1Mesa 20.0.2 Release Notes / 2020-03-18 2====================================== 3 4Mesa 20.0.2 is a bug fix release which fixes bugs found since the 20.0.1 5release. 6 7Mesa 20.0.2 implements the OpenGL 4.6 API, but the version reported by 8glGetString(GL_VERSION) or glGetIntegerv(GL_MAJOR_VERSION) / 9glGetIntegerv(GL_MINOR_VERSION) depends on the particular driver being 10used. Some drivers don't support all the features required in OpenGL 114.6. OpenGL 4.6 is **only** available if requested at context creation. 12Compatibility contexts may report a lower version depending on each 13driver. 14 15Mesa 20.0.2 implements the Vulkan 1.2 API, but the version reported by 16the apiVersion property of the VkPhysicalDeviceProperties struct depends 17on the particular driver being used. 18 19SHA256 checksum 20--------------- 21 22:: 23 24 aa54f1cb669550606aab8ceb475105d15aeb814fca5a778ce70d0fd10e98e86f mesa-20.0.2.tar.xz 25 26New features 27------------ 28 29Bug fixes 30--------- 31 32- RPCS3 / Persona 5 - Performance regression [RADV / Navi] 33- [CTS] dEQP-VK.descriptor_indexing.\* fails on RADV/LLVM 34- [RadeonSI][gfx10/navi] Kerbal Space Program crash: si_draw_vbo: 35 Assertion \`0' failed 36- src/compiler/glsl/glcpp/glcpp-parse.y:1297: \_token_print: Assertion 37 \`!"Error: Don't know how to print token."' failed. 38- Budget Cuts hits VK_AMD_shader_fragment_mask assert 39 40Changes 41------- 42 43Andreas Baierl (1): 44 45- gitlab-ci: Add add a set of lima flakes 46 47Bas Nieuwenhuizen (2): 48 49- amd/llvm: Fix divergent descriptor indexing. (v3) 50- amd/llvm: Fix divergent descriptor regressions with radeonsi. 51 52Danylo Piliaiev (2): 53 54- glsl: do not crash if string literal is used outside of 55 #include/#line 56- st/mesa: Fix signed integer overflow when using 57 util_throttle_memory_usage 58 59Dave Airlie (1): 60 61- gallium: fix build with latest meson and gcc10 62 63Dylan Baker (8): 64 65- docs: Add sha256sums for 20.0.1 66- .pick_status.json: Update to 07f1ef5656e0721282d01a8421eaca056348137d 67- .pick_status.json: Update to 70341d7746c177a4cd7377ef633e9f85afd11d54 68- .pick_status.json: Update to 625d8705f02e211e2733c3fe12845505725c37d4 69- .pick_status.json: Mark b83c9aca4a5fd02d920c90c1799137fed52dc1d9 as 70 backported 71- .pick_status.json: Update to ee9e0d1ecae307fa48200d2604d3114070253299 72- .pick_status.json: Update to 3dd0d12aa5fefa94123269a541c94cdf57599e34 73- .pick_status.json: Update to 94e37859a96cc56cf0c5418a5af00a3e9f5a1bf5 74 75Eric Anholt (1): 76 77- glsl/tests: Fix waiting for disk_cache_put() to finish. 78 79Eric Engestrom (7): 80 81- bin/gen_release_notes.py: fix commit list command 82- .pick_status.json: Update to 24db276d11976905b2e8a44965c684bb48c3d49f 83- gen_release_notes: fix vulkan version reported 84- docs/relnotes/20.0: fix vulkan version reported 85- .pick_status.json: Update to ba03e308b66b0b88f60b99d9d47851a5e1522e6e 86- vulkan/wsi: fix cleanup when dup() fails 87- gen_release_notes: fix version in "you should wait" message 88 89Francisco Jerez (1): 90 91- intel/fs: Fix workaround for VxH indirect addressing bug under 92 control flow. 93 94Jason Ekstrand (9): 95 96- isl: Set 3DSTATE_DEPTH_BUFFER::Depth correctly for 3D surfaces 97- iris: Don't skip fast depth clears if the color changed 98- anv: Parse VkPhysicalDeviceFeatures2 in CreateDevice 99- vulkan/wsi: Don't leak the FD when 100 GetImageDrmFormatModifierProperties fails 101- vulkan/wsi: Return an error if dup() fails 102- anv: Use the PIPE_CONTROL instead of bits for the CS stall W/A 103- anv: Use a proper end-of-pipe sync instead of just CS stall 104- anv: Do end-of-pipe sync around MCS/CCS ops instead of CS stall 105- anv: Do an end-of-pipe sync before updating AUX table entries 106 107José Fonseca (1): 108 109- meson: Avoid duplicate symbols. 110 111Kristian Høgsberg (2): 112 113- Revert "glsl: Use a simpler formula for tanh" 114- Revert "spirv: Use a simpler and more correct implementaiton of 115 tanh()" 116 117Marek Olšák (4): 118 119- Revert "mesa: check for z=0 in \_mesa_Vertex3dv()" 120- radeonsi: add a bug workaround for NGG - LATE_ALLOC_GS 121- ac: add a bug workaround for the 100% NGG culling case 122- gallium/cso_context: remove cso_delete_xxx_shader helpers to fix the 123 live cache 124 125Martin Fuzzey (3): 126 127- freedreno: android: fix build failure on android due to python 128 version 129- freedreno: android: add a6xx-pack.xml.h generation to android build 130- freedreno: android: fix build of perfcounters. 131 132Michel Dänzer (1): 133 134- llvmpipe: Use uintptr_t for pointer values 135 136Rafael Antognolli (3): 137 138- anv: Wait for the GPU to be idle before invalidating the aux table. 139- iris: Split aux map initialization from invalidation. 140- iris: Wait for the GPU to be idle before invalidating the aux table. 141 142Rob Clark (1): 143 144- freedreno: fix FD_MESA_DEBUG=inorder 145 146Samuel Pitoiset (5): 147 148- aco: fix image load/store with lod and 1D images 149- nir/lower_input_attachments: remove bogus assert in 150 try_lower_input_texop() 151- ac/llvm: add missing optimization barrier for 64-bit readlanes 152- radv: only inject implicit subpass dependencies if necessary 153- radv: fix random depth range unrestricted failures due to a cache 154 issue 155 156Timur Kristóf (2): 157 158- nir: Add ability to lower non-const quad broadcasts to const ones. 159- radv: Enable lowering dynamic quad broadcasts. 160 161Vinson Lee (1): 162 163- st/nine: Fix incompatible-pointer-types-discards-qualifiers errors. 164