search

iree-org / iree

A retargetable MLIR-based machine learning compiler and runtime toolkit.
http://iree.dev/
Apache License 2.0
2.56k stars 568 forks source link

(HIP) Lowering issues with SDPA when compiling for gfx1100. #17032

Open monorimet opened 4 months ago

monorimet commented 4 months ago

What happened?

I am experiencing issues trying to lower the linalg_ext.attention op for gfx1100 target, notably on SDXL unet. (Radeon 7900xtx)

The failing attention dispatch is at index 78 (save as dispatch_78.mlir for repro):

hal.executable public @main$async_dispatch_78 {
  hal.executable.variant public @rocm_hsaco_fb target(<"rocm", "rocm-hsaco-fb", {mma_intrinsics = [#iree_gpu.mma_layout<WMMA_F16_16x16x16_F32>], target_arch = "gfx1100", ukernels = "none"}>) {
    hal.executable.export public @main$async_dispatch_78_attention_20x4096x64xf16 ordinal(0) layout(#hal.pipeline.layout<push_constants = 4, sets = [<0, bindings = [<0, storage_buffer, ReadOnly>, <1, storage_buffer>]>]>) attributes {hal.interface.bindings = [#hal.interface.binding<0, 0>, #hal.interface.binding<0, 1>]} {
    ^bb0(%arg0: !hal.device):
      %x, %y, %z = flow.dispatch.workgroup_count_from_slice 
      hal.return %x, %y, %z : index, index, index
    }
    builtin.module {
      func.func @main$async_dispatch_78_attention_20x4096x64xf16() {
        %cst = arith.constant 1.250000e-01 : f16
        %0 = hal.interface.constant.load[0] : i32
        %1 = hal.interface.constant.load[1] : i32
        %2 = hal.interface.constant.load[2] : i32
        %3 = hal.interface.constant.load[3] : i32
        %4 = arith.index_castui %0 : i32 to index
        %5 = arith.index_castui %1 : i32 to index
        %6 = arith.index_castui %2 : i32 to index
        %7 = arith.index_castui %3 : i32 to index
        %8 = hal.interface.binding.subspan set(0) binding(0) type(storage_buffer) alignment(64) offset(%4) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>>
        %9 = hal.interface.binding.subspan set(0) binding(0) type(storage_buffer) alignment(64) offset(%5) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>>
        %10 = hal.interface.binding.subspan set(0) binding(0) type(storage_buffer) alignment(64) offset(%6) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>>
        %11 = hal.interface.binding.subspan set(0) binding(1) type(storage_buffer) alignment(64) offset(%7) : !flow.dispatch.tensor<writeonly:tensor<20x4096x64xf16>>
        %12 = flow.dispatch.tensor.load %8, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>> -> tensor<20x4096x64xf16>
        %13 = flow.dispatch.tensor.load %9, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>> -> tensor<20x4096x64xf16>
        %14 = flow.dispatch.tensor.load %10, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>> -> tensor<20x4096x64xf16>
        %15 = tensor.empty() : tensor<20x4096x64xf16>
        %16 = iree_linalg_ext.attention ins(%12, %13, %14, %cst : tensor<20x4096x64xf16>, tensor<20x4096x64xf16>, tensor<20x4096x64xf16>, f16) outs(%15 : tensor<20x4096x64xf16>) -> tensor<20x4096x64xf16>
        flow.dispatch.tensor.store %16, %11, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : tensor<20x4096x64xf16> -> !flow.dispatch.tensor<writeonly:tensor<20x4096x64xf16>>
        return
      }
    }
  }
}

The error:

iree.compiler.tools.binaries.CompilerToolError: Error invoking IREE compiler tool iree-compile.exe
Error code: 1
Diagnostics:
failed to translate executables
./unetdps\configured_compiled_unet_main$async_dispatch_78.mlir:9:6: error: Failures have been detected while processing an MLIR pass pipeline
      func.func @main$async_dispatch_78_attention_20x4096x64xf16() {
     ^
./unetdps\configured_compiled_unet_main$async_dispatch_78.mlir:9:6: note: Pipeline failed while executing [`TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_50, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_56, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_63, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_64, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_68, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_69, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_70, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_71, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_72, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `LLVMGPULowerExecutableTarget` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_75, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_78, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `ConvertToROCDL` on 'builtin.module' operation, `LLVMGPUVectorLowering` on 'func.func' operation: @main$async_dispatch_64_conv_2d_nchw_fchw_2x640x64x64x320x1x1_f16xf16xf32, `TranslateExecutablesPass` on 'hal.executable' operation: @main$async_dispatch_79, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `EmulateNarrowType` on 'builtin.module' operation, `GPUDistributeSharedMemoryCopy` on 'func.func' operation: @main$async_dispatch_72_matmul_transpose_b_8192x640x640_f16xf16xf32, `LLVMGPUVectorLowering` on 'func.func' operation: @main$async_dispatch_70_generic_8192x640_f32, `ConvertBf16ArithToF32` on 'builtin.module' operation, `LinalgExtToLoops` on 'func.func' operation: @main$async_dispatch_78_attention_20x4096x64xf16, `Canonicalizer` on 'builtin.module' operation, `PolynomialApproximationPass` on 'func.func' operation: @main$async_dispatch_71_generic_8192x640_f16xf32xf32xf16xf16xf16, `FoldTensorExtractOp` on 'builtin.module' operation]: reproducer generated at `./shark_tmp/core-reproducer.mlir`
./unetdps\configured_compiled_unet_main$async_dispatch_78.mlir:2:2: error: failed to run translation of source executable to target executable for backend #hal.executable.target<"rocm", "rocm-hsaco-fb", {mma_intrinsics = [#iree_gpu.mma_layout<WMMA_F16_16x16x16_F32>], target_arch = "gfx1100", ukernels = "none"}>
  hal.executable.variant public @rocm_hsaco_fb target(<"rocm", "rocm-hsaco-fb", {mma_intrinsics = [#iree_gpu.mma_layout<WMMA_F16_16x16x16_F32>], target_arch = "gfx1100", ukernels = "none"}>) {

Steps to reproduce your issue

iree-compile.exe dispatch_78.mlir --iree-input-type=torch --iree-vm-bytecode-module-output-format=flatbuffer-binary --iree-hal-target-backends=rocm --mlir-print-debuginfo --iree-hal-target-backends=rocm --iree-rocm-target-chip=gfx1100 --iree-opt-const-eval=false --iree-opt-data-tiling=False

What component(s) does this issue relate to?

No response

Version information

IREE (https://iree.dev): openxla/iree@5d8907e82f LLVM version 19.0.0git Optimized build with assertions

Additional context

No response

monorimet commented 4 months ago

unet_dump.txt Here is an IR dump obtained with --mlir-print-ir-after-all

Groverkss commented 4 months ago

Are you using the transform dialect spec for attention? Currently, attention can only lower using the transform dialect spec. The C++ pipeline is being worked out.

monorimet commented 4 months ago

Are you using the transform dialect spec for attention? Currently, attention can only lower using the transform dialect spec. The C++ pipeline is being worked out.

Without the transform spec the attention op should be tiled and decomposed in IREE. This path is verified to work on gfx90a, and IIUC should be the default path for ROCM targets where attention is not supported yet.

monorimet commented 4 months ago

Here's updated IR for config A unet, as well as the attention dispatch that fails to lower. It seems we have two options:

  1. have wmma implementation of attention, either by TD script or by C++ pipeline
  2. fix tiling and decomposition of attention ops for this target

The attention dispatch:

module attributes {hal.device.targets = [#hal.device.target<"rocm", {legacy_sync}, [#hal.executable.target<"rocm", "rocm-hsaco-fb", {mma_intrinsics = [#iree_gpu.mma_layout<WMMA_F16_16x16x16_F32>], target_arch = "gfx1100", ukernels = "none"}>]>]} {
  hal.executable private @run_forward$async_dispatch_73 {
    hal.executable.variant public @rocm_hsaco_fb target(<"rocm", "rocm-hsaco-fb", {mma_intrinsics = [#iree_gpu.mma_layout<WMMA_F16_16x16x16_F32>], target_arch = "gfx1100", ukernels = "none"}>) {
      hal.executable.export public @run_forward$async_dispatch_73_attention_20x4096x64xf16 ordinal(0) layout(#hal.pipeline.layout<push_constants = 4, sets = [<0, bindings = [<0, storage_buffer, ReadOnly>, <1, storage_buffer>]>]>) attributes {hal.interface.bindings = [#hal.interface.binding<0, 0>, #hal.interface.binding<0, 1>]} {
      ^bb0(%arg0: !hal.device):
        %x, %y, %z = flow.dispatch.workgroup_count_from_slice 
        hal.return %x, %y, %z : index, index, index
      }
      builtin.module {
        func.func @run_forward$async_dispatch_73_attention_20x4096x64xf16() attributes {translation_info = #iree_codegen.translation_info<LLVMGPUDistribute workgroup_size = [64, 1, 1] subgroup_size = 32>} {
          %cst = arith.constant 1.250000e-01 : f16
          %0 = hal.interface.constant.load[0] : i32
          %1 = hal.interface.constant.load[1] : i32
          %2 = hal.interface.constant.load[2] : i32
          %3 = hal.interface.constant.load[3] : i32
          %4 = arith.index_castui %0 : i32 to index
          %5 = arith.index_castui %1 : i32 to index
          %6 = arith.index_castui %2 : i32 to index
          %7 = arith.index_castui %3 : i32 to index
          %8 = hal.interface.binding.subspan set(0) binding(0) type(storage_buffer) alignment(64) offset(%4) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>>
          %9 = hal.interface.binding.subspan set(0) binding(0) type(storage_buffer) alignment(64) offset(%5) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>>
          %10 = hal.interface.binding.subspan set(0) binding(0) type(storage_buffer) alignment(64) offset(%6) flags(ReadOnly) : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>>
          %11 = hal.interface.binding.subspan set(0) binding(1) type(storage_buffer) alignment(64) offset(%7) : !flow.dispatch.tensor<writeonly:tensor<20x4096x64xf16>>
          %12 = flow.dispatch.tensor.load %8, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>> -> tensor<20x4096x64xf16>
          %13 = flow.dispatch.tensor.load %9, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>> -> tensor<20x4096x64xf16>
          %14 = flow.dispatch.tensor.load %10, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : !flow.dispatch.tensor<readonly:tensor<20x4096x64xf16>> -> tensor<20x4096x64xf16>
          %15 = tensor.empty() : tensor<20x4096x64xf16>
          %16 = iree_linalg_ext.attention {lowering_config = #iree_codegen.lowering_config<tile_sizes = [[1, 64]]>} ins(%12, %13, %14, %cst : tensor<20x4096x64xf16>, tensor<20x4096x64xf16>, tensor<20x4096x64xf16>, f16) outs(%15 : tensor<20x4096x64xf16>) -> tensor<20x4096x64xf16>
          flow.dispatch.tensor.store %16, %11, offsets = [0, 0, 0], sizes = [20, 4096, 64], strides = [1, 1, 1] : tensor<20x4096x64xf16> -> !flow.dispatch.tensor<writeonly:tensor<20x4096x64xf16>>
          return
        }
      }
    }
  }
  util.global private mutable @run_forward$async_dispatch_73_rocm_hsaco_fb_run_forward$async_dispatch_73_attention_20x4096x64xf16_buffer : !hal.buffer
  util.initializer {
    %c846622208 = arith.constant 846622208 : index
    %c-1_i64 = arith.constant -1 : i64
    %c0 = arith.constant 0 : index
    %device_0 = hal.devices.get %c0 : !hal.device
    %allocator = hal.device.allocator<%device_0 : !hal.device> : !hal.allocator
