Re: [PATCH] ./native2/common/jk_channel_apr_socket.c - use apr_port_t instead of short for port

2004-02-02 Thread Henri Gomez 
Commited, thanks

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [PATCH] ./native2/common/jk_channel_apr_socket.c - use apr_port_t instead of short for port

2004-02-02 Thread Günter Knauf 
 Commited, thanks
thanks, Henri!

Do you know what's about the ./native2/common/jk_channel_socket.c file? Does soemone 
still use it since APR is now mandatory, or can we remove it? If it stays then we 
should also patch the port there to 'unsigned short'...

Guenter.




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [PATCH] ./native2/common/jk_channel_apr_socket.c - use apr_port_t instead of short for port

2004-02-02 Thread Henri Gomez 
Günter Knauf a écrit :

Commited, thanks
thanks, Henri!

Do you know what's about the ./native2/common/jk_channel_socket.c file? Does soemone still use it since APR is now mandatory, or can we remove it? 
If it stays then we should also patch the port there to 'unsigned short'...
Yes, it should be removed but the conservative approach will be to
have a patch for unsigned short
-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [PATCH] ./native2/common/jk_channel_apr_socket.c - use apr_port_t instead of short for port

2004-02-02 Thread Günter Knauf 
Hi Henri,
 Do you know what's about the ./native2/common/jk_channel_socket.c file?
 Does soemone still use it since APR is now mandatory, or can we remove
 it?
If it stays then we should also patch the port there to 'unsigned
short'...

 Yes, it should be removed but the conservative approach will be to
 have a patch for unsigned short
here we go:
http://www.gknw.com/test/jk_channel_socket.c.diff

# Patch to solve http://nagoya.apache.org/bugzilla/show_bug.cgi?id=17579
#
--- ./jk/native2/common/jk_channel_socket.c.origThu Jan 29 18:23:28 2004
+++ ./jk/native2/common/jk_channel_socket.cMon Feb 02 18:19:28 2004
@@ -100,7 +100,7 @@
 int ndelay;
 struct sockaddr_in addr;
 char *host;
-short port; /* Should be unsigned - big ports will fail */
+unsigned short port;
 int keepalive;
 int timeout;
 };
@@ -116,7 +116,7 @@
 */
 
 static int JK_METHOD jk2_channel_socket_resolve(jk_env_t *env, char *host,
-   short port,
+   unsigned short port,
struct sockaddr_in *rc);
 
 static int JK_METHOD jk2_channel_socket_close(jk_env_t *env, jk_channel_t *ch,
@@ -276,7 +276,8 @@
 
 /** private: resolve the address on init
  */
-static int JK_METHOD jk2_channel_socket_resolve(jk_env_t *env, char *host, short port,
+static int JK_METHOD jk2_channel_socket_resolve(jk_env_t *env, char *host, 
+   unsigned short port,
struct sockaddr_in *rc)
 {
 int x;
@@ -285,7 +286,7 @@
 /* for now use the correct type, in_addr_t   */
 in_addr_t laddr;
 
-rc-sin_port   = htons((short)port);
+rc-sin_port   = htons((unsigned short)port);
 rc-sin_family = AF_INET;
 
 /* Check if we only have digits in the string */

Guenter.


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [PATCH] ./native2/common/jk_channel_apr_socket.c - use apr_port_t instead of short for port

2004-02-02 Thread Mike Anderson 
Committed.

Mike Anderson

 [EMAIL PROTECTED] 2/2/2004 10:29:26 AM 
Hi Henri,
 Do you know what's about the ./native2/common/jk_channel_socket.c
file?
 Does soemone still use it since APR is now mandatory, or can we
remove
 it?
If it stays then we should also patch the port there to 'unsigned
short'...

 Yes, it should be removed but the conservative approach will be to
 have a patch for unsigned short
here we go:
http://www.gknw.com/test/jk_channel_socket.c.diff 

# Patch to solve
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=17579 
#
--- ./jk/native2/common/jk_channel_socket.c.origThu Jan 29 18:23:28
2004
+++ ./jk/native2/common/jk_channel_socket.cMon Feb 02 18:19:28 2004
@@ -100,7 +100,7 @@
 int ndelay;
 struct sockaddr_in addr;
 char *host;
-short port; /* Should be unsigned - big ports will fail */
+unsigned short port;
 int keepalive;
 int timeout;
 };
@@ -116,7 +116,7 @@
 */
 
 static int JK_METHOD jk2_channel_socket_resolve(jk_env_t *env, char
*host,
-   short port,
+   unsigned short port,
struct sockaddr_in
*rc);
 
 static int JK_METHOD jk2_channel_socket_close(jk_env_t *env,
jk_channel_t *ch,
@@ -276,7 +276,8 @@
 
