[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added a subscriber: Jake-Egan.
Michael137 added a comment.

Oh looks like @Jake-Egan already did


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added a comment.

Looks like `gmodules` isn't supported on AIX (based on other tests that use 
`-dwarf-ext-refs`. So I'll just disable the new tests

Thanks for notifying @aaron.ballman , email notifications got lost in the inbox


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Aaron Ballman via Phabricator via cfe-commits]

aaron.ballman added a comment.

The changes in this patch seem to have caused 
https://lab.llvm.org/buildbot/#/builders/clang-ppc64-aix to go down for the 
past 4 days -- can you revert and investigate?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[NAKAMURA Takumi via Phabricator via cfe-commits]

chapuni added inline comments.



Comment at: clang/test/Modules/gmodules-preferred-name-alias.cpp:6
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s

I guess it made difficult to find out the actual error since it hid stderr.
In fact, I had to reproduce failures locally.

In addition, mixing stdout and stderr sometimes confuse line order due to 
buffering.

btw, did it really require stderr?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

This revision was not accepted when it landed; it landed in state "Needs 
Review".
This revision was automatically updated to reflect the committed changes.
Closed by commit rG711a64412749: [clang][DebugInfo] Emit DW_AT_type of 
preferred name if available (authored by Michael137).

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

Files:
  clang/lib/CodeGen/CGDebugInfo.cpp
  clang/lib/CodeGen/CGDebugInfo.h
  clang/test/CodeGen/preferred_name-chain.cpp
  clang/test/CodeGen/preferred_name.cpp
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.h
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.modulemap
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.h
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.modulemap
  clang/test/Modules/gmodules-preferred-name-alias.cpp
  clang/test/Modules/gmodules-preferred-name-typedef.cpp
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py

Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
@@ -22,7 +22,7 @@
 
 def make_expected_basic_string_ptr(self) -> str:
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-return f'std::unique_ptr >'
+return f'std::unique_ptr'
 else:
 return 'std::unique_ptr, std::allocator >, ' \
'std::default_delete, std::allocator > > >'
Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
@@ -59,13 +59,13 @@
 self.assertNotEqual(valobj.child[0].unsigned, 0)
 
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-string_type = "std::basic_string"
+string_type = "std::string"
 else:
-string_type = "std::basic_string, std::allocator >"
+string_type = "std::basic_string, std::allocator > "
 
 valobj = self.expect_var_path(
 "sp_str",
-type="std::shared_ptr<" + string_type + " >",
+type="std::shared_ptr<" + string_type + ">",
 children=[ValueCheck(name="__ptr_", summary='"hello"')],
 )
 self.assertRegex(valobj.summary, r'^"hello"( strong=1)? weak=1$')
Index: clang/test/Modules/gmodules-preferred-name-typedef.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-typedef.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-typedef.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-typedef.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
Index: clang/test/Modules/gmodules-preferred-name-alias.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-alias.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-alias.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-alias.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: 

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added a comment.

I'll submit later today then if nobody has any other objections

Will add a release note


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Wolfgang Pieb via Phabricator via cfe-commits]

wolfgangp added a comment.

> @probinson Sounds like Sony's solution also changes the `DW_AT_type` to a 
> non-canonical form. Do you still have concerns with the direction of this 
> patch? Would it cause any problems for you downstream?

Sorry for the late reply. Looking at the patch I don't think it will cause us 
any trouble. The 2 solutions should be able to live happily side by side.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added a comment.

ping


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added a comment.

In D145803#4220124 , @wolfgangp wrote:

> In D145803#4219894 , @probinson 
> wrote:
>
>> This is pretty different from the "always desugar to the canonical type" 
>> habit that has always been in place. Sony has done some downstream things to 
>> try to work around that in the past. @wolfgangp will remember it better than 
>> I do, but I think we make some effort to preserve the type-as-written. This 
>> goes in completely the other direction.
>
> The Sony solution does not rely on a user-specified attribute, but uniformly 
> preserves all typedefs, though only in template argument types. Similar to 
> this solution, it adds a second type to keep track of the preferred type, 
> though it does so in the TemplateArgument structure as a separate QualType. A 
> client can then either print the preferred type or the canonical type, 
> depending on a printing policy, which is controlled by an option. This leads 
> to a similar result in the resulting DWARF, i.e. the DW_AT_type node gets the 
> preferred type string. The preferred type can also be used for diagnostic 
> messages.

