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/vk-gl-cts/graphicsfuzz/spv-access-chains/0-opt.wgsl.expected.spvasm

764 lines
30 KiB
Plaintext
Raw Normal View History

Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
; SPIR-V
; Version: 1.3
; Generator: Google Tint Compiler; 0
spir-writer: handle break continuing block The continuing block can exit the loop in very constrained ways: When a break statement is placed such that it would exit from a loop’s § 7.3.8 Continuing Statement, then: - The break statement must appear as either: - The only statement in the if clause of an if statement that has: - no else clause or an empty else clause - no elseif clauses - The only statement in the else clause of an if statement that has an empty if clause and no elseif clauses. - That if statement must appear last in the continuing clause. By design, this allows a lossless round-trip from SPIR-V to WGSL and back to SPIR-V. But that requires this special case construct in WGSL to be translated to an OpBranchConditional with one target being the loop's megre block (which is where 'break' branches to), and the other targets the loop header (which is the loop backedge). That OpBranchConditional takes the place of the normal case of an unconditional backedge. Avoids errors like this: continue construct with the continue target X is not post dominated by the back-edge block Y Fixed: 1034 Change-Id: If472a179380b8d77af746a3cd8e279c8a5e56b37 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/59800 Auto-Submit: David Neto <dneto@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: David Neto <dneto@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-27 15:12:27 +00:00
; Bound: 506
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +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 %main "main" %gl_FragCoord_param_1 %x_GLF_color_1_1
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
OpExecutionMode %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 %gl_FragCoord_param_1 "gl_FragCoord_param_1"
OpName %x_GLF_color_1_1 "x_GLF_color_1_1"
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
OpName %gl_FragCoord "gl_FragCoord"
OpName %buf0 "buf0"
OpMemberName %buf0 0 "resolution"
OpName %x_7 "x_7"
OpName %map "map"
OpName %x_GLF_color "x_GLF_color"
OpName %main_1 "main_1"
OpName %pos "pos"
OpName %ipos "ipos"
OpName %i "i"
OpName %p "p"
OpName %canwalk "canwalk"
OpName %v "v"
OpName %directions "directions"
OpName %j "j"
OpName %d "d"
OpName %x_102 "x_102"
OpName %x_122 "x_122"
OpName %x_142 "x_142"
OpName %x_162 "x_162"
OpName %x_103_phi "x_103_phi"
OpName %x_123_phi "x_123_phi"
OpName %x_143_phi "x_143_phi"
OpName %x_163_phi "x_163_phi"
OpName %x_227 "x_227"
OpName %x_240 "x_240"
OpName %x_279 "x_279"
OpName %x_292 "x_292"
OpName %x_331 "x_331"
OpName %x_344 "x_344"
OpName %x_383 "x_383"
OpName %x_396 "x_396"
OpName %x_228_phi "x_228_phi"
OpName %x_241_phi "x_241_phi"
OpName %x_280_phi "x_280_phi"
OpName %x_293_phi "x_293_phi"
OpName %x_332_phi "x_332_phi"
OpName %x_345_phi "x_345_phi"
OpName %x_384_phi "x_384_phi"
OpName %x_397_phi "x_397_phi"
OpName %main_out "main_out"
OpMemberName %main_out 0 "x_GLF_color_1"
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 %main_inner "main_inner"
OpName %gl_FragCoord_param "gl_FragCoord_param"
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
OpName %main "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 %gl_FragCoord_param_1 BuiltIn FragCoord
OpDecorate %x_GLF_color_1_1 Location 0
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
OpDecorate %buf0 Block
OpMemberDecorate %buf0 0 Offset 0
OpDecorate %x_7 NonWritable
OpDecorate %x_7 DescriptorSet 0
OpDecorate %x_7 Binding 0
OpDecorate %_arr_int_uint_256 ArrayStride 4
OpMemberDecorate %main_out 0 Offset 0
%float = OpTypeFloat 32
%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
%gl_FragCoord_param_1 = OpVariable %_ptr_Input_v4float Input
%_ptr_Output_v4float = OpTypePointer Output %v4float
%7 = OpConstantNull %v4float
%x_GLF_color_1_1 = OpVariable %_ptr_Output_v4float Output %7
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
