maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/clang/test/CodeGenObjC/arc-literals.m

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

162 lines
7.1 KiB
Mathematica
Raw Normal View History

[Clang/Test]: Rename enable_noundef_analysis to disable-noundef-analysis and turn it off by default Turning on `enable_noundef_analysis` flag allows better codegen by removing freeze instructions. I modified clang by renaming `enable_noundef_analysis` flag to `disable-noundef-analysis` and turning it off by default. Test updates are made as a separate patch: D108453 Reviewed By: eugenis Differential Revision: https://reviews.llvm.org/D105169
2022-01-16 17:53:11 +08:00
// RUN: %clang_cc1 -I %S/Inputs -triple x86_64-apple-darwin10 -emit-llvm -fblocks -fobjc-arc -fobjc-runtime-has-weak -O2 -disable-llvm-passes -disable-noundef-analysis -o - %s | FileCheck %s
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
#include "literal-support.h"
// Check the various selector names we'll be using, in order.
// CHECK: c"numberWithInt:\00"
// CHECK: c"numberWithUnsignedInt:\00"
// CHECK: c"numberWithUnsignedLongLong:\00"
// CHECK: c"numberWithChar:\00"
Objective-C. revert patch for rdar://17554063. llvm-svn: 220812
2014-10-29 02:28:16 +08:00
// CHECK: c"arrayWithObjects:count:\00"
// CHECK: c"dictionaryWithObjects:forKeys:count:\00"
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// CHECK: c"prop\00"
[test] Add {{.*}} to make tests immune to dso_local/dso_preemptable/(none) differences For a definition (of most linkage types), dso_local is set for ELF -fno-pic/-fpie and COFF, but not for Mach-O. This nuance causes unneeded binary format differences. This patch replaces (function) `define ` with `define{{.*}} `, (variable/constant/alias) `= ` with `={{.*}} `, or inserts appropriate `{{.*}} ` if there is an explicit linkage. * Clang will set dso_local for Mach-O, which is currently implied by TargetMachine.cpp. This will make COFF/Mach-O and executable ELF similar. * Eventually I hope we can make dso_local the textual LLVM IR default (write explicit "dso_preemptable" when applicable) and -fpic ELF will be similar to everything else. This patch helps move toward that goal.
2020-12-31 12:45:56 +08:00
// CHECK-LABEL: define{{.*}} void @test_numeric()
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
void test_numeric() {
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK: {{call.*objc_msgSend.*i32 17.* [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]}}
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
id ilit = @17;
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK: {{call.*objc_msgSend.*i32 25.* [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]}}
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
id ulit = @25u;
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK: {{call.*objc_msgSend.*i64 42.* [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]}}
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
id ulllit = @42ull;
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK: {{call.*objc_msgSend.*i8 signext 97.* [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]}}
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
id charlit = @'a';
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
Revert "Revert r234581, it might have caused a few miscompiles in Chromium." This reverts commit r234700. It turns out that the lifetime markers were not the cause of Chromium failing but a bug which was uncovered by optimizations exposed by the markers. llvm-svn: 235553
2015-04-23 05:38:15 +08:00
// CHECK: call void @llvm.lifetime.end
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
Revert "Revert r234581, it might have caused a few miscompiles in Chromium." This reverts commit r234700. It turns out that the lifetime markers were not the cause of Chromium failing but a bug which was uncovered by optimizations exposed by the markers. llvm-svn: 235553
2015-04-23 05:38:15 +08:00
// CHECK: call void @llvm.lifetime.end
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
Revert "Revert r234581, it might have caused a few miscompiles in Chromium." This reverts commit r234700. It turns out that the lifetime markers were not the cause of Chromium failing but a bug which was uncovered by optimizations exposed by the markers. llvm-svn: 235553
2015-04-23 05:38:15 +08:00
// CHECK: call void @llvm.lifetime.end
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
Revert "Revert r234581, it might have caused a few miscompiles in Chromium." This reverts commit r234700. It turns out that the lifetime markers were not the cause of Chromium failing but a bug which was uncovered by optimizations exposed by the markers. llvm-svn: 235553
2015-04-23 05:38:15 +08:00
// CHECK: call void @llvm.lifetime.end
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// CHECK-NEXT: ret void
}
[test] Add {{.*}} to make tests immune to dso_local/dso_preemptable/(none) differences For a definition (of most linkage types), dso_local is set for ELF -fno-pic/-fpie and COFF, but not for Mach-O. This nuance causes unneeded binary format differences. This patch replaces (function) `define ` with `define{{.*}} `, (variable/constant/alias) `= ` with `={{.*}} `, or inserts appropriate `{{.*}} ` if there is an explicit linkage. * Clang will set dso_local for Mach-O, which is currently implied by TargetMachine.cpp. This will make COFF/Mach-O and executable ELF similar. * Eventually I hope we can make dso_local the textual LLVM IR default (write explicit "dso_preemptable" when applicable) and -fpic ELF will be similar to everything else. This patch helps move toward that goal.
