halide 17.0.0

[BrewTestBot]

Created by brew bump

---

Created with brew bump-formula-pr.

Details

release notes

## Changes Of Note

ParamMap has been removed entirely from the public API. All users of ParamMap should migrate to Callable instead.
Halide::Parameter has been moved to the public Halide API (it was formerly "internal" and not intended for public use).
New scheduling primitives:

Func::partition() and friends: Set the loop partition policy, which controls how/whether a loop is split into three loops (prologue/steady-state/epilogue). Loop partitioning can be useful to optimize boundary conditions (e.g. clamp_edge).
Func::hoist_storage() and friends: allows a functions's storage to be moved to a given loop level. Unlike Func::store_at(), no optimizations are triggered (e.g. sliding window).


New TailStrategy options for for existing scheduling directives:

ShiftInwardsAndBlend: Equivalent to ShiftInwards, but protects values that would be re-evaluated by loading the memory location that would be stored to, modifying only the elements not contained within the overlap, and then storing the blended result. Unlike ShiftInwards, this is valid to use in update definitions.
RoundUpAndBlend: Equivalent to RoundUp, but protects values that would be written beyond the end by loading the memory location that would be stored to, modifying only the elements within the region being computed, and then storing the blended result. Unlike RoundUp, this is valid to use on non-outermost splits in update definitions.


Substantially improved performance and display in the VizIR output.
Profiler improvements:

Substantially nicer text output
Injects timing into calls for copy_to_host and copy_to_device so you can measure host<->device copy overhead
Allows option sorting via HL_PROFILER_SORT env var


Substantially faster codegen for several GPU backends.
Experimental serialization/deserialization feature allows for saving of Halide IR code.
Various bug fixes and improvements in the Anderson2021 autoscheduler.
Improved ARM codegen, including: better patterns for sdot/udot; improved shift/mul codegen.
Support for Zen4 architecture in the x86 backend.
Updates to the ONNX app.
Various fixes and improvements to sliding-window and storage-folding.
Improvements to slow gather operations for some x86 variants.
Improvements to correctness for the .async() scheduling directive.
Improved codegen for float16 conversion, especially on x86.
Several compile-time warnings of dubious usefulness disabled.
WebAssembly codegen now defaults to assuming that saturating-float-to-int and sign-extension instructions sets are always available.
Target now does some reality-checking that it doesn't contain obviously nonsensical Feature combinations

