[polly] r253995 - ScopInfo: Add support for delinearizing fortran arrays
Tobias Grosser via llvm-commits
llvm-commits at lists.llvm.org
Tue Nov 24 09:06:39 PST 2015
Author: grosser
Date: Tue Nov 24 11:06:38 2015
New Revision: 253995
URL: http://llvm.org/viewvc/llvm-project?rev=253995&view=rev
Log:
ScopInfo: Add support for delinearizing fortran arrays
gfortran (and fortran in general?) does not compute the address of an array
element directly from the array sizes (e.g., %s0, %s1), but takes first the
maximum of the sizes and 0 (e.g., max(0, %s0)) before multiplying the resulting
value with the per-dimension array subscript expressions. To successfully
delinearize index expressions as we see them in fortran, we first filter 'smax'
expressions out of the SCEV expression, use them to guess array size parameters
and only then continue with the existing delinearization.
Added:
polly/trunk/test/ScopInfo/multidim_fortran_2d.ll
polly/trunk/test/ScopInfo/multidim_fortran_2d_params.ll
polly/trunk/test/ScopInfo/multidim_fortran_srem.ll
Modified:
polly/trunk/lib/Analysis/ScopDetection.cpp
Modified: polly/trunk/lib/Analysis/ScopDetection.cpp
URL: http://llvm.org/viewvc/llvm-project/polly/trunk/lib/Analysis/ScopDetection.cpp?rev=253995&r1=253994&r2=253995&view=diff
==============================================================================
--- polly/trunk/lib/Analysis/ScopDetection.cpp (original)
+++ polly/trunk/lib/Analysis/ScopDetection.cpp Tue Nov 24 11:06:38 2015
@@ -487,11 +487,103 @@ bool ScopDetection::isInvariant(const Va
MapInsnToMemAcc InsnToMemAcc;
+/// @brief Remove smax of smax(0, size) expressions from a SCEV expression and
+/// register the '...' components.
+///
+/// Array access expressions as they are generated by gfortran contain smax(0,
+/// size) expressions that confuse the 'normal' delinearization algorithm.
+/// However, if we extract such expressions before the normal delinearization
+/// takes place they can actually help to identify array size expressions in
+/// fortran accesses. For the subsequently following delinearization the smax(0,
+/// size) component can be replaced by just 'size'. This is correct as we will
+/// always add and verify the assumption that for all subscript expressions
+/// 'exp' the inequality 0 <= exp < size holds. Hence, we will also verify
+/// that 0 <= size, which means smax(0, size) == size.
+struct SCEVRemoveMax : public SCEVVisitor<SCEVRemoveMax, const SCEV *> {
+public:
+ static const SCEV *remove(ScalarEvolution &SE, const SCEV *Expr,
+ std::vector<const SCEV *> *Terms = nullptr) {
+
+ SCEVRemoveMax D(SE, Terms);
+ return D.visit(Expr);
+ }
+
+ SCEVRemoveMax(ScalarEvolution &SE, std::vector<const SCEV *> *Terms)
+ : SE(SE), Terms(Terms) {}
+
+ const SCEV *visitTruncateExpr(const SCEVTruncateExpr *Expr) { return Expr; }
+
+ const SCEV *visitZeroExtendExpr(const SCEVZeroExtendExpr *Expr) {
+ return Expr;
+ }
+
+ const SCEV *visitSignExtendExpr(const SCEVSignExtendExpr *Expr) {
+ return SE.getSignExtendExpr(visit(Expr->getOperand()), Expr->getType());
+ }
+
