swung0x48 d1b3a3325f [Test] (ServerLoop): arm the guilty-reset drain hold by the healthy record's seq so the previous drain cannot take it

2026-10-14 14:59:47 -04:00

..

MG_Backend

[Fix] (Espryt): keep the invalid-image placeholder on ANGLE instead of probing an empty unit it faults on

2026-10-14 10:30:17 -04:00

MG_Benchmark

[Fix] (Bench): make the blend-toggle gate go red when the case it names stops running

2026-09-07 23:18:10 -04:00

MG_Client

[Feat] (MG_Remote): a record can tell the server its answer will never be read - the bit works and cuts a frame's replies by 92%, and it proves the window's real blocker is the FIFO drain, not the pool depth

2026-09-26 02:00:44 +08:00

MG_Gbm

[Feature] (SharedImage): sample NV12/P010 dma-bufs through GL_OES_EGL_image_external on both backends, copying foreign buffers into server YUV images

2026-10-03 22:46:00 -04:00

MG_Impl

[Fix] (Espryt): invalid image units read zero through a shaped placeholder

2026-10-12 23:13:18 -04:00

MG_IntegrationTest

[Test] (Coverage): register ClipControlScenario on the gated split arms

2026-10-13 10:28:38 -04:00

MG_Pipe

[Fix] (Remote): stop the adoption's falling edge from overwriting writes made through the adopted map

2026-10-13 08:02:35 -04:00

MG_Protocol

[Feat] (MG_Remote): a record can tell the server its answer will never be read - the bit works and cuts a frame's replies by 92%, and it proves the window's real blocker is the FIFO drain, not the pool depth

2026-09-26 02:00:44 +08:00

MG_Remote

[Fix] (Remote): order buffer writebacks by producing seq so a stale capture writeback cannot retire a newer GPU-write mark

2026-10-12 06:46:04 -04:00

MG_State

[Fix] (Remote): stop the adoption's falling edge from overwriting writes made through the adopted map

2026-10-13 08:02:35 -04:00

MG_Test

[Test] (ServerLoop): arm the guilty-reset drain hold by the healthy record's seq so the previous drain cannot take it

2026-10-14 14:59:47 -04:00

MG_Util

[Fix] (Espryt): keep the invalid-image placeholder on ANGLE instead of probing an empty unit it faults on

2026-10-14 10:30:17 -04:00

Config.h

[Fix] (Espryt): keep the invalid-image placeholder on ANGLE instead of probing an empty unit it faults on

2026-10-14 10:30:17 -04:00

ConfigLoader.cpp

[Fix] (Espryt): probe empty image units and keep invalid units empty on drivers that read them as zero

2026-10-13 08:02:34 -04:00

Defines.h

[Feature] (EGL): MobileGL serves EGL through libglvnd

2026-10-01 12:22:55 -04:00

GlobalObjects.cpp

[Refactor] (Lifecycle): own MobileGL's lifecycle from the EGL layer instead of ELF static ctor/dtor - the first EGL/WGL entry point lazily initializes via a thread-safe, re-init-capable EnsureInitialized (AutoInit is gone), the last eglTerminate with no initialized display and nothing current tears the whole library down deterministically inside the EGL lifecycle, and the global singletons move to leak-at-exit heap storage so process exit runs no backend destructors at all (AutoDestroy and the Windows DllMain abandon hook are gone); fixes the exit-time SIGABRT from undefined static-destruction order - the DirectGLES buffer-pool mutex abort on Android clean exits, and the pre-existing macOS QueryTest/ProgramTest 'Subprocess aborted' gtest failures now pass (ctest 411/411)

2026-07-26 10:59:14 -04:00

Includes.h

[Refactor] (MG_Util, MG_Test): retire FastSTL for ska::flat_hash_map, the table MobileGlues settled on

2026-08-12 00:11:54 -04:00

Init.cpp

[Refactor] (P13 W5 S8): MOBILEGL_BUILD_RECORD_ARM is folded away - the record arm is unconditional - and retired_switches keeps it from coming back.

2026-10-06 15:03:11 -04:00

Init.h

Docs: MobileGL no longer steps aside for LIBGL_ALWAYS_SOFTWARE.

2026-10-09 14:23:56 -04:00