Use llvm-ar instead of llvm-link in emar

This makes the semantics of emar to be the same as the semantics of
system ar.

3 files changed, 29 insertions, 7 deletions

diff --git a/emar b/emar
index 49d995bd..e64cd8b9 100755
--- a/emar
+++ b/emar
@@ -12,11 +12,11 @@ from tools import shared
 
 DEBUG = os.environ.get('EMCC_DEBUG')
 
-newargs = [shared.LLVM_LINK] + sys.argv[3:] + ['-o='+sys.argv[2]]
+newargs = [shared.LLVM_AR] + sys.argv[1:]
 
 if DEBUG:
   print >> sys.stderr, 'emar:', sys.argv, '  ==>  ', newargs
 
 if len(newargs) > 2:
-  os.execvp(shared.LLVM_LINK, newargs)
+  os.execvp(shared.LLVM_AR, newargs)
 
diff --git a/emcc b/emcc
index b8e2489c..fa18968a 100755
--- a/emcc
+++ b/emcc
@@ -239,7 +239,7 @@ if EMMAKEN_CFLAGS: CC_ADDITIONAL_ARGS += EMMAKEN_CFLAGS.split(' ')
 # ---------------- Utilities ---------------
 
 SOURCE_SUFFIXES = ('.c', '.cpp', '.cxx', '.cc')
-BITCODE_SUFFIXES = ('.bc', '.o', '.a', '.dylib', '.so', '.dll')
+BITCODE_SUFFIXES = ('.bc', '.o', '.dylib', '.so', '.dll')
 ASSEMBLY_SUFFIXES = ('.ll',)
 
 def unsuffixed(name):
@@ -377,7 +377,7 @@ try:
   for i in range(len(newargs)): # find input files XXX this a simple heuristic. we should really analyze based on a full understanding of gcc params,
                                 # right now we just assume that what is left contains no more |-x OPT| things
     arg = newargs[i]
-    if arg.endswith(SOURCE_SUFFIXES + BITCODE_SUFFIXES + ASSEMBLY_SUFFIXES): # we already removed -o <target>, so all these should be inputs
+    if arg.endswith(SOURCE_SUFFIXES + BITCODE_SUFFIXES + ASSEMBLY_SUFFIXES) or shared.Building.is_ar(arg): # we already removed -o <target>, so all these should be inputs
       newargs[i] = ''
       if os.path.exists(arg):
         if arg.endswith(SOURCE_SUFFIXES):
@@ -440,6 +440,9 @@ try:
       if input_file.endswith(BITCODE_SUFFIXES):
         if DEBUG: print >> sys.stderr, 'emcc: copying bitcode file: ', input_file
         shutil.copyfile(input_file, in_temp(unsuffixed_basename(input_file) + '.o'))
+      elif shared.Building.is_ar(input_file):
+        if DEBUG: print >> sys.stderr, 'emcc: copying ar file: ', input_file
+        shutil.copyfile(input_file, in_temp(os.path.basename(input_file)))
       else: #.ll
         if not LEAVE_INPUTS_RAW:
           # Note that by assembling the .ll file, then disassembling it later, we will
diff --git a/tools/shared.py b/tools/shared.py
index 45400581..615f86e8 100644
--- a/tools/shared.py
+++ b/tools/shared.py
@@ -90,6 +90,7 @@ CLANG_CPP=os.path.expanduser(os.path.join(LLVM_ROOT, 'clang++'))
 CLANG=CLANG_CPP
 LLVM_LINK=os.path.join(LLVM_ROOT, 'llvm-link')
 LLVM_LD=os.path.join(LLVM_ROOT, 'llvm-ld')
+LLVM_AR=os.path.join(LLVM_ROOT, 'llvm-ar')
 LLVM_OPT=os.path.expanduser(os.path.join(LLVM_ROOT, 'opt'))
 LLVM_AS=os.path.expanduser(os.path.join(LLVM_ROOT, 'llvm-as'))
 LLVM_DIS=os.path.expanduser(os.path.join(LLVM_ROOT, 'llvm-dis'))
@@ -726,6 +727,17 @@ set(CMAKE_FIND_ROOT_PATH_MODE_PACKAGE ONLY)'''.replace('$EMSCRIPTEN_ROOT', path_
     return filename + '.cc.js'
 
   @staticmethod
+  def is_ar(filename):
+    try:
+      b = open(filename, 'r').read(8)
+      return b[0] == '!' and b[1] == '<' and \
+             b[2] == 'a' and b[3] == 'r' and \
+             b[4] == 'c' and b[5] == 'h' and \
+             b[6] == '>' and ord(b[7]) == 10
+    except:
+      return False
+
+  @staticmethod
   def is_bitcode(filename):
     # checks if a file contains LLVM bitcode
     # if the file doesn't exist or doesn't have valid symbols, it isn't bitcode
@@ -733,9 +745,13 @@ set(CMAKE_FIND_ROOT_PATH_MODE_PACKAGE ONLY)'''.replace('$EMSCRIPTEN_ROOT', path_
       defs = Building.llvm_nm(filename, stderr=PIPE)
       # If no symbols found, it might just be an empty bitcode file, try to dis it
       if len(defs.defs) + len(defs.undefs) + len(defs.commons) == 0:
-        test_ll = os.path.join(EMSCRIPTEN_TEMP_DIR, 'test.ll')
-        Building.llvm_dis(filename, test_ll)
-        assert os.path.exists(test_ll)
+        # llvm-nm 3.0 has a bug when reading symbols from ar files
+        # so try to see if we're dealing with an ar file, in which
+        # case we should try to dis it.
+        if not Building.is_ar(filename):
+          test_ll = os.path.join(EMSCRIPTEN_TEMP_DIR, 'test.ll')
+          Building.llvm_dis(filename, test_ll)
+          assert os.path.exists(test_ll)
     except:
       return False
 
@@ -743,6 +759,9 @@ set(CMAKE_FIND_ROOT_PATH_MODE_PACKAGE ONLY)'''.replace('$EMSCRIPTEN_ROOT', path_
     b = open(filename, 'r').read(4)
     if b[0] == 'B' and b[1] == 'C':
       return True
+    # look for ar signature
+    elif Building.is_ar(filename):
+      return True
     # on OS X, there is a 20-byte prefix
     elif ord(b[0]) == 222 and ord(b[1]) == 192 and ord(b[2]) == 23 and ord(b[3]) == 11:
       b = open(filename, 'r').read(24)