I hadn't had enough coffee when I wrote this. Currently, the final
increment of buf depends on the value loaded from the table, and
causes gcc to emit a cmov immediately before the return. It is smarter
to let it depend on r, since the increment can then be computed in
parallel with the final load/store pair. It also shaves 16 bytes of
.text.
Signed-off-by: Rasmus Villemoes <li...@rasmusvillemoes.dk>
---
lib/vsprintf.c | 10 ++++------
1 file changed, 4 insertions(+), 6 deletions(-)
diff --git a/lib/vsprintf.c b/lib/vsprintf.c
index e0bea9e5bbbf..7078d90c187b 100644
--- a/lib/vsprintf.c
+++ b/lib/vsprintf.c
@@ -164,9 +164,9 @@ static const u16 decpair[100] = {
/*
* This will print a single '0' even if r == 0, since we would
- * immediately jump to out_r where two 0s would be written and one of
- * them then discarded. This is needed by ip4_string below. All other
- * callers pass a non-zero value of r.
+ * immediately jump to out_r where two 0s would be written but only
+ * one of them accounted for in buf. This is needed by ip4_string
+ * below. All other callers pass a non-zero value of r.
*/
static noinline_for_stack
char *put_dec_trunc8(char *buf, unsigned r)
@@ -205,9 +205,7 @@ out_q:
out_r:
/* 1 <= r < 100 */
*((u16 *)buf) = decpair[r];
- buf += 2;
- if (buf[-1] == '0')
- buf--;
+ buf += r < 10 ? 1 : 2;
return buf;
}
--
2.1.3
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
