William A Rowe Jr wr...@rowe-clan.net writes:
unix with a still-broken iconv implementation (do any examples
still exist?)
The Subversion project has a SPARC buildbot running Solaris 10 and the
system iconv has problems with UTF-8. The buildbot uses GNU iconv as a
workaround.
--
Philip
TROY.LIU 劉春偉 troy@deltaww.com.cn writes:
In our practice, we found two threads get same address returned by
apr_palloc. It will happen about one hour later after our server
starts. We are using apr 1.4.5, the issue still exists in the
latest subversion.
From apr_pools.h:
*
William A. Rowe Jr. wr...@rowe-clan.net writes:
On Wed, 17 Apr 2013 15:14:31 + (UTC)
Igor Galić i.ga...@brainsware.org wrote:
caused by this projects creative use of APR:
https://code.google.com/p/unimrcp/issues/detail?id=29
Is there anything I can do other than compile a sepcial
Ben Reser b...@reser.org writes:
On Wed, Apr 3, 2013 at 12:29 PM, Jeff Trawick traw...@gmail.com wrote:
I'm not aware of any apr_file_lock() use within httpd, so you wouldn't
regress anything (unless you're using third-party modules that use
apr_file_lock). But I guess you've introduced a
Anthony Gargiulo anth...@agargiulo.com writes:
I am trying to compile arp-util 1.4.1 on scientific linux, but I can't seem
to get it to work.
As this is being compiled along with httpd 2.4.2, I used this to configure
httpd:
./configure --prefix=/opt/httpd --enable-so --enable-ssl
Jonathan Melissant jonathanmeliss...@gmail.com writes:
The problem here is that the host system (the c6x architecture) is not
known to the configure script.
In libtool, the host system is specified as follows:
# The host system.
host_alias=c6x
host=tic6x-unknown-coff
host_os=coff
Maybe
Jonathan Melissant jonathanmeliss...@gmail.com writes:
next, APR does not generate a static library (libapr.so file) but only a
static library (libapr.la). I have tried everything to change this but all
tries were unsuccesfull.
libapr.la is not a static library, it is a libtool library file.
Branko Čibej br...@apache.org writes:
The root of the problem is that automake
insists on recursive builds, which on top of all the other problems,
make builds an order of magnitude slower.
The manual claims it is possible to use automake non-recursively:
Laurent Charmet laurent.char...@atempo.com writes:
I have an issue with the function apr_strnatcasecmp(). Take a look at
gdb output:
5719if (apr_strnatcasecmp(type_str,
p_list[i].name) == 0) {
(gdb)
5720service =
Bojan Smojver bo...@rexursive.com writes:
On Tue, 2009-09-22 at 20:06 +, minf...@apache.org wrote:
URL: http://svn.apache.org/viewvc?rev=817809view=rev
Log:
The implementation of expand_array() in tables/apr_hash.c allocates
a new bucket array, without making any attempt to release the
Bojan Smojver bo...@rexursive.com writes:
On Mon, 2012-01-30 at 10:22 +, Philip Martin wrote:
It was reverted because it means the hash table data is not available
in pool cleanup handlers
I guess it should then be reverted on the 1.5.x branch as well.
Yes, I believe so.
--
Philip
Chris Darroch chr...@pearsoncmg.com writes:
The other question, I suppose, is whether there's a hit to performance
from apr_generate_random_bytes() call. Without having tested explicitly,
it looks like the default case for modern Linux is APR_HAS_RANDOM=1 and
DEV_RANDOM=/dev/random, with
Bert Huijben b...@qqmail.nl writes:
(We don't even have a single call to apr_pool_parent_get() in trunk)
We do, in update_editor.c.
--
Philip
* include/http_protocol.h: Use APR_INLINE instead of inline for C89
compatibility
Index: include/http_protocol.h
===
--- include/http_protocol.h (revision 1137943)
+++ include/http_protocol.h (working copy)
@@ -334,7 +334,7
apr trunk combines apr and apr-util into a single library and there is
only a single apr-2-config script. Subversion can be configured by
using this script for both apr and apr-util, but Subversion's Berkeley
DB detection doesn't work. This is because the apr-2-config script
doesn't support the
Philip Martin [EMAIL PROTECTED] writes:
Port some of the atomic code from 1.2.x to 0.9.x, in particular make
mutex operations that fail cause an abort and make the generic C
implementation of apr_atomic_cas work on 64 bit platforms.
