encounter
/
dawn-cmake
mirror of https://github.com/encounter/dawn-cmake.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
dawn-cmake/test/tint/shader_io/shared_struct_storage_buffe...

90 lines
3.8 KiB
Plaintext
Raw Normal View History

test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
; SPIR-V
; Version: 1.3
; Generator: Google Tint Compiler; 0
tint: Add PreservePadding transform This is used to ensure that assignments to host-visible memory do not modify padding bytes in structures and arrays. We decompose assignments of whole structure and array types into member-wise or element-wise copies, using helper functions. This is used in all backends except HLSL, which already decomposes memory accesses. Bug: tint:1571 Change-Id: Id6de2f917fb80151cc654a7e1c8413ae956f0d61 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/112720 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com>
2022-12-06 18:32:19 +00:00
; Bound: 46
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
; Schema: 0
OpCapability Shader
OpMemoryModel Logical GLSL450
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
OpEntryPoint Fragment %frag_main "frag_main" %f_1 %u_1 %v_1
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
OpExecutionMode %frag_main OriginUpperLeft
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
OpName %f_1 "f_1"
OpName %u_1 "u_1"
OpName %v_1 "v_1"
GLSL: Change Add[Spirv]BlockAttribute to support GLSL Modify the AddSpirvBlockAttribute transform to fix top-level structure access of uniform, storage and push-constant buffers for use in the GLSL backend. The small change to the transform makes the transform wrap host-sharable buffers, if they're also used as a non-host-sharable structure. Also rename the transform to AddBlockAttrbibute in order to reflect its wider applicability. Change-Id: Ib2bf4ebf6bce72790791dbae9387032be765e4b9 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101061 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Stephen White <senorblanco@chromium.org>
2022-09-02 19:19:10 +00:00
OpName %output_block "output_block"
OpMemberName %output_block 0 "inner"
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
OpName %S "S"
OpMemberName %S 0 "f"
OpMemberName %S 1 "u"
OpMemberName %S 2 "v"
OpName %output "output"
tint: Add PreservePadding transform This is used to ensure that assignments to host-visible memory do not modify padding bytes in structures and arrays. We decompose assignments of whole structure and array types into member-wise or element-wise copies, using helper functions. This is used in all backends except HLSL, which already decomposes memory accesses. Bug: tint:1571 Change-Id: Id6de2f917fb80151cc654a7e1c8413ae956f0d61 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/112720 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com>
2022-12-06 18:32:19 +00:00
OpName %assign_and_preserve_padding_output "assign_and_preserve_padding_output"
OpName %value "value"
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
OpName %frag_main_inner "frag_main_inner"
OpName %input "input"
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
OpName %frag_main "frag_main"
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
OpDecorate %f_1 Location 0
OpDecorate %u_1 Location 1
OpDecorate %u_1 Flat
OpDecorate %v_1 BuiltIn FragCoord
GLSL: Change Add[Spirv]BlockAttribute to support GLSL Modify the AddSpirvBlockAttribute transform to fix top-level structure access of uniform, storage and push-constant buffers for use in the GLSL backend. The small change to the transform makes the transform wrap host-sharable buffers, if they're also used as a non-host-sharable structure. Also rename the transform to AddBlockAttrbibute in order to reflect its wider applicability. Change-Id: Ib2bf4ebf6bce72790791dbae9387032be765e4b9 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101061 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Stephen White <senorblanco@chromium.org>
2022-09-02 19:19:10 +00:00
OpDecorate %output_block Block
OpMemberDecorate %output_block 0 Offset 0
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
OpMemberDecorate %S 0 Offset 0
OpMemberDecorate %S 1 Offset 4
OpMemberDecorate %S 2 Offset 128
OpDecorate %output DescriptorSet 0
OpDecorate %output Binding 0
%float = OpTypeFloat 32
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
%_ptr_Input_float = OpTypePointer Input %float
%f_1 = OpVariable %_ptr_Input_float Input
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
%uint = OpTypeInt 32 0
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
%_ptr_Input_uint = OpTypePointer Input %uint
%u_1 = OpVariable %_ptr_Input_uint Input
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
%v4float = OpTypeVector %float 4
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
%_ptr_Input_v4float = OpTypePointer Input %v4float
%v_1 = OpVariable %_ptr_Input_v4float Input
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
%S = OpTypeStruct %float %uint %v4float
GLSL: Change Add[Spirv]BlockAttribute to support GLSL Modify the AddSpirvBlockAttribute transform to fix top-level structure access of uniform, storage and push-constant buffers for use in the GLSL backend. The small change to the transform makes the transform wrap host-sharable buffers, if they're also used as a non-host-sharable structure. Also rename the transform to AddBlockAttrbibute in order to reflect its wider applicability. Change-Id: Ib2bf4ebf6bce72790791dbae9387032be765e4b9 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101061 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Stephen White <senorblanco@chromium.org>
2022-09-02 19:19:10 +00:00
%output_block = OpTypeStruct %S
%_ptr_StorageBuffer_output_block = OpTypePointer StorageBuffer %output_block
%output = OpVariable %_ptr_StorageBuffer_output_block StorageBuffer
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
%void = OpTypeVoid
