maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/clang/test/CodeGenCXX/microsoft-abi-sret-and-byva...

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

455 lines
23 KiB
C++
Raw Normal View History

[MS] Fix 'this' type when calling virtual methods with inalloca If the virtual method comes from a secondary vtable, then the type of the 'this' parameter should be i8*, and not a pointer to the complete class. In the MS ABI, the 'this' parameter on entry points to the vptr containing the virtual method that was called, so we use i8* instead of the normal type. We had a mismatch where the CGFunctionInfo of the call didn't match the CGFunctionInfo of the declaration, and this resulted in some assertions, but now both sides agree the type of 'this' is i8*. Fixes one issue raised in PR30293 llvm-svn: 280815
2016-09-07 23:15:51 +08:00
// RUN: %clang_cc1 -std=c++11 -emit-llvm %s -o - -triple=i386-pc-linux | FileCheck -check-prefix LINUX %s
// RUN: %clang_cc1 -std=c++11 -emit-llvm %s -o - -triple=i386-pc-win32 -mconstructor-aliases -fno-rtti | FileCheck -check-prefix WIN32 %s
// RUN: %clang_cc1 -std=c++11 -emit-llvm %s -o - -triple=thumb-pc-win32 -mconstructor-aliases -fno-rtti | FileCheck -check-prefix WOA %s
// RUN: %clang_cc1 -std=c++11 -emit-llvm %s -o - -triple=x86_64-pc-win32 -mconstructor-aliases -fno-rtti | FileCheck -check-prefix WIN64 %s
[ARM64] [Windows] Follow MS X86_64 C++ ABI when passing structs Summary: Microsoft's C++ object model for ARM64 is the same as that for X86_64. For example, small structs with non-trivial copy constructors or virtual function tables are passed indirectly. Currently, they are passed in registers when compiled with clang. Reviewers: rnk, mstorsjo, TomTan, haripul, javed.absar Reviewed By: rnk, mstorsjo Subscribers: kristof.beyls, chrib, llvm-commits, cfe-commits Differential Revision: https://reviews.llvm.org/D49770 llvm-svn: 338076
2018-07-27 06:18:28 +08:00
// RUN: %clang_cc1 -std=c++11 -emit-llvm %s -o - -triple=aarch64-windows-msvc -mconstructor-aliases -fno-rtti | FileCheck -check-prefix WOA64 %s
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
struct Empty {};
struct EmptyWithCtor {
EmptyWithCtor() {}
};
struct Small {
int x;
};
// This is a C++11 trivial and standard-layout struct but not a C++03 POD.
struct SmallCpp11NotCpp03Pod : Empty {
int x;
};
struct SmallWithCtor {
SmallWithCtor() {}
int x;
};
Expand aggregate arguments more often on 32-bit Windows Before this change, we would pass all non-HFA record arguments on Windows with byval. Byval often blocks optimizations and results in bad code generation. Windows now uses the existing workaround that other x86_32 platforms use. I also expanded the workaround to handle C++ records with constructors on Windows. On non-Windows platforms, we have to keep generating the same LLVM IR prototypes if we want our bitcode to be ABI compatible. Otherwise we will encounter mismatch issues like PR21573. Essentially fixes PR27522 in Clang instead of LLVM. Reviewers: hans Differential Revision: http://reviews.llvm.org/D19756 llvm-svn: 268261
2016-05-03 01:41:07 +08:00
struct Multibyte {
char a, b, c, d;
};
struct Packed {
short a;
int b;
short c;
};
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
struct SmallWithDtor {
SmallWithDtor();
~SmallWithDtor();
int x;
};
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
struct SmallWithVftable {
int x;
virtual void foo();
};
struct Medium {
int x, y;
};
struct MediumWithCopyCtor {
MediumWithCopyCtor();
MediumWithCopyCtor(const struct MediumWithCopyCtor &);
int x, y;
};
struct Big {
int a, b, c, d, e, f;
};
MS ABI x64: Pass small objects with dtors but no copy ctors directly Passing objects directly (in registers or memory) creates a second copy of the object in the callee. The callee always destroys its copy, but we also have to destroy any temporary created in the caller. In other words, copy elision of these kinds of objects is impossible. Objects larger than 8 bytes with non-trivial dtors and trivial copy ctors are still passed indirectly, and we can still elide copies of them. Fixes PR19640. llvm-svn: 207889
2014-05-03 08:33:28 +08:00
struct BigWithDtor {
BigWithDtor();
~BigWithDtor();
int a, b, c, d, e, f;
};
Use less byval on 32-bit Windows x86 for classes with bases This comes up in V8, which has a Handle template class that wraps a typed pointer, and is frequently passed by value. The pointer is stored in the base, HandleBase. This change allows us to pass the struct as a pointer instead of using byval. This avoids creating tons of temporary allocas that we copy from during call lowering. Eventually, it would be good to use FCAs here instead. llvm-svn: 291917
2017-01-14 01:18:19 +08:00
struct BaseNoByval : Small {
int bb;
};
[COFF, ARM64] Fix ABI implementation of struct returns Summary: Related llvm patch: D60348. Patch co-authored by Sanjin Sijaric. Reviewers: rnk, efriedma, TomTan, ssijaric, ostannard Reviewed By: efriedma Subscribers: dmajor, richard.townsend.arm, ostannard, javed.absar, kristof.beyls, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D60349 llvm-svn: 359932
2019-05-04 05:12:24 +08:00
struct SmallWithPrivate {
private:
int i;
};
Bring r325915 back. The tests that failed on a windows host have been fixed. Original message: Start setting dso_local for COFF. With this there are still some GVs where we don't set dso_local because setGVProperties is never called. I intend to fix that in followup commits. This is just the bare minimum to teach shouldAssumeDSOLocal what it should do for COFF. llvm-svn: 325940
2018-02-24 03:30:48 +08:00
// WIN32: declare dso_local void @"{{.*take_bools_and_chars.*}}"
[ms-cxxabi] Use inalloca on win32 when passing non-trivial C++ objects When a non-trivial parameter is present, clang now gathers up all the parameters that lack inreg and puts them into a packed struct. MSVC always aligns each parameter to 4 bytes and no more, so this is a pretty simple struct to lay out. On win64, non-trivial records are passed indirectly. Prior to this change, clang was incorrectly using byval on win64. I'm able to self-host a working clang with this change and additional LLVM patches. Reviewers: rsmith Differential Revision: http://llvm-reviews.chandlerc.com/D2636 llvm-svn: 200597
2014-02-01 08:04:45 +08:00
// WIN32: (<{ i8, [3 x i8], i8, [3 x i8], %struct.SmallWithDtor,
inalloca: Pad the struct *after* inserting each arg This ensures that the overall struct size will be a multiple of 4, as required by the ABI. llvm-svn: 205981
2014-04-11 03:09:43 +08:00
// WIN32: i8, [3 x i8], i8, [3 x i8], i32, i8, [3 x i8] }>* inalloca)
[ms-cxxabi] Use inalloca on win32 when passing non-trivial C++ objects When a non-trivial parameter is present, clang now gathers up all the parameters that lack inreg and puts them into a packed struct. MSVC always aligns each parameter to 4 bytes and no more, so this is a pretty simple struct to lay out. On win64, non-trivial records are passed indirectly. Prior to this change, clang was incorrectly using byval on win64. I'm able to self-host a working clang with this change and additional LLVM patches. Reviewers: rsmith Differential Revision: http://llvm-reviews.chandlerc.com/D2636 llvm-svn: 200597
2014-02-01 08:04:45 +08:00
void take_bools_and_chars(char a, char b, SmallWithDtor c, char d, bool e, int f, bool g);
void call_bools_and_chars() {
take_bools_and_chars('A', 'B', SmallWithDtor(), 'D', true, 13, false);
}
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
// Returning structs that fit into a register.