A less radical change to the code is simply to truncate
prev = *mem;
if (prev == cmp) {
*mem = with;
}
#endif /* APR_HAS_THREADS */
return prev;
}
--
Philip Martin
= *(void **)mem;
if (prev == cmp) {
*mem = with;
}
-return prev;
#endif /* APR_HAS_THREADS */
+return prev;
}
#endif /*!defined(apr_atomic_cas) !defined(APR_OVERRIDE_ATOMIC_CAS) */
--
Philip Martin
, \, args[i], \, NULL);
}
else {
--
Philip Martin
-threaded
libapr?
Glibc doesn't provide no-op functions for things like
pthread_mutex_create, those are only available when the full threading
library replaces all the no-op functions.
--
Philip Martin
fewer mutexes. Of
course it's likely that the thread-once functions would be used to
create mutexes so this may not be much of a saving.
--
Philip Martin
reject these dates.
This inconsistency means that Subversion is capable of writing data
that it is subsequently unable to read :-(
--
Philip Martin
Roy T. Fielding [EMAIL PROTECTED] writes:
On Monday, March 22, 2004, at 04:00 PM, Philip Martin wrote:
This code in apr_time_exp_get:
year = xt-tm_year;
if (year 70 || ((sizeof(time_t) = 4) (year = 138))) {
return APR_EBADDATE;
}
rejects all 2038 dates even though
an operation that could SEGV.
--
Philip Martin
] =
{306, 337, 0, 31, 61, 92, 122, 153, 184, 214, 245, 275};
-
-year = xt-tm_year;
-
-if (year 70 || ((sizeof(time_t) = 4) (year = 138))) {
-return APR_EBADDATE;
-}
/* shift new year to 1st March in order to make leap year calc easy */
--
Philip Martin
with a big disclaimer that hotcopy
does not work on ext3? Implement our own dot-first stuff in
svn_io_dir_walk? Wait for an APR fix and require that version? I've
had a look at the Subversion code and I think hotcopy is the only
thing affected.
--
Philip Martin
of apr_dir_read
guarantee it?
Questions for Subversion:
- Is svn_io_dir_walk supposed to be dot-first?
- If it is, are we going to rely on APR to provide it or implement it
locally?
--
Philip Martin
' in filesystem
'/home/pm/sw/subversion/obj/repostress/db' [500, #160014]
I am using the prefork MPM on Linux, and Subversion HEAD (r7287). The
error occurs on both single processor and dual processor machines, and
with both Apache 2.0.47 and STRIKER_2_0_48_PRE4.
--
Philip Martin
and it breaks Subversion.
--
Philip Martin
2 pm pm 48 Dec 18 18:21 foo
-rwxr-xr-x1 pm pm 5147 Dec 18 18:22 z
-rw-r--r--1 pm pm262 Dec 18 18:22 z.c
$ ./z
. 0
.. 0
z 0
foo 0
z.c 0
--
Philip Martin
+ ], [
Why should we check for db_create_4000 first? Why not wait until
after db_create is not found? -- justin
No real reason, it just happens to be the way I wrote it.
--
Philip Martin
Philip Martin [EMAIL PROTECTED] writes:
Eeek!
I've just upgraded to apache/apr/apr-util to HEAD and now I can
reproduce this.
$ svnadmin create repo
$ svn mkdir file://`pwd`/repo/foo
$ svn co file://`pwd`/repo wc
$ svn up wc
../svn/subversion/libsvn_wc/adm_crawler.c:315: (apr_err
=db.h
apu_db_lib=db
apu_db_version=4
+])], [
+ AC_CHECK_LIB(db, db_create_4000, [
+ apu_db_header=db.h
+ apu_db_lib=db
+ apu_db_version=4
])])])
fi
])
--
Philip Martin
apu_db_version=4
-])])])
+])])])])
fi
])
--
Philip Martin
on my system at the end of the last sentence...
