2 * Copyright © 2015 Intel Corporation
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
11 * The above copyright notice and this permission notice (including the next
12 * paragraph) shall be included in all copies or substantial portions of the
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
18 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
20 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
28 #include <sys/sysinfo.h>
32 #include "drm-uapi/drm_fourcc.h"
34 #include "anv_private.h"
35 #include "util/debug.h"
36 #include "util/build_id.h"
37 #include "util/disk_cache.h"
38 #include "util/mesa-sha1.h"
39 #include "util/os_file.h"
40 #include "util/u_atomic.h"
41 #include "util/u_string.h"
42 #include "util/xmlpool.h"
45 #include "common/gen_aux_map.h"
46 #include "common/gen_defines.h"
47 #include "compiler/glsl_types.h"
49 #include "genxml/gen7_pack.h"
51 static const char anv_dri_options_xml
[] =
53 DRI_CONF_SECTION_PERFORMANCE
54 DRI_CONF_VK_X11_OVERRIDE_MIN_IMAGE_COUNT(0)
55 DRI_CONF_VK_X11_STRICT_IMAGE_COUNT("false")
58 DRI_CONF_SECTION_DEBUG
59 DRI_CONF_ALWAYS_FLUSH_CACHE("false")
60 DRI_CONF_VK_WSI_FORCE_BGRA8_UNORM_FIRST("false")
64 /* This is probably far to big but it reflects the max size used for messages
65 * in OpenGLs KHR_debug.
67 #define MAX_DEBUG_MESSAGE_LENGTH 4096
70 compiler_debug_log(void *data
, const char *fmt
, ...)
72 char str
[MAX_DEBUG_MESSAGE_LENGTH
];
73 struct anv_device
*device
= (struct anv_device
*)data
;
74 struct anv_instance
*instance
= device
->physical
->instance
;
76 if (list_is_empty(&instance
->debug_report_callbacks
.callbacks
))
81 (void) vsnprintf(str
, MAX_DEBUG_MESSAGE_LENGTH
, fmt
, args
);
84 vk_debug_report(&instance
->debug_report_callbacks
,
85 VK_DEBUG_REPORT_DEBUG_BIT_EXT
,
86 VK_DEBUG_REPORT_OBJECT_TYPE_UNKNOWN_EXT
,
91 compiler_perf_log(void *data
, const char *fmt
, ...)
96 if (unlikely(INTEL_DEBUG
& DEBUG_PERF
))
97 intel_logd_v(fmt
, args
);
103 anv_compute_heap_size(int fd
, uint64_t gtt_size
)
105 /* Query the total ram from the system */
109 uint64_t total_ram
= (uint64_t)info
.totalram
* (uint64_t)info
.mem_unit
;
111 /* We don't want to burn too much ram with the GPU. If the user has 4GiB
112 * or less, we use at most half. If they have more than 4GiB, we use 3/4.
114 uint64_t available_ram
;
115 if (total_ram
<= 4ull * 1024ull * 1024ull * 1024ull)
116 available_ram
= total_ram
/ 2;
118 available_ram
= total_ram
* 3 / 4;
120 /* We also want to leave some padding for things we allocate in the driver,
121 * so don't go over 3/4 of the GTT either.
123 uint64_t available_gtt
= gtt_size
* 3 / 4;
125 return MIN2(available_ram
, available_gtt
);
129 anv_physical_device_init_heaps(struct anv_physical_device
*device
, int fd
)
131 if (anv_gem_get_context_param(fd
, 0, I915_CONTEXT_PARAM_GTT_SIZE
,
132 &device
->gtt_size
) == -1) {
133 /* If, for whatever reason, we can't actually get the GTT size from the
134 * kernel (too old?) fall back to the aperture size.
136 anv_perf_warn(NULL
, NULL
,
137 "Failed to get I915_CONTEXT_PARAM_GTT_SIZE: %m");
139 if (gen_get_aperture_size(fd
, &device
->gtt_size
) == -1) {
140 return vk_errorfi(device
->instance
, NULL
,
141 VK_ERROR_INITIALIZATION_FAILED
,
142 "failed to get aperture size: %m");
146 /* We only allow 48-bit addresses with softpin because knowing the actual
147 * address is required for the vertex cache flush workaround.
149 device
->supports_48bit_addresses
= (device
->info
.gen
>= 8) &&
150 device
->has_softpin
&&
151 device
->gtt_size
> (4ULL << 30 /* GiB */);
153 uint64_t heap_size
= anv_compute_heap_size(fd
, device
->gtt_size
);
155 if (heap_size
> (2ull << 30) && !device
->supports_48bit_addresses
) {
156 /* When running with an overridden PCI ID, we may get a GTT size from
157 * the kernel that is greater than 2 GiB but the execbuf check for 48bit
158 * address support can still fail. Just clamp the address space size to
159 * 2 GiB if we don't have 48-bit support.
161 intel_logw("%s:%d: The kernel reported a GTT size larger than 2 GiB but "
162 "not support for 48-bit addresses",
164 heap_size
= 2ull << 30;
167 device
->memory
.heap_count
= 1;
168 device
->memory
.heaps
[0] = (struct anv_memory_heap
) {
170 .flags
= VK_MEMORY_HEAP_DEVICE_LOCAL_BIT
,
173 uint32_t type_count
= 0;
174 for (uint32_t heap
= 0; heap
< device
->memory
.heap_count
; heap
++) {
175 if (device
->info
.has_llc
) {
176 /* Big core GPUs share LLC with the CPU and thus one memory type can be
177 * both cached and coherent at the same time.
179 device
->memory
.types
[type_count
++] = (struct anv_memory_type
) {
180 .propertyFlags
= VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT
|
181 VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT
|
182 VK_MEMORY_PROPERTY_HOST_COHERENT_BIT
|
183 VK_MEMORY_PROPERTY_HOST_CACHED_BIT
,
187 /* The spec requires that we expose a host-visible, coherent memory
188 * type, but Atom GPUs don't share LLC. Thus we offer two memory types
189 * to give the application a choice between cached, but not coherent and
190 * coherent but uncached (WC though).
192 device
->memory
.types
[type_count
++] = (struct anv_memory_type
) {
193 .propertyFlags
= VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT
|
194 VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT
|
195 VK_MEMORY_PROPERTY_HOST_COHERENT_BIT
,
198 device
->memory
.types
[type_count
++] = (struct anv_memory_type
) {
199 .propertyFlags
= VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT
|
200 VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT
|
201 VK_MEMORY_PROPERTY_HOST_CACHED_BIT
,
206 device
->memory
.type_count
= type_count
;
212 anv_physical_device_init_uuids(struct anv_physical_device
*device
)
214 const struct build_id_note
*note
=
215 build_id_find_nhdr_for_addr(anv_physical_device_init_uuids
);
217 return vk_errorfi(device
->instance
, NULL
,
218 VK_ERROR_INITIALIZATION_FAILED
,
219 "Failed to find build-id");
222 unsigned build_id_len
= build_id_length(note
);
223 if (build_id_len
< 20) {
224 return vk_errorfi(device
->instance
, NULL
,
225 VK_ERROR_INITIALIZATION_FAILED
,
226 "build-id too short. It needs to be a SHA");
229 memcpy(device
->driver_build_sha1
, build_id_data(note
), 20);
231 struct mesa_sha1 sha1_ctx
;
233 STATIC_ASSERT(VK_UUID_SIZE
<= sizeof(sha1
));
235 /* The pipeline cache UUID is used for determining when a pipeline cache is
236 * invalid. It needs both a driver build and the PCI ID of the device.
238 _mesa_sha1_init(&sha1_ctx
);
239 _mesa_sha1_update(&sha1_ctx
, build_id_data(note
), build_id_len
);
240 _mesa_sha1_update(&sha1_ctx
, &device
->info
.chipset_id
,
241 sizeof(device
->info
.chipset_id
));
242 _mesa_sha1_update(&sha1_ctx
, &device
->always_use_bindless
,
243 sizeof(device
->always_use_bindless
));
244 _mesa_sha1_update(&sha1_ctx
, &device
->has_a64_buffer_access
,
245 sizeof(device
->has_a64_buffer_access
));
246 _mesa_sha1_update(&sha1_ctx
, &device
->has_bindless_images
,
247 sizeof(device
->has_bindless_images
));
248 _mesa_sha1_update(&sha1_ctx
, &device
->has_bindless_samplers
,
249 sizeof(device
->has_bindless_samplers
));
250 _mesa_sha1_final(&sha1_ctx
, sha1
);
251 memcpy(device
->pipeline_cache_uuid
, sha1
, VK_UUID_SIZE
);
253 /* The driver UUID is used for determining sharability of images and memory
254 * between two Vulkan instances in separate processes. People who want to
255 * share memory need to also check the device UUID (below) so all this
256 * needs to be is the build-id.
258 memcpy(device
->driver_uuid
, build_id_data(note
), VK_UUID_SIZE
);
260 /* The device UUID uniquely identifies the given device within the machine.
261 * Since we never have more than one device, this doesn't need to be a real
262 * UUID. However, on the off-chance that someone tries to use this to
263 * cache pre-tiled images or something of the like, we use the PCI ID and
264 * some bits of ISL info to ensure that this is safe.
266 _mesa_sha1_init(&sha1_ctx
);
267 _mesa_sha1_update(&sha1_ctx
, &device
->info
.chipset_id
,
268 sizeof(device
->info
.chipset_id
));
269 _mesa_sha1_update(&sha1_ctx
, &device
->isl_dev
.has_bit6_swizzling
,
270 sizeof(device
->isl_dev
.has_bit6_swizzling
));
271 _mesa_sha1_final(&sha1_ctx
, sha1
);
272 memcpy(device
->device_uuid
, sha1
, VK_UUID_SIZE
);
278 anv_physical_device_init_disk_cache(struct anv_physical_device
*device
)
280 #ifdef ENABLE_SHADER_CACHE
282 ASSERTED
int len
= snprintf(renderer
, sizeof(renderer
), "anv_%04x",
283 device
->info
.chipset_id
);
284 assert(len
== sizeof(renderer
) - 2);
287 _mesa_sha1_format(timestamp
, device
->driver_build_sha1
);
289 const uint64_t driver_flags
=
290 brw_get_compiler_config_value(device
->compiler
);
291 device
->disk_cache
= disk_cache_create(renderer
, timestamp
, driver_flags
);
293 device
->disk_cache
= NULL
;
298 anv_physical_device_free_disk_cache(struct anv_physical_device
*device
)
300 #ifdef ENABLE_SHADER_CACHE
301 if (device
->disk_cache
)
302 disk_cache_destroy(device
->disk_cache
);
304 assert(device
->disk_cache
== NULL
);
309 get_available_system_memory()
311 char *meminfo
= os_read_file("/proc/meminfo", NULL
);
315 char *str
= strstr(meminfo
, "MemAvailable:");
321 uint64_t kb_mem_available
;
322 if (sscanf(str
, "MemAvailable: %" PRIx64
, &kb_mem_available
) == 1) {
324 return kb_mem_available
<< 10;
332 anv_physical_device_try_create(struct anv_instance
*instance
,
333 drmDevicePtr drm_device
,
334 struct anv_physical_device
**device_out
)
336 const char *primary_path
= drm_device
->nodes
[DRM_NODE_PRIMARY
];
337 const char *path
= drm_device
->nodes
[DRM_NODE_RENDER
];
342 brw_process_intel_debug_variable();
344 fd
= open(path
, O_RDWR
| O_CLOEXEC
);
346 return vk_error(VK_ERROR_INCOMPATIBLE_DRIVER
);
348 struct gen_device_info devinfo
;
349 if (!gen_get_device_info_from_fd(fd
, &devinfo
)) {
350 result
= vk_error(VK_ERROR_INCOMPATIBLE_DRIVER
);
354 const char *device_name
= gen_get_device_name(devinfo
.chipset_id
);
356 if (devinfo
.is_haswell
) {
357 intel_logw("Haswell Vulkan support is incomplete");
358 } else if (devinfo
.gen
== 7 && !devinfo
.is_baytrail
) {
359 intel_logw("Ivy Bridge Vulkan support is incomplete");
360 } else if (devinfo
.gen
== 7 && devinfo
.is_baytrail
) {
361 intel_logw("Bay Trail Vulkan support is incomplete");
362 } else if (devinfo
.gen
>= 8 && devinfo
.gen
<= 11) {
363 /* Gen8-11 fully supported */
364 } else if (devinfo
.gen
== 12) {
365 intel_logw("Vulkan is not yet fully supported on gen12");
367 result
= vk_errorfi(instance
, NULL
, VK_ERROR_INCOMPATIBLE_DRIVER
,
368 "Vulkan not yet supported on %s", device_name
);
372 struct anv_physical_device
*device
=
373 vk_alloc(&instance
->alloc
, sizeof(*device
), 8,
374 VK_SYSTEM_ALLOCATION_SCOPE_INSTANCE
);
375 if (device
== NULL
) {
376 result
= vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
380 vk_object_base_init(NULL
, &device
->base
, VK_OBJECT_TYPE_PHYSICAL_DEVICE
);
381 device
->instance
= instance
;
383 assert(strlen(path
) < ARRAY_SIZE(device
->path
));
384 snprintf(device
->path
, ARRAY_SIZE(device
->path
), "%s", path
);
386 device
->info
= devinfo
;
387 device
->name
= device_name
;
389 device
->no_hw
= device
->info
.no_hw
;
390 if (getenv("INTEL_NO_HW") != NULL
)
391 device
->no_hw
= true;
393 device
->pci_info
.domain
= drm_device
->businfo
.pci
->domain
;
394 device
->pci_info
.bus
= drm_device
->businfo
.pci
->bus
;
395 device
->pci_info
.device
= drm_device
->businfo
.pci
->dev
;
396 device
->pci_info
.function
= drm_device
->businfo
.pci
->func
;
398 device
->cmd_parser_version
= -1;
399 if (device
->info
.gen
== 7) {
400 device
->cmd_parser_version
=
401 anv_gem_get_param(fd
, I915_PARAM_CMD_PARSER_VERSION
);
402 if (device
->cmd_parser_version
== -1) {
403 result
= vk_errorfi(device
->instance
, NULL
,
404 VK_ERROR_INITIALIZATION_FAILED
,
405 "failed to get command parser version");
410 if (!anv_gem_get_param(fd
, I915_PARAM_HAS_WAIT_TIMEOUT
)) {
411 result
= vk_errorfi(device
->instance
, NULL
,
412 VK_ERROR_INITIALIZATION_FAILED
,
413 "kernel missing gem wait");
417 if (!anv_gem_get_param(fd
, I915_PARAM_HAS_EXECBUF2
)) {
418 result
= vk_errorfi(device
->instance
, NULL
,
419 VK_ERROR_INITIALIZATION_FAILED
,
420 "kernel missing execbuf2");
424 if (!device
->info
.has_llc
&&
425 anv_gem_get_param(fd
, I915_PARAM_MMAP_VERSION
) < 1) {
426 result
= vk_errorfi(device
->instance
, NULL
,
427 VK_ERROR_INITIALIZATION_FAILED
,
428 "kernel missing wc mmap");
432 device
->has_softpin
= anv_gem_get_param(fd
, I915_PARAM_HAS_EXEC_SOFTPIN
);
433 device
->has_exec_async
= anv_gem_get_param(fd
, I915_PARAM_HAS_EXEC_ASYNC
);
434 device
->has_exec_capture
= anv_gem_get_param(fd
, I915_PARAM_HAS_EXEC_CAPTURE
);
435 device
->has_exec_fence
= anv_gem_get_param(fd
, I915_PARAM_HAS_EXEC_FENCE
);
436 device
->has_syncobj
= anv_gem_get_param(fd
, I915_PARAM_HAS_EXEC_FENCE_ARRAY
);
437 device
->has_syncobj_wait
= device
->has_syncobj
&&
438 anv_gem_supports_syncobj_wait(fd
);
439 device
->has_context_priority
= anv_gem_has_context_priority(fd
);
441 result
= anv_physical_device_init_heaps(device
, fd
);
442 if (result
!= VK_SUCCESS
)
445 device
->use_softpin
= device
->has_softpin
&&
446 device
->supports_48bit_addresses
;
448 device
->has_context_isolation
=
449 anv_gem_get_param(fd
, I915_PARAM_HAS_CONTEXT_ISOLATION
);
451 device
->always_use_bindless
=
452 env_var_as_boolean("ANV_ALWAYS_BINDLESS", false);
454 /* We first got the A64 messages on broadwell and we can only use them if
455 * we can pass addresses directly into the shader which requires softpin.
457 device
->has_a64_buffer_access
= device
->info
.gen
>= 8 &&
460 /* We first get bindless image access on Skylake and we can only really do
461 * it if we don't have any relocations so we need softpin.
463 device
->has_bindless_images
= device
->info
.gen
>= 9 &&
466 /* We've had bindless samplers since Ivy Bridge (forever in Vulkan terms)
467 * because it's just a matter of setting the sampler address in the sample
468 * message header. However, we've not bothered to wire it up for vec4 so
469 * we leave it disabled on gen7.
471 device
->has_bindless_samplers
= device
->info
.gen
>= 8;
473 device
->has_implicit_ccs
= device
->info
.has_aux_map
;
475 device
->has_mem_available
= get_available_system_memory() != 0;
477 device
->always_flush_cache
=
478 driQueryOptionb(&instance
->dri_options
, "always_flush_cache");
480 device
->has_mmap_offset
=
481 anv_gem_get_param(fd
, I915_PARAM_MMAP_GTT_VERSION
) >= 4;
483 /* GENs prior to 8 do not support EU/Subslice info */
484 if (device
->info
.gen
>= 8) {
485 device
->subslice_total
= anv_gem_get_param(fd
, I915_PARAM_SUBSLICE_TOTAL
);
486 device
->eu_total
= anv_gem_get_param(fd
, I915_PARAM_EU_TOTAL
);
488 /* Without this information, we cannot get the right Braswell
489 * brandstrings, and we have to use conservative numbers for GPGPU on
490 * many platforms, but otherwise, things will just work.
492 if (device
->subslice_total
< 1 || device
->eu_total
< 1) {
493 intel_logw("Kernel 4.1 required to properly query GPU properties");
495 } else if (device
->info
.gen
== 7) {
496 device
->subslice_total
= 1 << (device
->info
.gt
- 1);
499 if (device
->info
.is_cherryview
&&
500 device
->subslice_total
> 0 && device
->eu_total
> 0) {
501 /* Logical CS threads = EUs per subslice * num threads per EU */
502 uint32_t max_cs_threads
=
503 device
->eu_total
/ device
->subslice_total
* device
->info
.num_thread_per_eu
;
505 /* Fuse configurations may give more threads than expected, never less. */
506 if (max_cs_threads
> device
->info
.max_cs_threads
)
507 device
->info
.max_cs_threads
= max_cs_threads
;
510 device
->compiler
= brw_compiler_create(NULL
, &device
->info
);
511 if (device
->compiler
== NULL
) {
512 result
= vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
515 device
->compiler
->shader_debug_log
= compiler_debug_log
;
516 device
->compiler
->shader_perf_log
= compiler_perf_log
;
517 device
->compiler
->supports_pull_constants
= false;
518 device
->compiler
->constant_buffer_0_is_relative
=
519 device
->info
.gen
< 8 || !device
->has_context_isolation
;
520 device
->compiler
->supports_shader_constants
= true;
521 device
->compiler
->compact_params
= false;
523 /* Broadwell PRM says:
525 * "Before Gen8, there was a historical configuration control field to
526 * swizzle address bit[6] for in X/Y tiling modes. This was set in three
527 * different places: TILECTL[1:0], ARB_MODE[5:4], and
528 * DISP_ARB_CTL[14:13].
530 * For Gen8 and subsequent generations, the swizzle fields are all
531 * reserved, and the CPU's memory controller performs all address
532 * swizzling modifications."
535 device
->info
.gen
< 8 && anv_gem_get_bit6_swizzle(fd
, I915_TILING_X
);
537 isl_device_init(&device
->isl_dev
, &device
->info
, swizzled
);
539 result
= anv_physical_device_init_uuids(device
);
540 if (result
!= VK_SUCCESS
)
543 anv_physical_device_init_disk_cache(device
);
545 if (instance
->enabled_extensions
.KHR_display
) {
546 master_fd
= open(primary_path
, O_RDWR
| O_CLOEXEC
);
547 if (master_fd
>= 0) {
548 /* prod the device with a GETPARAM call which will fail if
549 * we don't have permission to even render on this device
551 if (anv_gem_get_param(master_fd
, I915_PARAM_CHIPSET_ID
) == 0) {
557 device
->master_fd
= master_fd
;
559 result
= anv_init_wsi(device
);
560 if (result
!= VK_SUCCESS
)
561 goto fail_disk_cache
;
563 device
->perf
= anv_get_perf(&device
->info
, fd
);
565 anv_physical_device_get_supported_extensions(device
,
566 &device
->supported_extensions
);
569 device
->local_fd
= fd
;
571 *device_out
= device
;
576 anv_physical_device_free_disk_cache(device
);
578 ralloc_free(device
->compiler
);
580 vk_free(&instance
->alloc
, device
);
589 anv_physical_device_destroy(struct anv_physical_device
*device
)
591 anv_finish_wsi(device
);
592 anv_physical_device_free_disk_cache(device
);
593 ralloc_free(device
->compiler
);
594 ralloc_free(device
->perf
);
595 close(device
->local_fd
);
596 if (device
->master_fd
>= 0)
597 close(device
->master_fd
);
598 vk_object_base_finish(&device
->base
);
599 vk_free(&device
->instance
->alloc
, device
);
603 default_alloc_func(void *pUserData
, size_t size
, size_t align
,
604 VkSystemAllocationScope allocationScope
)
610 default_realloc_func(void *pUserData
, void *pOriginal
, size_t size
,
611 size_t align
, VkSystemAllocationScope allocationScope
)
613 return realloc(pOriginal
, size
);
617 default_free_func(void *pUserData
, void *pMemory
)
622 static const VkAllocationCallbacks default_alloc
= {
624 .pfnAllocation
= default_alloc_func
,
625 .pfnReallocation
= default_realloc_func
,
626 .pfnFree
= default_free_func
,
629 VkResult
anv_EnumerateInstanceExtensionProperties(
630 const char* pLayerName
,
631 uint32_t* pPropertyCount
,
632 VkExtensionProperties
* pProperties
)
634 VK_OUTARRAY_MAKE(out
, pProperties
, pPropertyCount
);
636 for (int i
= 0; i
< ANV_INSTANCE_EXTENSION_COUNT
; i
++) {
637 if (anv_instance_extensions_supported
.extensions
[i
]) {
638 vk_outarray_append(&out
, prop
) {
639 *prop
= anv_instance_extensions
[i
];
644 return vk_outarray_status(&out
);
647 VkResult
anv_CreateInstance(
648 const VkInstanceCreateInfo
* pCreateInfo
,
649 const VkAllocationCallbacks
* pAllocator
,
650 VkInstance
* pInstance
)
652 struct anv_instance
*instance
;
655 assert(pCreateInfo
->sType
== VK_STRUCTURE_TYPE_INSTANCE_CREATE_INFO
);
657 struct anv_instance_extension_table enabled_extensions
= {};
658 for (uint32_t i
= 0; i
< pCreateInfo
->enabledExtensionCount
; i
++) {
660 for (idx
= 0; idx
< ANV_INSTANCE_EXTENSION_COUNT
; idx
++) {
661 if (strcmp(pCreateInfo
->ppEnabledExtensionNames
[i
],
662 anv_instance_extensions
[idx
].extensionName
) == 0)
666 if (idx
>= ANV_INSTANCE_EXTENSION_COUNT
)
667 return vk_error(VK_ERROR_EXTENSION_NOT_PRESENT
);
669 if (!anv_instance_extensions_supported
.extensions
[idx
])
670 return vk_error(VK_ERROR_EXTENSION_NOT_PRESENT
);
672 enabled_extensions
.extensions
[idx
] = true;
675 instance
= vk_alloc2(&default_alloc
, pAllocator
, sizeof(*instance
), 8,
676 VK_SYSTEM_ALLOCATION_SCOPE_INSTANCE
);
678 return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
680 vk_object_base_init(NULL
, &instance
->base
, VK_OBJECT_TYPE_INSTANCE
);
683 instance
->alloc
= *pAllocator
;
685 instance
->alloc
= default_alloc
;
687 instance
->app_info
= (struct anv_app_info
) { .api_version
= 0 };
688 if (pCreateInfo
->pApplicationInfo
) {
689 const VkApplicationInfo
*app
= pCreateInfo
->pApplicationInfo
;
691 instance
->app_info
.app_name
=
692 vk_strdup(&instance
->alloc
, app
->pApplicationName
,
693 VK_SYSTEM_ALLOCATION_SCOPE_INSTANCE
);
694 instance
->app_info
.app_version
= app
->applicationVersion
;
696 instance
->app_info
.engine_name
=
697 vk_strdup(&instance
->alloc
, app
->pEngineName
,
698 VK_SYSTEM_ALLOCATION_SCOPE_INSTANCE
);
699 instance
->app_info
.engine_version
= app
->engineVersion
;
701 instance
->app_info
.api_version
= app
->apiVersion
;
704 if (instance
->app_info
.api_version
== 0)
705 instance
->app_info
.api_version
= VK_API_VERSION_1_0
;
707 instance
->enabled_extensions
= enabled_extensions
;
709 for (unsigned i
= 0; i
< ARRAY_SIZE(instance
->dispatch
.entrypoints
); i
++) {
710 /* Vulkan requires that entrypoints for extensions which have not been
711 * enabled must not be advertised.
