Page MenuHomePhabricator

Please use GitHub pull requests for new patches. Phabricator shutdown timeline

Differential D142894 Diff 501488 llvm/test/Transforms/LoopVectorize/RISCV/uniform-load-store.ll

Changeset View

Standalone View

View Options

llvm/test/Transforms/LoopVectorize/RISCV/uniform-load-store.ll

Show All 40 Lines
; SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; SCALABLE: for.body: ; SCALABLE: for.body:
; SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; SCALABLE-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8 ; SCALABLE-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8
; SCALABLE-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]] ; SCALABLE-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]]
; SCALABLE-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8 ; SCALABLE-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8
; SCALABLE-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; SCALABLE-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; SCALABLE-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; SCALABLE-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; SCALABLE-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP2:![0-9]+]] ; SCALABLE-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; SCALABLE: for.end: ; SCALABLE: for.end:
; SCALABLE-NEXT: ret void ; SCALABLE-NEXT: ret void
; ;
; FIXEDLEN-LABEL: @uniform_load( ; FIXEDLEN-LABEL: @uniform_load(
; FIXEDLEN-NEXT: entry: ; FIXEDLEN-NEXT: entry:
; FIXEDLEN-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]] ; FIXEDLEN-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; FIXEDLEN: vector.ph: ; FIXEDLEN: vector.ph:
; FIXEDLEN-NEXT: br label [[VECTOR_BODY:%.*]] ; FIXEDLEN-NEXT: br label [[VECTOR_BODY:%.*]]
Show All 21 Lines
; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] ; FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]]
; FIXEDLEN: for.body: ; FIXEDLEN: for.body:
; FIXEDLEN-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; FIXEDLEN-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; FIXEDLEN-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8 ; FIXEDLEN-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8
; FIXEDLEN-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]] ; FIXEDLEN-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]]
; FIXEDLEN-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8 ; FIXEDLEN-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8
; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP2:![0-9]+]] ; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; FIXEDLEN: for.end: ; FIXEDLEN: for.end:
; FIXEDLEN-NEXT: ret void ; FIXEDLEN-NEXT: ret void
; ;
; TF-SCALABLE-LABEL: @uniform_load( ; TF-SCALABLE-LABEL: @uniform_load(
; TF-SCALABLE-NEXT: entry: ; TF-SCALABLE-NEXT: entry:
; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE-NEXT: [[TMP1:%.*]] = icmp ult i64 -1025, [[TMP0]]
; TF-SCALABLE-NEXT: br i1 [[TMP1]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE: vector.ph: ; TF-SCALABLE: vector.ph:
; TF-SCALABLE-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP4:%.*]] = sub i64 [[TMP3]], 1 ; TF-SCALABLE-NEXT: [[TMP2:%.*]] = sub i64 [[TMP1]], 1
; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP4]] ; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP2]]
; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP2]] ; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP0]]
; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]] ; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; TF-SCALABLE: vector.body: ; TF-SCALABLE: vector.body:
; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[TMP5:%.*]] = add i64 [[INDEX]], 0 ; TF-SCALABLE-NEXT: [[TMP3:%.*]] = add i64 [[INDEX]], 0
; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP5]], i64 1024) ; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP3]], i64 1024)
; TF-SCALABLE-NEXT: [[TMP6:%.*]] = load i64, ptr [[B:%.*]], align 8 ; TF-SCALABLE-NEXT: [[TMP4:%.*]] = load i64, ptr [[B:%.*]], align 8
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP6]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP4]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: [[TMP7:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP5]] ; TF-SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP3]]
; TF-SCALABLE-NEXT: [[TMP8:%.*]] = getelementptr inbounds i64, ptr [[TMP7]], i32 0 ; TF-SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i64, ptr [[TMP5]], i32 0
; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP8]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP6]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]])
; TF-SCALABLE-NEXT: [[TMP9:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP7:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP9]] ; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP7]]
