daisytuner / docc

64%

LAST BUILD ON BRANCH main

branch: main

CHANGE BRANCH

x

Reset

• main
• BatchMatmul
• CFGPruner
• CUDACorrelationFix
• CUDATiling
• Cublas4Python
• DebugInfoRework
• ExtendedNPBench
• FixArgCapturing
• FixLiftreport
• HIPtarget
• InferenceAnalysis
• InstrumentationInfo
• KLSSubsetUndefined
• LibNodeOffloading
• LiftReport
• LiftingReport
• LinalgBroadcast
• MACOSASANFix
• MigrateOffloading
• Modelzoo
• MoveLoopReport
• MultiDimGPUKernels
• OLSTiling
• OffloadOpaque
• OptionalOffloadReport
• PythonWorkflow
• Roc
• ScheduleTypeUpdate
• SchedulerStreamlining
• SkipNestedTiling
• SupportPoolingLayers
• TensorReshape
• TimedDotGraph
• accelerator-dtypes
• activate-hpccg-cuda
• add-filter-in-scheduler
• add-normalization
• add-opt-report
• add-opt-report-to-transfertuning
• add-python-transfertuning
• add-transfertune-daisy-test
• add-transfertuning-tests
• addTTpolybench
• advanced-loop-tiling
• allocation-hoisting
• arg-capture-ci
• arg-capture-fixes
• arg-capture-scope
• arg-captures
• arg-expand
• arg-expansion
• assums-perf
• astype
• attributes
• avail-external-linkage
• backend-docc
• bindings-builder
• bindings-nodes
• bit-widths
• blacklist-vararg
• blas-lifting
• block-sorting-and-hoisting
• bplustree
• bytes
• byval-fix
• cast-attributes
• cfd-local-array
• change-default-docc-folder
• change-features
• change-model-location
• ci-levels
• ci-rollback
• cloudsc
• cloudsc2
• cmath-nodes-refactor
• code-style-cleanup
• compile-time-opt
• compound-loads
• compound-stores
• constant-prop
• copilot/add-onnx-torch-mlir-bindings
• copilot/extend-mlir-frontend-with-sdfg-dialect
• copilot/extend-python-frontend-casts
• copilot/extend-python-frontend-classes
• copilot/extend-python-frontend-ones-support
• cpack-sdfglib
• cross-library-inlining
• cuda-arch-opt
• cuda-partition-incubation-test
• cuda-tests
• cuda-transforms
• cuda_autotuning
• daisy-workflows
• dangling-nodes-ref-prop
• data-dependency-bugfixes
• dbg-functions
• dde-analysis-invalidation-on-modify
• dde-perf
• deactivate-mlir-tests-cuda
• dead-data-elimination
• deallocation-management
• debug-info
• debug-prints
• depthwise2DConv
• deterministic-topo-sort
• dimension-wise-intersect
• disjoint-namespaces
• distro-agnostic
• docc-ai-torch-mlir
• docc-analysis
• docc-noop
• docc-path-discovery
• docc-py
• docc-tmp
• dump-conditional-sdfgs
• dump-passes
• einsum
• einsum-tt
• einsum2gemm
• enable-transfertuning
• energy-measurements
• etsoc
• etsoc-ci
• etsoc-parallel
• ext-prefix
• extend-arith2sdfg
• extend-bindings
• extend-transfertuning
• faster-block-sorting
• filter-by-loopinfo
• fix-arg-capturing
• fix-argcapture-on-cuda
• fix-c++20-import
• fix-cfd-cuda
• fix-dde
• fix-linkonce
• fix-mlir-tests-workflow
• fix-return-test
• fix-sdfg-replacement
• fixed-deb-deps
• fixes
• fixing-dot-diz-crash
• fixrodiniacuda
• float-hpccg
• flop-analysis
• flop-analysis-extension
• fmod-fix
• for-each
• for2map-improvs
• frontend-hpccg
• func-blacklist-rettype
• function-lifting
• function-whitelisting
• generic-2nd-link-rebase
• globals-linkage
• graceful-validate
• hand-cuda-blocksize-into-transform
• hpccg-arm
• hpccg-diag
• hppcg-csr
• icmp-cast-fix
• icmp-fix
• ignore-not-implemented
• improving-global-cfg
• improving-tt-transform
• inline-daisy-functions
• instrument-python-code
• instrumentation
• instrumentation-format
• int-abs
• integrate-python-docc-frontends
• introduce-nightlies
• isl-printer-pow
• iterator-to-indvar
• library-nodes-refactor