713 if (!anv_instance_entrypoint_is_enabled(i
, instance
->app_info
.api_version
,
714 &instance
->enabled_extensions
)) {
715 instance
->dispatch
.entrypoints
[i
] = NULL
;
717 instance
->dispatch
.entrypoints
[i
] =
718 anv_instance_dispatch_table
.entrypoints
[i
];
722 for (unsigned i
= 0; i
< ARRAY_SIZE(instance
->physical_device_dispatch
.entrypoints
); i
++) {
723 /* Vulkan requires that entrypoints for extensions which have not been
724 * enabled must not be advertised.
726 if (!anv_physical_device_entrypoint_is_enabled(i
, instance
->app_info
.api_version
,
727 &instance
->enabled_extensions
)) {
728 instance
->physical_device_dispatch
.entrypoints
[i
] = NULL
;
730 instance
->physical_device_dispatch
.entrypoints
[i
] =
731 anv_physical_device_dispatch_table
.entrypoints
[i
];
735 for (unsigned i
= 0; i
< ARRAY_SIZE(instance
->device_dispatch
.entrypoints
); i
++) {
736 /* Vulkan requires that entrypoints for extensions which have not been
737 * enabled must not be advertised.
739 if (!anv_device_entrypoint_is_enabled(i
, instance
->app_info
.api_version
,
740 &instance
->enabled_extensions
, NULL
)) {
741 instance
->device_dispatch
.entrypoints
[i
] = NULL
;
743 instance
->device_dispatch
.entrypoints
[i
] =
744 anv_device_dispatch_table
.entrypoints
[i
];
748 instance
->physical_devices_enumerated
= false;
749 list_inithead(&instance
->physical_devices
);
751 result
= vk_debug_report_instance_init(&instance
->debug_report_callbacks
);
752 if (result
!= VK_SUCCESS
) {
753 vk_free2(&default_alloc
, pAllocator
, instance
);
754 return vk_error(result
);
757 instance
->pipeline_cache_enabled
=
758 env_var_as_boolean("ANV_ENABLE_PIPELINE_CACHE", true);
760 glsl_type_singleton_init_or_ref();
762 VG(VALGRIND_CREATE_MEMPOOL(instance
, 0, false));
764 driParseOptionInfo(&instance
->available_dri_options
, anv_dri_options_xml
);
765 driParseConfigFiles(&instance
->dri_options
, &instance
->available_dri_options
,
767 instance
->app_info
.engine_name
,
768 instance
->app_info
.engine_version
);
770 *pInstance
= anv_instance_to_handle(instance
);
775 void anv_DestroyInstance(
776 VkInstance _instance
,
777 const VkAllocationCallbacks
* pAllocator
)
779 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
784 list_for_each_entry_safe(struct anv_physical_device
, pdevice
,
785 &instance
->physical_devices
, link
)
786 anv_physical_device_destroy(pdevice
);
788 vk_free(&instance
->alloc
, (char *)instance
->app_info
.app_name
);
789 vk_free(&instance
->alloc
, (char *)instance
->app_info
.engine_name
);
791 VG(VALGRIND_DESTROY_MEMPOOL(instance
));
793 vk_debug_report_instance_destroy(&instance
->debug_report_callbacks
);
795 glsl_type_singleton_decref();
797 driDestroyOptionCache(&instance
->dri_options
);
798 driDestroyOptionInfo(&instance
->available_dri_options
);
800 vk_object_base_finish(&instance
->base
);
801 vk_free(&instance
->alloc
, instance
);
805 anv_enumerate_physical_devices(struct anv_instance
*instance
)
807 if (instance
->physical_devices_enumerated
)
810 instance
->physical_devices_enumerated
= true;
812 /* TODO: Check for more devices ? */
813 drmDevicePtr devices
[8];
816 max_devices
= drmGetDevices2(0, devices
, ARRAY_SIZE(devices
));
820 VkResult result
= VK_SUCCESS
;
821 for (unsigned i
= 0; i
< (unsigned)max_devices
; i
++) {
822 if (devices
[i
]->available_nodes
& 1 << DRM_NODE_RENDER
&&
823 devices
[i
]->bustype
== DRM_BUS_PCI
&&
824 devices
[i
]->deviceinfo
.pci
->vendor_id
== 0x8086) {
826 struct anv_physical_device
*pdevice
;
827 result
= anv_physical_device_try_create(instance
, devices
[i
],
829 /* Incompatible DRM device, skip. */
830 if (result
== VK_ERROR_INCOMPATIBLE_DRIVER
) {
835 /* Error creating the physical device, report the error. */
836 if (result
!= VK_SUCCESS
)
839 list_addtail(&pdevice
->link
, &instance
->physical_devices
);
842 drmFreeDevices(devices
, max_devices
);
844 /* If we successfully enumerated any devices, call it success */
848 VkResult
anv_EnumeratePhysicalDevices(
849 VkInstance _instance
,
850 uint32_t* pPhysicalDeviceCount
,
851 VkPhysicalDevice
* pPhysicalDevices
)
853 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
854 VK_OUTARRAY_MAKE(out
, pPhysicalDevices
, pPhysicalDeviceCount
);
856 VkResult result
= anv_enumerate_physical_devices(instance
);
857 if (result
!= VK_SUCCESS
)
860 list_for_each_entry(struct anv_physical_device
, pdevice
,
861 &instance
->physical_devices
, link
) {
862 vk_outarray_append(&out
, i
) {
863 *i
= anv_physical_device_to_handle(pdevice
);
867 return vk_outarray_status(&out
);
870 VkResult
anv_EnumeratePhysicalDeviceGroups(
871 VkInstance _instance
,
872 uint32_t* pPhysicalDeviceGroupCount
,
873 VkPhysicalDeviceGroupProperties
* pPhysicalDeviceGroupProperties
)
875 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
876 VK_OUTARRAY_MAKE(out
, pPhysicalDeviceGroupProperties
,
877 pPhysicalDeviceGroupCount
);
879 VkResult result
= anv_enumerate_physical_devices(instance
);
880 if (result
!= VK_SUCCESS
)
883 list_for_each_entry(struct anv_physical_device
, pdevice
,
884 &instance
->physical_devices
, link
) {
885 vk_outarray_append(&out
, p
) {
886 p
->physicalDeviceCount
= 1;
887 memset(p
->physicalDevices
, 0, sizeof(p
->physicalDevices
));
888 p
->physicalDevices
[0] = anv_physical_device_to_handle(pdevice
);
889 p
->subsetAllocation
= false;
891 vk_foreach_struct(ext
, p
->pNext
)
892 anv_debug_ignored_stype(ext
->sType
);
896 return vk_outarray_status(&out
);
899 void anv_GetPhysicalDeviceFeatures(
900 VkPhysicalDevice physicalDevice
,
901 VkPhysicalDeviceFeatures
* pFeatures
)
903 ANV_FROM_HANDLE(anv_physical_device
, pdevice
, physicalDevice
);
905 *pFeatures
= (VkPhysicalDeviceFeatures
) {
906 .robustBufferAccess
= true,
907 .fullDrawIndexUint32
= true,
908 .imageCubeArray
= true,
909 .independentBlend
= true,
910 .geometryShader
= true,
911 .tessellationShader
= true,
912 .sampleRateShading
= true,
913 .dualSrcBlend
= true,
915 .multiDrawIndirect
= true,
916 .drawIndirectFirstInstance
= true,
918 .depthBiasClamp
= true,
919 .fillModeNonSolid
= true,
920 .depthBounds
= pdevice
->info
.gen
>= 12,
924 .multiViewport
= true,
925 .samplerAnisotropy
= true,
926 .textureCompressionETC2
= pdevice
->info
.gen
>= 8 ||
927 pdevice
->info
.is_baytrail
,
928 .textureCompressionASTC_LDR
= pdevice
->info
.gen
>= 9, /* FINISHME CHV */
929 .textureCompressionBC
= true,
930 .occlusionQueryPrecise
= true,
931 .pipelineStatisticsQuery
= true,
932 .fragmentStoresAndAtomics
= true,
933 .shaderTessellationAndGeometryPointSize
= true,
934 .shaderImageGatherExtended
= true,
935 .shaderStorageImageExtendedFormats
= true,
936 .shaderStorageImageMultisample
= false,
937 .shaderStorageImageReadWithoutFormat
= false,
938 .shaderStorageImageWriteWithoutFormat
= true,
939 .shaderUniformBufferArrayDynamicIndexing
= true,
940 .shaderSampledImageArrayDynamicIndexing
= true,
941 .shaderStorageBufferArrayDynamicIndexing
= true,
942 .shaderStorageImageArrayDynamicIndexing
= true,
943 .shaderClipDistance
= true,
944 .shaderCullDistance
= true,
945 .shaderFloat64
= pdevice
->info
.gen
>= 8 &&
946 pdevice
->info
.has_64bit_float
,
947 .shaderInt64
= pdevice
->info
.gen
>= 8 &&
948 pdevice
->info
.has_64bit_int
,
949 .shaderInt16
= pdevice
->info
.gen
>= 8,
950 .shaderResourceMinLod
= pdevice
->info
.gen
>= 9,
951 .variableMultisampleRate
= true,
952 .inheritedQueries
= true,
955 /* We can't do image stores in vec4 shaders */
956 pFeatures
->vertexPipelineStoresAndAtomics
=
957 pdevice
->compiler
->scalar_stage
[MESA_SHADER_VERTEX
] &&
958 pdevice
->compiler
->scalar_stage
[MESA_SHADER_GEOMETRY
];
960 struct anv_app_info
*app_info
= &pdevice
->instance
->app_info
;
962 /* The new DOOM and Wolfenstein games require depthBounds without
963 * checking for it. They seem to run fine without it so just claim it's
964 * there and accept the consequences.
966 if (app_info
->engine_name
&& strcmp(app_info
->engine_name
, "idTech") == 0)
967 pFeatures
->depthBounds
= true;
971 anv_get_physical_device_features_1_1(struct anv_physical_device
*pdevice
,
972 VkPhysicalDeviceVulkan11Features
*f
)
974 assert(f
->sType
== VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_1_FEATURES
);
976 f
->storageBuffer16BitAccess
= pdevice
->info
.gen
>= 8;
977 f
->uniformAndStorageBuffer16BitAccess
= pdevice
->info
.gen
>= 8;
978 f
->storagePushConstant16
= pdevice
->info
.gen
>= 8;
979 f
->storageInputOutput16
= false;
981 f
->multiviewGeometryShader
= true;
982 f
->multiviewTessellationShader
= true;
983 f
->variablePointersStorageBuffer
= true;
984 f
->variablePointers
= true;
985 f
->protectedMemory
= false;
986 f
->samplerYcbcrConversion
= true;
987 f
->shaderDrawParameters
= true;
991 anv_get_physical_device_features_1_2(struct anv_physical_device
*pdevice
,
992 VkPhysicalDeviceVulkan12Features
*f
)
994 assert(f
->sType
== VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_FEATURES
);
996 f
->samplerMirrorClampToEdge
= true;
997 f
->drawIndirectCount
= true;
998 f
->storageBuffer8BitAccess
= pdevice
->info
.gen
>= 8;
999 f
->uniformAndStorageBuffer8BitAccess
= pdevice
->info
.gen
>= 8;
1000 f
->storagePushConstant8
= pdevice
->info
.gen
>= 8;
1001 f
->shaderBufferInt64Atomics
= pdevice
->info
.gen
>= 9 &&
1002 pdevice
->use_softpin
;
1003 f
->shaderSharedInt64Atomics
= false;
1004 f
->shaderFloat16
= pdevice
->info
.gen
>= 8;
1005 f
->shaderInt8
= pdevice
->info
.gen
>= 8;
1007 bool descIndexing
= pdevice
->has_a64_buffer_access
&&
1008 pdevice
->has_bindless_images
;
1009 f
->descriptorIndexing
= descIndexing
;
1010 f
->shaderInputAttachmentArrayDynamicIndexing
= false;
1011 f
->shaderUniformTexelBufferArrayDynamicIndexing
= descIndexing
;
1012 f
->shaderStorageTexelBufferArrayDynamicIndexing
= descIndexing
;
1013 f
->shaderUniformBufferArrayNonUniformIndexing
= false;
1014 f
->shaderSampledImageArrayNonUniformIndexing
= descIndexing
;
1015 f
->shaderStorageBufferArrayNonUniformIndexing
= descIndexing
;
1016 f
->shaderStorageImageArrayNonUniformIndexing
= descIndexing
;
1017 f
->shaderInputAttachmentArrayNonUniformIndexing
= false;
1018 f
->shaderUniformTexelBufferArrayNonUniformIndexing
= descIndexing
;
1019 f
->shaderStorageTexelBufferArrayNonUniformIndexing
= descIndexing
;
1020 f
->descriptorBindingUniformBufferUpdateAfterBind
= false;
1021 f
->descriptorBindingSampledImageUpdateAfterBind
= descIndexing
;
1022 f
->descriptorBindingStorageImageUpdateAfterBind
= descIndexing
;
1023 f
->descriptorBindingStorageBufferUpdateAfterBind
= descIndexing
;
1024 f
->descriptorBindingUniformTexelBufferUpdateAfterBind
= descIndexing
;
1025 f
->descriptorBindingStorageTexelBufferUpdateAfterBind
= descIndexing
;
1026 f
->descriptorBindingUpdateUnusedWhilePending
= descIndexing
;
1027 f
->descriptorBindingPartiallyBound
= descIndexing
;
1028 f
->descriptorBindingVariableDescriptorCount
= false;
1029 f
->runtimeDescriptorArray
= descIndexing
;
1031 f
->samplerFilterMinmax
= pdevice
->info
.gen
>= 9;
1032 f
->scalarBlockLayout
= true;
1033 f
->imagelessFramebuffer
= true;
1034 f
->uniformBufferStandardLayout
= true;
1035 f
->shaderSubgroupExtendedTypes
= true;
1036 f
->separateDepthStencilLayouts
= true;
1037 f
->hostQueryReset
= true;
1038 f
->timelineSemaphore
= true;
1039 f
->bufferDeviceAddress
= pdevice
->has_a64_buffer_access
;
1040 f
->bufferDeviceAddressCaptureReplay
= pdevice
->has_a64_buffer_access
;
1041 f
->bufferDeviceAddressMultiDevice
= false;
1042 f
->vulkanMemoryModel
= true;
1043 f
->vulkanMemoryModelDeviceScope
= true;
1044 f
->vulkanMemoryModelAvailabilityVisibilityChains
= true;
1045 f
->shaderOutputViewportIndex
= true;
1046 f
->shaderOutputLayer
= true;
1047 f
->subgroupBroadcastDynamicId
= true;
1050 void anv_GetPhysicalDeviceFeatures2(
1051 VkPhysicalDevice physicalDevice
,
1052 VkPhysicalDeviceFeatures2
* pFeatures
)
1054 ANV_FROM_HANDLE(anv_physical_device
, pdevice
, physicalDevice
);
1055 anv_GetPhysicalDeviceFeatures(physicalDevice
, &pFeatures
->features
);
1057 VkPhysicalDeviceVulkan11Features core_1_1
= {
1058 .sType
= VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_1_FEATURES
,
1060 anv_get_physical_device_features_1_1(pdevice
, &core_1_1
);
1062 VkPhysicalDeviceVulkan12Features core_1_2
= {
1063 .sType
= VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_FEATURES
,
1065 anv_get_physical_device_features_1_2(pdevice
, &core_1_2
);
1067 #define CORE_FEATURE(major, minor, feature) \
1068 features->feature = core_##major##_##minor.feature
1071 vk_foreach_struct(ext
, pFeatures
->pNext
) {
1072 switch (ext
->sType
) {
1073 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_8BIT_STORAGE_FEATURES_KHR
: {
1074 VkPhysicalDevice8BitStorageFeaturesKHR
*features
=
1075 (VkPhysicalDevice8BitStorageFeaturesKHR
*)ext
;
1076 CORE_FEATURE(1, 2, storageBuffer8BitAccess
);
1077 CORE_FEATURE(1, 2, uniformAndStorageBuffer8BitAccess
);
1078 CORE_FEATURE(1, 2, storagePushConstant8
);
1082 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_16BIT_STORAGE_FEATURES
: {
1083 VkPhysicalDevice16BitStorageFeatures
*features
=
1084 (VkPhysicalDevice16BitStorageFeatures
*)ext
;
1085 CORE_FEATURE(1, 1, storageBuffer16BitAccess
);
1086 CORE_FEATURE(1, 1, uniformAndStorageBuffer16BitAccess
);
1087 CORE_FEATURE(1, 1, storagePushConstant16
);
1088 CORE_FEATURE(1, 1, storageInputOutput16
);
1092 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_BUFFER_DEVICE_ADDRESS_FEATURES_EXT
: {
1093 VkPhysicalDeviceBufferDeviceAddressFeaturesEXT
*features
= (void *)ext
;
1094 features
->bufferDeviceAddress
= pdevice
->has_a64_buffer_access
;
1095 features
->bufferDeviceAddressCaptureReplay
= false;
1096 features
->bufferDeviceAddressMultiDevice
= false;
1100 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_BUFFER_DEVICE_ADDRESS_FEATURES_KHR
: {
1101 VkPhysicalDeviceBufferDeviceAddressFeaturesKHR
*features
= (void *)ext
;
1102 CORE_FEATURE(1, 2, bufferDeviceAddress
);
1103 CORE_FEATURE(1, 2, bufferDeviceAddressCaptureReplay
);
1104 CORE_FEATURE(1, 2, bufferDeviceAddressMultiDevice
);
1108 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_COMPUTE_SHADER_DERIVATIVES_FEATURES_NV
: {
1109 VkPhysicalDeviceComputeShaderDerivativesFeaturesNV
*features
=
1110 (VkPhysicalDeviceComputeShaderDerivativesFeaturesNV
*)ext
;
1111 features
->computeDerivativeGroupQuads
= true;
1112 features
->computeDerivativeGroupLinear
= true;
1116 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_CONDITIONAL_RENDERING_FEATURES_EXT
: {
1117 VkPhysicalDeviceConditionalRenderingFeaturesEXT
*features
=
1118 (VkPhysicalDeviceConditionalRenderingFeaturesEXT
*)ext
;
1119 features
->conditionalRendering
= pdevice
->info
.gen
>= 8 ||
1120 pdevice
->info
.is_haswell
;
1121 features
->inheritedConditionalRendering
= pdevice
->info
.gen
>= 8 ||
1122 pdevice
->info
.is_haswell
;
1126 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_CUSTOM_BORDER_COLOR_FEATURES_EXT
: {
1127 VkPhysicalDeviceCustomBorderColorFeaturesEXT
*features
=
1128 (VkPhysicalDeviceCustomBorderColorFeaturesEXT
*)ext
;
1129 features
->customBorderColors
= pdevice
->info
.gen
>= 8;
1130 features
->customBorderColorWithoutFormat
= pdevice
->info
.gen
>= 8;
1134 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_DEPTH_CLIP_ENABLE_FEATURES_EXT
: {
1135 VkPhysicalDeviceDepthClipEnableFeaturesEXT
*features
=
1136 (VkPhysicalDeviceDepthClipEnableFeaturesEXT
*)ext
;
1137 features
->depthClipEnable
= true;
1141 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FLOAT16_INT8_FEATURES_KHR
: {
1142 VkPhysicalDeviceFloat16Int8FeaturesKHR
*features
= (void *)ext
;
1143 CORE_FEATURE(1, 2, shaderFloat16
);
1144 CORE_FEATURE(1, 2, shaderInt8
);
1148 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FRAGMENT_SHADER_INTERLOCK_FEATURES_EXT
: {
1149 VkPhysicalDeviceFragmentShaderInterlockFeaturesEXT
*features
=
1150 (VkPhysicalDeviceFragmentShaderInterlockFeaturesEXT
*)ext
;
1151 features
->fragmentShaderSampleInterlock
= pdevice
->info
.gen
>= 9;
1152 features
->fragmentShaderPixelInterlock
= pdevice
->info
.gen
>= 9;
1153 features
->fragmentShaderShadingRateInterlock
= false;
1157 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_HOST_QUERY_RESET_FEATURES_EXT
: {
1158 VkPhysicalDeviceHostQueryResetFeaturesEXT
*features
=
1159 (VkPhysicalDeviceHostQueryResetFeaturesEXT
*)ext
;
1160 CORE_FEATURE(1, 2, hostQueryReset
);
1164 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_DESCRIPTOR_INDEXING_FEATURES_EXT
: {
1165 VkPhysicalDeviceDescriptorIndexingFeaturesEXT
*features
=
1166 (VkPhysicalDeviceDescriptorIndexingFeaturesEXT
*)ext
;
1167 CORE_FEATURE(1, 2, shaderInputAttachmentArrayDynamicIndexing
);
1168 CORE_FEATURE(1, 2, shaderUniformTexelBufferArrayDynamicIndexing
);
1169 CORE_FEATURE(1, 2, shaderStorageTexelBufferArrayDynamicIndexing
);
1170 CORE_FEATURE(1, 2, shaderUniformBufferArrayNonUniformIndexing
);
1171 CORE_FEATURE(1, 2, shaderSampledImageArrayNonUniformIndexing
);
1172 CORE_FEATURE(1, 2, shaderStorageBufferArrayNonUniformIndexing
);
1173 CORE_FEATURE(1, 2, shaderStorageImageArrayNonUniformIndexing
);
1174 CORE_FEATURE(1, 2, shaderInputAttachmentArrayNonUniformIndexing
);
1175 CORE_FEATURE(1, 2, shaderUniformTexelBufferArrayNonUniformIndexing
);
1176 CORE_FEATURE(1, 2, shaderStorageTexelBufferArrayNonUniformIndexing
);
1177 CORE_FEATURE(1, 2, descriptorBindingUniformBufferUpdateAfterBind
);
1178 CORE_FEATURE(1, 2, descriptorBindingSampledImageUpdateAfterBind
);
1179 CORE_FEATURE(1, 2, descriptorBindingStorageImageUpdateAfterBind
);
1180 CORE_FEATURE(1, 2, descriptorBindingStorageBufferUpdateAfterBind
);
1181 CORE_FEATURE(1, 2, descriptorBindingUniformTexelBufferUpdateAfterBind
);
1182 CORE_FEATURE(1, 2, descriptorBindingStorageTexelBufferUpdateAfterBind
);
1183 CORE_FEATURE(1, 2, descriptorBindingUpdateUnusedWhilePending
);
1184 CORE_FEATURE(1, 2, descriptorBindingPartiallyBound
);
1185 CORE_FEATURE(1, 2, descriptorBindingVariableDescriptorCount
);
1186 CORE_FEATURE(1, 2, runtimeDescriptorArray
);
1190 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_INDEX_TYPE_UINT8_FEATURES_EXT
: {
1191 VkPhysicalDeviceIndexTypeUint8FeaturesEXT
*features
=
1192 (VkPhysicalDeviceIndexTypeUint8FeaturesEXT
*)ext
;
1193 features
->indexTypeUint8
= true;
1197 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_INLINE_UNIFORM_BLOCK_FEATURES_EXT
: {
1198 VkPhysicalDeviceInlineUniformBlockFeaturesEXT
*features
=
1199 (VkPhysicalDeviceInlineUniformBlockFeaturesEXT
*)ext
;
1200 features
->inlineUniformBlock
= true;
1201 features
->descriptorBindingInlineUniformBlockUpdateAfterBind
= true;
1205 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_LINE_RASTERIZATION_FEATURES_EXT
: {
1206 VkPhysicalDeviceLineRasterizationFeaturesEXT
*features
=
1207 (VkPhysicalDeviceLineRasterizationFeaturesEXT
*)ext
;
1208 features
->rectangularLines
= true;
1209 features
->bresenhamLines
= true;
1210 /* Support for Smooth lines with MSAA was removed on gen11. From the
1211 * BSpec section "Multisample ModesState" table for "AA Line Support
1214 * GEN10:BUG:######## NUM_MULTISAMPLES == 1
1216 * Fortunately, this isn't a case most people care about.