2020-12-31 12:45:56 +08:00
// CHECK-LABEL: define{{.*}} void @test_array
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
void test_array(id a, id b) {
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK: [[A:%.*]] = alloca i8*,
// CHECK: [[B:%.*]] = alloca i8*,
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Retaining parameters
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call i8* @llvm.objc.retain(i8*
// CHECK: call i8* @llvm.objc.retain(i8*
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Constructing the array
Compute and preserve alignment more faithfully in IR-generation. Introduce an Address type to bundle a pointer value with an alignment. Introduce APIs on CGBuilderTy to work with Address values. Change core APIs on CGF/CGM to traffic in Address where appropriate. Require alignments to be non-zero. Update a ton of code to compute and propagate alignment information. As part of this, I've promoted CGBuiltin's EmitPointerWithAlignment helper function to CGF and made use of it in a number of places in the expression emitter. The end result is that we should now be significantly more correct when performing operations on objects that are locally known to be under-aligned. Since alignment is not reliably tracked in the type system, there are inherent limits to this, but at least we are no longer confused by standard operations like derived-to-base conversions and array-to-pointer decay. I've also fixed a large number of bugs where we were applying the complete-object alignment to a pointer instead of the non-virtual alignment, although most of these were hidden by the very conservative approach we took with member alignment. Also, because IRGen now reliably asserts on zero alignments, we should no longer be subject to an absurd but frustrating recurring bug where an incomplete type would report a zero alignment and then we'd naively do a alignmentAtOffset on it and emit code using an alignment equal to the largest power-of-two factor of the offset. We should also now be emitting much more aggressive alignment attributes in the presence of over-alignment. In particular, field access now uses alignmentAtOffset instead of min. Several times in this patch, I had to change the existing code-generation pattern in order to more effectively use the Address APIs. For the most part, this seems to be a strict improvement, like doing pointer arithmetic with GEPs instead of ptrtoint. That said, I've tried very hard to not change semantics, but it is likely that I've failed in a few places, for which I apologize. ABIArgInfo now always carries the assumed alignment of indirect and indirect byval arguments. In order to cut down on what was already a dauntingly large patch, I changed the code to never set align attributes in the IR on non-byval indirect arguments. That is, we still generate code which assumes that indirect arguments have the given alignment, but we don't express this information to the backend except where it's semantically required (i.e. on byvals). This is likely a minor regression for those targets that did provide this information, but it'll be trivial to add it back in a later patch. I partially punted on applying this work to CGBuiltin. Please do not add more uses of the CreateDefaultAligned{Load,Store} APIs; they will be going away eventually. llvm-svn: 246985
2015-09-08 16:05:57 +08:00
// CHECK: [[T0:%.*]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[OBJECTS:%[A-Za-z0-9]+]], i64 0, i64 0
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[V0:%.*]] = load i8*, i8** [[A]],
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: store i8* [[V0]], i8** [[T0]]
Compute and preserve alignment more faithfully in IR-generation. Introduce an Address type to bundle a pointer value with an alignment. Introduce APIs on CGBuilderTy to work with Address values. Change core APIs on CGF/CGM to traffic in Address where appropriate. Require alignments to be non-zero. Update a ton of code to compute and propagate alignment information. As part of this, I've promoted CGBuiltin's EmitPointerWithAlignment helper function to CGF and made use of it in a number of places in the expression emitter. The end result is that we should now be significantly more correct when performing operations on objects that are locally known to be under-aligned. Since alignment is not reliably tracked in the type system, there are inherent limits to this, but at least we are no longer confused by standard operations like derived-to-base conversions and array-to-pointer decay. I've also fixed a large number of bugs where we were applying the complete-object alignment to a pointer instead of the non-virtual alignment, although most of these were hidden by the very conservative approach we took with member alignment. Also, because IRGen now