Merge branch '4598' into improve_remove

CMakeLists.txt

@@ -405,6 +405,7 @@ if(EMSCRIPTEN)
   target_link_libraries(binaryen_wasm "-s NODERAWFS=0")
   target_link_libraries(binaryen_wasm "-s EXPORT_NAME=Binaryen")
   target_link_libraries(binaryen_wasm "-s EXPORT_ES6=1")
+  target_link_libraries(binaryen_wasm "-sEXPORTED_RUNTIME_METHODS=allocate")
   target_link_libraries(binaryen_wasm "--post-js ${CMAKE_CURRENT_SOURCE_DIR}/src/js/binaryen.js-post.js")
   target_link_libraries(binaryen_wasm "--extern-pre-js ${CMAKE_CURRENT_SOURCE_DIR}/src/js/binaryen.js-extern-pre.js")
   target_link_libraries(binaryen_wasm optimized "--closure 1")
@@ -439,6 +440,7 @@ if(EMSCRIPTEN)
   else()
     target_link_libraries(binaryen_js "-s EXPORT_ES6=1")
   endif()
+  target_link_libraries(binaryen_js "-sEXPORTED_RUNTIME_METHODS=allocate")
   target_link_libraries(binaryen_js "--post-js ${CMAKE_CURRENT_SOURCE_DIR}/src/js/binaryen.js-post.js")
   # js_of_ocaml needs a specified variable with special comment to provide the library to consumers
   if(JS_OF_OCAML)

README.md

@@ -561,6 +561,10 @@ The `check.py` script supports some options:
    tool cannot be found, and you'll see a warning.
  * We have tests from upstream in `tests/spec`, in git submodules. Running
    `./check.py` should update those.
+
+Note that we are trying to gradually port the legacy wasm-opt tests to use `lit` and `filecheck` as we modify them.
+For `passes` tests that output wast, this can be done automatically with `scripts/port_passes_tests_to_lit.py` and for non-`passes` tests that output wast, see 
+https://github.com/WebAssembly/binaryen/pull/4779 for an example of how to do a simple manual port.
 
 ### Setting up dependencies
 

scripts/fuzz_opt.py

@@ -263,7 +263,18 @@ def is_git_repo():
     # not all relaxed SIMD instructions are implemented in the interpreter
     'relaxed-simd.wast',
     # TODO fuzzer and interpreter support for strings
-    'strings.wast'
+    'strings.wast',
+    # ignore DWARF because it is incompatible with multivalue atm
+    'zlib.wasm',
+    'cubescript.wasm',
+    'class_with_dwarf_noprint.wasm',
+    'fib2_dwarf.wasm',
+    'fib_nonzero-low-pc_dwarf.wasm',
+    'inlined_to_start_dwarf.wasm',
+    'fannkuch3_manyopts_dwarf.wasm',
+    'fib2_emptylocspan_dwarf.wasm',
+    'fannkuch3_dwarf.wasm',
+    'multi_unit_abbrev_noprint.wasm',
 ]
 
 
@@ -350,9 +361,6 @@ def pick_initial_contents():
         #  )
         # )
         '--disable-multivalue',
-        # DWARF is incompatible with multivalue atm; it's more important to
-        # fuzz multivalue since we aren't actually fuzzing DWARF here
-        '--strip-dwarf',
     ]
 
     # the given wasm may not work with the chosen feature opts. for example, if
@@ -823,8 +831,8 @@ def handle_pair(self, input, before_wasm, after_wasm, opts):
         b1 = open('b1.wasm', 'rb').read()
         b2 = open('b2.wasm', 'rb').read()
         if (b1 != b2):
-            run([in_bin('wasm-dis'), 'b1.wasm', '-o', 'b1.wat', FEATURE_OPTS])
-            run([in_bin('wasm-dis'), 'b2.wasm', '-o', 'b2.wat', FEATURE_OPTS])
+            run([in_bin('wasm-dis'), 'b1.wasm', '-o', 'b1.wat'] + FEATURE_OPTS)
+            run([in_bin('wasm-dis'), 'b2.wasm', '-o', 'b2.wat'] + FEATURE_OPTS)
             t1 = open('b1.wat', 'r').read()
             t2 = open('b2.wat', 'r').read()
             compare(t1, t2, 'Output must be deterministic.', verbose=False)

