[libclc] r315170 - ldexp: Fix double precision function return type
Author: jvesely Date: Sat Oct 7 23:56:14 2017 New Revision: 315170 URL: http://llvm.org/viewvc/llvm-project?rev=315170&view=rev Log: ldexp: Fix double precision function return type Fixes ~1200 external calls from nvtpx library. Reviewer: Jeroen Ketema Signed-off-by: Jan Vesely Modified: libclc/trunk/generic/include/math/clc_ldexp.h Modified: libclc/trunk/generic/include/math/clc_ldexp.h URL: http://llvm.org/viewvc/llvm-project/libclc/trunk/generic/include/math/clc_ldexp.h?rev=315170&r1=315169&r2=315170&view=diff == --- libclc/trunk/generic/include/math/clc_ldexp.h (original) +++ libclc/trunk/generic/include/math/clc_ldexp.h Sat Oct 7 23:56:14 2017 @@ -2,5 +2,5 @@ _CLC_DEF _CLC_OVERLOAD float __clc_ldexp #ifdef cl_khr_fp64 #pragma OPENCL EXTENSION cl_khr_fp64 : enable - _CLC_DEF _CLC_OVERLOAD float __clc_ldexp(double, int); + _CLC_DEF _CLC_OVERLOAD double __clc_ldexp(double, int); #endif ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D36562: [Bitfield] Make the bitfield a separate location if it has width of legal integer type and its bit offset is naturally aligned for the type
hfinkel added inline comments.
Comment at: include/clang/Basic/DiagnosticDriverKinds.td:335
+def warn_drv_fine_grained_bitfield_accesses_ignored : Warning<
+ "option '-ffine-grained-bitfield-accesses' cannot be enabled together with
sanitizer; flag ignored">,
+ InGroup;
with a sanitizer
Comment at: include/clang/Driver/Options.td:1041
+ "ffine-grained-bitfield-accesses">, Group, Flags<[CC1Option]>,
+ HelpText<"Use separate access for bitfields with legal widths and
alignments.">;
+def fno_fine_grained_bitfield_accesses : Flag<["-"],
access -> accesses
Comment at: include/clang/Driver/Options.td:1044
+ "fno-fine-grained-bitfield-accesses">, Group,
Flags<[CC1Option]>,
+ HelpText<"Use a big integer wrap for a consecutive run of bitfields.">;
+
Use large-integer access for consecutive bitfield runs.
Comment at: include/clang/Frontend/CodeGenOptions.def:182
CODEGENOPT(SoftFloat , 1, 0) ///< -soft-float.
+CODEGENOPT(FineGrainedBitfieldAccesses, 1, 0) ///< Use separate access for
bitfields
+ ///< with legal widths and
alignments.
These lines are too long.
Comment at: lib/CodeGen/CGRecordLayoutBuilder.cpp:449
+// Otherwise, try to add bitfields to the run.
+if (Run != FieldEnd && !IsBetterAsSingleFieldRun(Run) &&
+Field != FieldEnd && !IsBetterAsSingleFieldRun(Field) &&
Why do you have the `IsBetterAsSingleFieldRun(Run)` check here (where we'll
evaluate it multiple times (for all of the fields in the run)). Can't you make
the predicate above directly?
// Any non-zero-length bitfield can start a new run.
if (Field->getBitWidthValue(Context) != 0 &&
!IsBetterAsSingleFieldRun(Field)) {
Run = Field;
StartBitOffset = getFieldBitOffset(*Field);
...