1218 features
->smoothLines
= pdevice
->info
.gen
< 10;
1219 features
->stippledRectangularLines
= false;
1220 features
->stippledBresenhamLines
= true;
1221 features
->stippledSmoothLines
= false;
1225 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_MULTIVIEW_FEATURES
: {
1226 VkPhysicalDeviceMultiviewFeatures
*features
=
1227 (VkPhysicalDeviceMultiviewFeatures
*)ext
;
1228 CORE_FEATURE(1, 1, multiview
);
1229 CORE_FEATURE(1, 1, multiviewGeometryShader
);
1230 CORE_FEATURE(1, 1, multiviewTessellationShader
);
1234 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_IMAGELESS_FRAMEBUFFER_FEATURES_KHR
: {
1235 VkPhysicalDeviceImagelessFramebufferFeaturesKHR
*features
=
1236 (VkPhysicalDeviceImagelessFramebufferFeaturesKHR
*)ext
;
1237 CORE_FEATURE(1, 2, imagelessFramebuffer
);
1241 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PERFORMANCE_QUERY_FEATURES_KHR
: {
1242 VkPhysicalDevicePerformanceQueryFeaturesKHR
*feature
=
1243 (VkPhysicalDevicePerformanceQueryFeaturesKHR
*)ext
;
1244 feature
->performanceCounterQueryPools
= true;
1245 /* HW only supports a single configuration at a time. */
1246 feature
->performanceCounterMultipleQueryPools
= false;
1250 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PIPELINE_EXECUTABLE_PROPERTIES_FEATURES_KHR
: {
1251 VkPhysicalDevicePipelineExecutablePropertiesFeaturesKHR
*features
=
1252 (VkPhysicalDevicePipelineExecutablePropertiesFeaturesKHR
*)ext
;
1253 features
->pipelineExecutableInfo
= true;
1257 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PRIVATE_DATA_FEATURES_EXT
: {
1258 VkPhysicalDevicePrivateDataFeaturesEXT
*features
= (void *)ext
;
1259 features
->privateData
= true;
1263 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PROTECTED_MEMORY_FEATURES
: {
1264 VkPhysicalDeviceProtectedMemoryFeatures
*features
= (void *)ext
;
1265 CORE_FEATURE(1, 1, protectedMemory
);
1269 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_ROBUSTNESS_2_FEATURES_EXT
: {
1270 VkPhysicalDeviceRobustness2FeaturesEXT
*features
= (void *)ext
;
1271 features
->robustBufferAccess2
= true;
1272 features
->robustImageAccess2
= true;
1273 features
->nullDescriptor
= true;
1277 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SAMPLER_YCBCR_CONVERSION_FEATURES
: {
1278 VkPhysicalDeviceSamplerYcbcrConversionFeatures
*features
=
1279 (VkPhysicalDeviceSamplerYcbcrConversionFeatures
*) ext
;
1280 CORE_FEATURE(1, 1, samplerYcbcrConversion
);
1284 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SCALAR_BLOCK_LAYOUT_FEATURES_EXT
: {
1285 VkPhysicalDeviceScalarBlockLayoutFeaturesEXT
*features
=
1286 (VkPhysicalDeviceScalarBlockLayoutFeaturesEXT
*)ext
;
1287 CORE_FEATURE(1, 2, scalarBlockLayout
);
1291 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SEPARATE_DEPTH_STENCIL_LAYOUTS_FEATURES_KHR
: {
1292 VkPhysicalDeviceSeparateDepthStencilLayoutsFeaturesKHR
*features
=
1293 (VkPhysicalDeviceSeparateDepthStencilLayoutsFeaturesKHR
*)ext
;
1294 CORE_FEATURE(1, 2, separateDepthStencilLayouts
);
1298 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SHADER_ATOMIC_INT64_FEATURES_KHR
: {
1299 VkPhysicalDeviceShaderAtomicInt64FeaturesKHR
*features
= (void *)ext
;
1300 CORE_FEATURE(1, 2, shaderBufferInt64Atomics
);
1301 CORE_FEATURE(1, 2, shaderSharedInt64Atomics
);
1305 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SHADER_DEMOTE_TO_HELPER_INVOCATION_FEATURES_EXT
: {
1306 VkPhysicalDeviceShaderDemoteToHelperInvocationFeaturesEXT
*features
= (void *)ext
;
1307 features
->shaderDemoteToHelperInvocation
= true;
1311 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SHADER_CLOCK_FEATURES_KHR
: {
1312 VkPhysicalDeviceShaderClockFeaturesKHR
*features
=
1313 (VkPhysicalDeviceShaderClockFeaturesKHR
*)ext
;
1314 features
->shaderSubgroupClock
= true;
1315 features
->shaderDeviceClock
= false;
1319 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SHADER_DRAW_PARAMETERS_FEATURES
: {
1320 VkPhysicalDeviceShaderDrawParametersFeatures
*features
= (void *)ext
;
1321 CORE_FEATURE(1, 1, shaderDrawParameters
);
1325 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SHADER_SUBGROUP_EXTENDED_TYPES_FEATURES_KHR
: {
1326 VkPhysicalDeviceShaderSubgroupExtendedTypesFeaturesKHR
*features
=
1327 (VkPhysicalDeviceShaderSubgroupExtendedTypesFeaturesKHR
*)ext
;
1328 CORE_FEATURE(1, 2, shaderSubgroupExtendedTypes
);
1332 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SUBGROUP_SIZE_CONTROL_FEATURES_EXT
: {
1333 VkPhysicalDeviceSubgroupSizeControlFeaturesEXT
*features
=
1334 (VkPhysicalDeviceSubgroupSizeControlFeaturesEXT
*)ext
;
1335 features
->subgroupSizeControl
= true;
1336 features
->computeFullSubgroups
= true;
1340 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_TEXEL_BUFFER_ALIGNMENT_FEATURES_EXT
: {
1341 VkPhysicalDeviceTexelBufferAlignmentFeaturesEXT
*features
=
1342 (VkPhysicalDeviceTexelBufferAlignmentFeaturesEXT
*)ext
;
1343 features
->texelBufferAlignment
= true;
1347 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_TIMELINE_SEMAPHORE_FEATURES_KHR
: {
1348 VkPhysicalDeviceTimelineSemaphoreFeaturesKHR
*features
=
1349 (VkPhysicalDeviceTimelineSemaphoreFeaturesKHR
*) ext
;
1350 CORE_FEATURE(1, 2, timelineSemaphore
);
1354 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VARIABLE_POINTERS_FEATURES
: {
1355 VkPhysicalDeviceVariablePointersFeatures
*features
= (void *)ext
;
1356 CORE_FEATURE(1, 1, variablePointersStorageBuffer
);
1357 CORE_FEATURE(1, 1, variablePointers
);
1361 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_TRANSFORM_FEEDBACK_FEATURES_EXT
: {
1362 VkPhysicalDeviceTransformFeedbackFeaturesEXT
*features
=
1363 (VkPhysicalDeviceTransformFeedbackFeaturesEXT
*)ext
;
1364 features
->transformFeedback
= true;
1365 features
->geometryStreams
= true;
1369 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_UNIFORM_BUFFER_STANDARD_LAYOUT_FEATURES_KHR
: {
1370 VkPhysicalDeviceUniformBufferStandardLayoutFeaturesKHR
*features
=
1371 (VkPhysicalDeviceUniformBufferStandardLayoutFeaturesKHR
*)ext
;
1372 CORE_FEATURE(1, 2, uniformBufferStandardLayout
);
1376 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VERTEX_ATTRIBUTE_DIVISOR_FEATURES_EXT
: {
1377 VkPhysicalDeviceVertexAttributeDivisorFeaturesEXT
*features
=
1378 (VkPhysicalDeviceVertexAttributeDivisorFeaturesEXT
*)ext
;
1379 features
->vertexAttributeInstanceRateDivisor
= true;
1380 features
->vertexAttributeInstanceRateZeroDivisor
= true;
1384 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_1_FEATURES
:
1385 anv_get_physical_device_features_1_1(pdevice
, (void *)ext
);
1388 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_FEATURES
:
1389 anv_get_physical_device_features_1_2(pdevice
, (void *)ext
);
1392 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_MEMORY_MODEL_FEATURES_KHR
: {
1393 VkPhysicalDeviceVulkanMemoryModelFeaturesKHR
*features
= (void *)ext
;
1394 CORE_FEATURE(1, 2, vulkanMemoryModel
);
1395 CORE_FEATURE(1, 2, vulkanMemoryModelDeviceScope
);
1396 CORE_FEATURE(1, 2, vulkanMemoryModelAvailabilityVisibilityChains
);
1400 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_YCBCR_IMAGE_ARRAYS_FEATURES_EXT
: {
1401 VkPhysicalDeviceYcbcrImageArraysFeaturesEXT
*features
=
1402 (VkPhysicalDeviceYcbcrImageArraysFeaturesEXT
*)ext
;
1403 features
->ycbcrImageArrays
= true;
1408 anv_debug_ignored_stype(ext
->sType
);
1416 #define MAX_PER_STAGE_DESCRIPTOR_UNIFORM_BUFFERS 64
1418 #define MAX_PER_STAGE_DESCRIPTOR_INPUT_ATTACHMENTS 64
1419 #define MAX_DESCRIPTOR_SET_INPUT_ATTACHMENTS 256
1421 #define MAX_CUSTOM_BORDER_COLORS 4096
1423 void anv_GetPhysicalDeviceProperties(
1424 VkPhysicalDevice physicalDevice
,
1425 VkPhysicalDeviceProperties
* pProperties
)
1427 ANV_FROM_HANDLE(anv_physical_device
, pdevice
, physicalDevice
);
1428 const struct gen_device_info
*devinfo
= &pdevice
->info
;
1430 /* See assertions made when programming the buffer surface state. */
1431 const uint32_t max_raw_buffer_sz
= devinfo
->gen
>= 7 ?
1432 (1ul << 30) : (1ul << 27);
1434 const uint32_t max_ssbos
= pdevice
->has_a64_buffer_access
? UINT16_MAX
: 64;
1435 const uint32_t max_textures
=
1436 pdevice
->has_bindless_images
? UINT16_MAX
: 128;
1437 const uint32_t max_samplers
=
1438 pdevice
->has_bindless_samplers
? UINT16_MAX
:
1439 (devinfo
->gen
>= 8 || devinfo
->is_haswell
) ? 128 : 16;
1440 const uint32_t max_images
=
1441 pdevice
->has_bindless_images
? UINT16_MAX
: MAX_IMAGES
;
1443 /* If we can use bindless for everything, claim a high per-stage limit,
1444 * otherwise use the binding table size, minus the slots reserved for
1445 * render targets and one slot for the descriptor buffer. */
1446 const uint32_t max_per_stage
=
1447 pdevice
->has_bindless_images
&& pdevice
->has_a64_buffer_access
1448 ? UINT32_MAX
: MAX_BINDING_TABLE_SIZE
- MAX_RTS
- 1;
1450 /* Limit max_threads to 64 for the GPGPU_WALKER command */
1451 const uint32_t max_workgroup_size
= 32 * MIN2(64, devinfo
->max_cs_threads
);
1453 VkSampleCountFlags sample_counts
=
1454 isl_device_get_sample_counts(&pdevice
->isl_dev
);
1457 VkPhysicalDeviceLimits limits
= {
1458 .maxImageDimension1D
= (1 << 14),
1459 .maxImageDimension2D
= (1 << 14),
1460 .maxImageDimension3D
= (1 << 11),
1461 .maxImageDimensionCube
= (1 << 14),
1462 .maxImageArrayLayers
= (1 << 11),
1463 .maxTexelBufferElements
= 128 * 1024 * 1024,
1464 .maxUniformBufferRange
= (1ul << 27),
1465 .maxStorageBufferRange
= max_raw_buffer_sz
,
1466 .maxPushConstantsSize
= MAX_PUSH_CONSTANTS_SIZE
,
1467 .maxMemoryAllocationCount
= UINT32_MAX
,
1468 .maxSamplerAllocationCount
= 64 * 1024,
1469 .bufferImageGranularity
= 64, /* A cache line */
1470 .sparseAddressSpaceSize
= 0,
1471 .maxBoundDescriptorSets
= MAX_SETS
,
1472 .maxPerStageDescriptorSamplers
= max_samplers
,
1473 .maxPerStageDescriptorUniformBuffers
= MAX_PER_STAGE_DESCRIPTOR_UNIFORM_BUFFERS
,
1474 .maxPerStageDescriptorStorageBuffers
= max_ssbos
,
1475 .maxPerStageDescriptorSampledImages
= max_textures
,
1476 .maxPerStageDescriptorStorageImages
= max_images
,
1477 .maxPerStageDescriptorInputAttachments
= MAX_PER_STAGE_DESCRIPTOR_INPUT_ATTACHMENTS
,
1478 .maxPerStageResources
= max_per_stage
,
1479 .maxDescriptorSetSamplers
= 6 * max_samplers
, /* number of stages * maxPerStageDescriptorSamplers */
1480 .maxDescriptorSetUniformBuffers
= 6 * MAX_PER_STAGE_DESCRIPTOR_UNIFORM_BUFFERS
, /* number of stages * maxPerStageDescriptorUniformBuffers */
1481 .maxDescriptorSetUniformBuffersDynamic
= MAX_DYNAMIC_BUFFERS
/ 2,
1482 .maxDescriptorSetStorageBuffers
= 6 * max_ssbos
, /* number of stages * maxPerStageDescriptorStorageBuffers */
1483 .maxDescriptorSetStorageBuffersDynamic
= MAX_DYNAMIC_BUFFERS
/ 2,
1484 .maxDescriptorSetSampledImages
= 6 * max_textures
, /* number of stages * maxPerStageDescriptorSampledImages */
1485 .maxDescriptorSetStorageImages
= 6 * max_images
, /* number of stages * maxPerStageDescriptorStorageImages */
1486 .maxDescriptorSetInputAttachments
= MAX_DESCRIPTOR_SET_INPUT_ATTACHMENTS
,
1487 .maxVertexInputAttributes
= MAX_VBS
,
1488 .maxVertexInputBindings
= MAX_VBS
,
1489 .maxVertexInputAttributeOffset
= 2047,
1490 .maxVertexInputBindingStride
= 2048,
1491 .maxVertexOutputComponents
= 128,
1492 .maxTessellationGenerationLevel
= 64,
1493 .maxTessellationPatchSize
= 32,
1494 .maxTessellationControlPerVertexInputComponents
= 128,
1495 .maxTessellationControlPerVertexOutputComponents
= 128,
1496 .maxTessellationControlPerPatchOutputComponents
= 128,
1497 .maxTessellationControlTotalOutputComponents
= 2048,
1498 .maxTessellationEvaluationInputComponents
= 128,
1499 .maxTessellationEvaluationOutputComponents
= 128,
1500 .maxGeometryShaderInvocations
= 32,
1501 .maxGeometryInputComponents
= 64,
1502 .maxGeometryOutputComponents
= 128,
1503 .maxGeometryOutputVertices
= 256,
1504 .maxGeometryTotalOutputComponents
= 1024,
1505 .maxFragmentInputComponents
= 116, /* 128 components - (PSIZ, CLIP_DIST0, CLIP_DIST1) */
1506 .maxFragmentOutputAttachments
= 8,
1507 .maxFragmentDualSrcAttachments
= 1,
1508 .maxFragmentCombinedOutputResources
= 8,
1509 .maxComputeSharedMemorySize
= 64 * 1024,
1510 .maxComputeWorkGroupCount
= { 65535, 65535, 65535 },
1511 .maxComputeWorkGroupInvocations
= max_workgroup_size
,
1512 .maxComputeWorkGroupSize
= {
1517 .subPixelPrecisionBits
= 8,
1518 .subTexelPrecisionBits
= 8,
1519 .mipmapPrecisionBits
= 8,
1520 .maxDrawIndexedIndexValue
= UINT32_MAX
,
1521 .maxDrawIndirectCount
= UINT32_MAX
,
1522 .maxSamplerLodBias
= 16,
1523 .maxSamplerAnisotropy
= 16,
1524 .maxViewports
= MAX_VIEWPORTS
,
1525 .maxViewportDimensions
= { (1 << 14), (1 << 14) },
1526 .viewportBoundsRange
= { INT16_MIN
, INT16_MAX
},
1527 .viewportSubPixelBits
= 13, /* We take a float? */
1528 .minMemoryMapAlignment
= 4096, /* A page */
1529 /* The dataport requires texel alignment so we need to assume a worst
1530 * case of R32G32B32A32 which is 16 bytes.
1532 .minTexelBufferOffsetAlignment
= 16,
1533 .minUniformBufferOffsetAlignment
= ANV_UBO_ALIGNMENT
,
1534 .minStorageBufferOffsetAlignment
= 4,
1535 .minTexelOffset
= -8,
1536 .maxTexelOffset
= 7,
1537 .minTexelGatherOffset
= -32,
1538 .maxTexelGatherOffset
= 31,
1539 .minInterpolationOffset
= -0.5,
1540 .maxInterpolationOffset
= 0.4375,
1541 .subPixelInterpolationOffsetBits
= 4,
1542 .maxFramebufferWidth
= (1 << 14),
1543 .maxFramebufferHeight
= (1 << 14),
1544 .maxFramebufferLayers
= (1 << 11),
1545 .framebufferColorSampleCounts
= sample_counts
,
1546 .framebufferDepthSampleCounts
= sample_counts
,
1547 .framebufferStencilSampleCounts
= sample_counts
,
1548 .framebufferNoAttachmentsSampleCounts
= sample_counts
,
1549 .maxColorAttachments
= MAX_RTS
,
1550 .sampledImageColorSampleCounts
= sample_counts
,
1551 .sampledImageIntegerSampleCounts
= sample_counts
,
1552 .sampledImageDepthSampleCounts
= sample_counts
,
1553 .sampledImageStencilSampleCounts
= sample_counts
,
1554 .storageImageSampleCounts
= VK_SAMPLE_COUNT_1_BIT
,
1555 .maxSampleMaskWords
= 1,
1556 .timestampComputeAndGraphics
= true,
1557 .timestampPeriod
= 1000000000.0 / devinfo
->timestamp_frequency
,
1558 .maxClipDistances
= 8,
1559 .maxCullDistances
= 8,
1560 .maxCombinedClipAndCullDistances
= 8,
1561 .discreteQueuePriorities
= 2,
1562 .pointSizeRange
= { 0.125, 255.875 },
1565 (devinfo
->gen
>= 9 || devinfo
->is_cherryview
) ?
1566 2047.9921875 : 7.9921875,
1568 .pointSizeGranularity
= (1.0 / 8.0),
1569 .lineWidthGranularity
= (1.0 / 128.0),
1570 .strictLines
= false,
1571 .standardSampleLocations
= true,
1572 .optimalBufferCopyOffsetAlignment
= 128,
1573 .optimalBufferCopyRowPitchAlignment
= 128,
1574 .nonCoherentAtomSize
= 64,
1577 *pProperties
= (VkPhysicalDeviceProperties
) {
1578 .apiVersion
= anv_physical_device_api_version(pdevice
),
1579 .driverVersion
= vk_get_driver_version(),
1581 .deviceID
= pdevice
->info
.chipset_id
,
1582 .deviceType
= VK_PHYSICAL_DEVICE_TYPE_INTEGRATED_GPU
,
1584 .sparseProperties
= {0}, /* Broadwell doesn't do sparse. */
1587 snprintf(pProperties
->deviceName
, sizeof(pProperties
->deviceName
),
1588 "%s", pdevice
->name
);
1589 memcpy(pProperties
->pipelineCacheUUID
,
1590 pdevice
->pipeline_cache_uuid
, VK_UUID_SIZE
);
1594 anv_get_physical_device_properties_1_1(struct anv_physical_device
*pdevice
,
1595 VkPhysicalDeviceVulkan11Properties
*p
)
1597 assert(p
->sType
== VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_1_PROPERTIES
);
1599 memcpy(p
->deviceUUID
, pdevice
->device_uuid
, VK_UUID_SIZE
);
1600 memcpy(p
->driverUUID
, pdevice
->driver_uuid
, VK_UUID_SIZE
);
1601 memset(p
->deviceLUID
, 0, VK_LUID_SIZE
);
1602 p
->deviceNodeMask
= 0;
1603 p
->deviceLUIDValid
= false;
1605 p
->subgroupSize
= BRW_SUBGROUP_SIZE
;
1606 VkShaderStageFlags scalar_stages
= 0;
1607 for (unsigned stage
= 0; stage
< MESA_SHADER_STAGES
; stage
++) {
1608 if (pdevice
->compiler
->scalar_stage
[stage
])
1609 scalar_stages
|= mesa_to_vk_shader_stage(stage
);
1611 p
->subgroupSupportedStages
= scalar_stages
;
1612 p
->subgroupSupportedOperations
= VK_SUBGROUP_FEATURE_BASIC_BIT
|
1613 VK_SUBGROUP_FEATURE_VOTE_BIT
|
1614 VK_SUBGROUP_FEATURE_BALLOT_BIT
|
1615 VK_SUBGROUP_FEATURE_SHUFFLE_BIT
|
1616 VK_SUBGROUP_FEATURE_SHUFFLE_RELATIVE_BIT
|
1617 VK_SUBGROUP_FEATURE_QUAD_BIT
;
1618 if (pdevice
->info
.gen
>= 8) {
1619 /* TODO: There's no technical reason why these can't be made to
1620 * work on gen7 but they don't at the moment so it's best to leave
1621 * the feature disabled than enabled and broken.
