Lines Matching full:nvk

29 - NVK: A Vulkan driver for Nvidia hardware
102 - nvk: Implement VK_EXT_dynamic_rendering_unused_attachments
105 - nvk: Implement VK_EXT_primitive_topology_list_restart
106 - nvk: Implement VK_EXT_image_sliced_view_of_3d
107 - nvk: Implement VK_KHR_workgroup_memory_explicit_layout
111 - nvk: Implement VK_EXT_depth_bias_control
864 - nvk: Fix segfault when opening DRI device file returns error
1370 - nvk: add missing finish calls
1371 - nvk: add some initial wsi framework.
1372 - nvk: fix header guards to be less generic.
1373 - nvk: add bind buffer memory
1374 - nvk: Add initial queue
1375 - nvk: add cmd buffer framework
1376 - nvk: Reset pushbufs on command buffer reset
1377 - nvk: reindent descriptor sets to mesa std.
1378 - nvk: add initial descriptor pool framework.
1379 - nvk: some boilerplate for descriptor sets
1380 - nvk: add descriptor set bo allocation.
1381 - nvk: implement buffer address.
1382 - nvk: descriptor set freeing fix
1383 - nvk: move to new command stream generator.
1384 - nvk: port the blit and copy code to new command submission.
1386 - nvk: link in codegen without gallium bits.
1387 - nvk: Initial wiring in of the compiler
1388 - nvk: Basic descriptor binding
1390 - nvk: retrieve gpc/mp counts from kernel.
1391 - nvk: add support for preamble and tls allocation.
1392 - nvk: add record result to cmd_buffer.
1393 - nvk: add command stream upload buffer.
1395 - nvk: add some basic format wrapping framework
1396 - nvk: add some compute limits
1397 - nvk: add basic nve4+ compute support.
1398 - nvk: fix empty cmd submission.
1401 - nvk: add initial 8/16 byte clears.
1402 - nvk: fix pipeline pushbuf sizing
1403 - nvk: increase graphics cpu push buffer
1404 - nvk: fix depth emission ordering.
1405 - nvk: add some limits/features from binary driver.
1406 - nvk: add indexed draw support.
1407 - nvk: assign vertex locations according to input attrib index
1408 - nvk: lower io to temps to avoid output reads in vertex shaders
1409 - nvk: handle NULL to destroy descriptor pool
1410 - nvk: add basic primitive restart
1411 - nvk: fix copy lower address extraction
1412 - nvk: fix multiple pipelines failure allocation case.
1413 - nvk: init dev->physical_device earlier.
1414 - nvk/winsys: store device ptr into bo instead of ptr
1415 - nvk: set the device fd
1417 - nvk: Report image alignments from NIL
1419 - nvk/nil: don't ask for compressed image kind
1420 - nvk/barrier: handle host bit.
1421 - nvk: add compute support for ampere
1422 - nvk: add min_lod to spirv caps.
1423 - nvk: fix r32_sint format support
1424 - nvk: expose EXT_sampler_filter_minmax
1425 - nvk: fix transform feedback crash when optimiser removes things.
1426 - nvk: merge tess info between tcs/tes.
1427 - nvk: introduce an optimisation loop.
1428 - nvk: add support for D32_SFLOAT_S8_UINT
1429 - nvk/query: fix push buffer size for copy pool results.
1430 - nvk: init image fields for requirements
1431 - nvk: handle alignments in device memory
1432 - nvk/tess: don't emit patch control points in pipeline
1433 - nvk: align geometry clip setting with nvc0
1434 - nvk: fix independent color write masks.
1435 - nvk: enable rgb32 texel buffer support
1436 - nvk: enable EXT_depth_clip_control
1437 - nvk: enable EXT_depth_clip_enable
1438 - nvk: always sync internal cmd bufs for vma lifetimes.
1440 - nvk: Add support for sparse buffers
1441 - nvk: Add support for sparse images
1442 - nvk/queue: add support for syncobjs and sparse binds
1443 - nvk: Handle pre-turing indirect buffers with sparse
1444 - nvk: enable sparse features
1445 - nvk: enable a bunch of external fence/semaphore bits
1446 - nvk: enable sparse residency buffer on maxwell+
1447 - nvk: add new internal bo allocation flag.
1448 - docs: add two nvk exts to features.txt
1450 - nvk: align sampler allocation counts with nvidia.
1452 - nvk: add gart forced cmd pool side buffer.
1453 - nvk: add cond render upload buffer.
1454 - nvk: enable KHR_shader_clock.
1455 - nvk: NOUVEAU_WS_BO_LOCAL is a trap.
1469 - nvk: add build_id linker argument.
1728 - nvk: Fix some cast defines
1729 - nvk: Add A8B8G8R8_*_PACK32 format support
1730 - nvk: Add bufferImageGranularity limit
1731 - nvk: Reset offset value in ResetDescriptorPool
1745 - nvk: Avoid strict aliasing warning in the pushbuffer encoding.
1746 - nvk: Fix uninitialized result usage in NVK_DEBUG_ZERO_MEMORY.
1747 - nvk: Fix unused result warnings in pushbuf resets.
1748 - nvk: Remove duplicate (disabled) point sprite setup.
1749 - nvk: Fix missing init of the stages to sync against.
1750 - nvk: Use depth_clamp_enable to select PIXEL_*_Z_CLAMP.
1752 - nvk: Quiet a compiler warning.
1753 - nvk: Clean up redundant vendor checking for physical device creation.
1754 - nvk: Add support for probing as a platform device.
1755 - nvk: Disable shaderStorageImageReadWithoutFormat pre-Maxwell.
