@objects="`find $(SUBDIRS) -name '[EMAIL PROTECTED]@' | grep -v
expat/lib`" ; \
$(LINK) @lib_target@
delete-exports:
There seem to be other spots were we search for the .lo files...
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
* Implement code to detect if a pipe is still alive on Windows.
> + */
> +if (other_children == NULL)
> +return;
and how does it work differently if you remove these two lines of
code?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
ht
ntics. If strtok[_r] will do the job, then it is
preferable.
Any concerns?
*mod_mime_magic and mod_proxy
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
ce the strchr() calls with inline
code. The only compiler I'm really familiar with w.r.t. code
generation will replace strchr() with inline code but I suspect that
is not the case with most compilers.
The MSVC .dsp changes are untested (duck!). I'm somewhat sorry about
that :)
Oh, an
Jeff Trawick <[EMAIL PROTECTED]> writes:
> > Add apr_strtok(), a thread-safe flavor of strtok() which has the
> > same interface as strtok_r().
...
> Some eager person may wish to replace the strchr() calls with inline
> code. The only compiler I'm rea
pthread_rwlock_t rwlock;
This is what fails to compile with RH 6.0. We found
pthread_rwlock_init() but didn't test for pthread_rwlock_t
without/with __USE_UNIX98.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
wait until this point in autoconfiguration to
>start turning on libc feature test macros. Maybe it will work on
>some glibc version with the current set of tests, but I don't think it
>is cool in general. It seems to me that as soon as you change libc
>feature test macros you n
oo before we test for threads because feature foo is where we
figure out that we must enable some libc feature test macro.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Greg Stein <[EMAIL PROTECTED]> writes:
> On Fri, Jun 01, 2001 at 07:55:52PM -0400, Jeff Trawick wrote:
> > Greg Stein <[EMAIL PROTECTED]> writes:
> >
> > > No. apr_hints is a last resort. Actual tests like the AC_TRY_COMPILE()
> > > that
> >
ID_SOURCE is for.
It seems to me that we have very specific information about a certain
libc's requirements but we're ashamed to admit it and so we're
dressing it up in autoconf for political correctness. Unfortunately,
the resulting code risks turning on unnecessary or undesirable
ower for all production uses.
-1
> +if (!context)
> +return APR_EINVAL;
ditto
> @@ -203,6 +209,9 @@
>apr_size_t inbytes_left, outbytes_left;
>#endif
>
> +if (!context)
> +return APR_EINVAL;
ditto
> @@ -283,6 +292,9 @@
With the message from
such an assert from some random program somebody still has to go find
the source code to apr-util and see what the assert means. The assert
doesn't tell how they got to that point so the core dump is still
required to debug it. And we get the core dump for free.
--
Jef
s initialize
finfo.filetype to APR_NOFILE since a caller's memset(,0,)/*calloc() of
the storage didn't initialize it to the right value?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
of printf support in APR affect whether or not a
program which uses APR can use printf?
Note that we rely on APR-provided format strings to be compatible with
printf format strings whenever possible so that gcc can keep us honest.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key
termination.
It might also be useful to register a routine to be called when a new
APR thread is created.
comments?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
just committed something along the same lines but with a couple of
tweaks:
1) apr_lock_destroy() has the wrong signature for a cleanup routine
2) the cleanup needs to be removed when the lock is destroyed
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
apr_sms_blocks.c
> > Log:
> > can't add to void *; pretend it is char *
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
designed to need the nesting then
allowing a nested mutex acquire to succeed would seem to hide an
application problem.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Jeff Trawick <[EMAIL PROTECTED]> writes:
> 1. It looks to me that on Unix APR mutexes are always nesting if
>APR_HAS_THREADS and APR mutexes are never nesting (subject to their
>underlying implementation of course). This is not cool.
I meant to say "It looks to
e request_pool is
> cleared. We have to dup, or the file won't be available to us, and the
> original bug will be back.
Isn't it just a matter of killing the cleanup associated with one pool
and registering the cleanup with the new pool?
