llvm-project/clang/test/CodeGen/inline.c

151 lines
5.3 KiB
C
Raw Normal View History

// REQUIRES: x86-registered-target
//
// RUN: echo "GNU89 tests:"
// RUN: %clang_cc1 %s -triple i386-unknown-unknown -O1 -disable-llvm-passes -emit-llvm -o - -std=gnu89 | FileCheck %s --check-prefix=CHECK1
// RUN: %clang_cc1 %s -triple i386-unknown-unknown -fexperimental-new-pass-manager -O1 -disable-llvm-passes -emit-llvm -o - -std=gnu89 | FileCheck %s --check-prefix=CHECK1
// CHECK1-LABEL: define i32 @foo()
// CHECK1-LABEL: define i32 @bar()
// CHECK1-LABEL: define void @unreferenced1()
// CHECK1-NOT: unreferenced2
// CHECK1-LABEL: define void @gnu_inline()
// CHECK1-LABEL: define i32 @test1
// CHECK1-LABEL: define i32 @test2
// CHECK1-LABEL: define void @test3()
// CHECK1-LABEL: define available_externally i32 @test4
// CHECK1-LABEL: define available_externally i32 @test5
// CHECK1-LABEL: define i32 @test6
// CHECK1-LABEL: define void @test7
// CHECK1: define i{{..}} @strlcpy
// CHECK1-NOT: test9
// CHECK1-LABEL: define void @testA
// CHECK1-LABEL: define void @testB
// CHECK1-LABEL: define void @testC
// CHECK1-LABEL: define available_externally i32 @ei()
// CHECK1-LABEL: define available_externally void @gnu_ei_inline()
// RUN: echo "C99 tests:"
// RUN: %clang_cc1 %s -triple i386-unknown-unknown -O1 -disable-llvm-passes -emit-llvm -o - -std=gnu99 | FileCheck %s --check-prefix=CHECK2
// RUN: %clang_cc1 %s -triple i386-unknown-unknown -fexperimental-new-pass-manager -O1 -disable-llvm-passes -emit-llvm -o - -std=gnu99 | FileCheck %s --check-prefix=CHECK2
// CHECK2-LABEL: define i32 @ei()
// CHECK2-LABEL: define i32 @bar()
// CHECK2-NOT: unreferenced1
// CHECK2-LABEL: define void @unreferenced2()
// CHECK2-LABEL: define void @gnu_inline()
// CHECK2-LABEL: define i32 @test1
// CHECK2-LABEL: define i32 @test2
// CHECK2-LABEL: define void @test3
// CHECK2-LABEL: define available_externally i32 @test4
// CHECK2-LABEL: define available_externally i32 @test5
// CHECK2-LABEL: define i32 @test6
// CHECK2-LABEL: define void @test7
// CHECK2: define available_externally i{{..}} @strlcpy
// CHECK2-LABEL: define void @test9
// CHECK2-LABEL: define void @testA
// CHECK2-LABEL: define void @testB
// CHECK2-LABEL: define void @testC
// CHECK2-LABEL: define available_externally i32 @foo()
// CHECK2-LABEL: define available_externally void @gnu_ei_inline()
// RUN: echo "C++ tests:"
// RUN: %clang_cc1 -x c++ %s -triple i386-unknown-unknown -O1 -disable-llvm-passes -emit-llvm -o - -std=c++98 | FileCheck %s --check-prefix=CHECK3
// RUN: %clang_cc1 -x c++ %s -triple i386-unknown-unknown -fexperimental-new-pass-manager -O1 -disable-llvm-passes -emit-llvm -o - -std=c++98 | FileCheck %s --check-prefix=CHECK3
// CHECK3-LABEL: define i32 @_Z3barv()
// CHECK3-LABEL: define linkonce_odr i32 @_Z3foov()
// CHECK3-NOT: unreferenced
// CHECK3-LABEL: define void @_Z10gnu_inlinev()
// CHECK3-LABEL: define available_externally void @_Z13gnu_ei_inlinev()
// CHECK3-NOT: @_Z5testCv
// CHECK3-LABEL: define linkonce_odr i32 @_Z2eiv()
// RUN: echo "MS C Mode tests:"
// RUN: %clang_cc1 %s -triple i386-pc-win32 -O1 -disable-llvm-passes -emit-llvm -o - -std=c99 | FileCheck %s --check-prefix=CHECK4
// RUN: %clang_cc1 %s -triple i386-pc-win32 -fexperimental-new-pass-manager -O1 -disable-llvm-passes -emit-llvm -o - -std=c99 | FileCheck %s --check-prefix=CHECK4
// CHECK4-NOT: define weak_odr void @_Exit(
// CHECK4-LABEL: define weak_odr dso_local i32 @ei()
// CHECK4-LABEL: define dso_local i32 @bar()
// CHECK4-NOT: unreferenced1
// CHECK4-LABEL: define weak_odr dso_local void @unreferenced2()
// CHECK4-LABEL: define dso_local void @gnu_inline()
// CHECK4-LABEL: define linkonce_odr dso_local i32 @foo()
// CHECK4-LABEL: define available_externally dso_local void @gnu_ei_inline()
__attribute__((noreturn)) void __cdecl _exit(int _Code);
__inline void __cdecl _Exit(int status) { _exit(status); }
Simplify the scheme used for keywords, and change the classification scheme to be more useful. The new scheme introduces a set of categories that should be more readable, and also reflects what we want to consider as an extension more accurately. Specifically, it makes the "what is a keyword" determination accurately reflect whether the keyword is a GNU or Microsoft extension. I also introduced separate flags for keyword aliases; this is useful because the classification of the aliases is mostly unrelated to the classification of the original keyword. This patch treats anything that's in the implementation namespace (prefixed with "__", or "_X" where "X" is any upper-case letter) as a keyword without marking it as an extension. This is consistent with the standards in that an implementation is allowed to define arbitrary extensions in the implementation namespace without violating the standard. This gets rid of all the nasty "extension used" warnings for stuff like __attribute__ in -pedantic mode. We still warn for extensions outside of the the implementation namespace, like typeof. If someone wants to implement -Wextensions or something like that, we could add additional information to the keyword table. This also removes processing for the unused "Boolean" language option; such an extension isn't supported on any other C implementation, so I don't see any point to adding it. The changes to test/CodeGen/inline.c are required because previously, we weren't actually disabling the "inline" keyword in -std=c89 mode. I'll remove Boolean and NoExtensions from LangOptions in a follow-up commit. llvm-svn: 70281