1917 - ci: build nvk
2138 - nvk: Implement EnumerateInstanceVersion
2139 - nvk: Add stub implementations of VkImage and VkImageView
2140 - nvk: Add stub implementation of VkSampler
2141 - nvk: Add a stub implementation of VkBuffer
2142 - nvk: Implement VkDescriptorSetLayout
2143 - nvk: Implement VkPipelineLayout
2144 - nvk: Add initial descriptor set lowering
2145 - nvk: Implement vkUpdateDescriptorSets
2146 - nvk: Expose nvk_descriptor_stride_align_for_type
2147 - nvk: Re-format descriptor set layouts
2148 - nvk: Re-format pipeline layouts
2149 - nvk: Re-format descriptor sets some more
2150 - nvk/buffer: Take an offset in nvk_buffer_address
2151 - nvk/buffer: Add a push_buffer_ref helper
2152 - nvk/copy: Use nvk_buffer_address in CmdCopyBuffer
2153 - nvk/image: Add image address helpers
2154 - nvk/copy: Use nvk_image_base_address()
2155 - nvk: Add an nvk_device_physical helper
2156 - nvk: Add a skeleton for pipelines
2157 - nvk: Re-arrange nvk_descriptor_set.h a bit
2158 - nvk: Reformat nvk_nir_lower_descriptors
2159 - nvk: Add a couple descriptor set address helpers
2160 - nvk: Move nvk_cmd_pool cast definitions
2161 - nvk: Rework whitespace in nvk_cmd_buffer.c
2162 - nvk: Add a root descriptor table
2163 - nvk: Fetch descriptor set addresses from the root table
2164 - nvk: Re-arrange nir_lower_explicit_io a bit
2165 - nvk: Lower load_global_constant_offset
2166 - nvk: Drop image_view_init
2167 - nvk: Stop returning VK_ERROR_FORMAT_NOT_SUPPORTED for non-blitable
2168 - nvk: Allow R32_UINT
2169 - nvk: Mark nvk_push_descriptor_set_ref() inline
2170 - nvk: Add a descriptor table data structure
2171 - nvk: Copy in the nouveau TIC format table
2172 - nvk/image_view: Reformat and fix Create/DestroyImageView
2173 - nvk: Add an image descriptor table to the device
2174 - nvk: Fill out TIC table entries for image views
2175 - nvk: Set b->cursor when lowering image intrinsics
2176 - nvk: Unify descriptor loading in lower_descriptors
2177 - nvk: Re-format nvk_image_view.h a bit
2178 - nvk: Re-format nvk_buffer.c a bit
2179 - nvk: Add a stub implementation of buffer views
2180 - nvk: Make texture descriptors a bit more acceptable to codegen
2181 - nvk: GART os host-cache-coherent
2182 - nvk: Reserve a null image descriptor
2183 - nvk: Rework descriptor writes
2188 - nvk: Add an nvk_get_format helper
2189 - nvk: Use helpers for push_ref
2190 - nvk: Align arguments consistently in copy/blit code
2191 - nvk: Move Fill/UpdateBuffer to nvk_cmd_copy
2192 - Revert "nvk: Stop returning VK_ERROR_FORMAT_NOT_SUPPORTED for non-blitable"
2193 - nvk: Manually offset for array layers in copy/blit
2194 - nvk: Convert to using NIL for image layout
2195 - nvk: Re-indent image entrypoints
2196 - nvk: Implement VkGetImageSubresourceLyout
2197 - nvk: Reset and properly clean up command buffer upload areas
2198 - nvk: Rework format features queries
2199 - nvk: Add a more competent GetPhysicalDeviceImageFormatProperties
2200 - nvk: Support compressed images in copy commands
2201 - nvk: Drop vk_sync BO refs after push_submit
2206 - nvk: Teture pool sizes are maximums not sizes
2207 - nvk: Re-format nvk_sampler.c
2208 - nvk: Implement samplers
2210 - nvk: Move is_storage_image_format to nvk_format.c
2211 - nvk: Implement buffer views
2212 - nvk: Advertise KHR_dedicated_allocation
2213 - nvk: Use the correct root descriptor table size for CmdDispatch
2214 - nvk: Add support for dynamic buffers
2215 - nvk: Better advertise image format features
2216 - nvk: Advertise descriptor array indexing
2217 - nvk: Advertise non-zero descriptor set limits
2218 - nvk: Use a descriptor type instead of a hand-rolled thing
2219 - nvk: Handle cube storage images properly
2220 - nvk: Load the requested descriptor size
2221 - nvk: Implement push constants
2222 - nvk: Properly indent a comment
2223 - nvk: Fix descriptor offset alignment
2224 - nvk: Use a switch for descriptor types in load_descriptor
2225 - nvk: Support inline uniform blocks
2226 - nvk: Delete the storage TIC in nvk_image_view_destroy
2227 - nvk: Assert that we don't double-free descriptors
2228 - nvk: Initial vkCmdClearImage support
2229 - nvk: Unconditionally zero image format properties
2230 - nvk: No-op sparse image format properties
2231 - nvk: Advertise minUniformBufferOffsetAlignment
2232 - nvk: Rework OOM handling for descriptor pools
2233 - nvk: Bind immutable samplers on descriptor set creation
2234 - nvk: Padd shader BOs by 4K to avoid I-cache overflow
2235 - nvk: Include nvk_private.h in everything
2236 - nvk: Make image/buffer address helpers const
2238 - nvk: Init WSI after setting up supported_sync_types
2246 - nvk: Add graphics state to command buffers
2247 - nvk: Split pipeline binding into helpers
2248 - nvk: Switch to vk_pipeline_shader_stage_to_nir
2249 - nvk: Don't free the NIR in nvk_compile_nir
2250 - nvk: Add an nvk_shader_address helper
2251 - nvk: Free pipeline shader BOs
2252 - nvk: Expose pipeline alloc/free functions
2253 - nvk: Make shader_upload take an nvk_device
2254 - nvk/shader: Assign I/O locations and gather info
2255 - nvk/shader: Populate headers for vertex and fragment shaders
2256 - nvk: Add a nvk_cmd_buffer_device() helper
2257 - nvk: Import 3D context init code from nouveau
2259 - nvk: Add boilerplate for Begin/EndRendering
2260 - nvk: Misc. additional state setup
2261 - nvk: Emit dynamic graphics state
2262 - nvk: Implement push constants and descriptors for graphics
2264 - nvk: Graphics pipelines
2265 - nvk: Implement vkCmdDraw()
2266 - nvk: Color attachments clears via image clears
2273 - nvk: Add an attachment format even for secondaries
2274 - nvk: Add an addr field to nvk_buffer
2275 - nvk: Expose a bind_vertex_buffer helper
2276 - nvk: Use vk_meta for CmdClearAttachments
2277 - nvk: Stop using vk_cmd_set_dynamic_graphics_state in meta_end()
2278 - nvk: Enable all the dynamic state features
2280 - nvk: add linear image creation support.
