[Object] Beginnings of SFrame parser and dumper

This PR adds the SFrameParser class and uses it from llvm-readobj to
dump the section contents. Currently, it only supports parsing the
SFrame section header. Other parts of the section will be added in
follow-up patches.

llvm-readobj uses the same sframe flag syntax as GNU readelf, but I have
not attempted match the output format of the tool. I'm starting with the
"llvm" output format because it's easier to generate and lets us
tweak the format to make it useful for testing the generation code. If
needed, support for the GNU format could be added by overriding this
functionality in the GNU ELF Dumper.

Author: labath

llvm/include/llvm/BinaryFormat/SFrame.h

@@ -15,31 +15,35 @@
 #ifndef LLVM_BINARYFORMAT_SFRAME_H
 #define LLVM_BINARYFORMAT_SFRAME_H
 
+#include "llvm/ADT/ArrayRef.h"
 #include "llvm/ADT/BitmaskEnum.h"
 #include "llvm/Support/DataTypes.h"
 #include "llvm/Support/Endian.h"
 
-namespace llvm::sframe {
+namespace llvm {
+
+template <typename T> struct EnumEntry;
+
+namespace sframe {
 
 LLVM_ENABLE_BITMASK_ENUMS_IN_NAMESPACE();
 
 constexpr uint16_t Magic = 0xdee2;
 
 enum class Version : uint8_t {
-  V1 = 1,
-  V2 = 2,
+#define HANDLE_SFRAME_VERSION(CODE, NAME) NAME = CODE,
+#include "llvm/BinaryFormat/SFrameConstants.def"
 };
 
 enum class Flags : uint8_t {
-  FDESorted = 0x01,
-  FramePointer = 0x02,
+#define HANDLE_SFRAME_FLAG(CODE, NAME) NAME = CODE,
+#include "llvm/BinaryFormat/SFrameConstants.def"
   LLVM_MARK_AS_BITMASK_ENUM(/*LargestValue=*/0xff),
 };
 
 enum class ABI : uint8_t {
-  AArch64EndianBig = 1,
-  AArch64EndianLittle = 2,
-  AMD64EndianLittle = 3,
+#define HANDLE_SFRAME_ABI(CODE, NAME) NAME = CODE,
+#include "llvm/BinaryFormat/SFrameConstants.def"
 };
 
 /// SFrame FRE Types. Bits 0-3 of FuncDescEntry.Info.
@@ -158,6 +162,11 @@ template <endianness E> using FrameRowEntryAddr1 = FrameRowEntry<uint8_t, E>;
 template <endianness E> using FrameRowEntryAddr2 = FrameRowEntry<uint16_t, E>;
 template <endianness E> using FrameRowEntryAddr4 = FrameRowEntry<uint32_t, E>;
 
-} // namespace llvm::sframe
+ArrayRef<EnumEntry<Version>> getVersions();
+ArrayRef<EnumEntry<Flags>> getFlags();
+ArrayRef<EnumEntry<ABI>> getABIs();
+
+} // namespace sframe
+} // namespace llvm
 
 #endif // LLVM_BINARYFORMAT_SFRAME_H

llvm/include/llvm/BinaryFormat/SFrameConstants.def

@@ -0,0 +1,38 @@
+//===- SFrameConstants.def --------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#if !(defined(HANDLE_SFRAME_VERSION) || defined(HANDLE_SFRAME_FLAG) ||  \
+      defined(HANDLE_SFRAME_ABI))
+#error "Missing HANDLE_SFRAME definition"
+#endif
+
+#ifndef HANDLE_SFRAME_VERSION
+#define HANDLE_SFRAME_VERSION(CODE, NAME)
+#endif
+
+#ifndef HANDLE_SFRAME_FLAG
+#define HANDLE_SFRAME_FLAG(CODE, NAME)
+#endif
+
+#ifndef HANDLE_SFRAME_ABI
+#define HANDLE_SFRAME_ABI(CODE, NAME)
+#endif
+
+HANDLE_SFRAME_VERSION(0x01, V1)
+HANDLE_SFRAME_VERSION(0x02, V2)
+
+HANDLE_SFRAME_FLAG(0x01, FDESorted)
+HANDLE_SFRAME_FLAG(0x02, FramePointer)
+
+HANDLE_SFRAME_ABI(0x01, AArch64EndianBig)
+HANDLE_SFRAME_ABI(0x02, AArch64EndianLittle)
+HANDLE_SFRAME_ABI(0x03, AMD64EndianLittle)
+
+#undef HANDLE_SFRAME_VERSION
+#undef HANDLE_SFRAME_FLAG
+#undef HANDLE_SFRAME_ABI

