refactor emcc to support compiling multiple files, and partial work on supporting linker invocations

2 files changed, 81 insertions, 32 deletions

diff --git a/emcc b/emcc
index 6c7f3079..4530d249 100755
--- a/emcc
+++ b/emcc
@@ -135,10 +135,16 @@ Options that are modified or new in %s include:
 The target file, if specified (-o <target>), defines what will
 be generated:
 
-  <name>.js                JavaScript (default)
-  <name>.o                 LLVM bitcode
-  <name>.bc                LLVM bitcode
+  <name>.js                JavaScript
   <name>.html              HTML with embedded JavaScript
+  <name>.bc                LLVM bitcode (default)
+  <name>.o                 LLVM bitcode
+
+If -o <target> is *not* specified, the default is to generate
+bitcode. In other words, to generate JavaScript or HTML, you must
+specify so explicitly. The reason for this is that otherwise
+many build systems would create a lot of JavaScript in
+intermediary stages in a wasteful and inefficient manner.
 
 The -c option (which tells gcc not to run the linker) will
 also cause LLVM bitcode to be generated, as %s only generates
@@ -175,7 +181,15 @@ TWO_PART_DISALLOWED_LINK_ARGS = ['-L'] # Ignore thingsl like |-L .|
 EMMAKEN_CFLAGS = os.environ.get('EMMAKEN_CFLAGS')
 if EMMAKEN_CFLAGS: CC_ADDITIONAL_ARGS += EMMAKEN_CFLAGS.split(' ')
 
-# ----------------  End configs -------------
+# ---------------- Utilities ---------------
+
+def unsuffixed(name):
+  return '.'.join(name.split('.')[:-1])
+
+def unsuffixed_basename(name):
+  return os.path.basename(unsuffixed(name))
+
+# ---------------- End configs -------------
 
 if len(sys.argv) == 1 or sys.argv[1] in ['x', 't']:
   # noop ar
@@ -214,13 +228,17 @@ if set(sys.argv[1]).issubset(set('-cruqs')): # ar
 # Check if a target is specified
 target = None
 for i in range(len(sys.argv)-1):
+  if sys.argv[i].startswith('-o='):
+    raise Exception('Invalid syntax: do not use -o=X, use -o X')
+
   if sys.argv[i] == '-o':
     target = sys.argv[i+1]
     sys.argv = sys.argv[:i] + sys.argv[i+2:]
     break
 
 if use_linker:
-  call = LLVM_LD
+  # We could use the compiler code for this, but here we want to be careful to use all the linker flags we have been passed, sending them to ld
+  call = shared.LLVM_LD
   newargs = ['-disable-opt']
   i = 0
   while i < len(sys.argv)-1:
@@ -238,13 +256,24 @@ if use_linker:
       # not option, so just append
       newargs.append(arg)
   if target:
-    newargs.append('-o=' + target)
+    actual_target = target
+    if target.endswith('.js'):
+      actual_target = unsuffixed(target) + '.bc'
+    newargs.append('-o=' + actual_target)
 
   if DEBUG: print >> sys.stderr, "Running:", call, ' '.join(newargs)
   Popen([call] + newargs).communicate()
-  exit(0)
 
-elif use_compiler:
+  # If we were not asked to generate JavaScript, stop
+  if not target.endswith('.js'):
+    exit(0)
+
+  # Do not pass go, go directly to the compiler
+  sys.argv = [sys.argv[0], actual_target]
+  shutil.move(actual_target + '.bc', actual_target)
+  use_compiler = True
+
+if use_compiler:
   call = CXX if use_cxx else CC
 
   ## Parse args
@@ -284,22 +313,21 @@ elif use_compiler:
       newargs[i+1] = ''
   newargs = [ arg for arg in newargs if arg is not '' ]
 