+ const SCEV *visitUDivExpr(const SCEVUDivExpr *Expr) { return Expr; }
+
+ const SCEV *visitSMaxExpr(const SCEVSMaxExpr *Expr) {
+ if (Expr->getOperand(0)->isZero()) {
+ auto Res = visit(Expr->getOperand(1));
+ if (Terms)
+ (*Terms).push_back(Res);
+ return Res;
+ }
+
+ return Expr;
+ }
+
+ const SCEV *visitUMaxExpr(const SCEVUMaxExpr *Expr) { return visit(Expr); }
+
+ const SCEV *visitUnknown(const SCEVUnknown *Expr) { return Expr; }
+
+ const SCEV *visitCouldNotCompute(const SCEVCouldNotCompute *Expr) {
+ return Expr;
+ }
+
+ const SCEV *visitConstant(const SCEVConstant *Expr) { return Expr; }
+
+ const SCEV *visitAddRecExpr(const SCEVAddRecExpr *Expr) {
+ SmallVector<const SCEV *, 5> NewOps;
+ for (const SCEV *Op : Expr->operands())
+ NewOps.push_back(visit(Op));
+
+ return SE.getAddRecExpr(NewOps, Expr->getLoop(), Expr->getNoWrapFlags());
+ }
+
+ const SCEV *visitAddExpr(const SCEVAddExpr *Expr) {
+ SmallVector<const SCEV *, 5> NewOps;
+ for (const SCEV *Op : Expr->operands())
+ NewOps.push_back(visit(Op));
+
+ return SE.getAddExpr(NewOps);
+ }
+
+ const SCEV *visitMulExpr(const SCEVMulExpr *Expr) {
+ SmallVector<const SCEV *, 5> NewOps;
+ for (const SCEV *Op : Expr->operands())
+ NewOps.push_back(visit(Op));
+
+ return SE.getMulExpr(NewOps);
+ }
+
+private:
+ ScalarEvolution &SE;
+ std::vector<const SCEV *> *Terms;
+};
+
SmallVector<const SCEV *, 4>
ScopDetection::getDelinearizationTerms(DetectionContext &Context,
const SCEVUnknown *BasePointer) const {
SmallVector<const SCEV *, 4> Terms;
for (const auto &Pair : Context.Accesses[BasePointer]) {
+ std::vector<const SCEV *> MaxTerms;
+ SCEVRemoveMax::remove(*SE, Pair.second, &MaxTerms);
+ if (MaxTerms.size() > 0) {
+ Terms.insert(Terms.begin(), MaxTerms.begin(), MaxTerms.end());
+ continue;
+ }
// In case the outermost expression is a plain add, we check if any of its
// terms has the form 4 * %inst * %param * %param ..., aka a term that
// contains a product between a parameter and an instruction that is
@@ -592,6 +684,7 @@ bool ScopDetection::computeAccessFunctio
for (const auto &Pair : Context.Accesses[BasePointer]) {
const Instruction *Insn = Pair.first;
auto *AF = Pair.second;
+ AF = SCEVRemoveMax::remove(*SE, AF);
bool IsNonAffine = false;
TempMemoryAccesses.insert(std::make_pair(Insn, MemAcc(Insn, Shape)));
MemAcc *Acc = &TempMemoryAccesses.find(Insn)->second;
Added: polly/trunk/test/ScopInfo/multidim_fortran_2d.ll
URL: http://llvm.org/viewvc/llvm-project/polly/trunk/test/ScopInfo/multidim_fortran_2d.ll?rev=253995&view=auto
==============================================================================
--- polly/trunk/test/ScopInfo/multidim_fortran_2d.ll (added)
+++ polly/trunk/test/ScopInfo/multidim_fortran_2d.ll Tue Nov 24 11:06:38 2015
@@ -0,0 +1,76 @@
+; RUN: opt %loadPolly -polly-scops -analyze < %s | FileCheck %s
+
+; subroutine init_array(ni, nj, pi, pj, a)
+; implicit none
+;
+; double precision, dimension(nj, ni) :: a
+; integer*8 :: ni, nj
+; integer*8 :: pi, pj
+; integer*8 :: i, j
+;
+; do i = 1, ni
+; do j = 1, nj
+; a(j, i) = i + j
+; end do
+; end do
