mirrors/freetype
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix Savannah bug #35660.

Browse Source 
For some divisions, we use casts to 32bit entities.  Always guard
against division by zero with these casts also.

* src/base/ftcalc.c (ft_div64by32): Remove redundant cast.
(FT_MulDiv, FT_MulDiv_No_Round): Add 32bit cast.
(FT_DivFix): Add 32bit cast (this omission triggered the bug).
This commit is contained in:
Werner Lemberg 2012-03-03 12:27:18 +01:00
parent 847502a217
commit ba67957d5e
2 changed files with 20 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
ChangeLog
View File

@ -1,3 +1,14 @@
2012-03-03 Werner Lemberg <wl@gnu.org>
Fix Savannah bug #35660.
For some divisions, we use casts to 32bit entities. Always guard
against division by zero with these casts also.
* src/base/ftcalc.c (ft_div64by32): Remove redundant cast.
(FT_MulDiv, FT_MulDiv_No_Round): Add 32bit cast.
(FT_DivFix): Add 32bit cast (this omission triggered the bug).
2012-03-03 Werner Lemberg <wl@gnu.org>
[psaux] Fix handling of track kerning.

17
src/base/ftcalc.c
View File

@ -4,7 +4,7 @@
/* */
/* Arithmetic computations (body). */
/* */
/* Copyright 1996-2001, 2002, 2003, 2004, 2005, 2006, 2008 by */
/* Copyright 1996-2006, 2008, 2012 by */
/* David Turner, Robert Wilhelm, and Werner Lemberg. */
/* */
/* This file is part of the FreeType project, and may only be used, */
@ -307,7 +307,7 @@
q <<= 1;
r |= lo >> 31;
if ( r >= (FT_UInt32)y )
if ( r >= y )
{
r -= y;
q |= 1;
@ -373,7 +373,7 @@
if ( a <= 46340L && b <= 46340L && c <= 176095L && c > 0 )
a = ( a * b + ( c >> 1 ) ) / c;
else if ( c > 0 )
else if ( (FT_Int32)c > 0 )
{
FT_Int64 temp, temp2;
@ -412,7 +412,7 @@
if ( a <= 46340L && b <= 46340L && c > 0 )
a = a * b / c;
else if ( c > 0 )
else if ( (FT_Int32)c > 0 )
{
FT_Int64 temp;
@ -544,7 +544,7 @@
s = (FT_Int32)a; a = FT_ABS( a );
s ^= (FT_Int32)b; b = FT_ABS( b );
if ( b == 0 )
if ( (FT_UInt32)b == 0 )
{
/* check for division by 0 */
q = (FT_UInt32)0x7FFFFFFFL;
@ -552,15 +552,16 @@
else if ( ( a >> 16 ) == 0 )
{
/* compute result directly */
q = (FT_UInt32)( (a << 16) + (b >> 1) ) / (FT_UInt32)b;
q = (FT_UInt32)( ( a << 16 ) + ( b >> 1 ) ) / (FT_UInt32)b;
}
else
{
/* we need more bits; we have to do it by hand */
FT_Int64 temp, temp2;
temp.hi = (FT_Int32) (a >> 16);
temp.lo = (FT_UInt32)(a << 16);
temp.hi = (FT_Int32) ( a >> 16 );
temp.lo = (FT_UInt32)( a << 16 );
temp2.hi = 0;
temp2.lo = (FT_UInt32)( b >> 1 );
FT_Add64( &temp, &temp2, &temp );