[lld-macho] create __TEXT,__unwind_info from __LD,__compact_unwind

Digest the input `__LD,__compact_unwind` and produce the output `__TEXT,__unwind_info`. This is the initial commit with the major functionality.

Successor commits will add handling for ...
* `__TEXT,__eh_frame`
* personalities & LSDA
* `-r` pass-through

Differential Revision: https://reviews.llvm.org/D86805

lld/MachO/CMakeLists.txt

@@ -2,8 +2,11 @@ set(LLVM_TARGET_DEFINITIONS Options.td)
 tablegen(LLVM Options.inc -gen-opt-parser-defs)
 add_public_tablegen_target(MachOOptionsTableGen)
 
+include_directories(${LLVM_MAIN_SRC_DIR}/../libunwind/include)
+
 add_lld_library(lldMachO2
   Arch/X86_64.cpp
+  UnwindInfoSection.cpp
   Driver.cpp
   DriverUtils.cpp
   ExportTrie.cpp

lld/MachO/OutputSegment.h

@@ -22,6 +22,7 @@ constexpr const char text[] = "__TEXT";
 constexpr const char data[] = "__DATA";
 constexpr const char linkEdit[] = "__LINKEDIT";
 constexpr const char dataConst[] = "__DATA_CONST";
+constexpr const char ld[] = "__LD"; // output only with -r
 
 } // namespace segment_names
 

lld/MachO/SyntheticSections.h

@@ -35,6 +35,10 @@ constexpr const char symbolTable[] = "__symbol_table";
 constexpr const char stringTable[] = "__string_table";
 constexpr const char got[] = "__got";
 constexpr const char threadPtrs[] = "__thread_ptrs";
+constexpr const char unwindInfo[] = "__unwind_info";
+// these are not synthetic, but in service of synthetic __unwind_info
+constexpr const char compactUnwind[] = "__compact_unwind";
+constexpr const char ehFrame[] = "__eh_frame";
 
