Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix calling convention and other small bugs #16

Merged
merged 12 commits into from
Oct 15, 2019
Merged

Fix calling convention and other small bugs #16

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
d551bb1
Fixing code for call gradient
wsmoses Oct 10, 2019
1a613ad
fixing vector bug
wsmoses Oct 10, 2019
3741228
Fix various calling convention issues
wsmoses Oct 12, 2019
aa4f76d
further calling convention fixes
wsmoses Oct 12, 2019
a025080
lookat ifcrash
wsmoses Oct 12, 2019
098aa5b
resolve fn tests
wsmoses Oct 12, 2019
60820fd
look at weird fail
wsmoses Oct 12, 2019
d04d3e4
fix subtle if bug
wsmoses Oct 12, 2019
3412e9c
fix julia
wsmoses Oct 12, 2019
1498c0e
update tests
wsmoses Oct 12, 2019
a8bdf3b
revamp loop calc
wsmoses Oct 12, 2019
ae6bee0
add more caching layers to speed compilation
wsmoses Oct 12, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
revamp loop calc
  • Loading branch information
@wsmoses
wsmoses committed Oct 12, 2019
commit a8bdf3b1b5dce14911f871bf10c6060033b8baf9
57 changes: 33 additions & 24 deletions enzyme/Enzyme/GradientUtils.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -504,7 +504,7 @@ std::pair<PHINode*,Value*> insertNewCanonicalIV(Loop* L, Type* Ty) {
return std::pair<PHINode*,Value*>(CanonicalIV,inc);
}

void removeRedundantIVs(BasicBlock* Header, BasicBlock* Preheader, PHINode* CanonicalIV, ScalarEvolution &SE, GradientUtils &gutils, Value* increment=nullptr) {
void removeRedundantIVs(const Loop* L, BasicBlock* Header, BasicBlock* Preheader, PHINode* CanonicalIV, ScalarEvolution &SE, GradientUtils &gutils, Value* increment, const SmallVectorImpl<BasicBlock*>&& latches) {
assert(Header);
assert(CanonicalIV);

Expand Down Expand Up @@ -535,26 +535,30 @@ void removeRedundantIVs(BasicBlock* Header, BasicBlock* Preheader, PHINode* Cano
gutils.erase(PN);
}

if (latches.size() == 1 && isa<BranchInst>(latches[0]->getTerminator()) && cast<BranchInst>(latches[0]->getTerminator())->isConditional())
for (auto use : CanonicalIV->users()) {
if (auto cmp = dyn_cast<ICmpInst>(use)) {
if (cmp->isUnsigned()) {
// Force i to be on LHS
if (cmp->getOperand(0) != CanonicalIV) {
//Below also swaps predicate correctly
cmp->swapOperands();
}
assert(cmp->getOperand(0) == CanonicalIV);
if (cast<BranchInst>(latches[0]->getTerminator())->getCondition() != cmp) continue;
// Force i to be on LHS
if (cmp->getOperand(0) != CanonicalIV) {
//Below also swaps predicate correctly
cmp->swapOperands();
}
assert(cmp->getOperand(0) == CanonicalIV);

auto scv = SE.getSCEVAtScope(cmp->getOperand(1), L);
if (cmp->isUnsigned() || (scv != SE.getCouldNotCompute() && SE.isKnownNonNegative(scv)) ) {

// valid replacements (since unsigned comparison and i starts at 0 counting up)

// * i < n => i != n, valid since first time i >= n occurs at i == n
if (cmp->getPredicate() == ICmpInst::ICMP_ULT) {
if (cmp->getPredicate() == ICmpInst::ICMP_ULT || cmp->getPredicate() == ICmpInst::ICMP_SLT) {
cmp->setPredicate(ICmpInst::ICMP_NE);
goto cend;
}