2281 - nvk: Use max alignment for descriptor pool sizes
2283 - nvk: Use meta for CmdClear*Image
2284 - nvk: Zero client memory objects
2285 - nvk: Bind texture and sampler header pools for 3D
2286 - nvk: Use the new headers for samplers
2287 - nvk: Implement nir_intrinsic_load_frag_coord
2289 - nvk: Don't assert when there are no attachments
2290 - nvk: Track and reference all device memory objects
2292 - nvk/copy: Mape bpp part of nouveau_copy_buffer
2293 - nvk: Implement copies for D24_UNORM_S8_UINT images
2294 - nvk: Drop sample locations structs
2295 - nvk/meta: Save and restore VI state
2296 - nvk: Re-initialize dynamic_graphics_state.vi when recycling
2297 - nvk: Move the vertex format table into nvk_format.h
2298 - nvk: Advertise vertex buffer format featues
2299 - nvk: Clean up try_create_physical_device error handling
2301 - nvk: Depth bounds need fui()
2311 - nvk: Advertise more format features
2312 - nvk: Clear dynamic state dirty after flushing it all
2314 - nvk: Depth buffers don't allow Z-tiling
2315 - nvk: Disable sparse Z on Maxwell+
2318 - nvk: Add internal helpers for device memory allocation
2319 - nvk: Do internal dedicated allocations for ZS images
2320 - nvk: Fix depth/stencil render pass clears
2321 - nvk: Fix viewport Z scale
2322 - nvk: Enable two-sided stencil
2323 - nvk: Flip the front-face setting
2324 - nvk: Advertise depth/stencil support
2325 - nvk: Don't destroy NULL descriptor pool BOs
2326 - nvk: Call nir_lower_input_attachments
2327 - nvk: Set GEOMETRY_SHADER_SELECTS_LAYER properly
2328 - nvk: Return OUT_OF_DEVICE_MEMORY if bo_new fails
2330 - nvk: Add nvk_queue_init/finish() helpers
2331 - nvk: Align descriptor buffers to NVK_MIN_UBO_ALIGNMENT
2332 - nvk: Re-flow a couple function prototypes
2333 - nvk: Assert samples == 1
2334 - nvk: Allocate descriptors for input attachments
2335 - nvk: Wire up early z and post depth coverage
2336 - nvk: Save/restore push constants around meta ops
2338 - nvk: Use hardware clears for attachment clears
2339 - nvk: Add image_view_init/finish functions
2340 - nvk: Implement vkCmdClear*Image directly
2341 - nvk: Use a UINT format to clear non-renderable images
2342 - nvk: Don't advertise tiling on non-power-of-two formats
2343 - nvk: Fix max anisotropy
2344 - nvk: Assert on CmdExecuteCommands
2345 - nvk: VkSamplerCreateInfo::mipLodBias is signed
2346 - nvk: Fix border color alpha
2350 - nvk: Set up clip and cull distances
2351 - nvk: Fix dynamic buffer descriptor copies
2352 - nvk: Inline nouveau_copy_linear
2353 - nvk/copy: Rename push to p
2354 - nvk/blit: Rename push to p
2355 - nvk/dispatch: Rename push to p
2356 - nvk: Drop most buffer tracking
2357 - nvk: Rework TLS/SLM and image/sampler table handling
2358 - nvk: Invalidate texture header and sampler caches each submit
2359 - nvk/sampler: Free descriptor table entries
2360 - nvk: Rework nvk_descriptor_table_add/remove
2361 - nvk: Implement descriptor table growing
2362 - nvk: Zero unused descriptors
2363 - nvk: Add some asserts for nv50 compiler image restrictions
2364 - nvk: Update to the new command buffer infrastructure
2365 - nvk: Split nvk_queue into its own file
2366 - nvk: Start every command buffer with a nop
2367 - nvk: Initialize fixed draw/default state once
2371 - nvk: Use f for extension features
2372 - nvk: Drop a TODO
2373 - nvk: Use VK_IMAGE_USAGE_*_ATTACHMENT_BIT for image clears
2374 - nvk: Increase the graphics pipeline push space
2386 - nvk: Add MME infrastructure
2387 - nvk: Use MME for clears
2389 - nvk: Use MME for vkCmdDraw[Indexed]()
2390 - nvk: Implement vkCmdDraw[Indexed]Indirect()
2391 - nvk: Use p for the nouveau_ws_push_buffer in zero_vram
2398 - nvk: Use nv_push directly for graphics pipelines
2400 - nvk: Use bo_new_mapped for the zero page
2401 - nvk: Always allocate empty_push
2402 - nvk: Move queue_sumbit to nvk_queue_drm_nouveau.c
2403 - nvk: Submit pushbufs directly
2404 - nvk: Use a regular BO for the empty push
2405 - nvk: Use a regular BO for the queue state push
2406 - nvk: Add an nvk_queue_submit_simple helper
2407 - nvk: Initialize the queue later in device setup
2408 - nvk: Use submit_simple for draw state init
2409 - nvk: Use queue_submit_simple for zero_vram
2410 - nvk: Break nvk_cmd_pool into its own file
2411 - nvk: Use cmd instead of cmd_buffer
2412 - nvk: Add BO recycling to the command pool
2413 - nvk: Return VkResult from nvk_cmd_buffer_upload_alloc
2414 - nvk: memcpy root descriptors for compute instead of doing a DMA
