api: autogenerate api trace print/endian

In addition to the external vppapitrace tool, VPP itself supports dumping of API trace files.
In two formats, "custom-dump" and "dump". "dump" gives a human friendly list,
and "custom-dump" is meant to give a list of commands that can be fed to VAT.
This patch only deals with "dump".
Prior to this fix, auto-generation was only done for the basic types.
This fix adds support for any type, including lists, and supports pretty-printing
of enums, strings, IP addresses, MAC addresses and so on.

Usage: api trace dump <api-trace-file>

For example

Change-Id: I4e485680e6dcfce7489299ae6cf31d835071ac40

---------- trace 48 -----------
vl_api_sw_interface_set_flags_t:
  _vl_msg_id: 75
  client_index: 0
  context: 10
  sw_if_index: 1
  flags: IF_STATUS_API_FLAG_ADMIN_UP
---------- trace 49 -----------
vl_api_sw_interface_add_del_address_t:
  _vl_msg_id: 88
  client_index: 0
  context: 11
  sw_if_index: 1
  is_add: 1
  del_all: 0
  prefix: 172.16.1.1/24
---------- trace 51 -----------
vl_api_cli_inband_t:
  _vl_msg_id: 819
  client_index: 0
  context: 13
  cmd: packet-generator capture pg0 pcap /tmp/vpp-unittest-TestMAP-YhcmDX/pg0_out.pcap disable
---------- trace 58 -----------
vl_api_ip_neighbor_add_del_t:
  _vl_msg_id: 199
  client_index: 0
  context: 20
  is_add: 1
  neighbor:
    sw_if_index: 2
    flags: IP_API_NEIGHBOR_FLAG_NONE
    mac_address: 0202.0000.ff02
    ip_address: fd01:2::2

Signed-off-by: Ole Troan <ot@cisco.com>
Change-Id: I5556d06008de2762e7c2d35a8b0963ae670b3db1
Type: fix
Signed-off-by: Ole Troan <ot@cisco.com>
Signed-off-by: Ole Troan <ot@cisco.com>
Signed-off-by: Ole Troan <ot@cisco.com>
Signed-off-by: Ole Troan <ot@cisco.com>
Signed-off-by: Ole Troan <ot@cisco.com>
Signed-off-by: Ole Troan <ot@cisco.com>
Signed-off-by: Ole Troan <ot@cisco.com>
Signed-off-by: Ole Troan <ot@cisco.com>

4 files changed, 434 insertions, 86 deletions

diff --git a/src/tools/vppapigen/vppapigen.py b/src/tools/vppapigen/vppapigen.py
index 2a939647c67..fa7e47afb73 100755
--- a/src/tools/vppapigen/vppapigen.py
+++ b/src/tools/vppapigen/vppapigen.py
@@ -8,6 +8,7 @@ import keyword
 import logging
 import binascii
 import os
+import sys
 
 log = logging.getLogger('vppapigen')
 
@@ -78,6 +79,16 @@ class VPPAPILexer(object):
 
     t_ignore_LINE_COMMENT = '//.*'
 
+    def t_FALSE(self, t):
+        r'false'
+        t.value = False
+        return t
+
+    def t_TRUE(self, t):
+        r'false'
+        t.value = True
+        return t
+
     def t_NUM(self, t):
         r'0[xX][0-9a-fA-F]+|-?\d+\.?\d*'
         base = 16 if t.value.startswith('0x') else 10
@@ -150,6 +161,14 @@ class Typedef():
                 self.manual_print = True
             elif f == 'manual_endian':
                 self.manual_endian = True
+        for b in block:
+            # Tag length field of a VLA
+            if isinstance(b, Array):
+                if b.lengthfield:
+                    for b2 in block:
+                        if b2.fieldname == b.lengthfield:
+                            b2.vla_len = True
+
         global_type_add(name, self)
 
         self.vla = False
@@ -176,10 +195,18 @@ class Typedef():
 
 
 class Using():
-    def __init__(self, name, alias):
+    def __init__(self, name, flags, alias):
         self.name = name
         self.vla = False
 