// * i <= n => i != n+1, valid since first time i > n occurs at i == n+1 [ which we assert is in bitrange as not infinite loop ]
if (cmp->getPredicate() == ICmpInst::ICMP_ULE) {
if (cmp->getPredicate() == ICmpInst::ICMP_ULE || cmp->getPredicate() == ICmpInst::ICMP_SLE) {
IRBuilder <>builder (Preheader->getTerminator());
if (auto inst = dyn_cast<Instruction>(cmp->getOperand(1))) {
builder.SetInsertPoint(inst->getNextNode());
Expand All @@ -565,13 +569,13 @@ void removeRedundantIVs(BasicBlock* Header, BasicBlock* Preheader, PHINode* Cano
}

// * i >= n => i == n, valid since first time i >= n occurs at i == n
if (cmp->getPredicate() == ICmpInst::ICMP_UGE) {
if (cmp->getPredicate() == ICmpInst::ICMP_UGE || cmp->getPredicate() == ICmpInst::ICMP_SGE) {
cmp->setPredicate(ICmpInst::ICMP_EQ);
goto cend;
}

// * i > n => i == n+1, valid since first time i > n occurs at i == n+1 [ which we assert is in bitrange as not infinite loop ]
if (cmp->getPredicate() == ICmpInst::ICMP_UGT) {
if (cmp->getPredicate() == ICmpInst::ICMP_UGT || cmp->getPredicate() == ICmpInst::ICMP_SGT) {
IRBuilder <>builder (Preheader->getTerminator());
if (auto inst = dyn_cast<Instruction>(cmp->getOperand(1))) {
builder.SetInsertPoint(inst->getNextNode());
Expand Down Expand Up @@ -618,39 +622,44 @@ void removeRedundantIVs(BasicBlock* Header, BasicBlock* Preheader, PHINode* Cano
gutils.erase(inst);
}

if (latches.size() == 1 && isa<BranchInst>(latches[0]->getTerminator()) && cast<BranchInst>(latches[0]->getTerminator())->isConditional())
for (auto use : increment->users()) {
if (auto cmp = dyn_cast<ICmpInst>(use)) {
if (cmp->isUnsigned()) {
// Force i+1 to be on LHS
if (cmp->getOperand(0) != increment) {
//Below also swaps predicate correctly
cmp->swapOperands();
}
assert(cmp->getOperand(0) == increment);
if (cast<BranchInst>(latches[0]->getTerminator())->getCondition() != cmp) continue;

// Force i+1 to be on LHS
if (cmp->getOperand(0) != increment) {
//Below also swaps predicate correctly
cmp->swapOperands();
}
assert(cmp->getOperand(0) == increment);

auto scv = SE.getSCEVAtScope(cmp->getOperand(1), L);
if (cmp->isUnsigned() || (scv != SE.getCouldNotCompute() && SE.isKnownNonNegative(scv)) ) {

// valid replacements (since unsigned comparison and i starts at 0 counting up)

// * i+1 < n => i+1 != n, valid since first time i+1 >= n occurs at i+1 == n
if (cmp->getPredicate() == ICmpInst::ICMP_ULT) {
if (cmp->getPredicate() == ICmpInst::ICMP_ULT || cmp->getPredicate() == ICmpInst::ICMP_SLT) {
cmp->setPredicate(ICmpInst::ICMP_NE);
continue;
}

// * i+1 <= n => i != n, valid since first time i+1 > n occurs at i+1 == n+1 => i == n
if (cmp->getPredicate() == ICmpInst::ICMP_ULE) {
if (cmp->getPredicate() == ICmpInst::ICMP_ULE || cmp->getPredicate() == ICmpInst::ICMP_SLE) {
cmp->setOperand(0, CanonicalIV);
cmp->setPredicate(ICmpInst::ICMP_NE);
continue;
}

// * i+1 >= n => i+1 == n, valid since first time i+1 >= n occurs at i+1 == n
if (cmp->getPredicate() == ICmpInst::ICMP_UGE) {
if (cmp->getPredicate() == ICmpInst::ICMP_UGE || cmp->getPredicate() == ICmpInst::ICMP_SGE) {
cmp->setPredicate(ICmpInst::ICMP_EQ);
continue;
}

