Re: [Mesa-dev] size of LP_MAX_VBUF_SIZE
Am 20.02.20 um 02:45 schrieb Dave Airlie: > Hey, > > Anyone know why LP_MAX_VBUF_SIZE is only 4K? > > Tess submits > 100K verts to the draw pipeline, which start to split > them, due to the 4K size of the above it splits 50 vertices per vbuf, > however it then calls draw_reset_vertex_ids which iterates over all > 100K vertices each time. > > I might try fixing the reset, but I wondered why this was only sending > 50 vertices at a time to the rasterizer. > > Dave. > Dave, I don't recall, I think this even predates me working on llvmpipe... That said, I think in general splitting into smaller chunks is done so things are more cache friendly (though the limit is so low it would fit into l1 cache even back then...). And probably the overhead of invoking things multiple times just wasn't all that large compared to the execution time of the vs (and the setup code in llvmpipe). I don't know if that was actually measured though at some point, and it is quite possible the average vertex size got quite a bit larger since then (hence max vertices per split lower), as everything was geared towards quite simple apps back then. So I think increasing the limit is probably quite fine, but splitting still needs to work correctly. Roland ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [ANNOUNCE] mesa 20.0.0
Please note that Mesa 20.0 now defaults to the 'iris' driver for Intel Gen8+ GPUs, and so you may want to set the -Dprefer-iris build option. See this email for more details: https://lists.freedesktop.org/archives/mesa-maintainers/2020-January/000150.html --Ken On Wednesday, February 19, 2020 2:07:58 PM PST Dylan Baker wrote: > Hi list, > > I'd like to announce mesa 20.0.0 as available for download immediately. I'm > very > pleased that we could get all of the issues blocking the release nailed down > quickly and make a release on time for once! > > This is a .0 release, and you may want to continue to to track 19.3.x until > 20.0.1 comes out in two weeks. 19.3.5 is planned to be the final 19.3 release > and is planned for next Wednesday. > > Dylan > > Shortlog > > > Alyssa Rosenzweig (3): > pan/midgard: Fix missing prefixes > pan/midgard: Don't crash with constants on unknown ops > pan/midgard: Use fprintf instead of printf for constants > > Danylo Piliaiev (1): > st/nir: Unify inputs_read/outputs_written before serializing NIR > > Dylan Baker (6): > .pick_status.json: Update to 2a98cf3b2ecea43cea148df7f77d2abadfd1c9db > .pick_status.json: Update to 946eacbafb47c8b94d47e7c9d2a8b02fff5a22fa > .pick_status.json: Update to bee5c9b0dc13dbae0ccf124124eaccebf7f2a435 > Docs: Add 20.0.0 release notes > docs: Empty new_features.txt > VERSION: bump for 20.0.0 release > > Erik Faye-Lund (1): > Revert "nir: Add a couple trivial abs optimizations" > > Francisco Jerez (6): > intel/fs/cse: Make HALT instruction act as CSE barrier. > intel/fs/gen7: Fix fs_inst::flags_written() for > SHADER_OPCODE_FIND_LIVE_CHANNEL. > intel/fs: Add virtual instruction to load mask of live channels into > flag register. > intel/fs/gen12: Workaround unwanted SEND execution due to broken NoMask > control flow. > intel/fs/gen12: Fixup/simplify SWSB annotations of SIMD32 scratch > writes. > intel/fs/gen12: Workaround data coherency issues due to broken NoMask > control flow. > > Krzysztof Raszkowski (1): > gallium/swr: simplify environmental variabled expansion code > > Marek Olšák (1): > radeonsi: don't wait for shader compilation to finish when destroying a > context > > Mathias Fröhlich (1): > egl: Implement getImage/putImage on pbuffer swrast. > > Peng Huang (1): > radeonsi: make si_fence_server_signal flush pipe without work > > Pierre-Eric Pelloux-Prayer (1): > radeonsi/ngg: add VGT_FLUSH when enabling fast launch > > Tapani Pälli (2): > glsl: fix a memory leak with resource_set > iris: fix aux buf map failure in 32bits app on Android > > Thong Thai (1): > Revert "st/va: Convert interlaced NV12 to progressive" > > Timothy Arceri (1): > glsl: fix gl_nir_set_uniform_initializers() for image arrays > > luc (1): > zink: confused compilation macro usage for zink in target helpers. > > > > git tag: mesa-20.0.0 > > https://mesa.freedesktop.org/archive/mesa-20.0.0.tar.xz > SHA256: bb6db3e54b608d2536d4000b3de7dd3ae115fc114e8acbb5afff4b3bbed04b34 > mesa-20.0.0.tar.xz > SHA512: > 3968820029434682fb6644947c76e12b6bb991a3d05cb519a6ea971ecc4aa9b21d03f84b9a452cc21cc77a7981db9e40dd2dfd3ea3dee85b3dd88a59b4842bbb > mesa-20.0.0.tar.xz > PGP: https://mesa.freedesktop.org/archive/mesa-20.0.0.tar.xz.sig > > signature.asc Description: This is a digitally signed message part. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
