isa/mutter
1
0
Fork 0
mirror of https://github.com/brl/mutter.git synced 2024-11-12 17:27:03 -05:00
Code Issues Packages Projects Releases Wiki Activity

work around punned-pointer problem

Browse Source
This commit is contained in:
Tomas Frydrych 2007-01-19 20:00:41 +00:00
parent ee6b583ec1
commit b8b0d01c91
2 changed files with 22 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
ChangeLog
View File

@ -1,3 +1,10 @@
2007-01-19 Tomas Frydrych <tf@openedhand.com>
* clutter/clutter-fixed.c: (clutter_sqrti):
Use union instead of casting int <-> float to get rid of
punned-pointer warning and avoid gcc optimatisation breaking
the function on arm.
2007-01-19 Emmanuele Bassi <ebassi@openedhand.com>
Allow the ClutterGroup subclasses to override the add and

27
clutter/clutter-fixed.c
View File

@ -470,17 +470,21 @@ clutter_sqrti (gint number)
* elsewhere in clutter is not good enough, and 10.22 is used instead.
*/
ClutterFixed x;
unsigned long y, y1; /* 10.22 fixed point */
unsigned long y1; /* 10.22 fixed point */
unsigned long f = 0x600000; /* '1.5' as 10.22 fixed */
float flt = number;
float flt2;
union
{
float f;
unsigned long i;
} flt, flt2;
flt.f = number;
x = CLUTTER_INT_TO_FIXED (number) / 2;
/* The QIII initial estimate */
y = * ( unsigned long * ) &flt;
y = 0x5f3759df - ( y >> 1 );
flt = * ( float * ) &y;
flt.i = 0x5f3759df - ( flt.i >> 1 );
/* Now, we convert the float to 10.22 fixed. We exploit the mechanism
* described at http://www.d6.com/users/checker/pdfs/gdmfp.pdf.
@ -494,22 +498,21 @@ clutter_sqrti (gint number)
* addition out, and it all goes pear shape, since without it, the bits
* in the float will not be correctly aligned.
*/
flt2 = flt + 2.0;
y = * ( long * ) &flt2;
y &= 0x7FFFFF;
flt2.f = flt.f + 2.0;
flt2.i &= 0x7FFFFF;
/* Now we correct the estimate, only single iterration is needed */
y1 = (y >> 11) * (y >> 11);
y1 = (flt2.i >> 11) * (flt2.i >> 11);
y1 = (y1 >> 8) * (x >> 8);
y1 = f - y1;
y = (y >> 11) * (y1 >> 11);
flt2.i = (flt2.i >> 11) * (y1 >> 11);
/* Invert, round and convert from 10.22 to an integer
* 0x1e3c68 is a magical rounding constant that produces slightly
* better results than 0x200000.
*/
return (number * y + 0x1e3c68) >> 22;
return (number * flt2.i + 0x1e3c68) >> 22;
}
/* <private> */