1623 p
->subgroupSupportedOperations
|= VK_SUBGROUP_FEATURE_ARITHMETIC_BIT
|
1624 VK_SUBGROUP_FEATURE_CLUSTERED_BIT
;
1626 p
->subgroupQuadOperationsInAllStages
= pdevice
->info
.gen
>= 8;
1628 p
->pointClippingBehavior
= VK_POINT_CLIPPING_BEHAVIOR_USER_CLIP_PLANES_ONLY
;
1629 p
->maxMultiviewViewCount
= 16;
1630 p
->maxMultiviewInstanceIndex
= UINT32_MAX
/ 16;
1631 p
->protectedNoFault
= false;
1632 /* This value doesn't matter for us today as our per-stage descriptors are
1635 p
->maxPerSetDescriptors
= 1024;
1636 p
->maxMemoryAllocationSize
= MAX_MEMORY_ALLOCATION_SIZE
;
1640 anv_get_physical_device_properties_1_2(struct anv_physical_device
*pdevice
,
1641 VkPhysicalDeviceVulkan12Properties
*p
)
1643 assert(p
->sType
== VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_PROPERTIES
);
1645 p
->driverID
= VK_DRIVER_ID_INTEL_OPEN_SOURCE_MESA_KHR
;
1646 memset(p
->driverName
, 0, sizeof(p
->driverName
));
1647 snprintf(p
->driverName
, VK_MAX_DRIVER_NAME_SIZE_KHR
,
1648 "Intel open-source Mesa driver");
1649 memset(p
->driverInfo
, 0, sizeof(p
->driverInfo
));
1650 snprintf(p
->driverInfo
, VK_MAX_DRIVER_INFO_SIZE_KHR
,
1651 "Mesa " PACKAGE_VERSION MESA_GIT_SHA1
);
1652 p
->conformanceVersion
= (VkConformanceVersionKHR
) {
1659 p
->denormBehaviorIndependence
=
1660 VK_SHADER_FLOAT_CONTROLS_INDEPENDENCE_ALL_KHR
;
1661 p
->roundingModeIndependence
=
1662 VK_SHADER_FLOAT_CONTROLS_INDEPENDENCE_NONE_KHR
;
1664 /* Broadwell does not support HF denorms and there are restrictions
1665 * other gens. According to Kabylake's PRM:
1667 * "math - Extended Math Function
1669 * Restriction : Half-float denorms are always retained."
1671 p
->shaderDenormFlushToZeroFloat16
= false;
1672 p
->shaderDenormPreserveFloat16
= pdevice
->info
.gen
> 8;
1673 p
->shaderRoundingModeRTEFloat16
= true;
1674 p
->shaderRoundingModeRTZFloat16
= true;
1675 p
->shaderSignedZeroInfNanPreserveFloat16
= true;
1677 p
->shaderDenormFlushToZeroFloat32
= true;
1678 p
->shaderDenormPreserveFloat32
= true;
1679 p
->shaderRoundingModeRTEFloat32
= true;
1680 p
->shaderRoundingModeRTZFloat32
= true;
1681 p
->shaderSignedZeroInfNanPreserveFloat32
= true;
1683 p
->shaderDenormFlushToZeroFloat64
= true;
1684 p
->shaderDenormPreserveFloat64
= true;
1685 p
->shaderRoundingModeRTEFloat64
= true;
1686 p
->shaderRoundingModeRTZFloat64
= true;
1687 p
->shaderSignedZeroInfNanPreserveFloat64
= true;
1689 /* It's a bit hard to exactly map our implementation to the limits
1690 * described here. The bindless surface handle in the extended
1691 * message descriptors is 20 bits and it's an index into the table of
1692 * RENDER_SURFACE_STATE structs that starts at bindless surface base
1693 * address. Given that most things consume two surface states per
1694 * view (general/sampled for textures and write-only/read-write for
1695 * images), we claim 2^19 things.
1697 * For SSBOs, we just use A64 messages so there is no real limit
1698 * there beyond the limit on the total size of a descriptor set.
1700 const unsigned max_bindless_views
= 1 << 19;
1701 p
->maxUpdateAfterBindDescriptorsInAllPools
= max_bindless_views
;
1702 p
->shaderUniformBufferArrayNonUniformIndexingNative
= false;
1703 p
->shaderSampledImageArrayNonUniformIndexingNative
= false;
1704 p
->shaderStorageBufferArrayNonUniformIndexingNative
= true;
1705 p
->shaderStorageImageArrayNonUniformIndexingNative
= false;
1706 p
->shaderInputAttachmentArrayNonUniformIndexingNative
= false;
1707 p
->robustBufferAccessUpdateAfterBind
= true;
1708 p
->quadDivergentImplicitLod
= false;
1709 p
->maxPerStageDescriptorUpdateAfterBindSamplers
= max_bindless_views
;
1710 p
->maxPerStageDescriptorUpdateAfterBindUniformBuffers
= MAX_PER_STAGE_DESCRIPTOR_UNIFORM_BUFFERS
;
1711 p
->maxPerStageDescriptorUpdateAfterBindStorageBuffers
= UINT32_MAX
;
1712 p
->maxPerStageDescriptorUpdateAfterBindSampledImages
= max_bindless_views
;
1713 p
->maxPerStageDescriptorUpdateAfterBindStorageImages
= max_bindless_views
;
1714 p
->maxPerStageDescriptorUpdateAfterBindInputAttachments
= MAX_PER_STAGE_DESCRIPTOR_INPUT_ATTACHMENTS
;
1715 p
->maxPerStageUpdateAfterBindResources
= UINT32_MAX
;
1716 p
->maxDescriptorSetUpdateAfterBindSamplers
= max_bindless_views
;
1717 p
->maxDescriptorSetUpdateAfterBindUniformBuffers
= 6 * MAX_PER_STAGE_DESCRIPTOR_UNIFORM_BUFFERS
;
1718 p
->maxDescriptorSetUpdateAfterBindUniformBuffersDynamic
= MAX_DYNAMIC_BUFFERS
/ 2;
1719 p
->maxDescriptorSetUpdateAfterBindStorageBuffers
= UINT32_MAX
;
1720 p
->maxDescriptorSetUpdateAfterBindStorageBuffersDynamic
= MAX_DYNAMIC_BUFFERS
/ 2;
1721 p
->maxDescriptorSetUpdateAfterBindSampledImages
= max_bindless_views
;
1722 p
->maxDescriptorSetUpdateAfterBindStorageImages
= max_bindless_views
;
1723 p
->maxDescriptorSetUpdateAfterBindInputAttachments
= MAX_DESCRIPTOR_SET_INPUT_ATTACHMENTS
;
1725 /* We support all of the depth resolve modes */
1726 p
->supportedDepthResolveModes
= VK_RESOLVE_MODE_SAMPLE_ZERO_BIT_KHR
|
1727 VK_RESOLVE_MODE_AVERAGE_BIT_KHR
|
1728 VK_RESOLVE_MODE_MIN_BIT_KHR
|
1729 VK_RESOLVE_MODE_MAX_BIT_KHR
;
1730 /* Average doesn't make sense for stencil so we don't support that */
1731 p
->supportedStencilResolveModes
= VK_RESOLVE_MODE_SAMPLE_ZERO_BIT_KHR
;
1732 if (pdevice
->info
.gen
>= 8) {
1733 /* The advanced stencil resolve modes currently require stencil
1734 * sampling be supported by the hardware.
1736 p
->supportedStencilResolveModes
|= VK_RESOLVE_MODE_MIN_BIT_KHR
|
1737 VK_RESOLVE_MODE_MAX_BIT_KHR
;
1739 p
->independentResolveNone
= true;
1740 p
->independentResolve
= true;
1742 p
->filterMinmaxSingleComponentFormats
= pdevice
->info
.gen
>= 9;
1743 p
->filterMinmaxImageComponentMapping
= pdevice
->info
.gen
>= 9;
1745 p
->maxTimelineSemaphoreValueDifference
= UINT64_MAX
;
1747 p
->framebufferIntegerColorSampleCounts
=
1748 isl_device_get_sample_counts(&pdevice
->isl_dev
);
1751 void anv_GetPhysicalDeviceProperties2(
1752 VkPhysicalDevice physicalDevice
,
1753 VkPhysicalDeviceProperties2
* pProperties
)
1755 ANV_FROM_HANDLE(anv_physical_device
, pdevice
, physicalDevice
);
1757 anv_GetPhysicalDeviceProperties(physicalDevice
, &pProperties
->properties
);
1759 VkPhysicalDeviceVulkan11Properties core_1_1
= {
1760 .sType
= VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_1_PROPERTIES
,
1762 anv_get_physical_device_properties_1_1(pdevice
, &core_1_1
);
1764 VkPhysicalDeviceVulkan12Properties core_1_2
= {
1765 .sType
= VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_PROPERTIES
,
1767 anv_get_physical_device_properties_1_2(pdevice
, &core_1_2
);
1769 #define CORE_RENAMED_PROPERTY(major, minor, ext_property, core_property) \
1770 memcpy(&properties->ext_property, &core_##major##_##minor.core_property, \
1771 sizeof(core_##major##_##minor.core_property))
1773 #define CORE_PROPERTY(major, minor, property) \
1774 CORE_RENAMED_PROPERTY(major, minor, property, property)
1776 vk_foreach_struct(ext
, pProperties
->pNext
) {
1777 switch (ext
->sType
) {
1778 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_CUSTOM_BORDER_COLOR_PROPERTIES_EXT
: {
1779 VkPhysicalDeviceCustomBorderColorPropertiesEXT
*properties
=
1780 (VkPhysicalDeviceCustomBorderColorPropertiesEXT
*)ext
;
1781 properties
->maxCustomBorderColorSamplers
= MAX_CUSTOM_BORDER_COLORS
;
1785 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_DEPTH_STENCIL_RESOLVE_PROPERTIES_KHR
: {
1786 VkPhysicalDeviceDepthStencilResolvePropertiesKHR
*properties
=
1787 (VkPhysicalDeviceDepthStencilResolvePropertiesKHR
*)ext
;
1788 CORE_PROPERTY(1, 2, supportedDepthResolveModes
);
1789 CORE_PROPERTY(1, 2, supportedStencilResolveModes
);
1790 CORE_PROPERTY(1, 2, independentResolveNone
);
1791 CORE_PROPERTY(1, 2, independentResolve
);
1795 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_DESCRIPTOR_INDEXING_PROPERTIES_EXT
: {
1796 VkPhysicalDeviceDescriptorIndexingPropertiesEXT
*properties
=
1797 (VkPhysicalDeviceDescriptorIndexingPropertiesEXT
*)ext
;
1798 CORE_PROPERTY(1, 2, maxUpdateAfterBindDescriptorsInAllPools
);
1799 CORE_PROPERTY(1, 2, shaderUniformBufferArrayNonUniformIndexingNative
);
1800 CORE_PROPERTY(1, 2, shaderSampledImageArrayNonUniformIndexingNative
);
1801 CORE_PROPERTY(1, 2, shaderStorageBufferArrayNonUniformIndexingNative
);
1802 CORE_PROPERTY(1, 2, shaderStorageImageArrayNonUniformIndexingNative
);
1803 CORE_PROPERTY(1, 2, shaderInputAttachmentArrayNonUniformIndexingNative
);
1804 CORE_PROPERTY(1, 2, robustBufferAccessUpdateAfterBind
);
1805 CORE_PROPERTY(1, 2, quadDivergentImplicitLod
);
1806 CORE_PROPERTY(1, 2, maxPerStageDescriptorUpdateAfterBindSamplers
);
1807 CORE_PROPERTY(1, 2, maxPerStageDescriptorUpdateAfterBindUniformBuffers
);
1808 CORE_PROPERTY(1, 2, maxPerStageDescriptorUpdateAfterBindStorageBuffers
);
1809 CORE_PROPERTY(1, 2, maxPerStageDescriptorUpdateAfterBindSampledImages
);
1810 CORE_PROPERTY(1, 2, maxPerStageDescriptorUpdateAfterBindStorageImages
);
1811 CORE_PROPERTY(1, 2, maxPerStageDescriptorUpdateAfterBindInputAttachments
);
1812 CORE_PROPERTY(1, 2, maxPerStageUpdateAfterBindResources
);
1813 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindSamplers
);
1814 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindUniformBuffers
);
1815 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindUniformBuffersDynamic
);
1816 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindStorageBuffers
);
1817 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindStorageBuffersDynamic
);
1818 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindSampledImages
);
1819 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindStorageImages
);
1820 CORE_PROPERTY(1, 2, maxDescriptorSetUpdateAfterBindInputAttachments
);
1824 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_DRIVER_PROPERTIES_KHR
: {
1825 VkPhysicalDeviceDriverPropertiesKHR
*properties
=
1826 (VkPhysicalDeviceDriverPropertiesKHR
*) ext
;
1827 CORE_PROPERTY(1, 2, driverID
);
1828 CORE_PROPERTY(1, 2, driverName
);
1829 CORE_PROPERTY(1, 2, driverInfo
);
1830 CORE_PROPERTY(1, 2, conformanceVersion
);
1834 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_EXTERNAL_MEMORY_HOST_PROPERTIES_EXT
: {
1835 VkPhysicalDeviceExternalMemoryHostPropertiesEXT
*props
=
1836 (VkPhysicalDeviceExternalMemoryHostPropertiesEXT
*) ext
;
1837 /* Userptr needs page aligned memory. */
1838 props
->minImportedHostPointerAlignment
= 4096;
1842 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_ID_PROPERTIES
: {
1843 VkPhysicalDeviceIDProperties
*properties
=
1844 (VkPhysicalDeviceIDProperties
*)ext
;
1845 CORE_PROPERTY(1, 1, deviceUUID
);
1846 CORE_PROPERTY(1, 1, driverUUID
);
1847 CORE_PROPERTY(1, 1, deviceLUID
);
1848 CORE_PROPERTY(1, 1, deviceLUIDValid
);
1852 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_INLINE_UNIFORM_BLOCK_PROPERTIES_EXT
: {
1853 VkPhysicalDeviceInlineUniformBlockPropertiesEXT
*props
=
1854 (VkPhysicalDeviceInlineUniformBlockPropertiesEXT
*)ext
;
1855 props
->maxInlineUniformBlockSize
= MAX_INLINE_UNIFORM_BLOCK_SIZE
;
1856 props
->maxPerStageDescriptorInlineUniformBlocks
=
1857 MAX_INLINE_UNIFORM_BLOCK_DESCRIPTORS
;
1858 props
->maxPerStageDescriptorUpdateAfterBindInlineUniformBlocks
=
1859 MAX_INLINE_UNIFORM_BLOCK_DESCRIPTORS
;
1860 props
->maxDescriptorSetInlineUniformBlocks
=
1861 MAX_INLINE_UNIFORM_BLOCK_DESCRIPTORS
;
1862 props
->maxDescriptorSetUpdateAfterBindInlineUniformBlocks
=
1863 MAX_INLINE_UNIFORM_BLOCK_DESCRIPTORS
;
1867 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_LINE_RASTERIZATION_PROPERTIES_EXT
: {
1868 VkPhysicalDeviceLineRasterizationPropertiesEXT
*props
=
1869 (VkPhysicalDeviceLineRasterizationPropertiesEXT
*)ext
;
1870 /* In the Skylake PRM Vol. 7, subsection titled "GIQ (Diamond)
1871 * Sampling Rules - Legacy Mode", it says the following:
1873 * "Note that the device divides a pixel into a 16x16 array of
1874 * subpixels, referenced by their upper left corners."
1876 * This is the only known reference in the PRMs to the subpixel
1877 * precision of line rasterization and a "16x16 array of subpixels"
1878 * implies 4 subpixel precision bits. Empirical testing has shown
1879 * that 4 subpixel precision bits applies to all line rasterization
1882 props
->lineSubPixelPrecisionBits
= 4;
1886 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_MAINTENANCE_3_PROPERTIES
: {
1887 VkPhysicalDeviceMaintenance3Properties
*properties
=
1888 (VkPhysicalDeviceMaintenance3Properties
*)ext
;
1889 /* This value doesn't matter for us today as our per-stage
1890 * descriptors are the real limit.
1892 CORE_PROPERTY(1, 1, maxPerSetDescriptors
);
1893 CORE_PROPERTY(1, 1, maxMemoryAllocationSize
);
1897 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_MULTIVIEW_PROPERTIES
: {
1898 VkPhysicalDeviceMultiviewProperties
*properties
=
1899 (VkPhysicalDeviceMultiviewProperties
*)ext
;
1900 CORE_PROPERTY(1, 1, maxMultiviewViewCount
);
1901 CORE_PROPERTY(1, 1, maxMultiviewInstanceIndex
);
1905 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PCI_BUS_INFO_PROPERTIES_EXT
: {
1906 VkPhysicalDevicePCIBusInfoPropertiesEXT
*properties
=
1907 (VkPhysicalDevicePCIBusInfoPropertiesEXT
*)ext
;
1908 properties
->pciDomain
= pdevice
->pci_info
.domain
;
1909 properties
->pciBus
= pdevice
->pci_info
.bus
;
1910 properties
->pciDevice
= pdevice
->pci_info
.device
;
1911 properties
->pciFunction
= pdevice
->pci_info
.function
;
1915 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PERFORMANCE_QUERY_PROPERTIES_KHR
: {
1916 VkPhysicalDevicePerformanceQueryPropertiesKHR
*properties
=
1917 (VkPhysicalDevicePerformanceQueryPropertiesKHR
*)ext
;
1918 /* We could support this by spawning a shader to do the equation
1921 properties
->allowCommandBufferQueryCopies
= false;
1925 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_POINT_CLIPPING_PROPERTIES
: {
1926 VkPhysicalDevicePointClippingProperties
*properties
=
1927 (VkPhysicalDevicePointClippingProperties
*) ext
;
1928 CORE_PROPERTY(1, 1, pointClippingBehavior
);
1932 #pragma GCC diagnostic push
1933 #pragma GCC diagnostic ignored "-Wswitch"
1934 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PRESENTATION_PROPERTIES_ANDROID
: {
1935 VkPhysicalDevicePresentationPropertiesANDROID
*props
=
1936 (VkPhysicalDevicePresentationPropertiesANDROID
*)ext
;
1937 props
->sharedImage
= VK_FALSE
;
1940 #pragma GCC diagnostic pop
1942 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PROTECTED_MEMORY_PROPERTIES
: {
1943 VkPhysicalDeviceProtectedMemoryProperties
*properties
=
1944 (VkPhysicalDeviceProtectedMemoryProperties
*)ext
;
1945 CORE_PROPERTY(1, 1, protectedNoFault
);
1949 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_PUSH_DESCRIPTOR_PROPERTIES_KHR
: {
1950 VkPhysicalDevicePushDescriptorPropertiesKHR
*properties
=
1951 (VkPhysicalDevicePushDescriptorPropertiesKHR
*) ext
;
1952 properties
->maxPushDescriptors
= MAX_PUSH_DESCRIPTORS
;
1956 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_ROBUSTNESS_2_PROPERTIES_EXT
: {
1957 VkPhysicalDeviceRobustness2PropertiesEXT
*properties
= (void *)ext
;
1958 properties
->robustStorageBufferAccessSizeAlignment
=
1959 ANV_SSBO_BOUNDS_CHECK_ALIGNMENT
;
1960 properties
->robustUniformBufferAccessSizeAlignment
=
1965 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SAMPLER_FILTER_MINMAX_PROPERTIES_EXT
: {
1966 VkPhysicalDeviceSamplerFilterMinmaxPropertiesEXT
*properties
=
1967 (VkPhysicalDeviceSamplerFilterMinmaxPropertiesEXT
*)ext
;
1968 CORE_PROPERTY(1, 2, filterMinmaxImageComponentMapping
);
1969 CORE_PROPERTY(1, 2, filterMinmaxSingleComponentFormats
);
1973 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SUBGROUP_PROPERTIES
: {
1974 VkPhysicalDeviceSubgroupProperties
*properties
= (void *)ext
;
1975 CORE_PROPERTY(1, 1, subgroupSize
);
1976 CORE_RENAMED_PROPERTY(1, 1, supportedStages
,
1977 subgroupSupportedStages
);
1978 CORE_RENAMED_PROPERTY(1, 1, supportedOperations
,
1979 subgroupSupportedOperations
);
1980 CORE_RENAMED_PROPERTY(1, 1, quadOperationsInAllStages
,
1981 subgroupQuadOperationsInAllStages
);
1985 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_SUBGROUP_SIZE_CONTROL_PROPERTIES_EXT
: {
1986 VkPhysicalDeviceSubgroupSizeControlPropertiesEXT
*props
=
1987 (VkPhysicalDeviceSubgroupSizeControlPropertiesEXT
*)ext
;
1988 STATIC_ASSERT(8 <= BRW_SUBGROUP_SIZE
&& BRW_SUBGROUP_SIZE
<= 32);
1989 props
->minSubgroupSize
= 8;
1990 props
->maxSubgroupSize
= 32;
1991 props
->maxComputeWorkgroupSubgroups
= pdevice
->info
.max_cs_threads
;
1992 props
->requiredSubgroupSizeStages
= VK_SHADER_STAGE_COMPUTE_BIT
;
1995 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FLOAT_CONTROLS_PROPERTIES_KHR
: {
1996 VkPhysicalDeviceFloatControlsPropertiesKHR
*properties
= (void *)ext
;
1997 CORE_PROPERTY(1, 2, denormBehaviorIndependence
);
1998 CORE_PROPERTY(1, 2, roundingModeIndependence
);
1999 CORE_PROPERTY(1, 2, shaderDenormFlushToZeroFloat16
);
2000 CORE_PROPERTY(1, 2, shaderDenormPreserveFloat16
);
2001 CORE_PROPERTY(1, 2, shaderRoundingModeRTEFloat16
);
2002 CORE_PROPERTY(1, 2, shaderRoundingModeRTZFloat16
);
2003 CORE_PROPERTY(1, 2, shaderSignedZeroInfNanPreserveFloat16
);
2004 CORE_PROPERTY(1, 2, shaderDenormFlushToZeroFloat32
);
2005 CORE_PROPERTY(1, 2, shaderDenormPreserveFloat32
);
2006 CORE_PROPERTY(1, 2, shaderRoundingModeRTEFloat32
);
2007 CORE_PROPERTY(1, 2, shaderRoundingModeRTZFloat32
);
2008 CORE_PROPERTY(1, 2, shaderSignedZeroInfNanPreserveFloat32
);
2009 CORE_PROPERTY(1, 2, shaderDenormFlushToZeroFloat64
);
2010 CORE_PROPERTY(1, 2, shaderDenormPreserveFloat64
);
2011 CORE_PROPERTY(1, 2, shaderRoundingModeRTEFloat64
);
2012 CORE_PROPERTY(1, 2, shaderRoundingModeRTZFloat64
);
2013 CORE_PROPERTY(1, 2, shaderSignedZeroInfNanPreserveFloat64
);
2017 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_TEXEL_BUFFER_ALIGNMENT_PROPERTIES_EXT
: {
2018 VkPhysicalDeviceTexelBufferAlignmentPropertiesEXT
*props
=
2019 (VkPhysicalDeviceTexelBufferAlignmentPropertiesEXT
*)ext
;
2021 /* From the SKL PRM Vol. 2d, docs for RENDER_SURFACE_STATE::Surface
2024 * "For SURFTYPE_BUFFER non-rendertarget surfaces, this field
2025 * specifies the base address of the first element of the surface,
2026 * computed in software by adding the surface base address to the
2027 * byte offset of the element in the buffer. The base address must
2028 * be aligned to element size."
2030 * The typed dataport messages require that things be texel aligned.
2031 * Otherwise, we may just load/store the wrong data or, in the worst
2032 * case, there may be hangs.