; TF-SCALABLE-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] ; TF-SCALABLE-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; TF-SCALABLE-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; TF-SCALABLE: middle.block: ; TF-SCALABLE: middle.block:
; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]]
; TF-SCALABLE: scalar.ph: ; TF-SCALABLE: scalar.ph:
; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; TF-SCALABLE: for.body: ; TF-SCALABLE: for.body:
; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; TF-SCALABLE-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8 ; TF-SCALABLE-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8
; TF-SCALABLE-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]] ; TF-SCALABLE-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]]
; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8 ; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8
; TF-SCALABLE-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; TF-SCALABLE-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; TF-SCALABLE-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; TF-SCALABLE-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; TF-SCALABLE-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP2:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; TF-SCALABLE: for.end: ; TF-SCALABLE: for.end:
; TF-SCALABLE-NEXT: ret void ; TF-SCALABLE-NEXT: ret void
; ;
; TF-FIXEDLEN-LABEL: @uniform_load( ; TF-FIXEDLEN-LABEL: @uniform_load(
; TF-FIXEDLEN-NEXT: entry: ; TF-FIXEDLEN-NEXT: entry:
; TF-FIXEDLEN-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]] ; TF-FIXEDLEN-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-FIXEDLEN: vector.ph: ; TF-FIXEDLEN: vector.ph:
; TF-FIXEDLEN-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-FIXEDLEN-NEXT: br label [[VECTOR_BODY:%.*]]
Show All 17 Lines
; TF-FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]] ; TF-FIXEDLEN-NEXT: br label [[FOR_BODY:%.*]]
; TF-FIXEDLEN: for.body: ; TF-FIXEDLEN: for.body:
; TF-FIXEDLEN-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; TF-FIXEDLEN-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; TF-FIXEDLEN-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8 ; TF-FIXEDLEN-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 8
; TF-FIXEDLEN-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]] ; TF-FIXEDLEN-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 [[IV]]
; TF-FIXEDLEN-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8 ; TF-FIXEDLEN-NEXT: store i64 [[V]], ptr [[ARRAYIDX]], align 8
; TF-FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; TF-FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; TF-FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; TF-FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; TF-FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP2:![0-9]+]] ; TF-FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; TF-FIXEDLEN: for.end: ; TF-FIXEDLEN: for.end:
; TF-FIXEDLEN-NEXT: ret void ; TF-FIXEDLEN-NEXT: ret void
; ;
entry: entry:
br label %for.body br label %for.body
for.body: for.body:
%iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ] %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
▲ Show 20 Lines • Show All 269 Lines • ▼ Show 20 Lines
; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]] ; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]]
; FIXEDLEN: for.end: ; FIXEDLEN: for.end:
; FIXEDLEN-NEXT: ret void ; FIXEDLEN-NEXT: ret void
; ;
; TF-SCALABLE-LABEL: @conditional_uniform_load( ; TF-SCALABLE-LABEL: @conditional_uniform_load(
; TF-SCALABLE-NEXT: entry: ; TF-SCALABLE-NEXT: entry:
; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE-NEXT: [[TMP1:%.*]] = icmp ult i64 -1025, [[TMP0]]
; TF-SCALABLE-NEXT: br i1 [[TMP1]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE: vector.ph: ; TF-SCALABLE: vector.ph:
; TF-SCALABLE-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP4:%.*]] = sub i64 [[TMP3]], 1 ; TF-SCALABLE-NEXT: [[TMP2:%.*]] = sub i64 [[TMP1]], 1
; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP4]] ; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP2]]
; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP2]] ; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP0]]
; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]] ; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
; TF-SCALABLE-NEXT: [[TMP5:%.*]] = call <vscale x 1 x i64> @llvm.experimental.stepvector.nxv1i64() ; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call <vscale x 1 x i64> @llvm.experimental.stepvector.nxv1i64()
; TF-SCALABLE-NEXT: [[TMP6:%.*]] = add <vscale x 1 x i64> [[TMP5]], zeroinitializer ; TF-SCALABLE-NEXT: [[TMP4:%.*]] = add <vscale x 1 x i64> [[TMP3]], zeroinitializer
