summaryrefslogtreecommitdiffstats
path: root/src/intel/vulkan
Commit message (Collapse)AuthorAgeFilesLines
* anv: Rework pipeline cachingJason Ekstrand2016-08-308-432/+320
| | | | | | | | | | | | | | | | | | | The original pipeline cache the Kristian wrote was based on a now-false premise that the shaders can be stored in the pipeline cache. The Vulkan 1.0 spec explicitly states that the pipeline cache object is transiant and you are allowed to delete it after using it to create a pipeline with no ill effects. As nice as Kristian's design was, it doesn't jive with the expectation provided by the Vulkan spec. The new pipeline cache uses reference-counted anv_shader_bin objects that are backed by a large state pool. The cache itself is just a hash table mapping keys hashes to anv_shader_bin objects. This has the added advantage of removing one more hand-rolled hash table from mesa. Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97476 Acked-by: Kristian Høgsberg Kristensen <[email protected]>
* anv: Add a struct for storing a compiled shaderJason Ekstrand2016-08-303-0/+160
| | | | | | | | | | | | | | | | This new anv_shader_bin struct stores the compiled kernel (as an anv_state) as well as all of the metadata that is generated at shader compile time. The struct is very similar to the old cache_entry struct except that it is reference counted and stores the actual pipeline_bind_map. Similarly to cache_entry, much of the actual data is floating-size and stored after the main struct. Unlike cache_entry, which was storred in GPU-accessable memory, the storage for anv_shader_bin kernels comes from a state pool. The struct itself is reference-counted so that it can be used by multiple pipelines at a time without fear of allocation issues. Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]> Acked-by: Kristian Høgsberg Kristensen <[email protected]>
* anv: Add pipeline_has_stage guards a few placesJason Ekstrand2016-08-304-10/+40
| | | | | | | | | All of these worked before because they were depending on prog_data to be null. Soon, we won't be able to depend on a nice prog_data pointer and it's nice to be more explicit anyway. Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]>
* anv: Remove unused fields from anv_pipeline_bind_mapJason Ekstrand2016-08-301-2/+0
| | | | | Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]>
* anv/pipeline: Properly handle OOM during shader compilationJason Ekstrand2016-08-302-15/+30
| | | | | Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]>
* anv/allocator: Correctly set the number of bucketsJason Ekstrand2016-08-301-1/+1
| | | | | | | | | | | | The range from ANV_MIN_STATE_SIZE_LOG2 to ANV_MAX_STATE_SIZE_LOG2 should be inclusive and we have asserts that ensure that you never try to allocate a state larger than (1 << ANV_MAX_STATE_SIZE_LOG2). However, without adding 1 to the difference, we allocate 1 too few bucckts and so, even though we have an assert, anything landing in the last bucket will fail to allocate properly.. Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]>
* anv/pipeline: Fix bind maps for fragment output arraysJason Ekstrand2016-08-301-1/+1
| | | | | | | Found by inspection. Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]>
* anv/descriptor_set: memset anv_descriptor_set_layoutJason Ekstrand2016-08-301-2/+1
| | | | | | | | We hash this data structure so we can't afford to have uninitialized data even if it is just structure padding. Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]>
* anv/wayland: fix assert typoEric Engestrom2016-08-301-1/+1
| | | | | Signed-off-by: Eric Engestrom <[email protected]> Reviewed-by: Anuj Phogat <[email protected]>
* anv/meta: fix unreachable() typoEric Engestrom2016-08-301-1/+1
| | | | | Signed-off-by: Eric Engestrom <[email protected]> Reviewed-by: Anuj Phogat <[email protected]>
* nir: Change nir_shader_get_entrypoint to return an impl.Kenneth Graunke2016-08-251-2/+3
| | | | | | | | | | | | | | | | | Jason suggested adding an assert(function->impl) here. All callers of this function actually want ->impl, so I decided just to change the API. We also change the nir_lower_io_to_temporaries API here. All but one caller passed nir_shader_get_entrypoint(), and with the previous commit, it now uses a nir_function_impl internally. Folding this change in avoids the need to change it and change it back. v2: Fix one call I missed in ir3_compiler (caught by Eric). Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Connor Abbott <[email protected]>
* anv: Include the pipeline layout in the shader hashJason Ekstrand2016-08-244-4/+40