What's Changed

Misc changes and fixes to RISCV codegen
Revise LLVM fix to work when no V8 or WABT available by @steven-johnson in Revise LLVM fix to work when no V8 or WABT available halide/Halide#7635
Be more careful about overflow in trim_bounds_using_alignment by @abadams in Be more careful about overflow in trim_bounds_using_alignment halide/Halide#7645
Add a compositing example app by @abadams in Add a compositing example app halide/Halide#7646
Get the ASAN toolchain working again by @steven-johnson in Get the ASAN toolchain working again halide/Halide#7604
Upgrade clang-format and clang-tidy to use v16 by @steven-johnson in Upgrade clang-format and clang-tidy to use v16 halide/Halide#7660
Enable the misc-use-anonymous-namespace clang-tidy check by @steven-johnson in Enable the misc-use-anonymous-namespace clang-tidy check halide/Halide#7661
Enable clang-tidy's modernize-use-default-member-init check by @steven-johnson in Enable clang-tidy's modernize-use-default-member-init check halide/Halide#7662
Update onnx app to Adams2019 autoscheduler and new autoscheduler API by @abadams in Update onnx app to Adams2019 autoscheduler and new autoscheduler API halide/Halide#7673
Remove ParamMap by @steven-johnson in Remove ParamMap halide/Halide#7675
Fix correctness_float16_t for ASAN builds by @steven-johnson in Fix correctness_float16_t for ASAN builds halide/Halide#7687
Add a select overload for tuples by @abadams in Add a select overload for tuples halide/Halide#7672
Add Sanitizer details to README_cmake.md by @steven-johnson in Add Sanitizer details to README_cmake.md halide/Halide#7688
Fix quadratic algorithm in simplify_correlated_differences by @abadams in Fix quadratic algorithm in simplify_correlated_differences halide/Halide#7686
Fix float16 under asan, attempt #2 by @steven-johnson in Fix float16 under asan, attempt #2 halide/Halide#7691
Add a warning if a Generator declares any Outputs before the final Input (Fixes #7669) by @steven-johnson in Add a warning if a Generator declares any Outputs before the final Input (Fixes #7669) halide/Halide#7697
Fixed the regularization for BGU. by @mcourteaux in Fixed the regularization for BGU. halide/Halide#7684
Fix clang and llvm versions in scripts by @TH3CHARLie in Fix clang and llvm versions in scripts halide/Halide#7702
Fix leaks caused by self-referential parameter constraints by @abadams in Fix leaks caused by self-referential parameter constraints halide/Halide#7700
Fix float16 warning for older clangs by @abadams in Fix float16 warning for older clangs halide/Halide#7701
Upgrade Halide main branch for LLVM18 by @steven-johnson in Upgrade Halide main branch for LLVM18 halide/Halide#7710
Improved profiler result printing. by @mcourteaux in Improved profiler result printing. halide/Halide#7709
Default WITH_TEST_FUZZ to OFF by @steven-johnson in Default WITH_TEST_FUZZ to OFF halide/Halide#7695
Throw an erorr if split is called with the same older and inner var name by @TH3CHARLie in Throw an erorr if split is called with the same older and inner var name halide/Halide#7715
Making HLSL code-gen a couple orders of magnitude faster... by @slomp in Making HLSL code-gen a couple orders of magnitude faster... halide/Halide#7719
Making Metal code-gen a bit faster by @slomp in Making Metal code-gen a bit faster halide/Halide#7720
Fix handling of thread features for scalars in Anderson2021 by @aekul in Fix handling of thread features for scalars in Anderson2021 halide/Halide#7726
Change default generator timeout to infinite by @abadams in Change default generator timeout to infinite halide/Halide#7718
Remove unused using decl by @abadams in Remove unused using decl halide/Halide#7730
[Hexagon] - Fix problems in sim_host.cpp by @pranavb-ca in [Hexagon] - Fix problems in sim_host.cpp halide/Halide#7725
Fix RDom usage in anderson2021_test_apps_autoscheduler (Fixes #7729) by @steven-johnson in Fix RDom usage in anderson2021_test_apps_autoscheduler (Fixes #7729) halide/Halide#7734
Fix leak on cloning functions with update defs by @abadams in Fix leak on cloning functions with update defs halide/Halide#7735
Ignore code in src/runtime/hexagon_remote/bin/src for clang-format by @steven-johnson in Ignore code in src/runtime/hexagon_remote/bin/src for clang-format halide/Halide#7736
