Florian Hahn 77d3f8a925 [TBAA] Don't emit pointer-tbaa for void pointers. (#122116) ...

While there are no special rules in the standards regarding void
pointers and strict aliasing, emitting distinct tags for void pointers
break some common idioms and there is no good alternative to re-write
the code without strict-aliasing violations. An example is to count the
entries in an array of pointers:

    int count_elements(void * values) {
      void **seq = values;
      int count;
      for (count = 0; seq && seq[count]; count++);
      return count;
    }

https://clang.godbolt.org/z/8dTv51v8W

An example in the wild is from
https://github.com/llvm/llvm-project/issues/119099

This patch avoids emitting distinct tags for void pointers, to avoid
those idioms causing mis-compiles for now.

Fixes https://github.com/llvm/llvm-project/issues/119099.
Fixes https://github.com/llvm/llvm-project/issues/122537.

PR: https://github.com/llvm/llvm-project/pull/122116

2025-01-31 11:38:14 +00:00

..

2011-04-15-vec-init-from-vec.cl

…

addr-space-struct-arg.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

address-space-constant-initializers.cl

…

address-spaces-conversions.cl

…

address-spaces-mangling.cl

…

address-spaces.cl

…

amdgcn-automatic-variable.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

amdgcn-buffer-rsrc-type.cl

[IRBuilder] Generate nuw GEPs for struct member accesses (#99538)

2024-08-09 13:25:04 +01:00

amdgcn-flat-scratch-name.cl

[clang][SPIR-V] Add support for AMDGCN flavoured SPIRV (#89796)

2024-06-07 11:50:23 +01:00

amdgcn-large-globals.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

amdgcn-non-temporal-store.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

amdgpu-abi-struct-arg-byref.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

amdgpu-abi-struct-coerce.cl

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

amdgpu-alignment.cl

[Clang] Remove 3-element vector load and store special handling (#104661)

2025-01-21 09:18:16 -05:00

amdgpu-attrs.cl

[AMDGPU] Adding the amdgpu_num_work_groups function attribute (#79035)

2024-03-12 10:30:39 -07:00

amdgpu-call-kernel.cl

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

amdgpu-calling-conv.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

amdgpu-debug-info-pointer-address-space.cl

…

amdgpu-debug-info-variable-expression.cl

[RemoveDIs] Print IR with debug records by default (#91724)

2024-06-14 15:07:27 +01:00

amdgpu-enqueue-kernel.cl

[TBAA] Don't emit pointer-tbaa for void pointers. (#122116)

2025-01-31 11:38:14 +00:00

amdgpu-env-amdgcn.cl

…

amdgpu-features-illegal.cl

[AMDGPU] Report error in clang if wave32 is requested where unsupported (#97633)

2024-07-09 14:25:58 -07:00

amdgpu-features-readonly.cl

…

amdgpu-features.cl

AMDGPU: Builtin & CodeGen support for v_cvt_sr_{bf16|f16}_f32 instructions (#117824)

2024-11-26 23:37:05 -05:00

amdgpu-ieee.cl

…

amdgpu-nullptr.cl

[InstCombine] Infer nusw + nneg -> nuw for getelementptr (#111144)

2024-12-05 14:36:40 +01:00

amdgpu-printf.cl

[clang][CodeGen] Allow memcpy replace with trivial auto var init

2024-03-21 09:55:04 +01:00

amdgpu-sizeof-alignof.cl

…

arm-integer-dot-product.cl

…

as_type.cl

…

atomic-builtins-default-to-device-scope.cl

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

atomic-ops-libcall.cl

[clang][CodeGen] Emit atomic IR in place of optimized libcalls. (#73176)

2024-02-12 09:33:09 -08:00

atomic-ops.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

atomics-cas-remarks-gfx90a.cl

clang/AMDGPU: Update test message order

2024-09-06 21:18:41 +04:00

atomics-unsafe-hw-remarks-gfx90a.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

backend-unsupported-warning.ll

…

blocks.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

bool_cast.cl

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

builtins-alloca.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

builtins-amdgcn-ci.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-dl-insts-err-clamp.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-dl-insts-err.cl

AMDGPU: Add support for v_dot2c_f32_bf16 instruction for gfx950 (#117598)

2024-11-25 19:51:01 -08:00

builtins-amdgcn-dl-insts-gfx11.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-dl-insts-gfx12.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-dl-insts.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-flat-address-space.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-fp8.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-fp-atomics-gfx7-err.cl

…

builtins-amdgcn-fp-atomics-gfx11-err.cl

…

builtins-amdgcn-fp-atomics-gfx90a-err.cl

…

builtins-amdgcn-fp-atomics-gfx908-err.cl

…

builtins-amdgcn-gfx9-4-generic-err.cl

[AMDGPU] Introduce a new generic target gfx9-4-generic (#115190)