// * i+1 > n => i == n, valid since first time i+1 > n occurs at i+1 == n+1 => i == n
if (cmp->getPredicate() == ICmpInst::ICMP_UGT) {
if (cmp->getPredicate() == ICmpInst::ICMP_UGT || cmp->getPredicate() == ICmpInst::ICMP_SGT) {
cmp->setOperand(0, CanonicalIV);
cmp->setPredicate(ICmpInst::ICMP_EQ);
continue;
Expand Down Expand Up @@ -686,7 +695,7 @@ bool getContextM(BasicBlock *BB, LoopContext &loopContext, std::map<Loop*,LoopCo
auto pair = insertNewCanonicalIV(L, Type::getInt64Ty(BB->getContext()));
PHINode* CanonicalIV = pair.first;
assert(CanonicalIV);
removeRedundantIVs(loopContexts[L].header, loopContexts[L].preheader, CanonicalIV, SE, gutils, pair.second);
removeRedundantIVs(L, loopContexts[L].header, loopContexts[L].preheader, CanonicalIV, SE, gutils, pair.second, fake::SCEVExpander::getLatches(L, loopContexts[L].exitBlocks));
loopContexts[L].var = CanonicalIV;
loopContexts[L].antivar = PHINode::Create(CanonicalIV->getType(), CanonicalIV->getNumIncomingValues(), CanonicalIV->getName()+"'phi");

Expand Down
10 changes: 5 additions & 5 deletions enzyme/test/Enzyme/llist.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -172,17 +172,17 @@ attributes #4 = { nounwind }
; CHECK-NEXT: %"'ipl" = load %struct.n*, %struct.n** %"next'ipg", align 8
; CHECK-NEXT: %[[loadst]] = load %struct.n*, %struct.n** %next, align 8, !tbaa !8
; CHECK-NEXT: %cmp = icmp eq %struct.n* %[[loadst]], null
; CHECK-NEXT: br i1 %cmp, label %invertfor.body, label %for.body
; CHECK-NEXT: br i1 %cmp, label %[[antiloop:.+]], label %for.body

; CHECK: invertentry:
; CHECK-NEXT: ret {} undef

; CHECK: invertfor.body.preheader: ; preds = %invertfor.body
; CHECK: invertfor.body.preheader:
; CHECK-NEXT: tail call void @free(i8* nonnull %_realloccache)
; CHECK-NEXT: br label %invertentry

; CHECK: invertfor.body:
; CHECK-NEXT: %[[antivar:.+]] = phi i64 [ %[[subidx:.+]], %invertfor.body ], [ %[[preidx]], %for.body ]
; CHECK: [[antiloop]]:
; CHECK-NEXT: %[[antivar:.+]] = phi i64 [ %[[subidx:.+]], %[[antiloop]] ], [ %[[preidx]], %for.body ]
; CHECK-NEXT: %[[subidx]] = add i64 %[[antivar]], -1
; CHECK-NEXT: %[[structptr:.+]] = getelementptr %struct.n*, %struct.n** %[[bcalloc]], i64 %[[antivar]]
; CHECK-NEXT: %[[struct:.+]] = load %struct.n*, %struct.n** %[[structptr]]
Expand All @@ -191,5 +191,5 @@ attributes #4 = { nounwind }
; CHECK-NEXT: %[[addval:.+]] = fadd fast double %[[val0]], %[[differet]]
; CHECK-NEXT: store double %[[addval]], double* %"value'ipg"
; CHECK-NEXT: %[[cmpeq:.+]] = icmp eq i64 %[[antivar]], 0
; CHECK-NEXT: br i1 %[[cmpeq]], label %invertfor.body.preheader, label %invertfor.body
; CHECK-NEXT: br i1 %[[cmpeq]], label %invertfor.body.preheader, label %[[antiloop]]
; CHECK-NEXT: }
8 changes: 4 additions & 4 deletions enzyme/test/Enzyme/nllist.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -332,7 +332,7 @@ attributes #4 = { nounwind }
; CHECK-NEXT: %[[dstructload]] = load %struct.n*, %struct.n** %"next'ipg", align 8
; CHECK-NEXT: %[[nextstruct]] = load %struct.n*, %struct.n** %next, align 8, !tbaa !7
; CHECK-NEXT: %[[mycmp:.+]] = icmp eq %struct.n* %[[nextstruct]], null
; CHECK-NEXT: br i1 %[[mycmp]], label %invertfor.cond.cleanup4, label %for.cond1.preheader
; CHECK-NEXT: br i1 %[[mycmp]], label %[[invertforcondcleanup:.+]], label %for.cond1.preheader