2009-04-28 11:13:54 +08:00
extern __inline int ei() { return 123; }
Simplify the scheme used for keywords, and change the classification scheme to be more useful. The new scheme introduces a set of categories that should be more readable, and also reflects what we want to consider as an extension more accurately. Specifically, it makes the "what is a keyword" determination accurately reflect whether the keyword is a GNU or Microsoft extension. I also introduced separate flags for keyword aliases; this is useful because the classification of the aliases is mostly unrelated to the classification of the original keyword. This patch treats anything that's in the implementation namespace (prefixed with "__", or "_X" where "X" is any upper-case letter) as a keyword without marking it as an extension. This is consistent with the standards in that an implementation is allowed to define arbitrary extensions in the implementation namespace without violating the standard. This gets rid of all the nasty "extension used" warnings for stuff like __attribute__ in -pedantic mode. We still warn for extensions outside of the the implementation namespace, like typeof. If someone wants to implement -Wextensions or something like that, we could add additional information to the keyword table. This also removes processing for the unused "Boolean" language option; such an extension isn't supported on any other C implementation, so I don't see any point to adding it. The changes to test/CodeGen/inline.c are required because previously, we weren't actually disabling the "inline" keyword in -std=c89 mode. I'll remove Boolean and NoExtensions from LangOptions in a follow-up commit. llvm-svn: 70281
2009-04-28 11:13:54 +08:00
__inline int foo() {
return ei();
}
int bar() { return foo(); }
Simplify the scheme used for keywords, and change the classification scheme to be more useful. The new scheme introduces a set of categories that should be more readable, and also reflects what we want to consider as an extension more accurately. Specifically, it makes the "what is a keyword" determination accurately reflect whether the keyword is a GNU or Microsoft extension. I also introduced separate flags for keyword aliases; this is useful because the classification of the aliases is mostly unrelated to the classification of the original keyword. This patch treats anything that's in the implementation namespace (prefixed with "__", or "_X" where "X" is any upper-case letter) as a keyword without marking it as an extension. This is consistent with the standards in that an implementation is allowed to define arbitrary extensions in the implementation namespace without violating the standard. This gets rid of all the nasty "extension used" warnings for stuff like __attribute__ in -pedantic mode. We still warn for extensions outside of the the implementation namespace, like typeof. If someone wants to implement -Wextensions or something like that, we could add additional information to the keyword table. This also removes processing for the unused "Boolean" language option; such an extension isn't supported on any other C implementation, so I don't see any point to adding it. The changes to test/CodeGen/inline.c are required because previously, we weren't actually disabling the "inline" keyword in -std=c89 mode. I'll remove Boolean and NoExtensions from LangOptions in a follow-up commit. llvm-svn: 70281
2009-04-28 11:13:54 +08:00
__inline void unreferenced1() {}
extern __inline void unreferenced2() {}
__inline __attribute((__gnu_inline__)) void gnu_inline() {}
// PR3988
Simplify the scheme used for keywords, and change the classification scheme to be more useful. The new scheme introduces a set of categories that should be more readable, and also reflects what we want to consider as an extension more accurately. Specifically, it makes the "what is a keyword" determination accurately reflect whether the keyword is a GNU or Microsoft extension. I also introduced separate flags for keyword aliases; this is useful because the classification of the aliases is mostly unrelated to the classification of the original keyword. This patch treats anything that's in the implementation namespace (prefixed with "__", or "_X" where "X" is any upper-case letter) as a keyword without marking it as an extension. This is consistent with the standards in that an implementation is allowed to define arbitrary extensions in the implementation namespace without violating the standard. This gets rid of all the nasty "extension used" warnings for stuff like __attribute__ in -pedantic mode. We still warn for extensions outside of the the implementation namespace, like typeof. If someone wants to implement -Wextensions or something like that, we could add additional information to the keyword table. This also removes processing for the unused "Boolean" language option; such an extension isn't supported on any other C implementation, so I don't see any point to adding it. The changes to test/CodeGen/inline.c are required because previously, we weren't actually disabling the "inline" keyword in -std=c89 mode. I'll remove Boolean and NoExtensions from LangOptions in a follow-up commit. llvm-svn: 70281
