[babl] extensions/cairo: regain some lost speed for 32bit linear float display

[Øyvind Kolås <ok src gnome org>]

commit 67ce109d119f01907ed90dba43d231a63cca8520
Author: Øyvind Kolås <pippin gimp org>
Date:   Thu Dec 15 02:08:46 2016 +0100

    extensions/cairo: regain some lost speed for 32bit linear float display

 extensions/cairo.c |   27 +++++++++++++++++++++++++++
 1 files changed, 27 insertions(+), 0 deletions(-)
---
diff --git a/extensions/cairo.c b/extensions/cairo.c
index 6bfc574..cda5eaa 100644
--- a/extensions/cairo.c
+++ b/extensions/cairo.c
@@ -138,6 +138,30 @@ conv_yA16_cairo32_le (unsigned char *src, unsigned char *dst, long samples)
 }
 
 static long
+conv_rgbA_gamma_float_cairo32_le (unsigned char *src,
+                                  unsigned char *dst,
+                                  long           samples)
+{
+  float *fsrc = (float *) src;
+  unsigned char *cdst = (unsigned char *) dst;
+  int n = samples;
+
+  while (n--)
+    {
+      int val = fsrc[2] * 255.0f  + 0.5f;
+      *cdst++ = val >= 0xff ? 0xff : val <= 0 ? 0 : val;
+      val = fsrc[1] * 255.0f + 0.5f;
+      *cdst++ = val >= 0xff ? 0xff : val <= 0 ? 0 : val;
+      val = fsrc[0] * 255.0f + 0.5f;
+      *cdst++ = val >= 0xff ? 0xff : val <= 0 ? 0 : val;
+      val = fsrc[3] * 255.0f + 0.5f;
+      *cdst++ = val >= 0xff ? 0xff : val <= 0 ? 0 : val;
+      fsrc+=4;
+    }
+  return samples;
+}
+
+static long
 conv_rgbafloat_cairo32_le (unsigned char *src,
                            unsigned char *dst,
                            long           samples)
@@ -226,6 +250,9 @@ init (void)
       babl_conversion_new (babl_format ("RGBA float"), f32, "linear",
                            conv_rgbafloat_cairo32_le, NULL);
 
+      babl_conversion_new (babl_format ("R'aG'aB'aA float"), f32, "linear",
+                           conv_rgbA_gamma_float_cairo32_le, NULL);
+
       babl_conversion_new (babl_format ("R'G'B'A u8"), f24, "linear", 
                            conv_rgba8_cairo24_le, NULL);
       babl_conversion_new (babl_format ("R'G'B' u8"), f24, "linear",