Hash :
869dde72
Author :
Date :
2025-02-07T15:44:12
Remove code paths for invoking the offline Metal compiler We experimented with invoking the Metal shader compiler directly to create cacheable shader libraries but the project did not show useful gains. It is currently all dead code, remove it. Bug: b/391990604 Change-Id: I6d13278d1a27d8b3ae961a0846f31292f2506b23 Reviewed-on: https://chromium-review.googlesource.com/c/angle/angle/+/6245258 Reviewed-by: Kimmo Kinnunen <kkinnunen@apple.com> Commit-Queue: Geoff Lang <geofflang@chromium.org>
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423
{
"description": [
"Copyright 2022 The ANGLE Project Authors. All rights reserved.",
"Use of this source code is governed by a BSD-style license that can be",
"found in the LICENSE file.",
"",
"mtl_features.json: Optional features for the Metal renderer."
],
"features": [
{
"name": "has_base_vertex_instanced_draw",
"category": "Features",
"description": [
"The renderer supports base vertex instanced draw"
]
},
{
"name": "has_explicit_mem_barrier",
"category": "Features",
"description": [
"The renderer supports explicit memory barrier"
]
},
{
"name": "has_cheap_render_pass",
"category": "Features",
"description": [
"The renderer can cheaply break a render pass."
]
},
{
"name": "has_non_uniform_dispatch",
"category": "Features",
"description": [
"The renderer supports non uniform compute shader dispatch's group size"
]
},
{
"name": "has_shader_stencil_output",
"category": "Features",
"description": [
"The renderer supports stencil output from fragment shader"
]
},
{
"name": "has_texture_swizzle",
"category": "Features",
"description": [
"The renderer supports texture swizzle"
]
},
{
"name": "has_depth_auto_resolve",
"category": "Features",
"description": [
"The renderer supports MSAA depth auto resolve at the end of render pass"
]
},
{
"name": "has_stencil_auto_resolve",
"category": "Features",
"description": [
"The renderer supports MSAA stencil auto resolve at the end of render pass"
]
},
{
"name": "has_events",
"category": "Features",
"description": [
"The renderer supports MTL(Shared)Event"
]
},
{
"name": "allow_inline_const_vertex_data",
"category": "Features",
"description": [
"The renderer supports using inline constant data for small client vertex data"
]
},
{
"name": "allow_separate_depth_stencil_buffers",
"category": "Features",
"description": [
"Some Apple platforms such as iOS allows separate depth and stencil buffers, ",
"whereas others such as macOS don't"
]
},
{
"name": "allow_runtime_sampler_compare_mode",
"category": "Features",
"description": [
"The renderer supports changing sampler's compare mode outside shaders"
]
},
{
"name": "allow_sampler_compare_gradient",
"category": "Features",
"description": [
"The renderer supports sample_compare with gradients"
]
},
{
"name": "allow_buffer_read_write",
"category": "Features",
"description": [
"The renderer supports buffer read and write in the same shader"
]
},
{
"name": "allow_multisample_store_and_resolve",
"category": "Features",
"description": [
"The renderer supports MSAA store and resolve in the same pass"
]
},
{
"name": "allow_gen_multiple_mips_per_pass",
"category": "Features",
"description": [
"The renderer supports generating multiple mipmaps per pass"
]
},
{
"name": "force_D24S8_as_unsupported",
"category": "Features",
"description": [
"Force Depth24Stencil8 format as unsupported."
]
},
{
"name": "force_buffer_GPU_storage",
"category": "Features",
"description": [
"On systems that support both buffer' memory allocation on GPU and shared memory (such as ",
"macOS), force using GPU memory allocation for buffers everytime or not."
]
},
{
"name": "force_non_CS_base_mipmap_generation",
"category": "Features",
"description": [
"Turn this feature on to disallow Compute Shader based mipmap generation. Compute Shader ",
"based mipmap generation might cause GPU hang on some older iOS devices."
]
},
{
"name": "emulate_transform_feedback",
"category": "Features",
"description": [
"Turn this on to allow transform feedback in Metal using a 2-pass VS for GLES3."
]
},
{
"name": "intel_explicit_bool_cast_workaround",
"category": "Workarounds",
"description": [
"Insert explicit casts for float/double/unsigned/signed int on macOS 10.15 with Intel ",
"driver"
]
},
{
"name": "intel_disable_fast_math",
"category": "Workarounds",
"description": [
"Disable fast math in atan and invariance cases when running below macOS 12.0"
]
},
{
"name": "allow_renderpass_without_attachment",
"category": "Features",
"description": [
"Allow creation of render passes without any attachments"
]
},
{
"name": "avoid_stencil_texture_swizzle",
"category": "Features",
"description": [
"Do not create swizzled views of stencil textures"
]
},
{
"name": "emulate_alpha_to_coverage",
"category": "Workarounds",
"description": [
"Some GPUs ignore alpha-to-coverage when [[sample_mask]] is written"
]
},
{
"name": "write_helper_sample_mask",
"category": "Workarounds",
"description": [
"Some GPUs produce incorrect derivatives unless [[sample_mask]] is written"
]
},
{
"name": "pre_transform_texture_cube_grad_derivatives",
"category": "Workarounds",
"description": [
"Apply a vendor-specific transformation to explicit cubemap derivatives"
]
},
{
"name": "multisample_color_format_shader_read_workaround",
"category": "Workarounds",
"description": [
"Add shaderRead usage to some multisampled texture formats"
],
"issue": "http://anglebug.com/42265518"
},
{
"name": "copy_IOSurface_to_non_IOSurface_for_read_optimization",
"category": "Workarounds",
"description": [
"some GPUs are faster to read an IOSurface texture by first copying the texture to a ",
"non-IOSurface texture"
],
"issue": "http://anglebug.com/40096835 http://anglebug.com/40096850"
},
{
"name": "copy_texture_to_buffer_for_read_optimization",
"category": "Workarounds",
"description": [
"some GPUs are faster to read a texture by first copying the texture to a buffer"
],
"issue": "http://anglebug.com/40096835"
},
{
"name": "limit_max_draw_buffers_for_testing",
"category": "Features",
"description": [
"Used to check the backend works when the device's advertized limit is less than the code's limit"
],
"issue": "http://anglebug.com/42265754"
},
{
"name": "limit_max_color_target_bits_for_testing",
"category": "Features",
"description": [
"Metal iOS has a limit on the number of color target bits per pixel."