--
Jeff Trawick | [EMAIL PROTECTED] |
ptor?
otherwise, the doc for APR_BUCKET_FLAGSENDFILE should mention that if
the bucket has this flag then the apr_file_t * has to be at the magic
offset
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Justin Erenkrantz <[EMAIL PROTECTED]> writes:
> On Fri, Jun 15, 2001 at 05:59:14AM -0400, Jeff Trawick wrote:
> > Jeff Trawick <[EMAIL PROTECTED]> writes:
> >
> > > 1. It looks to me that on Unix APR mutexes are always nesting if
> > >APR_
at I think should happen:
1) replace your strncpy with memcpy
2) add code after the loop to '\0'-terminate the resulting string
3) test, 'cause I sure didn't
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
p = '\0';
cp doesn't point to the end of the string to be built yet. If you had
*(cp + len) = '\0';
then I'd believe you.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
ode;
}
thang and I'll consider using array notation in code like that :)
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
ata it's trying
> to protect?
I think you need to be very explicit about what you see that you don't
like. I don't really know what your talking about.
It would seem that for the lock operations supported by APR there is a
platform-independent interface for the application to use. I
sometimes think it is missing some cool stuff but that is another
issue.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
committed... thanks!
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
e works fine.
If APR has the code to back down to the default then there has to be a
special return code from apr_lock_create() to indicate that the app
got the default and/or some other way for the app to find out that it
didn't get what it asked for.
--
Jeff Trawick | [EMAIL PROTECTED] |
getting to the right low-level
code fits well with Jim Jag's suggestion for allowing one APR build to
be able to use multiple os-provided lock mechanisms.
If we were willing to expose the function ptrs in public header files
then apr_lock_acquire() and apr_lock_release() could be macros
intraprocess lock mechanism*.
*though currently APR never has any choices to make
w.r.t. intraprocess lock mechanism
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
ock_t **lock, const char *fname,
apr_pool_t *cont)
{
-apr_status_t stat;
-if ((*lock)->scope != APR_INTRAPROCESS) {
-if ((stat = apr_unix_child_init_lock(lock, cont, fname)) !=
APR_SUCCESS) {
-return stat;
-}
-}
+if ((
e same function calls*.
If you don't have them both in the same function calls then how do you
implement APR_LOCK_ALL? It is nice to have APR_LOCK_ALL 'cause that
handles a portability/performance issue on behalf of the application.
--
Jeff Trawick | [EMAIL PROTECTED] | PG
ame), so the
> who thing seems to hinge on APR_PROCESS_LOCK_IS_GLOBAL. (I could be
> wrong here :)
That simply means that on other platforms and interprocess lock blocks
out other threads in the same process (or that the APR code for those
platforms has a bug :) ).
--
Jeff Trawick | [EMAIL
@
printf("Testing multiple locking\n");
printf("%-60s","Creating the lock we'll use");
-if ((rv = apr_lock_create(&multi, APR_MUTEX, APR_LOCKALL,"multi.lock",
-pool)) != APR_SUCCESS) {
+if ((rv = apr_lock_create(&multi, APR_MUTEX, APR_LOCKALL,APR_LOCK_CREATE,
+ "multi.lock", pool)) != APR_SUCCESS) {
printf("Failed!\n");
return rv;
}
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Justin Erenkrantz <[EMAIL PROTECTED]> writes:
> On Mon, Jun 25, 2001 at 04:30:45PM -0400, Jeff Trawick wrote:
> > Did anybody want this done differently?
>
> What about Read/write locks? Those would use the same function for
> creation, but none of the values for th
pr_lock_create_np() and
tell it to use the default mechanism.)
Okay, it's just a special name but it is clearly not in the set of
interfaces which folks would expect to work the same everywhere.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
ementation, and this goes against the
> "portable" part of APR.
I don't have a problem with this. Sorry for being stupid before :)
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
et of tests, but I don't think it
is cool in general. It seems to me that as soon as you change libc
feature test macros you need to start over at ground zero.
I think I'd feel safest with something like Victor's change but which
looks at the glibc version instead of the kernel version.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
che did a home directory
> lookup on an invalid user name. This isn't cool on the part of libc,
> but oh well.
This really sucks, but whatcha gonna do?
