• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1Mesa 22.2.1 Release Notes / 2022-10-11
2======================================
3
4Mesa 22.2.1 is a bug fix release which fixes bugs found since the 22.2.0 release.
5
6Mesa 22.2.1 implements the OpenGL 4.6 API, but the version reported by
7glGetString(GL_VERSION) or glGetIntegerv(GL_MAJOR_VERSION) /
8glGetIntegerv(GL_MINOR_VERSION) depends on the particular driver being used.
9Some drivers don't support all the features required in OpenGL 4.6. OpenGL
104.6 is **only** available if requested at context creation.
11Compatibility contexts may report a lower version depending on each driver.
12
13Mesa 22.2.1 implements the Vulkan 1.3 API, but the version reported by
14the apiVersion property of the VkPhysicalDeviceProperties struct
15depends on the particular driver being used.
16
17SHA256 checksum
18---------------
19
20::
21
22    0079beac0a33f45e7e0aec59e6913eafbc4268a3f1e2e330017440494f91b13c  mesa-22.2.1.tar.xz
23
24
25
26New features
27------------
28
29- None
30
31
32Bug fixes
33---------
34
35- r600: Shadows incorrecly rendered on Counter-Strike: Global Offensive with the NIR backend
36- radeonsi/radv: Discrepancy in handling wd_switch_on_eop
37- Shadows rendering issue in Deathloop
38- RADV: Regression with RPCS3 where nothing renders
39- [RADV][bisected] commit 3714c89d0ebf0c541d8f7a7a807ce7c1bfe3cfea at least lead to GPU hang when I run the game "Quake II RTX"
40
41
42Changes
43-------
44
45Caio Oliveira (1):
46
47- nir/lower_task_shader: Don't fail adding a launch when last instruction is a jump
48
49Connor Abbott (3):
50
51- tu: Fix maxPerStageDescriptorUpdateAfterBindInputAttachments
52- ir3/analyze_ubo_ranges: Account for reserved consts
53- ir3/analyze_ubo_ranges: Account for reserved consts better
54
55Daniel Schürmann (1):
56
57- aco/optimizer: disallow can_eliminate_and_exec() with s_not
58
59Dave Airlie (5):
60
61- lavapipe: fix 3d depth stencil image clearing.
62- lavapipe: take layers into account around render surface creation.
63- lavapipe: handle view mask layer count properly for begin rendering.
64- llvmpipe: memset empty images in the key
65- llvmpipe: bind sampler views/images properly.
66
67Dylan Baker (15):
68
69- docs: add release notes for 22.2.0
70- docs: Add sha256 sum for 22.2.0
71- .pick_status.json: Update to 3730be9873d556918a20795b37399b7e2e9630b5
72- .pick_status.json: Update to c268ae9efe602129b848c800dca83411d55c3aab
73- .pick_status.json: Mark 49efa73ba11c4cacaed0052b984e1fb884cf7600 as denominated
74- .pick_status.json: Mark 5b7c2dffc43eb7ada61807a6e679cc2c1b3b2682 as denominated
75- .pick_status.json: Mark 1aacd9492de42412c6c9ca79d48a40c4ffebcd79 as backported
76- .pick_status.json: Mark cbad4adc133b16c803ec9445c8dd144bc5023a62 as denominated
77- .pick_status.json: Mark 40a235c9a82027bcb92d34b7386ea6b1a2e1d587 as denominated
78- .pick_status.json: Update to 3ed712f7483791aa0a256472d276ff5a83c643bc
79- .pick_status.json: Update to e5631add0d1c97da4686f474735d45705facd308
80- .pick_status.json: Mark 5af6dad179dc4140964ab444a2a7f8cf8d16c561 as denominated
81- .pick_status.json: Update to 68f3c38c8011e3ff304a67b6ffb33fd21ee65b0c
82- .pick_status.json: Mark ff933485b7e328c5041050d5e41a2dd28a032a92 as denominated
83- docs: remove new_features.txt
84
85Emma Anholt (2):
86
87- turnip: Fix the "written stencil is unmodified" check.
88- turnip: Fix busy-waiting on syncobjs with OS_TIMEOUT_INFINITE.