2415 - nvk: Fully populate QMDs before uploading
2416 - nvk: Constant buffer alignment is actually 64B
2417 - nvk: Rework side-band data upload
2418 - nvk: Add an nvk_cmd_buffer_push helper
2419 - nvk: Add an nvk_cmd_buffer_ref_bo helper
2420 - nvk: Allocate upload buffers from the command pool
2421 - nvk: Use nvk_cmd_bo for push bufs
2422 - nvk: Implement vkCmdExecuteCommands()
2423 - nvk: Remove remaining references to nouveau_push.h
2426 - nvk: Re-indent vk_instance.c
2427 - nvk: Use vk_object_zalloc/free for descriptor pools/sets
2428 - nvk: Fix up whitespace in nvk_descriptor_set.c
2429 - nvk: Implement VK_KHR_push_descriptor
2430 - nvk: Reference descriptor set layouts in the sets themselves
2431 - nvk: Embed a nv_device_info in nvk_physical_device
2432 - nvk: Add an nvk_queue_submit wrapper
2433 - nvk: Also store the push BO map in nvk_queue_state
2434 - nvk: Bring back push sync and dumping
2435 - nvk: drop nvk_nir.h
2436 - nvk: Add lowering for load_global_constant_bounded
2437 - nvk: Properly implement robustBufferAccess
2443 - nvk: Support load_layer_id
2444 - nvk/meta: Save/restore descriptor set 0
2445 - nvk: Use meta for doing blits with the 3D hardware
2446 - nvk: WFI in pipeline barriers
2448 - nvk: Rework nvk_queue_submit_simple()
2449 - nvk: Add a heap data structure
2450 - nvk: Return a VkResult from nvk_shader_upload()
2451 - nvk: Add a shader heap to nvk_device
2452 - nvk: Allocate shaders from a heap
2453 - nvk: Rework whitespace in nvk_device_memory.c
2454 - nvk: Style fixes in nvk_physical_device.c
2455 - nvk: Reset semaphore syncs on wait
2456 - nvk/wsi: Style fixes
2457 - nvk/wsi: Use the common present implementation
2459 - nvk: Add a query pool object
2460 - nvk: Implement timestamp queries
2461 - nvk: Implement pipeline statistics and occlusion queries
2467 - nvk: Use mme_tu104_read_fifoed()
2468 - nvk: Implement vkCmdCopyQueryPoolResults()
2469 - nvk: Handle large command buffer uploads better
2470 - nvk: Use a normal DMA for CmdUpdateBuffer
2472 - nvk: Use mme_load_addr64()
2473 - nvk: Use poll for BO waits
2474 - nvk: Events
2475 - nvk: Don't crash if we fail to allocate a push BO
2476 - nvk: Stop leaking command pool BOs
2477 - nvk: Enable VK_KHR_create_renderpass2
2478 - nvk: Advertise VK_KHR_imageless_framebuffer
2479 - nvk: Flush the current pushbuf before allocating a new one
2480 - nvk: Advertise VK_KHR_separate_depth_stencil_layout
2481 - nvk: Tell WSI we don't support legacy scanout
2483 - nvk: Implement VK_EXT_pci_bus_info
2484 - nvk: Bind 3D images as 3D for clears
2485 - nvk: Support copies between 3D and 2D images
2487 - nvk: Support 2D views of 3D images
2488 - nvk: Advertise VK_KHR_maintenance1
2489 - nvk: Use 2D array views for 3D storage images
2491 - nvk: Advertise custom border color features
2493 - nvk: Implement VK_EXT_border_color_swizzle
2494 - nvk: Advertise VK_EXT_extended_dynamic_state3
2495 - nvk: Move more states to dynamic
2496 - nvk: Advertise VK_KHR_storage_buffer_storage_class
2497 - nvk: Add a helper for pushing descriptors
2499 - nvk: Implement VK_EXT/KHR_buffer_device_address
2500 - nvk: Break the guts of CmdDispatch into a helper
2501 - nvk: Implement DispatchIndirect
2503 - nvk: Implement CS invocations statistics queries
2509 - nvk: Stop asserting on MSAA
2510 - nvk: Handle zero color attachments better
2511 - nvk: Handle multisampled render targets properly
2512 - nvk: Support copies of MSAA images
2513 - nvk: Use the right view format for stencil texturing
2514 - nvk: Pass through a shader key for fragment shaders and MSAA
2515 - nvk: Set correct multisample regs for graphics pipelines
2516 - nvk: Stop creating a new upload BO every time
2517 - nvk: Fill out sample locations on Maxwell B+
2524 - nvk/meta: Fix restore for descriptor set 0
2525 - nvk: Use meta for MSAA resolves
2526 - nvk: Replace gl_SamplePosition with fract(gl_FragCoord.xy)
2527 - nvk: Stop advertising higher framebufferNoAttachmentsSampleCounts
2528 - nvk: Advertise MSAA via image format properties
2529 - nvk: Advertise VK_KHR_depth_stencil_resolve
2530 - nvk: Assert that descriptor buffer access stays in-bounds
2531 - nvk: Add a bo size to nvk_descriptor_set
2532 - nvk/format: Style fix for VkFormatProperties3KHR
2533 - nvk: Support VK_FORMAT_B10G11R11_UFLOAT_PACK32 for vertex buffers
2534 - nvk: Add a devenv ICD json file
2535 - nvk: Advertise EXT_vertex_attribute_divisor
2536 - nvk: Lower image_size to txs
