core, modules: like r1657897 but for core and other modules than mod_proxy. More uses of ap_map_http_request_error() and AP_FILTER_ERROR so that we never return an HTTP error status from a handler if some filter generated a response already. That is, from a handler, either ap_get_brigade() (an input filter) returned AP_FILTER_ERROR and we must forward it to ap_die(), or ap_pass_brigade() (an output filter) failed with any status and we must return AP_FILTER_ERROR in any case for ap_die() to determine whether a response is needed or not.

mod_cgi: log cgi script stderr to ScriptLog, use APLOGNO for log_scripterror errors.

Strip useless apr_brigade_cleanup() calls.

core: Stop the HTTP_IN filter from attempting to write error buckets to the output filters, which is bogus in the proxy case. Create a clean mapping from APR codes to HTTP status codes, and use it where needed.

Use %pm available since apr 1.3 instead of an extra call to apr_strerror

Replace an incorrect #if by the correct corresponding #if defined(...) PR 54133 reported by Teodor Milkov ()

Code clean up (remove useless memory allocation) Submitted by: Christophe JAILLET <christophe jaillet wanadoo fr> PR: 52648

Add lots of unique tags to error log messages

Remove more log message prefixes that are now redundant as the the error log format includes the module name.

More cleanup: Expand tabs and some more indentation fixes No functional change

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

PR51371: cross-compile issues with cgi and apr_procattr_limit_*.

Add more (trace) logging to the ap_scan_script_header*() functions Add ap_scan_script_header*_ex() functions that take a module index for logging. Make mod_cgi, mod_cgid, mod_proxy_fcgi, mod_proxy_scgi, mod_isapi use the new functions.

Introduce ap_(get|set)_core_module_config() functions/macros and use them everywhere. We know that the core module has module_index 0. Therefore we can save some pointer operations in ap_get_module_config(cv, &core_module) and ap_set_module_config(cv, &core_module, val). As these are called rather often, this may actually have some (small) measurable effect.

In comments: s@ XX @ XXX: @ so they can actually be found seen.

Avoid unnecessariy initialisation before we test mod_cgi's handler name.

Intel's compiler warns about assigning the non-enum value 0 to the enumeration apr_finfo_t.filetype, which is done when forgetting the previously-derived file type use the appropriate enum value APR_NOFILE instead Also change comparisons of the field with 0 to use APR_NOFILE instead, as is the practice of some existing code.

Use the new APLOG_USE_MODULE/AP_DECLARE_MODULE macros everywhere to take advantage of per-module loglevels

fix potential file descriptor leak with DEBUG_CGI

Bring back OS/2 support. Reverses r758929 with a little bit of conflict resolution.

* Do better checking of pollset operations in various places to avoid segfaults. PR: 46467 Submitted by: Stefan Fritsch <sf sfritsch.de> Reviewed by: rpluem

remove OS/2 platform support

Remove all references to CORE_PRIVATE.

* Prevent running through the error stack by returning OK and setting r->status accordingly if ret is HTTP_NOT_MODIFIED as this breaks mod_cache validating a stale entity.

PR#39710 - badly broken errordocuments for CGI We've just had another duplicate report of this on bugzilla. We've got a simple patch, and people asking WTF is going on with inaction. Noone seems clear on why the patch shouldn't be applied (http://marc.info/?l=apache-httpd-dev&m=117760311129386&w=2).

Follow up to r451006, use APR_STATUS_IS_TIMEUP instead of comparing against APR_TIMEUP directly. Reported by: jorton * modules/generators/mod_cgi.c (cgi_handler): Use APR_STATUS_IS_TIMEUP. * modules/generators/mod_cgid.c (cgid_handler): Ditto.

* Notice that reading of the request entity body failed due to a timeout

PR 31759 (mutated) - reported by Jo Rhett Don't return apr_status_t error value from input filter chain.

update license header text

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

No functional Change: Removing trailing whitespace. This also means that "blank" lines consisting of just spaces or tabs are now really blank lines

Remove CGI block on OPTIONS method so that scripts can respond to OPTIONS directly rather than via server default. PR: 15242 Reviewed-by: Paul Querna, Andre Malo, William A. Rowe, Jr.

