[clang] [PowerPC] Support local-dynamic TLS relocation on AIX (PR #66316)

[Felix via cfe-commits]

orcguru wrote:

I saw comments being folded, paste here for notice, sorry about the duplication:

So I did have another version which can essentially remove the assert check 
with some side effects:

(1) It did generated extra external symbol reference, which looks like no harm 
with simple HelloWorld. (Notice: xlC does not generate this symbol reference.)
```
.extern _Renamed..5f24__TLSML[UA]
.rename _Renamed..5f24__TLSML[UA],"_$TLSML"
```
(2) Inside `XCOFFObjectWriter::recordRelocation()` logic I have to apply a hack 
to get the correct symbol, because now there are two symbols: one is 
`_Renamed..5f24__TLSML[UA]`, the other is `_Renamed..5f24__TLSML[TC]`. Without 
hack it will choose the [UA] one which will cause linker failure.

Attached the draft patch: 
[ReplaceAssertHack.diff.txt](https://github.com/llvm/llvm-project/files/12723368/ReplaceAssertHack.diff.txt)

I think current version has the advantage that it does not generate any extra 
external symbol reference or any extra relocation entry, although it looks ugly 
with regard to those asserts.

Is it possible that we fix the hack implementation with some future version (I 
will open an issue and work on it afterwards)? How about let's move on with 
current approach?

@amy-kwan @stephenpeckham @bzEq @chenzheng1030 Appreciate your comments. Thank 
you!

https://github.com/llvm/llvm-project/pull/66316
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang] Fix pretty-printing assume_aligned attributes (PR #67331)

[Balazs Benics via cfe-commits]

https://github.com/steakhal updated 
https://github.com/llvm/llvm-project/pull/67331

>From 988dcf4f568a58d9b1127b2adf8890a7c7867ac5 Mon Sep 17 00:00:00 2001
From: Balazs Benics 
Date: Mon, 25 Sep 2023 15:37:34 +0200
Subject: [PATCH] [clang] Fix pretty-printing assume_aligned attributes

Inside `writePrettyPrintFunction()`, we check if we need to emit the
given argument:
```C++
if (!arg->isOptional() || arg->getIsOmitted() == "false") {
FoundNonOptArg = true;
continue;
}
```
For the `AssumeAligned` attribute, the second argument was optional, but
the `getIsOmitted()` returned `false`, thus we treated this argument as
**non-optional** in the end because of that disjunction.

It was because `getIsOmitted()` did not account for `Expr *` type, and
returned `false` on the fallthrough branch.

Fixes #67156
---
 clang/test/AST/attr-print-emit.cpp| 6 ++
 clang/utils/TableGen/ClangAttrEmitter.cpp | 7 ---
 2 files changed, 10 insertions(+), 3 deletions(-)

diff --git a/clang/test/AST/attr-print-emit.cpp 
b/clang/test/AST/attr-print-emit.cpp
index cc7413baf10e87a..8c48eb92daba5ee 100644
--- a/clang/test/AST/attr-print-emit.cpp
+++ b/clang/test/AST/attr-print-emit.cpp
@@ -2,6 +2,12 @@
 // RUN: %clang -emit-ast -o %t.ast %s
 // RUN: %clang_cc1 %t.ast -ast-print | FileCheck %s
 
+// CHECK: void *aa() __attribute__((assume_aligned(64)));
+void *aa() __attribute__((assume_aligned(64)));
+
+// CHECK: void *aa2() __attribute__((assume_aligned(64, 8)));
+void *aa2() __attribute__((assume_aligned(64, 8)));
+
 // CHECK: void xla(int a) __attribute__((xray_log_args(1)));
 void xla(int a) __attribute__((xray_log_args(1)));
 
diff --git a/clang/utils/TableGen/ClangAttrEmitter.cpp 
b/clang/utils/TableGen/ClangAttrEmitter.cpp
index 7ea09058c3d39f2..f036656a7685fd6 100644
--- a/clang/utils/TableGen/ClangAttrEmitter.cpp
+++ b/clang/utils/TableGen/ClangAttrEmitter.cpp
@@ -320,11 +320,12 @@ namespace {
 }
 
 std::string getIsOmitted() const override {
-  if (type == "IdentifierInfo *")
+  StringRef T = type;
+  if (T == "IdentifierInfo *" || T == "Expr *")
 return "!get" + getUpperName().str() + "()";
-  if (type == "TypeSourceInfo *")
+  if (T == "TypeSourceInfo *")
 return "!get" + getUpperName().str() + "Loc()";
-  if (type == "ParamIdx")
+  if (T == "ParamIdx")
 return "!get" + getUpperName().str() + "().isValid()";
   return "false";
 }

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang] Fix pretty-printing assume_aligned attributes (PR #67331)

[via cfe-commits]

github-actions[bot] wrote:




:warning: C/C++ code formatter, clang-format found issues in your code. 
:warning:



You can test this locally with the following command:


``bash
git-clang-format --diff 9f276d4ddd0efa2e323d674a35317c253ab66d58 
67e19382fb45fe5e06a5c8de2e7b1434c8b1c68f -- clang/test/AST/attr-print-emit.cpp 
clang/utils/TableGen/ClangAttrEmitter.cpp
``





View the diff from clang-format here.


``diff
diff --git a/clang/utils/TableGen/ClangAttrEmitter.cpp 
b/clang/utils/TableGen/ClangAttrEmitter.cpp
index effeba51e99e..a572810c3cf4 100644
--- a/clang/utils/TableGen/ClangAttrEmitter.cpp
+++ b/clang/utils/TableGen/ClangAttrEmitter.cpp
@@ -3176,26 +3176,26 @@ namespace clang {
 
 // Emits the enumeration list for attributes.
 void EmitClangAttrList(RecordKeeper &Records, raw_ostream &OS) {
-  emitSourceFileHeader("List of all attributes that Clang recognizes", OS);
+emitSourceFileHeader("List of all attributes that Clang recognizes", OS);
 
-  AttrClassHierarchy Hierarchy(Records);
+AttrClassHierarchy Hierarchy(Records);
 
-  // Add defaulting macro definitions.
-  Hierarchy.emitDefaultDefines(OS);
-  emitDefaultDefine(OS, "PRAGMA_SPELLING_ATTR", nullptr);
+// Add defaulting macro definitions.
+Hierarchy.emitDefaultDefines(OS);
+emitDefaultDefine(OS, "PRAGMA_SPELLING_ATTR", nullptr);
 
-  std::vector Attrs = Records.getAllDerivedDefinitions("Attr");
-  std::vector PragmaAttrs;
-  for (auto *Attr : Attrs) {
-if (!Attr->getValueAsBit("ASTNode"))
-  continue;
+std::vector Attrs = Records.getAllDerivedDefinitions("Attr");
+std::vector PragmaAttrs;
+for (auto *Attr : Attrs) {
+  if (!Attr->getValueAsBit("ASTNode"))
+continue;
 
-// Add the attribute to the ad-hoc groups.
-if (AttrHasPragmaSpelling(Attr))
-  PragmaAttrs.push_back(Attr);
+  // Add the attribute to the ad-hoc groups.
+  if (AttrHasPragmaSpelling(Attr))
+PragmaAttrs.push_back(Attr);
 
-// Place it in the hierarchy.
-Hierarchy.classifyAttr(Attr);
+  // Place it in the hierarchy.
+  Hierarchy.classifyAttr(Attr);
   }
 
   // Emit the main attribute list.
@@ -3594,7 +3594,8 @@ void EmitClangAttrHasAttrImpl(RecordKeeper &Records, 
raw_ostream &OS) {
 
 void EmitClangAttrSpellingListIndex(RecordKeeper &Records, raw_ostream &OS) {
   emitSourceFileHeader("Code to translate different attribute spellings "
-   "into internal identifiers", OS);
+   "into internal identifiers",
+   OS);
 
   OS << "  switch (getParsedKind()) {\n";
   OS << "case IgnoredAttribute:\n";
@@ -4687,8 +4688,7 @@ void EmitClangAttrNodeTraverse(RecordKeeper &Records, 
raw_ostream &OS) {
   }
 }
 
-void EmitClangAttrParserStringSwitches(RecordKeeper &Records,
-   raw_ostream &OS) {
+void EmitClangAttrParserStringSwitches(RecordKeeper &Records, raw_ostream &OS) 
{
   emitSourceFileHeader("Parser-related llvm::StringSwitch cases", OS);
   emitClangAttrArgContextList(Records, OS);
   emitClangAttrIdentifierArgList(Records, OS);

``




https://github.com/llvm/llvm-project/pull/67331
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D158156: [analyzer] Add C++ array delete checker

[Discookie via Phabricator via cfe-commits]

Discookie added a comment.

@steakhal gentle ping for one last round of reviews


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D158156/new/

https://reviews.llvm.org/D158156

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [PowerPC] Support local-dynamic TLS relocation on AIX (PR #66316)

[Chen Zheng via cfe-commits]

chenzheng1030 wrote:

> Is it possible that we fix the hack implementation with some future version 
> (I will open an issue and work on it afterwards)? How about let's move on 
> with current approach?
> @amy-kwan @stephenpeckham @bzEq @chenzheng1030 Appreciate your comments. 
> Thank you!

I am not against this proposal given the internal timeline.

https://github.com/llvm/llvm-project/pull/66316
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [PowerPC] Support local-dynamic TLS relocation on AIX (PR #66316)

[Chen Zheng via cfe-commits]

chenzheng1030 wrote:

The code formatting check fails. 

https://github.com/llvm/llvm-project/pull/66316
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[libunwind] [runtimes] Remove explicit -isysroot from the testing configurations on macOS (PR #66265)

[Martin Storsjö via cfe-commits]

mstorsjo wrote:

> But even regardless of #67201, the intended way to use the libc++ test suite 
> is to create a `.cfg.in` file that suits your needs if the general ones 
> don't, not to add more configuration options to the general ones.

IMO, I think this is something I fundamentally disagree with.

Yes, the previous test setup with long intricate python classes was probably 
too messy/complex. But the other extreme - only ever having entirely static 
`.cfg` files isn't the best either (if that's required, I'm quite sure there'll 
be cases where people just skip caring about running the tests, rather than 
adding a config).

Some level of configurability for allowing injecting compiler flags in the 
existing test configs would be very welcome, IMO. It shouldn't evolve into the 
old mess of course, but just allowing adding something like 
`LIBCXX_TEST_CXXFLAGS`, for adding flags specifically to the test runs, would 
be quite useful for some test configs I would expect. (I currently don't have a 
need for the sysroot stuff touched upon here, but I get the demand for it.) 
E.g. kinda like building libc++ with `-std=c++23` but testing compiling the 
user code with `-std=c++17`. I guess that particular case is supported via some 
of the testsuite parameters, but users may want to build/test in other 
configurations as well - and as long as it's fundamentally the same as existing 
configs, having to create a new test config for it feels very much unnecessary.

https://github.com/llvm/llvm-project/pull/66265
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang][TSA] Make RequiresCapability a DeclOrType attribute (PR #67095)

[Timm Baeder via cfe-commits]

Timm =?utf-8?q?Bäder?= 
Message-ID:
In-Reply-To: 


https://github.com/tbaederr updated 
https://github.com/llvm/llvm-project/pull/67095

>From 174144b3d2101964a438a22644cbfa4c2314700b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= 
Date: Fri, 22 Sep 2023 08:42:05 +0200
Subject: [PATCH 1/2] [clang][TSA] Make RequiresCapability a DeclOrType
 attribute

---
 clang/include/clang/Basic/Attr.td |  6 +++---
 clang/test/Sema/warn-thread-safety-analysis.c | 11 +++
 2 files changed, 14 insertions(+), 3 deletions(-)

diff --git a/clang/include/clang/Basic/Attr.td 
b/clang/include/clang/Basic/Attr.td
index 21a3b5226623cf2..7eab87dac6921f1 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -3305,7 +3305,7 @@ def ReleaseCapability : InheritableAttr {
   let Documentation = [ReleaseCapabilityDocs];
 }
 
-def RequiresCapability : InheritableAttr {
+def RequiresCapability : DeclOrTypeAttr {
   let Spellings = [Clang<"requires_capability", 0>,
Clang<"exclusive_locks_required", 0>,
Clang<"requires_shared_capability", 0>,
@@ -3314,8 +3314,8 @@ def RequiresCapability : InheritableAttr {
   let LateParsed = 1;
   let TemplateDependent = 1;
   let ParseArgumentsAsUnevaluated = 1;
-  let InheritEvenIfAlreadyPresent = 1;
-  let Subjects = SubjectList<[Function]>;
+  /*let InheritEvenIfAlreadyPresent = 1;*/
+  /*let Subjects = SubjectList<[Function]>;*/
   let Accessors = [Accessor<"isShared", [Clang<"requires_shared_capability", 
0>,
  Clang<"shared_locks_required", 0>]>];
   let Documentation = [Undocumented];
diff --git a/clang/test/Sema/warn-thread-safety-analysis.c 
b/clang/test/Sema/warn-thread-safety-analysis.c
index 642ea88ec3c96f7..cd852efac21cb81 100644
--- a/clang/test/Sema/warn-thread-safety-analysis.c
+++ b/clang/test/Sema/warn-thread-safety-analysis.c
@@ -136,6 +136,17 @@ int main(void) {
 // Cleanup happens automatically -> no warning.
   }
 
+  /// Function pointers
+  {
+int __attribute__((requires_capability(&mu1))) (*function_ptr)(int) = 
Foo_fun1;
+
+function_ptr(5); // expected-warning {{calling function 'function_ptr' 
requires holding mutex 'mu1'}}
+
+mutex_exclusive_lock(&mu1);
+int five = function_ptr(5);
+mutex_exclusive_unlock(&mu1);
+  }
+
   return 0;
 }
 

>From 2d0d268acef6e85325f656d2bf94cb205cc1e371 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= 
Date: Tue, 26 Sep 2023 09:57:41 +0200
Subject: [PATCH 2/2] Restrict to functions and function pointer decls

---
 clang/include/clang/Basic/Attr.td | 2 +-
 clang/lib/Sema/SemaDeclAttr.cpp   | 9 +
 2 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/clang/include/clang/Basic/Attr.td 
b/clang/include/clang/Basic/Attr.td
index 7eab87dac6921f1..dd2625aaabf16bd 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -3314,7 +3314,7 @@ def RequiresCapability : DeclOrTypeAttr {
   let LateParsed = 1;
   let TemplateDependent = 1;
   let ParseArgumentsAsUnevaluated = 1;
-  /*let InheritEvenIfAlreadyPresent = 1;*/
+  let InheritEvenIfAlreadyPresent = 1;
   /*let Subjects = SubjectList<[Function]>;*/
   let Accessors = [Accessor<"isShared", [Clang<"requires_shared_capability", 
0>,
  Clang<"shared_locks_required", 0>]>];
diff --git a/clang/lib/Sema/SemaDeclAttr.cpp b/clang/lib/Sema/SemaDeclAttr.cpp
index 090a54eedaa07d0..48fef83791369f5 100644
--- a/clang/lib/Sema/SemaDeclAttr.cpp
+++ b/clang/lib/Sema/SemaDeclAttr.cpp
@@ -8141,6 +8141,15 @@ static void handleRequiresCapabilityAttr(Sema &S, Decl 
*D,
   if (!AL.checkAtLeastNumArgs(S, 1))
 return;
 
+  // We allow this on function declaration as well as
+  // variable declarations of function pointer type.
+  if (!D->isFunctionPointerType() && !isa(D)) {
+// FIXME: Diagnostic should say "functions and function pointer decls" now 
I guess.
+S.Diag(D->getLocation(), diag::warn_attribute_wrong_decl_type)
+<< AL << AL.isRegularKeywordAttribute() << ExpectedFunction;
+return;
+  }
+
   // check that all arguments are lockable objects
   SmallVector Args;
   checkAttrArgsAreCapabilityObjs(S, D, AL, Args);

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang][TSA] Make RequiresCapability a DeclOrType attribute (PR #67095)

[Timm Baeder via cfe-commits]

Timm =?utf-8?q?Bäder?= 
Message-ID:
In-Reply-To: 


https://github.com/tbaederr updated 
https://github.com/llvm/llvm-project/pull/67095

>From 174144b3d2101964a438a22644cbfa4c2314700b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= 
Date: Fri, 22 Sep 2023 08:42:05 +0200
Subject: [PATCH 1/2] [clang][TSA] Make RequiresCapability a DeclOrType
 attribute

---
 clang/include/clang/Basic/Attr.td |  6 +++---
 clang/test/Sema/warn-thread-safety-analysis.c | 11 +++
 2 files changed, 14 insertions(+), 3 deletions(-)

diff --git a/clang/include/clang/Basic/Attr.td 
b/clang/include/clang/Basic/Attr.td
index 21a3b5226623cf2..7eab87dac6921f1 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -3305,7 +3305,7 @@ def ReleaseCapability : InheritableAttr {
   let Documentation = [ReleaseCapabilityDocs];
 }
 
-def RequiresCapability : InheritableAttr {
+def RequiresCapability : DeclOrTypeAttr {
   let Spellings = [Clang<"requires_capability", 0>,
Clang<"exclusive_locks_required", 0>,
Clang<"requires_shared_capability", 0>,
@@ -3314,8 +3314,8 @@ def RequiresCapability : InheritableAttr {
   let LateParsed = 1;
   let TemplateDependent = 1;
   let ParseArgumentsAsUnevaluated = 1;
-  let InheritEvenIfAlreadyPresent = 1;
-  let Subjects = SubjectList<[Function]>;
+  /*let InheritEvenIfAlreadyPresent = 1;*/
+  /*let Subjects = SubjectList<[Function]>;*/
   let Accessors = [Accessor<"isShared", [Clang<"requires_shared_capability", 
0>,
  Clang<"shared_locks_required", 0>]>];
   let Documentation = [Undocumented];
diff --git a/clang/test/Sema/warn-thread-safety-analysis.c 
b/clang/test/Sema/warn-thread-safety-analysis.c
index 642ea88ec3c96f7..cd852efac21cb81 100644
--- a/clang/test/Sema/warn-thread-safety-analysis.c
+++ b/clang/test/Sema/warn-thread-safety-analysis.c
@@ -136,6 +136,17 @@ int main(void) {
 // Cleanup happens automatically -> no warning.
   }
 
+  /// Function pointers
+  {
+int __attribute__((requires_capability(&mu1))) (*function_ptr)(int) = 
Foo_fun1;
+
+function_ptr(5); // expected-warning {{calling function 'function_ptr' 
requires holding mutex 'mu1'}}
+
+mutex_exclusive_lock(&mu1);
+int five = function_ptr(5);
+mutex_exclusive_unlock(&mu1);
+  }
+
   return 0;
 }
 

>From 5dd9266c4e0b0a02cbb02a29bf2606224ccf2820 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Timm=20B=C3=A4der?= 
Date: Tue, 26 Sep 2023 09:57:41 +0200
Subject: [PATCH 2/2] Restrict to functions and function pointer decls

---
 clang/include/clang/Basic/Attr.td |  2 +-
 clang/lib/Sema/SemaDeclAttr.cpp   | 10 ++
 2 files changed, 11 insertions(+), 1 deletion(-)

diff --git a/clang/include/clang/Basic/Attr.td 
b/clang/include/clang/Basic/Attr.td
index 7eab87dac6921f1..dd2625aaabf16bd 100644
--- a/clang/include/clang/Basic/Attr.td
+++ b/clang/include/clang/Basic/Attr.td
@@ -3314,7 +3314,7 @@ def RequiresCapability : DeclOrTypeAttr {
   let LateParsed = 1;
   let TemplateDependent = 1;
   let ParseArgumentsAsUnevaluated = 1;
-  /*let InheritEvenIfAlreadyPresent = 1;*/
+  let InheritEvenIfAlreadyPresent = 1;
   /*let Subjects = SubjectList<[Function]>;*/
   let Accessors = [Accessor<"isShared", [Clang<"requires_shared_capability", 
0>,
  Clang<"shared_locks_required", 0>]>];
diff --git a/clang/lib/Sema/SemaDeclAttr.cpp b/clang/lib/Sema/SemaDeclAttr.cpp
index 090a54eedaa07d0..7499b3dfc4b2995 100644
--- a/clang/lib/Sema/SemaDeclAttr.cpp
+++ b/clang/lib/Sema/SemaDeclAttr.cpp
@@ -8141,6 +8141,16 @@ static void handleRequiresCapabilityAttr(Sema &S, Decl 
*D,
   if (!AL.checkAtLeastNumArgs(S, 1))
 return;
 
+  // We allow this on function declaration as well as
+  // variable declarations of function pointer type.
+  if (!D->isFunctionPointerType() && !isa(D)) {
+// FIXME: Diagnostic should say "functions and function pointer decls" now 
I
+// guess.
+S.Diag(D->getLocation(), diag::warn_attribute_wrong_decl_type)
+<< AL << AL.isRegularKeywordAttribute() << ExpectedFunction;
+return;
+  }
+
   // check that all arguments are lockable objects
   SmallVector Args;
   checkAttrArgsAreCapabilityObjs(S, D, AL, Args);

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86][FP16] Add missing handling for FP16 constrained cmp intrinsics (PR #67400)

[Phoebe Wang via cfe-commits]

https://github.com/phoebewang created 
https://github.com/llvm/llvm-project/pull/67400

None

>From 08a386acd5e14a7af905ae3ea819e5d29d84e4c1 Mon Sep 17 00:00:00 2001
From: Phoebe Wang 
Date: Tue, 26 Sep 2023 16:07:17 +0800
Subject: [PATCH] [X86][FP16] Add missing handling for FP16 constrained cmp
 intrinsics