2537 - nvk: Fix a comment
2538 - nvk: Add an nvk_buffer_addr_range helper
2539 - nvk: Use nvk_buffer_addr_range for buffer descriptors
2540 - nvk: Re-order Vulkan 1.0 feature bits
2541 - nvk: Enable inheritedQueries
2542 - nvk: Enable VK_EXT_provoking_vertex
2543 - nvk: Advertise samplerMirrorClampToEdge via 1.2 features
2544 - nvk: Advertise VK_KHR_bind_memory2
2545 - nvk: Enable KHR_dynamic_rendering
2546 - nvk: Advertise KHR_uniform_buffer_standard_layout
2547 - nvk: Advertise EXT_index_type_uint8
2548 - nvk: Advertise VK_EXT_separate_stencil_usage
2549 - nvk: Capitalize NVK in user exposed strings
2550 - nvk: Rename grid_size to group_count
2551 - nvk: Lower load_num_workgroups ourselves
2552 - nvk: Drop block_size from the root descriptor table
2553 - nvk: Add a helper for loading resource_index-based descriptors
2554 - nvk: Set maxMemoryAllocationCount
2577 - nvk: Add support for contiguous heaps to nvk_heap
2578 - nvk: Use a contiguous shader heap pre-Volta
2579 - nvk: Disable indirect draw/dispatch and query copy MMEs for now
2580 - nvk: Free a couple regs in nvk_mme_build_draw_*()
2581 - nvk: Properly align root descriptor tables for pre-Pascal
2582 - nvk: Compile all NIR before running codegen
2584 - nvk: Use a GS for layerered rendering pre-MaxwellB
2585 - nvk: Handle zero-size index and vertex buffers pre-Turing
2586 - nvk: Cosmetic clean-ups to Create/DestroyDevice
2596 - nvk/drm: Take a byte offset/range in push_add_push
2597 - nvk: Rework nvk_cmd_push a bit
2598 - nvk: Add a helper for pushing indirect data
2599 - nvk: Make some MME builder names more consistent
2601 - nvk: Reduce register pressure in nvk_mme_build_draw*()
2603 - nvk: Implement Draw*Indirect on pre-Turing
2605 - nvk: Add a build test for MMEs
2606 - nvk: Don't over-size push descriptor sets
2607 - nvk: Return VK_ERROR_INCOMPATIBLE_DRIVER if the PCI vendor isn't NVIDIA
2608 - nvk: Bump init context batch size
2611 - nvk: Rework extension enables
2612 - nvk: Rework features enables
2613 - nvk: Advertise shaderImageGatherExtended
2616 - nvk: De-duplicate MME code for setting draw params
2617 - nvk: Clamp viewport clip to max range
2618 - nvk: Use the same lock for the submit and the memory objects list
2619 - nvk: Advertise ICD/loader interface version 4
2620 - nvk: Add instace WSI entrypoints
2623 - nvk: Add begin to mme_scratch
2624 - nvk: Use the new load/store_scratch helpers for DRAW_PAD_DW
2626 - nvk: Rework spill helpers and DRAW_COUNT spilling
2627 - nvk: Spill DRAW_IDX pre-Turing
2628 - nvk: Break the inner MME draw loop into a helper
2629 - nvk: Increase the push runout to 512 dwords
2632 - nvk: Implement uncompressed views of compressed images
2633 - nvk: Set pointClippingBehavior
2634 - nvk: Expose VK_KHR_maintenance2
2635 - nvk: Add a separate #define for SSBO alignment
2636 - nvk: Set spirv_to_nir_options::min_*_alignment
2637 - nvk: Use vk_device_memory
2638 - nvk: Implement VK_KHR_map_memory2
2639 - nvk: Sort SPIR-V caps
2640 - nvk: Advertise EXT_shader_viewport_index_layer on MaxwellB+
2641 - nvk: Only use view_id for layer in multiview
2642 - nvk/heap: Set the right pitch for heap resize copies
2643 - nvk: Advertise shaderStorageImageReadWithoutFormat
2644 - nvk: Fix the NO_PREFETCH assert for CmdDrawIndirect
2645 - nvk: Advertise KHR_spirv_1_4
2646 - nvk: s/device/dev in nvk_image.c
2647 - nvk: Add helpers for binding image planes
2648 - nvk: Take an nvk_image_plane in nouveau_copy_rect_image
2649 - nvk: Use the max descriptor alignemtn in GetDescriptorSetLayoutSupport
2650 - nvk: Use NVIDIA_VENDOR_ID in pdev try_create()
2651 - nvk: Use abbreviated names in nvk_device_memory.c
2652 - nvk: Add device and driver UUIDs
2653 - nvk: Add external memory queries
2654 - nvk: Dedicated allocations override internal
2655 - nvk: Require dedicated allocations for external images
2657 - nvk: Support dma-buf import
2658 - nvk: Support dma-buf export
2659 - nvk: Enable external memory extensions
2660 - nvk: Reformat nvk_buffer.c
2661 - nvk: Add a buffer alignment helper
2662 - nvk: Add an addr field to nvk_image_plane
2663 - nvk: Use canonical variable names in nvk_physical_device.c
2664 - nvk: Use canonical variable names in nvk_shader.c
2665 - nvk: Use canonical variable names in nvk_bo_sync.c
2666 - nvk: Use canonical variable names in nvk_sampler.c
2667 - nvk: Drop nvk_physical_device::instance
2668 - nvk: Only advertise EXT_pci_bus_info on discrete GPUs