2034 props
->storageTexelBufferOffsetAlignmentBytes
= 16;
2035 props
->storageTexelBufferOffsetSingleTexelAlignment
= true;
2037 /* The sampler, however, is much more forgiving and it can handle
2038 * arbitrary byte alignment for linear and buffer surfaces. It's
2039 * hard to find a good PRM citation for this but years of empirical
2040 * experience demonstrate that this is true.
2042 props
->uniformTexelBufferOffsetAlignmentBytes
= 1;
2043 props
->uniformTexelBufferOffsetSingleTexelAlignment
= false;
2047 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_TIMELINE_SEMAPHORE_PROPERTIES_KHR
: {
2048 VkPhysicalDeviceTimelineSemaphorePropertiesKHR
*properties
=
2049 (VkPhysicalDeviceTimelineSemaphorePropertiesKHR
*) ext
;
2050 CORE_PROPERTY(1, 2, maxTimelineSemaphoreValueDifference
);
2054 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_TRANSFORM_FEEDBACK_PROPERTIES_EXT
: {
2055 VkPhysicalDeviceTransformFeedbackPropertiesEXT
*props
=
2056 (VkPhysicalDeviceTransformFeedbackPropertiesEXT
*)ext
;
2058 props
->maxTransformFeedbackStreams
= MAX_XFB_STREAMS
;
2059 props
->maxTransformFeedbackBuffers
= MAX_XFB_BUFFERS
;
2060 props
->maxTransformFeedbackBufferSize
= (1ull << 32);
2061 props
->maxTransformFeedbackStreamDataSize
= 128 * 4;
2062 props
->maxTransformFeedbackBufferDataSize
= 128 * 4;
2063 props
->maxTransformFeedbackBufferDataStride
= 2048;
2064 props
->transformFeedbackQueries
= true;
2065 props
->transformFeedbackStreamsLinesTriangles
= false;
2066 props
->transformFeedbackRasterizationStreamSelect
= false;
2067 props
->transformFeedbackDraw
= true;
2071 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VERTEX_ATTRIBUTE_DIVISOR_PROPERTIES_EXT
: {
2072 VkPhysicalDeviceVertexAttributeDivisorPropertiesEXT
*props
=
2073 (VkPhysicalDeviceVertexAttributeDivisorPropertiesEXT
*)ext
;
2074 /* We have to restrict this a bit for multiview */
2075 props
->maxVertexAttribDivisor
= UINT32_MAX
/ 16;
2079 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_1_PROPERTIES
:
2080 anv_get_physical_device_properties_1_1(pdevice
, (void *)ext
);
2083 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_VULKAN_1_2_PROPERTIES
:
2084 anv_get_physical_device_properties_1_2(pdevice
, (void *)ext
);
2088 anv_debug_ignored_stype(ext
->sType
);
2093 #undef CORE_RENAMED_PROPERTY
2094 #undef CORE_PROPERTY
2097 /* We support exactly one queue family. */
2098 static const VkQueueFamilyProperties
2099 anv_queue_family_properties
= {
2100 .queueFlags
= VK_QUEUE_GRAPHICS_BIT
|
2101 VK_QUEUE_COMPUTE_BIT
|
2102 VK_QUEUE_TRANSFER_BIT
,
2104 .timestampValidBits
= 36, /* XXX: Real value here */
2105 .minImageTransferGranularity
= { 1, 1, 1 },
2108 void anv_GetPhysicalDeviceQueueFamilyProperties(
2109 VkPhysicalDevice physicalDevice
,
2111 VkQueueFamilyProperties
* pQueueFamilyProperties
)
2113 VK_OUTARRAY_MAKE(out
, pQueueFamilyProperties
, pCount
);
2115 vk_outarray_append(&out
, p
) {
2116 *p
= anv_queue_family_properties
;
2120 void anv_GetPhysicalDeviceQueueFamilyProperties2(
2121 VkPhysicalDevice physicalDevice
,
2122 uint32_t* pQueueFamilyPropertyCount
,
2123 VkQueueFamilyProperties2
* pQueueFamilyProperties
)
2126 VK_OUTARRAY_MAKE(out
, pQueueFamilyProperties
, pQueueFamilyPropertyCount
);
2128 vk_outarray_append(&out
, p
) {
2129 p
->queueFamilyProperties
= anv_queue_family_properties
;
2131 vk_foreach_struct(s
, p
->pNext
) {
2132 anv_debug_ignored_stype(s
->sType
);
2137 void anv_GetPhysicalDeviceMemoryProperties(
2138 VkPhysicalDevice physicalDevice
,
2139 VkPhysicalDeviceMemoryProperties
* pMemoryProperties
)
2141 ANV_FROM_HANDLE(anv_physical_device
, physical_device
, physicalDevice
);
2143 pMemoryProperties
->memoryTypeCount
= physical_device
->memory
.type_count
;
2144 for (uint32_t i
= 0; i
< physical_device
->memory
.type_count
; i
++) {
2145 pMemoryProperties
->memoryTypes
[i
] = (VkMemoryType
) {
2146 .propertyFlags
= physical_device
->memory
.types
[i
].propertyFlags
,
2147 .heapIndex
= physical_device
->memory
.types
[i
].heapIndex
,
2151 pMemoryProperties
->memoryHeapCount
= physical_device
->memory
.heap_count
;
2152 for (uint32_t i
= 0; i
< physical_device
->memory
.heap_count
; i
++) {
2153 pMemoryProperties
->memoryHeaps
[i
] = (VkMemoryHeap
) {
2154 .size
= physical_device
->memory
.heaps
[i
].size
,
2155 .flags
= physical_device
->memory
.heaps
[i
].flags
,
2161 anv_get_memory_budget(VkPhysicalDevice physicalDevice
,
2162 VkPhysicalDeviceMemoryBudgetPropertiesEXT
*memoryBudget
)
2164 ANV_FROM_HANDLE(anv_physical_device
, device
, physicalDevice
);
2165 uint64_t sys_available
= get_available_system_memory();
2166 assert(sys_available
> 0);
2168 VkDeviceSize total_heaps_size
= 0;
2169 for (size_t i
= 0; i
< device
->memory
.heap_count
; i
++)
2170 total_heaps_size
+= device
->memory
.heaps
[i
].size
;
2172 for (size_t i
= 0; i
< device
->memory
.heap_count
; i
++) {
2173 VkDeviceSize heap_size
= device
->memory
.heaps
[i
].size
;
2174 VkDeviceSize heap_used
= device
->memory
.heaps
[i
].used
;
2175 VkDeviceSize heap_budget
;
2177 double heap_proportion
= (double) heap_size
/ total_heaps_size
;
2178 VkDeviceSize sys_available_prop
= sys_available
* heap_proportion
;
2181 * Let's not incite the app to starve the system: report at most 90% of
2182 * available system memory.
2184 uint64_t heap_available
= sys_available_prop
* 9 / 10;
2185 heap_budget
= MIN2(heap_size
, heap_used
+ heap_available
);
2188 * Round down to the nearest MB
2190 heap_budget
&= ~((1ull << 20) - 1);
2193 * The heapBudget value must be non-zero for array elements less than
2194 * VkPhysicalDeviceMemoryProperties::memoryHeapCount. The heapBudget
2195 * value must be less than or equal to VkMemoryHeap::size for each heap.
2197 assert(0 < heap_budget
&& heap_budget
<= heap_size
);
2199 memoryBudget
->heapUsage
[i
] = heap_used
;
2200 memoryBudget
->heapBudget
[i
] = heap_budget
;
2203 /* The heapBudget and heapUsage values must be zero for array elements
2204 * greater than or equal to VkPhysicalDeviceMemoryProperties::memoryHeapCount
2206 for (uint32_t i
= device
->memory
.heap_count
; i
< VK_MAX_MEMORY_HEAPS
; i
++) {
2207 memoryBudget
->heapBudget
[i
] = 0;
2208 memoryBudget
->heapUsage
[i
] = 0;
2212 void anv_GetPhysicalDeviceMemoryProperties2(
2213 VkPhysicalDevice physicalDevice
,
2214 VkPhysicalDeviceMemoryProperties2
* pMemoryProperties
)
2216 anv_GetPhysicalDeviceMemoryProperties(physicalDevice
,
2217 &pMemoryProperties
->memoryProperties
);
2219 vk_foreach_struct(ext
, pMemoryProperties
->pNext
) {
2220 switch (ext
->sType
) {
2221 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_MEMORY_BUDGET_PROPERTIES_EXT
:
2222 anv_get_memory_budget(physicalDevice
, (void*)ext
);
2225 anv_debug_ignored_stype(ext
->sType
);
2232 anv_GetDeviceGroupPeerMemoryFeatures(
2235 uint32_t localDeviceIndex
,
2236 uint32_t remoteDeviceIndex
,
2237 VkPeerMemoryFeatureFlags
* pPeerMemoryFeatures
)
2239 assert(localDeviceIndex
== 0 && remoteDeviceIndex
== 0);
2240 *pPeerMemoryFeatures
= VK_PEER_MEMORY_FEATURE_COPY_SRC_BIT
|
2241 VK_PEER_MEMORY_FEATURE_COPY_DST_BIT
|
2242 VK_PEER_MEMORY_FEATURE_GENERIC_SRC_BIT
|
2243 VK_PEER_MEMORY_FEATURE_GENERIC_DST_BIT
;
2246 PFN_vkVoidFunction
anv_GetInstanceProcAddr(
2247 VkInstance _instance
,
2250 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
2252 /* The Vulkan 1.0 spec for vkGetInstanceProcAddr has a table of exactly
2253 * when we have to return valid function pointers, NULL, or it's left
2254 * undefined. See the table for exact details.
2259 #define LOOKUP_ANV_ENTRYPOINT(entrypoint) \
2260 if (strcmp(pName, "vk" #entrypoint) == 0) \
2261 return (PFN_vkVoidFunction)anv_##entrypoint
2263 LOOKUP_ANV_ENTRYPOINT(EnumerateInstanceExtensionProperties
);
2264 LOOKUP_ANV_ENTRYPOINT(EnumerateInstanceLayerProperties
);
2265 LOOKUP_ANV_ENTRYPOINT(EnumerateInstanceVersion
);
2266 LOOKUP_ANV_ENTRYPOINT(CreateInstance
);
2268 /* GetInstanceProcAddr() can also be called with a NULL instance.
2269 * See https://gitlab.khronos.org/vulkan/vulkan/issues/2057
2271 LOOKUP_ANV_ENTRYPOINT(GetInstanceProcAddr
);
2273 #undef LOOKUP_ANV_ENTRYPOINT
2275 if (instance
== NULL
)
2278 int idx
= anv_get_instance_entrypoint_index(pName
);
2280 return instance
->dispatch
.entrypoints
[idx
];
2282 idx
= anv_get_physical_device_entrypoint_index(pName
);
2284 return instance
->physical_device_dispatch
.entrypoints
[idx
];
2286 idx
= anv_get_device_entrypoint_index(pName
);
2288 return instance
->device_dispatch
.entrypoints
[idx
];
2293 /* With version 1+ of the loader interface the ICD should expose
2294 * vk_icdGetInstanceProcAddr to work around certain LD_PRELOAD issues seen in apps.
2297 VKAPI_ATTR PFN_vkVoidFunction VKAPI_CALL
vk_icdGetInstanceProcAddr(
2298 VkInstance instance
,
2302 VKAPI_ATTR PFN_vkVoidFunction VKAPI_CALL
vk_icdGetInstanceProcAddr(
2303 VkInstance instance
,
2306 return anv_GetInstanceProcAddr(instance
, pName
);
2309 PFN_vkVoidFunction
anv_GetDeviceProcAddr(
2313 ANV_FROM_HANDLE(anv_device
, device
, _device
);
2315 if (!device
|| !pName
)
2318 int idx
= anv_get_device_entrypoint_index(pName
);
2322 return device
->dispatch
.entrypoints
[idx
];
2325 /* With version 4+ of the loader interface the ICD should expose
2326 * vk_icdGetPhysicalDeviceProcAddr()
2329 VKAPI_ATTR PFN_vkVoidFunction VKAPI_CALL
vk_icdGetPhysicalDeviceProcAddr(
2330 VkInstance _instance
,
2333 PFN_vkVoidFunction
vk_icdGetPhysicalDeviceProcAddr(
2334 VkInstance _instance
,
2337 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
2339 if (!pName
|| !instance
)
2342 int idx
= anv_get_physical_device_entrypoint_index(pName
);
2346 return instance
->physical_device_dispatch
.entrypoints
[idx
];
2351 anv_CreateDebugReportCallbackEXT(VkInstance _instance
,
2352 const VkDebugReportCallbackCreateInfoEXT
* pCreateInfo
,
2353 const VkAllocationCallbacks
* pAllocator
,
2354 VkDebugReportCallbackEXT
* pCallback
)
2356 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
2357 return vk_create_debug_report_callback(&instance
->debug_report_callbacks
,
2358 pCreateInfo
, pAllocator
, &instance
->alloc
,
2363 anv_DestroyDebugReportCallbackEXT(VkInstance _instance
,
2364 VkDebugReportCallbackEXT _callback
,
2365 const VkAllocationCallbacks
* pAllocator
)
2367 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
2368 vk_destroy_debug_report_callback(&instance
->debug_report_callbacks
,
2369 _callback
, pAllocator
, &instance
->alloc
);
2373 anv_DebugReportMessageEXT(VkInstance _instance
,
2374 VkDebugReportFlagsEXT flags
,
2375 VkDebugReportObjectTypeEXT objectType
,
2378 int32_t messageCode
,
2379 const char* pLayerPrefix
,
2380 const char* pMessage
)
2382 ANV_FROM_HANDLE(anv_instance
, instance
, _instance
);
2383 vk_debug_report(&instance
->debug_report_callbacks
, flags
, objectType
,
2384 object
, location
, messageCode
, pLayerPrefix
, pMessage
);
2387 static struct anv_state
2388 anv_state_pool_emit_data(struct anv_state_pool
*pool
, size_t size
, size_t align
, const void *p
)
2390 struct anv_state state
;
2392 state
= anv_state_pool_alloc(pool
, size
, align
);
2393 memcpy(state
.map
, p
, size
);
2399 anv_device_init_border_colors(struct anv_device
*device
)
2401 if (device
->info
.is_haswell
) {
2402 static const struct hsw_border_color border_colors
[] = {
2403 [VK_BORDER_COLOR_FLOAT_TRANSPARENT_BLACK
] = { .float32
= { 0.0, 0.0, 0.0, 0.0 } },
2404 [VK_BORDER_COLOR_FLOAT_OPAQUE_BLACK
] = { .float32
= { 0.0, 0.0, 0.0, 1.0 } },
2405 [VK_BORDER_COLOR_FLOAT_OPAQUE_WHITE
] = { .float32
= { 1.0, 1.0, 1.0, 1.0 } },
2406 [VK_BORDER_COLOR_INT_TRANSPARENT_BLACK
] = { .uint32
= { 0, 0, 0, 0 } },
2407 [VK_BORDER_COLOR_INT_OPAQUE_BLACK
] = { .uint32
= { 0, 0, 0, 1 } },
2408 [VK_BORDER_COLOR_INT_OPAQUE_WHITE
] = { .uint32
= { 1, 1, 1, 1 } },
2411 device
->border_colors
=
2412 anv_state_pool_emit_data(&device
->dynamic_state_pool
,
2413 sizeof(border_colors
), 512, border_colors
);
2415 static const struct gen8_border_color border_colors
[] = {
2416 [VK_BORDER_COLOR_FLOAT_TRANSPARENT_BLACK
] = { .float32
= { 0.0, 0.0, 0.0, 0.0 } },
2417 [VK_BORDER_COLOR_FLOAT_OPAQUE_BLACK
] = { .float32
= { 0.0, 0.0, 0.0, 1.0 } },
2418 [VK_BORDER_COLOR_FLOAT_OPAQUE_WHITE
] = { .float32
= { 1.0, 1.0, 1.0, 1.0 } },
2419 [VK_BORDER_COLOR_INT_TRANSPARENT_BLACK
] = { .uint32
= { 0, 0, 0, 0 } },
2420 [VK_BORDER_COLOR_INT_OPAQUE_BLACK
] = { .uint32
= { 0, 0, 0, 1 } },
2421 [VK_BORDER_COLOR_INT_OPAQUE_WHITE
] = { .uint32
= { 1, 1, 1, 1 } },
2424 device
->border_colors
=
2425 anv_state_pool_emit_data(&device
->dynamic_state_pool
,
2426 sizeof(border_colors
), 64, border_colors
);
2431 anv_device_init_trivial_batch(struct anv_device
*device
)
2433 VkResult result
= anv_device_alloc_bo(device
, 4096,
2434 ANV_BO_ALLOC_MAPPED
,
2435 0 /* explicit_address */,
2436 &device
->trivial_batch_bo
);
2437 if (result
!= VK_SUCCESS
)
2440 struct anv_batch batch
= {
2441 .start
= device
->trivial_batch_bo
->map
,
2442 .next
= device
->trivial_batch_bo
->map
,
2443 .end
= device
->trivial_batch_bo
->map
+ 4096,
2446 anv_batch_emit(&batch
, GEN7_MI_BATCH_BUFFER_END
, bbe
);
2447 anv_batch_emit(&batch
, GEN7_MI_NOOP
, noop
);
2449 if (!device
->info
.has_llc
)
2450 gen_clflush_range(batch
.start
, batch
.next
- batch
.start
);
2455 VkResult
anv_EnumerateDeviceExtensionProperties(
2456 VkPhysicalDevice physicalDevice
,
2457 const char* pLayerName
,
2458 uint32_t* pPropertyCount
,
2459 VkExtensionProperties
* pProperties
)
2461 ANV_FROM_HANDLE(anv_physical_device
, device
, physicalDevice
);
2462 VK_OUTARRAY_MAKE(out
, pProperties
, pPropertyCount
);
2464 for (int i
= 0; i
< ANV_DEVICE_EXTENSION_COUNT
; i
++) {
2465 if (device
->supported_extensions
.extensions
[i
]) {
2466 vk_outarray_append(&out
, prop
) {
2467 *prop
= anv_device_extensions
[i
];
2472 return vk_outarray_status(&out
);
2476 anv_device_init_dispatch(struct anv_device
*device
)
2478 const struct anv_instance
*instance
= device
->physical
->instance
;
2480 const struct anv_device_dispatch_table
*genX_table
;
2481 switch (device
->info
.gen
) {
2483 genX_table
= &gen12_device_dispatch_table
;
2486 genX_table
= &gen11_device_dispatch_table
;
2489 genX_table
= &gen10_device_dispatch_table
;
2492 genX_table
= &gen9_device_dispatch_table
;
2495 genX_table
= &gen8_device_dispatch_table
;
2498 if (device
->info
.is_haswell
)
2499 genX_table
= &gen75_device_dispatch_table
;
2501 genX_table
= &gen7_device_dispatch_table
;
2504 unreachable("unsupported gen\n");
2507 for (unsigned i
= 0; i
< ARRAY_SIZE(device
->dispatch
.entrypoints
); i
++) {
2508 /* Vulkan requires that entrypoints for extensions which have not been
2509 * enabled must not be advertised.
2511 if (!anv_device_entrypoint_is_enabled(i
, instance
->app_info
.api_version
,
2512 &instance
->enabled_extensions
,
2513 &device
->enabled_extensions
)) {
2514 device
->dispatch
.entrypoints
[i
] = NULL
;
2515 } else if (genX_table
->entrypoints
[i
]) {
2516 device
->dispatch
.entrypoints
[i
] = genX_table
->entrypoints
[i
];
2518 device
->dispatch
.entrypoints
[i
] =
2519 anv_device_dispatch_table
.entrypoints
[i
];
2525 vk_priority_to_gen(int priority
)
2528 case VK_QUEUE_GLOBAL_PRIORITY_LOW_EXT
:
2529 return GEN_CONTEXT_LOW_PRIORITY
;
2530 case VK_QUEUE_GLOBAL_PRIORITY_MEDIUM_EXT
:
2531 return GEN_CONTEXT_MEDIUM_PRIORITY
;
2532 case VK_QUEUE_GLOBAL_PRIORITY_HIGH_EXT
:
2533 return GEN_CONTEXT_HIGH_PRIORITY
;
2534 case VK_QUEUE_GLOBAL_PRIORITY_REALTIME_EXT
:
2535 return GEN_CONTEXT_REALTIME_PRIORITY
;
2537 unreachable("Invalid priority");
2542 anv_device_init_hiz_clear_value_bo(struct anv_device
*device
)
2544 VkResult result
= anv_device_alloc_bo(device
, 4096,
2545 ANV_BO_ALLOC_MAPPED
,
2546 0 /* explicit_address */,
2547 &device
->hiz_clear_bo
);
2548 if (result
!= VK_SUCCESS
)
2551 union isl_color_value hiz_clear
= { .u32
= { 0, } };
2552 hiz_clear
.f32
[0] = ANV_HZ_FC_VAL
;
2554 memcpy(device
->hiz_clear_bo
->map
, hiz_clear
.u32
, sizeof(hiz_clear
.u32
));
2556 if (!device
->info
.has_llc
)
2557 gen_clflush_range(device
->hiz_clear_bo
->map
, sizeof(hiz_clear
.u32
));
2563 get_bo_from_pool(struct gen_batch_decode_bo
*ret
,
2564 struct anv_block_pool
*pool
,
2567 anv_block_pool_foreach_bo(bo
, pool
) {
2568 uint64_t bo_address
= gen_48b_address(bo
->offset
);
2569 if (address
>= bo_address
&& address
< (bo_address
+ bo
->size
)) {
2570 *ret
= (struct gen_batch_decode_bo
) {
2581 /* Finding a buffer for batch decoding */
2582 static struct gen_batch_decode_bo
2583 decode_get_bo(void *v_batch
, bool ppgtt
, uint64_t address
)
2585 struct anv_device
*device
= v_batch
;
2586 struct gen_batch_decode_bo ret_bo
= {};
2590 if (get_bo_from_pool(&ret_bo
, &device
->dynamic_state_pool
.block_pool
, address
))
2592 if (get_bo_from_pool(&ret_bo
, &device
->instruction_state_pool
.block_pool
, address
))
2594 if (get_bo_from_pool(&ret_bo
, &device
->binding_table_pool
.block_pool
, address
))
2596 if (get_bo_from_pool(&ret_bo
, &device
->surface_state_pool
.block_pool
, address
))
2599 if (!device
->cmd_buffer_being_decoded
)
2600 return (struct gen_batch_decode_bo
) { };
2602 struct anv_batch_bo
**bo
;
2604 u_vector_foreach(bo
, &device
->cmd_buffer_being_decoded
->seen_bbos
) {
2605 /* The decoder zeroes out the top 16 bits, so we need to as well */
2606 uint64_t bo_address
= (*bo
)->bo
->offset
& (~0ull >> 16);
2608 if (address
>= bo_address
&& address
< bo_address
+ (*bo
)->bo
->size
) {
2609 return (struct gen_batch_decode_bo
) {
2611 .size
= (*bo
)->bo
->size
,
2612 .map
= (*bo
)->bo
->map
,
2617 return (struct gen_batch_decode_bo
) { };
2620 struct gen_aux_map_buffer
{
2621 struct gen_buffer base
;
2622 struct anv_state state
;
2625 static struct gen_buffer
*
2626 gen_aux_map_buffer_alloc(void *driver_ctx
, uint32_t size
)
2628 struct gen_aux_map_buffer
*buf
= malloc(sizeof(struct gen_aux_map_buffer
));
2632 struct anv_device
*device
= (struct anv_device
*)driver_ctx
;
2633 assert(device
->physical
->supports_48bit_addresses
&&
2634 device
->physical
->use_softpin
);
2636 struct anv_state_pool
*pool
= &device
->dynamic_state_pool
;
2637 buf
->state
= anv_state_pool_alloc(pool
, size
, size
);
2639 buf
->base
.gpu
= pool
->block_pool
.bo
->offset
+ buf
->state
.offset
;
2640 buf
->base
.gpu_end
= buf
->base
.gpu
+ buf
->state
.alloc_size
;
2641 buf
->base
.map
= buf
->state
.map
;
2642 buf
->base
.driver_bo
= &buf
->state
;
2647 gen_aux_map_buffer_free(void *driver_ctx
, struct gen_buffer
*buffer
)
2649 struct gen_aux_map_buffer
*buf
= (struct gen_aux_map_buffer
*)buffer
;
2650 struct anv_device
*device
= (struct anv_device
*)driver_ctx
;
2651 struct anv_state_pool
*pool
= &device
->dynamic_state_pool
;
2652 anv_state_pool_free(pool
, buf
->state
);
2656 static struct gen_mapped_pinned_buffer_alloc aux_map_allocator
= {
2657 .alloc
= gen_aux_map_buffer_alloc
,
2658 .free
= gen_aux_map_buffer_free
,
2662 check_physical_device_features(VkPhysicalDevice physicalDevice
,
2663 const VkPhysicalDeviceFeatures
*features
)
2665 VkPhysicalDeviceFeatures supported_features
;
2666 anv_GetPhysicalDeviceFeatures(physicalDevice
, &supported_features
);
2667 VkBool32
*supported_feature
= (VkBool32
*)&supported_features
;
2668 VkBool32
*enabled_feature
= (VkBool32
*)features
;
2669 unsigned num_features
= sizeof(VkPhysicalDeviceFeatures
) / sizeof(VkBool32
);
2670 for (uint32_t i
= 0; i
< num_features
; i
++) {
2671 if (enabled_feature
[i
] && !supported_feature
[i
])
2672 return vk_error(VK_ERROR_FEATURE_NOT_PRESENT
);
2678 VkResult
anv_CreateDevice(
2679 VkPhysicalDevice physicalDevice
,
2680 const VkDeviceCreateInfo
* pCreateInfo
,
2681 const VkAllocationCallbacks
* pAllocator
,
2684 ANV_FROM_HANDLE(anv_physical_device
, physical_device
, physicalDevice
);
2686 struct anv_device
*device
;
2688 assert(pCreateInfo
->sType
== VK_STRUCTURE_TYPE_DEVICE_CREATE_INFO
);
2690 struct anv_device_extension_table enabled_extensions
= { };
2691 for (uint32_t i
= 0; i
< pCreateInfo
->enabledExtensionCount
; i
++) {
2693 for (idx
= 0; idx
< ANV_DEVICE_EXTENSION_COUNT
; idx
++) {
2694 if (strcmp(pCreateInfo
->ppEnabledExtensionNames
[i
],
2695 anv_device_extensions
[idx
].extensionName
) == 0)
2699 if (idx
>= ANV_DEVICE_EXTENSION_COUNT
)
2700 return vk_error(VK_ERROR_EXTENSION_NOT_PRESENT
);
2702 if (!physical_device
->supported_extensions
.extensions
[idx
])
2703 return vk_error(VK_ERROR_EXTENSION_NOT_PRESENT
);
2705 enabled_extensions
.extensions
[idx
] = true;
2708 /* Check enabled features */
2709 bool robust_buffer_access
= false;
2710 if (pCreateInfo
->pEnabledFeatures
) {
2711 result
= check_physical_device_features(physicalDevice
,
2712 pCreateInfo
->pEnabledFeatures
);
2713 if (result
!= VK_SUCCESS
)
2716 if (pCreateInfo
->pEnabledFeatures
->robustBufferAccess
)
2717 robust_buffer_access
= true;
2720 vk_foreach_struct_const(ext
, pCreateInfo
->pNext
) {
2721 switch (ext
->sType
) {
2722 case VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_FEATURES_2
: {
2723 const VkPhysicalDeviceFeatures2
*features
= (const void *)ext
;
2724 result
= check_physical_device_features(physicalDevice
,
2725 &features
->features
);
2726 if (result
!= VK_SUCCESS
)
2729 if (features
->features
.robustBufferAccess
)
2730 robust_buffer_access
= true;
2740 /* Check requested queues and fail if we are requested to create any
2741 * queues with flags we don't support.
