There is, first update ! Grab it from os4depot in upload query.
Change log:
1.0 (21.08.2026):
- Initial first public release.
1.1 (01.09.2026):
CORE
====
- Fixed big-endian serialization of 64-bit uniforms and bindless texture
/image handles into GPU constant buffers.
- Fixed a big-endian bug that caused scissored glClear operations on FBOs
to write (0,0,0,0) instead of the requested clear color.
Thanks to AfxGroup for the report!
- Added Mesa's pb_cache-based GPU buffer reuse, reducing allocation over-
head and improving FPS.
- Handled cases where GART buffer allocations fall back to VRAM, fixing
intermittent slowdowns and rendering flicker.
Thanks to Martin0873 and dirk67 for reports!
- Removed the hardcoded requirement for 32-bit screen modes. Three paths
are now supported:
- Native RGB565 on 16-bit screens.
- Native ARGB8888 on 32-bit screens.
- A fallback path for apps requesting ARGB8888 on a 16-bit screen,
with internal conversion to RGB565 (slower).
Thanks to Peb for the report!
- Worked around a Radeon HD bug that failed to clear all graphics state
when switching between 3D rendering and screen updates. Mesa apps using
programmable point sprites no longer hang the GPU.
- Work around a Radeon HD 5.x bug that groups Oland (chip ID 0x42) with
Hainan (0x43), causing Hainan's two-tile-pipe table to be programmed on
Oland GPUs, which normally use 4 tile pipes. At Mesa startup, affected
Oland configs are detected and reprogrammed with the correct tiling
table for the lifetime of the session. This fixes visual corruption
affecting 1D tiling, antialiasing, and related rendering paths.
- Fixed black output on RadeonHD 3.x in legacy applications using glBegin
glEnd. NOTE: RadeonHD 3.x is still not recommended due to its very poor
performance.
- The ATOM parser now tolerates option-ROM checksum mismatches instead of
failing, as some cards have a VBIOS with a stale container checksum.
Thanks to 328gts for report and tests!
- GL_RENDERER now reports the Radeon driver's Resource Manager (RM) API
version instead of the irrelevant Linux DRM version 0.0.
- Added automatic build dates to mesa.library version information.
- Improved runtime error reporting by replacing cryptic internal enum
values with meaningful descriptions.
TOOLS AND BENCHMARKS
====================
- c:radeon_gpu_info now also reports the mesa.library version and date.
- Improved glMark2:
- Added mesa.library add driver version information.
- Removed misleading EGL state errors.
- Fixed AOS4 bitmap detection for correct operation on both 16-bit and
32-bit screens.
- Added FixedGL, a benchmark for testing the legacy OpenGL fixed-function
pipeline: glBegin/glEnd, immediate mode, client arrays, display lists,
fixed matrices, fog, etc.
- Improved ShaderToy benchs:
- Added window resizing.
- Added double-click switching between windowed and fullscreen.
- Added proper support for both 16-bit and 32-bit screen modes.
- Added average FPS information on exit.
- Added four new high-quality shaders.
mesa.library 1.1 ( 1-09-2026) Tried FixedGL on my SAM460ex + RX500 and some test I only get a black window on some tests:
#d
shadertoy (dir)
glmark2 (dir)
FixedGL
#FixedGL
------------------------------------------------------------
FixedGL - Fixed-Function OpenGL Benchmark
------------------------------------------------------------
GL Vendor : AMD
GL Renderer : POLARIS12 (radeonsi, polaris12, ACO, RM 3.0, 54.76 (31.10.2025) Sam460 release)
GL Version : 4.3 (Compatibility Profile) Mesa 26.1.5 (git-2e7650ee6e)
System Info : mesa.library 1.1 (1.9.2026), RadeonRX 2.23
Surface : 640x480 windowed
Workloads : 29 (6 seconds each)
------------------------------------------------------------
[batched-triangles] vertices=3072:calls=6146: FPS: 379.13 FrameTime: 2.638 ms VertexRate: 1.16 M/s CallRate: 2.33 M/s
[split-triangles] vertices=3072:calls=8192: FPS: 117.05 FrameTime: 8.544 ms VertexRate: 0.36 M/s CallRate: 0.96 M/s
[legacy-quads] vertices=2048:calls=4098: FPS: 444.61 FrameTime: 2.249 ms VertexRate: 0.91 M/s CallRate: 1.82 M/s
[attribute-rich-quads] vertices=2048:calls=8194: FPS: 306.11 FrameTime: 3.267 ms VertexRate: 0.63 M/s CallRate: 2.51 M/s
---BLACK WINDOW---
[matrix-object-quads] vertices=2048:calls=6144: FPS: 170.87 FrameTime: 5.852 ms VertexRate: 0.35 M/s CallRate: 1.05 M/s
[medium-batch-triangles] vertices=3072:calls=6208: FPS: 541.41 FrameTime: 1.847 ms VertexRate: 1.66 M/s CallRate: 3.36 M/s
[points] vertices=2048:calls=4098: FPS: 697.19 FrameTime: 1.434 ms VertexRate: 1.43 M/s CallRate: 2.86 M/s
[lines] vertices=2048:calls=4098: FPS: 694.56 FrameTime: 1.440 ms VertexRate: 1.42 M/s CallRate: 2.85 M/s
[line-strip] vertices=528:calls=1088: FPS: 1229.64 FrameTime: 0.813 ms VertexRate: 0.65 M/s CallRate: 1.34 M/s