; TF-SCALABLE-NEXT: [[TMP7:%.*]] = mul <vscale x 1 x i64> [[TMP6]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 1, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer) ; TF-SCALABLE-NEXT: [[TMP5:%.*]] = mul <vscale x 1 x i64> [[TMP4]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 1, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer)
; TF-SCALABLE-NEXT: [[INDUCTION:%.*]] = add <vscale x 1 x i64> zeroinitializer, [[TMP7]] ; TF-SCALABLE-NEXT: [[INDUCTION:%.*]] = add <vscale x 1 x i64> zeroinitializer, [[TMP5]]
; TF-SCALABLE-NEXT: [[TMP8:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP9:%.*]] = mul i64 1, [[TMP8]] ; TF-SCALABLE-NEXT: [[TMP7:%.*]] = mul i64 1, [[TMP6]]
; TF-SCALABLE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP9]], i64 0 ; TF-SCALABLE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP7]], i64 0
; TF-SCALABLE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[DOTSPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[DOTSPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x ptr> poison, ptr [[B:%.*]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x ptr> poison, ptr [[B:%.*]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x ptr> [[BROADCAST_SPLATINSERT]], <vscale x 1 x ptr> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x ptr> [[BROADCAST_SPLATINSERT]], <vscale x 1 x ptr> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; TF-SCALABLE: vector.body: ; TF-SCALABLE: vector.body:
; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[VEC_IND:%.*]] = phi <vscale x 1 x i64> [ [[INDUCTION]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[VEC_IND:%.*]] = phi <vscale x 1 x i64> [ [[INDUCTION]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[TMP10:%.*]] = add i64 [[INDEX]], 0 ; TF-SCALABLE-NEXT: [[TMP8:%.*]] = add i64 [[INDEX]], 0
; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP10]], i64 1024) ; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP8]], i64 1024)
; TF-SCALABLE-NEXT: [[TMP11:%.*]] = icmp ugt <vscale x 1 x i64> [[VEC_IND]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 10, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer) ; TF-SCALABLE-NEXT: [[TMP9:%.*]] = icmp ugt <vscale x 1 x i64> [[VEC_IND]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 10, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer)
; TF-SCALABLE-NEXT: [[TMP10:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP9]], <vscale x 1 x i1> zeroinitializer
; TF-SCALABLE-NEXT: [[WIDE_MASKED_GATHER:%.*]] = call <vscale x 1 x i64> @llvm.masked.gather.nxv1i64.nxv1p0(<vscale x 1 x ptr> [[BROADCAST_SPLAT]], i32 8, <vscale x 1 x i1> [[TMP10]], <vscale x 1 x i64> poison)
; TF-SCALABLE-NEXT: [[TMP11:%.*]] = xor <vscale x 1 x i1> [[TMP9]], shufflevector (<vscale x 1 x i1> insertelement (<vscale x 1 x i1> poison, i1 true, i64 0), <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer)
; TF-SCALABLE-NEXT: [[TMP12:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP11]], <vscale x 1 x i1> zeroinitializer ; TF-SCALABLE-NEXT: [[TMP12:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP11]], <vscale x 1 x i1> zeroinitializer
; TF-SCALABLE-NEXT: [[WIDE_MASKED_GATHER:%.*]] = call <vscale x 1 x i64> @llvm.masked.gather.nxv1i64.nxv1p0(<vscale x 1 x ptr> [[BROADCAST_SPLAT]], i32 8, <vscale x 1 x i1> [[TMP12]], <vscale x 1 x i64> poison) ; TF-SCALABLE-NEXT: [[PREDPHI:%.*]] = select <vscale x 1 x i1> [[TMP10]], <vscale x 1 x i64> [[WIDE_MASKED_GATHER]], <vscale x 1 x i64> zeroinitializer
; TF-SCALABLE-NEXT: [[TMP13:%.*]] = xor <vscale x 1 x i1> [[TMP11]], shufflevector (<vscale x 1 x i1> insertelement (<vscale x 1 x i1> poison, i1 true, i64 0), <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer) ; TF-SCALABLE-NEXT: [[TMP13:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP8]]
; TF-SCALABLE-NEXT: [[TMP14:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP13]], <vscale x 1 x i1> zeroinitializer ; TF-SCALABLE-NEXT: [[TMP14:%.*]] = or <vscale x 1 x i1> [[TMP10]], [[TMP12]]
; TF-SCALABLE-NEXT: [[PREDPHI:%.*]] = select <vscale x 1 x i1> [[TMP12]], <vscale x 1 x i64> [[WIDE_MASKED_GATHER]], <vscale x 1 x i64> zeroinitializer ; TF-SCALABLE-NEXT: [[TMP15:%.*]] = getelementptr inbounds i64, ptr [[TMP13]], i32 0
; TF-SCALABLE-NEXT: [[TMP15:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP10]] ; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[PREDPHI]], ptr [[TMP15]], i32 8, <vscale x 1 x i1> [[TMP14]])
; TF-SCALABLE-NEXT: [[TMP16:%.*]] = or <vscale x 1 x i1> [[TMP12]], [[TMP14]] ; TF-SCALABLE-NEXT: [[TMP16:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP17:%.*]] = getelementptr inbounds i64, ptr [[TMP15]], i32 0 ; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP16]]
; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[PREDPHI]], ptr [[TMP17]], i32 8, <vscale x 1 x i1> [[TMP16]])
; TF-SCALABLE-NEXT: [[TMP18:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP18]]