Small small_return() { return Small(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX-LABEL: define void @_Z12small_returnv(%struct.Small* noalias sret align 4 %agg.result)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"()
// WIN64: define dso_local i32 @"?small_return@@YA?AUSmall@@XZ"()
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
Medium medium_return() { return Medium(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX-LABEL: define void @_Z13medium_returnv(%struct.Medium* noalias sret align 4 %agg.result)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"()
// WIN64: define dso_local i64 @"?medium_return@@YA?AUMedium@@XZ"()
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
// Returning structs that fit into a register but are not POD.
SmallCpp11NotCpp03Pod small_non_pod_return() { return SmallCpp11NotCpp03Pod(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX-LABEL: define void @_Z20small_non_pod_returnv(%struct.SmallCpp11NotCpp03Pod* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(%struct.SmallCpp11NotCpp03Pod* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?small_non_pod_return@@YA?AUSmallCpp11NotCpp03Pod@@XZ"(%struct.SmallCpp11NotCpp03Pod* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
SmallWithCtor small_with_ctor_return() { return SmallWithCtor(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX-LABEL: define void @_Z22small_with_ctor_returnv(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
[MS] Pass non-trivially-copyable objects indirectly on Windows ARM This isn't exactly what MSVC does, unfortunately. MSVC does not pass objects with destructors but no copy constructors by address. More ARM expertise is required to really understand what should be done here. Fixes PR29136. llvm-svn: 279764
2016-08-26 02:23:28 +08:00
// FIXME: The 'sret' mark here doesn't seem to be enough to convince LLVM to
// preserve the hidden sret pointer in R0 across the function.
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// WOA: define dso_local arm_aapcs_vfpcc void @"?small_with_ctor_return@@YA?AUSmallWithCtor@@XZ"(%struct.SmallWithCtor* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
SmallWithVftable small_with_vftable_return() { return SmallWithVftable(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX-LABEL: define void @_Z25small_with_vftable_returnv(%struct.SmallWithVftable* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(%struct.SmallWithVftable* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?small_with_vftable_return@@YA?AUSmallWithVftable@@XZ"(%struct.SmallWithVftable* noalias sret align 8 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
MediumWithCopyCtor medium_with_copy_ctor_return() { return MediumWithCopyCtor(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX-LABEL: define void @_Z28medium_with_copy_ctor_returnv(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
// WOA: define dso_local arm_aapcs_vfpcc void @"?medium_with_copy_ctor_return@@YA?AUMediumWithCopyCtor@@XZ"(%struct.MediumWithCopyCtor* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
// Returning a large struct that doesn't fit into a register.
Big big_return() { return Big(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX-LABEL: define void @_Z10big_returnv(%struct.Big* noalias sret align 4 %agg.result)
// WIN32: define dso_local void @"?big_return@@YA?AUBig@@XZ"(%struct.Big* noalias sret align 4 %agg.result)
// WIN64: define dso_local void @"?big_return@@YA?AUBig@@XZ"(%struct.Big* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void small_arg(Small s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z9small_arg5Small(i32 %s.0)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?small_arg@@YAXUSmall@@@Z"(i32 %s.0)
// WIN64: define dso_local void @"?small_arg@@YAXUSmall@@@Z"(i32 %s.coerce)
// WOA: define dso_local arm_aapcs_vfpcc void @"?small_arg@@YAXUSmall@@@Z"([1 x i32] %s.coerce)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void medium_arg(Medium s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z10medium_arg6Medium(i32 %s.0, i32 %s.1)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?medium_arg@@YAXUMedium@@@Z"(i32 %s.0, i32 %s.1)
// WIN64: define dso_local void @"?medium_arg@@YAXUMedium@@@Z"(i64 %s.coerce)
// WOA: define dso_local arm_aapcs_vfpcc void @"?medium_arg@@YAXUMedium@@@Z"([2 x i32] %s.coerce)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
Use less byval on 32-bit Windows x86 for classes with bases This comes up in V8, which has a Handle template class that wraps a typed pointer, and is frequently passed by value. The pointer is stored in the base, HandleBase. This change allows us to pass the struct as a pointer instead of using byval. This avoids creating tons of temporary allocas that we copy from during call lowering. Eventually, it would be good to use FCAs here instead. llvm-svn: 291917
2017-01-14 01:18:19 +08:00
void base_no_byval_arg(BaseNoByval s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z17base_no_byval_arg11BaseNoByval(%struct.BaseNoByval* byval(%struct.BaseNoByval) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?base_no_byval_arg@@YAXUBaseNoByval@@@Z"(i32 %s.0, i32 %s.1)
// WIN64: define dso_local void @"?base_no_byval_arg@@YAXUBaseNoByval@@@Z"(i64 %s.coerce)
// WOA: define dso_local arm_aapcs_vfpcc void @"?base_no_byval_arg@@YAXUBaseNoByval@@@Z"([2 x i32] %s.coerce)
Use less byval on 32-bit Windows x86 for classes with bases This comes up in V8, which has a Handle template class that wraps a typed pointer, and is frequently passed by value. The pointer is stored in the base, HandleBase. This change allows us to pass the struct as a pointer instead of using byval. This avoids creating tons of temporary allocas that we copy from during call lowering. Eventually, it would be good to use FCAs here instead. llvm-svn: 291917
2017-01-14 01:18:19 +08:00
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void small_arg_with_ctor(SmallWithCtor s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z19small_arg_with_ctor13SmallWithCtor(%struct.SmallWithCtor* byval(%struct.SmallWithCtor) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?small_arg_with_ctor@@YAXUSmallWithCtor@@@Z"(i32 %s.0)
// WIN64: define dso_local void @"?small_arg_with_ctor@@YAXUSmallWithCtor@@@Z"(i32 %s.coerce)
// WOA: define dso_local arm_aapcs_vfpcc void @"?small_arg_with_ctor@@YAXUSmallWithCtor@@@Z"([1 x i32] %s.coerce)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
Expand aggregate arguments more often on 32-bit Windows Before this change, we would pass all non-HFA record arguments on Windows with byval. Byval often blocks optimizations and results in bad code generation. Windows now uses the existing workaround that other x86_32 platforms use. I also expanded the workaround to handle C++ records with constructors on Windows. On non-Windows platforms, we have to keep generating the same LLVM IR prototypes if we want our bitcode to be ABI compatible. Otherwise we will encounter mismatch issues like PR21573. Essentially fixes PR27522 in Clang instead of LLVM. Reviewers: hans Differential Revision: http://reviews.llvm.org/D19756 llvm-svn: 268261
2016-05-03 01:41:07 +08:00
// FIXME: We could coerce to a series of i32s here if we wanted to.