; CHECK: for.body5: ; preds = %for.body5, %for.cond1.preheader
; CHECK-NEXT: %[[iv:.+]] = phi i64 [ %[[ivnext:.+]], %for.body5 ], [ 0, %for.cond1.preheader ]
Expand All @@ -349,17 +349,17 @@ attributes #4 = { nounwind }

; CHECK: invertfor.cond1.preheader: ; preds = %invertfor.body5
; CHECK-NEXT: %[[icmp:.+]] = icmp eq i64 %[[antivar:.+]], 0
; CHECK-NEXT: br i1 %[[icmp]], label %invertfor.cond1.preheader.preheader, label %invertfor.cond.cleanup4
; CHECK-NEXT: br i1 %[[icmp]], label %invertfor.cond1.preheader.preheader, label %[[invertforcondcleanup]]

; CHECK: invertfor.cond.cleanup4:
; CHECK: [[invertforcondcleanup]]:
; CHECK-NEXT: %[[antivar]] = phi i64 [ %[[isub:.+]], %invertfor.cond1.preheader ], [ %[[preidx]], %for.cond.cleanup4 ]
; CHECK-NEXT: %[[isub]] = add i64 %[[antivar]], -1
; CHECK-NEXT: %[[toload:.+]] = getelementptr double*, double** %[[todoublep]], i64 %[[antivar]]
; CHECK-NEXT: %[[loadediv:.+]] = load double*, double** %[[toload]], align 8, !invariant.load
; CHECK-NEXT: br label %invertfor.body5

; CHECK: invertfor.body5:
; CHECK-NEXT: %[[mantivar:.+]] = phi i64 [ %times, %invertfor.cond.cleanup4 ], [ %[[idxsub:.+]], %invertfor.body5 ]
; CHECK-NEXT: %[[mantivar:.+]] = phi i64 [ %times, %[[invertforcondcleanup]] ], [ %[[idxsub:.+]], %invertfor.body5 ]
; CHECK-NEXT: %[[idxsub]] = add i64 %[[mantivar]], -1
; CHECK-NEXT: %"arrayidx'ipg" = getelementptr double, double* %[[loadediv]], i64 %[[mantivar]]
; CHECK-NEXT: %[[arrayload:.+]] = load double, double* %"arrayidx'ipg"
Expand Down
8 changes: 4 additions & 4 deletions enzyme/test/Enzyme/sumbr2.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
; RUN: opt < %s %loadEnzyme -enzyme -enzyme_preopt=false -inline -mem2reg -instsimplify -adce -loop-deletion -correlated-propagation -instcombine -simplifycfg -S | FileCheck %s
; RUN: opt < %s %loadEnzyme -enzyme -enzyme_preopt=false -inline -mem2reg -instsimplify -adce -loop-deletion -correlated-propagation -instcombine -simplifycfg -S -jump-threading -instsimplify -simplifycfg -adce -loop-deletion -simplifycfg | FileCheck %s

; Function Attrs: norecurse nounwind readonly uwtable
define dso_local double @sum(double* nocapture readonly %x, i64 %n) #0 {
Expand Down Expand Up @@ -45,15 +45,15 @@ attributes #2 = { nounwind }
; CHECK-NEXT: br i1 %[[exists]], label %diffesum.exit, label %[[antiloop:.+]]