2671 - nvk: Set deviceType based on nv_device_info::type
2674 - nvk: Use nv_device_info for class checks
2675 - nvk: Rename nvk_device::ctx to ws_ctx
2676 - nvk: Add a ws_dev to nvk_device and use it
2677 - nvk: Move the winsys device to nvk_device
2678 - nvk: Don't enumerate pre-Kepler GPUs
2679 - nvk: Implement VK_EXT_physical_device_drm
2680 - nvk: Require an environment variable for poorly tested hardware
2681 - nvk: Use the new core vk_sampler struct
2685 - nvk: Bind 3D depth/stencil images as 2D arrays
2686 - nvk: Flush more state on VI_BINDINGS_VALID dirty
2687 - nvk: Don't skip zero-size bindings in GetDescriptorSetLayoutSupport
2688 - docs: Add a docs page for NVK
2689 - docs: Add NVK to features.txt
2690 - docs/relnotes: Stick something about NVK in new_features.txt
2693 - nvk: Use the new NIL helper for image level extents for copies
2694 - nvk: Improve image format properties and limits
2695 - nvk: Rework multi-plane format features a bit
2696 - nvk: Use nvk_root_descriptor_offset for drawInfoBase
2697 - nvk: Add a root_desc_addr to the root descriptor table
2698 - nvk: Add support for variable pointers
2699 - nvk: Enable the SPIR-V DeviceGroup capability
2700 - nvk: Separate the MME query copy code out a bit
2701 - nvk: Implement CopyQueryPoolResults with a compute shader
2702 - nvk: Misc. style nits
2703 - nvk: Rework memory requirements to handle aspects correctly
2704 - nvk: Implement the maintenance5 image layout queries
2705 - nvk: Use VkBufferUsageFlags2
2706 - nvk: Implement CmdBindIndexBuffer2KHR
2707 - nvk: Implement GetRenderingAreaGranularityKHR
2708 - nvk: Decorate CmdBegin/EndRendering entrypoints
2712 - nvk: Use the imported nouveau_drm.h headers
2715 - nvk: Use an empty EXEC for the empty submit case
2717 - nvk: Print an error message if VM_BIND support is missing
2718 - nvk: Enable the new UAPI
2719 - nvk: Use more consistent device variable names
2720 - nvk: Call nir_lower_int64
2722 - nvk: Remove plane sources from tex instructions
2723 - nvk: Use common physical device properties
2738 - nvk: Don't use nir_ssa_for_src()
2802 - nvk: Plumb no_prefetch through to the DRM back-end
2820 - nvk: Clean up includes
2821 - nvk: Add include guards to nvk_bo_sync.h
2822 - nvk: SPDX everything
2825 - nvk: Don't add a dummy attachment when gl_SampleMask is written
2826 - nvk: Set the discard bit for Z/S self-deps
2827 - nvk: Invalidate the texture cache in PipelineBarrier
2828 - nvk: Lower interp_at_sample to interp_at_offset
2829 - nvk: Disable statistics around meta ops
2830 - nvk: Clean up viewport math
2831 - nvk: Fix depth clipping parameters
2832 - nvk: Enable dynamic clip/clamp enable
2833 - nvk: Set GUARDBAND_Z_SCALE_1 when Z-clipping
2836 - nvk: Re-order physical device limits
2837 - nvk: Advertise maxMemoryAllocationCount = 4096
2838 - nvk: Advertise discreteQueuePriorities = 2
2839 - nvk: Rip out old UAPI support
2840 - nvk/drm: Drop the push_add_push_bo() helper
2841 - nvk/drm: Drop the push_add_bo() helper
2842 - nvk: Drop command buffer BO tracking
2843 - nvk: Drop memory object tracking
2844 - nvk: Drop the device-level mutex
2845 - nvk: Get rid of the tiled memory allocation helpers
2846 - nvk/drm: Restructure nvk_queue_submit_drm_nouveau()
2847 - nvk/drm: Split exec as needed for large command buffers
2848 - nvk: Don't store the descriptor pool BO in the set
2849 - nvk: Store a 20-bit driver_build_sha in nvk_instance
2850 - nvk: Hook up the disk cache
2851 - nvk: Re-structure early shader compilation a bit
2852 - nvk: Add a default pipeline cache
2853 - nvk: Cache NIR shaders
2854 - nvk: Init pipelineCacheUUID
2856 - nvk: Take GETPARAM_EXEC_PUSH_MAX into account
2857 - nvk: Handle zero-sized sparse buffers
2858 - nvk: Use align() and align64() instead of ALIGN_POT
2860 - nvk: Add a nvk_cmd_buffer_compute_cls() helper
2861 - nvk: Invalidate sampler/texture header caches in BeginCommandBuffer()
2862 - nvk: Invalidate SKED caches at the top of command buffers
2863 - nvk: Advertise more inline uniform block limits
2864 - nvk: Emit MME_DMA_SYSMEMBAR before indirect draw/dispatch
2865 - nvk: Set max descriptors to 2^20 for most descriptor types
2866 - nvk: Reset descriptor pool allocator when all sets are destroyed
2869 - nvk: Disable depth or stencil tests when unbound
2870 - nvk: Always emit at least one color attachment
2871 - nvk: Improve address space and buffer size limits
2872 - nvk: Always set pixel_min/max_Z to CLAMP
