aboutsummaryrefslogtreecommitdiffstats
path: root/lib/Sema/SemaCUDA.cpp
blob: 203c09c57112b7232868a6cd65d8e542ea7160f0 (plain) (blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
//===--- SemaCUDA.cpp - Semantic Analysis for CUDA constructs -------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
/// \file
/// This file implements semantic analysis for CUDA constructs.
///
//===----------------------------------------------------------------------===//

#include "clang/AST/ASTContext.h"
#include "clang/AST/Decl.h"
#include "clang/AST/ExprCXX.h"
#include "clang/Basic/Cuda.h"
#include "clang/Lex/Preprocessor.h"
#include "clang/Sema/Lookup.h"
#include "clang/Sema/Sema.h"
#include "clang/Sema/SemaDiagnostic.h"
#include "clang/Sema/SemaInternal.h"
#include "clang/Sema/Template.h"
#include "llvm/ADT/Optional.h"
#include "llvm/ADT/SmallVector.h"
using namespace clang;

void Sema::PushForceCUDAHostDevice() {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");
  ForceCUDAHostDeviceDepth++;
}

bool Sema::PopForceCUDAHostDevice() {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");
  if (ForceCUDAHostDeviceDepth == 0)
    return false;
  ForceCUDAHostDeviceDepth--;
  return true;
}

ExprResult Sema::ActOnCUDAExecConfigExpr(Scope *S, SourceLocation LLLLoc,
                                         MultiExprArg ExecConfig,
                                         SourceLocation GGGLoc) {
  FunctionDecl *ConfigDecl = Context.getcudaConfigureCallDecl();
  if (!ConfigDecl)
    return ExprError(Diag(LLLLoc, diag::err_undeclared_var_use)
                     << getCudaConfigureFuncName());
  QualType ConfigQTy = ConfigDecl->getType();

  DeclRefExpr *ConfigDR = new (Context)
      DeclRefExpr(Context, ConfigDecl, false, ConfigQTy, VK_LValue, LLLLoc);
  MarkFunctionReferenced(LLLLoc, ConfigDecl);

  return BuildCallExpr(S, ConfigDR, LLLLoc, ExecConfig, GGGLoc, nullptr,
                       /*IsExecConfig=*/true);
}

Sema::CUDAFunctionTarget
Sema::IdentifyCUDATarget(const ParsedAttributesView &Attrs) {
  bool HasHostAttr = false;
  bool HasDeviceAttr = false;
  bool HasGlobalAttr = false;
  bool HasInvalidTargetAttr = false;
  for (const ParsedAttr &AL : Attrs) {
    switch (AL.getKind()) {
    case ParsedAttr::AT_CUDAGlobal:
      HasGlobalAttr = true;
      break;
    case ParsedAttr::AT_CUDAHost:
      HasHostAttr = true;
      break;
    case ParsedAttr::AT_CUDADevice:
      HasDeviceAttr = true;
      break;
    case ParsedAttr::AT_CUDAInvalidTarget:
      HasInvalidTargetAttr = true;
      break;
    default:
      break;
    }
  }

  if (HasInvalidTargetAttr)
    return CFT_InvalidTarget;

  if (HasGlobalAttr)
    return CFT_Global;

  if (HasHostAttr && HasDeviceAttr)
    return CFT_HostDevice;

  if (HasDeviceAttr)
    return CFT_Device;

  return CFT_Host;
}

template <typename A>
static bool hasAttr(const FunctionDecl *D, bool IgnoreImplicitAttr) {
  return D->hasAttrs() && llvm::any_of(D->getAttrs(), [&](Attr *Attribute) {
           return isa<A>(Attribute) &&
                  !(IgnoreImplicitAttr && Attribute->isImplicit());
         });
}