 } // namespace section_names
 

lld/MachO/UnwindInfoSection.cpp

@@ -0,0 +1,284 @@
+//===- UnwindInfoSection.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 "UnwindInfoSection.h"
+#include "Config.h"
+#include "InputSection.h"
+#include "MergedOutputSection.h"
+#include "OutputSection.h"
+#include "OutputSegment.h"
+#include "Symbols.h"
+#include "SyntheticSections.h"
+#include "Target.h"
+
+#include "lld/Common/ErrorHandler.h"
+#include "llvm/ADT/SmallVector.h"
+#include "llvm/BinaryFormat/MachO.h"
+
+using namespace llvm;
+using namespace llvm::MachO;
+using namespace lld;
+using namespace lld::macho;
+
+// Compact Unwind format is a Mach-O evolution of DWARF Unwind that
+// optimizes space and exception-time lookup.  Most DWARF unwind
+// entries can be replaced with Compact Unwind entries, but the ones
+// that cannot are retained in DWARF form.
+//
+// This comment will address macro-level organization of the pre-link
+// and post-link compact unwind tables. For micro-level organization
+// pertaining to the bitfield layout of the 32-bit compact unwind
+// entries, see libunwind/include/mach-o/compact_unwind_encoding.h
+//
+// Important clarifying factoids:
+//
+// * __LD,__compact_unwind is the compact unwind format for compiler
+// output and linker input. It is never a final output. It could be
+// an intermediate output with the `-r` option which retains relocs.
+//
+// * __TEXT,__unwind_info is the compact unwind format for final
+// linker output. It is never an input.
+//
+// * __TEXT,__eh_frame is the DWARF format for both linker input and output.
+//
+// * __TEXT,__unwind_info entries are divided into 4 KiB pages (2nd
+// level) by ascending address, and the pages are referenced by an
+// index (1st level) in the section header.
+//
+// * Following the headers in __TEXT,__unwind_info, the bulk of the
+// section contains a vector of compact unwind entries
+// `{functionOffset, encoding}` sorted by ascending `functionOffset`.
+// Adjacent entries with the same encoding can be folded to great
+// advantage, achieving a 3-order-of-magnitude reduction in the
+// number of entries.
+//
+// * The __TEXT,__unwind_info format can accommodate up to 127 unique
+// encodings for the space-efficient compressed format. In practice,
+// fewer than a dozen unique encodings are used by C++ programs of
+// all sizes. Therefore, we don't even bother implementing the regular
+// non-compressed format. Time will tell if anyone in the field ever
+// overflows the 127-encodings limit.
+
+// TODO(gkm): prune __eh_frame entries superseded by __unwind_info
+// TODO(gkm): how do we align the 2nd-level pages?
+
+UnwindInfoSection::UnwindInfoSection()
+    : SyntheticSection(segment_names::text, section_names::unwindInfo) {}
+
+bool UnwindInfoSection::isNeeded() const {
+  return (compactUnwindSection != nullptr);
+}
+
+// Scan the __LD,__compact_unwind entries and compute the space needs of
+// __TEXT,__unwind_info and __TEXT,__eh_frame
+
+void UnwindInfoSection::finalize() {
+  if (compactUnwindSection == nullptr)
+    return;
+
+  // At this point, the address space for __TEXT,__text has been
+  // assigned, so we can relocate the __LD,__compact_unwind entries
+  // into a temporary buffer. Relocation is necessary in order to sort
+  // the CU entries by function address. Sorting is necessary so that
+  // we can fold adjacent CU entries with identical
+  // encoding+personality+lsda. Folding is necessary because it reduces
+  // the number of CU entries by as much as 3 orders of magnitude!
+  compactUnwindSection->finalize();
+  assert(compactUnwindSection->getSize() % sizeof(CompactUnwindEntry64) == 0);
+  size_t cuCount =
+      compactUnwindSection->getSize() / sizeof(CompactUnwindEntry64);
+  cuVector.resize(cuCount);
+  // Relocate all __LD,__compact_unwind entries
+  compactUnwindSection->writeTo(reinterpret_cast<uint8_t *>(cuVector.data()));
+
+  // Rather than sort & fold the 32-byte entries directly, we create a
+  // vector of pointers to entries and sort & fold that instead.
+  cuPtrVector.reserve(cuCount);
+  for (const auto &cuEntry : cuVector)