+        self.manual_print = False
+        self.manual_endian = False
+        for f in flags:
+            if f == 'manual_print':
+                self.manual_print = True
+            elif f == 'manual_endian':
+                self.manual_endian = True
+
         if isinstance(alias, Array):
             a = {'type': alias.fieldtype,
                  'length': alias.length}
@@ -194,11 +221,20 @@ class Using():
 
 
 class Union():
-    def __init__(self, name, block):
+    def __init__(self, name, flags, block):
         self.type = 'Union'
         self.manual_print = False
         self.manual_endian = False
         self.name = name
+
+        self.manual_print = False
+        self.manual_endian = False
+        for f in flags:
+            if f == 'manual_print':
+                self.manual_print = True
+            elif f == 'manual_endian':
+                self.manual_endian = True
+
         self.block = block
         self.crc = str(block).encode()
         global_type_add(name, self)
@@ -250,6 +286,12 @@ class Define():
                         'VLA field "{}" must be the last '
                         'field in message "{}"'
                         .format(b.fieldname, name))
+            # Tag length field of a VLA
+            if isinstance(b, Array):
+                if b.lengthfield:
+                    for b2 in block:
+                        if b2.fieldname == b.lengthfield:
+                            b2.vla_len = True
 
     def __repr__(self):
         return self.name + str(self.flags) + str(self.block)
@@ -298,8 +340,10 @@ class Import():
 
 
 class Option():
-    def __init__(self, option):
+    def __init__(self, option, value):
+        self.type = 'Option'
         self.option = option
+        self.value = value
         self.crc = str(option).encode()
 
     def __repr__(self):
@@ -524,9 +568,17 @@ class VPPAPIParser(object):
         '''typedef : TYPEDEF ID '{' block_statements_opt '}' ';' '''
         p[0] = Typedef(p[2], [], p[4])
 
+    def p_typedef_flist(self, p):
+        '''typedef : flist TYPEDEF ID '{' block_statements_opt '}' ';' '''
+        p[0] = Typedef(p[3], p[1], p[5])
+
     def p_typedef_alias(self, p):
         '''typedef : TYPEDEF declaration '''
-        p[0] = Using(p[2].fieldname, p[2])
+        p[0] = Using(p[2].fieldname, [], p[2])
+
+    def p_typedef_alias_flist(self, p):
+        '''typedef : flist TYPEDEF declaration '''
+        p[0] = Using(p[3].fieldname, p[1], p[3])
 
     def p_block_statements_opt(self, p):
         '''block_statements_opt : block_statements '''
@@ -620,7 +672,7 @@ class VPPAPIParser(object):
 
     def p_option(self, p):
         '''option : OPTION ID '=' assignee ';' '''
-        p[0] = Option([p[1], p[2], p[4]])
+        p[0] = Option(p[2], p[4])
 
     def p_assignee(self, p):
         '''assignee : NUM
@@ -653,7 +705,11 @@ class VPPAPIParser(object):
 
     def p_union(self, p):
         '''union : UNION ID '{' block_statements_opt '}' ';' '''
-        p[0] = Union(p[2], p[4])
+        p[0] = Union(p[2], [], p[4])
+
+    def p_union_flist(self, p):
+        '''union : flist UNION ID '{' block_statements_opt '}' ';' '''
+        p[0] = Union(p[3], p[1], p[5])
 
     # Error rule for syntax errors
     def p_error(self, p):
@@ -716,7 +772,7 @@ class VPPAPI(object):
                   isinstance(o, Union)):
                 s['types'].append(o)
             elif isinstance(o, Using):
-                s['Alias'][o.name] = o.alias
+                s['Alias'][o.name] = o
             else:
                 if tname not in s:
                     raise ValueError('Unknown class type: {} {}'
@@ -802,6 +858,7 @@ class VPPAPI(object):
                                   isinstance(o, Using)):
                 continue
             if isinstance(o, Import):
+                result.append(o)
                 self.process_imports(o.result, True, result)
             else:
                 result.append(o)
@@ -894,8 +951,12 @@ def main():
 