; CHECK: [[antiloop]]:
; CHECK-NEXT: %"add'de.0.i" = phi double [ %[[m0dadd:.+]], %[[antiloop]] ], [ 1.000000e+00, %entry ]
; CHECK-NEXT: %[[dadd:.+]] = phi double [ %[[m0dadd:.+]], %[[antiloop]] ], [ 1.000000e+00, %entry ]
; CHECK-NEXT: %[[antivar:.+]] = phi i64 [ %[[sub:.+]], %[[antiloop]] ], [ %n, %entry ]
; CHECK-NEXT: %[[sub]] = add i64 %[[antivar]], -1
; CHECK-NEXT: %"arrayidx'ipg.i" = getelementptr double, double* %xp, i64 %[[antivar]]
; CHECK-NEXT: %[[toload:.+]] = load double, double* %"arrayidx'ipg.i", align 8
; CHECK-NEXT: %[[tostore:.+]] = fadd fast double %[[toload]], %"add'de.0.i"
; CHECK-NEXT: %[[tostore:.+]] = fadd fast double %[[toload]], %[[dadd]]
; CHECK-NEXT: store double %[[tostore]], double* %"arrayidx'ipg.i", align 8
; CHECK-NEXT: %res_unwrap.i = uitofp i64 %[[sub]] to double
; CHECK-NEXT: %[[m0dadd]] = fmul fast double %"add'de.0.i", %res_unwrap.i
; CHECK-NEXT: %[[m0dadd]] = fmul fast double %[[dadd]], %res_unwrap.i
; CHECK-NEXT: %[[itercmp:.+]] = icmp eq i64 %[[sub]], 0
; CHECK-NEXT: br i1 %[[itercmp]], label %diffesum.exit, label %invertextra.i

Expand Down
2 changes: 1 addition & 1 deletion enzyme/test/Enzyme/sumsimple.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
; RUN: opt < %s %loadEnzyme -enzyme -enzyme_preopt=false -inline -mem2reg -instsimplify -adce -loop-deletion -correlated-propagation -simplifycfg -S -early-cse | FileCheck %s
; RUN: opt < %s %loadEnzyme -enzyme -enzyme_preopt=false -inline -mem2reg -instsimplify -adce -loop-deletion -correlated-propagation -simplifycfg -S -early-cse -simplifycfg | FileCheck %s

; Function Attrs: noinline nounwind uwtable
define dso_local void @f(double* %x, double** %y, i64 %n) #0 {
Expand Down
6 changes: 3 additions & 3 deletions enzyme/test/Enzyme/sumwithbreak.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
; RUN: opt < %s %loadEnzyme -enzyme -enzyme_preopt=false -mem2reg -instcombine -correlated-propagation -adce -instcombine -simplifycfg -early-cse -simplifycfg -loop-unroll -instcombine -simplifycfg -gvn -jump-threading -instcombine -S | FileCheck %s
; RUN: opt < %s %loadEnzyme -enzyme -enzyme_preopt=false -mem2reg -instcombine -correlated-propagation -adce -instcombine -simplifycfg -early-cse -simplifycfg -loop-unroll -instcombine -simplifycfg -gvn -jump-threading -instcombine -simplifycfg -S | FileCheck %s

; Function Attrs: noinline nounwind uwtable
define dso_local double @f(double* nocapture readonly %x, i64 %n) #0 {
Expand Down Expand Up @@ -57,8 +57,8 @@ attributes #0 = { noinline nounwind uwtable }
; CHECK-NEXT: %arrayidx4 = getelementptr inbounds double, double* %x, i64 %iv
; CHECK-NEXT: %0 = load double, double* %arrayidx4, align 8
; CHECK-NEXT: %add5 = fadd fast double %0, %data.016
; CHECK-NEXT: %cmp = icmp eq i64 %iv, %n
; CHECK-NEXT: br i1 %cmp, label %invertif.end.peel, label %for.body
; CHECK-NEXT: %cmp = icmp ult i64 %iv, %n
; CHECK-NEXT: br i1 %cmp, label %for.body, label %invertif.end.peel

; CHECK: invertentry:
; CHECK-NEXT: ret {} undef
Expand Down