llvm/include/llvm/Object/SFrameParser.h

@@ -0,0 +1,48 @@
+//===- SFrameParser.h -------------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_OBJECT_SFRAME_H
+#define LLVM_OBJECT_SFRAME_H
+
+#include "llvm/ADT/ArrayRef.h"
+#include "llvm/BinaryFormat/SFrame.h"
+#include "llvm/Support/Error.h"
+#include <cstdint>
+
+namespace llvm {
+namespace object {
+
+template <endianness E> class SFrameParser {
+public:
+  static Expected<SFrameParser> create(ArrayRef<uint8_t> Contents);
+
+  const sframe::Preamble<E> &getPreamble() const { return Header.Preamble; }
+  const sframe::Header<E> &getHeader() const { return Header; }
+
+  bool usesFixedRAOffset() const {
+    return getHeader().ABIArch == sframe::ABI::AMD64EndianLittle;
+  }
+  bool usesFixedFPOffset() const {
+    return false; // Not used in any currently defined ABI.
+  }
+
+private:
+  ArrayRef<uint8_t> Data;
+  const sframe::Header<E> &Header;
+
+  SFrameParser(ArrayRef<uint8_t> Data, const sframe::Header<E> &Header)
+      : Data(Data), Header(Header) {}
+};
+
+extern template class SFrameParser<endianness::big>;
+extern template class SFrameParser<endianness::little>;
+
+} // end namespace object
+} // end namespace llvm
+
+#endif // LLVM_OBJECT_SFRAME_H

llvm/lib/BinaryFormat/CMakeLists.txt

@@ -11,6 +11,7 @@ add_llvm_component_library(LLVMBinaryFormat
   MsgPackDocumentYAML.cpp
   MsgPackReader.cpp
   MsgPackWriter.cpp
+  SFrame.cpp
   Wasm.cpp
   XCOFF.cpp
 

llvm/lib/BinaryFormat/SFrame.cpp

@@ -0,0 +1,37 @@
+//===-- SFrame.cpp -----------------------------------------------*- C++-*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "llvm/BinaryFormat/SFrame.h"
+#include "llvm/Support/ScopedPrinter.h"
+
+using namespace llvm;
+
+ArrayRef<EnumEntry<sframe::Version>> sframe::getVersions() {
+  static constexpr EnumEntry<Version> Versions[] = {
+#define HANDLE_SFRAME_VERSION(CODE, NAME) {#NAME, sframe::Version::NAME},
+#include "llvm/BinaryFormat/SFrameConstants.def"
+  };
+
+  return ArrayRef(Versions);
+}
+
+ArrayRef<EnumEntry<sframe::Flags>> sframe::getFlags() {
+  static constexpr EnumEntry<sframe::Flags> Flags[] = {
+#define HANDLE_SFRAME_FLAG(CODE, NAME) {#NAME, sframe::Flags::NAME},
+#include "llvm/BinaryFormat/SFrameConstants.def"
+  };
+  return ArrayRef(Flags);
+}
+
+ArrayRef<EnumEntry<sframe::ABI>> sframe::getABIs() {
+  static constexpr EnumEntry<sframe::ABI> ABIs[] = {
+#define HANDLE_SFRAME_ABI(CODE, NAME) {#NAME, sframe::ABI::NAME},
+#include "llvm/BinaryFormat/SFrameConstants.def"
+  };
+  return ArrayRef(ABIs);
+}

llvm/lib/Object/CMakeLists.txt

@@ -25,6 +25,7 @@ add_llvm_component_library(LLVMObject
   OffloadBundle.cpp
   RecordStreamer.cpp
   RelocationResolver.cpp
+  SFrameParser.cpp
   SymbolicFile.cpp
   SymbolSize.cpp
   TapiFile.cpp

