| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
| |
Not highly sophisticated or efficient but passes every test I know of for
unpredictability & generates mod_auth_digest's secret in negligible time.
Critical comment from crypto experts is more than welcome.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61290 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61235 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
can portably get descriptions for them.
*) move signal stuff from apr.h to (new) apr_signal.h
*) os2/signals.c was collapsed into unix/signals.c
*) APR initialization will init the signal names (if needed)
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61207 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
| |
see: http://apr.apache.org/~dougm/apr_rename.pl
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61194 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
| |
but apr_errno.h is public. It also called in <os.h> with possibly
inappropriate headers, and if os2.h follows any other header's usual
conventions, calling it the second time for more declarations can be
dicey. Relocated <os2.h> into apr.h.in (accessable to all, since it's
later leveraged by most), which simplifies a bunch of code, and relocate
all of the os2errno values into apr_errno.h for APR_STATUS_IS_FOO tests.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61156 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61148 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61147 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61073 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
APR_DECLARES() to the sources, based on compilation emits (there
are many that should be changed eventually, but the compiler will
emit errors if those sources are added for win32).
This change also splits libapr from apr, so the two projects are
compiled seperately. Both .dsp files must be kept up-to-date with
source file revisions.
Finally, libapr.def is no longer needed - so it is gone.
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61072 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
To simplify the task, I also shifted the Makefiles to include a rules.mk
(based on APRUTIL's with a few tweaks). Still needs some work to remove the
INCLUDES setup in all the Makefiles (these can be shared).
buildconf now does more work (and generates some output)
aclocal.m4 is based on a number of M4 files, rather than standalone
apr/test/ has been updated but is probably broken in a few ways.
objs/ is now gone. we link directly from the .lo files.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@61037 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
problem to solve, is that we need an apr_lock in order to allocate
pools, so that we can lock things out when allocating. So, how do we
allocate locks without a pool to allocate from? The answer is to create
a global_apr_pool, which is a bootstrapping pool. There should NEVER
be a sub-pool off this pool, and it is static to an APR file. This is
only used to allow us to allocate the locks cleanly, without using the
NULL pool hack.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60997 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
str* and mem* functions. This fixes all of them. APR_HAVE_STRINGS_H is now
defined in apr.h.
Submitted by: Dale Ghent <daleg@elemental.org>
Reviewed by: Jeff Trawick
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60987 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
Submitted by: Dale Ghent <daleg@elemental.org>
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60978 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
| |
Modules are named mod_foo.so
Dynamic Libraries are named libfoo.dll, and are stored in bin/
The former ApacheCoreDll is now libhttpd.dll
Apache.exe moves to bin/
The make install now copies include, lib, and libexec
All build options are normalized, filenames adjusted appropriately
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60975 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
2000-12-21 Branko Cibej <brane@xbc.nu>
* aprlib.def: Export apr_setprocattr_childin,
apr_setprocattr_childout, and apr_setprocattr_childerr.
* threadproc/win32/proc.c: Folded pipe creation out into helper
function. Moved inheritable-ness mucking from apr_create_process
into apr_setprocattr_* functions.
Added apr_setprocattr_child{in,out,err}.
Note that APR_SHELLCMD won't work (and has never worked) on Win9x.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60973 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60905 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60902 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60897 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
the helper to apr_os_strerror for OS2.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60896 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60894 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
function
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60851 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
| |
destroyed prematurely.
What a great day - mod_info is really back! Thanks to two Ryans and a Jeff.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60815 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
of the data, so the const is proper. This also allows clients to pass const
data in.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60812 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60810 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
| |
it currently uses.
Submitted by: Jon Travis <jtravis@covalent.net>
Reviewed by: Ryan Bloom
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60806 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
| |
apr_terminate should only tear the locks down if this is the final call.
This allows multiple stand-alone programs that all use APR to be combined
cleanly without requiring a lot of if statements. Each program just calls
apr_initialize and apr_terminate, but only the first and last calls
respectively do anything.
Submitted by: Doug MacEachern <dougm@covalent.net>
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60801 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
Hudson <ghudson@mit.edu>.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60792 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
from Karl Fogel <kfogel@collab.net>:
(apr_getopt_option_t): rename from apr_longopt_t, because more accurate.
(apr_initopt, apr_getopt_long, reverse): take "char *const *argv", act
on a copy instead of the original, since we may permute the array.
