[Git][xorg-team/lib/libglvnd][upstream-unstable] 70 commits: include: install GL headers when GL is enabled

[Timo Aaltonen <tjaalton@debian.org>]

Title: GitLab

Timo Aaltonen pushed to branch upstream-unstable at X Strike Force / lib / libglvnd

Commits:

• 6f52473d
  by Eric Engestrom at 2019-09-30T15:08:37Z

  include: install GL headers when GL is enabled
  
  A typo made it depend on EGL instead.
  
  Fixes: ab9b5fcc3bf90064418f ("Install the GL/GLES/GLX/EGL header files.")
  

• 879a453d
  by Kyle Brenneman at 2019-09-30T21:35:25Z

  Merge branch 'headers-fix-typo' into 'master'
  
  include: install GL headers when GL is enabled
  
  See merge request glvnd/libglvnd!190

• 3a1fb32c
  by Laurent Carlier at 2019-10-08T16:34:16Z

  Add a configure option to disable glesv1 or glesv2
  
  Because mesa can be built without glesv1 so it breaks autodetection.
  
  Fixes: https://bugs.archlinux.org/task/64032 ('mesa-demos doesn't build, unable to find GLES/gl.h')
  

• 6fc91b93
  by Kyle Brenneman at 2019-10-08T16:34:16Z

  Merge branch 'gles-enable-fix' into 'master'
  
  Add a configure option to disable glesv1 or glesv2
  
  See merge request glvnd/libglvnd!191

• 51233cc5
  by Adam Jackson at 2019-10-17T15:53:50Z

  egl: Sync with Khronos
  
      commit de3a5e867d906a04a5c37ee0d89e7f01d3598eb9
      Author: Ken Russell <kbrussel@alum.mit.edu>
      Date:   Sat Oct 12 05:44:43 2019 -0700
  
          Reserve enums 0x34A0..0x34AF for ANGLE project. (#93)
  
  Closes: https://gitlab.freedesktop.org/glvnd/libglvnd/issues/193
  

• a1a73db6
  by Kyle Brenneman at 2019-10-18T15:17:32Z

  Merge branch 'egl-registry-sync' into 'master'
  
  egl: Sync with Khronos
  
  Update egl.h and egl.xml to upstream commit de3a5e867d906a04a5c37ee0d89e7f01d3598eb9.
  
  Closes #193
  
  See merge request glvnd/libglvnd!192

• 5dfdc5a6
  by Kyle Brenneman at 2019-10-22T22:48:29Z

  Update GL/gl.h to match Mesa.
  
  Updated GL/gl.h to match the copy from Mesa at commit
  a0829cf23b307ca44ab8c4505974fb7c8d71a35a.
  

• 6e310f72
  by Lepton Wu at 2019-10-24T21:57:46Z

  GLdispatch: Improve the x86 tsd stubs performance.
  
  This skips touching %ebx most times and the same change for mesa shows that
  glGetString performance increased from 114M/s to 120M/s on my desktop.
  

• 1dea90a9
  by Lepton Wu at 2019-10-26T02:13:11Z

  GLdispatch: Improve the x86 TLS stubs performance.
  
  This save one call frame and a similar change for mesa shows that
  glGetString performance increased from 118M/s to 128M/s on my desktop.
  

• 58f1c0db
  by Kyle Brenneman at 2019-10-28T19:55:32Z

  Merge branch 'copy-mesa-gl-h' into 'master'
  
  Update GL/gl.h to match Mesa.
  
  Closes #194
  
  See merge request glvnd/libglvnd!195

• 931791bf
  by Kyle Brenneman at 2019-10-28T22:01:45Z

  Add GitLab CI configuration.
  
  Add a .gitlab-ci.yml file to run libglvnd's unit tests.
  
  The "build-distcheck" target will simply run "make distcheck". That should also
  cover running the unit tests for the x86-64 TLS build.
  
