You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by rb...@apache.org on 2002/10/15 19:46:13 UTC
Final patch for a long time.
The recent conversations on this list have made me finally realize that I
have been here too long. I need a project that is not the Apache web
server. So, this is my good-bye. I will be unsubscribing from the Apache
web server development lists in the next day or two. I will still be
involved in APR development work, so I will still be reachable at
rbb@apache.org.
There are two projects that I started that have not been finished. I
would hope that one of two things would happen to both of them. Either
somebody else should pick them up and run with them, or they should be
removed from the server.
The first project is the Perchild MPM. It basically works, but there are
bugs.
The second is SSL upgrade. I have the patches, they haven't been
committed yet. I have attached them at the bottom of this message. The
reason they haven't been committed, is that I don't have a client to test
them with, and I haven't had time to create one. The responses are
correct I have checked them in plain text. The place that bugs most
likely exist, is the actual upgrade code that does the handshake. This is
an important feature, and I would really like to see it in 2.0.
It has been a lot of fun the last four years working on Apache, but I have
been here too long, and it isn't fun anymore. It isn't worth doing if it
isn't fun.
Ryan
_______________________________________________________________________________
Ryan Bloom rbb@apache.org
550 Jean St
Oakland CA 94610
-------------------------------------------------------------------------------
? build.err
? build.log
? output.log
? sslupgrade.patch
? modules/new
? srclib/apr/APRVARS
? srclib/apr/build.err
? srclib/apr/build.log
? srclib/apr/newpoll2.tar.gz
? srclib/apr/i18n/unix/Makefile
? srclib/apr/test/build.err
? srclib/apr/test/build.log
? srclib/apr/test/garg
? srclib/apr/test/testall
Index: modules/ssl/mod_ssl.c
===================================================================
RCS file: /home/cvs/httpd-2.0/modules/ssl/mod_ssl.c,v
retrieving revision 1.72
diff -u -d -b -w -r1.72 mod_ssl.c
--- modules/ssl/mod_ssl.c 14 Oct 2002 04:15:57 -0000 1.72
+++ modules/ssl/mod_ssl.c 15 Oct 2002 16:49:07 -0000
@@ -105,7 +105,7 @@
/*
* Per-server context configuration directives
*/
- SSL_CMD_SRV(Engine, FLAG,
+ SSL_CMD_SRV(Engine, TAKE1,
"SSL switch for the protocol engine "
"(`on', `off')")
SSL_CMD_ALL(CipherSuite, TAKE1,
@@ -274,7 +274,7 @@
return 1;
}
-static int ssl_hook_pre_connection(conn_rec *c, void *csd)
+int ssl_init_ssl_connection(conn_rec *c)
{
SSLSrvConfigRec *sc = mySrvConfig(c->base_server);
SSL *ssl;
@@ -283,40 +283,14 @@
modssl_ctx_t *mctx;
/*
- * Immediately stop processing if SSL is disabled for this connection
+ * Seed the Pseudo Random Number Generator (PRNG)
*/
- if (!(sc && (sc->enabled ||
- (sslconn && sslconn->is_proxy))))
- {
- return DECLINED;
- }
+ ssl_rand_seed(c->base_server, c->pool, SSL_RSCTX_CONNECT, "");
- /*
- * Create SSL context
- */
if (!sslconn) {
sslconn = ssl_init_connection_ctx(c);
}
- if (sslconn->disabled) {
- return DECLINED;
- }
-
- /*
- * Remember the connection information for
- * later access inside callback functions
- */
-
- ap_log_error(APLOG_MARK, APLOG_INFO, 0, c->base_server,
- "Connection to child %ld established "