Neat, thanks for outlining your approach. IIUC, this basically just replicates 
the `PrintingPolicy::UsePreferredNames` when printing the names of template 
specializations. But it works for all typedefs, not just those annotated with 
the attribute.

@probinson Sounds like Sony's solution also changes the `DW_AT_type` to a 
non-canonical form. Do you still have concerns with the direction of this 
patch? Would it cause any problems for you downstream?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[David Blaikie via Phabricator via cfe-commits]

dblaikie added inline comments.



Comment at: clang/test/CodeGen/preferred_name.cpp:49
+
+Foo> varFooInt;
+

Michael137 wrote:
> dblaikie wrote:
> > Michael137 wrote:
> > > probinson wrote:
> > > > This doesn't become `Foo` ?
> > > The name stays as `Foo>` but the type of the template parameter 
> > > becomes `BarInt`. So the dwarf would look like:
> > > ```
> > > DW_TAG_structure_type
> > >   DW_AT_name  ("Foo >")
> > > 
> > >   DW_TAG_template_type_parameter
> > > DW_AT_type(0x0095 "BarInt")
> > > DW_AT_name("T")
> > > 
> > > ```
> > > Will add the parameter metadata to the test
> > Hmm, that seems buggy - why doesn't `Foo >` become `Foo`? 
> > (is the preferred name ever used in the DW_AT_name? I'd have thought it 
> > should be unless it's explicitly avoided to ensure type 
> > compatibility/collapsing with debug info built without this feature 
> > enabled?)
> I suspect it's because the name is constructed using the clang TypePrinter. 
> And we turn off `PrintingPolicy::UsePreferredNames` by default to avoid 
> divergence from GCC.
> 
> Will double check though
Yeah, that sounds right/OK to me. Sorry for the diversion.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 marked an inline comment as not done.
Michael137 added inline comments.



Comment at: clang/test/CodeGen/preferred_name.cpp:49
+
+Foo> varFooInt;
+

dblaikie wrote:
> Michael137 wrote:
> > probinson wrote:
> > > This doesn't become `Foo` ?
> > The name stays as `Foo>` but the type of the template parameter 
> > becomes `BarInt`. So the dwarf would look like:
> > ```
> > DW_TAG_structure_type
> >   DW_AT_name  ("Foo >")
> > 
> >   DW_TAG_template_type_parameter
> > DW_AT_type(0x0095 "BarInt")
> > DW_AT_name("T")
> > 
> > ```
> > Will add the parameter metadata to the test
> Hmm, that seems buggy - why doesn't `Foo >` become `Foo`? 
> (is the preferred name ever used in the DW_AT_name? I'd have thought it 
> should be unless it's explicitly avoided to ensure type 
> compatibility/collapsing with debug info built without this feature enabled?)
I suspect it's because the name is constructed using the clang TypePrinter. And 
we turn off `PrintingPolicy::UsePreferredNames` by default to avoid divergence 
from GCC.

Will double check though


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[David Blaikie via Phabricator via cfe-commits]

dblaikie added inline comments.



Comment at: clang/test/CodeGen/preferred_name.cpp:49
+
+Foo> varFooInt;
+

Michael137 wrote:
> probinson wrote:
> > This doesn't become `Foo` ?
> The name stays as `Foo>` but the type of the template parameter 
> becomes `BarInt`. So the dwarf would look like:
> ```
> DW_TAG_structure_type
>   DW_AT_name  ("Foo >")
> 
>   DW_TAG_template_type_parameter
> DW_AT_type(0x0095 "BarInt")
> DW_AT_name("T")
> 
> ```
> Will add the parameter metadata to the test
Hmm, that seems buggy - why doesn't `Foo >` become `Foo`? (is 
the preferred name ever used in the DW_AT_name? I'd have thought it should be 
unless it's explicitly avoided to ensure type compatibility/collapsing with 
debug info built without this feature enabled?)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 updated this revision to Diff 508563.
Michael137 added a comment.