---
 clang/lib/CodeGen/CGBuiltin.cpp   |   9 +
 .../X86/avx512fp16-builtins-constrained-cmp.c | 399 ++
 2 files changed, 408 insertions(+)
 create mode 100644 clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c

diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index d14cf0dccb09982..04c0325c7fd038b 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -15594,6 +15594,15 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned 
BuiltinID,
   case X86::BI__builtin_ia32_cmppd256:
 IID = Intrinsic::x86_avx_cmp_pd_256;
 break;
+  case X86::BI__builtin_ia32_cmpph128_mask:
+IID = Intrinsic::x86_avx512fp16_mask_cmp_ph_128;
+break;
+  case X86::BI__builtin_ia32_cmpph256_mask:
+IID = Intrinsic::x86_avx512fp16_mask_cmp_ph_256;
+break;
+  case X86::BI__builtin_ia32_cmpph512_mask:
+IID = Intrinsic::x86_avx512fp16_mask_cmp_ph_512;
+break;
   case X86::BI__builtin_ia32_cmpps512_mask:
 IID = Intrinsic::x86_avx512_mask_cmp_ps_512;
 break;
diff --git a/clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c 
b/clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c
new file mode 100644
index 000..1a164ff57fda1b1
--- /dev/null
+++ b/clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c
@@ -0,0 +1,399 @@
+// RUN: %clang_cc1 -flax-vector-conversions=none -ffreestanding %s 
-triple=x86_64-apple-darwin -target-feature +avx512fp16 -emit-llvm 
-ffp-exception-behavior=strict -o - -Wall -Werror | FileCheck %s
+
+#include 
+
+__mmask32 test_mm512_cmp_round_ph_mask(__m512h a, __m512h b) {
+  // CHECK-LABEL: @test_mm512_cmp_round_ph_mask
+  // CHECK: call <32 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.512(<32 x half> 
%{{.*}}, <32 x half> %{{.*}}, i32 0, <32 x i1> {{.*}}, i32 8)
+  return _mm512_cmp_round_ph_mask(a, b, _CMP_EQ_OQ, _MM_FROUND_NO_EXC);
+}
+
+__mmask32 test_mm512_mask_cmp_round_ph_mask(__mmask32 m, __m512h a, __m512h b) 
{
+  // CHECK-LABEL: @test_mm512_mask_cmp_round_ph_mask
+  // CHECK: call <32 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.512(<32 x half> 
%{{.*}}, <32 x half> %{{.*}}, i32 0, <32 x i1> {{.*}}, i32 8)
+  return _mm512_mask_cmp_round_ph_mask(m, a, b, _CMP_EQ_OQ, _MM_FROUND_NO_EXC);
+}
+
+__mmask8 test_mm_cmp_ph_mask_eq_oq(__m128h a, __m128h b) {
+  // CHECK-LABEL: @test_mm_cmp_ph_mask_eq_oq
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 0, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_EQ_OQ);
+}
+
+__mmask8 test_mm_cmp_ph_mask_lt_os(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_lt_os
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 1, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_LT_OS);
+}
+
+__mmask8 test_mm_cmp_ph_mask_le_os(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_le_os
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 2, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_LE_OS);
+}
+
+__mmask8 test_mm_cmp_ph_mask_unord_q(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_unord_q
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 3, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_UNORD_Q);
+}
+
+__mmask8 test_mm_cmp_ph_mask_neq_uq(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_neq_uq
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 4, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_NEQ_UQ);
+}
+
+__mmask8 test_mm_cmp_ph_mask_nlt_us(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_nlt_us
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 5, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_NLT_US);
+}
+
+__mmask8 test_mm_cmp_ph_mask_nle_us(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_nle_us
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 6, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_NLE_US);
+}
+
+__mmask8 test_mm_cmp_ph_mask_ord_q(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_ord_q
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 7, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_ORD_Q);
+}
+
+__mmask8 test_mm_cmp_ph_mask_eq_uq(__m128h a, __m128h b) 

[clang] [clang][RISCV] Fix the condition of checking signature in getIndex (PR #67403)

[Jianjian Guan via cfe-commits]

https://github.com/jacquesguan created 
https://github.com/llvm/llvm-project/pull/67403

The current condition causes assert failing if try to add a new vendor vector 
file which only contains the same type signature.

>From 3ad50442210c7303f0ac55339db778aa777532c8 Mon Sep 17 00:00:00 2001
From: Jianjian GUAN 
Date: Tue, 26 Sep 2023 16:32:14 +0800
Subject: [PATCH] [clang][RISCV] Fix the condition of checking signature in
 getIndex

The current condition causes assert failing if try to add a new vendor vector 
file which only contains the same type signature.
---
 clang/utils/TableGen/RISCVVEmitter.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/clang/utils/TableGen/RISCVVEmitter.cpp 
b/clang/utils/TableGen/RISCVVEmitter.cpp
index 41025926058ed07..c08e48b3f44dfe3 100644
--- a/clang/utils/TableGen/RISCVVEmitter.cpp
+++ b/clang/utils/TableGen/RISCVVEmitter.cpp
@@ -290,7 +290,7 @@ unsigned 
SemaSignatureTable::getIndex(ArrayRef Signature) {
 return 0;
 
   // Checking Signature already in table or not.
-  if (Signature.size() < SignatureTable.size()) {
+  if (Signature.size() <= SignatureTable.size()) {
 size_t Bound = SignatureTable.size() - Signature.size() + 1;
 for (size_t Index = 0; Index < Bound; ++Index) {
   if (equal(Signature.begin(), Signature.end(),

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [CodeGen] Avoid potential sideeffects from XOR (PR #67193)

[Simon Pilgrim via cfe-commits]

@@ -9066,21 +9066,18 @@ bool 
AArch64InstrInfo::shouldOutlineFromFunctionByDefault(
 
 void AArch64InstrInfo::buildClearRegister(Register Reg, MachineBasicBlock &MBB,
   MachineBasicBlock::iterator Iter,
-  DebugLoc &DL) const {
+  DebugLoc &DL,
+  bool NoSideEffects) const {
   const MachineFunction &MF = *MBB.getParent();
   const AArch64Subtarget &STI = MF.getSubtarget();
   const AArch64RegisterInfo &TRI = *STI.getRegisterInfo();
 
   if (TRI.isGeneralPurposeRegister(MF, Reg)) {
-BuildMI(MBB, Iter, DL, get(AArch64::MOVi64imm), Reg)
-  .addImm(0);
+BuildMI(MBB, Iter, DL, get(AArch64::MOVZXi), Reg).addImm(0).addImm(0);
   } else if (STI.hasSVE()) {
-BuildMI(MBB, Iter, DL, get(AArch64::DUP_ZI_D), Reg)
-  .addImm(0)
-  .addImm(0);
+BuildMI(MBB, Iter, DL, get(AArch64::DUP_ZI_D), Reg).addImm(0).addImm(0);
   } else {
-BuildMI(MBB, Iter, DL, get(AArch64::MOVIv2d_ns), Reg)
-  .addImm(0);
+BuildMI(MBB, Iter, DL, get(AArch64::MOVIv2d_ns), Reg).addImm(0);

RKSimon wrote:

NFC clang-format change - remove this from the patch

https://github.com/llvm/llvm-project/pull/67193
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang-tools-extra] [CodeGen] Avoid potential sideeffects from XOR (PR #67193)

[Simon Pilgrim via cfe-commits]

@@ -9066,21 +9066,18 @@ bool 
AArch64InstrInfo::shouldOutlineFromFunctionByDefault(
 
 void AArch64InstrInfo::buildClearRegister(Register Reg, MachineBasicBlock &MBB,
   MachineBasicBlock::iterator Iter,
-  DebugLoc &DL) const {
+  DebugLoc &DL,
+  bool NoSideEffects) const {
   const MachineFunction &MF = *MBB.getParent();
   const AArch64Subtarget &STI = MF.getSubtarget();
   const AArch64RegisterInfo &TRI = *STI.getRegisterInfo();
 
   if (TRI.isGeneralPurposeRegister(MF, Reg)) {
-BuildMI(MBB, Iter, DL, get(AArch64::MOVi64imm), Reg)
-  .addImm(0);
+BuildMI(MBB, Iter, DL, get(AArch64::MOVZXi), Reg).addImm(0).addImm(0);
   } else if (STI.hasSVE()) {
-BuildMI(MBB, Iter, DL, get(AArch64::DUP_ZI_D), Reg)
-  .addImm(0)
-  .addImm(0);
+BuildMI(MBB, Iter, DL, get(AArch64::DUP_ZI_D), Reg).addImm(0).addImm(0);
   } else {
-BuildMI(MBB, Iter, DL, get(AArch64::MOVIv2d_ns), Reg)
-  .addImm(0);
+BuildMI(MBB, Iter, DL, get(AArch64::MOVIv2d_ns), Reg).addImm(0);

RKSimon wrote:

NFC clang-format change - remove this from the patch

https://github.com/llvm/llvm-project/pull/67193
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86][FP16] Add missing handling for FP16 constrained cmp intrinsics (PR #67400)

[Simon Pilgrim via cfe-commits]

https://github.com/RKSimon approved this pull request.

LGTM

https://github.com/llvm/llvm-project/pull/67400
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

https://github.com/vgvassilev unresolved 
https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [analyzer] Add std::variant checker (PR #66481)

[Gábor Spaits via cfe-commits]

https://github.com/spaits edited https://github.com/llvm/llvm-project/pull/66481
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Freddy Ye via cfe-commits]

https://github.com/FreddyLeaf created 
https://github.com/llvm/llvm-project/pull/67410

None

>From 3ae0fa2592d3a11dd084d3aefb54c26819016781 Mon Sep 17 00:00:00 2001
From: Freddy Ye 
Date: Tue, 26 Sep 2023 16:44:01 +0800
Subject: [PATCH] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx
 into sse/sse2

---
 clang/include/clang/Basic/BuiltinsX86.def|   8 +-
 clang/lib/Headers/avxintrin.h| 272 -
 clang/lib/Headers/emmintrin.h| 119 ++
 clang/lib/Headers/xmmintrin.h| 152 +++
 clang/test/CodeGen/X86/avx-builtins.c| 396 ---
 clang/test/CodeGen/X86/sse-builtins.c| 198 ++
 clang/test/CodeGen/X86/sse2-builtins.c   | 198 ++
 clang/test/CodeGen/target-features-error-2.c |   4 +-
 8 files changed, 673 insertions(+), 674 deletions(-)

diff --git a/clang/include/clang/Basic/BuiltinsX86.def 
b/clang/include/clang/Basic/BuiltinsX86.def
index e4802f8ab1c1562..6bbbc4bbdd75c68 100644
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ b/clang/include/clang/Basic/BuiltinsX86.def
@@ -233,6 +233,8 @@ TARGET_BUILTIN(__builtin_ia32_minps, "V4fV4fV4f", 
"ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_maxps, "V4fV4fV4f", "ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_minss, "V4fV4fV4f", "ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_maxss, "V4fV4fV4f", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "sse")
 
 TARGET_BUILTIN(__builtin_ia32_cmpeqpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpltpd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -250,6 +252,8 @@ TARGET_BUILTIN(__builtin_ia32_cmpneqsd, "V2dV2dV2d", 
"ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpnltsd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpnlesd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpordsd, "V2dV2dV2d", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_minpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_maxpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_minsd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -469,12 +473,8 @@ TARGET_BUILTIN(__builtin_ia32_blendvps256, "V8fV8fV8fV8f", 
"ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_shufpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_shufps256, "V8fV8fV8fIi", "ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_dpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_cmppd256, "V4dV4dV4dIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_cmpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_pd256, "V2dV4dIi", "ncV:256:", 
"avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_ps256, "V4fV8fIi", "ncV:256:", 
"avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_si256, "V4iV8iIi", "ncV:256:", 
"avx")
diff --git a/clang/lib/Headers/avxintrin.h b/clang/lib/Headers/avxintrin.h
index b796bb773ec11f0..afbf26403f201b6 100644
--- a/clang/lib/Headers/avxintrin.h
+++ b/clang/lib/Headers/avxintrin.h
@@ -1569,160 +1569,6 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
   ((__m256d)__builtin_ia32_shufpd256((__v4df)(__m256d)(a), \
  (__v4df)(__m256d)(b), (int)(mask)))
 
-/* Compare */
-#define _CMP_EQ_OQ0x00 /* Equal (ordered, non-signaling)  */
-#define _CMP_LT_OS0x01 /* Less-than (ordered, signaling)  */
-#define _CMP_LE_OS0x02 /* Less-than-or-equal (ordered, signaling)  */
-#define _CMP_UNORD_Q  0x03 /* Unordered (non-signaling)  */
-#define _CMP_NEQ_UQ   0x04 /* Not-equal (unordered, non-signaling)  */
-#define _CMP_NLT_US   0x05 /* Not-less-than (unordered, signaling)  */
-#define _CMP_NLE_US   0x06 /* Not-less-than-or-equal (unordered, signaling)  */
-#define _CMP_ORD_Q0x07 /* Ordered (non-signaling)   */
-#define _CMP_EQ_UQ0x08 /* Equal (unordered, non-signaling)  */
-#define _CMP_NGE_US   0x09 /* Not-greater-than-or-equal (unordered, signaling) 
 */
-#define _CMP_NGT_US   0x0a /* Not-greater-than (unordered, signaling)  */
-#define _CMP_FALSE_OQ 0x0b /* False (ordered, non-signaling)  */
-#define _CMP_NEQ_OQ   0x0c /* Not-equal (ordered, non-signaling)  */
-#define _CMP_GE_OS0x0d /* Greater-than-or-equal (ordered, signaling)  */
-#define _CMP_GT_OS0x0e /* Greater-than (ordered, signaling)  */
-#define _CMP_TRUE_UQ  0x0f /* True (unordered, non-signaling)  */
-#define _CMP_EQ_OS0x10 /* Equal (ordered, signalin

[clang] [Clang] Fix missing diagnostic for non-standard layout type in `offsetof` (PR #65246)

[via cfe-commits]

kasuga-fj wrote:

> Do we have any Github workflow documentation that mentions we shouldn't 
> force-push to update PRs?

According to the [LLVM GitHub User 
Guide](https://llvm.org/docs/GitHub.html#rebasing-pull-requests-and-force-pushes),
 we should generally avoid rebasing and pushing force. But I trust it will be 
allowed in this case.

> My question is about the desired end state of libc++, not whether that end 
> state should be achieved in this specific patch or not.

I misunderstood, sorry. I'm not an expert of libcxx, so I cannot determine what 
is should be (although it is only what an amateur thinks, at least it seems to 
be working fine now, so just suppressing warning would not be a big problem, I 
think). 
Incidentally, as you may have already noticed, it may be worth noting that a 
relatively recent commit introduced `__libcpp_datasizeof` 
(https://github.com/llvm/llvm-project/commit/c4e98722ca79c827dd57b809e0ef16b3b8da8c72).

> Certainly _FirstPaddingByte<> can't be standard layout if the T it holds is 
> not standard layout?

Yes, you are correct. I was only checking with standard-layout class... I tried 
passing a non-standard layout class in the type parameter and 
`_FirstPaddingByte<>` became non standard-layout. By the way, I checked the 
CI's failure log and it looks that perhaps there is no test which uses 
`__libcpp_datasizeof` with non standard-layout type (I may have missed 
something). I think it would be better adding such one.

https://github.com/llvm/llvm-project/pull/65246
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Freddy Ye via cfe-commits]

FreddyLeaf wrote:

This PR is to align with icc: https://godbolt.org/z/EzbfzTrzr We can modify 
intrinsic guide if required.

