summaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)AuthorFilesLines
2015-02-02Merge branch 'maint-1.5' into masterHEADmasterChad Versace1-1/+2
* maint-1.5: wflinfo: Fix MSVC compilation error.
2015-02-02wflinfo: Fix MSVC compilation error.Jose Fonseca1-1/+2
Workaround what seems to be a bug in MSVC parser for C99. http://lists.freedesktop.org/archives/waffle/2015-January/000975.html Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@intel.com>
2015-01-27Merge branch 'maint-1.5' into masterChad Versace2-21/+21
* maint-1.5: wflinfo: find glGetStringi on both Mali and WGL waffle: return accurate result from wegl_platform_init()
2015-01-27wflinfo: find glGetStringi on both Mali and WGLFrank Henigman1-21/+19
Do the glGetStringi lookup after making context current so it works on WGL. Remove an incorrect glGetStringi lookup, which returned NULL on Mali. Signed-off-by: Frank Henigman <fjhenigman@google.com> Reviewed-by: Chad Versace <chad.versace@intel.com>
2015-01-27waffle: return accurate result from wegl_platform_init()Frank Henigman1-0/+2
wegl_platform_init() now returns false if dlopen or dlsym fails. Signed-off-by: Frank Henigman <fjhenigman@google.com> Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@intel.com>
2015-01-27waffle: Add public func waffle_window_create2()Chad Versace3-9/+71
Today, waffle_window() has only two parameters: width and height. Frank Henigman wants to extend Waffle's GBM backend with the ability to post window contents to the display. Multiple methods exist for posting content to the screen with the drm API, and that method should be configurable per waffle_window. Therefore, we need to be able to pass additional attributes to waffle_window_create(). It would also be nice to specify at time of creation that the waffle_window should be full screen. Again, we need to pass additional attributes to waffle_window_create(). The new function waffle_window_create2() is conceptually equivalent to the original waffle_window_create() with the addition of an attrib_list parameter. The only supported attributes are currently WAFFLE_WINDOW_WIDTH and WAFFLE_WINDOW_HEIGHT. See the manpage changes for more details. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Add func wcore_attrib_list_pop()Chad Versace2-0/+46
This is useful for removing attributes as they are parsed. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Add func wcore_attrib_list_copy()Chad Versace2-0/+40
This is useful for making a writable copy of a read-only attribute list. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Add attrib_list param to func wcore_platform::window::createChad Versace15-18/+74
This prepares for adding waffle_window_create2() to Waffle's public API, which will have an attrib_list parameter. No attributes are supported yet. Therefore this patch validates, at the top of each ${PLATFORM}_window_create(), that the attrib_list is empty. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Add func wcore_error_bad_attributeChad Versace2-0/+12
This little helper function emits WAFFLE_ERROR_BAD_ATTRIBUTE with a default error message. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27waffle: Fix signature of wcore_platform::window::create()Chad Versace14-33/+33
Fix signature to match that of waffle_window_create(). The types of parameters 'width' and 'height' didn't match the public header. As a consequence, also fix the signature of ${PLATFORM}_window_create() for all platforms. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27waffle: Fix mismatch in waffle_window_create prototypeChad Versace1-1/+1
The prototype differed between the public header and file waffle_window.c. This is solely a cosmetic change, because 'int' and 'int32_t' are the same on all platforms supported by Waffle. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Add func wcore_attrib_list_from_int32Chad Versace2-0/+52
This function converts an int32_t attribute list to an intptr_t attribute list. It will be needed in the future when updating waffle_config_choose() to accept an intptr_t attribute list. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Add arithmetic functions that detect overflowChad Versace2-0/+49
Define the function below. All act on size_t inputs. wcore_add_size wcore_iadd_size : in-place addition wcore_mul_size wcore_imul_size : in-place multiplication Future patches will use the functions to safely calculate the 'size' value given to malloc. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Define intptr_t variants of wcore_attrib_list functionsChad Versace2-0/+29
When designing the original Waffle 1.0 API, I made a mistake when I chose to declare attribute lists as arrays of int32_t. Instead, they should have been arrays of intptr_t. A new public function, waffle_window_create2, will have a `const intptr_t attrib_list[]` parameter. Therefore waffle needs intptr_t variants of the wcore_attrib_list functions. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Define wcore_attrib_list funcs with type-generic templateChad Versace1-70/+82