Clean up really long line lengths in Anderson2021 by @steven-johnson in Clean up really long line lengths in Anderson2021 halide/Halide#7728
Revise labels on autoscheduler tests by @steven-johnson in Revise labels on autoscheduler tests halide/Halide#7732
Speedup the VizIR HTML. by @mcourteaux in Speedup the VizIR HTML. halide/Halide#7713
Run clang-tidy on macOS runners instead of Linux by @steven-johnson in Run clang-tidy on macOS runners instead of Linux halide/Halide#7746
Fix infinite recursion in loop partitioning by @abadams in Fix infinite recursion in loop partitioning halide/Halide#7743
Fix leaks in test/correctness/memoize.cpp by @abadams in Fix leaks in test/correctness/memoize.cpp halide/Halide#7705
Allow optional sorting of profiler output via HL_PROFILER_SORT env var (Fixes #7638) by @steven-johnson in Allow optional sorting of profiler output via HL_PROFILER_SORT env var (Fixes #7638) halide/Halide#7639
Permit llvm 15 on windows by @abadams in Permit llvm 15 on windows halide/Halide#7744
Revert accidental typo change in #7746 by @steven-johnson in Revert accidental typo change in #7746 halide/Halide#7747
[vulkan] Fix heap buffer overflow in Vulkan extension handling discovered by ASAN by @derek-gerstmann in [vulkan] Fix heap buffer overflow in Vulkan extension handling discovered by ASAN halide/Halide#7740
[vulkan] Fix SPIR-V IR references causing leaks by @derek-gerstmann in [vulkan] Fix SPIR-V IR references causing leaks halide/Halide#7739
Improve error-handling in Anderson2021, and ensure build deps are cor… by @steven-johnson in Improve error-handling in Anderson2021, and ensure build deps are cor… halide/Halide#7748
StmtViz: Search for tooltip only in the child node by @antonysigma in StmtViz: Search for tooltip only in the child node halide/Halide#7754
Experimental serializer by @TH3CHARLie in Experimental serializer halide/Halide#7594
Define cast<i32>(u32) overflow behavior by @rootjalex in Define cast<i32>(u32) overflow behavior halide/Halide#7769
Fix vector reduce HTML by @mcourteaux in Fix vector reduce HTML halide/Halide#7773
Remove fragile simd_op_check test for mlal/mlsl on ARM by @rootjalex in Remove fragile simd_op_check test for mlal/mlsl on ARM halide/Halide#7775
Speedup page loading of VizStmt. by @mcourteaux in Speedup page loading of VizStmt. halide/Halide#7755
Try to fix remaining ASAN-reported leaks by @steven-johnson in Try to fix remaining ASAN-reported leaks halide/Halide#7767
Fix out of bounds access in anderson2021_test_apps_autoscheduler by @aekul in Fix out of bounds access in anderson2021_test_apps_autoscheduler halide/Halide#7771
Don't introduce reinterprets in find/lower intrinsics by @rootjalex in Don't introduce reinterprets in find/lower intrinsics halide/Halide#7776
[Hexagon] -Build Hexagon runtime components using the Hexagon SDK (Clone of #7671) by @pranavb-ca in [Hexagon] -Build Hexagon runtime components using the Hexagon SDK (Clone of #7671) halide/Halide#7741
slice IRMatcher should only match on slices by @abadams in slice IRMatcher should only match on slices halide/Halide#7772
Don't inject undef() in the simplifier by @abadams in Don't inject undef() in the simplifier halide/Halide#7791
Fix for top-of-tree LLVM by @steven-johnson in Fix for top-of-tree LLVM halide/Halide#7798
[ARM] Distribute shifts as muls by @rootjalex in [ARM] Distribute shifts as muls halide/Halide#7790
[ARM] support new udot/sdot patterns by @rootjalex in [ARM] support new udot/sdot patterns halide/Halide#7800
Remove some unused includes by @abadams in Remove some unused includes halide/Halide#7799
Add support to the makefile for serialization by @abadams in Add support to the makefile for serialization halide/Halide#7762
[wasm] Enable PIC for WebAssembly on LLVM v18.x by @derek-gerstmann in [wasm] Enable PIC for WebAssembly on LLVM v18.x halide/Halide#7803
Update WebGPU to latest Emscripten/Dawn API by @steven-johnson in Update WebGPU to latest Emscripten/Dawn API halide/Halide#7804
Add jump-buttons to get fro Stmt directly to Assembly by @mcourteaux in Add jump-buttons to get fro Stmt directly to Assembly halide/Halide#7793
