mirror of
https://github.com/godotengine/godot.git
synced 2024-12-27 11:24:59 +08:00
fa33e0f62d
License is LGPLv2.1 + static linking exception, version is 2.4.1
1629 lines
42 KiB
C
1629 lines
42 KiB
C
/*
|
|
* libwebsockets - small server side websockets and web server implementation
|
|
*
|
|
* Copyright (C) 2010-2017 Andy Green <andy@warmcat.com>
|
|
*
|
|
* This library is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
* License as published by the Free Software Foundation:
|
|
* version 2.1 of the License.
|
|
*
|
|
* This library is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
* Lesser General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
* License along with this library; if not, write to the Free Software
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
|
|
* MA 02110-1301 USA
|
|
*/
|
|
|
|
#include "private-libwebsockets.h"
|
|
|
|
#ifndef LWS_BUILD_HASH
|
|
#define LWS_BUILD_HASH "unknown-build-hash"
|
|
#endif
|
|
|
|
static const char *library_version = LWS_LIBRARY_VERSION " " LWS_BUILD_HASH;
|
|
|
|
/**
|
|
* lws_get_library_version: get version and git hash library built from
|
|
*
|
|
* returns a const char * to a string like "1.1 178d78c"
|
|
* representing the library version followed by the git head hash it
|
|
* was built from
|
|
*/
|
|
LWS_VISIBLE const char *
|
|
lws_get_library_version(void)
|
|
{
|
|
return library_version;
|
|
}
|
|
|
|
static const char * const mount_protocols[] = {
|
|
"http://",
|
|
"https://",
|
|
"file://",
|
|
"cgi://",
|
|
">http://",
|
|
">https://",
|
|
"callback://"
|
|
};
|
|
|
|
#if defined(LWS_WITH_HTTP2)
|
|
/*
|
|
* These are the standardized defaults.
|
|
* Override what actually goes in the vhost settings in platform or user code.
|
|
* Leave these alone because they are used to determine "what is different
|
|
* from the protocol defaults".
|
|
*/
|
|
const struct http2_settings lws_h2_defaults = { {
|
|
1,
|
|
/* H2SET_HEADER_TABLE_SIZE */ 4096,
|
|
/* *** This controls how many entries in the dynamic table ***
|
|
* Allows the sender to inform the remote endpoint of the maximum
|
|
* size of the header compression table used to decode header
|
|
* blocks, in octets. The encoder can select any size equal to or
|
|
* less than this value by using signaling specific to the header
|
|
* compression format inside a header block (see [COMPRESSION]).
|
|
* The initial value is 4,096 octets.
|
|
*/
|
|
/* H2SET_ENABLE_PUSH */ 1,
|
|
/* H2SET_MAX_CONCURRENT_STREAMS */ 0x7fffffff,
|
|
/* H2SET_INITIAL_WINDOW_SIZE */ 65535,
|
|
/* H2SET_MAX_FRAME_SIZE */ 16384,
|
|
/* H2SET_MAX_HEADER_LIST_SIZE */ 0x7fffffff,
|
|
/*< This advisory setting informs a peer of the maximum size of
|
|
* header list that the sender is prepared to accept, in octets.
|
|
* The value is based on the uncompressed size of header fields,
|
|
* including the length of the name and value in octets plus an
|
|
* overhead of 32 octets for each header field.
|
|
*/
|
|
|
|
}};
|
|
|
|
const struct http2_settings lws_h2_stock_settings = { {
|
|
1,
|
|
/* H2SET_HEADER_TABLE_SIZE */ 4096,
|
|
/* *** This controls how many entries in the dynamic table ***
|
|
* Allows the sender to inform the remote endpoint of the maximum
|
|
* size of the header compression table used to decode header
|
|
* blocks, in octets. The encoder can select any size equal to or
|
|
* less than this value by using signaling specific to the header
|
|
* compression format inside a header block (see [COMPRESSION]).
|
|
* The initial value is 4,096 octets.
|
|
*
|
|
* Can't pass h2spec with less than 4096 here...
|
|
*/
|
|
/* H2SET_ENABLE_PUSH */ 1,
|
|
/* H2SET_MAX_CONCURRENT_STREAMS */ 24,
|
|
/* H2SET_INITIAL_WINDOW_SIZE */ 65535,
|
|
/* H2SET_MAX_FRAME_SIZE */ 16384,
|
|
/* H2SET_MAX_HEADER_LIST_SIZE */ 4096,
|
|
/*< This advisory setting informs a peer of the maximum size of
|
|
* header list that the sender is prepared to accept, in octets.
|
|
* The value is based on the uncompressed size of header fields,
|
|
* including the length of the name and value in octets plus an
|
|
* overhead of 32 octets for each header field.
|
|
*/
|
|
|
|
}};
|
|
#endif
|
|
|
|
LWS_VISIBLE void *
|
|
lws_protocol_vh_priv_zalloc(struct lws_vhost *vhost,
|
|
const struct lws_protocols *prot, int size)
|
|
{
|
|
int n = 0;
|
|
|
|
/* allocate the vh priv array only on demand */
|
|
if (!vhost->protocol_vh_privs) {
|
|
vhost->protocol_vh_privs = (void **)lws_zalloc(
|
|
vhost->count_protocols * sizeof(void *), "protocol_vh_privs");
|
|
if (!vhost->protocol_vh_privs)
|
|
return NULL;
|
|
}
|
|
|
|
while (n < vhost->count_protocols && &vhost->protocols[n] != prot)
|
|
n++;
|
|
|
|
if (n == vhost->count_protocols) {
|
|
n = 0;
|
|
while (n < vhost->count_protocols &&
|
|
strcmp(vhost->protocols[n].name, prot->name))
|
|
n++;
|
|
|
|
if (n == vhost->count_protocols)
|
|
return NULL;
|
|
}
|
|
|
|
vhost->protocol_vh_privs[n] = lws_zalloc(size, "vh priv");
|
|
return vhost->protocol_vh_privs[n];
|
|
}
|
|
|
|
LWS_VISIBLE void *
|
|
lws_protocol_vh_priv_get(struct lws_vhost *vhost,
|
|
const struct lws_protocols *prot)
|
|
{
|
|
int n = 0;
|
|
|
|
if (!vhost || !vhost->protocol_vh_privs)
|
|
return NULL;
|
|
|
|
while (n < vhost->count_protocols && &vhost->protocols[n] != prot)
|
|
n++;
|
|
|
|
if (n == vhost->count_protocols) {
|
|
n = 0;
|
|
while (n < vhost->count_protocols &&
|
|
strcmp(vhost->protocols[n].name, prot->name))
|
|
n++;
|
|
|
|
if (n == vhost->count_protocols) {
|
|
lwsl_err("%s: unknown protocol %p\n", __func__, prot);
|
|
return NULL;
|
|
}
|
|
}
|
|
|
|
return vhost->protocol_vh_privs[n];
|
|
}
|
|
|
|
static const struct lws_protocol_vhost_options *
|
|
lws_vhost_protocol_options(struct lws_vhost *vh, const char *name)
|
|
{
|
|
const struct lws_protocol_vhost_options *pvo = vh->pvo;
|
|
|
|
if (!name)
|
|
return NULL;
|
|
|
|
while (pvo) {
|
|
if (!strcmp(pvo->name, name))
|
|
return pvo;
|
|
pvo = pvo->next;
|
|
}
|
|
|
|
return NULL;
|
|
}
|
|
|
|
/*
|
|
* inform every vhost that hasn't already done it, that
|
|
* his protocols are initializing
|
|
*/
|
|
LWS_VISIBLE int
|
|
lws_protocol_init(struct lws_context *context)
|
|
{
|
|
struct lws_vhost *vh = context->vhost_list;
|
|
const struct lws_protocol_vhost_options *pvo, *pvo1;
|
|
struct lws wsi;
|
|
int n;
|
|
|
|
if (context->doing_protocol_init)
|
|
return 0;
|
|
|
|
context->doing_protocol_init = 1;
|
|
|
|
memset(&wsi, 0, sizeof(wsi));
|
|
wsi.context = context;
|
|
|
|
lwsl_info("%s\n", __func__);
|
|
|
|
while (vh) {
|
|
wsi.vhost = vh;
|
|
|
|
/* only do the protocol init once for a given vhost */
|
|
if (vh->created_vhost_protocols)
|
|
goto next;
|
|
|
|
/* initialize supported protocols on this vhost */
|
|
|
|
for (n = 0; n < vh->count_protocols; n++) {
|
|
wsi.protocol = &vh->protocols[n];
|
|
if (!vh->protocols[n].name)
|
|
continue;
|
|
pvo = lws_vhost_protocol_options(vh,
|
|
vh->protocols[n].name);
|
|
if (pvo) {
|
|
/*
|
|
* linked list of options specific to
|
|
* vh + protocol
|
|
*/
|
|
pvo1 = pvo;
|
|
pvo = pvo1->options;
|
|
|
|
while (pvo) {
|
|
lwsl_notice(
|
|
" vhost \"%s\", protocol \"%s\", option \"%s\"\n",
|
|
vh->name,
|
|
vh->protocols[n].name,
|
|
pvo->name);
|
|
|
|
if (!strcmp(pvo->name, "default")) {
|
|
lwsl_notice("Setting default "
|
|
"protocol for vh %s to %s\n",
|
|
vh->name,
|
|
vh->protocols[n].name);
|
|
vh->default_protocol_index = n;
|
|
}
|
|
if (!strcmp(pvo->name, "raw")) {
|
|
lwsl_notice("Setting raw "
|
|
"protocol for vh %s to %s\n",
|
|
vh->name,
|
|
vh->protocols[n].name);
|
|
vh->raw_protocol_index = n;
|
|
}
|
|
pvo = pvo->next;
|
|
}
|
|
|
|
pvo = pvo1->options;
|
|
}
|
|
|
|
/*
|
|
* inform all the protocols that they are doing their
|
|
* one-time initialization if they want to.
|
|
*
|
|
* NOTE the wsi is all zeros except for the context, vh
|
|
* + protocol ptrs so lws_get_context(wsi) etc can work
|
|
*/
|
|
if (vh->protocols[n].callback(&wsi,
|
|
LWS_CALLBACK_PROTOCOL_INIT, NULL,
|
|
(void *)pvo, 0)) {
|
|
lwsl_err("%s: vhost %s failed init\n", __func__,
|
|
vh->protocols[n].name);
|
|
context->doing_protocol_init = 0;
|
|
return 1;
|
|
}
|
|
}
|
|
|
|
vh->created_vhost_protocols = 1;
|
|
next:
|
|
vh = vh->vhost_next;
|
|
}
|
|
|
|
context->doing_protocol_init = 0;
|
|
|
|
if (!context->protocol_init_done)
|
|
lws_finalize_startup(context);
|
|
|
|
context->protocol_init_done = 1;
|
|
|
|
return 0;
|
|
}
|
|
|
|
LWS_VISIBLE int
|
|
lws_callback_http_dummy(struct lws *wsi, enum lws_callback_reasons reason,
|
|
void *user, void *in, size_t len)
|
|
{
|
|
struct lws_ssl_info *si;
|
|
#ifdef LWS_WITH_CGI
|
|
struct lws_cgi_args *args;
|
|
#endif
|
|
#if defined(LWS_WITH_CGI) || defined(LWS_WITH_HTTP_PROXY)
|
|
char buf[512];
|
|
int n;
|
|
#endif
|
|
|
|
switch (reason) {
|
|
case LWS_CALLBACK_HTTP:
|
|
#ifndef LWS_NO_SERVER
|
|
if (lws_return_http_status(wsi, HTTP_STATUS_NOT_FOUND, NULL))
|
|
return -1;
|
|
|
|
if (lws_http_transaction_completed(wsi))
|
|
#endif
|
|
return -1;
|
|
break;
|
|
#if !defined(LWS_NO_SERVER)
|
|
case LWS_CALLBACK_HTTP_FILE_COMPLETION:
|
|
if (lws_http_transaction_completed(wsi))
|
|
return -1;
|
|
break;
|
|
#endif
|
|
|
|
case LWS_CALLBACK_HTTP_WRITEABLE:
|
|
#ifdef LWS_WITH_CGI
|
|
if (wsi->reason_bf & (LWS_CB_REASON_AUX_BF__CGI_HEADERS |
|
|
LWS_CB_REASON_AUX_BF__CGI)) {
|
|
n = lws_cgi_write_split_stdout_headers(wsi);
|
|
if (n < 0) {
|
|
lwsl_debug("LWS_CB_REASON_AUX_BF__CGI forcing close\n");
|
|
return -1;
|
|
}
|
|
if (!n)
|
|
lws_rx_flow_control(wsi->cgi->stdwsi[LWS_STDOUT], 1);
|
|
|
|
if (wsi->reason_bf & LWS_CB_REASON_AUX_BF__CGI_HEADERS)
|
|
wsi->reason_bf &= ~LWS_CB_REASON_AUX_BF__CGI_HEADERS;
|
|
else
|
|
wsi->reason_bf &= ~LWS_CB_REASON_AUX_BF__CGI;
|
|
break;
|
|
}
|
|
|
|
if (wsi->reason_bf & LWS_CB_REASON_AUX_BF__CGI_CHUNK_END) {
|
|
if (!wsi->http2_substream) {
|
|
memcpy(buf + LWS_PRE, "0\x0d\x0a\x0d\x0a", 5);
|
|
lwsl_debug("writing chunk terminator and exiting\n");
|
|
n = lws_write(wsi, (unsigned char *)buf + LWS_PRE,
|
|
5, LWS_WRITE_HTTP);
|
|
} else
|
|
n = lws_write(wsi, (unsigned char *)buf + LWS_PRE,
|
|
0, LWS_WRITE_HTTP_FINAL);
|
|
|
|
/* always close after sending it */
|
|
return -1;
|
|
}
|
|
#endif
|
|
#if defined(LWS_WITH_HTTP_PROXY)
|
|
if (wsi->reason_bf & LWS_CB_REASON_AUX_BF__PROXY) {
|
|
char *px = buf + LWS_PRE;
|
|
int lenx = sizeof(buf) - LWS_PRE;
|
|
|
|
/*
|
|
* our sink is writeable and our source has something
|
|
* to read. So read a lump of source material of
|
|
* suitable size to send or what's available, whichever
|
|
* is the smaller.
|
|
*/
|
|
wsi->reason_bf &= ~LWS_CB_REASON_AUX_BF__PROXY;
|
|
if (!lws_get_child(wsi))
|
|
break;
|
|
if (lws_http_client_read(lws_get_child(wsi), &px, &lenx) < 0)
|
|
return -1;
|
|
break;
|
|
}
|
|
#endif
|
|
break;
|
|
|
|
#if defined(LWS_WITH_HTTP_PROXY)
|
|
case LWS_CALLBACK_RECEIVE_CLIENT_HTTP:
|
|
assert(lws_get_parent(wsi));
|
|
if (!lws_get_parent(wsi))
|
|
break;
|
|
lws_get_parent(wsi)->reason_bf |= LWS_CB_REASON_AUX_BF__PROXY;
|
|
lws_callback_on_writable(lws_get_parent(wsi));
|
|
break;
|
|
|
|
case LWS_CALLBACK_RECEIVE_CLIENT_HTTP_READ:
|
|
assert(lws_get_parent(wsi));
|
|
n = lws_write(lws_get_parent(wsi), (unsigned char *)in,
|
|
len, LWS_WRITE_HTTP);
|
|
if (n < 0)
|
|
return -1;
|
|
break;
|
|
|
|
case LWS_CALLBACK_ESTABLISHED_CLIENT_HTTP: {
|
|
unsigned char *p, *end;
|
|
char ctype[64], ctlen = 0;
|
|
|
|
p = (unsigned char *)buf + LWS_PRE;
|
|
end = p + sizeof(buf) - LWS_PRE;
|
|
|
|
if (lws_add_http_header_status(lws_get_parent(wsi),
|
|
HTTP_STATUS_OK, &p, end))
|
|
return 1;
|
|
if (lws_add_http_header_by_token(lws_get_parent(wsi),
|
|
WSI_TOKEN_HTTP_SERVER,
|
|
(unsigned char *)"libwebsockets",
|
|
13, &p, end))
|
|
return 1;
|
|
|
|
ctlen = lws_hdr_copy(wsi, ctype, sizeof(ctype),
|
|
WSI_TOKEN_HTTP_CONTENT_TYPE);
|
|
if (ctlen > 0) {
|
|
if (lws_add_http_header_by_token(lws_get_parent(wsi),
|
|
WSI_TOKEN_HTTP_CONTENT_TYPE,
|
|
(unsigned char *)ctype, ctlen, &p, end))
|
|
return 1;
|
|
}
|
|
|
|
if (lws_finalize_http_header(lws_get_parent(wsi), &p, end))
|
|
return 1;
|
|
|
|
*p = '\0';
|
|
n = lws_write(lws_get_parent(wsi),
|
|
(unsigned char *)buf + LWS_PRE,
|
|
p - ((unsigned char *)buf + LWS_PRE),
|
|
LWS_WRITE_HTTP_HEADERS);
|
|
if (n < 0)
|
|
return -1;
|
|
|
|
break; }
|
|
|
|
#endif
|
|
|
|
#ifdef LWS_WITH_CGI
|
|
/* CGI IO events (POLLIN/OUT) appear here, our default policy is:
|
|
*
|
|
* - POST data goes on subprocess stdin
|
|
* - subprocess stdout goes on http via writeable callback
|
|
* - subprocess stderr goes to the logs
|
|
*/
|
|
case LWS_CALLBACK_CGI:
|
|
args = (struct lws_cgi_args *)in;
|
|
switch (args->ch) { /* which of stdin/out/err ? */
|
|
case LWS_STDIN:
|
|
/* TBD stdin rx flow control */
|
|
break;
|
|
case LWS_STDOUT:
|
|
/* quench POLLIN on STDOUT until MASTER got writeable */
|
|
lws_rx_flow_control(args->stdwsi[LWS_STDOUT], 0);
|
|
wsi->reason_bf |= LWS_CB_REASON_AUX_BF__CGI;
|
|
/* when writing to MASTER would not block */
|
|
lws_callback_on_writable(wsi);
|
|
break;
|
|
case LWS_STDERR:
|
|
n = lws_get_socket_fd(args->stdwsi[LWS_STDERR]);
|
|
if (n < 0)
|
|
break;
|
|
n = read(n, buf, sizeof(buf) - 2);
|
|
if (n > 0) {
|
|
if (buf[n - 1] != '\n')
|
|
buf[n++] = '\n';
|
|
buf[n] = '\0';
|
|
lwsl_notice("CGI-stderr: %s\n", buf);
|
|
}
|
|
break;
|
|
}
|
|
break;
|
|
|
|
case LWS_CALLBACK_CGI_TERMINATED:
|
|
lwsl_debug("LWS_CALLBACK_CGI_TERMINATED: %d %" PRIu64 "\n",
|
|
wsi->cgi->explicitly_chunked,
|
|
(uint64_t)wsi->cgi->content_length);
|
|
if (!wsi->cgi->explicitly_chunked &&
|
|
!wsi->cgi->content_length) {
|
|
/* send terminating chunk */
|
|
lwsl_debug("LWS_CALLBACK_CGI_TERMINATED: ending\n");
|
|
wsi->reason_bf |= LWS_CB_REASON_AUX_BF__CGI_CHUNK_END;
|
|
lws_callback_on_writable(wsi);
|
|
lws_set_timeout(wsi, PENDING_TIMEOUT_CGI, 3);
|
|
break;
|
|
}
|
|
return -1;
|
|
|
|
case LWS_CALLBACK_CGI_STDIN_DATA: /* POST body for stdin */
|
|
args = (struct lws_cgi_args *)in;
|
|
args->data[args->len] = '\0';
|
|
n = lws_get_socket_fd(args->stdwsi[LWS_STDIN]);
|
|
if (n < 0)
|
|
return -1;
|
|
n = write(n, args->data, args->len);
|
|
if (n < args->len)
|
|
lwsl_notice("LWS_CALLBACK_CGI_STDIN_DATA: "
|
|
"sent %d only %d went", n, args->len);
|
|
return n;
|
|
#endif
|
|
|
|
case LWS_CALLBACK_SSL_INFO:
|
|
si = in;
|
|
|
|
(void)si;
|
|
lwsl_notice("LWS_CALLBACK_SSL_INFO: where: 0x%x, ret: 0x%x\n",
|
|
si->where, si->ret);
|
|
break;
|
|
|
|
default:
|
|
break;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
/* list of supported protocols and callbacks */
|
|
|
|
static const struct lws_protocols protocols_dummy[] = {
|
|
/* first protocol must always be HTTP handler */
|
|
|
|
{
|
|
"http-only", /* name */
|
|
lws_callback_http_dummy, /* callback */
|
|
0, /* per_session_data_size */
|
|
0, /* max frame size / rx buffer */
|
|
0, NULL, 0
|
|
},
|
|
/*
|
|
* the other protocols are provided by lws plugins
|
|
*/
|
|
{ NULL, NULL, 0, 0, 0, NULL, 0} /* terminator */
|
|
};
|
|
|
|
#ifdef LWS_PLAT_OPTEE
|
|
#undef LWS_HAVE_GETENV
|
|
#endif
|
|
|
|
LWS_VISIBLE struct lws_vhost *
|
|
lws_create_vhost(struct lws_context *context,
|
|
struct lws_context_creation_info *info)
|
|
{
|
|
struct lws_vhost *vh = lws_zalloc(sizeof(*vh), "create vhost"),
|
|
**vh1 = &context->vhost_list;
|
|
const struct lws_http_mount *mounts;
|
|
const struct lws_protocol_vhost_options *pvo;
|
|
#ifdef LWS_WITH_PLUGINS
|
|
struct lws_plugin *plugin = context->plugin_list;
|
|
#endif
|
|
struct lws_protocols *lwsp;
|
|
int m, f = !info->pvo;
|
|
#ifdef LWS_HAVE_GETENV
|
|
char *p;
|
|
#endif
|
|
int n;
|
|
|
|
if (!vh)
|
|
return NULL;
|
|
|
|
if (!info->protocols)
|
|
info->protocols = &protocols_dummy[0];
|
|
|
|
vh->context = context;
|
|
if (!info->vhost_name)
|
|
vh->name = "default";
|
|
else
|
|
vh->name = info->vhost_name;
|
|
|
|
if (info->options & LWS_SERVER_OPTION_ONLY_RAW)
|
|
lwsl_info("%s set to only support RAW\n", vh->name);
|
|
|
|
#if defined(LWS_WITH_HTTP2)
|
|
vh->set = context->set;
|
|
if (info->http2_settings[0])
|
|
for (n = 1; n < LWS_H2_SETTINGS_LEN; n++)
|
|
vh->set.s[n] = info->http2_settings[n];
|
|
#endif
|
|
|
|
vh->iface = info->iface;
|
|
#if !defined(LWS_WITH_ESP8266) && !defined(LWS_WITH_ESP32) && !defined(OPTEE_TA) && !defined(WIN32)
|
|
vh->bind_iface = info->bind_iface;
|
|
#endif
|
|
|
|
for (vh->count_protocols = 0;
|
|
info->protocols[vh->count_protocols].callback;
|
|
vh->count_protocols++)
|
|
;
|
|
|
|
vh->options = info->options;
|
|
vh->pvo = info->pvo;
|
|
vh->headers = info->headers;
|
|
vh->user = info->user;
|
|
vh->ssl_info_event_mask = info->ssl_info_event_mask;
|
|
if (info->keepalive_timeout)
|
|
vh->keepalive_timeout = info->keepalive_timeout;
|
|
else
|
|
vh->keepalive_timeout = 5;
|
|
|
|
if (info->timeout_secs_ah_idle)
|
|
vh->timeout_secs_ah_idle = info->timeout_secs_ah_idle;
|
|
else
|
|
vh->timeout_secs_ah_idle = 10;
|
|
|
|
/*
|
|
* give the vhost a unified list of protocols including the
|
|
* ones that came from plugins
|
|
*/
|
|
lwsp = lws_zalloc(sizeof(struct lws_protocols) *
|
|
(vh->count_protocols +
|
|
context->plugin_protocol_count + 1), "vhost-specific plugin table");
|
|
if (!lwsp) {
|
|
lwsl_err("OOM\n");
|
|
return NULL;
|
|
}
|
|
|
|
m = vh->count_protocols;
|
|
memcpy(lwsp, info->protocols, sizeof(struct lws_protocols) * m);
|
|
|
|
/* for compatibility, all protocols enabled on vhost if only
|
|
* the default vhost exists. Otherwise only vhosts who ask
|
|
* for a protocol get it enabled.
|
|
*/
|
|
|
|
if (context->options & LWS_SERVER_OPTION_EXPLICIT_VHOSTS)
|
|
f = 0;
|
|
(void)f;
|
|
#ifdef LWS_WITH_PLUGINS
|
|
if (plugin) {
|
|
|
|
while (plugin) {
|
|
for (n = 0; n < plugin->caps.count_protocols; n++) {
|
|
/*
|
|
* for compatibility's sake, no pvo implies
|
|
* allow all protocols
|
|
*/
|
|
if (f || lws_vhost_protocol_options(vh,
|
|
plugin->caps.protocols[n].name)) {
|
|
memcpy(&lwsp[m],
|
|
&plugin->caps.protocols[n],
|
|
sizeof(struct lws_protocols));
|
|
m++;
|
|
vh->count_protocols++;
|
|
}
|
|
}
|
|
plugin = plugin->list;
|
|
}
|
|
}
|
|
#endif
|
|
|
|
if (
|
|
#ifdef LWS_WITH_PLUGINS
|
|
(context->plugin_list) ||
|
|
#endif
|
|
context->options & LWS_SERVER_OPTION_EXPLICIT_VHOSTS)
|
|
vh->protocols = lwsp;
|
|
else {
|
|
vh->protocols = info->protocols;
|
|
lws_free(lwsp);
|
|
}
|
|
|
|
vh->same_vh_protocol_list = (struct lws **)
|
|
lws_zalloc(sizeof(struct lws *) * vh->count_protocols, "same vh list");
|
|
|
|
vh->mount_list = info->mounts;
|
|
|
|
#ifdef LWS_WITH_UNIX_SOCK
|
|
if (LWS_UNIX_SOCK_ENABLED(context)) {
|
|
lwsl_notice("Creating Vhost '%s' path \"%s\", %d protocols\n",
|
|
vh->name, info->iface, vh->count_protocols);
|
|
} else
|
|
#endif
|
|
lwsl_notice("Creating Vhost '%s' port %d, %d protocols, IPv6 %s\n",
|
|
vh->name, info->port, vh->count_protocols,
|
|
LWS_IPV6_ENABLED(vh) ? "on" : "off");
|
|
|
|
mounts = info->mounts;
|
|
while (mounts) {
|
|
(void)mount_protocols[0];
|
|
lwsl_notice(" mounting %s%s to %s\n",
|
|
mount_protocols[mounts->origin_protocol],
|
|
mounts->origin, mounts->mountpoint);
|
|
|
|
/* convert interpreter protocol names to pointers */
|
|
pvo = mounts->interpret;
|
|
while (pvo) {
|
|
for (n = 0; n < vh->count_protocols; n++)
|
|
if (!strcmp(pvo->value, vh->protocols[n].name)) {
|
|
((struct lws_protocol_vhost_options *)pvo)->value =
|
|
(const char *)(lws_intptr_t)n;
|
|
break;
|
|
}
|
|
if (n == vh->count_protocols)
|
|
lwsl_err("ignoring unknown interpret protocol %s\n",
|
|
pvo->value);
|
|
pvo = pvo->next;
|
|
}
|
|
|
|
mounts = mounts->mount_next;
|
|
}
|
|
|
|
#ifndef LWS_NO_EXTENSIONS
|
|
#ifdef LWS_WITH_PLUGINS
|
|
if (context->plugin_extension_count) {
|
|
|
|
m = 0;
|
|
while (info->extensions && info->extensions[m].callback)
|
|
m++;
|
|
|
|
/*
|
|
* give the vhost a unified list of extensions including the
|
|
* ones that came from plugins
|
|
*/
|
|
vh->extensions = lws_zalloc(sizeof(struct lws_extension) *
|
|
(m +
|
|
context->plugin_extension_count + 1), "extensions");
|
|
if (!vh->extensions)
|
|
return NULL;
|
|
|
|
memcpy((struct lws_extension *)vh->extensions, info->extensions,
|
|
sizeof(struct lws_extension) * m);
|
|
plugin = context->plugin_list;
|
|
while (plugin) {
|
|
memcpy((struct lws_extension *)&vh->extensions[m],
|
|
plugin->caps.extensions,
|
|
sizeof(struct lws_extension) *
|
|
plugin->caps.count_extensions);
|
|
m += plugin->caps.count_extensions;
|
|
plugin = plugin->list;
|
|
}
|
|
} else
|
|
#endif
|
|
vh->extensions = info->extensions;
|
|
#endif
|
|
|
|
vh->listen_port = info->port;
|
|
#if !defined(LWS_WITH_ESP8266)
|
|
vh->http_proxy_port = 0;
|
|
vh->http_proxy_address[0] = '\0';
|
|
#if defined(LWS_WITH_SOCKS5)
|
|
vh->socks_proxy_port = 0;
|
|
vh->socks_proxy_address[0] = '\0';
|
|
#endif
|
|
|
|
/* either use proxy from info, or try get it from env var */
|
|
|
|
/* http proxy */
|
|
if (info->http_proxy_address) {
|
|
/* override for backwards compatibility */
|
|
if (info->http_proxy_port)
|
|
vh->http_proxy_port = info->http_proxy_port;
|
|
lws_set_proxy(vh, info->http_proxy_address);
|
|
} else {
|
|
#ifdef LWS_HAVE_GETENV
|
|
p = getenv("http_proxy");
|
|
if (p)
|
|
lws_set_proxy(vh, p);
|
|
#endif
|
|
}
|
|
#if defined(LWS_WITH_SOCKS5)
|
|
/* socks proxy */
|
|
if (info->socks_proxy_address) {
|
|
/* override for backwards compatibility */
|
|
if (info->socks_proxy_port)
|
|
vh->socks_proxy_port = info->socks_proxy_port;
|
|
lws_set_socks(vh, info->socks_proxy_address);
|
|
} else {
|
|
#ifdef LWS_HAVE_GETENV
|
|
p = getenv("socks_proxy");
|
|
if (p)
|
|
lws_set_socks(vh, p);
|
|
#endif
|
|
}
|
|
#endif
|
|
#endif
|
|
|
|
vh->ka_time = info->ka_time;
|
|
vh->ka_interval = info->ka_interval;
|
|
vh->ka_probes = info->ka_probes;
|
|
|
|
if (vh->options & LWS_SERVER_OPTION_STS)
|
|
lwsl_notice(" STS enabled\n");
|
|
|
|
#ifdef LWS_WITH_ACCESS_LOG
|
|
if (info->log_filepath) {
|
|
vh->log_fd = open(info->log_filepath,
|
|
O_CREAT | O_APPEND | O_RDWR, 0600);
|
|
if (vh->log_fd == (int)LWS_INVALID_FILE) {
|
|
lwsl_err("unable to open log filepath %s\n",
|
|
info->log_filepath);
|
|
goto bail;
|
|
}
|
|
#ifndef WIN32
|
|
if (context->uid != -1)
|
|
if (chown(info->log_filepath, context->uid,
|
|
context->gid) == -1)
|
|
lwsl_err("unable to chown log file %s\n",
|
|
info->log_filepath);
|
|
#endif
|
|
} else
|
|
vh->log_fd = (int)LWS_INVALID_FILE;
|
|
#endif
|
|
if (lws_context_init_server_ssl(info, vh))
|
|
goto bail;
|
|
if (lws_context_init_client_ssl(info, vh))
|
|
goto bail;
|
|
if (lws_context_init_server(info, vh)) {
|
|
lwsl_err("init server failed\n");
|
|
goto bail;
|
|
}
|
|
|
|
while (1) {
|
|
if (!(*vh1)) {
|
|
*vh1 = vh;
|
|
break;
|
|
}
|
|
vh1 = &(*vh1)->vhost_next;
|
|
};
|
|
/* for the case we are adding a vhost much later, after server init */
|
|
|
|
if (context->protocol_init_done)
|
|
lws_protocol_init(context);
|
|
|
|
return vh;
|
|
|
|
bail:
|
|
lws_free(vh);
|
|
|
|
return NULL;
|
|
}
|
|
|
|
LWS_VISIBLE int
|
|
lws_init_vhost_client_ssl(const struct lws_context_creation_info *info,
|
|
struct lws_vhost *vhost)
|
|
{
|
|
struct lws_context_creation_info i;
|
|
|
|
memcpy(&i, info, sizeof(i));
|
|
i.port = CONTEXT_PORT_NO_LISTEN;
|
|
|
|
return lws_context_init_client_ssl(&i, vhost);
|
|
}
|
|
|
|
LWS_VISIBLE struct lws_context *
|
|
lws_create_context(struct lws_context_creation_info *info)
|
|
{
|
|
struct lws_context *context = NULL;
|
|
struct lws_plat_file_ops *prev;
|
|
#ifndef LWS_NO_DAEMONIZE
|
|
int pid_daemon = get_daemonize_pid();
|
|
#endif
|
|
int n;
|
|
#if defined(__ANDROID__)
|
|
struct rlimit rt;
|
|
#endif
|
|
|
|
lwsl_info("Initial logging level %d\n", log_level);
|
|
lwsl_info("Libwebsockets version: %s\n", library_version);
|
|
#if defined(GCC_VER)
|
|
lwsl_info("Compiled with %s\n", GCC_VER);
|
|
#endif
|
|
#if LWS_POSIX
|
|
#ifdef LWS_WITH_IPV6
|
|
if (!lws_check_opt(info->options, LWS_SERVER_OPTION_DISABLE_IPV6))
|
|
lwsl_info("IPV6 compiled in and enabled\n");
|
|
else
|
|
lwsl_info("IPV6 compiled in but disabled\n");
|
|
#else
|
|
lwsl_info("IPV6 not compiled in\n");
|
|
#endif
|
|
#if !defined(LWS_PLAT_OPTEE) && !defined(LWS_PLAT_ESP32)
|
|
lws_feature_status_libev(info);
|
|
lws_feature_status_libuv(info);
|
|
#endif
|
|
#endif
|
|
lwsl_info(" LWS_DEF_HEADER_LEN : %u\n", LWS_DEF_HEADER_LEN);
|
|
lwsl_info(" LWS_MAX_PROTOCOLS : %u\n", LWS_MAX_PROTOCOLS);
|
|
lwsl_info(" LWS_MAX_SMP : %u\n", LWS_MAX_SMP);
|
|
lwsl_info(" sizeof (*info) : %ld\n", (long)sizeof(*info));
|
|
#if defined(LWS_WITH_STATS)
|
|
lwsl_info(" LWS_WITH_STATS : on\n");
|
|
#endif
|
|
#if LWS_POSIX
|
|
lwsl_info(" SYSTEM_RANDOM_FILEPATH: '%s'\n", SYSTEM_RANDOM_FILEPATH);
|
|
#endif
|
|
#if defined(LWS_WITH_HTTP2)
|
|
lwsl_info(" HTTP2 support : available\n");
|
|
#else
|
|
lwsl_info(" HTTP2 support : not configured");
|
|
#endif
|
|
if (lws_plat_context_early_init())
|
|
return NULL;
|
|
|
|
context = lws_zalloc(sizeof(struct lws_context), "context");
|
|
if (!context) {
|
|
lwsl_err("No memory for websocket context\n");
|
|
return NULL;
|
|
}
|
|
if (info->pt_serv_buf_size)
|
|
context->pt_serv_buf_size = info->pt_serv_buf_size;
|
|
else
|
|
context->pt_serv_buf_size = 4096;
|
|
|
|
#if defined(LWS_WITH_HTTP2)
|
|
context->set = lws_h2_stock_settings;
|
|
#endif
|
|
|
|
#if LWS_MAX_SMP > 1
|
|
pthread_mutex_init(&context->lock, NULL);
|
|
#endif
|
|
|
|
#if defined(LWS_WITH_ESP32)
|
|
context->last_free_heap = esp_get_free_heap_size();
|
|
#endif
|
|
|
|
/* default to just the platform fops implementation */
|
|
|
|
context->fops_platform.LWS_FOP_OPEN = _lws_plat_file_open;
|
|
context->fops_platform.LWS_FOP_CLOSE = _lws_plat_file_close;
|
|
context->fops_platform.LWS_FOP_SEEK_CUR = _lws_plat_file_seek_cur;
|
|
context->fops_platform.LWS_FOP_READ = _lws_plat_file_read;
|
|
context->fops_platform.LWS_FOP_WRITE = _lws_plat_file_write;
|
|
context->fops_platform.fi[0].sig = NULL;
|
|
|
|
/*
|
|
* arrange a linear linked-list of fops starting from context->fops
|
|
*
|
|
* platform fops
|
|
* [ -> fops_zip (copied into context so .next settable) ]
|
|
* [ -> info->fops ]
|
|
*/
|
|
|
|
context->fops = &context->fops_platform;
|
|
prev = (struct lws_plat_file_ops *)context->fops;
|
|
|
|
#if defined(LWS_WITH_ZIP_FOPS)
|
|
/* make a soft copy so we can set .next */
|
|
context->fops_zip = fops_zip;
|
|
prev->next = &context->fops_zip;
|
|
prev = (struct lws_plat_file_ops *)prev->next;
|
|
#endif
|
|
|
|
/* if user provided fops, tack them on the end of the list */
|
|
if (info->fops)
|
|
prev->next = info->fops;
|
|
|
|
context->reject_service_keywords = info->reject_service_keywords;
|
|
if (info->external_baggage_free_on_destroy)
|
|
context->external_baggage_free_on_destroy =
|
|
info->external_baggage_free_on_destroy;
|
|
|
|
context->time_up = time(NULL);
|
|
|
|
context->simultaneous_ssl_restriction = info->simultaneous_ssl_restriction;
|
|
|
|
#ifndef LWS_NO_DAEMONIZE
|
|
if (pid_daemon) {
|
|
context->started_with_parent = pid_daemon;
|
|
lwsl_info(" Started with daemon pid %d\n", pid_daemon);
|
|
}
|
|
#endif
|
|
#if defined(__ANDROID__)
|
|
n = getrlimit ( RLIMIT_NOFILE,&rt);
|
|
if (-1 == n) {
|
|
lwsl_err("Get RLIMIT_NOFILE failed!\n");
|
|
return NULL;
|
|
}
|
|
context->max_fds = rt.rlim_cur;
|
|
#else
|
|
context->max_fds = getdtablesize();
|
|
#endif
|
|
|
|
if (info->count_threads)
|
|
context->count_threads = info->count_threads;
|
|
else
|
|
context->count_threads = 1;
|
|
|
|
if (context->count_threads > LWS_MAX_SMP)
|
|
context->count_threads = LWS_MAX_SMP;
|
|
|
|
context->token_limits = info->token_limits;
|
|
|
|
context->options = info->options;
|
|
|
|
if (info->timeout_secs)
|
|
context->timeout_secs = info->timeout_secs;
|
|
else
|
|
context->timeout_secs = AWAITING_TIMEOUT;
|
|
|
|
context->ws_ping_pong_interval = info->ws_ping_pong_interval;
|
|
|
|
lwsl_info(" default timeout (secs): %u\n", context->timeout_secs);
|
|
|
|
if (info->max_http_header_data)
|
|
context->max_http_header_data = info->max_http_header_data;
|
|
else
|
|
if (info->max_http_header_data2)
|
|
context->max_http_header_data =
|
|
info->max_http_header_data2;
|
|
else
|
|
context->max_http_header_data = LWS_DEF_HEADER_LEN;
|
|
if (info->max_http_header_pool)
|
|
context->max_http_header_pool = info->max_http_header_pool;
|
|
else
|
|
context->max_http_header_pool = LWS_DEF_HEADER_POOL;
|
|
|
|
/*
|
|
* Allocate the per-thread storage for scratchpad buffers,
|
|
* and header data pool
|
|
*/
|
|
for (n = 0; n < context->count_threads; n++) {
|
|
context->pt[n].serv_buf = lws_malloc(context->pt_serv_buf_size,
|
|
"pt_serv_buf");
|
|
if (!context->pt[n].serv_buf) {
|
|
lwsl_err("OOM\n");
|
|
return NULL;
|
|
}
|
|
|
|
#ifdef LWS_WITH_LIBUV
|
|
context->pt[n].context = context;
|
|
#endif
|
|
context->pt[n].tid = n;
|
|
context->pt[n].ah_list = NULL;
|
|
context->pt[n].ah_pool_length = 0;
|
|
|
|
lws_pt_mutex_init(&context->pt[n]);
|
|
}
|
|
|
|
if (info->fd_limit_per_thread)
|
|
context->fd_limit_per_thread = info->fd_limit_per_thread;
|
|
else
|
|
context->fd_limit_per_thread = context->max_fds /
|
|
context->count_threads;
|
|
|
|
lwsl_info(" Threads: %d each %d fds\n", context->count_threads,
|
|
context->fd_limit_per_thread);
|
|
|
|
if (!info->ka_interval && info->ka_time > 0) {
|
|
lwsl_err("info->ka_interval can't be 0 if ka_time used\n");
|
|
return NULL;
|
|
}
|
|
|
|
#ifdef LWS_WITH_LIBEV
|
|
/* (Issue #264) In order to *avoid breaking backwards compatibility*, we
|
|
* enable libev mediated SIGINT handling with a default handler of
|
|
* lws_sigint_cb. The handler can be overridden or disabled
|
|
* by invoking lws_sigint_cfg after creating the context, but
|
|
* before invoking lws_initloop:
|
|
*/
|
|
context->use_ev_sigint = 1;
|
|
context->lws_ev_sigint_cb = &lws_ev_sigint_cb;
|
|
#endif /* LWS_WITH_LIBEV */
|
|
#ifdef LWS_WITH_LIBUV
|
|
/* (Issue #264) In order to *avoid breaking backwards compatibility*, we
|
|
* enable libev mediated SIGINT handling with a default handler of
|
|
* lws_sigint_cb. The handler can be overridden or disabled
|
|
* by invoking lws_sigint_cfg after creating the context, but
|
|
* before invoking lws_initloop:
|
|
*/
|
|
context->use_ev_sigint = 1;
|
|
context->lws_uv_sigint_cb = &lws_uv_sigint_cb;
|
|
#endif
|
|
#ifdef LWS_WITH_LIBEVENT
|
|
/* (Issue #264) In order to *avoid breaking backwards compatibility*, we
|
|
* enable libev mediated SIGINT handling with a default handler of
|
|
* lws_sigint_cb. The handler can be overridden or disabled
|
|
* by invoking lws_sigint_cfg after creating the context, but
|
|
* before invoking lws_initloop:
|
|
*/
|
|
context->use_ev_sigint = 1;
|
|
context->lws_event_sigint_cb = &lws_event_sigint_cb;
|
|
#endif /* LWS_WITH_LIBEVENT */
|
|
|
|
#if defined(LWS_WITH_PEER_LIMITS)
|
|
/* scale the peer hash table according to the max fds for the process,
|
|
* so that the max list depth averages 16. Eg, 1024 fd -> 64,
|
|
* 102400 fd -> 6400
|
|
*/
|
|
context->pl_hash_elements =
|
|
(context->count_threads * context->fd_limit_per_thread) / 16;
|
|
context->pl_hash_table = lws_zalloc(sizeof(struct lws_peer *) *
|
|
context->pl_hash_elements, "peer limits hash table");
|
|
context->ip_limit_ah = info->ip_limit_ah;
|
|
context->ip_limit_wsi = info->ip_limit_wsi;
|
|
#endif
|
|
|
|
lwsl_info(" mem: context: %5lu bytes (%ld ctx + (%ld thr x %d))\n",
|
|
(long)sizeof(struct lws_context) +
|
|
(context->count_threads * context->pt_serv_buf_size),
|
|
(long)sizeof(struct lws_context),
|
|
(long)context->count_threads,
|
|
context->pt_serv_buf_size);
|
|
|
|
lwsl_info(" mem: http hdr rsvd: %5lu bytes (%u thr x (%u + %lu) x %u))\n",
|
|
(long)(context->max_http_header_data +
|
|
sizeof(struct allocated_headers)) *
|
|
context->max_http_header_pool * context->count_threads,
|
|
context->count_threads,
|
|
context->max_http_header_data,
|
|
(long)sizeof(struct allocated_headers),
|
|
context->max_http_header_pool);
|
|
n = sizeof(struct lws_pollfd) * context->count_threads *
|
|
context->fd_limit_per_thread;
|
|
context->pt[0].fds = lws_zalloc(n, "fds table");
|
|
if (context->pt[0].fds == NULL) {
|
|
lwsl_err("OOM allocating %d fds\n", context->max_fds);
|
|
goto bail;
|
|
}
|
|
lwsl_info(" mem: pollfd map: %5u\n", n);
|
|
|
|
if (info->server_string) {
|
|
context->server_string = info->server_string;
|
|
context->server_string_len = (short)
|
|
strlen(context->server_string);
|
|
}
|
|
|
|
#if LWS_MAX_SMP > 1
|
|
/* each thread serves his own chunk of fds */
|
|
for (n = 1; n < (int)info->count_threads; n++)
|
|
context->pt[n].fds = context->pt[n - 1].fds +
|
|
context->fd_limit_per_thread;
|
|
#endif
|
|
|
|
if (lws_plat_init(context, info))
|
|
goto bail;
|
|
|
|
#if defined(LWS_WITH_HTTP2)
|
|
/*
|
|
* let the user code see what the platform default SETTINGS were, he
|
|
* can modify them when he creates the vhosts.
|
|
*/
|
|
for (n = 1; n < LWS_H2_SETTINGS_LEN; n++)
|
|
info->http2_settings[n] = context->set.s[n];
|
|
#endif
|
|
|
|
lws_context_init_ssl_library(info);
|
|
|
|
context->user_space = info->user;
|
|
|
|
/*
|
|
* if he's not saying he'll make his own vhosts later then act
|
|
* compatibly and make a default vhost using the data in the info
|
|
*/
|
|
if (!lws_check_opt(info->options, LWS_SERVER_OPTION_EXPLICIT_VHOSTS))
|
|
if (!lws_create_vhost(context, info)) {
|
|
lwsl_err("Failed to create default vhost\n");
|
|
return NULL;
|
|
}
|
|
|
|
lws_context_init_extensions(info, context);
|
|
|
|
lwsl_info(" mem: per-conn: %5lu bytes + protocol rx buf\n",
|
|
(unsigned long)sizeof(struct lws));
|
|
|
|
strcpy(context->canonical_hostname, "unknown");
|
|
lws_server_get_canonical_hostname(context, info);
|
|
|
|
context->uid = info->uid;
|
|
context->gid = info->gid;
|
|
|
|
#if defined(LWS_HAVE_SYS_CAPABILITY_H) && defined(LWS_HAVE_LIBCAP)
|
|
memcpy(context->caps, info->caps, sizeof(context->caps));
|
|
context->count_caps = info->count_caps;
|
|
#endif
|
|
|
|
/*
|
|
* drop any root privs for this process
|
|
* to listen on port < 1023 we would have needed root, but now we are
|
|
* listening, we don't want the power for anything else
|
|
*/
|
|
if (!lws_check_opt(info->options, LWS_SERVER_OPTION_EXPLICIT_VHOSTS))
|
|
lws_plat_drop_app_privileges(info);
|
|
|
|
/*
|
|
* give all extensions a chance to create any per-context
|
|
* allocations they need
|
|
*/
|
|
if (info->port != CONTEXT_PORT_NO_LISTEN) {
|
|
if (lws_ext_cb_all_exts(context, NULL,
|
|
LWS_EXT_CB_SERVER_CONTEXT_CONSTRUCT, NULL, 0) < 0)
|
|
goto bail;
|
|
} else
|
|
if (lws_ext_cb_all_exts(context, NULL,
|
|
LWS_EXT_CB_CLIENT_CONTEXT_CONSTRUCT, NULL, 0) < 0)
|
|
goto bail;
|
|
|
|
return context;
|
|
|
|
bail:
|
|
lws_context_destroy(context);
|
|
return NULL;
|
|
}
|
|
|
|
LWS_VISIBLE LWS_EXTERN void
|
|
lws_context_deprecate(struct lws_context *context, lws_reload_func cb)
|
|
{
|
|
struct lws_vhost *vh = context->vhost_list, *vh1;
|
|
struct lws *wsi;
|
|
|
|
/*
|
|
* "deprecation" means disable the context from accepting any new
|
|
* connections and free up listen sockets to be used by a replacement
|
|
* context.
|
|
*
|
|
* Otherwise the deprecated context remains operational, until its
|
|
* number of connected sockets falls to zero, when it is deleted.
|
|
*/
|
|
|
|
/* for each vhost, close his listen socket */
|
|
|
|
while (vh) {
|
|
wsi = vh->lserv_wsi;
|
|
if (wsi) {
|
|
wsi->socket_is_permanently_unusable = 1;
|
|
lws_close_free_wsi(wsi, LWS_CLOSE_STATUS_NOSTATUS);
|
|
wsi->context->deprecation_pending_listen_close_count++;
|
|
/*
|
|
* other vhosts can share the listen port, they
|
|
* point to the same wsi. So zap those too.
|
|
*/
|
|
vh1 = context->vhost_list;
|
|
while (vh1) {
|
|
if (vh1->lserv_wsi == wsi)
|
|
vh1->lserv_wsi = NULL;
|
|
vh1 = vh1->vhost_next;
|
|
}
|
|
}
|
|
vh = vh->vhost_next;
|
|
}
|
|
|
|
context->deprecated = 1;
|
|
context->deprecation_cb = cb;
|
|
}
|
|
|
|
LWS_VISIBLE LWS_EXTERN int
|
|
lws_context_is_deprecated(struct lws_context *context)
|
|
{
|
|
return context->deprecated;
|
|
}
|
|
|
|
LWS_VISIBLE void
|
|
lws_context_destroy2(struct lws_context *context);
|
|
|
|
|
|
static void
|
|
lws_vhost_destroy1(struct lws_vhost *vh)
|
|
{
|
|
const struct lws_protocols *protocol = NULL;
|
|
struct lws_context_per_thread *pt;
|
|
int n, m = vh->context->count_threads;
|
|
struct lws_context *context = vh->context;
|
|
struct lws wsi;
|
|
|
|
lwsl_info("%s\n", __func__);
|
|
|
|
if (vh->being_destroyed)
|
|
return;
|
|
|
|
vh->being_destroyed = 1;
|
|
|
|
/*
|
|
* Are there other vhosts that are piggybacking on our listen socket?
|
|
* If so we need to hand the listen socket off to one of the others
|
|
* so it will remain open. If not, leave it attached to the closing
|
|
* vhost and it will get closed.
|
|
*/
|
|
|
|
if (vh->lserv_wsi)
|
|
lws_start_foreach_ll(struct lws_vhost *, v, context->vhost_list) {
|
|
if (v != vh &&
|
|
!v->being_destroyed &&
|
|
v->listen_port == vh->listen_port &&
|
|
((!v->iface && !vh->iface) ||
|
|
(v->iface && vh->iface &&
|
|
!strcmp(v->iface, vh->iface)))) {
|
|
/*
|
|
* this can only be a listen wsi, which is
|
|
* restricted... it has no protocol or other
|
|
* bindings or states. So we can simply
|
|
* swap it to a vhost that has the same
|
|
* iface + port, but is not closing.
|
|
*/
|
|
assert(v->lserv_wsi == NULL);
|
|
v->lserv_wsi = vh->lserv_wsi;
|
|
vh->lserv_wsi = NULL;
|
|
if (v->lserv_wsi)
|
|
v->lserv_wsi->vhost = v;
|
|
|
|
lwsl_notice("%s: listen skt from %s to %s\n",
|
|
__func__, vh->name, v->name);
|
|
break;
|
|
}
|
|
} lws_end_foreach_ll(v, vhost_next);
|
|
|
|
/*
|
|
* Forcibly close every wsi assoicated with this vhost. That will
|
|
* include the listen socket if it is still associated with the closing
|
|
* vhost.
|
|
*/
|
|
|
|
while (m--) {
|
|
pt = &context->pt[m];
|
|
|
|
for (n = 0; (unsigned int)n < context->pt[m].fds_count; n++) {
|
|
struct lws *wsi = wsi_from_fd(context, pt->fds[n].fd);
|
|
if (!wsi)
|
|
continue;
|
|
if (wsi->vhost != vh)
|
|
continue;
|
|
|
|
lws_close_free_wsi(wsi,
|
|
LWS_CLOSE_STATUS_NOSTATUS_CONTEXT_DESTROY
|
|
/* no protocol close */);
|
|
n--;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* let the protocols destroy the per-vhost protocol objects
|
|
*/
|
|
|
|
memset(&wsi, 0, sizeof(wsi));
|
|
wsi.context = vh->context;
|
|
wsi.vhost = vh;
|
|
protocol = vh->protocols;
|
|
if (protocol) {
|
|
n = 0;
|
|
while (n < vh->count_protocols) {
|
|
wsi.protocol = protocol;
|
|
protocol->callback(&wsi, LWS_CALLBACK_PROTOCOL_DESTROY,
|
|
NULL, NULL, 0);
|
|
protocol++;
|
|
n++;
|
|
}
|
|
}
|
|
|
|
/*
|
|
* remove vhost from context list of vhosts
|
|
*/
|
|
|
|
lws_start_foreach_llp(struct lws_vhost **, pv, context->vhost_list) {
|
|
if (*pv == vh) {
|
|
*pv = vh->vhost_next;
|
|
break;
|
|
}
|
|
} lws_end_foreach_llp(pv, vhost_next);
|
|
|
|
/* add ourselves to the pending destruction list */
|
|
|
|
vh->vhost_next = vh->context->vhost_pending_destruction_list;
|
|
vh->context->vhost_pending_destruction_list = vh;
|
|
}
|
|
|
|
static void
|
|
lws_vhost_destroy2(struct lws_vhost *vh)
|
|
{
|
|
const struct lws_protocols *protocol = NULL;
|
|
struct lws_context *context = vh->context;
|
|
struct lws_deferred_free *df;
|
|
int n;
|
|
|
|
lwsl_info("%s: %p\n", __func__, vh);
|
|
|
|
/* if we are still on deferred free list, remove ourselves */
|
|
|
|
lws_start_foreach_llp(struct lws_deferred_free **, pdf,
|
|
context->deferred_free_list) {
|
|
if ((*pdf)->payload == vh) {
|
|
df = *pdf;
|
|
*pdf = df->next;
|
|
lws_free(df);
|
|
break;
|
|
}
|
|
} lws_end_foreach_llp(pdf, next);
|
|
|
|
/* remove ourselves from the pending destruction list */
|
|
|
|
lws_start_foreach_llp(struct lws_vhost **, pv,
|
|
context->vhost_pending_destruction_list) {
|
|
if ((*pv) == vh) {
|
|
*pv = (*pv)->vhost_next;
|
|
break;
|
|
}
|
|
} lws_end_foreach_llp(pv, vhost_next);
|
|
|
|
/*
|
|
* Free all the allocations associated with the vhost
|
|
*/
|
|
|
|
protocol = vh->protocols;
|
|
if (protocol) {
|
|
n = 0;
|
|
while (n < vh->count_protocols) {
|
|
if (vh->protocol_vh_privs &&
|
|
vh->protocol_vh_privs[n]) {
|
|
lws_free(vh->protocol_vh_privs[n]);
|
|
vh->protocol_vh_privs[n] = NULL;
|
|
}
|
|
protocol++;
|
|
n++;
|
|
}
|
|
}
|
|
if (vh->protocol_vh_privs)
|
|
lws_free(vh->protocol_vh_privs);
|
|
lws_ssl_SSL_CTX_destroy(vh);
|
|
lws_free(vh->same_vh_protocol_list);
|
|
#ifdef LWS_WITH_PLUGINS
|
|
if (LWS_LIBUV_ENABLED(context)) {
|
|
if (context->plugin_list)
|
|
lws_free((void *)vh->protocols);
|
|
} else
|
|
#endif
|
|
{
|
|
if (context->options & LWS_SERVER_OPTION_EXPLICIT_VHOSTS)
|
|
lws_free((void *)vh->protocols);
|
|
}
|
|
|
|
#ifdef LWS_WITH_PLUGINS
|
|
#ifndef LWS_NO_EXTENSIONS
|
|
if (context->plugin_extension_count)
|
|
lws_free((void *)vh->extensions);
|
|
#endif
|
|
#endif
|
|
#ifdef LWS_WITH_ACCESS_LOG
|
|
if (vh->log_fd != (int)LWS_INVALID_FILE)
|
|
close(vh->log_fd);
|
|
#endif
|
|
|
|
/*
|
|
* although async event callbacks may still come for wsi handles with
|
|
* pending close in the case of asycn event library like libuv,
|
|
* they do not refer to the vhost. So it's safe to free.
|
|
*/
|
|
|
|
lwsl_info(" %s: Freeing vhost %p\n", __func__, vh);
|
|
|
|
memset(vh, 0, sizeof(*vh));
|
|
lws_free(vh);
|
|
}
|
|
|
|
int
|
|
lws_check_deferred_free(struct lws_context *context, int force)
|
|
{
|
|
struct lws_deferred_free *df;
|
|
time_t now = lws_now_secs();
|
|
|
|
lws_start_foreach_llp(struct lws_deferred_free **, pdf,
|
|
context->deferred_free_list) {
|
|
if (now > (*pdf)->deadline || force) {
|
|
df = *pdf;
|
|
*pdf = df->next;
|
|
/* finalize vh destruction */
|
|
lwsl_notice("deferred vh %p destroy\n", df->payload);
|
|
lws_vhost_destroy2(df->payload);
|
|
lws_free(df);
|
|
continue; /* after deletion we already point to next */
|
|
}
|
|
} lws_end_foreach_llp(pdf, next);
|
|
|
|
return 0;
|
|
}
|
|
|
|
LWS_VISIBLE void
|
|
lws_vhost_destroy(struct lws_vhost *vh)
|
|
{
|
|
struct lws_deferred_free *df = lws_malloc(sizeof(*df), "deferred free");
|
|
|
|
if (!df)
|
|
return;
|
|
|
|
lws_vhost_destroy1(vh);
|
|
|
|
/* part 2 is deferred to allow all the handle closes to complete */
|
|
|
|
df->next = vh->context->deferred_free_list;
|
|
df->deadline = lws_now_secs() + 5;
|
|
df->payload = vh;
|
|
vh->context->deferred_free_list = df;
|
|
}
|
|
|
|
LWS_VISIBLE void
|
|
lws_context_destroy(struct lws_context *context)
|
|
{
|
|
struct lws_context_per_thread *pt;
|
|
struct lws_vhost *vh = NULL;
|
|
struct lws wsi;
|
|
int n, m;
|
|
|
|
if (!context) {
|
|
lwsl_notice("%s: ctx %p\n", __func__, context);
|
|
return;
|
|
}
|
|
if (context->being_destroyed1) {
|
|
lwsl_notice("%s: ctx %p: already being destroyed\n",
|
|
__func__, context);
|
|
return;
|
|
}
|
|
|
|
lwsl_info("%s: ctx %p\n", __func__, context);
|
|
|
|
m = context->count_threads;
|
|
context->being_destroyed = 1;
|
|
context->being_destroyed1 = 1;
|
|
|
|
memset(&wsi, 0, sizeof(wsi));
|
|
wsi.context = context;
|
|
|
|
#ifdef LWS_LATENCY
|
|
if (context->worst_latency_info[0])
|
|
lwsl_notice("Worst latency: %s\n", context->worst_latency_info);
|
|
#endif
|
|
|
|
while (m--) {
|
|
pt = &context->pt[m];
|
|
|
|
for (n = 0; (unsigned int)n < context->pt[m].fds_count; n++) {
|
|
struct lws *wsi = wsi_from_fd(context, pt->fds[n].fd);
|
|
if (!wsi)
|
|
continue;
|
|
|
|
lws_close_free_wsi(wsi,
|
|
LWS_CLOSE_STATUS_NOSTATUS_CONTEXT_DESTROY
|
|
/* no protocol close */);
|
|
n--;
|
|
}
|
|
lws_pt_mutex_destroy(pt);
|
|
}
|
|
|
|
/*
|
|
* give all extensions a chance to clean up any per-context
|
|
* allocations they might have made
|
|
*/
|
|
|
|
n = lws_ext_cb_all_exts(context, NULL,
|
|
LWS_EXT_CB_SERVER_CONTEXT_DESTRUCT, NULL, 0);
|
|
|
|
n = lws_ext_cb_all_exts(context, NULL,
|
|
LWS_EXT_CB_CLIENT_CONTEXT_DESTRUCT, NULL, 0);
|
|
|
|
/*
|
|
* inform all the protocols that they are done and will have no more
|
|
* callbacks.
|
|
*
|
|
* We can't free things until after the event loop shuts down.
|
|
*/
|
|
if (context->protocol_init_done)
|
|
vh = context->vhost_list;
|
|
while (vh) {
|
|
struct lws_vhost *vhn = vh->vhost_next;
|
|
lws_vhost_destroy1(vh);
|
|
vh = vhn;
|
|
}
|
|
|
|
for (n = 0; n < context->count_threads; n++) {
|
|
pt = &context->pt[n];
|
|
|
|
lws_libev_destroyloop(context, n);
|
|
lws_libuv_destroyloop(context, n);
|
|
lws_libevent_destroyloop(context, n);
|
|
|
|
lws_free_set_NULL(context->pt[n].serv_buf);
|
|
|
|
while (pt->ah_list)
|
|
_lws_destroy_ah(pt, pt->ah_list);
|
|
}
|
|
lws_plat_context_early_destroy(context);
|
|
|
|
if (context->pt[0].fds)
|
|
lws_free_set_NULL(context->pt[0].fds);
|
|
|
|
if (!LWS_LIBUV_ENABLED(context))
|
|
lws_context_destroy2(context);
|
|
}
|
|
|
|
/*
|
|
* call the second one after the event loop has been shut down cleanly
|
|
*/
|
|
|
|
LWS_VISIBLE void
|
|
lws_context_destroy2(struct lws_context *context)
|
|
{
|
|
struct lws_vhost *vh = NULL, *vh1;
|
|
#if defined(LWS_WITH_PEER_LIMITS)
|
|
uint32_t n;
|
|
#endif
|
|
|
|
lwsl_info("%s: ctx %p\n", __func__, context);
|
|
|
|
/*
|
|
* free all the per-vhost allocations
|
|
*/
|
|
|
|
vh = context->vhost_list;
|
|
while (vh) {
|
|
vh1 = vh->vhost_next;
|
|
lws_vhost_destroy2(vh);
|
|
vh = vh1;
|
|
}
|
|
|
|
/* remove ourselves from the pending destruction list */
|
|
|
|
while (context->vhost_pending_destruction_list)
|
|
/* removes itself from list */
|
|
lws_vhost_destroy2(context->vhost_pending_destruction_list);
|
|
|
|
|
|
lws_stats_log_dump(context);
|
|
|
|
lws_ssl_context_destroy(context);
|
|
lws_plat_context_late_destroy(context);
|
|
|
|
#if defined(LWS_WITH_PEER_LIMITS)
|
|
for (n = 0; n < context->pl_hash_elements; n++) {
|
|
lws_start_foreach_llp(struct lws_peer **, peer,
|
|
context->pl_hash_table[n]) {
|
|
struct lws_peer *df = *peer;
|
|
*peer = df->next;
|
|
lws_free(df);
|
|
continue;
|
|
} lws_end_foreach_llp(peer, next);
|
|
}
|
|
lws_free(context->pl_hash_table);
|
|
#endif
|
|
|
|
if (context->external_baggage_free_on_destroy)
|
|
free(context->external_baggage_free_on_destroy);
|
|
|
|
lws_check_deferred_free(context, 1);
|
|
|
|
#if LWS_MAX_SMP > 1
|
|
pthread_mutex_destroy(&context->lock, NULL);
|
|
#endif
|
|
|
|
lws_free(context);
|
|
}
|