Repository:
rL LLVM
https://reviews.llvm.org/D36562
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D24933: Enable configuration files in clang
hfinkel added a comment. Some suggested improvements to the English in the documentation... Comment at: docs/UsersManual.rst:700 + +Configuration files group command line options and allow to specify all of +them just by referencing the configuration file. They may be used, for Configuration files group command-line options and allow all of them to be specified just by ... Comment at: docs/UsersManual.rst:702 +them just by referencing the configuration file. They may be used, for +instance, to collect options required to tune compilation for particular +target, such as -L, -I, -l, --sysroot, codegen options etc. instance -> example "for instance" and "for example" are essentially the same, but for consistency, "for example" is much more common in this document, so we should standardize. Comment at: docs/UsersManual.rst:703 +instance, to collect options required to tune compilation for particular +target, such as -L, -I, -l, --sysroot, codegen options etc. + options, etc. Comment at: docs/UsersManual.rst:706 +The command line option `--config` can be used to specify configuration +file in a Clang invocation. For instance: + instance -> example Comment at: docs/UsersManual.rst:714 +If the provided argument contains a directory separator, it is considered as +a file path, options are read from that file. Otherwise the argument is treated +as a file name and is searched for sequentially in the directories: and options are Comment at: docs/UsersManual.rst:719 +- the directory where Clang executable resides. +Both user and system directory for configuration files are specified during +clang build using cmake parameters, CLANG_CONFIG_FILE_USER_DIR and directory -> directories Comment at: docs/UsersManual.rst:720 +Both user and system directory for configuration files are specified during +clang build using cmake parameters, CLANG_CONFIG_FILE_USER_DIR and +CLANG_CONFIG_FILE_SYSTEM_DIR respectively. The first found file is used. It is cmake -> CMake Comment at: docs/UsersManual.rst:721 +clang build using cmake parameters, CLANG_CONFIG_FILE_USER_DIR and +CLANG_CONFIG_FILE_SYSTEM_DIR respectively. The first found file is used. It is +an error if the required file cannot be found. The first file found is used. Comment at: docs/UsersManual.rst:724 + +Another way to specify configuration file is to encode it in executable name. For +instance, if Clang executable is named `armv7l-clang` (it may be a symbolic link specify a configuration file Comment at: docs/UsersManual.rst:725 +Another way to specify configuration file is to encode it in executable name. For +instance, if Clang executable is named `armv7l-clang` (it may be a symbolic link +to `clang`), then Clang will search file `armv7l.cfg` in the directory where Clang if the Clang executable Comment at: docs/UsersManual.rst:726 +instance, if Clang executable is named `armv7l-clang` (it may be a symbolic link +to `clang`), then Clang will search file `armv7l.cfg` in the directory where Clang +resides. will search for file Comment at: docs/UsersManual.rst:729 + +If driver mode is specified in invocation, Clang tries to find file specific for +the specified mode. For instance, if executable file is `x86_64-clang-cl`, Clang If a driver mode Comment at: docs/UsersManual.rst:729 + +If driver mode is specified in invocation, Clang tries to find file specific for +the specified mode. For instance, if executable file is `x86_64-clang-cl`, Clang hfinkel wrote: > If a driver mode to find a file Comment at: docs/UsersManual.rst:730 +If driver mode is specified in invocation, Clang tries to find file specific for +the specified mode. For instance, if executable file is `x86_64-clang-cl`, Clang +first looks for `x86_64-cl.cfg` and if it is not found, looks for `x86_64.cfg'. if the executable is named `x86_64-clang-cl` Comment at: docs/UsersManual.rst:733 + +If command line contains options that effectively changes target architecture +(these are -m32, -EL and some other) and configuration file starts with architecture If the command line Comment at: docs/UsersManual.rst:733 + +If command line contains options that effectively changes target architecture +(these are -m32, -EL and some other) and configuration file starts with architecture hfinkel wrote: > If the command line changes -> change Comment at: docs/UsersManual.rst:734 +If command line contains options that effecti
[PATCH] D38101: [Sema] Diagnose tautological comparison with type's min/max values
lebedev.ri updated this revision to Diff 118143.
lebedev.ri added a comment.
Rework `AnalyzeComparison()` and `DiagnoseOutOfRangeComparison()` - avoid that
fast-return path in `DiagnoseOutOfRangeComparison()`
`AnalyzeComparison()` is now a bit smarter, and it
1. calls `CheckTautologicalComparison()`
2. returns if it diagnosed anything
3. else, calls `DiagnoseOutOfRangeComparison()`
4. returns if it diagnosed anything.
`check-clang-sema` and `check-clang-semacxx` are good.
Stage-2 build still good.
Repository:
rL LLVM
https://reviews.llvm.org/D38101
Files:
docs/ReleaseNotes.rst
include/clang/Basic/DiagnosticGroups.td
include/clang/Basic/DiagnosticSemaKinds.td
lib/Sema/SemaChecking.cpp
test/Sema/outof-range-constant-compare.c
test/Sema/tautological-constant-compare.c
test/Sema/tautological-unsigned-zero-compare.c
Index: test/Sema/tautological-unsigned-zero-compare.c
===
--- test/Sema/tautological-unsigned-zero-compare.c
+++ test/Sema/tautological-unsigned-zero-compare.c
@@ -1,47 +1,370 @@
// RUN: %clang_cc1 -fsyntax-only -DTEST -verify %s
// RUN: %clang_cc1 -fsyntax-only -Wno-tautological-unsigned-zero-compare -verify %s
+// RUN: %clang_cc1 -fsyntax-only -DTEST -verify -x c++ %s
+// RUN: %clang_cc1 -fsyntax-only -Wno-tautological-unsigned-zero-compare -verify -x c++ %s
-unsigned value(void);
+unsigned uvalue(void);
+signed int svalue(void);
-int main() {
- unsigned un = value();
+#define macro(val) val
+
+#ifdef __cplusplus
+template
+void TFunc() {
+ // Make sure that we do warn for normal variables in template functions !