/// IdentifyCUDATarget - Determine the CUDA compilation target for this function
Sema::CUDAFunctionTarget Sema::IdentifyCUDATarget(const FunctionDecl *D,
                                                  bool IgnoreImplicitHDAttr) {
  // Code that lives outside a function is run on the host.
  if (D == nullptr)
    return CFT_Host;

  if (D->hasAttr<CUDAInvalidTargetAttr>())
    return CFT_InvalidTarget;

  if (D->hasAttr<CUDAGlobalAttr>())
    return CFT_Global;

  if (hasAttr<CUDADeviceAttr>(D, IgnoreImplicitHDAttr)) {
    if (hasAttr<CUDAHostAttr>(D, IgnoreImplicitHDAttr))
      return CFT_HostDevice;
    return CFT_Device;
  } else if (hasAttr<CUDAHostAttr>(D, IgnoreImplicitHDAttr)) {
    return CFT_Host;
  } else if (D->isImplicit() && !IgnoreImplicitHDAttr) {
    // Some implicit declarations (like intrinsic functions) are not marked.
    // Set the most lenient target on them for maximal flexibility.
    return CFT_HostDevice;
  }

  return CFT_Host;
}

// * CUDA Call preference table
//
// F - from,
// T - to
// Ph - preference in host mode
// Pd - preference in device mode
// H  - handled in (x)
// Preferences: N:native, SS:same side, HD:host-device, WS:wrong side, --:never.
//
// | F  | T  | Ph  | Pd  |  H  |
// |----+----+-----+-----+-----+
// | d  | d  | N   | N   | (c) |
// | d  | g  | --  | --  | (a) |
// | d  | h  | --  | --  | (e) |
// | d  | hd | HD  | HD  | (b) |
// | g  | d  | N   | N   | (c) |
// | g  | g  | --  | --  | (a) |
// | g  | h  | --  | --  | (e) |
// | g  | hd | HD  | HD  | (b) |
// | h  | d  | --  | --  | (e) |
// | h  | g  | N   | N   | (c) |
// | h  | h  | N   | N   | (c) |
// | h  | hd | HD  | HD  | (b) |
// | hd | d  | WS  | SS  | (d) |
// | hd | g  | SS  | --  |(d/a)|
// | hd | h  | SS  | WS  | (d) |
// | hd | hd | HD  | HD  | (b) |

Sema::CUDAFunctionPreference
Sema::IdentifyCUDAPreference(const FunctionDecl *Caller,
                             const FunctionDecl *Callee) {
  assert(Callee && "Callee must be valid.");
  CUDAFunctionTarget CallerTarget = IdentifyCUDATarget(Caller);
  CUDAFunctionTarget CalleeTarget = IdentifyCUDATarget(Callee);

  // If one of the targets is invalid, the check always fails, no matter what
  // the other target is.
  if (CallerTarget == CFT_InvalidTarget || CalleeTarget == CFT_InvalidTarget)
    return CFP_Never;

  // (a) Can't call global from some contexts until we support CUDA's
  // dynamic parallelism.
  if (CalleeTarget == CFT_Global &&
      (CallerTarget == CFT_Global || CallerTarget == CFT_Device))
    return CFP_Never;

  // (b) Calling HostDevice is OK for everyone.
  if (CalleeTarget == CFT_HostDevice)
    return CFP_HostDevice;

  // (c) Best case scenarios
  if (CalleeTarget == CallerTarget ||
      (CallerTarget == CFT_Host && CalleeTarget == CFT_Global) ||
      (CallerTarget == CFT_Global && CalleeTarget == CFT_Device))
    return CFP_Native;

  // (d) HostDevice behavior depends on compilation mode.
  if (CallerTarget == CFT_HostDevice) {
    // It's OK to call a compilation-mode matching function from an HD one.
    if ((getLangOpts().CUDAIsDevice && CalleeTarget == CFT_Device) ||
        (!getLangOpts().CUDAIsDevice &&
         (CalleeTarget == CFT_Host || CalleeTarget == CFT_Global)))
      return CFP_SameSide;

    // Calls from HD to non-mode-matching functions (i.e., to host functions
    // when compiling in device mode or to device functions when compiling in
    // host mode) are allowed at the sema level, but eventually rejected if
    // they're ever codegened.  TODO: Reject said calls earlier.
    return CFP_WrongSide;
  }

  // (e) Calling across device/host boundary is not something you should do.
  if ((CallerTarget == CFT_Host && CalleeTarget == CFT_Device) ||
      (CallerTarget == CFT_Device && CalleeTarget == CFT_Host) ||
      (CallerTarget == CFT_Global && CalleeTarget == CFT_Host))
    return CFP_Never;

  llvm_unreachable("All cases should've been handled by now.");
}

void Sema::EraseUnwantedCUDAMatches(
    const FunctionDecl *Caller,
    SmallVectorImpl<std::pair<DeclAccessPair, FunctionDecl *>> &Matches) {
  if (Matches.size() <= 1)
    return;

  using Pair = std::pair<DeclAccessPair, FunctionDecl*>;