- "(server %s, client %s)", c->id, sc->vhost_id,
- c->remote_ip ? c->remote_ip : "unknown");
-
- /*
- * Seed the Pseudo Random Number Generator (PRNG)
- */
- ssl_rand_seed(c->base_server, c->pool, SSL_RSCTX_CONNECT, "");
-
mctx = sslconn->is_proxy ? sc->proxy : sc->server;
/*
@@ -368,6 +342,44 @@
return APR_SUCCESS;
}
+static int ssl_hook_pre_connection(conn_rec *c, void *csd)
+{
+ SSLSrvConfigRec *sc = mySrvConfig(c->base_server);
+ SSLConnRec *sslconn = myConnConfig(c);
+
+ /*
+ * Immediately stop processing if SSL is disabled for this connection
+ */
+ if (!(sc && (sc->enabled == TRUE ||
+ (sslconn && sslconn->is_proxy))))
+ {
+ return DECLINED;
+ }
+
+ /*
+ * Create SSL context
+ */
+ if (!sslconn) {
+ sslconn = ssl_init_connection_ctx(c);
+ }
+
+ if (sslconn->disabled) {
+ return DECLINED;
+ }
+
+ /*
+ * Remember the connection information for
+ * later access inside callback functions
+ */
+
+ ap_log_error(APLOG_MARK, APLOG_INFO, 0, c->base_server,
+ "Connection to child %ld established "
+ "(server %s, client %s)", c->id, sc->vhost_id,
+ c->remote_ip ? c->remote_ip : "unknown");
+
+ return ssl_init_ssl_connection(c);
+}
+
static apr_status_t ssl_abort(SSLFilterRec *filter, conn_rec *c)
{
SSLConnRec *sslconn = myConnConfig(c);
@@ -572,6 +584,15 @@
return 443;
}
+static void ssl_hook_Insert_Filter(request_rec *r)
+{
+ SSLSrvConfigRec *sc = mySrvConfig(r->server);
+
+ if (sc->enabled == UNSET) {
+ ap_add_output_filter("UPGRADE_FILTER", NULL, r, r->connection);
+ }
+}
+
/*
* the module registration phase
*/
@@ -592,6 +613,8 @@
ap_hook_access_checker(ssl_hook_Access, NULL,NULL, APR_HOOK_MIDDLE);
ap_hook_auth_checker (ssl_hook_Auth, NULL,NULL, APR_HOOK_MIDDLE);
ap_hook_post_read_request(ssl_hook_ReadReq, NULL,NULL, APR_HOOK_MIDDLE);
+ ap_hook_insert_filter (ssl_hook_Insert_Filter, NULL,NULL, APR_HOOK_MIDDLE);
+/* ap_hook_handler (ssl_hook_Upgrade, NULL,NULL, APR_HOOK_MIDDLE); */
ssl_var_register();
Index: modules/ssl/mod_ssl.h
===================================================================
RCS file: /home/cvs/httpd-2.0/modules/ssl/mod_ssl.h,v
retrieving revision 1.121
diff -u -d -b -w -r1.121 mod_ssl.h
--- modules/ssl/mod_ssl.h 14 Oct 2002 04:15:58 -0000 1.121
+++ modules/ssl/mod_ssl.h 15 Oct 2002 16:49:07 -0000
@@ -549,7 +549,7 @@
const char *ssl_cmd_SSLPassPhraseDialog(cmd_parms *, void *, const char *);
const char *ssl_cmd_SSLCryptoDevice(cmd_parms *, void *, const char *);
const char *ssl_cmd_SSLRandomSeed(cmd_parms *, void *, const char *, const char *, const char *);
-const char *ssl_cmd_SSLEngine(cmd_parms *, void *, int);
+const char *ssl_cmd_SSLEngine(cmd_parms *, void *, const char *);
const char *ssl_cmd_SSLCipherSuite(cmd_parms *, void *, const char *);
const char *ssl_cmd_SSLCertificateFile(cmd_parms *, void *, const char *);
const char *ssl_cmd_SSLCertificateKeyFile(cmd_parms *, void *, const char *);
@@ -601,6 +601,7 @@
int ssl_hook_Fixup(request_rec *);
int ssl_hook_ReadReq(request_rec *);
int ssl_hook_Handler(request_rec *);
+int ssl_hook_Upgrade(request_rec *);
/* OpenSSL callbacks */
RSA *ssl_callback_TmpRSA(SSL *, int, int);
@@ -722,6 +723,8 @@
char *ssl_util_algotypestr(ssl_algo_t);
char *ssl_util_ptxtsub(apr_pool_t *, const char *, const char *, char *);
void ssl_util_thread_setup(apr_pool_t *);
+int ssl_init_ssl_connection(conn_rec *c);
+
#define APR_SHM_MAXSIZE (64 * 1024 * 1024)
#endif /* __MOD_SSL_H__ */
