llvm-project/llvm/test/Transforms/LICM/sink_multiple.ll

20 lines
666 B
LLVM

; The loop sinker was running from the bottom of the loop to the top, causing
; it to miss opportunities to sink instructions that depended on sinking other
; instructions from the loop. Instead they got hoisted, which is better than
; leaving them in the loop, but increases register pressure pointlessly.
; RUN: llvm-upgrade < %s | llvm-as | opt -licm | llvm-dis | %prcontext getelementptr 1 | grep Out:
%Ty = type { int, int }
%X = external global %Ty
int %test() {
br label %Loop
Loop:
%dead = getelementptr %Ty* %X, long 0, uint 0
%sunk2 = load int* %dead
br bool false, label %Loop, label %Out
Out:
ret int %sunk2
}