89
90Erik Faye-Lund (4):
91
92- mesa/st: always use normalized coords for samplers
93- mesa/st: always use normalized coords for samplers
94- mesa/st: always use normalized coords for samplers
95- docs: fixup broken rst syntax
96
97Feng Jiang (1):
98
99- virgl/vtest: fix memory overwrite problem in virgl_vtest_send_get_caps()
100
101Gert Wollny (2):
102
103- r600/sfn: Handle shifts on Cayman
104- nir_lower_to_source_mods: Don't sneek in an abs modifier from parent
105
106James Park (1):
107
108- meson,amd: Remove Windows libelf wrap
109
110Jason Ekstrand (3):
111
112- iris: Use a larger alignment for buffer allocations
113- iris: Fix more BO alignments
114- iris: Handle resource offsets in buffer copies
115
116Jesse Natalie (1):
117
118- u_atomic: Add a helper for pointer compare-exchange
119
120Jonathan (2):
121
122- gallium/u_threaded: Prepare for buffer sharedness tracking
123- gallium/u_threaded: buffer sharedness tracking
124
125Jonathan Gray (1):
126
127- iris: check i915 features after hw gen
128
129Lionel Landwerlin (12):
130
131- anv: clear descriptorsets if AllocateDescriptorSets fails
132- anv: add missing wokraround for texture cache invalidate
133- intel/nir/rt: fixup generate hit
134- intel/nir/rt: spill/fill the entire ray query data
135- intel/nir/rt: fix ray query proceed level
136- intel/nir/rt: change scratch check validation
137- intel/nir/rt: store ray query state in scratch
138- intel/fs: fixup a64 messages
139- intel/nir: fixup preserved metadata in rayquery lowering
140- intel/nir: fix potential invalid function impl ptr usage
141- intel/nir: disable assert on async stack id
142- anv: remove HDC flush from invalidate bits
143
144Marcin Ślusarz (1):
145
146- intel/compiler: remove second shading rate lowering for mesh
147
148Marek Olšák (1):
149
150- radeonsi: don't flush asynchronously for fence_server_signal
151
152Martin Krastev (1):
153
154- meson: add svga gallium driver to 'auto' group on arm/aarch64
155
156Mike Blumenkrantz (23):
157
158- zink: uncap ssbo size
159- zink: fix atomic ssbo indexing with non-32bit values
160- zink: split up get_clear_data()
161- zink: make void clears more robust
162- zink: fix/relax resolve geometry check
163- zink: fix atomic ssbo fadd offsets
164- st/bitmap: use normalized coords in samplers with lowered rects
165- st/drawpixels: use normalized coords in samplers with lowered rects
166- zink: handle culldistance xfb outputs like clipdistance
167- lavapipe: set writemask for compute shader buffers
168- lavapipe: propagate shader access info across libraries
169- zink: free vertex_state_cache on shutdown
170- zink: always unset var->data.explicit_xfb_buffer
171- zink: avoid overflow when clamping bufferviews
172- zink: always run optimize_nir after lower_64bit_vars
173- zink: don't flatten 64bit arrays during rewrite
174- mesa: set normalized_coords for bindless texture buffer samplers
175- aux/trace: dump depth_clamp for rasterizer state
176- zink: set layouts before possibly reordering image copies
177- zink: add a draw barrier when unbinding a sampler-bound fb surface
178- glthread: handle DeleteBuffers(n=-1) gracefully
179- zink: fix/improve handling for multi-component bitfield ops
180- pipe-loader: fix zink driinfo header path
181
182Pierre-Eric Pelloux-Prayer (3):
183
184- radeonsi/gfx11: don't set VERTS_PER_SUBGRP to 0
185- tc: do a GPU->CPU copy to initialize cpu_storage
186- tc: don't use CPU storage for glBufferData
187
188Qiang Yu (1):
189
190- nir: fix nir_xfb_info buffer_to_stream length
191
192Rhys Perry (1):
193
194- aco: fix VMEMtoScalarWriteHazard s_waitcnt mitigation
195
196Samuel Pitoiset (1):
197
198- radv: make sure to initialize wd_switch_on_eop before checking its value
199
200SoroushIMG (1):
201
202- zink: fix nir_op_unpack_64_2x32 emission
203
204Tapani Pälli (1):
205
206- mesa/st: fix a set_sampler_views call parameter order
207
208Timur Kristóf (4):
209
210- aco/optimizer_postRA: Mark a register overwritten when predecessors disagree.
211- aco/optimizer_postRA: Don't assume all operand registers were written by same instr.
212- aco/optimizer_postRA: Fix logical control flow handling.
213- radv: Use a fallback for marketing name when libdrm doesn't know it.
214
215Yiwei Zhang (1):
216
217- mesa/st: emit string marker based on PIPE_CAP_STRING_MARKER
218
219Yonggang Luo (3):
220
221- drm-shim: drop gnu99 override
222- ci/docs: Stick to specific version of python packages
223- util: Disable usage of __attribute__((__const__)) when the compiler is clang
224
225nyanmisaka (1):
226
227- gallium: Do not include offscreen winsys if glx is disabled
228