     # Build a list of objects. Hash of lists.
     result = []
-    parser.process_imports(parsed_objects, False, result)
-    s = parser.process(result)
+
+    if args.output_module == 'C':
+        s = parser.process(parsed_objects)
+    else:
+        parser.process_imports(parsed_objects, False, result)
+        s = parser.process(result)
 
     # Add msg_id field
     s['Define'] = add_msg_id(s['Define'])
diff --git a/src/tools/vppapigen/vppapigen_c.py b/src/tools/vppapigen/vppapigen_c.py
index c1bc11d4a12..d6cdda8baac 100644
--- a/src/tools/vppapigen/vppapigen_c.py
+++ b/src/tools/vppapigen/vppapigen_c.py
@@ -2,6 +2,8 @@
 import datetime
 import os
 import time
+import sys
+from io import StringIO
 
 datestring = datetime.datetime.utcfromtimestamp(
     int(os.environ.get('SOURCE_DATE_EPOCH', time.time())))
@@ -86,16 +88,6 @@ def msg_name_crc_list(s, suffix):
     return output
 
 
-def duplicate_wrapper_head(name):
-    s = "#ifndef _vl_api_defined_%s\n" % name
-    s += "#define _vl_api_defined_%s\n" % name
-    return s
-
-
-def duplicate_wrapper_tail():
-    return '#endif\n\n'
-
-
 def api2c(fieldtype):
     mappingtable = {'string': 'vl_api_string_t', }
     if fieldtype in mappingtable:
@@ -111,22 +103,20 @@ def typedefs(objs, aliases, filename):
 /****** Typedefs ******/
 
 #ifdef vl_typedefs
-#ifndef included_{module}
-#define included_{module}
+#ifndef included_{module}_typedef
+#define included_{module}_typedef
 '''
     output = output.format(module=name)
 
     for k, v in aliases.items():
-        output += duplicate_wrapper_head(k)
-        if 'length' in v:
-            output +=  'typedef %s vl_api_%s_t[%s];\n' % (v['type'], k, v['length'])
+        if 'length' in v.alias:
+            output += ('typedef %s vl_api_%s_t[%s];\n'
+                       % (v.alias['type'], k, v.alias['length']))
         else:
-            output += 'typedef %s vl_api_%s_t;\n' % (v['type'], k)
-        output += duplicate_wrapper_tail()
+            output += 'typedef %s vl_api_%s_t;\n' % (v.alias['type'], k)
 
     for o in objs:
         tname = o.__class__.__name__
-        output += duplicate_wrapper_head(o.name)
         if tname == 'Enum':
             if o.enumtype == 'u32':
                 output += "typedef enum {\n"
@@ -140,33 +130,43 @@ def typedefs(objs, aliases, filename):
                 size1 = 'sizeof(vl_api_%s_t)' % o.name
                 size2 = 'sizeof(%s)' % o.enumtype
                 err_str = 'size of API enum %s is wrong' % o.name
-                output += 'STATIC_ASSERT(%s == %s, "%s");\n' % (size1, size2, err_str)
+                output += ('STATIC_ASSERT(%s == %s, "%s");\n'
+                           % (size1, size2, err_str))
         else:
             if tname == 'Union':
                 output += "typedef VL_API_PACKED(union _vl_api_%s {\n" % o.name
             else:
-                output += "typedef VL_API_PACKED(struct _vl_api_%s {\n" % o.name
+                output += ("typedef VL_API_PACKED(struct _vl_api_%s {\n"
+                           % o.name)
             for b in o.block:
+                if b.type == 'Option':
+                    continue
                 if b.type == 'Field':
-                    output += "    %s %s;\n" % (api2c(b.fieldtype), b.fieldname)
+                    output += "    %s %s;\n" % (api2c(b.fieldtype),
+                                                b.fieldname)
                 elif b.type == 'Array':
                     if b.lengthfield:
-                        output += "    %s %s[0];\n" % (api2c(b.fieldtype), b.fieldname)
+                        output += "    %s %s[0];\n" % (api2c(b.fieldtype),
+                                                       b.fieldname)
                     else:
                         # Fixed length strings decay to nul terminated u8
                         if b.fieldtype == 'string':
                             if b.modern_vla:
-                                output += '    {} {};\n'.format(api2c(b.fieldtype), b.fieldname)
+                                output += ('    {} {};\n'
+                                           .format(api2c(b.fieldtype),
+                                                   b.fieldname))
                             else:
-                                output += '    u8 {}[{}];\n'.format(b.fieldname, b.length)
+                                output += ('    u8 {}[{}];\n'
+                                           .format(b.fieldname, b.length))
                         else:
-                            output += "    %s %s[%s];\n" % (api2c(b.fieldtype), b.fieldname,
-                                                            b.length)
+                            output += ("    %s %s[%s];\n" %
+                                       (api2c(b.fieldtype), b.fieldname,
+                                        b.length))
                 else:
-                    raise ValueError("Error in processing array type %s" % b)
+                    raise ValueError("Error in processing type {} for {}"
+                                     .format(b, o.name))
 