2743 assert(pCreateInfo
->queueCreateInfoCount
> 0);
2744 for (uint32_t i
= 0; i
< pCreateInfo
->queueCreateInfoCount
; i
++) {
2745 if (pCreateInfo
->pQueueCreateInfos
[i
].flags
!= 0)
2746 return vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2749 /* Check if client specified queue priority. */
2750 const VkDeviceQueueGlobalPriorityCreateInfoEXT
*queue_priority
=
2751 vk_find_struct_const(pCreateInfo
->pQueueCreateInfos
[0].pNext
,
2752 DEVICE_QUEUE_GLOBAL_PRIORITY_CREATE_INFO_EXT
);
2754 VkQueueGlobalPriorityEXT priority
=
2755 queue_priority
? queue_priority
->globalPriority
:
2756 VK_QUEUE_GLOBAL_PRIORITY_MEDIUM_EXT
;
2758 device
= vk_alloc2(&physical_device
->instance
->alloc
, pAllocator
,
2760 VK_SYSTEM_ALLOCATION_SCOPE_DEVICE
);
2762 return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
2764 vk_device_init(&device
->vk
, pCreateInfo
,
2765 &physical_device
->instance
->alloc
, pAllocator
);
2767 if (INTEL_DEBUG
& DEBUG_BATCH
) {
2768 const unsigned decode_flags
=
2769 GEN_BATCH_DECODE_FULL
|
2770 ((INTEL_DEBUG
& DEBUG_COLOR
) ? GEN_BATCH_DECODE_IN_COLOR
: 0) |
2771 GEN_BATCH_DECODE_OFFSETS
|
2772 GEN_BATCH_DECODE_FLOATS
;
2774 gen_batch_decode_ctx_init(&device
->decoder_ctx
,
2775 &physical_device
->info
,
2776 stderr
, decode_flags
, NULL
,
2777 decode_get_bo
, NULL
, device
);
2780 device
->physical
= physical_device
;
2781 device
->no_hw
= physical_device
->no_hw
;
2782 device
->_lost
= false;
2784 /* XXX(chadv): Can we dup() physicalDevice->fd here? */
2785 device
->fd
= open(physical_device
->path
, O_RDWR
| O_CLOEXEC
);
2786 if (device
->fd
== -1) {
2787 result
= vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2791 device
->context_id
= anv_gem_create_context(device
);
2792 if (device
->context_id
== -1) {
2793 result
= vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2797 result
= anv_queue_init(device
, &device
->queue
);
2798 if (result
!= VK_SUCCESS
)
2799 goto fail_context_id
;
2801 if (physical_device
->use_softpin
) {
2802 if (pthread_mutex_init(&device
->vma_mutex
, NULL
) != 0) {
2803 result
= vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2807 /* keep the page with address zero out of the allocator */
2808 util_vma_heap_init(&device
->vma_lo
,
2809 LOW_HEAP_MIN_ADDRESS
, LOW_HEAP_SIZE
);
2811 util_vma_heap_init(&device
->vma_cva
, CLIENT_VISIBLE_HEAP_MIN_ADDRESS
,
2812 CLIENT_VISIBLE_HEAP_SIZE
);
2814 /* Leave the last 4GiB out of the high vma range, so that no state
2815 * base address + size can overflow 48 bits. For more information see
2816 * the comment about Wa32bitGeneralStateOffset in anv_allocator.c
2818 util_vma_heap_init(&device
->vma_hi
, HIGH_HEAP_MIN_ADDRESS
,
2819 physical_device
->gtt_size
- (1ull << 32) -
2820 HIGH_HEAP_MIN_ADDRESS
);
2823 list_inithead(&device
->memory_objects
);
2825 /* As per spec, the driver implementation may deny requests to acquire
2826 * a priority above the default priority (MEDIUM) if the caller does not
2827 * have sufficient privileges. In this scenario VK_ERROR_NOT_PERMITTED_EXT
2830 if (physical_device
->has_context_priority
) {
2831 int err
= anv_gem_set_context_param(device
->fd
, device
->context_id
,
2832 I915_CONTEXT_PARAM_PRIORITY
,
2833 vk_priority_to_gen(priority
));
2834 if (err
!= 0 && priority
> VK_QUEUE_GLOBAL_PRIORITY_MEDIUM_EXT
) {
2835 result
= vk_error(VK_ERROR_NOT_PERMITTED_EXT
);
2840 device
->info
= physical_device
->info
;
2841 device
->isl_dev
= physical_device
->isl_dev
;
2843 /* On Broadwell and later, we can use batch chaining to more efficiently
2844 * implement growing command buffers. Prior to Haswell, the kernel
2845 * command parser gets in the way and we have to fall back to growing
2848 device
->can_chain_batches
= device
->info
.gen
>= 8;
2850 device
->robust_buffer_access
= robust_buffer_access
;
2851 device
->enabled_extensions
= enabled_extensions
;
2853 anv_device_init_dispatch(device
);
2855 if (pthread_mutex_init(&device
->mutex
, NULL
) != 0) {
2856 result
= vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2860 pthread_condattr_t condattr
;
2861 if (pthread_condattr_init(&condattr
) != 0) {
2862 result
= vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2865 if (pthread_condattr_setclock(&condattr
, CLOCK_MONOTONIC
) != 0) {
2866 pthread_condattr_destroy(&condattr
);
2867 result
= vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2870 if (pthread_cond_init(&device
->queue_submit
, &condattr
) != 0) {
2871 pthread_condattr_destroy(&condattr
);
2872 result
= vk_error(VK_ERROR_INITIALIZATION_FAILED
);
2875 pthread_condattr_destroy(&condattr
);
2877 result
= anv_bo_cache_init(&device
->bo_cache
);
2878 if (result
!= VK_SUCCESS
)
2879 goto fail_queue_cond
;
2881 anv_bo_pool_init(&device
->batch_bo_pool
, device
);
2883 result
= anv_state_pool_init(&device
->dynamic_state_pool
, device
,
2884 DYNAMIC_STATE_POOL_MIN_ADDRESS
, 0, 16384);
2885 if (result
!= VK_SUCCESS
)
2886 goto fail_batch_bo_pool
;
2888 if (device
->info
.gen
>= 8) {
2889 /* The border color pointer is limited to 24 bits, so we need to make
2890 * sure that any such color used at any point in the program doesn't
2891 * exceed that limit.
2892 * We achieve that by reserving all the custom border colors we support
2893 * right off the bat, so they are close to the base address.
2895 anv_state_reserved_pool_init(&device
->custom_border_colors
,
2896 &device
->dynamic_state_pool
,
2897 sizeof(struct gen8_border_color
),
2898 MAX_CUSTOM_BORDER_COLORS
, 64);
2901 result
= anv_state_pool_init(&device
->instruction_state_pool
, device
,
2902 INSTRUCTION_STATE_POOL_MIN_ADDRESS
, 0, 16384);
2903 if (result
!= VK_SUCCESS
)
2904 goto fail_dynamic_state_pool
;
2906 result
= anv_state_pool_init(&device
->surface_state_pool
, device
,
2907 SURFACE_STATE_POOL_MIN_ADDRESS
, 0, 4096);
2908 if (result
!= VK_SUCCESS
)
2909 goto fail_instruction_state_pool
;
2911 if (physical_device
->use_softpin
) {
2912 int64_t bt_pool_offset
= (int64_t)BINDING_TABLE_POOL_MIN_ADDRESS
-
2913 (int64_t)SURFACE_STATE_POOL_MIN_ADDRESS
;
2914 assert(INT32_MIN
< bt_pool_offset
&& bt_pool_offset
< 0);
2915 result
= anv_state_pool_init(&device
->binding_table_pool
, device
,
2916 SURFACE_STATE_POOL_MIN_ADDRESS
,
2917 bt_pool_offset
, 4096);
2918 if (result
!= VK_SUCCESS
)
2919 goto fail_surface_state_pool
;
2922 if (device
->info
.gen
>= 12) {
2923 device
->aux_map_ctx
= gen_aux_map_init(device
, &aux_map_allocator
,
2924 &physical_device
->info
);
2925 if (!device
->aux_map_ctx
)
2926 goto fail_binding_table_pool
;
2929 result
= anv_device_alloc_bo(device
, 4096,
2930 ANV_BO_ALLOC_CAPTURE
| ANV_BO_ALLOC_MAPPED
/* flags */,
2931 0 /* explicit_address */,
2932 &device
->workaround_bo
);
2933 if (result
!= VK_SUCCESS
)
2934 goto fail_surface_aux_map_pool
;
2936 device
->workaround_address
= (struct anv_address
) {
2937 .bo
= device
->workaround_bo
,
2938 .offset
= align_u32(
2939 intel_debug_write_identifiers(device
->workaround_bo
->map
,
2940 device
->workaround_bo
->size
,
2944 if (!device
->info
.has_llc
) {
2945 gen_clflush_range(device
->workaround_bo
->map
,
2946 device
->workaround_address
.offset
);
2949 result
= anv_device_init_trivial_batch(device
);
2950 if (result
!= VK_SUCCESS
)
2951 goto fail_workaround_bo
;
2953 /* Allocate a null surface state at surface state offset 0. This makes
2954 * NULL descriptor handling trivial because we can just memset structures
2955 * to zero and they have a valid descriptor.
2957 device
->null_surface_state
=
2958 anv_state_pool_alloc(&device
->surface_state_pool
,
2959 device
->isl_dev
.ss
.size
,
2960 device
->isl_dev
.ss
.align
);
2961 isl_null_fill_state(&device
->isl_dev
, device
->null_surface_state
.map
,
2962 isl_extent3d(1, 1, 1) /* This shouldn't matter */);
2963 assert(device
->null_surface_state
.offset
== 0);
2965 if (device
->info
.gen
>= 10) {
2966 result
= anv_device_init_hiz_clear_value_bo(device
);
2967 if (result
!= VK_SUCCESS
)
2968 goto fail_trivial_batch_bo
;
2971 anv_scratch_pool_init(device
, &device
->scratch_pool
);
2973 switch (device
->info
.gen
) {
2975 if (!device
->info
.is_haswell
)
2976 result
= gen7_init_device_state(device
);
2978 result
= gen75_init_device_state(device
);
2981 result
= gen8_init_device_state(device
);
2984 result
= gen9_init_device_state(device
);
2987 result
= gen10_init_device_state(device
);
2990 result
= gen11_init_device_state(device
);
2993 result
= gen12_init_device_state(device
);
2996 /* Shouldn't get here as we don't create physical devices for any other
2998 unreachable("unhandled gen");
3000 if (result
!= VK_SUCCESS
)
3001 goto fail_clear_value_bo
;
3003 anv_pipeline_cache_init(&device
->default_pipeline_cache
, device
, true);
3005 anv_device_init_blorp(device
);
3007 anv_device_init_border_colors(device
);
3009 anv_device_perf_init(device
);
3011 *pDevice
= anv_device_to_handle(device
);
3015 fail_clear_value_bo
:
3016 if (device
->info
.gen
>= 10)
3017 anv_device_release_bo(device
, device
->hiz_clear_bo
);
3018 anv_scratch_pool_finish(device
, &device
->scratch_pool
);
3019 fail_trivial_batch_bo
:
3020 anv_device_release_bo(device
, device
->trivial_batch_bo
);
3022 anv_device_release_bo(device
, device
->workaround_bo
);
3023 fail_surface_aux_map_pool
:
3024 if (device
->info
.gen
>= 12) {
3025 gen_aux_map_finish(device
->aux_map_ctx
);
3026 device
->aux_map_ctx
= NULL
;
3028 fail_binding_table_pool
:
3029 if (physical_device
->use_softpin
)
3030 anv_state_pool_finish(&device
->binding_table_pool
);
3031 fail_surface_state_pool
:
3032 anv_state_pool_finish(&device
->surface_state_pool
);
3033 fail_instruction_state_pool
:
3034 anv_state_pool_finish(&device
->instruction_state_pool
);
3035 fail_dynamic_state_pool
:
3036 if (device
->info
.gen
>= 8)
3037 anv_state_reserved_pool_finish(&device
->custom_border_colors
);
3038 anv_state_pool_finish(&device
->dynamic_state_pool
);
3040 anv_bo_pool_finish(&device
->batch_bo_pool
);
3041 anv_bo_cache_finish(&device
->bo_cache
);
3043 pthread_cond_destroy(&device
->queue_submit
);
3045 pthread_mutex_destroy(&device
->mutex
);
3047 if (physical_device
->use_softpin
) {
3048 util_vma_heap_finish(&device
->vma_hi
);
3049 util_vma_heap_finish(&device
->vma_cva
);
3050 util_vma_heap_finish(&device
->vma_lo
);
3053 anv_queue_finish(&device
->queue
);
3055 anv_gem_destroy_context(device
, device
->context_id
);
3059 vk_free(&device
->vk
.alloc
, device
);
3064 void anv_DestroyDevice(
3066 const VkAllocationCallbacks
* pAllocator
)
3068 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3073 anv_device_finish_blorp(device
);
3075 anv_pipeline_cache_finish(&device
->default_pipeline_cache
);
3077 anv_queue_finish(&device
->queue
);
3079 #ifdef HAVE_VALGRIND
3080 /* We only need to free these to prevent valgrind errors. The backing
3081 * BO will go away in a couple of lines so we don't actually leak.
3083 if (device
->info
.gen
>= 8)
3084 anv_state_reserved_pool_finish(&device
->custom_border_colors
);
3085 anv_state_pool_free(&device
->dynamic_state_pool
, device
->border_colors
);
3086 anv_state_pool_free(&device
->dynamic_state_pool
, device
->slice_hash
);
3089 anv_scratch_pool_finish(device
, &device
->scratch_pool
);
3091 anv_device_release_bo(device
, device
->workaround_bo
);
3092 anv_device_release_bo(device
, device
->trivial_batch_bo
);
3093 if (device
->info
.gen
>= 10)
3094 anv_device_release_bo(device
, device
->hiz_clear_bo
);
3096 if (device
->info
.gen
>= 12) {
3097 gen_aux_map_finish(device
->aux_map_ctx
);
3098 device
->aux_map_ctx
= NULL
;
3101 if (device
->physical
->use_softpin
)
3102 anv_state_pool_finish(&device
->binding_table_pool
);
3103 anv_state_pool_finish(&device
->surface_state_pool
);
3104 anv_state_pool_finish(&device
->instruction_state_pool
);
3105 anv_state_pool_finish(&device
->dynamic_state_pool
);
3107 anv_bo_pool_finish(&device
->batch_bo_pool
);
3109 anv_bo_cache_finish(&device
->bo_cache
);
3111 if (device
->physical
->use_softpin
) {
3112 util_vma_heap_finish(&device
->vma_hi
);
3113 util_vma_heap_finish(&device
->vma_cva
);
3114 util_vma_heap_finish(&device
->vma_lo
);
3117 pthread_cond_destroy(&device
->queue_submit
);
3118 pthread_mutex_destroy(&device
->mutex
);
3120 anv_gem_destroy_context(device
, device
->context_id
);
3122 if (INTEL_DEBUG
& DEBUG_BATCH
)
3123 gen_batch_decode_ctx_finish(&device
->decoder_ctx
);
3127 vk_device_finish(&device
->vk
);
3128 vk_free(&device
->vk
.alloc
, device
);
3131 VkResult
anv_EnumerateInstanceLayerProperties(
3132 uint32_t* pPropertyCount
,
3133 VkLayerProperties
* pProperties
)
3135 if (pProperties
== NULL
) {
3136 *pPropertyCount
= 0;
3140 /* None supported at this time */
3141 return vk_error(VK_ERROR_LAYER_NOT_PRESENT
);
3144 VkResult
anv_EnumerateDeviceLayerProperties(
3145 VkPhysicalDevice physicalDevice
,
3146 uint32_t* pPropertyCount
,
3147 VkLayerProperties
* pProperties
)
3149 if (pProperties
== NULL
) {
3150 *pPropertyCount
= 0;
3154 /* None supported at this time */
3155 return vk_error(VK_ERROR_LAYER_NOT_PRESENT
);
3158 void anv_GetDeviceQueue(
3160 uint32_t queueNodeIndex
,
3161 uint32_t queueIndex
,
3164 const VkDeviceQueueInfo2 info
= {
3165 .sType
= VK_STRUCTURE_TYPE_DEVICE_QUEUE_INFO_2
,
3168 .queueFamilyIndex
= queueNodeIndex
,
3169 .queueIndex
= queueIndex
,
3172 anv_GetDeviceQueue2(_device
, &info
, pQueue
);
3175 void anv_GetDeviceQueue2(
3177 const VkDeviceQueueInfo2
* pQueueInfo
,
3180 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3182 assert(pQueueInfo
->queueIndex
== 0);
3184 if (pQueueInfo
->flags
== device
->queue
.flags
)
3185 *pQueue
= anv_queue_to_handle(&device
->queue
);
3191 _anv_device_set_lost(struct anv_device
*device
,
3192 const char *file
, int line
,
3193 const char *msg
, ...)
3198 p_atomic_inc(&device
->_lost
);
3201 err
= __vk_errorv(device
->physical
->instance
, device
,
3202 VK_DEBUG_REPORT_OBJECT_TYPE_DEVICE_EXT
,
3203 VK_ERROR_DEVICE_LOST
, file
, line
, msg
, ap
);
3206 if (env_var_as_boolean("ANV_ABORT_ON_DEVICE_LOSS", false))
3213 _anv_queue_set_lost(struct anv_queue
*queue
,
3214 const char *file
, int line
,
3215 const char *msg
, ...)
3220 p_atomic_inc(&queue
->device
->_lost
);
3223 err
= __vk_errorv(queue
->device
->physical
->instance
, queue
->device
,
3224 VK_DEBUG_REPORT_OBJECT_TYPE_DEVICE_EXT
,
3225 VK_ERROR_DEVICE_LOST
, file
, line
, msg
, ap
);
3228 if (env_var_as_boolean("ANV_ABORT_ON_DEVICE_LOSS", false))
3235 anv_device_query_status(struct anv_device
*device
)
3237 /* This isn't likely as most of the callers of this function already check
3238 * for it. However, it doesn't hurt to check and it potentially lets us
3241 if (anv_device_is_lost(device
))
3242 return VK_ERROR_DEVICE_LOST
;
3244 uint32_t active
, pending
;
3245 int ret
= anv_gem_gpu_get_reset_stats(device
, &active
, &pending
);
3247 /* We don't know the real error. */
3248 return anv_device_set_lost(device
, "get_reset_stats failed: %m");
3252 return anv_device_set_lost(device
, "GPU hung on one of our command buffers");
3253 } else if (pending
) {
3254 return anv_device_set_lost(device
, "GPU hung with commands in-flight");
3261 anv_device_bo_busy(struct anv_device
*device
, struct anv_bo
*bo
)
3263 /* Note: This only returns whether or not the BO is in use by an i915 GPU.
3264 * Other usages of the BO (such as on different hardware) will not be
3265 * flagged as "busy" by this ioctl. Use with care.
3267 int ret
= anv_gem_busy(device
, bo
->gem_handle
);
3269 return VK_NOT_READY
;
3270 } else if (ret
== -1) {
3271 /* We don't know the real error. */
3272 return anv_device_set_lost(device
, "gem wait failed: %m");
3275 /* Query for device status after the busy call. If the BO we're checking
3276 * got caught in a GPU hang we don't want to return VK_SUCCESS to the
3277 * client because it clearly doesn't have valid data. Yes, this most
3278 * likely means an ioctl, but we just did an ioctl to query the busy status
3279 * so it's no great loss.
