Move longjmp into native code

src/library.js

@@ -1524,27 +1524,33 @@ LibraryManager.library = {
   // setjmp.h
   // ==========================================================================
 
-  longjmp__sig: 'vii',
-#if SUPPORT_LONGJMP
-  longjmp: function(env, value) {
-    _setThrew(env, value || 1);
-    throw 'longjmp';
-  },
-#else
+  _emscripten_throw_longjmp__sig: 'v',
+  _emscripten_throw_longjmp: function() { throw 'longjmp'; },
+#if !SUPPORT_LONGJMP
+  // These are in order to print helpful error messages when either longjmp of
+  // setjmp is used.
   longjmp__deps: [function() {
     error('longjmp support was disabled (SUPPORT_LONGJMP=0), but it is required by the code (either set SUPPORT_LONGJMP=1, or remove uses of it in the project)');
   }],
+  get setjmp__deps() {
+    return this.longjmp__deps;
+  },
+  // This is to print the correct error message when a program is built with
+  // SUPPORT_LONGJMP=1 but linked with SUPPORT_LONGJMP=0. When a program is
+  // built with SUPPORT_LONGJMP=1, the object file contains references of not
+  // longjmp but _emscripten_throw_longjmp, which is called from
+  // emscripten_longjmp.
+  get _emscripten_throw_longjmp__deps() {
+    return this.longjmp__deps;
+  },
   // will never be emitted, as the dep errors at compile time
   longjmp: function(env, value) {
     abort('longjmp not supported');
   },
+  setjmp: function(env, value) {
+    abort('setjmp not supported');
+  },
 #endif
-  // TODO: remove these aliases if/when the LLVM backend can stop emitting them
-  // (it emits them atm as they are generated by an IR pass, at at that time
-  // they each have a different signature - it is only at the wasm level that
-  // they become identical).
-  emscripten_longjmp__sig: 'vii',
-  emscripten_longjmp: 'longjmp',
 
   // ==========================================================================
   // sys/wait.h

src/library_signals.js

@@ -137,22 +137,6 @@ var funs = {
     setErrNo(ERRNO_CODES.EINTR);
     return -1;
   },
-#if SUPPORT_LONGJMP
-#if ASSERTIONS
-  siglongjmp__deps: ['longjmp'],
-  siglongjmp: function(env, value) {
-    // We cannot wrap the sigsetjmp, but I hope that
-    // in most cases siglongjmp will be called later.
-
-    // siglongjmp can be called very many times, so don't flood the stderr.
-    warnOnce("Calling longjmp() instead of siglongjmp()");
-    _longjmp(env, value);
-  },
-#else
-  siglongjmp__sig: 'vii',
-  siglongjmp: 'longjmp',
-#endif
-#endif
 
   sigpending: function(set) {
     {{{ makeSetValue('set', 0, 0, 'i32') }}};

system/lib/compiler-rt/emscripten_setjmp.c

@@ -16,6 +16,8 @@ typedef struct TableEntry {
 } TableEntry;
 
 extern void setTempRet0(uint32_t value);
+extern void setThrew(uintptr_t threw, int value);
+extern void _emscripten_throw_longjmp(); // defined in src/library.js
 
 TableEntry* saveSetjmp(uint32_t* env, uint32_t label, TableEntry* table, uint32_t size) {
   // Not particularly fast: slow table lookup of setjmpId to label. But setjmp
@@ -55,3 +57,8 @@ uint32_t testSetjmp(uint32_t id, TableEntry* table, uint32_t size) {
   }
   return 0;
 }
+
+void emscripten_longjmp(uintptr_t env, int val) {
+  setThrew(env, val);
+  _emscripten_throw_longjmp();
+}

system/lib/libc/musl/include/setjmp.h

@@ -22,10 +22,11 @@ typedef jmp_buf sigjmp_buf;
 /* XXX EMSCRIPTEN: No signals support, alias sigsetjmp and siglongjmp to their non-signals counterparts. */
 #if __EMSCRIPTEN__
 #define sigsetjmp(buf, x) setjmp((buf))
+#define siglongjmp(buf, val) longjmp(buf, val)
 #else
 int sigsetjmp (sigjmp_buf, int);
+_Noreturn int siglongjmp (sigjmp_buf, int);
 #endif
-_Noreturn void siglongjmp (sigjmp_buf, int);
 #endif
 