+; end subroutine
+;
+; Verify we correctly delinearize accesses
+
+; CHECK: MustWriteAccess := [Reduction Type: NONE] [Scalar: 0]
+; CHECK: [tmp9, tmp14] -> { Stmt_bb17[i0, i1] -> MemRef_arg4[i0, i1] };
+
+target datalayout = "e-p:64:64:64-S128-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f16:16:16-f32:32:32-f64:64:64-f128:128:128-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64"
+
+define void @ham(i64* noalias %arg, i64* noalias %arg1, i64* noalias %arg2, i64* noalias %arg3, [0 x double]* noalias %arg4) unnamed_addr {
+bb:
+ br label %bb5
+
+bb5: ; preds = %bb
+ %tmp = load i64, i64* %arg1, align 8
+ %tmp6 = icmp slt i64 %tmp, 0
+ %tmp7 = select i1 %tmp6, i64 0, i64 %tmp
+ %tmp8 = xor i64 %tmp7, -1
+ %tmp9 = load i64, i64* %arg, align 8
+ %tmp10 = icmp sgt i64 %tmp9, 0
+ br i1 %tmp10, label %bb11, label %bb32
+
+bb11: ; preds = %bb5
+ br label %bb12
+
+bb12: ; preds = %bb28, %bb11
+ %tmp13 = phi i64 [ %tmp30, %bb28 ], [ 1, %bb11 ]
+ %tmp14 = load i64, i64* %arg1, align 8
+ %tmp15 = icmp sgt i64 %tmp14, 0
+ br i1 %tmp15, label %bb16, label %bb28
+
+bb16: ; preds = %bb12
+ br label %bb17
+
+bb17: ; preds = %bb17, %bb16
+ %tmp18 = phi i64 [ %tmp26, %bb17 ], [ 1, %bb16 ]
+ %tmp19 = mul i64 %tmp13, %tmp7
+ %tmp20 = add i64 %tmp19, %tmp8
+ %tmp21 = add i64 %tmp20, %tmp18
+ %tmp22 = add i64 %tmp18, %tmp13
+ %tmp23 = sitofp i64 %tmp22 to double
+ %tmp24 = getelementptr [0 x double], [0 x double]* %arg4, i64 0, i64 %tmp21
+ store double %tmp23, double* %tmp24, align 8
+ %tmp25 = icmp eq i64 %tmp18, %tmp14
+ %tmp26 = add i64 %tmp18, 1
+ br i1 %tmp25, label %bb27, label %bb17
+
+bb27: ; preds = %bb17
+ br label %bb28
+
+bb28: ; preds = %bb27, %bb12
+ %tmp29 = icmp eq i64 %tmp13, %tmp9
+ %tmp30 = add i64 %tmp13, 1
+ br i1 %tmp29, label %bb31, label %bb12
+
+bb31: ; preds = %bb28
+ br label %bb32
+
+bb32: ; preds = %bb31, %bb5
+ ret void
+}
Added: polly/trunk/test/ScopInfo/multidim_fortran_2d_params.ll
URL: http://llvm.org/viewvc/llvm-project/polly/trunk/test/ScopInfo/multidim_fortran_2d_params.ll?rev=253995&view=auto
==============================================================================
--- polly/trunk/test/ScopInfo/multidim_fortran_2d_params.ll (added)
+++ polly/trunk/test/ScopInfo/multidim_fortran_2d_params.ll Tue Nov 24 11:06:38 2015
@@ -0,0 +1,84 @@
+; RUN: opt %loadPolly -polly-scops -analyze < %s | FileCheck %s
+
+; subroutine init_array(ni, nj, pi, pj, a)
+; implicit none
+
+; double precision, dimension(nj, ni) :: a
+; integer*8 :: ni, nj
+; integer*8 :: pi, pj
+; integer*8 :: i, j
+
+; do i = 1, ni
+; do j = 1, nj
+; a(j + pi, i + pj) = i + j
+; end do
+; end do
+; end subroutine
+
+; CHECK: [tmp9, nj_loaded2, tmp20, tmp19] -> { Stmt_bb17[i0, i1] -> MemRef_a[-1 + tmp20 + i0, nj_loaded2 + tmp19 + i1] : i1 <= -1 - tmp19; Stmt_bb17[i0, i1] -> MemRef_a[tmp20 + i0, tmp19 + i1] : i1 >= -tmp19 };
+
+
+target datalayout = "e-p:64:64:64-S128-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f16:16:16-f32:32:32-f64:64:64-f128:128:128-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64"