+ unsigned char c = svalue();
+#ifdef TEST
+ if (c < 0) // expected-warning {{comparison of unsigned expression < 0 is always false}}
+ return;
+#else
+ if (c < 0)
+ return;
+#endif
+
+ if (c < macro(0))
+ return;
+
+ T v = svalue();
+ if (v < 0)
+ return;
+}
+#endif
+
+int main()
+{
+#ifdef __cplusplus
+ TFunc();
+ TFunc();
+#endif
+
+ unsigned un = uvalue();
#ifdef TEST
+ if (un == 0)
+ return 0;
+ if (un != 0)
+ return 0;
if (un < 0) // expected-warning {{comparison of unsigned expression < 0 is always false}}
-return 0;
+ return 0;
+ if (un <= 0)
+ return 0;
+ if (un > 0)
+ return 0;
if (un >= 0) // expected-warning {{comparison of unsigned expression >= 0 is always true}}
-return 0;
+ return 0;
+
+ if (0 == un)
+ return 0;
+ if (0 != un)
+ return 0;
+ if (0 < un)
+ return 0;
if (0 <= un) // expected-warning {{comparison of 0 <= unsigned expression is always true}}
-return 0;
+ return 0;
if (0 > un) // expected-warning {{comparison of 0 > unsigned expression is always false}}
-return 0;
- if (un < 0U) // expected-warning {{comparison of unsigned expression < 0 is always false}}
-return 0;
- if (un >= 0U) // expected-warning {{comparison of unsigned expression >= 0 is always true}}
-return 0;
- if (0U <= un) // expected-warning {{comparison of 0 <= unsigned expression is always true}}
-return 0;
- if (0U > un) // expected-warning {{comparison of 0 > unsigned expression is always false}}
-return 0;
+ return 0;
+ if (0 >= un)
+ return 0;
+
+ if (un == 0UL)
+ return 0;
+ if (un != 0UL)
+ return 0;
+ if (un < 0UL) // expected-warning {{comparison of unsigned expression < 0 is always false}}
+ return 0;
+ if (un <= 0UL)
+ return 0;
+ if (un > 0UL)
+ return 0;
+ if (un >= 0UL) // expected-warning {{comparison of unsigned expression >= 0 is always true}}
+ return 0;
+
+ if (0UL == un)
+ return 0;
+ if (0UL != un)
+ return 0;
+ if (0UL < un)
+ return 0;
+ if (0UL <= un) // expected-warning {{comparison of 0 <= unsigned expression is always true}}
+ return 0;
+ if (0UL > un) // expected-warning {{comparison of 0 > unsigned expression is always false}}
+ return 0;
+ if (0UL >= un)
+ return 0;
#else
// expected-no-diagnostics
+ if (un == 0)
+ return 0;
+ if (un != 0)
+ return 0;
if (un < 0)
-return 0;
+ return 0;
+ if (un <= 0)
+ return 0;
+ if (un > 0)
+ return 0;
if (un >= 0)
-return 0;
+ return 0;
+
+ if (0 == un)
+ return 0;
+ if (0 != un)
+ return 0;
+ if (0 < un)
+ return 0;
if (0 <= un)
-return 0;
+ return 0;
if (0 > un)
-return 0;
- if (un < 0U)
-return 0;
- if (un >= 0U)
-return 0;
- if (0U <= un)
-return 0;
- if (0U > un)
-return 0;
+ return 0;
+ if (0 >= un)
+ return 0;
+
+ if (un == 0UL)
+ return 0;
+ if (un != 0UL)
+ return 0;
+ if (un < 0UL)
+ return 0;
+ if (un <= 0UL)
+ return 0;
+ if (un > 0UL)
+ return 0;
+ if (un >= 0UL)
+ return 0;
+
+ if (0UL == un)
+ return 0;
+ if (0UL != un)
+ return 0;
+ if (0UL < un)
+ return 0;
+ if (0UL <= un)
+ return 0;
+ if (0UL > un)
+ return 0;
+ if (0UL
[PATCH] D37436: Initial implementation of C attributes (WG14 N2137)
aaron.ballman added inline comments.
Comment at: clang/lib/Parse/ParseDecl.cpp:5973-5974
+
+// If there are attributes following the identifier list, parse them and
+// prohibit them.
+MaybeParseCXX11Attributes(FnAttrs);
rsmith wrote:
> Do you anticipate people trying this? Is the concern here that a function
> declarator can normally be followed by attributes, and so consistency might
> imply that we allow attributes on the function declarator after an
> identifier-list instead?
