bde 2008-02-14 15:10:34 UTC
FreeBSD src repository
Modified files:
lib/msun/src s_ceil.c s_ceill.c s_floor.c s_floorl.c
s_trunc.c s_truncl.c
Log:
Oops, the weak reference for ceill(), floorl() and truncl() was in the
wrong file. This broke
bde 2008-02-15 07:01:40 UTC
FreeBSD src repository
Modified files:
lib/msun/src s_ceil.c s_floor.c s_trunc.c
Log:
Sigh, the weak reference for ceill(), floorl() and truncl() was in
unreachable code due to a missing include. This kept arm and powerpc
broken.
bde 2008-02-18 14:02:12 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c s_cos.c s_sin.c s_tan.c
Log:
Inline __ieee754__rem_pio2(). With gcc4-2, this gives an average
optimization of about 10% for cos(x), sin(x) and tan(x) on
|x| < 2**19*pi/2.
bde 2008-02-18 14:09:41 UTC
FreeBSD src repository
Modified files:
lib/msun/ld80k_tanl.c
Log:
Fix a typo which broke k_tanl.c on !(amd64 || i386).
Revision ChangesPath
1.2 +1 -1 src/lib/msun/ld80/k_tanl.c
bde 2008-02-18 15:39:52 UTC
FreeBSD src repository
Modified files:
lib/msun/ld80k_tanl.c
Log:
2 long double constants were missing L suffixes. This helped break tanl()
on !(amd64 || i386). It gave slightly worse than double precision in some
cases. tanl() now p
bde 2008-02-19 12:54:14 UTC
FreeBSD src repository
Modified files:
lib/msun/src k_cos.c k_sin.c
Log:
Rearrange the polynomial evaluation for better parallelism. This
saves an average of about 8 cycles or 5% on A64 (amd64 and i386 --
more in cycles but about the s
bde 2008-02-19 15:42:46 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2f.c
Log:
Merge cosmetic changes from e_rem_pio2.c 1.10 (convert to __FBSDID();
fix indentation and return type of __ieee754_rem_pio2()).
Remove unused variables.
Revi
bde 2008-02-19 15:30:58 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c
Log:
Optimize for 3pi/4 <= |x| <= 9pi/4 in much the same way as for
pi/4 <= |x| <= 3pi/4. Use the same branch ladder as for float precision.
Remove the optimization for |
bde 2008-02-22 09:21:14 UTC
FreeBSD src repository
Modified files:
lib/msun/src s_rintl.c
Log:
Fix rintl() on signaling NaNs and unsupported formats.
Revision ChangesPath
1.3 +3 -5 src/lib/msun/src/s_rintl.c
_
bde 2008-02-22 10:04:53 UTC
FreeBSD src repository
Modified files:
lib/msun/src s_rintl.c
Log:
Optimize the fixup for +-0 by using better classification for this case
and by using a table lookup to avoid a branch when this case occurs.
On i386, this saves 1-4 cycl
bde 2008-02-22 11:59:05 UTC
FreeBSD src repository
Modified files:
lib/msun/src s_rintl.c
Log:
Optimize the conversion to bits a little (by about 11 cycles or 16%
on i386 (A64), 5 cycles on amd64 (A64), and 3 cycles on ia64). gcc
tends to generate very bad code f
bde 2008-02-22 14:11:03 UTC
FreeBSD src repository
Modified files:
lib/msun/src math_private.h
Log:
Add an irint() function in inline asm for amd64 and i386. irint() is
the same as lrint() except it returns int instead of long. Though the
extern lrint() is fairl
bde 2008-02-22 15:55:15 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c e_rem_pio2f.c
Log:
Optimize the 9pi/2 < |x| <= 2**19pi/2 case on amd64 and i386 by avoiding
the the double to int conversion operation which is very slow on these
arches.
bde 2008-02-22 17:26:24 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c
Log:
Remove the "quick check no cancellation" optimization for
9pi/2 < |x| < 32pi/2 since it is only a small or negative optimation
and it gets in the way of further optim
bde 2008-02-22 18:43:23 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c e_rem_pio2f.c
Log:
Avoid using FP-to-integer conversion for !(amd64 || i386) too. Use the
FP-to-FP method to round to an integer on all arches, and convert this
to an int
bde 2008-02-23 12:53:21 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c e_rem_pio2f.c
Log:
Optimize the 9pi/2 < |x| <= 2**19pi/2 case some more by avoiding an
fabs(), a conditional branch, and sign adjustments of 3 variables for
x < 0 when the
bde 2008-02-25 11:43:20 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c e_rem_pio2f.c k_rem_pio2.c
Log:
Fix some off-by-1 errors.
e_rem_pio2.c:
Float and double precision didn't work because init_jk[] was 1 too small.
It needs to be 2 lar
bde 2008-02-25 13:33:20 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2f.c math_private.h s_cosf.c
s_sinf.c s_tanf.c
Log:
Change __ieee754_rem_pio2f() to return double instead of float so that
this function and its caller
bde 2008-02-25 18:28:58 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c e_rem_pio2f.c
Log:
Use a temporary array instead of the arg array y[] for calling
__kernel_rem_pio2(). This simplifies analysis of aliasing and thus
results in better cod
bde 2008-02-25 22:19:17 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2f.c s_cosf.c s_sinf.c s_tanf.c
Log:
Inline __ieee754__rem_pio2f(). On amd64 (A64) and i386 (A64), this
gives an average speedup of about 12 cycles or 17% for
9pi/4 < |x| <=
bde 2008-02-28 16:22:36 UTC
FreeBSD src repository
Modified files:
lib/msun/src e_rem_pio2.c e_rem_pio2f.c
Log:
Fix and improve some magic numbers for the "medium size" case.
e_rem_pio2.c:
This case goes up to about 2**20pi/2, but the comment about it said that
501 - 521 of 521 matches
Mail list logo