2009-04-28 11:13:54 +08:00
extern __inline __attribute__((gnu_inline)) void gnu_ei_inline() {}
void (*P)() = gnu_ei_inline;
// <rdar://problem/6818429>
int test1();
Simplify the scheme used for keywords, and change the classification scheme to be more useful. The new scheme introduces a set of categories that should be more readable, and also reflects what we want to consider as an extension more accurately. Specifically, it makes the "what is a keyword" determination accurately reflect whether the keyword is a GNU or Microsoft extension. I also introduced separate flags for keyword aliases; this is useful because the classification of the aliases is mostly unrelated to the classification of the original keyword. This patch treats anything that's in the implementation namespace (prefixed with "__", or "_X" where "X" is any upper-case letter) as a keyword without marking it as an extension. This is consistent with the standards in that an implementation is allowed to define arbitrary extensions in the implementation namespace without violating the standard. This gets rid of all the nasty "extension used" warnings for stuff like __attribute__ in -pedantic mode. We still warn for extensions outside of the the implementation namespace, like typeof. If someone wants to implement -Wextensions or something like that, we could add additional information to the keyword table. This also removes processing for the unused "Boolean" language option; such an extension isn't supported on any other C implementation, so I don't see any point to adding it. The changes to test/CodeGen/inline.c are required because previously, we weren't actually disabling the "inline" keyword in -std=c89 mode. I'll remove Boolean and NoExtensions from LangOptions in a follow-up commit. llvm-svn: 70281
2009-04-28 11:13:54 +08:00
__inline int test1() { return 4; }
__inline int test2() { return 5; }
__inline int test2();
int test2();
void test_test1() { test1(); }
void test_test2() { test2(); }
// PR3989
Simplify the scheme used for keywords, and change the classification scheme to be more useful. The new scheme introduces a set of categories that should be more readable, and also reflects what we want to consider as an extension more accurately. Specifically, it makes the "what is a keyword" determination accurately reflect whether the keyword is a GNU or Microsoft extension. I also introduced separate flags for keyword aliases; this is useful because the classification of the aliases is mostly unrelated to the classification of the original keyword. This patch treats anything that's in the implementation namespace (prefixed with "__", or "_X" where "X" is any upper-case letter) as a keyword without marking it as an extension. This is consistent with the standards in that an implementation is allowed to define arbitrary extensions in the implementation namespace without violating the standard. This gets rid of all the nasty "extension used" warnings for stuff like __attribute__ in -pedantic mode. We still warn for extensions outside of the the implementation namespace, like typeof. If someone wants to implement -Wextensions or something like that, we could add additional information to the keyword table. This also removes processing for the unused "Boolean" language option; such an extension isn't supported on any other C implementation, so I don't see any point to adding it. The changes to test/CodeGen/inline.c are required because previously, we weren't actually disabling the "inline" keyword in -std=c89 mode. I'll remove Boolean and NoExtensions from LangOptions in a follow-up commit. llvm-svn: 70281
2009-04-28 11:13:54 +08:00
extern __inline void test3() __attribute__((gnu_inline));
__inline void __attribute__((gnu_inline)) test3() {}
extern int test4(void);
extern __inline __attribute__ ((__gnu_inline__)) int test4(void)
{
2009-07-22 04:52:43 +08:00
return 0;
}
void test_test4() { test4(); }
extern __inline int test5(void) __attribute__ ((__gnu_inline__));
extern __inline int __attribute__ ((__gnu_inline__)) test5(void)
{
2009-07-22 04:52:43 +08:00
return 0;
}
void test_test5() { test5(); }
// PR10233
__inline int test6() { return 0; }
extern int test6();
2012-02-02 14:06:34 +08:00
// No PR#, but this once crashed clang in C99 mode due to buggy extern inline
// redeclaration detection.
void test7() { }
void test7();
// PR11062; the fact that the function is named strlcpy matters here.
inline __typeof(sizeof(int)) strlcpy(char *dest, const char *src, __typeof(sizeof(int)) size) { return 3; }
void test8() { strlcpy(0,0,0); }
// PR10657; the test crashed in C99 mode
extern inline void test9() { }
void test9();
inline void testA() {}
void testA();
void testB();
inline void testB() {}
extern void testB();
extern inline void testC() {}
inline void testC();