- Fix function docstring


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

Files:
  clang/lib/CodeGen/CGDebugInfo.cpp
  clang/lib/CodeGen/CGDebugInfo.h
  clang/test/CodeGen/preferred_name-chain.cpp
  clang/test/CodeGen/preferred_name.cpp
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.h
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.modulemap
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.h
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.modulemap
  clang/test/Modules/gmodules-preferred-name-alias.cpp
  clang/test/Modules/gmodules-preferred-name-typedef.cpp
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py

Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
@@ -22,7 +22,7 @@
 
 def make_expected_basic_string_ptr(self) -> str:
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-return f'std::unique_ptr >'
+return f'std::unique_ptr'
 else:
 return 'std::unique_ptr, std::allocator >, ' \
'std::default_delete, std::allocator > > >'
Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
@@ -59,13 +59,13 @@
 self.assertNotEqual(valobj.child[0].unsigned, 0)
 
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-string_type = "std::basic_string"
+string_type = "std::string"
 else:
-string_type = "std::basic_string, std::allocator >"
+string_type = "std::basic_string, std::allocator > "
 
 valobj = self.expect_var_path(
 "sp_str",
-type="std::shared_ptr<" + string_type + " >",
+type="std::shared_ptr<" + string_type + ">",
 children=[ValueCheck(name="__ptr_", summary='"hello"')],
 )
 self.assertRegex(valobj.summary, r'^"hello"( strong=1)? weak=1$')
Index: clang/test/Modules/gmodules-preferred-name-typedef.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-typedef.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-typedef.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-typedef.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
Index: clang/test/Modules/gmodules-preferred-name-alias.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-alias.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-alias.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-alias.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
Index: clang/test/Modules/Inputs/gmodules-preferred-name-typedef.modulemap
===
--- /dev/null
+++ 

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 updated this revision to Diff 508562.
Michael137 added a comment.

- Refine tests:
  - Check for type of template parameter
  - Add test for chained typedefs.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

Files:
  clang/lib/CodeGen/CGDebugInfo.cpp
  clang/lib/CodeGen/CGDebugInfo.h
  clang/test/CodeGen/preferred_name-chain.cpp
  clang/test/CodeGen/preferred_name.cpp
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.h
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.modulemap
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.h
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.modulemap
  clang/test/Modules/gmodules-preferred-name-alias.cpp
  clang/test/Modules/gmodules-preferred-name-typedef.cpp
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py

Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
@@ -22,7 +22,7 @@
 
 def make_expected_basic_string_ptr(self) -> str:
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-return f'std::unique_ptr >'
+return f'std::unique_ptr'
 else:
 return 'std::unique_ptr, std::allocator >, ' \
'std::default_delete, std::allocator > > >'
Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
@@ -59,13 +59,13 @@
 self.assertNotEqual(valobj.child[0].unsigned, 0)
 
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-string_type = "std::basic_string"
+string_type = "std::string"
 else:
-string_type = "std::basic_string, std::allocator >"
+string_type = "std::basic_string, std::allocator > "
 
 valobj = self.expect_var_path(
 "sp_str",
-type="std::shared_ptr<" + string_type + " >",
+type="std::shared_ptr<" + string_type + ">",
 children=[ValueCheck(name="__ptr_", summary='"hello"')],
 )
 self.assertRegex(valobj.summary, r'^"hello"( strong=1)? weak=1$')
Index: clang/test/Modules/gmodules-preferred-name-typedef.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-typedef.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-typedef.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-typedef.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
Index: clang/test/Modules/gmodules-preferred-name-alias.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-alias.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-alias.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-alias.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
Index: clang/test/Modules/Inputs/gmodules-preferred-name-typedef.modulemap

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added inline comments.



Comment at: clang/test/CodeGen/preferred_name.cpp:49
+
+Foo> varFooInt;
+

probinson wrote:
> This doesn't become `Foo` ?
The name stays as `Foo>` but the type of the template parameter 
becomes `BarInt`. So the dwarf would look like:
```
DW_TAG_structure_type
  DW_AT_name  ("Foo >")

  DW_TAG_template_type_parameter
DW_AT_type(0x0095 "BarInt")
DW_AT_name("T")

```
Will add the parameter metadata to the test


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Wolfgang Pieb via Phabricator via cfe-commits]

wolfgangp added a comment.

In D145803#4219894 , @probinson wrote:

> This is pretty different from the "always desugar to the canonical type" 
> habit that has always been in place. Sony has done some downstream things to 
> try to work around that in the past. @wolfgangp will remember it better than 
> I do, but I think we make some effort to preserve the type-as-written. This 
> goes in completely the other direction.

The Sony solution does not rely on a user-specified attribute, but uniformly 
preserves all typedefs, though only in template argument types. Similar to this 
solution, it adds a second type to keep track of the preferred type, though it 
does so in the TemplateArgument structure as a separate QualType. A client can 
then either print the preferred type or the canonical type, depending on a 
printing policy, which is controlled by an option. This leads to a similar 
result in the resulting DWARF, i.e. the DW_AT_type node gets the preferred type 
string. The preferred type can also be used for diagnostic messages.




