configure: Recognize the host_cpu value "armv8l"

configure: Allow specifying the page size.

Fixes https://github.com/NVIDIA/libglvnd/issues/132

Document EGL vendor ICD search paths

The behaviour of a relative `library_path` is deliberately left
unspecified here. As an implementation detail, it's currently
passed directly to `dlopen()`, so in practice it will be interpreted as
relative to the process's current working directory, but that doesn't
seem practically useful. Vulkan treats relative paths as being relative
to the JSON file in which they were found, which seems more likely to
be a helpful behaviour.

Otherwise this documents the current behaviour, without any of the
possible changes suggested on #187. If any of those changes are made,
then this document should be updated at the same time.

Signed-off-by: Simon McVittie <smcv@collabora.com>

The configure script now checks the environment variable GLDISPATCH_PAGE_SIZE.
If it's set, then it overrides the GLDISPATCH_PAGE_SIZE macro used in
libGLdispatch to align the assembly dispatch stubs.

This is mainly useful on ppc64le and aarch64 builds, where the page size can
vary.

Fix InternalMakeCurrentDispatch to use the constant EGL_FALSE instead of the
Xlib-specific False.

If $host_cpu is "armv8l", then that means the 32-bit compatibility mode on an
aarch64 system. Treat that the same as armv7.

Make x11 optional

Fixes https://github.com/NVIDIA/libglvnd/issues/183

Defining a target-specific CFLAGS variable overrides AM_CFLAGS, so add the same
include paths for the GLX tests that we used to add using AM_CFLAGS.

Also remove the -Wno-error flags, since they aren't needed anymore.

In the makefiles, use the CFLAGS and LIBS variables set by the
PKG_CHECK_MODULES macros for X11 and Xext.

Add a new configure option, --disable-x11. That option will disable GLX, and
will #ifdef out the X11 platform detection in EGL, but also remove the
dependency on the X11 libraries and headers.

In addition, only check for the xext and glproto packages if GLX is enabled,
since that's the only component that needs them.

Add pkgconfig data

Add a --disable-headers option, which tells it not to install the
GL/GLES/GLX/EGL header files.

Since we've got pkg-config files for the various libraries, we should also
install the corresponding header files so that the include directory matches
up.

Add copies of glcorearb.h, and of the GLES2 and GLES3 headers.

Updated the EGL headers and XML file to the Khronos repository, from commit
598f20e3b7b7eec3e8d8a83e64b9592a21c55bb6.

Updated the OpenGL and GLX header and XML files to the Khronos repository, from
commit 08749e62826f34ba2e6e3733af545895f23e784b.

This preserves the names from Mesa's gl, glesv1_cm, glesv2, and egl pkgconfig
data, and for compatibility with that, gl.pc gives you libGL.

Also added glx.pc and opengl.pc for libglvnd's libGLX and libOpenGL libraries.
If newer applications want to explicitly depend on libGLX and the appropriate
rendering API they can ask for the new pkgconfig names.

Note that the version numbers in each file are the API versions that each
library exposes, not the libglvnd package version.

src/util/trace.c calls glvnd_vasprintf, which is defined in
src/util/utils_misc.c, so any tests that use trace.c also need to
use utils_misc.c.

Add a check in the configure script to check if the dirent struct has a d_type
member, since that's not available on all systems.

vasprintf isn't available on all systems, so use the local implementation
instead.

AARCH64 : Fix address passed to clear cache

The entrypointExec address is modified before being passed to
clear_cache function. We do a small add/subtract for ARMv7 in
order to enable Thumb mode. This is not required for Aarch64 as
it does not support Thumb mode. Since we were not adding anything,
do not subtract as well before passing the address to clear_cache.

This was causing a crash while calling glXGetProcAddress.

GLdispatch: Fix the TOC save in the PPC64LE TSD stubs (v2)

We don't need to save and restore r12 across the call to _glapi_get_current.
It's only used for function linkage, and the dispatch stub overwrites the value
a few instructions later anyway.

We can't use a simple ld instruction for a slot greater than or equal to 4096,
because the offset no longer fits in the signed 16-bit offset field.

Instead, use an addis instruction to deal with the high-order bits in the
offset.

Call _glapi_get_current through the PLT call stub. The PLT call stub takes care
of saving the TOC pointer.

Also rearranged the assembly code to remove the need for a backwards jump after
calling _glapi_get_current.

In GL/gl.h, remove PFNGLBLENDCOLORPROC, PFNGLBLENDEQUATIONPROC, and the
GL_OES_EGL_image section.

PFNGLBLENDCOLORPROC and PFNGLBLENDEQUATIONPROC are defined in GL/glext.h in the
GL_VERSION_1_4 section.

GL_OES_EGL_image is defined in GLES/glext.h and GLES2/gl2ext.h, and the
GLeglImageOES typedef is also defined in GL/glext.h under
GL_EXT_EGL_image_storage.

This is important for testing the TSD stubs, which have a separate fast path
for single-threaded programs.

The TSD stubs start with a simple global variable to hold the dispatch table
so that a single-threaded program doesn't have to deal with the overhead of
calling pthread_getspecific. When a second thread comes along, it sets that
variable to NULL, which makes the stubs call pthread_getspecific instead.

This change adds a flag to the testgldispatch test program that tells it to
call into libGLdispatch from two threads to force it into its multi-threaded
mode.

It also adds three new test scripts, which are the same testgldispatch tests
but for the multithreaded path.

In __glDispatchCheckMultithreaded, don't unlock the mutex until after calling
stubCurrentPatchCb->threadAttach.

Without that lock, another thread could come along and change
stubCurrentPatchCb out from under it.

When you run "make dist", the resulting tarball includes all of the generated
files. When building from that, we don't need to generate anything.

The $(VM_V_GEN) variable should be $(AM_V_GEN).

build: Find Python the Autotools way

An added advantage to this is that it makes it really easy to build with
Python 3, the same way other Autotools projects support it:

    $ export PYTHON=/usr/bin/python3
    $ ./configure

Update the gl.xml, GL/glext.h, and the GLES headers to commit
42f61786696df5102625d9b046976ee857645704 from the Khronos registry.

Remove FindProcIndex, __eglSetDispatchIndex, and
__eglDispatchFindDispatchFunction from egldispatchstubs.c.

The dispatch indices are allocated and assigned during initialization, so none
of those functions are ever called.

Add a separate version script export_list_tsd.ver for the TSD build of
libGLdispatch, since the two builds don't have the same set of symbols.