  // Gets the CUDA function preference for a call from Caller to Match.
  auto GetCFP = [&](const Pair &Match) {
    return IdentifyCUDAPreference(Caller, Match.second);
  };

  // Find the best call preference among the functions in Matches.
  CUDAFunctionPreference BestCFP = GetCFP(*std::max_element(
      Matches.begin(), Matches.end(),
      [&](const Pair &M1, const Pair &M2) { return GetCFP(M1) < GetCFP(M2); }));

  // Erase all functions with lower priority.
  llvm::erase_if(Matches,
                 [&](const Pair &Match) { return GetCFP(Match) < BestCFP; });
}

/// When an implicitly-declared special member has to invoke more than one
/// base/field special member, conflicts may occur in the targets of these
/// members. For example, if one base's member __host__ and another's is
/// __device__, it's a conflict.
/// This function figures out if the given targets \param Target1 and
/// \param Target2 conflict, and if they do not it fills in
/// \param ResolvedTarget with a target that resolves for both calls.
/// \return true if there's a conflict, false otherwise.
static bool
resolveCalleeCUDATargetConflict(Sema::CUDAFunctionTarget Target1,
                                Sema::CUDAFunctionTarget Target2,
                                Sema::CUDAFunctionTarget *ResolvedTarget) {
  // Only free functions and static member functions may be global.
  assert(Target1 != Sema::CFT_Global);
  assert(Target2 != Sema::CFT_Global);

  if (Target1 == Sema::CFT_HostDevice) {
    *ResolvedTarget = Target2;
  } else if (Target2 == Sema::CFT_HostDevice) {
    *ResolvedTarget = Target1;
  } else if (Target1 != Target2) {
    return true;
  } else {
    *ResolvedTarget = Target1;
  }

  return false;
}

bool Sema::inferCUDATargetForImplicitSpecialMember(CXXRecordDecl *ClassDecl,
                                                   CXXSpecialMember CSM,
                                                   CXXMethodDecl *MemberDecl,
                                                   bool ConstRHS,
                                                   bool Diagnose) {
  llvm::Optional<CUDAFunctionTarget> InferredTarget;

  // We're going to invoke special member lookup; mark that these special
  // members are called from this one, and not from its caller.
  ContextRAII MethodContext(*this, MemberDecl);

  // Look for special members in base classes that should be invoked from here.
  // Infer the target of this member base on the ones it should call.
  // Skip direct and indirect virtual bases for abstract classes.
  llvm::SmallVector<const CXXBaseSpecifier *, 16> Bases;
  for (const auto &B : ClassDecl->bases()) {
    if (!B.isVirtual()) {
      Bases.push_back(&B);
    }
  }

  if (!ClassDecl->isAbstract()) {
    for (const auto &VB : ClassDecl->vbases()) {
      Bases.push_back(&VB);
    }
  }

  for (const auto *B : Bases) {
    const RecordType *BaseType = B->getType()->getAs<RecordType>();
    if (!BaseType) {
      continue;
    }

    CXXRecordDecl *BaseClassDecl = cast<CXXRecordDecl>(BaseType->getDecl());
    Sema::SpecialMemberOverloadResult SMOR =
        LookupSpecialMember(BaseClassDecl, CSM,
                            /* ConstArg */ ConstRHS,
                            /* VolatileArg */ false,
                            /* RValueThis */ false,
                            /* ConstThis */ false,
                            /* VolatileThis */ false);

    if (!SMOR.getMethod())
      continue;

    CUDAFunctionTarget BaseMethodTarget = IdentifyCUDATarget(SMOR.getMethod());
    if (!InferredTarget.hasValue()) {
      InferredTarget = BaseMethodTarget;
    } else {
      bool ResolutionError = resolveCalleeCUDATargetConflict(
          InferredTarget.getValue(), BaseMethodTarget,
          InferredTarget.getPointer());
      if (ResolutionError) {
        if (Diagnose) {
          Diag(ClassDecl->getLocation(),
               diag::note_implicit_member_target_infer_collision)
              << (unsigned)CSM << InferredTarget.getValue() << BaseMethodTarget;
        }
        MemberDecl->addAttr(CUDAInvalidTargetAttr::CreateImplicit(Context));
        return true;
      }
    }
  }

