summaryrefslogtreecommitdiff
path: root/platform/linux
Commit message (Collapse)AuthorAgeFilesLines
* [android,darwin,ios,qt] Remove unused codeThiago Marcos P. Santos2019-02-121-11/+0
| | | | These files are no longer needed after the new GL abstraction.
* [core] Switch to the new OpenGL abstractionThiago Marcos P. Santos2019-02-121-0/+3
| | | | | | This patch will make Mapbox GL Core never use OpenGL directly. We should consider locking into OpenGL ES 2.0 to simplify the code path and remove #ifdefs.
* [android,linux,darwin,qt] Load OpenGL for the platformThiago Marcos P. Santos2019-02-121-0/+166
| | | | | | | | These loaders will provide pointers to the platform implementation of OpenGL. On Android, Linux and Darwin, we simply link with the OpenGL implementation that is expected to exist on the platform. Qt has its own GL loader, so we just connect the dots.
* [android] Implement google benchmark runner for Android platformAlexander Shalamov2019-01-111-2/+2
|
* [build] rework platform/default directory and add -files.txt for vendored libsKonstantin Käfer2018-12-141-33/+33
|
* Switch code coverage system to CodeCov (#13297)Jordan Kiley2018-11-081-28/+0
|
* Refer corresponding LayerFactory instance from the Layer::ImplMikhail Pozdnyakov2018-11-071-0/+1
|
* [build] Add build system for submodulesThiago Marcos P. Santos2018-10-311-3/+0
| | | | | | | | | Vendorize the submodules and a simple CMake build system for all them. The dependencies will inherit compilation options for core. The goal is to make Mapbox GL Core completely self contained with a simple offline build.
* [linux] add map_snapshotter.*pp to linux core buildear7h2018-10-091-0/+4
|
* [linux,qt] Remove remainder of WebP supportKonstantin Käfer2018-09-131-2/+0
|
* [android] add support for gnustlKonstantin Käfer2018-09-111-1/+2
|
* [build] use vendored version of ICUKonstantin Käfer2018-09-051-1/+2
|
* [build] use vendored SQLite for LTO benefits + custom compile flagsKonstantin Käfer2018-09-051-2/+2
|
* [offline] Add option to pass geojson input fileIvo van Dongen2018-08-201-0/+1
|
* [build] Move all mason_use statements to a single fileJohn Firebaugh2018-07-131-12/+0
| | | | So its checksum can be used in a CI cache key name.
* [build] Record binary size via GitHub checkJohn Firebaugh2018-07-031-2/+1
|
* [core] Default "collator" implementationChris Loer2018-07-031-0/+3
| | | | | | - Based on nunicode - Not locale-aware - Used by linux and Qt builds
* [core, vendor] Create vendored nunicode 1.8.Chris Loer2018-07-031-2/+3
| | | | | - Version bump to 1.8 necessary for "unaccent" functionality - Qt now depends on nunicode, ruling out use of precompiled binaries
* [linux] Documentation update to reflect Ubuntu 18.04 packages (#12272)Antonio Zugaldia2018-07-021-3/+1
| | | | | | * [linux] update libllvm version for ubuntu 18.04 * [linux] cmake 3.x is now bundled with ubuntu 18.04
* [build] GLFW version 2018-06-27-0be4f3fBruno de Oliveira Abinader2018-06-271-1/+1
|
* [node] hide all symbols from the Node.js addonsKonstantin Käfer2018-06-041-1/+3
|
* [linux] link EGL backend against SwiftShader to avoid Mesa dependencyKonstantin Käfer2018-06-041-7/+2
|
* [linux] load cURL dynamically for wider compatibilityKonstantin Käfer2018-06-041-1/+3
|
* [linux] statically link the C++ standard libraryKonstantin Käfer2018-06-041-0/+8
| | | | Most of the STL is header-only anyway, and statically linking it only has a small overhead. On the other hand, it allows us to build more portable executables.