    %buffer = hal.allocator.allocate<%allocator : !hal.allocator> affinity(%c-1_i64) type("DeviceVisible|DeviceLocal") usage("TransferSource|TransferTarget|Transfer|DispatchStorageRead|DispatchStorageWrite|DispatchStorage") : !hal.buffer{%c846622208}
    util.global.store %buffer, @run_forward$async_dispatch_73_rocm_hsaco_fb_run_forward$async_dispatch_73_attention_20x4096x64xf16_buffer : !hal.buffer
    util.return
  }
  util.func public @run_forward$async_dispatch_73_rocm_hsaco_fb_run_forward$async_dispatch_73_attention_20x4096x64xf16(%arg0: i32) attributes {iree.abi.stub, iree.reflection = {iree.benchmark = "dispatch"}} {
    %c-1_i32 = arith.constant -1 : i32
    %c-1_i64 = arith.constant -1 : i64
    %c423311104 = arith.constant 423311104 : index
    %c1 = arith.constant 1 : index
    %c423310976 = arith.constant 423310976 : index
    %c106756736_i32 = arith.constant 106756736 : i32
    %c180157056_i32 = arith.constant 180157056 : i32
    %c169671296_i32 = arith.constant 169671296 : i32
    %c10490944_i32 = arith.constant 10490944 : i32
    %c0 = arith.constant 0 : index
    %0 = arith.index_cast %arg0 : i32 to index
    %device_0 = hal.devices.get %c0 : !hal.device
    %cmd = hal.command_buffer.create device(%device_0 : !hal.device) mode("OneShot|AllowInlineExecution") categories(Dispatch) : !hal.command_buffer
    %pipeline_layout = hal.pipeline_layout.lookup device(%device_0 : !hal.device) layout(<push_constants = 4, sets = [<0, bindings = [<0, storage_buffer, ReadOnly>, <1, storage_buffer>]>]>) : !hal.pipeline_layout
    hal.command_buffer.push_constants<%cmd : !hal.command_buffer> layout(%pipeline_layout : !hal.pipeline_layout) offset(0) values([%c10490944_i32, %c169671296_i32, %c180157056_i32, %c106756736_i32]) : i32, i32, i32, i32
    %run_forward$async_dispatch_73_rocm_hsaco_fb_run_forward$async_dispatch_73_attention_20x4096x64xf16_buffer = util.global.load @run_forward$async_dispatch_73_rocm_hsaco_fb_run_forward$async_dispatch_73_attention_20x4096x64xf16_buffer : !hal.buffer
    hal.command_buffer.push_descriptor_set<%cmd : !hal.command_buffer> layout(%pipeline_layout : !hal.pipeline_layout)[%c0] bindings([
      %c0 = (%run_forward$async_dispatch_73_rocm_hsaco_fb_run_forward$async_dispatch_73_attention_20x4096x64xf16_buffer : !hal.buffer)[%c0, %c423310976], 
      %c1 = (%run_forward$async_dispatch_73_rocm_hsaco_fb_run_forward$async_dispatch_73_attention_20x4096x64xf16_buffer : !hal.buffer)[%c423311104, %c423310976]
    ])
    %workgroup_x, %workgroup_y, %workgroup_z = hal.executable.calculate_workgroups device(%device_0 : !hal.device) target(@run_forward$async_dispatch_73::@rocm_hsaco_fb::@run_forward$async_dispatch_73_attention_20x4096x64xf16) : index, index, index
    %exe = hal.executable.lookup device(%device_0 : !hal.device) executable(@run_forward$async_dispatch_73) : !hal.executable
    %ordinal = hal.executable.export.ordinal target(@run_forward$async_dispatch_73::@rocm_hsaco_fb::@run_forward$async_dispatch_73_attention_20x4096x64xf16) : index
    scf.for %arg1 = %c0 to %0 step %c1 {
      hal.command_buffer.dispatch<%cmd : !hal.command_buffer> target(%exe : !hal.executable)[%ordinal] workgroups([%workgroup_x, %workgroup_y, %workgroup_z])
      hal.command_buffer.execution_barrier<%cmd : !hal.command_buffer> source("Dispatch|CommandRetire") target("CommandIssue|Dispatch") flags("None")
    }
    hal.command_buffer.finalize<%cmd : !hal.command_buffer>
    %1 = util.null : !hal.fence
    %fence = hal.fence.create device(%device_0 : !hal.device) flags("None") : !hal.fence
    hal.device.queue.execute<%device_0 : !hal.device> affinity(%c-1_i64) wait(%1) signal(%fence) commands([%cmd])
    %status = hal.fence.await until([%fence]) timeout_millis(%c-1_i32) : i32
    util.status.check_ok %status, "failed to wait on timepoint"
    util.return
  }
}