llvm/lib/Object/SFrameParser.cpp

@@ -0,0 +1,56 @@
+//===- SFrameParser.cpp ---------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "llvm/Object/SFrameParser.h"
+#include "llvm/BinaryFormat/SFrame.h"
+#include "llvm/Object/Error.h"
+#include "llvm/Support/FormatVariadic.h"
+
+using namespace llvm;
+using namespace llvm::object;
+
+template <typename T>
+static Expected<const T &> getDataSliceAs(ArrayRef<uint8_t> Data,
+                                          uint64_t Offset) {
+  static_assert(std::is_trivial_v<T>);
+  if (Data.size() < Offset + sizeof(T)) {
+    return make_error<GenericBinaryError>(
+        formatv("unexpected end of data at offset {0:x} while reading [{1:x}, "
+                "{2:x})",
+                Data.size(), Offset, Offset + sizeof(T))
+            .str(),
+        object_error::unexpected_eof);
+  }
+  return *reinterpret_cast<const T *>(Data.data() + Offset);
+}
+
+template <endianness E>
+Expected<SFrameParser<E>> SFrameParser<E>::create(ArrayRef<uint8_t> Contents) {
+  Expected<const sframe::Preamble<E> &> Preamble =
+      getDataSliceAs<sframe::Preamble<E>>(Contents, 0);
+  if (!Preamble)
+    return Preamble.takeError();
+
+  if (Preamble->Magic != sframe::Magic) {
+    return make_error<GenericBinaryError>("invalid magic number",
+                                          object_error::parse_failed);
+  }
+  if (Preamble->Version != sframe::Version::V2) {
+    return make_error<GenericBinaryError>("invalid/unsupported version number",
+                                          object_error::parse_failed);
+  }
+
+  Expected<const sframe::Header<E> &> Header =
+      getDataSliceAs<sframe::Header<E>>(Contents, 0);
+  if (!Header)
+    return Header.takeError();
+  return SFrameParser(Contents, *Header);
+}
+
+template class llvm::object::SFrameParser<endianness::big>;
+template class llvm::object::SFrameParser<endianness::little>;