Index: modules/ssl/ssl_engine_config.c
===================================================================
RCS file: /home/cvs/httpd-2.0/modules/ssl/ssl_engine_config.c,v
retrieving revision 1.69
diff -u -d -b -w -r1.69 ssl_engine_config.c
--- modules/ssl/ssl_engine_config.c 14 Oct 2002 04:15:58 -0000 1.69
+++ modules/ssl/ssl_engine_config.c 15 Oct 2002 16:49:07 -0000
@@ -205,7 +205,7 @@
SSLSrvConfigRec *sc = apr_palloc(p, sizeof(*sc));
sc->mc = NULL;
- sc->enabled = UNSET;
+ sc->enabled = FALSE;
sc->proxy_enabled = UNSET;
sc->vhost_id = NULL; /* set during module init */
sc->vhost_id_len = 0; /* set during module init */
@@ -581,13 +581,24 @@
return NULL;
}
-const char *ssl_cmd_SSLEngine(cmd_parms *cmd, void *dcfg, int flag)
+const char *ssl_cmd_SSLEngine(cmd_parms *cmd, void *dcfg, const char *arg)
{
SSLSrvConfigRec *sc = mySrvConfig(cmd->server);
- sc->enabled = flag ? TRUE : FALSE;
-
+ if (!strcasecmp(arg, "On")) {
+ sc->enabled = TRUE;
return NULL;
+ }
+ else if (!strcasecmp(arg, "Off")) {
+ sc->enabled = FALSE;
+ return NULL;
+ }
+ else if (!strcasecmp(arg, "Optional")) {
+ sc->enabled = UNSET;
+ return NULL;
+ }
+
+ return "Argument must be On, Off, or Optional";
}
const char *ssl_cmd_SSLCipherSuite(cmd_parms *cmd,
Index: modules/ssl/ssl_engine_init.c
===================================================================
RCS file: /home/cvs/httpd-2.0/modules/ssl/ssl_engine_init.c,v
retrieving revision 1.104
diff -u -d -b -w -r1.104 ssl_engine_init.c
--- modules/ssl/ssl_engine_init.c 14 Oct 2002 04:15:58 -0000 1.104
+++ modules/ssl/ssl_engine_init.c 15 Oct 2002 16:49:07 -0000
@@ -247,11 +247,13 @@
sc->vhost_id = ssl_util_vhostid(p, s);
sc->vhost_id_len = strlen(sc->vhost_id);
+#if 0
+ /* If sc->enabled is UNSET, then SSL is optional on this vhost */
/* Fix up stuff that may not have been set */
if (sc->enabled == UNSET) {
sc->enabled = FALSE;
}
-
+#endif
if (sc->proxy_enabled == UNSET) {
sc->proxy_enabled = FALSE;
}
@@ -981,6 +983,9 @@
apr_pool_t *ptemp,
SSLSrvConfigRec *sc)
{
+ /* A bit of a hack, but initialize the server if SSL is optional or
+ * not.
+ */
if (sc->enabled) {
ap_log_error(APLOG_MARK, APLOG_INFO, 0, s,
"Configuring server for SSL protocol");
@@ -1009,7 +1014,7 @@
for (s = base_server; s; s = s->next) {
sc = mySrvConfig(s);
- if (sc->enabled && (s->port == DEFAULT_HTTP_PORT)) {
+ if ((sc->enabled == TRUE) && (s->port == DEFAULT_HTTP_PORT)) {
ap_log_error(APLOG_MARK, APLOG_WARNING, 0,
base_server,
"Init: (%s) You configured HTTPS(%d) "
Index: modules/ssl/ssl_engine_io.c
===================================================================
RCS file: /home/cvs/httpd-2.0/modules/ssl/ssl_engine_io.c,v
retrieving revision 1.81
diff -u -d -b -w -r1.81 ssl_engine_io.c
--- modules/ssl/ssl_engine_io.c 14 Oct 2002 04:15:58 -0000 1.81
+++ modules/ssl/ssl_engine_io.c 15 Oct 2002 16:49:07 -0000
@@ -577,6 +577,85 @@
return APR_SUCCESS;
}
+static apr_status_t ssl_io_filter_Upgrade(ap_filter_t *f,
+ apr_bucket_brigade *bb)
+
+{
+#define SWITCH_STATUS_LINE "101 Switching Protocols"
+#define UPGRADE_HEADER "Upgrade: TLS/1.0 HTTP/1.1"
+#define CONNECTION_HEADER "Conenction: Upgrade"
+ const char *upgrade;
+ const char *connection;
+ apr_bucket_brigade *upgradebb;
+ request_rec *r = f->r;
+ SSLConnRec *sslconn;
+ SSL *ssl;
+
+ /* Just remove the filter, if it doesn't work the first time, it won't
+ * work at all for this request.
+ */
+ ap_remove_output_filter(f);
+
+ /* No need to ensure that this is a server with optional SSL, the filter
+ * is only inserted if that is true.
+ */
+
+ upgrade = apr_table_get(r->headers_in, "Upgrade");
+ if (upgrade == NULL) {
+ return ap_pass_brigade(f->next, bb);
+ }
+ connection = apr_table_get(r->headers_in, "Connection");
+
+ apr_table_unset(r->headers_out, "Upgrade");
+
+ if (strcmp(connection, "Upgrade") || strcmp(upgrade, "TLS/1.0")) {
+ return ap_pass_brigade(f->next, bb);
+ }
+
+ if (r->method_number == M_OPTIONS) {
+ apr_bucket *b = NULL;
+ /* This is a mandatory SSL upgrade. */
+
+ upgradebb = apr_brigade_create(r->pool, f->c->bucket_alloc);
+
+ ap_fputstrs(f->next, upgradebb, SWITCH_STATUS_LINE, CRLF,
+ UPGRADE_HEADER, CRLF, CONNECTION_HEADER, CRLF, CRLF, NULL);
+
+ b = apr_bucket_flush_create(f->c->bucket_alloc);
+ APR_BRIGADE_INSERT_TAIL(upgradebb, b);
+
+ ap_pass_brigade(f->next, upgradebb);
+ }
+ else {
+ /* This is optional, and should be configurable, for now don't bother
+ * doing anything.
+ */
+ return ap_pass_brigade(f->next, bb);
+ }
+
+ ssl_init_ssl_connection(f->c);
+
+ ap_log_error(APLOG_MARK, APLOG_INFO, 0, r->server,
+ "Awaiting re-negotiation handshake");
+
+ sslconn = myConnConfig(f->c);
+ ssl = sslconn->ssl;
+
+ SSL_set_state(ssl, SSL_ST_ACCEPT);
+ SSL_do_handshake(ssl);
+
+ if (SSL_get_state(ssl) != SSL_ST_OK) {
+ ap_log_error(APLOG_MARK, APLOG_ERR, 0, r->server,
+ "Re-negotiation handshake failed: "
+ "Not accepted by client!?");
+
+ return AP_FILTER_ERROR;
+ }
+
+ return OK;
+
+}
+
static apr_status_t ssl_io_filter_Output(ap_filter_t *f,
apr_bucket_brigade *bb)
{
@@ -943,6 +1022,11 @@
void ssl_io_filter_register(apr_pool_t *p)
{
+ /* This filter MUST be after the HTTP_HEADER filter, but it also must be
+ * a resource-level filter so it has the request_rec.
+ */
+ ap_register_output_filter ("UPGRADE_FILTER", ssl_io_filter_Upgrade, NULL, AP_FTYPE_PROTOCOL + 5);
+
ap_register_input_filter (ssl_io_filter, ssl_io_filter_Input, NULL, AP_FTYPE_CONNECTION + 5);
ap_register_output_filter (ssl_io_filter, ssl_io_filter_Output, NULL, AP_FTYPE_CONNECTION + 5);
return;
Index: modules/ssl/ssl_engine_kernel.c
===================================================================
RCS file: /home/cvs/httpd-2.0/modules/ssl/ssl_engine_kernel.c,v
retrieving revision 1.78
diff -u -d -b -w -r1.78 ssl_engine_kernel.c
--- modules/ssl/ssl_engine_kernel.c 14 Oct 2002 04:15:58 -0000 1.78
+++ modules/ssl/ssl_engine_kernel.c 15 Oct 2002 16:49:08 -0000
@@ -322,6 +322,16 @@
* Support for SSLRequireSSL directive
*/
if (dc->bSSLRequired && !ssl) {
+ if (sc->enabled == UNSET) {
+ /* This vhost was configured for optional SSL, just tell the
+ * client that we need to upgrade.
+ */
+ apr_table_setn(r->err_headers_out, "Upgrade", "TLS/1.0, HTTP/1.1");
+ apr_table_setn(r->err_headers_out, "Connection", "Upgrade");
+
+ return HTTP_UPGRADE_REQUIRED;
+ }
+
ap_log_rerror(APLOG_MARK, APLOG_ERR, 0, r,
"access to %s failed, reason: %s",
r->filename, "SSL connection required");
@@ -1109,6 +1119,10 @@
STACK_OF(X509) *peer_certs;
SSL *ssl;
int i;
+
+ if (sc->enabled == UNSET) {
+ apr_table_setn(r->headers_out, "Upgrade", "TLS/1.0, HTTP/1.1");
+ }
/*
* Check to see if SSL is on
Index: modules/ssl/ssl_util.c
===================================================================
RCS file: /home/cvs/httpd-2.0/modules/ssl/ssl_util.c,v
retrieving revision 1.35
diff -u -d -b -w -r1.35 ssl_util.c
--- modules/ssl/ssl_util.c 14 Oct 2002 04:15:58 -0000 1.35
+++ modules/ssl/ssl_util.c 15 Oct 2002 16:49:08 -0000
@@ -84,7 +84,7 @@
port = s->port;
else {
sc = mySrvConfig(s);
- if (sc->enabled)
+ if (sc->enabled == TRUE)
port = DEFAULT_HTTPS_PORT;
else
port = DEFAULT_HTTP_PORT;
Re: Final patch for a long time.
Posted by rb...@apache.org.
On Tue, 15 Oct 2002, Johannes Erdfelt wrote:
> On Tue, Oct 15, 2002, rbb@apache.org <rb...@apache.org> wrote:
> > The first project is the Perchild MPM. It basically works, but there are
> > bugs.
The module works for simple cases. However, I have seen a couple of bugs:
1) The second request in a pipelined request doesn't usually get passed
to a new child.
2) Configuration is complex, and should be made simpler.
3) The thread starting/stopping is a bit buggy.
4) Thread stopping should probably just be removed, it doesn't really
help anything, becasue threads are so light-weight.
5) testing testing testing. There are more bugs, but they need to be
found. Set up the server, and test away. The more requests you send, the
more likely you are to find the bugs.
Ryan
>
> Can you give some more information?
>
> I'm interested in using the Perchild MPM myself, but haven't because of
> the reports that it's buggy.
>
> I'm willing to help debug it and make it usable, but the 3 bug reports I
> can find in bugzilla aren't very verbose.
>
> Any hints or direction where to look? Anything not in bugzilla that you
> know about?
>
> JE
>
--
_______________________________________________________________________________
Ryan Bloom rbb@apache.org
550 Jean St
Oakland CA 94610
-------------------------------------------------------------------------------
Re: Final patch for a long time.
Posted by Johannes Erdfelt <jo...@erdfelt.com>.
On Tue, Oct 15, 2002, rbb@apache.org <rb...@apache.org> wrote:
> The first project is the Perchild MPM. It basically works, but there are
> bugs.
Can you give some more information?
I'm interested in using the Perchild MPM myself, but haven't because of
the reports that it's buggy.
I'm willing to help debug it and make it usable, but the 3 bug reports I
can find in bugzilla aren't very verbose.
Any hints or direction where to look? Anything not in bugzilla that you
know about?
JE
Re: SSL upgrade [was: Final patch for a long time]
Posted by Joe Orton <jo...@redhat.com>.
On Thu, Dec 12, 2002 at 01:08:08AM -0600, William Rowe wrote:
> My proposed solution is to review the patch and apply it to cvs HEAD. Get it
> committed. Of course there are no test suites right now, and there won't be
> for a little while yet. But once the code exists, it will be simpler to keep the
> SSL upgrade facility maintained, and debug it as the clients become available
> (most especially, libwww exercises through perl-framework.)