* [linux] Don't compile linux mbgl-loop-uv against node headersJohn Firebaugh2018-05-041-11/+16
| | | | Previously, `target_include_directories(mbgl-loop-uv PUBLIC ${NODEJS_INCLUDE_DIRS})` always applied. Compiles were working by accident because node is always installed in CI. Instead, split and inline loop-uv.cmake contents: linux uses `target_add_mason_package(mbgl-loop-uv PUBLIC libuv)`, node uses `target_include_directories(mbgl-loop-node PUBLIC ${NODEJS_INCLUDE_DIRS})`.
* [docs] Make per-platform installation docs self-containedJohn Firebaugh2018-04-301-4/+16
| | | | I've seen several issues where users followed platform-specific install docs, but were unaware of additional prerequisites spelled out in the top-level INSTALL.md. So let's try making each platform's installation documentation self contained.
* [build] Remove unused files, and cleanup directoriesKonstantin Käfer2018-01-031-1/+1
|
* [build] Replace boost_libprogram_options with argsBruno de Oliveira Abinader2017-12-181-1/+1
|
* [core] Enable local glyph generation using TinySDF.Chris Loer2017-12-111-0/+1
| | | | | | | - Platform-specific LocalGlyphRasterizer is responsible for deciding which glyphs to rasterize locally and for implementing the rasterization. - Default platform implementation doesn't locally generate any glyphs -> no behavior change - Unit test uses StubLocalGlyphRasterizer, which returns a single fixed bitmap for all CJK glyphs - Rename glyph_loader.test to glyph_manager.test
* [build] standardize on -fvisibility=hidden for all targetsKonstantin Käfer2017-11-291-1/+0
| | | | | | | Enables -fvisibility=hidden for iOS and Linux, and adds a workaround for GCC 6.3-7.1 Adds a GCC 6 build Enables diagnostics for C files Fixes a shadow warning in parsedate.c
* [core] move HeadlessBackend extension initialization code into ImplKonstantin Käfer2017-11-292-20/+22
|
* [core] fold HeadlessDisplay into the headless RenderBackend implementationKonstantin Käfer2017-11-295-246/+180
|
* [core] Make the OpenGL implementation platform specificThiago Marcos P. Santos2017-10-252-0/+16
| | | | | Split the headers, so each platform includes the most appropriated header.
* [build] split out DefaultFileSource and dependents to a separate targetKonstantin Käfer2017-09-281-20/+32
| | | | We don't want to link it into the node bindings, so keep it in a separate target
* [linux] headless display - support opengl es 2Ivo van Dongen2017-08-301-0/+3
|
* Add {Source,CompositeCamera}Function benchmarks (#9838)Anand Thakker2017-08-291-1/+1
|
* [node, tests] Consolidate headless rendering logic in HeadlessFrontendJohn Firebaugh2017-07-261-2/+2
|
* [build] Downgrade GCC support from 5.0 to 4.9Bruno de Oliveira Abinader2017-07-191-3/+6
|
* [test] allow only a single shared display to avoid egl errorsIvo van Dongen2017-07-182-2/+2
|
* [build] Update GLFW to 2017-07-13-67c9155Bruno de Oliveira Abinader2017-07-131-1/+1
|
* [core] Isolate pthread-based tls implementationThiago Marcos P. Santos2017-07-071-0/+1
|
* [build] Switch from Travis CI to CircleCIBruno de Oliveira Abinader2017-06-221-4/+11
|
* [tidy] modernize-make-uniqueBruno de Oliveira Abinader2017-05-121-1/+1
|
* [tidy] modernize-use-equals-defaultBruno de Oliveira Abinader2017-05-121-2/+1
|
* [tidy] modernize-use-overrideBruno de Oliveira Abinader2017-05-121-1/+1
|
* [tidy] modernize-use-autoBruno de Oliveira Abinader2017-05-122-3/+3
|
* [build] Added 'shared_thread_pool.cpp' to 'mbgl-core' target on LinuxBruno de Oliveira Abinader2017-05-021-0/+1
|
* [core] always create a pixelbuffer surface in the EGL backendKonstantin Käfer2017-04-052-9/+6
| | | | We need a surface to activate a context. Some implementation don't seem to error when passing EGL_NO_SURFACE, but it doesn't seem standards compliant.
* [core] destroy surface before context in EGL backendKonstantin Käfer2017-04-051-4/+4
|
* [build] don't advertise headers that aren't used in public headersKonstantin Käfer2017-03-281-1/+1
|