WinNT MPM: Exit the child if the parent process crashes or is terminated. Submitted by: Oracle, via trawick The original modification was made some years ago for Oracle HTTP Server by an Oracle employee. trawick made additional changes for style and for trunk/2.4.x changes.

minor cleanups - don't duplicate prototype for child_main() - make a few things static

correction to r1374780: mpm_winnt.h now needs scoreboard.h for ap_generation_t

WinNT MPM: Store pid and generation for each thread in scoreboard to allow tracking of threads from exiting children via mod_status or other such mechanisms.

Replace ap_create_core_ctx()/ap_core_ctx_get_bb() with a hook

Cleanup effort in prep for GA push: Trim trailing whitespace... no func change

axe the cast in the call to our mpm_nt_eventlog_stderr_open() by fixing the function

cleanups to function signatures, prototypes, visibility

closer to compiling

apr_proc_mutex.h is called for by this very header

reverting 735080, starting over, sorry

apr_proc_mutex.h is called for by this very header

reorganize some internal 'externals'

rip child.c-specific macros away from mpm_winnt.h

rename private COMP_CONTEXT to winnt_conn_ctx_t and make static within child.c

Axe Win9x codepath, including Win32DisableAcceptEx logic. Starting clean.

Track the internal stack_res_flag for properly constraining win32 stacksize on windows 2003 (xp) and 2008 (vista).

Similar change to r483553 to help developers work in parallel on win32, install default command-line build to \Apache2x directory, and name service Apache2.x (which means this doesn't need to change over and over as trunk becomes 2.5, 2.7 etc.)

update license header text

Update the copyright year in all .c, .h and .xml files

Doxygen fixup / cleanup submited by: Neale Ranns neale ranns.org reviewed by: Ian Holsman

Fix Windows CompContext buff size miscalculation

Fix some symbols which should -not- be exported, and decorate real_exit_code with ap_ (this symbol must be exported.) If the others were desired to be public, they should have been ap_xxx decorated.

Update copyright year to 2005 and standardize on current copyright owner line.

enable IPv6 for Windows XP and 2003. In addition this this we need a way to set APR_HAVE_IPV6 at build time in apr.hw

fix name of The Apache Software Foundation

fix copyright date according to the first check in

apply Apache License, Version 2.0

update license to 2004.

add support for querying MPM state to ap_mpm_query() for the WinNT MPM

Win32: Rename WindowsSocketsWorkaround directive to Win32DisableAcceptEx. Clean up code paths.

Added the WindowsSocketsWorkaroud directive for Windows NT/2000/XP to work around problems with certain VPN and Firewall products that have buggy AcceptEx implementations

finished that boring job: update license to 2003. Happy New Year! ;-))

pconf global factors out nicely. The one other pconf appears to be eqivilant to pchild.

Move the last little useful bit of registry.c into services.c, and blast it. All Win32 registry foo is going through ap_regkey now.

Finish the -w keep-console-open-on-error patch.

Win32: Move apr_bucket_alloc() to a more reasonable location to fix memory leak.

Port the -k config option to Apache 2.0. This allows users to reconfigure their service after it has been installed.

My [sixth?] major revamp of service.c. Traded an event for a mutex to the service_init completion, expanded timeouts, moved SERVICE_STOPPED message posting to the main thread since sometimes, in odd cirumstances, our SCM thread wasn't resumed prior to termination, and ripped the code for the stderr logs to use nt_eventlog.c instead. And generally tried to make the code just a little bit more grokable [as if such a thing is really possible.]

Cleanup the eventlogger code, and split it into its own source. [got the copyright correct, no less.]

Update our copyright for this year.

Upon advice of the FirstBill, I began looking for exceptions that occur only in the service-mode. SetServiceDescription was borked, now fixed. Since we don't use the posix/libc style stderr, I've also pitched that code, which was causing an exception.

Externalize the functions to manage the mpm's queue of completion_contexts. Now a module can produce to the queue of contexts that the worker threads consume from. Insomnia tonight in Cary NC...

Win32: Deprecate ap_start_shutdown/ap_start_restart in favor of ap_signal_parent(enum). This simplifies the code a bit for some more MPM cleanup work forthcoming.

Since we expect many users to start experimenting with the next release, and the 'Apache' default server name definately conflicts with a stable 1.3.20 install, changed the AP_DEFAULT_SERVICE_NAME to "Apache2" (unless the user overrides in the compilation.)

get the MPMs consistent w.r.t. visibility of ap_max_requests_per_child

Clean up some exposed data that should be internal to service.c

*) Reimplement the Windows MPM (mpm_winnt.c) to eliminate calling DuplicateHandle on an IOCompletionPort (a practice which MS "discourages"). The new model does not rely on associating the completion port with the listening sockets, thus the completion port can be completely managed within the child process. A dedicated thread accepts connections off the network, then calls PostQueuedCompletionStatus() to wake up worker threads blocked on the completion port. [Bill Stoddard]

Update copyright to 2001

Begin restructuring scoreboard code to enable adding back in the ability to use IPC other than shared memory. Get mod_status working on Windows again. Still to do: Rename some of the function APIs. Replace all calls to reinit_scoreboard with ap_create_scoreboard. Add back in support for scoreboard files.

