Copy head@r302406 to stable/11 as part of the 11.0-RELEASE cycle.
Prune svn:mergeinfo from the new branch, as nothing has been merged
here.

Additional commits post-branch will follow.

Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation

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| <= 2**19pi/2 and a smaller speedup for larger x, and a
small speeddown for |x| <= 9pi/4 (only 1-2 cycles average, but that
is 4%).

Inlining this is less likely to bust caches than inlining the float
version since it is much smaller (about 220 bytes text and rodata) and
has many fewer branches. However, the float version was already large
due to its manual inlining of the branches and also the polynomial
evaluations.

Change __ieee754_rem_pio2f() to return double instead of float so that
this function and its callers cosf(), sinf() and tanf() don't waste time
converting values from doubles to floats and back for |x| > 9pi/4.
All these functions were optimized a few years ago to mostly use doubles
internally and across the __kernel*() interfaces but not across the
__ieee754_rem_pio2f() interface.

This saves about 40 cycles in cosf(), sinf() and tanf() for |x| > 9pi/4
on amd64 (A64), and about 20 cycles on i386 (A64) (except for cosf()
and sinf() in the upper range). 40 cycles is about 35% for |x| < 9pi/4
<= 2**19pi/2 and about 5% for |x| > 2**19pi/2. The saving is much
larger on amd64 than on i386 since the conversions are not easy to
optimize except on i386 where some of them are automatic and others
are optimized invalidly. amd64 is still about 10% slower in cosf()
and tanf() in the lower range due to conversion overhead.

This also gives a tiny speedup for |x| <= 9pi/4 on amd64 (by simplifying
the code). It also avoids compiler bugs and/or additional slowness
in the conversions on (not yet supported) machines where double_t !=
double.

s/rcsid/__FBSDID/

Fixed cosf(x) when x is a "negative" NaNs. I broke this in rev.1.10.
cosf(x) is supposed to return something like x when x is a NaN, and
we actually fairly consistently return x-x which is normally very like
x (on i386 and and it is x if x is a quiet NaN and x with the quiet bit
set if x is a signaling NaN. Rev.1.10 broke this by normalising x to
fabsf(x). It's not clear if fabsf(x) is should preserve x if x is a NaN,
but it actually clears the sign bit, and other parts of the code depended
on this.

The bugs can be fixed by saving x before normalizing it, and using the
saved x only for NaNs, and using uint32_t instead of int32_t for ix
so that negative NaNs are not misclassified even if fabsf() doesn't
clear their sign bit, but gcc pessimizes the saving very well, especially
on Athlon XPs (it generates extra loads and stores, and mixes use of
the SSE and i387, and this somehow messes up pipelines). Normalizing
x is not a very good optimization anyway, so stop doing it. (It adds
latency to the FPU pipelines, but in previous versions it helped except
for |x| <= 3pi/4 by simplifying the integer pipelines.) Use the same
organization as in s_sinf.c and s_tanf.c with some branches reordered.
These changes combined recover most of the performance of the unfixed
version on A64 but still lose 10% on AXP with gcc-3.4 -O1 but not with
gcc-3.3 -O1.

Exploit skew-symmetry to avoid an operation: -sin(x-A) = sin(A-x). This
gives a tiny but hopefully always free optimization in the 2 quadrants
to which it applies. On Athlons, it reduces maximum latency by 4 cycles
in these quadrants but has usually has a smaller effect on total time
(typically ~2 cycles (~5%), but sometimes 8 cycles when the compiler
generates poor code).

Try to use the "proximity" (~) operator consistently in comments
(x ~<= a, not x <= ~a). This got messed up in some places when the
comments were moved from e_rem_pio2f.c.

Added my (non-)copyright.

Use only double precision for "kernel" cosf and sinf (except for
returning float). The functions are renamed from __kernel_{cos,sin}f()
to __kernel_{cos,sin}df() so that misuses of them will cause link errors
and not crashes.

This version is an almost-routine translation with no special optimizations
for accuracy or efficiency. The not-quite-routine part is that in
__kernel_cosf(), regenerating the minimax polynomial with double
precision coefficients gives a coefficient for the x**2 term that is
not quite -0.5, so the literal 0.5 in the code and the related `hz'
variable need to be modified; also, the special code for reducing the
error in 1.0-x**2*0.5 is no longer needed, so it is convenient to
adjust all the logic for the x**2 term a little. Note that without
extra precision, it would be very bad to use a coefficient of other
than -0.5 for the x**2 term -- the old version depends on multiplication
by -0.5 being infinitely precise so as not to need even more special
code for reducing the error in 1-x**2*0.5.

This gives an unimportant increase in accuracy, from ~0.8 to ~0.501
ulps. Almost all of the error is from the final rounding step, since
the choice of the minimax polynomials so that their contribution to the
error is a bit less than 0.5 ulps just happens to give contributions that
are significantly less (~.001 ulps).

An Athlons, for uniformly distributed args in [-2pi, 2pi], this gives
overall speed increases in the 10-20% range, despite giving a speed
decrease of typically 19% (from 31 cycles up to 37) for sinf() on args
in [-pi/4, pi/4].