Fix bug where non-200 CGI responses will not send anything down filter chain. This is most notable when mod_cache is used. This has been used in production on wiki.apache.org for a while now. * modules/generators/mod_cgi.c (cgi_handler): When a non-zero value is returned by scan_script, set the status field and ensure that we have an EOS to send down the filer stack.

* modules/generators/mod_cgi.c (cgi_bucket_read): Log a specific error if a timeout occurs from a blocking read; more helpful than a generic "apr_bucket_read failed" error from some higher-up filter.

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

* modules/generators/mod_cgi.c (cgi_handler): Ensure all stderr from the script is logged if it generated a non-local redirect. PR: 20111

* server/util_filter.c (ap_save_brigade): Handle an ENOTIMPL setaside function correctly. * modules/generators/mod_cgi.c (cgi_handler): Revert r1.169, unnecessary CGI bucket lifetime kludge. PR: 31247

* modules/generators/mod_cgi.c (cgi_handler): Fix possible segfaults in CGI bucket methods if invoked from a subrequest. PR: 31247

#ifndef AP_UNSAFE_ERROR_LOG_UNESCAPED the stderr to errorlog stuff for those who don't want to escape (as in server/log.c)

Escape bytes returned by the errfn because it might be from an untrusted source

Remove deprecated APR_STATUS_IS_SUCCESS() macro in favor of direct test against APR_SUCCESS.

rev back to previous version, removing chnges to overload detach field , putting back addrspace field

Overloading detached field in cgi_exec_info_t stuct in order to use same field to set detached and addrspace values for proc to be started.

Added new field in apr_procattr_t and cgi_exec_info_t structures to support loading in current or new address space (NetWare ONLY). Replaced changes that added APR_PROGRAM_ADDRSPACE committed 6/11/04. Reviewed by Brad Nicholes

* modules/generators/mod_cgi.c (cgi_handler): Soak up stderr from nph- scripts correctly. PR: 18348 Submitted by: Jeff Trawick (in mod_jcgi)

* modules/generators/mod_cgi.c (cgi_handler): Combine common code between nph and non-nph handling; use a CGI bucket for both cases if APR_FILES_AS_SOCKETS.

Implement the "CGI bucket" in mod_cgi to handle stderr output during script execution, preventing deadlock if stderr output fills the pipe buffer: * modules/generators/mod_cgi.c (log_script_err): Return a read error. (cgi_bucket_create, cgi_bucket_dup, cgi_read_stdout, cgi_bucket_read): New functions. (cgi_handler): Use new CGI bucket rather than a pipe bucket if APR_FILES_AS_SOCKETS; use zero read timeout from stdout/stderr during script execution. PR: 22030 Submitted by: Joe Orton, Jeff Trawick

fix name of The Apache Software Foundation

fix copyright dates according to the first check in

apply Apache License, 2.0

update license to 2004.

get rid of _FOREACH

switch to APR 1.0 API (which is still in flux) because of the changes to the argument lists of apr_mmap_dup and apr_socket_create, 2.1-dev won't build with apr and apr-util's 0.9 branch anymore

drop superfluous argument from ap_ssi_parse_string. The new API should now be stable.

adjust mod_cgi and mod_cgid bump MMN for API change.

Fix the inability to log errors like exec failure in mod_ext_filter/mod_cgi script children (e.g., exec failure). This was broken after such children stopped inheriting the error log handle. (For this to work with mod_ext_filter, that module's LogStderr option has to be on.)

mod_cgi, mod_cgid, mod_ext_filter: Log errors when scripts cannot be started on Unix because of such problems as bad permissions, bad shebang line, etc. Some minor points: If mod_ext_filter debug is enabled, we go ahead and burn cycles in the parent to try to discover when we won't be able to exec the script. The mod_cgid handler wasn't communicating the right log level to the daemon; that was required to get the child errfn to spit out the right message, and it may fix an existing problem.

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

Fix the building of cgi command lines when the query string contains '='. PR: 13914 Submitted by: Ville Skytt� <ville.skytta@iki.fi> (mod_cgi) Jeff Trawick (mod_cgid)

Terminate CGI scripts when the client connection drops. This fix only applies to some normal paths in mod_cgi. mod_cgid is still busted. A prereq for this fix is the change to content-length filter to notice c->aborted: http://cvs.apache.org/viewcvs.cgi/httpd-2.0/server/protoc ol.c.diff?r1=1.119&r2=1.120&diff_format=h PR: 8388

Log error message to the client without the path name.

