openxla/iree issues and pull requests

#15636 - Fix intermittent failure - functions with `_try_` in their name may fail spuriously.

Pull Request - State: closed - Opened by bjacob about 1 year ago - 1 comment

#15636 - Fix intermittent failure - functions with `_try_` in their name may fail spuriously.

Pull Request - State: closed - Opened by bjacob about 1 year ago - 1 comment

#15635 - Convert MetalSPIRV compiler target into a plugin.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: platform/macos 🍎, cleanup 🧹

#15635 - Convert MetalSPIRV compiler target into a plugin.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: platform/macos 🍎, cleanup 🧹

#15634 - [Flow] Update fusion heuristics to enable non adhoc dequant-matvec f…

Pull Request - State: open - Opened by Max191 about 1 year ago - 2 comments

#15634 - [Flow] Update fusion heuristics to enable non adhoc dequant-matvec f…

Pull Request - State: open - Opened by Max191 about 1 year ago - 2 comments

#15633 - [LLVMGPU] Enable dynamic 1-D subgroup reduce

Pull Request - State: closed - Opened by qedawkins about 1 year ago

#15633 - [LLVMGPU] Enable dynamic 1-D subgroup reduce

Pull Request - State: closed - Opened by qedawkins about 1 year ago

#15632 - Rename `emitc.call` to `emitc.call_opaque`

Pull Request - State: closed - Opened by marbre about 1 year ago - 1 comment

#15632 - Rename `emitc.call` to `emitc.call_opaque`

Pull Request - State: open - Opened by marbre about 1 year ago

#15632 - Rename `emitc.call` to `emitc.call_opaque`

Pull Request - State: open - Opened by marbre about 1 year ago

#15631 - Compiler error from migrating the StableHLO input conversion to a compiler plugin.

Issue - State: open - Opened by LLITCHEV about 1 year ago - 4 comments
Labels: bug 🐞, support

#15631 - Compiler error from migrating the StableHLO input conversion to a compiler plugin.

Issue - State: open - Opened by LLITCHEV about 1 year ago - 4 comments
Labels: bug 🐞, support

#15630 - (CI) Fix SHARK benchmark workflow.

Pull Request - State: open - Opened by monorimet about 1 year ago - 3 comments

#15630 - (CI) Fix SHARK benchmark workflow.

Pull Request - State: open - Opened by monorimet about 1 year ago - 3 comments

#15629 - [EPIC][CPU] Enable predictable performance on mixed-types GEMM using data-tiling

Issue - State: open - Opened by hanhanW about 1 year ago - 2 comments
Labels: codegen, codegen/llvm, epic

#15629 - [EPIC][CPU] Enable predictable performance on mixed-types GEMM using data-tiling

Issue - State: open - Opened by hanhanW about 1 year ago - 2 comments
Labels: codegen, codegen/llvm, epic

#15629 - [EPIC][CPU] Enable predictable performance on mixed-types GEMM using data-tiling

Issue - State: open - Opened by hanhanW about 1 year ago - 2 comments
Labels: codegen, codegen/llvm, epic

#15628 - Enable check_initialization_order on ASan CI.

Pull Request - State: open - Opened by ScottTodd about 1 year ago - 1 comment
Labels: infrastructure

#15628 - Enable check_initialization_order on ASan CI.

Pull Request - State: open - Opened by ScottTodd about 1 year ago - 1 comment
Labels: infrastructure

#15628 - Enable check_initialization_order on ASan CI.

Pull Request - State: open - Opened by ScottTodd about 1 year ago - 1 comment
Labels: infrastructure

#15627 - Fix experimental/webgpu build by setting binary dir.

Pull Request - State: open - Opened by ScottTodd about 1 year ago
Labels: hal/webgpu

#15627 - Fix experimental/webgpu build by setting binary dir.

Pull Request - State: open - Opened by ScottTodd about 1 year ago
Labels: hal/webgpu

#15626 - Only include conv tests in winograd test suites.

Pull Request - State: closed - Opened by hanhanW about 1 year ago - 1 comment

#15626 - Only include conv tests in winograd test suites.

