[PATCH -tip] ftrace: Correctly calculate the first function in the .text section

[Matt Fleming]

This patch fixes a bug whereby recordmcount.pl didn't stop searching
once it had correctly detected the function at the beginning of the
.text section. To stop it searching, I needed to reset $read_function.
The effect of this bug was that some entries in __mcount_loc section
were created with the negative reloc addends. The last text section
found was used as the base and all mcount calls were at a relative
offset to that, so at final link time the addresses were fixed up to
point to somewhere completely bogus.

This all resulted in ftrace dynamically modifying addresses that weren't
actually mcount callsites.

I also noticed another bug and fixed up the condition from,

	if (!defined($ref_func) || !defined($weak{$text})) {
to
	if (!defined($ref_func) && !defined($weak{$text})) {

This now matches the comment above the conditional.

Signed-off-by: Matt Fleming <matt@xxxxxxxxxxxxxxxxx>
---
 scripts/recordmcount.pl |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/scripts/recordmcount.pl b/scripts/recordmcount.pl
index 7109e2b..356ff6a 100755
--- a/scripts/recordmcount.pl
+++ b/scripts/recordmcount.pl
@@ -414,8 +414,9 @@ while (<IN>) {
 	    $read_function = 0;
 	} else {
 	    # if we already have a function, and this is weak, skip it
-	    if (!defined($ref_func) || !defined($weak{$text})) {
+	    if (!defined($ref_func) && !defined($weak{$text})) {
 		$ref_func = $text;
+		$read_function = 0;
 	    }
 	}
     } elsif ($read_headers && /$mcount_section/) {
-- 
1.6.4.rc0

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/