reliably asserts on zero alignments, we should no longer be subject to an absurd but frustrating recurring bug where an incomplete type would report a zero alignment and then we'd naively do a alignmentAtOffset on it and emit code using an alignment equal to the largest power-of-two factor of the offset. We should also now be emitting much more aggressive alignment attributes in the presence of over-alignment. In particular, field access now uses alignmentAtOffset instead of min. Several times in this patch, I had to change the existing code-generation pattern in order to more effectively use the Address APIs. For the most part, this seems to be a strict improvement, like doing pointer arithmetic with GEPs instead of ptrtoint. That said, I've tried very hard to not change semantics, but it is likely that I've failed in a few places, for which I apologize. ABIArgInfo now always carries the assumed alignment of indirect and indirect byval arguments. In order to cut down on what was already a dauntingly large patch, I changed the code to never set align attributes in the IR on non-byval indirect arguments. That is, we still generate code which assumes that indirect arguments have the given alignment, but we don't express this information to the backend except where it's semantically required (i.e. on byvals). This is likely a minor regression for those targets that did provide this information, but it'll be trivial to add it back in a later patch. I partially punted on applying this work to CGBuiltin. Please do not add more uses of the CreateDefaultAligned{Load,Store} APIs; they will be going away eventually. llvm-svn: 246985
2015-09-08 16:05:57 +08:00
// CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[OBJECTS]], i64 0, i64 1
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[V1:%.*]] = load i8*, i8** [[B]],
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: store i8* [[V1]], i8** [[T0]]
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[T0:%.*]] = load [[CLASS_T:%.*]]*, [[CLASS_T:%.*]]** @"OBJC_CLASSLIST
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: [[T1:%.*]] = bitcast [[CLASS_T]]* [[T0]] to i8*
[ObjC] Emit selector load right before msgSend call. We currently emit the selector load early, but only because we need it to compute the signature (so that we know which msgSend variant to call). We can prepare the signature with a plain undef, and replace it with the materialized selector value if (and only if) needed, later. Concretely, this usually doesn't have an effect, but tests need updating because we reordered the receiver bitcast and the selector load, which is always fine. There is one notable change: with this, when a msgSend needs a receiver null check, the selector is now loaded in the non-null block, instead of before the null check. That should be a mild improvement.
2022-01-27 09:31:26 +08:00
// CHECK-NEXT: [[SEL:%.*]] = load i8*, i8** @OBJC_SELECTOR_REFERENCES
Objective-C. revert patch for rdar://17554063. llvm-svn: 220812
2014-10-29 02:28:16 +08:00
// CHECK-NEXT: [[T2:%.*]] = bitcast [2 x i8*]* [[OBJECTS]] to i8**
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK-NEXT: [[T3:%.*]] = call i8* bitcast ({{.*@objc_msgSend.*}})(i8* [[T1]], i8* [[SEL]], i8** [[T2]], i64 2) [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]
Use @llvm.objc.clang.arc.use intrinsic instead of clang.arc.use function. Calls to this function are deleted in the ARC optimizer. However when the ARC optimizer was updated to use intrinsics instead of functions (r349534), the corresponding clang change (r349535) to use intrinsics missed this one so it wasn't being deleted. llvm-svn: 349782
2018-12-21 02:05:41 +08:00
// CHECK: call void (...) @llvm.objc.clang.arc.use(i8* [[V0]], i8* [[V1]])
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
id arr = @[a, b];
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
// CHECK: call void @llvm.objc.release
// CHECK: call void @llvm.objc.release
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// CHECK-NEXT: ret void
}
[test] Add {{.*}} to make tests immune to dso_local/dso_preemptable/(none) differences For a definition (of most linkage types), dso_local is set for ELF -fno-pic/-fpie and COFF, but not for Mach-O. This nuance causes unneeded binary format differences. This patch replaces (function) `define ` with `define{{.*}} `, (variable/constant/alias) `= ` with `={{.*}} `, or inserts appropriate `{{.*}} ` if there is an explicit linkage. * Clang will set dso_local for Mach-O, which is currently implied by TargetMachine.cpp. This will make COFF/Mach-O and executable ELF similar. * Eventually I hope we can make dso_local the textual LLVM IR default (write explicit "dso_preemptable" when applicable) and -fpic ELF will be similar to everything else. This patch helps move toward that goal.
