babl: adjust default BABL_TOLERANCE

This as a start of fixing issue #49, lower precision code gets generated
on AMD EPYC, due to use of rcpps to get a reciprocal - which has lower
precision on AMD EPYC. The conversions with AMD EPYC gets included with
a small margin - so we should not be shedding many other fast conversions
due to this.

diff --git a/babl/babl-fish-path.c b/babl/babl-fish-path.c
index c8911d192b64563814cc15d79de1d31dda155d10..8d86e627e78447535c1e44aaff51422ba54b7071 100644 (file)
--- a/babl/babl-fish-path.c
+++ b/babl/babl-fish-path.c
@@ -21,7 +21,7 @@
 #include "babl-internal.h"
 #include "babl-ref-pixels.h"
 
-#define BABL_TOLERANCE             0.0000047
+#define BABL_TOLERANCE             0.0000041
 #define BABL_MAX_COST_VALUE        2000000
 #define BABL_HARD_MAX_PATH_LENGTH  8
 #define BABL_MAX_NAME_LEN          1024