Ellerman wrote:

quoted

In __ftrace_make_nop() (the 64-bit version), we have code to deal with
two ftrace ABIs. There is the original ABI, which looks mostly like a
function call, and then the mprofile-kernel ABI which is just a branch.

...

quoted

diff --git a/arch/powerpc/kernel/ftrace.c b/arch/powerpc/kernel/ftrace.c
index 7af6c4de044b..438442dac44c 100644
--- a/arch/powerpc/kernel/ftrace.c
+++ b/arch/powerpc/kernel/ftrace.c

@@ -144,6 +144,21 @@ __ftrace_make_nop(struct module *mod,
  		return -EINVAL;
  	}
  
+#ifdef CC_USING_MPROFILE_KERNEL
+	/* When using -mkernel_profile there is no load to jump over */
+	pop = PPC_INST_NOP;
+
+	if (probe_kernel_read(&op, (void *)(ip - 4), 4)) {
+		pr_err("Fetching instruction at %lx failed.\n", ip - 4);
+		return -EFAULT;
+	}
+
+	/* We expect either a mlfr r0, or a std r0, LRSAVE(r1) */

nit..  "mflr" and not "mlfr"

Yep, copied from the old comment, but I'll fix it up. Thanks.

cheers