Swung0x48
eb71b669b3
[Fix] (Build): make the pull and verify builds compile again
2026-10-04 06:30:24 -04:00
swung0x48
bb31f374a3
[Perf] (DirectVulkan): keep the wire draw's render pass open across draws to the same attachments, closing it only for work that cannot sit inside one
2026-10-04 06:18:35 -04:00
swung0x48
e5ee31b022
[Build] (Anland): build the anland server for android-30, the APK's minSdk, so thread_local uses native TLS
2026-10-04 06:02:45 -04:00
swung0x48
915882b293
[Perf] (Server): time records only when the per-frame statistics are on
2026-10-04 06:02:44 -04:00
swung0x48
963f4d8177
[Perf] (DirectVulkan): reuse completed command buffers instead of freeing and allocating one per submission
2026-10-04 06:02:43 -04:00
swung0x48
d627fca387
[Perf] (Transport): the server rings a parked client only once the sequence it waits for is applied
2026-10-04 05:40:14 -04:00
swung0x48
2de0d6dd25
[Perf] (Pipe): reuse client-array buffers instead of a blocking create per draw, and copy contiguous ranges in one go
2026-10-04 04:54:54 -04:00
swung0x48
66b47a101b
[Fix] (Remote): glFinish waits for the GPU, Magma bounds unsubmitted clear objects, ES contexts advertise OES_mapbuffer/rgb8_rgba8/depth24
2026-10-04 04:54:53 -04:00
swung0x48
4722fa38bb
[Docs] (Performance): glmark2 baseline of Espryt and Magma against the stock kgsl stack on anland, with profiles and ranked bottlenecks
2026-10-04 04:33:24 -04:00
swung0x48
3a289222af
[Docs] (MobileVA): plan a vendor-agnostic VA-API driver on Android MediaCodec, H.264 first
2026-10-03 23:46:13 -04:00
swung0x48
18a5dba0e4
[Docs] (Anland): Chrome decodes video in hardware through the YUV shared images, with probes to check it
2026-10-03 22:46:01 -04:00
swung0x48
7218bacb49
[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
swung0x48
4633ca223c
[Fix] (Server): log the per-frame P65ServerFrame line only when MOBILEGL_SERVER_FRAME_STATS asks for it
2026-10-03 14:20:11 -04:00
swung0x48
0ef13ab0a6
[Docs] (Anland): Chrome decodes video in software because MobileGL cannot import the container VA-API driver's NV12 frames.
2026-10-03 14:16:04 -04:00
swung0x48
287e1df8dc
[Fix] (Server, Magma, Espryt): end the session whose GPU work hangs the GPU so it cannot stall the desktop again.
2026-10-03 12:28:39 -04:00
swung0x48
4f6f4125ac
[Fix] (EGL, SharedImage): real EGL native fences publish a dma-buf producer's writes, so Chrome's GPU process frames reach the compositor.
2026-10-03 10:59:02 -04:00
swung0x48
2a1889b817
Discover gtest cases at ctest start so parallel links stop sharing one discovery JSON file.
2026-10-03 09:38:51 -04:00
swung0x48
07e7730085
[Docs] (Anland): kwin-recover.sh reads the latch from the server log file, not the too-small logcat buffer.
2026-10-03 09:05:55 -04:00
swung0x48
03648c2783
[Docs] (Anland): add a device test that loses KWin's own session and checks the desktop keeps compositing.
2026-10-03 09:05:54 -04:00
swung0x48
f22ae625cb
[Fix] (Server): a lost session's window lease is waited for by the fresh session, and a window wait holds no backend turn.
2026-10-03 09:05:54 -04:00
swung0x48
f52a546707
[Fix] (EGL, Remote): a server-owned surface is re-created at the size it asked for, and a lost session's image reference is never released on the fresh one.
2026-10-03 09:05:53 -04:00
swung0x48
77c604061e
[Docs] (Anland): glxgears is fixed-function, so it measures presentation only
2026-10-03 07:53:11 -04:00
swung0x48
36190b7cc0
[Fix] (EGL, Wayland): a swap that runs into a device loss reports it and still commits the window, instead of showing an empty frame.
2026-10-03 07:17:08 -04:00
swung0x48
cf763b4081
[Fix] (EGL, Remote): a thread that released a lost context keeps reporting the reset until it binds another.
2026-10-03 07:17:07 -04:00
swung0x48
2acc6c41bb
[Fix] (Pipe): an evicted sampler CSO is deleted from every share group of its session.
2026-10-03 05:37:53 -04:00
swung0x48
a84104a40f
[Fix] (Pipe): the default framebuffer's record names no placeholder surface that only one share group holds.
2026-10-03 05:37:53 -04:00
swung0x48
1dd888cb21
[Fix] (EGL, DirectGLES): keep a lost context's release off the fresh session, bind no surfaceless context natively, and keep the display while surfaces remain.
2026-10-03 05:37:52 -04:00
swung0x48
393002a46e
[Fix] (Pipe): give each context's applier its own serial range so backend memos never match another context.
2026-10-03 05:37:51 -04:00
swung0x48
25a76c1b20
[Fix] (Pipe): drain a texture's pending texels only under a context of the share group that wrote them.
