X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=jcc.c;h=06ce04ae5524522cb898560d105cbf063ad42be0;hp=ad884096f354cbc801129fed46d39d2395523630;hb=0037521e356511ffd45f769e732e9bd4c120da0f;hpb=355b51d46c75908f843ede4f4c9d629fdd2a1781 diff --git a/jcc.c b/jcc.c index ad884096..06ce04ae 100644 --- a/jcc.c +++ b/jcc.c @@ -1,4 +1,4 @@ -const char jcc_rcs[] = "$Id: jcc.c,v 1.130 2007/04/19 13:47:34 fabiankeil Exp $"; +const char jcc_rcs[] = "$Id: jcc.c,v 1.177 2008/05/10 11:51:12 fabiankeil Exp $"; /********************************************************************* * * File : $Source: /cvsroot/ijbswa/current/jcc.c,v $ @@ -6,7 +6,7 @@ const char jcc_rcs[] = "$Id: jcc.c,v 1.130 2007/04/19 13:47:34 fabiankeil Exp $" * Purpose : Main file. Contains main() method, main loop, and * the main connection-handling function. * - * Copyright : Written by and Copyright (C) 2001-2007 the SourceForge + * Copyright : Written by and Copyright (C) 2001-2008 the SourceForge * Privoxy team. http://www.privoxy.org/ * * Based on the Internet Junkbuster originally written @@ -33,6 +33,208 @@ const char jcc_rcs[] = "$Id: jcc.c,v 1.130 2007/04/19 13:47:34 fabiankeil Exp $" * * Revisions : * $Log: jcc.c,v $ + * Revision 1.177 2008/05/10 11:51:12 fabiankeil + * Make the "read the rest of the headers" loop a bit more readable. + * + * Revision 1.176 2008/05/10 11:37:57 fabiankeil + * - Instead of logging when the IIS5 hack is enabled, log when it fails. + * - Remove useless comment. + * + * Revision 1.175 2008/05/09 18:53:59 fabiankeil + * Fix comment grammar. + * + * Revision 1.174 2008/05/07 18:05:53 fabiankeil + * Remove the pointless buffer in client_protocol_is_unsupported(). + * + * Revision 1.173 2008/05/06 15:09:00 fabiankeil + * Least-effort fix for bug #1821930 (reported by Lee): + * If the response doesn't look like HTTP, + * tell the client and log the problem. + * + * Revision 1.172 2008/04/16 16:38:21 fabiankeil + * Don't pass the whole csp structure to flush_socket() + * when it only needs a file descriptor and a buffer. + * + * Revision 1.171 2008/03/27 18:27:25 fabiankeil + * Remove kill-popups action. + * + * Revision 1.170 2008/03/06 16:33:46 fabiankeil + * If limit-connect isn't used, don't limit CONNECT requests to port 443. + * + * Revision 1.169 2008/03/04 18:30:39 fabiankeil + * Remove the treat-forbidden-connects-like-blocks action. We now + * use the "blocked" page for forbidden CONNECT requests by default. + * + * Revision 1.168 2008/03/02 12:25:25 fabiankeil + * Also use shiny new connect_port_is_forbidden() in jcc.c. + * + * Revision 1.167 2008/02/23 16:57:12 fabiankeil + * Rename url_actions() to get_url_actions() and let it + * use the standard parameter ordering. + * + * Revision 1.166 2008/02/23 16:33:43 fabiankeil + * Let forward_url() use the standard parameter ordering + * and mark its second parameter immutable. + * + * Revision 1.165 2008/02/02 19:36:56 fabiankeil + * Remove the "Listening ... for local connections only" log message. + * Whether or not remote connections are able to reach Privoxy is up + * to the operating system. + * + * Revision 1.164 2007/12/16 18:32:46 fabiankeil + * Prevent the log messages for CONNECT requests to unacceptable + * ports from printing the limit-connect argument as [null] if + * limit-connect hasn't been explicitly enabled. + * + * Revision 1.163 2007/12/13 01:47:11 david__schmidt + * Make sure all console-mode apps get a usage() instance + * + * Revision 1.162 2007/12/06 17:54:57 fabiankeil + * Reword NO_SERVER_DATA_RESPONSE to make it harder + * to misunderstand what the message is all about. + * + * Revision 1.161 2007/12/04 19:44:22 fabiankeil + * Unbreak trustfile which previously didn't work without + * FEATURE_TOGGLE. Fixes BR#1843585, reported by Lee. + * + * Revision 1.160 2007/11/29 18:00:29 fabiankeil + * Plug memory leak. Spotted by Valgrind, triggered by + * Privoxy-Regression-Test feeding proxyfuzz.py. + * + * Revision 1.159 2007/11/24 14:34:09 fabiankeil + * In the HTTP snipplets, refer to the client as client. + * + * Revision 1.158 2007/11/11 16:44:17 fabiankeil + * Emit a log message when activating the MS IIS5 hack. + * + * Revision 1.157 2007/11/03 17:34:49 fabiankeil + * Log the "weak randomization factor" warning only + * once for mingw32 and provide some more details. + * + * Revision 1.156 2007/11/01 18:20:58 fabiankeil + * Initialize log module after initializing mutexes, future + * deadlocks in that code should now work cross-platform. + * + * Revision 1.155 2007/10/23 20:12:45 fabiankeil + * Fix first CSUCCEED line to end in \r\n as required by RFC1945. + * Reported by Bert van Leeuwen in BR#1818808. + * + * Revision 1.154 2007/10/19 17:00:08 fabiankeil + * Downgrade "Flushing header and buffers" message to LOG_LEVEL_INFO. + * + * Revision 1.153 2007/10/14 14:12:41 fabiankeil + * When in daemon mode, close stderr after the configuration file has been + * parsed the first time. If logfile isn't set, stop logging. Fixes BR#897436. + * + * Revision 1.152 2007/10/04 18:03:34 fabiankeil + * - Fix a crash when parsing invalid requests whose first header + * is rejected by get_header(). Regression (re?)introduced + * in r1.143 by yours truly. + * - Move ACTION_VANILLA_WAFER handling into parsers.c's + * client_cookie_adder() to make sure send-vanilla-wafer can be + * controlled through tags (and thus regression-tested). + * + * Revision 1.151 2007/09/29 10:21:16 fabiankeil + * - Move get_filter_function() from jcc.c to filters.c + * so the filter functions can be static. + * - Don't bother filtering body-less responses. + * + * Revision 1.150 2007/09/28 16:39:29 fabiankeil + * Execute content filters through execute_content_filter(). + * + * Revision 1.149 2007/09/04 15:08:48 fabiankeil + * Initialize req to NULL to make sure it's defined if the + * first read_socket() call fails. Reported by icmp30. + * + * Revision 1.148 2007/08/26 16:47:13 fabiankeil + * Add Stephen Gildea's --pre-chroot-nslookup patch [#1276666], + * extensive comments moved to user manual. + * + * Revision 1.147 2007/08/25 14:42:40 fabiankeil + * Don't crash if a broken header filter wiped out the request line. + * + * Revision 1.146 2007/08/20 17:09:32 fabiankeil + * Fix byte_count calculation in case of flushes + * and don't parse the server headers a second time. + * + * Revision 1.145 2007/08/19 13:13:31 fabiankeil + * - If there's a connection problem after we already forwarded + * parts of the original content, just hang up. Fixes BR#1776724. + * - Fix warnings about unused code on mingw32. + * - In case of flushes, calculate the byte count + * less incorrectly (I think). + * + * Revision 1.144 2007/08/11 14:43:22 fabiankeil + * Add some more prototypes for static functions. + * + * Revision 1.143 2007/08/05 13:58:19 fabiankeil + * Comment out request_contains_null_bytes() until it's used again. + * + * Revision 1.142 2007/08/05 13:50:26 fabiankeil + * #1763173 from Stefan Huehner: s@const static@static const@ + * and declare some more functions static. + * + * Revision 1.141 2007/08/04 09:56:23 fabiankeil + * - Log rejected CONNECT requests with LOG_LEVEL_INFO + * and explain why they were rejected in the first place. + * - Fix the LOG_LEVEL_CLF message for crunches of unallowed + * CONNECT requests. The request line was missing. + * - Add two more XXX reminders as we don't have enough already. + * + * Revision 1.140 2007/07/21 11:51:36 fabiankeil + * As Hal noticed, checking dispatch_cgi() as the last cruncher + * looks like a bug if CGI requests are blocked unintentionally, + * so don't do it unless the user enabled the new config option + * "allow-cgi-request-crunching". + * + * Revision 1.139 2007/07/14 07:46:41 fabiankeil + * - Allow to rewrite the request destination behind the client's back. + * - Turn the weird-looking unconditional for loop that + * reads the client request into a conditional while loop. + * Move the stuff that only runs once out of the loop. + * - Move parts of chat(), server_content_type() and the + * necessary stuff to fix BR#1750917 into get_filter_function(). + * + * Revision 1.138 2007/06/03 18:45:18 fabiankeil + * Temporary workaround for BR#1730105. + * + * Revision 1.137 2007/06/01 18:16:36 fabiankeil + * Use the same mutex for gethostbyname() and gethostbyaddr() to prevent + * deadlocks and crashes on OpenBSD and possibly other OS with neither + * gethostbyname_r() nor gethostaddr_r(). Closes BR#1729174. + * Thanks to Ralf Horstmann for report and solution. + * + * Revision 1.136 2007/06/01 16:41:11 fabiankeil + * Add forward-override{} to change the forwarding settings through + * action sections. This is mainly interesting to forward different + * clients differently (for example based on User-Agent or request + * origin). + * + * Revision 1.135 2007/05/24 17:03:50 fabiankeil + * - Let usage() mention the --chroot parameter. + * - Use read_socket() consistently and always leave + * the last buffer byte alone, even in cases where + * null termination (currently) doesn't matter. + * + * Revision 1.134 2007/05/16 14:59:46 fabiankeil + * - Fix config file loading on Unix if no config file is specified. + * Since r1.97 Privoxy would always interpret the last argument as + * config file, even if it's a valid command line option. + * - Abort in case of unrecognized command line options. Closes #1719696. + * - Remove a bunch of unnecessary strcpy() calls (yay for c&p without thinking). + * - Replace the remaining strcpy() and strcat() calls with strlcpy() and strcat(). + * + * Revision 1.133 2007/05/04 11:23:19 fabiankeil + * - Don't rerun crunchers that only depend on the request URL. + * - Don't count redirects and CGI requests as "blocked requests". + * + * Revision 1.132 2007/04/25 15:15:17 fabiankeil + * Support crunching based on tags created by server-header taggers. + * + * Revision 1.131 2007/04/22 13:24:50 fabiankeil + * Make HTTP snippets static (again). Add a Content-Type for those + * with content so the browser doesn't guess it based on the URL. + * * Revision 1.130 2007/04/19 13:47:34 fabiankeil * Move crunching and request line rebuilding out of chat(). * @@ -870,7 +1072,6 @@ http://www.fabiankeil.de/sourcecode/privoxy/ #include "filters.h" #include "loaders.h" #include "parsers.h" -#include "killpopup.h" #include "miscutil.h" #include "errlog.h" #include "jbsockets.h" @@ -896,8 +1097,28 @@ int urls_rejected = 0; /* total nr of urls rejected */ int g_terminate = 0; #endif -static void listen_loop(void); +#if !defined(_WIN32) && !defined(__OS2__) && !defined(AMIGA) +static void sig_handler(int the_signal); +#endif +static int client_protocol_is_unsupported(const struct client_state *csp, char *req); +static jb_err get_request_destination_elsewhere(struct client_state *csp, struct list *headers); +static jb_err get_server_headers(struct client_state *csp); +static const char *crunch_reason(const struct http_response *rsp); +static void send_crunch_response(struct client_state *csp, struct http_response *rsp); +/* + * static int request_contains_null_bytes(const struct client_state *csp, char *buf, int len); + */ +static void build_request_line(struct client_state *csp, const struct forward_spec *fwd, char **request_line); +static jb_err change_request_destination(struct client_state *csp); static void chat(struct client_state *csp); +static void serve(struct client_state *csp); +#if !defined(_WIN32) || defined(_WIN_CONSOLE) +static void usage(const char *myname); +#endif +static void initialize_mutexes(void); +static jb_socket bind_port_helper(struct configuration_spec *config); +static void listen_loop(void); + #ifdef AMIGA void serve(struct client_state *csp); #else /* ifndef AMIGA */ @@ -920,6 +1141,10 @@ static int32 server_thread(void *data); pthread_mutex_t log_mutex; pthread_mutex_t log_init_mutex; +#if !defined(HAVE_GETHOSTBYADDR_R) || !defined(HAVE_GETHOSTBYNAME_R) +pthread_mutex_t resolver_mutex; +#endif /* !defined(HAVE_GETHOSTBYADDR_R) || !defined(HAVE_GETHOSTBYNAME_R) */ + #ifndef HAVE_GMTIME_R pthread_mutex_t gmtime_mutex; #endif /* ndef HAVE_GMTIME_R */ @@ -928,14 +1153,6 @@ pthread_mutex_t gmtime_mutex; pthread_mutex_t localtime_mutex; #endif /* ndef HAVE_GMTIME_R */ -#ifndef HAVE_GETHOSTBYADDR_R -pthread_mutex_t gethostbyaddr_mutex; -#endif /* ndef HAVE_GETHOSTBYADDR_R */ - -#ifndef HAVE_GETHOSTBYNAME_R -pthread_mutex_t gethostbyname_mutex; -#endif /* ndef HAVE_GETHOSTBYNAME_R */ - #ifndef HAVE_RANDOM pthread_mutex_t rand_mutex; #endif /* ndef HAVE_RANDOM */ @@ -948,70 +1165,111 @@ const char *pidfile = NULL; int received_hup_signal = 0; #endif /* defined unix */ -/* The vanilla wafer. */ -static const char VANILLA_WAFER[] = - "NOTICE=TO_WHOM_IT_MAY_CONCERN_" - "Do_not_send_me_any_copyrighted_information_other_than_the_" - "document_that_I_am_requesting_or_any_of_its_necessary_components._" - "In_particular_do_not_send_me_any_cookies_that_" - "are_subject_to_a_claim_of_copyright_by_anybody._" - "Take_notice_that_I_refuse_to_be_bound_by_any_license_condition_" - "(copyright_or_otherwise)_applying_to_any_cookie._"; - /* HTTP snipplets. */ -const static char CSUCCEED[] = - "HTTP/1.0 200 Connection established\n" +static const char CSUCCEED[] = + "HTTP/1.0 200 Connection established\r\n" "Proxy-Agent: Privoxy/" VERSION "\r\n\r\n"; -const static char CHEADER[] = - "HTTP/1.0 400 Invalid header received from browser\r\n" +static const char CHEADER[] = + "HTTP/1.0 400 Invalid header received from client\r\n" "Proxy-Agent: Privoxy " VERSION "\r\n" "Content-Type: text/plain\r\n" "Connection: close\r\n\r\n" - "Invalid header received from browser.\r\n"; + "Invalid header received from client.\r\n"; -const static char CFORBIDDEN[] = - "HTTP/1.0 403 Connection not allowable\r\n" - "Proxy-Agent: Privoxy " VERSION "\r\n" - "X-Hint: If you read this message interactively, then you know why this happens ,-)\r\n" - "Connection: close\r\n\r\n"; - -const static char FTP_RESPONSE[] = - "HTTP/1.0 400 Invalid request received from browser\r\n" +static const char FTP_RESPONSE[] = + "HTTP/1.0 400 Invalid request received from client\r\n" "Content-Type: text/plain\r\n" "Connection: close\r\n\r\n" "Invalid request. Privoxy doesn't support FTP.\r\n"; -const static char GOPHER_RESPONSE[] = - "HTTP/1.0 400 Invalid request received from browser\r\n" +static const char GOPHER_RESPONSE[] = + "HTTP/1.0 400 Invalid request received from client\r\n" "Content-Type: text/plain\r\n" "Connection: close\r\n\r\n" "Invalid request. Privoxy doesn't support gopher.\r\n"; /* XXX: should be a template */ -const static char MISSING_DESTINATION_RESPONSE[] = - "HTTP/1.0 400 Bad request received from browser\r\n" +static const char MISSING_DESTINATION_RESPONSE[] = + "HTTP/1.0 400 Bad request received from client\r\n" "Proxy-Agent: Privoxy " VERSION "\r\n" "Content-Type: text/plain\r\n" "Connection: close\r\n\r\n" "Bad request. Privoxy was unable to extract the destination.\r\n"; /* XXX: should be a template */ -const static char NO_SERVER_DATA_RESPONSE[] = +static const char NO_SERVER_DATA_RESPONSE[] = "HTTP/1.0 502 Server or forwarder response empty\r\n" "Proxy-Agent: Privoxy " VERSION "\r\n" "Content-Type: text/plain\r\n" "Connection: close\r\n\r\n" "Empty server or forwarder response.\r\n" - "The connection was closed without sending any data.\r\n"; + "The connection has been closed but Privoxy didn't receive any data.\r\n"; /* XXX: should be a template */ -const static char NULL_BYTE_RESPONSE[] = - "HTTP/1.0 400 Bad request received from browser\r\n" +static const char INVALID_SERVER_HEADERS_RESPONSE[] = + "HTTP/1.0 502 Server or forwarder response invalid\r\n" + "Proxy-Agent: Privoxy " VERSION "\r\n" + "Content-Type: text/plain\r\n" + "Connection: close\r\n\r\n" + "Bad response. The server or forwarder response doesn't look like HTTP.\r\n"; + +#if 0 +/* XXX: should be a template */ +static const char NULL_BYTE_RESPONSE[] = + "HTTP/1.0 400 Bad request received from client\r\n" "Proxy-Agent: Privoxy " VERSION "\r\n" "Content-Type: text/plain\r\n" "Connection: close\r\n\r\n" "Bad request. Null byte(s) before end of request.\r\n"; +#endif + +/* XXX: should be a template */ +static const char MESSED_UP_REQUEST_RESPONSE[] = + "HTTP/1.0 400 Malformed request after rewriting\r\n" + "Proxy-Agent: Privoxy " VERSION "\r\n" + "Content-Type: text/plain\r\n" + "Connection: close\r\n\r\n" + "Bad request. Messed up with header filters.\r\n"; + +/* A function to crunch a response */ +typedef struct http_response *(*crunch_func_ptr)(struct client_state *); + +/* Crunch function flags */ +#define CF_NO_FLAGS 0 +/* Cruncher applies to forced requests as well */ +#define CF_IGNORE_FORCE 1 +/* Crunched requests are counted for the block statistics */ +#define CF_COUNT_AS_REJECT 2 + +/* A crunch function and its flags */ +struct cruncher +{ + const crunch_func_ptr cruncher; + const int flags; +}; + +static int crunch_response_triggered(struct client_state *csp, const struct cruncher crunchers[]); + +/* Complete list of cruncher functions */ +static const struct cruncher crunchers_all[] = { + { direct_response, CF_COUNT_AS_REJECT|CF_IGNORE_FORCE}, + { block_url, CF_COUNT_AS_REJECT }, +#ifdef FEATURE_TRUST + { trust_url, CF_COUNT_AS_REJECT }, +#endif /* def FEATURE_TRUST */ + { redirect_url, CF_NO_FLAGS }, + { dispatch_cgi, CF_IGNORE_FORCE}, + { NULL, 0 } +}; + +/* Light version, used after tags are applied */ +static const struct cruncher crunchers_light[] = { + { block_url, CF_COUNT_AS_REJECT }, + { redirect_url, CF_NO_FLAGS }, + { NULL, 0 } +}; + #if !defined(_WIN32) && !defined(__OS2__) && !defined(AMIGA) /********************************************************************* @@ -1078,10 +1336,8 @@ static void sig_handler(int the_signal) * FALSE if the request doesn't look invalid. * *********************************************************************/ -int client_protocol_is_unsupported(const struct client_state *csp, char *req) +static int client_protocol_is_unsupported(const struct client_state *csp, char *req) { - char buf[BUFFER_SIZE]; - /* * If it's a FTP or gopher request, we don't support it. * @@ -1097,21 +1353,26 @@ int client_protocol_is_unsupported(const struct client_state *csp, char *req) */ if (!strncmpic(req, "GET ftp://", 10) || !strncmpic(req, "GET gopher://", 13)) { + const char *response = NULL; + const char *protocol = NULL; + if (!strncmpic(req, "GET ftp://", 10)) { - strcpy(buf, FTP_RESPONSE); - log_error(LOG_LEVEL_ERROR, "%s tried to use Privoxy as FTP proxy: %s", - csp->ip_addr_str, req); + response = FTP_RESPONSE; + protocol = "FTP"; } else { - strcpy(buf, GOPHER_RESPONSE); - log_error(LOG_LEVEL_ERROR, "%s tried to use Privoxy as gopher proxy: %s", - csp->ip_addr_str, req); + response = GOPHER_RESPONSE; + protocol = "GOPHER"; } - log_error(LOG_LEVEL_CLF, "%s - - [%T] \"%s\" 400 0", csp->ip_addr_str, req); + log_error(LOG_LEVEL_ERROR, + "%s tried to use Privoxy as %s proxy: %s", + csp->ip_addr_str, protocol, req); + log_error(LOG_LEVEL_CLF, + "%s - - [%T] \"%s\" 400 0", csp->ip_addr_str, req); freez(req); - write_socket(csp->cfd, buf, strlen(buf)); + write_socket(csp->cfd, response, strlen(response)); return TRUE; } @@ -1147,9 +1408,8 @@ int client_protocol_is_unsupported(const struct client_state *csp, char *req) * JB_ERR_PARSE if it isn't. * *********************************************************************/ -jb_err get_request_destination_elsewhere(struct client_state *csp, struct list *headers) +static jb_err get_request_destination_elsewhere(struct client_state *csp, struct list *headers) { - char buf[BUFFER_SIZE]; char *req; if (!(csp->config->feature_flags & RUNTIME_FEATURE_ACCEPT_INTERCEPTED_REQUESTS)) @@ -1161,8 +1421,7 @@ jb_err get_request_destination_elsewhere(struct client_state *csp, struct list * log_error(LOG_LEVEL_CLF, "%s - - [%T] \"%s\" 400 0", csp->ip_addr_str, csp->http->cmd); - strcpy(buf, CHEADER); - write_socket(csp->cfd, buf, strlen(buf)); + write_socket(csp->cfd, CHEADER, strlen(CHEADER)); destroy_list(headers); return JB_ERR_PARSE; @@ -1188,8 +1447,7 @@ jb_err get_request_destination_elsewhere(struct client_state *csp, struct list * csp->ip_addr_str, csp->http->cmd, req); freez(req); - strcpy(buf, MISSING_DESTINATION_RESPONSE); - write_socket(csp->cfd, buf, strlen(buf)); + write_socket(csp->cfd, MISSING_DESTINATION_RESPONSE, strlen(MISSING_DESTINATION_RESPONSE)); destroy_list(headers); return JB_ERR_PARSE; @@ -1217,12 +1475,12 @@ jb_err get_request_destination_elsewhere(struct client_state *csp, struct list * * JB_ERR_PARSE if the headers were incomplete. * *********************************************************************/ -jb_err get_server_headers(struct client_state *csp) +static jb_err get_server_headers(struct client_state *csp) { int continue_hack_in_da_house = 0; char * header; - while (((header = get_header(csp)) != NULL) || continue_hack_in_da_house) + while (((header = get_header(csp->iob)) != NULL) || continue_hack_in_da_house) { if (header == NULL) { @@ -1273,7 +1531,6 @@ jb_err get_server_headers(struct client_state *csp) return JB_ERR_PARSE; } - /* Enlist header */ if (JB_ERR_MEMORY == enlist(csp->headers, header)) { /* @@ -1303,7 +1560,7 @@ jb_err get_server_headers(struct client_state *csp) * Returns : A string with the crunch reason or an error description. * *********************************************************************/ -const char *crunch_reason(const struct http_response *rsp) +static const char *crunch_reason(const struct http_response *rsp) { char * reason = NULL; @@ -1366,7 +1623,7 @@ const char *crunch_reason(const struct http_response *rsp) * Returns : Nothing. * *********************************************************************/ -void send_crunch_response(struct client_state *csp, struct http_response *rsp) +static void send_crunch_response(struct client_state *csp, struct http_response *rsp) { const struct http_request *http = csp->http; char status_code[4]; @@ -1429,6 +1686,7 @@ void send_crunch_response(struct client_state *csp, struct http_response *rsp) } +#if 0 /********************************************************************* * * Function : request_contains_null_bytes @@ -1436,6 +1694,8 @@ void send_crunch_response(struct client_state *csp, struct http_response *rsp) * Description : Checks for NULL bytes in the request and sends * an error message to the client if any were found. * + * XXX: currently not used, see comment in chat(). + * * Parameters : * 1 : csp = Current client state (buffers, headers, etc...) * 2 : buf = Data from the client's request to check. @@ -1445,7 +1705,7 @@ void send_crunch_response(struct client_state *csp, struct http_response *rsp) * FALSE otherwise. * *********************************************************************/ -int request_contains_null_bytes(const struct client_state *csp, char *buf, int len) +static int request_contains_null_bytes(const struct client_state *csp, char *buf, int len) { size_t c_len; /* Request lenght when treated as C string */ @@ -1475,8 +1735,7 @@ int request_contains_null_bytes(const struct client_state *csp, char *buf, int l log_error(LOG_LEVEL_HEADER, "Offending request data with NULL bytes turned into \'°\' characters: %s", buf); - strcpy(buf, NULL_BYTE_RESPONSE); - write_socket(csp->cfd, buf, strlen(buf)); + write_socket(csp->cfd, NULL_BYTE_RESPONSE, strlen(NULL_BYTE_RESPONSE)); /* XXX: Log correct size */ log_error(LOG_LEVEL_CLF, "%s - - [%T] \"Invalid request\" 400 0", csp->ip_addr_str); @@ -1486,6 +1745,7 @@ int request_contains_null_bytes(const struct client_state *csp, char *buf, int l return FALSE; } +#endif /********************************************************************* @@ -1497,69 +1757,56 @@ int request_contains_null_bytes(const struct client_state *csp, char *buf, int l * * Parameters : * 1 : csp = Current client state (buffers, headers, etc...) + * 2 : crunchers = list of cruncher functions to run * * Returns : TRUE if the request was answered with a crunch response * FALSE otherwise. * *********************************************************************/ -int crunch_response_triggered(struct client_state *csp) +static int crunch_response_triggered(struct client_state *csp, const struct cruncher crunchers[]) { -/* - * This next lines are a little ugly, but they simplify the if statements - * below. Basically if TOGGLE, then we want the if to test if the - * CSP_FLAG_TOGGLED_ON flag ist set, else we don't. And if FEATURE_FORCE_LOAD, - * then we want the if to test for CSP_FLAG_FORCED, else we don't. - */ -#ifdef FEATURE_TOGGLE -# define IS_TOGGLED_ON_AND (csp->flags & CSP_FLAG_TOGGLED_ON) && -#else /* ifndef FEATURE_TOGGLE */ -# define IS_TOGGLED_ON_AND -#endif /* ndef FEATURE_TOGGLE */ -#ifdef FEATURE_FORCE_LOAD -# define IS_NOT_FORCED_AND !(csp->flags & CSP_FLAG_FORCED) && -#else /* ifndef FEATURE_FORCE_LOAD */ -# define IS_NOT_FORCED_AND -#endif /* def FEATURE_FORCE_LOAD */ - -#define IS_ENABLED_AND IS_TOGGLED_ON_AND IS_NOT_FORCED_AND - struct http_response *rsp = NULL; + const struct cruncher *c; - if ( - /* We may not forward the request by rfc2616 sect 14.31 */ - (NULL != (rsp = direct_response(csp))) - - /* or we are enabled and... */ - || (IS_ENABLED_AND ( - - /* ..the request was blocked */ - ( NULL != (rsp = block_url(csp))) - - /* ..or untrusted */ -#ifdef FEATURE_TRUST - || ( NULL != (rsp = trust_url(csp))) -#endif /* def FEATURE_TRUST */ - - /* ..or a redirect kicked in */ - || ( NULL != (rsp = redirect_url(csp))) - )) - /* - * .. or a CGI call was detected and answered. - * - * This check comes last to give the user the power - * to deny acces to some (or all) of the cgi pages. - */ - || (NULL != (rsp = dispatch_cgi(csp))) - - ) + /* + * If CGI request crunching is disabled, + * check the CGI dispatcher out of order to + * prevent unintentional blocks or redirects. + */ + if (!(csp->config->feature_flags & RUNTIME_FEATURE_CGI_CRUNCHING) + && (NULL != (rsp = dispatch_cgi(csp)))) { /* Deliver, log and free the interception response. */ send_crunch_response(csp, rsp); + return TRUE; + } + + for (c = crunchers; c->cruncher != NULL; c++) + { + /* + * Check the cruncher if either Privoxy is toggled + * on and the request isn't forced, or if the cruncher + * applies to forced requests as well. + */ + if (((csp->flags & CSP_FLAG_TOGGLED_ON) && + !(csp->flags & CSP_FLAG_FORCED)) || + (c->flags & CF_IGNORE_FORCE)) + { + rsp = c->cruncher(csp); + if (NULL != rsp) + { + /* Deliver, log and free the interception response. */ + send_crunch_response(csp, rsp); #ifdef FEATURE_STATISTICS - csp->flags |= CSP_FLAG_REJECTED; + if (c->flags & CF_COUNT_AS_REJECT) + { + csp->flags |= CSP_FLAG_REJECTED; + } #endif /* def FEATURE_STATISTICS */ - return TRUE; + return TRUE; + } + } } return FALSE; @@ -1578,11 +1825,13 @@ int crunch_response_triggered(struct client_state *csp) * Parameters : * 1 : csp = Current client state (buffers, headers, etc...) * 2 : fwd = The forwarding spec used for the request + * XXX: Should use http->fwd instead. + * 3 : request_line = The old request line which will be replaced. * * Returns : Nothing. Terminates in case of memory problems. * *********************************************************************/ -void build_request_line(struct client_state *csp, const struct forward_spec *fwd) +static void build_request_line(struct client_state *csp, const struct forward_spec *fwd, char **request_line) { struct http_request *http = csp->http; @@ -1606,30 +1855,64 @@ void build_request_line(struct client_state *csp, const struct forward_spec *fwd /* * Rebuild the request line. - * XXX: If a http forwarder is used and the HTTP version - * wasn't downgraded, we don't have to rebuild anything. */ - freez(http->cmd); - - http->cmd = strdup(http->gpc); - string_append(&http->cmd, " "); + freez(*request_line); + *request_line = strdup(http->gpc); + string_append(request_line, " "); if (fwd->forward_host) { - string_append(&http->cmd, http->url); + string_append(request_line, http->url); } else { - string_append(&http->cmd, http->path); + string_append(request_line, http->path); } - string_append(&http->cmd, " "); - string_append(&http->cmd, http->ver); + string_append(request_line, " "); + string_append(request_line, http->ver); - if (http->cmd == NULL) + if (*request_line == NULL) { log_error(LOG_LEVEL_FATAL, "Out of memory writing HTTP command"); } - log_error(LOG_LEVEL_HEADER, "New HTTP Request-Line: %s", http->cmd); + log_error(LOG_LEVEL_HEADER, "New HTTP Request-Line: %s", *request_line); +} + + +/********************************************************************* + * + * Function : change_request_destination + * + * Description : Parse a (rewritten) request line and regenerate + * the http request data. + * + * Parameters : + * 1 : csp = Current client state (buffers, headers, etc...) + * + * Returns : Forwards the parse_http_request() return code. + * Terminates in case of memory problems. + * + *********************************************************************/ +static jb_err change_request_destination(struct client_state *csp) +{ + struct http_request *http = csp->http; + jb_err err; + + log_error(LOG_LEVEL_INFO, "Rewrite detected: %s", csp->headers->first->str); + free_http_request(http); + err = parse_http_request(csp->headers->first->str, http, csp); + if (JB_ERR_OK != err) + { + log_error(LOG_LEVEL_ERROR, "Couldn't parse rewritten request: %s.", + jb_err_to_string(err)); + } + http->ocmd = strdup(http->cmd); /* XXX: ocmd is a misleading name */ + if (http->ocmd == NULL) + { + log_error(LOG_LEVEL_FATAL, "Out of memory copying rewritten HTTP request line"); + } + + return err; } @@ -1657,7 +1940,7 @@ static void chat(struct client_state *csp) char buf[BUFFER_SIZE]; char *hdr; char *p; - char *req; + char *req = NULL; fd_set rfds; int n; jb_socket maxfd; @@ -1669,17 +1952,10 @@ static void chat(struct client_state *csp) const struct forward_spec * fwd; struct http_request *http; int len; /* for buffer sizes (and negative error codes) */ -#ifdef FEATURE_KILL_POPUPS - int block_popups; /* bool, 1==will block popups */ - int block_popups_now = 0; /* bool, 1==currently blocking popups */ -#endif /* def FEATURE_KILL_POPUPS */ - - int pcrs_filter; /* bool, 1==will filter through pcrs */ - int gif_deanimate; /* bool, 1==will deanimate gifs */ - int jpeg_inspect; /* bool, 1==will inspect jpegs */ + jb_err err; /* Function that does the content filtering for the current request */ - char *(*content_filter)() = NULL; + filter_function_ptr content_filter = NULL; /* Skeleton for HTTP response, if we should intercept the request */ struct http_response *rsp; @@ -1697,18 +1973,12 @@ static void chat(struct client_state *csp) * could get blocked here if a client connected, then didn't say anything! */ - for (;;) + do { - len = read_socket(csp->cfd, buf, sizeof(buf)-1); + len = read_socket(csp->cfd, buf, sizeof(buf) - 1); if (len <= 0) break; /* error! */ - if (request_contains_null_bytes(csp, buf, len)) - { - /* NULL bytes found and dealt with, just hang up. */ - return; - } - /* * If there is no memory left for buffering the * request, there is nothing we can do but hang up @@ -1718,17 +1988,37 @@ static void chat(struct client_state *csp) return; } - req = get_header(csp); + req = get_header(csp->iob); - if (req == NULL) - { - break; /* no HTTP request! */ - } + } while ((NULL != req) && ('\0' == *req)); - if (*req == '\0') + if ((NULL != req) && ('\0' != *req)) + { + /* Request received. Validate and parse it. */ + +#if 0 + /* + * XXX: Temporary disabled to prevent problems + * with POST requests whose bodies are allowed to + * contain NULL bytes. BR#1730105. + * + * The main purpose of this check is to properly + * log stuff like BitTorrent traffic and other junk + * that hits public proxies. It's not required for + * Privoxy to functions as those requests are discarded + * later on anyway. + * + * It probably should be rewritten to only check + * the head of the request. Another option would + * be to let all POST requests pass, although that + * may not be good enough. + */ + if (request_contains_null_bytes(csp, buf, len)) { - continue; /* more to come! */ + /* NULL bytes found and dealt with, just hang up. */ + return; } +#endif /* Does the request line look invalid? */ if (client_protocol_is_unsupported(csp, req)) @@ -1762,25 +2052,18 @@ static void chat(struct client_state *csp) } #endif /* def FEATURE_FORCE_LOAD */ - - switch( parse_http_request(req, http, csp) ) + err = parse_http_request(req, http, csp); + if (JB_ERR_OK != err) { - case JB_ERR_MEMORY: - log_error(LOG_LEVEL_ERROR, "Out of memory while parsing request."); - break; - case JB_ERR_PARSE: - log_error(LOG_LEVEL_ERROR, "Couldn't parse request: %s.", req); - break; + log_error(LOG_LEVEL_ERROR, "Couldn't parse request: %s.", jb_err_to_string(err)); } freez(req); - break; } if (http->cmd == NULL) { - strcpy(buf, CHEADER); - write_socket(csp->cfd, buf, strlen(buf)); + write_socket(csp->cfd, CHEADER, strlen(CHEADER)); /* XXX: Use correct size */ log_error(LOG_LEVEL_CLF, "%s - - [%T] \"Invalid request\" 400 0", csp->ip_addr_str); log_error(LOG_LEVEL_ERROR, "Invalid header received from %s.", csp->ip_addr_str); @@ -1793,31 +2076,47 @@ static void chat(struct client_state *csp) init_list(headers); for (;;) { - if ( ( ( p = get_header(csp) ) != NULL) && ( *p == '\0' ) ) + p = get_header(csp->iob); + + if (p == NULL) { - len = read_socket(csp->cfd, buf, sizeof(buf)); + /* There are no additional headers to read. */ + break; + } + + if (*p == '\0') + { + /* + * We didn't receive a complete header + * line yet, get the rest of it. + */ + len = read_socket(csp->cfd, buf, sizeof(buf) - 1); if (len <= 0) { log_error(LOG_LEVEL_ERROR, "read from client failed: %E"); + destroy_list(headers); return; } - /* - * If there is no memory left for buffering the - * request, there is nothing we can do but hang up - */ if (add_to_iob(csp, buf, len)) { + /* + * If there is no memory left for buffering the + * request, there is nothing we can do but hang up + */ + destroy_list(headers); return; } - continue; } - - if (p == NULL) break; - - enlist(headers, p); - freez(p); - + else + { + /* + * We were able to read a complete + * header and can finaly enlist it. + */ + enlist(headers, p); + freez(p); + } } if (http->host == NULL) @@ -1837,19 +2136,76 @@ static void chat(struct client_state *csp) * An error response has already been send * and we're done here. */ - return; + return; } } - /* decide how to route the HTTP request */ + /* + * Determine the actions for this URL + */ +#ifdef FEATURE_TOGGLE + if (!(csp->flags & CSP_FLAG_TOGGLED_ON)) + { + /* Most compatible set of actions (i.e. none) */ + init_current_action(csp->action); + } + else +#endif /* ndef FEATURE_TOGGLE */ + { + get_url_actions(csp, http); + } + + /* + * Save a copy of the original request for logging + */ + http->ocmd = strdup(http->cmd); + + if (http->ocmd == NULL) + { + log_error(LOG_LEVEL_FATAL, "Out of memory copying HTTP request line"); + } + + enlist(csp->headers, http->cmd); + + /* Append the previously read headers */ + list_append_list_unique(csp->headers, headers); + destroy_list(headers); + + err = sed(client_patterns, add_client_headers, csp); + if (JB_ERR_OK != err) + { + assert(err == JB_ERR_PARSE); + log_error(LOG_LEVEL_FATAL, "Failed to parse client headers"); + } + csp->flags |= CSP_FLAG_CLIENT_HEADER_PARSING_DONE; - if ((fwd = forward_url(http, csp)) == NULL) + /* Check request line for rewrites. */ + if ((NULL == csp->headers->first->str) + || (strcmp(http->cmd, csp->headers->first->str) && + (JB_ERR_OK != change_request_destination(csp)))) + { + /* + * A header filter broke the request line - bail out. + */ + write_socket(csp->cfd, MESSED_UP_REQUEST_RESPONSE, strlen(MESSED_UP_REQUEST_RESPONSE)); + /* XXX: Use correct size */ + log_error(LOG_LEVEL_CLF, "%s - - [%T] \"Invalid request generated\" 500 0", csp->ip_addr_str); + log_error(LOG_LEVEL_ERROR, "Invalid request line after applying header filters."); + + free_http_request(http); + return; + } + + /* decide how to route the HTTP request */ + fwd = forward_url(csp, http); + if (NULL == fwd) { log_error(LOG_LEVEL_FATAL, "gateway spec is NULL!?!? This can't happen!"); /* Never get here - LOG_LEVEL_FATAL causes program exit */ } - /* build the http request to send to the server + /* + * build the http request to send to the server * we have to do one of the following: * * create = use the original HTTP request to create a new @@ -1885,112 +2241,35 @@ static void chat(struct client_state *csp) * */ - /* - * Determine the actions for this URL - */ -#ifdef FEATURE_TOGGLE - if (!(csp->flags & CSP_FLAG_TOGGLED_ON)) + if (http->ssl && connect_port_is_forbidden(csp)) { - /* Most compatible set of actions (i.e. none) */ - init_current_action(csp->action); + const char *acceptable_connect_ports = + csp->action->string[ACTION_STRING_LIMIT_CONNECT]; + assert(NULL != acceptable_connect_ports); + log_error(LOG_LEVEL_INFO, "Request from %s marked for blocking. " + "limit-connect{%s} doesn't allow CONNECT requests to port %d.", + csp->ip_addr_str, acceptable_connect_ports, csp->http->port); + csp->action->flags |= ACTION_BLOCK; + http->ssl = 0; } - else -#endif /* ndef FEATURE_TOGGLE */ - { - url_actions(http, csp); - } - - - /* - * Check if a CONNECT request is allowable: - * In the absence of a +limit-connect action, allow only port 443. - * If there is an action, allow whatever matches the specificaton. - */ - if(http->ssl) - { - if( ( !(csp->action->flags & ACTION_LIMIT_CONNECT) && csp->http->port != 443) - || (csp->action->flags & ACTION_LIMIT_CONNECT - && !match_portlist(csp->action->string[ACTION_STRING_LIMIT_CONNECT], csp->http->port)) ) - { - if (csp->action->flags & ACTION_TREAT_FORBIDDEN_CONNECTS_LIKE_BLOCKS) - { - /* - * The response may confuse some clients, - * but makes unblocking easier. - */ - log_error(LOG_LEVEL_ERROR, "Marking suspicious CONNECT request from %s for blocking.", - csp->ip_addr_str); - csp->action->flags |= ACTION_BLOCK; - http->ssl = 0; - } - else - { - strcpy(buf, CFORBIDDEN); - write_socket(csp->cfd, buf, strlen(buf)); - log_error(LOG_LEVEL_CONNECT, "Denying suspicious CONNECT request from %s", csp->ip_addr_str); - log_error(LOG_LEVEL_CLF, "%s - - [%T] \" \" 403 0", csp->ip_addr_str); - return; - } - } - } - - - /* - * Save a copy of the original request for logging - */ - http->ocmd = strdup(http->cmd); - if (http->ocmd == NULL) - { - log_error(LOG_LEVEL_FATAL, "Out of memory copying HTTP request line"); - } - - /* - * (Re)build the HTTP request for non-SSL requests. - */ if (http->ssl == 0) { - build_request_line(csp, fwd); + freez(csp->headers->first->str); + build_request_line(csp, fwd, &csp->headers->first->str); } - enlist(csp->headers, http->cmd); - /* Append the previously read headers */ - list_append_list_unique(csp->headers, headers); - destroy_list(headers); - - /* - * If the user has not supplied any wafers, and the user has not - * told us to suppress the vanilla wafer, then send the vanilla wafer. - */ - if (list_is_empty(csp->action->multi[ACTION_MULTI_WAFER]) - && ((csp->action->flags & ACTION_VANILLA_WAFER) != 0)) - { - enlist(csp->action->multi[ACTION_MULTI_WAFER], VANILLA_WAFER); - } - - hdr = sed(client_patterns, add_client_headers, csp); + hdr = list_to_text(csp->headers); if (hdr == NULL) { /* FIXME Should handle error properly */ log_error(LOG_LEVEL_FATAL, "Out of memory parsing client header"); } - csp->flags |= CSP_FLAG_CLIENT_HEADER_PARSING_DONE; - -#ifdef FEATURE_KILL_POPUPS - block_popups = ((csp->action->flags & ACTION_NO_POPUPS) != 0); -#endif /* def FEATURE_KILL_POPUPS */ - - pcrs_filter = (csp->rlist != NULL) && /* There are expressions to be used */ - (!list_is_empty(csp->action->multi[ACTION_MULTI_FILTER])); - - gif_deanimate = ((csp->action->flags & ACTION_DEANIMATE) != 0); - - jpeg_inspect = ((csp->action->flags & ACTION_JPEG_INSPECT) != 0); /* * We have a request. Check if one of the crunchers wants it. */ - if (crunch_response_triggered(csp)) + if (crunch_response_triggered(csp, crunchers_all)) { /* * Yes. The client got the crunch response @@ -2053,7 +2332,7 @@ static void chat(struct client_state *csp) /* Write the answer to the client */ - if(rsp != NULL) + if (rsp != NULL) { send_crunch_response(csp, rsp); } @@ -2069,7 +2348,7 @@ static void chat(struct client_state *csp) */ if (write_socket(csp->sfd, hdr, strlen(hdr)) - || (flush_socket(csp->sfd, csp) < 0)) + || (flush_socket(csp->sfd, csp->iob) < 0)) { log_error(LOG_LEVEL_CONNECT, "write header to: %s failed: %E", http->hostport); @@ -2092,7 +2371,7 @@ static void chat(struct client_state *csp) * so just send the "connect succeeded" message to the * client, flush the rest, and get out of the way. */ - if (write_socket(csp->cfd, CSUCCEED, sizeof(CSUCCEED)-1)) + if (write_socket(csp->cfd, CSUCCEED, strlen(CSUCCEED))) { freez(hdr); return; @@ -2141,7 +2420,7 @@ static void chat(struct client_state *csp) if (FD_ISSET(csp->cfd, &rfds)) { - len = read_socket(csp->cfd, buf, sizeof(buf)); + len = read_socket(csp->cfd, buf, sizeof(buf) - 1); if (len <= 0) { @@ -2183,10 +2462,23 @@ static void chat(struct client_state *csp) "CONNECT already confirmed. Unable to tell the client about the problem."); return; } + else if (byte_count) + { + /* + * Just hang up. We already transmitted the original headers + * and parts of the original content and therefore missed the + * chance to send an error message (without risking data corruption). + * + * XXX: we could retry with a fancy range request here. + */ + log_error(LOG_LEVEL_ERROR, "Already forwarded the original headers. " + "Unable to tell the client about the problem."); + return; + } rsp = error_response(csp, "connect-failed", errno); - if(rsp) + if (rsp) { send_crunch_response(csp, rsp); } @@ -2199,14 +2491,6 @@ static void chat(struct client_state *csp) */ buf[len] = '\0'; -#ifdef FEATURE_KILL_POPUPS - /* Filter the popups on this read. */ - if (block_popups_now) - { - filter_popups(buf, csp); - } -#endif /* def FEATURE_KILL_POPUPS */ - /* Normally, this would indicate that we've read * as much as the server has sent us and we can * close the client connection. However, Microsoft @@ -2237,24 +2521,35 @@ static void chat(struct client_state *csp) */ if (content_filter) { + p = execute_content_filter(csp, content_filter); /* * If the content filter fails, use the original * buffer and length. * (see p != NULL ? p : csp->iob->cur below) */ - if (NULL == (p = (*content_filter)(csp))) + if (NULL == p) { csp->content_length = (size_t)(csp->iob->eod - csp->iob->cur); } - hdr = sed(server_patterns_light, NULL, csp); + if (JB_ERR_OK != sed(server_patterns_light, NULL, csp)) + { + log_error(LOG_LEVEL_FATAL, "Failed to parse server headers."); + } + hdr = list_to_text(csp->headers); if (hdr == NULL) { /* FIXME Should handle error properly */ log_error(LOG_LEVEL_FATAL, "Out of memory parsing server header"); } + /* + * Shouldn't happen because this was the second sed run + * and tags are only created for the first one. + */ + assert(!crunch_response_triggered(csp, crunchers_all)); + if (write_socket(csp->cfd, hdr, strlen(hdr)) || write_socket(csp->cfd, p != NULL ? p : csp->iob->cur, csp->content_length)) { @@ -2306,9 +2601,9 @@ static void chat(struct client_state *csp) size_t hdrlen; int flushed; - log_error(LOG_LEVEL_ERROR, "Flushing header and buffers. Stepping back from filtering."); + log_error(LOG_LEVEL_INFO, "Flushing header and buffers. Stepping back from filtering."); - hdr = sed(server_patterns, add_server_headers, csp); + hdr = list_to_text(csp->headers); if (hdr == NULL) { /* @@ -2321,11 +2616,10 @@ static void chat(struct client_state *csp) return; } - hdrlen = strlen(hdr); if (write_socket(csp->cfd, hdr, hdrlen) - || ((flushed = flush_socket(csp->cfd, csp)) < 0) + || ((flushed = flush_socket(csp->cfd, csp->iob)) < 0) || (write_socket(csp->cfd, buf, (size_t)len))) { log_error(LOG_LEVEL_CONNECT, "Flush header and buffers to client failed: %E"); @@ -2334,11 +2628,15 @@ static void chat(struct client_state *csp) return; } - byte_count += hdrlen + (size_t)flushed + (size_t)len; + /* + * Reset the byte_count to the amount of bytes + * we just flushed. len will be added a few lines below, + * hdrlen doesn't matter for LOG_LEVEL_CLF. + */ + byte_count = (size_t)flushed; freez(hdr); content_filter = NULL; server_body = 1; - } } else @@ -2384,6 +2682,8 @@ static void chat(struct client_state *csp) * and there isn't anything * we can do about it. */ + log_error(LOG_LEVEL_INFO, + "MS IIS5 hack didn't produce valid headers."); break; } else @@ -2402,8 +2702,30 @@ static void chat(struct client_state *csp) { log_error(LOG_LEVEL_ERROR, "Empty server or forwarder response."); log_error(LOG_LEVEL_CLF, "%s - - [%T] \"%s\" 502 0", csp->ip_addr_str, http->cmd); - strcpy(buf, NO_SERVER_DATA_RESPONSE); - write_socket(csp->cfd, buf, strlen(buf)); + write_socket(csp->cfd, NO_SERVER_DATA_RESPONSE, strlen(NO_SERVER_DATA_RESPONSE)); + free_http_request(http); + return; + } + + assert(csp->headers->first->str); + assert(!http->ssl); + if (strncmpic(csp->headers->first->str, "HTTP", 4)) + { + /* + * It doesn't look like a HTTP response: + * tell the client and log the problem. + */ + if (strlen(csp->headers->first->str) > 30) + { + csp->headers->first->str[30] = '\0'; + } + log_error(LOG_LEVEL_ERROR, + "Invalid server or forwarder response. Starts with: %s", + csp->headers->first->str); + log_error(LOG_LEVEL_CLF, + "%s - - [%T] \"%s\" 502 0", csp->ip_addr_str, http->cmd); + write_socket(csp->cfd, INVALID_SERVER_HEADERS_RESPONSE, + strlen(INVALID_SERVER_HEADERS_RESPONSE)); free_http_request(http); return; } @@ -2411,58 +2733,34 @@ static void chat(struct client_state *csp) /* we have now received the entire header. * filter it and send the result to the client */ - - hdr = sed(server_patterns, add_server_headers, csp); + if (JB_ERR_OK != sed(server_patterns, add_server_headers, csp)) + { + log_error(LOG_LEVEL_FATAL, "Failed to parse server headers."); + } + hdr = list_to_text(csp->headers); if (hdr == NULL) { /* FIXME Should handle error properly */ log_error(LOG_LEVEL_FATAL, "Out of memory parsing server header"); } -#ifdef FEATURE_KILL_POPUPS - /* Start blocking popups if appropriate. */ - - if ((csp->content_type & CT_TEXT) && /* It's a text / * MIME-Type */ - !http->ssl && /* We talk plaintext */ - block_popups) /* Policy allows */ + if (crunch_response_triggered(csp, crunchers_light)) { - block_popups_now = 1; /* - * Filter the part of the body that came in the same read - * as the last headers: + * One of the tags created by a server-header + * tagger triggered a crunch. We already + * delivered the crunch response to the client + * and are done here after cleaning up. */ - filter_popups(csp->iob->cur, csp); + freez(hdr); + return; } - -#endif /* def FEATURE_KILL_POPUPS */ - /* Buffer and pcrs filter this if appropriate. */ - if ((csp->content_type & CT_TEXT) && /* It's a text / * MIME-Type */ - !http->ssl && /* We talk plaintext */ - pcrs_filter) /* Policy allows */ - { - content_filter = pcrs_filter_response; - } - - /* Buffer and gif_deanimate this if appropriate. */ - - if ((csp->content_type & CT_GIF) && /* It's an image/gif MIME-Type */ - !http->ssl && /* We talk plaintext */ - gif_deanimate) /* Policy allows */ - { - content_filter = gif_deanimate_response; - } - - /* Buffer and jpg_inspect this if appropriate. */ - - if ((csp->content_type & CT_JPEG) && /* It's an image/jpeg MIME-Type */ - !http->ssl && /* We talk plaintext */ - jpeg_inspect) /* Policy allows */ + if (!http->ssl) /* We talk plaintext */ { - content_filter = jpeg_inspect_response; + content_filter = get_filter_function(csp); } - /* * Only write if we're not buffering for content modification */ @@ -2474,7 +2772,7 @@ static void chat(struct client_state *csp) */ if (write_socket(csp->cfd, hdr, strlen(hdr)) - || ((len = flush_socket(csp->cfd, csp)) < 0)) + || ((len = flush_socket(csp->cfd, csp->iob)) < 0)) { log_error(LOG_LEVEL_CONNECT, "write header to client failed: %E"); @@ -2501,6 +2799,8 @@ static void chat(struct client_state *csp) */ if (ms_iis5_hack) { + log_error(LOG_LEVEL_INFO, + "Closed server connection detected with MS IIS5 hack enabled."); break; } } @@ -2579,6 +2879,7 @@ static int32 server_thread(void *data) #endif +#if !defined(_WIN32) || defined(_WIN_CONSOLE) /********************************************************************* * * Function : usage @@ -2590,19 +2891,24 @@ static int32 server_thread(void *data) * Returns : No. ,-) * *********************************************************************/ -void usage(const char *myname) +static void usage(const char *myname) { printf("Privoxy version " VERSION " (" HOME_PAGE_URL ")\n" -#if !defined(unix) - "Usage: %s [--help] [--version] [configfile]\n" -#else - "Usage: %s [--help] [--version] [--no-daemon] [--pidfile pidfile] [--user user[.group]] [configfile]\n" -#endif - "Aborting.\n", myname); - + "Usage: %s " +#if defined(unix) + "[--chroot] " +#endif /* defined(unix) */ + "[--help] " +#if defined(unix) + "[--no-daemon] [--pidfile pidfile] [--pre-chroot-nslookup hostname] [--user user[.group]] " +#endif /* defined(unix) */ + "[--version] [configfile]\n" + "Aborting\n", myname); + exit(2); } +#endif /* #if !defined(_WIN32) || defined(_WIN_CONSOLE) */ /********************************************************************* @@ -2616,7 +2922,7 @@ void usage(const char *myname) * Returns : Void, exits in case of errors. * *********************************************************************/ -void initialize_mutexes() +static void initialize_mutexes(void) { int err = 0; @@ -2634,7 +2940,14 @@ void initialize_mutexes() * * For example older FreeBSD versions (< 6.x?) * have no gethostbyname_r, but gethostbyname is - * thead safe. + * thread safe. + */ +#if !defined(HAVE_GETHOSTBYADDR_R) || !defined(HAVE_GETHOSTBYNAME_R) + if (!err) err = pthread_mutex_init(&resolver_mutex, 0); +#endif /* !defined(HAVE_GETHOSTBYADDR_R) || !defined(HAVE_GETHOSTBYNAME_R) */ + /* + * XXX: should we use a single mutex for + * localtime() and gmtime() as well? */ #ifndef HAVE_GMTIME_R if (!err) err = pthread_mutex_init(&gmtime_mutex, 0); @@ -2644,14 +2957,6 @@ void initialize_mutexes() if (!err) err = pthread_mutex_init(&localtime_mutex, 0); #endif /* ndef HAVE_GMTIME_R */ -#ifndef HAVE_GETHOSTBYADDR_R - if (!err) err = pthread_mutex_init(&gethostbyaddr_mutex, 0); -#endif /* ndef HAVE_GETHOSTBYADDR_R */ - -#ifndef HAVE_GETHOSTBYNAME_R - if (!err) err = pthread_mutex_init(&gethostbyname_mutex, 0); -#endif /* ndef HAVE_GETHOSTBYNAME_R */ - #ifndef HAVE_RANDOM if (!err) err = pthread_mutex_init(&rand_mutex, 0); #endif /* ndef HAVE_RANDOM */ @@ -2708,6 +3013,7 @@ int main(int argc, const char *argv[]) struct group *grp = NULL; char *p; int do_chroot = 0; + char *pre_chroot_nslookup_to_load_resolver = NULL; #endif Argc = argc; @@ -2723,6 +3029,9 @@ int main(int argc, const char *argv[]) /* * Parse the command line arguments + * + * XXX: simply printing usage information in case of + * invalid arguments isn't particularly user friendly. */ while (++argc_pos < argc) { @@ -2799,12 +3108,29 @@ int main(int argc, const char *argv[]) if (p != NULL) *--p = '\0'; } + else if (strcmp(argv[argc_pos], "--pre-chroot-nslookup" ) == 0) + { + if (++argc_pos == argc) usage(argv[0]); + pre_chroot_nslookup_to_load_resolver = strdup(argv[argc_pos]); + } + else if (strcmp(argv[argc_pos], "--chroot" ) == 0) { do_chroot = 1; } - #endif /* defined(unix) */ + + else if (argc_pos + 1 != argc) + { + /* + * This is neither the last command line + * option, nor was it recognized before, + * therefore it must be invalid. + */ + usage(argv[0]); + } + else + #endif /* defined(_WIN32) && !defined(_WIN_CONSOLE) */ { configfile = argv[argc_pos]; @@ -2815,24 +3141,30 @@ int main(int argc, const char *argv[]) #if defined(unix) if ( *configfile != '/' ) { - char *abs_file, cwd[1024]; + char cwd[BUFFER_SIZE]; + char *abs_file; + size_t abs_file_size; /* make config-filename absolute here */ - if ( !(getcwd(cwd, sizeof(cwd)))) + if (NULL == getcwd(cwd, sizeof(cwd))) { - perror("get working dir failed"); + perror("failed to get current working directory"); exit( 1 ); } - if (!(basedir = strdup(cwd)) - || (!(abs_file = malloc( strlen( basedir ) + strlen( configfile ) + 5 )))) + /* XXX: why + 5? */ + abs_file_size = strlen(cwd) + strlen(configfile) + 5; + basedir = strdup(cwd); + + if (NULL == basedir || + NULL == (abs_file = malloc(abs_file_size))) { perror("malloc failed"); exit( 1 ); } - strcpy( abs_file, basedir ); - strcat( abs_file, "/" ); - strcat( abs_file, configfile ); + strlcpy(abs_file, basedir, abs_file_size); + strlcat(abs_file, "/", abs_file_size ); + strlcat(abs_file, configfile, abs_file_size); configfile = abs_file; } #endif /* defined unix */ @@ -2841,6 +3173,7 @@ int main(int argc, const char *argv[]) files->next = NULL; clients->next = NULL; + /* XXX: factor out initialising after the next stable release. */ #ifdef AMIGA InitAmiga(); #elif defined(_WIN32) @@ -2850,9 +3183,21 @@ int main(int argc, const char *argv[]) /* Prepare mutexes if supported and necessary. */ initialize_mutexes(); + /* Enable logging until further notice. */ + init_log_module(Argv[0]); + random_seed = (unsigned int)time(NULL); #ifdef HAVE_RANDOM srandom(random_seed); +#elif defined (_WIN32) + /* + * See pick_from_range() in miscutil.c for details. + */ + log_error(LOG_LEVEL_INFO, + "No thread-safe PRNG implemented for your platform. " + "Using weak \'randomization\' factor which will " + "limit the already questionable usefulness of " + "header-time-randomizing actions (disabled by default)."); #else srand(random_seed); #endif /* ifdef HAVE_RANDOM */ @@ -2960,10 +3305,9 @@ int main(int argc, const char *argv[]) close ( fd ); } #endif /* 1 */ - /* FIXME: should close stderr (fd 2) here too, but the test - * for existence - * and load config file is done in listen_loop() and puts - * some messages on stderr there. + /* + * stderr (fd 2) will be closed later on, when the + * log file has been parsed. */ close( 0 ); @@ -2990,6 +3334,14 @@ int main(int argc, const char *argv[]) { log_error(LOG_LEVEL_FATAL, "Home directory for %s undefined", pw->pw_name); } + /* Read the time zone file from /etc before doing chroot. */ + tzset(); + if (NULL != pre_chroot_nslookup_to_load_resolver + && '\0' != pre_chroot_nslookup_to_load_resolver[0]) + { + /* Initialize resolver library. */ + (void) resolve_hostname_to_ip(pre_chroot_nslookup_to_load_resolver); + } if (chroot(pw->pw_dir) < 0) { log_error(LOG_LEVEL_FATAL, "Cannot chroot to %s", pw->pw_dir); @@ -3007,13 +3359,13 @@ int main(int argc, const char *argv[]) { char putenv_dummy[64]; - strcpy(putenv_dummy, "HOME=/"); + strlcpy(putenv_dummy, "HOME=/", sizeof(putenv_dummy)); if (putenv(putenv_dummy) != 0) { log_error(LOG_LEVEL_FATAL, "Cannot putenv(): HOME"); } - snprintf(putenv_dummy, 64, "USER=%s", pw->pw_name); + snprintf(putenv_dummy, sizeof(putenv_dummy), "USER=%s", pw->pw_name); if (putenv(putenv_dummy) != 0) { log_error(LOG_LEVEL_FATAL, "Cannot putenv(): USER"); @@ -3085,16 +3437,7 @@ static jb_socket bind_port_helper(struct configuration_spec * config) int result; jb_socket bfd; - if ( (config->haddr != NULL) - && (config->haddr[0] == '1') - && (config->haddr[1] == '2') - && (config->haddr[2] == '7') - && (config->haddr[3] == '.') ) - { - log_error(LOG_LEVEL_INFO, "Listening on port %d for local connections only", - config->hport); - } - else if (config->haddr == NULL) + if (config->haddr == NULL) { log_error(LOG_LEVEL_INFO, "Listening on port %d on all IP addresses", config->hport); @@ -3194,7 +3537,7 @@ static void listen_loop(void) */ if (received_hup_signal) { - init_error_log(Argv[0], config->logfile, config->debug); + init_error_log(Argv[0], config->logfile); received_hup_signal = 0; } #endif @@ -3269,10 +3612,10 @@ static void listen_loop(void) #ifdef FEATURE_TOGGLE if (global_toggle_state) +#endif /* def FEATURE_TOGGLE */ { csp->flags |= CSP_FLAG_TOGGLED_ON; } -#endif /* def FEATURE_TOGGLE */ if (run_loader(csp)) {