libroot: Replace most of libm with musl's.

The glibc libm code was showing its age, and has recently been
the subject of a number of tickets about its inaccuracy.
Additionally, some developers have complained about
how convoluted the headers are, and thus how hard it is
to add support for new architectures (and how flaky
the support for the existing architectures is.)

So, with this commit, nearly the entire glibc libm has been
gutted and replaced with the one from musl 1.1.24.

The complex functions from glibc are retained (as they
are more mature than musl's), as are some glibc-internal
libm functions.

This also has the advantage that these functions are
actually using our <math.h>, whereas GCC used its own,
which was rather dangerous for obvious reasons.

Additionally, the new math functions are always compiled
with GCC 8 (even on x86_gcc2), as it seems GCC 2 does
not quite understand some of the union-aliasing they
use (a lot of which was added in C99, I suppose.)
FFmpeg on x86_gcc2 is already compiled with GCC 8
and that has so far worked out well, so there should
not be any problems caused by this.

I did verify that ARM and PPC at least still compile,
though other architectures may require a bit more work
(they are not bootstrapped so I could not do much.)

Should fix #14933 among other issues.

Change-Id: Ifeea0ddab23a8d0480fc26dece1b0192afc263bd

Synchronize sysdep.h with current glibc

So that I can more easily import sparc headers from it.

NO_UNDERSCORES is gone and is now always the default.
Some previously arch-specific defines are now valid for all platforms.

Change-Id: If9876b241719559bdcb5cd9d8b1dc97e5e3d96b3
Reviewed-on: https://review.haiku-os.org/c/1141
Reviewed-by: waddlesplash <waddlesplash@gmail.com>

* Removed libroot/posix/math completely and added the respective
functions from glibc 2.3.2. Adding those one by one is extremely
boring and time-consuming work. We should probably just remove the
math stuff again, add a verbatim copy of the glibc directories, and
shuffle things around afterwards.
* Note that for x86 holds sizeof(long double) == 96. I removed the
generic 128 bit long double functions that were in use before and
added respective 96 bit versions to arch/x86 (or left them out, if
they weren't used). For PPC the System V ABI spec supplement specifies
sizeof(long double) == 128, but I had trouble configuring gcc to
actually use that setting (looks like there simply is no support for
TI mode for PPC 32). So we have to live with sizeof(long double) ==
sizeof(double) == 64 for the time being.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15477 a95241bf-73f2-0310-859d-f6bbb57e9c96

Added PPC specific and some generic glibc headers. Taken from
glibc 2.3.2, which seems to be the version used in other places,
though I also found older files.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15402 a95241bf-73f2-0310-859d-f6bbb57e9c96

* Removed libroot/posix/math completely and added the respective
functions from glibc 2.3.2. Adding those one by one is extremely
boring and time-consuming work. We should probably just remove the
math stuff again, add a verbatim copy of the glibc directories, and
shuffle things around afterwards.
* Note that for x86 holds sizeof(long double) == 96. I removed the
generic 128 bit long double functions that were in use before and
added respective 96 bit versions to arch/x86 (or left them out, if
they weren't used). For PPC the System V ABI spec supplement specifies
sizeof(long double) == 128, but I had trouble configuring gcc to
actually use that setting (looks like there simply is no support for
TI mode for PPC 32). So we have to live with sizeof(long double) ==
sizeof(double) == 64 for the time being.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15477 a95241bf-73f2-0310-859d-f6bbb57e9c96

Added PPC specific and some generic glibc headers. Taken from
glibc 2.3.2, which seems to be the version used in other places,
though I also found older files.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15402 a95241bf-73f2-0310-859d-f6bbb57e9c96