  The "build-i386", "build-i386-tsd", and "build-x86-64-tsd" targets will build
  and test the x86 TLS and TSD builds and the x86-64 TSD build.
  
  The pure C stubs aren't covered yet, but they'll be easy to add once the unit
  tests are fixed for them.
  
  ARM and PPC builds also aren't covered yet.
  

• f4668353
  by Matt Turner at 2019-11-07T01:41:08Z

  Provide an empty GLES3/gl3ext.h header
  
  The one in Mesa is empty but appears to come from some Khronos
  repository, but gl3ext.h has never existed in the OpenGL-Registry
  repository on github. Provide it for backwards compatibility.
  
  Closes: https://gitlab.freedesktop.org/glvnd/libglvnd/issues/195
  

• b581ecfa
  by Kyle Brenneman at 2019-11-07T14:39:59Z

  Merge branch 'gl3ext.h' into 'master'
  
  Provide an empty GLES3/gl3ext.h header
  
  Closes #195
  
  See merge request glvnd/libglvnd!200

• 2a8e4011
  by Kyle Brenneman at 2019-11-14T15:14:46Z

  Merge branch 'add-gitlab-ci' into 'master'
  
  Add GitLab CI configuration
  
  See merge request glvnd/libglvnd!198

• fc2ab74c
  by Kyle Brenneman at 2019-11-14T15:34:46Z

  Add a USE_DISPATCH_ASM macro to config.h.
  
  Define a new USE_DISPATCH_ASM macro in config.h if assembly dispatch stubs are
  enabled. Update the GLX entrypoint generation to use that macro instead of
  checking all of the USE_*_ASM macros.
  
  This is less likely to break if we add assembly support for another
  architecture. In addition, it'll make it easier to skip any unit tests
  that depend on being able to generate new dispatch stubs.
  

• 33e1fffb
  by Kyle Brenneman at 2019-11-14T15:34:47Z

  tests: Check for assembly support in testgldispatch.
  
  In testgldispatch, if it's testing generating or patching entrypoints, then
  check if assembly support is enabled, and skip the test if it's not.
  

• 8bb82da3
  by Kyle Brenneman at 2019-11-14T15:34:47Z

  tests: Add some threading tests for libGLdispatch.
  
  Added a new test, testgldispatchthread, which tests some of the thread-handling
  stuff in libGLdispatch.
  

• d1ce26f9
  by Kyle Brenneman at 2019-11-14T15:34:48Z

  tests: Add a new test for GLX entrypoint generation.
  
  Added a separate test program testglxgetprocaddress_genentry to test the GLX
  entrypoint generation by itself instead of lumping that in with
  testglxgetprocaddress.
  

• 4261cb96
  by Kyle Brenneman at 2019-11-14T15:34:48Z

  tests: Rework testglxgetprocaddress.
  
  It now calls glXGetClientString up front to load the dummy vendor library. The
  GLX entrypoint generation path has its own test now, so this test can now work
  on builds without assembly support.
  
  It now uses glXQueryServerString to test calling a core GLX function. That's a
  better test because it returns a predictable value, so we can make sure that it
  actually gets dispatched to the vendor library, instead of just checking that
  it doesn't crash.
  
  For testing looking up an OpenGL function, it'll call glXGetProcAddress and
  directly call __glDispatchGetProcAddress, and make sure that it gets the same
  pointer from both. It doesn't try to call the function, since OpenGL
  dispatching has its own tests.
  

• bc674f79
  by Kyle Brenneman at 2019-11-14T15:34:49Z

  tests: Use __GLX_FORCE_VENDOR_LIBRARY_0 for all of the GLX tests.
  
  Set the environment variable __GLX_FORCE_VENDOR_LIBRARY_0 for all of the GLX
  tests instead of __GLX_VENDOR_LIBRARY_NAME.
  
  Individual tests can force libGLX to load a vendor library early by calling
  glXGetClientString or similar.
  

• 17ecd2f1
  by Kyle Brenneman at 2019-11-14T15:34:49Z

  tests: Make the glXMakeCurrent tests work without assembly support.
  