  // Same as for bases, but now for special members of fields.
  for (const auto *F : ClassDecl->fields()) {
    if (F->isInvalidDecl()) {
      continue;
    }

    const RecordType *FieldType =
        Context.getBaseElementType(F->getType())->getAs<RecordType>();
    if (!FieldType) {
      continue;
    }

    CXXRecordDecl *FieldRecDecl = cast<CXXRecordDecl>(FieldType->getDecl());
    Sema::SpecialMemberOverloadResult SMOR =
        LookupSpecialMember(FieldRecDecl, CSM,
                            /* ConstArg */ ConstRHS && !F->isMutable(),
                            /* VolatileArg */ false,
                            /* RValueThis */ false,
                            /* ConstThis */ false,
                            /* VolatileThis */ false);

    if (!SMOR.getMethod())
      continue;

    CUDAFunctionTarget FieldMethodTarget =
        IdentifyCUDATarget(SMOR.getMethod());
    if (!InferredTarget.hasValue()) {
      InferredTarget = FieldMethodTarget;
    } else {
      bool ResolutionError = resolveCalleeCUDATargetConflict(
          InferredTarget.getValue(), FieldMethodTarget,
          InferredTarget.getPointer());
      if (ResolutionError) {
        if (Diagnose) {
          Diag(ClassDecl->getLocation(),
               diag::note_implicit_member_target_infer_collision)
              << (unsigned)CSM << InferredTarget.getValue()
              << FieldMethodTarget;
        }
        MemberDecl->addAttr(CUDAInvalidTargetAttr::CreateImplicit(Context));
        return true;
      }
    }
  }

  if (InferredTarget.hasValue()) {
    if (InferredTarget.getValue() == CFT_Device) {
      MemberDecl->addAttr(CUDADeviceAttr::CreateImplicit(Context));
    } else if (InferredTarget.getValue() == CFT_Host) {
      MemberDecl->addAttr(CUDAHostAttr::CreateImplicit(Context));
    } else {
      MemberDecl->addAttr(CUDADeviceAttr::CreateImplicit(Context));
      MemberDecl->addAttr(CUDAHostAttr::CreateImplicit(Context));
    }
  } else {
    // If no target was inferred, mark this member as __host__ __device__;
    // it's the least restrictive option that can be invoked from any target.
    MemberDecl->addAttr(CUDADeviceAttr::CreateImplicit(Context));
    MemberDecl->addAttr(CUDAHostAttr::CreateImplicit(Context));
  }

  return false;
}

bool Sema::isEmptyCudaConstructor(SourceLocation Loc, CXXConstructorDecl *CD) {
  if (!CD->isDefined() && CD->isTemplateInstantiation())
    InstantiateFunctionDefinition(Loc, CD->getFirstDecl());

  // (E.2.3.1, CUDA 7.5) A constructor for a class type is considered
  // empty at a point in the translation unit, if it is either a
  // trivial constructor
  if (CD->isTrivial())
    return true;

  // ... or it satisfies all of the following conditions:
  // The constructor function has been defined.
  // The constructor function has no parameters,
  // and the function body is an empty compound statement.
  if (!(CD->hasTrivialBody() && CD->getNumParams() == 0))
    return false;

  // Its class has no virtual functions and no virtual base classes.
  if (CD->getParent()->isDynamicClass())
    return false;

  // The only form of initializer allowed is an empty constructor.
  // This will recursively check all base classes and member initializers
  if (!llvm::all_of(CD->inits(), [&](const CXXCtorInitializer *CI) {
        if (const CXXConstructExpr *CE =
                dyn_cast<CXXConstructExpr>(CI->getInit()))
          return isEmptyCudaConstructor(Loc, CE->getConstructor());
        return false;
      }))
    return false;

  return true;
}

bool Sema::isEmptyCudaDestructor(SourceLocation Loc, CXXDestructorDecl *DD) {
  // No destructor -> no problem.
  if (!DD)
    return true;

  if (!DD->isDefined() && DD->isTemplateInstantiation())
    InstantiateFunctionDefinition(Loc, DD->getFirstDecl());

  // (E.2.3.1, CUDA 7.5) A destructor for a class type is considered
  // empty at a point in the translation unit, if it is either a
  // trivial constructor
  if (DD->isTrivial())
    return true;