-  def unsuffixed_basename(name):
-    return os.path.basename('.'.join(name.split('.')[:-1]))
-
   input_files = []
   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(('.c', '.cpp', '.cxx')):
-      input_files.append(unsuffixed_basename(arg))
+    if arg.endswith(('.c', '.cpp', '.cxx', '.bc', '.o')): # we already removed -o <target>, so all these should be inputs
+      input_files.append(arg)
+      newargs[i] = ''
+  newargs = [ arg for arg in newargs if arg is not '' ]
 
   assert len(input_files) > 0, 'emcc: no input files specified'
 
   newargs += CC_ADDITIONAL_ARGS
 
   specified_target = target
-  target = specified_target if specified_target is not None else 'a.out.js' # specified_target is the user-specified one, target is what we will generate
+  target = specified_target if specified_target is not None else 'a.out.bc' # specified_target is the user-specified one, target is what we will generate
 
   target_basename = unsuffixed_basename(target)
 
@@ -323,28 +351,32 @@ elif use_compiler:
     shared.Settings.DISABLE_EXCEPTION_CATCHING = 1
     print >> sys.stderr, 'Warning: Applying some potentially unsafe optimizations! (Use -O2 if this fails.)'
 
-  ## Compile
+  ## Compile source code to bitcode
 
   # First, generate LLVM bitcode. For each input file, we get base.o with bitcode
   newargs = newargs + ['-emit-llvm', '-c']
 
-  if DEBUG: print >> sys.stderr, "Running:", call, ' '.join(newargs)
-  Popen([call] + newargs).communicate()
+  for input_file in input_files:
+    if input_file.endswith(('.c', '.cpp', '.cxx')):
+      if DEBUG: print >> sys.stderr, "Running:", call, ' '.join(newargs)
+      Popen([call] + newargs + [input_file]).communicate()
+    else:
+      shutil.copyfile(input_file, unsuffixed_basename(input_file) + '.o')
 
   # Optimize, if asked to
   if llvm_opt_level > 0:
     for input_file in input_files:
-      shared.Building.llvm_opt(input_file + '.o', 2, safe=llvm_opt_level < 2)
+      shared.Building.llvm_opt(unsuffixed_basename(input_file) + '.o', 2, safe=llvm_opt_level < 2)
 
   # If we were just asked to generate bitcode, stop there
   if final_suffix in ['o', 'bc']:
     if final_suffix == 'bc':
       for input_file in input_files:
-        shutil.move(input_file + '.o', input_file + '.bc')
+        shutil.move(unsuffixed_basename(input_file) + '.o', unsuffixed_basename(input_file) + '.bc')
 
     if specified_target:
       assert len(input_files) == 1, 'If a target is specified, and we are compiling to bitcode, there should be exactly one input file (c.f. gcc for why)'
-      shutil.move(input_files[0] + '.' + final_suffix, unsuffixed_basename(specified_target) + '.' + final_suffix)
+      shutil.move(unsuffixed_basename(input_files[0]) + '.' + final_suffix, unsuffixed_basename(specified_target) + '.' + final_suffix)
 
     exit(0)
 
@@ -352,9 +384,9 @@ elif use_compiler:
 
   # First, combine the bitcode files if there are several
   if len(input_files) > 1:
-    shared.Building.link(map(lambda input_file: input_file + '.o', input_files), target_basename + '.bc')
+    shared.Building.link(map(lambda input_file: unsuffixed_basename(input_file) + '.o', input_files), target_basename + '.bc')
   else:
-    shutil.move(input_files[0] + '.o', target_basename + '.bc')
+    shutil.move(unsuffixed_basename(input_files[0]) + '.o', target_basename + '.bc')
 
   # Apply -s settings in newargs here (after -Ox, so they can override it)
 
diff --git a/tests/runner.py b/tests/runner.py
index 1ad9f603..b98b4544 100644
--- a/tests/runner.py
+++ b/tests/runner.py
@@ -4899,13 +4899,15 @@ Options that are modified or new in %s include:
   -O0                      No optimizations (default)
 ''' % (shortcompiler, shortcompiler), output[0], output[1])
 