I don't anticipate this being a common thing for people to try, but it is to
prevent an ambiguity:
```
void fp(a, b) [[foo]] int a, b; { }
```
Does [[foo]] appertain to the function type or to the declarations `a` and `b`.
N2165 prohibits attributes from appearing in this location.
Comment at: clang/lib/Parse/ParseTentative.cpp:592
bool OuterMightBeMessageSend) {
- if (Tok.is(tok::kw_alignas))
+ if (Tok.is(tok::kw_alignas) && getLangOpts().CPlusPlus11)
return CAK_AttributeSpecifier;
rsmith wrote:
> This change is redundant. We wouldn't lex a `kw_alignas` token outside C++11.
Ah, I thought that we would hit this case for `_Alignas` as well, but I see
that's a different keyword. Just to double-check -- we don't expect
`-fdouble-square-bracket-attributes` to enable `alignas` in C++98, correct?
https://reviews.llvm.org/D37436
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D37436: Initial implementation of C attributes (WG14 N2137)
aaron.ballman updated this revision to Diff 118141.
aaron.ballman marked 5 inline comments as done.
aaron.ballman added a comment.
Corrected review feedback from Richard.
- Changed the cc1 option to -fdouble-square-bracket-attributes
- Corrected regression with opaque-enum-declarations
https://reviews.llvm.org/D37436
Files:
../llvm/tools/clang/include/clang/Basic/Attr.td
../llvm/tools/clang/include/clang/Basic/Attributes.h
../llvm/tools/clang/include/clang/Basic/LangOptions.def
../llvm/tools/clang/include/clang/Driver/Options.td
../llvm/tools/clang/include/clang/Parse/Parser.h
../llvm/tools/clang/include/clang/Sema/AttributeList.h
../llvm/tools/clang/lib/Frontend/CompilerInvocation.cpp
../llvm/tools/clang/lib/Lex/Lexer.cpp
../llvm/tools/clang/lib/Parse/ParseDecl.cpp
../llvm/tools/clang/lib/Parse/ParseDeclCXX.cpp
../llvm/tools/clang/lib/Sema/AttributeList.cpp
../llvm/tools/clang/test/Misc/ast-dump-c-attr.c
../llvm/tools/clang/test/Parser/c2x-attributes.c
../llvm/tools/clang/test/Parser/c2x-attributes.m
../llvm/tools/clang/test/Sema/attr-deprecated-c2x.c
../llvm/tools/clang/utils/TableGen/ClangAttrEmitter.cpp
Index: ../llvm/tools/clang/utils/TableGen/ClangAttrEmitter.cpp
===
--- ../llvm/tools/clang/utils/TableGen/ClangAttrEmitter.cpp
+++ ../llvm/tools/clang/utils/TableGen/ClangAttrEmitter.cpp
@@ -58,7 +58,7 @@
assert(V != "GCC" && "Given a GCC spelling, which means this hasn't been"
"flattened!");
-if (V == "CXX11" || V == "Pragma")
+if (V == "CXX11" || V == "C2x" || V == "Pragma")
NS = Spelling.getValueAsString("Namespace");
bool Unset;
K = Spelling.getValueAsBitOrUnset("KnownToGCC", Unset);
@@ -1326,7 +1326,7 @@
if (Variety == "GNU") {
Prefix = " __attribute__((";
Suffix = "))";
-} else if (Variety == "CXX11") {
+} else if (Variety == "CXX11" || Variety == "C2x") {
Prefix = " [[";
Suffix = "]]";
std::string Namespace = Spellings[I].nameSpace();
@@ -2716,10 +2716,14 @@
// If this is the C++11 variety, also add in the LangOpts test.
if (Variety == "CXX11")
Test += " && LangOpts.CPlusPlus11";
+ else if (Variety == "C2x")
+Test += " && LangOpts.DoubleSquareBracketAttributes";
} else if (Variety == "CXX11")
// C++11 mode should be checked against LangOpts, which is presumed to be
// present in the caller.
Test = "LangOpts.CPlusPlus11";
+else if (Variety == "C2x")
+ Test = "LangOpts.DoubleSquareBracketAttributes";
std::string TestStr =
!Test.empty() ? Test + " ? " + llvm::itostr(Version) + " : 0" : "1";
@@ -2740,7 +2744,7 @@
// and declspecs. Then generate a big switch statement for each of them.
std::vector Attrs = Records.getAllDerivedDefinitions("Attr");
std::vector Declspec, Microsoft, GNU, Pragma;
- std::map> CXX;
+ std::map> CXX, C2x;
// Walk over the list of all attributes, and split them out based on the
// spelling variety.
