mturk 2003/01/23 04:13:03
Modified:jk/native2/common jk_channel_jni.c
Log:
Use the macro instead of strdup. We don't need to dupe the
strings on non EBCID machines.
Revision ChangesPath
1.39 +13 -14jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
mturk 2003/01/23 04:27:45
Modified:jk/native2/common jk_channel_jni.c
Log:
Fix the SSTRDUP_ASCII macro for EBDIC systems.
Revision ChangesPath
1.40 +2 -2 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
mturk 2002/09/26 03:39:57
Modified:jk/native2/common jk_channel_jni.c
Log:
Some Henri's leftovers :)
Revision ChangesPath
1.35 +2 -2 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
[EMAIL PROTECTED] wrote:
mturk 2002/09/26 03:39:57
Modified:jk/native2/common jk_channel_jni.c
Log:
Some Henri's leftovers :)
Revision ChangesPath
1.35 +2 -2 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
mturk 2002/07/14 06:36:03
Modified:jk/native2/common jk_channel_jni.c
Log:
Remove the jvm detaching from channel ajp to the service.
Revision ChangesPath
1.33 +7 -3 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
mturk 2002/07/08 06:41:47
Modified:jk/native2/common jk_channel_jni.c
Log:
no message
Revision ChangesPath
1.32 +9 -10 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
mturk 2002/07/02 05:08:25
Modified:jk/native2/common jk_channel_jni.c
Log:
Call the vm detach on 'afterRequest' and on
channel destoy.
Revision ChangesPath
1.31 +5 -1 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
mturk 2002/06/29 00:51:20
Modified:jk/native2/common jk_channel_jni.c
Log:
Add 'init' param that if set disables the channel until the
TC startup finishes, or lb timeout occurs.
Revision ChangesPath
1.29 +38 -3
mturk 2002/06/29 11:31:01
Modified:jk/native2/common jk_channel_jni.c
Log:
Make default 60 second initialization.
Revision ChangesPath
1.30 +7 -8 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
mturk 2002/06/26 04:06:58
Modified:jk/native2/common jk_channel_jni.c
Log:
Fix the multithreading bug in channel open, setting worker to the
disabled state, that caused already opened enpoint to fail.
Revision ChangesPath
1.28 +27 -18
costin 2002/06/10 16:57:24
Modified:jk/native2/common jk_channel_jni.c
Log:
Remove the dispatch ( which remains to be used only for incoming AJP messages,
until we finish merging the 2 mechanisms ).
Invoke() will be used instead - that's the new mechanism, used to support
nacho 02/05/19 14:41:25
Modified:jk/native2/common jk_channel_jni.c
Log:
* Better close, try to release memory and Java Objects..
Revision ChangesPath
1.21 +34 -9 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
costin 02/05/10 17:02:53
Modified:jk/native2/common jk_channel_jni.c
Log:
Display a small warning if getArray() doesn't pin, it's an indication
of possible performance problems.
Revision ChangesPath
1.17 +14 -4
costin 02/04/25 11:48:13
Modified:jk/native2/common jk_channel_jni.c
Log:
Fix error handling.
Remove duplicated code ( now all dispatching is done consistently in jni/ ).
Revision ChangesPath
1.10 +48 -160
costin 02/04/15 16:52:41
Modified:jk/native2/common jk_channel_jni.c
Log:
Use a more consistent and flexible mechanism for callbacks from C to java.
Also 'consolidate' all native code in one class, to simplify loading of the .so
and avoid ClassLoader problems ( yes, there are
costin 02/04/12 16:07:11
Modified:jk/native2/common jk_channel_jni.c
Log:
Upgrade the jni channel.
Revision ChangesPath
1.8 +9 -8 jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
costin 02/04/11 12:30:24
Modified:jk/native2/common jk_channel_jni.c jk_channel_socket.c
Log:
Fix initialization.
Revision ChangesPath
1.6 +17 -14jakarta-tomcat-connectors/jk/native2/common/jk_channel_jni.c
Index: jk_channel_jni.c
costin 02/01/25 22:55:17
Modified:jk/native2/common jk_channel_jni.c
Log:
A number of bug fixes - make sure we save the global ref, not the ref.
We now get an Endpoint from the java side - and cache/reuse it with the
jni endpoint.
The jni channel is 'interesting' - it's
costin 02/01/11 20:59:19
Added: jk/native2/common jk_channel_jni.c
Log:
I'm changing the way the jni worker operates, to use the same code
as ajp.
Making enhancements to jni and ajp in parallel is too difficult, and the current
jni code has some problems - it's using
19 matches
Mail list logo