%_ptr_Private_v4float = OpTypePointer Private %v4float
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
%gl_FragCoord = OpVariable %_ptr_Private_v4float Private %7
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
%v2float = OpTypeVector %float 2
%buf0 = OpTypeStruct %v2float
%_ptr_Uniform_buf0 = OpTypePointer Uniform %buf0
%x_7 = OpVariable %_ptr_Uniform_buf0 Uniform
%int = OpTypeInt 32 1
%uint = OpTypeInt 32 0
%uint_256 = OpConstant %uint 256
%_arr_int_uint_256 = OpTypeArray %int %uint_256
%_ptr_Private__arr_int_uint_256 = OpTypePointer Private %_arr_int_uint_256
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
%20 = OpConstantNull %_arr_int_uint_256
%map = OpVariable %_ptr_Private__arr_int_uint_256 Private %20
%x_GLF_color = OpVariable %_ptr_Private_v4float Private %7
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
%void = OpTypeVoid
%22 = OpTypeFunction %void
%_ptr_Function_v2float = OpTypePointer Function %v2float
%28 = OpConstantNull %v2float
%v2int = OpTypeVector %int 2
%_ptr_Function_v2int = OpTypePointer Function %v2int
%32 = OpConstantNull %v2int
%_ptr_Function_int = OpTypePointer Function %int
%35 = OpConstantNull %int
%bool = OpTypeBool
%_ptr_Function_bool = OpTypePointer Function %bool
%40 = OpConstantNull %bool
%uint_0 = OpConstant %uint 0
%_ptr_Uniform_v2float = OpTypePointer Uniform %v2float
%_ptr_Function_float = OpTypePointer Function %float
%uint_1 = OpConstant %uint 1
%float_16 = OpConstant %float 16
%int_0 = OpConstant %int 0
%int_256 = OpConstant %int 256
%_ptr_Private_int = OpTypePointer Private %int
%int_1 = OpConstant %int 1
%83 = OpConstantComposite %v2int %int_0 %int_0
%true = OpConstantTrue %bool
%int_2 = OpConstant %int 2
%int_16 = OpConstant %int 16
%int_14 = OpConstant %int 14
%false = OpConstantFalse %bool
%int_8 = OpConstant %int 8
%float_1 = OpConstant %float 1
%489 = OpConstantComposite %v4float %float_1 %float_1 %float_1 %float_1
%float_0 = OpConstant %float 0
spir-writer: handle break continuing block The continuing block can exit the loop in very constrained ways: When a break statement is placed such that it would exit from a loop’s § 7.3.8 Continuing Statement, then: - The break statement must appear as either: - The only statement in the if clause of an if statement that has: - no else clause or an empty else clause - no elseif clauses - The only statement in the else clause of an if statement that has an empty if clause and no elseif clauses. - That if statement must appear last in the continuing clause. By design, this allows a lossless round-trip from SPIR-V to WGSL and back to SPIR-V. But that requires this special case construct in WGSL to be translated to an OpBranchConditional with one target being the loop's megre block (which is where 'break' branches to), and the other targets the loop header (which is the loop backedge). That OpBranchConditional takes the place of the normal case of an unconditional backedge. Avoids errors like this: continue construct with the continue target X is not post dominated by the back-edge block Y Fixed: 1034 Change-Id: If472a179380b8d77af746a3cd8e279c8a5e56b37 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/59800 Auto-Submit: David Neto <dneto@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: David Neto <dneto@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-27 15:12:27 +00:00
%492 = OpConstantComposite %v4float %float_0 %float_0 %float_0 %float_1
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
%main_out = OpTypeStruct %v4float
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
%493 = OpTypeFunction %main_out %v4float
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
%main_1 = OpFunction %void None %22
%25 = OpLabel
%pos = OpVariable %_ptr_Function_v2float Function %28
%ipos = OpVariable %_ptr_Function_v2int Function %32
%i = OpVariable %_ptr_Function_int Function %35
%p = OpVariable %_ptr_Function_v2int Function %32
%canwalk = OpVariable %_ptr_Function_bool Function %40
%v = OpVariable %_ptr_Function_int Function %35
%directions = OpVariable %_ptr_Function_int Function %35