This is a compromise. The parameter must be compatible with the argv
value passed to main, since that is the primary purpose of the
function. But as has been pointed out, people might want to use the
function with arrays other than argv, and we shouldn't touch the
caller's data.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60790 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Greg Stein <gstein@lyra.org> and Karl Fogel <kfogel@collab.net>):
Simplify apr_getopt_long interface and add support for interleaving
options with arguments.
(apr_getopt_t, apr_initopt): Add "interleave" settable flag and
"skip_start" and "skip_end" fields to keep track of skipped non-option
arguments. Change argv to a char ** so that we can permute it.
(apr_long_option_t): Renamed to apr_longopt_t.
(apr_longopt_t): Rename "val" to "optch" and use for the short option
name as well as the result. "name" can now be NULL, and a structure
with an "optch" of 0 now terminates the list.
(reverse, permute, serr, cerr): New helper functions.
(apr_getopt_long): Rewrite for simplified interface and for
interleaved option support.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60788 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60779 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
FormatMessage on all Win32 platforms. This patch addresses the problem,
and leaves room for other obsure errors to be added. Obviously, the
apr_os_strerror doesn't belong in a single source file, but I'm not
fixing that before the a8 release.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60773 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
*) Added new struct and typedef for apr_getopt_long_t.
*) Added prototype and inline documentation for new apr_getopt_long
function.
misc/unix/getopt.c:
*) Added new apr_getopt_long function to handle long option
processing.
*) Added static method pretty_path() to avoid duplicating code
*) modified apr_getopt to use pretty_path. Removed cut-and-paste code
as well as char *p.
Submitted by: B. W. Fitzpatrick <fitz@red-bean.com>
Reviewed by: Greg Stein
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60752 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
will not be checked into CVS, as they are added to Makefile when the
user runs "make depend." The exact mechanism for building
dependencies is moved to a script called mkdep.sh in the APR helpers
directory. Folks are free to make the mechanism more general (i.e.,
work on systems without gcc -MM), but for now it still requires gcc
-MM. The patch also removes some commented out variable definitions
and rules.
BeOS- and OS/2-specific makefiles have not been updated. I'll post a
patch to those or go ahead and commit them later, but David and Brian
will get to test them.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60727 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
area. This also messes with some unix stuff where it's shared with OS/2.
The strategy used to get the right platform specific include files is:
- Every Makefile.in contains something like
OSDIR=$(INCDIR)/arch/@OSDIR@
DEFOSDIR=$(INCDIR)/arch/@DEFAULT_OSDIR@
INCLUDES=-I$(INCDIR) -I$(OSDIR) -I$(DEFOSDIR)
- all APR private includes look like #include "fileio.h", ie no leading
arch directory so no #ifdef'ing needed in shared .c files.
This ensures that the include file for the target platform is always used
if it exists, otherwise the default is used.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60720 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
| |
the include directory. All private header files for APR are being moved.
This allows platforms that only implement some of the APR types to compile
cleanly.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60656 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
| |
* removed cruft that did nothing (yet)
* add prototypes for the apr_canon... fn's
* Consider: apr_compare_canonical (need to count matching elements)
* Consider: Manoj's suggestion to handle symlinks within here.
* Consider: fanf's concept to use dev/inode for comparison w/o strs.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60607 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
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.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60587 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
Submitted by: Victor J. Orlikowski <v.j.orlikowski@gte.net>
Reviewed by: Ryan Bloom
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60572 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60562 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60560 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The reason is really, really simple. If we ever choose to mix clib and
dos error codes, they criss-cross and don't line up, but they share the
same number space. As I wrote the new APR_IS_ERROR macros, I realized
we were about to shoot ourselves in the foot.
These changes nearly entirely affect Win32 only. The next big patch will
affect all of the rv == APR_ENOENT type problems throughout the system.
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60553 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60552 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60550 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60549 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60537 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60536 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60500 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
| |
(Note: apr_initopt() and/or its callers still need tweaking.)
Submitted by: Bill Rowe
Reviewed by: Jeff Trawick
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60496 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
apr_getopt_t->cont is not really necessary, but I don't yet see a reason
to kill it outright. Might come in handy later.
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60494 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
or multiple argument parsers to play with the argument list.
This change forces the user to call apr_initopt followed by the usual
calls to apr_getopt, with one exception. The argument to a switch (-x)
option is passed to apr_getopt as the &((const char *) arg). Also, the
old int rv status/switch value is now simply the switch value, and the
status is returned as the ap_status_t result.
PR:
Obtained from:
Submitted by:
Reviewed by:
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@60491 13f79535-47bb-0310-9956-ffa450edef68
|