intel.com/SDM shows:
```
CMPSD xmm1, xmm2/m64, imm8
SSE2: Compare low double-precision floating-point value in xmm2/m64 and xmm1 
using bits 2:0 of imm8 as comparison predicate
Intel C/C++ Compiler Intrinsic Equivalent
(V)CMPSD _m128d _mm_cmp_sd(_m128d a, __m128d b, const int imm)

https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[via cfe-commits]

github-actions[bot] wrote:




:warning: C/C++ code formatter, clang-format found issues in your code. 
:warning:



You can test this locally with the following command:


``bash
git-clang-format --diff 7675f541f75baa20e8ec007cd625a837e89fc01f 
3ae0fa2592d3a11dd084d3aefb54c26819016781 -- clang/lib/Headers/avxintrin.h 
clang/lib/Headers/emmintrin.h clang/lib/Headers/xmmintrin.h 
clang/test/CodeGen/X86/avx-builtins.c clang/test/CodeGen/X86/sse-builtins.c 
clang/test/CodeGen/X86/sse2-builtins.c 
clang/test/CodeGen/target-features-error-2.c
``





View the diff from clang-format here.


``diff
diff --git a/clang/lib/Headers/emmintrin.h b/clang/lib/Headers/emmintrin.h
index 3e103ee7a24b..f857b9864d08 100644
--- a/clang/lib/Headers/emmintrin.h
+++ b/clang/lib/Headers/emmintrin.h
@@ -4798,9 +4798,9 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS 
_mm_castsi128_pd(__m128i __a) {
 ///0x1E: Greater-than (ordered, non-signaling) \n
 ///0x1F: True (unordered, signaling)
 /// \returns A 128-bit vector of [2 x double] containing the comparison 
results.
-#define _mm_cmp_pd(a, b, c) \
-  ((__m128d)__builtin_ia32_cmppd((__v2df)(__m128d)(a), \
- (__v2df)(__m128d)(b), (c)))
+#define _mm_cmp_pd(a, b, c)
\
+  ((__m128d)__builtin_ia32_cmppd((__v2df)(__m128d)(a), (__v2df)(__m128d)(b),   
\
+ (c)))
 
 /// Compares each of the corresponding scalar double-precision values of
 ///two 128-bit vectors of [2 x double], using the operation specified by 
the
@@ -4857,9 +4857,9 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS 
_mm_castsi128_pd(__m128i __a) {
 ///0x1E: Greater-than (ordered, non-signaling) \n
 ///0x1F: True (unordered, signaling)
 /// \returns A 128-bit vector of [2 x double] containing the comparison 
results.
-#define _mm_cmp_sd(a, b, c) \
-  ((__m128d)__builtin_ia32_cmpsd((__v2df)(__m128d)(a), \
- (__v2df)(__m128d)(b), (c)))
+#define _mm_cmp_sd(a, b, c)
\
+  ((__m128d)__builtin_ia32_cmpsd((__v2df)(__m128d)(a), (__v2df)(__m128d)(b),   
\
+ (c)))
 
 #if defined(__cplusplus)
 extern "C" {
diff --git a/clang/lib/Headers/xmmintrin.h b/clang/lib/Headers/xmmintrin.h
index 7b238d8ba000..fbca88a2823f 100644
--- a/clang/lib/Headers/xmmintrin.h
+++ b/clang/lib/Headers/xmmintrin.h
@@ -2937,38 +2937,41 @@ _mm_movemask_ps(__m128 __a)
 }
 
 /* Compare */
-#define _CMP_EQ_OQ0x00 /* Equal (ordered, non-signaling)  */
-#define _CMP_LT_OS0x01 /* Less-than (ordered, signaling)  */
-#define _CMP_LE_OS0x02 /* Less-than-or-equal (ordered, signaling)  */
-#define _CMP_UNORD_Q  0x03 /* Unordered (non-signaling)  */
-#define _CMP_NEQ_UQ   0x04 /* Not-equal (unordered, non-signaling)  */
-#define _CMP_NLT_US   0x05 /* Not-less-than (unordered, signaling)  */
-#define _CMP_NLE_US   0x06 /* Not-less-than-or-equal (unordered, signaling)  */
-#define _CMP_ORD_Q0x07 /* Ordered (non-signaling)   */
-#define _CMP_EQ_UQ0x08 /* Equal (unordered, non-signaling)  */
-#define _CMP_NGE_US   0x09 /* Not-greater-than-or-equal (unordered, signaling) 
 */
-#define _CMP_NGT_US   0x0a /* Not-greater-than (unordered, signaling)  */
+#define _CMP_EQ_OQ 0x00/* Equal (ordered, non-signaling)  */
+#define _CMP_LT_OS 0x01/* Less-than (ordered, signaling)  */
+#define _CMP_LE_OS 0x02/* Less-than-or-equal (ordered, signaling)  */
+#define _CMP_UNORD_Q 0x03  /* Unordered (non-signaling)  */
+#define _CMP_NEQ_UQ 0x04   /* Not-equal (unordered, non-signaling)  */
+#define _CMP_NLT_US 0x05   /* Not-less-than (unordered, signaling)  */
+#define _CMP_NLE_US 0x06   /* Not-less-than-or-equal (unordered, signaling)  */
+#define _CMP_ORD_Q 0x07/* Ordered (non-signaling)   */
+#define _CMP_EQ_UQ 0x08/* Equal (unordered, non-signaling)  */
+#define _CMP_NGE_US 0x09   /* Not-greater-than-or-equal (unordered, signaling) 
\
+*/
+#define _CMP_NGT_US 0x0a   /* Not-greater-than (unordered, signaling)  */
 #define _CMP_FALSE_OQ 0x0b /* False (ordered, non-signaling)  */
-#define _CMP_NEQ_OQ   0x0c /* Not-equal (ordered, non-signaling)  */
-#define _CMP_GE_OS0x0d /* Greater-than-or-equal (ordered, signaling)  */
-#define _CMP_GT_OS0x0e /* Greater-than (ordered, signaling)  */
-#define _CMP_TRUE_UQ  0x0f /* True (unordered, non-signaling)  */
-#define _CMP_EQ_OS0x10 /* Equal (ordered, signaling)  */
-#define _CMP_LT_OQ0x11 /* Less-than (ordered, non-signaling)  */
-#define _CMP_LE_OQ0x12 /* Less-than-or-equal (ordered, non-signaling)  */
-#define _CMP_UNORD_S  0x13 /* Unordered (signaling)  */
-#define _CMP_NEQ_US   0x14 /* Not-equal (unordered, signaling)  */
-#define _CMP_NLT_UQ   0x15 /* Not-less-than (unordered, non-signaling)  */
-#define _CMP_NLE_UQ   0x16 /* Not-less-than-or-equal (unordered, 
non-signaling)  */
-#define _CMP_

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Freddy Ye via cfe-commits]

https://github.com/FreddyLeaf updated 
https://github.com/llvm/llvm-project/pull/67410

>From 3ae0fa2592d3a11dd084d3aefb54c26819016781 Mon Sep 17 00:00:00 2001
From: Freddy Ye 
Date: Tue, 26 Sep 2023 16:44:01 +0800
Subject: [PATCH 1/2] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from
 avx into sse/sse2

---
 clang/include/clang/Basic/BuiltinsX86.def|   8 +-
 clang/lib/Headers/avxintrin.h| 272 -
 clang/lib/Headers/emmintrin.h| 119 ++
 clang/lib/Headers/xmmintrin.h| 152 +++
 clang/test/CodeGen/X86/avx-builtins.c| 396 ---
 clang/test/CodeGen/X86/sse-builtins.c| 198 ++
 clang/test/CodeGen/X86/sse2-builtins.c   | 198 ++
 clang/test/CodeGen/target-features-error-2.c |   4 +-
 8 files changed, 673 insertions(+), 674 deletions(-)

diff --git a/clang/include/clang/Basic/BuiltinsX86.def 
b/clang/include/clang/Basic/BuiltinsX86.def
index e4802f8ab1c1562..6bbbc4bbdd75c68 100644
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ b/clang/include/clang/Basic/BuiltinsX86.def
@@ -233,6 +233,8 @@ TARGET_BUILTIN(__builtin_ia32_minps, "V4fV4fV4f", 
"ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_maxps, "V4fV4fV4f", "ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_minss, "V4fV4fV4f", "ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_maxss, "V4fV4fV4f", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "sse")
 
 TARGET_BUILTIN(__builtin_ia32_cmpeqpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpltpd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -250,6 +252,8 @@ TARGET_BUILTIN(__builtin_ia32_cmpneqsd, "V2dV2dV2d", 
"ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpnltsd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpnlesd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpordsd, "V2dV2dV2d", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_minpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_maxpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_minsd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -469,12 +473,8 @@ TARGET_BUILTIN(__builtin_ia32_blendvps256, "V8fV8fV8fV8f", 
"ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_shufpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_shufps256, "V8fV8fV8fIi", "ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_dpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_cmppd256, "V4dV4dV4dIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_cmpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_pd256, "V2dV4dIi", "ncV:256:", 
"avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_ps256, "V4fV8fIi", "ncV:256:", 
"avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_si256, "V4iV8iIi", "ncV:256:", 
"avx")
diff --git a/clang/lib/Headers/avxintrin.h b/clang/lib/Headers/avxintrin.h
index b796bb773ec11f0..afbf26403f201b6 100644
--- a/clang/lib/Headers/avxintrin.h
+++ b/clang/lib/Headers/avxintrin.h
@@ -1569,160 +1569,6 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
   ((__m256d)__builtin_ia32_shufpd256((__v4df)(__m256d)(a), \
  (__v4df)(__m256d)(b), (int)(mask)))
 
-/* Compare */
-#define _CMP_EQ_OQ0x00 /* Equal (ordered, non-signaling)  */
-#define _CMP_LT_OS0x01 /* Less-than (ordered, signaling)  */
-#define _CMP_LE_OS0x02 /* Less-than-or-equal (ordered, signaling)  */
-#define _CMP_UNORD_Q  0x03 /* Unordered (non-signaling)  */
-#define _CMP_NEQ_UQ   0x04 /* Not-equal (unordered, non-signaling)  */
-#define _CMP_NLT_US   0x05 /* Not-less-than (unordered, signaling)  */
-#define _CMP_NLE_US   0x06 /* Not-less-than-or-equal (unordered, signaling)  */
-#define _CMP_ORD_Q0x07 /* Ordered (non-signaling)   */
-#define _CMP_EQ_UQ0x08 /* Equal (unordered, non-signaling)  */
-#define _CMP_NGE_US   0x09 /* Not-greater-than-or-equal (unordered, signaling) 
 */
-#define _CMP_NGT_US   0x0a /* Not-greater-than (unordered, signaling)  */
-#define _CMP_FALSE_OQ 0x0b /* False (ordered, non-signaling)  */
-#define _CMP_NEQ_OQ   0x0c /* Not-equal (ordered, non-signaling)  */
-#define _CMP_GE_OS0x0d /* Greater-than-or-equal (ordered, signaling)  */
-#define _CMP_GT_OS0x0e /* Greater-than (ordered, signaling)  */
-#define _CMP_TRUE_UQ  0x0f /* True (unordered, non-signaling)  */
-#define _CMP_EQ_OS0x10 /* Equal (ordered, signaling)

[clang] [Clang] Fix missing diagnostic for non-standard layout type in `offsetof` (PR #65246)

[via cfe-commits]

philnik777 wrote:

Assuming that Clang doesn't plan on optimizing anything based on this, I think 
it should be fine to just silence the warning. We could also add a 
`__datasizeof` builtin in a separate patch if we want. That would probably be 
less error-prone and cheaper than our current solution. FWIW we'll need some 
version of this when we want to replace `__compressed_pair`, so a cheaper 
solution would probably be a good idea.

https://github.com/llvm/llvm-project/pull/65246
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang-format] Correctly annotate return type of function pointer (PR #66893)

[Pierre Jolivet via cfe-commits]

prj- wrote:

Pinging @owenca.

https://github.com/llvm/llvm-project/pull/66893
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [ARM] fix "+fp.dp" in multilib selection (PR #67412)

[Dominik Wójt via cfe-commits]

https://github.com/domin144 created 
https://github.com/llvm/llvm-project/pull/67412

When the FPU was selected with "+(no)fp(.dp)" extensions in "-march" or "-mcpu" 
options, the FPU used for multilib selection was still the default one for 
given architecture or CPU.

From 4cc2328982cc8260263aa803f2b02bff7d7b4a85 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Dominik=20W=C3=B3jt?= 
Date: Wed, 20 Sep 2023 16:39:27 +0200
Subject: [PATCH] [ARM] fix "+fp.dp" in multilib selection

When the FPU was selected with "+(no)fp(.dp)" extensions in "-march" or
"-mcpu" options, the FPU used for multilib selection was still the
default one for given architecture or CPU.
---
 clang/lib/Driver/ToolChains/Arch/ARM.cpp  |  5 ++
 .../test/Driver/print-multi-selection-flags.c | 10 +++
 .../llvm/TargetParser/ARMTargetParser.h   |  8 +++
 llvm/lib/TargetParser/ARMTargetParser.cpp | 69 +++
 4 files changed, 77 insertions(+), 15 deletions(-)

diff --git a/clang/lib/Driver/ToolChains/Arch/ARM.cpp 
b/clang/lib/Driver/ToolChains/Arch/ARM.cpp
index bb66db5feae8c3b..9a042d351bda0b6 100644
--- a/clang/lib/Driver/ToolChains/Arch/ARM.cpp
+++ b/clang/lib/Driver/ToolChains/Arch/ARM.cpp
@@ -626,6 +626,11 @@ llvm::ARM::FPUKind arm::getARMTargetFeatures(const Driver 
&D,
 if (!llvm::ARM::getFPUFeatures(FPUKind, Features))
   D.Diag(clang::diag::err_drv_clang_unsupported)
   << std::string("-mfpu=") + AndroidFPU;
+  } else if (ArchArgFPUKind != llvm::ARM::FK_INVALID ||
+ CPUArgFPUKind != llvm::ARM::FK_INVALID) {
+FPUKind =
+CPUArgFPUKind != llvm::ARM::FK_INVALID ? CPUArgFPUKind : 
ArchArgFPUKind;
+(void)llvm::ARM::getFPUFeatures(FPUKind, Features);
   } else {
 if (!ForAS) {
   std::string CPU = arm::getARMTargetCPU(CPUName, ArchName, Triple);
diff --git a/clang/test/Driver/print-multi-selection-flags.c 
b/clang/test/Driver/print-multi-selection-flags.c
index 819ff242ced0026..f020a22ea128aac 100644
--- a/clang/test/Driver/print-multi-selection-flags.c
+++ b/clang/test/Driver/print-multi-selection-flags.c
@@ -35,6 +35,16 @@
 // CHECK-MVENOFP-NOT: -march=thumbv8.1m.main{{.*}}+mve.fp{{.*}}
 // CHECK-MVENOFP: -mfpu=none
 
+// RUN: %clang -print-multi-flags-experimental --target=arm-none-eabihf 
-march=armv8.1m.main+fp.dp | FileCheck --check-prefix=CHECK-V8_1_FP_DP %s
+// CHECK-V8_1_FP_DP: -march=thumbv8.1m.main{{.*}}
+// CHECK-V8_1_FP_DP: -mfloat-abi=hard
+// CHECK-V8_1_FP_DP: -mfpu=fp-armv8-fullfp16-d16
+
+// RUN: %clang -print-multi-flags-experimental --target=arm-none-eabihf 
-mcpu=cortex-m85+nofp.dp | FileCheck --check-prefix=CHECK-M85_NO_FP_DP %s
+// CHECK-M85_NO_FP_DP: -march=thumbv8.1m.main{{.*}}
+// CHECK-M85_NO_FP_DP: -mfloat-abi=hard
+// CHECK-M85_NO_FP_DP: -mfpu=fp-armv8-fullfp16-sp-d16
+
 // RUN: %clang -print-multi-flags-experimental --target=aarch64-none-elf 
-march=armv8-a+lse | FileCheck --check-prefix=CHECK-LSE %s
 // CHECK-LSE: -march=aarch64{{.*}}+lse{{.*}}
 
diff --git a/llvm/include/llvm/TargetParser/ARMTargetParser.h 
b/llvm/include/llvm/TargetParser/ARMTargetParser.h
index b893eab1902f81b..c42d66f048fccc0 100644
--- a/llvm/include/llvm/TargetParser/ARMTargetParser.h
+++ b/llvm/include/llvm/TargetParser/ARMTargetParser.h
@@ -143,6 +143,14 @@ enum class FPURestriction {
   SP_D16///< Only single-precision instructions, with 16 D registers
 };
 
+inline bool isDoublePrecision(const FPURestriction restriction) {
+  return restriction != FPURestriction::SP_D16;
+}
+
+inline bool has32Regs(const FPURestriction restriction) {
+  return restriction == FPURestriction::None;
+}
+
 // An FPU name implies one of three levels of Neon support:
 enum class NeonSupportLevel {
   None = 0, ///< No Neon
diff --git a/llvm/lib/TargetParser/ARMTargetParser.cpp 
b/llvm/lib/TargetParser/ARMTargetParser.cpp
index c84928eeb07b140..524bb4e0fd0db22 100644
--- a/llvm/lib/TargetParser/ARMTargetParser.cpp
+++ b/llvm/lib/TargetParser/ARMTargetParser.cpp
@@ -366,26 +366,50 @@ StringRef ARM::getArchExtFeature(StringRef ArchExt) {
 }
 
 static ARM::FPUKind findDoublePrecisionFPU(ARM::FPUKind InputFPUKind) {
+  if (InputFPUKind == ARM::FK_INVALID || InputFPUKind == ARM::FK_NONE)
+return ARM::FK_INVALID;
+
+  const ARM::FPUName &InputFPU = ARM::FPUNames[InputFPUKind];
+
+  if (ARM::isDoublePrecision(InputFPU.Restriction))
+return InputFPUKind;
+
+  // Otherwise, look for an FPU entry with all the same fields, except
+  // that it supports double precision.
+  for (const ARM::FPUName &CandidateFPU : ARM::FPUNames) {
+if (CandidateFPU.FPUVer == InputFPU.FPUVer &&
+CandidateFPU.NeonSupport == InputFPU.NeonSupport &&
+ARM::has32Regs(CandidateFPU.Restriction) ==
+ARM::has32Regs(InputFPU.Restriction) &&
+ARM::isDoublePrecision(CandidateFPU.Restriction)) {
+  return CandidateFPU.ID;
+}
+  }
+
+  // nothing found
+  return ARM::FK_INVALID;
+}
+
+static ARM::FPUKind findSinglePrecisionFPU(ARM::FPUKind InputFPUKind) {
+  if (InputFPUKin

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -1625,6 +1625,39 @@ SmallVector has grown a few other minor advantages over 
std::vector, causing
and is no longer "private to the implementation". A name like
``SmallVectorHeader`` might be more appropriate.
 
+.. _dss_pagedvector:
+
+llvm/ADT/PagedVector.h
+^^
+
+``PagedVector`` is a random access container that allocates
+(PageSize) elements of type Type when the first element of a page is accessed

vgvassilev wrote:

```suggestion
``PageSize`` elements of type ``Type`` when the first element of a page is 
accessed
```

https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -0,0 +1,323 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/iterator_range.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+
+namespace llvm {
+/// A vector that allocates memory in pages.
+///
+/// Order is kept, but memory is allocated only when one element of the page is
+/// accessed. This introduces a level of indirection, but it is useful when you
+/// have a sparsely initialised vector where the full size is allocated 
upfront.
+///
+/// As a side effect the elements are initialised later than in a normal 
vector.
+/// On the first access to one of the elements of a given page all, the 
elements
+/// of the page are initialised. This also means that the elements of the page
+/// are initialised beyond the size of the vector.
+///
+/// Similarly on destruction the elements are destroyed only when the page is
+/// not needed anymore, delaying invoking the destructor of the elements.
+///
+/// Notice that this does not have iterators, because if you have iterators it
+/// probably means you are going to touch all the memory in any case, so better
+/// use a std::vector in the first place.
+template  class PagedVector {
+  static_assert(PageSize > 1, "PageSize must be greater than 0. Most likely "
+  "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.
+  size_t Size = 0;
+
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable SmallVector PageToDataPtrs;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The order of
+  // the elements however depends on the order of access of the pages.
+  PointerIntPair Allocator;
+
+  constexpr static T *InvalidPage = nullptr;
+
+public:
+  using value_type = T;
+
+  /// Default constructor. We build our own allocator and mark it as such with
+  /// `true` in the second pair element.
+  PagedVector() : Allocator(new BumpPtrAllocator, true) {}
+  PagedVector(BumpPtrAllocator *A) : Allocator(A, false) {
+assert(A != nullptr && "Allocator cannot be null");
+  }
+
+  ~PagedVector() {
+clear();
+// If we own the allocator, delete it.
+if (Allocator.getInt())
+  delete Allocator.getPointer();
+  }
+
+  /// Look up an element at position `Index`.
+  /// If the associated page is not filled, it will be filled with default
+  /// constructed elements. If the associated page is filled, return the
+  /// element.
+  T &operator[](size_t Index) const {
+assert(Index < Size);
+assert(Index / PageSize < PageToDataPtrs.size());
+T *&PagePtr = PageToDataPtrs[Index / PageSize];
+// If the page was not yet allocated, allocate it.
+if (PagePtr == InvalidPage) {
+  T *NewPagePtr = Allocator.getPointer()->template Allocate(PageSize);
+  // We need to invoke the default constructor on all the elements of the
+  // page.
+  for (size_t I = 0; I < PageSize; ++I)
+new (NewPagePtr + I) T();
+
+  PagePtr = NewPagePtr;
+}
+// Dereference the element in the page.
+return *((Index % PageSize) + PagePtr);
+  }
+
+  /// Return the capacity of the vector. I.e. the maximum size it can be
+  /// expanded to with the resize method without allocating more pages.
+  [[nodiscard]] size_t capacity() const {
+return PageToDataPtrs.size() * PageSize;
+  }
+
+  /// Return the size of the vector. I.e. the maximum index that can be
+  /// accessed, i.e. the maximum value which was used as argument of the
+  /// resize method.
+  [[nodiscard]] size_t size() const { return Size; }
+
+  /// Resize the vector. Notice that the constructor of the elements will not
+  /// be invoked until an element of a given page is accessed, at which point
+  /// all the elements of the page will be constructed.
+  ///
+  /// If the new size is smaller than the current size, the elements of the
+  /// pages that are not needed anymore will be destroyed, however, elements of
+  /// the last page will not be destroyed.
+  ///
+  /// For these reason the usage of this vector is discouraged if you rely
+  /// on the construction / destructor of the elements to be 

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -1625,6 +1625,38 @@ SmallVector has grown a few other minor advantages over 
std::vector, causing
and is no longer "private to the implementation". A name like
``SmallVectorHeader`` might be more appropriate.
 
+.. _dss_pagedvector:
+
+llvm/ADT/PagedVector.h
+^^
+
+``PagedVector`` is a random access container that allocates
+(PageSize) elements of type Type when the first element of a page is accessed
+via the ``operator[]``.  This is useful for the case in which the number of
+elements is known in advance and their actual initialization is expensive and
+sparse so that it's only done lazily when the element is accessed. When the
+number of used pages is small significant memory savings can be achieved.
+
+The main advantage is that a ``PagedVector`` allows to delay the actual 
allocation
+of the page until it's needed, at the extra cost of one integer per page and 
one
+extra indirection when accessing elements with their positional index. 
+
+In order to maximise the memory footprint of this container, it's important to
+balance the PageSize so that it's not too small (otherwise the overhead of the
+integer per page might become too high) and not too big (otherwise the memory 
is
+wasted if the page is not fully used).
+
+Moreover, while retaining the oder of the elements based on their insertion

vgvassilev wrote:

That seems unresolved.

https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -0,0 +1,323 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/iterator_range.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+
+namespace llvm {
+/// A vector that allocates memory in pages.
+///
+/// Order is kept, but memory is allocated only when one element of the page is
+/// accessed. This introduces a level of indirection, but it is useful when you
+/// have a sparsely initialised vector where the full size is allocated 
upfront.
+///
+/// As a side effect the elements are initialised later than in a normal 
vector.
+/// On the first access to one of the elements of a given page all, the 
elements
+/// of the page are initialised. This also means that the elements of the page
+/// are initialised beyond the size of the vector.
+///
+/// Similarly on destruction the elements are destroyed only when the page is
+/// not needed anymore, delaying invoking the destructor of the elements.
+///
+/// Notice that this does not have iterators, because if you have iterators it
+/// probably means you are going to touch all the memory in any case, so better
+/// use a std::vector in the first place.
+template  class PagedVector {
+  static_assert(PageSize > 1, "PageSize must be greater than 0. Most likely "
+  "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.

vgvassilev wrote:

Maybe we should transform these into doxygen-style comments too. We do that 
often in the llvm codebase.

https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -1625,6 +1625,39 @@ SmallVector has grown a few other minor advantages over 
std::vector, causing
and is no longer "private to the implementation". A name like
``SmallVectorHeader`` might be more appropriate.
 
+.. _dss_pagedvector:
+
+llvm/ADT/PagedVector.h
+^^
+
+``PagedVector`` is a random access container that allocates
+(PageSize) elements of type Type when the first element of a page is accessed
+via the ``operator[]``.  This is useful for the case in which the number of
+elements is known in advance and their actual initialization is expensive and
+sparse so that it's only done lazily when the element is accessed. When the
+number of used pages is small significant memory savings can be achieved.

vgvassilev wrote:

```suggestion
via the ``operator[]``.  This is useful for cases where the number of
elements is known in advance; their actual initialization is expensive; and 
they are
sparsely used. This utility uses page-granular lazily initialization when the 
element is accessed. When the
number of used pages is small significant memory savings can be achieved.
```

Something along these lines. I am not sure if I understand the last sentence. 
Do you mean that if the page size is small? Or you mean if the usage is sparse 
then we can achieve savings. If the latter, I guess we say that in the previous 
sentence.

https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -0,0 +1,323 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/iterator_range.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+
+namespace llvm {
+/// A vector that allocates memory in pages.
+///
+/// Order is kept, but memory is allocated only when one element of the page is
+/// accessed. This introduces a level of indirection, but it is useful when you
+/// have a sparsely initialised vector where the full size is allocated 
upfront.
+///
+/// As a side effect the elements are initialised later than in a normal 
vector.
+/// On the first access to one of the elements of a given page all, the 
elements
+/// of the page are initialised. This also means that the elements of the page
+/// are initialised beyond the size of the vector.
+///
+/// Similarly on destruction the elements are destroyed only when the page is
+/// not needed anymore, delaying invoking the destructor of the elements.
+///
+/// Notice that this does not have iterators, because if you have iterators it
+/// probably means you are going to touch all the memory in any case, so better
+/// use a std::vector in the first place.
+template  class PagedVector {
+  static_assert(PageSize > 1, "PageSize must be greater than 0. Most likely "
+  "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.
+  size_t Size = 0;
+
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable SmallVector PageToDataPtrs;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The order of
+  // the elements however depends on the order of access of the pages.
+  PointerIntPair Allocator;
+
+  constexpr static T *InvalidPage = nullptr;
+
+public:
+  using value_type = T;
+
+  /// Default constructor. We build our own allocator and mark it as such with
+  /// `true` in the second pair element.
+  PagedVector() : Allocator(new BumpPtrAllocator, true) {}
+  PagedVector(BumpPtrAllocator *A) : Allocator(A, false) {
+assert(A != nullptr && "Allocator cannot be null");
+  }
+
+  ~PagedVector() {
+clear();
+// If we own the allocator, delete it.
+if (Allocator.getInt())
+  delete Allocator.getPointer();
+  }
+
+  /// Look up an element at position `Index`.
+  /// If the associated page is not filled, it will be filled with default
+  /// constructed elements. If the associated page is filled, return the
+  /// element.
+  T &operator[](size_t Index) const {
+assert(Index < Size);
+assert(Index / PageSize < PageToDataPtrs.size());
+T *&PagePtr = PageToDataPtrs[Index / PageSize];
+// If the page was not yet allocated, allocate it.
+if (PagePtr == InvalidPage) {
+  T *NewPagePtr = Allocator.getPointer()->template Allocate(PageSize);
+  // We need to invoke the default constructor on all the elements of the
+  // page.
+  for (size_t I = 0; I < PageSize; ++I)
+new (NewPagePtr + I) T();
+
+  PagePtr = NewPagePtr;
+}
+// Dereference the element in the page.
+return *((Index % PageSize) + PagePtr);
+  }
+
+  /// Return the capacity of the vector. I.e. the maximum size it can be
+  /// expanded to with the resize method without allocating more pages.
+  [[nodiscard]] size_t capacity() const {
+return PageToDataPtrs.size() * PageSize;
+  }
+
+  /// Return the size of the vector. I.e. the maximum index that can be
+  /// accessed, i.e. the maximum value which was used as argument of the
+  /// resize method.
+  [[nodiscard]] size_t size() const { return Size; }
+
+  /// Resize the vector. Notice that the constructor of the elements will not
+  /// be invoked until an element of a given page is accessed, at which point
+  /// all the elements of the page will be constructed.
+  ///
+  /// If the new size is smaller than the current size, the elements of the
+  /// pages that are not needed anymore will be destroyed, however, elements of
+  /// the last page will not be destroyed.
+  ///
+  /// For these reason the usage of this vector is discouraged if you rely
+  /// on the construction / destructor of the elements to be 

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -1625,6 +1625,39 @@ SmallVector has grown a few other minor advantages over 
std::vector, causing
and is no longer "private to the implementation". A name like
``SmallVectorHeader`` might be more appropriate.
 
+.. _dss_pagedvector:
+
+llvm/ADT/PagedVector.h
+^^
+
+``PagedVector`` is a random access container that allocates
+(PageSize) elements of type Type when the first element of a page is accessed
+via the ``operator[]``.  This is useful for the case in which the number of
+elements is known in advance and their actual initialization is expensive and
+sparse so that it's only done lazily when the element is accessed. When the
+number of used pages is small significant memory savings can be achieved.
+
+The main advantage is that a ``PagedVector`` allows to delay the actual
+allocation of the page until it's needed, at the extra cost of one integer per
+page and one extra indirection when accessing elements with their positional
+index. 
+
+In order to maximise the memory footprint of this container, it's important to

vgvassilev wrote:

```suggestion
In order to minimise the memory footprint of this container, it's important to
```

https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -0,0 +1,322 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+#include 
+
+namespace llvm {
+// A vector that allocates memory in pages.
+// Order is kept, but memory is allocated only when one element of the page is
+// accessed. This introduces a level of indirection, but it is useful when you
+// have a sparsely initialised vector where the full size is allocated upfront
+// with the default constructor and elements are initialised later, on first
+// access.
+//
+// Notice that this does not have iterators, because if you
+// have iterators it probably means you are going to touch
+// all the memory in any case, so better use a std::vector in
+// the first place.
+//
+// Pages are allocated in SLAB_SIZE chunks, using the BumpPtrAllocator.
+template 
+class PagedVector {
+  static_assert(PAGE_SIZE > 0, "PAGE_SIZE must be greater than 0. Most likely "
+   "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.
+  std::size_t Size = 0;
+
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable std::vector PageToDataIdx;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The order of
+  // the elements however depends on the order of access of the pages.
+  PointerIntPair Allocator;
+
+  constexpr static uintptr_t InvalidPage = SIZE_MAX;
+
+public:
+  using value_type = T;
+
+  // Default constructor. We build our own allocator.
+  PagedVector() : Allocator(new BumpPtrAllocator, true) {}
+  PagedVector(BumpPtrAllocator *A) : Allocator(A, false) {}
+
+  ~PagedVector() {
+// If we own the allocator, delete it.
+if (Allocator.getInt() == true)
+  delete Allocator.getPointer();
+  }
+
+  // Lookup an element at position i.
+  // If the associated page is not filled, it will be filled with default
+  // constructed elements. If the associated page is filled, return the 
element.
+  T &operator[](std::size_t Index) const {
+assert(Index < Size);
+assert(Index / PAGE_SIZE < PageToDataIdx.size());
+uintptr_t &PagePtr = PageToDataIdx[Index / PAGE_SIZE];
+// If the page was not yet allocated, allocate it.
+if (PagePtr == InvalidPage) {
+  T *NewPagePtr = Allocator.getPointer()->template Allocate(PAGE_SIZE);
+  // We need to invoke the default constructor on all the elements of the
+  // page.
+  for (std::size_t I = 0; I < PAGE_SIZE; ++I)
+new (NewPagePtr + I) T();
+
+  PagePtr = reinterpret_cast(NewPagePtr);
+}
+// Dereference the element in the page.
+return *((Index % PAGE_SIZE) + reinterpret_cast(PagePtr));
+  }
+
+  // Return the capacity of the vector. I.e. the maximum size it can be 
expanded
+  // to with the resize method without allocating more pages.
+  [[nodiscard]] std::size_t capacity() const {
+return PageToDataIdx.size() * PAGE_SIZE;
+  }
+
+  // Return the size of the vector. I.e. the maximum index that can be
+  // accessed, i.e. the maximum value which was used as argument of the
+  // resize method.
+  [[nodiscard]] std::size_t size() const { return Size; }
+
+  // Expands the vector to the given NewSize number of elements.
+  // If the vector was smaller, allocates new pages as needed.
+  // It should be called only with NewSize >= Size.
+  void resize(std::size_t NewSize) {
+// Handle shrink case: delete the pages and update the size.
+if (NewSize < Size) {
+  std::size_t NewLastPage = (NewSize - 1) / PAGE_SIZE;
+  for (std::size_t I = NewLastPage + 1; I < PageToDataIdx.size(); ++I) {
+uintptr_t PagePtr = PageToDataIdx[I];
+if (PagePtr == InvalidPage)
+  continue;
+T *Page = reinterpret_cast(PagePtr);
+// We need to invoke the destructor on all the elements of the page.
+for (std::size_t J = 0; J < PAGE_SIZE; ++J)
+  Page[J].~T();
+Allocator.getPointer()->Deallocate(Page);
+  }
+  // Delete the extra ones in the new last page.
+  uintptr_t PagePtr = PageToDataIdx[NewLastPage];
+  if (PagePtr != InvalidPage) {
+T *Page = reinterpret_cast(PagePtr)

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -0,0 +1,323 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/iterator_range.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+
+namespace llvm {
+/// A vector that allocates memory in pages.
+///
+/// Order is kept, but memory is allocated only when one element of the page is
+/// accessed. This introduces a level of indirection, but it is useful when you
+/// have a sparsely initialised vector where the full size is allocated 
upfront.
+///
+/// As a side effect the elements are initialised later than in a normal 
vector.
+/// On the first access to one of the elements of a given page all, the 
elements
+/// of the page are initialised. This also means that the elements of the page
+/// are initialised beyond the size of the vector.
+///
+/// Similarly on destruction the elements are destroyed only when the page is
+/// not needed anymore, delaying invoking the destructor of the elements.
+///
+/// Notice that this does not have iterators, because if you have iterators it
+/// probably means you are going to touch all the memory in any case, so better
+/// use a std::vector in the first place.
+template  class PagedVector {
+  static_assert(PageSize > 1, "PageSize must be greater than 0. Most likely "
+  "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.
+  size_t Size = 0;
+
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable SmallVector PageToDataPtrs;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The order of
+  // the elements however depends on the order of access of the pages.
+  PointerIntPair Allocator;
+
+  constexpr static T *InvalidPage = nullptr;
+
+public:
+  using value_type = T;
+
+  /// Default constructor. We build our own allocator and mark it as such with
+  /// `true` in the second pair element.
+  PagedVector() : Allocator(new BumpPtrAllocator, true) {}
+  PagedVector(BumpPtrAllocator *A) : Allocator(A, false) {
+assert(A != nullptr && "Allocator cannot be null");
+  }
+
+  ~PagedVector() {
+clear();
+// If we own the allocator, delete it.
+if (Allocator.getInt())
+  delete Allocator.getPointer();
+  }
+
+  /// Look up an element at position `Index`.
+  /// If the associated page is not filled, it will be filled with default
+  /// constructed elements. If the associated page is filled, return the
+  /// element.
+  T &operator[](size_t Index) const {
+assert(Index < Size);
+assert(Index / PageSize < PageToDataPtrs.size());
+T *&PagePtr = PageToDataPtrs[Index / PageSize];
+// If the page was not yet allocated, allocate it.
+if (PagePtr == InvalidPage) {
+  T *NewPagePtr = Allocator.getPointer()->template Allocate(PageSize);
+  // We need to invoke the default constructor on all the elements of the
+  // page.
+  for (size_t I = 0; I < PageSize; ++I)
+new (NewPagePtr + I) T();
+
+  PagePtr = NewPagePtr;
+}
+// Dereference the element in the page.
+return *((Index % PageSize) + PagePtr);
+  }
+
+  /// Return the capacity of the vector. I.e. the maximum size it can be
+  /// expanded to with the resize method without allocating more pages.
+  [[nodiscard]] size_t capacity() const {
+return PageToDataPtrs.size() * PageSize;
+  }
+
+  /// Return the size of the vector. I.e. the maximum index that can be
+  /// accessed, i.e. the maximum value which was used as argument of the
+  /// resize method.
+  [[nodiscard]] size_t size() const { return Size; }
+
+  /// Resize the vector. Notice that the constructor of the elements will not
+  /// be invoked until an element of a given page is accessed, at which point
+  /// all the elements of the page will be constructed.
+  ///
+  /// If the new size is smaller than the current size, the elements of the
+  /// pages that are not needed anymore will be destroyed, however, elements of
+  /// the last page will not be destroyed.
+  ///
+  /// For these reason the usage of this vector is discouraged if you rely
+  /// on the construction / destructor of the elements to be 

[clang] [clang-format] Correctly annotate return type of function pointer (PR #66893)

[Owen Pan via cfe-commits]

owenca wrote:

> Pinging @owenca.

Waiting for one of @mydeveloperday @HazardyKnusperkeks @rymiel 

https://github.com/llvm/llvm-project/pull/66893
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Freddy Ye via cfe-commits]

https://github.com/FreddyLeaf updated 
https://github.com/llvm/llvm-project/pull/67410

>From 3ae0fa2592d3a11dd084d3aefb54c26819016781 Mon Sep 17 00:00:00 2001
From: Freddy Ye 
Date: Tue, 26 Sep 2023 16:44:01 +0800
Subject: [PATCH 1/3] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from
 avx into sse/sse2

---
 clang/include/clang/Basic/BuiltinsX86.def|   8 +-
 clang/lib/Headers/avxintrin.h| 272 -
 clang/lib/Headers/emmintrin.h| 119 ++
 clang/lib/Headers/xmmintrin.h| 152 +++
 clang/test/CodeGen/X86/avx-builtins.c| 396 ---
 clang/test/CodeGen/X86/sse-builtins.c| 198 ++
 clang/test/CodeGen/X86/sse2-builtins.c   | 198 ++
 clang/test/CodeGen/target-features-error-2.c |   4 +-
 8 files changed, 673 insertions(+), 674 deletions(-)

diff --git a/clang/include/clang/Basic/BuiltinsX86.def 
b/clang/include/clang/Basic/BuiltinsX86.def
index e4802f8ab1c1562..6bbbc4bbdd75c68 100644
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ b/clang/include/clang/Basic/BuiltinsX86.def
@@ -233,6 +233,8 @@ TARGET_BUILTIN(__builtin_ia32_minps, "V4fV4fV4f", 
"ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_maxps, "V4fV4fV4f", "ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_minss, "V4fV4fV4f", "ncV:128:", "sse")
 TARGET_BUILTIN(__builtin_ia32_maxss, "V4fV4fV4f", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "sse")
 
 TARGET_BUILTIN(__builtin_ia32_cmpeqpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpltpd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -250,6 +252,8 @@ TARGET_BUILTIN(__builtin_ia32_cmpneqsd, "V2dV2dV2d", 
"ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpnltsd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpnlesd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_cmpordsd, "V2dV2dV2d", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_minpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_maxpd, "V2dV2dV2d", "ncV:128:", "sse2")
 TARGET_BUILTIN(__builtin_ia32_minsd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -469,12 +473,8 @@ TARGET_BUILTIN(__builtin_ia32_blendvps256, "V8fV8fV8fV8f", 
"ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_shufpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_shufps256, "V8fV8fV8fIi", "ncV:256:", "avx")
 TARGET_BUILTIN(__builtin_ia32_dpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_cmppd256, "V4dV4dV4dIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_cmpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_pd256, "V2dV4dIi", "ncV:256:", 
"avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_ps256, "V4fV8fIi", "ncV:256:", 
"avx")
 TARGET_BUILTIN(__builtin_ia32_vextractf128_si256, "V4iV8iIi", "ncV:256:", 
"avx")
diff --git a/clang/lib/Headers/avxintrin.h b/clang/lib/Headers/avxintrin.h
index b796bb773ec11f0..afbf26403f201b6 100644
--- a/clang/lib/Headers/avxintrin.h
+++ b/clang/lib/Headers/avxintrin.h
@@ -1569,160 +1569,6 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
   ((__m256d)__builtin_ia32_shufpd256((__v4df)(__m256d)(a), \
  (__v4df)(__m256d)(b), (int)(mask)))
 
-/* Compare */
-#define _CMP_EQ_OQ0x00 /* Equal (ordered, non-signaling)  */
-#define _CMP_LT_OS0x01 /* Less-than (ordered, signaling)  */
-#define _CMP_LE_OS0x02 /* Less-than-or-equal (ordered, signaling)  */
-#define _CMP_UNORD_Q  0x03 /* Unordered (non-signaling)  */
-#define _CMP_NEQ_UQ   0x04 /* Not-equal (unordered, non-signaling)  */
-#define _CMP_NLT_US   0x05 /* Not-less-than (unordered, signaling)  */
-#define _CMP_NLE_US   0x06 /* Not-less-than-or-equal (unordered, signaling)  */
-#define _CMP_ORD_Q0x07 /* Ordered (non-signaling)   */
-#define _CMP_EQ_UQ0x08 /* Equal (unordered, non-signaling)  */
-#define _CMP_NGE_US   0x09 /* Not-greater-than-or-equal (unordered, signaling) 
 */
-#define _CMP_NGT_US   0x0a /* Not-greater-than (unordered, signaling)  */
-#define _CMP_FALSE_OQ 0x0b /* False (ordered, non-signaling)  */
-#define _CMP_NEQ_OQ   0x0c /* Not-equal (ordered, non-signaling)  */
-#define _CMP_GE_OS0x0d /* Greater-than-or-equal (ordered, signaling)  */
-#define _CMP_GT_OS0x0e /* Greater-than (ordered, signaling)  */
-#define _CMP_TRUE_UQ  0x0f /* True (unordered, non-signaling)  */
-#define _CMP_EQ_OS0x10 /* Equal (ordered, signaling)

[clang] [PowerPC] Support local-dynamic TLS relocation on AIX (PR #66316)

[Felix via cfe-commits]

orcguru wrote:

> The code formatting check fails.

That is intentional: 
https://github.com/llvm/llvm-project/pull/66316#discussion_r1333195506

https://github.com/llvm/llvm-project/pull/66316
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang-analysis]Fix false positive in mutation check when using pointer to member function (PR #66846)

[via cfe-commits]

https://github.com/mzyKi resolved 
https://github.com/llvm/llvm-project/pull/66846
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang-tools-extra] [clang-analysis]Fix false positive in mutation check when using pointer to member function (PR #66846)

[via cfe-commits]

https://github.com/mzyKi resolved 
https://github.com/llvm/llvm-project/pull/66846
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Simon Pilgrim via cfe-commits]

https://github.com/RKSimon edited 
https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Simon Pilgrim via cfe-commits]

@@ -4742,6 +4742,125 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS 
_mm_castsi128_pd(__m128i __a) {
   return (__m128d)__a;
 }
 
+/// Compares each of the corresponding double-precision values of two
+///128-bit vectors of [2 x double], using the operation specified by the
+///immediate integer operand.
+///
+///Returns a [2 x double] vector consisting of two doubles corresponding to
+///the two comparison results: zero if the comparison is false, and all 1's
+///if the comparison is true.
+///
+/// \headerfile 
+///
+/// \code
+/// __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the  VCMPPD  instruction.

RKSimon wrote:

(V)CMPPD ?

https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Simon Pilgrim via cfe-commits]

@@ -4742,6 +4742,125 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS 
_mm_castsi128_pd(__m128i __a) {
   return (__m128d)__a;
 }
 
+/// Compares each of the corresponding double-precision values of two
+///128-bit vectors of [2 x double], using the operation specified by the
+///immediate integer operand.
+///
+///Returns a [2 x double] vector consisting of two doubles corresponding to
+///the two comparison results: zero if the comparison is false, and all 1's
+///if the comparison is true.
+///
+/// \headerfile 
+///
+/// \code
+/// __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the  VCMPPD  instruction.
+///
+/// \param a
+///A 128-bit vector of [2 x double].
+/// \param b
+///A 128-bit vector of [2 x double].
+/// \param c
+///An immediate integer operand, with bits [4:0] specifying which 
comparison

RKSimon wrote:

What should we do about non-AVX targets only supporting 3-bit imm modes (vs AVX 
5-bit imm nodes)?

https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Simon Pilgrim via cfe-commits]

https://github.com/RKSimon commented:

No strong objections, but I think the documentation needs to be tweaked to 
better explain SSE vs AVX handling.

https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Simon Pilgrim via cfe-commits]

https://github.com/RKSimon edited 
https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] 31631d3 - [X86][FP16] Add missing handling for FP16 constrained cmp intrinsics (#67400)

[via cfe-commits]

Author: Phoebe Wang
Date: 2023-09-26T19:27:57+08:00
New Revision: 31631d307fe065966c512f8383948fafa23a5bd8

URL: 
https://github.com/llvm/llvm-project/commit/31631d307fe065966c512f8383948fafa23a5bd8
DIFF: 
https://github.com/llvm/llvm-project/commit/31631d307fe065966c512f8383948fafa23a5bd8.diff

LOG: [X86][FP16] Add missing handling for FP16 constrained cmp intrinsics 
(#67400)

Added: 
clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c

Modified: 
clang/lib/CodeGen/CGBuiltin.cpp

Removed: 




diff  --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index d14cf0dccb09982..04c0325c7fd038b 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -15594,6 +15594,15 @@ Value *CodeGenFunction::EmitX86BuiltinExpr(unsigned 
BuiltinID,
   case X86::BI__builtin_ia32_cmppd256:
 IID = Intrinsic::x86_avx_cmp_pd_256;
 break;
+  case X86::BI__builtin_ia32_cmpph128_mask:
+IID = Intrinsic::x86_avx512fp16_mask_cmp_ph_128;
+break;
+  case X86::BI__builtin_ia32_cmpph256_mask:
+IID = Intrinsic::x86_avx512fp16_mask_cmp_ph_256;
+break;
+  case X86::BI__builtin_ia32_cmpph512_mask:
+IID = Intrinsic::x86_avx512fp16_mask_cmp_ph_512;
+break;
   case X86::BI__builtin_ia32_cmpps512_mask:
 IID = Intrinsic::x86_avx512_mask_cmp_ps_512;
 break;

diff  --git a/clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c 
b/clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c
new file mode 100644
index 000..1a164ff57fda1b1
--- /dev/null
+++ b/clang/test/CodeGen/X86/avx512fp16-builtins-constrained-cmp.c
@@ -0,0 +1,399 @@
+// RUN: %clang_cc1 -flax-vector-conversions=none -ffreestanding %s 
-triple=x86_64-apple-darwin -target-feature +avx512fp16 -emit-llvm 
-ffp-exception-behavior=strict -o - -Wall -Werror | FileCheck %s
+
+#include 
+
+__mmask32 test_mm512_cmp_round_ph_mask(__m512h a, __m512h b) {
+  // CHECK-LABEL: @test_mm512_cmp_round_ph_mask
+  // CHECK: call <32 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.512(<32 x half> 
%{{.*}}, <32 x half> %{{.*}}, i32 0, <32 x i1> {{.*}}, i32 8)
+  return _mm512_cmp_round_ph_mask(a, b, _CMP_EQ_OQ, _MM_FROUND_NO_EXC);
+}
+
+__mmask32 test_mm512_mask_cmp_round_ph_mask(__mmask32 m, __m512h a, __m512h b) 
{
+  // CHECK-LABEL: @test_mm512_mask_cmp_round_ph_mask
+  // CHECK: call <32 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.512(<32 x half> 
%{{.*}}, <32 x half> %{{.*}}, i32 0, <32 x i1> {{.*}}, i32 8)
+  return _mm512_mask_cmp_round_ph_mask(m, a, b, _CMP_EQ_OQ, _MM_FROUND_NO_EXC);
+}
+
+__mmask8 test_mm_cmp_ph_mask_eq_oq(__m128h a, __m128h b) {
+  // CHECK-LABEL: @test_mm_cmp_ph_mask_eq_oq
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 0, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_EQ_OQ);
+}
+
+__mmask8 test_mm_cmp_ph_mask_lt_os(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_lt_os
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 1, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_LT_OS);
+}
+
+__mmask8 test_mm_cmp_ph_mask_le_os(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_le_os
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 2, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_LE_OS);
+}
+
+__mmask8 test_mm_cmp_ph_mask_unord_q(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_unord_q
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 3, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_UNORD_Q);
+}
+
+__mmask8 test_mm_cmp_ph_mask_neq_uq(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_neq_uq
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 4, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_NEQ_UQ);
+}
+
+__mmask8 test_mm_cmp_ph_mask_nlt_us(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_nlt_us