%j = OpVariable %_ptr_Function_int Function %35
%d = OpVariable %_ptr_Function_int Function %35
%x_102 = OpVariable %_ptr_Function_bool Function %40
%x_122 = OpVariable %_ptr_Function_bool Function %40
%x_142 = OpVariable %_ptr_Function_bool Function %40
%x_162 = OpVariable %_ptr_Function_bool Function %40
%x_103_phi = OpVariable %_ptr_Function_bool Function %40
%x_123_phi = OpVariable %_ptr_Function_bool Function %40
%x_143_phi = OpVariable %_ptr_Function_bool Function %40
%x_163_phi = OpVariable %_ptr_Function_bool Function %40
%x_227 = OpVariable %_ptr_Function_bool Function %40
%x_240 = OpVariable %_ptr_Function_bool Function %40
%x_279 = OpVariable %_ptr_Function_bool Function %40
%x_292 = OpVariable %_ptr_Function_bool Function %40
%x_331 = OpVariable %_ptr_Function_bool Function %40
%x_344 = OpVariable %_ptr_Function_bool Function %40
%x_383 = OpVariable %_ptr_Function_bool Function %40
%x_396 = OpVariable %_ptr_Function_bool Function %40
%x_228_phi = OpVariable %_ptr_Function_bool Function %40
%x_241_phi = OpVariable %_ptr_Function_bool Function %40
%x_280_phi = OpVariable %_ptr_Function_bool Function %40
%x_293_phi = OpVariable %_ptr_Function_bool Function %40
%x_332_phi = OpVariable %_ptr_Function_bool Function %40
%x_345_phi = OpVariable %_ptr_Function_bool Function %40
%x_384_phi = OpVariable %_ptr_Function_bool Function %40
%x_397_phi = OpVariable %_ptr_Function_bool Function %40
%45 = OpLoad %v4float %gl_FragCoord
%48 = OpAccessChain %_ptr_Uniform_v2float %x_7 %uint_0
%49 = OpLoad %v2float %48
%50 = OpCompositeExtract %float %45 0
%51 = OpCompositeExtract %float %45 1
%52 = OpCompositeConstruct %v2float %50 %51
%53 = OpFDiv %v2float %52 %49
OpStore %pos %53
%55 = OpAccessChain %_ptr_Function_float %pos %uint_0
%56 = OpLoad %float %55
%58 = OpAccessChain %_ptr_Function_float %pos %uint_1
%59 = OpLoad %float %58
%62 = OpFMul %float %56 %float_16
%60 = OpConvertFToS %int %62
%64 = OpFMul %float %59 %float_16
%63 = OpConvertFToS %int %64
%65 = OpCompositeConstruct %v2int %60 %63
OpStore %ipos %65
OpStore %i %int_0
OpBranch %67
%67 = OpLabel
OpLoopMerge %68 %69 None
OpBranch %70
%70 = OpLabel
%71 = OpLoad %int %i
%73 = OpSLessThan %bool %71 %int_256
OpSelectionMerge %74 None
OpBranchConditional %73 %75 %76
%75 = OpLabel
OpBranch %74
%76 = OpLabel
OpBranch %68
%74 = OpLabel
%77 = OpLoad %int %i
%79 = OpAccessChain %_ptr_Private_int %map %77
OpStore %79 %int_0
OpBranch %69
%69 = OpLabel
%80 = OpLoad %int %i
%82 = OpIAdd %int %80 %int_1
OpStore %i %82
OpBranch %67
%68 = OpLabel
OpStore %p %83
OpStore %canwalk %true
OpStore %v %int_0
OpBranch %85
%85 = OpLabel
OpLoopMerge %86 %87 None
OpBranch %88
%88 = OpLabel
%97 = OpLoad %int %v
%98 = OpIAdd %int %97 %int_1
OpStore %v %98
OpStore %directions %int_0
%99 = OpAccessChain %_ptr_Function_int %p %uint_0
%100 = OpLoad %int %99
%101 = OpSGreaterThan %bool %100 %int_0
OpStore %x_103_phi %101
OpSelectionMerge %102 None
OpBranchConditional %101 %103 %102
%103 = OpLabel
%104 = OpAccessChain %_ptr_Function_int %p %uint_0
%105 = OpLoad %int %104
%106 = OpAccessChain %_ptr_Function_int %p %uint_1
%107 = OpLoad %int %106
%109 = OpISub %int %105 %int_2
%111 = OpIMul %int %107 %int_16
%112 = OpIAdd %int %109 %111
%113 = OpAccessChain %_ptr_Private_int %map %112
%114 = OpLoad %int %113
%115 = OpIEqual %bool %114 %int_0
OpStore %x_102 %115
%116 = OpLoad %bool %x_102
OpStore %x_103_phi %116
OpBranch %102
%102 = OpLabel
%117 = OpLoad %bool %x_103_phi
OpSelectionMerge %118 None
OpBranchConditional %117 %119 %118
%119 = OpLabel
%120 = OpLoad %int %directions
%121 = OpIAdd %int %120 %int_1
OpStore %directions %121
OpBranch %118
%118 = OpLabel
%122 = OpAccessChain %_ptr_Function_int %p %uint_1
%123 = OpLoad %int %122
%124 = OpSGreaterThan %bool %123 %int_0
OpStore %x_123_phi %124
OpSelectionMerge %125 None
OpBranchConditional %124 %126 %125
%126 = OpLabel
%127 = OpAccessChain %_ptr_Function_int %p %uint_0
%128 = OpLoad %int %127
%129 = OpAccessChain %_ptr_Function_int %p %uint_1
%130 = OpLoad %int %129
%131 = OpISub %int %130 %int_2
%132 = OpIMul %int %131 %int_16
%133 = OpIAdd %int %128 %132
%134 = OpAccessChain %_ptr_Private_int %map %133