+target triple = "x86_64-unknown-linux-gnu"
+
+module asm "\09.ident\09\22GCC: (GNU) 4.6.4 LLVM: 3.3.1\22"
+
+; Function Attrs: nounwind uwtable
+define void @blam(i64* noalias %arg, i64* noalias %nj, i64* noalias %arg2, i64* noalias %arg3, [0 x double]* noalias %a) unnamed_addr #0 {
+bb:
+ br label %bb5
+
+bb5: ; preds = %bb
+ %nj_loaded = load i64, i64* %nj, align 8
+ %tmp6 = icmp slt i64 %nj_loaded, 0
+ %tmp7 = select i1 %tmp6, i64 0, i64 %nj_loaded
+ %tmp8 = xor i64 %tmp7, -1
+ %tmp9 = load i64, i64* %arg, align 8
+ %tmp10 = icmp sgt i64 %tmp9, 0
+ br i1 %tmp10, label %bb11, label %bb36
+
+bb11: ; preds = %bb5
+ br label %bb12
+
+bb12: ; preds = %bb32, %bb11
+ %tmp13 = phi i64 [ %tmp34, %bb32 ], [ 1, %bb11 ]
+ %nj_loaded2 = load i64, i64* %nj, align 8
+ %tmp15 = icmp sgt i64 %nj_loaded2, 0
+ br i1 %tmp15, label %bb16, label %bb32
+
+bb16: ; preds = %bb12
+ br label %bb17
+
+bb17: ; preds = %bb17, %bb16
+ %tmp18 = phi i64 [ %tmp30, %bb17 ], [ 1, %bb16 ]
+ %tmp19 = load i64, i64* %arg2, align 8
+ %tmp20 = load i64, i64* %arg3, align 8
+ %tmp21 = add i64 %tmp20, %tmp13
+ %tmp22 = mul i64 %tmp21, %tmp7
+ %tmp23 = add i64 %tmp18, %tmp8
+ %tmp24 = add i64 %tmp23, %tmp19
+ %tmp25 = add i64 %tmp24, %tmp22
+ %tmp26 = add i64 %tmp18, %tmp13
+ %tmp27 = sitofp i64 %tmp26 to double
+ %tmp28 = getelementptr [0 x double], [0 x double]* %a, i64 0, i64 %tmp25
+ store double %tmp27, double* %tmp28, align 8
+ %tmp29 = icmp eq i64 %tmp18, %nj_loaded2
+ %tmp30 = add i64 %tmp18, 1
+ br i1 %tmp29, label %bb31, label %bb17
+
+bb31: ; preds = %bb17
+ br label %bb32
+
+bb32: ; preds = %bb31, %bb12
+ %tmp33 = icmp eq i64 %tmp13, %tmp9
+ %tmp34 = add i64 %tmp13, 1
+ br i1 %tmp33, label %bb35, label %bb12
+
+bb35: ; preds = %bb32
+ br label %bb36
+
+bb36: ; preds = %bb35, %bb5
+ ret void
+}
+
+attributes #0 = { nounwind uwtable }
Added: polly/trunk/test/ScopInfo/multidim_fortran_srem.ll
URL: http://llvm.org/viewvc/llvm-project/polly/trunk/test/ScopInfo/multidim_fortran_srem.ll?rev=253995&view=auto
==============================================================================
--- polly/trunk/test/ScopInfo/multidim_fortran_srem.ll (added)
+++ polly/trunk/test/ScopInfo/multidim_fortran_srem.ll Tue Nov 24 11:06:38 2015
@@ -0,0 +1,169 @@
+; RUN: opt %loadPolly -polly-scops -analyze < %s | FileCheck %s
+target datalayout = "e-p:64:64:64-S128-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f16:16:16-f32:32:32-f64:64:64-f128:128:128-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64"
+
+; CHECK: ReadAccess := [Reduction Type: NONE] [Scalar: 0]
+; CHECK: [tmp180, tmp177, tmp183, tmp162, tmp157, tmp150, tmp146, tmp140, tmp] -> { Stmt_bb203[i0, i1, i2] -> MemRef_tmp173[o0, 1 + i1, 1 + i2] : exists (e0 = floor((-i0 + o0)/3): 3e0 = -i0 + o0 and o0 <= 2 and o0 >= 0) };
+; CHECK: ReadAccess := [Reduction Type: NONE] [Scalar: 0]
+; CHECK: [tmp180, tmp177, tmp183, tmp162, tmp157, tmp150, tmp146, tmp140, tmp] -> { Stmt_bb203[i0, i1, i2] -> MemRef_tmp173[o0, 1 + i1, 1 + i2] : exists (e0 = floor((-2 - i0 + o0)/3): 3e0 = -2 - i0 + o0 and o0 <= 2 and o0 >= 0) };