             output += '}) vl_api_%s_t;\n' % o.name
-        output += duplicate_wrapper_tail()
 
     output += "\n#endif"
     output += "\n#endif\n\n"
@@ -175,6 +175,7 @@ def typedefs(objs, aliases, filename):
 
 
 format_strings = {'u8': '%u',
+                  'bool': '%u',
                   'i8': '%d',
                   'u16': '%u',
                   'i16': '%d',
@@ -182,13 +183,136 @@ format_strings = {'u8': '%u',
                   'i32': '%ld',
                   'u64': '%llu',
                   'i64': '%llu',
-                  'f64': '%.2f', }
+                  'f64': '%.2f'}
+
+noprint_fields = {'_vl_msg_id': None,
+                  'client_index': None,
+                  'context': None}
+
+
+class Printfun():
+    _dispatch = {}
+
+    def __init__(self, stream):
+        self.stream = stream
+
+    def print_string(self, o, stream):
+        write = stream.write
+        if o.modern_vla:
+            write('    if (vl_api_string_len(&a->{f}) > 0) {{\n'
+                  .format(f=o.fieldname))
+            write('        s = format(s, "\\n%U{f}: %.*s", '
+                  'format_white_space, indent, '
+                  'vl_api_string_len(&a->{f}) - 1, '
+                  'vl_api_from_api_string(&a->{f}));\n'.format(f=o.fieldname))
+            write('    } else {\n')
+            write('        s = format(s, "\\n%U{f}:", '
+                  'format_white_space, indent);\n'.format(f=o.fieldname))
+            write('    }\n')
+        else:
+            write('    s = format(s, "\\n%U{f}: %s", '
+                  'format_white_space, indent, a->{f});\n'
+                  .format(f=o.fieldname))
+
+    def print_field(self, o, stream):
+        write = stream.write
+        if o.fieldname in noprint_fields:
+            return
+        if o.fieldtype in format_strings:
+            f = format_strings[o.fieldtype]
+            write('    s = format(s, "\\n%U{n}: {f}", '
+                  'format_white_space, indent, a->{n});\n'
+                  .format(n=o.fieldname, f=f))
+        else:
+            write('    s = format(s, "\\n%U{n}: %U", '
+                  'format_white_space, indent, '
+                  'format_{t}, &a->{n}, indent);\n'
+                  .format(n=o.fieldname, t=o.fieldtype))
 