| | | | | | | | | | | | The pipeline layout affects shader compilation because it is what determines binding table locations as well as whether or not a particular buffer has dynamic offsets. Since this affects the generated shader, it needs to be in the hash. This fixes a bunch of CTS tests now that the CTS is using a pipeline cache. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kristian Høgsberg <[email protected]> Cc: "12.0" <[email protected]>
* anv: Add a --disable-vulkan-icd-full-driver-path optionJason Ekstrand2016-08-252-2/+8
| | | | | | | | This option makes installed Vulkan ICD files contain only a driver library name and not a path. This is intended for distros to help them work around multi-arch issues. Reviewed-by: Dave Airlie <[email protected]>
* anv: meta_blit2d: adapt texel fetch pitch for fake w-tiledLionel Landwerlin2016-08-241-1/+3
| | | | | | | | | | | | | We need to compute detiling coordinates using the physical size of W tiling (128x32) rather than the logical size (64x64). v2: Correct comment (Jason) Fixes dEQP-VK.api.copy_and_blit.image_to_image_stencil Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97448 Signed-off-by: Lionel Landwerlin <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/clear: Clear E5B9G9R9 images as R32_UINTJason Ekstrand2016-08-231-2/+14
| | | | | | | | | | | | | We can't actually clear these images normally because we can't render to them. Instead, we have to manually unpack the rgb9e5 color value on the CPU and clear it as R32_UINT. We still have a bit of work to do to clear non-power-of-two images, but this should get all of the power-of-two clears working on at least Haswell. This fixes three of the new Vulkan CTS tests in the dEQP-VK.api.image_clearing.clear_color_image.* group. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Nanley Chery <[email protected]> Cc: "12.0" <[email protected]>
* anv/clear: Make cmd_clear_image take an actual VkClearValueJason Ekstrand2016-08-231-5/+5
| | | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Nanley Chery <[email protected]> Cc: "12.0" <[email protected]>
* anv/blit2d: Add support for RGB destinationsJason Ekstrand2016-08-231-2/+214
| | | | | | | | This fixes 104 of the new image_clearing and copy_and_blit Vulkan CTS tests. Signed-off-by: Jason Ekstrand <[email protected]> Cc: "12.0" <[email protected]>
* anv/blit2d: Add a format parameter to bind_dst and create_iviewJason Ekstrand2016-08-231-5/+10
| | | | | | Signed-off-by: Jasosn Ekstrand <[email protected]> Reviewed-by: Nanley Chery <[email protected]> Cc: "12.0" <[email protected]>
* anv/image: Don't create invalid render target surfacesJason Ekstrand2016-08-231-1/+12
| | | | Signed-off-by: Jason Ekstrand <[email protected]>
* anv/formats: Don't use an RGBX format if it isn't renderableJason Ekstrand2016-08-231-1/+2
| | | | | | | | | | The whole point of using RGBX is so that we can render to it so if it isn't renderable, that kind-of defeats the purpose. Some formats (one example is R32G32B32X32_SFLOAT) exist in the format table but aren't actually renderable. Eventually, we'd like to get away from RGBX entirely, but this fixes hangs on BDW today. Signed-off-by: Jason Ekstrand <[email protected]>
* anv: Throw INCOMPATIBLE_DRIVER for non-fatal initialization errorsJason Ekstrand2016-08-221-5/+3
| | | | | | | | | | | The only reason we should throw INITIALIZATION_FAILED is if we have found useable intel hardware but have failed to bring it up for some reason. Otherwise, we should just throw INCOMPATIBLE_DRIVER which will turn into successfully advertising 0 physical devices Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Bas Nieuwenhuizen <[email protected]> Tested-by: Bas Nieuwenhuizen <[email protected]>
* anv: GetDeviceImageFormatProperties: fix TRANSFER formatsLionel Landwerlin2016-08-221-16/+6
| | | | | | | | | | | | | We let the user believe we support some transfer formats which we don't. This can lead to crashes when actually trying to use those formats for example on dEQP-VK.api.copy_and_blit.image_to_image.* tests. Let all formats we can render to or sample from as meta implements transfers using attachments. Signed-off-by: Lionel Landwerlin <[email protected]> Cc: Jason Ekstrand <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv: Give the installed intel_icd.json file an absolute pathJason Ekstrand2016-08-203-2/+8
| | | | | | | | | | | | Not providing a path allows the ICD to work on multi-arch systems but breaks it if you install anywhere other than /usr/lib. Given that users may be installing locally in .local or similar, we probably do want to provide a filename. Distros can carry a revert of this commit if they want an intel_icd.json file without the path. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Eric Engestrom <[email protected]> Reviewed-by: Chad Versace <[email protected]>