Oops, my mistake, I failed to apply the complete patch. It works here
too.
I find disallowing apr_terminate an ugly solution :-( I suppose it's
valid for the testsuite problem, but it doesn't look good for
applications in general.
--
Philip
. The handler
should cleanup only those resources local to the process, leaving the
proc_mutex in a working state. It should be the application's
responsibility to call (or fail to call :) apr_proc_mutex_destroy at
the appropriate place.
--
Philip Martin
that
erroneously deleted a temporary file when forking a child process, I
know I would prefer the latter!
--
Philip Martin
)++ with read, wait, increment, write.
Had the test been written like this originally it would not have
passed erroneously. It may also help when/if someone ports APR to a
new platform.
--
Philip Martin
Aaron Bannert [EMAIL PROTECTED] writes:
On Sun, Nov 17, 2002 at 12:25:46PM +, Philip Martin wrote:
$ ./testprocmutex
APR Proc Mutex Test
==
Exclusive lock test
Initializing the lock OK
Starting all
Philip Martin [EMAIL PROTECTED] writes:
I just ripped out that code, want to give this another try on your
SMP Linux box?
It still fails.
As I said, I don't know much about semaphores, but I think the problem
is something to do with process exit.
The pool cleanup handler
static
Philip Martin [EMAIL PROTECTED] writes:
The first problem is the line
if (apr_os_thread_equal(mutex-owner, apr_os_thread_current())) {
where there is access to the shared data mutex-owner without any sort
of synchronization. Now mutex-owner may not be an atomic type, in
which case
mutex contention).
--
Philip Martin
Hello
Running the latest valgrind on the Subversion executables generates
errors that are caused by APR.
==4022== pthread_mutex_unlock: mutex is not locked
==4022==at 0x404EE4BF: (within /usr/lib/valgrind/libpthread.so)
==4022==by 0x403B7043: thread_mutex_cleanup (thread_mutex.c:67)
Sander Striker [EMAIL PROTECTED] writes:
Essentially the Subversion code
does this
apr_pool_clear_debug (some_pool, file_line);
apr_pool_create_ex(error_pool, some_pool, abort_on_pool_failure, NULL);
The problem is that when some_pool was originally created by
[EMAIL PROTECTED] writes:
Author: striker
Date: 2002-05-26 09:27 GMT
New Revision: 2017
Modified:
trunk/subversion/libsvn_subr/svn_error.c
Log:
* subversion/libsvn_subr/svn_error.c
(svn_pool_create): Use the new functionality in APR to set
a maximum on the total amount of
Hi
Subversion's ac-helpers/aprutil.m4 has the code
ac_configure_args_save=$ac_configure_args
ac_configure_args=$ac_configure_args MKDIR=\$MKDIR\
SVN_SUBDIR_CONFIG(apr-util, --with-apr=../apr)
ac_configure_args=$ac_configure_args_save
This causes MKDIR to be set for the apr-util
Cliff Woolley [EMAIL PROTECTED] writes:
On Sat, 2 Feb 2002, Kevin Pilch-Bisson wrote:
On Sat, Feb 02, 2002 at 08:34:44PM +, Philip Martin wrote:
+#ifdef APR_HAS_THREADS
+status = apr_thread_mutex_lock(umask_mutex);
+if (!APR_STATUS_IS_SUCCESS(status))
+return
Hello
I am trying to get Subversion to support builddir != srcdir when
building apr/apr-util from source in the Subversion tree. This used to
work until Subversion switched to using find_apr.m4 and find_apu.m4
and started using apr-util.
The first problem is that I am having trouble working out
Greg Stein [EMAIL PROTECTED] writes:
(btw, for completeness, we'd also want ways to turn off readonly and to the
executable state)
Here's a patch that does that, and provides a way to query the
attributes. It also makes the permission setting by apr_file_attrs_set
respect the umask of the
Philip Martin [EMAIL PROTECTED] writes:
Greg Stein [EMAIL PROTECTED] writes:
I'll apply the APR side of this. Thanks for the patch!
Thanks for applying it, however it's got a bit mangled. In
apr_file_info.h the magic number APR_FILE_SOURCE_PERMS needs to be
different from all possible
50 matches
Mail list logo