2020-12-31 12:45:56 +08:00
// CHECK-LABEL: define{{.*}} void @test_dictionary
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
void test_dictionary(id k1, id o1, id k2, id o2) {
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK: [[K1:%.*]] = alloca i8*,
// CHECK: [[O1:%.*]] = alloca i8*,
// CHECK: [[K2:%.*]] = alloca i8*,
// CHECK: [[O2:%.*]] = alloca i8*,
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Retaining parameters
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call i8* @llvm.objc.retain(i8*
// CHECK: call i8* @llvm.objc.retain(i8*
// CHECK: call i8* @llvm.objc.retain(i8*
// CHECK: call i8* @llvm.objc.retain(i8*
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Constructing the arrays
Compute and preserve alignment more faithfully in IR-generation. Introduce an Address type to bundle a pointer value with an alignment. Introduce APIs on CGBuilderTy to work with Address values. Change core APIs on CGF/CGM to traffic in Address where appropriate. Require alignments to be non-zero. Update a ton of code to compute and propagate alignment information. As part of this, I've promoted CGBuiltin's EmitPointerWithAlignment helper function to CGF and made use of it in a number of places in the expression emitter. The end result is that we should now be significantly more correct when performing operations on objects that are locally known to be under-aligned. Since alignment is not reliably tracked in the type system, there are inherent limits to this, but at least we are no longer confused by standard operations like derived-to-base conversions and array-to-pointer decay. I've also fixed a large number of bugs where we were applying the complete-object alignment to a pointer instead of the non-virtual alignment, although most of these were hidden by the very conservative approach we took with member alignment. Also, because IRGen now reliably asserts on zero alignments, we should no longer be subject to an absurd but frustrating recurring bug where an incomplete type would report a zero alignment and then we'd naively do a alignmentAtOffset on it and emit code using an alignment equal to the largest power-of-two factor of the offset. We should also now be emitting much more aggressive alignment attributes in the presence of over-alignment. In particular, field access now uses alignmentAtOffset instead of min. Several times in this patch, I had to change the existing code-generation pattern in order to more effectively use the Address APIs. For the most part, this seems to be a strict improvement, like doing pointer arithmetic with GEPs instead of ptrtoint. That said, I've tried very hard to not change semantics, but it is likely that I've failed in a few places, for which I apologize. ABIArgInfo now always carries the assumed alignment of indirect and indirect byval arguments. In order to cut down on what was already a dauntingly large patch, I changed the code to never set align attributes in the IR on non-byval indirect arguments. That is, we still generate code which assumes that indirect arguments have the given alignment, but we don't express this information to the backend except where it's semantically required (i.e. on byvals). This is likely a minor regression for those targets that did provide this information, but it'll be trivial to add it back in a later patch. I partially punted on applying this work to CGBuiltin. Please do not add more uses of the CreateDefaultAligned{Load,Store} APIs; they will be going away eventually. llvm-svn: 246985
2015-09-08 16:05:57 +08:00
// CHECK: [[T0:%.*]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[KEYS:%[A-Za-z0-9]+]], i64 0, i64 0
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[V0:%.*]] = load i8*, i8** [[K1]],
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: store i8* [[V0]], i8** [[T0]]
Compute and preserve alignment more faithfully in IR-generation. Introduce an Address type to bundle a pointer value with an alignment. Introduce APIs on CGBuilderTy to work with Address values. Change core APIs on CGF/CGM to traffic in Address where appropriate. Require alignments to be non-zero. Update a ton of code to compute and propagate alignment information. As part of this, I've promoted CGBuiltin's EmitPointerWithAlignment helper function to CGF and made use of it in a number of places in the expression emitter. The end result is that we should now be significantly more correct when performing operations on objects that are locally known to be under-aligned. Since alignment is not reliably tracked in the type system, there are inherent limits to this, but at least we are no longer confused by standard operations like derived-to-base conversions and array-to-pointer decay. I've also fixed a large number of bugs where we were applying the complete-object alignment to a pointer instead of the non-virtual alignment, although most of these were hidden by the very conservative approach we took with member alignment. Also, because IRGen now reliably asserts on zero alignments, we should no longer be subject to an absurd but frustrating recurring bug where an incomplete type would report a zero alignment and then we'd naively do a alignmentAtOffset on it and emit code using an alignment equal to the largest power-of-two factor of the offset. We should also now be emitting much more aggressive alignment attributes in the presence of over-alignment. In particular, field access now uses alignmentAtOffset instead of