• linkage
• linker-test
• llvm-infra-fixes
• llvm-tests
• long-names-bug
• loop-info-recursive
• loop-info-update
• loop-normalization-hotfix
• loop-normalizations
• loop-rotate
• loop-shift
• loop-skewing-fors
• lulesh
• lulesh-cuda
• macos
• main-revert
• mantevo
• map-fusion
• map-fusion-multi-dims
• map-optimizations
• mean-std-nodes
• measurements
• min-data-transfers-hpccg
• miniAMR-maps
• minor-bug-fixes
• minor-bugs
• missing-ubuntu-deb
• mlir-benchmarks
• mlir-cf-assert
• mlir-export-sdfg
• mlir-frontend
• mlir-integration
• mlir-malloc
• mlir-math
• mlir-opt-linear
• mlir-target-plugins
• mlir-target-tests
• mlir-tensor-ops-and-index
• more-hoisting
• more-tests
• move-token-to-secret
• moves-python-frontend
• native-ndarray
• nightlies
• nightly-fix
• node-bindings-transfer-tuning
• norm-exp
• np-clip
• npbench
• npbench-crc16
• npbench-ludcmp
• npbench-mlp
• offloading-nodes
• omp-plugin
• onnx-target
• opaque-references
• openfoam
• openfoam-2
• papi-fix
• perf-improvs
• performance-improvements
• pipeline-update
• plugin-sdfg-lookup
• pointer-symbolic-cmp
• polybench-precision
• polybench-tt-test
• pow-sym
• prepping-for-sdfgrpc
• ptrtoint
• python-ast-refactor
• python-debug-info
• python-optional-cuda
• python-transpose-conv
• python-wheels
• quick-fix-dead-data-elim
• range-based-trunc-elimination
• readd-error-handling
• readonly-hoisting
• recipes-cleanup
• reenable-feature-dump
• ref-measurement
• refactor-apps-test
• refactor-cutout
• refactor-transfertuning
• refactor-workflows
• regions-renaming
• release-builds
• remove-cutout-sequence
• rename-mlir-tools
• restore-global-cfg
• returns
• revert-228-DebugInfoRework
• rework-mlir-frontend
• robustness
• rodinia-debugging
• rodinia-verifier-shl
• rodinia-workflows
• rtl
• rtl-fixes
• rtl-reuse
• rtl-tests-exit-codes
• run-polybench-on-tansy
• sanitizers
• schedule-dumping
• sext-i1-fix
• siblings-failure
• stable-sdfg-registry
• static-boost
• statistics
• stdlib-nodes
• storage-types
• subcompile-opts
• submodules-update
• symbol-alias
• symbol-evolution-fix
• target-lib-node-mapping
• targetTutorial
• tasklet-fusion
• tasklet-tensor-node
• tensor-type
• tensor-type-2
• tenstorrent/prep-gemm-test
• test-measurement
• test-type
• tile-fusion
• topo-det
• torch-matmul-net-test
• towards-cuda-transfertuning
• transfer-inlining
• transfer-minimization
• transferMinimization
• transformations-api
• transitions-bug
• trap-nodes
• trisolv-fix
• trunc-analysis
• tt-blas-dot
• tt-dot-instrument
• tt-loop-classes
• tt-matvec
• tt-python-tests
• unify-loop-exits
• unsigned-constant-promotion
• upload-to-backend
• use-stable-pytorch
• version-string
• workflows
• zero-loops-tt-pass
• zext-analysis

Build # 23201411123

Build Type

push

github

Committed by web-flow

Commit Message

Mlir target plugins (#593)

* Made dumping of SDFGs for python-based frontends optional (DOCC_DEBUG and DOCC_CI).

 ~ ensure correct json dump is designated for arg_capture & instrumentation
 + enable mlir front end to use python target-overrides and make the same debug outputs as regular python front end

* Dump MLIR code and translated SDFG string on parsing-failure

* Removing dead code that assumed presence of an optional metadata entry pointing to an SDFG Json dump that does not exist.

Run Details

12 of 17 new or added lines in 1 file covered. (70.59%)

1 existing line in 1 file now uncovered.

26052 of 40718 relevant lines covered (63.98%)

403.73 hits per line