+    _dispatch['Field'] = print_field
 
-def printfun(objs):
-    output = '''\
+    def print_array(self, o, stream):
+        write = stream.write
+
+        forloop = '''\
+    for (i = 0; i < {lfield}; i++) {{
+        s = format(s, "\\n%U{n}: %U",
+                   format_white_space, indent, format_{t}, &a->{n}[i], indent);
+    }}
+'''
+
+        forloop_format = '''\
+    for (i = 0; i < {lfield}; i++) {{
+        s = format(s, "\\n%U{n}: {t}",
+                   format_white_space, indent, a->{n}[i]);
+    }}
+'''
+
+        if o.fieldtype == 'string':
+            return self.print_string(o, stream)
+
+        if o.fieldtype == 'u8':
+            if o.lengthfield:
+                write('    s = format(s, "\\n%U{n}: %U", format_white_space, '
+                      'indent, format_hex_bytes, a->{n}, a->{lfield});\n'
+                      .format(n=o.fieldname, lfield=o.lengthfield))
+            else:
+                write('    s = format(s, "\\n%U{n}: %U", format_white_space, '
+                      'indent, format_hex_bytes, a, {lfield});\n'
+                      .format(n=o.fieldname, lfield=o.length))
+            return
+
+        lfield = 'a->' + o.lengthfield if o.lengthfield else o.length
+        if o.fieldtype in format_strings:
+            write(forloop_format.format(lfield=lfield,
+                                        t=format_strings[o.fieldtype],
+                                        n=o.fieldname))
+        else:
+            write(forloop.format(lfield=lfield, t=o.fieldtype, n=o.fieldname))
+
+    _dispatch['Array'] = print_array
+
+    def print_alias(self, k, v, stream):
+        write = stream.write
+        if ('length' in v.alias and v.alias['length'] and
+                v.alias['type'] == 'u8'):
+            write('    return format(s, "%U", format_hex_bytes, a, {});\n'
+                  .format(v.alias['length']))
+        elif v.alias['type'] in format_strings:
+            write('    return format(s, "{}", *a);\n'
+                  .format(format_strings[v.alias['type']]))
+        else:
+            write('    return format(s, "{} (print not implemented)"'
+                  .format(k))
+
+    def print_enum(self, o, stream):
+        write = stream.write
+        write("    switch(*a) {\n")
+        for b in o:
+            write("    case %s:\n" % b[1])
+            write('        return format(s, "{}");\n'.format(b[0]))
+        write('    }\n')
+
+    _dispatch['Enum'] = print_enum
+
+    def print_obj(self, o, stream):
+        write = stream.write
+
+        if o.type in self._dispatch:
+            self._dispatch[o.type](self, o, stream)
+        else:
+            write('    s = format(s, "\\n{} {} {} (print not implemented");\n'
+                  .format(o.type, o.fieldtype, o.fieldname))
+
+
+def printfun(objs, stream, modulename):
+    write = stream.write
+
+    h = '''\
 /****** Print functions *****/
 #ifdef vl_printfun
+#ifndef included_{module}_printfun
+#define included_{module}_printfun
 
 #ifdef LP64
 #define _uword_fmt \"%lld\"
@@ -199,32 +323,103 @@ def printfun(objs):
 #endif
 
 '''
+
+    signature = '''\
+static inline void *vl_api_{name}_t_print (vl_api_{name}_t *a, void *handle)
+{{
+    u8 *s = 0;
+    u32 indent __attribute__((unused)) = 2;
+    int i __attribute__((unused));
+'''
+
+    h = h.format(module=modulename)
+    write(h)
+
+    pp = Printfun(stream)
+    for t in objs:
+        if t.manual_print:
+            write("/***** manual: vl_api_%s_t_print  *****/\n\n" % t.name)
+            continue
+        write(signature.format(name=t.name))
+        write('    /* Message definition: vl_api_{}_t: */\n'.format(t.name))
+        write("    s = format(s, \"vl_api_%s_t:\");\n" % t.name)
+        for o in t.block:
+            pp.print_obj(o, stream)
+        write('    vec_add1(s, 0);\n')
+        write('    vl_print (handle, (char *)s);\n')
+        write('    vec_free (s);\n')
+        write('    return handle;\n')
+        write('}\n\n')
+
+    write("\n#endif")
+    write("\n#endif /* vl_printfun */\n")
+
+    return ''
+
+
+def printfun_types(objs, aliases, stream, modulename):
+    write = stream.write
+    pp = Printfun(stream)
+
+    h = '''\
+/****** Print functions *****/
+#ifdef vl_printfun
+#ifndef included_{module}_printfun_types
+#define included_{module}_printfun_types
+
+'''
+    h = h.format(module=modulename)
+    write(h)
+
+    signature = '''\
+static inline u8 *format_vl_api_{name}_t (u8 *s, va_list * args)
+{{
+    vl_api_{name}_t *a = va_arg (*args, vl_api_{name}_t *);
+    u32 indent __attribute__((unused)) = va_arg (*args, u32);
+    int i __attribute__((unused));
+    indent += 2;
+'''
+
+    for k, v in aliases.items():
+        if v.manual_print:
+            write("/***** manual: vl_api_%s_t_print  *****/\n\n" % k)
+            continue
+
+        write(signature.format(name=k))
+        pp.print_alias(k, v, stream)
+        write('}\n\n')
+
     for t in objs:
         if t.__class__.__name__ == 'Enum':
+            write(signature.format(name=t.name))
+            pp.print_enum(t.block, stream)
+            write('    return s;\n')
+            write('}\n\n')
             continue
+
         if t.manual_print:
-            output += "/***** manual: vl_api_%s_t_print  *****/\n\n" % t.name
+            write("/***** manual: vl_api_%s_t_print  *****/\n\n" % t.name)
             continue
-        output += duplicate_wrapper_head(t.name + '_t_print')
-        output += "static inline void *vl_api_%s_t_print (vl_api_%s_t *a," % \
-                  (t.name, t.name)
-        output += "void *handle)\n{\n"
-        output += "    vl_print(handle, \"vl_api_%s_t:\\n\");\n" % t.name
 
+        write(signature.format(name=t.name))
         for o in t.block:
-            if o.type != 'Field':
-                continue
-            if o.fieldtype in format_strings:
-                output += "    vl_print(handle, \"%s: %s\\n\", a->%s);\n" % \
-                          (o.fieldname, format_strings[o.fieldtype],
-                           o.fieldname)
-
-        output += '    return handle;\n'
-        output += '}\n\n'
-        output += duplicate_wrapper_tail()
+            pp.print_obj(o, stream)
+
+        write('    return s;\n')
+        write('}\n\n')
+
+    write("\n#endif")
+    write("\n#endif /* vl_printfun_types */\n")
 
-    output += "\n#endif /* vl_printfun */\n"
 
+def imports(imports):
+    output = '/* Imported API files */\n'
+    output += '#ifndef vl_api_version\n'
+
+    for i in imports:
+        s = i.filename.replace('plugins/', '')
+        output += '#include <{}.h>\n'.format(s)
+    output += '#endif\n'
     return output
 
 
@@ -239,11 +434,64 @@ endian_strings = {
 }
 
 
-def endianfun(objs):
+def endianfun_array(o):
+    forloop = '''\
+    for (i = 0; i < {length}; i++) {{
+        a->{name}[i] = {format}(a->{name}[i]);
+    }}
+'''
+
+    forloop_format = '''\
+    for (i = 0; i < {length}; i++) {{
+        {type}_endian(&a->{name}[i]);
+    }}
+'''
+
+    output = ''
+    if o.fieldtype == 'u8' or o.fieldtype == 'string':
+        output += '    /* a->{n} = a->{n} (no-op) */\n'.format(n=o.fieldname)
+    else:
+        lfield = 'a->' + o.lengthfield if o.lengthfield else o.length
+        if o.fieldtype in endian_strings:
+            output += (forloop
+                       .format(length=lfield,
+                               format=endian_strings[o.fieldtype],
+                               name=o.fieldname))
+        else:
+            output += (forloop_format
+                       .format(length=lfield, type=o.fieldtype,
+                               name=o.fieldname))
+    return output
+
+
+def endianfun_obj(o):
+    output = ''
+    if o.type == 'Array':
+        return endianfun_array(o)
+    elif o.type != 'Field':
+        output += ('    s = format(s, "\\n{} {} {} (print not implemented");\n'
+                   .format(o.type, o.fieldtype, o.fieldname))
+        return output
+    if o.fieldtype in endian_strings:
+        output += ('    a->{name} = {format}(a->{name});\n'
+                   .format(name=o.fieldname,
+                           format=endian_strings[o.fieldtype]))
+    elif o.fieldtype.startswith('vl_api_'):
+        output += ('    {type}_endian(&a->{name});\n'
+                   .format(type=o.fieldtype, name=o.fieldname))
+    else:
+        output += '    /* a->{n} = a->{n} (no-op) */\n'.format(n=o.fieldname)
+
+    return output
+
+
+def endianfun(objs, aliases, modulename):
     output = '''\
 