+
+
+define void @pluto(i32* noalias %arg, [0 x i32]* noalias %arg2, [0 x i32]* noalias %arg3, [0 x i32]* noalias %arg4, [0 x i32]* noalias %arg5, [0 x i32]* noalias %arg6, [0 x i32]* noalias %arg7, [0 x i32]* noalias %arg8, [0 x i32]* noalias %arg9, double* noalias %arg10, double* noalias %arg11, [0 x double]* noalias %arg12, [0 x double]* noalias %arg13, [0 x i32]* noalias %arg14, i32* noalias %arg15, [0 x i32]* noalias %arg16, i32* noalias %arg17, i32* noalias %arg18, i32* noalias %arg19, i32* noalias %arg20, i32* noalias %arg21, i32* noalias %arg22, i32* noalias %arg23, i32* noalias %arg24, i32* noalias %arg25, i32* noalias %arg26, i32* noalias %arg27, [0 x double]* noalias %arg28, [0 x double]* noalias %arg29, [0 x double]* noalias %arg30, [0 x double]* noalias %arg31, [0 x double]* noalias %arg32, [0 x double]* noalias %arg33, [0 x double]* noalias %arg34, [0 x double]* noalias %arg35, [0 x double]* noalias %arg36, [0 x double]* noalias %arg37, [0 x double]* noalias %arg38, [0 x do
uble]* noalias %arg39, [0 x double]* noalias %arg40, [0 x double]* noalias %arg41, [0 x double]* noalias %arg42, [0 x double]* noalias %arg43, [0 x double]* noalias %arg44, [0 x double]* noalias %arg45, [0 x double]* noalias %arg46, [0 x double]* noalias %arg47, [0 x double]* noalias %arg48, [0 x double]* noalias %arg49, [0 x double]* noalias %arg50, [0 x double]* noalias %arg51, [0 x double]* noalias %arg52, [0 x double]* noalias %arg53, [0 x double]* noalias %arg54, [0 x double]* noalias %arg55, [0 x double]* noalias %arg56, [0 x double]* noalias %arg57, [0 x double]* noalias %arg58, [0 x double]* noalias %arg59, [0 x double]* noalias %arg60, [0 x double]* noalias %arg61, [0 x double]* noalias %arg62, [0 x double]* noalias %arg63, [0 x double]* noalias %arg64, [0 x double]* noalias %arg65, [0 x double]* noalias %arg66, [0 x double]* noalias %arg67, [0 x double]* noalias %arg68, [0 x double]* noalias %arg69, i32* noalias %arg70, i32* noalias %arg71, i32* noalias %arg72, i32* noalia
s %arg73, i32* noalias %arg74, i32* noalias %arg75, i32* noalias %arg76, i32* noalias %arg77, i32* noalias %arg78, i32* noalias %arg79, i32* noalias %arg80, i32* noalias %arg81, i32* noalias %arg82, i32* noalias %arg83, i32* noalias %arg84, i32* noalias %arg85, i32* noalias %arg86, i32* noalias %arg87, i32* noalias %arg88, i32* noalias %arg89, i32* noalias %arg90, i32* noalias %arg91, i32* noalias %arg92, i32* noalias %arg93, i32* noalias %arg94, i32* noalias %arg95, i32* noalias %arg96, i32* noalias %arg97, [0 x double]* noalias %arg98, [0 x double]* noalias %arg99, [0 x double]* noalias %arg100, [0 x double]* noalias %arg101, double* noalias %arg102, double* noalias %arg103, double* noalias %arg104, i32* noalias %arg105, double* noalias %arg106, double* noalias %arg107, [0 x double]* noalias %arg108, [0 x double]* noalias %arg109, [0 x double]* noalias %arg110, [0 x double]* noalias %arg111, [0 x double]* noalias %arg112, [0 x double]* noalias %arg113, [0 x double]* noalias %arg11