Update clang-tidy action to stop breaking by @steven-johnson in Update clang-tidy action to stop breaking halide/Halide#7808
[serialization] Add serialization support to generator interface  by @derek-gerstmann in [serialization] Add serialization support to generator interface  halide/Halide#7792
Ensure that multitarget AOT builds have consistent random sequence by @steven-johnson in Ensure that multitarget AOT builds have consistent random sequence halide/Halide#7717
Move clang-tidy checks back to Linux by @steven-johnson in Move clang-tidy checks back to Linux halide/Halide#7817
Update 'Check CMake file lists' action by @steven-johnson in Update 'Check CMake file lists' action halide/Halide#7809
Remove dead auto-schedule label in CMake by @steven-johnson in Remove dead auto-schedule label in CMake halide/Halide#7818
Don't return an undefined Stmt() from IfThenElse visitor by @abadams in Don't return an undefined Stmt() from IfThenElse visitor halide/Halide#7816
Avoid generating name collisions in CSE by @abadams in Avoid generating name collisions in CSE halide/Halide#7821
Add a check that PredicateLoads must be used in the outermost split of a dimension by @TH3CHARLie in Add a check that PredicateLoads must be used in the outermost split of a dimension halide/Halide#7788
Enable emission of float16/32 casts on x86 by @abadams in Enable emission of float16/32 casts on x86 halide/Halide#7837
Iterate over lets in the correct order in VectorizeLoops by @vksnk in Iterate over lets in the correct order in VectorizeLoops halide/Halide#7830
Zen4 support by @abadams in Zen4 support halide/Halide#7840
Update arguments in driver.cpp to match what correctness/simd_op_check has by @vksnk in Update arguments in driver.cpp to match what correctness/simd_op_check has halide/Halide#7842
[tutorials] Add tutorial on JIT compile/execute performance by @derek-gerstmann in [tutorials] Add tutorial on JIT compile/execute performance halide/Halide#7838
[api] Promote Internal::Parameter to Halide::Parameter  by @derek-gerstmann in [api] Promote Internal::Parameter to Halide::Parameter  halide/Halide#7829
[Hexagon] - Fix 8-bit unsigned saturating downcasts for HVX (Fixes #7806) by @pranavb-ca in [Hexagon] - Fix 8-bit unsigned saturating downcasts for HVX (Fixes #7806) halide/Halide#7825
Handle nested vectorization in store predicates by @abadams in Handle nested vectorization in store predicates halide/Halide#7864
Respect input buffer constraints in root-level bounds inference exprs by @abadams in Respect input buffer constraints in root-level bounds inference exprs halide/Halide#7865
Prevent use of uninitialized scalar Parameters in JIT code (#7847, partial) by @steven-johnson in Prevent use of uninitialized scalar Parameters in JIT code (#7847, partial) halide/Halide#7853
Handle unreachable code in bounds inference by @abadams in Handle unreachable code in bounds inference halide/Halide#7866
[serialization] Add support to serialize to memory, and a basic serialization tutorial by @derek-gerstmann in [serialization] Add support to serialize to memory, and a basic serialization tutorial halide/Halide#7760
Don't deduce unreachability from predicated out of bounds stores by @abadams in Don't deduce unreachability from predicated out of bounds stores halide/Halide#7874
Validate for types when fusing Vars with RVars by @abadams in Validate for types when fusing Vars with RVars halide/Halide#7877
Consider all dimensions before deciding to slide over a new dimension by @abadams in Consider all dimensions before deciding to slide over a new dimension halide/Halide#7875
Update onnx app to work with newer versions of protobuf by @abadams in Update onnx app to work with newer versions of protobuf halide/Halide#7879
HTML Stmt IR with conceptual code and device code. by @mcourteaux in HTML Stmt IR with conceptual code and device code. halide/Halide#7843
Update README.md to include RISCV in llvm build instructions by @abadams in Update README.md to include RISCV in llvm build instructions halide/Halide#7878
Implement elementwise complex value division by @antonysigma in Implement elementwise complex value division halide/Halide#7848