%135 = OpLoad %int %134
%136 = OpIEqual %bool %135 %int_0
OpStore %x_122 %136
%137 = OpLoad %bool %x_122
OpStore %x_123_phi %137
OpBranch %125
%125 = OpLabel
%138 = OpLoad %bool %x_123_phi
OpSelectionMerge %139 None
OpBranchConditional %138 %140 %139
%140 = OpLabel
%141 = OpLoad %int %directions
%142 = OpIAdd %int %141 %int_1
OpStore %directions %142
OpBranch %139
%139 = OpLabel
%143 = OpAccessChain %_ptr_Function_int %p %uint_0
%144 = OpLoad %int %143
%146 = OpSLessThan %bool %144 %int_14
OpStore %x_143_phi %146
OpSelectionMerge %147 None
OpBranchConditional %146 %148 %147
%148 = OpLabel
%149 = OpAccessChain %_ptr_Function_int %p %uint_0
%150 = OpLoad %int %149
%151 = OpAccessChain %_ptr_Function_int %p %uint_1
%152 = OpLoad %int %151
%153 = OpIAdd %int %150 %int_2
%154 = OpIMul %int %152 %int_16
%155 = OpIAdd %int %153 %154
%156 = OpAccessChain %_ptr_Private_int %map %155
%157 = OpLoad %int %156
%158 = OpIEqual %bool %157 %int_0
OpStore %x_142 %158
%159 = OpLoad %bool %x_142
OpStore %x_143_phi %159
OpBranch %147
%147 = OpLabel
%160 = OpLoad %bool %x_143_phi
OpSelectionMerge %161 None
OpBranchConditional %160 %162 %161
%162 = OpLabel
%163 = OpLoad %int %directions
%164 = OpIAdd %int %163 %int_1
OpStore %directions %164
OpBranch %161
%161 = OpLabel
%165 = OpAccessChain %_ptr_Function_int %p %uint_1
%166 = OpLoad %int %165
%167 = OpSLessThan %bool %166 %int_14
OpStore %x_163_phi %167
OpSelectionMerge %168 None
OpBranchConditional %167 %169 %168
%169 = OpLabel
%170 = OpAccessChain %_ptr_Function_int %p %uint_0
%171 = OpLoad %int %170
%172 = OpAccessChain %_ptr_Function_int %p %uint_1
%173 = OpLoad %int %172
%174 = OpIAdd %int %173 %int_2
%175 = OpIMul %int %174 %int_16
%176 = OpIAdd %int %171 %175
%177 = OpAccessChain %_ptr_Private_int %map %176
%178 = OpLoad %int %177
%179 = OpIEqual %bool %178 %int_0
OpStore %x_162 %179
%180 = OpLoad %bool %x_162
OpStore %x_163_phi %180
OpBranch %168
%168 = OpLabel
%181 = OpLoad %bool %x_163_phi
OpSelectionMerge %182 None
OpBranchConditional %181 %183 %182
%183 = OpLabel
%184 = OpLoad %int %directions
%185 = OpIAdd %int %184 %int_1
OpStore %directions %185
OpBranch %182
%182 = OpLabel
%202 = OpLoad %int %directions
%203 = OpIEqual %bool %202 %int_0
OpSelectionMerge %204 None
OpBranchConditional %203 %205 %206
%205 = OpLabel
OpStore %canwalk %false
OpStore %i %int_0
OpBranch %208
%208 = OpLabel
OpLoopMerge %209 %210 None
OpBranch %211
%211 = OpLabel
%212 = OpLoad %int %i
%214 = OpSLessThan %bool %212 %int_8
OpSelectionMerge %215 None
OpBranchConditional %214 %216 %217
%216 = OpLabel
OpBranch %215
%217 = OpLabel
OpBranch %209
%215 = OpLabel
OpStore %j %int_0
OpBranch %218
%218 = OpLabel
OpLoopMerge %219 %220 None
OpBranch %221
%221 = OpLabel
%222 = OpLoad %int %j
%223 = OpSLessThan %bool %222 %int_8
OpSelectionMerge %224 None
OpBranchConditional %223 %225 %226
%225 = OpLabel
OpBranch %224
%226 = OpLabel
OpBranch %219
%224 = OpLabel
%227 = OpLoad %int %j
%228 = OpLoad %int %i
%229 = OpIMul %int %227 %int_2
%230 = OpIMul %int %228 %int_2
%231 = OpIMul %int %230 %int_16
%232 = OpIAdd %int %229 %231
%233 = OpAccessChain %_ptr_Private_int %map %232
%234 = OpLoad %int %233
%235 = OpIEqual %bool %234 %int_0
OpSelectionMerge %236 None
OpBranchConditional %235 %237 %236
%237 = OpLabel
%238 = OpLoad %int %j
%239 = OpAccessChain %_ptr_Function_int %p %uint_0
%240 = OpIMul %int %238 %int_2
OpStore %239 %240
%241 = OpLoad %int %i
%242 = OpAccessChain %_ptr_Function_int %p %uint_1
%243 = OpIMul %int %241 %int_2
OpStore %242 %243
OpStore %canwalk %true
OpBranch %236
%236 = OpLabel
OpBranch %220
%220 = OpLabel
%244 = OpLoad %int %j
%245 = OpIAdd %int %244 %int_1
OpStore %j %245
OpBranch %218
%219 = OpLabel
OpBranch %210
%210 = OpLabel
%246 = OpLoad %int %i
%247 = OpIAdd %int %246 %int_1
OpStore %i %247
OpBranch %208
%209 = OpLabel
%248 = OpAccessChain %_ptr_Function_int %p %uint_0
%249 = OpLoad %int %248
%250 = OpAccessChain %_ptr_Function_int %p %uint_1
%251 = OpLoad %int %250
%252 = OpIMul %int %251 %int_16
%253 = OpIAdd %int %249 %252
%254 = OpAccessChain %_ptr_Private_int %map %253