Pull Request - State: closed - Opened by hanhanW about 1 year ago - 1 comment

#15625 - Set `iree-llvmcpu-target-cpu` to `cortex-x1` for benchmarking Pixel 6

Pull Request - State: open - Opened by pzread about 1 year ago - 6 comments
Labels: benchmarks:android-cpu

#15625 - Set `iree-llvmcpu-target-cpu` to `cortex-x1` for benchmarking Pixel 6

Pull Request - State: open - Opened by pzread about 1 year ago - 6 comments
Labels: benchmarks:android-cpu

#15625 - Set `iree-llvmcpu-target-cpu` to `cortex-x1` for benchmarking Pixel 6

Pull Request - State: open - Opened by pzread about 1 year ago - 6 comments
Labels: benchmarks:android-cpu

#15624 - Cleanup ci.yml docs that referenced old code.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: infrastructure, cleanup 🧹

#15624 - Cleanup ci.yml docs that referenced old code.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: infrastructure, cleanup 🧹

#15624 - Cleanup ci.yml docs that referenced old code.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: infrastructure, cleanup 🧹

#15624 - Cleanup ci.yml docs that referenced old code.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: infrastructure, cleanup 🧹

#15623 - The run_shark_tank.yml workflow is broken

Issue - State: open - Opened by ScottTodd about 1 year ago
Labels: bug 🐞

#15623 - The run_shark_tank.yml workflow is broken

Issue - State: closed - Opened by ScottTodd about 1 year ago
Labels: bug 🐞

#15623 - The run_shark_tank.yml workflow is broken

Issue - State: open - Opened by ScottTodd about 1 year ago
Labels: bug 🐞

#15623 - The run_shark_tank.yml workflow is broken

Issue - State: open - Opened by ScottTodd about 1 year ago
Labels: bug 🐞

#15622 - Add support for checking accuracy of BF16 and INT values

Issue - State: open - Opened by mariecwhite about 1 year ago - 2 comments
Labels: bug 🐞

#15621 - [CPU][DT] Improve vector lowering for mmt4d

Pull Request - State: open - Opened by dcaballe about 1 year ago - 6 comments
Labels: benchmarks:android-cpu

#15621 - [CPU][DT] Improve vector lowering for mmt4d

Pull Request - State: open - Opened by dcaballe about 1 year ago - 6 comments
Labels: benchmarks:android-cpu

#15621 - [CPU][DT] Improve vector lowering for mmt4d

Pull Request - State: open - Opened by dcaballe about 1 year ago - 6 comments
Labels: benchmarks:android-cpu

#15621 - [CPU][DT] Improve vector lowering for mmt4d

Pull Request - State: open - Opened by dcaballe about 1 year ago - 6 comments
Labels: benchmarks:android-cpu

#15620 - Fix bug that incorrectly determined narrow tile sizes for batch_matmul

Pull Request - State: closed - Opened by NatashaKnk about 1 year ago

#15620 - Fix bug that incorrectly determined narrow tile sizes for batch_matmul

Pull Request - State: closed - Opened by NatashaKnk about 1 year ago

#15620 - Fix bug that incorrectly determined narrow tile sizes for batch_matmul

Pull Request - State: closed - Opened by NatashaKnk about 1 year ago

#15619 - Re-enable `regression_test_amdgpu_rocm` build.

Pull Request - State: open - Opened by ScottTodd about 1 year ago
Labels: infrastructure, hal/rocm

#15619 - Re-enable `regression_test_amdgpu_rocm` build.

Pull Request - State: open - Opened by ScottTodd about 1 year ago
Labels: infrastructure, hal/rocm

#15618 - Warn instead of assert on unknown jobs.

Pull Request - State: open - Opened by ScottTodd about 1 year ago
Labels: infrastructure

#15618 - Warn instead of assert on unknown jobs.

Pull Request - State: open - Opened by ScottTodd about 1 year ago
Labels: infrastructure

#15618 - Warn instead of assert on unknown jobs.

Pull Request - State: open - Opened by ScottTodd about 1 year ago
Labels: infrastructure

#15617 - Disable `regression_test_amdgpu_rocm` build for now.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: infrastructure, hal/rocm

#15616 - [Reducer] Generalize operand to result delta pass

Pull Request - State: closed - Opened by Groverkss about 1 year ago

#15615 - iree-import-tf fails with error: missing entry functions

Issue - State: open - Opened by monowaranjum about 1 year ago - 1 comment
Labels: bug 🐞, integrations/tensorflow, support

#15615 - iree-import-tf fails with error: missing entry functions

Issue - State: open - Opened by monowaranjum about 1 year ago - 1 comment
Labels: bug 🐞, integrations/tensorflow, support