Comment at: clang/lib/CodeGen/CGDebugInfo.h:289
+  /// PreferredNameAttr attribute on \ref RD refers to. If no such
+  /// attribute exists, returns nullptr.
+  llvm::DIType *GetPreferredNameType(const CXXRecordDecl *RD,

This comment seems a bit garbled. Also, this is just a helper function, no?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Paul Robinson via Phabricator via cfe-commits]

probinson added subscribers: wolfgangp, probinson.
probinson added a comment.

This is pretty different from the "always desugar to the canonical type" habit 
that has always been in place. Sony has done some downstream things to try to 
work around that in the past. @wolfgangp will remember it better than I do, but 
I think we make some effort to preserve the type-as-written. This goes in 
completely the other direction.




Comment at: clang/test/CodeGen/preferred_name.cpp:49
+
+Foo> varFooInt;
+

This doesn't become `Foo` ?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added inline comments.



Comment at: clang/lib/CodeGen/CGDebugInfo.cpp:2640
   if (!D || !D->isCompleteDefinition())
-return FwdDecl;
+return {FwdDecl, nullptr};
 

aprantl wrote:
> I'm curious if this works if we encounter a forward declaration, early exit 
> here, then encounter a use of the type, and then the definition, since 
> completeClass effectively also ignores the preferred name?
Good point. If we ever take this branch we won't end up emitting the preferred 
name. For my `-gmodules` test cases this works out fine because the modules 
that contain the instantiations would see the preferred name. But I can indeed 
construct a non-modules test case where we end up using a forward declared 
structure where the preferred name gets ignored here. Not sure we can do much 
here for such cases.

The alternative I considered where we create some sort of 
`PreferredNameCache[TagType*] => DIDerivedType` and use it when replacing 
forward declarations in `finalize()` doesn't work for the normal case because 
we don't have any forward declarations to replace


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Adrian Prantl via Phabricator via cfe-commits]

aprantl added inline comments.



Comment at: clang/lib/CodeGen/CGDebugInfo.cpp:2640
   if (!D || !D->isCompleteDefinition())
-return FwdDecl;
+return {FwdDecl, nullptr};
 

I'm curious if this works if we encounter a forward declaration, early exit 
here, then encounter a use of the type, and then the definition, since 
completeClass effectively also ignores the preferred name?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 updated this revision to Diff 506588.
Michael137 added a comment.
Herald added a project: LLDB.
Herald added a subscriber: lldb-commits.

- Fix cycle in `DIDerivedType` when we deal with forward declarations and 
complete explicitly via `completeClassData` (e.g., with `-gmodules`)
- Add tests for `-gmodules`


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

Files:
  clang/lib/CodeGen/CGDebugInfo.cpp
  clang/lib/CodeGen/CGDebugInfo.h
  clang/test/CodeGen/preferred_name.cpp
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.h
  clang/test/Modules/Inputs/gmodules-preferred-name-alias.modulemap
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.h
  clang/test/Modules/Inputs/gmodules-preferred-name-typedef.modulemap
  clang/test/Modules/gmodules-preferred-name-alias.cpp
  clang/test/Modules/gmodules-preferred-name-typedef.cpp
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
  
lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py

Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/unique_ptr/TestDataFormatterLibcxxUniquePtr.py
@@ -22,7 +22,7 @@
 
 def make_expected_basic_string_ptr(self) -> str:
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-return f'std::unique_ptr >'
+return f'std::unique_ptr'
 else:
 return 'std::unique_ptr, std::allocator >, ' \
'std::default_delete, std::allocator > > >'
Index: lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
===
--- lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
+++ lldb/test/API/functionalities/data-formatter/data-formatter-stl/libcxx/shared_ptr/TestDataFormatterLibcxxSharedPtr.py
@@ -59,13 +59,13 @@
 self.assertNotEqual(valobj.child[0].unsigned, 0)
 
 if self.expectedCompiler(["clang"]) and self.expectedCompilerVersion(['>', '16.0']):
-string_type = "std::basic_string"
+string_type = "std::string"
 else:
-string_type = "std::basic_string, std::allocator >"
+string_type = "std::basic_string, std::allocator > "
 
