16f9c5da45
This PR resolves the issue of the SPIR-V specification, requiring Shader-coupled capabilities to read/write images in the OpenCL SPIR-V environment, from the perspective of the LLVM SPIR-V backend. See https://github.com/KhronosGroup/SPIRV-Headers/issues/487 for details and discussion. Current implementation correctly reproduces requirements of the SPIR-V specification, however, since the requirements are problematic, out current implementation blocks generation of valid SPIR-V code for compute environments. This PR is to implement a solution discussed at the SPIR-V WG to allow proceeding with generation of valid SPIR-V code for the OpenCL environment and do not impact Vulkan environment at the same time.
58 lines
3.2 KiB
LLVM
58 lines
3.2 KiB
LLVM
; RUN: llc -O0 -mtriple=spirv32-unknown-unknown %s -o - | FileCheck %s --check-prefix=CHECK-SPIRV
|
|
|
|
; StorageImageReadWithoutFormat/StorageImageWriteWithoutFormat implicitly
|
|
; declare Shader, causing a SPIR-V module to be rejected by the OpenCL
|
|
; run-time. See https://github.com/KhronosGroup/SPIRV-Headers/issues/487
|
|
; De-facto, OpImageRead and OpImageWrite are allowed to use Unknown Image
|
|
; Formats when the Kernel capability is declared. We reflect this behavior
|
|
; in the test case, and leave the check under CHECK-SPIRV-NOT to track
|
|
; the issue and follow-up its final resolution when ready.
|
|
; CHECK-SPIRV-NOT: OpCapability StorageImageReadWithoutFormat
|
|
|
|
; CHECK-SPIRV: %[[#]] = OpImageRead %[[#]] %[[#]] %[[#]] Sample %[[#]]
|
|
|
|
define spir_kernel void @sample_test(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0) %source, i32 %sampler, <4 x float> addrspace(1)* nocapture %results) {
|
|
entry:
|
|
%call = tail call spir_func i32 @_Z13get_global_idj(i32 0)
|
|
%call1 = tail call spir_func i32 @_Z13get_global_idj(i32 1)
|
|
%call2 = tail call spir_func i32 @_Z15get_image_width19ocl_image2d_msaa_ro(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0) %source)
|
|
%call3 = tail call spir_func i32 @_Z16get_image_height19ocl_image2d_msaa_ro(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0) %source)
|
|
%call4 = tail call spir_func i32 @_Z21get_image_num_samples19ocl_image2d_msaa_ro(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0) %source)
|
|
%cmp20 = icmp eq i32 %call4, 0
|
|
br i1 %cmp20, label %for.end, label %for.body.lr.ph
|
|
|
|
for.body.lr.ph: ; preds = %entry
|
|
%vecinit = insertelement <2 x i32> undef, i32 %call, i32 0
|
|
%vecinit8 = insertelement <2 x i32> %vecinit, i32 %call1, i32 1
|
|
br label %for.body
|
|
|
|
for.body: ; preds = %for.body.lr.ph, %for.body
|
|
%sample.021 = phi i32 [ 0, %for.body.lr.ph ], [ %inc, %for.body ]
|
|
%mul5 = mul i32 %sample.021, %call3
|
|
%tmp = add i32 %mul5, %call1
|
|
%tmp19 = mul i32 %tmp, %call2
|
|
%add7 = add i32 %tmp19, %call
|
|
%call9 = tail call spir_func <4 x float> @_Z11read_imagef19ocl_image2d_msaa_roDv2_ii(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0) %source, <2 x i32> %vecinit8, i32 %sample.021)
|
|
%arrayidx = getelementptr inbounds <4 x float>, <4 x float> addrspace(1)* %results, i32 %add7
|
|
store <4 x float> %call9, <4 x float> addrspace(1)* %arrayidx, align 16
|
|
%inc = add nuw i32 %sample.021, 1
|
|
%cmp = icmp ult i32 %inc, %call4
|
|
br i1 %cmp, label %for.body, label %for.end.loopexit
|
|
|
|
for.end.loopexit: ; preds = %for.body
|
|
br label %for.end
|
|
|
|
for.end: ; preds = %for.end.loopexit, %entry
|
|
ret void
|
|
}
|
|
|
|
declare spir_func i32 @_Z13get_global_idj(i32)
|
|
|
|
declare spir_func i32 @_Z15get_image_width19ocl_image2d_msaa_ro(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0))
|
|
|
|
declare spir_func i32 @_Z16get_image_height19ocl_image2d_msaa_ro(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0))
|
|
|
|
declare spir_func i32 @_Z21get_image_num_samples19ocl_image2d_msaa_ro(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0))
|
|
|
|
declare spir_func <4 x float> @_Z11read_imagef19ocl_image2d_msaa_roDv2_ii(target("spirv.Image", void, 1, 0, 0, 1, 0, 0, 0), <2 x i32>, i32)
|