2873 - nvk: Use nouveau_ws_bo_unmap() instead of munmap()
2874 - nvk: Free the disk cache
2875 - nvk: Add an nvk_shader_finish() helper
2876 - nvk: Handle unbinding images and buffers
2877 - nvk: Clean up the disk cache on physical device create fail path
2879 - nvk: Add a nvk_cmd_buffer_dirty_render_pass() helper
2880 - nvk: Re-sort device features
2881 - nvk: Implement VK_EXT_depth_bias_control
2882 - nvk: Advertise VK_KHR_workgroup_memory_explicit_layout
2883 - nvk: Implement VK_EXT_image_sliced_view_of_3d
2884 - nvk: Advertise VK_EXT_primitive_topology_list_restart
2885 - nvk: Advertise VK_EXT_attachment_feedback_loop_layout
2886 - features: Mark VK_EXT_attachment_feedback_loop_layout done for NVK
2887 - nvk: Re-arrange Vulkan 1.2 features to match the header
2888 - nvk: Advertise shaderOutputLayer and shaderOutputViewportIndex
2889 - nvk: Enable descriptorIndexing
2890 - nvk: Implement VK_EXT_dynamic_rendering_unused_attachments
3061 - nvk: Do not increment instance id across draws
3062 - nvk: Add a macro for root descriptor table byte offsets
3063 - nvk: Set base vertex state in sequential mme draw
3064 - nvk: Support base instance in instanced draw calls
3065 - nvk: Switch point rasterization to point sprites
3066 - nvk: Support large points
3067 - nvk: Compile geometry shaders
3069 - nvk: Simplify mme build function argument
3070 - nvk: Support VK_KHR_shader_draw_parameters
3071 - nvk: Support for vertex shader transform feedback
3072 - nvk: Support transform feedback indirect draws
3073 - nvk: Support transform feedback geometry streams
3074 - nvk: Support transform feedback queries
3075 - nvk: Support vertex shader transform feedback on Fermi
3076 - nvk: Disable PRIMITIVE_RESTART_VERTEX_ARRAY by default
3077 - nvk: Fix geometry shader active stream mask
3078 - nvk: Support geometry shaders
3079 - nvk: Basic tessellation shader support
3080 - nvk: Assign locations correctly for arrayed IO
3081 - nvk: Enable multiview with tessellation shader
3082 - nvk: Fix cases where execution mode is specified in the tesc shader.
3083 - nvk: Respect tessellation domain origin state
3084 - nvk: Lower io to temporaries for tessellation evaluation nir
3085 - nvk: Support VkDescriptorSetVariableDescriptorCountLayoutSupport
3086 - nvk: Handle cases of descriptor bindings with variable counts
3087 - nvk: Add nir non-uniform optimization pass
3088 - nvk: Enable descriptor indexing
3089 - nvk: Do not keep redundant info for tessellation domain
3091 - nvk: Enable dynamic line rasterization mode state
3092 - nvk: Fix support for VK_EXT_sample_locations
3093 - nvk: Support dynamic state for enabling sample locations
3711 - nvk: add vulkan skeleton
3713 - nvk: use winsys lib
3714 - nvk: fix nvk_buffer include guards
3717 - nvk: implement GetPhysicalDeviceQueueFamilyProperties2 to make the CTS happy
3718 - nvk: advertize memory heaps and types
3722 - nvk: set nonCoherentAtomSize as the CTS divides with this value
3724 - nvk: add basic device memory support
3732 - nvk: allocate a GPU context for each VkDevice
3733 - nvk: add nvk_bo_sync
3734 - nvk: add nvk_CmdPipelineBarrier2 stub
3735 - nvk: impl nvk_CmdCopyBuffer
3740 - nvk: use fermi class definitions
3741 - nvk: add basic support for images
3742 - nvk: simple format table
3743 - nvk: add support for blits
3744 - nvk: report maxMipLevels as 1
3745 - nvk: optimize blit command buffer gen
3746 - nvk: implement CmdFillBuffer
3747 - nvk: implement CmdUpdateBuffer
3748 - nvk: implement CmdCopyBuffer2
3749 - nvk: advertise VK_KHR_copy_commands2
3750 - nvk: implicitly reset the command buffer
3752 - nvk: reduce pitch even further in CmdFillBuffer
3753 - nvk: support multiple miplevels
3754 - nvk: support array blits over multiple layers
3755 - nvk: tiling prep work for VK_EXT_image_2d_view_of_3d
3758 - nvk/blit: assert that formats are supported
3764 - nvk: set deviceName
3771 - nvk: clean up bo mappings
3775 - nvk: fix OOB read inside nvk_get_va_format
3776 - nvk: alloc a zero page and use it for vertex runouts
3777 - nvk: fix zero page refing
3778 - nvk: support exporting buffers
3779 - nvk: fix some class version checks
3780 - nvk: properly align shaders pre Turing
3781 - nvk: rework QMD handling to support pre Turing
3782 - nvk: align desc root table
3783 - nvk: Use SET_PIPELINE_PROGRAM pre-Volta
3784 - nvk: properly align slm size
3785 - nvk: use remaps for image copies
3786 - nvk: reduce pitch for FillBuffer
3787 - nvk: bind more subchans in init_context_state