min. Several times in this patch, I had to change the existing code-generation pattern in order to more effectively use the Address APIs. For the most part, this seems to be a strict improvement, like doing pointer arithmetic with GEPs instead of ptrtoint. That said, I've tried very hard to not change semantics, but it is likely that I've failed in a few places, for which I apologize. ABIArgInfo now always carries the assumed alignment of indirect and indirect byval arguments. In order to cut down on what was already a dauntingly large patch, I changed the code to never set align attributes in the IR on non-byval indirect arguments. That is, we still generate code which assumes that indirect arguments have the given alignment, but we don't express this information to the backend except where it's semantically required (i.e. on byvals). This is likely a minor regression for those targets that did provide this information, but it'll be trivial to add it back in a later patch. I partially punted on applying this work to CGBuiltin. Please do not add more uses of the CreateDefaultAligned{Load,Store} APIs; they will be going away eventually. llvm-svn: 246985
2015-09-08 16:05:57 +08:00
// CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[OBJECTS:%[A-Za-z0-9]+]], i64 0, i64 0
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[V1:%.*]] = load i8*, i8** [[O1]],
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: store i8* [[V1]], i8** [[T0]]
Compute and preserve alignment more faithfully in IR-generation. Introduce an Address type to bundle a pointer value with an alignment. Introduce APIs on CGBuilderTy to work with Address values. Change core APIs on CGF/CGM to traffic in Address where appropriate. Require alignments to be non-zero. Update a ton of code to compute and propagate alignment information. As part of this, I've promoted CGBuiltin's EmitPointerWithAlignment helper function to CGF and made use of it in a number of places in the expression emitter. The end result is that we should now be significantly more correct when performing operations on objects that are locally known to be under-aligned. Since alignment is not reliably tracked in the type system, there are inherent limits to this, but at least we are no longer confused by standard operations like derived-to-base conversions and array-to-pointer decay. I've also fixed a large number of bugs where we were applying the complete-object alignment to a pointer instead of the non-virtual alignment, although most of these were hidden by the very conservative approach we took with member alignment. Also, because IRGen now reliably asserts on zero alignments, we should no longer be subject to an absurd but frustrating recurring bug where an incomplete type would report a zero alignment and then we'd naively do a alignmentAtOffset on it and emit code using an alignment equal to the largest power-of-two factor of the offset. We should also now be emitting much more aggressive alignment attributes in the presence of over-alignment. In particular, field access now uses alignmentAtOffset instead of min. Several times in this patch, I had to change the existing code-generation pattern in order to more effectively use the Address APIs. For the most part, this seems to be a strict improvement, like doing pointer arithmetic with GEPs instead of ptrtoint. That said, I've tried very hard to not change semantics, but it is likely that I've failed in a few places, for which I apologize. ABIArgInfo now always carries the assumed alignment of indirect and indirect byval arguments. In order to cut down on what was already a dauntingly large patch, I changed the code to never set align attributes in the IR on non-byval indirect arguments. That is, we still generate code which assumes that indirect arguments have the given alignment, but we don't express this information to the backend except where it's semantically required (i.e. on byvals). This is likely a minor regression for those targets that did provide this information, but it'll be trivial to add it back in a later patch. I partially punted on applying this work to CGBuiltin. Please do not add more uses of the CreateDefaultAligned{Load,Store} APIs; they will be going away eventually. llvm-svn: 246985
2015-09-08 16:05:57 +08:00
// CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[KEYS]], i64 0, i64 1
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[V2:%.*]] = load i8*, i8** [[K2]],
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: store i8* [[V2]], i8** [[T0]]
Compute and preserve alignment more faithfully in IR-generation. Introduce an Address type to bundle a pointer value with an alignment. Introduce APIs on CGBuilderTy to work with Address values. Change core APIs on CGF/CGM to traffic in Address where appropriate. Require alignments to be non-zero. Update a ton of code to compute and propagate alignment information. As part of this, I've promoted CGBuiltin's EmitPointerWithAlignment helper function to CGF and made use of it in a number of places in the expression emitter. The end result is that we should now be significantly more correct when performing operations on objects that are locally known to be under-aligned. Since alignment is not reliably tracked in the type system, there are inherent limits to this, but at least we are no longer confused by standard operations like derived-to-base conversions and array-to-pointer decay. I've also fixed a large number of