  // ... or it satisfies all of the following conditions:
  // The destructor function has been defined.
  // and the function body is an empty compound statement.
  if (!DD->hasTrivialBody())
    return false;

  const CXXRecordDecl *ClassDecl = DD->getParent();

  // Its class has no virtual functions and no virtual base classes.
  if (ClassDecl->isDynamicClass())
    return false;

  // Only empty destructors are allowed. This will recursively check
  // destructors for all base classes...
  if (!llvm::all_of(ClassDecl->bases(), [&](const CXXBaseSpecifier &BS) {
        if (CXXRecordDecl *RD = BS.getType()->getAsCXXRecordDecl())
          return isEmptyCudaDestructor(Loc, RD->getDestructor());
        return true;
      }))
    return false;

  // ... and member fields.
  if (!llvm::all_of(ClassDecl->fields(), [&](const FieldDecl *Field) {
        if (CXXRecordDecl *RD = Field->getType()
                                    ->getBaseElementTypeUnsafe()
                                    ->getAsCXXRecordDecl())
          return isEmptyCudaDestructor(Loc, RD->getDestructor());
        return true;
      }))
    return false;

  return true;
}

void Sema::checkAllowedCUDAInitializer(VarDecl *VD) {
  if (VD->isInvalidDecl() || !VD->hasInit() || !VD->hasGlobalStorage())
    return;
  const Expr *Init = VD->getInit();
  if (VD->hasAttr<CUDADeviceAttr>() || VD->hasAttr<CUDAConstantAttr>() ||
      VD->hasAttr<CUDASharedAttr>()) {
    assert(!VD->isStaticLocal() || VD->hasAttr<CUDASharedAttr>());
    bool AllowedInit = false;
    if (const CXXConstructExpr *CE = dyn_cast<CXXConstructExpr>(Init))
      AllowedInit =
          isEmptyCudaConstructor(VD->getLocation(), CE->getConstructor());
    // We'll allow constant initializers even if it's a non-empty
    // constructor according to CUDA rules. This deviates from NVCC,
    // but allows us to handle things like constexpr constructors.
    if (!AllowedInit &&
        (VD->hasAttr<CUDADeviceAttr>() || VD->hasAttr<CUDAConstantAttr>()))
      AllowedInit = VD->getInit()->isConstantInitializer(
          Context, VD->getType()->isReferenceType());

    // Also make sure that destructor, if there is one, is empty.
    if (AllowedInit)
      if (CXXRecordDecl *RD = VD->getType()->getAsCXXRecordDecl())
        AllowedInit =
            isEmptyCudaDestructor(VD->getLocation(), RD->getDestructor());

    if (!AllowedInit) {
      Diag(VD->getLocation(), VD->hasAttr<CUDASharedAttr>()
                                  ? diag::err_shared_var_init
                                  : diag::err_dynamic_var_init)
          << Init->getSourceRange();
      VD->setInvalidDecl();
    }
  } else {
    // This is a host-side global variable.  Check that the initializer is
    // callable from the host side.
    const FunctionDecl *InitFn = nullptr;
    if (const CXXConstructExpr *CE = dyn_cast<CXXConstructExpr>(Init)) {
      InitFn = CE->getConstructor();
    } else if (const CallExpr *CE = dyn_cast<CallExpr>(Init)) {
      InitFn = CE->getDirectCallee();
    }
    if (InitFn) {
      CUDAFunctionTarget InitFnTarget = IdentifyCUDATarget(InitFn);
      if (InitFnTarget != CFT_Host && InitFnTarget != CFT_HostDevice) {
        Diag(VD->getLocation(), diag::err_ref_bad_target_global_initializer)
            << InitFnTarget << InitFn;
        Diag(InitFn->getLocation(), diag::note_previous_decl) << InitFn;
        VD->setInvalidDecl();
      }
    }
  }
}