void multibyte_arg(Multibyte s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z13multibyte_arg9Multibyte(%struct.Multibyte* byval(%struct.Multibyte) align 4 %s)
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// WIN32: define dso_local void @"?multibyte_arg@@YAXUMultibyte@@@Z"(%struct.Multibyte* byval(%struct.Multibyte) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define dso_local void @"?multibyte_arg@@YAXUMultibyte@@@Z"(i32 %s.coerce)
// WOA: define dso_local arm_aapcs_vfpcc void @"?multibyte_arg@@YAXUMultibyte@@@Z"([1 x i32] %s.coerce)
Expand aggregate arguments more often on 32-bit Windows Before this change, we would pass all non-HFA record arguments on Windows with byval. Byval often blocks optimizations and results in bad code generation. Windows now uses the existing workaround that other x86_32 platforms use. I also expanded the workaround to handle C++ records with constructors on Windows. On non-Windows platforms, we have to keep generating the same LLVM IR prototypes if we want our bitcode to be ABI compatible. Otherwise we will encounter mismatch issues like PR21573. Essentially fixes PR27522 in Clang instead of LLVM. Reviewers: hans Differential Revision: http://reviews.llvm.org/D19756 llvm-svn: 268261
2016-05-03 01:41:07 +08:00
void packed_arg(Packed s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z10packed_arg6Packed(%struct.Packed* byval(%struct.Packed) align 4 %s)
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// WIN32: define dso_local void @"?packed_arg@@YAXUPacked@@@Z"(%struct.Packed* byval(%struct.Packed) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define dso_local void @"?packed_arg@@YAXUPacked@@@Z"(%struct.Packed* %s)
Expand aggregate arguments more often on 32-bit Windows Before this change, we would pass all non-HFA record arguments on Windows with byval. Byval often blocks optimizations and results in bad code generation. Windows now uses the existing workaround that other x86_32 platforms use. I also expanded the workaround to handle C++ records with constructors on Windows. On non-Windows platforms, we have to keep generating the same LLVM IR prototypes if we want our bitcode to be ABI compatible. Otherwise we will encounter mismatch issues like PR21573. Essentially fixes PR27522 in Clang instead of LLVM. Reviewers: hans Differential Revision: http://reviews.llvm.org/D19756 llvm-svn: 268261
2016-05-03 01:41:07 +08:00
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// Test that dtors are invoked in the callee.
void small_arg_with_dtor(SmallWithDtor s) {}
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
// WIN32: define dso_local void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(<{ %struct.SmallWithDtor }>* inalloca %0) {{.*}} {
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: call x86_thiscallcc void @"??1SmallWithDtor@@QAE@XZ"
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// WIN32: }
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define dso_local void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(i32 %s.coerce) {{.*}} {
// WIN64: call void @"??1SmallWithDtor@@QEAA@XZ"
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// WIN64: }
[ARM64] [Windows] Follow MS X86_64 C++ ABI when passing structs Summary: Microsoft's C++ object model for ARM64 is the same as that for X86_64. For example, small structs with non-trivial copy constructors or virtual function tables are passed indirectly. Currently, they are passed in registers when compiled with clang. Reviewers: rnk, mstorsjo, TomTan, haripul, javed.absar Reviewed By: rnk, mstorsjo Subscribers: kristof.beyls, chrib, llvm-commits, cfe-commits Differential Revision: https://reviews.llvm.org/D49770 llvm-svn: 338076
2018-07-27 06:18:28 +08:00
// WOA64: define dso_local void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(i64 %s.coerce) {{.*}} {
[COFF, ARM64] Fix ABI implementation of struct returns Summary: Related llvm patch: D60348. Patch co-authored by Sanjin Sijaric. Reviewers: rnk, efriedma, TomTan, ssijaric, ostannard Reviewed By: efriedma Subscribers: dmajor, richard.townsend.arm, ostannard, javed.absar, kristof.beyls, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D60349 llvm-svn: 359932
2019-05-04 05:12:24 +08:00
// WOA64: call void @"??1SmallWithDtor@@QEAA@XZ"(%struct.SmallWithDtor* %s)
[ARM64] [Windows] Follow MS X86_64 C++ ABI when passing structs Summary: Microsoft's C++ object model for ARM64 is the same as that for X86_64. For example, small structs with non-trivial copy constructors or virtual function tables are passed indirectly. Currently, they are passed in registers when compiled with clang. Reviewers: rnk, mstorsjo, TomTan, haripul, javed.absar Reviewed By: rnk, mstorsjo Subscribers: kristof.beyls, chrib, llvm-commits, cfe-commits Differential Revision: https://reviews.llvm.org/D49770 llvm-svn: 338076
2018-07-27 06:18:28 +08:00
// WOA64: }
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
[MS] Pass non-trivially-copyable objects indirectly on Windows ARM This isn't exactly what MSVC does, unfortunately. MSVC does not pass objects with destructors but no copy constructors by address. More ARM expertise is required to really understand what should be done here. Fixes PR29136. llvm-svn: 279764
2016-08-26 02:23:28 +08:00
// FIXME: MSVC incompatible!
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WOA: define dso_local arm_aapcs_vfpcc void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(%struct.SmallWithDtor* %s) {{.*}} {
// WOA: call arm_aapcs_vfpcc void @"??1SmallWithDtor@@QAA@XZ"(%struct.SmallWithDtor* %s)
[MS] Pass non-trivially-copyable objects indirectly on Windows ARM This isn't exactly what MSVC does, unfortunately. MSVC does not pass objects with destructors but no copy constructors by address. More ARM expertise is required to really understand what should be done here. Fixes PR29136. llvm-svn: 279764
2016-08-26 02:23:28 +08:00
// WOA: }
[COFF, ARM64] Fix ABI implementation of struct returns Summary: Related llvm patch: D60348. Patch co-authored by Sanjin Sijaric. Reviewers: rnk, efriedma, TomTan, ssijaric, ostannard Reviewed By: efriedma Subscribers: dmajor, richard.townsend.arm, ostannard, javed.absar, kristof.beyls, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D60349 llvm-svn: 359932
2019-05-04 05:12:24 +08:00
// Test that the eligible non-aggregate is passed directly, but returned
// indirectly on ARM64 Windows.