@@ -2756,6 +2760,8 @@
Microsoft.push_back(R);
else if (Variety == "CXX11")
CXX[SI.nameSpace()].push_back(R);
+ else if (Variety == "C2x")
+C2x[SI.nameSpace()].push_back(R);
else if (Variety == "Pragma")
Pragma.push_back(R);
}
@@ -2775,20 +2781,25 @@
OS << "case AttrSyntax::Pragma:\n";
OS << " return llvm::StringSwitch(Name)\n";
GenerateHasAttrSpellingStringSwitch(Pragma, OS, "Pragma");
- OS << "case AttrSyntax::CXX: {\n";
- // C++11-style attributes are further split out based on the Scope.
- for (auto I = CXX.cbegin(), E = CXX.cend(); I != E; ++I) {
-if (I != CXX.begin())
- OS << " else ";
-if (I->first.empty())
- OS << "if (!Scope || Scope->getName() == \"\") {\n";
-else
- OS << "if (Scope->getName() == \"" << I->first << "\") {\n";
-OS << " return llvm::StringSwitch(Name)\n";
-GenerateHasAttrSpellingStringSwitch(I->second, OS, "CXX11", I->first);
-OS << "}";
- }
- OS << "\n}\n";
+ auto fn = [&OS](const char *Spelling, const char *Variety,
+ const std::map> &List) {
+OS << "case AttrSyntax::" << Variety << ": {\n";
+// C++11-style attributes are further split out based on the Scope.
+for (auto I = List.cbegin(), E = List.cend(); I != E; ++I) {
+ if (I != List.cbegin())
+OS << " else ";
+ if (I->first.empty())
+OS << "if (!Scope || Scope->getName() == \"\") {\n";
+ else
+OS << "if (Scope->getName() == \"" << I->first << "\") {\n";
+ OS << " return llvm::StringSwitch(Name)\n";
+ GenerateHasAttrSpellingStringSwitch(I->second, OS, Spelling, I->first);
+ OS << "}";
+}
+OS << "\n}\n";
+ };
+ fn("CXX11", "CXX", CXX);
+ fn("C2x", "C", C2x);
OS << "}\n";
}
@@ -2809,10 +2820,11 @@
<< StringSwitch(Spellings[I].variety())
.Case("GNU", 0)
.Case(
[PATCH] D38667: AMDGPU: Parse r600 CPU name early and expose FMAF capability
jvesely created this revision.
Herald added subscribers: t-tye, tpr, dstuttard, nhaehnle, wdng, kzhuravl.
Improve amdgcn macro test
Repository:
rL LLVM
https://reviews.llvm.org/D38667
Files:
lib/Basic/Targets/AMDGPU.cpp
test/Preprocessor/predefined-arch-macros.c
Index: test/Preprocessor/predefined-arch-macros.c
===
--- test/Preprocessor/predefined-arch-macros.c
+++ test/Preprocessor/predefined-arch-macros.c
@@ -2325,10 +2325,20 @@
// RUN: -target amdgcn-unknown-unknown \
// RUN: | FileCheck -match-full-lines %s -check-prefix=CHECK_AMDGCN
// CHECK_AMDGCN: #define __AMDGCN__ 1
+// CHECK_AMDGCN: #define __HAS_FMAF__ 1
+// CHECK_AMDGCN: #define __HAS_FP64__ 1
+// CHECK_AMDGCN: #define __HAS_LDEXPF__ 1
// Begin r600 tests
//
// RUN: %clang -march=amdgcn -E -dM %s -o - 2>&1 \
// RUN: -target r600-unknown-unknown \
// RUN: | FileCheck -match-full-lines %s -check-prefix=CHECK_R600
// CHECK_R600: #define __R600__ 1
+// CHECK_R600-NOT: #define __HAS_FMAF__ 1
+
+// RUN: %clang -march=amdgcn -mcpu=cypress -E -dM %s -o - 2>&1 \
+// RUN: -target r600-unknown-unknown \
+// RUN: | FileCheck -match-full-lines %s -check-prefix=CHECK_R600_FP64
+// CHECK_R600_FP64-DAG: #define __R600__ 1
+// CHECK_R600_FP64-DAG: #define __HAS_FMAF__ 1
Index: lib/Basic/Targets/AMDGPU.cpp
===
--- lib/Basic/Targets/AMDGPU.cpp
+++ lib/Basic/Targets/AMDGPU.cpp
@@ -304,14 +304,20 @@
AMDGPUTargetInfo::AMDGPUTargetInfo(const llvm::Triple &Triple,
const TargetOptions &Opts)
-: TargetInfo(Triple), GPU(isAMDGCN(Triple) ? GK_GFX6 : GK_R600),
+: TargetInfo(Triple),