// With -fcuda-host-device-constexpr, an unattributed constexpr function is
// treated as implicitly __host__ __device__, unless:
//  * it is a variadic function (device-side variadic functions are not
//    allowed), or
//  * a __device__ function with this signature was already declared, in which
//    case in which case we output an error, unless the __device__ decl is in a
//    system header, in which case we leave the constexpr function unattributed.
//
// In addition, all function decls are treated as __host__ __device__ when
// ForceCUDAHostDeviceDepth > 0 (corresponding to code within a
//   #pragma clang force_cuda_host_device_begin/end
// pair).
void Sema::maybeAddCUDAHostDeviceAttrs(FunctionDecl *NewD,
                                       const LookupResult &Previous) {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");

  if (ForceCUDAHostDeviceDepth > 0) {
    if (!NewD->hasAttr<CUDAHostAttr>())
      NewD->addAttr(CUDAHostAttr::CreateImplicit(Context));
    if (!NewD->hasAttr<CUDADeviceAttr>())
      NewD->addAttr(CUDADeviceAttr::CreateImplicit(Context));
    return;
  }

  if (!getLangOpts().CUDAHostDeviceConstexpr || !NewD->isConstexpr() ||
      NewD->isVariadic() || NewD->hasAttr<CUDAHostAttr>() ||
      NewD->hasAttr<CUDADeviceAttr>() || NewD->hasAttr<CUDAGlobalAttr>())
    return;

  // Is D a __device__ function with the same signature as NewD, ignoring CUDA
  // attributes?
  auto IsMatchingDeviceFn = [&](NamedDecl *D) {
    if (UsingShadowDecl *Using = dyn_cast<UsingShadowDecl>(D))
      D = Using->getTargetDecl();
    FunctionDecl *OldD = D->getAsFunction();
    return OldD && OldD->hasAttr<CUDADeviceAttr>() &&
           !OldD->hasAttr<CUDAHostAttr>() &&
           !IsOverload(NewD, OldD, /* UseMemberUsingDeclRules = */ false,
                       /* ConsiderCudaAttrs = */ false);
  };
  auto It = llvm::find_if(Previous, IsMatchingDeviceFn);
  if (It != Previous.end()) {
    // We found a __device__ function with the same name and signature as NewD
    // (ignoring CUDA attrs).  This is an error unless that function is defined
    // in a system header, in which case we simply return without making NewD
    // host+device.
    NamedDecl *Match = *It;
    if (!getSourceManager().isInSystemHeader(Match->getLocation())) {
      Diag(NewD->getLocation(),
           diag::err_cuda_unattributed_constexpr_cannot_overload_device)
          << NewD;
      Diag(Match->getLocation(),
           diag::note_cuda_conflicting_device_function_declared_here);
    }
    return;
  }

  NewD->addAttr(CUDAHostAttr::CreateImplicit(Context));
  NewD->addAttr(CUDADeviceAttr::CreateImplicit(Context));
}

// Do we know that we will eventually codegen the given function?
static bool IsKnownEmitted(Sema &S, FunctionDecl *FD) {
  // Templates are emitted when they're instantiated.
  if (FD->isDependentContext())
    return false;

  // When compiling for device, host functions are never emitted.  Similarly,
  // when compiling for host, device and global functions are never emitted.
  // (Technically, we do emit a host-side stub for global functions, but this
  // doesn't count for our purposes here.)
  Sema::CUDAFunctionTarget T = S.IdentifyCUDATarget(FD);
  if (S.getLangOpts().CUDAIsDevice && T == Sema::CFT_Host)
    return false;
  if (!S.getLangOpts().CUDAIsDevice &&
      (T == Sema::CFT_Device || T == Sema::CFT_Global))
    return false;

  // Check whether this function is externally visible -- if so, it's
  // known-emitted.
  //
  // We have to check the GVA linkage of the function's *definition* -- if we
  // only have a declaration, we don't know whether or not the function will be
  // emitted, because (say) the definition could include "inline".
  FunctionDecl *Def = FD->getDefinition();

  if (Def &&
      !isDiscardableGVALinkage(S.getASTContext().GetGVALinkageForFunction(Def)))
    return true;

  // Otherwise, the function is known-emitted if it's in our set of
  // known-emitted functions.
  return S.DeviceKnownEmittedFns.count(FD) > 0;
}