  Change glMakeCurrentTestResults to a GLX function instead of a GL function, so
  that it can be dispatched using a vendor-provided stub instead of having to
  generate one.
  
  Remove the testglxmclate test, and the --late option in testglxmakecurrent,
  since that only makes sense when MakeCurrentTestResults was a GL function.
  

• 84d4d662
  by Kyle Brenneman at 2019-11-14T15:34:50Z

  tests: Remove testpatchentrypoints
  
  The testgldispatch tests cover the entrypoint patching code, so the
  testpatchentrypoints test is redundant.
  

• 207b0bf8
  by Kyle Brenneman at 2019-11-14T15:38:54Z

  Add GitLab CI coverage for C stubs.
  
  Now that the unit tests can handle builds without assembly support, add builds
  with C dispatch stubs to .gitlab-ci.yml.
  

• 1613facc
  by Kyle Brenneman at 2019-11-14T15:47:40Z

  Merge branch 'github/fork/kbrenneman/fix-tests-no-assembly' into 'master'
  
  Fix for unit tests with no assembly support
  
  Closes #154
  
  See merge request glvnd/libglvnd!180

• 4269cdcf
  by Kyle Brenneman at 2019-11-22T15:38:16Z

  Move glvnd_genentry into the GLX directory.
  
  The glvnd_genentry code isn't used outside of GLX.
  

• 86a9ed57
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Define dynamic dispatch stubs at compile time.
  
  Instead of allocating pages of memory for the dynamic dispatch stubs, define
  them at build time like the static stubs.
  
  The dynamic stubs are identical to the static ones, so we can use the same code
  to define both the static and dynamic stubs at compile time.
  
  Removed entry_generate(). entry_generate_default_code is now only used to
  restore entrypoints after patching.
  
  Aside from simplifying things by not having a separate allocation for the
  dynamic stubs, this should also allow the dispatch stubs to function properly
  in both processes after a call to fork.
  

• 59b622a7
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Remove writeEntry variables in entry_generate_default_code.
  
  Remove the writeEntry variables in the various entry_generate_default_code
  functions, since it doesn't have separate writable and executable mappings.
  

• 78321673
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Simplify the dispatch stub bookkeeping.
  
  Changed the interface in stub.h to work with function indexes instead of
  mapi_stub pointers. The mapi_stub structure is now an internal detail in
  stub.c.
  
  The mapi_stub struct now only contains a name and a slot.
  
  Removed the dynamic_stubs array. Instead, it just keeps an array of the
  function names for each dynamic stub. It now looks up the function pointers
  with entry_get_public, and the slot number is always equal to the index, so the
  name is the only thing it needs to keep track of.
  

• 9bde8ba6
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Change entry_get_patch_addresses to take an index.
  
  entry_get_patch_addresses now takes an index instead of a function pointer.
  

• 5222e64d
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Move entry_get_patch_addresses into entry_common.c.
  
  Move the function entry_get_patch_addresses into entry_common.c for all
  architectures.
  
  Now that it takes an index instead of a pointer, ther's no longer a difference
  between ARMv7 and everything else.
  

• 44ea7c35
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Change entry_get_patch_address to return a single address.
  
  entry_get_patch_address now returns a single address, since it doesn't have
  separate writable and executable mappings anymore.
  

• 13df5329
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Change entry_generate_default_code to take an index.
  
  For better consistency with the rest of GLdispatch, entry_generate_default_code
  now takes the index of the stub instead of a function pointer.
  

• 0c08a092
  by Kyle Brenneman at 2019-11-22T17:18:55Z

  GLdispatch: Use all 4096 dynamic dispatch stubs.
  
  In stub_add_dynamic, use all 4096 dynamic dispatch stubs. There's no longer any
  need to leave the last one empty.
  

• 50a0db24
  by Kyle Brenneman at 2019-11-22T20:16:44Z

  GLX: Define the GLX dispatch stubs at compile time.
  