The current wcore_attrib_list functions work on int32_t attribute lists. A future commit will add variants that work on intptr_t attribute lists. To prevent code duplication, this patch moves the definition of all wcore_attrib_list32 functions into a type-generic macro, WCORE_ATTRIB_LIST_COMMON_FUNCS, which will also be used to define the upcoming intptr_t functions. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Change return type of wcore_attrib_list32_lengthChad Versace2-3/+4
Change from int32_t to size_t, because - Length is a measurement of unsigned "size". - In general, length values are often used in subsequent calculation of memory size which is passed to functions like malloc, whose input is size_t not int. - The cast from ptrdiff_t to int32_t in wcore_attrib_list_length just looked wrong. A cast to size_t makes more sense. As precedent, strlen() also returns size_t. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-27core: Rename functions s/wcore_attrib_list/wcore_attrib_list32/Chad Versace5-74/+74
When designing the original Waffle 1.0 API, I made a mistake when I chose to declare attribute lists as arrays of int32_t. Instead, they should have been arrays of intptr_t. A new public function, waffle_window_create2, will have a `const intptr_t attrib_list[]` parameter. Therefore waffle needs intptr_t variants of the wcore_attrib_list functions. This patch makes room for the new intptr_t wcore_attrib_list functions by renaming the old ones. Signed-off-by: Chad Versace <chad.versace@intel.com> Tested-by: Emil Velikov <emil.l.velikov@gmail.com> (msvc/wgl) Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2015-01-19waffle: initial empty implementation of nacl backendTapani Pälli15-1/+814
Patch adds nacl platform skeleton. Only thing it does is that it creates a container that holds the 3D context object which is responsible for any communication required with browser. v2: cleanups, remove unnecessary casts, use c99 initializer (Emil Velikov) v3: additional code cleanups (Chad Versace) Signed-off-by: Tapani Pälli <tapani.palli@intel.com> Reviewed-by: Chad Versace <chad.versace@intel.com>
2015-01-19waffle: add support for building Waffle using NaCl toolchainTapani Pälli2-0/+11
Patch adds changes required to use NaCl compiler and libraries to build Waffle. Build can be configured to use specific version of the NaCl SDK, toolchain for the build needs to be selected with cmake variable CMAKE_TOOLCHAIN_FILE. Example command line to configure a build: cmake -Dwaffle_has_nacl=ON \ -Dnacl_sdk_path=/home/tpalli/nacl/nacl_sdk \ -Dnacl_version=pepper_39 \ -DCMAKE_TOOLCHAIN_FILE=cmake/toolchain-nacl-x86_64-glibc.cmake \ -DCMAKE_BUILD_TYPE=Release \ . v2: cmake fixes, create toolchain files for nacl (Emil Velikov) v3: rename toolchain files, cmake fixes (Chad Versace, Emil Velikov) Signed-off-by: Tapani Pälli <tapani.palli@intel.com> Reviewed-by: Chad Versace <chad.versace@intel.com>
2014-12-31wgl: Verify the client area size matches the required size on window ↵José Fonseca1-1/+11
creation too. By default, Windows will limit windows too large to theu desktop size, and windows too small to be big enough for the titlebar. Waffle's windows don't get affected as their style is WS_POPUPWINDOW, which doesn't include the WS_CAPTION style. This change adds more assertion, just in case this ever changes, as many piglit tests rely on large/tiny windows to have the requested size. Also replace `#ifdef DEBUG` with `#ifndef NDEBUG`, as NDEBUG is the define that controls assert macro. That said, I wonder if we should call `waffle_errorf(WAFFLE_ERROR_INTERNAL, ...)` and verify this on release builds too. Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@intel.com>
2014-12-31wgl: Fix requirements for creation of ES2 contextChad Versace1-1/+3
To create an ES2 context, Waffle required WGL_EXT_create_context_es2_profile. Fix Waffle to require either WGL_EXT_create_context_es_profile *or* WGL_EXT_create_context_es2_profile, because WGL_EXT_create_context_es_profile is an updated variant of WGL_EXT_create_context_es2_profile that supercedes it. Fixes #23: https://github.com/waffle-gl/waffle/issues/23 Reviewed-by: José Fonseca <jfonseca@vmware.com> Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
2014-12-09Merge branch 'maint-1.4'Chad Versace2-2/+24
* maint-1.4: doc: Add release notes for Waffle 1.4.3 waffle: Bump version to 1.4.3 glx: Fix requirements for creation of ES2 context cmake: Fix building with CMake 2.8.11 examples: Remove bitrotted XCode project files debian: Release 1.4.2-1 Conflicts: Android.mk cmake/Modules/WaffleDefineVersion.cmake Trivial version conflicts. debian/changelog Resolve conflicts by ignoring incoming changes.
