[Mesa-dev] [ANNOUNCE] Mesa 18.3.3 release candidate

Thu Jan 31 23:40:04 UTC 2019

On Wed, 30 Jan 2019 18:33:35 +0000 Emil Velikov <emil.l.velikov at gmail.com> said:

You might want to hold off on this. My bugfix was actually patched out by partly
removing some of it. The void ptr math should never have been there and wasn't
in the final patch.

I'm talking about:

+                void *cpu2 = cpu + 8;

In 300d3ae8b1445b5060f92c77c0f577f4b7b2c7d6

At least with gcc8 mesa is a dud on Raspberry Pi (can't upload/downlaod
textures without crashing) without the fixes. I moved the secondary ptr math
into the ASM chunk because the C compiler seemed to just mess up cpu2 ptr
content/value for me on gcc8 (it also kept the parameter inputs/outputs cleaner
and consistent with other ASM chunks). Keeping this as void ptr math alone is
just wrong and asking for trouble and as it unfixed a fix I already had in
submitted patches.

Being at FOSDEM I now no longer have access to my OS image with all of this set
up to test and won't until next week. I can't dig in and verify. Without my
fixes at all it's a dead man walking with gcc8, and thus Arch Linux is broken
entirely on Rpi without it (and has been for a while now).

> Hello list,
> 
> The candidate for the Mesa 18.3.3 is now available. Currently we have:
>  - 45 queued
>  - 4 nominated (outstanding)
>  - and 3 rejected patches
> 
> 
> In the current release candidate we have fix for a memory leak in the etnaviv
> driver, better NEON assembly code in vc4 and couple of stability improvements
> to the radeonsi driver. A memory leak affecting all gallium drivers have been
> addressed.
> 
> The time required to compile GLSL shaders with large amount of uniforms, such
> as Godot, has been improved.
> 
> GLX and swrast have also been improved for better. 
> 
> On the Vulkan side, the ANV driver adjusted the number of images supported
> for gen8 and earlier to 8, while for newer hardware it remains the same 64.
> The RADV driver has seens a collection of stability improvements and fix for
> the Vulkan version advertised in some corner cases.
> 
> To top it all up, the meson build system has seen a steady amount of fixes:
> the DSO version for the VDPAU drivers is now included, ICC compatibility
> patches, opencl handling and relaxed handling when building osmesa.
> 
> 
> Take a look at section "Mesa stable queue" for more information.
> 
> 
> Testing reports/general approval
> --------------------------------
> Any testing reports (or general approval of the state of the branch) will be
> greatly appreciated.
> 
> The plan is to have 18.3.3 this thursday (31st January 2019), around or
> shortly after 18:00 GMT.
> 
> If you have any questions or suggestions - be that about the current patch
> queue or otherwise, please go ahead.
> 
> 
> Trivial merge conflicts
> -----------------------
> 
> commit b280cdb59e38af5d10e148fb5f2ff5e29503bf10
> Author: Carsten Haitzler (Rasterman) <raster at rasterman.com>
> 
>     vc4: Use named parameters for the NEON inline asm.
> 
>     (cherry picked from commiti 522f68847152e9111def094f7fb35b44f3d0fc80)
> 
> 
> commit 813f0a82960ed637e862ee596cef23c6574a7888
> Author: Carsten Haitzler (Rasterman) <raster at rasterman.com>
> 
>     vc4: Declare the cpu pointers as being modified in NEON asm.
> 
>     (cherry picked from commit 300d3ae8b1445b5060f92c77c0f577f4b7b2c7d6)
>     (cherry picked from commit 385843ac3ce1b868d9e24fcb2dbc0c8d5f5a7c99)
> 
> 
> commit 5e8af9e609632f7062bf9dc81b44c6877a7d8ee4
> Author: Tomeu Vizoso <tomeu.vizoso at collabora.com>
> 
>     etnaviv: Consolidate buffer references from framebuffers
> 
>     (cherry picked from commit bf1dfcc3e8120400c9a78d03dd914a786728b5f7)
> 
> 
> 
> Cheers,
> Emil
> 
> 
> Mesa stable queue
> -----------------
> 
> Nominated (4)
> ==============
> 
> Ernestas Kulik (2):
>       f6e49d5ad0f vc4: Fix leak in HW queries error path
>       90458bef544 v3d: Fix leak in resource setup error path
> 
> Jason Ekstrand (1):
>       cf42b0f9e2d intel/fs: Handle IMAGE_SIZE in size_read() and