3788 - nvk: support pre Maxwell Texture Headers
3789 - nvk/device: fix order of error handling
3790 - nvk: allocate VAB memory area
3791 - nvk: wire up M2MF for Fermi
3794 - nvk: Add a macro to set MMIO registers via falcons
3796 - nvk: fix num_gprs for Volta+
3797 - nvk: replace mp with tpc
3798 - nvk: properly calculate SLM region by taking per arch limits into account
3800 - nvk: enable fp helper invocations loads on more gens
4360 - nvk: Remove reference to genUserClip
4943 - nvk: implement vkGetDeviceImageMemoryRequirementsKHR()
4944 - nvk: add stub for vkGetDeviceImageSparseMemoryRequirementsKHR()
4945 - nvk: implement vkGetDeviceBufferMemoryRequirementsKHR()
4946 - nvk: advertise VK_KHR_maintenance4
4947 - nvk: advertise DemoteToHelperInvocation
4948 - nvk: Enable multiplane images and image views
4949 - nouveau/nvk: Add YCbCr sampler NIR lowering pass
4950 - nouveau/nvk: Support multi-plane descriptors in nvk_nir_lower_descriptors.c
4951 - nouveau/nvk: Create helper function for sampler creation
4952 - nouveau/nvk: Add multiple sampler planes for CONVERSION_SEPARATE_RECONSTRUCTION_FILTER_BIT
4953 - nouveau/nvk: Enable VK_KHR_sampler_ycbcr
4956 - nvk: Enable SEPARATE_RECONSTRUCTION_FILTER_BIT for multi-planar formats only
4957 - nvk: Enable MIDPOINT_CHROMA_SAMPLES_BIT for multi-planar formats only
5189 - nvk: Convert system values for gl_PointCoord and PointCoord into inputs
5190 - nvk: Add base_group to root descriptor table
5191 - nvk: Lower base_workgroup_id
5192 - nvk: Implement nvk_CmdDispatchBase and delete nvk_CmdDispatch
5193 - nvk: Advertise KHR_device_group
5194 - nvk: Add VK_FORMAT_B4G4R4A4_UNORM_PACK16 format to nil_format_info table
5195 - nvk: Add A4B4G4R4 formats to nil_format_info table
5196 - nvk: Advertise EXT_4444_formats
5197 - nvk: Enable shadow sampling
5198 - nvk: Implement VK_EXT_non_seamless_cube_map
5200 - nvk: Implement VK_EXT_image_view_min_lod
5201 - nvk: Update mutable descriptor struct type
5202 - nvk: Replace asserts with conditional that sets type_list = NULL
5203 - nvk: Implement nvk_GetDescriptorSetLayoutSupport
5204 - nvk: Enable VK_KHR_maintenance3
5205 - nvk: Advertise VK_EXT_mutable_descriptor_type
5206 - nvk: Set image index to zero for NULL nvk_buffer_view
5207 - nvk: Advertise VK_EXT_image_robustness
5208 - nvk: Advertise VK_EXT_robustness2
5209 - nvk: Add view_index to root descriptor table
5210 - nvk: Lower nir_intrinsic_load_view_index
5211 - nvk: Add draw support for multiview
5212 - nvk: Add query support for multiview
5213 - nvk: Add input attachments support for multiview
5214 - nvk: Advertise VK_KHR_multiview
5215 - nvk: Load view_mask to shadow scratch in nvk_CmdBeginRendering
5216 - nvk: Combine CLEAR_VIEWS and CLEAR_LAYERS MME macros
5217 - nvk: Move code inside view mask loops to a helper function
5850 - nvk: Enable VK_EXT_load_store_op_none
6002 - nvk: use common physical device enumeration
6003 - nvk: fix implicit-fallthrough warnings with clang
6004 - nvk: delete commented code
6005 - nvk: fix mem leaks
6006 - nvk: use common descriptor set layout code
6007 - nvk: use common pipeline layout code
6008 - nvk: advertise KHR_shader_non_semantic_info
6009 - nvk: advertise KHR_image_format_list
6010 - nvk: advertise EXT_private_data
6011 - nvk: advertise KHR_sampler_mirror_clamp_to_edge
6012 - nvk: KHR_descriptor_update_template
6013 - nvk: CmdPushDescriptorSetWithTemplateKHR
6014 - nvk: drop dead assignment
6015 - nvk: drop dead assignment
6016 - nvk: fix initialization override
6017 - nvk: sort extensions
6018 - nvk: advertize KHR_relaxed_block_layout
6019 - nvk: add check for VK_IMAGE_CREATE_2D_VIEW_COMPATIBLE_BIT_EXT
6020 - nvk: advertise EXT_image_2d_view_of_3d
6021 - nvk: fix maxPushDescriptors
6022 - nvk: call correct macro to clear views
6024 - nvk: add warning on non-nouveau drm driver
6025 - nvk: Implement VK_KHR_draw_indirect_count on Turing+
6026 - nvk: set device info before use in nvk_get_device_extensions
6027 - nvk: simplify code by using new helpers
6028 - nvk: remove duplicated device features
6029 - nvk: EXT_conditional_rendering
6030 - nvk: advertise VK_EXT_tooling_info
6031 - nvk: set optimization level to 3
6120 - nvk: Fix assert
6269 - nvk: Should use alignment instead of align
6274 - nvk: implement GetDeviceMemoryCommitment
6275 - nvk: support GetImageSparseMemoryRequirements2
6276 - nvk: expose KHR_driver_properties
6305 - nvk: Support VK_EXT_line_rasterization