+ GPU(isAMDGCN(Triple) ? GK_GFX6 : parseR600Name(Opts.CPU)),
hasFP64(false), hasFMAF(false), hasLDEXPF(false),
AS(isGenericZero(Triple)) {
if (getTriple().getArch() == llvm::Triple::amdgcn) {
hasFP64 = true;
hasFMAF = true;
hasLDEXPF = true;
}
+ if (getTriple().getArch() == llvm::Triple::r600) {
+if (GPU == GK_EVERGREEN_DOUBLE_OPS || GPU == GK_CAYMAN) {
+ hasFMAF = true;
+}
+ }
auto IsGenericZero = isGenericZero(Triple);
resetDataLayout(getTriple().getArch() == llvm::Triple::amdgcn
? (IsGenericZero ? DataLayoutStringSIGenericIsZero
Index: test/Preprocessor/predefined-arch-macros.c
===
--- test/Preprocessor/predefined-arch-macros.c
+++ test/Preprocessor/predefined-arch-macros.c
@@ -2325,10 +2325,20 @@
// RUN: -target amdgcn-unknown-unknown \
// RUN: | FileCheck -match-full-lines %s -check-prefix=CHECK_AMDGCN
// CHECK_AMDGCN: #define __AMDGCN__ 1
+// CHECK_AMDGCN: #define __HAS_FMAF__ 1
+// CHECK_AMDGCN: #define __HAS_FP64__ 1
+// CHECK_AMDGCN: #define __HAS_LDEXPF__ 1
// Begin r600 tests
//
// RUN: %clang -march=amdgcn -E -dM %s -o - 2>&1 \
// RUN: -target r600-unknown-unknown \
// RUN: | FileCheck -match-full-lines %s -check-prefix=CHECK_R600
// CHECK_R600: #define __R600__ 1
+// CHECK_R600-NOT: #define __HAS_FMAF__ 1
+
+// RUN: %clang -march=amdgcn -mcpu=cypress -E -dM %s -o - 2>&1 \
+// RUN: -target r600-unknown-unknown \
+// RUN: | FileCheck -match-full-lines %s -check-prefix=CHECK_R600_FP64
+// CHECK_R600_FP64-DAG: #define __R600__ 1
+// CHECK_R600_FP64-DAG: #define __HAS_FMAF__ 1
Index: lib/Basic/Targets/AMDGPU.cpp
===
--- lib/Basic/Targets/AMDGPU.cpp
+++ lib/Basic/Targets/AMDGPU.cpp
@@ -304,14 +304,20 @@
AMDGPUTargetInfo::AMDGPUTargetInfo(const llvm::Triple &Triple,
const TargetOptions &Opts)
-: TargetInfo(Triple), GPU(isAMDGCN(Triple) ? GK_GFX6 : GK_R600),
+: TargetInfo(Triple),
+ GPU(isAMDGCN(Triple) ? GK_GFX6 : parseR600Name(Opts.CPU)),
hasFP64(false), hasFMAF(false), hasLDEXPF(false),
AS(isGenericZero(Triple)) {
if (getTriple().getArch() == llvm::Triple::amdgcn) {
hasFP64 = true;
hasFMAF = true;
hasLDEXPF = true;
}
+ if (getTriple().getArch() == llvm::Triple::r600) {
+if (GPU == GK_EVERGREEN_DOUBLE_OPS || GPU == GK_CAYMAN) {
+ hasFMAF = true;
+}
+ }
auto IsGenericZero = isGenericZero(Triple);
resetDataLayout(getTriple().getArch() == llvm::Triple::amdgcn
? (IsGenericZero ? DataLayoutStringSIGenericIsZero
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[PATCH] D38101: [Sema] Diagnose tautological comparison with type's min/max values
lebedev.ri added inline comments.
Comment at: lib/Sema/SemaChecking.cpp:8719
+ // Type limit values are handled later by CheckTautologicalComparison().
+ if (IsTypeLimit(S, Other, Constant, ConstValue) && (!OtherIsBooleanType))
return;
lebedev.ri wrote:
> lebedev.ri wrote:
> > rsmith wrote:
> > > Is this necessary? (Aren't the type limit values always within the range
> > > of the type in question?)
> > >
> > > Can we avoid evaluating `Constant` a extra time here? (We already have
> > > its value in `Value`.)
> > Uhm, good question :)
> > If i remove this, `check-clang-sema` and `check-clang-semacxx` still pass.
> > I agree that it does not make much sense. Initially it was only checking
> > for `Value == 0`.
> > Git suggests that initially this branch was added by @rtrieu, maybe can
> > help.
