[PATCH] D132634: [AArch64] Add index check before lowerInterleavedStore() uses ShuffleVectorInst's mask

[Peter Rong via Phabricator via llvm-commits]

Peter added inline comments.


================
Comment at: llvm/test/CodeGen/AArch64/aarch64-shuffle-undef.ll:24
+  %B = urem <32 x i16> <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1>, <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1>
+  %S = shufflevector <32 x i16> %B, <32 x i16> %B, <32 x i32> undef
+  store <32 x i16> %S, <32 x i16>* %A2, align 64
----------------
Peter wrote:
> fhahn wrote:
> > can the vectors be shortened here? Does it also crash with a `poison` mask?
> From my experience, it seems only lengths of 32 will trigger the bug.
> 
> I just tested again. `undef` crashes both Debug and Release build. `poison` only crashes Debug build. This just adds the mystery for me...
> 
> I checked the failed testcases in Transform, it seems they only considered when part of the Mask is undef or poison, not the whole Mask.
Ok I managed to find a testcase with only vector length of 16.

```
; ModuleID = 'PoC.ll'
source_filename = "M"

define void @f() {
BB:
  %A2 = alloca <16 x i32>, align 64
  %B = urem <16 x i32> <i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1>, <i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1>
  %S = shufflevector <16 x i32> %B, <16 x i32> %B, <16 x i32> undef
  store <16 x i32> %S, <16 x i32>* %A2, align 64
  ret void
}
```

What I don't understand is, if we change this case's integer type from i32 to i16, `llc` will complaint, even though the instruction seems legit to me

```
./build-debug/bin/llc: error: ./build-debug/bin/llc: PoC.ll:8:22: error: invalid shufflevector operands
  %S = shufflevector <16 x i16> %B, <16 x i16> %B, <16 x i16> undef
```

But when switching to `<16 x i32>`, it crashes...


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D132634/new/

https://reviews.llvm.org/D132634