Fix a long-standing bug in 2.0, CGI scripts were being called with relative paths instead of absolute paths. Apache 1.3 used absolute paths for everything except for SuExec, this brings back that standard.

Simplify add_ssi_vars() and move the ssi/cgi vars into their respective functions [so we may address r->subprocess_env in the ap_cgi_build_command win32 handler.]

Note the changed meaning of the NULL next_filter argument to the ap_sub_req_lookup() family, and fix a few oddball cases (those are, PATH_TRANSLATED reference issues.)

Per RFC 2616 section 9.4, we SHOULD return the content-length if possible on HEAD requests (even for CGI requests). Also do it because it fixes the cause of .37's demise. This also fixes a condition seen on FreeBSD-4.6 where HEAD requests on CGI scripts would hang because the OS would return EAGAIN in log_script_err (via apr_file_gets). Since we never read the script output, this caused FreeBSD to not let us read stderr.

Solve the 80/20 by initializing and storing server_rec->timeout and server_rec->keep_alive_timeout in apr_time_interval_t format (in apr units, whatever they be), as both values exist to pass into APR, and all APR timeouts are in apr_time_t. Reviewed by: Cliff Woolley

Cleanup an emit

Consolidated the startup attributes for building the command line for a CGI binary into a single structure. This allows platform integrator to manipulate the startup parameters much more easily. Also added the detach attribute so that each CGI binary can be spawn appropriately for the OS.

Rewrite mod_cgi's input handling to use brigades and input filters directly. This removes the unnecessary overhead of the *_client_block functions. Optimize usage of the dbuf char* (don't do pcalloc on it!)

- Remove APR_STATUS_IS_SUCCESS calls in favor of directly testing APR_SUCCESS. - Minor style nit with spaces. (No functional changes.)

Call apr_file_write_full instead to achieve the same effect. apr_file_write_full is *guaranteed* to either write everything or return an error. It'll only write short if it received an error.

Stylistic changes - remove tabs, clean up long lines, etc. (No functional changes.) (Functional optimizations/cleanups coming later...)

stop using APLOG_NOERRNO in calls to ap_log_[pr]error()

AcceptPathInfo was totally backwards... it would reject when set to on and by default and accept when set to off for the default handler, and would reject only if set to accept for mod_cgi(d) and mod_isapi. PR: 8234

process_cgi is more descriptive than replace_cmd. It also affects query_string args.

Wasted at least 20 minutes reparsing this code till I determined it's correct. But make the sucker legible so noone else repeats the experience.

In order for Win32 to accept the shebang line of #!perl so we 1) search the PATH and 2) resolve perl.exe for perl, we must add the progtype so we can override APR_PROGRAM with APR_PROGRAM_PATH. This is the API change only for the cgi build command line option function.

Fixed script error logging, which I'd left commented out after making the changes to use a brigade to read the script headers Obtained from: Submitted by: Reviewed by:

BUCKET FREELISTS Add an allocator-passing mechanism throughout the bucket brigades API. From Apache's standpoint, the apr_bucket_alloc_t* used throughout a given connection is stored in the conn_rec by the create_connection hook. That means it's the MPM's job to optimize recycling of apr_bucket_alloc_t's -- the MPM must ensure that no two threads can ever use the same one at the same time, for instance.

eliminate unused variables

Changed mod_cgi to not do single-byte reads to consume the script headers

An error is an error. Since we replace the message with 'failed to invoke command; ...' we aught to log it at the right level.

Trace the invoked command when we use CreateProcess()-style emulation of execxx() for apr_proc_create().

Commit 1 of 2 to: 1. rename ap_rset_content_type to ap_set_content_type 2. reverse the arguments on the call to aligh with ap_set_content_length

Get nph- CGI scripts working again by avoiding filters that can't possibly deal with a script that generates the protocol header. PR: 8902, 8907, 9983

First commit to introduce accessor function to set r->content_type..

Eliminate potential ap_server_root_relative segfaults, with the input of Jeff Trawick's style changes to the first patches. Doesn't include the fixes to ssl [more complex], and we won't trap errors that involve ap_serverroot, since we presume that was normalized on the way in. Therefore, testing ap_server_root_relative(DEFAULT_FOO) cases should never become necessary.

deal with the rename of kill_after_timeout to APR_KILL_AFTER_TIMEOUT

Update our copyright for this year.

Fixup DEBUG_CGI code paths to compile and not leak fds. PR: 9670, 9671 Submitted by: David MacKenzie <djm@pix.net> Reviewed by: Justin Erenkrantz

Correctly check script_in variable. (script_out and script_err are already correct.) PR: 9669

Ever tried to extract some value of QUERY_STRING from within SSI? Now you can :) Provides $0..$9 results for any regex evaluation, and strengthens the expr parser for regex expressions.

Related to Brian's commit to fix the XXX comments I'd added a while back. These lines were broken too, I just hadn't commented them.

Fixed calculation of bucket lengths

Apply the converse of the AcceptPathInfo patch for isapi's/cgi's. The default behavior remains, accept PATH_INFO, but it may be expressly revoked with AcceptPathInfo Off

Fix the cmd command for mod_include. When we are processing a cmd command, we do not want to use the r->filename to set the command name. The command comes from the SSI tag. To do this, I added a variable to the function that builds the command line in mod_cgi. This allows the include_cmd function to specify the command line itself. PR: 8772

Modify post_config hook so that it can return a error, causing the server not to start. previous method was to call exit(1) which would not fail gracefully PR: Obtained from: Submitted by: Reviewed by: (Idea only Jeff Trawick)

Performance: Move the is_included strcmp to after the check to see if the request is to be served by the cgi handler.

Port Justin's mod_cgid change to mod_cgi. mod_cgi wasn't actually broken in the way that mod_cgid was, but it's still good do have these checks be against APR_SUCCESS rather than 0 for clarity.

Remove the returned-length "w" parameter from apr_bucket_heap_create() and apr_bucket_heap_make(). It was useless, since the length is invariant from the length passed in and from the resulting bucket's b->length. This takes care of a long-standing issue first brought up in February and discussed on the dev@apr list. (Issue #2 from the "Bucket API Cleanup Issues" thread.) See http://marc.theaimsgroup.com/?l=apr-dev&m=98324983126666&w=2 Reviewed by: Ryan Bloom (concept)

This patch changes the apr_table_elts macro so that it provides access to the internals of an apr_table_t via a const pointer instead of the current non-const pointer. Submitted by: Brian Pane <BPane@pacbell.net> Reviewed by: Ian Holsman

Fix constness problem reported by gcc: mod_cgi.c: In function `include_cmd': mod_cgi.c:812: warning: passing arg 1 of pointer to function from incompatible pointer type Reported by: Ian's autobuild

Here's the declaration for all to share. Don't expect many to use it.

Remove the Win32 script-processing exception from mod_cgi, and roll build_command_line/build_argv_list into a unified, overrideable ap_cgi_build_command optional function. Eliminates a ton of Win32 cruft from core.c for registry parsing. Win32 (through the default handler, and newest changes to the apr_proc_create fn) continues to serve .bat/.exe files. This is in preparation for adding modules/arch/win32/mod_win32 for scripts. Please review the mod_cgi.c behavior very carefully.

Moved split_and_pass_pretag_buckets back to being a macro at Ryans's request. Removed the return from it by setting and returning a return code instead. Updated the code to check the return code from teh macro and do the right thing.

fix what would seem to be a serious problem in split_and_pass_pretag_buckets(): the retcode is not specified on one of the paths

Optimize out what is a very expensive in-request call to ap_server_root_relative (and is becoming more expensive). Now the call happens in the config phase where it belongs. Note someone can hop in and transpose the stat and open with an open and getfileinfo if you are a performance hound.

Fix for mod_include. Ryan's patch to check error codes put a return in the wrong place. Also, the include handler return code wasn't being checked. I don't like macros with returns, so I converted SPLIT_AND_PASS_PRETAG_BUCKETS into a function.

Fix the new method code. We need to cast 1 to an apr_int64_t or it will be treated as a 32-bit integer, and it will wrap after being shifted 32 times. Submitted by: Cody Sherr <csherr@covalent.net> and Ryan Morgan <rmorgan@covalent.net>

We are done with this subrequest (before we even got started...) so dump it early.