 /****** Endian swap functions *****/\n\
 #ifdef vl_endianfun
+#ifndef included_{module}_endianfun
+#define included_{module}_endianfun
 
 #undef clib_net_to_host_uword
 #ifdef LP64
@@ -253,30 +501,55 @@ def endianfun(objs):
 #endif
 
 '''
+    output = output.format(module=modulename)
+
+    signature = '''\
+static inline void vl_api_{name}_t_endian (vl_api_{name}_t *a)
+{{
+    int i __attribute__((unused));
+'''
+
+    for k, v in aliases.items():
+        if v.manual_endian:
+            output += "/***** manual: vl_api_%s_t_endian  *****/\n\n" % k
+            continue
+
+        output += signature.format(name=k)
+        if ('length' in v.alias and v.alias['length'] and
+                v.alias['type'] == 'u8'):
+            output += ('    /* a->{name} = a->{name} (no-op) */\n'
+                       .format(name=k))
+        elif v.alias['type'] in format_strings:
+            output += ('    *a = {}(*a);\n'
+                       .format(endian_strings[v.alias['type']]))
+        else:
+            output += '    /* Not Implemented yet {} */'.format(k)
+        output += '}\n\n'
 
     for t in objs:
         if t.__class__.__name__ == 'Enum':
+            output += signature.format(name=t.name)
+            if t.enumtype in endian_strings:
+                output += ('    *a = {}(*a);\n'
+                           .format(endian_strings[t.enumtype]))
+            else:
+                output += ('    /* a->{name} = a->{name} (no-op) */\n'
+                           .format(name=t.name))
+
+            output += '}\n\n'
             continue
+
         if t.manual_endian:
             output += "/***** manual: vl_api_%s_t_endian  *****/\n\n" % t.name
             continue
-        output += duplicate_wrapper_head(t.name + '_t_endian')
-        output += "static inline void vl_api_%s_t_endian (vl_api_%s_t *a)" % \
-                  (t.name, t.name)
-        output += "\n{\n"
 
-        for o in t.block:
-            if o.type != 'Field':
-                continue
-            if o.fieldtype in endian_strings:
-                output += "    a->%s = %s(a->%s);\n" % \
-                        (o.fieldname, endian_strings[o.fieldtype], o.fieldname)
-            else:
-                output += "    /* a->%s = a->%s (no-op) */\n" % \
-                                          (o.fieldname, o.fieldname)
+        output += signature.format(name=t.name)
 
+        for o in t.block:
+            output += endianfun_obj(o)
         output += '}\n\n'
-        output += duplicate_wrapper_tail()
+
+    output += "\n#endif"
     output += "\n#endif /* vl_endianfun */\n\n"
 
     return output
@@ -304,16 +577,23 @@ def version_tuple(s, module):
 # Plugin entry point
 #
 def run(input_filename, s):
+    stream = StringIO()
     basename = os.path.basename(input_filename)
     filename, file_extension = os.path.splitext(basename)
+    modulename = filename.replace('.', '_')
+
     output = top_boilerplate.format(datestring=datestring,
                                     input_filename=basename)
+    output += imports(s['Import'])
     output += msg_ids(s)
     output += msg_names(s)
     output += msg_name_crc_list(s, filename)
-    output += typedefs(s['types'] + s['Define'], s['Alias'], filename + file_extension)
-    output += printfun(s['types'] + s['Define'])
-    output += endianfun(s['types'] + s['Define'])
+    output += typedefs(s['types'] + s['Define'], s['Alias'],
+                       filename + file_extension)
+    printfun_types(s['types'], s['Alias'], stream, modulename)
+    printfun(s['Define'], stream, modulename)
+    output += stream.getvalue()
+    output += endianfun(s['types'] + s['Define'], s['Alias'],  modulename)
     output += version_tuple(s, basename)
     output += bottom_boilerplate.format(input_filename=basename,
                                         file_crc=s['file_crc'])
diff --git a/src/tools/vppapigen/vppapigen_json.py b/src/tools/vppapigen/vppapigen_json.py
index 124c0d3a0bd..ef1c9823c33 100644
--- a/src/tools/vppapigen/vppapigen_json.py
+++ b/src/tools/vppapigen/vppapigen_json.py
@@ -26,12 +26,14 @@ def walk_services(s):
     return r
 