scripts/gen-s-parser.py

@@ -617,6 +617,13 @@
     ("string.new_wtf8",      "makeStringNew(s, StringNewWTF8)"),
     ("string.new_wtf16",     "makeStringNew(s, StringNewWTF16)"),
     ("string.const",         "makeStringConst(s)"),
+    ("string.measure_wtf8",  "makeStringMeasure(s, StringMeasureWTF8)"),
+    ("string.measure_wtf16", "makeStringMeasure(s, StringMeasureWTF16)"),
+    ("string.is_usv_sequence", "makeStringMeasure(s, StringMeasureIsUSV)"),
+    ("string.encode_wtf8",   "makeStringEncode(s, StringEncodeWTF8)"),
+    ("string.encode_wtf16",  "makeStringEncode(s, StringEncodeWTF16)"),
+    ("string.concat",        "makeStringConcat(s)"),
+    ("string.eq",            "makeStringEq(s)"),
 ]
 
 
@@ -686,7 +693,7 @@ def insert(self, inst, expr):
         self.do_insert(inst, inst, expr)
 
 
-def instruction_parser():
+def instruction_parser(new_parser=False):
     """Build a trie out of all the instructions, then emit it as C++ code."""
     trie = Node()
     inst_length = 0
@@ -696,12 +703,23 @@ def instruction_parser():
 
     printer = CodePrinter()
 
-    printer.print_line("char op[{}] = {{'\\0'}};".format(inst_length + 1))
-    printer.print_line("strncpy(op, s[0]->c_str(), {});".format(inst_length))
+    if not new_parser:
+        printer.print_line("char op[{}] = {{'\\0'}};".format(inst_length + 1))
+        printer.print_line("strncpy(op, s[0]->c_str(), {});".format(inst_length))
 
     def print_leaf(expr, inst):
-        printer.print_line("if (strcmp(op, \"{inst}\") == 0) {{ return {expr}; }}"
-                           .format(inst=inst, expr=expr))
+        if new_parser:
+            expr = expr.replace("()", "(ctx)")
+            expr = expr.replace("(s", "(ctx, in")
+            printer.print_line("if (op == \"{inst}\"sv) {{".format(inst=inst))
+            with printer.indent():
+                printer.print_line("auto ret = {expr};".format(expr=expr))
+                printer.print_line("CHECK_ERR(ret);")
+                printer.print_line("return *ret;")
+            printer.print_line("}")
+        else:
+            printer.print_line("if (strcmp(op, \"{inst}\") == 0) {{ return {expr}; }}"
+                               .format(inst=inst, expr=expr))
         printer.print_line("goto parse_error;")
 
     def emit(node, idx=0):
@@ -730,7 +748,10 @@ def emit(node, idx=0):
     emit(trie)
     printer.print_line("parse_error:")
     with printer.indent():
-        printer.print_line("throw ParseException(std::string(op), s.line, s.col);")
+        if new_parser:
+            printer.print_line("return in.err(\"unrecognized instruction\");")
+        else:
+            printer.print_line("throw ParseException(std::string(op), s.line, s.col);")
 
 
 def print_header():
@@ -756,6 +777,8 @@ def main():
         sys.exit(1)
     print_header()
     generate_with_guard(instruction_parser, "INSTRUCTION_PARSER")
+    print()
+    generate_with_guard(lambda: instruction_parser(True), "NEW_INSTRUCTION_PARSER")
     print_footer()
 
 

scripts/update_lit_checks.py

@@ -40,7 +40,7 @@
 ALL_ITEMS = '|'.join(['type', 'import', 'global', 'memory', 'data', 'table',
                       'elem', 'tag', 'export', 'start', 'func'])
 ITEM_NAME = r'\$?[^\s()]*|"[^\s()]*"'
-ITEM_RE = re.compile(r'(^\s*)\((' + ALL_ITEMS + r')\s+(' + ITEM_NAME + ').*$',
+ITEM_RE = re.compile(r'(?:^\s*\(rec\s*)?(^\s*)\((' + ALL_ITEMS + r')\s+(' + ITEM_NAME + ').*$',
                      re.MULTILINE)
 
 FUZZ_EXEC_FUNC = re.compile(r'^\[fuzz-exec\] calling (?P<name>\S*)$')