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 5, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_NLT_US);
+}
+
+__mmask8 test_mm_cmp_ph_mask_nle_us(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_nle_us
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 6, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_NLE_US);
+}
+
+__mmask8 test_mm_cmp_ph_mask_ord_q(__m128h a, __m128h b) {
+  // CHECK-LABEL: test_mm_cmp_ph_mask_ord_q
+  // CHECK: call <8 x i1> @llvm.x86.avx512fp16.mask.cmp.ph.128(<8 x half> 
%{{.*}}, <8 x half> %{{.*}}, i32 7, <8 x i1> {{.*}})
+  return _mm_cmp_ph_mask(a, b, _CMP_ORD_Q);
+}
+
+__mmask8 tes

[clang] [X86][FP16] Add missing handling for FP16 constrained cmp intrinsics (PR #67400)

[Phoebe Wang via cfe-commits]

https://github.com/phoebewang closed 
https://github.com/llvm/llvm-project/pull/67400
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf resolved https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From 4b02b55cccd73b0fe3efea52054735fca6e6c159 Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH 1/2] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 323 +
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 712 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf resolved https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From 4b02b55cccd73b0fe3efea52054735fca6e6c159 Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH 1/3] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 323 +
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 712 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf resolved https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From 4b02b55cccd73b0fe3efea52054735fca6e6c159 Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH 1/4] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 323 +
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 712 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf resolved https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From 4b02b55cccd73b0fe3efea52054735fca6e6c159 Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH 1/5] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 323 +
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 712 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf resolved https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From 4b02b55cccd73b0fe3efea52054735fca6e6c159 Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH 1/6] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 323 +
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 712 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

@@ -0,0 +1,323 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/iterator_range.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+
+namespace llvm {
+/// A vector that allocates memory in pages.
+///
+/// Order is kept, but memory is allocated only when one element of the page is
+/// accessed. This introduces a level of indirection, but it is useful when you
+/// have a sparsely initialised vector where the full size is allocated 
upfront.
+///
+/// As a side effect the elements are initialised later than in a normal 
vector.
+/// On the first access to one of the elements of a given page all, the 
elements
+/// of the page are initialised. This also means that the elements of the page
+/// are initialised beyond the size of the vector.
+///
+/// Similarly on destruction the elements are destroyed only when the page is
+/// not needed anymore, delaying invoking the destructor of the elements.
+///
+/// Notice that this does not have iterators, because if you have iterators it
+/// probably means you are going to touch all the memory in any case, so better
+/// use a std::vector in the first place.
+template  class PagedVector {
+  static_assert(PageSize > 1, "PageSize must be greater than 0. Most likely "
+  "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.
+  size_t Size = 0;
+
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable SmallVector PageToDataPtrs;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The order of
+  // the elements however depends on the order of access of the pages.
+  PointerIntPair Allocator;
+
+  constexpr static T *InvalidPage = nullptr;
+
+public:
+  using value_type = T;
+
+  /// Default constructor. We build our own allocator and mark it as such with
+  /// `true` in the second pair element.
+  PagedVector() : Allocator(new BumpPtrAllocator, true) {}
+  PagedVector(BumpPtrAllocator *A) : Allocator(A, false) {
+assert(A != nullptr && "Allocator cannot be null");
+  }
+
+  ~PagedVector() {
+clear();
+// If we own the allocator, delete it.
+if (Allocator.getInt())
+  delete Allocator.getPointer();
+  }
+
+  /// Look up an element at position `Index`.
+  /// If the associated page is not filled, it will be filled with default
+  /// constructed elements. If the associated page is filled, return the
+  /// element.
+  T &operator[](size_t Index) const {
+assert(Index < Size);
+assert(Index / PageSize < PageToDataPtrs.size());
+T *&PagePtr = PageToDataPtrs[Index / PageSize];
+// If the page was not yet allocated, allocate it.
+if (PagePtr == InvalidPage) {
+  T *NewPagePtr = Allocator.getPointer()->template Allocate(PageSize);
+  // We need to invoke the default constructor on all the elements of the
+  // page.
+  for (size_t I = 0; I < PageSize; ++I)
+new (NewPagePtr + I) T();
+
+  PagePtr = NewPagePtr;
+}
+// Dereference the element in the page.
+return *((Index % PageSize) + PagePtr);
+  }
+
+  /// Return the capacity of the vector. I.e. the maximum size it can be
+  /// expanded to with the resize method without allocating more pages.
+  [[nodiscard]] size_t capacity() const {
+return PageToDataPtrs.size() * PageSize;
+  }
+
+  /// Return the size of the vector. I.e. the maximum index that can be
+  /// accessed, i.e. the maximum value which was used as argument of the
+  /// resize method.
+  [[nodiscard]] size_t size() const { return Size; }
+
+  /// Resize the vector. Notice that the constructor of the elements will not
+  /// be invoked until an element of a given page is accessed, at which point
+  /// all the elements of the page will be constructed.
+  ///
+  /// If the new size is smaller than the current size, the elements of the
+  /// pages that are not needed anymore will be destroyed, however, elements of
+  /// the last page will not be destroyed.
+  ///
+  /// For these reason the usage of this vector is discouraged if you rely
+  /// on the construction / destructor of the elements to be 

[clang] [Driver] -include: deprecate probing .gch (PR #67084)

[Nico Weber via cfe-commits]

nico wrote:

Can we add a dedicated `-Wdeprecated-gch` diag group or something for this? 
This is a very disruptive change, and it'd be good if we didn't have to disable 
all deprecation warnings while we rearrange things.

Reading the linked bugs, it's not clear that removing this would've helped all 
these issues all that much and it introduces a bunch of toil. So maybe we don't 
want to do this at all? But if we do, please put it behind a dedicated warning 
flag.

https://github.com/llvm/llvm-project/pull/67084
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Freddy Ye via cfe-commits]

FreddyLeaf wrote:

> No strong objections, but I think the documentation needs to be tweaked to 
> better explain SSE vs AVX handling.

I have an idea to show two results for the `_mm_cmp_sd`, one's CPUID is SSE2, 
the other one's is AVX.

https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang analysis][NFCI] Preparatory work for D153131. (PR #67420)

[Clement Courbet via cfe-commits]

https://github.com/legrosbuffle created 
https://github.com/llvm/llvm-project/pull/67420

This was ported over from phabricator.

Review https://reviews.llvm.org/D153131.

>From 3581fc00f690194ac4bac631311ecdb18593b7ba Mon Sep 17 00:00:00 2001
From: Clement Courbet 
Date: Tue, 26 Sep 2023 14:02:44 +0200
Subject: [PATCH] [clang analysis][NFCI] Preparatory work for D153131.

This was ported over from phabricator.i
Review https://reviews.llvm.org/D153131.
---
 clang/lib/Analysis/ThreadSafety.cpp | 133 +++-
 1 file changed, 72 insertions(+), 61 deletions(-)

diff --git a/clang/lib/Analysis/ThreadSafety.cpp 
b/clang/lib/Analysis/ThreadSafety.cpp
index 3e6ceb7d54c427a..f160cf4d013c78d 100644
--- a/clang/lib/Analysis/ThreadSafety.cpp
+++ b/clang/lib/Analysis/ThreadSafety.cpp
@@ -1015,6 +1015,19 @@ class ThreadSafetyAnalyzer {
 
   BeforeSet *GlobalBeforeSet;
 
+  void warnIfMutexNotHeld(const FactSet &FSet, const NamedDecl *D,
+  const Expr *Exp, AccessKind AK, Expr *MutexExp,
+  ProtectedOperationKind POK, til::LiteralPtr *Self,
+  SourceLocation Loc);
+  void warnIfMutexHeld(const FactSet &FSet, const NamedDecl *D, const Expr 
*Exp,
+   Expr *MutexExp, til::LiteralPtr *Self,
+   SourceLocation Loc);
+
+  void checkAccess(const FactSet &FSet, const Expr *Exp, AccessKind AK,
+   ProtectedOperationKind POK);
+  void checkPtAccess(const FactSet &FSet, const Expr *Exp, AccessKind AK,
+ ProtectedOperationKind POK);
+
 public:
   ThreadSafetyAnalyzer(ThreadSafetyHandler &H, BeforeSet* Bset)
   : Arena(&Bpa), SxBuilder(Arena), Handler(H), GlobalBeforeSet(Bset) {}
@@ -1534,16 +1547,15 @@ class BuildLockset : public 
ConstStmtVisitor {
   unsigned CtxIndex;
 
   // helper functions
-  void warnIfMutexNotHeld(const NamedDecl *D, const Expr *Exp, AccessKind AK,
-  Expr *MutexExp, ProtectedOperationKind POK,
-  til::LiteralPtr *Self, SourceLocation Loc);
-  void warnIfMutexHeld(const NamedDecl *D, const Expr *Exp, Expr *MutexExp,
-   til::LiteralPtr *Self, SourceLocation Loc);
 
   void checkAccess(const Expr *Exp, AccessKind AK,
-   ProtectedOperationKind POK = POK_VarAccess);
+   ProtectedOperationKind POK = POK_VarAccess) {
+Analyzer->checkAccess(FSet, Exp, AK, POK);
+  }
   void checkPtAccess(const Expr *Exp, AccessKind AK,
- ProtectedOperationKind POK = POK_VarAccess);
+ ProtectedOperationKind POK = POK_VarAccess) {
+Analyzer->checkPtAccess(FSet, Exp, AK, POK);
+  }
 
   void handleCall(const Expr *Exp, const NamedDecl *D,
   til::LiteralPtr *Self = nullptr,
@@ -1571,17 +1583,14 @@ class BuildLockset : public 
ConstStmtVisitor {
 
 /// Warn if the LSet does not contain a lock sufficient to protect access
 /// of at least the passed in AccessKind.
-void BuildLockset::warnIfMutexNotHeld(const NamedDecl *D, const Expr *Exp,
-  AccessKind AK, Expr *MutexExp,
-  ProtectedOperationKind POK,
-  til::LiteralPtr *Self,
-  SourceLocation Loc) {
+void ThreadSafetyAnalyzer::warnIfMutexNotHeld(
+const FactSet &FSet, const NamedDecl *D, const Expr *Exp, AccessKind AK,
+Expr *MutexExp, ProtectedOperationKind POK, til::LiteralPtr *Self,
+SourceLocation Loc) {
   LockKind LK = getLockKindFromAccessKind(AK);
-
-  CapabilityExpr Cp =
-  Analyzer->SxBuilder.translateAttrExpr(MutexExp, D, Exp, Self);
+  CapabilityExpr Cp = SxBuilder.translateAttrExpr(MutexExp, D, Exp, Self);
   if (Cp.isInvalid()) {
-warnInvalidLock(Analyzer->Handler, MutexExp, D, Exp, Cp.getKind());
+warnInvalidLock(Handler, MutexExp, D, Exp, Cp.getKind());
 return;
   } else if (Cp.shouldIgnore()) {
 return;
@@ -1589,68 +1598,67 @@ void BuildLockset::warnIfMutexNotHeld(const NamedDecl 
*D, const Expr *Exp,
 
   if (Cp.negative()) {
 // Negative capabilities act like locks excluded
-const FactEntry *LDat = FSet.findLock(Analyzer->FactMan, !Cp);
+const FactEntry *LDat = FSet.findLock(FactMan, !Cp);
 if (LDat) {
-  Analyzer->Handler.handleFunExcludesLock(
-  Cp.getKind(), D->getNameAsString(), (!Cp).toString(), Loc);
-  return;
+Handler.handleFunExcludesLock(Cp.getKind(), D->getNameAsString(),
+  (!Cp).toString(), Loc);
+return;
 }
 
 // If this does not refer to a negative capability in the same class,
 // then stop here.
-if (!Analyzer->inCurrentScope(Cp))
-  return;
+if (!inCurrentScope(Cp))
+return;
 
 // Otherwise the negative requirement must be propagated to the caller.
-LDat = FSet.findLock(Analyzer->FactMan, Cp);
+LDat = FSet.findLo

[clang] [clang analysis][NFCI] Preparatory work for D153131. (PR #67420)

[via cfe-commits]

llvmbot wrote:




@llvm/pr-subscribers-clang


Changes

This was ported over from phabricator.

Review https://reviews.llvm.org/D153131.

---
Full diff: https://github.com/llvm/llvm-project/pull/67420.diff


1 Files Affected:

- (modified) clang/lib/Analysis/ThreadSafety.cpp (+72-61) 


``diff
diff --git a/clang/lib/Analysis/ThreadSafety.cpp 
b/clang/lib/Analysis/ThreadSafety.cpp
index 3e6ceb7d54c427a..f160cf4d013c78d 100644
--- a/clang/lib/Analysis/ThreadSafety.cpp
+++ b/clang/lib/Analysis/ThreadSafety.cpp
@@ -1015,6 +1015,19 @@ class ThreadSafetyAnalyzer {
 
   BeforeSet *GlobalBeforeSet;
 
+  void warnIfMutexNotHeld(const FactSet &FSet, const NamedDecl *D,
+  const Expr *Exp, AccessKind AK, Expr *MutexExp,
+  ProtectedOperationKind POK, til::LiteralPtr *Self,
+  SourceLocation Loc);
+  void warnIfMutexHeld(const FactSet &FSet, const NamedDecl *D, const Expr 
*Exp,
+   Expr *MutexExp, til::LiteralPtr *Self,
+   SourceLocation Loc);
+
+  void checkAccess(const FactSet &FSet, const Expr *Exp, AccessKind AK,
+   ProtectedOperationKind POK);
+  void checkPtAccess(const FactSet &FSet, const Expr *Exp, AccessKind AK,
+ ProtectedOperationKind POK);
+
 public:
   ThreadSafetyAnalyzer(ThreadSafetyHandler &H, BeforeSet* Bset)
   : Arena(&Bpa), SxBuilder(Arena), Handler(H), GlobalBeforeSet(Bset) {}
@@ -1534,16 +1547,15 @@ class BuildLockset : public 
ConstStmtVisitor {
   unsigned CtxIndex;
 
   // helper functions
-  void warnIfMutexNotHeld(const NamedDecl *D, const Expr *Exp, AccessKind AK,
-  Expr *MutexExp, ProtectedOperationKind POK,
-  til::LiteralPtr *Self, SourceLocation Loc);
-  void warnIfMutexHeld(const NamedDecl *D, const Expr *Exp, Expr *MutexExp,
-   til::LiteralPtr *Self, SourceLocation Loc);
 
   void checkAccess(const Expr *Exp, AccessKind AK,
-   ProtectedOperationKind POK = POK_VarAccess);
+   ProtectedOperationKind POK = POK_VarAccess) {
+Analyzer->checkAccess(FSet, Exp, AK, POK);
+  }
   void checkPtAccess(const Expr *Exp, AccessKind AK,
- ProtectedOperationKind POK = POK_VarAccess);
+ ProtectedOperationKind POK = POK_VarAccess) {
+Analyzer->checkPtAccess(FSet, Exp, AK, POK);
+  }
 
   void handleCall(const Expr *Exp, const NamedDecl *D,
   til::LiteralPtr *Self = nullptr,
@@ -1571,17 +1583,14 @@ class BuildLockset : public 
ConstStmtVisitor {
 
 /// Warn if the LSet does not contain a lock sufficient to protect access
 /// of at least the passed in AccessKind.
-void BuildLockset::warnIfMutexNotHeld(const NamedDecl *D, const Expr *Exp,
-  AccessKind AK, Expr *MutexExp,
-  ProtectedOperationKind POK,
-  til::LiteralPtr *Self,
-  SourceLocation Loc) {
+void ThreadSafetyAnalyzer::warnIfMutexNotHeld(
+const FactSet &FSet, const NamedDecl *D, const Expr *Exp, AccessKind AK,
+Expr *MutexExp, ProtectedOperationKind POK, til::LiteralPtr *Self,
+SourceLocation Loc) {
   LockKind LK = getLockKindFromAccessKind(AK);
-
-  CapabilityExpr Cp =
-  Analyzer->SxBuilder.translateAttrExpr(MutexExp, D, Exp, Self);
+  CapabilityExpr Cp = SxBuilder.translateAttrExpr(MutexExp, D, Exp, Self);
   if (Cp.isInvalid()) {
-warnInvalidLock(Analyzer->Handler, MutexExp, D, Exp, Cp.getKind());
+warnInvalidLock(Handler, MutexExp, D, Exp, Cp.getKind());
 return;
   } else if (Cp.shouldIgnore()) {
 return;
@@ -1589,68 +1598,67 @@ void BuildLockset::warnIfMutexNotHeld(const NamedDecl 
*D, const Expr *Exp,
 
   if (Cp.negative()) {
 // Negative capabilities act like locks excluded
-const FactEntry *LDat = FSet.findLock(Analyzer->FactMan, !Cp);
+const FactEntry *LDat = FSet.findLock(FactMan, !Cp);
 if (LDat) {
-  Analyzer->Handler.handleFunExcludesLock(
-  Cp.getKind(), D->getNameAsString(), (!Cp).toString(), Loc);
-  return;
+Handler.handleFunExcludesLock(Cp.getKind(), D->getNameAsString(),
+  (!Cp).toString(), Loc);
+return;
 }
 
 // If this does not refer to a negative capability in the same class,
 // then stop here.
-if (!Analyzer->inCurrentScope(Cp))
-  return;
+if (!inCurrentScope(Cp))
+return;
 
 // Otherwise the negative requirement must be propagated to the caller.
-LDat = FSet.findLock(Analyzer->FactMan, Cp);
+LDat = FSet.findLock(FactMan, Cp);
 if (!LDat) {
-  Analyzer->Handler.handleNegativeNotHeld(D, Cp.toString(), Loc);
+Handler.handleNegativeNotHeld(D, Cp.toString(), Loc);
 }
 return;
   }
 
-  const FactEntry *LDat = FSet.findLockUniv(Analyzer->FactMan, Cp);
+  const FactEnt

[clang] 39f36d5 - [Driver] Give warn_drv_include_probe_gch a dedicated flag

[Hans Wennborg via cfe-commits]

Author: Hans Wennborg
Date: 2023-09-26T14:14:37+02:00
New Revision: 39f36d5d5c88f49eaa946a3ee9ef887c04275a33

URL: 
https://github.com/llvm/llvm-project/commit/39f36d5d5c88f49eaa946a3ee9ef887c04275a33
DIFF: 
https://github.com/llvm/llvm-project/commit/39f36d5d5c88f49eaa946a3ee9ef887c04275a33.diff

LOG: [Driver] Give warn_drv_include_probe_gch a dedicated flag

See discussion on https://github.com/llvm/llvm-project/pull/67084

Added: 


Modified: 
clang/include/clang/Basic/DiagnosticDriverKinds.td
clang/include/clang/Basic/DiagnosticGroups.td

Removed: 




diff  --git a/clang/include/clang/Basic/DiagnosticDriverKinds.td 
b/clang/include/clang/Basic/DiagnosticDriverKinds.td
index 839e7930b51efcd..30d8f05dc92f9f0 100644
--- a/clang/include/clang/Basic/DiagnosticDriverKinds.td
+++ b/clang/include/clang/Basic/DiagnosticDriverKinds.td
@@ -433,7 +433,7 @@ def warn_drv_treating_input_as_cxx : Warning<
   InGroup;
 def warn_drv_include_probe_gch : Warning<
   "'%0' probing .gch is deprecated. Use '-include-pch %1' or switch to .pch 
instead">,
-  InGroup;
+  InGroup;
 def warn_drv_pch_not_first_include : Warning<
   "precompiled header '%0' was ignored because '%1' is not first '-include'">;
 def warn_missing_sysroot : Warning<"no such sysroot directory: '%0'">,

diff  --git a/clang/include/clang/Basic/DiagnosticGroups.td 
b/clang/include/clang/Basic/DiagnosticGroups.td
index afbf0f0ed22e57f..0b09c002191848a 100644
--- a/clang/include/clang/Basic/DiagnosticGroups.td
+++ b/clang/include/clang/Basic/DiagnosticGroups.td
@@ -209,6 +209,7 @@ def DeprecatedWritableStr : 
DiagGroup<"deprecated-writable-strings",
   [CXX11CompatDeprecatedWritableStr]>;
 def DeprecatedPragma : DiagGroup<"deprecated-pragma">;
 def DeprecatedType : DiagGroup<"deprecated-type">;
+def DeprecatedIncludeGch : DiagGroup<"deprecated-include-gch">;
 // FIXME: Why is DeprecatedImplementations not in this group?
 def Deprecated : DiagGroup<"deprecated", [DeprecatedAnonEnumEnumConversion,
   DeprecatedArrayCompare,
@@ -232,6 +233,7 @@ def Deprecated : DiagGroup<"deprecated", 
[DeprecatedAnonEnumEnumConversion,
   DeprecatedVolatile,
   DeprecatedWritableStr,
   
DeprecatedRedundantConstexprStaticDef,
+  DeprecatedIncludeGch
   ]>,
  DiagCategory<"Deprecations">;
 



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang-format] Fix a bug in NamespaceEndCommentsFixer (PR #67422)

[Owen Pan via cfe-commits]

https://github.com/owenca created 
https://github.com/llvm/llvm-project/pull/67422

Fixed #67407.

>From 1c74983f9d2a16ab26d64624448df89e5fff87c4 Mon Sep 17 00:00:00 2001
From: Owen Pan 
Date: Tue, 26 Sep 2023 05:17:25 -0700
Subject: [PATCH] [clang-format] Fix a bug in NamespaceEndCommentsFixer

Fixed #67407.
---
 .../lib/Format/NamespaceEndCommentsFixer.cpp  |  4 +-