3281 return anv_device_query_status(device
);
3285 anv_device_wait(struct anv_device
*device
, struct anv_bo
*bo
,
3288 int ret
= anv_gem_wait(device
, bo
->gem_handle
, &timeout
);
3289 if (ret
== -1 && errno
== ETIME
) {
3291 } else if (ret
== -1) {
3292 /* We don't know the real error. */
3293 return anv_device_set_lost(device
, "gem wait failed: %m");
3296 /* Query for device status after the wait. If the BO we're waiting on got
3297 * caught in a GPU hang we don't want to return VK_SUCCESS to the client
3298 * because it clearly doesn't have valid data. Yes, this most likely means
3299 * an ioctl, but we just did an ioctl to wait so it's no great loss.
3301 return anv_device_query_status(device
);
3304 VkResult
anv_DeviceWaitIdle(
3307 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3309 if (anv_device_is_lost(device
))
3310 return VK_ERROR_DEVICE_LOST
;
3312 return anv_queue_submit_simple_batch(&device
->queue
, NULL
);
3316 anv_vma_alloc(struct anv_device
*device
,
3317 uint64_t size
, uint64_t align
,
3318 enum anv_bo_alloc_flags alloc_flags
,
3319 uint64_t client_address
)
3321 pthread_mutex_lock(&device
->vma_mutex
);
3325 if (alloc_flags
& ANV_BO_ALLOC_CLIENT_VISIBLE_ADDRESS
) {
3326 if (client_address
) {
3327 if (util_vma_heap_alloc_addr(&device
->vma_cva
,
3328 client_address
, size
)) {
3329 addr
= client_address
;
3332 addr
= util_vma_heap_alloc(&device
->vma_cva
, size
, align
);
3334 /* We don't want to fall back to other heaps */
3338 assert(client_address
== 0);
3340 if (!(alloc_flags
& ANV_BO_ALLOC_32BIT_ADDRESS
))
3341 addr
= util_vma_heap_alloc(&device
->vma_hi
, size
, align
);
3344 addr
= util_vma_heap_alloc(&device
->vma_lo
, size
, align
);
3347 pthread_mutex_unlock(&device
->vma_mutex
);
3349 assert(addr
== gen_48b_address(addr
));
3350 return gen_canonical_address(addr
);
3354 anv_vma_free(struct anv_device
*device
,
3355 uint64_t address
, uint64_t size
)
3357 const uint64_t addr_48b
= gen_48b_address(address
);
3359 pthread_mutex_lock(&device
->vma_mutex
);
3361 if (addr_48b
>= LOW_HEAP_MIN_ADDRESS
&&
3362 addr_48b
<= LOW_HEAP_MAX_ADDRESS
) {
3363 util_vma_heap_free(&device
->vma_lo
, addr_48b
, size
);
3364 } else if (addr_48b
>= CLIENT_VISIBLE_HEAP_MIN_ADDRESS
&&
3365 addr_48b
<= CLIENT_VISIBLE_HEAP_MAX_ADDRESS
) {
3366 util_vma_heap_free(&device
->vma_cva
, addr_48b
, size
);
3368 assert(addr_48b
>= HIGH_HEAP_MIN_ADDRESS
);
3369 util_vma_heap_free(&device
->vma_hi
, addr_48b
, size
);
3372 pthread_mutex_unlock(&device
->vma_mutex
);
3375 VkResult
anv_AllocateMemory(
3377 const VkMemoryAllocateInfo
* pAllocateInfo
,
3378 const VkAllocationCallbacks
* pAllocator
,
3379 VkDeviceMemory
* pMem
)
3381 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3382 struct anv_physical_device
*pdevice
= device
->physical
;
3383 struct anv_device_memory
*mem
;
3384 VkResult result
= VK_SUCCESS
;
3386 assert(pAllocateInfo
->sType
== VK_STRUCTURE_TYPE_MEMORY_ALLOCATE_INFO
);
3388 /* The Vulkan 1.0.33 spec says "allocationSize must be greater than 0". */
3389 assert(pAllocateInfo
->allocationSize
> 0);
3391 VkDeviceSize aligned_alloc_size
=
3392 align_u64(pAllocateInfo
->allocationSize
, 4096);
3394 if (aligned_alloc_size
> MAX_MEMORY_ALLOCATION_SIZE
)
3395 return vk_error(VK_ERROR_OUT_OF_DEVICE_MEMORY
);
3397 assert(pAllocateInfo
->memoryTypeIndex
< pdevice
->memory
.type_count
);
3398 struct anv_memory_type
*mem_type
=
3399 &pdevice
->memory
.types
[pAllocateInfo
->memoryTypeIndex
];
3400 assert(mem_type
->heapIndex
< pdevice
->memory
.heap_count
);
3401 struct anv_memory_heap
*mem_heap
=
3402 &pdevice
->memory
.heaps
[mem_type
->heapIndex
];
3404 uint64_t mem_heap_used
= p_atomic_read(&mem_heap
->used
);
3405 if (mem_heap_used
+ aligned_alloc_size
> mem_heap
->size
)
3406 return vk_error(VK_ERROR_OUT_OF_DEVICE_MEMORY
);
3408 mem
= vk_alloc2(&device
->vk
.alloc
, pAllocator
, sizeof(*mem
), 8,
3409 VK_SYSTEM_ALLOCATION_SCOPE_OBJECT
);
3411 return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
3413 assert(pAllocateInfo
->memoryTypeIndex
< pdevice
->memory
.type_count
);
3414 vk_object_base_init(&device
->vk
, &mem
->base
, VK_OBJECT_TYPE_DEVICE_MEMORY
);
3415 mem
->type
= mem_type
;
3419 mem
->host_ptr
= NULL
;
3421 enum anv_bo_alloc_flags alloc_flags
= 0;
3423 const VkExportMemoryAllocateInfo
*export_info
= NULL
;
3424 const VkImportAndroidHardwareBufferInfoANDROID
*ahw_import_info
= NULL
;
3425 const VkImportMemoryFdInfoKHR
*fd_info
= NULL
;
3426 const VkImportMemoryHostPointerInfoEXT
*host_ptr_info
= NULL
;
3427 const VkMemoryDedicatedAllocateInfo
*dedicated_info
= NULL
;
3428 VkMemoryAllocateFlags vk_flags
= 0;
3429 uint64_t client_address
= 0;
3431 vk_foreach_struct_const(ext
, pAllocateInfo
->pNext
) {
3432 switch (ext
->sType
) {
3433 case VK_STRUCTURE_TYPE_EXPORT_MEMORY_ALLOCATE_INFO
:
3434 export_info
= (void *)ext
;
3437 case VK_STRUCTURE_TYPE_IMPORT_ANDROID_HARDWARE_BUFFER_INFO_ANDROID
:
3438 ahw_import_info
= (void *)ext
;
3441 case VK_STRUCTURE_TYPE_IMPORT_MEMORY_FD_INFO_KHR
:
3442 fd_info
= (void *)ext
;
3445 case VK_STRUCTURE_TYPE_IMPORT_MEMORY_HOST_POINTER_INFO_EXT
:
3446 host_ptr_info
= (void *)ext
;
3449 case VK_STRUCTURE_TYPE_MEMORY_ALLOCATE_FLAGS_INFO
: {
3450 const VkMemoryAllocateFlagsInfo
*flags_info
= (void *)ext
;
3451 vk_flags
= flags_info
->flags
;
3455 case VK_STRUCTURE_TYPE_MEMORY_DEDICATED_ALLOCATE_INFO
:
3456 dedicated_info
= (void *)ext
;
3459 case VK_STRUCTURE_TYPE_MEMORY_OPAQUE_CAPTURE_ADDRESS_ALLOCATE_INFO_KHR
: {
3460 const VkMemoryOpaqueCaptureAddressAllocateInfoKHR
*addr_info
=
3461 (const VkMemoryOpaqueCaptureAddressAllocateInfoKHR
*)ext
;
3462 client_address
= addr_info
->opaqueCaptureAddress
;
3467 anv_debug_ignored_stype(ext
->sType
);
3472 /* By default, we want all VkDeviceMemory objects to support CCS */
3473 if (device
->physical
->has_implicit_ccs
)
3474 alloc_flags
|= ANV_BO_ALLOC_IMPLICIT_CCS
;
3476 if (vk_flags
& VK_MEMORY_ALLOCATE_DEVICE_ADDRESS_BIT_KHR
)
3477 alloc_flags
|= ANV_BO_ALLOC_CLIENT_VISIBLE_ADDRESS
;
3479 if ((export_info
&& export_info
->handleTypes
) ||
3480 (fd_info
&& fd_info
->handleType
) ||
3481 (host_ptr_info
&& host_ptr_info
->handleType
)) {
3482 /* Anything imported or exported is EXTERNAL */
3483 alloc_flags
|= ANV_BO_ALLOC_EXTERNAL
;
3485 /* We can't have implicit CCS on external memory with an AUX-table.
3486 * Doing so would require us to sync the aux tables across processes
3487 * which is impractical.
3489 if (device
->info
.has_aux_map
)
3490 alloc_flags
&= ~ANV_BO_ALLOC_IMPLICIT_CCS
;
3493 /* Check if we need to support Android HW buffer export. If so,
3494 * create AHardwareBuffer and import memory from it.
3496 bool android_export
= false;
3497 if (export_info
&& export_info
->handleTypes
&
3498 VK_EXTERNAL_MEMORY_HANDLE_TYPE_ANDROID_HARDWARE_BUFFER_BIT_ANDROID
)
3499 android_export
= true;
3501 if (ahw_import_info
) {
3502 result
= anv_import_ahw_memory(_device
, mem
, ahw_import_info
);
3503 if (result
!= VK_SUCCESS
)
3507 } else if (android_export
) {
3508 result
= anv_create_ahw_memory(_device
, mem
, pAllocateInfo
);
3509 if (result
!= VK_SUCCESS
)
3512 const VkImportAndroidHardwareBufferInfoANDROID import_info
= {
3515 result
= anv_import_ahw_memory(_device
, mem
, &import_info
);
3516 if (result
!= VK_SUCCESS
)
3522 /* The Vulkan spec permits handleType to be 0, in which case the struct is
3525 if (fd_info
&& fd_info
->handleType
) {
3526 /* At the moment, we support only the below handle types. */
3527 assert(fd_info
->handleType
==
3528 VK_EXTERNAL_MEMORY_HANDLE_TYPE_OPAQUE_FD_BIT
||
3529 fd_info
->handleType
==
3530 VK_EXTERNAL_MEMORY_HANDLE_TYPE_DMA_BUF_BIT_EXT
);
3532 result
= anv_device_import_bo(device
, fd_info
->fd
, alloc_flags
,
3533 client_address
, &mem
->bo
);
3534 if (result
!= VK_SUCCESS
)
3537 /* For security purposes, we reject importing the bo if it's smaller
3538 * than the requested allocation size. This prevents a malicious client
3539 * from passing a buffer to a trusted client, lying about the size, and
3540 * telling the trusted client to try and texture from an image that goes
3541 * out-of-bounds. This sort of thing could lead to GPU hangs or worse
3542 * in the trusted client. The trusted client can protect itself against
3543 * this sort of attack but only if it can trust the buffer size.
3545 if (mem
->bo
->size
< aligned_alloc_size
) {
3546 result
= vk_errorf(device
, device
, VK_ERROR_INVALID_EXTERNAL_HANDLE
,
3547 "aligned allocationSize too large for "
3548 "VK_EXTERNAL_MEMORY_HANDLE_TYPE_OPAQUE_FD_BIT: "
3549 "%"PRIu64
"B > %"PRIu64
"B",
3550 aligned_alloc_size
, mem
->bo
->size
);
3551 anv_device_release_bo(device
, mem
->bo
);
3555 /* From the Vulkan spec:
3557 * "Importing memory from a file descriptor transfers ownership of
3558 * the file descriptor from the application to the Vulkan
3559 * implementation. The application must not perform any operations on
3560 * the file descriptor after a successful import."
3562 * If the import fails, we leave the file descriptor open.
3568 if (host_ptr_info
&& host_ptr_info
->handleType
) {
3569 if (host_ptr_info
->handleType
==
3570 VK_EXTERNAL_MEMORY_HANDLE_TYPE_HOST_MAPPED_FOREIGN_MEMORY_BIT_EXT
) {
3571 result
= vk_error(VK_ERROR_INVALID_EXTERNAL_HANDLE
);
3575 assert(host_ptr_info
->handleType
==
3576 VK_EXTERNAL_MEMORY_HANDLE_TYPE_HOST_ALLOCATION_BIT_EXT
);
3578 result
= anv_device_import_bo_from_host_ptr(device
,
3579 host_ptr_info
->pHostPointer
,
3580 pAllocateInfo
->allocationSize
,
3584 if (result
!= VK_SUCCESS
)
3587 mem
->host_ptr
= host_ptr_info
->pHostPointer
;
3591 /* Regular allocate (not importing memory). */
3593 result
= anv_device_alloc_bo(device
, pAllocateInfo
->allocationSize
,
3594 alloc_flags
, client_address
, &mem
->bo
);
3595 if (result
!= VK_SUCCESS
)
3598 if (dedicated_info
&& dedicated_info
->image
!= VK_NULL_HANDLE
) {
3599 ANV_FROM_HANDLE(anv_image
, image
, dedicated_info
->image
);
3601 /* Some legacy (non-modifiers) consumers need the tiling to be set on
3602 * the BO. In this case, we have a dedicated allocation.
3604 if (image
->needs_set_tiling
) {
3605 const uint32_t i915_tiling
=
3606 isl_tiling_to_i915_tiling(image
->planes
[0].surface
.isl
.tiling
);
3607 int ret
= anv_gem_set_tiling(device
, mem
->bo
->gem_handle
,
3608 image
->planes
[0].surface
.isl
.row_pitch_B
,
3611 anv_device_release_bo(device
, mem
->bo
);
3612 result
= vk_errorf(device
, device
, VK_ERROR_OUT_OF_DEVICE_MEMORY
,
3613 "failed to set BO tiling: %m");
3620 mem_heap_used
= p_atomic_add_return(&mem_heap
->used
, mem
->bo
->size
);
3621 if (mem_heap_used
> mem_heap
->size
) {
3622 p_atomic_add(&mem_heap
->used
, -mem
->bo
->size
);
3623 anv_device_release_bo(device
, mem
->bo
);
3624 result
= vk_errorf(device
, device
, VK_ERROR_OUT_OF_DEVICE_MEMORY
,
3625 "Out of heap memory");
3629 pthread_mutex_lock(&device
->mutex
);
3630 list_addtail(&mem
->link
, &device
->memory_objects
);
3631 pthread_mutex_unlock(&device
->mutex
);
3633 *pMem
= anv_device_memory_to_handle(mem
);
3638 vk_free2(&device
->vk
.alloc
, pAllocator
, mem
);
3643 VkResult
anv_GetMemoryFdKHR(
3645 const VkMemoryGetFdInfoKHR
* pGetFdInfo
,
3648 ANV_FROM_HANDLE(anv_device
, dev
, device_h
);
3649 ANV_FROM_HANDLE(anv_device_memory
, mem
, pGetFdInfo
->memory
);
3651 assert(pGetFdInfo
->sType
== VK_STRUCTURE_TYPE_MEMORY_GET_FD_INFO_KHR
);
3653 assert(pGetFdInfo
->handleType
== VK_EXTERNAL_MEMORY_HANDLE_TYPE_OPAQUE_FD_BIT
||
3654 pGetFdInfo
->handleType
== VK_EXTERNAL_MEMORY_HANDLE_TYPE_DMA_BUF_BIT_EXT
);
3656 return anv_device_export_bo(dev
, mem
->bo
, pFd
);
3659 VkResult
anv_GetMemoryFdPropertiesKHR(
3661 VkExternalMemoryHandleTypeFlagBits handleType
,
3663 VkMemoryFdPropertiesKHR
* pMemoryFdProperties
)
3665 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3667 switch (handleType
) {
3668 case VK_EXTERNAL_MEMORY_HANDLE_TYPE_DMA_BUF_BIT_EXT
:
3669 /* dma-buf can be imported as any memory type */
3670 pMemoryFdProperties
->memoryTypeBits
=
3671 (1 << device
->physical
->memory
.type_count
) - 1;
3675 /* The valid usage section for this function says:
3677 * "handleType must not be one of the handle types defined as
3680 * So opaque handle types fall into the default "unsupported" case.
3682 return vk_error(VK_ERROR_INVALID_EXTERNAL_HANDLE
);
3686 VkResult
anv_GetMemoryHostPointerPropertiesEXT(
3688 VkExternalMemoryHandleTypeFlagBits handleType
,
3689 const void* pHostPointer
,
3690 VkMemoryHostPointerPropertiesEXT
* pMemoryHostPointerProperties
)
3692 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3694 assert(pMemoryHostPointerProperties
->sType
==
3695 VK_STRUCTURE_TYPE_MEMORY_HOST_POINTER_PROPERTIES_EXT
);
3697 switch (handleType
) {
3698 case VK_EXTERNAL_MEMORY_HANDLE_TYPE_HOST_ALLOCATION_BIT_EXT
:
3699 /* Host memory can be imported as any memory type. */
3700 pMemoryHostPointerProperties
->memoryTypeBits
=
3701 (1ull << device
->physical
->memory
.type_count
) - 1;
3706 return VK_ERROR_INVALID_EXTERNAL_HANDLE
;
3710 void anv_FreeMemory(
3712 VkDeviceMemory _mem
,
3713 const VkAllocationCallbacks
* pAllocator
)
3715 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3716 ANV_FROM_HANDLE(anv_device_memory
, mem
, _mem
);
3721 pthread_mutex_lock(&device
->mutex
);
3722 list_del(&mem
->link
);
3723 pthread_mutex_unlock(&device
->mutex
);
3726 anv_UnmapMemory(_device
, _mem
);
3728 p_atomic_add(&device
->physical
->memory
.heaps
[mem
->type
->heapIndex
].used
,
3731 anv_device_release_bo(device
, mem
->bo
);
3733 #if defined(ANDROID) && ANDROID_API_LEVEL >= 26
3735 AHardwareBuffer_release(mem
->ahw
);
3738 vk_object_base_finish(&mem
->base
);
3739 vk_free2(&device
->vk
.alloc
, pAllocator
, mem
);
3742 VkResult
anv_MapMemory(
3744 VkDeviceMemory _memory
,
3745 VkDeviceSize offset
,
3747 VkMemoryMapFlags flags
,
3750 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3751 ANV_FROM_HANDLE(anv_device_memory
, mem
, _memory
);
3758 if (mem
->host_ptr
) {
3759 *ppData
= mem
->host_ptr
+ offset
;
3763 if (size
== VK_WHOLE_SIZE
)
3764 size
= mem
->bo
->size
- offset
;
3766 /* From the Vulkan spec version 1.0.32 docs for MapMemory:
3768 * * If size is not equal to VK_WHOLE_SIZE, size must be greater than 0
3769 * assert(size != 0);
3770 * * If size is not equal to VK_WHOLE_SIZE, size must be less than or
3771 * equal to the size of the memory minus offset
3774 assert(offset
+ size
<= mem
->bo
->size
);
3776 /* FIXME: Is this supposed to be thread safe? Since vkUnmapMemory() only
3777 * takes a VkDeviceMemory pointer, it seems like only one map of the memory
3778 * at a time is valid. We could just mmap up front and return an offset
3779 * pointer here, but that may exhaust virtual memory on 32 bit
3782 uint32_t gem_flags
= 0;
3784 if (!device
->info
.has_llc
&&
3785 (mem
->type
->propertyFlags
& VK_MEMORY_PROPERTY_HOST_COHERENT_BIT
))
3786 gem_flags
|= I915_MMAP_WC
;
3788 /* GEM will fail to map if the offset isn't 4k-aligned. Round down. */
3789 uint64_t map_offset
;
3790 if (!device
->physical
->has_mmap_offset
)
3791 map_offset
= offset
& ~4095ull;
3794 assert(offset
>= map_offset
);
3795 uint64_t map_size
= (offset
+ size
) - map_offset
;
3797 /* Let's map whole pages */
3798 map_size
= align_u64(map_size
, 4096);
3800 void *map
= anv_gem_mmap(device
, mem
->bo
->gem_handle
,
3801 map_offset
, map_size
, gem_flags
);
3802 if (map
== MAP_FAILED
)
3803 return vk_error(VK_ERROR_MEMORY_MAP_FAILED
);
3806 mem
->map_size
= map_size
;
3808 *ppData
= mem
->map
+ (offset
- map_offset
);
3813 void anv_UnmapMemory(
3815 VkDeviceMemory _memory
)
3817 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3818 ANV_FROM_HANDLE(anv_device_memory
, mem
, _memory
);
3820 if (mem
== NULL
|| mem
->host_ptr
)
3823 anv_gem_munmap(device
, mem
->map
, mem
->map_size
);
3830 clflush_mapped_ranges(struct anv_device
*device
,
3832 const VkMappedMemoryRange
*ranges
)
3834 for (uint32_t i
= 0; i
< count
; i
++) {
3835 ANV_FROM_HANDLE(anv_device_memory
, mem
, ranges
[i
].memory
);
3836 if (ranges
[i
].offset
>= mem
->map_size
)
3839 gen_clflush_range(mem
->map
+ ranges
[i
].offset
,
3840 MIN2(ranges
[i
].size
, mem
->map_size
- ranges
[i
].offset
));
3844 VkResult
anv_FlushMappedMemoryRanges(
3846 uint32_t memoryRangeCount
,
3847 const VkMappedMemoryRange
* pMemoryRanges
)
3849 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3851 if (device
->info
.has_llc
)
3854 /* Make sure the writes we're flushing have landed. */
3855 __builtin_ia32_mfence();
3857 clflush_mapped_ranges(device
, memoryRangeCount
, pMemoryRanges
);
3862 VkResult
anv_InvalidateMappedMemoryRanges(
3864 uint32_t memoryRangeCount
,
3865 const VkMappedMemoryRange
* pMemoryRanges
)
3867 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3869 if (device
->info
.has_llc
)
3872 clflush_mapped_ranges(device
, memoryRangeCount
, pMemoryRanges
);
3874 /* Make sure no reads get moved up above the invalidate. */
3875 __builtin_ia32_mfence();
3880 void anv_GetBufferMemoryRequirements(
3883 VkMemoryRequirements
* pMemoryRequirements
)
3885 ANV_FROM_HANDLE(anv_buffer
, buffer
, _buffer
);
3886 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3888 /* The Vulkan spec (git aaed022) says:
3890 * memoryTypeBits is a bitfield and contains one bit set for every
3891 * supported memory type for the resource. The bit `1<<i` is set if and
3892 * only if the memory type `i` in the VkPhysicalDeviceMemoryProperties
3893 * structure for the physical device is supported.
3895 uint32_t memory_types
= (1ull << device
->physical
->memory
.type_count
) - 1;
3897 /* Base alignment requirement of a cache line */
3898 uint32_t alignment
= 16;
3900 if (buffer
->usage
& VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT
)
3901 alignment
= MAX2(alignment
, ANV_UBO_ALIGNMENT
);
3903 pMemoryRequirements
->size
= buffer
->size
;
3904 pMemoryRequirements
->alignment
= alignment
;
3906 /* Storage and Uniform buffers should have their size aligned to
3907 * 32-bits to avoid boundary checks when last DWord is not complete.