bugs where we were applying the complete-object alignment to a pointer instead of the non-virtual alignment, although most of these were hidden by the very conservative approach we took with member alignment. Also, because IRGen now reliably asserts on zero alignments, we should no longer be subject to an absurd but frustrating recurring bug where an incomplete type would report a zero alignment and then we'd naively do a alignmentAtOffset on it and emit code using an alignment equal to the largest power-of-two factor of the offset. We should also now be emitting much more aggressive alignment attributes in the presence of over-alignment. In particular, field access now uses alignmentAtOffset instead of min. Several times in this patch, I had to change the existing code-generation pattern in order to more effectively use the Address APIs. For the most part, this seems to be a strict improvement, like doing pointer arithmetic with GEPs instead of ptrtoint. That said, I've tried very hard to not change semantics, but it is likely that I've failed in a few places, for which I apologize. ABIArgInfo now always carries the assumed alignment of indirect and indirect byval arguments. In order to cut down on what was already a dauntingly large patch, I changed the code to never set align attributes in the IR on non-byval indirect arguments. That is, we still generate code which assumes that indirect arguments have the given alignment, but we don't express this information to the backend except where it's semantically required (i.e. on byvals). This is likely a minor regression for those targets that did provide this information, but it'll be trivial to add it back in a later patch. I partially punted on applying this work to CGBuiltin. Please do not add more uses of the CreateDefaultAligned{Load,Store} APIs; they will be going away eventually. llvm-svn: 246985
2015-09-08 16:05:57 +08:00
// CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[OBJECTS]], i64 0, i64 1
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[V3:%.*]] = load i8*, i8** [[O2]],
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: store i8* [[V3]], i8** [[T0]]
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Constructing the dictionary
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[T0:%.*]] = load [[CLASS_T:%.*]]*, [[CLASS_T:%.*]]** @"OBJC_CLASSLIST
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: [[T1:%.*]] = bitcast [[CLASS_T]]* [[T0]] to i8*
[ObjC] Emit selector load right before msgSend call. We currently emit the selector load early, but only because we need it to compute the signature (so that we know which msgSend variant to call). We can prepare the signature with a plain undef, and replace it with the materialized selector value if (and only if) needed, later. Concretely, this usually doesn't have an effect, but tests need updating because we reordered the receiver bitcast and the selector load, which is always fine. There is one notable change: with this, when a msgSend needs a receiver null check, the selector is now loaded in the non-null block, instead of before the null check. That should be a mild improvement.
2022-01-27 09:31:26 +08:00
// CHECK-NEXT: [[SEL:%.*]] = load i8*, i8** @OBJC_SELECTOR_REFERENCES
Objective-C. revert patch for rdar://17554063. llvm-svn: 220812
2014-10-29 02:28:16 +08:00
// CHECK-NEXT: [[T2:%.*]] = bitcast [2 x i8*]* [[OBJECTS]] to i8**
// CHECK-NEXT: [[T3:%.*]] = bitcast [2 x i8*]* [[KEYS]] to i8**
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK-NEXT: [[T4:%.*]] = call i8* bitcast ({{.*@objc_msgSend.*}})(i8* [[T1]], i8* [[SEL]], i8** [[T2]], i8** [[T3]], i64 2) [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]
// CHECK-NEXT: call void (...) @llvm.objc.clang.arc.noop.use(i8* [[T4]])
Use @llvm.objc.clang.arc.use intrinsic instead of clang.arc.use function. Calls to this function are deleted in the ARC optimizer. However when the ARC optimizer was updated to use intrinsics instead of functions (r349534), the corresponding clang change (r349535) to use intrinsics missed this one so it wasn't being deleted. llvm-svn: 349782
2018-12-21 02:05:41 +08:00
// CHECK-NEXT: call void (...) @llvm.objc.clang.arc.use(i8* [[V0]], i8* [[V1]], i8* [[V2]], i8* [[V3]])
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
id dict = @{ k1 : o1, k2 : o2 };
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
// CHECK: call void @llvm.objc.release
// CHECK: call void @llvm.objc.release
// CHECK: call void @llvm.objc.release
// CHECK: call void @llvm.objc.release
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// CHECK-NEXT: ret void
}
@interface A
@end
@interface B
@property (retain) A* prop;
@end
[test] Add {{.*}} to make tests immune to dso_local/dso_preemptable/(none) differences For a definition (of most linkage types), dso_local is set for ELF -fno-pic/-fpie and COFF, but not for Mach-O. This nuance causes unneeded binary format differences. This patch replaces (function) `define ` with `define{{.*}} `, (variable/constant/alias) `= ` with `={{.*}} `, or inserts appropriate `{{.*}} ` if there is an explicit linkage. * Clang will set dso_local for Mach-O, which is currently implied by TargetMachine.cpp. This will make COFF/Mach-O and executable ELF similar. * Eventually I hope we can make dso_local the textual LLVM IR default (write explicit "dso_preemptable" when applicable) and -fpic ELF will be similar to everything else. This patch helps move toward that goal.