* genxml/gen9: Make 3DSTATE_SBE::AttributeActiveComponentFormat an arrayJason Ekstrand2016-08-191-35/+3
| | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* anv: remove dummy VK_DEBUG_MARKER_EXT entry pointsmil Velikov2016-08-181-20/+0
| | | | | | | | | | | | The vkCmdDbgMarker{Begin,End} symbols are exported, yet the json does no advertise that the driver supports the extension. Furthermore the functions are empty stubs. Remove those until we get a proper implementation and json notation. Cc: "12.0" <[email protected]> Cc: Jason Ekstrand <[email protected]> Signed-off-by: Emil Velikov <[email protected]>
* anv: do not export the Vulkan APIEmil Velikov2016-08-181-11/+0
| | | | | | | | | | | | | | | | | | | | | | | | | With version 1 of the Loader interface there is an internal/private symbol (vk_icdGetInstanceProcAddr) which is used to retrieve all the API from the Vulkan entrypoints from the ICD. Implying that exposing the Vulkan API is not recommended. Version 2 goes a step further explicitly forbiding the ICD from exposing Vulkan symbols (and adding a negotiation API) As a reference: - Nvidia 367.35 Missing negotiation API - version 1. Exposes only vk_icdGetInstanceProcAddr. - AMD 16.30.3.306809 Have negotiation API - version 2, Exposes vk_icdGetInstanceProcAddr. Exposes a couple of Vulkan entry points - seems to be in violation with the spec. Cc: "12.0" <[email protected]> Cc: Christian König <[email protected]> Signed-off-by: Emil Velikov <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv: automake: build with -BsymbolicEmil Velikov2016-08-181-0/+1
| | | | | | | | | Explicitly suggested in the Loader interface version 2 section, but it's good idea either way. It essentially, ensures that our symbols are not interposed. Cc: "12.0" <[email protected]> Cc: Jason Ekstrand <[email protected]> Signed-off-by: Emil Velikov <[email protected]>
* anv: automake: use VISIBILITY_CFLAGS to restrict symbol visibilityEmil Velikov2016-08-182-3/+7
| | | | | | | | | | | Hide the internal symbols and annotate the vk_icdGetInstanceProcAddr as public since the loader needs it (since v1 of the loader interface). v2: Add VISIBILITY_CFLAGS to AM_CFLAGS (Ken) Cc: "12.0" <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> (v1) Signed-off-by: Emil Velikov <[email protected]>
* anv: remove internal 'validate' layerEmil Velikov2016-08-182-110/+11
| | | | | | | | | | | Presently the layer has only a single entry point. As mentioned by Jason the function does not validate anything that isn't checked elsewhere, thus we can drop the whole thing. Cc: "12.0" <[email protected]> Cc: Jason Ekstrand <[email protected]> Suggested-by: Jason Ekstrand <[email protected]> Signed-off-by: Emil Velikov <[email protected]>
* anv: fix writemask on blit fragment shader.Dave Airlie2016-08-161-1/+1
| | | | | | | | | I'm not sure if anything even uses this, but I found this on radv, so just fix it on anv for consistency. Reviewed-by: Jason Ekstrand <[email protected]> Cc: [email protected] Signed-off-by: Dave Airlie <[email protected]>
* anv: add genX_multisample.h to the sources list(s).Emil Velikov2016-08-151-0/+1
| | | | | | Otherwise it won't end up in the release tarball. Signed-off-by: Emil Velikov <[email protected]>
* anv/x11: Add support for Xlib platformKevin Strasser2016-08-153-15/+80
| | | | | | | | | | | Some applications continue to use the Xlib client library and expect that VK_KHR_xlib_surface will be available in the driver. Service these applications by converting the Display pointer to xcb_connection_t and use the existing xcb code in the driver. Signed-off-by: Kevin Strasser <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Emil Velikov <[email protected]>
* anv: pipeline: gen7: fix assert in debug modeLionel Landwerlin2016-08-121-0/+4
| | | | | | | | | | SampleMask is only 8bits long on gen7. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97278 Signed-off-by: Lionel Landwerlin <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Anuj Phogat <[email protected]>
* anv/device: Add limits for InterpolationOffsetAnuj Phogat2016-08-121-3/+3
| | | | | | | | Fixes the vulkan cts regression in test dEQP-VK.api.info.device.properties Cc: Mark Janes <[email protected]> Signed-off-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv: Use macro to avoid code duplication for sample positionsAnuj Phogat2016-08-123-92/+106