I think there were a couple of mistakes in the patch:
> --- modules/ssl/ssl_engine_io.c 23 Nov 2002 21:19:03 -0000 1.101
> +++ modules/ssl/ssl_engine_io.c 12 Dec 2002 07:06:46 -0000
> @@ -1181,6 +1181,84 @@
> return APR_SUCCESS;
> }
>
> +static apr_status_t ssl_io_filter_Upgrade(ap_filter_t *f,
> + apr_bucket_brigade *bb)
> +
> +{
> +#define SWITCH_STATUS_LINE "101 Switching Protocols"
Should be "HTTP/1.1 101 Swiching Protocols" unless the prefix is added
somewhere I missed, otherwise this isn't a valid status-line.
> +#define UPGRADE_HEADER "Upgrade: TLS/1.0 HTTP/1.1"
> +#define CONNECTION_HEADER "Conenction: Upgrade"
Spot the typo :)
> + connection = apr_table_get(r->headers_in, "Connection");
> +
> + apr_table_unset(r->headers_out, "Upgrade");
> +
> + if (strcmp(connection, "Upgrade") || strcmp(upgrade, "TLS/1.0")) {
> + return ap_pass_brigade(f->next, bb);
> + }
I don't think the requirement that the client sends exactly "Connection:
Upgrade" is correct; the only requirement here is on the client to send
a Connection header including the "upgrade" token.
joe
RE: SSL upgrade [was: Final patch for a long time]
Posted by Sander Striker <st...@apache.org>.
> From: William A. Rowe, Jr. [mailto:wrowe@apache.org]
> Sent: Thursday, December 12, 2002 8:08 AM
> At 11:46 AM 10/15/2002, rbb@apache.org wrote:
> [snip]
> >The second is SSL upgrade. I have the patches, they haven't been
> >committed yet. I have attached them at the bottom of this message. The
> >reason they haven't been committed, is that I don't have a client to test
> >them with, and I haven't had time to create one. The responses are
> >correct I have checked them in plain text. The place that bugs most
> >likely exist, is the actual upgrade code that does the handshake. This is
> >an important feature, and I would really like to see it in 2.0.
>
> I see a couple of very important aspects to this patch:
[...]
> My proposed solution is to review the patch and apply it to cvs HEAD. Get it
> committed. Of course there are no test suites right now, and there won't be
> for a little while yet. But once the code exists, it will be simpler to keep the
> SSL upgrade facility maintained, and debug it as the clients become available
> (most especially, libwww exercises through perl-framework.)
>
> Any disagreement?
Not from me. +1 on committing it.
Sander
SSL upgrade [was: Final patch for a long time]
Posted by "William A. Rowe, Jr." <wr...@apache.org>.
At 11:46 AM 10/15/2002, rbb@apache.org wrote:
[snip]
>The second is SSL upgrade. I have the patches, they haven't been
>committed yet. I have attached them at the bottom of this message. The
>reason they haven't been committed, is that I don't have a client to test
>them with, and I haven't had time to create one. The responses are
>correct I have checked them in plain text. The place that bugs most
>likely exist, is the actual upgrade code that does the handshake. This is
>an important feature, and I would really like to see it in 2.0.
I see a couple of very important aspects to this patch:
1) we must have an implementation of connection: upgrade for libwww, since
that is the mechanism we use for httpd-test/perl-framework. I don't have
such a fix, so I'm just asking the community if anyone has explored that
avenue.
2) it has to be maintained. I've looked at this patch, it appears quite correct.
I'm going to begin working on applying it to cvs HEAD. I'm not concerned
about it quickly appearing in 2.0 since there are no clients right now. I'm
much more concerned about it's availability once clients can support it.
3) right now, the ssl code (ssl_engine_io) was already pretty heavily refactored.
The patch wasn't easy to apply. We are discussing other refactorings that
will make SSL much simpler to follow and far less error-prone. Those will
effectively invalidate the effort Ryan has already invested.
My proposed solution is to review the patch and apply it to cvs HEAD. Get it
committed. Of course there are no test suites right now, and there won't be
for a little while yet. But once the code exists, it will be simpler to keep the
SSL upgrade facility maintained, and debug it as the clients become available
(most especially, libwww exercises through perl-framework.)
Any disagreement?
The current patch that applies to cvs HEAD is attached.
Bill