Fix crash on scaling on OS X.

[dcpomatic.git] / src / lib / image.cc

diff --git a/src/lib/image.cc b/src/lib/image.cc
index 5c04f70e6ab40f0e4c5ee1afd32f0ff6a14b2a55..dbea62d26b3b7363d7c33ce7df19dff170430a5f 100644 (file)
--- a/src/lib/image.cc
+++ b/src/lib/image.cc
@@ -420,9 +420,19 @@ Image::allocate ()
        _stride[0] = _stride[1] = _stride[2] = _stride[3] = 0;
 
        for (int i = 0; i < components(); ++i) {
-               _line_size[i] = _size.width * bytes_per_pixel(i);
+               _line_size[i] = ceil (_size.width * bytes_per_pixel(i));
                _stride[i] = stride_round_up (i, _line_size, _aligned ? 32 : 1);
-               _data[i] = (uint8_t *) av_malloc (_stride[i] * lines (i));
+
+               /* The assembler function ff_rgb24ToY_avx (in libswscale/x86/input.asm)
+                  uses a 16-byte fetch to read three bytes (R/G/B) of image data.
+                  Hence on the last pixel of the last line it reads over the end of
+                  the actual data by 1 byte.  If the width of an image is a multiple
+                  of the stride alignment there will be no padding at the end of image lines.
+                  OS X crashes on this illegal read, though other operating systems don't
+                  seem to mind.  The nasty + 1 in this malloc makes sure there is always a byte
+                  for that instruction to read safely.
+               */
+               _data[i] = (uint8_t *) av_malloc (_stride[i] * lines (i) + 1);
        }
 }