  Rewrote the generated dispatch stubs for GLX so that they're all defined at
  compile time, instead of generating them at runtime.
  
  Insted of patching the address of the vendor-provided functions into the
  assembly stubs, it will now keep a separate array of function pointers. The
  assembly stubs will look up the approprite entry in that array and jump to it.
  
  As a result, the assembly code is entirely static, so we don't need to deal
  with generating or modifying executable code at runtime.
  

• 6811a4bc
  by Kyle Brenneman at 2019-11-22T20:16:48Z

  GLX: Add labels to the GLX entrypoint stubs.
  

• 67aa9ac5
  by Kyle Brenneman at 2019-11-22T20:39:13Z

  Merge branch 'github/fork/kbrenneman/compile-time-stubs-glx-v2' into 'master'
  
  GLX: Define the GLX dispatch stubs at compile time (v2)
  
  See merge request glvnd/libglvnd!146

• 02b8b933
  by Kyle Brenneman at 2019-11-22T20:40:27Z

  Merge branch 'github/fork/kbrenneman/compile-time-stubs-gldispatch-v2' into 'master'
  
  GLdispatch: Define all dispatch stubs at compile time (v2)
  
  See merge request glvnd/libglvnd!145

• 79b1b893
  by Kyle Brenneman at 2019-11-22T21:48:56Z

  Merge branch 'tls' into 'master'
  
  GLdispatch: Improve the x86 TLS stubs performance.
  
  See merge request glvnd/libglvnd!196

• 005d2568
  by Kyle Brenneman at 2019-11-22T21:49:07Z

  Merge branch 'fasm2' into 'master'
  
  GLdispatch: Improve the x86 tsd stubs performance.
  
  See merge request glvnd/libglvnd!197

• 77c4c27e
  by Kyle Brenneman at 2019-11-22T22:03:30Z

  tests: Fix some missing libraries.
  
  Fixed some missing static libraries for the EGL dummy vendor library.
  
  Added Xlib to the GLX dummy library.
  
  Added the --no-undefined flag for both of the dummy vendor libraries.
  

• aeb905d8
  by Kyle Brenneman at 2019-11-22T22:58:11Z

  util: Remove the executable allocation functions.
  
  Remove all of the functions related to allocating executable memory, since
  they're no longer needed.
  

• 3c27c981
  by Kyle Brenneman at 2019-11-22T22:58:11Z

  GLdispatch: Remove runtime assembly code generation.
  
  Removed the rest of the assembly code generation from libGLdispatch. Since we
  don't need to generate dispatch stubs at runtime for extension functions
  anymore, the code generation is now only used to restore the dispatch stubs
  after a vendor library patches them.
  
  Now, instead of reconstructing each dispatch stub, libGLdispatch will just
  allocate a copy of all of the entrypoints before patching them, and then it
  uses memcpy to restore them.
  

• 3465c7c2
  by Kyle Brenneman at 2019-11-22T22:58:11Z

  GLdispatch: Remove entry_init_public().
  
  entry_init_public() is empty on all platforms, so just remove it.
  

• 9ba775ea
  by Kyle Brenneman at 2019-11-22T23:39:46Z

  configure: Define EGL_NO_X11 unconditionally.
  
  Change the configure script to define the EGL_NO_X11 macro unconditionally, so
  that the EGL headers don't try to include the Xlib headers.
  
  Regardless of whether --disable-x11 is used, libglvnd doesn't need X11-specific
  typedefs for EGLNativeDisplayType et. al.
  

• e9b5e559
  by Kyle Brenneman at 2019-11-22T23:56:40Z

  Merge branch 'egl-no-x11-macro' into 'master'
  
  EGL: Set EGL_NO_X11 macro if --disable-x11 is used.
  