OpStore %254 %int_1
OpBranch %204
%206 = OpLabel
%255 = OpLoad %int %v
%256 = OpLoad %int %directions
%257 = OpSMod %int %255 %256
OpStore %d %257
%258 = OpLoad %int %directions
%259 = OpLoad %int %v
%260 = OpIAdd %int %259 %258
OpStore %v %260
%261 = OpLoad %int %d
%262 = OpSGreaterThanEqual %bool %261 %int_0
OpStore %x_228_phi %262
OpSelectionMerge %263 None
OpBranchConditional %262 %264 %263
%264 = OpLabel
%265 = OpAccessChain %_ptr_Function_int %p %uint_0
%266 = OpLoad %int %265
%267 = OpSGreaterThan %bool %266 %int_0
OpStore %x_227 %267
%268 = OpLoad %bool %x_227
OpStore %x_228_phi %268
OpBranch %263
%263 = OpLabel
%269 = OpLoad %bool %x_228_phi
OpStore %x_241_phi %269
OpSelectionMerge %270 None
OpBranchConditional %269 %271 %270
%271 = OpLabel
%272 = OpAccessChain %_ptr_Function_int %p %uint_0
%273 = OpLoad %int %272
%274 = OpAccessChain %_ptr_Function_int %p %uint_1
%275 = OpLoad %int %274
%276 = OpISub %int %273 %int_2
%277 = OpIMul %int %275 %int_16
%278 = OpIAdd %int %276 %277
%279 = OpAccessChain %_ptr_Private_int %map %278
%280 = OpLoad %int %279
%281 = OpIEqual %bool %280 %int_0
OpStore %x_240 %281
%282 = OpLoad %bool %x_240
OpStore %x_241_phi %282
OpBranch %270
%270 = OpLabel
%283 = OpLoad %bool %x_241_phi
OpSelectionMerge %284 None
OpBranchConditional %283 %285 %284
%285 = OpLabel
%286 = OpLoad %int %d
%287 = OpISub %int %286 %int_1
OpStore %d %287
%288 = OpAccessChain %_ptr_Function_int %p %uint_0
%289 = OpLoad %int %288
%290 = OpAccessChain %_ptr_Function_int %p %uint_1
%291 = OpLoad %int %290
%292 = OpIMul %int %291 %int_16
%293 = OpIAdd %int %289 %292
%294 = OpAccessChain %_ptr_Private_int %map %293
OpStore %294 %int_1
%295 = OpAccessChain %_ptr_Function_int %p %uint_0
%296 = OpLoad %int %295
%297 = OpAccessChain %_ptr_Function_int %p %uint_1
%298 = OpLoad %int %297
%299 = OpISub %int %296 %int_1
%300 = OpIMul %int %298 %int_16
%301 = OpIAdd %int %299 %300
%302 = OpAccessChain %_ptr_Private_int %map %301
OpStore %302 %int_1
%303 = OpAccessChain %_ptr_Function_int %p %uint_0
%304 = OpLoad %int %303
%305 = OpAccessChain %_ptr_Function_int %p %uint_1
%306 = OpLoad %int %305
%307 = OpISub %int %304 %int_2
%308 = OpIMul %int %306 %int_16
%309 = OpIAdd %int %307 %308
%310 = OpAccessChain %_ptr_Private_int %map %309
OpStore %310 %int_1
%311 = OpAccessChain %_ptr_Function_int %p %uint_0
%312 = OpLoad %int %311
%313 = OpAccessChain %_ptr_Function_int %p %uint_0
%314 = OpISub %int %312 %int_2
OpStore %313 %314
OpBranch %284
%284 = OpLabel
%315 = OpLoad %int %d
%316 = OpSGreaterThanEqual %bool %315 %int_0
OpStore %x_280_phi %316
OpSelectionMerge %317 None
OpBranchConditional %316 %318 %317
%318 = OpLabel
%319 = OpAccessChain %_ptr_Function_int %p %uint_1
%320 = OpLoad %int %319
%321 = OpSGreaterThan %bool %320 %int_0
OpStore %x_279 %321
%322 = OpLoad %bool %x_279
OpStore %x_280_phi %322
OpBranch %317
%317 = OpLabel
%323 = OpLoad %bool %x_280_phi
OpStore %x_293_phi %323
OpSelectionMerge %324 None
OpBranchConditional %323 %325 %324
%325 = OpLabel
%326 = OpAccessChain %_ptr_Function_int %p %uint_0
%327 = OpLoad %int %326
%328 = OpAccessChain %_ptr_Function_int %p %uint_1
%329 = OpLoad %int %328
%330 = OpISub %int %329 %int_2
%331 = OpIMul %int %330 %int_16
%332 = OpIAdd %int %327 %331
%333 = OpAccessChain %_ptr_Private_int %map %332
%334 = OpLoad %int %333
%335 = OpIEqual %bool %334 %int_0
OpStore %x_292 %335
%336 = OpLoad %bool %x_292
OpStore %x_293_phi %336
OpBranch %324
%324 = OpLabel
%337 = OpLoad %bool %x_293_phi
OpSelectionMerge %338 None
OpBranchConditional %337 %339 %338
%339 = OpLabel
%340 = OpLoad %int %d
%341 = OpISub %int %340 %int_1
OpStore %d %341
%342 = OpAccessChain %_ptr_Function_int %p %uint_0
%343 = OpLoad %int %342
%344 = OpAccessChain %_ptr_Function_int %p %uint_1
%345 = OpLoad %int %344
%346 = OpIMul %int %345 %int_16
%347 = OpIAdd %int %343 %346
%348 = OpAccessChain %_ptr_Private_int %map %347
OpStore %348 %int_1
%349 = OpAccessChain %_ptr_Function_int %p %uint_0
%350 = OpLoad %int %349
%351 = OpAccessChain %_ptr_Function_int %p %uint_1
%352 = OpLoad %int %351
%353 = OpISub %int %352 %int_1
%354 = OpIMul %int %353 %int_16