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// WOA64: define dso_local void @"?small_arg_with_private_member@@YA?AUSmallWithPrivate@@U1@@Z"(%struct.SmallWithPrivate* inreg noalias sret align 4 %agg.result, i64 %s.coerce) {{.*}} {
[COFF, ARM64] Fix ABI implementation of struct returns Summary: Related llvm patch: D60348. Patch co-authored by Sanjin Sijaric. Reviewers: rnk, efriedma, TomTan, ssijaric, ostannard Reviewed By: efriedma Subscribers: dmajor, richard.townsend.arm, ostannard, javed.absar, kristof.beyls, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D60349 llvm-svn: 359932
2019-05-04 05:12:24 +08:00
SmallWithPrivate small_arg_with_private_member(SmallWithPrivate s) { return s; }
MS ABI x64: Pass small objects with dtors but no copy ctors directly Passing objects directly (in registers or memory) creates a second copy of the object in the callee. The callee always destroys its copy, but we also have to destroy any temporary created in the caller. In other words, copy elision of these kinds of objects is impossible. Objects larger than 8 bytes with non-trivial dtors and trivial copy ctors are still passed indirectly, and we can still elide copies of them. Fixes PR19640. llvm-svn: 207889
2014-05-03 08:33:28 +08:00
void call_small_arg_with_dtor() {
small_arg_with_dtor(SmallWithDtor());
}
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64-LABEL: define dso_local void @"?call_small_arg_with_dtor@@YAXXZ"()
// WIN64: call %struct.SmallWithDtor* @"??0SmallWithDtor@@QEAA@XZ"
// WIN64: call void @"?small_arg_with_dtor@@YAXUSmallWithDtor@@@Z"(i32 %{{.*}})
MS ABI x64: Pass small objects with dtors but no copy ctors directly Passing objects directly (in registers or memory) creates a second copy of the object in the callee. The callee always destroys its copy, but we also have to destroy any temporary created in the caller. In other words, copy elision of these kinds of objects is impossible. Objects larger than 8 bytes with non-trivial dtors and trivial copy ctors are still passed indirectly, and we can still elide copies of them. Fixes PR19640. llvm-svn: 207889
2014-05-03 08:33:28 +08:00
// WIN64: ret void
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// Test that references aren't destroyed in the callee.
void ref_small_arg_with_dtor(const SmallWithDtor &s) { }
[CGCall] Annotate references with "align" attribute. If we're going to assume references are dereferenceable, we should also assume they're aligned: otherwise, we can't actually dereference them. See also D80072. Differential Revision: https://reviews.llvm.org/D80166
2020-05-19 02:29:11 +08:00
// WIN32: define dso_local void @"?ref_small_arg_with_dtor@@YAXABUSmallWithDtor@@@Z"(%struct.SmallWithDtor* nonnull align {{[0-9]+}} dereferenceable({{[0-9]+}}) %s) {{.*}} {
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32-NOT: call x86_thiscallcc void @"??1SmallWithDtor@@QAE@XZ"
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// WIN32: }
[CGCall] Annotate references with "align" attribute. If we're going to assume references are dereferenceable, we should also assume they're aligned: otherwise, we can't actually dereference them. See also D80072. Differential Revision: https://reviews.llvm.org/D80166
2020-05-19 02:29:11 +08:00
// WIN64-LABEL: define dso_local void @"?ref_small_arg_with_dtor@@YAXAEBUSmallWithDtor@@@Z"(%struct.SmallWithDtor* nonnull align {{[0-9]+}} dereferenceable({{[0-9]+}}) %s)
MS ABI x64: Pass small objects with dtors but no copy ctors directly Passing objects directly (in registers or memory) creates a second copy of the object in the callee. The callee always destroys its copy, but we also have to destroy any temporary created in the caller. In other words, copy elision of these kinds of objects is impossible. Objects larger than 8 bytes with non-trivial dtors and trivial copy ctors are still passed indirectly, and we can still elide copies of them. Fixes PR19640. llvm-svn: 207889
2014-05-03 08:33:28 +08:00
void big_arg_with_dtor(BigWithDtor s) {}
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64-LABEL: define dso_local void @"?big_arg_with_dtor@@YAXUBigWithDtor@@@Z"(%struct.BigWithDtor* %s)
// WIN64: call void @"??1BigWithDtor@@QEAA@XZ"
MS ABI x64: Pass small objects with dtors but no copy ctors directly Passing objects directly (in registers or memory) creates a second copy of the object in the callee. The callee always destroys its copy, but we also have to destroy any temporary created in the caller. In other words, copy elision of these kinds of objects is impossible. Objects larger than 8 bytes with non-trivial dtors and trivial copy ctors are still passed indirectly, and we can still elide copies of them. Fixes PR19640. llvm-svn: 207889
2014-05-03 08:33:28 +08:00
// WIN64: }
void call_big_arg_with_dtor() {
big_arg_with_dtor(BigWithDtor());
}
// We can elide the copy of the temporary in the caller, because this object is
// larger than 8 bytes and is passed indirectly.
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64-LABEL: define dso_local void @"?call_big_arg_with_dtor@@YAXXZ"()
// WIN64: call %struct.BigWithDtor* @"??0BigWithDtor@@QEAA@XZ"
// WIN64: call void @"?big_arg_with_dtor@@YAXUBigWithDtor@@@Z"(%struct.BigWithDtor* %{{.*}})
// WIN64-NOT: call void @"??1BigWithDtor@@QEAA@XZ"
MS ABI x64: Pass small objects with dtors but no copy ctors directly Passing objects directly (in registers or memory) creates a second copy of the object in the callee. The callee always destroys its copy, but we also have to destroy any temporary created in the caller. In other words, copy elision of these kinds of objects is impossible. Objects larger than 8 bytes with non-trivial dtors and trivial copy ctors are still passed indirectly, and we can still elide copies of them. Fixes PR19640. llvm-svn: 207889
2014-05-03 08:33:28 +08:00
// WIN64: ret void
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// Test that temporaries passed by reference are destroyed in the caller.
void temporary_ref_with_dtor() {
ref_small_arg_with_dtor(SmallWithDtor());
}
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?temporary_ref_with_dtor@@YAXXZ"() {{.*}} {
// WIN32: call x86_thiscallcc %struct.SmallWithDtor* @"??0SmallWithDtor@@QAE@XZ"
// WIN32: call void @"?ref_small_arg_with_dtor@@YAXABUSmallWithDtor@@@Z"
// WIN32: call x86_thiscallcc void @"??1SmallWithDtor@@QAE@XZ"
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// WIN32: }
void takes_two_by_val_with_dtor(SmallWithDtor a, SmallWithDtor b);
void eh_cleanup_arg_with_dtor() {
takes_two_by_val_with_dtor(SmallWithDtor(), SmallWithDtor());
}
// When exceptions are off, we don't have any cleanups. See
// microsoft-abi-exceptions.cpp for these cleanups.
