[llvm] Support CSA for floating-point & pointer types (PR #182313)

via llvm-commits llvm-commits at lists.llvm.org
Thu Feb 19 09:17:01 PST 2026


github-actions[bot] wrote:

<!--PREMERGE ADVISOR COMMENT: Linux-->
# :penguin: Linux x64 Test Results

* 169836 tests passed
* 3071 tests skipped
* 2 tests failed

## Failed Tests
(click on a test name to see its output)

### LLVM
<details>
<summary>LLVM.Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll</summary>

```
Exit Code: 1

Command Output (stdout):
--
# RUN: at line 2
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/opt -p loop-vectorize -mtriple riscv64 -mattr=+v -S < /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/opt -p loop-vectorize -mtriple riscv64 -mattr=+v -S
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
# .---command stderr------------
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll:247:15: error: CHECK-NEXT: is not on the line after the previous match
# | ; CHECK-NEXT: br label %[[FOR_BODY:.*]]
# |               ^
# | <stdin>:178:2: note: 'next' match was here
# |  br label %vector.body
# |  ^
# | <stdin>:171:7: note: previous match ended here
# | entry:
# |       ^
# | <stdin>:172:1: note: non-matching line after previous match is here
# |  %min.iters.check = icmp ult i64 %n, 16
# | ^
# | 
# | Input file: <stdin>
# | Check file: /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/RISCV/select-cmp-reduction.ll
# | 
# | -dump-input=help explains the following input dump.
# | 
# | Input was:
# | <<<<<<
# |           .
# |           .
# |           .
# |         173:  br i1 %min.iters.check, label %scalar.ph, label %vector.ph 
# |         174:  
# |         175: vector.ph: ; preds = %entry 
# |         176:  %n.mod.vf = urem i64 %n, 8 
# |         177:  %n.vec = sub i64 %n, %n.mod.vf 
# |         178:  br label %vector.body 
# | next:247      !~~~~~~~~~~~~~~~~~~~~  error: match on wrong line
# |         179:  
# |         180: vector.body: ; preds = %vector.body, %vector.ph 
# |         181:  %index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ] 
# |         182:  %vec.phi = phi <8 x float> [ splat (float 3.000000e+00), %vector.ph ], [ %6, %vector.body ] 
# |         183:  %0 = phi <8 x i1> [ zeroinitializer, %vector.ph ], [ %5, %vector.body ] 
# |           .
# |           .
# |           .
# | >>>>>>
# `-----------------------------
# error: command failed with exit status: 1

--

```
</details>
<details>
<summary>LLVM.Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll</summary>

```
Exit Code: 1

Command Output (stdout):
--
# RUN: at line 2
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/opt -passes=loop-vectorize -S < /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll 2>&1 | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll --check-prefix=X86
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/opt -passes=loop-vectorize -S
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll --check-prefix=X86
# note: command had no output on stdout or stderr
# RUN: at line 3
/home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/opt -passes=loop-vectorize -mattr=+avx512f -S < /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll 2>&1 | /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll --check-prefix=AVX512
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/opt -passes=loop-vectorize -mattr=+avx512f -S
# note: command had no output on stdout or stderr
# executed command: /home/gha/actions-runner/_work/llvm-project/llvm-project/build/bin/FileCheck /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll --check-prefix=AVX512
# .---command stderr------------
# | /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll:144:16: error: AVX512-NEXT: is not on the line after the previous match
# | ; AVX512-NEXT: br label %[[LOOP:.*]]
# |                ^
# | <stdin>:71:2: note: 'next' match was here
# |  br label %vector.body
# |  ^
# | <stdin>:60:7: note: previous match ended here
# | entry:
# |       ^
# | <stdin>:61:1: note: non-matching line after previous match is here
# |  %min.iters.check = icmp ult i64 %N, 24
# | ^
# | 
# | Input file: <stdin>
# | Check file: /home/gha/actions-runner/_work/llvm-project/llvm-project/llvm/test/Transforms/LoopVectorize/X86/conditional-scalar-assignment.ll
# | 
# | -dump-input=help explains the following input dump.
# | 
# | Input was:
# | <<<<<<
# |           .
# |           .
# |           .
# |          66:  %n.vec = sub i64 %N, %n.mod.vf 
# |          67:  %broadcast.splatinsert = insertelement <8 x i64> poison, i64 %a, i64 0 
# |          68:  %broadcast.splat = shufflevector <8 x i64> %broadcast.splatinsert, <8 x i64> poison, <8 x i32> zeroinitializer 
# |          69:  %broadcast.splatinsert1 = insertelement <8 x ptr> poison, ptr %init, i64 0 
# |          70:  %broadcast.splat2 = shufflevector <8 x ptr> %broadcast.splatinsert1, <8 x ptr> poison, <8 x i32> zeroinitializer 
# |          71:  br label %vector.body 
# | next:144      !~~~~~~~~~~~~~~~~~~~~  error: match on wrong line
# |          72:  
# |          73: vector.body: ; preds = %vector.body, %vector.ph 
# |          74:  %index = phi i64 [ 0, %vector.ph ], [ %index.next, %vector.body ] 
# |          75:  %vec.phi = phi <8 x ptr> [ %broadcast.splat2, %vector.ph ], [ %7, %vector.body ] 
# |          76:  %0 = phi <8 x i1> [ zeroinitializer, %vector.ph ], [ %6, %vector.body ] 
# |           .
# |           .
# |           .
# | >>>>>>
# `-----------------------------
# error: command failed with exit status: 1

--

```
</details>

If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label.

https://github.com/llvm/llvm-project/pull/182313


More information about the llvm-commits mailing list