%355 = OpIAdd %int %350 %354
%356 = OpAccessChain %_ptr_Private_int %map %355
OpStore %356 %int_1
%357 = OpAccessChain %_ptr_Function_int %p %uint_0
%358 = OpLoad %int %357
%359 = OpAccessChain %_ptr_Function_int %p %uint_1
%360 = OpLoad %int %359
%361 = OpISub %int %360 %int_2
%362 = OpIMul %int %361 %int_16
%363 = OpIAdd %int %358 %362
%364 = OpAccessChain %_ptr_Private_int %map %363
OpStore %364 %int_1
%365 = OpAccessChain %_ptr_Function_int %p %uint_1
%366 = OpLoad %int %365
%367 = OpAccessChain %_ptr_Function_int %p %uint_1
%368 = OpISub %int %366 %int_2
OpStore %367 %368
OpBranch %338
%338 = OpLabel
%369 = OpLoad %int %d
%370 = OpSGreaterThanEqual %bool %369 %int_0
OpStore %x_332_phi %370
OpSelectionMerge %371 None
OpBranchConditional %370 %372 %371
%372 = OpLabel
%373 = OpAccessChain %_ptr_Function_int %p %uint_0
%374 = OpLoad %int %373
%375 = OpSLessThan %bool %374 %int_14
OpStore %x_331 %375
%376 = OpLoad %bool %x_331
OpStore %x_332_phi %376
OpBranch %371
%371 = OpLabel
%377 = OpLoad %bool %x_332_phi
OpStore %x_345_phi %377
OpSelectionMerge %378 None
OpBranchConditional %377 %379 %378
%379 = OpLabel
%380 = OpAccessChain %_ptr_Function_int %p %uint_0
%381 = OpLoad %int %380
%382 = OpAccessChain %_ptr_Function_int %p %uint_1
%383 = OpLoad %int %382
%384 = OpIAdd %int %381 %int_2
%385 = OpIMul %int %383 %int_16
%386 = OpIAdd %int %384 %385
%387 = OpAccessChain %_ptr_Private_int %map %386
%388 = OpLoad %int %387
%389 = OpIEqual %bool %388 %int_0
OpStore %x_344 %389
%390 = OpLoad %bool %x_344
OpStore %x_345_phi %390
OpBranch %378
%378 = OpLabel
%391 = OpLoad %bool %x_345_phi
OpSelectionMerge %392 None
OpBranchConditional %391 %393 %392
%393 = OpLabel
%394 = OpLoad %int %d
%395 = OpISub %int %394 %int_1
OpStore %d %395
%396 = OpAccessChain %_ptr_Function_int %p %uint_0
%397 = OpLoad %int %396
%398 = OpAccessChain %_ptr_Function_int %p %uint_1
%399 = OpLoad %int %398
%400 = OpIMul %int %399 %int_16
%401 = OpIAdd %int %397 %400
%402 = OpAccessChain %_ptr_Private_int %map %401
OpStore %402 %int_1
%403 = OpAccessChain %_ptr_Function_int %p %uint_0
%404 = OpLoad %int %403
%405 = OpAccessChain %_ptr_Function_int %p %uint_1
%406 = OpLoad %int %405
%407 = OpIAdd %int %404 %int_1
%408 = OpIMul %int %406 %int_16
%409 = OpIAdd %int %407 %408
%410 = OpAccessChain %_ptr_Private_int %map %409
OpStore %410 %int_1
%411 = OpAccessChain %_ptr_Function_int %p %uint_0
%412 = OpLoad %int %411
%413 = OpAccessChain %_ptr_Function_int %p %uint_1
%414 = OpLoad %int %413
%415 = OpIAdd %int %412 %int_2
%416 = OpIMul %int %414 %int_16
%417 = OpIAdd %int %415 %416
%418 = OpAccessChain %_ptr_Private_int %map %417
OpStore %418 %int_1
%419 = OpAccessChain %_ptr_Function_int %p %uint_0
%420 = OpLoad %int %419
%421 = OpAccessChain %_ptr_Function_int %p %uint_0
%422 = OpIAdd %int %420 %int_2
OpStore %421 %422
OpBranch %392
%392 = OpLabel
%423 = OpLoad %int %d
%424 = OpSGreaterThanEqual %bool %423 %int_0
OpStore %x_384_phi %424
OpSelectionMerge %425 None
OpBranchConditional %424 %426 %425
%426 = OpLabel
%427 = OpAccessChain %_ptr_Function_int %p %uint_1
%428 = OpLoad %int %427
%429 = OpSLessThan %bool %428 %int_14
OpStore %x_383 %429
%430 = OpLoad %bool %x_383
OpStore %x_384_phi %430
OpBranch %425
%425 = OpLabel
%431 = OpLoad %bool %x_384_phi
OpStore %x_397_phi %431
OpSelectionMerge %432 None
OpBranchConditional %431 %433 %432
%433 = OpLabel
%434 = OpAccessChain %_ptr_Function_int %p %uint_0
%435 = OpLoad %int %434
%436 = OpAccessChain %_ptr_Function_int %p %uint_1
%437 = OpLoad %int %436
%438 = OpIAdd %int %437 %int_2
%439 = OpIMul %int %438 %int_16
%440 = OpIAdd %int %435 %439
%441 = OpAccessChain %_ptr_Private_int %map %440
%442 = OpLoad %int %441
%443 = OpIEqual %bool %442 %int_0
OpStore %x_396 %443
%444 = OpLoad %bool %x_396
OpStore %x_397_phi %444
OpBranch %432
%432 = OpLabel
%445 = OpLoad %bool %x_397_phi
OpSelectionMerge %446 None
OpBranchConditional %445 %447 %446
%447 = OpLabel
%448 = OpLoad %int %d
%449 = OpISub %int %448 %int_1
OpStore %d %449
%450 = OpAccessChain %_ptr_Function_int %p %uint_0
%451 = OpLoad %int %450
%452 = OpAccessChain %_ptr_Function_int %p %uint_1