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?eh_cleanup_arg_with_dtor@@YAXXZ"() {{.*}} {
// WIN32: call x86_thiscallcc %struct.SmallWithDtor* @"??0SmallWithDtor@@QAE@XZ"
// WIN32: call x86_thiscallcc %struct.SmallWithDtor* @"??0SmallWithDtor@@QAE@XZ"
// WIN32: call void @"?takes_two_by_val_with_dtor@@YAXUSmallWithDtor@@0@Z"
// WIN32-NOT: call x86_thiscallcc void @"??1SmallWithDtor@@QAE@XZ"
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// WIN32: }
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void small_arg_with_vftable(SmallWithVftable s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z22small_arg_with_vftable16SmallWithVftable(%struct.SmallWithVftable* %s)
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
// WIN32: define dso_local void @"?small_arg_with_vftable@@YAXUSmallWithVftable@@@Z"(<{ %struct.SmallWithVftable }>* inalloca %0)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define dso_local void @"?small_arg_with_vftable@@YAXUSmallWithVftable@@@Z"(%struct.SmallWithVftable* %s)
[ARM64] [Windows] Follow MS X86_64 C++ ABI when passing structs Summary: Microsoft's C++ object model for ARM64 is the same as that for X86_64. For example, small structs with non-trivial copy constructors or virtual function tables are passed indirectly. Currently, they are passed in registers when compiled with clang. Reviewers: rnk, mstorsjo, TomTan, haripul, javed.absar Reviewed By: rnk, mstorsjo Subscribers: kristof.beyls, chrib, llvm-commits, cfe-commits Differential Revision: https://reviews.llvm.org/D49770 llvm-svn: 338076
2018-07-27 06:18:28 +08:00
// WOA64: define dso_local void @"?small_arg_with_vftable@@YAXUSmallWithVftable@@@Z"(%struct.SmallWithVftable* %s)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void medium_arg_with_copy_ctor(MediumWithCopyCtor s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z25medium_arg_with_copy_ctor18MediumWithCopyCtor(%struct.MediumWithCopyCtor* %s)
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
// WIN32: define dso_local void @"?medium_arg_with_copy_ctor@@YAXUMediumWithCopyCtor@@@Z"(<{ %struct.MediumWithCopyCtor }>* inalloca %0)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define dso_local void @"?medium_arg_with_copy_ctor@@YAXUMediumWithCopyCtor@@@Z"(%struct.MediumWithCopyCtor* %s)
// WOA: define dso_local arm_aapcs_vfpcc void @"?medium_arg_with_copy_ctor@@YAXUMediumWithCopyCtor@@@Z"(%struct.MediumWithCopyCtor* %s)
[ARM64] [Windows] Follow MS X86_64 C++ ABI when passing structs Summary: Microsoft's C++ object model for ARM64 is the same as that for X86_64. For example, small structs with non-trivial copy constructors or virtual function tables are passed indirectly. Currently, they are passed in registers when compiled with clang. Reviewers: rnk, mstorsjo, TomTan, haripul, javed.absar Reviewed By: rnk, mstorsjo Subscribers: kristof.beyls, chrib, llvm-commits, cfe-commits Differential Revision: https://reviews.llvm.org/D49770 llvm-svn: 338076
2018-07-27 06:18:28 +08:00
// WOA64: define dso_local void @"?medium_arg_with_copy_ctor@@YAXUMediumWithCopyCtor@@@Z"(%struct.MediumWithCopyCtor* %s)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void big_arg(Big s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z7big_arg3Big(%struct.Big* byval(%struct.Big) align 4 %s)
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// WIN32: define dso_local void @"?big_arg@@YAXUBig@@@Z"(%struct.Big* byval(%struct.Big) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define dso_local void @"?big_arg@@YAXUBig@@@Z"(%struct.Big* %s)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
Fix argument expansion of reference fields of structs r268261 made Clang "expand" more struct arguments on Windows. It removed the check for 'RD->isCLike()', which was preventing us from attempting to expand structs with reference type fields. Our expansion code was attempting to load and pass each field of the type in turn. We were accidentally doing one to many loads on reference type fields. On the function prologue side, we can use EmitLValueForFieldInitialization, which obviously gets the address of the field. On the call side, I tweaked EmitRValueForField directly, since this is the only use of this method. Fixes PR27607 llvm-svn: 268321
2016-05-03 06:42:34 +08:00
// PR27607: We would attempt to load i32 value out of the reference instead of
// just loading the pointer from the struct during argument expansion.
struct RefField {
RefField(int &x);
int &x;
};
void takes_ref_field(RefField s) {}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z15takes_ref_field8RefField(%struct.RefField* byval(%struct.RefField) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?takes_ref_field@@YAXURefField@@@Z"(i32* %s.0)
// WIN64: define dso_local void @"?takes_ref_field@@YAXURefField@@@Z"(i64 %s.coerce)
Fix argument expansion of reference fields of structs r268261 made Clang "expand" more struct arguments on Windows. It removed the check for 'RD->isCLike()', which was preventing us from attempting to expand structs with reference type fields. Our expansion code was attempting to load and pass each field of the type in turn. We were accidentally doing one to many loads on reference type fields. On the function prologue side, we can use EmitLValueForFieldInitialization, which obviously gets the address of the field. On the call side, I tweaked EmitRValueForField directly, since this is the only use of this method. Fixes PR27607 llvm-svn: 268321
2016-05-03 06:42:34 +08:00
void pass_ref_field() {
int x;
takes_ref_field(RefField(x));
}
Revert "[CodeGenModule] Assume dso_local for -fpic -fno-semantic-interposition" This reverts commit 789a46f2d742e11edaade28cb59a0f4d2a1d770e. Accidentally committed.
2020-02-04 02:09:39 +08:00
// LINUX-LABEL: define void @_Z14pass_ref_fieldv()
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// LINUX: call void @_Z15takes_ref_field8RefField(%struct.RefField* byval(%struct.RefField) align 4 %{{.*}})
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32-LABEL: define dso_local void @"?pass_ref_field@@YAXXZ"()
// WIN32: call void @"?takes_ref_field@@YAXURefField@@@Z"(i32* %{{.*}})
// WIN64-LABEL: define dso_local void @"?pass_ref_field@@YAXXZ"()
// WIN64: call void @"?takes_ref_field@@YAXURefField@@@Z"(i64 %{{.*}})
Fix argument expansion of reference fields of structs r268261 made Clang "expand" more struct arguments on Windows. It removed the check for 'RD->isCLike()', which was preventing us from attempting to expand structs with reference type fields. Our expansion code was attempting to load and pass each field of the type in turn. We were accidentally doing one to many loads on reference type fields. On the function prologue side, we can use EmitLValueForFieldInitialization, which obviously gets the address of the field. On the call side, I tweaked EmitRValueForField directly, since this is the only use of this method. Fixes PR27607 llvm-svn: 268321
2016-05-03 06:42:34 +08:00
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
class Class {
public:
Small thiscall_method_small() { return Small(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX: define {{.*}} void @_ZN5Class21thiscall_method_smallEv(%struct.Small* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small@Class@@QAE?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_small@Class@@QEAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
SmallWithCtor thiscall_method_small_with_ctor() { return SmallWithCtor(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX: define {{.*}} void @_ZN5Class31thiscall_method_small_with_ctorEv(%struct.SmallWithCtor* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} x86_thiscallcc void @"?thiscall_method_small_with_ctor@Class@@QAE?AUSmallWithCtor@@XZ"(%class.Class* %this, %struct.SmallWithCtor* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_small_with_ctor@Class@@QEAA?AUSmallWithCtor@@XZ"(%class.Class* %this, %struct.SmallWithCtor* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
Small __cdecl cdecl_method_small() { return Small(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX: define {{.*}} void @_ZN5Class18cdecl_method_smallEv(%struct.Small* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} void @"?cdecl_method_small@Class@@QAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?cdecl_method_small@Class@@QEAA?AUSmall@@XZ"(%class.Class* %this, %struct.Small* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
Big __cdecl cdecl_method_big() { return Big(); }
Revert "OpaquePtr: Add type to sret attribute" This reverts commit 55c4ff91bd820d72014f63dcf7f3d5a0d3397986. Issues were introduced as discussed in https://reviews.llvm.org/D88241 where this change made previous bugs in the linker and BitCodeWriter visible.