+    cuPtrVector.emplace_back(&cuEntry);
+  std::sort(cuPtrVector.begin(), cuPtrVector.end(),
+            [](const CompactUnwindEntry64 *a, const CompactUnwindEntry64 *b) {
+              return a->functionAddress < b->functionAddress;
+            });
+
+  // Fold adjacent entries with matching encoding+personality+lsda
+  // We use three iterators on the same cuPtrVector to fold in-situ:
+  // (1) `foldBegin` is the first of a potential sequence of matching entries
+  // (2) `foldEnd` is the first non-matching entry after `foldBegin`.
+  // The semi-open interval [ foldBegin .. foldEnd ) contains a range
+  // entries that can be folded into a single entry and written to ...
+  // (3) `foldWrite`
+  auto foldWrite = cuPtrVector.begin();
+  for (auto foldBegin = cuPtrVector.begin(); foldBegin < cuPtrVector.end();) {
+    auto foldEnd = foldBegin;
+    while (++foldEnd < cuPtrVector.end() &&
+           (*foldBegin)->encoding == (*foldEnd)->encoding &&
+           (*foldBegin)->personality == (*foldEnd)->personality &&
+           (*foldBegin)->lsda == (*foldEnd)->lsda)
+      ;
+    *foldWrite++ = *foldBegin;
+    foldBegin = foldEnd;
+  }
+  cuPtrVector.erase(foldWrite, cuPtrVector.end());
+
+  // Count frequencies of the folded encodings
+  llvm::DenseMap<compact_unwind_encoding_t, size_t> encodingFrequencies;
+  for (auto cuPtrEntry : cuPtrVector)
+    encodingFrequencies[cuPtrEntry->encoding]++;
+  if (encodingFrequencies.size() > UNWIND_INFO_COMMON_ENCODINGS_MAX)
+    error("TODO(gkm): handle common encodings table overflow");
+
+  // Make a table of encodings, sorted by descending frequency
+  for (const auto &frequency : encodingFrequencies)
+    commonEncodings.emplace_back(frequency);
+  std::sort(commonEncodings.begin(), commonEncodings.end(),
+            [](const std::pair<compact_unwind_encoding_t, size_t> &a,
+               const std::pair<compact_unwind_encoding_t, size_t> &b) {
+              if (a.second == b.second)
+                // When frequencies match, secondarily sort on encoding
+                // to maintain parity with validate-unwind-info.py
+                return a.first > b.first;
+              return a.second > b.second;
+            });
+
+  // Split folded encodings into pages, limited by capacity of a page
+  // and the 24-bit range of function offset
+  //
+  // Record the page splits as a vector of iterators on cuPtrVector
+  // such that successive elements form a semi-open interval. E.g.,
+  // page X's bounds are thus: [ pageBounds[X] .. pageBounds[X+1] )
+  //
+  // Note that pageBounds.size() is one greater than the number of
+  // pages, and pageBounds.back() holds the sentinel cuPtrVector.cend()
+  pageBounds.push_back(cuPtrVector.cbegin());
+  // TODO(gkm): cut 1st page entries short to accommodate section headers ???
+  CompactUnwindEntry64 cuEntryKey;
+  for (size_t i = 0;;) {
+    // Limit the search to entries that can fit within a 4 KiB page.
+    const auto pageBegin = pageBounds[0] + i;
+    const auto pageMax =
+        pageBounds[0] +
+        std::min(i + UNWIND_INFO_COMPRESSED_SECOND_LEVEL_ENTRIES_MAX,
+                 cuPtrVector.size());
+    // Exclude entries with functionOffset that would overflow 24 bits
+    cuEntryKey.functionAddress = (*pageBegin)->functionAddress +
+                                 UNWIND_INFO_COMPRESSED_ENTRY_FUNC_OFFSET_MASK;
+    const auto pageBreak = std::lower_bound(
+        pageBegin, pageMax, &cuEntryKey,
+        [](const CompactUnwindEntry64 *a, const CompactUnwindEntry64 *b) {
+          return a->functionAddress < b->functionAddress;
+        });
+    pageBounds.push_back(pageBreak);
+    if (pageBreak == cuPtrVector.cend())
+      break;
+    i = pageBreak - cuPtrVector.cbegin();
+  }
+
+  // compute size of __TEXT,__unwind_info section
+  level2PagesOffset =
+      sizeof(unwind_info_section_header) +
+      commonEncodings.size() * sizeof(uint32_t) +
+      personalities.size() * sizeof(uint32_t) +
+      pageBounds.size() * sizeof(unwind_info_section_header_index_entry) +
+      lsdaEntries.size() * sizeof(unwind_info_section_header_lsda_index_entry);
+  unwindInfoSize = level2PagesOffset +
+                   (pageBounds.size() - 1) *
+                       sizeof(unwind_info_compressed_second_level_page_header) +
+                   cuPtrVector.size() * sizeof(uint32_t);
+}
+
+// All inputs are relocated and output adddresses are known, so write!
+
+void UnwindInfoSection::writeTo(uint8_t *buf) const {
+  // section header
+  auto *uip = reinterpret_cast<unwind_info_section_header *>(buf);
+  uip->version = 1;
+  uip->commonEncodingsArraySectionOffset = sizeof(unwind_info_section_header);
+  uip->commonEncodingsArrayCount = commonEncodings.size();
+  uip->personalityArraySectionOffset =
+      uip->commonEncodingsArraySectionOffset +
+      (uip->commonEncodingsArrayCount * sizeof(uint32_t));
+  uip->personalityArrayCount = personalities.size();
+  uip->indexSectionOffset = uip->personalityArraySectionOffset +
+                            (uip->personalityArrayCount * sizeof(uint32_t));
+  uip->indexCount = pageBounds.size();
+
+  // Common encodings
+  auto *i32p = reinterpret_cast<uint32_t *>(&uip[1]);
+  for (const auto &encoding : commonEncodings)
+    *i32p++ = encoding.first;
+
+  // Personalities
+  for (const auto &personality : personalities)
+    *i32p++ = personality;
+
+  // Level-1 index
+  uint32_t lsdaOffset =
+      uip->indexSectionOffset +
+      uip->indexCount * sizeof(unwind_info_section_header_index_entry);
+  uint64_t l2PagesOffset = level2PagesOffset;
+  auto *iep = reinterpret_cast<unwind_info_section_header_index_entry *>(i32p);
+  for (size_t i = 0; i < pageBounds.size() - 1; i++) {
+    iep->functionOffset = (*pageBounds[i])->functionAddress;
+    iep->secondLevelPagesSectionOffset = l2PagesOffset;
+    iep->lsdaIndexArraySectionOffset = lsdaOffset;
+    iep++;
+    // TODO(gkm): pad to 4 KiB page boundary ???
+    size_t entryCount = pageBounds[i + 1] - pageBounds[i];
+    uint64_t pageSize = sizeof(unwind_info_section_header_index_entry) +
+                        entryCount * sizeof(uint32_t);
+    l2PagesOffset += pageSize;
+  }
+  // Level-1 sentinel
+  const CompactUnwindEntry64 &cuEnd = cuVector.back();
+  iep->functionOffset = cuEnd.functionAddress + cuEnd.functionLength;
+  iep->secondLevelPagesSectionOffset = 0;
+  iep->lsdaIndexArraySectionOffset = lsdaOffset;
+  iep++;
+
+  // LSDAs
+  auto *lep =
+      reinterpret_cast<unwind_info_section_header_lsda_index_entry *>(iep);
+  for (const auto &lsda : lsdaEntries) {
+    lep->functionOffset = lsda.functionOffset;
+    lep->lsdaOffset = lsda.lsdaOffset;
+  }
+
+  // create map from encoding to common-encoding-table index compact
+  // encoding entries use 7 bits to index the common-encoding table
+  size_t i = 0;
+  llvm::DenseMap<compact_unwind_encoding_t, size_t> commonEncodingIndexes;
+  for (const auto &encoding : commonEncodings)
+    commonEncodingIndexes[encoding.first] = i++;
+
+  // Level-2 pages
+  auto *p2p =
+      reinterpret_cast<unwind_info_compressed_second_level_page_header *>(lep);
+  for (size_t i = 0; i < pageBounds.size() - 1; i++) {
+    p2p->kind = UNWIND_SECOND_LEVEL_COMPRESSED;
+    p2p->entryPageOffset =
+        sizeof(unwind_info_compressed_second_level_page_header);
+    p2p->entryCount = pageBounds[i + 1] - pageBounds[i];
+    p2p->encodingsPageOffset =
+        p2p->entryPageOffset + p2p->entryCount * sizeof(uint32_t);
+    p2p->encodingsCount = 0;
+    auto *ep = reinterpret_cast<uint32_t *>(&p2p[1]);
+    auto cuPtrVectorIt = pageBounds[i];
+    uintptr_t functionAddressBase = (*cuPtrVectorIt)->functionAddress;
+    while (cuPtrVectorIt < pageBounds[i + 1]) {
+      const CompactUnwindEntry64 *cuep = *cuPtrVectorIt++;
+      size_t cueIndex = commonEncodingIndexes.lookup(cuep->encoding);
+      *ep++ = ((cueIndex << UNWIND_INFO_COMPRESSED_ENTRY_FUNC_OFFSET_BITS) |
+               (cuep->functionAddress - functionAddressBase));
+    }
+    p2p =
+        reinterpret_cast<unwind_info_compressed_second_level_page_header *>(ep);
+  }
+  assert(getSize() ==
+         static_cast<size_t>((reinterpret_cast<uint8_t *>(p2p) - buf)));
+}