[line-loop] vertices=2048:calls=5120: FPS: 637.08 FrameTime: 1.570 ms VertexRate: 1.30 M/s CallRate: 3.26 M/s
[triangle-strip] vertices=1056:calls=2144: FPS: 951.93 FrameTime: 1.050 ms VertexRate: 1.01 M/s CallRate: 2.04 M/s
[triangle-fan] vertices=3072:calls=6144: FPS: 528.29 FrameTime: 1.893 ms VertexRate: 1.62 M/s CallRate: 3.25 M/s
[quad-strip] vertices=1056:calls=2144: FPS: 948.59 FrameTime: 1.054 ms VertexRate: 1.00 M/s CallRate: 2.03 M/s
[polygons] vertices=2048:calls=5120: FPS: 643.69 FrameTime: 1.554 ms VertexRate: 1.32 M/s CallRate: 3.30 M/s
[texture-only-quads] vertices=2048:calls=6146: FPS: 558.54 FrameTime: 1.790 ms VertexRate: 1.14 M/s CallRate: 3.43 M/s
[lighting-only-quads] vertices=2048:calls=6146: FPS: 558.37 FrameTime: 1.791 ms VertexRate: 1.14 M/s CallRate: 3.43 M/s
[vector-entrypoint-rich-quads] vertices=2048:calls=8194: FPS: 507.41 FrameTime: 1.971 ms VertexRate: 1.04 M/s CallRate: 4.16 M/s
[multitexture-lightmap-quads] vertices=2048:calls=10242: FPS: 403.60 FrameTime: 2.478 ms VertexRate: 0.83 M/s CallRate: 4.13 M/s
[fog-secondary-texture-quads] vertices=2048:calls=10242: FPS: 396.30 FrameTime: 2.523 ms VertexRate: 0.81 M/s CallRate: 4.06 M/s
[per-vertex-color-material-quads] vertices=2048:calls=6146: FPS: 555.96 FrameTime: 1.799 ms VertexRate: 1.14 M/s CallRate: 3.42 M/s
[array-element-rich-quads] vertices=2048:calls=2050: FPS: 1282.99 FrameTime: 0.779 ms VertexRate: 2.63 M/s CallRate: 2.63 M/s
---WINDOW WITH OBJECTS/SQUARES/CLORS---
[non-begin-end-draw-arrays-rich-triangles] vertices=3072:calls=1: FPS: 496.13 FrameTime: 2.016 ms VertexRate: 1.52 M/s CallRate: 0.00 M/s
[non-begin-end-draw-elements-rich-triangles] vertices=3072:calls=1: FPS: 387.56 FrameTime: 2.580 ms VertexRate: 1.19 M/s CallRate: 0.00 M/s
[large-display-list-rich-quads] vertices=2048:calls=1: FPS: 896.32 FrameTime: 1.116 ms VertexRate: 1.84 M/s CallRate: 0.00 M/s
[small-display-lists-rich-quads] vertices=2048:calls=512: FPS: 24.36 FrameTime: 41.056 ms VertexRate: 0.05 M/s CallRate: 0.01 M/s
---AGAIN BLACK WINDOW---
[mixed-primitive-surfaces] vertices=2254:calls=5532: FPS: 599.46 FrameTime: 1.668 ms VertexRate: 1.35 M/s CallRate: 3.32 M/s
---AGAIN WINDOW WITH OBJECTS/SQUARES/CLORS---
[static-vbo-draw-arrays-rich-triangles] vertices=3072:calls=1: FPS: 961.17 FrameTime: 1.040 ms VertexRate: 2.95 M/s CallRate: 0.00 M/s
[static-vbo-ebo-draw-elements-rich-triangles] vertices=3072:calls=1: FPS: 939.95 FrameTime: 1.064 ms VertexRate: 2.89 M/s CallRate: 0.00 M/s
[dynamic-vbo-subdata-ebo-rich-triangles] vertices=3072:calls=2: FPS: 633.30 FrameTime: 1.579 ms VertexRate: 1.95 M/s CallRate: 0.00 M/s
------------------------------------------------------------
FixedGL Score: 603
------------------------------------------------------------
Does Mesa v. 1.1 work with Radeon HD 6450 (Caicos)?
As i aware Caicos is pre RadeonSI, is R600. This one i didn't touch in mesa at all. Maybe some old version of RadeonHD like 3.7 and pre did support it ? If so, i may need to enable r600 too, but then as i have no such card i can't test it properly..
@Jabirulo Will have a look at, thanks. For sake of tests try it plz after reboot, to see if it random or stable reproduced.
@kas1e since you stated that for HD cards the driver version 5.14 is raccomanded I updated from 3.7 to 5.14.
But with glmark2 almost all tests run at lower FPS and the final score is with 3.7 3040 with 5.14 2317 big difference do you know what can causing them?
So 3.7 can be faster in those tests glMark2 tests , but for example Q3 timedemo Verde Mob on 5.20 : 127 fps, but on 3.7 115 fps.
But what is for sure, that it looks like card depends too. At least for Oland on x1000 3.7 for Oland is no go (and same results i have without mesa too, like just with warp3dnova/ogles2).
Quote:
FixedGL crash entire machine with 5.14, works with 3.7
Need crash log, and if tried you after reboot too ?
@pvanni I do check also gl4es version of glmark2 on my x1000 (some tests dind't compiles because of shaders, but most are), and so result with Verde card :
RadeonHD 3.7 : 1287 RadeonHD 5.20 : 1149
So actually, it start to looks like it family based. At least with Oland Mob. 5.20 is much better. For Verde, not so, and 3.7 there good enough as well, and while in some tests 5.20 better, in other ones 3.7 is better.
But that we tell only about glmark2. Game results compiled over gl4es so ogles2/warp3dnova are:
So see, while glmark even over gl4es is worse a bit, in games difference is quite noticeable.
But that all to be tested with mesa too. But at least glmark2 be it gl4es or mesa, show a bit worse results in 5.20, while reallife show 5.20 as better.
But it's not THAT bad as i think before because of Oland only where difference is so big.