Full model in torch dialect IR: stable_diffusion_xl_base_1_0_PNDM_64_1024x1024_fp16_unet_30.mlir.txt

Error output:

iree.compiler.tools.binaries.CompilerToolError: Error invoking IREE compiler tool iree-compile.exe
Error code: 1
Diagnostics:
failed to translate executables
./unet_gfx1100_a_dps\configured_compiled_scheduled_unet_run_forward$async_dispatch_73.mlir:9:6: error: Failures have been detected while processing an MLIR pass pipeline
      func.func @run_forward$async_dispatch_73_attention_20x4096x64xf16() attributes {translation_info = #iree_codegen.translation_info<LLVMGPUDistribute workgroup_size = [64, 1, 1] subgroup_size = 32>} {
     ^
./unet_gfx1100_a_dps\configured_compiled_scheduled_unet_run_forward$async_dispatch_73.mlir:9:6: note: Pipeline failed while executing [`TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_15, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_21, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_51, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_58, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_59, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_60, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_64, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `LLVMGPULowerExecutableTarget` on 'func.func' operation: @run_forward$async_dispatch_64_matmul_transpose_b_8192x640x640_f16xf16xf32, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_65, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `LowerExecutableUsingTransformDialect` on 'builtin.module' operation, `TransformDialectInterpreter` on 'builtin.module' operation, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_66, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_67, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `LLVMGPULowerExecutableTarget` on 'func.func' operation: @run_forward$async_dispatch_67_matmul_transpose_b_8192x640x640_f16xf16xf32, `LLVMGPUVectorLowering` on 'func.func' operation: @run_forward$async_dispatch_51_conv_2d_nhwc_hwcf_2x64x64x640x3x3x320_f16xf16xf32, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_70, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `TranslateExecutablesPass` on 'hal.executable' operation: @run_forward$async_dispatch_73, `TranslateTargetExecutableVariantsPass` on 'hal.executable.variant' operation: @rocm_hsaco_fb, `LLVMGPULowerExecutableTarget` on 'func.func' operation: @run_forward$async_dispatch_70_generic_2x10x4096x64_f32xf16, `FoldMemRefAliasOps` on 'func.func' operation: @run_forward$async_dispatch_64_matmul_transpose_b_8192x640x640_f16xf16xf32, `LLVMGPUVectorDistribute` on 'func.func' operation: @run_forward$async_dispatch_67_matmul_transpose_b_8192x640x640_f16xf16xf32, `ConvertToROCDL` on 'builtin.module' operation, `ConvertToROCDL` on 'builtin.module' operation, `ExpandStridedMetadata` on 'func.func' operation: @run_forward$async_dispatch_21_conv_2d_nhwc_hwcf_2x128x128x320x3x3x320_f16xf16xf32, `LinalgExtToLoops` on 'func.func' operation: @run_forward$async_dispatch_73_attention_20x4096x64xf16, `LinalgExtToLoops` on 'func.func' operation: @run_forward$async_dispatch_59_conv_2d_nhwc_hwcf_2x64x64x640x1x1x320_f16xf16xf32, `Canonicalizer` on 'func.func' operation: @run_forward$async_dispatch_58_conv_2d_nhwc_hwcf_2x64x64x640x3x3x640_f16xf16xf32, `Canonicalizer` on 'func.func' operation: @run_forward$async_dispatch_70_generic_2x10x4096x64_f32xf16, `ConvertAffineToStandard` on 'func.func' operation: @run_forward$async_dispatch_66_generic_8192x640_f16xf32xf32xf16xf16xf16]: reproducer generated at `./shark_tmp/core-reproducer.mlir`
./unet_gfx1100_a_dps\configured_compiled_scheduled_unet_run_forward$async_dispatch_73.mlir:2:2: error: failed to run translation of source executable to target executable for backend #hal.executable.target<"rocm", "rocm-hsaco-fb", {mma_intrinsics = [#iree_gpu.mma_layout<WMMA_F16_16x16x16_F32>], target_arch = "gfx1100", ukernels = "none"}>
  hal.executable.variant public @rocm_hsaco_fb target(<"rocm", "rocm-hsaco-fb", {mma_intrinsics = [#iree_gpu.mma_layout<WMMA_F16_16x16x16_F32>], target_arch = "gfx1100", ukernels = "none"}>) {
 ^

Invoked with:
 iree-compile.exe C:\V\iree-build\compiler\bindings\python\iree\compiler\tools\..\_mlir_libs\iree-compile.exe stable_diffusion_xl_base_1_0_PNDM_64_1024x1024_fp16_unet_30.mlir --iree-input-type=torch --iree-vm-bytecode-module-output-format=flatbuffer-binary --iree-hal-target-backends=rocm --mlir-print-debuginfo --mlir-print-op-on-diagnostic=false --mlir-pass-pipeline-crash-reproducer=./shark_tmp/core-reproducer.mlir --iree-hal-target-backends=rocm --iree-rocm-target-chip=gfx1100 --iree-opt-const-eval=false --iree-opt-data-tiling=False --iree-hal-dump-executable-files-to=./unet_gfx1100_a_dps --mlir-print-debuginfo=false --iree-codegen-gpu-native-math-precision=true --iree-codegen-llvmgpu-use-vector-distribution --iree-global-opt-propagate-transposes=true --iree-opt-outer-dim-concat=true --iree-vm-target-truncate-unsupported-floats --iree-llvmgpu-enable-prefetch=true --verify=false --iree-opt-data-tiling=false --iree-preprocessing-pass-pipeline=builtin.module(iree-preprocessing-transpose-convolution-pipeline, iree-preprocessing-pad-to-intrinsics)

Need more information? Set IREE_SAVE_TEMPS=/some/dir in your environment to save all artifacts and reproducers.