4, [0 x double]* noalias %arg115, [0 x double]* noalias %arg116, [0 x double]* noalias %arg117, [0 x double]* noalias %arg118, [0 x double]* noalias %arg119, [0 x double]* noalias %arg120, [0 x double]* noalias %arg121, [0 x double]* noalias %arg122, [0 x double]* noalias %arg123, [0 x double]* noalias %arg124, [0 x double]* noalias %arg125, [0 x double]* noalias %arg126, [0 x double]* noalias %arg127, [0 x double]* noalias %arg128, [0 x double]* noalias %arg129, [0 x double]* noalias %arg130, [0 x double]* noalias %arg131, i32* noalias %arg132, [0 x double]* noalias %arg133, [0 x double]* noalias %arg134, [0 x double]* noalias %arg135) {
+bb:
+ br label %bb136
+
+bb136: ; preds = %bb
+ %tmp = load i32, i32* %arg19, align 4
+ %tmp137 = sext i32 %tmp to i64
+ %tmp138 = icmp slt i64 %tmp137, 0
+ %tmp139 = select i1 %tmp138, i64 0, i64 %tmp137
+ %tmp140 = load i32, i32* %arg20, align 4
+ %tmp141 = sext i32 %tmp140 to i64
+ %tmp142 = mul nsw i64 %tmp139, %tmp141
+ %tmp143 = icmp slt i64 %tmp142, 0
+ %tmp144 = select i1 %tmp143, i64 0, i64 %tmp142
+ %tmp145 = xor i64 %tmp139, -1
+ %tmp146 = load i32, i32* %arg19, align 4
+ %tmp147 = sext i32 %tmp146 to i64
+ %tmp148 = icmp slt i64 %tmp147, 0
+ %tmp149 = select i1 %tmp148, i64 0, i64 %tmp147
+ %tmp150 = load i32, i32* %arg20, align 4
+ %tmp151 = sext i32 %tmp150 to i64
+ %tmp152 = mul nsw i64 %tmp149, %tmp151
+ %tmp153 = icmp slt i64 %tmp152, 0
+ %tmp154 = select i1 %tmp153, i64 0, i64 %tmp152
+ %tmp155 = xor i64 %tmp149, -1
+ %tmp156 = getelementptr inbounds [0 x i32], [0 x i32]* %arg3, i64 0, i64 0
+ %tmp157 = load i32, i32* %tmp156, align 4
+ %tmp158 = sext i32 %tmp157 to i64
+ %tmp159 = icmp slt i64 %tmp158, 0
+ %tmp160 = select i1 %tmp159, i64 0, i64 %tmp158
+ %tmp161 = getelementptr [0 x i32], [0 x i32]* %arg3, i64 0, i64 1
+ %tmp162 = load i32, i32* %tmp161, align 4
+ %tmp163 = sext i32 %tmp162 to i64
+ %tmp164 = mul nsw i64 %tmp160, %tmp163
+ %tmp165 = icmp slt i64 %tmp164, 0
+ %tmp166 = select i1 %tmp165, i64 0, i64 %tmp164
+ %tmp167 = mul i64 %tmp166, 3
+ %tmp168 = icmp slt i64 %tmp167, 0
+ %tmp169 = select i1 %tmp168, i64 0, i64 %tmp167
+ %tmp170 = shl i64 %tmp169, 3
+ %tmp171 = icmp ne i64 %tmp170, 0
+ %tmp172 = select i1 %tmp171, i64 %tmp170, i64 1
+ %tmp173 = tail call noalias i8* @wobble(i64 %tmp172) #1
+ %tmp174 = xor i64 %tmp160, -1
+ %tmp175 = sub i64 %tmp174, %tmp166
+ %tmp176 = getelementptr inbounds [0 x i32], [0 x i32]* %arg3, i64 0, i64 0
+ %tmp177 = load i32, i32* %tmp176, align 4
+ %tmp178 = sext i32 %tmp177 to i64
+ %tmp179 = getelementptr [0 x i32], [0 x i32]* %arg3, i64 0, i64 1
+ %tmp180 = load i32, i32* %tmp179, align 4
+ %tmp181 = sext i32 %tmp180 to i64
+ %tmp182 = getelementptr [0 x i32], [0 x i32]* %arg3, i64 0, i64 2
+ %tmp183 = load i32, i32* %tmp182, align 4
+ %tmp184 = sext i32 %tmp183 to i64
+ %tmp185 = add nsw i64 %tmp184, -1
+ %tmp186 = icmp sgt i64 %tmp185, 1
+ br i1 %tmp186, label %bb187, label %bb249
+
+bb187: ; preds = %bb136
+ br label %bb188
+
+bb188: ; preds = %bb187, %bb245
+ %tmp189 = phi i64 [ %tmp247, %bb245 ], [ 2, %bb187 ]
+ %tmp190 = add i64 %tmp189, -2
+ %tmp191 = srem i64 %tmp190, 3