  See merge request glvnd/libglvnd!193

• 2d5d9d6c
  by Aaron Plattner at 2019-12-03T21:39:43Z

  Merge branch 'tests-fix-missing-libs' into 'master'
  
  tests: Fix some missing libraries
  
  See merge request glvnd/libglvnd!201

• 3aec7f09
  by Dylan Baker at 2019-12-03T23:51:34Z

  gitlab-ci: Update the wayland-template version
  
  This resolves issues with creating a docker image.
  

• eade458a
  by Eric Engestrom at 2019-12-04T00:09:20Z

  Merge branch 'gitlab-ci-fixes' into 'master'
  
  gitlab-ci: Update the wayland-template version
  
  See merge request glvnd/libglvnd!207

• fc32d568
  by Kyle Brenneman at 2019-12-04T15:06:08Z

  EGL: Fix two exported extension functions
  
  Fix a bug in gen_egl_dispatch.py which caused libEGL.so to export two extension
  functions, eglCreatePlatformWindowSurfaceEXT and
  eglCreatePlatformPixmapSurfaceEXT.
  
  Fixes https://gitlab.freedesktop.org/glvnd/libglvnd/issues/196
  

• 7b9e9ad6
  by Kyle Brenneman at 2019-12-04T15:10:29Z

  Merge branch 'egl-fix-ext-export' into 'master'
  
  EGL: Fix two exported extension functions
  
  Closes #196
  
  See merge request glvnd/libglvnd!204

• 000a0f2b
  by Dylan Baker at 2019-12-04T16:51:22Z

  Add .editorconfig file
  
  Many project in the freedesktop graphics space use editorconfig files as
  a way to standardize style in an editor agnostic way. This one
  configures all of the files current in libglvnd (I looked at the .c and
  .h files and they seem to all use 4 spaces for indent). Because it
  seemed everything used 4 spaces I just set that to the default and
  overrode as necessary.
  

• ad2c870a
  by Dylan Baker at 2019-12-04T16:53:00Z

  tests: Add -t to testgldispatch patched thr test
  
  Since right now it tests the same thing as the patch test
  

• 12caca77
  by Eric Engestrom at 2019-12-04T17:33:53Z

  Merge branch 'editorconfig' into 'master'
  
  Add .editorconfig file
  
  See merge request glvnd/libglvnd!205

• 3a7aed00
  by Dylan Baker at 2019-12-04T18:02:00Z

  tests: Add symbol check tests
  
  These use the symbol-check.py script from mesa, and should help to
  ensure that glvnd doesn't start exposing/hiding new symbols. It's also
  useful to double check that meson and autotools are doing the same
  thing.
  

• db89fc74
  by Kyle Brenneman at 2019-12-04T18:52:39Z

  Merge branch 'meson-test-fixes' into 'master'
  
  Test changes from the meson MR
  
  See merge request glvnd/libglvnd!206

• 3f6730ca
  by Kyle Brenneman at 2019-12-04T19:28:03Z

  tests: Add tests for the end of the dispatch table.
  
  Add a '-l' flag to testgldispatch, which tells it to test a generated stub
  at the very end of the dispatch table.
  
  On PPC, looking up a function past index 4096 requires more than just a simple
  register+immediate load, so this makes sure that it's correct.
  

• fbcce35d
  by Kyle Brenneman at 2019-12-05T18:13:17Z

  Merge branch 'cleanup-assembly' into 'master'
  
  Remove runtime assembly generation
  
  See merge request glvnd/libglvnd!202

• cbd5f13f
  by Matt Turner at 2019-12-05T18:22:38Z

  Merge branch 'tests-gldispatch-check-last' into 'master'
  
  Add unit tests for the end of GLdispatch's dispatch table.
  
  See merge request glvnd/libglvnd!203

• a9060bb6
  by Kyle Brenneman at 2019-12-05T19:50:56Z

  tests: Add tests for the end of the dispatch table.
  
  Add a '-l' flag to testgldispatch, which tells it to test a generated stub
  at the very end of the dispatch table.
  
  On PPC, looking up a function past index 4096 requires more than just a simple
  register+immediate load, so this makes sure that it's correct.
  