Mess up the "kernel" float trig function .c files with ifdefs so that
they can be #included in other .c files to give inline functions, and
use them to inline the functions in most callers (not in e_lgammaf_r.c).
__kernel_tanf() is too large and complicated for gcc to inline very well.

An athlons, this gives a speed increase under favourable pipeline
conditions of about 10% overall (larger for AXP, smaller for A64).
E.g., on AXP, sinf() on uniformly distributed args in [-2Pi, 2Pi]
now takes 30-56 cycles; it used to take 45-61 cycles; hardware fsin
takes 65-129.

Moved all the optimizations for |x| <= 9pi/2 from
__ieee754_rem_pio2f() to its 3 callers and manually inline them.

On Athlons, with favourable compiler flags and optimizations and
favourable pipeline conditions, this gives a speedup of 30-40 cycles
for cosf(), sinf() and tanf() on the range pi/4 < |x| <= 9pi/4, so
thes functions are now signifcantly faster than the hardware trig
functions in many cases. E.g., in a benchmark with uniformly distributed
x in [-2pi, 2pi], A64 hardware fcos took 72-129 cycles and cosf() took
37-55 cycles. Out-of-order execution is needed to get both of these
times. The optimizations in this commit apparently work more by
removing 1 serialization point than by reducing latency.

Minor cleanups:

s_cosf.c and s_sinf.c:
Use a non-bogus magic constant for the threshold of pi/4. It was 2 ulps
smaller than pi/4 rounded down, but its value is not critical so it should
be the result of natural rounding.

s_cosf.c and s_tanf.c:
Use a literal 0.0 instead of an unnecessary variable initialized to
[(float)]0.0. Let the function prototype convert to 0.0F.

Improved wording in some comments.

Attempted to improve indentation of comments.

Moved the optimization for tiny x from __kernel_{cos,sin}[f](x) to
{cos_sin}[f](x) so that x doesn't need to be reclassified in the
"kernel" functions to determine if it is tiny (it still needs to be
reclassified in the cosine case for other reasons that will go away).

This optimization is quite large for exponentially distributed x, since
x is tiny for almost half of the domain, but it is a pessimization for
uniformally distributed x since it takes a little time for all cases
but rarely applies. Arg reduction on exponentially distributed x
rarely gives a tiny x unless the reduction is null, so it is best to
only do the optimization if the initial x is tiny, which is what this
commit arranges. The imediate result is an average optimization of
1.4% relative to the previous version in a case that doesn't favour
the optimization (double cos(x) on all float x) and a large
pessimization for the relatively unimportant cases of lgamma[f][_r](x)
on tiny, negative, exponentially distributed x. The optimization should
be recovered for lgamma*() as part of fixing lgamma*()'s low-quality
arg reduction.

Fixed various wrong constants for the cutoff for "tiny". For cosine,
the cutoff is when x**2/2! == {FLT or DBL}_EPSILON/2. We round down
to an integral power of 2 (and for cos() reduce the power by another
1) because the exact cutoff doesn't matter and would take more work
to determine. For sine, the exact cutoff is larger due to the ration
of terms being x**2/3! instead of x**2/2!, but we use the same cutoff
as for cosine. We now use a cutoff of 2**-27 for double precision and
2**-12 for single precision. 2**-27 was used in all cases but was
misspelled 2**27 in comments. Wrong and sloppy cutoffs just cause
missed optimizations (provided the rounding mode is to nearest --
other modes just aren't supported).

Fix formatting, this is hard to explain, so I'll show one example.

- float ynf(int n, float x) /* wrapper ynf */
+float
+ynf(int n, float x) /* wrapper ynf */

This is because the __STDC__ stuff was indented.

Reviewed by: md5

Assume __STDC__, remove non-__STDC__ code.

Reviewed by: md5

$Id$ -> $FreeBSD$

Revert $FreeBSD$ to $Id$

Make the long-awaited change from $Id$ to $FreeBSD$

This will make a number of things easier in the future, as well as (finally!)
avoiding the Id-smashing problem which has plagued developers for so long.

Boy, I'm glad we're not using sup anymore. This update would have been
insane otherwise.

Remove trailing whitespace.

This commit was generated by cvs2svn to compensate for changes in r2116,
which included commits to RCS files with non-trunk default branches.

J.T. Conklin's latest version of the Sun math library.

-- Begin comments from J.T. Conklin:
The most significant improvement is the addition of "float" versions
of the math functions that take float arguments, return floats, and do
all operations in floating point. This doesn't help (performance)
much on the i386, but they are still nice to have.

The float versions were orginally done by Cygnus' Ian Taylor when
fdlibm was integrated into the libm we support for embedded systems.
I gave Ian a copy of my libm as a starting point since I had already
fixed a lot of bugs & problems in Sun's original code. After he was
done, I cleaned it up a bit and integrated the changes back into my
libm.
-- End comments

Reviewed by: jkh
Submitted by: jtc