[truetype] Tweak the previous commit.

* src/truetype/ttinterp.c (Current_Ratio): Put unit vector
components as the second TT_MulFix14 arguments. This is required
on 16-bit systems.
2.6.5
Alexei Podtelezhnikov 12 years ago
parent 5ad05f97b6
commit cece4d3b5d
  1. 8
      ChangeLog
  2. 10
      src/truetype/ttinterp.c

@ -1,3 +1,11 @@
2012-12-06 Alexei Podtelezhnikov <apodtele@gmail.com>
[truetype] Tweak the previous commit.
* src/truetype/ttinterp.c (Current_Ratio): Put unit vector
components as the second TT_MulFix14 arguments. This is required
on 16-bit systems.
2012-12-06 Alexei Podtelezhnikov <apodtele@gmail.com>
[truetype] Microoptimizations in bytecode interpreter.

@ -233,7 +233,7 @@
/*************************************************************************/
/* */
/* This macro computes (a*2^14)/b and compliments TT_MulFix14. */
/* This macro computes (a*2^14)/b and complements TT_MulFix14. */
/* */
#define TT_DivFix14( a, b ) \
FT_DivFix( a, (b) << 2 )
@ -1671,10 +1671,10 @@
FT_F26Dot6 x, y;
x = TT_MulFix14( CUR.GS.projVector.x,
CUR.tt_metrics.x_ratio );
y = TT_MulFix14( CUR.GS.projVector.y,
CUR.tt_metrics.y_ratio );
x = TT_MulFix14( CUR.tt_metrics.x_ratio,
CUR.GS.projVector.x );
y = TT_MulFix14( CUR.tt_metrics.y_ratio,
CUR.GS.projVector.y );
CUR.tt_metrics.ratio = TT_VecLen( x, y );
}
}

Loading…
Cancel
Save