llvm/test/tools/llvm-readobj/ELF/sframe-header.test

@@ -0,0 +1,118 @@
+## Check parsing and dumping of the SFrame header.
+# RUN: yaml2obj --docnum=1 %s -o %t.1
+# RUN: llvm-readobj --sframe=.sframe_1b --sframe=.sframe_bad_magic \
+# RUN:   --sframe=.sframe_bad_version --sframe=.sframe_6b \
+# RUN:   --sframe=.sframe_header %t.1 2>&1 | FileCheck %s --check-prefix=CASE1
+
+## Check big-endian support and the handling of --sframe argument default.
+# RUN: yaml2obj --docnum=2 %s -o %t.2
+# RUN: llvm-readobj --sframe %t.2 2>&1 | FileCheck %s --check-prefix=CASE2
+
+--- !ELF
+FileHeader:
+  Class:   ELFCLASS64
+  Data:    ELFDATA2LSB
+  Type:    ET_EXEC
+  Machine: EM_X86_64
+Sections:
+  - Name:  .sframe_1b
+    Type:  SHT_PROGBITS
+    Flags: [ SHF_ALLOC ]
+    ContentArray: [ 0x00 ]
+# CASE1-LABEL: Contents of SFrame section '.sframe_1b':
+# CASE1: warning: {{.*}}: unexpected end of data at offset 0x1 while reading [0x0, 0x4)
+
+  - Name:  .sframe_bad_magic
+    Type:  SHT_PROGBITS
+    Flags: [ SHF_ALLOC ]
+    ContentArray: [ 0xde, 0xad, 0xbe, 0xef]
+# CASE1-LABEL: Contents of SFrame section '.sframe_bad_magic':
+# CASE1: warning: {{.*}}: invalid magic number
+
+  - Name:  .sframe_bad_version
+    Type:  SHT_PROGBITS
+    Flags: [ SHF_ALLOC ]
+    ContentArray: [
+      0xe2, 0xde, 0x01, 0x00  # Preamble (magic, version, flags)
+    ]
+# CASE1-LABEL: Contents of SFrame section '.sframe_bad_version':
+# CASE1: warning: {{.*}}: invalid/unsupported version number
+
+  - Name:  .sframe_6b
+    Type:  SHT_PROGBITS
+    Flags: [ SHF_ALLOC ]
+    ContentArray: [
+      0xe2, 0xde, 0x02, 0x00,  # Preamble (magic, version, flags)
+      0x01, 0x02
+    ]
+
+# CASE1-LABEL: Contents of SFrame section '.sframe_6b':
+# CASE1: warning: {{.*}}: unexpected end of data at offset 0x6 while reading [0x0, 0x1c)
+
+  - Name:  .sframe_header
+    Type:  SHT_PROGBITS
+    Flags: [ SHF_ALLOC ]
+    ContentArray: [
+      0xe2, 0xde, 0x02, 0x00,  # Preamble (magic, version, flags)
+      # Header:
+      0x03, 0x42, 0x47, 0x00,  # ABI, Fixed FP offset, Fixed RA Offset, AUX header length
+      0x01, 0x00, 0x00, 0x00,  # Number of FDEs
+      0x10, 0x00, 0x00, 0x00,  # Number of FREs
+      0x00, 0x10, 0x00, 0x00,  # FRE length
+      0x04, 0x00, 0x00, 0x00,  # FDE offset
+      0x00, 0x01, 0x00, 0x00,  # FRE offset
+    ]
+# CASE1-LABEL: Contents of SFrame section '.sframe_header':
+# CASE1:      Header {
+# CASE1-NEXT:    Magic: 0xDEE2
+# CASE1-NEXT:    Version: V2 (0x2)
+# CASE1-NEXT:    Flags [ (0x0)
+# CASE1-NEXT:    ]
+# CASE1-NEXT:    ABI: AMD64EndianLittle (0x3)
+# CASE1-NEXT:    CFA fixed RA offset: 71
+# CASE1-NEXT:    CFA fixed FP offset (unused): 66
+# CASE1-NEXT:    Auxiliary header length: 0
+# CASE1-NEXT:    Num FDEs: 1
+# CASE1-NEXT:    Num FREs: 16
+# CASE1-NEXT:    FRE subsection length: 4096
+# CASE1-NEXT:    FDE subsection offset: 4
+# CASE1-NEXT:    FRE subsection offset: 256
+# CASE1-NEXT: }
+
+--- !ELF
+FileHeader:
+  Class: ELFCLASS64
+  Data:  ELFDATA2MSB
+  Type:  ET_EXEC
+  Machine: EM_AARCH64
+Sections:
+  - Name:  .sframe
+    Type:  SHT_PROGBITS
+    Flags: [ SHF_ALLOC ]
+    ContentArray: [
+      0xde, 0xe2, 0x02, 0x01,  # Preamble (magic, version, flags)
+      # Header:
+      0x01, 0x42, 0x47, 0x00,  # ABI, Fixed FP offset, Fixed RA Offset, AUX header length
+      0x00, 0x00, 0x00, 0x01,  # Number of FDEs
+      0x00, 0x00, 0x00, 0x10,  # Number of FREs
+      0x00, 0x00, 0x10, 0x00,  # FRE length
+      0x00, 0x00, 0x00, 0x04,  # FDE offset
+      0x00, 0x00, 0x01, 0x00,  # FRE offset
+    ]
+# CASE2-LABEL: Contents of SFrame section '.sframe':
+# CASE2:      Header {
+# CASE2-NEXT:    Magic: 0xDEE2
+# CASE2-NEXT:    Version: V2 (0x2)
+# CASE2-NEXT:    Flags [ (0x1)
+# CASE2-NEXT:    FDESorted (0x1)
+# CASE2-NEXT:    ]
+# CASE2-NEXT:    ABI: AArch64EndianBig (0x1)
+# CASE2-NEXT:    CFA fixed RA offset (unused): 71
+# CASE2-NEXT:    CFA fixed FP offset (unused): 66
+# CASE2-NEXT:    Auxiliary header length: 0
+# CASE2-NEXT:    Num FDEs: 1
+# CASE2-NEXT:    Num FREs: 16
+# CASE2-NEXT:    FRE subsection length: 4096
+# CASE2-NEXT:    FDE subsection offset: 4
+# CASE2-NEXT:    FRE subsection offset: 256
+# CASE2-NEXT: }