Merge branch 'master' into i6495-syscall-templ-invariants

.github/workflows/ci-package.yml

@@ -471,7 +471,7 @@ jobs:
         7z x c:\projects\install\doxygen.zip -oc:\projects\install\doxygen > nul
         set PATH=c:\projects\install\doxygen;%PATH%
         dir "c:\Program Files (x86)\WiX Toolset"*
-        set PATH=C:\Program Files (x86)\WiX Toolset v3.11\bin;%PATH%
+        set PATH=C:\Program Files (x86)\WiX Toolset v3.14\bin;%PATH%
         call "C:/Program Files (x86)/Microsoft Visual Studio/2019/Enterprise/VC/Auxiliary/Build/vcvars32.bat"
         echo ------ Running suite ------
         echo PATH is "%PATH%"

api/docs/CMake_rundoxygen.cmake

@@ -256,6 +256,11 @@ if (embeddable)
   endforeach ()
 
   # We leverage the javascript menu from the non-embedded version.
+  # Menu hierarchy changes:
+  #   (1) Remove the outer "DynamoRIO" layer.
+  #   (2) Remove home page at the end.
+  #   (3) Insert an "API Reference" layer around API details.
+  #   (4) Move the "Extension API" menu inside "Build Your Own Tool".
   file(GLOB all_js ${CMAKE_CURRENT_BINARY_DIR}/../html/*.js)
   set(found_user_docs OFF)
   foreach (js ${all_js})
@@ -286,7 +291,31 @@ if (embeddable)
         message(FATAL_ERROR "Cannot find menu entry for page \"Extension API\"")
       endif ()
       string(REGEX MATCH "\n[^\n]+\"Extension API\"[^\n]+\n" ext_entry "${string}")
-      string(REPLACE "${ext_entry}" "\n" string "${string}")
+      # If ext_entry ends with a [, the menu items are expanded inline and the
+      # whole Extension API menu can be moved within this file.
+      string(REGEX MATCH "\\[\n" ext_multiline "${ext_entry}")
+      if (ext_multiline)
+        # Extract everything up to but not including the newly-inserted API Reference.
+        string(REGEX REPLACE
+               "(\n[^\n]+\"Extension API\",.+)(\n\\[ \"API Reference\")"
+               "\\2" string "${string}")
+        # Remember the full Extension API match.
+        set(ext_entry ${CMAKE_MATCH_1})
+        if (NOT ext_entry)
+          message(FATAL_ERROR "Cannot move menu entry for \"Extension API\"")
+        endif ()
+        # Now insert the Extension API menu right above Disassembly Library.
+        if (NOT string MATCHES "Disassembly Library")
+          message(FATAL_ERROR "Cannot find menu entry for page \"Disassembly Library\"")
+        endif ()
+        string(REGEX REPLACE "(\n[^\n]+\"Disassembly Library)" "${ext_entry}\\1"
+          string "${string}")
+        # Set the user_docs flag since we found its contents here.
+        set(found_user_docs ON)
+      else ()
+        # Remove the 1-line Extension API string from this file.
+        string(REPLACE "${ext_entry}" "\n" string "${string}")
+      endif ()
     else ()
       # Remove name so we can inline.
       string(REGEX REPLACE "var [^\n]* =\n" "" string "${string}")
@@ -319,6 +348,9 @@ if (embeddable)
     if (js MATCHES "page_user_docs.js")
       # CMake 3.6+ guarantees the glob is sorted lexicographically, so we've already
       # seen navtreedata.js.
+      if (found_user_docs)
+        message(FATAL_ERROR "Found unexpected \"page_user_docs\" menu file")
+      endif ()
       set(found_user_docs ON)
       if (ext_entry)
         if (NOT string MATCHES "Disassembly Library")

api/docs/bt.dox

@@ -1,5 +1,5 @@
 /* **********************************************************
- * Copyright (c) 2011-2023 Google, Inc.  All rights reserved.
+ * Copyright (c) 2011-2024 Google, Inc.  All rights reserved.
  * Copyright (c) 2007-2009 VMware, Inc.  All rights reserved.
  * **********************************************************/
 
@@ -64,7 +64,8 @@ following sections:
 \section sec_IR Instruction Representation
 