2014-12-09glx: Fix requirements for creation of ES2 contextChad Versace1-1/+3
To create an ES2 context, Waffle required GLX_EXT_create_context_es2_profile. Fix Waffle to require either GLX_EXT_create_context_es_profile *or* GLX_EXT_create_context_es2_profile, because GLX_EXT_create_context_es_profile is an updated variant of GLX_EXT_create_context_es2_profile that supercedes it. Fixes #24: https://github.com/waffle-gl/waffle/issues/24 Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
2014-12-09cmake: Fix building with CMake 2.8.11Chad Versace1-1/+21
Regressed by commit 92116dae. Waffle's CMakeLists declares that 2.8.11 is required, but commit 92116dae begin using a 2.8.12 feature: the PRIVATE keyword to target_link_libraries(). We should not bump Waffle's CMake requirement to 2.8.12, though, because RHEL-7 and Chrome OS still have only 2.8.11. This patch fixes Waffle to use the PRIVATE keyword only if cmake >= 2.8.12. Otherwise, it falls back to the LINK_INTERFACE_LIBRARIES hack removed by 92116dae. Reported-by: Matej Cepl <mcepl@cepl.eu> Reviewed-by: Jordan Justen <jordan.l.justen@intel.com> Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
2014-12-09Merge branch 'liuyq/fix-android-5.0-build'Chad Versace1-4/+6
* cooking/liuyq/fix-android-5.0-build: Android.mk: fix problem when compile with 64bit version droid_surfaceflingerlink.cpp: fix compiling error for android5.0
2014-12-04cmake: Find and set include path for wglext.hJosé Fonseca1-0/+1
GL/*glext.h is not provided by the system (be it MinGW or MSVC), so its path must be specified separately. I used GLEXT_INCLUDE_DIR, which is the name also used for Piglit, so that existing CMake initial caches might already provide it. Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
2014-11-23Merge branch 'evelikov/remove-libgbm-libgl-libdeps-v2'Chad Versace14-60/+198
* cooking/evelikov/remove-libgbm-libgl-libdeps-v2: glx: drop explicit linking against libGL.so glx: fetch the libGL function pointers at glx_platform_create pkg/archlinux: demote libgbm to (make|opt)depend gbm: drop explicit linking against libgbm.so (v2) gbm: fetch the libgbm function pointers at wgbm_platform_init pkg/archlinux: waffle does not link against libegl pkg/archlinux: use configure autodetection Conflicts: src/waffle/gbm/wgbm_display.c Minor conflict due to 9a4f7c0, the libglapi workaround.
2014-11-23Merge branch 'maint-1.4'Chad Versace1-0/+23
* maint-1.4: wflinfo: Fix glGetStringi on Mali
2014-11-23wflinfo: Fix glGetStringi on MaliChad Versace1-1/+22
On Mali EGL 1.4, you can't obtain non-extension functions (including OpenGL ES 3.0 functions) with eglGetProcAddress. You must use dlsym. This patch carefully avoids breaking glGetStringi on other platforms, which may expose glGetStringi dynamically (as in eglGetProcAddress) but not statically (as in dlsym). Fixes #20: https://github.com/waffle-gl/waffle/pull/20 Fixes: chromium:428061 [https://code.google.com/p/chromium/issues/detail?id=428061] See: https://github.com/anholt/libepoxy/issues/21 Reported-by: Frank Henigman <fjhenigman@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-19wflinfo: add GLSL version information to wflinfoDylan Baker1-1/+19
For versions that do not provide a shader version (GL ES 1.x, GL < 2.0) print "None", otherwise print the value of GL_SHADING_LANGUAGE_VERSION. None was picked because it is simple, obvious, and unlike N/A can't be expanded into multiple meanings v2: - Don't print for gles1, since gles1 doesn't have a shading language and will always return FLINFO_GL_ERROR v3: - Print 'None' for GL ES 1.x and GL < 2.0 - Only print shader_version in verbose mode Fixes: https://github.com/waffle-gl/waffle/issues/16 Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com> Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-14droid_surfaceflingerlink.cpp: fix compiling error for android5.0Yongqin Liu1-4/+6
update the version check for android 5.0 in file droid_surfaceflingerlink.cpp Change-Id: Iee3f294dfd4951d4b2fff9cc2afeebc0d2d24802 Signed-off-by: Yongqin Liu <yongqin.liu@linaro.org>
2014-11-12Merge branch 'maint-1.4'Chad Versace1-0/+2
* maint-1.4: gbm: work around mesa linkage issue
2014-11-13glx: drop explicit linking against libGL.soEmil Velikov8-40/+52
Use the function pointers stored in glx_platform, to communicate with the library. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-13glx: fetch the libGL function pointers at glx_platform_createEmil Velikov2-4/+66
Thus way with a follow up commit we can use them and eliminate the link dependency from waffle. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-13gbm: drop explicit linking against libgbm.so (v2)Emil Velikov4-15/+21
Use the function pointers stored in wgbm_platform, to communicate with the library. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-13gbm: fetch the libgbm function pointers at wgbm_platform_initEmil Velikov2-1/+59
Thus way with a follow up commit we can use them and eliminate the link dependency from waffle. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-12Merge branch 'cooking/evelikov/wgl'Chad Versace27-45/+1897
* cooking/evelikov/wgl: (53 commits) man: spelling fix man: spelling/grammar fix README: fix spelling typo wgl: s/glX/wgl/ in error message README: spelling fixes wflinfo: call get_proc_address after make_current cmake: ensure waffle-static name differs from the shared one cmake: Set default location for all artifacts to top-level directories core/tests: silence the final MSVC warning core/tests: correctly store the threads exit codes wgl: use correct dtor in wgl_window_priv_create error path core: silence MSVC compiler warning wgl: strings.h does not exist in MSVC cmake: misc fixes and cleanups test/gl_basic_test: implement windows version for run_testsuite() test/gl_basic_test: add PLATFORM_WGL support/tests test/gl_basic_test: deduplicate gl_(debug,fwdcompat) tests test/gl_basic_test: introduce linux_gles* tests test/gl_basic_test: introduce linux_gl(30_core - 43_compat) tests test/gl_basic_test: use all_gl_(10-21) tests for all platforms ...