Get files including mpm_winnt.h compiling again by removing the unnecessary references to buff.h and BUFF.

Return buff.h to the winnt mpm till Bill (or I, if I must) remove the acceptex style contexts into apr.

Renamed all MODULE_EXPORT symbols to AP_MODULE_DECLARE and all symbols for CORE_EXPORT to AP_CORE_DECLARE (namespace protecting the wrapper) and retitled API_EXPORT as AP_DECLARE and APR_EXPORT as APR_DECLARE. All _VAR_ flavors changes to _DATA to be absolutely clear. Thank you Greg, for the most obvious suggestion.

Adjust the WinNT MPM/SCM code for apr_initopt/apr_getopt; mostly just a bunch of constness changes. PR: Obtained from: Submitted by: Reviewed by:

Convert all of the apr_listen.h to ap_listen.h. This was a side effect of Doug M. big change last night.

prefix libapr functions and types with apr_

Advisory: This patch is still anywhere but complete. However, it resolves most of the big bugs of the services to mpm relocation. I apologize that it's taken this long to get back on this. Brought across; Keep console open (needs a bit of cleanup and could be done better now for 2.0), service error event logging, and some other minor threading cleanup. Big issues remain: logged errors (even APLOG_STARTUP) drop into the logfile in spite of our interest in getting them to the user's console. Some threading sequences may yet be out of sorts, and those will be cleaned up by the end of the week. PR: Obtained from: Submitted by: Reviewed by:

PR: Obtained from: Submitted by: Reviewed by: First checkin of Win32 service control structure merged into the mpm. Project files to be updated with this 'experimental' implementation once peer review is complete and the module maintainer accepts the patch.

PR: Obtained from: Submitted by: Reviewed by: Reverse out all _EXPORT_VAR changes back to their original _VAR_EXPORT names for linkage (API_, CORE_, and MODULE_).

This patch corrects the issues from the AP_EXPORT and linkage specification arguments to the ap_hooks.h declarations. As with the APR_ and AP_ patches, API_VAR_EXPORT becomes API_EXPORT_VAR, and MODULE_VAR_EXPORT becomes MODULE_EXPORT_VAR. I will be happy to revert the inclusion of ap_config.h from httpd.h if this bothers anyone. More individual modules need to be patched if we do so. The API_EXPORTs all moved into central storage in the ap_config.h header. Without WIN32 or API_STATIC compile time declarations, these macros remain no-ops. This patch also moves the following data from http_main to http_config: const char *ap_server_argv0; const char *ap_server_root; ap_array_header_t *ap_server_pre_read_config; ap_array_header_t *ap_server_post_read_config; ap_array_header_t *ap_server_config_defines; And the following variables had already moved into ap_hooks.c: ap_pool_t *g_pHookPool; (initialized now in http_config) int g_bDebugHooks; (out of http_config) const char *g_szCurrentHookName; (out of http_config) The changes to http_main.c are in preparation for that module to move out to a seperate .exe for win32. Other platforms will be unaffected, outside of these changes.

s/IMPORT/EXPORT/g; typo i assume.

Fixes to allow Apache to run as a Win95 service... highlights main_win32.h : Moved delarations to a header, by request ap_listen.h : References types declared in http_config.h http_main.h : Add the Win32 flavor entry point declaration apr.hw : Cleanup the redundancy department of redundancy win32/proc.c : Double null termination was required here Everything else should be obvious and isolated to Win32. Build files will be committed seperately.

Change ap_context_t to ap_pool_t. This compiles, runs, and serves pages on Linux, but probably breaks somewhere.

Update to Apache Software License version 1.1

Fix all the License issues. Including: s/Apache Group/Apache Software Foundation/ s/1999/2000/ s/Sascha's license/ASF license

More sendfile work. Use new sendfile API in Apache, update Windows MPM to begin using APR socket API. Note: sendfile on Unix side is broken. Need to detect for NULL hdtr. I'll do it later this week if no one else steps up.

Fix some warnings when configured with --enable-maintainer-mode. ap_start_(shutdown|restart) are no longer static in dexter, because we explicitly state we are creating them to be called from places other than the parent. This is the first in a series of patches to get the 2.0 code to compile cleanly again.

Fix GetAcceptExSockaddrs plus a few tweaks.

Use AcceptEx with an async completion port. What we are gaining here is LIFO dispatching out of the worker thread pool.

Changed pools to contexts. Tested with prefork and pthread mpm's. I'll check this out tomorrow and make sure everything was checked in correctly.

Use AcceptEx (a.k.a. accept and receive) on windows NT. Begin work to make the WIN9x code co-exist with the NT specific code. Since I started this work, I've managed to improve the static page serving performance of Apache on NT by almost 50%. Still a long way from Apache on Linux though. TODO: - Reuse the AcceptEx accept socket (performance) - Clean up the organization of the winnt mpm - Implement Async AcceptEx and Receive. Will this break modules? - File handle cache (file i/o accounts for more than 50% of the time per request) - TransmitFile (I have prototype code implemented). Should we add a send_file API to the I/O control layer? Maybe Apache should always try to send_file. Platforms that don't support send_file could in turn, invoke the appropriate buff code which would then invoke socket iol write/writev.

Win32: Start work on the winnt mpm. This is all horribly broken and hacked right now