],
"issue": "http://anglebug.com/42265754"
},
{
"name": "preemptively_start_provoking_vertex_command_buffer",
"category": "Features",
"description": [
"AMD Metal Drivers appear to have a bug this works around"
],
"issue": "http://anglebug.com/42266101"
},
{
"name": "upload_data_to_iosurfaces_with_staging_buffers",
"category": "Workarounds",
"description": [
"When uploading data to IOSurface-backed textures, use a staging buffer."
],
"issue": "http://anglebug.com/40096850"
},
{
"name": "always_use_staged_buffer_updates",
"category": "Features",
"description": [
"Always update buffers by copying the data to a staging buffer and then blitting it to the actual buffer"
],
"issue": "http://anglebug.com/40644888"
},
{
"name": "use_shadow_buffers_when_appropriate",
"category": "Features",
"description": [
"On some architectures using a shadow buffer can be faster for certain size buffers"
],
"issue": "http://anglebug.com/40644888"
},
{
"name": "always_use_managed_storage_mode_for_buffers",
"category": "Features",
"description": [
"Metal buffers can be managed, shared, or private. Sometimes managed is fastest"
],
"issue": "http://anglebug.com/40644888"
},
{
"name": "always_use_shared_storage_mode_for_buffers",
"category": "Features",
"description": [
"Metal buffers can be managed, shared, or private. Sometimes shared is fastest"
],
"issue": "http://anglebug.com/40644888"
},
{
"name": "prefer_cpu_for_buffersubdata",
"category": "Features",
"description": [
"Makes bufferSubData always update via CPU"
],
"issue": "http://anglebug.com/40644888"
},
{
"name": "disable_programmable_blending",
"category": "Features",
"description": [
"Disable programmable blending in order to test read_write pixel local storage textures"
],
"issue": "http://anglebug.com/40096838"
},
{
"name": "disable_r_w_texture_tier2_support",
"category": "Features",
"description": [
"Disable tier2 read_write textures in order to test tier1 support"
],
"issue": "http://anglebug.com/40096838"
},
{
"name": "disable_raster_order_groups",
"category": "Features",
"description": [
"Disable raster order groups in order to test pixel local storage memory barriers"
],
"issue": "http://anglebug.com/40096838"
},
{
"name": "enable_in_memory_mtl_library_cache",
"category": "Features",
"description": [
"Cache MTLLibrary objects in memory."
],
"issue": "http://crbug.com/1385510"
},
{
"name": "enable_parallel_mtl_library_compilation",
"category": "Features",
"description": [
"Compile MTLLibrary in multiple threads."
],
"issue": "http://crbug.com/1385510"
},
{
"name": "always_prefer_staged_texture_uploads",
"category": "Features",
"description": [
"Always prefer to upload texture data via a staging buffer and avoid MTLTexture::replaceRegion."
],
"issue": "http://crbug.com/1380790"
},
{
"name": "disable_staged_initialization_of_packed_texture_formats",
"category": "Features",
"description": [
"Staged GPU upload of some packed texture formats such as RGB9_E5 fail on Intel GPUs."
],
"issue": "http://anglebug.com/40644905"
},
{
"name": "generate_shareable_shaders",
"category": "Features",
"description": [
"Attempts to generate shaders that are shareable. More specifically, shaders",
" end up with conditionals that are decided at run time via input parameters vs",
" compile time. This results in bigger shaders."
],
"issue": "http://crbug.com/1423136"
},
{
"name": "disable_metal_on_nvidia",
"category": "Features",
"description": [
"NVIDIA GPUs are unsupported due to scarcity of the hardware."
],
"issue": "http://anglebug.com/42266609"
},
{
"name": "flush_after_stream_vertex_data",
"category": "Features",
"description": [
"Flush after calls to StreamVertexData to work around driver bugs."
]
},
{
"name": "require_gpu_family_2",
"category": "Features",
"description": [
"Mac GPU Family 2 is required to support all the features of OpenGL ES 2.0"
],
"issue": "http://anglebug.com/40096869"
},
{
"name": "rescope_global_variables",
"category": "Features",
"description": [
"Rescope global variables that are only used in one function to be function-local."
],
"issue": "http://anglebug.com/42266744"
},
{
"name": "enable_multisampled_render_to_texture_on_non_tilers",
"category": "Features",
"description": [
"Enable EXT_multisampled_render_to_texture on non tiled GPUs."
],
"issue": "http://anglebug.com/42261786"
},
{
"name": "emulate_dont_care_load_with_random_clear",
"category": "Features",
"description": [
"Emulate DontCare loadAction with Clear loadAction. The clear values will be randomized."
],
"issue": "http://anglebug.com/42261786"
},
{
"name": "inject_asm_statement_into_loop_bodies",
"category": "Features",
"description": [
"Inject asm(\"\") statements into loop bodies to force all loops to be treated as ",
"having side effects and not optimized out."
],
"issue": "http://crbug.com/1513738"
}
]
}