Sema::DeviceDiagBuilder Sema::CUDADiagIfDeviceCode(SourceLocation Loc,
                                                   unsigned DiagID) {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");
  DeviceDiagBuilder::Kind DiagKind = [this] {
    switch (CurrentCUDATarget()) {
    case CFT_Global:
    case CFT_Device:
      return DeviceDiagBuilder::K_Immediate;
    case CFT_HostDevice:
      // An HD function counts as host code if we're compiling for host, and
      // device code if we're compiling for device.  Defer any errors in device
      // mode until the function is known-emitted.
      if (getLangOpts().CUDAIsDevice) {
        return IsKnownEmitted(*this, dyn_cast<FunctionDecl>(CurContext))
                   ? DeviceDiagBuilder::K_ImmediateWithCallStack
                   : DeviceDiagBuilder::K_Deferred;
      }
      return DeviceDiagBuilder::K_Nop;

    default:
      return DeviceDiagBuilder::K_Nop;
    }
  }();
  return DeviceDiagBuilder(DiagKind, Loc, DiagID,
                           dyn_cast<FunctionDecl>(CurContext), *this);
}

Sema::DeviceDiagBuilder Sema::CUDADiagIfHostCode(SourceLocation Loc,
                                                 unsigned DiagID) {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");
  DeviceDiagBuilder::Kind DiagKind = [this] {
    switch (CurrentCUDATarget()) {
    case CFT_Host:
      return DeviceDiagBuilder::K_Immediate;
    case CFT_HostDevice:
      // An HD function counts as host code if we're compiling for host, and
      // device code if we're compiling for device.  Defer any errors in device
      // mode until the function is known-emitted.
      if (getLangOpts().CUDAIsDevice)
        return DeviceDiagBuilder::K_Nop;

      return IsKnownEmitted(*this, dyn_cast<FunctionDecl>(CurContext))
                 ? DeviceDiagBuilder::K_ImmediateWithCallStack
                 : DeviceDiagBuilder::K_Deferred;
    default:
      return DeviceDiagBuilder::K_Nop;
    }
  }();
  return DeviceDiagBuilder(DiagKind, Loc, DiagID,
                           dyn_cast<FunctionDecl>(CurContext), *this);
}

bool Sema::CheckCUDACall(SourceLocation Loc, FunctionDecl *Callee) {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");
  assert(Callee && "Callee may not be null.");

  auto &ExprEvalCtx = ExprEvalContexts.back();
  if (ExprEvalCtx.isUnevaluated() || ExprEvalCtx.isConstantEvaluated())
    return true;

  // FIXME: Is bailing out early correct here?  Should we instead assume that
  // the caller is a global initializer?
  FunctionDecl *Caller = dyn_cast<FunctionDecl>(CurContext);
  if (!Caller)
    return true;

  // If the caller is known-emitted, mark the callee as known-emitted.
  // Otherwise, mark the call in our call graph so we can traverse it later.
  bool CallerKnownEmitted = IsKnownEmitted(*this, Caller);
  if (CallerKnownEmitted) {
    // Host-side references to a __global__ function refer to the stub, so the
    // function itself is never emitted and therefore should not be marked.
    if (getLangOpts().CUDAIsDevice || IdentifyCUDATarget(Callee) != CFT_Global)
      markKnownEmitted(*this, Caller, Callee, Loc, IsKnownEmitted);
  } else {
    // If we have
    //   host fn calls kernel fn calls host+device,
    // the HD function does not get instantiated on the host.  We model this by
    // omitting at the call to the kernel from the callgraph.  This ensures
    // that, when compiling for host, only HD functions actually called from the
    // host get marked as known-emitted.
    if (getLangOpts().CUDAIsDevice || IdentifyCUDATarget(Callee) != CFT_Global)
      DeviceCallGraph[Caller].insert({Callee, Loc});
  }

  DeviceDiagBuilder::Kind DiagKind = [this, Caller, Callee,
                                      CallerKnownEmitted] {
    switch (IdentifyCUDAPreference(Caller, Callee)) {
    case CFP_Never:
      return DeviceDiagBuilder::K_Immediate;
    case CFP_WrongSide:
      assert(Caller && "WrongSide calls require a non-null caller");
      // If we know the caller will be emitted, we know this wrong-side call
      // will be emitted, so it's an immediate error.  Otherwise, defer the
      // error until we know the caller is emitted.
      return CallerKnownEmitted ? DeviceDiagBuilder::K_ImmediateWithCallStack
                                : DeviceDiagBuilder::K_Deferred;
    default:
      return DeviceDiagBuilder::K_Nop;
    }
  }();

  if (DiagKind == DeviceDiagBuilder::K_Nop)
    return true;