 valobj = self.expect_var_path(
 "sp_str",
-type="std::shared_ptr<" + string_type + " >",
+type="std::shared_ptr<" + string_type + ">",
 children=[ValueCheck(name="__ptr_", summary='"hello"')],
 )
 self.assertRegex(valobj.summary, r'^"hello"( strong=1)? weak=1$')
Index: clang/test/Modules/gmodules-preferred-name-typedef.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-typedef.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-typedef.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-typedef.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
Index: clang/test/Modules/gmodules-preferred-name-alias.cpp
===
--- /dev/null
+++ clang/test/Modules/gmodules-preferred-name-alias.cpp
@@ -0,0 +1,12 @@
+// RUN: rm -rf %t
+// RUN: %clang_cc1 -std=c++11 -dwarf-ext-refs -fmodule-format=obj \
+// RUN: -fmodule-map-file=%S/Inputs/gmodules-preferred-name-alias.modulemap \
+// RUN: -fmodules-cache-path=%t -debug-info-kind=standalone -debugger-tuning=lldb \
+// RUN: -fmodules -mllvm -debug-only=pchcontainer -x c++ \
+// RUN: -I %S/Inputs %s &> %t.ll
+// RUN: cat %t.ll | FileCheck %s
+
+#include "gmodules-preferred-name-alias.h"
+
+// CHECK: ![[#]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", scope: ![[#]], file: ![[#]], line: [[#]], baseType: ![[PREF_BASE:[0-9]+]])
+// CHECK: ![[PREF_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, 

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added a comment.

Apparently some Objective-C/gmodules LLDB tests aren't happy with this change
Investigating...


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 added a comment.

In D145803#4186805 , @dblaikie wrote:

> Yeah, can't say this had occurred to me - but totally makes sense/reckon it's 
> OK. Any reason to limit this to lldb? I'd expect it'd probably "Just 
> Work(tm)" on any DWARF consumer?

No particular reason other than being on the safe side and get field experience 
before enabling it for all consumers. But I agree, I don't see why this 
couldn't be enabled always.

> it doesn't hit any recursion issues? (I guess maybe skirts it due to the 
> existing recursion handling in the decl/def structure type stuff - so it 
> creates a declaration for the type, then creates the typedef, which can find 
> that existing declaration, then return the typedef from the create type 
> query?)

Yup that's the intention

> I guess it means that references to the type even for like, the type of the 
> "this" parameter - would refer to the typedef? That's /probably/ OK if a bit 
> surprising to some people/consumers sometimes?

Good point, it does also repoint the `this` pointer to the typedef. LLDB seems 
to handle this fine. Can add some test cases for this (in Clang and LLDB)


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[David Blaikie via Phabricator via cfe-commits]

dblaikie added a comment.

Yeah, can't say this had occurred to me - but totally makes sense/reckon it's 
OK. Any reason to limit this to lldb? I'd expect it'd probably "Just Work(tm)" 
on any DWARF consumer?

it doesn't hit any recursion issues? (I guess maybe skirts it due to the 
existing recursion handling in the decl/def structure type stuff - so it 
creates a declaration for the type, then creates the typedef, which can find 
that existing declaration, then return the typedef from the create type query?)

I guess it means that references to the type even for like, the type of the 
"this" parameter - would refer to the typedef? That's /probably/ OK if a bit 
surprising to some people/consumers sometimes?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D145803/new/

https://reviews.llvm.org/D145803

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

[PATCH] D145803: [clang][DebugInfo] Emit DW_AT_type of preferred name if available

[Michael Buch via Phabricator via cfe-commits]

Michael137 created this revision.
Michael137 added a reviewer: aprantl.
Herald added a project: All.
Michael137 added a comment.
Michael137 updated this revision to Diff 504274.
Michael137 updated this revision to Diff 504276.
Michael137 edited the summary of this revision.
Michael137 added a reviewer: dblaikie.
Michael137 published this revision for review.
Herald added a project: clang.
Herald added a subscriber: cfe-commits.

Yet another alternative for emitting preferred_names.

Makes the attribute introduced in https://reviews.llvm.org/D145077 redundant. 
Also doesn't require any changes to LLDB (making 
https://reviews.llvm.org/D145078 redundant, which got a bit hairy)


aprantl added a comment.

This seems to be a much simpler implementation that just automatically works. I 
think I'd prefer that over adding new DWARF attributes.


Michael137 added a comment.

- Add test


Michael137 added a comment.