#15614 - [spirv] Use `gpu.subgrup_reduce` in SPIR-V vector reduction pipeline

Pull Request - State: closed - Opened by kuhar about 1 year ago - 1 comment
Labels: codegen/spirv, benchmarks:android-gpu, benchmarks:vulkan-nvidia

#15613 - [EPIC][CPU] int4 sub-type emulation

Issue - State: open - Opened by hanhanW about 1 year ago
Labels: codegen/llvm, epic

#15612 - Convert WebGPU compiler target into a plugin.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago - 1 comment
Labels: cleanup 🧹, platform/web 🌐

#15612 - Convert WebGPU compiler target into a plugin.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago - 1 comment
Labels: cleanup 🧹, platform/web 🌐

#15611 - [CPU][DT] i8.i8.i32 mmt4d codegen kernel is 10x slower than ukernels

Issue - State: open - Opened by hanhanW about 1 year ago - 8 comments
Labels: codegen, codegen/llvm

#15610 - [stablehlo] Implement `cholesky` composition using `linalg` operations

Pull Request - State: open - Opened by rsuderman about 1 year ago
Labels: integrations/stablehlo

#15610 - [stablehlo] Implement `cholesky` composition using `linalg` operations

Pull Request - State: open - Opened by rsuderman about 1 year ago
Labels: integrations/stablehlo

#15609 - Adjust pkgci logging.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: quality of life 😊

#15608 - Assertion failure when compiling generic op with i4 narrow type

Issue - State: open - Opened by pzread about 1 year ago - 1 comment

#15608 - Assertion failure when compiling generic op with i4 narrow type

Issue - State: open - Opened by pzread about 1 year ago - 1 comment

#15608 - Assertion failure when compiling generic op with i4 narrow type

Issue - State: closed - Opened by pzread about 1 year ago - 2 comments

#15607 - Removing io_parameters.read/write in favor of gather/scatter.

Pull Request - State: closed - Opened by benvanik about 1 year ago

#15607 - Removing io_parameters.read/write in favor of gather/scatter.

Pull Request - State: closed - Opened by benvanik about 1 year ago

#15606 - Use setup.yml in pkgci, allowing for CI configuration/skipping.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago - 1 comment
Labels: infrastructure

#15606 - Use setup.yml in pkgci, allowing for CI configuration/skipping.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago - 1 comment
Labels: infrastructure

#15606 - Use setup.yml in pkgci, allowing for CI configuration/skipping.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago - 1 comment
Labels: infrastructure

#15606 - Use setup.yml in pkgci, allowing for CI configuration/skipping.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago - 1 comment
Labels: infrastructure

#15605 - Add custom edit URLs to generated website pages.

Pull Request - State: closed - Opened by ScottTodd about 1 year ago
Labels: documentation, compiler/dialects

#15604 - [EPIC][DT] Have better data-tiling coverage and program graph

Issue - State: open - Opened by hanhanW about 1 year ago
Labels: codegen, codegen/llvm, epic

#15603 - SDXL Unet ROCM runtime issues

Issue - State: open - Opened by Abhishek-Varma about 1 year ago - 1 comment
Labels: bug 🐞, hal/rocm

#15601 - [VMVX] Rework on vmvx lowering strategy.

Pull Request - State: closed - Opened by hanhanW about 1 year ago - 1 comment
Labels: benchmarks:x86_64, benchmarks:android-cpu

#15601 - [VMVX] Rework on vmvx lowering strategy.

Pull Request - State: closed - Opened by hanhanW about 1 year ago - 1 comment
Labels: benchmarks:x86_64, benchmarks:android-cpu

#15601 - [VMVX] Rework on vmvx lowering strategy.

Pull Request - State: closed - Opened by hanhanW about 1 year ago - 1 comment
Labels: benchmarks:x86_64, benchmarks:android-cpu

#15600 - Need support for tensor.pack direct vectorization

Issue - State: open - Opened by hanhanW about 1 year ago - 11 comments
Labels: codegen

#15600 - Need support for tensor.pack direct vectorization

Issue - State: open - Opened by hanhanW about 1 year ago - 11 comments
Labels: codegen

#15600 - Need support for tensor.pack direct vectorization

Issue - State: open - Opened by hanhanW about 1 year ago - 11 comments
Labels: codegen

#15599 - Add vector_ext dialect

Pull Request - State: closed - Opened by harsh-nod about 1 year ago - 13 comments

#15599 - Add vector_ext dialect

Pull Request - State: closed - Opened by harsh-nod about 1 year ago - 13 comments