> [[
> https://github.com/llvm-mirror/clang/commit/526e627d2bd7e8cbf630526d315c90864898d9ff#diff-93faf32157a807b1b7953f3747db08b6R4332
> | The most original version of this code ]]
> After some though i think the initial check `Value == 0` was simply to
> quickly bail out
> out of `DiagnoseOutOfRangeComparison()`, and not waste any time for the
> obvious case
> (`0`), which can't be out-of-range, ever. So i think the right behaviour
> could be:
> 1. Either simply restore the original check:
> ```
> // 0 values are handled later by CheckTautologicalComparison().
> if ((Value == 0) && (!OtherIsBooleanType))
> return;
> ```
> And add a comment there about it
> 2. Or, drop it completely
> 3. Or, perhaps refactor `CheckTautologicalComparison()`, and more or less
> call it from
> function `AnalyzeComparison()`, before calling
> `DiagnoseOutOfRangeComparison()`,
> thus completely avoiding the need to re-evaluate `Constant` there later
> on,
> and simplify the logic in the process.
>
> I personally think the `3.` *might* be best.
> WDYT?
Tried implementing `3.`.
It won't work, because `DiagnoseOutOfRangeComparison()` needs the `{L,R}HS`
after `IgnoreParenImpCasts()`, and `CheckTautologicalComparison()` is not ok
with that.
It seems that at most, i could re-use the detection of `RhsConstant`.
So, new options:
1. Either simply restore the original check, and add a comment there about the
logic behind it
2. Or, drop the check completely
3. Or, move the `CheckTautologicalComparison()` call before
`DiagnoseOutOfRangeComparison()`
And if `DiagnoseOutOfRangeComparison()` has already emitted diagnostic,
return.
Much like what `CheckTautologicalComparison()` already does.
So i think `3.` is still the best option :)
(tried implementing it, appears to work)
Repository:
rL LLVM
https://reviews.llvm.org/D38101
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[clang-tools-extra] r315149 - Fix signed/unsigned warning
Author: rksimon
Date: Sat Oct 7 05:24:10 2017
New Revision: 315149
URL: http://llvm.org/viewvc/llvm-project?rev=315149&view=rev
Log:
Fix signed/unsigned warning
Modified:
clang-tools-extra/trunk/clangd/ClangdUnit.cpp
Modified: clang-tools-extra/trunk/clangd/ClangdUnit.cpp
URL:
http://llvm.org/viewvc/llvm-project/clang-tools-extra/trunk/clangd/ClangdUnit.cpp?rev=315149&r1=315148&r2=315149&view=diff
==
--- clang-tools-extra/trunk/clangd/ClangdUnit.cpp (original)
+++ clang-tools-extra/trunk/clangd/ClangdUnit.cpp Sat Oct 7 05:24:10 2017
@@ -616,7 +616,7 @@ public:
// Right now the overloaded candidates seem to be provided in a "best fit"
// order, so I'm not too worried about this.
SigHelp.activeSignature = 0;
-assert(CurrentArg <= std::numeric_limits::max() &&
+assert(CurrentArg <= (unsigned)std::numeric_limits::max() &&
"too many arguments");
SigHelp.activeParameter = static_cast(CurrentArg);
for (unsigned I = 0; I < NumCandidates; ++I) {
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
[libclc] r315147 - Delete empy directory
Author: jketema Date: Sat Oct 7 02:10:44 2017 New Revision: 315147 URL: http://llvm.org/viewvc/llvm-project?rev=315147&view=rev Log: Delete empy directory Removed: libclc/trunk/ptx/lib/integer/ ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Re: r314571 - [Analyzer] Synthesize function body for std::call_once
This revision might be the cause of
https://bugs.llvm.org/show_bug.cgi?id=34869. I'm still working on a reduced
test case.