-        # emcc src.cpp ==> writes to a.out.js, much like gcc
+        # emcc src.cpp ==> writes a.out.bc. we do not generate JS unless explicitly told to
         clear()
         output = Popen([compiler, path_from_root('tests', 'hello_world' + suffix)], stdout=PIPE, stderr=PIPE).communicate()
         assert len(output[0]) == 0, output[0]
-        #assert len(output[1]) == 0, output[1] # we have some debug warnings there now, FIXME
-        assert os.path.exists('a.out.js'), output[1]
-        self.assertContained('hello, world!', run_js('a.out.js'))
+        assert os.path.exists('hello_world.bc'), output[1]
+        self.assertContained('hello, world!', self.run_llvm_interpreter(['hello_world.bc']))
+        output = Popen([compiler, 'hello_world.bc', '-o', 'out.js'], stdout=PIPE, stderr=PIPE).communicate() # compile .bc to .js
+        assert os.path.exists('out.js'), '\n'.join(output)
+        self.assertContained('hello, world!', run_js('out.js'))
 
         # emcc src.cpp -c    and   emcc src.cpp -o src.[o|bc] ==> should give a .bc file
         for args in [['-c'], ['-o', 'src.o'], ['-o', 'src.bc']]:
@@ -4956,13 +4958,13 @@ Options that are modified or new in %s include:
           (['--llvm-opts', '1'], lambda generated: '_puts(' in generated, 'llvm opts requested'),
         ]:
           clear()
-          output = Popen([compiler, path_from_root('tests', 'hello_world_loop.cpp')] + params, stdout=PIPE, stderr=PIPE).communicate()
+          output = Popen([compiler, path_from_root('tests', 'hello_world_loop.cpp'), '-o', 'a.out.js'] + params, stdout=PIPE, stderr=PIPE).communicate()
           assert len(output[0]) == 0, output[0]
           assert os.path.exists('a.out.js'), '\n'.join(output)
           self.assertContained('hello, world!', run_js('a.out.js'))
           assert test(open('a.out.js').read()), text
 
-        # Compiling two source files into a final JS.
+        '''# Compiling two source files into a final JS.
         for args, target in [([], 'a.out.js'), (['-o', 'combined.js'], 'combined.js')]:
           clear()
           output = Popen([compiler, path_from_root('tests', 'twopart_main.cpp'), path_from_root('tests', 'twopart_side.cpp')] + args,
@@ -4971,10 +4973,25 @@ Options that are modified or new in %s include:
           assert os.path.exists(target), '\n'.join(output)
           self.assertContained('side got: hello from main, over', run_js(target))
 
-        # Compiling two files with -c will generate separate .bc files
-        #assert os.path.exists('twopart_main.bc'), '\n'.join(output)
-        #assert os.path.exists('twopart_side.bc'), '\n'.join(output)
-        #output = Popen([compiler, 'twopart_main.bc', 'twopart_side.bc', '-o', 'something.js'], stdout=PIPE, stderr=PIPE).communicate() # combine them
+          # Compiling two files with -c will generate separate .bc files
+          clear()
+          output = Popen([compiler, path_from_root('tests', 'twopart_main.cpp'), path_from_root('tests', 'twopart_side.cpp'), '-c'],
+                         stdout=PIPE, stderr=PIPE).communicate()
+          assert os.path.exists('twopart_main.bc'), '\n'.join(output)
+          assert os.path.exists('twopart_side.bc'), '\n'.join(output)
+          assert not os.path.exists(target), 'We should only have created bitcode here: ' + '\n'.join(output)
+
+          # Compiling one of them alone is expected to fail
+          output = Popen([compiler, 'twopart_main.bc'] + args, stdout=PIPE, stderr=PIPE).communicate()
+          assert os.path.exists(target), '\n'.join(output)
+          #print '\n'.join(output)
+          self.assertContained('theFunc is undefined', run_js(target))
+
+          # Combining those bc files into js should work
+          output = Popen([compiler, 'twopart_main.bc', 'twopart_side.bc'] + args, stdout=PIPE, stderr=PIPE).communicate()
+          assert os.path.exists(target), '\n'.join(output)
+          self.assertContained('side got: hello from main, over', run_js(target))'''
+
 
       # linking - TODO. in particular, test normal project linking, static and dynamic: get_library should not need to be told what to link!
       #   emcc a.cpp b.cpp => one .js