- Update commit message


[clang][DebugInfo] Emit DW_AT_type of preferred name if available

With this patch, whenever we emit a `DW_AT_type` for some declaration
and the type is a template class with a `clang::PreferredNameAttr`, we
will emit the typedef that the attribute refers to instead. I.e.,

  0x123 DW_TAG_variable
  DW_AT_name "var"
  DW_AT_type (0x123 "basic_string")
  
  0x124 DW_TAG_structure_type
  DW_AT_name "basic_string"

...becomes

  0x123 DW_TAG_variable
  DW_AT_name "var"
  DW_AT_type (0x124 "std::string")
  
  0x124 DW_TAG_structure_type
  DW_AT_name "basic_string"
  
  0x125 DW_TAG_typedef
  DW_AT_name "std::string"
  DW_AT_type (0x124 "basic_string")

For now we keep this behind LLDB tuning.

**Testing**

- Added clang unit-test
- `check-llvm`, `check-clang` pass
- Confirmed that this change correctly repoints

`basic_string` references in some of my test programs.

- Will add follow-up LLDB API tests


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D145803

Files:
  clang/lib/CodeGen/CGDebugInfo.cpp
  clang/lib/CodeGen/CGDebugInfo.h
  clang/test/CodeGen/preferred_name.cpp

Index: clang/test/CodeGen/preferred_name.cpp
===
--- /dev/null
+++ clang/test/CodeGen/preferred_name.cpp
@@ -0,0 +1,85 @@
+// RUN: %clang_cc1 -triple x86_64-unk-unk -o - -emit-llvm -debug-info-kind=standalone -debugger-tuning=lldb %s | FileCheck %s --check-prefixes=COMMON,LLDB
+// RUN: %clang_cc1 -triple x86_64-unk-unk -o - -emit-llvm -debug-info-kind=standalone -debugger-tuning=gdb %s | FileCheck %s --check-prefixes=COMMON,GDB
+
+template 
+struct Foo;
+
+typedef Foo BarInt;
+typedef Foo BarDouble;
+
+template 
+using Bar = Foo;
+
+template 
+struct [[clang::preferred_name(BarInt),
+ clang::preferred_name(BarDouble),
+ clang::preferred_name(Bar),
+ clang::preferred_name(Bar)]] Foo{
+ };
+
+int main() {
+Foo varInt;
+
+// COMMON: !DILocalVariable(name: "varInt", scope: ![[#]], file: ![[#]], line: [[#]], type: ![[BAR_INT_TY:[0-9]+]])
+// LLDB:   ![[BAR_INT_TY]] = !DIDerivedType(tag: DW_TAG_typedef, name: "BarInt", file: ![[#]], line: [[#]], baseType: ![[BAR_INT_BASE:[0-9]+]])
+// LLDB:   ![[BAR_INT_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo", file: ![[#]], line: [[#]], size: [[#]]
+// GDB:![[BAR_INT_TY]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo", file: ![[#]], line: [[#]], size: [[#]]
+
+Foo varDouble;
+
+// COMMON: !DILocalVariable(name: "varDouble", scope: ![[#]], file: ![[#]], line: [[#]], type: ![[BAR_DOUBLE_TY:[0-9]+]])
+// LLDB:   ![[BAR_DOUBLE_TY]] = !DIDerivedType(tag: DW_TAG_typedef, name: "BarDouble", file: ![[#]], line: [[#]], baseType: ![[BAR_DOUBLE_BASE:[0-9]+]])
+// LLDB:   ![[BAR_DOUBLE_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
+// GDB:![[BAR_DOUBLE_TY]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
+
+Foo varShort;
+
+// COMMON: !DILocalVariable(name: "varShort", scope: ![[#]], file: ![[#]], line: [[#]], type: ![[BAR_SHORT_TY:[0-9]+]])
+// LLDB:   ![[BAR_SHORT_TY]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", file: ![[#]], line: [[#]], baseType: ![[BAR_SHORT_BASE:[0-9]+]])
+// LLDB:   ![[BAR_SHORT_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
+// GDB:![[BAR_SHORT_TY]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
+
+Foo varChar;
+
+// COMMON: !DILocalVariable(name: "varChar", scope: ![[#]], file: ![[#]], line: [[#]], type: ![[BAR_CHAR_TY:[0-9]+]])
+// LLDB:   ![[BAR_CHAR_TY]] = !DIDerivedType(tag: DW_TAG_typedef, name: "Bar", file: ![[#]], line: [[#]], baseType: ![[BAR_CHAR_BASE:[0-9]+]])
+// LLDB:   ![[BAR_CHAR_BASE]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
+// GDB:![[BAR_CHAR_TY]] = distinct !DICompositeType(tag: DW_TAG_structure_type, name: "Foo"
+
+Foo> varFooInt;
+
+// COMMON: !DILocalVariable(name: "varFooInt", scope: ![[#]], file: ![[#]], line: