Re: Running Apache in the foreground
On Thu, Dec 27, 2001 at 05:44:22PM -0800, Jos Backus wrote: > I'm willing to code a patch which allows Apache to run in the foreground in > its own session. Currently it kills the pgrp it is in even though it didn't > create it (bad practice imo - only destroy what you create). That is, if there > is any interest. A previous bugreport in this area was ignored so I may be > wasting my time. I'll interpret no response as ``no interest, go away''. I'm not sure what version you are talking about (1.3 or 2.0), but on 2.0 Unix-based MPMs, you can run httpd with -DNO_DETACH and it will not detach from the terminal. ./httpd -DNO_DETACH The only place I'm seeing us killing the process group is when you send a SIGTERM. Is this what you are talking about? Do you have a better mechanism in mind? AIUI, some OSes do not deliver signals sent to the parent to the children and others do. I think we were attempting to obtain consistency. I imagine we could have httpd set it own process group, but that might not be good either. Thoughts? -- justin
Re: cvs commit: apache-1.3/src/os/netware ApacheCore.imp
Applied, but this needs to be double-checked on Netware. > wrowe 01/12/27 21:03:08 > > Modified:src ApacheCore.def >src/ap ap_snprintf.c >src/include ap.h ap_alloc.h buff.h http_conf_globals.h > http_config.h http_core.h http_log.h http_main.h > http_protocol.h http_request.h http_vhost.h httpd.h > rfc1413.h >src/main alloc.c buff.c http_config.c http_core.c http_log.c > http_main.c http_protocol.c http_request.c > http_vhost.c rfc1413.c util.c >src/os/netware ApacheCore.imp > Log: > Normalize symbol exports for Win32/Netware to the httpd.exp reference. > Diff tags pre_win_nw_syms/post_win_nw_syms for complete edit. > > Note I've corrected _SEVERAL_ badly declared symbols on Win32 into > API_EXPORT_NONSTD flavors (e.g. those using (...) args). This may, > or may not, break binary compatibility depending on how those args > are addressed, and if those functions were used. > > I've further tested by setting aside the .def file and rebuilding, > and validated that our symbols list corresponds to the API_DECLARE() > macros at this moment. > > Submitted by: Thomas Eibner <[EMAIL PROTECTED]> > Reviewed by: Stoddard, Rowe Here is the list of badly declared, now _NONSTD functions (with a list like this, I'm afraid much binary compatibility will be lost); -API_EXPORT(int) ap_snprintf(char *buf, size_t len, const char *format,...) +API_EXPORT_NONSTD(int) ap_snprintf(char *buf, size_t len, const char *format,...) -API_EXPORT(void) ap_table_do(int (*comp) (void *, const char *, const char *), void *rec, - const table *t,...); +API_EXPORT_NONSTD(void) ap_table_do(int (*comp) (void *, const char *, const char *), +void *rec, const table *t,...); -API_EXPORT(int) ap_bvputs(BUFF *fb,...); +API_EXPORT_NONSTD(int) ap_bvputs(BUFF *fb,...); -API_EXPORT(int) ap_rprintf(request_rec *r, const char *fmt,...) +API_EXPORT_NONSTD(int) ap_rprintf(request_rec *r, const char *fmt,...) -API_EXPORT(void) ap_log_error(const char *file, int line, int level, +API_EXPORT_NONSTD(void) ap_log_error(const char *file, int line, int level, const server_rec *s, const char *fmt, ...) -API_EXPORT(void) ap_log_rerror(const char *file, int line, int level, +API_EXPORT_NONSTD(void) ap_log_rerror(const char *file, int line, int level, const request_rec *s, const char *fmt, ...) -API_EXPORT(void) ap_log_printf(const server_rec *s, const char *fmt, ...) +API_EXPORT_NONSTD(void) ap_log_printf(const server_rec *s, const char *fmt, ...) And finally, here is a list of the differences remaining between httpd.exp and the win32 exported symbols. I'm posting the list mostly because someone might want to go back over httpd.exp, since some win32 inclusions should fall in the httpd.exp list. Also, there are a few more missing symbols in the win32 list yet to be fixed. --- httpd.exp Thu Dec 27 23:07:54 2001 +++ declared.ref Thu Dec 27 23:45:46 2001 +access_module +alias_module +ap_acquire_mutex +ap_add_loaded_module +ap_bpushh +ap_check_alarm -ap_child_terminate +ap_create_mutex +ap_destroy_mutex -ap_dummy_mutex +ap_get_limit_req_body +ap_get_service_key +ap_get_win32_interpreter -ap_listeners +ap_loaded_modules +ap_md5_binary +ap_note_cleanups_for_h +ap_open_mutex +ap_os_canonical_filename +ap_os_case_canonical_filename +ap_os_dso_error +ap_os_dso_load -ap_os_is_path_absolute +ap_os_is_filename_valid +ap_os_systemcase_filename +ap_pcloseh -ap_prelinked_modules -ap_preloaded_modules -ap_register_other_child +ap_registry_get_server_root +ap_release_mutex +ap_remove_loaded_module -ap_rfc1413_timeout +ap_scan_script_header_err_core -ap_signal -ap_slack -ap_sys_siglist -ap_unregister_other_child -ap_util_init -ap_util_uri_init +ap_uuencode +apache_main +asis_module +auth_module +autoindex_module +cgi_module +closedir +config_log_module +dir_module +env_module +imap_module +includes_module +mime_module +negotiation_module +opendir +os_spawnle +os_spawnv +os_spawnve +os_stat +os_strftime +readdir +regcomp +regerror +regexec +regfree +setenvif_module +so_module -XML_DefaultCurrent -XML_ErrorString -XML_ExternalEntityParserCreate -XML_GetBase -XML_GetBuffer -XML_GetCurrentByteCount -XML_GetCurrentByteIndex -XML_GetCurrentColumnNumber -XML_GetCurrentLineNumber -XML_GetErrorCode -XML_GetSpecifiedAttributeCount -XML_Parse -XML_ParseBuffer -XML_ParserCreate -XML_ParserCreateNS -XML_ParserFree -XML_SetBase -XML_SetCdataSectionHandler -XML_SetCharacterDataHandler -XML_SetCommentHandler -XML_SetDefaultHandler -XML_SetDefaultHandlerExpand -XML_SetElementHandler -XML_SetEncoding -XML_SetExternalEntityRefHandler -XML_SetExternalEntityRefHandlerArg -XML_SetNamespaceDeclHandler -XML_SetNotationDeclHandler -XML_SetNotStandaloneHandler -XML_SetProcessingInstructi
[PATCH] Re: f->c and f->r->connection
On Thu, Dec 27, 2001 at 11:07:22PM -0500, Cliff Woolley wrote:
>
> Why are there some circumstances when f->c is NULL but f->r and
> f->r->connection are valid? I could understand f->r being NULL while f->c
> would not be, but the other way around took me totally by surprise.
I agree. I wonder if this patch would break anything. It seems
sane. -- justin
Index: server/util_filter.c
===
RCS file: /home/cvs/httpd-2.0/server/util_filter.c,v
retrieving revision 1.73
diff -u -r1.73 util_filter.c
--- server/util_filter.c2001/12/28 05:46:21 1.73
+++ server/util_filter.c2001/12/28 05:49:22
@@ -277,7 +277,7 @@
f->frec = node->frec;
f->ctx = ctx;
f->r = r;
-f->c = c;
+f->c = r ? r->connection : c;
if (INSERT_BEFORE(f, *outf)) {
f->next = *outf;
Re: conn_rec apr_bucket_alloc_t* field
On Thu, Dec 27, 2001 at 08:29:39PM -0800, Ian Holsman wrote: > I personally like bucket_alloc > It might be a bit long, but it is descriptive, and people will can > understand exactly what it is about by looking at the name +1. =) -- justin
Re: conn_rec apr_bucket_alloc_t* field
Cliff Woolley wrote: > Anybody have preferences as to what I name the field I'm adding to the > conn_rec to store the apr_bucket_alloc_t*? It's used _everywhere_, so > short and sweet would be preferable. Right now in the patch it's called > c->balloc, but that seems a bit wonky. At least it's unique enough that I > can search-and-replace it with something else easily. ;-) Ideas? A > couple of people have suggested the more verbose c->bucket_alloc or > something like that, but this really is used _everywhere_. The less > obtrusive, the better. Too bad there's nothing as succinct as "c->pool" > that we could name it... that would be ideal. What about c->bmemsys or > something? Ideas? I personally like bucket_alloc It might be a bit long, but it is descriptive, and people will can understand exactly what it is about by looking at the name > > --Cliff > > -- >Cliff Woolley >[EMAIL PROTECTED] >Charlottesville, VA > > >
conn_rec apr_bucket_alloc_t* field
Anybody have preferences as to what I name the field I'm adding to the conn_rec to store the apr_bucket_alloc_t*? It's used _everywhere_, so short and sweet would be preferable. Right now in the patch it's called c->balloc, but that seems a bit wonky. At least it's unique enough that I can search-and-replace it with something else easily. ;-) Ideas? A couple of people have suggested the more verbose c->bucket_alloc or something like that, but this really is used _everywhere_. The less obtrusive, the better. Too bad there's nothing as succinct as "c->pool" that we could name it... that would be ideal. What about c->bmemsys or something? Ideas? --Cliff -- Cliff Woolley [EMAIL PROTECTED] Charlottesville, VA
f->c and f->r->connection
Why are there some circumstances when f->c is NULL but f->r and f->r->connection are valid? I could understand f->r being NULL while f->c would not be, but the other way around took me totally by surprise. Thanks, --Cliff -- Cliff Woolley [EMAIL PROTECTED] Charlottesville, VA
Re: Time for and Apache 2.0 Tag and Roll?
On Thu, 27 Dec 2001, Brian Pane wrote: > The tag+roll sounds okay with me, unless you want to wait for > the bucket free list code (the stuff that Cliff is working on)-- > since that's the last significant design change that I've heard > of for 2.0. That code is close (will be done within the week), but it touches a lot of things and will need a good bit of review/discussion no doubt before being committed. So I say go ahead and roll another version within the next few days before I have a chance to break things. ;-) Plus that way we'll hopefully have a good "before and after" set of versions to compare to see how much good the buckets freelist stuff actually does us. --Cliff -- Cliff Woolley [EMAIL PROTECTED] Charlottesville, VA
Running Apache in the foreground
I'm willing to code a patch which allows Apache to run in the foreground in its own session. Currently it kills the pgrp it is in even though it didn't create it (bad practice imo - only destroy what you create). That is, if there is any interest. A previous bugreport in this area was ignored so I may be wasting my time. I'll interpret no response as ``no interest, go away''. Thanks, -- Jos Backus _/ _/_/_/Santa Clara, CA _/ _/ _/ _/ _/_/_/ _/ _/ _/_/ [EMAIL PROTECTED] _/_/ _/_/_/use Std::Disclaimer;
Re: [PATCH] update apr_thread_exit's prototype and fix usage in worker
From: <[EMAIL PROTECTED]> Sent: Thursday, December 27, 2001 9:31 AM > +1 from me. > > david > > > > ** Will: Please verify my win32 changes. I don't recall if this patch > > contains the changes we talked about for win32 from my original posting > > of this patch. I'll test, but presume it does and go ahead and commit [sometimes it's simpler just to 'mop-up' :]
Re: Time for and Apache 2.0 Tag and Roll?
Brian Pane <[EMAIL PROTECTED]> writes: > >Things seem pretty stable right now. I'd like to see us push out a couple > >more big betas with wide publicity (and usage) before we go GA. > Sounds good to me > > --Brian +1! -- Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site: http://www.geocities.com/SiliconValley/Park/9289/ Born in Roswell... married an alien...
Re: Time for and Apache 2.0 Tag and Roll?
Aaron Bannert wrote: >On Thu, Dec 27, 2001 at 01:07:48PM -0800, Brian Pane wrote: > >>Bill Stoddard wrote: >> >>>I'll tag in a few days if I don't hear any dissent. >>> >>>Bill >>> >>The tag+roll sounds okay with me, unless you want to wait for >>the bucket free list code (the stuff that Cliff is working on)-- >>since that's the last significant design change that I've heard >>of for 2.0. >> > >There's no reason we can't tag now and push out another alpha/beta, and >do the same thing in a week or so when the bucket freelist gets committed. >The more often we do a beta the more people we'll have testing our code. > >Things seem pretty stable right now. I'd like to see us push out a couple >more big betas with wide publicity (and usage) before we go GA. > Sounds good to me --Brian
Re: cvs commit: httpd-2.0/os/unix unixd.c
Greg Ames <[EMAIL PROTECTED]> writes: > Jeff Trawick wrote: > > > > [EMAIL PROTECTED] writes: > > > > > gregames01/12/27 11:53:37 > > > > > > Modified:.STATUS > > >os/unix unixd.c > > > Log: > > > change ap_unix_create_privileged_process() to allocate enough space in > > > the newargs array to allow for a null terminator. This allows suexec and > > > mod_cgid to work together reliably. > > > > What did David Bowie say? > > Ah, wham bam, thank you Ma'am ??? CH-CH-CH-CH-CHANGES -- Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site: http://www.geocities.com/SiliconValley/Park/9289/ Born in Roswell... married an alien...
Re: Time for and Apache 2.0 Tag and Roll?
On Thu, Dec 27, 2001 at 01:07:48PM -0800, Brian Pane wrote: > Bill Stoddard wrote: > > >I'll tag in a few days if I don't hear any dissent. > > > >Bill > > > > The tag+roll sounds okay with me, unless you want to wait for > the bucket free list code (the stuff that Cliff is working on)-- > since that's the last significant design change that I've heard > of for 2.0. There's no reason we can't tag now and push out another alpha/beta, and do the same thing in a week or so when the bucket freelist gets committed. The more often we do a beta the more people we'll have testing our code. Things seem pretty stable right now. I'd like to see us push out a couple more big betas with wide publicity (and usage) before we go GA. -aaron
Re: Time for and Apache 2.0 Tag and Roll?
Bill Stoddard wrote: >I'll tag in a few days if I don't hear any dissent. > >Bill > The tag+roll sounds okay with me, unless you want to wait for the bucket free list code (the stuff that Cliff is working on)-- since that's the last significant design change that I've heard of for 2.0. --Brian
Re: cvs commit: httpd-2.0/os/unix unixd.c
Jeff Trawick wrote: > > [EMAIL PROTECTED] writes: > > > gregames01/12/27 11:53:37 > > > > Modified:.STATUS > >os/unix unixd.c > > Log: > > change ap_unix_create_privileged_process() to allocate enough space in > > the newargs array to allow for a null terminator. This allows suexec and > > mod_cgid to work together reliably. > > What did David Bowie say? Ah, wham bam, thank you Ma'am ???
Re: cvs commit: httpd-2.0/os/unix unixd.c
On Thu, Dec 27, 2001 at 03:42:22PM -0500, Jeff Trawick wrote: > > change ap_unix_create_privileged_process() to allocate enough space in > > the newargs array to allow for a null terminator. This allows suexec and > > mod_cgid to work together reliably. > > What did David Bowie say? Ground Control to Major Tom? ;) -a
Re: cvs commit: httpd-2.0/os/unix unixd.c
[EMAIL PROTECTED] writes: > gregames01/12/27 11:53:37 > > Modified:.STATUS >os/unix unixd.c > Log: > change ap_unix_create_privileged_process() to allocate enough space in > the newargs array to allow for a null terminator. This allows suexec and > mod_cgid to work together reliably. What did David Bowie say? -- Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site: http://www.geocities.com/SiliconValley/Park/9289/ Born in Roswell... married an alien...
Re: Time for and Apache 2.0 Tag and Roll?
On Thu, Dec 27, 2001 at 11:39:39AM -0500, Dale Ghent wrote: > mod_ssl still identifies itself as alpha code. Is this not the case > anymore? Well, does it work for you? The only outstanding issue I'm aware of is that shmem session caching isn't implemented. And, it really is waiting for someone to actually code it up. (For some reason, I never get around to it.) -- justin
Re: Time for and Apache 2.0 Tag and Roll?
On Wed, Dec 26, 2001 at 08:39:57PM -0500, Bill Stoddard wrote: > I'll tag in a few days if I don't hear any dissent. I was about to suggest the same thing. =) -- justin
Re: New LICENSE file...
At 2:38 PM -0500 12/20/01, Bill Stoddard wrote: >I'd like to propose that we extend our LICENSE file to include references to/licenses >of >all the other components we include in the server. This gets all the license >information >in one place. > +1 -- === Jim Jagielski [|] [EMAIL PROTECTED] [|] http://www.jaguNET.com/ "A society that will trade a little liberty for a little order will lose both and deserve neither"
Re: Time for and Apache 2.0 Tag and Roll?
Aaron Bannert <[EMAIL PROTECTED]> writes: > On Thu, Dec 27, 2001 at 11:24:33AM -0500, Jeff Trawick wrote: > > "Bill Stoddard" <[EMAIL PROTECTED]> writes: > > > > > I'll tag in a few days if I don't hear any dissent. > > > > Hopefully Aaron can get his thread-exit fixes in soon? > > I can commit them right now if we would prefer C-T-R. As I understand it, you have tested on the platforms you have access to and worst-case scenario for the other platforms is that a minor tweak might be needed. If that is accurate, then I'm comfortable with you committing it right now. -- Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site: http://www.geocities.com/SiliconValley/Park/9289/ Born in Roswell... married an alien...
Re: Time for and Apache 2.0 Tag and Roll?
Commit it and we'll review. I have a patch that I'll try to get to the list tomorrow, though it probably won't be suitable for commit prior to the next roll. With it the worker MPM builds on BeOS simply by using --with-mpm=worker on the configure line :) With some work the same would apply to the other MPM's as well. david > On Thu, Dec 27, 2001 at 11:24:33AM -0500, Jeff Trawick wrote: >> "Bill Stoddard" <[EMAIL PROTECTED]> writes: >> >> > I'll tag in a few days if I don't hear any dissent. >> >> Hopefully Aaron can get his thread-exit fixes in soon? > > I can commit them right now if we would prefer C-T-R. > > -aaron
Re: Time for and Apache 2.0 Tag and Roll?
On Wed, 26 Dec 2001, Bill Stoddard wrote: | I'll tag in a few days if I don't hear any dissent. mod_ssl still identifies itself as alpha code. Is this not the case anymore? /dale
Re: Time for and Apache 2.0 Tag and Roll?
On Thu, Dec 27, 2001 at 11:24:33AM -0500, Jeff Trawick wrote: > "Bill Stoddard" <[EMAIL PROTECTED]> writes: > > > I'll tag in a few days if I don't hear any dissent. > > Hopefully Aaron can get his thread-exit fixes in soon? I can commit them right now if we would prefer C-T-R. -aaron
Re: Time for and Apache 2.0 Tag and Roll?
Greg Ames <[EMAIL PROTECTED]> writes: > Bill Stoddard wrote: > > > > I'll tag in a few days if I don't hear any dissent. > > sounds good. Hopefully we'll have a quorum available to pound on the > tagged code on various platforms. We haven't heard from Madhu in a > while, but if we roll on daedalus HP-UX should be OK. FWIW, I haven't had problems lately with my regression builds on HP-UX 11. -- Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site: http://www.geocities.com/SiliconValley/Park/9289/ Born in Roswell... married an alien...
Re: Time for and Apache 2.0 Tag and Roll?
"Bill Stoddard" <[EMAIL PROTECTED]> writes: > I'll tag in a few days if I don't hear any dissent. Hopefully Aaron can get his thread-exit fixes in soon? -- Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site: http://www.geocities.com/SiliconValley/Park/9289/ Born in Roswell... married an alien...
Re: Time for and Apache 2.0 Tag and Roll?
Bill Stoddard wrote: > > I'll tag in a few days if I don't hear any dissent. sounds good. Hopefully we'll have a quorum available to pound on the tagged code on various platforms. We haven't heard from Madhu in a while, but if we roll on daedalus HP-UX should be OK. Greg
Time for and Apache 2.0 Tag and Roll?
I'll tag in a few days if I don't hear any dissent. Bill
Re: [PATCH] update apr_thread_exit's prototype and fix usage in worker
+1 from me. david > The following patch does the trick. I've actually attached two patches, > the first is the APR change and the second is the update to the worker > MPM to use the new function signature. What's changed in APR: > apr_thread_exit no longer takes a pointer to an apr_status_t, which was > very ackward. It now takes a simple pass-by-copy apr_status_t parameter > which is then stored in the thread structure on all platforms and > pulled back out with a call to apr_thread_join. > > * I did my best to get this to work on non-unix platforms, but I > will need help verifying my changes (especially netware). > > ** Will: Please verify my win32 changes. I don't recall if this patch > contains the changes we talked about for win32 from my original posting > of this patch. > > p.s. I liked the name "exitval" that David put in Beos better than what > I had before "rv". Other than that this patch is pretty much > the same thing (minus what David already implemented on Beos) I > posted a couple months ago. > > -aaron
Re: cvs commit: httpd-2.0 STATUS
[EMAIL PROTECTED] wrote: > >RELEASE SHOWSTOPPERS: [snip] >* Test suite failures: > o perchild doesn't even build Well, it builds on Red Hat 5.2 now, but doesn't start (or at least keep running) with the perl-framework's default configuration. And it doesn't even build on Red Hat 6.2; ./configure fails with Configuring Apache Portable Runtime Utility library... configuring package in srclib/apr-util now loading cache ../.././config.cache configure: error: can not find install-sh or install.sh in ../apr/build /tmp/ap2/httpd-2.0/srclib/apr-util/../apr/build configure failed for srclib/apr-util For details see http://Source-Zone.Org/Apache/regression/>. -- #kenP-)} Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/ Author, developer, opinionist http://Apache-Server.Com/ "All right everyone! Step away from the glowing hamburger!"
Re: Benchmark: 29 vs 30
Aaron Bannert wrote: > > Having some problems with the images, I think the spaces are freaking > out my Linux/Netscape browser. Indeed. Embedded spaces are illegal; the page that links to the images needs to use encoding like 'Transport%20Error%20Analysis.gif'. Curiously enough, '%20' encoding of spaces works, but '+' encoding doesn't. I thought that historically '+' was a shorthand for '%20', though 2396 doesn't acknowledge this; 1630 does, but only applies it to the query-string component of the URI. Maybe Roy can clear up my fuzzy memory on this..? Any road, those embedded spaces in the links need to be encoded. -- #kenP-)} Ken Coar, Sanagendamgagwedweinini http://Golux.Com/coar/ Author, developer, opinionist http://Apache-Server.Com/ "All right everyone! Step away from the glowing hamburger!"