GLSL: Change Add[Spirv]BlockAttribute to support GLSL Modify the AddSpirvBlockAttribute transform to fix top-level structure access of uniform, storage and push-constant buffers for use in the GLSL backend. The small change to the transform makes the transform wrap host-sharable buffers, if they're also used as a non-host-sharable structure. Also rename the transform to AddBlockAttrbibute in order to reflect its wider applicability. Change-Id: Ib2bf4ebf6bce72790791dbae9387032be765e4b9 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101061 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Stephen White <senorblanco@chromium.org>
2022-09-02 19:19:10 +00:00
%14 = OpTypeFunction %void %S
%uint_0 = OpConstant %uint 0
tint: Add PreservePadding transform This is used to ensure that assignments to host-visible memory do not modify padding bytes in structures and arrays. We decompose assignments of whole structure and array types into member-wise or element-wise copies, using helper functions. This is used in all backends except HLSL, which already decomposes memory accesses. Bug: tint:1571 Change-Id: Id6de2f917fb80151cc654a7e1c8413ae956f0d61 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/112720 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com>
2022-12-06 18:32:19 +00:00
%_ptr_StorageBuffer_float = OpTypePointer StorageBuffer %float
%uint_1 = OpConstant %uint 1
%_ptr_StorageBuffer_uint = OpTypePointer StorageBuffer %uint
%uint_2 = OpConstant %uint 2
%_ptr_StorageBuffer_v4float = OpTypePointer StorageBuffer %v4float
%38 = OpTypeFunction %void
%assign_and_preserve_padding_output = OpFunction %void None %14
%value = OpFunctionParameter %S
%18 = OpLabel
%21 = OpAccessChain %_ptr_StorageBuffer_float %output %uint_0 %uint_0
%22 = OpCompositeExtract %float %value 0
OpStore %21 %22
%25 = OpAccessChain %_ptr_StorageBuffer_uint %output %uint_0 %uint_1
%26 = OpCompositeExtract %uint %value 1
OpStore %25 %26
%29 = OpAccessChain %_ptr_StorageBuffer_v4float %output %uint_0 %uint_2
%30 = OpCompositeExtract %v4float %value 2
OpStore %29 %30
OpReturn
OpFunctionEnd
GLSL: Change Add[Spirv]BlockAttribute to support GLSL Modify the AddSpirvBlockAttribute transform to fix top-level structure access of uniform, storage and push-constant buffers for use in the GLSL backend. The small change to the transform makes the transform wrap host-sharable buffers, if they're also used as a non-host-sharable structure. Also rename the transform to AddBlockAttrbibute in order to reflect its wider applicability. Change-Id: Ib2bf4ebf6bce72790791dbae9387032be765e4b9 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/101061 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Stephen White <senorblanco@chromium.org>
2022-09-02 19:19:10 +00:00
%frag_main_inner = OpFunction %void None %14
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
%input = OpFunctionParameter %S
tint: Add PreservePadding transform This is used to ensure that assignments to host-visible memory do not modify padding bytes in structures and arrays. We decompose assignments of whole structure and array types into member-wise or element-wise copies, using helper functions. This is used in all backends except HLSL, which already decomposes memory accesses. Bug: tint:1571 Change-Id: Id6de2f917fb80151cc654a7e1c8413ae956f0d61 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/112720 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com>
2022-12-06 18:32:19 +00:00
%33 = OpLabel
%34 = OpCompositeExtract %float %input 0
%35 = OpCompositeExtract %uint %input 1
%36 = OpCompositeExtract %v4float %input 2
%37 = OpFunctionCall %void %assign_and_preserve_padding_output %input
spirv: Use generic transform to process shader IO The refactored CanonicalizeEntryPointIO transform makes it much easier to handle SPIR-V style IO as well, and doing this removes a lot of duplicated code. Remove all of the SPIR-V transform code for shader IO and vertex point size. Bug: tint:920 Change-Id: Id1b97517619b4d2fd09b45d5aee848259f3dfa77 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60840 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com> Auto-Submit: James Price <jrprice@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-08-05 17:34:19 +00:00
OpReturn
OpFunctionEnd
tint: Add PreservePadding transform This is used to ensure that assignments to host-visible memory do not modify padding bytes in structures and arrays. We decompose assignments of whole structure and array types into member-wise or element-wise copies, using helper functions. This is used in all backends except HLSL, which already decomposes memory accesses. Bug: tint:1571 Change-Id: Id6de2f917fb80151cc654a7e1c8413ae956f0d61 Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/112720 Reviewed-by: Ben Clayton <bclayton@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: James Price <jrprice@google.com>
2022-12-06 18:32:19 +00:00
%frag_main = OpFunction %void None %38
%40 = OpLabel
%42 = OpLoad %float %f_1
%43 = OpLoad %uint %u_1
%44 = OpLoad %v4float %v_1
%45 = OpCompositeConstruct %S %42 %43 %44
%41 = OpFunctionCall %void %frag_main_inner %45
test: Add E2E test coverage for shader IO This provides much more complete coverage than what we have in the unit tests. We now test: - All builtins, for all stages, both struct and non-struct - Multiple location attributes for vertex and fragment stages, both struct and non-struct - Mixing builtins and location attributes, whilst mixing struct and non-struct - A few "interesting" cases of IO structs being shared between different functions, stages, and with an SSBO variable There are 7 skipped tests for MSL due to two different MSL bugs which will be fixed in upcoming patches. Change-Id: I8b802591762c8ff018e01bf37838551e353162b1 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53120 Kokoro: Kokoro <noreply+kokoro@google.com> Reviewed-by: Ben Clayton <bclayton@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Auto-Submit: James Price <jrprice@google.com>
2021-06-03 09:38:34 +00:00
OpReturn
OpFunctionEnd