2020-09-29 16:24:54 +08:00
// LINUX: define {{.*}} void @_ZN5Class16cdecl_method_bigEv(%struct.Big* noalias sret align 4 %agg.result, %class.Class* %this)
// WIN32: define {{.*}} void @"?cdecl_method_big@Class@@QAA?AUBig@@XZ"(%class.Class* %this, %struct.Big* noalias sret align 4 %agg.result)
// WIN64: define linkonce_odr dso_local void @"?cdecl_method_big@Class@@QEAA?AUBig@@XZ"(%class.Class* %this, %struct.Big* noalias sret align 4 %agg.result)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void thiscall_method_arg(Empty s) {}
// LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE5Empty(%class.Class* %this)
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// WIN32: define {{.*}} void @"?thiscall_method_arg@Class@@QAEXUEmpty@@@Z"(%class.Class* %this, %struct.Empty* byval(%struct.Empty) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_arg@Class@@QEAAXUEmpty@@@Z"(%class.Class* %this, i8 %s.coerce)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void thiscall_method_arg(EmptyWithCtor s) {}
// LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE13EmptyWithCtor(%class.Class* %this)
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// WIN32: define {{.*}} void @"?thiscall_method_arg@Class@@QAEXUEmptyWithCtor@@@Z"(%class.Class* %this, %struct.EmptyWithCtor* byval(%struct.EmptyWithCtor) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_arg@Class@@QEAAXUEmptyWithCtor@@@Z"(%class.Class* %this, i8 %s.coerce)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void thiscall_method_arg(Small s) {}
[i386 ABI] expand small C like structs in C++, just like how we handle small C structs. This comes up when we have a function that takes a struct and is defined in a C++ file and used in a C file. Before this commit, we will generate byval for C++ and will expand the struct for C, thus causing difference at IR level. We will use bitcast of function type at the callsite, which causes the inliner to not inline the function. This commit changes how we handle small C like structs at IR level, but at backend, we should generate the same argument passing before and after the commit. Note that the condition for expanding is still over conservative. We should be able to expand type that is spelled with “class” and types that are not C-like. But this commit fixes the inconsistent argument passing between C/C++. Reviewed by John. rdar://20121030 llvm-svn: 234033
2015-04-04 02:10:29 +08:00
// LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE5Small(%class.Class* %this, i32 %s.0)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define {{.*}} void @"?thiscall_method_arg@Class@@QAEXUSmall@@@Z"(%class.Class* %this, i32 %s.0)
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_arg@Class@@QEAAXUSmall@@@Z"(%class.Class* %this, i32 %s.coerce)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void thiscall_method_arg(SmallWithCtor s) {}
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE13SmallWithCtor(%class.Class* %this, %struct.SmallWithCtor* byval(%struct.SmallWithCtor) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define {{.*}} void @"?thiscall_method_arg@Class@@QAEXUSmallWithCtor@@@Z"(%class.Class* %this, i32 %s.0)
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_arg@Class@@QEAAXUSmallWithCtor@@@Z"(%class.Class* %this, i32 %s.coerce)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
void thiscall_method_arg(Big s) {}
LLVM IR: Generate new-style byval-with-Type from Clang LLVM IR recently added a Type parameter to the byval Attribute, so that when pointers become opaque and no longer have an element type the information will still be present in IR. For now the Type parameter is optional (which is why Clang didn't need this change at the time), but it will become mandatory soon. llvm-svn: 362652
2019-06-06 05:12:14 +08:00
// LINUX: define {{.*}} void @_ZN5Class19thiscall_method_argE3Big(%class.Class* %this, %struct.Big* byval(%struct.Big) align 4 %s)
// WIN32: define {{.*}} void @"?thiscall_method_arg@Class@@QAEXUBig@@@Z"(%class.Class* %this, %struct.Big* byval(%struct.Big) align 4 %s)
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64: define linkonce_odr dso_local void @"?thiscall_method_arg@Class@@QEAAXUBig@@@Z"(%class.Class* %this, %struct.Big* %s)
Set SRet flags properly in '-cxx-abi microsoft'. Also, - abstract out the indirect/in memory/in registers decisions into the CGCXXABI - fix handling of empty struct arguments for '-cxx-abi microsoft' - add/fix tests llvm-svn: 179681
2013-04-17 20:54:10 +08:00
};
void use_class() {
Class c;
c.thiscall_method_small();
c.thiscall_method_small_with_ctor();
c.cdecl_method_small();
c.cdecl_method_big();
c.thiscall_method_arg(Empty());
c.thiscall_method_arg(EmptyWithCtor());
c.thiscall_method_arg(Small());
c.thiscall_method_arg(SmallWithCtor());
c.thiscall_method_arg(Big());
}
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
struct X {
X();
~X();
};
void g(X) {
}
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
// WIN32: define dso_local void @"?g@@YAXUX@@@Z"(<{ %struct.X, [3 x i8] }>* inalloca %0) {{.*}} {
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: call x86_thiscallcc void @"??1X@@QAE@XZ"(%struct.X* {{.*}})
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// WIN32: }
void f() {
g(X());
}
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: define dso_local void @"?f@@YAXXZ"() {{.*}} {
// WIN32-NOT: call {{.*}} @"??1X@@QAE@XZ"
[ms-cxxabi] Destroy temporary record arguments in the callee Itanium destroys them in the caller at the end of the full expression, but MSVC destroys them in the callee. This is further complicated by the need to emit EH-only destructor cleanups in the caller. This should help clang compile MSVC's debug iterators more correctly. There is still an outstanding issue in PR5064 of a memcpy emitted by the LLVM backend, which is not correct for C++ records. Fixes PR16226. Reviewers: rjmccall Differential Revision: http://llvm-reviews.chandlerc.com/D929 llvm-svn: 184543
2013-06-21 20:45:15 +08:00
// WIN32: }
[ms-cxxabi] Use inalloca on win32 when passing non-trivial C++ objects When a non-trivial parameter is present, clang now gathers up all the parameters that lack inreg and puts them into a packed struct. MSVC always aligns each parameter to 4 bytes and no more, so this is a pretty simple struct to lay out. On win64, non-trivial records are passed indirectly. Prior to this change, clang was incorrectly using byval on win64. I'm able to self-host a working clang with this change and additional LLVM patches. Reviewers: rsmith Differential Revision: http://llvm-reviews.chandlerc.com/D2636 llvm-svn: 200597
2014-02-01 08:04:45 +08:00
namespace test2 {
// We used to crash on this due to the mixture of POD byval and non-trivial
// byval.