%453 = OpLoad %int %452
%454 = OpIMul %int %453 %int_16
%455 = OpIAdd %int %451 %454
%456 = OpAccessChain %_ptr_Private_int %map %455
OpStore %456 %int_1
%457 = OpAccessChain %_ptr_Function_int %p %uint_0
%458 = OpLoad %int %457
%459 = OpAccessChain %_ptr_Function_int %p %uint_1
%460 = OpLoad %int %459
%461 = OpIAdd %int %460 %int_1
%462 = OpIMul %int %461 %int_16
%463 = OpIAdd %int %458 %462
%464 = OpAccessChain %_ptr_Private_int %map %463
OpStore %464 %int_1
%465 = OpAccessChain %_ptr_Function_int %p %uint_0
%466 = OpLoad %int %465
%467 = OpAccessChain %_ptr_Function_int %p %uint_1
%468 = OpLoad %int %467
%469 = OpIAdd %int %468 %int_2
%470 = OpIMul %int %469 %int_16
%471 = OpIAdd %int %466 %470
%472 = OpAccessChain %_ptr_Private_int %map %471
OpStore %472 %int_1
%473 = OpAccessChain %_ptr_Function_int %p %uint_1
%474 = OpLoad %int %473
%475 = OpAccessChain %_ptr_Function_int %p %uint_1
%476 = OpIAdd %int %474 %int_2
OpStore %475 %476
OpBranch %446
%446 = OpLabel
OpBranch %204
%204 = OpLabel
%477 = OpAccessChain %_ptr_Function_int %ipos %uint_1
%478 = OpLoad %int %477
%479 = OpAccessChain %_ptr_Function_int %ipos %uint_0
%480 = OpLoad %int %479
%481 = OpIMul %int %478 %int_16
%482 = OpIAdd %int %481 %480
%483 = OpAccessChain %_ptr_Private_int %map %482
%484 = OpLoad %int %483
%485 = OpIEqual %bool %484 %int_1
OpSelectionMerge %486 None
OpBranchConditional %485 %487 %486
%487 = OpLabel
OpStore %x_GLF_color %489
OpReturn
%486 = OpLabel
OpBranch %87
%87 = OpLabel
%490 = OpLoad %bool %canwalk
spir-writer: handle break continuing block The continuing block can exit the loop in very constrained ways: When a break statement is placed such that it would exit from a loop’s § 7.3.8 Continuing Statement, then: - The break statement must appear as either: - The only statement in the if clause of an if statement that has: - no else clause or an empty else clause - no elseif clauses - The only statement in the else clause of an if statement that has an empty if clause and no elseif clauses. - That if statement must appear last in the continuing clause. By design, this allows a lossless round-trip from SPIR-V to WGSL and back to SPIR-V. But that requires this special case construct in WGSL to be translated to an OpBranchConditional with one target being the loop's megre block (which is where 'break' branches to), and the other targets the loop header (which is the loop backedge). That OpBranchConditional takes the place of the normal case of an unconditional backedge. Avoids errors like this: continue construct with the continue target X is not post dominated by the back-edge block Y Fixed: 1034 Change-Id: If472a179380b8d77af746a3cd8e279c8a5e56b37 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/59800 Auto-Submit: David Neto <dneto@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: David Neto <dneto@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-27 15:12:27 +00:00
OpBranchConditional %490 %85 %86
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
%86 = OpLabel
spir-writer: handle break continuing block The continuing block can exit the loop in very constrained ways: When a break statement is placed such that it would exit from a loop’s § 7.3.8 Continuing Statement, then: - The break statement must appear as either: - The only statement in the if clause of an if statement that has: - no else clause or an empty else clause - no elseif clauses - The only statement in the else clause of an if statement that has an empty if clause and no elseif clauses. - That if statement must appear last in the continuing clause. By design, this allows a lossless round-trip from SPIR-V to WGSL and back to SPIR-V. But that requires this special case construct in WGSL to be translated to an OpBranchConditional with one target being the loop's megre block (which is where 'break' branches to), and the other targets the loop header (which is the loop backedge). That OpBranchConditional takes the place of the normal case of an unconditional backedge. Avoids errors like this: continue construct with the continue target X is not post dominated by the back-edge block Y Fixed: 1034 Change-Id: If472a179380b8d77af746a3cd8e279c8a5e56b37 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/59800 Auto-Submit: David Neto <dneto@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: David Neto <dneto@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-27 15:12:27 +00:00