+ %tmp192 = add nsw i64 %tmp191, 1
+ %tmp193 = srem i64 %tmp189, 3
+ %tmp194 = add nsw i64 %tmp193, 1
+ %tmp195 = add nsw i64 %tmp181, -1
+ %tmp196 = icmp sgt i64 %tmp195, 1
+ br i1 %tmp196, label %bb197, label %bb245
+
+bb197: ; preds = %bb188
+ br label %bb198
+
+bb198: ; preds = %bb197, %bb241
+ %tmp199 = phi i64 [ %tmp243, %bb241 ], [ 2, %bb197 ]
+ %tmp200 = add nsw i64 %tmp178, -1
+ %tmp201 = icmp sgt i64 %tmp200, 1
+ br i1 %tmp201, label %bb202, label %bb241
+
+bb202: ; preds = %bb198
+ br label %bb203
+
+bb203: ; preds = %bb202, %bb203
+ %tmp204 = phi i64 [ %tmp239, %bb203 ], [ 2, %bb202 ]
+ %tmp205 = mul i64 %tmp199, %tmp160
+ %tmp206 = mul i64 %tmp192, %tmp166
+ %tmp207 = add i64 %tmp206, %tmp175
+ %tmp208 = add i64 %tmp207, %tmp205
+ %tmp209 = add i64 %tmp208, %tmp204
+ %tmp210 = bitcast i8* %tmp173 to double*
+ %tmp211 = getelementptr double, double* %tmp210, i64 %tmp209
+ %tmp212 = load double, double* %tmp211, align 8
+ %tmp213 = mul i64 %tmp199, %tmp160
+ %tmp214 = mul i64 %tmp194, %tmp166
+ %tmp215 = add i64 %tmp214, %tmp175
+ %tmp216 = add i64 %tmp215, %tmp213
+ %tmp217 = add i64 %tmp216, %tmp204
+ %tmp218 = bitcast i8* %tmp173 to double*
+ %tmp219 = getelementptr double, double* %tmp218, i64 %tmp217
+ %tmp220 = load double, double* %tmp219, align 8
+ %tmp221 = fadd double %tmp212, %tmp220
+ %tmp222 = mul i64 %tmp199, %tmp139
+ %tmp223 = mul i64 %tmp189, %tmp144
+ %tmp224 = sub i64 %tmp145, %tmp144
+ %tmp225 = add i64 %tmp224, %tmp223
+ %tmp226 = add i64 %tmp225, %tmp222
+ %tmp227 = add i64 %tmp226, %tmp204
+ %tmp228 = mul i64 %tmp199, %tmp149
+ %tmp229 = mul i64 %tmp189, %tmp154
+ %tmp230 = sub i64 %tmp155, %tmp154
+ %tmp231 = add i64 %tmp230, %tmp229
+ %tmp232 = add i64 %tmp231, %tmp228
+ %tmp233 = add i64 %tmp232, %tmp204
+ %tmp234 = getelementptr [0 x double], [0 x double]* %arg56, i64 0, i64 %tmp233
+ %tmp235 = load double, double* %tmp234, align 8
+ %tmp236 = fadd double %tmp235, %tmp221
+ %tmp237 = getelementptr [0 x double], [0 x double]* %arg55, i64 0, i64 %tmp227
+ store double %tmp236, double* %tmp237, align 8
+ %tmp238 = icmp eq i64 %tmp204, %tmp200
+ %tmp239 = add i64 %tmp204, 1
+ br i1 %tmp238, label %bb240, label %bb203
+
+bb240: ; preds = %bb203
+ br label %bb241
+
+bb241: ; preds = %bb240, %bb198
+ %tmp242 = icmp eq i64 %tmp199, %tmp195
+ %tmp243 = add i64 %tmp199, 1
+ br i1 %tmp242, label %bb244, label %bb198
+
+bb244: ; preds = %bb241
+ br label %bb245
+
+bb245: ; preds = %bb244, %bb188
+ %tmp246 = icmp eq i64 %tmp189, %tmp185
+ %tmp247 = add i64 %tmp189, 1
+ br i1 %tmp246, label %bb248, label %bb188
+
+bb248: ; preds = %bb245
+ br label %bb249
+
+bb249: ; preds = %bb248, %bb136
+ %tmp250 = icmp eq i8* %tmp173, null
+ br i1 %tmp250, label %bb252, label %bb251
+
+bb251: ; preds = %bb249
+ tail call void @snork(i8* %tmp173) #1
+ br label %bb252
+
+bb252: ; preds = %bb251, %bb249
+ ret void
+}
+
+; Function Attrs: nounwind
+declare noalias i8* @wobble(i64) #1
+
+; Function Attrs: nounwind
+declare void @snork(i8*) #1
More information about the llvm-commits
mailing list