  // Avoid emitting this error twice for the same location.  Using a hashtable
  // like this is unfortunate, but because we must continue parsing as normal
  // after encountering a deferred error, it's otherwise very tricky for us to
  // ensure that we only emit this deferred error once.
  if (!LocsWithCUDACallDiags.insert({Caller, Loc}).second)
    return true;

  DeviceDiagBuilder(DiagKind, Loc, diag::err_ref_bad_target, Caller, *this)
      << IdentifyCUDATarget(Callee) << Callee << IdentifyCUDATarget(Caller);
  DeviceDiagBuilder(DiagKind, Callee->getLocation(), diag::note_previous_decl,
                    Caller, *this)
      << Callee;
  return DiagKind != DeviceDiagBuilder::K_Immediate &&
         DiagKind != DeviceDiagBuilder::K_ImmediateWithCallStack;
}

void Sema::CUDASetLambdaAttrs(CXXMethodDecl *Method) {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");
  if (Method->hasAttr<CUDAHostAttr>() || Method->hasAttr<CUDADeviceAttr>())
    return;
  FunctionDecl *CurFn = dyn_cast<FunctionDecl>(CurContext);
  if (!CurFn)
    return;
  CUDAFunctionTarget Target = IdentifyCUDATarget(CurFn);
  if (Target == CFT_Global || Target == CFT_Device) {
    Method->addAttr(CUDADeviceAttr::CreateImplicit(Context));
  } else if (Target == CFT_HostDevice) {
    Method->addAttr(CUDADeviceAttr::CreateImplicit(Context));
    Method->addAttr(CUDAHostAttr::CreateImplicit(Context));
  }
}

void Sema::checkCUDATargetOverload(FunctionDecl *NewFD,
                                   const LookupResult &Previous) {
  assert(getLangOpts().CUDA && "Should only be called during CUDA compilation");
  CUDAFunctionTarget NewTarget = IdentifyCUDATarget(NewFD);
  for (NamedDecl *OldND : Previous) {
    FunctionDecl *OldFD = OldND->getAsFunction();
    if (!OldFD)
      continue;

    CUDAFunctionTarget OldTarget = IdentifyCUDATarget(OldFD);
    // Don't allow HD and global functions to overload other functions with the
    // same signature.  We allow overloading based on CUDA attributes so that
    // functions can have different implementations on the host and device, but
    // HD/global functions "exist" in some sense on both the host and device, so
    // should have the same implementation on both sides.
    if (NewTarget != OldTarget &&
        ((NewTarget == CFT_HostDevice) || (OldTarget == CFT_HostDevice) ||
         (NewTarget == CFT_Global) || (OldTarget == CFT_Global)) &&
        !IsOverload(NewFD, OldFD, /* UseMemberUsingDeclRules = */ false,
                    /* ConsiderCudaAttrs = */ false)) {
      Diag(NewFD->getLocation(), diag::err_cuda_ovl_target)
          << NewTarget << NewFD->getDeclName() << OldTarget << OldFD;
      Diag(OldFD->getLocation(), diag::note_previous_declaration);
      NewFD->setInvalidDecl();
      break;
    }
  }
}

template <typename AttrTy>
static void copyAttrIfPresent(Sema &S, FunctionDecl *FD,
                              const FunctionDecl &TemplateFD) {
  if (AttrTy *Attribute = TemplateFD.getAttr<AttrTy>()) {
    AttrTy *Clone = Attribute->clone(S.Context);
    Clone->setInherited(true);
    FD->addAttr(Clone);
  }
}

void Sema::inheritCUDATargetAttrs(FunctionDecl *FD,
                                  const FunctionTemplateDecl &TD) {
  const FunctionDecl &TemplateFD = *TD.getTemplatedDecl();
  copyAttrIfPresent<CUDAGlobalAttr>(*this, FD, TemplateFD);
  copyAttrIfPresent<CUDAHostAttr>(*this, FD, TemplateFD);
  copyAttrIfPresent<CUDADeviceAttr>(*this, FD, TemplateFD);
}

std::string Sema::getCudaConfigureFuncName() const {
  if (getLangOpts().HIP)
    return "hipConfigureCall";

  // New CUDA kernel launch sequence.
  if (CudaFeatureEnabled(Context.getTargetInfo().getSDKVersion(),
                         CudaFeature::CUDA_USES_NEW_LAUNCH))
    return "__cudaPushCallConfiguration";

  // Legacy CUDA kernel configuration call
  return "cudaConfigureCall";
}