OpStore %x_GLF_color %492
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
OpReturn
OpFunctionEnd
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
%main_inner = OpFunction %main_out None %493
%gl_FragCoord_param = OpFunctionParameter %v4float
spir-writer: handle break continuing block The continuing block can exit the loop in very constrained ways: When a break statement is placed such that it would exit from a loop’s § 7.3.8 Continuing Statement, then: - The break statement must appear as either: - The only statement in the if clause of an if statement that has: - no else clause or an empty else clause - no elseif clauses - The only statement in the else clause of an if statement that has an empty if clause and no elseif clauses. - That if statement must appear last in the continuing clause. By design, this allows a lossless round-trip from SPIR-V to WGSL and back to SPIR-V. But that requires this special case construct in WGSL to be translated to an OpBranchConditional with one target being the loop's megre block (which is where 'break' branches to), and the other targets the loop header (which is the loop backedge). That OpBranchConditional takes the place of the normal case of an unconditional backedge. Avoids errors like this: continue construct with the continue target X is not post dominated by the back-edge block Y Fixed: 1034 Change-Id: If472a179380b8d77af746a3cd8e279c8a5e56b37 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/59800 Auto-Submit: David Neto <dneto@google.com> Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: David Neto <dneto@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-27 15:12:27 +00:00
%497 = OpLabel
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
OpStore %gl_FragCoord %gl_FragCoord_param
%498 = OpFunctionCall %void %main_1
%499 = OpLoad %v4float %x_GLF_color
%500 = OpCompositeConstruct %main_out %499
OpReturnValue %500
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
OpFunctionEnd
%main = OpFunction %void None %22
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
%502 = OpLabel
%504 = OpLoad %v4float %gl_FragCoord_param_1
%503 = OpFunctionCall %main_out %main_inner %504
%505 = OpCompositeExtract %v4float %503 0
OpStore %x_GLF_color_1_1 %505
Add tests derived from VK-GL-CTS This adds SPIR-V assembly and WGSL tests derived from VK-GL-CTS commit 571256871c2e2f03995373e1e4a02958d8cd8cf5. The following procedure was followed: - Those .amber files in VK-GL-CTS wholly owned by Google were identified - All GLSL and SPIR-V shaders were extracted from the Amber files and converted into SPIR-V binaries - The compact-ids pass of spirv-opt was applied to each binary - Duplicate binaries were removed - spirv-opt -O was used to obtain an optimized version of each remaining binary, with duplicates discarded - Binaries that failed validation using spirv-val with target environment SPIR-V 1.3 were discarded - Those binaries that tint could not successfully convert into WGSL were put aside for further investigation - SPIR-V assembly versions of the remaining binaries are included in this CL - test-runner with -generate-expected and -generate-skip was used to generate expected .spvasm, .msl, .hlsl and .wgsl outputs for these SPIR-V assembly tests - Each successfully-generated .expected.wgsl is included in this CL again, as a WGLSL test - test-runner with -generate-expected and -generate-skip was used again, to generate expected outputs for these WGSL tests Change-Id: Ibe9baf2729cf97e0b633db9a426f53362a5de540 Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/58842 Kokoro: Kokoro <noreply+kokoro@google.com> Commit-Queue: Ben Clayton <bclayton@google.com> Reviewed-by: Ben Clayton <bclayton@google.com>
2021-07-23 13:10:12 +00:00
OpReturn
OpFunctionEnd