 .../Format/NamespaceEndCommentsFixerTest.cpp  | 57 +++
 2 files changed, 23 insertions(+), 38 deletions(-)

diff --git a/clang/lib/Format/NamespaceEndCommentsFixer.cpp 
b/clang/lib/Format/NamespaceEndCommentsFixer.cpp
index 4d3bd3b33f0f113..aed31db87495406 100644
--- a/clang/lib/Format/NamespaceEndCommentsFixer.cpp
+++ b/clang/lib/Format/NamespaceEndCommentsFixer.cpp
@@ -170,7 +170,7 @@ bool validEndComment(const FormatToken *RBraceTok, 
StringRef NamespaceName,
   // Valid namespace end comments don't need to be edited.
   static const llvm::Regex NamespaceCommentPattern =
   llvm::Regex("^/[/*] *(end (of )?)? *(anonymous|unnamed)? *"
-  "namespace( +([a-zA-Z0-9:_]+))?\\.? *(\\*/)?$",
+  "namespace( +([a-zA-Z0-9:_ ]+))?\\.? *(\\*/)?$",
   llvm::Regex::IgnoreCase);
   static const llvm::Regex NamespaceMacroCommentPattern =
   llvm::Regex("^/[/*] *(end (of )?)? *(anonymous|unnamed)? *"
@@ -189,7 +189,7 @@ bool validEndComment(const FormatToken *RBraceTok, 
StringRef NamespaceName,
 // Comment does not match regex.
 return false;
   }
-  StringRef NamespaceNameInComment = Groups.size() > 5 ? Groups[5] : "";
+  StringRef NamespaceNameInComment = Groups.size() > 5 ? Groups[5].rtrim() : 
"";
   // Anonymous namespace comments must not mention a namespace name.
   if (NamespaceName.empty() && !NamespaceNameInComment.empty())
 return false;
diff --git a/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp 
b/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp
index ec335c985ebba20..1ebcba551e4c9d2 100644
--- a/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp
+++ b/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp
@@ -658,42 +658,27 @@ TEST_F(NamespaceEndCommentsFixerTest,
 }
 
 TEST_F(NamespaceEndCommentsFixerTest, KeepsValidEndComment) {
-  EXPECT_EQ("namespace {\n"
-"int i;\n"
-"} // end anonymous namespace",
-fixNamespaceEndComments("namespace {\n"
-"int i;\n"
-"} // end anonymous namespace"));
-  EXPECT_EQ("namespace A {\n"
-"int i;\n"
-"} /* end of namespace A */",
-fixNamespaceEndComments("namespace A {\n"
-"int i;\n"
-"} /* end of namespace A */"));
-  EXPECT_EQ("namespace A {\n"
-"int i;\n"
-"}   //   namespace A",
-fixNamespaceEndComments("namespace A {\n"
-"int i;\n"
-"}   //   namespace A"));
-  EXPECT_EQ("namespace A::B {\n"
-"int i;\n"
-"} // end namespace A::B",
-fixNamespaceEndComments("namespace A::B {\n"
-"int i;\n"
-"} // end namespace A::B"));
-  EXPECT_EQ("namespace A {\n"
-"int i;\n"
-"}; // end namespace A",
-fixNamespaceEndComments("namespace A {\n"
-"int i;\n"
-"}; // end namespace A"));
-  EXPECT_EQ("namespace {\n"
-"int i;\n"
-"}; /* unnamed namespace */",
-fixNamespaceEndComments("namespace {\n"
-"int i;\n"
-"}; /* unnamed namespace */"));
+  EXPECT_TRUE(isFormatted("namespace {\n"
+  "int i;\n"
+  "} // end anonymous namespace"));
+  EXPECT_TRUE(isFormatted("namespace A {\n"
+  "int i;\n"
+  "} /* end of namespace A */"));
+  EXPECT_TRUE(isFormatted("namespace A {\n"
+  "int i;\n"
+  "}   //   namespace A"));
+  EXPECT_TRUE(isFormatted("namespace A::B {\n"
+  "int i;\n"
+  "} // end namespace A::B"));
+  EXPECT_TRUE(isFormatted("namespace A {\n"
+  "int i;\n"
+  "}; // end namespace A"));
+  EXPECT_TRUE(isFormatted("namespace {\n"
+  "int i;\n"
+  "}; /* unnamed namespace */"));
+  EXPECT_TRUE(isFormatted("namespace a::inline b {\n"
+  "int c;\n"
+  "}; // namespace a::inline b"));
 }
 
 TEST_F(NamespaceEndCommentsFixerTest, KeepsValidMacroEndComment) {

___
cfe-commits mailing list
cfe-commits@lists.llvm.o

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From c22ccc3153ccaac0e97f7ac41349fcc382508884 Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 321 
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 710 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised(), (

[clang] [clang-format] Fix a bug in NamespaceEndCommentsFixer (PR #67422)

[via cfe-commits]

llvmbot wrote:




@llvm/pr-subscribers-clang-format


Changes

Fixed #67407.

---
Full diff: https://github.com/llvm/llvm-project/pull/67422.diff


2 Files Affected:

- (modified) clang/lib/Format/NamespaceEndCommentsFixer.cpp (+2-2) 
- (modified) clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp (+21-36) 


``diff
diff --git a/clang/lib/Format/NamespaceEndCommentsFixer.cpp 
b/clang/lib/Format/NamespaceEndCommentsFixer.cpp
index 4d3bd3b33f0f113..aed31db87495406 100644
--- a/clang/lib/Format/NamespaceEndCommentsFixer.cpp
+++ b/clang/lib/Format/NamespaceEndCommentsFixer.cpp
@@ -170,7 +170,7 @@ bool validEndComment(const FormatToken *RBraceTok, 
StringRef NamespaceName,
   // Valid namespace end comments don't need to be edited.
   static const llvm::Regex NamespaceCommentPattern =
   llvm::Regex("^/[/*] *(end (of )?)? *(anonymous|unnamed)? *"
-  "namespace( +([a-zA-Z0-9:_]+))?\\.? *(\\*/)?$",
+  "namespace( +([a-zA-Z0-9:_ ]+))?\\.? *(\\*/)?$",
   llvm::Regex::IgnoreCase);
   static const llvm::Regex NamespaceMacroCommentPattern =
   llvm::Regex("^/[/*] *(end (of )?)? *(anonymous|unnamed)? *"
@@ -189,7 +189,7 @@ bool validEndComment(const FormatToken *RBraceTok, 
StringRef NamespaceName,
 // Comment does not match regex.
 return false;
   }
-  StringRef NamespaceNameInComment = Groups.size() > 5 ? Groups[5] : "";
+  StringRef NamespaceNameInComment = Groups.size() > 5 ? Groups[5].rtrim() : 
"";
   // Anonymous namespace comments must not mention a namespace name.
   if (NamespaceName.empty() && !NamespaceNameInComment.empty())
 return false;
diff --git a/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp 
b/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp
index ec335c985ebba20..1ebcba551e4c9d2 100644
--- a/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp
+++ b/clang/unittests/Format/NamespaceEndCommentsFixerTest.cpp
@@ -658,42 +658,27 @@ TEST_F(NamespaceEndCommentsFixerTest,
 }
 
 TEST_F(NamespaceEndCommentsFixerTest, KeepsValidEndComment) {
-  EXPECT_EQ("namespace {\n"
-"int i;\n"
-"} // end anonymous namespace",
-fixNamespaceEndComments("namespace {\n"
-"int i;\n"
-"} // end anonymous namespace"));
-  EXPECT_EQ("namespace A {\n"
-"int i;\n"
-"} /* end of namespace A */",
-fixNamespaceEndComments("namespace A {\n"
-"int i;\n"
-"} /* end of namespace A */"));
-  EXPECT_EQ("namespace A {\n"
-"int i;\n"
-"}   //   namespace A",
-fixNamespaceEndComments("namespace A {\n"
-"int i;\n"
-"}   //   namespace A"));
-  EXPECT_EQ("namespace A::B {\n"
-"int i;\n"
-"} // end namespace A::B",
-fixNamespaceEndComments("namespace A::B {\n"
-"int i;\n"
-"} // end namespace A::B"));
-  EXPECT_EQ("namespace A {\n"
-"int i;\n"
-"}; // end namespace A",
-fixNamespaceEndComments("namespace A {\n"
-"int i;\n"
-"}; // end namespace A"));
-  EXPECT_EQ("namespace {\n"
-"int i;\n"
-"}; /* unnamed namespace */",
-fixNamespaceEndComments("namespace {\n"
-"int i;\n"
-"}; /* unnamed namespace */"));
+  EXPECT_TRUE(isFormatted("namespace {\n"
+  "int i;\n"
+  "} // end anonymous namespace"));
+  EXPECT_TRUE(isFormatted("namespace A {\n"
+  "int i;\n"
+  "} /* end of namespace A */"));
+  EXPECT_TRUE(isFormatted("namespace A {\n"
+  "int i;\n"
+  "}   //   namespace A"));
+  EXPECT_TRUE(isFormatted("namespace A::B {\n"
+  "int i;\n"
+  "} // end namespace A::B"));
+  EXPECT_TRUE(isFormatted("namespace A {\n"
+  "int i;\n"
+  "}; // end namespace A"));
+  EXPECT_TRUE(isFormatted("namespace {\n"
+  "int i;\n"
+  "}; /* unnamed namespace */"));
+  EXPECT_TRUE(isFormatted("namespace a::inline b {\n"
+  "int c;\n"
+  "}; // namespace a::inline b"));
 }
 
 TEST_F(NamespaceEndCommentsFixerTest, KeepsValidMacroEndComment) {

``




https://github.com/llvm/llvm-project/pull/67422
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [Driver] -include: deprecate probing .gch (PR #67084)

[via cfe-commits]

zmodem wrote:

> Can we add a dedicated -Wdeprecated-gch diag group or something for this?

I gave it a flag in 39f36d5d5c88f49eaa946a3ee9ef887c04275a33

> (not sure what projects actually do this with Clang)

Turns out Chromium (and anyone else using gn) does: https://crbug.com/1486799
And probably other projects doing GCC-style pre-compiled header builds with 
Clang as a drop-in replacement for GCC.

In an environment like that, I'm not sure what action users should take when 
they see this warning: GCC doesn't understand `-include-pch`, and if `x.h.gch` 
is renamed to `x.h.pch`, GCC's `-include x.h` will not find it.

So it seems this feature really is necessary for drop-in compatibility with GCC.

https://github.com/llvm/llvm-project/pull/67084
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [Driver] -include: deprecate probing .gch (PR #67084)

[via cfe-commits]

zmodem wrote:

As an archaeological note, the gch probing was added a long time ago in 
a2aedc69435dd430925e77022cbd26cc51e30982 with the motivation

> It's wonky, but we include looking for .gch so we can support seamless 
> replacement into a build system already set up to be generating .gch files.

https://github.com/llvm/llvm-project/pull/67084
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

@@ -2352,26 +2352,61 @@ static void handleUnusedAttr(Sema &S, Decl *D, const 
ParsedAttr &AL) {
   D->addAttr(::new (S.Context) UnusedAttr(S.Context, AL));
 }
 
-static void handleConstructorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  uint32_t priority = ConstructorAttr::DefaultPriority;
+static bool diagnoseInvalidPriority(Sema &S, uint32_t Priority,
+const ParsedAttr &A,
+SourceLocation PriorityLoc) {
+  // Only perform the priority check if the attribute is outside of a system
+  // header. Values <= 100 are reserved for the implementation, and libc++
+  // benefits from being able to specify values in that range.
+  if ((Priority < 101 || Priority > 65535) &&
+  !S.getSourceManager().isInSystemHeader(A.getLoc())) {

AaronBallman wrote:

Ah, no, I can clarify the comment -- the 65535 is the documented upper bound, 
but we could reasonably extend that in the future if we wanted.

https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

@@ -2352,26 +2352,61 @@ static void handleUnusedAttr(Sema &S, Decl *D, const 
ParsedAttr &AL) {
   D->addAttr(::new (S.Context) UnusedAttr(S.Context, AL));
 }
 
-static void handleConstructorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  uint32_t priority = ConstructorAttr::DefaultPriority;
+static bool diagnoseInvalidPriority(Sema &S, uint32_t Priority,
+const ParsedAttr &A,
+SourceLocation PriorityLoc) {
+  // Only perform the priority check if the attribute is outside of a system
+  // header. Values <= 100 are reserved for the implementation, and libc++
+  // benefits from being able to specify values in that range.
+  if ((Priority < 101 || Priority > 65535) &&
+  !S.getSourceManager().isInSystemHeader(A.getLoc())) {
+S.Diag(A.getLoc(), diag::err_attribute_argument_out_of_range)
+<< PriorityLoc << A << 101 << 65535;

AaronBallman wrote:

Yup, GCC compat: https://godbolt.org/z/hYa3bPMPW

https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

@@ -2352,26 +2352,61 @@ static void handleUnusedAttr(Sema &S, Decl *D, const 
ParsedAttr &AL) {
   D->addAttr(::new (S.Context) UnusedAttr(S.Context, AL));
 }
 
-static void handleConstructorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  uint32_t priority = ConstructorAttr::DefaultPriority;
+static bool diagnoseInvalidPriority(Sema &S, uint32_t Priority,
+const ParsedAttr &A,
+SourceLocation PriorityLoc) {
+  // Only perform the priority check if the attribute is outside of a system
+  // header. Values <= 100 are reserved for the implementation, and libc++
+  // benefits from being able to specify values in that range.
+  if ((Priority < 101 || Priority > 65535) &&
+  !S.getSourceManager().isInSystemHeader(A.getLoc())) {
+S.Diag(A.getLoc(), diag::err_attribute_argument_out_of_range)
+<< PriorityLoc << A << 101 << 65535;
+A.setInvalid();
+return true;
+  }
+  return false;
+}
+
+template 
+static void handleCtorDtorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
+  uint32_t Priority = CtorDtorAttr::DefaultPriority;
   if (S.getLangOpts().HLSL && AL.getNumArgs()) {
 S.Diag(AL.getLoc(), diag::err_hlsl_init_priority_unsupported);
 return;
   }
-  if (AL.getNumArgs() &&
-  !checkUInt32Argument(S, AL, AL.getArgAsExpr(0), priority))
-return;
 
-  D->addAttr(::new (S.Context) ConstructorAttr(S.Context, AL, priority));
-}
+  // If we're given an argument for the priority, check that it's valid.
+  if (AL.getNumArgs()) {
+if (!checkUInt32Argument(S, AL, AL.getArgAsExpr(0), Priority))
+  return;
 
-static void handleDestructorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  uint32_t priority = DestructorAttr::DefaultPriority;
-  if (AL.getNumArgs() &&
-  !checkUInt32Argument(S, AL, AL.getArgAsExpr(0), priority))
+// Ensure the priority is in a reasonable range.
+if (diagnoseInvalidPriority(S, Priority, AL,
+AL.getArgAsExpr(0)->getExprLoc()))
+  return;
+  }
+
+  // Ensure the function we're attaching to is something that is sensible to
+  // automatically call before or after main(); it should accept no arguments
+  // and return no value (but it is not an error because it is theoretically

AaronBallman wrote:

Good catch! That comment is stale, I'll correct it.

https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

AaronBallman wrote:

> Should they be valid for `constexpr` functions?

Sure, but perhaps not for `consteval` functions. A `constexpr` function can be 
deferred to runtime, whereas a `consteval` one cannot. WDYT?

https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang][Diagnostics] Highlight code snippets (PR #66514)

[Konrad Kleine via cfe-commits]

Timm =?utf-8?q?Bäder?= ,
Timm =?utf-8?q?Bäder?= ,
Timm =?utf-8?q?Bäder?= ,
Timm =?utf-8?q?Bäder?= ,
Timm =?utf-8?q?Bäder?= 
Message-ID:
In-Reply-To: 


https://github.com/kwk resolved https://github.com/llvm/llvm-project/pull/66514
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf resolved https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From c22ccc3153ccaac0e97f7ac41349fcc382508884 Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH 1/2] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 321 
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 710 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised(

[clang] [libc++] P2770R0: "Stashing stashing iterators for proper flattening" (PR #66033)

[Jakub Mazurkiewicz via cfe-commits]

https://github.com/JMazurkiewicz updated 
https://github.com/llvm/llvm-project/pull/66033

>From bf7096f8f435268fd830c27056e2f24564f4356b Mon Sep 17 00:00:00 2001
From: Jakub Mazurkiewicz 
Date: Fri, 8 Sep 2023 18:20:59 +0200
Subject: [PATCH 1/5] [libc++] P2770R0: "Stashing stashing iterators for proper
 flattening"

* Parially implements P2770R0: "Stashing stashing iterators for proper 
flattening"
* `join_with_view` hasn't been done yet since this type isn't implemented yet
* Rename `test/libcxx/ranges/range.adaptors/range.adaptor.tuple` directory to 
`test/libcxx/ranges/range.adaptors/range.adaptor.helpers` to match the 
standard: http://eel.is/c++draft/range.adaptor.helpers (this change happened in 
P2770R0, see point 3 of wording).
* Rename `libcxx\test\std\ranges\range.adaptors\range.join.view` to 
`libcxx\test\std\ranges\range.adaptors\range.join` to match the standard
---
 libcxx/docs/Status/Cxx23.rst  |   1 +
 libcxx/docs/Status/Cxx23Papers.csv|   2 +-
 libcxx/docs/Status/RangesMajorFeatures.csv|   1 +
 libcxx/docs/Status/RangesViews.csv|   2 +-
 libcxx/include/CMakeLists.txt |   1 +
 libcxx/include/__ranges/helpers.h |  39 ++
 libcxx/include/__ranges/join_view.h   | 131 +++---
 libcxx/include/module.modulemap.in|   1 +
 libcxx/include/ranges |   1 +
 libcxx/include/regex  |   8 ++
 libcxx/modules/std/ranges.inc |   2 -
 .../as-lvalue.verify.cpp  |  47 +++
 .../tuple-for-each.pass.cpp   |   0
 .../segmented_iterator.compile.pass.cpp   |   1 -
 .../alg.copy/ranges.copy.segmented.pass.cpp   |   2 -
 .../alg.copy/ranges.copy_backward.pass.cpp|   2 -
 .../iterator/ctor.parent.outer.pass.cpp   |  57 
 .../adaptor.pass.cpp  |   1 -
 .../base.pass.cpp |   1 -
 .../begin.pass.cpp|  31 -
 .../ctad.compile.pass.cpp |   1 -
 .../ctad.verify.cpp   |   1 -
 .../ctor.default.pass.cpp |   1 -
 .../ctor.view.pass.cpp|   1 -
 .../end.pass.cpp  |  22 +--
 .../general.pass.cpp  |   1 -
 .../iterator/arrow.pass.cpp   |   1 -
 .../iterator/ctor.default.pass.cpp|  16 +--
 .../iterator/ctor.other.pass.cpp  |   3 +-
 .../iterator/decrement.pass.cpp   |  21 ++-
 .../iterator/eq.pass.cpp  |   7 +-
 .../iterator/increment.pass.cpp   |  18 ++-
 .../iterator/iter.move.pass.cpp   |   1 -
 .../iterator/iter.swap.pass.cpp   |   1 -
 .../iterator/member_types.compile.pass.cpp|   1 -
 .../iterator/star.pass.cpp|   1 -
 .../sentinel/ctor.default.pass.cpp|   1 -
 .../sentinel/ctor.other.pass.cpp  |   1 -
 .../sentinel/ctor.parent.pass.cpp |   1 -
 .../sentinel/eq.pass.cpp  |  19 ++-
 .../{range.join.view => range.join}/types.h   |  66 -
 ...rator_concept_conformance.compile.pass.cpp |   4 +-
 .../std/re/re.iter/re.regiter/types.pass.cpp  |   4 +
 ...rator_concept_conformance.compile.pass.cpp |   4 +-
 .../std/re/re.iter/re.tokiter/types.pass.cpp  |   4 +
 45 files changed, 350 insertions(+), 182 deletions(-)
 create mode 100644 libcxx/include/__ranges/helpers.h
 create mode 100644 
libcxx/test/libcxx/ranges/range.adaptors/range.adaptor.helpers/as-lvalue.verify.cpp
 rename libcxx/test/libcxx/ranges/range.adaptors/{range.adaptor.tuple => 
range.adaptor.helpers}/tuple-for-each.pass.cpp (100%)
 delete mode 100644 
libcxx/test/std/ranges/range.adaptors/range.join.view/iterator/ctor.parent.outer.pass.cpp
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/adaptor.pass.cpp (99%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/base.pass.cpp (98%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/begin.pass.cpp (83%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/ctad.compile.pass.cpp (98%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/ctad.verify.cpp (97%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/ctor.default.pass.cpp (97%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/ctor.view.pass.cpp (97%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/end.pass.cpp (95%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/general.pass.cpp (98%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/iterator/arrow.pass.cpp (99%)
 rename libcxx/test/std/ranges/range.adaptors/{range.join.view => 
range.join}/iterato

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Freddy Ye via cfe-commits]

@@ -4742,6 +4742,125 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS 
_mm_castsi128_pd(__m128i __a) {
   return (__m128d)__a;
 }
 
+/// Compares each of the corresponding double-precision values of two
+///128-bit vectors of [2 x double], using the operation specified by the
+///immediate integer operand.
+///
+///Returns a [2 x double] vector consisting of two doubles corresponding to
+///the two comparison results: zero if the comparison is false, and all 1's
+///if the comparison is true.
+///
+/// \headerfile 
+///
+/// \code
+/// __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the  VCMPPD  instruction.
+///
+/// \param a
+///A 128-bit vector of [2 x double].
+/// \param b
+///A 128-bit vector of [2 x double].
+/// \param c
+///An immediate integer operand, with bits [4:0] specifying which 
comparison

FreddyLeaf wrote:

After this PR, I'll get such results:
```
$ cat small.c 

