fivedev/harbour-core
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

2011-05-17 19:21 UTC+0200 Viktor Szakats (harbour.01 syenar.hu)

Browse Source 
* src/rtl/hbrandom.c
    ! fixed wrong random numbers being generated after:
        2011-03-11 09:42 UTC+0100 Przemyslaw Czerpak
      this is simple revert, Przemek pls recheck/refix original issue.
      Values 0x????FFFF and 0x????0000 were returned from these calls:
         hb_randomInt( 0xFFFFFFFF )
         hb_randomInt( 0xFFFFFFFE )
         hb_randomInt( 0x7FFFFFFF )
      under mingw32 and mingw64 (both win and linux builds)
This commit is contained in:
Viktor Szakats
2011-05-17 17:22:27 +00:00
parent 14d0d4161d
commit c5a73944e2
2 changed files with 25 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
harbour/ChangeLog
View File

@@ -16,13 +16,24 @@
The license applies to all entries newer than 2009-04-28.
*/
2011-05-17 19:21 UTC+0200 Viktor Szakats (harbour.01 syenar.hu)
* src/rtl/hbrandom.c
! fixed wrong random numbers being generated after:
2011-03-11 09:42 UTC+0100 Przemyslaw Czerpak
this is simple revert, Przemek pls recheck/refix original issue.
Values 0x????FFFF and 0x????0000 were returned from these calls:
hb_randomInt( 0xFFFFFFFF )
hb_randomInt( 0xFFFFFFFE )
hb_randomInt( 0x7FFFFFFF )
under mingw32 and mingw64 (both win and linux builds)
2011-05-17 19:02 UTC+0200 Viktor Szakats (harbour.01 syenar.hu)
* include/harbour.hbx
* src/rtl/hbrandom.c
+ HB_RAND32() -> <32-bit random number>
; TOFIX: random numbers are broken under mingw 4.5.2, only high 16bits are random.
in all HB_RAND*() functions
in all HB_RAND*() functions [DONE]
2011-05-17 17:48 UTC+0200 Viktor Szakats (harbour.01 syenar.hu)
* utils/hbmk2/hbmk2.prg

13
harbour/src/rtl/hbrandom.c
View File

@@ -141,7 +141,20 @@ double hb_random_num()
}
d1 = ( double ) rand();
#ifdef __HB_THIS_FIX_BREAKS_MINGW32_64__
d2 = ( double ) RAND_MAX + 1.0;
#else
d2 = ( double ) RAND_MAX;
#if defined( __BORLANDC__ )
/* It seems that on Windows platform there some weirdness about EPSILON value so
that a float division using an epsilon smaller than 1e-10 may be rounded.
Must dig if it's a borland lib bug or a windows problem.
*/
d2 += 0.001;
#else
d2 += DBL_EPSILON;
#endif
#endif
return d1 / d2;
}