 The primary data structures involved in instruction manipulation are
-the #instr_t, which represents a single instruction, and the \c
+the #opnd_t, which represents one operand; the #instr_t, which
+represents a single instruction; and the \c
 #instrlist_t, which is a linked list of instructions.  The header files
 dr_ir_instrlist.h and dr_ir_instr.h list a number of functions that
 operate on these data structures, including:
@@ -89,6 +90,20 @@ a primary contributor to DynamoRIO's efficiency.
 The instruction representation includes all of the operands, whether
 implicit or explicit, and the condition code effects of each instruction.
 This allows for analysis of liveness of registers and condition codes.
+The operands are split into sources and destinations.
+
+A memory reference is treated as one operand even when it uses
+registers to compute its address: those constituent registers are not
+listed as their own separate source operands (unless they are read for
+other reasons such as updating the index register).  This means that a
+store to memory will have that store as a destination operand without
+listing the store's addressing mode registers as source operands in
+their own right.  Tools interested in all registers inside such
+operands can use opnd_get_num_regs_used() and opnd_get_reg_used() to
+generically walk the registers inside an operand, or
+instr_reads_from_reg() to determine whether an instruction reads a
+register either as a source operand or as a component of a destination
+memory reference.
 
 DynamoRIO's IR is mostly opaque to clients.  Key data structures have their
 sizes exposed to allow for stack allocation, but their fields are opaque.  In

api/docs/test_suite.dox

@@ -94,7 +94,7 @@ failure email step".  Be sure to match the surrounding indentation as
 indentation matters for .yml files.
 
     - name: Setup tmate session
-      if: ${{ failure() }}
+      if: failure()
       uses: mxschmitt/action-tmate@v3
 
 Next, delete all the other workflow files in the ".github/workflows/" directory.

clients/drcachesim/reader/record_file_reader.h

@@ -141,6 +141,15 @@ class record_reader_t : public std::iterator<std::input_iterator_tag, trace_entr
         return cur_entry_;
     }
 
+    static bool
+    record_is_pre_instr(trace_entry_t *record)
+    {
+        return record->type == TRACE_TYPE_ENCODING ||
+            // The branch target marker sits between any encodings and the instr.
+            (record->type == TRACE_TYPE_MARKER &&
+             record->size == TRACE_MARKER_TYPE_BRANCH_TARGET);
+    }
+
     record_reader_t &
     operator++()
     {
@@ -149,7 +158,13 @@ class record_reader_t : public std::iterator<std::input_iterator_tag, trace_entr
         UNUSED(res);
         if (!eof_) {
             ++cur_ref_count_;
-            if (type_is_instr(static_cast<trace_type_t>(cur_entry_.type)))
+            // We increment the instr count at the encoding as that avoids multiple
+            // problems with separating encodings from instrs when skipping (including
+            // for scheduler regions of interest) and when replaying schedules: anything
+            // using instr ordinals as boundaries.
+            if (!prev_record_was_pre_instr_ &&
+                (record_is_pre_instr(&cur_entry_) ||
+                 type_is_instr(static_cast<trace_type_t>(cur_entry_.type))))
                 ++cur_instr_count_;
             else if (cur_entry_.type == TRACE_TYPE_MARKER) {
                 switch (cur_entry_.size) {
@@ -177,6 +192,7 @@ class record_reader_t : public std::iterator<std::input_iterator_tag, trace_entr
                     break;
                 }
             }
+            prev_record_was_pre_instr_ = record_is_pre_instr(&cur_entry_);
         }
         return *this;
     }
@@ -273,6 +289,10 @@ class record_reader_t : public std::iterator<std::input_iterator_tag, trace_entr
     uint64_t cur_instr_count_ = 0;
     uint64_t last_timestamp_ = 0;
     uint64_t first_timestamp_ = 0;
+    // Whether the prior record was a record that immediately precedes
+    // an instruction or another record of this type: an encoding or
+    // TRACE_MARKER_TYPE_BRANCH_TARGET.
+    bool prev_record_was_pre_instr_ = false;
 
     // Remember top-level headers for the memtrace_stream_t interface.
     uint64_t version_ = 0;