Re: [bisected] kernel panic 2.6.22 -> 2.6.26-rc9+

On Wed, 23 Jul 2008 06:52:31 +0200 Willy Tarreau [off-list ref] wrote:

quoted

On Sun, Jul 13, 2008 at 02:55:25PM -0700, Andrew Morton wrote:

quoted

On Sun, 13 Jul 2008 23:51:06 +0200 Krzysztof Halasa [off-list ref] wrote:

quoted

Andrew Morton [off-list ref] writes:

quoted

--- a/include/asm-arm/bitops.h~a
+++ a/include/asm-arm/bitops.h

@@ -277,9 +277,16 @@ static inline int constant_fls(int x)
  * the clz instruction for much better code efficiency.
  */
 
-#define fls(x) \
+#define __fls(x) \
 	( __builtin_constant_p(x) ? constant_fls(x) : \
 	  ({ int __r; asm("clz\t%0, %1" : "=r"(__r) : "r"(x) : "cc"); 32-__r; }) )
+
+/* Implement fls() in C so that 64-bit args are suitably truncated */
+static inline int fls(int x)
+{
+	return __fls(x);
+}
+

Well, I like it more as it fixes all possible places instead of only
fls64().

But... can't we just move the #define body into the inline fls(x)?
Will there be other users of __fls(x)? It seems the
__builtin_constant_p(x) works for inline functions.

Could.  That was a minimal&safe thing.

quoted

The above patch fixes the kernel panic, too.

OK, thanks.

Andrew,

have you sent your patch to Linus ? I haven't seen it merged yet, and
I'd like to get it into -stable too.

This one?

From: Andrew Morton <akpm@linux-foundation.org>

arm's fls() is implemented as a macro, causing it to misbehave when passed
64-bit arguments.  Fix.

Cc: Nickolay Vinogradov <redacted>
Tested-by: Krzysztof Halasa <khc@pm.waw.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 include/asm-arm/bitops.h |    9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff -puN include/asm-arm/bitops.h~arm-fix-fls-for-64-bit-arguments include/asm-arm/bitops.h

--- a/include/asm-arm/bitops.h~arm-fix-fls-for-64-bit-arguments
+++ a/include/asm-arm/bitops.h

@@ -277,9 +277,16 @@ static inline int constant_fls(int x)
  * the clz instruction for much better code efficiency.
  */
 
-#define fls(x) \
+#define __fls(x) \
 	( __builtin_constant_p(x) ? constant_fls(x) : \
 	  ({ int __r; asm("clz\t%0, %1" : "=r"(__r) : "r"(x) : "cc"); 32-__r; }) )
+
+/* Implement fls() in C so that 64-bit args are suitably truncated */
+static inline int fls(int x)
+{
+	return __fls(x);
+}
+
 #define ffs(x) ({ unsigned long __t = (x); fls(__t & -__t); })
 #define __ffs(x) (ffs(x) - 1)
 #define ffz(x) __ffs( ~(x) )

_