This is not C++, falling out of scope doesn't constitute destruction :)

Multiviews and mod_negotiation will already handle this case, it never belonged in mod_cgi itself.

Be a bit more explicit. If the filetype is not an APR_REG, the fail the request. This relys on directory_walk having already resolved symboloc links.

Win32: Get exec cgi tag working. finfo.protection == 0 on Windows from the apr_stat() call in directory walk (because discovering the permissions on Windows is ridiculously expensive). All we really need to know is whether the file exists.

Safe casts - we are assured that len_read is positive, but it must be stored as an int to initially check the ap_get_client_block() result.

Goodbye ap_send_http_header

Get mod_cgi building again after moving a bunch of code from the HTTP module to the core.

*) Introduce "ap_conf_vector_t" type to assist with legibility and provide some type safety. (unfortunately, our old "void*" is type-safe with the new one, but over time we should be better) *) Propagate the new type to all appropriate functions. *) Random cleaning, whitespace, stylistic nits.

Update copyright to 2001

s/apr_bucket_pipe_creat/apr_bucket_pipe_create/ typo

*) continued header revamping *) torch some headers (and some libs) from the autoconf stuff

*) remove obsolete http_conf_globals.h *) simplify all this #ifdef'ing: just nuke the unixd.h header. it isn't actually needed.

renaming various functions for consistency sake see: http://apr.apache.org/~dougm/apr_rename.pl PR: Obtained from: Submitted by: Reviewed by:

Make sure that mod_include's post_config hook always runs before mod_cgi(d)'s post_config hook. Mod_include creates a hash table in the post_config stage that mod_cgi(d) uses during it's post_config stage, so it must be done in this order. Submitted by: Brian Havard <brianh@kheldar.apana.org.au>

Move the logic for creating CGI processes from mod_include to mod_cgi(d). This removes a good deal of duplicate logic for creating CGI scripts.

Eliminate all assumptions that finfo.protection reflects the existance or absense of a file. finfo.filetype is defined as 0 if APR_NOFILE, or a non-zero value if the file could be apr_stat()'ed.

Another module to cooperate with APR_FINFO_

The changes required for the APR_FINFO_wanted argument to apr_stat/lstat/getfileinfo. These are -NOT- optimal, they are simply the required changes to get the server working. The size of the patch is a warning about how we need to really look at what we are trying to accomplish with all of these stat/lstat calls.

The big change. This is part 3 of the apr-util symbols rename, please see the first commit of srclib/apr-util/include (cvs apr-util/include) for the quick glance at symbols changed.

Provide apr_pool_t arg to register_hooks, since anything they do in that step -must- be done with a pool that will not outlive the cmd pool, from which they may have been dynamically loaded.

adjust remaining modules to use the new handler hook method (Alan Edwards) bring back the old handler prototype by reusing r->handler (dougm) PR: Obtained from: Submitted by: Reviewed by:

Make handlers use hooks.

mod_cgi: Fix some problems where the wrong error value was being traced (errno instead of apr_status). mod_cgid: Keep some of the code in synch with the version in mod_cgi.

On Unix, we do not want to use an absolute path for CGI scripts. SuEXEC will not let us use one, and when not using SuEXEC, we chdir() into the correct directory.

Get exe CGI's working again on Windows

*) sprinkle a little magic "const" dust around (specifically, to compensate for the prototype change to ap_os_create_privileged_process()) *) wrap some long lines

*) Compensate for recent changes in the APR headers. Specifically, some files need to specifically include stdio.h, or a particular apr_*.h header. *) Adjust callers of apr_create_process() to deal with the extra "const" *) Add "const" to args of ap_os_create_privileged_process()

Cleanup more compiler warnings.

Clean up some of the warnings from a Tru64 build, mostly due to the recent apr_ssize_t->apr_size_t changes. The problem in mod_cgid was an int function which didn't have a return statement.

Fix foobar

Add back suexec support.

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.

Document more bad Win32 behavior... if the url isn't .exe, then don't throw an .exe

Write all of the request body to the child, not just what the kernel would accept on the first write. Because the pipe is non-blocking (via setting an APR timeout), we have to send in a loop.

Change r->filters to r->output_filters. This sets things up for us to put input filters into Apache.

Use wrapper macros to avoid the verbosity of the AP_RING macros when doing things with bucket brigades.