| | | | | Signed-off-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/device: Enable sample shading on gen7+Anuj Phogat2016-08-091-1/+1
| | | | | | | Passes all 30 min_sample_shading tests in vulkan cts. Signed-off-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/gen7_pipeline: Set multisample state using shared functionAnuj Phogat2016-08-091-16/+4
| | | | | Signed-off-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/pipeline: Add sample locations for gen7-7.5Anuj Phogat2016-08-091-0/+47
| | | | | | | | | | V1: Add multisample positions (Nanley) V2: Fix 8x sample positions to match OpenGL (Anuj) V3: Vulkan has standard sample locations. They need not be same as in OpenGL. (Anuj) Signed-off-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/pipeline: Move emit_ms_state() to genX_pipeline_util.hAnuj Phogat2016-08-092-40/+40
| | | | | | | This will help sharing multisample state setting code. Signed-off-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/pipeline/gen7: Set multisample modesLionel Landwerlin2016-08-083-3/+14
| | | | | | | | | | | | | | | | | Fixes the following failures : dEQP-VK.api.copy_and_blit.resolve_image.whole_4_bit dEQP-VK.api.copy_and_blit.resolve_image.whole_8_bit dEQP-VK.api.copy_and_blit.resolve_image.partial_4_bit dEQP-VK.api.copy_and_blit.resolve_image.partial_8_bit dEQP-VK.api.copy_and_blit.resolve_image.with_regions_4_bit dEQP-VK.api.copy_and_blit.resolve_image.with_regions_8_bit Tested on IVB/HSW Signed-off-by: Lionel Landwerlin <[email protected]> Reviewed-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/pipeline: rename info to rs_info in emit_rs_stateLionel Landwerlin2016-08-081-8/+8
| | | | | | Signed-off-by: Lionel Landwerlin <[email protected]> Reviewed-by: Anuj Phogat <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv/pipeline/gen7: Set the depth format in 3DSTATE_SFJason Ekstrand2016-08-083-2/+25
| | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* anv/pipeline: Unify 3DSTATE_RASTER and 3DSTATE_SF setup between gen7 and gen8Jason Ekstrand2016-08-083-91/+62
| | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* anv/pipeline/gen8: Set 3DSTATE_SF::StatisticsEnableJason Ekstrand2016-08-081-0/+1
| | | | | | | | | We've been setting it in gen7 forever but never in gen8; best to make it consistent. This hasn't caused any problems yet because we don't advertise support for statistics queries yet. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* anv/pipeline/gen8: Unconditionally set DXMultisampleRasterizaitonEnableJason Ekstrand2016-08-081-9/+2
| | | | | | | | | | | | | | | | | | The multisample rasterization mode is computed based on this field, 3DSTATE_RASTER::DXMultisampleRasterizationMode (only for forced multisampling), 3DSTATE_RASTER::APIMode, and the number of samples. There are two tables in the SKL PRM that describe how the final multisample mode is calculated: "Windower (WM) Stage >> Multisampling >> Multisample ModeState >> Table 1" and the formula for "SF_INT::Multisample Rasterization Mode". The "DX Multisample Rasterization Enable" bit changes whether multisample mode is set to OFF_PIXEL or ON_PATTERN in the samples > 1 case. In the samples == 1 case, the bit has no effect. Since Vulkan has no concept of disabling multisampling for samples > 1, we can just set the bit. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* anv/pipeline/gen8: Use fewer designated initializers in emit_rs_stateJason Ekstrand2016-08-081-26/+27
| | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* genxml: Make 3DSTATE_SF more consistent between gen7 and gen8+Jason Ekstrand2016-08-081-2/+2
| | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* anv/pipeline/gen8: Remove an old commentJason Ekstrand2016-08-081-2/+0
| | | | | | | This is now handled in emit_3dstate_clip Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* anv/gen7_pipeline: Set PixelShaderKillPixel for discardsNanley Chery2016-08-051-0/+1
| | | | | | | | | | According to the IVB PRM Vol2 P1, this bit must be set if a pixel shader contains a discard instruction. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97207 Cc: "12.0" <[email protected]> Signed-off-by: Nanley Chery <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* anv: Perform rasterizer discard in the SOL stage instead of the clipper.Kenneth Graunke2016-07-304-3/+12
| | | | | | | | | See commit b0629e6894513a2c49a018bc3342a4e55435a236, where we discovered that the SOL stage's "Rendering Disable" feature is a lot faster at throwing away all geometry than the clipper's "reject all" mode. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-23 19:01:48 +0000