-
Notifications
You must be signed in to change notification settings - Fork 13.7k
[flang][debug] Support fixed size character type. #95462
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
This PR adds debug support for fixed size character type. The character type gets translated to DIStringType. As I have noticed in comments, currently DIStringType does not have a way to represent the underlying character type of the string. This restricts our ability to represent wide string. As an example, this is how the debugger shows 2 different type of string. Note that non-ascii characters work ok with default kind string. character(kind=4, len=5) :: str1 character(len=16) :: str2 str1 = 'hello' str2 = 'π = 3.14' (gdb) p str1 $1 = 'h\000\000\000e\000\000\000l\000\000\000l\000\000\000o\000\000\000' (gdb) p str2 $2 = 'π = 3.14 '
@llvm/pr-subscribers-flang-fir-hlfir Author: Abid Qadeer (abidh) ChangesThis PR adds debug support for fixed size character type. The character type gets translated to DIStringType. As I have noticed in comments, currently DIStringType does not have a way to represent the underlying character type of the string. This restricts our ability to represent wide string. As an example, this is how the debugger shows 2 different type of string. Note that non-ascii characters work ok with default kind string. character(kind=4, len=5) :: str1 (gdb) p str1 (gdb) p str2 Full diff: https://github.com/llvm/llvm-project/pull/95462.diff 4 Files Affected:
diff --git a/flang/lib/Optimizer/Transforms/DebugTypeGenerator.cpp b/flang/lib/Optimizer/Transforms/DebugTypeGenerator.cpp
index 53745d10fe9e4..abc79010f11cc 100644
--- a/flang/lib/Optimizer/Transforms/DebugTypeGenerator.cpp
+++ b/flang/lib/Optimizer/Transforms/DebugTypeGenerator.cpp
@@ -190,6 +190,33 @@ mlir::LLVM::DITypeAttr DebugTypeGenerator::convertSequenceType(
/* associated */ nullptr);
}
+mlir::LLVM::DITypeAttr DebugTypeGenerator::convertCharacterType(
+ fir::CharacterType charTy, mlir::LLVM::DIFileAttr fileAttr,
+ mlir::LLVM::DIScopeAttr scope, mlir::Location loc) {
+ mlir::MLIRContext *context = module.getContext();
+ if (!charTy.hasConstantLen())
+ return genPlaceholderType(context);
+
+ // DWARF 5 says the following about the character encoding in 5.1.1.2.
+ // "DW_ATE_ASCII and DW_ATE_UCS specify encodings for the Fortran 2003
+ // string kinds ASCII (ISO/IEC 646:1991) and ISO_10646 (UCS-4 in ISO/IEC
+ // 10646:2000)."
+ unsigned encoding = llvm::dwarf::DW_ATE_ASCII;
+ if (charTy.getFKind() != 1)
+ encoding = llvm::dwarf::DW_ATE_UCS;
+
+ // FIXME: Currently the DIStringType in llvm does not have the option to set
+ // type of the underlying character. This restricts out ability to represent
+ // string with non-default characters. Please see issue #95440 for more
+ // details.
+ return mlir::LLVM::DIStringTypeAttr::get(
+ context, llvm::dwarf::DW_TAG_string_type,
+ mlir::StringAttr::get(context, ""),
+ charTy.getLen() * kindMapping.getCharacterBitsize(charTy.getFKind()),
+ /* alignInBits */ 0, /* stringLength */ nullptr,
+ /* stringLengthExp */ nullptr, /* stringLocationExp */ nullptr, encoding);
+}
+
mlir::LLVM::DITypeAttr
DebugTypeGenerator::convertType(mlir::Type Ty, mlir::LLVM::DIFileAttr fileAttr,
mlir::LLVM::DIScopeAttr scope,
@@ -224,6 +251,8 @@ DebugTypeGenerator::convertType(mlir::Type Ty, mlir::LLVM::DIFileAttr fileAttr,
bitWidth * 2, llvm::dwarf::DW_ATE_complex_float);
} else if (auto seqTy = mlir::dyn_cast_or_null<fir::SequenceType>(Ty)) {
return convertSequenceType(seqTy, fileAttr, scope, loc);
+ } else if (auto charTy = mlir::dyn_cast_or_null<fir::CharacterType>(Ty)) {
+ return convertCharacterType(charTy, fileAttr, scope, loc);
} else if (auto boxTy = mlir::dyn_cast_or_null<fir::BoxType>(Ty)) {
auto elTy = boxTy.getElementType();
if (auto seqTy = mlir::dyn_cast_or_null<fir::SequenceType>(elTy))
diff --git a/flang/lib/Optimizer/Transforms/DebugTypeGenerator.h b/flang/lib/Optimizer/Transforms/DebugTypeGenerator.h
index 11515d11dfed6..aa26694ab5114 100644
--- a/flang/lib/Optimizer/Transforms/DebugTypeGenerator.h
+++ b/flang/lib/Optimizer/Transforms/DebugTypeGenerator.h
@@ -45,6 +45,10 @@ class DebugTypeGenerator {
mlir::LLVM::DIFileAttr fileAttr,
mlir::LLVM::DIScopeAttr scope, mlir::Location loc,
bool genAllocated, bool genAssociated);
+ mlir::LLVM::DITypeAttr convertCharacterType(fir::CharacterType charTy,
+ mlir::LLVM::DIFileAttr fileAttr,
+ mlir::LLVM::DIScopeAttr scope,
+ mlir::Location loc);
mlir::ModuleOp module;
KindMapping kindMapping;
std::uint64_t dimsSize;
diff --git a/flang/test/Integration/debug-char-type-1.f90 b/flang/test/Integration/debug-char-type-1.f90
new file mode 100644
index 0000000000000..a0aebd3125c6e
--- /dev/null
+++ b/flang/test/Integration/debug-char-type-1.f90
@@ -0,0 +1,21 @@
+! RUN: %flang_fc1 -emit-llvm -debug-info-kind=standalone %s -o - | FileCheck %s
+
+module helper
+ character(len=40) :: str
+end module helper
+
+program test
+ use helper
+ character(kind=4, len=8) :: first
+ character(len=10) :: second
+ first = '3.14 = π'
+ second = 'Fortran'
+ str = 'Hello World!'
