Pinned issues
Issues
- 2
oss-fuzz is of dubious value for our use-cases
#8171 opened by abadams - 4
Signed integer overflow occurred during constant-folding. Signed integer overflow for int32 and int64 is undefined behavior in Halide.
#8227 opened by jansel - 0
blank
#8233 opened by mehmetozsoy1 - 0
Does Halide support opencl on Android (Qualcomm)?
#8228 opened by wgy123-cell - 2
Python bindings do not support int64 constants
#8224 opened by jansel - 1
- 4
Why does having inconsistent input and output dimensions in a Halide generator significantly slow down performance?
#8218 opened by mhua97 - 0
- 2
- 1
[HVX] Blur example fails with: Failed to write shared object to 'libhalide_shared_runtime.so'
#8201 opened by FabianSchuetze - 1
- 6
[D3D12] HLSL 6?
#8164 opened by soufianekhiat - 0
- 4
We should have helper functions that return bool Exprs that query properties of the target
#8192 opened by abadams - 2
Improvements for Halide floating-point math functions.
#8199 opened by zvookin - 1
- 2
Different results on different GPUs (Stress test)
#8191 opened by soufianekhiat - 3
- 0
Implement nested loop with indexing coordinates different from looping coordinates
#8190 opened by mcoduoza - 7
Error message "Unsupported HVX type: float32x32"
#8170 opened by jxl1080 - 2
- 2
Some find intrinsics rewrite rules that convert widen_right_add to widening_add are wrong
#8172 opened by abadams - 1
Hoist storage undefined symbol
#8141 opened by abadams - 1
Scheduling the sum() of a large image is too hard
#8100 opened by steven-johnson - 3
Fuzzing Failures
#8165 opened by steven-johnson - 0
Support aliased input and output buffers
#8161 opened by abadams - 1
Fuzz-test failure
#8159 opened by steven-johnson - 3
- 2
Getting InternalError with a combination of compute_with, and reductions that use functions with update definitions
#8149 opened by sakehl - 1
[upstream] `generator_aot_cleanup_on_error`/`generator_aotcpp_cleanup_on_error` don't work with RustiCL
#8157 opened by LebedevRI - 5
Sliding window by backing up the loop min is busted
#8140 opened by abadams - 0
Three tutorials missing on the website.
#8146 opened by mcourteaux - 1
- 1
Support for newer Snapdragon DSP (v73) architectures
#8122 opened by adarsh044 - 1
- 5
names shouldn't be uniqued by string concatenation
#8131 opened by abadams - 9
Vulkan backend bug in region_allocator
#8079 opened by immortalsalomon - 4
Vulkan backend bug in casting boolean
#8064 opened by immortalsalomon - 0
- 0
- 0
Tracking bug for SVE2 scalable vector todos.
#8119 opened by zvookin - 0
- 0
Better discipline for interaction between Halide and LLVM types in CodeGen_LLVM.
#8117 opened by zvookin - 6
riscv64 build is broken in `17.0.0`
#8114 opened by LebedevRI - 1
Code completion with Python bindings?
#8112 opened by ziriax - 1
- 2
- 3
- 1
Investigate using vector predication intrinsics for SVE2 architecture specific cases in CodeGen_ARM.
#8088 opened by zvookin - 0
test/correctness/float16_t_neon_op_check.cpp is busted
#8083 opened by abadams