[WIN32SS][NTGDI] Don't use FLOAT in IntEscapeMatrix (#1572)

Appendum to 1a40284. Use a 16.16 fixed point value instead of FLOAT. CORE-15838
This commit is contained in:
Katayama Hirofumi MZ 2019-05-23 19:40:54 +09:00 committed by GitHub
parent e5bffe49da
commit 0137096449
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23

View file

@ -719,8 +719,8 @@ VOID FASTCALL IntWidthMatrix(FT_Face face, FT_Matrix *pmat, LONG lfWidth)
VOID FASTCALL IntEscapeMatrix(FT_Matrix *pmat, LONG lfEscapement)
{
FT_Vector vecAngle;
/* Convert from angle in tenths of degrees to 'FT_Angle' degrees */
FT_Angle angle = FT_FixedFromFloat((FLOAT)lfEscapement / 10);
/* Convert the angle in tenths of degrees into degrees as a 16.16 fixed-point value */
FT_Angle angle = INT_TO_FIXED(lfEscapement) / 10;
FT_Vector_Unit(&vecAngle, angle);
pmat->xx = vecAngle.x;
pmat->xy = -vecAngle.y;