+end program test
+
+! CHECK-DAG: !DIGlobalVariable(name: "str"{{.*}}type: ![[TY40:[0-9]+]]{{.*}})
+! CHECK-DAG: ![[TY40]] = !DIStringType(size: 320, encoding: DW_ATE_ASCII)
+! CHECK-DAG: !DILocalVariable(name: "first"{{.*}}type: ![[TY8:[0-9]+]])
+! CHECK-DAG: ![[TY8]] = !DIStringType(size: 256, encoding: DW_ATE_UCS)
+! CHECK-DAG: !DILocalVariable(name: "second"{{.*}}type: ![[TY10:[0-9]+]])
+! CHECK-DAG: ![[TY10]] = !DIStringType(size: 80, encoding: DW_ATE_ASCII)
\ No newline at end of file
diff --git a/flang/test/Transforms/debug-char-type-1.fir b/flang/test/Transforms/debug-char-type-1.fir
new file mode 100644
index 0000000000000..cdce3b7b8b334
--- /dev/null
+++ b/flang/test/Transforms/debug-char-type-1.fir
@@ -0,0 +1,19 @@
+// RUN: fir-opt --add-debug-info --mlir-print-debuginfo %s | FileCheck %s
+
+module attributes {dlti.dl_spec = #dlti.dl_spec<>} {
+ fir.global @_QMhelperEstr1 : !fir.char<1,40> {
+ %0 = fir.zero_bits !fir.char<1,40>
+ fir.has_value %0 : !fir.char<1,40>
+ } loc(#loc1)
+ fir.global @_QMhelperEstr2 : !fir.char<4,20> {
+ %0 = fir.zero_bits !fir.char<4,20>
+ fir.has_value %0 : !fir.char<4,20>
+ } loc(#loc1)
+}
+#loc1 = loc("string.f90":1:1)
+
+// CHECK-DAG: #[[TY1:.*]] = #llvm.di_string_type<tag = DW_TAG_string_type, name = "", sizeInBits = 320, encoding = DW_ATE_ASCII>
+// CHECK-DAG: #llvm.di_global_variable<{{.*}}name = "str1"{{.*}}type = #[[TY1]]{{.*}}>
+// CHECK-DAG: #[[TY2:.*]] = #llvm.di_string_type<tag = DW_TAG_string_type, name = "", sizeInBits = 640, encoding = DW_ATE_UCS>
+// CHECK-DAG: #llvm.di_global_variable<{{.*}}name = "str2"{{.*}}type = #[[TY2]]{{.*}}>
+
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, thanks!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. Thank you!
context, llvm::dwarf::DW_TAG_string_type, | ||
mlir::StringAttr::get(context, ""), | ||
charTy.getLen() * kindMapping.getCharacterBitsize(charTy.getFKind()), | ||
/* alignInBits */ 0, /* stringLength */ nullptr, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: the usual style for these comments is /*alignBits=*/
(no spaces and =
at the end).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I will fix it here and throughout the file in a separete PR. Thanks.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
As mentioned in [here](#95462 (comment)), the formatting of the comments have been fixed. Also added comments before literal arguments.
This PR adds debug support for fixed size character type. The character type gets translated to DIStringType.
As I have noticed in comments, currently DIStringType does not have a way to represent the underlying character type of the string. This restricts our ability to represent wide string. As an example, this is how the debugger shows 2 different type of string. Note that non-ascii characters work ok with default kind string.
character(kind=4, len=5) :: str1
character(len=16) :: str2
str1 = 'hello'
str2 = 'π = 3.14'
(gdb) p str1
$1 = 'h\000\000\000e\000\000\000l\000\000\000l\000\000\000o\000\000\000'
(gdb) p str2
$2 = 'π = 3.14 '