2024-11-12 23:11:05 -05:00

builtins-amdgcn-gfx9.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-gfx10.cl

[AMDGPU] Allow overload of __builtin_amdgcn_mov_dpp8 (#113610)

2024-10-31 02:19:20 -07:00

builtins-amdgcn-gfx11-err.cl

[LLVM][AMDGPU] Add Intrinsic and Builtin for ds_bpermute_fi_b32 (#124616)

2025-01-29 14:04:10 +01:00

builtins-amdgcn-gfx11-param-err.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-gfx11.cl

[AMDGPU] Add a new target for gfx1153 (#113138)

2024-10-23 12:56:58 +09:00

builtins-amdgcn-gfx12-err.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-gfx12-param-err.cl

[AMDGPU] modify named barrier builtins and intrinsics (#114550)

2024-11-06 10:37:22 -08:00

builtins-amdgcn-gfx12-wmma-w32.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-gfx12-wmma-w64.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-gfx12.cl

[LLVM][AMDGPU] Add Intrinsic and Builtin for ds_bpermute_fi_b32 (#124616)

2025-01-29 14:04:10 +01:00

builtins-amdgcn-gfx940.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

builtins-amdgcn-gfx950-err.cl

AMDGPU: Builtin & CodeGen support for v_cvt_sr_{bf16|f16}_f32 instructions (#117824)

2024-11-26 23:37:05 -05:00

builtins-amdgcn-gfx950-read-tr.cl

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

builtins-amdgcn-gfx950.cl

AMDGPU: Change bitop3 intrinsic operand to i32 (#118647)

2024-12-04 15:44:04 -05:00

builtins-amdgcn-global-load-tr-gfx11-err.cl

AMDGPU: Add back half and bfloat support for global_load_tr16 pats (#99540)

2024-07-18 11:23:35 -07:00

builtins-amdgcn-global-load-tr-gfx12-w32-err.cl

AMDGPU: Add back half and bfloat support for global_load_tr16 pats (#99540)

2024-07-18 11:23:35 -07:00

builtins-amdgcn-global-load-tr-gfx12-w64-err.cl

AMDGPU: Add back half and bfloat support for global_load_tr16 pats (#99540)

2024-07-18 11:23:35 -07:00

builtins-amdgcn-global-load-tr-w32.cl

AMDGPU: Add back half and bfloat support for global_load_tr16 pats (#99540)

2024-07-18 11:23:35 -07:00

builtins-amdgcn-global-load-tr-w64.cl

AMDGPU: Add back half and bfloat support for global_load_tr16 pats (#99540)

2024-07-18 11:23:35 -07:00

builtins-amdgcn-gws-insts.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-interp.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-make-buffer-rsrc.cl

[Clang][AMDGPU] Add a builtin for llvm.amdgcn.make.buffer.rsrc intrinsic (#95276)

2024-06-20 11:01:54 -04:00

builtins-amdgcn-mfma.cl

AMDGPU: Add v_smfmac_f32_32x32x64_fp8_fp8 for gfx950 (#117259)

2024-11-22 12:11:06 -08:00

builtins-amdgcn-raw-buffer-load.cl

[Clang][AMDGPU] Add builtins for instrinsic llvm.amdgcn.raw.ptr.buffer.load (#99258)

2024-07-18 15:33:03 -04:00

builtins-amdgcn-raw-buffer-store.cl

[Clang][AMDGPU] Use unsigned data type for __builtin_amdgcn_raw_buffer_store_* (#99546)

2024-07-18 16:34:59 -04:00

builtins-amdgcn-raytracing.cl

[test] %clang_cc1: remove redundant actions

2024-05-05 10:46:06 -07:00

builtins-amdgcn-swmmac-w32.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-swmmac-w64.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-vi.cl

[AMDGPU] Relax __builtin_amdgcn_update_dpp sema check (#113341)

2024-10-22 12:32:08 -07:00

builtins-amdgcn-wave32.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-wave64.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-wmma-w32-gfx10-err.cl

[AMDGPU] Check wavefrontsize for GFX11 WMMA builtins (#79980)

2024-02-01 10:49:42 +00:00

builtins-amdgcn-wmma-w32-param-err.cl

…

builtins-amdgcn-wmma-w32.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn-wmma-w64-gfx10-err.cl

[AMDGPU] Check wavefrontsize for GFX11 WMMA builtins (#79980)

2024-02-01 10:49:42 +00:00

builtins-amdgcn-wmma-w64-param-err.cl

…

builtins-amdgcn-wmma-w64.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-amdgcn.cl

[InstCombine] Infer nusw + nneg -> nuw for getelementptr (#111144)

2024-12-05 14:36:40 +01:00

builtins-f16.cl

Emit constrained atan2 intrinsic for clang builtin (#113636)

2024-11-12 13:34:29 -08:00

builtins-fp-atomics-gfx8.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

builtins-fp-atomics-gfx12.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

builtins-fp-atomics-gfx90a.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

builtins-fp-atomics-gfx940.cl

clang/AMDGPU: Emit atomicrmw for global/flat fadd v2bf16 builtins (#96875)

2024-08-20 23:20:03 +04:00

builtins-fp-atomics-gfx1030.cl

…

builtins-generic-amdgcn.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

builtins-r600.cl

[Clang] Swap range and noundef metadata to attribute for intrinsics. (#94851)

2024-06-19 17:23:53 +02:00

builtins.cl

…

byval.cl

…

cast_image.cl

…

cl20-device-side-enqueue-attributes.cl

[Clang] Avoid Using byval for ndrange_t when emitting __enqueue_kernel_basic (#116435)

2024-11-15 16:54:29 -05:00

cl20-device-side-enqueue.cl

[Clang] Avoid Using byval for ndrange_t when emitting __enqueue_kernel_basic (#116435)

2024-11-15 16:54:29 -05:00

cl-strict-aliasing.cl

…

cl-uniform-wg-size.cl

…

const-str-array-decay.cl

…

constant-addr-space-globals.cl

…

convergent.cl

…

debug-info-programming-language.cl

…

enqueue-kernel-non-entry-block.cl

[clang][OpenCL][CodeGen][AMDGPU] Do not use private as the default AS for when generic is available (#112442)

2024-10-22 12:05:48 +01:00

event_t.cl

…

ext-int-shift.cl

…

ext-vector-shuffle.cl

…

fdeclare-opencl-builtins.cl

…

fpmath.cl

…

func-call-dbg-loc.cl

…

half.cl

…

images.cl

…

inline-asm-amdgcn.cl

[clang][SPIR-V] Add support for AMDGCN flavoured SPIRV (#89796)

2024-06-07 11:50:23 +01:00

intel-subgroups-avc-ext-types.cl

…

kernel-arg-info-single-as.cl

…

kernel-arg-info.cl

…

kernel-attributes.cl

…

kernel-metadata.cl

…

kernel-param-alignment.cl

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

kernels-have-spir-cc-by-default.cl

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

lifetime.cl

…

local-initializer-undef.cl

…

local.cl

…

logical-ops.cl

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

memcpy.cl

…

no-half.cl

…

no-signed-zeros.cl

…

norecurse.cl

…

null_queue.cl

…

numbered-address-space.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

opencl_types.cl

AMDGPU: Fix libcall recognition of image array types (#119832)

2024-12-16 15:04:53 +09:00

overload.cl

…

partial_initializer.cl

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

pipe_builtin.cl

[OpenCL] Restore addrspacecast for pipe builtins (#112514)

2024-10-16 13:58:12 +02:00

pipe_types_mangling.cl

…

pipe_types.cl

…

preserve_vec3.cl

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

printf.cl

…

private-array-initialization.cl

…

ptx-calls.cl

[NVPTX] Switch front-ends and tests to ptx_kernel cc (#120806)

2025-01-07 18:24:50 -08:00

ptx-kernels.cl

[NVPTX] Switch front-ends and tests to ptx_kernel cc (#120806)

2025-01-07 18:24:50 -08:00

reflect.cl

[NVPTX] Switch front-ends and tests to ptx_kernel cc (#120806)

2025-01-07 18:24:50 -08:00

relaxed-fpmath.cl

Remove FiniteMathOnly and use only NoHonorINFs and NoHonorNANs. (#97342)

2024-07-26 08:16:38 -04:00

sampler.cl

…

shifts.cl

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

single-precision-constant.cl

…

size_t.cl

…

spir32_target.cl

…

spir64_target.cl

…

spir_version.cl

[OpenCL] Change default standard version to CL1.2

2021-07-26 15:04:34 +01:00

spir-calling-conv.cl

…

spir-debug-info-pointer-address-space.cl

…

spirv_target.cl

…

sqrt-fpmath.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

str_literals.cl

…

to_addr_builtin.cl

…

unroll-hint.cl

…

vector_literals.cl

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

vector_logops.cl

…

vector_odd.cl

…

vector_shufflevector.cl

…

vectorLoadStore.cl

…

visibility.cl

[test] %clang_cc1 -emit-llvm: remove redundant -S

2024-05-04 17:00:29 -07:00

vla.cl

…