struct NonTrivial {
NonTrivial();
NonTrivial(const NonTrivial &o);
~NonTrivial();
int a;
};
struct POD { int b; };
int foo(NonTrivial a, POD b);
void bar() {
POD b;
b.b = 13;
int c = foo(NonTrivial(), b);
}
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32-LABEL: define dso_local void @"?bar@test2@@YAXXZ"() {{.*}} {
Update clang to account for changes made to LLVM in r203376 llvm-svn: 203377
2014-03-09 15:09:06 +08:00
// WIN32: %[[argmem:[^ ]*]] = alloca inalloca [[argmem_ty:<{ %"struct.test2::NonTrivial", %"struct.test2::POD" }>]]
Update Clang tests to handle explicitly typed gep changes in LLVM. llvm-svn: 230783
2015-02-28 03:18:17 +08:00
// WIN32: getelementptr inbounds [[argmem_ty]], [[argmem_ty]]* %[[argmem]], i32 0, i32 1
[ms-cxxabi] Use inalloca on win32 when passing non-trivial C++ objects When a non-trivial parameter is present, clang now gathers up all the parameters that lack inreg and puts them into a packed struct. MSVC always aligns each parameter to 4 bytes and no more, so this is a pretty simple struct to lay out. On win64, non-trivial records are passed indirectly. Prior to this change, clang was incorrectly using byval on win64. I'm able to self-host a working clang with this change and additional LLVM patches. Reviewers: rsmith Differential Revision: http://llvm-reviews.chandlerc.com/D2636 llvm-svn: 200597
2014-02-01 08:04:45 +08:00
// WIN32: call void @llvm.memcpy
Update Clang tests to handle explicitly typed gep changes in LLVM. llvm-svn: 230783
2015-02-28 03:18:17 +08:00
// WIN32: getelementptr inbounds [[argmem_ty]], [[argmem_ty]]* %[[argmem]], i32 0, i32 0
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: call x86_thiscallcc %"struct.test2::NonTrivial"* @"??0NonTrivial@test2@@QAE@XZ"
// WIN32: call i32 @"?foo@test2@@YAHUNonTrivial@1@UPOD@1@@Z"([[argmem_ty]]* inalloca %argmem)
[ms-cxxabi] Use inalloca on win32 when passing non-trivial C++ objects When a non-trivial parameter is present, clang now gathers up all the parameters that lack inreg and puts them into a packed struct. MSVC always aligns each parameter to 4 bytes and no more, so this is a pretty simple struct to lay out. On win64, non-trivial records are passed indirectly. Prior to this change, clang was incorrectly using byval on win64. I'm able to self-host a working clang with this change and additional LLVM patches. Reviewers: rsmith Differential Revision: http://llvm-reviews.chandlerc.com/D2636 llvm-svn: 200597
2014-02-01 08:04:45 +08:00
// WIN32: ret void
// WIN32: }
}
MS ABI: Use the proper type for inalloca args Summary: The definition of a type later in a translation unit may change it's type from {}* to (%struct.foo*)*. Earlier function definitions may use the former while more recent definitions might use the later. This is fine until they interact with one another (like one calling the other). In these cases, a bitcast is needed because the inalloca must match the function call but the store to the lvalue which initializes the argument slot has to match the rvalue's type. This technique is along the same lines with what the other, non-inalloca, codepaths perform. This fixes PR19287. Reviewers: rnk CC: cfe-commits Differential Revision: http://llvm-reviews.chandlerc.com/D3224 llvm-svn: 205217
2014-04-01 00:12:47 +08:00
inalloca: Pad the struct *after* inserting each arg This ensures that the overall struct size will be a multiple of 4, as required by the ABI. llvm-svn: 205981
2014-04-11 03:09:43 +08:00
namespace test3 {
// Check that we padded the inalloca struct to a multiple of 4.
struct NonTrivial {
NonTrivial();
NonTrivial(const NonTrivial &o);
~NonTrivial();
int a;
};
void foo(NonTrivial a, bool b) { }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
// WIN32-LABEL: define dso_local void @"?foo@test3@@YAXUNonTrivial@1@_N@Z"(<{ %"struct.test3::NonTrivial", i8, [3 x i8] }>* inalloca %0)
inalloca: Pad the struct *after* inserting each arg This ensures that the overall struct size will be a multiple of 4, as required by the ABI. llvm-svn: 205981
2014-04-11 03:09:43 +08:00
}
MS ABI: Use the proper type for inalloca args Summary: The definition of a type later in a translation unit may change it's type from {}* to (%struct.foo*)*. Earlier function definitions may use the former while more recent definitions might use the later. This is fine until they interact with one another (like one calling the other). In these cases, a bitcast is needed because the inalloca must match the function call but the store to the lvalue which initializes the argument slot has to match the rvalue's type. This technique is along the same lines with what the other, non-inalloca, codepaths perform. This fixes PR19287. Reviewers: rnk CC: cfe-commits Differential Revision: http://llvm-reviews.chandlerc.com/D3224 llvm-svn: 205217
2014-04-01 00:12:47 +08:00
// We would crash here because the later definition of ForwardDeclare1 results
// in a different IR type for the value we want to store. However, the alloca's
// type will use the argument type selected by fn1.