#include 

__m128d cmp (__m128d a, __m128d b)
{
  return _mm_cmp_sd (a, b, 10);
}

$ clang -msse2 -O2 small.c -S -o - | grep cmpsd
cmpsd   $10, %xmm1, %xmm0
$ clang -mavx -O2 small.c -S -o - | grep vcmp
vcmpngtsd   %xmm1, %xmm0, %xmm0
```
Not sure if the higher bits could be omitted in encoding. More investigations 
need to be done. Will update.

https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf updated https://github.com/llvm/llvm-project/pull/66430

>From 47392e5996cccada1d638495a9376b0430142c8c Mon Sep 17 00:00:00 2001
From: Giulio Eulisse <10544+...@users.noreply.github.com>
Date: Thu, 14 Sep 2023 21:58:21 +0200
Subject: [PATCH] Introduce PagedVector class

The goal of the class is to be an (almost) drop in replacement for
SmallVector and std::vector when those are presized and filled later,
as it happens in SourceManager and ASTReader.

By splitting the actual vector in pages of the same size and allocating
the pages only when they are needed, using this containers reduces the
memory usage by a factor 4 for the cases relevant to the ALICE
experiment ROOT / cling usage.
---
 clang/include/clang/Basic/SourceManager.h |   3 +-
 clang/include/clang/Serialization/ASTReader.h |   5 +-
 clang/lib/Basic/SourceManager.cpp |  10 +-
 clang/lib/Serialization/ASTReader.cpp |   5 +-
 llvm/docs/ProgrammersManual.rst   |  33 ++
 llvm/include/llvm/ADT/PagedVector.h   | 323 +
 llvm/unittests/ADT/CMakeLists.txt |   1 +
 llvm/unittests/ADT/PagedVectorTest.cpp| 342 ++
 8 files changed, 712 insertions(+), 10 deletions(-)
 create mode 100644 llvm/include/llvm/ADT/PagedVector.h
 create mode 100644 llvm/unittests/ADT/PagedVectorTest.cpp

diff --git a/clang/include/clang/Basic/SourceManager.h 
b/clang/include/clang/Basic/SourceManager.h
index 2f846502d6f3327..e37caa2252532f9 100644
--- a/clang/include/clang/Basic/SourceManager.h
+++ b/clang/include/clang/Basic/SourceManager.h
@@ -43,6 +43,7 @@
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringRef.h"
@@ -699,7 +700,7 @@ class SourceManager : public RefCountedBase {
   ///
   /// Negative FileIDs are indexes into this table. To get from ID to an index,
   /// use (-ID - 2).
-  SmallVector LoadedSLocEntryTable;
+  llvm::PagedVector LoadedSLocEntryTable;
 
   /// The starting offset of the next local SLocEntry.
   ///
diff --git a/clang/include/clang/Serialization/ASTReader.h 
b/clang/include/clang/Serialization/ASTReader.h
index dc1eb21c27801fe..65e19c6e44cf571 100644
--- a/clang/include/clang/Serialization/ASTReader.h
+++ b/clang/include/clang/Serialization/ASTReader.h
@@ -38,6 +38,7 @@
 #include "llvm/ADT/DenseSet.h"
 #include "llvm/ADT/IntrusiveRefCntPtr.h"
 #include "llvm/ADT/MapVector.h"
+#include "llvm/ADT/PagedVector.h"
 #include "llvm/ADT/STLExtras.h"
 #include "llvm/ADT/SetVector.h"
 #include "llvm/ADT/SmallPtrSet.h"
@@ -487,7 +488,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the type with
   /// ID = (I + 1) << FastQual::Width has already been loaded
-  std::vector TypesLoaded;
+  llvm::PagedVector TypesLoaded;
 
   using GlobalTypeMapType =
   ContinuousRangeMap;
@@ -501,7 +502,7 @@ class ASTReader
   ///
   /// When the pointer at index I is non-NULL, the declaration with ID
   /// = I + 1 has already been loaded.
-  std::vector DeclsLoaded;
+  llvm::PagedVector DeclsLoaded;
 
   using GlobalDeclMapType =
   ContinuousRangeMap;
diff --git a/clang/lib/Basic/SourceManager.cpp 
b/clang/lib/Basic/SourceManager.cpp
index 0521ac7b30339ab..7fa8b8096ac4931 100644
--- a/clang/lib/Basic/SourceManager.cpp
+++ b/clang/lib/Basic/SourceManager.cpp
@@ -2344,11 +2344,11 @@ SourceManager::MemoryBufferSizes 
SourceManager::getMemoryBufferSizes() const {
 }
 
 size_t SourceManager::getDataStructureSizes() const {
-  size_t size = llvm::capacity_in_bytes(MemBufferInfos)
-+ llvm::capacity_in_bytes(LocalSLocEntryTable)
-+ llvm::capacity_in_bytes(LoadedSLocEntryTable)
-+ llvm::capacity_in_bytes(SLocEntryLoaded)
-+ llvm::capacity_in_bytes(FileInfos);
+  size_t size = llvm::capacity_in_bytes(MemBufferInfos) +
+llvm::capacity_in_bytes(LocalSLocEntryTable) +
+llvm::capacity_in_bytes(LoadedSLocEntryTable) +
+llvm::capacity_in_bytes(SLocEntryLoaded) +
+llvm::capacity_in_bytes(FileInfos);
 
   if (OverriddenFilesInfo)
 size += llvm::capacity_in_bytes(OverriddenFilesInfo->OverriddenFiles);
diff --git a/clang/lib/Serialization/ASTReader.cpp 
b/clang/lib/Serialization/ASTReader.cpp
index 0952244d037a77c..badd54987af18dd 100644
--- a/clang/lib/Serialization/ASTReader.cpp
+++ b/clang/lib/Serialization/ASTReader.cpp
@@ -7944,9 +7944,10 @@ void ASTReader::PrintStats() {
   std::fprintf(stderr, "*** AST File Statistics:\n");
 
   unsigned NumTypesLoaded =
-  TypesLoaded.size() - llvm::count(TypesLoaded, QualType());
+  TypesLoaded.size() - llvm::count(TypesLoaded.materialised(), QualType());
   unsigned NumDeclsLoaded =
-  DeclsLoaded.size() - llvm::count(DeclsLoaded, (Decl *)nullptr);
+  DeclsLoaded.size() -
+  llvm::count(DeclsLoaded.materialised(), 

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

@@ -0,0 +1,322 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+#include 
+
+namespace llvm {
+// A vector that allocates memory in pages.
+// Order is kept, but memory is allocated only when one element of the page is
+// accessed. This introduces a level of indirection, but it is useful when you
+// have a sparsely initialised vector where the full size is allocated upfront
+// with the default constructor and elements are initialised later, on first
+// access.
+//
+// Notice that this does not have iterators, because if you
+// have iterators it probably means you are going to touch
+// all the memory in any case, so better use a std::vector in
+// the first place.
+//
+// Pages are allocated in SLAB_SIZE chunks, using the BumpPtrAllocator.
+template 
+class PagedVector {
+  static_assert(PAGE_SIZE > 0, "PAGE_SIZE must be greater than 0. Most likely "
+   "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.
+  std::size_t Size = 0;
+
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable std::vector PageToDataIdx;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The order of
+  // the elements however depends on the order of access of the pages.
+  PointerIntPair Allocator;
+
+  constexpr static uintptr_t InvalidPage = SIZE_MAX;
+
+public:
+  using value_type = T;
+
+  // Default constructor. We build our own allocator.
+  PagedVector() : Allocator(new BumpPtrAllocator, true) {}
+  PagedVector(BumpPtrAllocator *A) : Allocator(A, false) {}
+
+  ~PagedVector() {
+// If we own the allocator, delete it.
+if (Allocator.getInt() == true)
+  delete Allocator.getPointer();
+  }
+
+  // Lookup an element at position i.
+  // If the associated page is not filled, it will be filled with default
+  // constructed elements. If the associated page is filled, return the 
element.
+  T &operator[](std::size_t Index) const {
+assert(Index < Size);
+assert(Index / PAGE_SIZE < PageToDataIdx.size());
+uintptr_t &PagePtr = PageToDataIdx[Index / PAGE_SIZE];
+// If the page was not yet allocated, allocate it.
+if (PagePtr == InvalidPage) {
+  T *NewPagePtr = Allocator.getPointer()->template Allocate(PAGE_SIZE);
+  // We need to invoke the default constructor on all the elements of the
+  // page.
+  for (std::size_t I = 0; I < PAGE_SIZE; ++I)
+new (NewPagePtr + I) T();
+
+  PagePtr = reinterpret_cast(NewPagePtr);
+}
+// Dereference the element in the page.
+return *((Index % PAGE_SIZE) + reinterpret_cast(PagePtr));
+  }
+
+  // Return the capacity of the vector. I.e. the maximum size it can be 
expanded
+  // to with the resize method without allocating more pages.
+  [[nodiscard]] std::size_t capacity() const {
+return PageToDataIdx.size() * PAGE_SIZE;
+  }
+
+  // Return the size of the vector. I.e. the maximum index that can be
+  // accessed, i.e. the maximum value which was used as argument of the
+  // resize method.
+  [[nodiscard]] std::size_t size() const { return Size; }
+
+  // Expands the vector to the given NewSize number of elements.
+  // If the vector was smaller, allocates new pages as needed.
+  // It should be called only with NewSize >= Size.
+  void resize(std::size_t NewSize) {
+// Handle shrink case: delete the pages and update the size.
+if (NewSize < Size) {
+  std::size_t NewLastPage = (NewSize - 1) / PAGE_SIZE;
+  for (std::size_t I = NewLastPage + 1; I < PageToDataIdx.size(); ++I) {
+uintptr_t PagePtr = PageToDataIdx[I];
+if (PagePtr == InvalidPage)
+  continue;
+T *Page = reinterpret_cast(PagePtr);
+// We need to invoke the destructor on all the elements of the page.
+for (std::size_t J = 0; J < PAGE_SIZE; ++J)
+  Page[J].~T();
+Allocator.getPointer()->Deallocate(Page);
+  }
+  // Delete the extra ones in the new last page.
+  uintptr_t PagePtr = PageToDataIdx[NewLastPage];
+  if (PagePtr != InvalidPage) {
+T *Page = reinterpret_cast(PagePtr)

[clang] Introduce paged vector (PR #66430)

[Giulio Eulisse via cfe-commits]

https://github.com/ktf resolved https://github.com/llvm/llvm-project/pull/66430
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

@@ -2352,26 +2352,61 @@ static void handleUnusedAttr(Sema &S, Decl *D, const 
ParsedAttr &AL) {
   D->addAttr(::new (S.Context) UnusedAttr(S.Context, AL));
 }
 
-static void handleConstructorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  uint32_t priority = ConstructorAttr::DefaultPriority;
+static bool diagnoseInvalidPriority(Sema &S, uint32_t Priority,
+const ParsedAttr &A,
+SourceLocation PriorityLoc) {
+  // Only perform the priority check if the attribute is outside of a system
+  // header. Values <= 100 are reserved for the implementation, and libc++
+  // benefits from being able to specify values in that range.
+  if ((Priority < 101 || Priority > 65535) &&

AaronBallman wrote:

Eh, I'm on the fence given that there are three different attributes which are 
unrelated except for having the notion of a priority. I'll make some local 
variables for this rather than try to associate it with a particular attribute.

https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2 (PR #67410)

[Phoebe Wang via cfe-commits]

@@ -4742,6 +4742,125 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS 
_mm_castsi128_pd(__m128i __a) {
   return (__m128d)__a;
 }
 
+/// Compares each of the corresponding double-precision values of two
+///128-bit vectors of [2 x double], using the operation specified by the
+///immediate integer operand.
+///
+///Returns a [2 x double] vector consisting of two doubles corresponding to
+///the two comparison results: zero if the comparison is false, and all 1's
+///if the comparison is true.
+///
+/// \headerfile 
+///
+/// \code
+/// __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the  VCMPPD  instruction.
+///
+/// \param a
+///A 128-bit vector of [2 x double].
+/// \param b
+///A 128-bit vector of [2 x double].
+/// \param c
+///An immediate integer operand, with bits [4:0] specifying which 
comparison

phoebewang wrote:

How about add sema check for SSE case? Users need to remind they may get 
unexpected result due to not enable AVX feature.

https://github.com/llvm/llvm-project/pull/67410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Introduce paged vector (PR #66430)

[Vassil Vassilev via cfe-commits]

@@ -0,0 +1,323 @@
+//===- llvm/ADT/PagedVector.h - 'Lazyly allocated' vectors *- C++
+//-*-===//
+//
+// 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
+//
+//===--===//
+//
+// This file defines the PagedVector class.
+//
+//===--===//
+#ifndef LLVM_ADT_PAGEDVECTOR_H
+#define LLVM_ADT_PAGEDVECTOR_H
+
+#include "llvm/ADT/PointerIntPair.h"
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/ADT/iterator_range.h"
+#include "llvm/Support/Allocator.h"
+#include 
+#include 
+
+namespace llvm {
+/// A vector that allocates memory in pages.
+///
+/// Order is kept, but memory is allocated only when one element of the page is
+/// accessed. This introduces a level of indirection, but it is useful when you
+/// have a sparsely initialised vector where the full size is allocated 
upfront.
+///
+/// As a side effect the elements are initialised later than in a normal 
vector.
+/// On the first access to one of the elements of a given page all, the 
elements
+/// of the page are initialised. This also means that the elements of the page
+/// are initialised beyond the size of the vector.
+///
+/// Similarly on destruction the elements are destroyed only when the page is
+/// not needed anymore, delaying invoking the destructor of the elements.
+///
+/// Notice that this does not have iterators, because if you have iterators it
+/// probably means you are going to touch all the memory in any case, so better
+/// use a std::vector in the first place.
+template  class PagedVector {
+  static_assert(PageSize > 1, "PageSize must be greater than 0. Most likely "
+  "you want it to be greater than 16.");
+  // The actual number of element in the vector which can be accessed.
+  size_t Size = 0;
+
+  // The position of the initial element of the page in the Data vector.
+  // Pages are allocated contiguously in the Data vector.
+  mutable SmallVector PageToDataPtrs;
+  // Actual page data. All the page elements are added to this vector on the
+  // first access of any of the elements of the page. Elements default
+  // constructed and elements of the page are stored contiguously. The order of
+  // the elements however depends on the order of access of the pages.
+  PointerIntPair Allocator;
+
+  constexpr static T *InvalidPage = nullptr;
+
+public:
+  using value_type = T;
+
+  /// Default constructor. We build our own allocator and mark it as such with
+  /// `true` in the second pair element.
+  PagedVector() : Allocator(new BumpPtrAllocator, true) {}
+  PagedVector(BumpPtrAllocator *A) : Allocator(A, false) {
+assert(A != nullptr && "Allocator cannot be null");
+  }
+
+  ~PagedVector() {
+clear();
+// If we own the allocator, delete it.
+if (Allocator.getInt())
+  delete Allocator.getPointer();
+  }
+
+  /// Look up an element at position `Index`.
+  /// If the associated page is not filled, it will be filled with default
+  /// constructed elements. If the associated page is filled, return the
+  /// element.
+  T &operator[](size_t Index) const {
+assert(Index < Size);
+assert(Index / PageSize < PageToDataPtrs.size());
+T *&PagePtr = PageToDataPtrs[Index / PageSize];
+// If the page was not yet allocated, allocate it.
+if (PagePtr == InvalidPage) {
+  T *NewPagePtr = Allocator.getPointer()->template Allocate(PageSize);
+  // We need to invoke the default constructor on all the elements of the
+  // page.
+  for (size_t I = 0; I < PageSize; ++I)
+new (NewPagePtr + I) T();
+
+  PagePtr = NewPagePtr;
+}
+// Dereference the element in the page.
+return *((Index % PageSize) + PagePtr);
+  }
+
+  /// Return the capacity of the vector. I.e. the maximum size it can be
+  /// expanded to with the resize method without allocating more pages.
+  [[nodiscard]] size_t capacity() const {
+return PageToDataPtrs.size() * PageSize;
+  }
+
+  /// Return the size of the vector. I.e. the maximum index that can be
+  /// accessed, i.e. the maximum value which was used as argument of the
+  /// resize method.
+  [[nodiscard]] size_t size() const { return Size; }
+
+  /// Resize the vector. Notice that the constructor of the elements will not
+  /// be invoked until an element of a given page is accessed, at which point
+  /// all the elements of the page will be constructed.
+  ///
+  /// If the new size is smaller than the current size, the elements of the
+  /// pages that are not needed anymore will be destroyed, however, elements of
+  /// the last page will not be destroyed.
+  ///
+  /// For these reason the usage of this vector is discouraged if you rely
+  /// on the construction / destructor of the elements to be 

[clang] 166074e - [clang analysis][NFCI] Preparatory work for D153131. (#67420)

[via cfe-commits]

Author: Clement Courbet
Date: 2023-09-26T15:05:14+02:00
New Revision: 166074eff2e9a5f79b791f1cc9b641a4e2968616

URL: 
https://github.com/llvm/llvm-project/commit/166074eff2e9a5f79b791f1cc9b641a4e2968616
DIFF: 
https://github.com/llvm/llvm-project/commit/166074eff2e9a5f79b791f1cc9b641a4e2968616.diff

LOG: [clang analysis][NFCI] Preparatory work for D153131. (#67420)

This was ported over from phabricator.

Review https://reviews.llvm.org/D153131.

Added: 


Modified: 
clang/lib/Analysis/ThreadSafety.cpp

Removed: 




diff  --git a/clang/lib/Analysis/ThreadSafety.cpp 
b/clang/lib/Analysis/ThreadSafety.cpp
index 3e6ceb7d54c427a..f160cf4d013c78d 100644
--- a/clang/lib/Analysis/ThreadSafety.cpp
+++ b/clang/lib/Analysis/ThreadSafety.cpp
@@ -1015,6 +1015,19 @@ class ThreadSafetyAnalyzer {
 
   BeforeSet *GlobalBeforeSet;
 
+  void warnIfMutexNotHeld(const FactSet &FSet, const NamedDecl *D,
+  const Expr *Exp, AccessKind AK, Expr *MutexExp,
+  ProtectedOperationKind POK, til::LiteralPtr *Self,
+  SourceLocation Loc);
+  void warnIfMutexHeld(const FactSet &FSet, const NamedDecl *D, const Expr 
*Exp,
+   Expr *MutexExp, til::LiteralPtr *Self,
+   SourceLocation Loc);
+
+  void checkAccess(const FactSet &FSet, const Expr *Exp, AccessKind AK,
+   ProtectedOperationKind POK);
+  void checkPtAccess(const FactSet &FSet, const Expr *Exp, AccessKind AK,
+ ProtectedOperationKind POK);
+
 public:
   ThreadSafetyAnalyzer(ThreadSafetyHandler &H, BeforeSet* Bset)
   : Arena(&Bpa), SxBuilder(Arena), Handler(H), GlobalBeforeSet(Bset) {}
@@ -1534,16 +1547,15 @@ class BuildLockset : public 
ConstStmtVisitor {
   unsigned CtxIndex;
 
   // helper functions
-  void warnIfMutexNotHeld(const NamedDecl *D, const Expr *Exp, AccessKind AK,
-  Expr *MutexExp, ProtectedOperationKind POK,
-  til::LiteralPtr *Self, SourceLocation Loc);
-  void warnIfMutexHeld(const NamedDecl *D, const Expr *Exp, Expr *MutexExp,
-   til::LiteralPtr *Self, SourceLocation Loc);
 
   void checkAccess(const Expr *Exp, AccessKind AK,
-   ProtectedOperationKind POK = POK_VarAccess);
+   ProtectedOperationKind POK = POK_VarAccess) {
+Analyzer->checkAccess(FSet, Exp, AK, POK);
+  }
   void checkPtAccess(const Expr *Exp, AccessKind AK,
- ProtectedOperationKind POK = POK_VarAccess);
+ ProtectedOperationKind POK = POK_VarAccess) {
+Analyzer->checkPtAccess(FSet, Exp, AK, POK);
+  }
 
   void handleCall(const Expr *Exp, const NamedDecl *D,
   til::LiteralPtr *Self = nullptr,
@@ -1571,17 +1583,14 @@ class BuildLockset : public 
ConstStmtVisitor {
 
 /// Warn if the LSet does not contain a lock sufficient to protect access
 /// of at least the passed in AccessKind.
-void BuildLockset::warnIfMutexNotHeld(const NamedDecl *D, const Expr *Exp,
-  AccessKind AK, Expr *MutexExp,
-  ProtectedOperationKind POK,
-  til::LiteralPtr *Self,
-  SourceLocation Loc) {
+void ThreadSafetyAnalyzer::warnIfMutexNotHeld(
+const FactSet &FSet, const NamedDecl *D, const Expr *Exp, AccessKind AK,
+Expr *MutexExp, ProtectedOperationKind POK, til::LiteralPtr *Self,
+SourceLocation Loc) {
   LockKind LK = getLockKindFromAccessKind(AK);
-
-  CapabilityExpr Cp =
-  Analyzer->SxBuilder.translateAttrExpr(MutexExp, D, Exp, Self);
+  CapabilityExpr Cp = SxBuilder.translateAttrExpr(MutexExp, D, Exp, Self);
   if (Cp.isInvalid()) {
-warnInvalidLock(Analyzer->Handler, MutexExp, D, Exp, Cp.getKind());
+warnInvalidLock(Handler, MutexExp, D, Exp, Cp.getKind());
 return;
   } else if (Cp.shouldIgnore()) {
 return;
@@ -1589,68 +1598,67 @@ void BuildLockset::warnIfMutexNotHeld(const NamedDecl 
*D, const Expr *Exp,
 
   if (Cp.negative()) {
 // Negative capabilities act like locks excluded
-const FactEntry *LDat = FSet.findLock(Analyzer->FactMan, !Cp);
+const FactEntry *LDat = FSet.findLock(FactMan, !Cp);
 if (LDat) {
-  Analyzer->Handler.handleFunExcludesLock(
-  Cp.getKind(), D->getNameAsString(), (!Cp).toString(), Loc);
-  return;
+Handler.handleFunExcludesLock(Cp.getKind(), D->getNameAsString(),
+  (!Cp).toString(), Loc);
+return;
 }
 
 // If this does not refer to a negative capability in the same class,
 // then stop here.
-if (!Analyzer->inCurrentScope(Cp))
-  return;
+if (!inCurrentScope(Cp))
+return;
 
 // Otherwise the negative requirement must be propagated to the caller.
-LDat = FSet.findLock(Analyzer->FactMa

[clang] [clang analysis][NFCI] Preparatory work for D153131. (PR #67420)

[Clement Courbet via cfe-commits]

https://github.com/legrosbuffle closed 
https://github.com/llvm/llvm-project/pull/67420
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

@@ -171,6 +171,11 @@ Attribute Changes in Clang
   automatic diagnostic to use parameters of types that the format style
   supports but that are never the result of default argument promotion, such as
   ``float``. (`#59824: `_)
+- The ``constructor`` and ``destructor`` attributes now diagnose when:
+  - the priority is not between 101 and 65535, inclusive,
+  - the function it is applied to accepts arguments or has a non-void return

AaronBallman wrote:

I was on the fence about diagnosing the return value, but ultimately decided it 
was a type error to have a non-void return value because of calling 
conventions. e.g., 
```
struct S {
  long double ld1, ld2;
};