Reduce the manualarity of managing bucket brigade lists by using the new AP_RING macros. Most of this commit is fairly pedestrian as you would expect, but I had to redo the chunking filter because of the amount of pointer juggling it did. I have done some minimal testing of this patch and it seems to work.

Call ap_pass_brigade() *before* soaking up stderr. The normal CGI is going to write a bunch of stuff to stdout and nothing to stderr. Reading stderr before processing the stdout pipe can cause deadlock if the CGI fills the kernel buffer for stdout because the CGI will stall writing to stdout while the httpd process will stall reading from the CGI's stderr. The NPH CGI behavior was unchanged. If this is still a problem for NPH CGIs, it was a problem before filtering too. Also, get rid of a couple of unused variables in cgi_handler.

Make CGI's use bucket brigades. This will only work with mod_cgi for the time being. The changes should apply equally to mod_cgid, but I haven't had a chance to really try it yet. This is definately sub-optimal as far as performance is concerned. The concepts are there, and the flags are available for performance tuning, but I didn't want to really get too enmeshed in it today.

Fix a small typo and an incorrect type. Submitted by: Victor J. Orlikowski <v.j.orlikowski@gte.net> Reviewed by: Ryan Bloom

Remove IOLs from Apache. They are no longer necessary, now that we have filtering beginning to work. There is a hack that has been repeated through this patch, we morph a pipe into a socket, and put the socket into the BUFF. Everytime we do that, we are working with a pipe from a CGI, and we should be creating a pipe bucket and passing that bucket back. Because we don't actually have pipe buckets yet, we are using this hack. When we get pipe buckets, this will be fixed.

Remaining cleanup of ap_ -> apr_ and AP_ -> APR_ transformation... see src/lib/apr/apr_compat.h for most details. Also a few minor nits to get Win32 to build. PR: Obtained from: Submitted by: Reviewed by:

prefix libapr functions and types with apr_

Move all APR functions related to strings to their own directory, and create a new header for those functions. This is the first step to removing the apr/lib directory completely, and moving those files/functions to descriptive directories.

Include strings.h for strcasecmp(), strncasecmp(), and bzero(). Include time.h for time(). This removes a bunch of compiler warnings with gcc -Wall on AIX. Submitted by: Jeff Trawick, Victor Orlikowski

blast the old names for the status codes

Convert module and fix warnings.

This patch solves several specific issues: 1.3.x truncated any open/command arguments following the %1 arg. so this patch adds the char** arguments to several functions 1.3.x did not expand environment strings (%userprofile% etc.) *) This patch may still not do so, if we are running with a subset of the 'normal' environment for security reasons. 1.3.x did not parse the extension itself (eg. the .pl key itself) for the command, failing the 'named' type (eg. perlscript), so this patch first tests the 'named' key, then the .ext key PR: Obtained from: Submitted by: Reviewed by:

Nothing in Apache should be calling stat anymore, so we can remove the check for sys/stat.h and all references to the resulting macro.

Add the resource limiting code back to Apache 2.0. This only works on Unix because I can't find any other platforms with rlimit. If there are other platforms that need this code, then some of the code needs to move. This has just barely been tested, so it could probably use some good testing.

Fix zombie process problem with mod_cgi. This problem was introduced as part of the "Convert ap_proc_t to a complete type." enhancement. mod_cgi previously declared a ptr to an ap_proc_t, storage was allocated by ap_create_process() from the request pool, and the ap_proc_t address was passed to ap_note_subprocess(). With the "Convert ap_proc_t to a complete type." change, ap_proc_t lived in autodata, but the address was still passed to ap_note_subprocess(). When the pool was cleaned up, the ap_proc_t in autodata had been used for something else, so the contents were garbage, but pool cleanup needed the contents, especially the os pid to pass to waidpid(). Since this was garbage, we never reaped status from the cgi child and thus the cgi child remained a zombie.

Make the string in a log message unique so we know what failed. Before, "couldn't create child process: " was used for a couple of different logs.

EBCDIC: Rearrange calls to ap_checkconv() so that most handlers won't need to call it. Submitted by: Greg Ames, Jeff Trawick

mod_cgi: Make ScriptLog directive work again by fixing the flags passed to ap_open(). TODO: Roll recent mod_cgi logging changes into mod_cgid.

mod_cgi: Restore logging of stderr from child process when ScriptLog isn't used (as in 1.3), except that on Unix it is now logged via ap_log_rerror() instead of by the child having STDERR_FILENO refer to the error log. Submitted by: Greg Ames Reviewed by: Jeff Trawick

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.