2026-10-03 05:37:51 -04:00
swung0x48
a9f22ebb63
[Fix] (EGL): create and end a context's objects under its own binding and the stream lock, and defer destroying a current context.
2026-10-03 05:37:50 -04:00
swung0x48
2fe011adf6
[Test] (Wire): enable the concurrent-context repro and add release, sibling, deferred-destroy and read-back-race shapes.
2026-10-03 05:37:49 -04:00
swung0x48
11668b02f2
[Docs] (Anland): Chrome runs its GPU in its own process by default and recovers from a device loss.
2026-10-03 05:21:53 -04:00
swung0x48
30977c181f
[Feature] (EGL+GL): an opt-in OpenGL ES identity for ES contexts with BGRA textures, and contexts take the dispatcher's API, so Chrome's GPU process runs on ANGLE.
2026-10-03 05:21:52 -04:00
swung0x48
e5f097f9e1
[Docs] (Anland): add a Chrome device-loss recovery test and the GPU-process launch mode.
2026-10-03 04:37:34 -04:00
swung0x48
90e63c9e2a
[Feature] (EGL+GL): advertise EGL_EXT_create_context_robustness and GL robustness, with per-context reset notification and a pid-targeted device-loss knob.
2026-10-03 04:37:33 -04:00
swung0x48
af2b9f4c1c
[Docs] (Anland): root cause of the X11 stale-region issue
2026-10-03 04:36:24 -04:00
BZLZHH
8302dd30bf
[Docs] (Anland): the lock screen behind the black desktop, and where the Magma text gap is not
2026-10-03 16:34:25 +08:00
swung0x48
cd58f048b6
[Fix] (Shader): an inactive fragment output no longer takes location 0 from the active one
2026-10-03 04:16:41 -04:00
BZLZHH
a6d788191f
[Docs] (Anland): record the Magma verification and its missing glyphs
2026-10-03 16:07:53 +08:00
BZLZHH
0d89db36dd
[Docs] (Anland): the D1100 + Mali changes and why each one was needed
2026-10-03 15:57:11 +08:00
BZLZHH
b7a6e6c538
[Docs] (Anland): the D1100 + Mali, kernel 4.14 bring-up and its three fixes
2026-10-03 15:55:53 +08:00
BZLZHH
f44888ca01
[Fix] (Espryt): present through eglSwapBuffers when the driver refuses the damage entry point
...
A vendor EGL can export eglSwapBuffersWithDamageEXT and then answer EGL_FALSE on
a window surface without setting an error (measured on an Android 4.14 vendor
kernel with a Mali driver; the surface is current, the context is current, the
ANativeWindow reports its full extent, and the plain entry point presents the
same surface happily). Trusting that pointer kept the frame: the window's buffer
queue never saw a buffer, the compositor's own window stayed a solid colour on
screen, and every client of the session fell back to wl_shm because the
compositor could not sample what it could not present.
A refused damage swap now falls back to eglSwapBuffers and says so once. A
driver without the extension presents through the plain call as before.
2026-10-03 15:54:55 +08:00
BZLZHH
8245221755
[Fix] (Remote): name a shared image by extent and format when the kernel gives descriptors one identity
...
(st_dev, st_ino) is how an incoming dma-buf descriptor names the image it was
exported from, and it is the only identity the registry had. Some kernels do not
keep those apart: on a 4.14 vendor kernel every descriptor of every image reports
dev 12 ino 10658 with st_size 0, and /proc/self/fdinfo carries nothing else, so
the registry correctly detected the ambiguity and then refused every import -
which is every zero-copy window a compositor ever tries to sample, leaving the
whole stack on the wl_shm fallback.
The identity stays the kernel's to give. When it is ambiguous, the import's own
extent and format select among the live images exported under that identity, in
allocation order and each claimed once, which is the order a producer allocates
and hands them over in. Identify() takes the extent and the fourcc the import
record already carries; nothing else changes.
2026-10-03 15:54:55 +08:00
swung0x48
b0dad54cf1
[Docs] (Anland): device numbers and the open stale-region issue for X11 on the GPU
2026-10-03 03:20:12 -04:00
swung0x48
95aa618d22
[Docs] (Anland): x11-verify measures 1280x720 glxgears and runs Qt apps on xcb
2026-10-03 02:58:03 -04:00
swung0x48
5de6ecd4e1
[Fix] (SharedImage): uploads into a texture bound to a shared image are written into the image, and a producer's flush sends them first
2026-10-03 02:58:02 -04:00
swung0x48
7fd965ceda
[Fix] (SharedImage): a session that writes images publishes every frame boundary as a write
2026-10-03 01:51:25 -04:00
swung0x48
9e7305b885
[Docs] (Anland): plan and device scripts for X11 apps on the GPU
2026-10-03 01:51:24 -04:00
swung0x48
bd4c5b5770
[Test] (Gbm): drive mobilegl_gbm through the system loader the way Xwayland's glamor does
2026-10-03 01:51:23 -04:00
swung0x48
198d6feaf3
[Feature] (GLX): present windows as shared images through DRI3 and Present, with MIT-SHM readback as the fallback
2026-10-03 01:51:22 -04:00