; TF-SCALABLE-NEXT: [[VEC_IND_NEXT]] = add <vscale x 1 x i64> [[VEC_IND]], [[DOTSPLAT]] ; TF-SCALABLE-NEXT: [[VEC_IND_NEXT]] = add <vscale x 1 x i64> [[VEC_IND]], [[DOTSPLAT]]
; TF-SCALABLE-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] ; TF-SCALABLE-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; TF-SCALABLE-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; TF-SCALABLE: middle.block: ; TF-SCALABLE: middle.block:
; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]]
; TF-SCALABLE: scalar.ph: ; TF-SCALABLE: scalar.ph:
; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; TF-SCALABLE: for.body: ; TF-SCALABLE: for.body:
; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LATCH:%.*]] ] ; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LATCH:%.*]] ]
; TF-SCALABLE-NEXT: [[CMP:%.*]] = icmp ugt i64 [[IV]], 10 ; TF-SCALABLE-NEXT: [[CMP:%.*]] = icmp ugt i64 [[IV]], 10
▲ Show 20 Lines • Show All 155 Lines • ▼ Show 20 Lines
; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP9:![0-9]+]] ; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP9:![0-9]+]]
; FIXEDLEN: for.end: ; FIXEDLEN: for.end:
; FIXEDLEN-NEXT: ret void ; FIXEDLEN-NEXT: ret void
; ;
; TF-SCALABLE-LABEL: @uniform_load_unaligned( ; TF-SCALABLE-LABEL: @uniform_load_unaligned(
; TF-SCALABLE-NEXT: entry: ; TF-SCALABLE-NEXT: entry:
; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE-NEXT: [[TMP1:%.*]] = icmp ult i64 -1025, [[TMP0]]
; TF-SCALABLE-NEXT: br i1 [[TMP1]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE: vector.ph: ; TF-SCALABLE: vector.ph:
; TF-SCALABLE-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP4:%.*]] = sub i64 [[TMP3]], 1 ; TF-SCALABLE-NEXT: [[TMP2:%.*]] = sub i64 [[TMP1]], 1
; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP4]] ; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP2]]
; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP2]] ; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP0]]
; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]] ; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; TF-SCALABLE: vector.body: ; TF-SCALABLE: vector.body:
; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[TMP5:%.*]] = add i64 [[INDEX]], 0 ; TF-SCALABLE-NEXT: [[TMP3:%.*]] = add i64 [[INDEX]], 0
; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP5]], i64 1024) ; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP3]], i64 1024)
; TF-SCALABLE-NEXT: [[TMP6:%.*]] = load i64, ptr [[B:%.*]], align 1 ; TF-SCALABLE-NEXT: [[TMP4:%.*]] = load i64, ptr [[B:%.*]], align 1
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP6]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP4]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: [[TMP7:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP5]] ; TF-SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP3]]
; TF-SCALABLE-NEXT: [[TMP8:%.*]] = getelementptr inbounds i64, ptr [[TMP7]], i32 0 ; TF-SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i64, ptr [[TMP5]], i32 0
; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP8]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP6]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]])
; TF-SCALABLE-NEXT: [[TMP9:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP7:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP9]] ; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP7]]
; TF-SCALABLE-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] ; TF-SCALABLE-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; TF-SCALABLE-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; TF-SCALABLE: middle.block: ; TF-SCALABLE: middle.block:
; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]]
; TF-SCALABLE: scalar.ph: ; TF-SCALABLE: scalar.ph:
; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; TF-SCALABLE: for.body: ; TF-SCALABLE: for.body:
; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; TF-SCALABLE-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 1 ; TF-SCALABLE-NEXT: [[V:%.*]] = load i64, ptr [[B]], align 1
▲ Show 20 Lines • Show All 133 Lines • ▼ Show 20 Lines
; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP11:![0-9]+]] ; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP11:![0-9]+]]
; FIXEDLEN: for.end: ; FIXEDLEN: for.end:
; FIXEDLEN-NEXT: ret void ; FIXEDLEN-NEXT: ret void
; ;
; TF-SCALABLE-LABEL: @uniform_store( ; TF-SCALABLE-LABEL: @uniform_store(
; TF-SCALABLE-NEXT: entry: ; TF-SCALABLE-NEXT: entry:
; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE-NEXT: [[TMP1:%.*]] = icmp ult i64 -1025, [[TMP0]]
; TF-SCALABLE-NEXT: br i1 [[TMP1]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE: vector.ph: ; TF-SCALABLE: vector.ph:
; TF-SCALABLE-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP4:%.*]] = sub i64 [[TMP3]], 1 ; TF-SCALABLE-NEXT: [[TMP2:%.*]] = sub i64 [[TMP1]], 1
; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP4]] ; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP2]]
; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP2]] ; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP0]]
; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]] ; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; TF-SCALABLE: vector.body: ; TF-SCALABLE: vector.body:
; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[TMP5:%.*]] = add i64 [[INDEX]], 0 ; TF-SCALABLE-NEXT: [[TMP3:%.*]] = add i64 [[INDEX]], 0
; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP5]], i64 1024) ; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP3]], i64 1024)
; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B:%.*]], align 8 ; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B:%.*]], align 8
; TF-SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP5]] ; TF-SCALABLE-NEXT: [[TMP4:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP3]]
; TF-SCALABLE-NEXT: [[TMP7:%.*]] = getelementptr inbounds i64, ptr [[TMP6]], i32 0 ; TF-SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i64, ptr [[TMP4]], i32 0
; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP7]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP5]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]])
; TF-SCALABLE-NEXT: [[TMP8:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP8]] ; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP6]]
; TF-SCALABLE-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] ; TF-SCALABLE-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; TF-SCALABLE-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP8:![0-9]+]]
; TF-SCALABLE: middle.block: ; TF-SCALABLE: middle.block:
; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]]
; TF-SCALABLE: scalar.ph: ; TF-SCALABLE: scalar.ph:
; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; TF-SCALABLE: for.body: ; TF-SCALABLE: for.body:
; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B]], align 8 ; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B]], align 8
▲ Show 20 Lines • Show All 150 Lines • ▼ Show 20 Lines
; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP13:![0-9]+]] ; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP13:![0-9]+]]
; FIXEDLEN: for.end: ; FIXEDLEN: for.end:
; FIXEDLEN-NEXT: ret void ; FIXEDLEN-NEXT: ret void
; ;
; TF-SCALABLE-LABEL: @uniform_store_of_loop_varying( ; TF-SCALABLE-LABEL: @uniform_store_of_loop_varying(
; TF-SCALABLE-NEXT: entry: ; TF-SCALABLE-NEXT: entry:
; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE-NEXT: [[TMP1:%.*]] = icmp ult i64 -1025, [[TMP0]]
; TF-SCALABLE-NEXT: br i1 [[TMP1]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE: vector.ph: ; TF-SCALABLE: vector.ph:
; TF-SCALABLE-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP4:%.*]] = sub i64 [[TMP3]], 1 ; TF-SCALABLE-NEXT: [[TMP2:%.*]] = sub i64 [[TMP1]], 1
; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP4]] ; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP2]]
; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP2]] ; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP0]]
; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]] ; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
; TF-SCALABLE-NEXT: [[TMP5:%.*]] = call <vscale x 1 x i64> @llvm.experimental.stepvector.nxv1i64() ; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call <vscale x 1 x i64> @llvm.experimental.stepvector.nxv1i64()
; TF-SCALABLE-NEXT: [[TMP6:%.*]] = add <vscale x 1 x i64> [[TMP5]], zeroinitializer ; TF-SCALABLE-NEXT: [[TMP4:%.*]] = add <vscale x 1 x i64> [[TMP3]], zeroinitializer
; TF-SCALABLE-NEXT: [[TMP7:%.*]] = mul <vscale x 1 x i64> [[TMP6]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 1, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer) ; TF-SCALABLE-NEXT: [[TMP5:%.*]] = mul <vscale x 1 x i64> [[TMP4]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 1, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer)
; TF-SCALABLE-NEXT: [[INDUCTION:%.*]] = add <vscale x 1 x i64> zeroinitializer, [[TMP7]] ; TF-SCALABLE-NEXT: [[INDUCTION:%.*]] = add <vscale x 1 x i64> zeroinitializer, [[TMP5]]
; TF-SCALABLE-NEXT: [[TMP8:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP9:%.*]] = mul i64 1, [[TMP8]] ; TF-SCALABLE-NEXT: [[TMP7:%.*]] = mul i64 1, [[TMP6]]
; TF-SCALABLE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP9]], i64 0 ; TF-SCALABLE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP7]], i64 0
; TF-SCALABLE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[DOTSPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[DOTSPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x ptr> poison, ptr [[B:%.*]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x ptr> poison, ptr [[B:%.*]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x ptr> [[BROADCAST_SPLATINSERT]], <vscale x 1 x ptr> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x ptr> [[BROADCAST_SPLATINSERT]], <vscale x 1 x ptr> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT1]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT1]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; TF-SCALABLE: vector.body: ; TF-SCALABLE: vector.body:
; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[VEC_IND:%.*]] = phi <vscale x 1 x i64> [ [[INDUCTION]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[VEC_IND:%.*]] = phi <vscale x 1 x i64> [ [[INDUCTION]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[TMP10:%.*]] = add i64 [[INDEX]], 0 ; TF-SCALABLE-NEXT: [[TMP8:%.*]] = add i64 [[INDEX]], 0
; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP10]], i64 1024) ; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP8]], i64 1024)
; TF-SCALABLE-NEXT: call void @llvm.masked.scatter.nxv1i64.nxv1p0(<vscale x 1 x i64> [[VEC_IND]], <vscale x 1 x ptr> [[BROADCAST_SPLAT]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.scatter.nxv1i64.nxv1p0(<vscale x 1 x i64> [[VEC_IND]], <vscale x 1 x ptr> [[BROADCAST_SPLAT]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]])
; TF-SCALABLE-NEXT: [[TMP11:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP10]] ; TF-SCALABLE-NEXT: [[TMP9:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP8]]
; TF-SCALABLE-NEXT: [[TMP12:%.*]] = getelementptr inbounds i64, ptr [[TMP11]], i32 0 ; TF-SCALABLE-NEXT: [[TMP10:%.*]] = getelementptr inbounds i64, ptr [[TMP9]], i32 0
; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT2]], ptr [[TMP12]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT2]], ptr [[TMP10]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]])
; TF-SCALABLE-NEXT: [[TMP13:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP11:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP13]] ; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP11]]
; TF-SCALABLE-NEXT: [[VEC_IND_NEXT]] = add <vscale x 1 x i64> [[VEC_IND]], [[DOTSPLAT]] ; TF-SCALABLE-NEXT: [[VEC_IND_NEXT]] = add <vscale x 1 x i64> [[VEC_IND]], [[DOTSPLAT]]
; TF-SCALABLE-NEXT: [[TMP14:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] ; TF-SCALABLE-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; TF-SCALABLE-NEXT: br i1 [[TMP14]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]]
; TF-SCALABLE: middle.block: ; TF-SCALABLE: middle.block:
; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]]
; TF-SCALABLE: scalar.ph: ; TF-SCALABLE: scalar.ph:
; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; TF-SCALABLE: for.body: ; TF-SCALABLE: for.body:
; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; TF-SCALABLE-NEXT: store i64 [[IV]], ptr [[B]], align 8 ; TF-SCALABLE-NEXT: store i64 [[IV]], ptr [[B]], align 8
▲ Show 20 Lines • Show All 170 Lines • ▼ Show 20 Lines
; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP15:![0-9]+]] ; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP15:![0-9]+]]
; FIXEDLEN: for.end: ; FIXEDLEN: for.end:
; FIXEDLEN-NEXT: ret void ; FIXEDLEN-NEXT: ret void
; ;
; TF-SCALABLE-LABEL: @conditional_uniform_store( ; TF-SCALABLE-LABEL: @conditional_uniform_store(
; TF-SCALABLE-NEXT: entry: ; TF-SCALABLE-NEXT: entry:
; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE-NEXT: [[TMP1:%.*]] = icmp ult i64 -1025, [[TMP0]]
; TF-SCALABLE-NEXT: br i1 [[TMP1]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE: vector.ph: ; TF-SCALABLE: vector.ph:
; TF-SCALABLE-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP4:%.*]] = sub i64 [[TMP3]], 1 ; TF-SCALABLE-NEXT: [[TMP2:%.*]] = sub i64 [[TMP1]], 1
; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP4]] ; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP2]]
; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP2]] ; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP0]]
; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]] ; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
; TF-SCALABLE-NEXT: [[TMP5:%.*]] = call <vscale x 1 x i64> @llvm.experimental.stepvector.nxv1i64() ; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call <vscale x 1 x i64> @llvm.experimental.stepvector.nxv1i64()
; TF-SCALABLE-NEXT: [[TMP6:%.*]] = add <vscale x 1 x i64> [[TMP5]], zeroinitializer ; TF-SCALABLE-NEXT: [[TMP4:%.*]] = add <vscale x 1 x i64> [[TMP3]], zeroinitializer
; TF-SCALABLE-NEXT: [[TMP7:%.*]] = mul <vscale x 1 x i64> [[TMP6]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 1, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer) ; TF-SCALABLE-NEXT: [[TMP5:%.*]] = mul <vscale x 1 x i64> [[TMP4]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 1, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer)
; TF-SCALABLE-NEXT: [[INDUCTION:%.*]] = add <vscale x 1 x i64> zeroinitializer, [[TMP7]] ; TF-SCALABLE-NEXT: [[INDUCTION:%.*]] = add <vscale x 1 x i64> zeroinitializer, [[TMP5]]