Rename stdin, stdout, stderr from new ap_proc_t to in, out, and err because Windows was having problems with the original names.

Convert ap_proc_t to a complete type. This lets us access the pid directly. Only the prefork MPM has been ported so far, the rest of the Unix MPM's are coming later today.

Remove all remaining instances in Apache code of ap_{block,unblock}_alarms.

Update a couple of modules to use ap_stat() rather than stat(). PR: Obtained from: Submitted by: Reviewed by:

Add logic to timeout CGI scripts properly. This requires that APR pipes can be timed out.

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

Make the server work without asking for buffered file I/O. This removes all APR_BUFFERED flags from the server, and it fixes a minor bug in the ungetchar patch (I forgot a file when comitting).

Update to Apache Software License version 1.1

Backout layered I/O changes.

Enabled layered I/O. Docs are forthcoming.

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

Add a test program for pipes. Add support for non-buffered CGI's to mod_cgi. This still doesn't terminate CGI's after a time limit, that's the next step.

Set up return codes correctly in the failure case.

Manoj has been pushing for this for a while, but I've been too dense to understand that he was right. :-) Basically, this makes the modules use ap_config to test for header file inclusion. This method is not to be used for larger modules that run autoconf theirselves. Including ap_config is only valid for modules which rely on Apache to do their configuration. Currently, this is only the core modules.

This is a huge change to the configure system. Basically, this name space protects most, if not all, of the Apache macros. This has been tested on Linux using all of the Unix MPM's. The biggest push was that Apache's httpd.h header file can no longer include the ap_config.h file. Most of the other files include this themselves now. Submitted by: Ryan Bloom and Manoj Kasichainula

Rework ap_finfo_t to split the file type out of the protection field. I've taken a stab at the unix implementation but tested only on OS/2.

time overhaul: - ap_time_t is a 64-bit scalar, microseconds since epoch - ap_exploded_time_t corresponds to struct tm with a few extras probably broken on anything except linux.

Separate the stat structure from the file structure and use ap_stat and ap_getfileinfo in apache.

Initialize all ap_file_t's to NULL. This allows ap_open and ap_stat to work together without causing memory leaks.

Make CGI's work correctly in Apache 2.0. APR wants the first element in the argument list to be the program name. This is exactly what POSIX requires, and what Windows wants, but Apache 1.3 would create the arg list, and then have another function squeeze the program name in later. This patch fixes that by having mod_cgi put the program name the correct place.

Get rid of more platform dependant code.

We may not always want to make the pipes between processes non-blocking. This patch allows us to set if both sides of the pipes are nonblocking, both sides are blocking, just the parent side blocks, or just the child side blocks for all three of the pipes created during create_process. I have also modified Apache to take advantage of this change.

Undo the ap_bgets errno patch. It will return status with ap_berror().

An errnoless interface for ap_bgets

ap_bwrite now exports an errnoless interface.

Make args to CGIs work.

Fix mod_cgi so it works on Unix systems. We need to be VERY careful about calling ap_cleanups_for_exec now. There is really no reason to call this from outside of APR, unless you are specifically doing a fork/exec. If you use ap_create_process, it will call ap_cleanups_for_exec for you in the right spot now. Also cleaned up some of the code because it helped with debugging. We need better error messages in mod_cgi now. iCVS:

Apply ap_log_error() errno parameter addition to a few more files.

A couple of arg swaps on ap_open

Fix argv list generation bug.

Still does not work reliably, but it's closer to being right.

Need to pass in just the directory, not including the filename. Note that CGI arguments are being completely ignored. suexec needs to be handled, the CGI arguments need to be canonicalized (e.g., quote args containing spaces, etc.) and ScriptInterpreterSource needs reimplementing. Whew!

Fix bug interpreting cgi_child return code. Do some clean-up.

Fix function arguments to respect APR's New World Order.

First cut at migrating mod_cgi over to APR. Still needs some work to make it Win32 friendly (use named pipes, etc.) Submitted by: Paul Reder Reviewed by: Bill Stoddard

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.