 
-def walk_defs(s, is_message = False):
+def walk_defs(s, is_message=False):
     r = []
     for t in s:
         d = []
         d.append(t.name)
         for b in t.block:
+            if b.type == 'Option':
+                continue
             if b.type == 'Field':
                 if b.limit:
                     d.append([b.fieldtype, b.fieldname, b.limit])
@@ -39,7 +41,8 @@ def walk_defs(s, is_message = False):
                     d.append([b.fieldtype, b.fieldname])
             elif b.type == 'Array':
                 if b.lengthfield:
-                    d.append([b.fieldtype, b.fieldname, b.length, b.lengthfield])
+                    d.append([b.fieldtype, b.fieldname,
+                              b.length, b.lengthfield])
                 else:
                     d.append([b.fieldtype, b.fieldname, b.length])
             elif b.type == 'Union':
@@ -62,12 +65,15 @@ def walk_defs(s, is_message = False):
 def run(filename, s):
     j = {}
 
-    j['types'] = walk_defs([o for o in s['types'] if o.__class__.__name__ == 'Typedef'])
+    j['types'] = (walk_defs([o for o in s['types']
+                             if o.__class__.__name__ == 'Typedef']))
     j['messages'] = walk_defs(s['Define'], True)
-    j['unions'] = walk_defs([o for o in s['types'] if o.__class__.__name__ == 'Union'])
-    j['enums'] = walk_enums([o for o in s['types'] if o.__class__.__name__ == 'Enum'])
+    j['unions'] = (walk_defs([o for o in s['types']
+                              if o.__class__.__name__ == 'Union']))
+    j['enums'] = (walk_enums([o for o in s['types']
+                              if o.__class__.__name__ == 'Enum']))
     j['services'] = walk_services(s['Service'])
     j['options'] = s['Option']
-    j['aliases'] = s['Alias']
+    j['aliases'] = {k: v.alias for k, v in s['Alias'].items()}
     j['vl_api_version'] = hex(s['file_crc'])
     return json.dumps(j, indent=4, separators=(',', ': '))
diff --git a/src/tools/vppapitrace/vppapitrace.py b/src/tools/vppapitrace/vppapitrace.py
index 640b820d994..8089b3a2236 100755
--- a/src/tools/vppapitrace/vppapitrace.py
+++ b/src/tools/vppapitrace/vppapitrace.py
@@ -102,6 +102,7 @@ def unserialize_msgtbl(data, offset):
 
 def serialize_msgtbl(messages):
     offset = 0
+    # XXX 100K?
     data = bytearray(100000)
     nmsg = len(messages)
     data = struct.pack(">I", nmsg)
@@ -380,8 +381,8 @@ def generate(args):
 
     filename, file_extension = os.path.splitext(args.input)
     input_type = JSON if file_extension == '.json' else APITRACE
-
     filename, file_extension = os.path.splitext(args.output)
+
     if args.todump:
         output_type = DUMP
     else:
@@ -395,7 +396,7 @@ def generate(args):
     if input_type == output_type:
         sys.exit("error: Nothing to convert between")
 
-    if input_type == JSON and output_type == APITRACE:
+    if input_type != JSON and output_type == APITRACE:
         sys.exit("error: Input file must be JSON file: {}".format(args.input))
 
     messages, services = init_api(args.apidir)
@@ -407,11 +408,11 @@ def generate(args):
             i += 1
 
         n, result = json2apitrace(messages, args.input)
+        msgtbl = serialize_msgtbl(messages)
+
         print('API messages: {}'.format(n))
-        header = struct.pack(">IIB", n, len(messages), 0)
+        header = struct.pack(">IIB", n, len(msgtbl), 0)
 
-        i = 0
-        msgtbl = serialize_msgtbl(messages)
         with open(args.output, 'wb') as outfile:
             outfile.write(header)
             outfile.write(msgtbl)