[PATCH] Fix order_base_2(0)

[David Howells <dhowells@redhat.com>]

The order_base_2() function is either wrongly documented or wrongly
implemented.  In the preceding comment, it says that:

	ob2(0) = 0

but this is not valid as roundup_pow_of_two()'s documentation asserts that:

	* - the result is undefined when n == 0

Fix order_base_2(n) to actually return 0 when n == 0.

However, should we just instead define that the result of order_base_2(0) is
undefined?

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Robert P. J. Day <rpjday@crashcourse.ca>
cc: Andrew Morton <akpm@linux-foundation.org>
---

 include/linux/log2.h |    3 +--
 1 files changed, 1 insertions(+), 2 deletions(-)

diff --git a/include/linux/log2.h b/include/linux/log2.h
index fd7ff3d..d6463d8 100644
--- a/include/linux/log2.h
+++ b/include/linux/log2.h
@@ -202,7 +202,6 @@ unsigned long __rounddown_pow_of_two(unsigned long n)
  *  ob2(5) = 3
  *  ... and so on.
  */
-
-#define order_base_2(n) ilog2(roundup_pow_of_two(n))
+#define order_base_2(n) ((n) == 0 ? 0 : ilog2(roundup_pow_of_two(n)))
 
 #endif /* _LINUX_LOG2_H */