2020-12-31 12:45:56 +08:00
// CHECK-LABEL: define{{.*}} void @test_property
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
void test_property(B *b) {
// Retain parameter
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call i8* @llvm.objc.retain
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
Compute and preserve alignment more faithfully in IR-generation. Introduce an Address type to bundle a pointer value with an alignment. Introduce APIs on CGBuilderTy to work with Address values. Change core APIs on CGF/CGM to traffic in Address where appropriate. Require alignments to be non-zero. Update a ton of code to compute and propagate alignment information. As part of this, I've promoted CGBuiltin's EmitPointerWithAlignment helper function to CGF and made use of it in a number of places in the expression emitter. The end result is that we should now be significantly more correct when performing operations on objects that are locally known to be under-aligned. Since alignment is not reliably tracked in the type system, there are inherent limits to this, but at least we are no longer confused by standard operations like derived-to-base conversions and array-to-pointer decay. I've also fixed a large number of bugs where we were applying the complete-object alignment to a pointer instead of the non-virtual alignment, although most of these were hidden by the very conservative approach we took with member alignment. Also, because IRGen now reliably asserts on zero alignments, we should no longer be subject to an absurd but frustrating recurring bug where an incomplete type would report a zero alignment and then we'd naively do a alignmentAtOffset on it and emit code using an alignment equal to the largest power-of-two factor of the offset. We should also now be emitting much more aggressive alignment attributes in the presence of over-alignment. In particular, field access now uses alignmentAtOffset instead of min. Several times in this patch, I had to change the existing code-generation pattern in order to more effectively use the Address APIs. For the most part, this seems to be a strict improvement, like doing pointer arithmetic with GEPs instead of ptrtoint. That said, I've tried very hard to not change semantics, but it is likely that I've failed in a few places, for which I apologize. ABIArgInfo now always carries the assumed alignment of indirect and indirect byval arguments. In order to cut down on what was already a dauntingly large patch, I changed the code to never set align attributes in the IR on non-byval indirect arguments. That is, we still generate code which assumes that indirect arguments have the given alignment, but we don't express this information to the backend except where it's semantically required (i.e. on byvals). This is likely a minor regression for those targets that did provide this information, but it'll be trivial to add it back in a later patch. I partially punted on applying this work to CGBuiltin. Please do not add more uses of the CreateDefaultAligned{Load,Store} APIs; they will be going away eventually. llvm-svn: 246985
2015-09-08 16:05:57 +08:00
// CHECK: [[T0:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[OBJECTS:%.*]], i64 0, i64 0
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Invoke 'prop'
[ObjC] Emit selector load right before msgSend call. We currently emit the selector load early, but only because we need it to compute the signature (so that we know which msgSend variant to call). We can prepare the signature with a plain undef, and replace it with the materialized selector value if (and only if) needed, later. Concretely, this usually doesn't have an effect, but tests need updating because we reordered the receiver bitcast and the selector load, which is always fine. There is one notable change: with this, when a msgSend needs a receiver null check, the selector is now loaded in the non-null block, instead of before the null check. That should be a mild improvement.
2022-01-27 09:31:26 +08:00
// CHECK: [[T1:%.*]] = bitcast
// CHECK-NEXT: [[SEL:%.*]] = load i8*, i8** @OBJC_SELECTOR_REFERENCES
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK-NEXT: [[V0:%.*]] = call [[B:%.*]]* bitcast ({{.*}} @objc_msgSend to {{.*}})(i8* [[T1]], i8* [[SEL]]) [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]
// CHECK-NEXT: call void (...) @llvm.objc.clang.arc.noop.use([[B]]* [[V0]])
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: [[V1:%.*]] = bitcast [[B]]* [[V0]] to i8*
// Store to array.