> is_send_from_grf()
> 
> Rob Clark (1):
>       c3baa077bf6 freedreno: stop frob'ing pipe_resource::nr_samples
> 
> 
> Queued (45)
> ===========
> 
> Andres Gomez (2):
>       bin/get-pick-list.sh: fix the oneline printing
>       bin/get-pick-list.sh: fix redirection in sh
> 
> Axel Davy (1):
>       st/nine: Immediately upload user provided textures
> 
> Bas Nieuwenhuizen (3):
>       radv: Only use 32 KiB per threadgroup on Stoney.
>       radv: Set partial_vs_wave for pipelines with just GS, not tess.
>       nir: Account for atomics in copy propagation.
> 
> Bruce Cherniak (1):
>       gallium/swr: Fix multi-context sync fence deadlock.
> 
> Carsten Haitzler (Rasterman) (2):
>       vc4: Use named parameters for the NEON inline asm.
>       vc4: Declare the cpu pointers as being modified in NEON asm.
> Squashed with:
>       vc4: Declare the last cpu pointer as being modified in NEON asm.
> 
> Dave Airlie (3):
>       dri_interface: add put shm image2 (v2)
>       glx: add support for putimageshm2 path (v2)
>       gallium: use put image shm2 path (v2)
> 
> Dylan Baker (4):
>       meson: allow building dri driver without window system if osmesa is
> classic meson: fix swr KNL build
>       meson: Fix compiler checks for SWR with ICC
>       meson: Add warnings and errors when using ICC
> 
> Emil Velikov (4):
>       docs: add sha256 checksums for 18.3.2
>       cherry-ignore: radv: Fix multiview depth clears
>       cherry-ignore: spirv: Handle arbitrary bit sizes for deref array indices
>       cherry-ignore: WARNING: Commit XXX lists invalid sha
> 
> Eric Anholt (2):
>       vc4: Don't leak the GPU fd for renderonly usage.
>       vc4: Enable NEON asm on meson cross-builds.
> 
> Eric Engestrom (2):
>       configure: EGL requirements only apply if EGL is built
>       meson/vdpau: add missing soversion
> 
> Iago Toral Quiroga (1):
>       anv/device: fix maximum number of images supported
> 
> Jason Ekstrand (3):
>       anv/nir: Rework arguments to apply_pipeline_layout
>       anv: Only parse pImmutableSamplers if the descriptor has samplers
>       nir/xfb: Fix offset accounting for dvec3/4
> 
> Karol Herbst (2):
>       nv50/ir: disable tryCollapseChainedMULs in ConstantFolding for precise
> instructions glsl/lower_output_reads: set invariant and precise flags on
> temporaries
> 
> Lionel Landwerlin (1):
>       anv: fix invalid binding table index computation
> 
> Marek Olšák (4):
>       radeonsi: also apply the GS hang workaround to draws without
> tessellation radeonsi: fix a u_blitter crash after a shader with FBFETCH
>       radeonsi: fix rendering to tiny viewports where the viewport center is
> > 8K st/mesa: purge framebuffers when unbinding a context
> 
> Niklas Haas (1):
>       radv: correctly use vulkan 1.0 by default
> 
> Pierre Moreau (1):
>       meson: Fix with_gallium_icd to with_opencl_icd
> 
> Rob Clark (1):
>       loader: fix the no-modifiers case
> 
> Samuel Pitoiset (1):
>       radv: clean up setting partial_es_wave for distributed tess on VI
> 
> Timothy Arceri (4):
>       ac/nir_to_llvm: fix interpolateAt* for arrays
>       ac/nir_to_llvm: fix clamp shadow reference for more hardware
>       radv/ac: fix some fp16 handling
>       glsl: use remap location when serialising uniform program resource data
> 
> Tomeu Vizoso (1):
>       etnaviv: Consolidate buffer references from framebuffers
> 
> Vinson Lee (1):
>       meson: Fix typo.
> 
> 
> Rejected (3)
> ============
> 
> Bas Nieuwenhuizen (1):
>       f67dea5e19e radv: Fix multiview depth clears
> 
> Reason: This commit requires commits aeaf8dbd097 and 7484bc894b9 which did not
>         land in branch.
> 
> Jason Ekstrand (2):
>       bfe31c5e461 nir/builder: Add nir_i2i and nir_u2u helpers which take a
> bit size abfe674c54b spirv: Handle arbitrary bit sizes for deref array indices
> 
> Reason: The commits aren't suitable in their present form. Backport
>          requested, silence from developer.
> 

-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
Carsten Haitzler - raster at rasterman.com