; TF-SCALABLE-NEXT: [[TMP8:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP9:%.*]] = mul i64 1, [[TMP8]] ; TF-SCALABLE-NEXT: [[TMP7:%.*]] = mul i64 1, [[TMP6]]
; TF-SCALABLE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP9]], i64 0 ; TF-SCALABLE-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[TMP7]], i64 0
; TF-SCALABLE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[DOTSPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[DOTSPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[DOTSPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 1 x ptr> poison, ptr [[B:%.*]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 1 x ptr> poison, ptr [[B:%.*]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 1 x ptr> [[BROADCAST_SPLATINSERT1]], <vscale x 1 x ptr> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 1 x ptr> [[BROADCAST_SPLATINSERT1]], <vscale x 1 x ptr> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; TF-SCALABLE: vector.body: ; TF-SCALABLE: vector.body:
; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[VEC_IND:%.*]] = phi <vscale x 1 x i64> [ [[INDUCTION]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[VEC_IND:%.*]] = phi <vscale x 1 x i64> [ [[INDUCTION]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[TMP10:%.*]] = add i64 [[INDEX]], 0 ; TF-SCALABLE-NEXT: [[TMP8:%.*]] = add i64 [[INDEX]], 0
; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP10]], i64 1024) ; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP8]], i64 1024)
; TF-SCALABLE-NEXT: [[TMP11:%.*]] = icmp ugt <vscale x 1 x i64> [[VEC_IND]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 10, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer) ; TF-SCALABLE-NEXT: [[TMP9:%.*]] = icmp ugt <vscale x 1 x i64> [[VEC_IND]], shufflevector (<vscale x 1 x i64> insertelement (<vscale x 1 x i64> poison, i64 10, i64 0), <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer)
; TF-SCALABLE-NEXT: [[TMP12:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP11]], <vscale x 1 x i1> zeroinitializer ; TF-SCALABLE-NEXT: [[TMP10:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP9]], <vscale x 1 x i1> zeroinitializer
; TF-SCALABLE-NEXT: call void @llvm.masked.scatter.nxv1i64.nxv1p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], <vscale x 1 x ptr> [[BROADCAST_SPLAT2]], i32 8, <vscale x 1 x i1> [[TMP12]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.scatter.nxv1i64.nxv1p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], <vscale x 1 x ptr> [[BROADCAST_SPLAT2]], i32 8, <vscale x 1 x i1> [[TMP10]])
; TF-SCALABLE-NEXT: [[TMP13:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP10]] ; TF-SCALABLE-NEXT: [[TMP11:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP8]]
; TF-SCALABLE-NEXT: [[TMP14:%.*]] = xor <vscale x 1 x i1> [[TMP11]], shufflevector (<vscale x 1 x i1> insertelement (<vscale x 1 x i1> poison, i1 true, i64 0), <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer) ; TF-SCALABLE-NEXT: [[TMP12:%.*]] = xor <vscale x 1 x i1> [[TMP9]], shufflevector (<vscale x 1 x i1> insertelement (<vscale x 1 x i1> poison, i1 true, i64 0), <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer)
; TF-SCALABLE-NEXT: [[TMP15:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP14]], <vscale x 1 x i1> zeroinitializer ; TF-SCALABLE-NEXT: [[TMP13:%.*]] = select <vscale x 1 x i1> [[ACTIVE_LANE_MASK]], <vscale x 1 x i1> [[TMP12]], <vscale x 1 x i1> zeroinitializer
; TF-SCALABLE-NEXT: [[TMP16:%.*]] = or <vscale x 1 x i1> [[TMP12]], [[TMP15]] ; TF-SCALABLE-NEXT: [[TMP14:%.*]] = or <vscale x 1 x i1> [[TMP10]], [[TMP13]]
; TF-SCALABLE-NEXT: [[TMP17:%.*]] = getelementptr inbounds i64, ptr [[TMP13]], i32 0 ; TF-SCALABLE-NEXT: [[TMP15:%.*]] = getelementptr inbounds i64, ptr [[TMP11]], i32 0
; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP17]], i32 8, <vscale x 1 x i1> [[TMP16]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP15]], i32 8, <vscale x 1 x i1> [[TMP14]])
; TF-SCALABLE-NEXT: [[TMP18:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP16:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP18]] ; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP16]]
; TF-SCALABLE-NEXT: [[VEC_IND_NEXT]] = add <vscale x 1 x i64> [[VEC_IND]], [[DOTSPLAT]] ; TF-SCALABLE-NEXT: [[VEC_IND_NEXT]] = add <vscale x 1 x i64> [[VEC_IND]], [[DOTSPLAT]]
; TF-SCALABLE-NEXT: [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] ; TF-SCALABLE-NEXT: [[TMP17:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; TF-SCALABLE-NEXT: br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP12:![0-9]+]]
; TF-SCALABLE: middle.block: ; TF-SCALABLE: middle.block:
; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]]
; TF-SCALABLE: scalar.ph: ; TF-SCALABLE: scalar.ph:
; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; TF-SCALABLE: for.body: ; TF-SCALABLE: for.body:
; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LATCH:%.*]] ] ; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LATCH:%.*]] ]
; TF-SCALABLE-NEXT: [[CMP:%.*]] = icmp ugt i64 [[IV]], 10 ; TF-SCALABLE-NEXT: [[CMP:%.*]] = icmp ugt i64 [[IV]], 10
▲ Show 20 Lines • Show All 154 Lines • ▼ Show 20 Lines
; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1 ; FIXEDLEN-NEXT: [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024 ; FIXEDLEN-NEXT: [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV_NEXT]], 1024
; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP17:![0-9]+]] ; FIXEDLEN-NEXT: br i1 [[EXITCOND_NOT]], label [[FOR_END]], label [[FOR_BODY]], !llvm.loop [[LOOP17:![0-9]+]]
; FIXEDLEN: for.end: ; FIXEDLEN: for.end:
; FIXEDLEN-NEXT: ret void ; FIXEDLEN-NEXT: ret void
; ;
; TF-SCALABLE-LABEL: @uniform_store_unaligned( ; TF-SCALABLE-LABEL: @uniform_store_unaligned(
; TF-SCALABLE-NEXT: entry: ; TF-SCALABLE-NEXT: entry:
; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE-NEXT: [[TMP1:%.*]] = icmp ult i64 -1025, [[TMP0]]
; TF-SCALABLE-NEXT: br i1 [[TMP1]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; TF-SCALABLE: vector.ph: ; TF-SCALABLE: vector.ph:
; TF-SCALABLE-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP0:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP3:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP1:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[TMP4:%.*]] = sub i64 [[TMP3]], 1 ; TF-SCALABLE-NEXT: [[TMP2:%.*]] = sub i64 [[TMP1]], 1
; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP4]] ; TF-SCALABLE-NEXT: [[N_RND_UP:%.*]] = add i64 1024, [[TMP2]]
; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP2]] ; TF-SCALABLE-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[N_RND_UP]], [[TMP0]]
; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]] ; TF-SCALABLE-NEXT: [[N_VEC:%.*]] = sub i64 [[N_RND_UP]], [[N_MOD_VF]]
; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0 ; TF-SCALABLE-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i64> poison, i64 [[V:%.*]], i64 0
; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer ; TF-SCALABLE-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i64> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[VECTOR_BODY:%.*]]
; TF-SCALABLE: vector.body: ; TF-SCALABLE: vector.body:
; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; TF-SCALABLE-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; TF-SCALABLE-NEXT: [[TMP5:%.*]] = add i64 [[INDEX]], 0 ; TF-SCALABLE-NEXT: [[TMP3:%.*]] = add i64 [[INDEX]], 0
; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP5]], i64 1024) ; TF-SCALABLE-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <vscale x 1 x i1> @llvm.get.active.lane.mask.nxv1i1.i64(i64 [[TMP3]], i64 1024)
; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B:%.*]], align 1 ; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B:%.*]], align 1
; TF-SCALABLE-NEXT: [[TMP6:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP5]] ; TF-SCALABLE-NEXT: [[TMP4:%.*]] = getelementptr inbounds i64, ptr [[A:%.*]], i64 [[TMP3]]
; TF-SCALABLE-NEXT: [[TMP7:%.*]] = getelementptr inbounds i64, ptr [[TMP6]], i32 0 ; TF-SCALABLE-NEXT: [[TMP5:%.*]] = getelementptr inbounds i64, ptr [[TMP4]], i32 0
; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP7]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]]) ; TF-SCALABLE-NEXT: call void @llvm.masked.store.nxv1i64.p0(<vscale x 1 x i64> [[BROADCAST_SPLAT]], ptr [[TMP5]], i32 8, <vscale x 1 x i1> [[ACTIVE_LANE_MASK]])
; TF-SCALABLE-NEXT: [[TMP8:%.*]] = call i64 @llvm.vscale.i64() ; TF-SCALABLE-NEXT: [[TMP6:%.*]] = call i64 @llvm.vscale.i64()
; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP8]] ; TF-SCALABLE-NEXT: [[INDEX_NEXT]] = add i64 [[INDEX]], [[TMP6]]
; TF-SCALABLE-NEXT: [[TMP9:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] ; TF-SCALABLE-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; TF-SCALABLE-NEXT: br i1 [[TMP9]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]] ; TF-SCALABLE-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]]
; TF-SCALABLE: middle.block: ; TF-SCALABLE: middle.block:
; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]] ; TF-SCALABLE-NEXT: br i1 true, label [[FOR_END:%.*]], label [[SCALAR_PH]]
; TF-SCALABLE: scalar.ph: ; TF-SCALABLE: scalar.ph:
; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ] ; TF-SCALABLE-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]] ; TF-SCALABLE-NEXT: br label [[FOR_BODY:%.*]]
; TF-SCALABLE: for.body: ; TF-SCALABLE: for.body:
; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ] ; TF-SCALABLE-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[FOR_BODY]] ]
; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B]], align 1 ; TF-SCALABLE-NEXT: store i64 [[V]], ptr [[B]], align 1
▲ Show 20 Lines • Show All 57 Lines • Show Last 20 Lines