refactor emcc in preparation for compiling multiple files at once

2 files changed, 47 insertions, 14 deletions

diff --git a/emcc b/emcc
index 0edc7b05..dad157a9 100755
--- a/emcc
+++ b/emcc
@@ -267,6 +267,7 @@ elif use_compiler:
     elif newargs[i].startswith('--llvm-opts'):
       assert '=' not in newargs[i], 'Invalid llvm opts parameter (do not use "=")'
       llvm_opt_level = eval(newargs[i+1])
+      assert 0 <= llvm_opt_level <= 1, 'Only two levels of LLVM optimizations are supported so far, 0 (none) and 1 (safe)'
       newargs[i] = ''
       newargs[i+1] = ''
   newargs = [ arg for arg in newargs if arg is not '' ]
@@ -283,13 +284,25 @@ 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))
+
+  assert len(input_files) > 0, 'emcc: no input files specified'
+
   newargs += CC_ADDITIONAL_ARGS
 
-  if target is None:
-    # No explicit -o specified, so do the most natural thing, compile to .js
-    target = 'a.out.js'
+  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
+  if specified_target: assert len(input_files) == 1, 'If a target is specified, there should be exactly one input file'
 
-  target_basename = '.'.join(target.split('.')[:-1])
+  target_basename = unsuffixed_basename(target)
 
   if '-c' in newargs: # -c means do not link in gcc, and for us, the parallel is to not go all the way to JS, but stop at bitcode
     target = target_basename + '.bc'
@@ -313,24 +326,36 @@ elif use_compiler:
 
   ## Compile
 
-  # First, generate LLVM bitcode TODO: handle |emcc a.cpp b.cpp -c| which generate *two* bitcode files
-  newargs = newargs + ['-emit-llvm', '-c', '-o', target_basename + '.bc']
+  # 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()
 
   # Optimize, if asked to
   if llvm_opt_level > 0:
-    shared.Building.llvm_opt(target_basename + '.bc', 2, safe=llvm_opt_level < 2)
+    for input_file in input_files:
+      shared.Building.llvm_opt(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 == 'o':
-      shutil.move(target_basename + '.bc', target_basename + '.o')
+    if final_suffix == 'bc':
+      for input_file in input_files:
+        shutil.move(input_file + '.o', input_file + '.bc')
+
+    if specified_target:
+      shutil.move(input_files[0] + '.' + final_suffix, unsuffixed_basename(specified_target) + '.' + final_suffix)
+
     exit(0)
 
   ## Continue on to create JavaScript
 
+  # 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')
+  else:
+    shutil.move(input_files[0] + '.o', target_basename + '.bc')
+
   # Apply -s settings in newargs here (after -Ox, so they can override it)
 
   for change in settings_changes:
diff --git a/tests/runner.py b/tests/runner.py
index 33b3687b..a41a0d27 100644
--- a/tests/runner.py
+++ b/tests/runner.py
@@ -4904,16 +4904,16 @@ Options that are modified or new in %s include:
         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
+        assert os.path.exists('a.out.js'), output[1]
         self.assertContained('hello, world!', run_js('a.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']]:
-          target = args[1] if len(args) == 2 else 'a.out.bc'
+          target = args[1] if len(args) == 2 else 'hello_world.bc'
           clear()
           output = Popen([compiler, path_from_root('tests', 'hello_world' + suffix)] + args, stdout=PIPE, stderr=PIPE).communicate()
           assert len(output[0]) == 0, output[0]
-          assert os.path.exists(target), output
+          assert os.path.exists(target), 'Expected %s to exist since args are %s : %s' % (target, str(args), output)
           self.assertContained('hello, world!', self.run_llvm_interpreter([target]))
 
         # emcc src.cpp -o something.js [-Ox]. -O0 is the same as not specifying any optimization setting
@@ -4962,9 +4962,17 @@ Options that are modified or new in %s include:
           self.assertContained('hello, world!', run_js('a.out.js'))
           assert test(open('a.out.js').read()), text
 
+        ## Compiling two source files TODO: with -o
+        #clear()
+        #output = Popen([compiler, path_from_root('tests', 'twopart_main.cpp'), path_from_root('tests', 'twopart_side.cpp')],
+        #               stdout=PIPE, stderr=PIPE).communicate()
+        #assert len(output[0]) == 0, output[0]
+        #assert os.path.exists('a.out.js'), '\n'.join(output)
+        #self.assertContained('side got: hello from main, over', run_js('a.out.js'))
+        ##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
 
-      # emcc --llvm-opts=x .. ==> pick level of LLVM optimizations (default is 0, to be safe?)
-      # We can run safe opts at each emcc invocation, now we are a gcc replacement. Removes the entire need for llvm opts only at the end.
       # 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
       #   emcc a.cpp b.cpp -c => two .o files