• 24fddd6e
  by Dylan Baker at 2019-12-05T19:50:56Z

  tests/dummy/glx: Remove config.h
  
  autotools force includes this with `-include config.h`, so it's
  unnecessary to make compilation work. Meson doesn't generate a config.h
  so this will break meson compilation.
  

• 16dc3235
  by Dylan Baker at 2019-12-05T20:01:36Z

  Add meson build system
  
  Theres a couple of things that this meson build system does differently
  than autotools. It doesn't use a config.h file, it just puts #defines on
  the command line with -D. It also does all of the code generation in the
  generated folder, simply because it's simpler to do that.
  
  On my 2 core / 4 thread KBL system:
  
  autotools (no ccache):
  sh -c "./autogen.sh&& ./configure && make -j6 check"  44.74s user 6.70s system 145% cpu 35.269 total
  
  autotools (warm ccache):
  sh -c "./autogen.sh&& ./configure && make -j6 check"  32.86s user 4.22s system 129% cpu 28.580 total
  
  meson (no ccache):
  sh -c "meson build; ninja -C build test"  23.48s user 3.71s system 236% cpu 11.487 total
  
  meson (warm ccache)
  sh -c "meson build; ninja -C build test"  16.06s user 2.31s system 210% cpu 8.727 total
  

• b308a876
  by Dylan Baker at 2019-12-05T20:01:36Z

  autotools: Include meson.build and meson_options.txt in dist
  

• 416e73a8
  by Dylan Baker at 2019-12-05T20:01:36Z

  README: add content about using meson
  

• 61ed0176
  by Dylan Baker at 2019-12-05T20:01:36Z

  gitlab-ci: Split some parts that are autotools specific
  

• e10e423c
  by Dylan Baker at 2019-12-05T20:01:36Z

  gitlab-ci: Install meson dependencies for meson building
  

• bc908ed8
  by Dylan Baker at 2019-12-05T20:01:36Z

  gitlab-ci: Add meson build support
  

• cdf63cfc
  by Dylan Baker at 2019-12-05T20:01:36Z

  gitlab-ci: Also run meson from the dist tarball
  

• a1350eef
  by Eric Engestrom at 2019-12-05T20:26:41Z

  Merge branch 'meson' into 'master'
  
  New meson build system
  
  See merge request glvnd/libglvnd!199

• 0f39337d
  by Matt Turner at 2019-12-05T21:16:00Z

  Drop unnecessary assembly macros
  
  These were presumably copied from Mesa where they enable assembly code
  that is not part of the dispatch layer. They are unused in libglvnd.
  

• 05e762ba
  by Matt Turner at 2019-12-05T21:44:43Z

  Bump the version number to 1.3.0.
  

29 changed files:

• + .editorconfig
• + .gitlab-ci.yml
• + .gitlab-ci/i686-linux-gnu
• + .gitlab-ci/run_build.sh
• + .gitlab-ci/run_distcheck.sh
• + .gitlab-ci/run_meson.sh
• + .gitlab-ci/ubuntu_install.sh
• Makefile.am
• README.md
• + bin/symbols-check.py
• configure.ac
• include/EGL/eglplatform.h
• include/GL/gl.h
• + include/GLES3/gl3ext.h
• include/Makefile.am
• + include/meson.build
• + meson.build
• + meson_options.txt
• src/EGL/Makefile.am
• + src/EGL/egl-symbol-check.sh
• + src/EGL/egl.symbols
• + src/EGL/meson.build
• src/GL/Makefile.am
• + src/GL/gl-symbol-check.sh
• + src/GL/gl.symbols
• src/GL/libgl.c
• + src/GL/meson.build
• src/GLESv1/Makefile.am
• + src/GLESv1/glesv1-symbol-check.sh

The diff was not included because it is too large.

—
View it on GitLab.
You're receiving this email because of your account on salsa.debian.org. If you'd like to receive fewer emails, you can adjust your notification settings.