__attribute__((constructor))
struct S func(void) {
  return (struct S){ 1.0l, 2.0l };
}
```
where the returned struct is actually returned via a hidden pointer passed to 
the function. I don't want to play whack-a-mole with signature checking unless 
there's some compelling use case in the wild where people actually are using 
these functions to obtain a value outside of ctor/dtor calls.

https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

@@ -2352,26 +2352,61 @@ static void handleUnusedAttr(Sema &S, Decl *D, const 
ParsedAttr &AL) {
   D->addAttr(::new (S.Context) UnusedAttr(S.Context, AL));
 }
 
-static void handleConstructorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  uint32_t priority = ConstructorAttr::DefaultPriority;
+static bool diagnoseInvalidPriority(Sema &S, uint32_t Priority,
+const ParsedAttr &A,
+SourceLocation PriorityLoc) {
+  // Only perform the priority check if the attribute is outside of a system
+  // header. Values <= 100 are reserved for the implementation, and libc++
+  // benefits from being able to specify values in that range.
+  if ((Priority < 101 || Priority > 65535) &&
+  !S.getSourceManager().isInSystemHeader(A.getLoc())) {
+S.Diag(A.getLoc(), diag::err_attribute_argument_out_of_range)
+<< PriorityLoc << A << 101 << 65535;
+A.setInvalid();
+return true;
+  }
+  return false;
+}
+
+template 
+static void handleCtorDtorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
+  uint32_t Priority = CtorDtorAttr::DefaultPriority;
   if (S.getLangOpts().HLSL && AL.getNumArgs()) {
 S.Diag(AL.getLoc(), diag::err_hlsl_init_priority_unsupported);
 return;
   }
-  if (AL.getNumArgs() &&
-  !checkUInt32Argument(S, AL, AL.getArgAsExpr(0), priority))
-return;
 
-  D->addAttr(::new (S.Context) ConstructorAttr(S.Context, AL, priority));
-}
+  // If we're given an argument for the priority, check that it's valid.
+  if (AL.getNumArgs()) {
+if (!checkUInt32Argument(S, AL, AL.getArgAsExpr(0), Priority))
+  return;
 
-static void handleDestructorAttr(Sema &S, Decl *D, const ParsedAttr &AL) {
-  uint32_t priority = DestructorAttr::DefaultPriority;
-  if (AL.getNumArgs() &&
-  !checkUInt32Argument(S, AL, AL.getArgAsExpr(0), priority))
+// Ensure the priority is in a reasonable range.
+if (diagnoseInvalidPriority(S, Priority, AL,
+AL.getArgAsExpr(0)->getExprLoc()))
+  return;
+  }
+
+  // Ensure the function we're attaching to is something that is sensible to
+  // automatically call before or after main(); it should accept no arguments
+  // and return no value (but it is not an error because it is theoretically
+  // possible to call the function during normal program execution and pass it
+  // valid values). It also cannot be a member function. We allow K&R C
+  // functions because that's a difficult edge case where it depends on how the
+  // function is defined as to whether it does or does not expect arguments.
+  auto *FD = cast(D);
+  if (!FD->getReturnType()->isVoidType() ||
+  (FD->hasPrototype() && FD->getNumParams() != 0)) {
+S.Diag(AL.getLoc(), diag::err_ctor_dtor_attr_on_non_void_func)
+<< AL << FD->getSourceRange();
 return;
+  } else if (auto *MD = dyn_cast(FD); MD && MD->isInstance()) {
+S.Diag(AL.getLoc(), diag::err_ctor_dtor_member_func)
+<< AL << FD->getSourceRange();
+return;
+  }
 
-  D->addAttr(::new (S.Context) DestructorAttr(S.Context, AL, priority));
+  D->addAttr(::new (S.Context) CtorDtorAttr(S.Context, AL, Priority));

AaronBallman wrote:

Sure, I can switch to that!

https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D153131: [clang analysis][thread-safety] Handle return-by-reference...

[Clement Courbet via Phabricator via cfe-commits]

courbet added a comment.

Thanks for the review.




Comment at: clang/lib/Analysis/ThreadSafety.cpp:2304-2305
 
+  CFGBlockInfo &Initial = BlockInfo[CFGraph->getEntry().getBlockID()];
+  CFGBlockInfo &Final = BlockInfo[CFGraph->getExit().getBlockID()];
+

aaronpuchert wrote:
> You might want to do the `*` -> `&` in a separate commit.
SG, done in https://github.com/llvm/llvm-project/pull/66750


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D153131/new/

https://reviews.llvm.org/D153131

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D153131: [clang analysis][thread-safety] Handle return-by-reference...

[Clement Courbet via Phabricator via cfe-commits]

courbet updated this revision to Diff 557354.
courbet added a comment.

rebase


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D153131/new/

https://reviews.llvm.org/D153131

Files:
  clang/include/clang/Analysis/Analyses/ThreadSafety.h
  clang/include/clang/Basic/DiagnosticSemaKinds.td
  clang/lib/Analysis/ThreadSafety.cpp
  clang/lib/Sema/AnalysisBasedWarnings.cpp
  clang/test/SemaCXX/warn-thread-safety-analysis.cpp

Index: clang/test/SemaCXX/warn-thread-safety-analysis.cpp
===
--- clang/test/SemaCXX/warn-thread-safety-analysis.cpp
+++ clang/test/SemaCXX/warn-thread-safety-analysis.cpp
@@ -5580,6 +5580,85 @@
   }
 };
 
+class Return {
+  Mutex mu;
+  Foo foo GUARDED_BY(mu);
+  Foo* foo_ptr PT_GUARDED_BY(mu);
+
+  Foo returns_value_locked() {
+MutexLock lock(&mu);
+return foo;
+  }
+
+  Foo returns_value_locks_required() EXCLUSIVE_LOCKS_REQUIRED(mu) {
+return foo;
+  }
+
+  Foo returns_value_releases_lock_after_return() UNLOCK_FUNCTION(mu) {
+MutexLock lock(&mu, true);
+return foo;
+  }
+
+  Foo returns_value_aquires_lock() EXCLUSIVE_LOCK_FUNCTION(mu) {
+mu.Lock();
+return foo;
+  }
+  
+  Foo returns_value_not_locked() {
+return foo;   // expected-warning {{reading variable 'foo' requires holding mutex 'mu'}}
+  }
+  
+  Foo returns_value_releases_lock_before_return() UNLOCK_FUNCTION(mu) {
+mu.Unlock();
+return foo;   // expected-warning {{reading variable 'foo' requires holding mutex 'mu'}}
+  }
+
+  Foo &returns_ref_not_locked() {
+return foo;   // expected-warning {{returning variable 'foo' by reference requires holding mutex 'mu'}}
+  }
+
+  Foo &returns_ref_locked() {
+MutexLock lock(&mu);
+return foo;   // expected-warning {{returning variable 'foo' by reference requires holding mutex 'mu'}}
+  }
+
+  Foo &returns_ref_shared_locks_required() SHARED_LOCKS_REQUIRED(mu) {
+return foo;   // expected-warning {{returning variable 'foo' by reference requires holding mutex 'mu' exclusively}}
+  }
+
+  Foo &returns_ref_exclusive_locks_required() EXCLUSIVE_LOCKS_REQUIRED(mu) {
+return foo;
+  }
+
+  Foo &returns_ref_releases_lock_after_return() UNLOCK_FUNCTION(mu) {
+MutexLock lock(&mu, true);
+return foo;   // expected-warning {{returning variable 'foo' by reference requires holding mutex 'mu' exclusively}}
+  }
+
+  Foo& returns_ref_releases_lock_before_return() UNLOCK_FUNCTION(mu) {
+mu.Unlock();
+return foo;   // // expected-warning {{returning variable 'foo' by reference requires holding mutex 'mu' exclusively}}
+  }
+  
+  Foo &returns_ref_aquires_lock() EXCLUSIVE_LOCK_FUNCTION(mu) {
+mu.Lock();
+return foo;
+  }
+  
+  const Foo &returns_constref_shared_locks_required() SHARED_LOCKS_REQUIRED(mu) {
+return foo;
+  }
+  
+  Foo *returns_ptr() {
+return &foo;  // FIXME -- Do we want to warn on this ?
+  }
+
+  Foo &returns_ref2() {
+return *foo_ptr;  // expected-warning {{returning the value that 'foo_ptr' points to by reference requires holding mutex 'mu' exclusively}}
+  }
+
+};
+
 
 }  // end namespace PassByRefTest
 
Index: clang/lib/Sema/AnalysisBasedWarnings.cpp
===
--- clang/lib/Sema/AnalysisBasedWarnings.cpp
+++ clang/lib/Sema/AnalysisBasedWarnings.cpp
@@ -1983,6 +1983,12 @@
 case POK_PtPassByRef:
   DiagID = diag::warn_pt_guarded_pass_by_reference;
   break;
+case POK_ReturnByRef:
+  DiagID = diag::warn_guarded_return_by_reference;
+  break;
+case POK_PtReturnByRef:
+  DiagID = diag::warn_pt_guarded_return_by_reference;
+  break;
   }
   PartialDiagnosticAt Warning(Loc, S.PDiag(DiagID) << Kind
<< D
@@ -2013,6 +2019,12 @@
 case POK_PtPassByRef:
   DiagID = diag::warn_pt_guarded_pass_by_reference;
   break;
+case POK_ReturnByRef:
+  DiagID = diag::warn_guarded_return_by_reference;
+  break;
+case POK_PtReturnByRef:
+  DiagID = diag::warn_pt_guarded_return_by_reference;
+  break;
   }
   PartialDiagnosticAt Warning(Loc, S.PDiag(DiagID) << Kind
<< D
Index: clang/lib/Analysis/ThreadSafety.cpp
===
--- clang/lib/Analysis/ThreadSafety.cpp
+++ clang/lib/Analysis/ThreadSafety.cpp
@@ -1008,7 +1008,7 @@
   threadSafety::SExprBuilder SxBuilder;
 
   ThreadSafetyHandler &Handler;
-  const CXXMethodDecl *CurrentMethod = nullptr;
+  const FunctionDecl *CurrentFunction;
   LocalVariableMap LocalVarMap;
   FactManager FactMan;
   std::vector BlockInfo;
@@ -1243,10 +1243,10 @@
 
   // Members are in scope from methods

[clang] [clang analysis][thread-safety] Handle return-by-reference... (PR #67428)

[Clement Courbet via cfe-commits]

https://github.com/legrosbuffle created 
https://github.com/llvm/llvm-project/pull/67428

...of guarded variables, when the function is not marked as requiring locks:

```
class Return {
  Mutex mu;
  Foo foo GUARDED_BY(mu);

  Foo &returns_ref_locked() {
MutexLock lock(&mu);
return foo;  // BAD
  }

  Foo &returns_ref_locks_required() SHARED_LOCKS_REQUIRED(mu) {
return foo;  // OK
  }
};
```

This is carried over from phabricator: https://reviews.llvm.org/D153131

>From c14b45843e0a72d96c4898fc7d7ee72bcd619988 Mon Sep 17 00:00:00 2001
From: Clement Courbet 
Date: Tue, 26 Sep 2023 15:22:09 +0200
Subject: [PATCH] [clang analysis][thread-safety] Handle return-by-reference...

...of guarded variables, when the function is not marked as requiring locks:

```
class Return {
  Mutex mu;
  Foo foo GUARDED_BY(mu);

  Foo &returns_ref_locked() {
MutexLock lock(&mu);
return foo;  // BAD
  }

  Foo &returns_ref_locks_required() SHARED_LOCKS_REQUIRED(mu) {
return foo;  // OK
  }
};
```

This is carried over from phabricator: https://reviews.llvm.org/D153131
---
 .../clang/Analysis/Analyses/ThreadSafety.h|  8 +-
 .../clang/Basic/DiagnosticSemaKinds.td| 10 ++-
 clang/lib/Analysis/ThreadSafety.cpp   | 80 +--
 clang/lib/Sema/AnalysisBasedWarnings.cpp  | 12 +++
 .../SemaCXX/warn-thread-safety-analysis.cpp   | 79 ++
 5 files changed, 161 insertions(+), 28 deletions(-)

diff --git a/clang/include/clang/Analysis/Analyses/ThreadSafety.h 
b/clang/include/clang/Analysis/Analyses/ThreadSafety.h
index 1808d1d71e05d2c..0866b09bab2995e 100644
--- a/clang/include/clang/Analysis/Analyses/ThreadSafety.h
+++ b/clang/include/clang/Analysis/Analyses/ThreadSafety.h
@@ -47,7 +47,13 @@ enum ProtectedOperationKind {
   POK_PassByRef,
 
   /// Passing a pt-guarded variable by reference.
-  POK_PtPassByRef
+  POK_PtPassByRef,
+
+  /// Returning a guarded variable by reference.
+  POK_ReturnByRef,
+
+  /// Returning a pt-guarded variable by reference.
+  POK_PtReturnByRef,
 };
 
 /// This enum distinguishes between different kinds of lock actions. For
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td 
b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index f4eb02fd9570c2f..8e423ea7691de88 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -3858,7 +3858,7 @@ def warn_fun_requires_negative_cap : Warning<
   "calling function %0 requires negative capability '%1'">,
   InGroup, DefaultIgnore;
 
-// Thread safety warnings on pass by reference
+// Thread safety warnings on pass/return by reference
 def warn_guarded_pass_by_reference : Warning<
   "passing variable %1 by reference requires holding %0 "
   "%select{'%2'|'%2' exclusively}3">,
@@ -3867,6 +3867,14 @@ def warn_pt_guarded_pass_by_reference : Warning<
   "passing the value that %1 points to by reference requires holding %0 "
   "%select{'%2'|'%2' exclusively}3">,
   InGroup, DefaultIgnore;
+def warn_guarded_return_by_reference : Warning<
+  "returning variable %1 by reference requires holding %0 "
+  "%select{'%2'|'%2' exclusively}3">,
+  InGroup, DefaultIgnore;
+def warn_pt_guarded_return_by_reference : Warning<
+  "returning the value that %1 points to by reference requires holding %0 "
+  "%select{'%2'|'%2' exclusively}3">,
+  InGroup, DefaultIgnore;
 
 // Imprecise thread safety warnings
 def warn_variable_requires_lock : Warning<
diff --git a/clang/lib/Analysis/ThreadSafety.cpp 
b/clang/lib/Analysis/ThreadSafety.cpp
index f160cf4d013c78d..77b12f750e18a45 100644
--- a/clang/lib/Analysis/ThreadSafety.cpp
+++ b/clang/lib/Analysis/ThreadSafety.cpp
@@ -1008,7 +1008,7 @@ class ThreadSafetyAnalyzer {
   threadSafety::SExprBuilder SxBuilder;
 
   ThreadSafetyHandler &Handler;
-  const CXXMethodDecl *CurrentMethod = nullptr;
+  const FunctionDecl *CurrentFunction;
   LocalVariableMap LocalVarMap;
   FactManager FactMan;
   std::vector BlockInfo;
@@ -1243,10 +1243,10 @@ bool ThreadSafetyAnalyzer::inCurrentScope(const 
CapabilityExpr &CapE) {
 
   // Members are in scope from methods of the same class.
   if (const auto *P = dyn_cast(SExp)) {
-if (!CurrentMethod)
+if (!isa_and_nonnull(CurrentFunction))
   return false;
 const ValueDecl *VD = P->clangDecl();
-return VD->getDeclContext() == CurrentMethod->getDeclContext();
+return VD->getDeclContext() == CurrentFunction->getDeclContext();
   }
 
   return false;
@@ -1541,6 +1541,8 @@ class BuildLockset : public 
ConstStmtVisitor {
 
   ThreadSafetyAnalyzer *Analyzer;
   FactSet FSet;
+  // The fact set for the function on exit.
+  const FactSet &FunctionExitFSet;
   /// Maps constructed objects to `this` placeholder prior to initialization.
   llvm::SmallDenseMap ConstructedObjects;
   LocalVariableMap::Context LVarCtx;
@@ -1566,9 +1568,11 @@ class BuildLockset : public 
ConstStmtVisitor {
 bool SkipFirstParam = false);
 
 public:
-  BuildLock

[clang] [clang][Interp] Three-way comparisons (PR #65901)

[Erich Keane via cfe-commits]

https://github.com/erichkeane commented:

This has been sitting a while, else I wouldn't feel comfortable reviewing (as I 
haven't been keeping up on these as much as I should have).  However, I saw 1 
issue with the comment I think, else I'm OK with this.  So if you can fix that, 
I'll approve.

https://github.com/llvm/llvm-project/pull/65901
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang][Interp] Three-way comparisons (PR #65901)

[Erich Keane via cfe-commits]

@@ -253,6 +253,29 @@ bool ByteCodeExprGen::VisitBinaryOperator(const 
BinaryOperator *BO) {
 return this->delegate(RHS);
   }
 
+  // Special case for C++'s three-way/spaceship operator <=>, which
+  // returns a std::strong_ordering (which is class, so doesn't have a

erichkeane wrote:

So this isn't really true, see the rules here: 
https://en.cppreference.com/w/cpp/language/default_comparisons

It can end up returning a `std::weak_ordering` or `std::partial_ordering` as 
well.  The rest here is true, but I wanted to make sure you know, since these 
have slightly different semantics that I don't see handled here?

https://github.com/llvm/llvm-project/pull/65901
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang][Interp] Three-way comparisons (PR #65901)

[Erich Keane via cfe-commits]

@@ -253,6 +253,29 @@ bool ByteCodeExprGen::VisitBinaryOperator(const 
BinaryOperator *BO) {
 return this->delegate(RHS);
   }
 
+  // Special case for C++'s three-way/spaceship operator <=>, which
+  // returns a std::strong_ordering (which is class, so doesn't have a

erichkeane wrote:

I already put that up, but I see lower you refer to all 3, so this is just a 
request to modify that comment as necessary.

https://github.com/llvm/llvm-project/pull/65901
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang][Interp] Three-way comparisons (PR #65901)

[Erich Keane via cfe-commits]

https://github.com/erichkeane edited 
https://github.com/llvm/llvm-project/pull/65901
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang analysis][thread-safety] Handle return-by-reference... (PR #67428)

[via cfe-commits]

llvmbot wrote:




@llvm/pr-subscribers-clang


Changes

...of guarded variables, when the function is not marked as requiring locks:

```
class Return {
  Mutex mu;
  Foo foo GUARDED_BY(mu);

  Foo &returns_ref_locked() {
MutexLock lock(&mu);
return foo;  // BAD
  }

  Foo &returns_ref_locks_required() SHARED_LOCKS_REQUIRED(mu) {
return foo;  // OK
  }
};
```

This is carried over from phabricator: https://reviews.llvm.org/D153131

---
Full diff: https://github.com/llvm/llvm-project/pull/67428.diff


5 Files Affected:

- (modified) clang/include/clang/Analysis/Analyses/ThreadSafety.h (+7-1) 
- (modified) clang/include/clang/Basic/DiagnosticSemaKinds.td (+9-1) 
- (modified) clang/lib/Analysis/ThreadSafety.cpp (+54-26) 
- (modified) clang/lib/Sema/AnalysisBasedWarnings.cpp (+12) 
- (modified) clang/test/SemaCXX/warn-thread-safety-analysis.cpp (+79) 


``diff
diff --git a/clang/include/clang/Analysis/Analyses/ThreadSafety.h 
b/clang/include/clang/Analysis/Analyses/ThreadSafety.h
index 1808d1d71e05d2c..0866b09bab2995e 100644
--- a/clang/include/clang/Analysis/Analyses/ThreadSafety.h
+++ b/clang/include/clang/Analysis/Analyses/ThreadSafety.h
@@ -47,7 +47,13 @@ enum ProtectedOperationKind {
   POK_PassByRef,
 
   /// Passing a pt-guarded variable by reference.
-  POK_PtPassByRef
+  POK_PtPassByRef,
+
+  /// Returning a guarded variable by reference.
+  POK_ReturnByRef,
+
+  /// Returning a pt-guarded variable by reference.
+  POK_PtReturnByRef,
 };
 
 /// This enum distinguishes between different kinds of lock actions. For
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td 
b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index f4eb02fd9570c2f..8e423ea7691de88 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -3858,7 +3858,7 @@ def warn_fun_requires_negative_cap : Warning<
   "calling function %0 requires negative capability '%1'">,
   InGroup, DefaultIgnore;
 
-// Thread safety warnings on pass by reference
+// Thread safety warnings on pass/return by reference
 def warn_guarded_pass_by_reference : Warning<
   "passing variable %1 by reference requires holding %0 "
   "%select{'%2'|'%2' exclusively}3">,
@@ -3867,6 +3867,14 @@ def warn_pt_guarded_pass_by_reference : Warning<
   "passing the value that %1 points to by reference requires holding %0 "
   "%select{'%2'|'%2' exclusively}3">,
   InGroup, DefaultIgnore;
+def warn_guarded_return_by_reference : Warning<
+  "returning variable %1 by reference requires holding %0 "
+  "%select{'%2'|'%2' exclusively}3">,
+  InGroup, DefaultIgnore;
+def warn_pt_guarded_return_by_reference : Warning<
+  "returning the value that %1 points to by reference requires holding %0 "
+  "%select{'%2'|'%2' exclusively}3">,
+  InGroup, DefaultIgnore;
 
 // Imprecise thread safety warnings
 def warn_variable_requires_lock : Warning<
diff --git a/clang/lib/Analysis/ThreadSafety.cpp 
b/clang/lib/Analysis/ThreadSafety.cpp
index f160cf4d013c78d..77b12f750e18a45 100644
--- a/clang/lib/Analysis/ThreadSafety.cpp
+++ b/clang/lib/Analysis/ThreadSafety.cpp
@@ -1008,7 +1008,7 @@ class ThreadSafetyAnalyzer {
   threadSafety::SExprBuilder SxBuilder;
 
   ThreadSafetyHandler &Handler;
-  const CXXMethodDecl *CurrentMethod = nullptr;
+  const FunctionDecl *CurrentFunction;
   LocalVariableMap LocalVarMap;
   FactManager FactMan;
   std::vector BlockInfo;
@@ -1243,10 +1243,10 @@ bool ThreadSafetyAnalyzer::inCurrentScope(const 
CapabilityExpr &CapE) {
 
   // Members are in scope from methods of the same class.
   if (const auto *P = dyn_cast(SExp)) {
-if (!CurrentMethod)
+if (!isa_and_nonnull(CurrentFunction))
   return false;
 const ValueDecl *VD = P->clangDecl();
-return VD->getDeclContext() == CurrentMethod->getDeclContext();
+return VD->getDeclContext() == CurrentFunction->getDeclContext();
   }
 
   return false;
@@ -1541,6 +1541,8 @@ class BuildLockset : public 
ConstStmtVisitor {
 
   ThreadSafetyAnalyzer *Analyzer;
   FactSet FSet;
+  // The fact set for the function on exit.
+  const FactSet &FunctionExitFSet;
   /// Maps constructed objects to `this` placeholder prior to initialization.
   llvm::SmallDenseMap ConstructedObjects;
   LocalVariableMap::Context LVarCtx;
@@ -1566,9 +1568,11 @@ class BuildLockset : public 
ConstStmtVisitor {
 bool SkipFirstParam = false);
 
 public:
-  BuildLockset(ThreadSafetyAnalyzer *Anlzr, CFGBlockInfo &Info)
+  BuildLockset(ThreadSafetyAnalyzer *Anlzr, CFGBlockInfo &Info,
+   const FactSet &FunctionExitFSet)
   : ConstStmtVisitor(), Analyzer(Anlzr), FSet(Info.EntrySet),
-LVarCtx(Info.EntryContext), CtxIndex(Info.EntryIndex) {}
+FunctionExitFSet(FunctionExitFSet), LVarCtx(Info.EntryContext),
+CtxIndex(Info.EntryIndex) {}
 
   void VisitUnaryOperator(const UnaryOperator *UO);
   void VisitBinaryOperator(const BinaryOperator *BO);
@@ -1577,6 +1581,7 @@ class BuildLockset : 

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

https://github.com/AaronBallman resolved 
https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

https://github.com/AaronBallman resolved 
https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

https://github.com/AaronBallman resolved 
https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

https://github.com/AaronBallman resolved 
https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

https://github.com/AaronBallman resolved 
https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] Diagnose problematic uses of constructor/destructor attribute (PR #67360)

[Aaron Ballman via cfe-commits]

https://github.com/AaronBallman resolved 
https://github.com/llvm/llvm-project/pull/67360
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang][Interp] Add IntegralAP for arbitrary-precision integers (PR #65844)

[Erich Keane via cfe-commits]

Timm =?utf-8?q?Bäder?= ,
Timm =?utf-8?q?Bäder?= ,
Timm =?utf-8?q?Bäder?= 
Message-ID:
In-Reply-To: 



@@ -1596,6 +1599,9 @@ bool 
ByteCodeExprGen::visitZeroInitializer(QualType QT,
 return this->emitZeroSint64(E);
   case PT_Uint64:
 return this->emitZeroUint64(E);
+  case PT_IntAP:
+  case PT_IntAPS:
+assert(false);

erichkeane wrote:

Should this be an `llvm_unreachable`?

https://github.com/llvm/llvm-project/pull/65844
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits