86813aa786
Epilogue vectorization currently relies on the resume phi for the canonical induction being always available, which is why VPPhi are considered to have side-effects, to prevent their removal. This patch adds a new ResumeForEpilogue opcode to mark the resume phi as used for epilogue vectorization. This allows treating VPPhis in general as not having side-effects, enabling removal of unused VPPhis.
106 lines
5.3 KiB
LLVM
106 lines
5.3 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
|
|
; RUN: opt -p loop-vectorize -force-vector-width=4 -force-vector-interleave=1 -prefer-predicate-over-epilogue=predicate-dont-vectorize -S %s | FileCheck %s
|
|
target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128-Fn32"
|
|
|
|
define void @tail_fold_switch(ptr %dst, i32 %0) {
|
|
; CHECK-LABEL: define void @tail_fold_switch(
|
|
; CHECK-SAME: ptr [[DST:%.*]], i32 [[TMP0:%.*]]) {
|
|
; CHECK-NEXT: [[ENTRY:.*:]]
|
|
; CHECK-NEXT: br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
|
|
; CHECK: [[VECTOR_PH]]:
|
|
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[TMP0]], i64 0
|
|
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer
|
|
; CHECK-NEXT: [[TMP2:%.*]] = icmp eq <4 x i32> [[BROADCAST_SPLAT]], splat (i32 1)
|
|
; CHECK-NEXT: br label %[[VECTOR_BODY:.*]]
|
|
; CHECK: [[VECTOR_BODY]]:
|
|
; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[PRED_STORE_CONTINUE6:.*]] ]
|
|
; CHECK-NEXT: [[VEC_IND:%.*]] = phi <4 x i8> [ <i8 0, i8 1, i8 2, i8 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[PRED_STORE_CONTINUE6]] ]
|
|
; CHECK-NEXT: [[TMP1:%.*]] = icmp ule <4 x i8> [[VEC_IND]], splat (i8 4)
|
|
; CHECK-NEXT: [[TMP3:%.*]] = select <4 x i1> [[TMP1]], <4 x i1> [[TMP2]], <4 x i1> zeroinitializer
|
|
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i1> [[TMP3]], i32 0
|
|
; CHECK-NEXT: br i1 [[TMP4]], label %[[PRED_STORE_IF:.*]], label %[[PRED_STORE_CONTINUE:.*]]
|
|
; CHECK: [[PRED_STORE_IF]]:
|
|
; CHECK-NEXT: [[TMP5:%.*]] = add i64 [[INDEX]], 0
|
|
; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[DST]], i64 [[TMP5]]
|
|
; CHECK-NEXT: store i32 0, ptr [[TMP6]], align 4
|
|
; CHECK-NEXT: br label %[[PRED_STORE_CONTINUE]]
|
|
; CHECK: [[PRED_STORE_CONTINUE]]:
|
|
; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x i1> [[TMP3]], i32 1
|
|
; CHECK-NEXT: br i1 [[TMP7]], label %[[PRED_STORE_IF1:.*]], label %[[PRED_STORE_CONTINUE2:.*]]
|
|
; CHECK: [[PRED_STORE_IF1]]:
|
|
; CHECK-NEXT: [[TMP8:%.*]] = add i64 [[INDEX]], 1
|
|
; CHECK-NEXT: [[TMP9:%.*]] = getelementptr inbounds i32, ptr [[DST]], i64 [[TMP8]]
|
|
; CHECK-NEXT: store i32 0, ptr [[TMP9]], align 4
|
|
; CHECK-NEXT: br label %[[PRED_STORE_CONTINUE2]]
|
|
; CHECK: [[PRED_STORE_CONTINUE2]]:
|
|
; CHECK-NEXT: [[TMP10:%.*]] = extractelement <4 x i1> [[TMP3]], i32 2
|
|
; CHECK-NEXT: br i1 [[TMP10]], label %[[PRED_STORE_IF3:.*]], label %[[PRED_STORE_CONTINUE4:.*]]
|
|
; CHECK: [[PRED_STORE_IF3]]:
|
|
; CHECK-NEXT: [[TMP11:%.*]] = add i64 [[INDEX]], 2
|
|
; CHECK-NEXT: [[TMP12:%.*]] = getelementptr inbounds i32, ptr [[DST]], i64 [[TMP11]]
|
|
; CHECK-NEXT: store i32 0, ptr [[TMP12]], align 4
|
|
; CHECK-NEXT: br label %[[PRED_STORE_CONTINUE4]]
|
|
; CHECK: [[PRED_STORE_CONTINUE4]]:
|
|
; CHECK-NEXT: [[TMP13:%.*]] = extractelement <4 x i1> [[TMP3]], i32 3
|
|
; CHECK-NEXT: br i1 [[TMP13]], label %[[PRED_STORE_IF5:.*]], label %[[PRED_STORE_CONTINUE6]]
|
|
; CHECK: [[PRED_STORE_IF5]]:
|
|
; CHECK-NEXT: [[TMP14:%.*]] = add i64 [[INDEX]], 3
|
|
; CHECK-NEXT: [[TMP15:%.*]] = getelementptr inbounds i32, ptr [[DST]], i64 [[TMP14]]
|
|
; CHECK-NEXT: store i32 0, ptr [[TMP15]], align 4
|
|
; CHECK-NEXT: br label %[[PRED_STORE_CONTINUE6]]
|
|
; CHECK: [[PRED_STORE_CONTINUE6]]:
|
|
; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
|
|
; CHECK-NEXT: [[VEC_IND_NEXT]] = add <4 x i8> [[VEC_IND]], splat (i8 4)
|
|
; CHECK-NEXT: [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], 8
|
|
; CHECK-NEXT: br i1 [[TMP16]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
|
|
; CHECK: [[MIDDLE_BLOCK]]:
|
|
; CHECK-NEXT: br label %[[EXIT:.*]]
|
|
; CHECK: [[SCALAR_PH]]:
|
|
; CHECK-NEXT: br label %[[LOOP_HEADER:.*]]
|
|
; CHECK: [[LOOP_HEADER]]:
|
|
; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP_LATCH:.*]] ]
|
|
; CHECK-NEXT: switch i32 [[TMP0]], label %[[LOOP_LATCH]] [
|
|
; CHECK-NEXT: i32 0, label %[[LOOP_LATCH]]
|
|
; CHECK-NEXT: i32 1, label %[[IF_THEN:.*]]
|
|
; CHECK-NEXT: ]
|
|
; CHECK: [[IF_THEN]]:
|
|
; CHECK-NEXT: [[GEP:%.*]] = getelementptr inbounds i32, ptr [[DST]], i64 [[IV]]
|
|
; CHECK-NEXT: store i32 0, ptr [[GEP]], align 4
|
|
; CHECK-NEXT: br label %[[LOOP_LATCH]]
|
|
; CHECK: [[LOOP_LATCH]]:
|
|
; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1
|
|
; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], 4
|
|
; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP3:![0-9]+]]
|
|
; CHECK: [[EXIT]]:
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
entry:
|
|
br label %loop.header
|
|
|
|
loop.header:
|
|
%iv = phi i64 [ 0, %entry ], [ %iv.next, %loop.latch ]
|
|
switch i32 %0, label %loop.latch [
|
|
i32 0, label %loop.latch
|
|
i32 1, label %if.then
|
|
]
|
|
|
|
if.then:
|
|
%gep = getelementptr inbounds i32, ptr %dst, i64 %iv
|
|
store i32 0, ptr %gep, align 4
|
|
br label %loop.latch
|
|
|
|
loop.latch:
|
|
%iv.next = add i64 %iv, 1
|
|
%ec = icmp eq i64 %iv, 4
|
|
br i1 %ec, label %exit, label %loop.header
|
|
|
|
exit:
|
|
ret void
|
|
}
|
|
;.
|
|
; CHECK: [[LOOP0]] = distinct !{[[LOOP0]], [[META1:![0-9]+]], [[META2:![0-9]+]]}
|
|
; CHECK: [[META1]] = !{!"llvm.loop.isvectorized", i32 1}
|
|
; CHECK: [[META2]] = !{!"llvm.loop.unroll.runtime.disable"}
|
|
; CHECK: [[LOOP3]] = distinct !{[[LOOP3]], [[META2]], [[META1]]}
|
|
;.
|