On Sat, Sep 30, 2017 at 2:03 AM, George Karpenkov via cfe-commits <
cfe-commits@lists.llvm.org> wrote:
> Author: george.karpenkov
> Date: Fri Sep 29 17:03:22 2017
> New Revision: 314571
>
> URL: http://llvm.org/viewvc/llvm-project?rev=314571&view=rev
> Log:
> [Analyzer] Synthesize function body for std::call_once
>
> Differential Revision: https://reviews.llvm.org/D37840
>
> Added:
> cfe/trunk/test/Analysis/call_once.cpp
> Modified:
> cfe/trunk/lib/Analysis/BodyFarm.cpp
>
> Modified: cfe/trunk/lib/Analysis/BodyFarm.cpp
> URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/
> Analysis/BodyFarm.cpp?rev=314571&r1=314570&r2=314571&view=diff
>
> ==
> --- cfe/trunk/lib/Analysis/BodyFarm.cpp (original)
> +++ cfe/trunk/lib/Analysis/BodyFarm.cpp Fri Sep 29 17:03:22 2017
> @@ -14,11 +14,18 @@
>
> #include "BodyFarm.h"
> #include "clang/AST/ASTContext.h"
> +#include "clang/AST/CXXInheritance.h"
> #include "clang/AST/Decl.h"
> #include "clang/AST/Expr.h"
> +#include "clang/AST/ExprCXX.h"
> #include "clang/AST/ExprObjC.h"
> +#include "clang/AST/NestedNameSpecifier.h"
> #include "clang/Analysis/CodeInjector.h"
> +#include "clang/Basic/OperatorKinds.h"
> #include "llvm/ADT/StringSwitch.h"
> +#include "llvm/Support/Debug.h"
> +
> +#define DEBUG_TYPE "body-farm"
>
> using namespace clang;
>
> @@ -55,7 +62,9 @@ public:
>CompoundStmt *makeCompound(ArrayRef);
>
>/// Create a new DeclRefExpr for the referenced variable.
> - DeclRefExpr *makeDeclRefExpr(const VarDecl *D);
> + DeclRefExpr *makeDeclRefExpr(const VarDecl *D,
> + bool RefersToEnclosingVariableOrCapture =
> false,
> + bool GetNonReferenceType = false);
>
>/// Create a new UnaryOperator representing a dereference.
>UnaryOperator *makeDereference(const Expr *Arg, QualType Ty);
> @@ -66,9 +75,24 @@ public:
>/// Create an implicit cast to a builtin boolean type.
>ImplicitCastExpr *makeIntegralCastToBoolean(const Expr *Arg);
>
> - // Create an implicit cast for lvalue-to-rvaluate conversions.
> + /// Create an implicit cast for lvalue-to-rvaluate conversions.
>ImplicitCastExpr *makeLvalueToRvalue(const Expr *Arg, QualType Ty);
>
> + /// Create an implicit cast for lvalue-to-rvaluate conversions.
> + ImplicitCastExpr *makeLvalueToRvalue(const Expr *Arg,
> + bool GetNonReferenceType = false);
> +
> + /// Make RValue out of variable declaration, creating a temporary
> + /// DeclRefExpr in the process.
> + ImplicitCastExpr *
> + makeLvalueToRvalue(const VarDecl *Decl,
> + bool RefersToEnclosingVariableOrCapture = false,
> + bool GetNonReferenceType = false);
> +
> + /// Create an implicit cast of the given type.
> + ImplicitCastExpr *makeImplicitCast(const Expr *Arg, QualType Ty,
> + CastKind CK = CK_LValueToRValue);
> +
>/// Create an Objective-C bool literal.
>ObjCBoolLiteralExpr *makeObjCBool(bool Val);
>
> @@ -78,6 +102,18 @@ public:
>/// Create a Return statement.
>ReturnStmt *makeReturn(const Expr *RetVal);
>
> + /// Create an integer literal.
> + IntegerLiteral *makeIntegerLiteral(uint64_t value);
> +
> + /// Create a member expression.
> + MemberExpr *makeMemberExpression(Expr *base, ValueDecl *MemberDecl,
> + bool IsArrow = false,
> + ExprValueKind ValueKind = VK_LValue);
> +
> + /// Returns a *first* member field of a record declaration with a given
> name.
> + /// \return an nullptr if no member with such a name exists.
> + NamedDecl *findMemberField(const CXXRecordDecl *RD, StringRef Name);
> +
> private:
>ASTContext &C;
> };
> @@ -106,16 +142,16 @@ CompoundStmt *ASTMaker::makeCompound(Arr
>return new (C) CompoundStmt(C, Stmts, SourceLocation(),
> SourceLocation());
> }
>
> -DeclRefExpr *ASTMaker::makeDeclRefExpr(const VarDecl *D) {
> - DeclRefExpr *DR =
> -DeclRefExpr::Create(/* Ctx = */ C,
> -/* QualifierLoc = */ NestedNameSpecifierLoc(),
> -/* TemplateKWLoc = */ SourceLocation(),
> -/* D = */ const_cast(D),
> -/* RefersToEnclosingVariableOrCapture = */ false,
> -/* NameLoc = */ SourceLocation(),
> -/* T = */ D->getType(),
> -/* VK = */ VK_LValue);
> +DeclRefExpr *ASTMaker::makeDeclRefExpr(const VarDecl *D,
> + bool RefersToEnclosingVariableOrCap
> ture,
> + bool GetNonReferenceType) {
> + auto Type = D->getType();
> + if (GetNonReferenceType)
> +Type = Type.getNonReferenceType();
> +