 #if defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) \

tests/test_other.py

@@ -10433,8 +10433,15 @@ def test_deps_info(self):
         cmd.append('-sUSE_WEBGPU')
       if function.startswith('__cxa_'):
         cmd.append('-fexceptions')
-      # Causes WebAssemblyLowerEmscriptenEHSjLj pass in llvm to crash
-      if function == 'setjmp':
+      # In WebAssemblyLowerEmscriptenEHSjLj pass in the LLVM backend, function
+      # calls that exist in the same function with setjmp are converted to some
+      # code sequence that includes emscripten_longjmp. emscripten_longjmp is
+      # included in deps_info.py because in non-LTO builds setjmp does not exist
+      # anymore in the object files. So the mere indirect reference of setjmp or
+      # emscripten_longjmp does not generate calls to its dependencies specified
+      # in deps_info.py. Also Emscripten EH has a known restriction that setjmp
+      # cannot be called or referenced indirectly anyway.
+      if function in ['emscripten_longjmp', 'setjmp']:
         continue
       print(shared.shlex_join(cmd))
       self.run_process(cmd)

tools/deps_info.py

@@ -93,7 +93,12 @@
   'emscripten_idb_load': ['malloc', 'free'],
   'emscripten_init_websocket_to_posix_socket_bridge': ['malloc', 'free'],
   'emscripten_log': ['strlen'],
-  'emscripten_longjmp': ['setThrew'],
+  # This list is the same as setjmp's dependencies. In non-LTO builds, setjmp
+  # does not exist in the object files; it is converted into a code sequence
+  # that includes several functions, one of which is emscripten_longjmp. This is
+  # a trick to include these dependencies for setjmp even when setjmp does not
+  # exist. Refer to setjmp's entry for more details.
+  'emscripten_longjmp': ['malloc', 'free', 'saveSetjmp', 'setThrew'],
   'emscripten_pc_get_file': ['emscripten_builtin_malloc', 'emscripten_builtin_free', 'emscripten_builtin_memalign', 'malloc', 'free'],
   'emscripten_pc_get_function': ['emscripten_builtin_malloc', 'emscripten_builtin_free', 'emscripten_builtin_memalign', 'malloc', 'free'],
   'emscripten_run_preload_plugins_data': ['malloc'],
@@ -167,7 +172,6 @@
   'gmtime_r':  ['malloc'],
   'localtime': ['_get_tzname', '_get_daylight', '_get_timezone', 'malloc'],
   'localtime_r': ['_get_tzname', '_get_daylight', '_get_timezone', 'malloc'],
-  'longjmp': ['setThrew'],
   'mktime': ['_get_tzname', '_get_daylight', '_get_timezone', 'malloc'],
   'mmap': ['memalign', 'memset', 'malloc'],
   'munmap': ['malloc', 'free'],
@@ -178,10 +182,15 @@
   'accept': ['htons'],
   'recvfrom': ['htons'],
   'send': ['ntohs'],
-  'setjmp': ['saveSetjmp'],
+  # In WebAssemblyLowerEmscriptenEHSjLj pass in the LLVM backend, function calls
+  # that exist in the same function with setjmp are converted to some code
+  # sequence that includes invokes, malloc, free, saveSetjmp, and
+  # emscripten_longjmp. setThrew is called from invokes, but there's no way to
+  # directly include invokes in deps_info.py, so we list it as a setjmp's
+  # dependency.
+  'setjmp': ['malloc', 'free', 'saveSetjmp', 'setThrew'],
   'setprotoent': ['malloc'],
   'setgroups': ['sysconf'],
-  'siglongjmp': ['setThrew'],
   'syslog': ['malloc', 'ntohs'],
   'timegm': ['_get_tzname', '_get_daylight', '_get_timezone', 'malloc'],
   'times': ['memset'],

tools/wasm2c/base.c

@@ -113,21 +113,10 @@ static jmp_buf setjmp_stack[MAX_SETJMP_STACK];
 
 static u32 next_setjmp = 0;
 
-// Declare an export that may be needed and may not be. For example if longjmp
-// is included then we need setThrew, but we must declare setThrew so that
-// the C compiler can build us without error if longjmp is not actually used.
-
-#define DECLARE_WEAK_EXPORT(ret, name, args) \
-__attribute__((weak)) \
-ret (*WASM_RT_ADD_PREFIX(name)) args = NULL;
-
-DECLARE_WEAK_EXPORT(void, Z_setThrewZ_vii, (u32, u32));
-
-IMPORT_IMPL(void, Z_envZ_emscripten_longjmpZ_vii, (u32 buf, u32 value), {
+IMPORT_IMPL(void, Z_envZ__emscripten_throw_longjmpZ_vv, (), {
   if (next_setjmp == 0) {
     abort_with_message("longjmp without setjmp");
   }
-  WASM_RT_ADD_PREFIX(Z_setThrewZ_vii)(buf, value ? value : 1);
   longjmp(setjmp_stack[next_setjmp - 1], 1);
 });