[PATCH] erofs-utils: use cmpsgn(x, y) for standardized large value comparsion

Gao Xiang hsiangkao at aol.com
Mon Oct 14 22:30:48 AEDT 2019


Previously, roundup(bb->buffers.off % EROFS_BLKSIZ, alignsize)
+ incr + extrasize is a unsigned 64bit value and sgn(x) didn't
work properly. Fix it.

Signed-off-by: Gao Xiang <hsiangkao at aol.com>
---
 include/erofs/defs.h | 5 +++--
 lib/cache.c          | 6 +++---
 2 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/include/erofs/defs.h b/include/erofs/defs.h
index 15db4e3..db51350 100644
--- a/include/erofs/defs.h
+++ b/include/erofs/defs.h
@@ -136,9 +136,10 @@ typedef int64_t         s64;
 	type __max2 = (y);			\
 	__max1 > __max2 ? __max1: __max2; })
 
-#define sgn(x) ({		\
+#define cmpsgn(x, y) ({		\
 	typeof(x) _x = (x);	\
-(_x > 0) - (_x < 0); })
+	typeof(y) _y = (y);	\
+(_x > _y) - (_x < _y); })
 
 #define ARRAY_SIZE(arr)	(sizeof(arr) / sizeof((arr)[0]))
 
diff --git a/lib/cache.c b/lib/cache.c
index 41d2d5d..e61b201 100644
--- a/lib/cache.c
+++ b/lib/cache.c
@@ -80,9 +80,9 @@ static int __erofs_battach(struct erofs_buffer_block *bb,
 			   bool dryrun)
 {
 	const erofs_off_t alignedoffset = roundup(bb->buffers.off, alignsize);
-	const int oob = sgn(roundup(bb->buffers.off % EROFS_BLKSIZ,
-				    alignsize) + incr + extrasize -
-			    EROFS_BLKSIZ);
+	const int oob = cmpsgn(roundup(bb->buffers.off % EROFS_BLKSIZ,
+				       alignsize) + incr + extrasize,
+			       EROFS_BLKSIZ);
 	bool tailupdate = false;
 	erofs_blk_t blkaddr;
 
-- 
2.17.1



More information about the Linux-erofs mailing list