324bdd662d
This change adds a pass to scalarize vectors in global scope into arrays. There are three distinct parts 1. find the globals that need to be updated and define what the new type should be 2. initialize that new type and copy over all the right attributes over from the old type. 3. Use the instruction visitor pattern to update the loads, stores, and geps for the layout of the new data structure. resolves https://github.com/llvm/llvm-project/issues/107920
13 lines
792 B
LLVM
13 lines
792 B
LLVM
; RUN: opt -S -dxil-data-scalarization -scalarizer -scalarize-load-store -dxil-op-lower -mtriple=dxil-pc-shadermodel6.3-library %s | FileCheck %s
|
|
; RUN: llc %s -mtriple=dxil-pc-shadermodel6.3-library --filetype=asm -o - | FileCheck %s
|
|
|
|
; Make sure we don't touch arrays without vectors and that can recurse multiple-dimension arrays of vectors
|
|
|
|
@staticArray = internal global [4 x i32] [i32 1, i32 2, i32 3, i32 4], align 4
|
|
@"groushared3dArrayofVectors" = local_unnamed_addr addrspace(3) global [3 x [3 x [3 x <4 x i32>]]] zeroinitializer, align 16
|
|
|
|
; CHECK @staticArray
|
|
; CHECK-NOT: @staticArray.scalarized
|
|
; CHECK: @groushared3dArrayofVectors.scalarized = local_unnamed_addr addrspace(3) global [3 x [3 x [3 x [4 x i32]]]] zeroinitializer, align 16
|
|
; CHECK-NOT: @groushared3dArrayofVectors
|