update http parser to 2.1

Makefile

@@ -4,7 +4,7 @@ LIBS    := -lpthread -lm
 TARGET  := $(shell uname -s | tr [A-Z] [a-z] 2>/dev/null || echo unknown)
 
 ifeq ($(TARGET), sunos)
-	CFLAGS += -D_PTHREADS
+	CFLAGS += -D_PTHREADS -D_POSIX_C_SOURCE=200112L
 	LIBS   += -lsocket
 endif
 

README

@@ -6,14 +6,14 @@ wrk - a HTTP benchmarking tool
 
 Basic Usage
 
-  wrk -t8 -c400 -r10m http://localhost:8080/index.html
+  wrk -t8 -c400 -d30 http://localhost:8080/index.html
 
-  This runs wrk with 8 threads, keeping 400 connections open, and making a
+  This runs a benchmark for 30 seconds, using 8 threads, and keeping
-  total of 10 million HTTP GET requests to http://localhost:8080/index.html
+  400 HTTP connections open.
 
   Output:
 
-  Making 10000000 requests to http://localhost:8080/index.html
+  Running 30s test @ http://localhost:8080/index.html
     8 threads and 400 connections
     Thread Stats   Avg      Stdev     Max   +/- Stdev
       Latency   439.75us  350.49us   7.60ms   92.88%

src/http_parser.c

@@ -37,19 +37,24 @@
 # define MIN(a,b) ((a) < (b) ? (a) : (b))
 #endif
 
+#ifndef ARRAY_SIZE
+# define ARRAY_SIZE(a) (sizeof(a) / sizeof((a)[0]))
+#endif
+
+#ifndef BIT_AT
+# define BIT_AT(a, i)                                                \
+  (!!((unsigned int) (a)[(unsigned int) (i) >> 3] &                  \
+   (1 << ((unsigned int) (i) & 7))))
+#endif
+
+#ifndef ELEM_AT
+# define ELEM_AT(a, i, v) ((unsigned int) (i) < ARRAY_SIZE(a) ? (a)[(i)] : (v))
+#endif
 
-#if HTTP_PARSER_DEBUG
-#define SET_ERRNO(e)                                                 \
-do {                                                                 \
-  parser->http_errno = (e);                                          \
-  parser->error_lineno = __LINE__;                                   \
-} while (0)
-#else
 #define SET_ERRNO(e)                                                 \
 do {                                                                 \
   parser->http_errno = (e);                                          \
 } while(0)
-#endif
 
 
 /* Run the notify callback FOR, returning ER if it fails */
@@ -123,31 +128,10 @@ do {                                                                 \
 
 
 static const char *method_strings[] =
-  { "DELETE"
+  {
-  , "GET"
+#define XX(num, name, string) #string,
-  , "HEAD"
+  HTTP_METHOD_MAP(XX)
-  , "POST"
+#undef XX
-  , "PUT"
-  , "CONNECT"
-  , "OPTIONS"
-  , "TRACE"
-  , "COPY"
-  , "LOCK"
-  , "MKCOL"
-  , "MOVE"
-  , "PROPFIND"
-  , "PROPPATCH"
-  , "UNLOCK"
-  , "REPORT"
-  , "MKACTIVITY"
-  , "CHECKOUT"
-  , "MERGE"
-  , "M-SEARCH"
-  , "NOTIFY"
-  , "SUBSCRIBE"
-  , "UNSUBSCRIBE"
-  , "PATCH"
-  , "PURGE"
   };
 
 
@@ -205,41 +189,49 @@ static const int8_t unhex[256] =
   };
 
 
-static const uint8_t normal_url_char[256] = {
+#if HTTP_PARSER_STRICT
-/*   0 nul    1 soh    2 stx    3 etx    4 eot    5 enq    6 ack    7 bel  */
+# define T(v) 0
-        0,       0,       0,       0,       0,       0,       0,       0,
+#else
-/*   8 bs     9 ht    10 nl    11 vt    12 np    13 cr    14 so    15 si   */
+# define T(v) v
-        0,       0,       0,       0,       0,       0,       0,       0,
+#endif
-/*  16 dle   17 dc1   18 dc2   19 dc3   20 dc4   21 nak   22 syn   23 etb */
-        0,       0,       0,       0,       0,       0,       0,       0,
-/*  24 can   25 em    26 sub   27 esc   28 fs    29 gs    30 rs    31 us  */
-        0,       0,       0,       0,       0,       0,       0,       0,
-/*  32 sp    33  !    34  "    35  #    36  $    37  %    38  &    39  '  */
-        0,       1,       1,       0,       1,       1,       1,       1,
-/*  40  (    41  )    42  *    43  +    44  ,    45  -    46  .    47  /  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/*  48  0    49  1    50  2    51  3    52  4    53  5    54  6    55  7  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/*  56  8    57  9    58  :    59  ;    60  <    61  =    62  >    63  ?  */
-        1,       1,       1,       1,       1,       1,       1,       0,
-/*  64  @    65  A    66  B    67  C    68  D    69  E    70  F    71  G  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/*  72  H    73  I    74  J    75  K    76  L    77  M    78  N    79  O  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/*  80  P    81  Q    82  R    83  S    84  T    85  U    86  V    87  W  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/*  88  X    89  Y    90  Z    91  [    92  \    93  ]    94  ^    95  _  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/*  96  `    97  a    98  b    99  c   100  d   101  e   102  f   103  g  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/* 104  h   105  i   106  j   107  k   108  l   109  m   110  n   111  o  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/* 112  p   113  q   114  r   115  s   116  t   117  u   118  v   119  w  */
-        1,       1,       1,       1,       1,       1,       1,       1,
-/* 120  x   121  y   122  z   123  {   124  |   125  }   126  ~   127 del */
-        1,       1,       1,       1,       1,       1,       1,       0, };
 
 
+static const uint8_t normal_url_char[32] = {
+/*   0 nul    1 soh    2 stx    3 etx    4 eot    5 enq    6 ack    7 bel  */
+        0    |   0    |   0    |   0    |   0    |   0    |   0    |   0,
+/*   8 bs     9 ht    10 nl    11 vt    12 np    13 cr    14 so    15 si   */
+        0    | T(2)   |   0    |   0    | T(16)  |   0    |   0    |   0,
+/*  16 dle   17 dc1   18 dc2   19 dc3   20 dc4   21 nak   22 syn   23 etb */
+        0    |   0    |   0    |   0    |   0    |   0    |   0    |   0,
+/*  24 can   25 em    26 sub   27 esc   28 fs    29 gs    30 rs    31 us  */
+        0    |   0    |   0    |   0    |   0    |   0    |   0    |   0,
+/*  32 sp    33  !    34  "    35  #    36  $    37  %    38  &    39  '  */
+        0    |   2    |   4    |   0    |   16   |   32   |   64   |  128,
+/*  40  (    41  )    42  *    43  +    44  ,    45  -    46  .    47  /  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/*  48  0    49  1    50  2    51  3    52  4    53  5    54  6    55  7  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/*  56  8    57  9    58  :    59  ;    60  <    61  =    62  >    63  ?  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |   0,
+/*  64  @    65  A    66  B    67  C    68  D    69  E    70  F    71  G  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/*  72  H    73  I    74  J    75  K    76  L    77  M    78  N    79  O  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/*  80  P    81  Q    82  R    83  S    84  T    85  U    86  V    87  W  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/*  88  X    89  Y    90  Z    91  [    92  \    93  ]    94  ^    95  _  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/*  96  `    97  a    98  b    99  c   100  d   101  e   102  f   103  g  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/* 104  h   105  i   106  j   107  k   108  l   109  m   110  n   111  o  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/* 112  p   113  q   114  r   115  s   116  t   117  u   118  v   119  w  */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |  128,
+/* 120  x   121  y   122  z   123  {   124  |   125  }   126  ~   127 del */
+        1    |   2    |   4    |   8    |   16   |   32   |   64   |   0, };
+
+#undef T
+
 enum state
   { s_dead = 1 /* important that this is > 0 */
 
@@ -266,13 +258,9 @@ enum state
   , s_req_schema
   , s_req_schema_slash
   , s_req_schema_slash_slash
-  , s_req_host_start
+  , s_req_server_start
-  , s_req_host_v6_start
+  , s_req_server
-  , s_req_host_v6
+  , s_req_server_with_at
-  , s_req_host_v6_end
-  , s_req_host
-  , s_req_port_start
-  , s_req_port
   , s_req_path
   , s_req_query_string_start
   , s_req_query_string
@@ -350,6 +338,19 @@ enum header_states
   , h_connection_close
   };
 
+enum http_host_state
+  {
+    s_http_host_dead = 1
+  , s_http_userinfo_start
+  , s_http_userinfo
+  , s_http_host_start
+  , s_http_host_v6_start
+  , s_http_host
+  , s_http_host_v6
+  , s_http_host_v6_end
+  , s_http_host_port_start
+  , s_http_host_port
+};
 
 /* Macros for character classes; depends on strict-mode  */
 #define CR                  '\r'
@@ -359,15 +360,21 @@ enum header_states
 #define IS_NUM(c)           ((c) >= '0' && (c) <= '9')
 #define IS_ALPHANUM(c)      (IS_ALPHA(c) || IS_NUM(c))
 #define IS_HEX(c)           (IS_NUM(c) || (LOWER(c) >= 'a' && LOWER(c) <= 'f'))
+#define IS_MARK(c)          ((c) == '-' || (c) == '_' || (c) == '.' || \
+  (c) == '!' || (c) == '~' || (c) == '*' || (c) == '\'' || (c) == '(' || \
+  (c) == ')')
+#define IS_USERINFO_CHAR(c) (IS_ALPHANUM(c) || IS_MARK(c) || (c) == '%' || \
+  (c) == ';' || (c) == ':' || (c) == '&' || (c) == '=' || (c) == '+' || \
+  (c) == '$' || (c) == ',')
 
 #if HTTP_PARSER_STRICT
 #define TOKEN(c)            (tokens[(unsigned char)c])
-#define IS_URL_CHAR(c)      (normal_url_char[(unsigned char) (c)])
+#define IS_URL_CHAR(c)      (BIT_AT(normal_url_char, (unsigned char)c))
 #define IS_HOST_CHAR(c)     (IS_ALPHANUM(c) || (c) == '.' || (c) == '-')
 #else
 #define TOKEN(c)            ((c == ' ') ? ' ' : tokens[(unsigned char)c])
 #define IS_URL_CHAR(c)                                                         \
-  (normal_url_char[(unsigned char) (c)] || ((c) & 0x80))
+  (BIT_AT(normal_url_char, (unsigned char)c) || ((c) & 0x80))
 #define IS_HOST_CHAR(c)                                                        \
   (IS_ALPHANUM(c) || (c) == '.' || (c) == '-' || (c) == '_')
 #endif
@@ -401,7 +408,7 @@ static struct {
 };
 #undef HTTP_STRERROR_GEN
 
-int http_message_needs_eof(http_parser *parser);
+int http_message_needs_eof(const http_parser *parser);
 
 /* Our URL parser.
  *
@@ -417,7 +424,15 @@ int http_message_needs_eof(http_parser *parser);
 static enum state
 parse_url_char(enum state s, const char ch)
 {
-  assert(!isspace(ch));
+  if (ch == ' ' || ch == '\r' || ch == '\n') {
+    return s_dead;
+  }
+
+#if HTTP_PARSER_STRICT
+  if (ch == '\t' || ch == '\f') {
+    return s_dead;
+  }
+#endif
 
   switch (s) {
     case s_req_spaces_before_url:
@@ -455,67 +470,33 @@ parse_url_char(enum state s, const char ch)
 
     case s_req_schema_slash_slash:
       if (ch == '/') {
-        return s_req_host_start;
+        return s_req_server_start;
       }
 
       break;
 
-    case s_req_host_start:
+    case s_req_server_with_at:
-      if (ch == '[') {
+      if (ch == '@') {
-        return s_req_host_v6_start;
+        return s_dead;
       }
 
-      if (IS_HOST_CHAR(ch)) {
+    /* FALLTHROUGH */
-        return s_req_host;
+    case s_req_server_start:
+    case s_req_server:
+      if (ch == '/') {
+        return s_req_path;
       }
 
-      break;
+      if (ch == '?') {
-
+        return s_req_query_string_start;
-    case s_req_host:
-      if (IS_HOST_CHAR(ch)) {
-        return s_req_host;
       }
 
-      /* FALLTHROUGH */
+      if (ch == '@') {
-    case s_req_host_v6_end:
+        return s_req_server_with_at;
-      switch (ch) {
-        case ':':
-          return s_req_port_start;
-
-        case '/':
-          return s_req_path;
-
-        case '?':
-          return s_req_query_string_start;
       }
 
-      break;
+      if (IS_USERINFO_CHAR(ch) || ch == '[' || ch == ']') {
-
+        return s_req_server;
-    case s_req_host_v6:
-      if (ch == ']') {
-        return s_req_host_v6_end;
-      }
-
-      /* FALLTHROUGH */
-    case s_req_host_v6_start:
-      if (IS_HEX(ch) || ch == ':') {
-        return s_req_host_v6;
-      }
-      break;
-
-    case s_req_port:
-      switch (ch) {
-        case '/':
-          return s_req_path;
-
-        case '?':
-          return s_req_query_string_start;
-      }
-
-      /* FALLTHROUGH */
-    case s_req_port_start:
-      if (IS_NUM(ch)) {
-        return s_req_port;
       }
 
       break;
@@ -637,13 +618,9 @@ size_t http_parser_execute (http_parser *parser,
   case s_req_schema:
   case s_req_schema_slash:
   case s_req_schema_slash_slash:
-  case s_req_host_start:
+  case s_req_server_start:
-  case s_req_host_v6_start:
+  case s_req_server:
-  case s_req_host_v6:
+  case s_req_server_with_at:
-  case s_req_host_v6_end:
-  case s_req_host:
-  case s_req_port_start:
-  case s_req_port:
   case s_req_query_string_start:
   case s_req_query_string:
   case s_req_fragment_start:
@@ -889,6 +866,7 @@ size_t http_parser_execute (http_parser *parser,
       case s_res_line_almost_done:
         STRICT_CHECK(ch != LF);
         parser->state = s_header_field_start;
+        CALLBACK_NOTIFY(status_complete);
         break;
 
       case s_start_req:
@@ -918,7 +896,7 @@ size_t http_parser_execute (http_parser *parser,
             /* or PROPFIND|PROPPATCH|PUT|PATCH|PURGE */
             break;
           case 'R': parser->method = HTTP_REPORT; break;
-          case 'S': parser->method = HTTP_SUBSCRIBE; break;
+          case 'S': parser->method = HTTP_SUBSCRIBE; /* or SEARCH */ break;
           case 'T': parser->method = HTTP_TRACE; break;
           case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE */ break;
           default:
@@ -965,6 +943,12 @@ size_t http_parser_execute (http_parser *parser,
           } else {
             goto error;
           }
+        } else if (parser->method == HTTP_SUBSCRIBE) {
+          if (parser->index == 1 && ch == 'E') {
+            parser->method = HTTP_SEARCH;
+          } else {
+            goto error;
+          }
         } else if (parser->index == 1 && parser->method == HTTP_POST) {
           if (ch == 'R') {
             parser->method = HTTP_PROPFIND; /* or HTTP_PROPPATCH */
@@ -998,7 +982,7 @@ size_t http_parser_execute (http_parser *parser,
 
         MARK(url);
         if (parser->method == HTTP_CONNECT) {
-          parser->state = s_req_host_start;
+          parser->state = s_req_server_start;
         }
 
         parser->state = parse_url_char((enum state)parser->state, ch);
@@ -1013,10 +997,7 @@ size_t http_parser_execute (http_parser *parser,
       case s_req_schema:
       case s_req_schema_slash:
       case s_req_schema_slash_slash:
-      case s_req_host_start:
+      case s_req_server_start:
-      case s_req_host_v6_start:
-      case s_req_host_v6:
-      case s_req_port_start:
       {
         switch (ch) {
           /* No whitespace allowed here */
@@ -1036,9 +1017,8 @@ size_t http_parser_execute (http_parser *parser,
         break;
       }
 
-      case s_req_host:
+      case s_req_server:
-      case s_req_host_v6_end:
+      case s_req_server_with_at:
-      case s_req_port:
       case s_req_path:
       case s_req_query_string_start:
       case s_req_query_string:
@@ -1867,7 +1847,7 @@ error:
 
 /* Does the parser need to see an EOF to find the end of the message? */
 int
-http_message_needs_eof (http_parser *parser)
+http_message_needs_eof (const http_parser *parser)
 {
   if (parser->type == HTTP_REQUEST) {
     return 0;
@@ -1890,7 +1870,7 @@ http_message_needs_eof (http_parser *parser)
 
 
 int
-http_should_keep_alive (http_parser *parser)
+http_should_keep_alive (const http_parser *parser)
 {
   if (parser->http_major > 0 && parser->http_minor > 0) {
     /* HTTP/1.1 */
@@ -1908,9 +1888,10 @@ http_should_keep_alive (http_parser *parser)
 }
 
 
-const char * http_method_str (enum http_method m)
+const char *
+http_method_str (enum http_method m)
 {
-  return method_strings[m];
+  return ELEM_AT(method_strings, m, "<unknown>");
 }
 
 
@@ -1937,6 +1918,144 @@ http_errno_description(enum http_errno err) {
   return http_strerror_tab[err].description;
 }
 
+static enum http_host_state
+http_parse_host_char(enum http_host_state s, const char ch) {
+  switch(s) {
+    case s_http_userinfo:
+    case s_http_userinfo_start:
+      if (ch == '@') {
+        return s_http_host_start;
+      }
+
+      if (IS_USERINFO_CHAR(ch)) {
+        return s_http_userinfo;
+      }
+      break;
+
+    case s_http_host_start:
+      if (ch == '[') {
+        return s_http_host_v6_start;
+      }
+
+      if (IS_HOST_CHAR(ch)) {
+        return s_http_host;
+      }
+
+      break;
+
+    case s_http_host:
+      if (IS_HOST_CHAR(ch)) {
+        return s_http_host;
+      }
+
+    /* FALLTHROUGH */
+    case s_http_host_v6_end:
+      if (ch == ':') {
+        return s_http_host_port_start;
+      }
+
+      break;
+
+    case s_http_host_v6:
+      if (ch == ']') {
+        return s_http_host_v6_end;
+      }
+
+    /* FALLTHROUGH */
+    case s_http_host_v6_start:
+      if (IS_HEX(ch) || ch == ':' || ch == '.') {
+        return s_http_host_v6;
+      }
+
+      break;
+
+    case s_http_host_port:
+    case s_http_host_port_start:
+      if (IS_NUM(ch)) {
+        return s_http_host_port;
+      }
+
+      break;
+
+    default:
+      break;
+  }
+  return s_http_host_dead;
+}
+
+static int
+http_parse_host(const char * buf, struct http_parser_url *u, int found_at) {
+  enum http_host_state s;
+
+  const char *p;
+  size_t buflen = u->field_data[UF_HOST].off + u->field_data[UF_HOST].len;
+
+  u->field_data[UF_HOST].len = 0;
+
+  s = found_at ? s_http_userinfo_start : s_http_host_start;
+
+  for (p = buf + u->field_data[UF_HOST].off; p < buf + buflen; p++) {
+    enum http_host_state new_s = http_parse_host_char(s, *p);
+
+    if (new_s == s_http_host_dead) {
+      return 1;
+    }
+
+    switch(new_s) {
+      case s_http_host:
+        if (s != s_http_host) {
+          u->field_data[UF_HOST].off = p - buf;
+        }
+        u->field_data[UF_HOST].len++;
+        break;
+
+      case s_http_host_v6:
+        if (s != s_http_host_v6) {
+          u->field_data[UF_HOST].off = p - buf;
+        }
+        u->field_data[UF_HOST].len++;
+        break;
+
+      case s_http_host_port:
+        if (s != s_http_host_port) {
+          u->field_data[UF_PORT].off = p - buf;
+          u->field_data[UF_PORT].len = 0;
+          u->field_set |= (1 << UF_PORT);
+        }
+        u->field_data[UF_PORT].len++;
+        break;
+
+      case s_http_userinfo:
+        if (s != s_http_userinfo) {
+          u->field_data[UF_USERINFO].off = p - buf ;
+          u->field_data[UF_USERINFO].len = 0;
+          u->field_set |= (1 << UF_USERINFO);
+        }
+        u->field_data[UF_USERINFO].len++;
+        break;
+
+      default:
+        break;
+    }
+    s = new_s;
+  }
+
+  /* Make sure we don't end somewhere unexpected */
+  switch (s) {
+    case s_http_host_start:
+    case s_http_host_v6_start:
+    case s_http_host_v6:
+    case s_http_host_port_start:
+    case s_http_userinfo:
+    case s_http_userinfo_start:
+      return 1;
+    default:
+      break;
+  }
+
+  return 0;
+}
+
 int
 http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
                       struct http_parser_url *u)
@@ -1944,9 +2063,10 @@ http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
   enum state s;
   const char *p;
   enum http_parser_url_fields uf, old_uf;
+  int found_at = 0;
 
   u->port = u->field_set = 0;
-  s = is_connect ? s_req_host_start : s_req_spaces_before_url;
+  s = is_connect ? s_req_server_start : s_req_spaces_before_url;
   uf = old_uf = UF_MAX;
 
   for (p = buf; p < buf + buflen; p++) {
@@ -1960,10 +2080,7 @@ http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
       /* Skip delimeters */
       case s_req_schema_slash:
       case s_req_schema_slash_slash:
-      case s_req_host_start:
+      case s_req_server_start:
-      case s_req_host_v6_start:
-      case s_req_host_v6_end:
-      case s_req_port_start:
       case s_req_query_string_start:
       case s_req_fragment_start:
         continue;
@@ -1972,13 +2089,12 @@ http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
         uf = UF_SCHEMA;
         break;
 
-      case s_req_host:
+      case s_req_server_with_at:
-      case s_req_host_v6:
+        found_at = 1;
-        uf = UF_HOST;
-        break;
 
-      case s_req_port:
+      /* FALLTROUGH */
-        uf = UF_PORT;
+      case s_req_server:
+        uf = UF_HOST;
         break;
 
       case s_req_path:
@@ -2011,23 +2127,19 @@ http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
     old_uf = uf;
   }
 
+  /* host must be present if there is a schema */
+  /* parsing http:///toto will fail */
+  if ((u->field_set & ((1 << UF_SCHEMA) | (1 << UF_HOST))) != 0) {
+    if (http_parse_host(buf, u, found_at) != 0) {
+      return 1;
+    }
+  }
+
   /* CONNECT requests can only contain "hostname:port" */
   if (is_connect && u->field_set != ((1 << UF_HOST)|(1 << UF_PORT))) {
     return 1;
   }
 
-  /* Make sure we don't end somewhere unexpected */
-  switch (s) {
-  case s_req_host_v6_start:
-  case s_req_host_v6:
-  case s_req_host_v6_end:
-  case s_req_host:
-  case s_req_port_start:
-    return 1;
-  default:
-    break;
-  }
-
   if (u->field_set & (1 << UF_PORT)) {
     /* Don't bother with endp; we've already validated the string */
     unsigned long v = strtoul(buf + u->field_data[UF_PORT].off, NULL, 10);
@@ -2056,3 +2168,8 @@ http_parser_pause(http_parser *parser, int paused) {
     assert(0 && "Attempting to pause parser in error state");
   }
 }
+
+int
+http_body_is_final(const struct http_parser *parser) {
+    return parser->state == s_message_done;
+}

src/http_parser.h

@@ -24,11 +24,13 @@
 extern "C" {
 #endif
 
-#define HTTP_PARSER_VERSION_MAJOR 1
+#define HTTP_PARSER_VERSION_MAJOR 2
-#define HTTP_PARSER_VERSION_MINOR 0
+#define HTTP_PARSER_VERSION_MINOR 1
 
 #include <sys/types.h>
 #if defined(_WIN32) && !defined(__MINGW32__) && (!defined(_MSC_VER) || _MSC_VER<1600)
+#include <BaseTsd.h>
+#include <stddef.h>
 typedef __int8 int8_t;
 typedef unsigned __int8 uint8_t;
 typedef __int16 int16_t;
@@ -37,9 +39,6 @@ typedef __int32 int32_t;
 typedef unsigned __int32 uint32_t;
 typedef __int64 int64_t;
 typedef unsigned __int64 uint64_t;
-
-typedef unsigned int size_t;
-typedef int ssize_t;
 #else
 #include <stdint.h>
 #endif
@@ -51,14 +50,6 @@ typedef int ssize_t;
 # define HTTP_PARSER_STRICT 1
 #endif
 
-/* Compile with -DHTTP_PARSER_DEBUG=1 to add extra debugging information to
- * the error reporting facility.
- */
-#ifndef HTTP_PARSER_DEBUG
-# define HTTP_PARSER_DEBUG 0
-#endif
-
-
 /* Maximium header size allowed */
 #define HTTP_MAX_HEADER_SIZE (80*1024)
 
@@ -77,7 +68,7 @@ typedef struct http_parser_settings http_parser_settings;
  * chunked' headers that indicate the presence of a body.
  *
  * http_data_cb does not return data chunks. It will be call arbitrarally
- * many times for each string. E.G. you might get 10 callbacks for "on_path"
+ * many times for each string. E.G. you might get 10 callbacks for "on_url"
  * each providing just a few characters more data.
  */
 typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);
@@ -85,43 +76,44 @@ typedef int (*http_cb) (http_parser*);
 
 
 /* Request Methods */
-#define HTTP_METHOD_MAP(XX) \
+#define HTTP_METHOD_MAP(XX)         \
-  XX(0, DELETE) \
+  XX(0,  DELETE,      DELETE)       \
-  XX(1, GET) \
+  XX(1,  GET,         GET)          \
-  XX(2, HEAD) \
+  XX(2,  HEAD,        HEAD)         \
-  XX(3, POST) \
+  XX(3,  POST,        POST)         \
-  XX(4, PUT) \
+  XX(4,  PUT,         PUT)          \
-  /* pathological */ \
+  /* pathological */                \
-  XX(5, CONNECT) \
+  XX(5,  CONNECT,     CONNECT)      \
-  XX(6, OPTIONS) \
+  XX(6,  OPTIONS,     OPTIONS)      \
-  XX(7, TRACE) \
+  XX(7,  TRACE,       TRACE)        \
-  /* webdav */ \
+  /* webdav */                      \
-  XX(8, COPY) \
+  XX(8,  COPY,        COPY)         \
-  XX(9, LOCK) \
+  XX(9,  LOCK,        LOCK)         \
-  XX(10, MKCOL) \
+  XX(10, MKCOL,       MKCOL)        \
-  XX(11, MOVE) \
+  XX(11, MOVE,        MOVE)         \
-  XX(12, PROPFIND) \
+  XX(12, PROPFIND,    PROPFIND)     \
-  XX(13, PROPPATCH) \
+  XX(13, PROPPATCH,   PROPPATCH)    \
-  XX(14, UNLOCK) \
+  XX(14, SEARCH,      SEARCH)       \
-  /* subversion */ \
+  XX(15, UNLOCK,      UNLOCK)       \
-  XX(15, REPORT) \
+  /* subversion */                  \
-  XX(16, MKACTIVITY) \
+  XX(16, REPORT,      REPORT)       \
-  XX(17, CHECKOUT) \
+  XX(17, MKACTIVITY,  MKACTIVITY)   \
-  XX(18, MERGE) \
+  XX(18, CHECKOUT,    CHECKOUT)     \
-  /* upnp */ \
+  XX(19, MERGE,       MERGE)        \
-  XX(19, MSEARCH) \
+  /* upnp */                        \
-  XX(20, NOTIFY) \
+  XX(20, MSEARCH,     M-SEARCH)     \
-  XX(21, SUBSCRIBE) \
+  XX(21, NOTIFY,      NOTIFY)       \
-  XX(22, UNSUBSCRIBE) \
+  XX(22, SUBSCRIBE,   SUBSCRIBE)    \
-  /* RFC-5789 */ \
+  XX(23, UNSUBSCRIBE, UNSUBSCRIBE)  \
-  XX(23, PATCH) \
+  /* RFC-5789 */                    \
-  XX(24, PURGE) \
+  XX(24, PATCH,       PATCH)        \
+  XX(25, PURGE,       PURGE)        \
 
 enum http_method
   {
-#define XX(num, name) HTTP_##name = num,
+#define XX(num, name, string) HTTP_##name = num,
   HTTP_METHOD_MAP(XX)
-#undef X
+#undef XX
   };
 
 
@@ -149,6 +141,7 @@ enum flags
                                                                      \
   /* Callback-related errors */                                      \
   XX(CB_message_begin, "the on_message_begin callback failed")       \
+  XX(CB_status_complete, "the on_status_complete callback failed")   \
   XX(CB_url, "the on_url callback failed")                           \
   XX(CB_header_field, "the on_header_field callback failed")         \
   XX(CB_header_value, "the on_header_value callback failed")         \
@@ -195,13 +188,6 @@ enum http_errno {
 /* Get an http_errno value from an http_parser */
 #define HTTP_PARSER_ERRNO(p)            ((enum http_errno) (p)->http_errno)
 
-/* Get the line number that generated the current error */
-#if HTTP_PARSER_DEBUG
-#define HTTP_PARSER_ERRNO_LINE(p)       ((p)->error_lineno)
-#else
-#define HTTP_PARSER_ERRNO_LINE(p)       0
-#endif
-
 
 struct http_parser {
   /** PRIVATE **/
@@ -228,10 +214,6 @@ struct http_parser {
    */
   unsigned char upgrade : 1;
 
-#if HTTP_PARSER_DEBUG
-  uint32_t error_lineno;
-#endif
-
   /** PUBLIC **/
   void *data; /* A pointer to get hook to the "connection" or "socket" object */
 };
@@ -240,6 +222,7 @@ struct http_parser {
 struct http_parser_settings {
   http_cb      on_message_begin;
   http_data_cb on_url;
+  http_cb      on_status_complete;
   http_data_cb on_header_field;
   http_data_cb on_header_value;
   http_cb      on_headers_complete;
@@ -255,7 +238,8 @@ enum http_parser_url_fields
   , UF_PATH             = 3
   , UF_QUERY            = 4
   , UF_FRAGMENT         = 5
-  , UF_MAX              = 6
+  , UF_USERINFO         = 6
+  , UF_MAX              = 7
   };
 
 
@@ -287,12 +271,12 @@ size_t http_parser_execute(http_parser *parser,
 
 
 /* If http_should_keep_alive() in the on_headers_complete or
- * on_message_complete callback returns true, then this will be should be
+ * on_message_complete callback returns 0, then this should be
  * the last message on the connection.
  * If you are the server, respond with the "Connection: close" header.
  * If you are the client, close the connection.
  */
-int http_should_keep_alive(http_parser *parser);
+int http_should_keep_alive(const http_parser *parser);
 
 /* Returns a string version of the HTTP method. */
 const char *http_method_str(enum http_method m);
@@ -311,6 +295,9 @@ int http_parser_parse_url(const char *buf, size_t buflen,
 /* Pause or un-pause the parser; a nonzero value pauses */
 void http_parser_pause(http_parser *parser, int paused);
 
+/* Checks if this is the final chunk of the body. */
+int http_body_is_final(const http_parser *parser);
+
 #ifdef __cplusplus
 }
 #endif

src/units.c

@@ -62,10 +62,10 @@ static int scan_units(char *s, uint64_t *n, units *m) {
 
     if ((c = sscanf(s, "%"SCNu64"%2s", &base, unit)) < 1) return -1;
 
-    if (c == 2) {
+    if (c == 2 && strncasecmp(unit, m->base, 3)) {
         for (i = 0; m->units[i] != NULL; i++) {
             scale *= m->scale;
-            if (!strncasecmp(unit, m->units[i], sizeof(unit))) break;
+            if (!strncasecmp(unit, m->units[i], 3)) break;
         }
         if (m->units[i] == NULL) return -1;
     }
@@ -91,6 +91,14 @@ char *format_time_us(long double n) {
     return format_units(n, units, 2);
 }
 
+char *format_time_s(long double n) {
+    return format_units(n, &time_units_s, 0);
+}
+
 int scan_metric(char *s, uint64_t *n) {
     return scan_units(s, n, &metric_units);
 }
+
+int scan_time(char *s, uint64_t *n) {
+    return scan_units(s, n, &time_units_s);
+}

src/units.h

@@ -4,7 +4,9 @@
 char *format_binary(long double);
 char *format_metric(long double);
 char *format_time_us(long double);
+char *format_time_s(long double);
 
 int scan_metric(char *, uint64_t *);
+int scan_time(char *, uint64_t *);
 
 #endif /* UNITS_H */

src/wrk.c

@@ -31,7 +31,7 @@ static struct config {
     struct addrinfo addr;
     uint64_t threads;
     uint64_t connections;
-    uint64_t requests;
+    uint64_t duration;
     uint64_t timeout;
 } cfg;
 
@@ -50,17 +50,25 @@ static const struct http_parser_settings parser_settings = {
     .on_message_complete = request_complete
 };
 
+static volatile sig_atomic_t stop = 0;
+
+static void handler(int sig) {
+    stop = 1;
+}
+
 static void usage() {
     printf("Usage: wrk <options> <url>                            \n"
            "  Options:                                            \n"
-           "    -c, --connections <n>  Connections to keep open   \n"
+           "    -c, --connections <N>  Connections to keep open   \n"
-           "    -r, --requests    <n>  Total requests to make     \n"
+           "    -d, --duration    <T>  Duration of test           \n"
-           "    -t, --threads     <n>  Number of threads to use   \n"
+           "    -t, --threads     <N>  Number of threads to use   \n"
            "                                                      \n"
-           "    -H, --header      <h>  Add header to request      \n"
+           "    -H, --header      <H>  Add header to request      \n"
+           "        --timeout     <T>  Socket/request timeout     \n"
            "    -v, --version          Print version details      \n"
            "                                                      \n"
-           "  Numeric arguments may include a SI unit (2k, 2M, 2G)\n");
+           "  Numeric arguments may include a SI unit (1k, 1M, 1G)\n"
+           "  Time arguments may include a time unit (2s, 2m, 2h)\n");
 }
 
 int main(int argc, char **argv) {
@@ -104,14 +112,9 @@ int main(int argc, char **argv) {
     for (addr = addrs; addr != NULL; addr = addr->ai_next) {
         int fd = socket(addr->ai_family, addr->ai_socktype, addr->ai_protocol);
         if (fd == -1) continue;
-        if (connect(fd, addr->ai_addr, addr->ai_addrlen) == -1) {
+        rc = connect(fd, addr->ai_addr, addr->ai_addrlen);
-            if (errno == EHOSTUNREACH || errno == ECONNREFUSED) {
-                close(fd);
-                continue;
-            }
-        }
         close(fd);
-        break;
+        if (rc == 0) break;
     }
 
     if (addr == NULL) {
@@ -121,6 +124,7 @@ int main(int argc, char **argv) {
     }
 
     signal(SIGPIPE, SIG_IGN);
+    signal(SIGINT,  SIG_IGN);
     cfg.addr     = *addr;
     request.buf  = format_request(host, port, path, headers);
     request.size = strlen(request.buf);
@@ -131,12 +135,12 @@ int main(int argc, char **argv) {
 
     thread *threads = zcalloc(cfg.threads * sizeof(thread));
     uint64_t connections = cfg.connections / cfg.threads;
-    uint64_t requests    = cfg.requests    / cfg.threads;
+    uint64_t stop_at     = time_us() + (cfg.duration * 1000000);
 
     for (uint64_t i = 0; i < cfg.threads; i++) {
         thread *t = &threads[i];
         t->connections = connections;
-        t->requests    = requests;
+        t->stop_at     = stop_at;
 
         if (pthread_create(&t->thread, NULL, &thread_main, t)) {
             char *msg = strerror(errno);
@@ -145,7 +149,15 @@ int main(int argc, char **argv) {
         }
     }
 
-    printf("Making %"PRIu64" requests to %s\n", cfg.requests, url);
+    struct sigaction sa = {
+        .sa_handler = handler,
+        .sa_flags   = 0,
+    };
+    sigfillset(&sa.sa_mask);
+    sigaction(SIGINT, &sa, NULL);
+
+    char *time = format_time_s(cfg.duration);
+    printf("Running %s test @ %s\n", time, url);
     printf("  %"PRIu64" threads and %"PRIu64" connections\n", cfg.threads, cfg.connections);
 
     uint64_t start    = time_us();
@@ -282,17 +294,20 @@ static int sample_rate(aeEventLoop *loop, long long id, void *data) {
 static int request_complete(http_parser *parser) {
     connection *c = parser->data;
     thread *thread = c->thread;
+    uint64_t now = time_us();
+
+    thread->complete++;
+    c->latency = now - c->start;
 
     if (parser->status_code > 399) {
         thread->errors.status++;
     }
 
-    if (++thread->complete >= thread->requests) {
+    if (now >= thread->stop_at) {
         aeStop(thread->loop);
         goto done;
     }
 
-    c->latency = time_us() - c->start;
     if (!http_should_keep_alive(parser)) goto reconnect;
 
     http_parser_init(parser, HTTP_RESPONSE);
@@ -310,9 +325,10 @@ static int request_complete(http_parser *parser) {
 
 static int check_timeouts(aeEventLoop *loop, long long id, void *data) {
     thread *thread = data;
-    connection *c = thread->cs;
+    connection *c  = thread->cs;
+    uint64_t now   = time_us();
 
-    uint64_t maxAge = time_us() - (cfg.timeout * 1000);
+    uint64_t maxAge = now - (cfg.timeout * 1000);
 
     for (uint64_t i = 0; i < thread->connections; i++, c++) {
         if (maxAge > c->start) {
@@ -320,6 +336,10 @@ static int check_timeouts(aeEventLoop *loop, long long id, void *data) {
         }
     }
 
+    if (stop || now >= thread->stop_at) {
+        aeStop(loop);
+    }
+
     return TIMEOUT_INTERVAL_MS;
 }
 
@@ -382,26 +402,33 @@ static char *extract_url_part(char *url, struct http_parser_url *parser_url, enu
 }
 
 static char *format_request(char *host, char *port, char *path, char **headers) {
-    char *req = NULL;
+    char *req  = NULL;
-
+    char *head = NULL;
-    aprintf(&req, "GET %s HTTP/1.1\r\n", path);
-    aprintf(&req, "Host: %s", host);
-    if (port) aprintf(&req, ":%s", port);
-    aprintf(&req, "\r\n");
 
     for (char **h = headers; *h != NULL; h++) {
-        aprintf(&req, "%s\r\n", *h);
+        aprintf(&head, "%s\r\n", *h);
+        if (!strncasecmp(*h, "Host:", 5)) {
+            host = NULL;
+            port = NULL;
+        }
     }
 
-    aprintf(&req, "\r\n");
+    aprintf(&req, "GET %s HTTP/1.1\r\n", path);
+    if (host) aprintf(&req, "Host: %s", host);
+    if (port) aprintf(&req, ":%s", port);
+    if (host) aprintf(&req, "\r\n");
+    aprintf(&req, "%s\r\n", head ? head : "");
+
+    free(head);
     return req;
 }
 
 static struct option longopts[] = {
     { "connections", required_argument, NULL, 'c' },
-    { "requests",    required_argument, NULL, 'r' },
+    { "duration",    required_argument, NULL, 'd' },
     { "threads",     required_argument, NULL, 't' },
     { "header",      required_argument, NULL, 'H' },
+    { "timeout",     required_argument, NULL, 'T' },
     { "help",        no_argument,       NULL, 'h' },
     { "version",     no_argument,       NULL, 'v' },
     { NULL,          0,                 NULL,  0  }
@@ -413,10 +440,10 @@ static int parse_args(struct config *cfg, char **url, char **headers, int argc,
     memset(cfg, 0, sizeof(struct config));
     cfg->threads     = 2;
     cfg->connections = 10;
-    cfg->requests    = 100;
+    cfg->duration    = 10;
     cfg->timeout     = SOCKET_TIMEOUT_MS;
 
-    while ((c = getopt_long(argc, argv, "t:c:r:H:v?", longopts, NULL)) != -1) {
+    while ((c = getopt_long(argc, argv, "t:c:d:H:T:rv?", longopts, NULL)) != -1) {
         switch (c) {
             case 't':
                 if (scan_metric(optarg, &cfg->threads)) return -1;
@@ -424,16 +451,22 @@ static int parse_args(struct config *cfg, char **url, char **headers, int argc,
             case 'c':
                 if (scan_metric(optarg, &cfg->connections)) return -1;
                 break;
-            case 'r':
+            case 'd':
-                if (scan_metric(optarg, &cfg->requests)) return -1;
+                if (scan_time(optarg, &cfg->duration)) return -1;
                 break;
             case 'H':
                 *header++ = optarg;
                 break;
+            case 'T':
+                if (scan_time(optarg, &cfg->timeout)) return -1;
+                cfg->timeout *= 1000;
+                break;
             case 'v':
                 printf("wrk %s [%s] ", VERSION, aeGetApiName());
                 printf("Copyright (C) 2012 Will Glozer\n");
                 break;
+            case 'r':
+                fprintf(stderr, "wrk 2.0.0+ uses -d instead of -r\n");
             case 'h':
             case '?':
             case ':':
@@ -442,7 +475,7 @@ static int parse_args(struct config *cfg, char **url, char **headers, int argc,
         }
     }
 
-    if (optind == argc || !cfg->threads || !cfg->requests) return -1;
+    if (optind == argc || !cfg->threads || !cfg->duration) return -1;
 
     if (!cfg->connections || cfg->connections < cfg->threads) {
         fprintf(stderr, "number of connections must be >= threads\n");

src/wrk.h

@@ -11,7 +11,7 @@
 #include "http_parser.h"
 #include "tinymt64.h"
 
-#define VERSION  "1.1.0"
+#define VERSION  "2.0.0"
 #define RECVBUF  8192
 #define SAMPLES  100000
 
@@ -31,7 +31,7 @@ typedef struct {
     pthread_t thread;
     aeEventLoop *loop;
     uint64_t connections;
-    uint64_t requests;
+    uint64_t stop_at;
     uint64_t complete;
     uint64_t bytes;
     uint64_t start;