 /** private: resolve the address on init
  */
-static int JK_METHOD jk2_channel_socket_resolve(jk_env_t *env, char
*host, short port,
+static int JK_METHOD jk2_channel_socket_resolve(jk_env_t *env, char
*host, 
+   unsigned short port,
struct sockaddr_in
*rc)
 {
 int x;
@@ -285,7 +286,7 @@
 /* for now use the correct type, in_addr_t   */
 in_addr_t laddr;
 
-rc-sin_port   = htons((short)port);
+rc-sin_port   = htons((unsigned short)port);
 rc-sin_family = AF_INET;
 
 /* Check if we only have digits in the string */

Guenter.


-
To unsubscribe, e-mail: [EMAIL PROTECTED] 
For additional commands, e-mail: [EMAIL PROTECTED] 


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [PATCH] ./native2/common/jk_channel_apr_socket.c - use apr_port_t instead of short for port

2004-02-01 Thread NormW 
Good morning All.
This would also allow closing of Bug 17579 since the recent changeover to
APR for mod_jk2 (now using jk_channel_apr_socket.c)

Norm

 Hi,
 a user just found that the port number is limited to less than 32768 with
mod_jk2;
 the patch below solves this:
 http://www.gknw.com/test/jk_channel_apr_socket.c.diff



==
 --- ./native2/common/jk_channel_apr_socket.c.orig Thu Jan 29 18:23:28 2004
 +++ ./native2/common/jk_channel_apr_socket.c Sat Jan 31 01:45:56 2004
 @@ -74,9 +74,9 @@
   */
  struct jk_channel_apr_private {
  int ndelay;
 +apr_port_t port;
  apr_sockaddr_t *addr;
  char *host;
 -short port;
  int keepalive;
  int timeout;
  };
 @@ -168,7 +168,7 @@
  jk_channel_apr_private_t *socketInfo=
  (jk_channel_apr_private_t *)(ch-_privatePtr);
  int rc;
 -short port=socketInfo-port;
 +apr_port_t port=socketInfo-port;

  if( socketInfo-host==NULL ) {
  char *localName=ch-mbean-localName;

 Guenter.


 -
 To unsubscribe, e-mail: [EMAIL PROTECTED]
 For additional commands, e-mail: [EMAIL PROTECTED]




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: [PATCH] ./native2/common/jk_channel_apr_socket.c - use apr_port_t instead of short for port

2004-02-01 Thread Günter Knauf 
Hi,
 This would also allow closing of Bug 17579 since the recent changeover to
 APR for mod_jk2 (now using jk_channel_apr_socket.c)
thanks for the pointer! Showed that my recent patch was incomplete...; so next try:

http://www.gknw.com/test/jk_channel_apr_socket.c.diff
==
# Patch to solve http://nagoya.apache.org/bugzilla/show_bug.cgi?id=17579
#
--- ./jk/native2/common/jk_channel_apr_socket.c.origThu Jan 29 18:23:28 2004
+++ ./jk/native2/common/jk_channel_apr_socket.c Mon Feb 02 00:49:56 2004
@@ -74,9 +74,9 @@
  */
 struct jk_channel_apr_private {
 int ndelay;
+apr_port_t port;
 apr_sockaddr_t *addr;
 char *host;
-short port;
 int keepalive;
 int timeout;
 };
@@ -92,7 +92,7 @@
 */
 
 static int JK_METHOD jk2_channel_apr_resolve(jk_env_t *env, char *host,
-short port,
+apr_port_t port,
 jk_channel_apr_private_t *rc);
 
 static int JK_METHOD jk2_channel_apr_close(jk_env_t *env, jk_channel_t *_this,
@@ -168,7 +168,7 @@
 jk_channel_apr_private_t *socketInfo=
 (jk_channel_apr_private_t *)(ch-_privatePtr);
 int rc;
-short port=socketInfo-port;
+apr_port_t port=socketInfo-port;
 
 if( socketInfo-host==NULL ) {
 char *localName=ch-mbean-localName;
@@ -227,7 +227,7 @@
 /** private: resolve the address on init
  */
 static int JK_METHOD jk2_channel_apr_resolve(jk_env_t *env,
- char *host, short port,
+ char *host, apr_port_t port,
  jk_channel_apr_private_t *rc)
 {
 int err;


Guenter.


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]