Explicitly name the allocgroups on GPU schedules "allocgroup__..." by @mcourteaux in Explicitly name the allocgroups on GPU schedules "allocgroup__..." halide/Halide#7883
Generate simpler LLVM IR for shuffles that recursively become broadcasts by @abadams in Generate simpler LLVM IR for shuffles that recursively become broadcasts halide/Halide#7902
Check for overflow in Type constructor by @abadams in Check for overflow in Type constructor halide/Halide#7889
Mutating if branches in isolation can break reachability analysis by @abadams in Mutating if branches in isolation can break reachability analysis halide/Halide#7895
Disable warning for mismatched new/delete by @abadams in Disable warning for mismatched new/delete halide/Halide#7897
Assignment is not associative by @abadams in Assignment is not associative halide/Halide#7894
Don't lift loop vars outside of their loops in sliding window by @abadams in Don't lift loop vars outside of their loops in sliding window halide/Halide#7896
Stop interleaver from expanding the scope of letstmts by @abadams in Stop interleaver from expanding the scope of letstmts halide/Halide#7908
Highlight groups for the HTML Stmt file and tooltips to reveal types. by @mcourteaux in Highlight groups for the HTML Stmt file and tooltips to reveal types. halide/Halide#7887
Static analysis (MSVC) fixes for device_buffer_utils.h by @slomp in Static analysis (MSVC) fixes for device_buffer_utils.h halide/Halide#7904
Check returned result in the test by @vksnk in Check returned result in the test halide/Halide#7911
Fix read-after-write hazard analysis in storage folding by @abadams in Fix read-after-write hazard analysis in storage folding halide/Halide#7910
Turn off SLP vectorization for avx512 only by @abadams in Turn off SLP vectorization for avx512 only halide/Halide#7918
Scheduling directive to hoist the storage of the function by @vksnk in Scheduling directive to hoist the storage of the function halide/Halide#7915
Improve the error message if you store_at without a compute_at by @vksnk in Improve the error message if you store_at without a compute_at halide/Halide#7923
Loop Partitioning Policy through Stage::partition(VarOrRVar, LoopPartitionPolicy) by @mcourteaux in Loop Partitioning Policy through Stage::partition(VarOrRVar, LoopPartitionPolicy) halide/Halide#7914
Remove use of dynamic_cast. by @zvookin in Remove use of dynamic_cast. halide/Halide#7931
Add special build for testing serialization via a serialization roundtrip in JIT compilation and fix serialization leaks by @TH3CHARLie in Add special build for testing serialization via a serialization roundtrip in JIT compilation and fix serialization leaks halide/Halide#7763
Add missing serialization of Dim::partition_policy by @TH3CHARLie in Add missing serialization of Dim::partition_policy halide/Halide#7935
Make sure all Halide arithmetic scalar types can be named from the Generator interface. by @zvookin in Make sure all Halide arithmetic scalar types can be named from the Generator interface. halide/Halide#7934
Remove the deprecated API llvm::Type::getInt8PtrTy usage. by @hokein in Remove the deprecated API llvm::Type::getInt8PtrTy usage. halide/Halide#7937
More targeted fix for gather instructions being slow on intel processors by @abadams in More targeted fix for gather instructions being slow on intel processors halide/Halide#7945
Track likely values through lets in loop partitioning by @abadams in Track likely values through lets in loop partitioning halide/Halide#7930
Add missing condition to if renesting rule by @abadams in Add missing condition to if renesting rule halide/Halide#7952
Always call lower_round_to_nearest_ties_to_even on arm32 by @vksnk in Always call lower_round_to_nearest_ties_to_even on arm32 halide/Halide#7957
Improve code size and compile time for local laplacian app by @abadams in Improve code size and compile time for local laplacian app halide/Halide#7927
[serialization] Serialize stub definitions of external parameters. by @derek-gerstmann in [serialization] Serialize stub definitions of external parameters. halide/Halide#7926
[WebGPU] Update to latest native headers by @jrprice in [WebGPU] Update to latest native headers halide/Halide#7932