2014-11-12gbm: work around mesa linkage issueFrank Henigman1-0/+2
Mesa drivers such as i965_dri.so use symbols from libglapi but do not pull in libglapi (https://bugs.freedesktop.org/show_bug.cgi?id=57702) so dlopen libglapi before we need it. Fixes #21: https://github.com/waffle-gl/waffle/pull/21 Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-09Merge branch 'evelikov/drop-egl-linking'Chad Versace16-100/+312
* cooking/evelikov/drop-egl-linking: egl: drop explicit linking against libEGL.so egl: retrieve the libEGL function pointers at wegl_platform_init() egl: add preliminary wegl_platform
2014-11-09Merge branch 'maint-1.4'Chad Versace3-1/+21
* maint-1.4: gbm: lock and release buffer in swap buffers
2014-11-09gbm: lock and release buffer in swap buffersFrank Henigman3-1/+21
Otherwise mesa prints "libEGL debug: EGL user error 0x300d (EGL_BAD_SURFACE) in dri2_swap_buffers" when environment variable EGL_LOG_LEVEL=debug. Fixes #17: https://github.com/waffle-gl/waffle/pull/17 Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-09wgl: s/glX/wgl/ in error messageEmil Velikov1-1/+1
Spotted while running through piglits and the test required OGL 4.0 context on my OGL 3.3 card. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-09wflinfo: call get_proc_address after make_currentEmil Velikov1-2/+2
WGL requires that an active context is available as wglGetProcAddress is called otherwise the function pointer returned will be NULL. v2: Rebase Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
2014-11-09cmake: ensure waffle-static name differs from the shared oneEmil Velikov1-1/+1
In our current state 'make check/check-func' will succeed, but plain 'make all' will fail. This is due to naming colision when building for Windows, as RUNTIME, LIBRARY and ARCHIVE have different meanings between Windows and everyone else. Essentially the waffle.lib from the shared library will clash with the one for static waffle. Cc: Chad Versace <chad.versace@linux.intel.com> Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
2014-11-09cmake: Set default location for all artifacts to top-level directoriesEmil Velikov2-7/+0
By default, put executables into $CMAKE_BINARY_DIR/bin and libraries into $CMAKE_BINARY_DIR/lib. Test executables still go into $CMAKE_BINARY_DIR/test. This patch's main purpose is to consolidate repetitive code. It doesn't actually move anything except 'libcmocka.a'. v2: [Emil] - Move gl_basic_test (dynamic linked against waffle) to bin, or the test will fail on Windows. Due to the DLL being in a different folder. Cc: Emil Velikov <emil.l.velikov@gmail.com> Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-09core/tests: silence the final MSVC warningEmil Velikov1-1/+1
Avoid forcing intptr_t down a variable of type int. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
2014-11-09core/tests: correctly store the threads exit codesEmil Velikov1-2/+2
The exit codes are defined as ints, while we were using an bool array. This resulted in stack corruption due to the difference of the size in the two types. Huge thanks to Jose for spotting this one. Spotted-by: Jose Fonseca <jfonseca@vmware.com> Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
2014-11-09wgl: use correct dtor in wgl_window_priv_create error pathEmil Velikov1-1/+1
This is the private ctor, as such use the private dtor if we fail at any point. Noticed while casing wgl_window_destroy() related bug in piglit. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
2014-11-09core: silence MSVC compiler warningEmil Velikov1-1/+1
The return type of wcore_attrib_list_length() is int32_t and at the same time MSVC promotes the (int32_t) math at the return line to int64_t thus causing a warning message. Explicitly cast it to make MSVC happy. Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-28 15:15:43 +0000