3908 * This would ensure that not internal padding would be needed for
3911 if (device
->robust_buffer_access
&&
3912 (buffer
->usage
& VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT
||
3913 buffer
->usage
& VK_BUFFER_USAGE_STORAGE_BUFFER_BIT
))
3914 pMemoryRequirements
->size
= align_u64(buffer
->size
, 4);
3916 pMemoryRequirements
->memoryTypeBits
= memory_types
;
3919 void anv_GetBufferMemoryRequirements2(
3921 const VkBufferMemoryRequirementsInfo2
* pInfo
,
3922 VkMemoryRequirements2
* pMemoryRequirements
)
3924 anv_GetBufferMemoryRequirements(_device
, pInfo
->buffer
,
3925 &pMemoryRequirements
->memoryRequirements
);
3927 vk_foreach_struct(ext
, pMemoryRequirements
->pNext
) {
3928 switch (ext
->sType
) {
3929 case VK_STRUCTURE_TYPE_MEMORY_DEDICATED_REQUIREMENTS
: {
3930 VkMemoryDedicatedRequirements
*requirements
= (void *)ext
;
3931 requirements
->prefersDedicatedAllocation
= false;
3932 requirements
->requiresDedicatedAllocation
= false;
3937 anv_debug_ignored_stype(ext
->sType
);
3943 void anv_GetImageMemoryRequirements(
3946 VkMemoryRequirements
* pMemoryRequirements
)
3948 ANV_FROM_HANDLE(anv_image
, image
, _image
);
3949 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3951 /* The Vulkan spec (git aaed022) says:
3953 * memoryTypeBits is a bitfield and contains one bit set for every
3954 * supported memory type for the resource. The bit `1<<i` is set if and
3955 * only if the memory type `i` in the VkPhysicalDeviceMemoryProperties
3956 * structure for the physical device is supported.
3958 * All types are currently supported for images.
3960 uint32_t memory_types
= (1ull << device
->physical
->memory
.type_count
) - 1;
3962 pMemoryRequirements
->size
= image
->size
;
3963 pMemoryRequirements
->alignment
= image
->alignment
;
3964 pMemoryRequirements
->memoryTypeBits
= memory_types
;
3967 void anv_GetImageMemoryRequirements2(
3969 const VkImageMemoryRequirementsInfo2
* pInfo
,
3970 VkMemoryRequirements2
* pMemoryRequirements
)
3972 ANV_FROM_HANDLE(anv_device
, device
, _device
);
3973 ANV_FROM_HANDLE(anv_image
, image
, pInfo
->image
);
3975 anv_GetImageMemoryRequirements(_device
, pInfo
->image
,
3976 &pMemoryRequirements
->memoryRequirements
);
3978 vk_foreach_struct_const(ext
, pInfo
->pNext
) {
3979 switch (ext
->sType
) {
3980 case VK_STRUCTURE_TYPE_IMAGE_PLANE_MEMORY_REQUIREMENTS_INFO
: {
3981 const VkImagePlaneMemoryRequirementsInfo
*plane_reqs
=
3982 (const VkImagePlaneMemoryRequirementsInfo
*) ext
;
3983 uint32_t plane
= anv_image_aspect_to_plane(image
->aspects
,
3984 plane_reqs
->planeAspect
);
3986 assert(image
->planes
[plane
].offset
== 0);
3988 /* The Vulkan spec (git aaed022) says:
3990 * memoryTypeBits is a bitfield and contains one bit set for every
3991 * supported memory type for the resource. The bit `1<<i` is set
3992 * if and only if the memory type `i` in the
3993 * VkPhysicalDeviceMemoryProperties structure for the physical
3994 * device is supported.
3996 * All types are currently supported for images.
3998 pMemoryRequirements
->memoryRequirements
.memoryTypeBits
=
3999 (1ull << device
->physical
->memory
.type_count
) - 1;
4001 pMemoryRequirements
->memoryRequirements
.size
= image
->planes
[plane
].size
;
4002 pMemoryRequirements
->memoryRequirements
.alignment
=
4003 image
->planes
[plane
].alignment
;
4008 anv_debug_ignored_stype(ext
->sType
);
4013 vk_foreach_struct(ext
, pMemoryRequirements
->pNext
) {
4014 switch (ext
->sType
) {
4015 case VK_STRUCTURE_TYPE_MEMORY_DEDICATED_REQUIREMENTS
: {
4016 VkMemoryDedicatedRequirements
*requirements
= (void *)ext
;
4017 if (image
->needs_set_tiling
|| image
->external_format
) {
4018 /* If we need to set the tiling for external consumers, we need a
4019 * dedicated allocation.
4021 * See also anv_AllocateMemory.
4023 requirements
->prefersDedicatedAllocation
= true;
4024 requirements
->requiresDedicatedAllocation
= true;
4026 requirements
->prefersDedicatedAllocation
= false;
4027 requirements
->requiresDedicatedAllocation
= false;
4033 anv_debug_ignored_stype(ext
->sType
);
4039 void anv_GetImageSparseMemoryRequirements(
4042 uint32_t* pSparseMemoryRequirementCount
,
4043 VkSparseImageMemoryRequirements
* pSparseMemoryRequirements
)
4045 *pSparseMemoryRequirementCount
= 0;
4048 void anv_GetImageSparseMemoryRequirements2(
4050 const VkImageSparseMemoryRequirementsInfo2
* pInfo
,
4051 uint32_t* pSparseMemoryRequirementCount
,
4052 VkSparseImageMemoryRequirements2
* pSparseMemoryRequirements
)
4054 *pSparseMemoryRequirementCount
= 0;
4057 void anv_GetDeviceMemoryCommitment(
4059 VkDeviceMemory memory
,
4060 VkDeviceSize
* pCommittedMemoryInBytes
)
4062 *pCommittedMemoryInBytes
= 0;
4066 anv_bind_buffer_memory(const VkBindBufferMemoryInfo
*pBindInfo
)
4068 ANV_FROM_HANDLE(anv_device_memory
, mem
, pBindInfo
->memory
);
4069 ANV_FROM_HANDLE(anv_buffer
, buffer
, pBindInfo
->buffer
);
4071 assert(pBindInfo
->sType
== VK_STRUCTURE_TYPE_BIND_BUFFER_MEMORY_INFO
);
4074 buffer
->address
= (struct anv_address
) {
4076 .offset
= pBindInfo
->memoryOffset
,
4079 buffer
->address
= ANV_NULL_ADDRESS
;
4083 VkResult
anv_BindBufferMemory(
4086 VkDeviceMemory memory
,
4087 VkDeviceSize memoryOffset
)
4089 anv_bind_buffer_memory(
4090 &(VkBindBufferMemoryInfo
) {
4091 .sType
= VK_STRUCTURE_TYPE_BIND_BUFFER_MEMORY_INFO
,
4094 .memoryOffset
= memoryOffset
,
4100 VkResult
anv_BindBufferMemory2(
4102 uint32_t bindInfoCount
,
4103 const VkBindBufferMemoryInfo
* pBindInfos
)
4105 for (uint32_t i
= 0; i
< bindInfoCount
; i
++)
4106 anv_bind_buffer_memory(&pBindInfos
[i
]);
4111 VkResult
anv_QueueBindSparse(
4113 uint32_t bindInfoCount
,
4114 const VkBindSparseInfo
* pBindInfo
,
4117 ANV_FROM_HANDLE(anv_queue
, queue
, _queue
);
4118 if (anv_device_is_lost(queue
->device
))
4119 return VK_ERROR_DEVICE_LOST
;
4121 return vk_error(VK_ERROR_FEATURE_NOT_PRESENT
);
4126 VkResult
anv_CreateEvent(
4128 const VkEventCreateInfo
* pCreateInfo
,
4129 const VkAllocationCallbacks
* pAllocator
,
4132 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4133 struct anv_event
*event
;
4135 assert(pCreateInfo
->sType
== VK_STRUCTURE_TYPE_EVENT_CREATE_INFO
);
4137 event
= vk_alloc2(&device
->vk
.alloc
, pAllocator
, sizeof(*event
), 8,
4138 VK_SYSTEM_ALLOCATION_SCOPE_OBJECT
);
4140 return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
4142 vk_object_base_init(&device
->vk
, &event
->base
, VK_OBJECT_TYPE_EVENT
);
4143 event
->state
= anv_state_pool_alloc(&device
->dynamic_state_pool
,
4144 sizeof(uint64_t), 8);
4145 *(uint64_t *)event
->state
.map
= VK_EVENT_RESET
;
4147 *pEvent
= anv_event_to_handle(event
);
4152 void anv_DestroyEvent(
4155 const VkAllocationCallbacks
* pAllocator
)
4157 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4158 ANV_FROM_HANDLE(anv_event
, event
, _event
);
4163 anv_state_pool_free(&device
->dynamic_state_pool
, event
->state
);
4165 vk_object_base_finish(&event
->base
);
4166 vk_free2(&device
->vk
.alloc
, pAllocator
, event
);
4169 VkResult
anv_GetEventStatus(
4173 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4174 ANV_FROM_HANDLE(anv_event
, event
, _event
);
4176 if (anv_device_is_lost(device
))
4177 return VK_ERROR_DEVICE_LOST
;
4179 return *(uint64_t *)event
->state
.map
;
4182 VkResult
anv_SetEvent(
4186 ANV_FROM_HANDLE(anv_event
, event
, _event
);
4188 *(uint64_t *)event
->state
.map
= VK_EVENT_SET
;
4193 VkResult
anv_ResetEvent(
4197 ANV_FROM_HANDLE(anv_event
, event
, _event
);
4199 *(uint64_t *)event
->state
.map
= VK_EVENT_RESET
;
4206 VkResult
anv_CreateBuffer(
4208 const VkBufferCreateInfo
* pCreateInfo
,
4209 const VkAllocationCallbacks
* pAllocator
,
4212 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4213 struct anv_buffer
*buffer
;
4215 /* Don't allow creating buffers bigger than our address space. The real
4216 * issue here is that we may align up the buffer size and we don't want
4217 * doing so to cause roll-over. However, no one has any business
4218 * allocating a buffer larger than our GTT size.
4220 if (pCreateInfo
->size
> device
->physical
->gtt_size
)
4221 return vk_error(VK_ERROR_OUT_OF_DEVICE_MEMORY
);
4223 assert(pCreateInfo
->sType
== VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO
);
4225 buffer
= vk_alloc2(&device
->vk
.alloc
, pAllocator
, sizeof(*buffer
), 8,
4226 VK_SYSTEM_ALLOCATION_SCOPE_OBJECT
);
4228 return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
4230 vk_object_base_init(&device
->vk
, &buffer
->base
, VK_OBJECT_TYPE_BUFFER
);
4231 buffer
->size
= pCreateInfo
->size
;
4232 buffer
->usage
= pCreateInfo
->usage
;
4233 buffer
->address
= ANV_NULL_ADDRESS
;
4235 *pBuffer
= anv_buffer_to_handle(buffer
);
4240 void anv_DestroyBuffer(
4243 const VkAllocationCallbacks
* pAllocator
)
4245 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4246 ANV_FROM_HANDLE(anv_buffer
, buffer
, _buffer
);
4251 vk_object_base_finish(&buffer
->base
);
4252 vk_free2(&device
->vk
.alloc
, pAllocator
, buffer
);
4255 VkDeviceAddress
anv_GetBufferDeviceAddress(
4257 const VkBufferDeviceAddressInfoKHR
* pInfo
)
4259 ANV_FROM_HANDLE(anv_buffer
, buffer
, pInfo
->buffer
);
4261 assert(!anv_address_is_null(buffer
->address
));
4262 assert(buffer
->address
.bo
->flags
& EXEC_OBJECT_PINNED
);
4264 return anv_address_physical(buffer
->address
);
4267 uint64_t anv_GetBufferOpaqueCaptureAddress(
4269 const VkBufferDeviceAddressInfoKHR
* pInfo
)
4274 uint64_t anv_GetDeviceMemoryOpaqueCaptureAddress(
4276 const VkDeviceMemoryOpaqueCaptureAddressInfoKHR
* pInfo
)
4278 ANV_FROM_HANDLE(anv_device_memory
, memory
, pInfo
->memory
);
4280 assert(memory
->bo
->flags
& EXEC_OBJECT_PINNED
);
4281 assert(memory
->bo
->has_client_visible_address
);
4283 return gen_48b_address(memory
->bo
->offset
);
4287 anv_fill_buffer_surface_state(struct anv_device
*device
, struct anv_state state
,
4288 enum isl_format format
,
4289 struct anv_address address
,
4290 uint32_t range
, uint32_t stride
)
4292 isl_buffer_fill_state(&device
->isl_dev
, state
.map
,
4293 .address
= anv_address_physical(address
),
4294 .mocs
= device
->isl_dev
.mocs
.internal
,
4297 .swizzle
= ISL_SWIZZLE_IDENTITY
,
4298 .stride_B
= stride
);
4301 void anv_DestroySampler(
4304 const VkAllocationCallbacks
* pAllocator
)
4306 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4307 ANV_FROM_HANDLE(anv_sampler
, sampler
, _sampler
);
4312 if (sampler
->bindless_state
.map
) {
4313 anv_state_pool_free(&device
->dynamic_state_pool
,
4314 sampler
->bindless_state
);
4317 if (sampler
->custom_border_color
.map
) {
4318 anv_state_reserved_pool_free(&device
->custom_border_colors
,
4319 sampler
->custom_border_color
);
4322 vk_object_base_finish(&sampler
->base
);
4323 vk_free2(&device
->vk
.alloc
, pAllocator
, sampler
);
4326 VkResult
anv_CreateFramebuffer(
4328 const VkFramebufferCreateInfo
* pCreateInfo
,
4329 const VkAllocationCallbacks
* pAllocator
,
4330 VkFramebuffer
* pFramebuffer
)
4332 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4333 struct anv_framebuffer
*framebuffer
;
4335 assert(pCreateInfo
->sType
== VK_STRUCTURE_TYPE_FRAMEBUFFER_CREATE_INFO
);
4337 size_t size
= sizeof(*framebuffer
);
4339 /* VK_KHR_imageless_framebuffer extension says:
4341 * If flags includes VK_FRAMEBUFFER_CREATE_IMAGELESS_BIT_KHR,
4342 * parameter pAttachments is ignored.
4344 if (!(pCreateInfo
->flags
& VK_FRAMEBUFFER_CREATE_IMAGELESS_BIT_KHR
)) {
4345 size
+= sizeof(struct anv_image_view
*) * pCreateInfo
->attachmentCount
;
4346 framebuffer
= vk_alloc2(&device
->vk
.alloc
, pAllocator
, size
, 8,
4347 VK_SYSTEM_ALLOCATION_SCOPE_OBJECT
);
4348 if (framebuffer
== NULL
)
4349 return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
4351 for (uint32_t i
= 0; i
< pCreateInfo
->attachmentCount
; i
++) {
4352 ANV_FROM_HANDLE(anv_image_view
, iview
, pCreateInfo
->pAttachments
[i
]);
4353 framebuffer
->attachments
[i
] = iview
;
4355 framebuffer
->attachment_count
= pCreateInfo
->attachmentCount
;
4357 framebuffer
= vk_alloc2(&device
->vk
.alloc
, pAllocator
, size
, 8,
4358 VK_SYSTEM_ALLOCATION_SCOPE_OBJECT
);
4359 if (framebuffer
== NULL
)
4360 return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY
);
4362 framebuffer
->attachment_count
= 0;
4365 vk_object_base_init(&device
->vk
, &framebuffer
->base
,
4366 VK_OBJECT_TYPE_FRAMEBUFFER
);
4368 framebuffer
->width
= pCreateInfo
->width
;
4369 framebuffer
->height
= pCreateInfo
->height
;
4370 framebuffer
->layers
= pCreateInfo
->layers
;
4372 *pFramebuffer
= anv_framebuffer_to_handle(framebuffer
);
4377 void anv_DestroyFramebuffer(
4380 const VkAllocationCallbacks
* pAllocator
)
4382 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4383 ANV_FROM_HANDLE(anv_framebuffer
, fb
, _fb
);
4388 vk_object_base_finish(&fb
->base
);
4389 vk_free2(&device
->vk
.alloc
, pAllocator
, fb
);
4392 static const VkTimeDomainEXT anv_time_domains
[] = {
4393 VK_TIME_DOMAIN_DEVICE_EXT
,
4394 VK_TIME_DOMAIN_CLOCK_MONOTONIC_EXT
,
4395 VK_TIME_DOMAIN_CLOCK_MONOTONIC_RAW_EXT
,
4398 VkResult
anv_GetPhysicalDeviceCalibrateableTimeDomainsEXT(
4399 VkPhysicalDevice physicalDevice
,
4400 uint32_t *pTimeDomainCount
,
4401 VkTimeDomainEXT
*pTimeDomains
)
4404 VK_OUTARRAY_MAKE(out
, pTimeDomains
, pTimeDomainCount
);
4406 for (d
= 0; d
< ARRAY_SIZE(anv_time_domains
); d
++) {
4407 vk_outarray_append(&out
, i
) {
4408 *i
= anv_time_domains
[d
];
4412 return vk_outarray_status(&out
);
4416 anv_clock_gettime(clockid_t clock_id
)
4418 struct timespec current
;
4421 ret
= clock_gettime(clock_id
, ¤t
);
4422 if (ret
< 0 && clock_id
== CLOCK_MONOTONIC_RAW
)
4423 ret
= clock_gettime(CLOCK_MONOTONIC
, ¤t
);
4427 return (uint64_t) current
.tv_sec
* 1000000000ULL + current
.tv_nsec
;
4430 #define TIMESTAMP 0x2358
4432 VkResult
anv_GetCalibratedTimestampsEXT(
4434 uint32_t timestampCount
,
4435 const VkCalibratedTimestampInfoEXT
*pTimestampInfos
,
4436 uint64_t *pTimestamps
,
4437 uint64_t *pMaxDeviation
)
4439 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4440 uint64_t timestamp_frequency
= device
->info
.timestamp_frequency
;
4443 uint64_t begin
, end
;
4444 uint64_t max_clock_period
= 0;
4446 begin
= anv_clock_gettime(CLOCK_MONOTONIC_RAW
);
4448 for (d
= 0; d
< timestampCount
; d
++) {
4449 switch (pTimestampInfos
[d
].timeDomain
) {
4450 case VK_TIME_DOMAIN_DEVICE_EXT
:
4451 ret
= anv_gem_reg_read(device
->fd
, TIMESTAMP
| I915_REG_READ_8B_WA
,
4455 return anv_device_set_lost(device
, "Failed to read the TIMESTAMP "
4458 uint64_t device_period
= DIV_ROUND_UP(1000000000, timestamp_frequency
);
4459 max_clock_period
= MAX2(max_clock_period
, device_period
);
4461 case VK_TIME_DOMAIN_CLOCK_MONOTONIC_EXT
:
4462 pTimestamps
[d
] = anv_clock_gettime(CLOCK_MONOTONIC
);
4463 max_clock_period
= MAX2(max_clock_period
, 1);
4466 case VK_TIME_DOMAIN_CLOCK_MONOTONIC_RAW_EXT
:
4467 pTimestamps
[d
] = begin
;
4475 end
= anv_clock_gettime(CLOCK_MONOTONIC_RAW
);
4478 * The maximum deviation is the sum of the interval over which we
4479 * perform the sampling and the maximum period of any sampled
4480 * clock. That's because the maximum skew between any two sampled
4481 * clock edges is when the sampled clock with the largest period is
4482 * sampled at the end of that period but right at the beginning of the
4483 * sampling interval and some other clock is sampled right at the
4484 * begining of its sampling period and right at the end of the
4485 * sampling interval. Let's assume the GPU has the longest clock
4486 * period and that the application is sampling GPU and monotonic:
4489 * w x y z 0 1 2 3 4 5 6 7 8 9 a b c d e f
4490 * Raw -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-
4494 * GPU -----_____-----_____-----_____-----_____
4497 * x y z 0 1 2 3 4 5 6 7 8 9 a b c
4498 * Monotonic -_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-
4500 * Interval <----------------->
4501 * Deviation <-------------------------->
4505 * m = read(monotonic) 2
4508 * We round the sample interval up by one tick to cover sampling error
4509 * in the interval clock
4512 uint64_t sample_interval
= end
- begin
+ 1;
4514 *pMaxDeviation
= sample_interval
+ max_clock_period
;
4519 /* vk_icd.h does not declare this function, so we declare it here to
4520 * suppress Wmissing-prototypes.
4522 PUBLIC VKAPI_ATTR VkResult VKAPI_CALL
4523 vk_icdNegotiateLoaderICDInterfaceVersion(uint32_t* pSupportedVersion
);
4525 PUBLIC VKAPI_ATTR VkResult VKAPI_CALL
4526 vk_icdNegotiateLoaderICDInterfaceVersion(uint32_t* pSupportedVersion
)
4528 /* For the full details on loader interface versioning, see
4529 * <https://github.com/KhronosGroup/Vulkan-LoaderAndValidationLayers/blob/master/loader/LoaderAndLayerInterface.md>.
4530 * What follows is a condensed summary, to help you navigate the large and
4531 * confusing official doc.
4533 * - Loader interface v0 is incompatible with later versions. We don't
4536 * - In loader interface v1:
4537 * - The first ICD entrypoint called by the loader is
4538 * vk_icdGetInstanceProcAddr(). The ICD must statically expose this
4540 * - The ICD must statically expose no other Vulkan symbol unless it is
4541 * linked with -Bsymbolic.
4542 * - Each dispatchable Vulkan handle created by the ICD must be
4543 * a pointer to a struct whose first member is VK_LOADER_DATA. The
4544 * ICD must initialize VK_LOADER_DATA.loadMagic to ICD_LOADER_MAGIC.
4545 * - The loader implements vkCreate{PLATFORM}SurfaceKHR() and
4546 * vkDestroySurfaceKHR(). The ICD must be capable of working with
4547 * such loader-managed surfaces.
4549 * - Loader interface v2 differs from v1 in:
4550 * - The first ICD entrypoint called by the loader is
4551 * vk_icdNegotiateLoaderICDInterfaceVersion(). The ICD must
4552 * statically expose this entrypoint.
4554 * - Loader interface v3 differs from v2 in:
4555 * - The ICD must implement vkCreate{PLATFORM}SurfaceKHR(),
4556 * vkDestroySurfaceKHR(), and other API which uses VKSurfaceKHR,
4557 * because the loader no longer does so.
4559 * - Loader interface v4 differs from v3 in:
4560 * - The ICD must implement vk_icdGetPhysicalDeviceProcAddr().
4562 *pSupportedVersion
= MIN2(*pSupportedVersion
, 4u);
4566 VkResult
anv_CreatePrivateDataSlotEXT(
4568 const VkPrivateDataSlotCreateInfoEXT
* pCreateInfo
,
4569 const VkAllocationCallbacks
* pAllocator
,
4570 VkPrivateDataSlotEXT
* pPrivateDataSlot
)
4572 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4573 return vk_private_data_slot_create(&device
->vk
, pCreateInfo
, pAllocator
,
4577 void anv_DestroyPrivateDataSlotEXT(
4579 VkPrivateDataSlotEXT privateDataSlot
,
4580 const VkAllocationCallbacks
* pAllocator
)
4582 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4583 vk_private_data_slot_destroy(&device
->vk
, privateDataSlot
, pAllocator
);
4586 VkResult
anv_SetPrivateDataEXT(
4588 VkObjectType objectType
,
4589 uint64_t objectHandle
,
4590 VkPrivateDataSlotEXT privateDataSlot
,
4593 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4594 return vk_object_base_set_private_data(&device
->vk
,
4595 objectType
, objectHandle
,
4596 privateDataSlot
, data
);
4599 void anv_GetPrivateDataEXT(
4601 VkObjectType objectType
,
4602 uint64_t objectHandle
,
4603 VkPrivateDataSlotEXT privateDataSlot
,
4606 ANV_FROM_HANDLE(anv_device
, device
, _device
);
4607 vk_object_base_get_private_data(&device
->vk
,
4608 objectType
, objectHandle
,
4609 privateDataSlot
, pData
);