Return values from stub functions in Deserialization by @steven-johnson in Return values from stub functions in Deserialization halide/Halide#7963
Make the fast inverse test throughput-limited rather than latency-limited by @abadams in Make the fast inverse test throughput-limited rather than latency-limited halide/Halide#7958
Attempt to fix nested vectorization gemm performance on new build bot by @abadams in Attempt to fix nested vectorization gemm performance on new build bot halide/Halide#7959
Update instructions to include generated schedules by @antonysigma in Update instructions to include generated schedules halide/Halide#7928
[serialization] Add Halide version and serialization version in serialization format by @TH3CHARLie in [serialization] Add Halide version and serialization version in serialization format halide/Halide#7905
Handle many more intrinsics in Bounds.cpp by @steven-johnson in Handle many more intrinsics in Bounds.cpp halide/Halide#7823
Disallow async nestings that violate read after write dependencies by @abadams in Disallow async nestings that violate read after write dependencies halide/Halide#7868
complete_x86_target() should enable F16C and FMA when AVX2 is present by @steven-johnson in complete_x86_target() should enable F16C and FMA when AVX2 is present halide/Halide#7971
Add two new tail strategies for update definitions by @abadams in Add two new tail strategies for update definitions halide/Halide#7949
Add appropriate mattrs for arm-32 extensions by @abadams in Add appropriate mattrs for arm-32 extensions halide/Halide#7978
Move canonical version numbers into source, not build system (#7980) by @steven-johnson in Move canonical version numbers into source, not build system (#7980) halide/Halide#7981
Silence useless "Insufficient parallelism" autoscheduler warning by @steven-johnson in Silence useless "Insufficient parallelism" autoscheduler warning halide/Halide#7990
Add a notebook with a visualization of the aprrox_* functions and their errors by @vksnk in Add a notebook with a visualization of the aprrox_* functions and their errors halide/Halide#7974
Make narrowing float->int casts on wasm go via wider ints by @abadams in Make narrowing float->int casts on wasm go via wider ints halide/Halide#7973
Fix handling of assert statements whose conditions get vectorized by @abadams in Fix handling of assert statements whose conditions get vectorized halide/Halide#7989
Fix all "unscheduled update()" warnings in our code by @steven-johnson in Fix all "unscheduled update()" warnings in our code halide/Halide#7991
Silence useless 'Outer dim vectorization of var' warning in Mullapudi… by @steven-johnson in Silence useless 'Outer dim vectorization of var' warning in Mullapudi… halide/Halide#7992
Make wasm +sign-ext and +nontrapping-fptoint the default by @steven-johnson in Make wasm +sign-ext and +nontrapping-fptoint the default halide/Halide#7995
Teach unrolling to exploit conditions in enclosing ifs by @abadams in Teach unrolling to exploit conditions in enclosing ifs halide/Halide#7969
Do some basic validation of Target Features (#7986) by @steven-johnson in Do some basic validation of Target Features (#7986) halide/Halide#7987
Inject profiling for function calls to 'halide_copy_to_host' and 'halide_copy_to_device'. by @mcourteaux in Inject profiling for function calls to 'halide_copy_to_host' and 'halide_copy_to_device'. halide/Halide#7913
bounds_of_nested_lanes assumed that one layer of nested vectorization could be removed at a time, but failed in situations with unusual nesting structures. by @abadams in Fix bounds_of_nested_lanes halide/Halide#8039 and 8055
we now track whether or not let expressions failed to solve in solver; failure to do this meant we did unhelpful transformations in some cases which let to exploding compile times. by @abadams in Track whether or not let expressions failed to solve in solver halide/Halide#7982

Full Changelog: halide/Halide@v16.0.0...v17.0.0

[github-actions]

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. To keep this pull request open, add a help wanted or in progress label.

[github-actions]

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. To keep this pull request open, add a help wanted or in progress label.

[chenrui333]

#164349