struct ForwardDeclare1;
typedef void (*FnPtr1)(ForwardDeclare1);
Fix type mismatch assertion related to inalloca and PR19287 Augment the test case from r205217 to catch this related bug. Fixes the Windows self-host which was failing on VariantValue.cpp. llvm-svn: 205378
2014-04-02 08:16:53 +08:00
void fn1(FnPtr1 a, SmallWithDtor b) { }
MS ABI: Use the proper type for inalloca args Summary: The definition of a type later in a translation unit may change it's type from {}* to (%struct.foo*)*. Earlier function definitions may use the former while more recent definitions might use the later. This is fine until they interact with one another (like one calling the other). In these cases, a bitcast is needed because the inalloca must match the function call but the store to the lvalue which initializes the argument slot has to match the rvalue's type. This technique is along the same lines with what the other, non-inalloca, codepaths perform. This fixes PR19287. Reviewers: rnk CC: cfe-commits Differential Revision: http://llvm-reviews.chandlerc.com/D3224 llvm-svn: 205217
2014-04-01 00:12:47 +08:00
struct ForwardDeclare1 {};
Fix type mismatch assertion related to inalloca and PR19287 Augment the test case from r205217 to catch this related bug. Fixes the Windows self-host which was failing on VariantValue.cpp. llvm-svn: 205378
2014-04-02 08:16:53 +08:00
void fn2(FnPtr1 a, SmallWithDtor b) { fn1(a, b); };
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32-LABEL: define dso_local void @"?fn2@@YAXP6AXUForwardDeclare1@@@ZUSmallWithDtor@@@Z"
Update Clang tests to handle explicitly typed gep changes in LLVM. llvm-svn: 230783
2015-02-28 03:18:17 +08:00
// WIN32: %[[a:[^ ]*]] = getelementptr inbounds [[argmem_ty:<{ {}\*, %struct.SmallWithDtor }>]], [[argmem_ty:<{ {}\*, %struct.SmallWithDtor }>]]* %{{.*}}, i32 0, i32 0
Fix type mismatch assertion related to inalloca and PR19287 Augment the test case from r205217 to catch this related bug. Fixes the Windows self-host which was failing on VariantValue.cpp. llvm-svn: 205378
2014-04-02 08:16:53 +08:00
// WIN32: %[[a1:[^ ]*]] = bitcast {}** %[[a]] to void [[dst_ty:\(%struct.ForwardDeclare1\*\)\*]]*
// WIN32: %[[argmem:[^ ]*]] = alloca inalloca [[argmem_ty]]
Update Clang tests to handle explicitly typed gep changes in LLVM. llvm-svn: 230783
2015-02-28 03:18:17 +08:00
// WIN32: %[[gep1:[^ ]*]] = getelementptr inbounds [[argmem_ty]], [[argmem_ty]]* %[[argmem]], i32 0, i32 1
Fix type mismatch assertion related to inalloca and PR19287 Augment the test case from r205217 to catch this related bug. Fixes the Windows self-host which was failing on VariantValue.cpp. llvm-svn: 205378
2014-04-02 08:16:53 +08:00
// WIN32: %[[bc1:[^ ]*]] = bitcast %struct.SmallWithDtor* %[[gep1]] to i8*
Change memcpy/memove/memset to have dest and source alignment attributes (Step 1). Summary: Upstream LLVM is changing the the prototypes of the @llvm.memcpy/memmove/memset intrinsics. This change updates the Clang tests for this change. The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument which is required to be a constant integer. It represents the alignment of the dest (and source), and so must be the minimum of the actual alignment of the two. This change removes the alignment argument in favour of placing the alignment attribute on the source and destination pointers of the memory intrinsic call. For example, code which used to read: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false) will now read call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false) At this time the source and destination alignments must be the same (Step 1). Step 2 of the change, to be landed shortly, will relax that contraint and allow the source and destination to have different alignments. llvm-svn: 322964
2018-01-20 01:12:54 +08:00
// WIN32: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %[[bc1]], i8* align 4 {{.*}}, i32 4, i1 false)
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// WIN32: %[[a2:[^ ]*]] = load void [[dst_ty]], void [[dst_ty]]* %[[a1]], align 4
Update Clang tests to handle explicitly typed gep changes in LLVM. llvm-svn: 230783
2015-02-28 03:18:17 +08:00
// WIN32: %[[gep2:[^ ]*]] = getelementptr inbounds [[argmem_ty]], [[argmem_ty]]* %[[argmem]], i32 0, i32 0
Fix type mismatch assertion related to inalloca and PR19287 Augment the test case from r205217 to catch this related bug. Fixes the Windows self-host which was failing on VariantValue.cpp. llvm-svn: 205378
2014-04-02 08:16:53 +08:00
// WIN32: %[[addr:[^ ]*]] = bitcast {}** %[[gep2]] to void [[dst_ty]]*
// WIN32: store void [[dst_ty]] %[[a2]], void [[dst_ty]]* %[[addr]], align 4
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32: call void @"?fn1@@YAXP6AXUForwardDeclare1@@@ZUSmallWithDtor@@@Z"([[argmem_ty]]* inalloca %[[argmem]])
[MS] Fix 'this' type when calling virtual methods with inalloca If the virtual method comes from a secondary vtable, then the type of the 'this' parameter should be i8*, and not a pointer to the complete class. In the MS ABI, the 'this' parameter on entry points to the vptr containing the virtual method that was called, so we use i8* instead of the normal type. We had a mismatch where the CGFunctionInfo of the call didn't match the CGFunctionInfo of the declaration, and this resulted in some assertions, but now both sides agree the type of 'this' is i8*. Fixes one issue raised in PR30293 llvm-svn: 280815
2016-09-07 23:15:51 +08:00
namespace pr30293 {
// Virtual methods living in a secondary vtable take i8* as their 'this'
// parameter because the 'this' parameter on entry points to the secondary
// vptr. We used to have a bug where we didn't apply this rule consistently,
// and it would cause assertion failures when used with inalloca.
struct A {
virtual void f();
};
struct B {
virtual void __cdecl h(SmallWithDtor);
};
struct C final : A, B {
void g();
void __cdecl h(SmallWithDtor);
void f();
};
void C::g() { return h(SmallWithDtor()); }
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN32-LABEL: define dso_local x86_thiscallcc void @"?g@C@pr30293@@QAEXXZ"(%"struct.pr30293::C"* %this)
// WIN32: call x86_thiscallcc %struct.SmallWithDtor* @"??0SmallWithDtor@@QAE@XZ"
// WIN32: call void @"?h@C@pr30293@@UAAXUSmallWithDtor@@@Z"(<{ i8*, %struct.SmallWithDtor }>* inalloca %{{[^,)]*}})
// WIN32: declare dso_local void @"?h@C@pr30293@@UAAXUSmallWithDtor@@@Z"(<{ i8*, %struct.SmallWithDtor }>* inalloca)
[MS] Fix 'this' type when calling virtual methods with inalloca If the virtual method comes from a secondary vtable, then the type of the 'this' parameter should be i8*, and not a pointer to the complete class. In the MS ABI, the 'this' parameter on entry points to the vptr containing the virtual method that was called, so we use i8* instead of the normal type. We had a mismatch where the CGFunctionInfo of the call didn't match the CGFunctionInfo of the declaration, and this resulted in some assertions, but now both sides agree the type of 'this' is i8*. Fixes one issue raised in PR30293 llvm-svn: 280815
2016-09-07 23:15:51 +08:00
[MS] Don't escape MS C++ names with \01 It is not needed after LLVM r327734. Now it will be easier to copy-paste IR symbol names from Clang. llvm-svn: 327738
2018-03-17 04:36:49 +08:00
// WIN64-LABEL: define dso_local void @"?g@C@pr30293@@QEAAXXZ"(%"struct.pr30293::C"* %this)
// WIN64: declare dso_local void @"?h@C@pr30293@@UEAAXUSmallWithDtor@@@Z"(i8*, i32)
[MS] Fix 'this' type when calling virtual methods with inalloca If the virtual method comes from a secondary vtable, then the type of the 'this' parameter should be i8*, and not a pointer to the complete class. In the MS ABI, the 'this' parameter on entry points to the vptr containing the virtual method that was called, so we use i8* instead of the normal type. We had a mismatch where the CGFunctionInfo of the call didn't match the CGFunctionInfo of the declaration, and this resulted in some assertions, but now both sides agree the type of 'this' is i8*. Fixes one issue raised in PR30293 llvm-svn: 280815
2016-09-07 23:15:51 +08:00
}