// CHECK-NEXT: store i8* [[V1]], i8** [[T0]]
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
Objective-C. revert patch for rdar://17554063. llvm-svn: 220812
2014-10-29 02:28:16 +08:00
// Invoke arrayWithObjects:count:
Update Clang tests to handle explicitly typed load changes in LLVM. llvm-svn: 230795
2015-02-28 05:19:58 +08:00
// CHECK-NEXT: [[T0:%.*]] = load [[CLASS_T]]*, [[CLASS_T]]** @"OBJC_CLASSLIST
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: [[T1:%.*]] = bitcast [[CLASS_T]]* [[T0]] to i8*
[ObjC] Emit selector load right before msgSend call. We currently emit the selector load early, but only because we need it to compute the signature (so that we know which msgSend variant to call). We can prepare the signature with a plain undef, and replace it with the materialized selector value if (and only if) needed, later. Concretely, this usually doesn't have an effect, but tests need updating because we reordered the receiver bitcast and the selector load, which is always fine. There is one notable change: with this, when a msgSend needs a receiver null check, the selector is now loaded in the non-null block, instead of before the null check. That should be a mild improvement.
2022-01-27 09:31:26 +08:00
// CHECK-NEXT: [[SEL:%.*]] = load i8*, i8** @OBJC_SELECTOR_REFERENCES
Objective-C. revert patch for rdar://17554063. llvm-svn: 220812
2014-10-29 02:28:16 +08:00
// CHECK-NEXT: [[T2:%.*]] = bitcast [1 x i8*]* [[OBJECTS]] to i8**
[ObjC][ARC] Use operand bundle "clang.arc.attachedcall" on x86-64 https://reviews.llvm.org/D92808 made clang use the operand bundle instead of emitting retainRV/claimRV calls on arm64. This commit makes changes to clang that are needed to use the operand bundle on x86-64. Differential Revision: https://reviews.llvm.org/D111331
2021-11-09 10:01:48 +08:00
// CHECK-NEXT: [[T3:%.*]] = call i8* bitcast ({{.*}} @objc_msgSend to {{.*}}(i8* [[T1]], i8* [[SEL]], i8** [[T2]], i64 1) [ "clang.arc.attachedcall"(i8* (i8*)* @llvm.objc.retainAutoreleasedReturnValue) ]
// CHECK-NEXT: call void (...) @llvm.objc.clang.arc.noop.use(i8* [[T3]])
Use @llvm.objc.clang.arc.use intrinsic instead of clang.arc.use function. Calls to this function are deleted in the ARC optimizer. However when the ARC optimizer was updated to use intrinsics instead of functions (r349534), the corresponding clang change (r349535) to use intrinsics missed this one so it wasn't being deleted. llvm-svn: 349782
2018-12-21 02:05:41 +08:00
// CHECK-NEXT: call void (...) @llvm.objc.clang.arc.use(i8* [[V1]])
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: bitcast
// CHECK-NEXT: bitcast
// CHECK-NEXT: store
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
id arr = @[ b.prop ];
// Release b.prop
Protect the values of array and dictionary literals from the ARC optimizer while they're held in local unsafe buffers. Based on a patch by Jesse Rusak! rdar://13573224 llvm-svn: 178721
2013-04-04 08:20:38 +08:00
// CHECK-NEXT: [[T0:%.*]] = bitcast [[B]]* [[V0]] to i8*
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK-NEXT: call void @llvm.objc.release(i8* [[T0]])
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Destroy arr
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// Destroy b
Generate objc intrinsics instead of runtime calls as the ARC optimizer now works only on intrinsics Differential Revision: https://reviews.llvm.org/D55802 Reviewers: rjmccall llvm-svn: 349535
2018-12-19 04:33:00 +08:00
// CHECK: call void @llvm.objc.release
Add clang support for new Objective-C literal syntax for NSDictionary, NSArray, NSNumber, and boolean literals. This includes both Sema and Codegen support. Included is also support for new Objective-C container subscripting. My apologies for the large patch. It was very difficult to break apart. The patch introduces changes to the driver as well to cause clang to link in additional runtime support when needed to support the new language features. Docs are forthcoming to document the implementation and behavior of these features. llvm-svn: 152137
2012-03-07 04:05:56 +08:00
// CHECK-NEXT: ret void
[ObjC] Emit selector load right before msgSend call. We currently emit the selector load early, but only because we need it to compute the signature (so that we know which msgSend variant to call). We can prepare the signature with a plain undef, and replace it with the materialized selector value if (and only if) needed, later. Concretely, this usually doesn't have an effect, but tests need updating because we reordered the receiver bitcast and the selector load, which is always fine. There is one notable change: with this, when a msgSend needs a receiver null check, the selector is now loaded in the non-null block, instead of before the null check. That should be a mild improvement.
2022-01-27 09:31:26 +08:00
}