See /usr/local/apache/corefiles/httpd.core.20 if you're curious.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at w
gt; -func:pthread_mutexattr_setpshared custom:with_pthread_cross,
> +func:pthread_mutexattr_setpshared,
> APR_DECIDE(USE_PROC_PTHREAD_SERIALIZE, [pthread mutex]))
> if test "x$apr_lock_method" != "x"; then
> APR_DECISION_FORCE($apr_lock_met
k_get():
#if APR_HAS_PROC_PTHREAD_SERIALIZE
os->pthread_crossproc = lock->pthread_interproc;
#endif
#if APR_HAS_SYSVSEM_SERIALIZE || APR_HAS_FCNTL_SERIALIZE ||
APR_HAS_FLOCK_SERIALIZE
os->int_crossproc = lock->interproc;
#endif
#if APR_USE_PTHREAD_SERIALIZE
os->
Justin Erenkrantz <[EMAIL PROTECTED]> writes:
> On Mon, Jul 02, 2001 at 08:21:10AM -0400, Jeff Trawick wrote:
> > Please don't commit just yet.
>
> Should I back this out? Ryan wants to T&R at 10AM PST. I guess the
> easy "fix" would be to place fc
Cliff Woolley <[EMAIL PROTECTED]> writes:
> On 2 Jul 2001, Jeff Trawick wrote:
>
> > In create_lock(), initialize as follows before filling in any lock handles:
> >
> > #if APR_HAS_PROC_PTHREAD_SERIALIZE
> > new->pthread_interproc = NULL;
>
Cliff Woolley <[EMAIL PROTECTED]> writes:
> On 2 Jul 2001, Jeff Trawick wrote:
>
> > > We didn't catch this part, though. The apr_lock_t is pcalloc'ed just
> > > prior to calling create_lock(). Do we really need to do this? If so,
> > > feel
vial.c, but that looks easy to fix.
Currently, we think APR_HAS_SYSVSEM_SERIALIZE if the platform has
semget() and semctl().
What is in your apr/include/arch/unix/apr_private.h for HAVE_SEMGET
and HAVE_SEMCTL?
Can you grep through the Darwin header files for the related
declarations (struct sembuf, IPC_PRIVATE, etc.)? Maybe they are in a
header file which isn't included from crossproc.c?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Jeff Trawick <[EMAIL PROTECTED]> writes:
> Chuck Murcko <[EMAIL PROTECTED]> writes:
>
> > On Sunday, July 1, 2001, at 06:46 AM, Pier P. Fumagalli wrote:
> >
> > > Aaron Bannert at [EMAIL PROTECTED] wrote:
> > >
> > >> On Sat, Jun 30, 2
many platforms (cross process and intra
process)
Isn't this going to bite prefork?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
gets the PROCESS_PRIVATE lock.
Note that APR currently doesn't have the right information now to set
APR_PROCESS_LOCK_MECH_IS_GLOBAL properly, so we end up getting two
locks instead of one on some platform/mechanisms where it isn't really
needed.
But assuming that we get more hints o
resumably we'd want such modules to
work with APR with Apache 2.0
I'm a bit surprised that none of the folks who were around when
CROSS_PROCESS vs. LOCKALL was invented have participated in the
discussion. I think I'm at least as concerned with that as with
losing a lock flavor.
Justin Erenkrantz <[EMAIL PROTECTED]> writes:
> On Tue, Jul 03, 2001 at 02:46:39PM -0400, Jeff Trawick wrote:
> > Note that APR currently doesn't have the right information now to set
> > APR_PROCESS_LOCK_MECH_IS_GLOBAL properly, so we end up getting two
> > locks
ere...
. I have no idea how important it is to have
APR_HAS_PROC_PTHREAD_SERIALIZE on HP-UX...
. I don't know of any other systems in this situation
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
%d\n", stat);
+exit(-1);
+}
+fprintf(stdout, "OK\n");
+
fprintf(stdout, "\tClient: Connecting to socket...");
stat = apr_connect(sock, remote_sa);
@@ -147,14 +155,6 @@
}
fprintf(stdout, "OK\n");
-fprintf(stdout,
ted :)
>
> messaging is cool: guaranteed transmission and data size:
> a combination of the best of TCP and the best of UDP.
>
> luke
>
> On Tue, Jul 10, 2001 at 01:43:49AM -0400, Jeff Trawick wrote:
> > (Unix at least; Win32 connect is a mess for non-blocking/time
xed with apr_ for
namespace protection reasons.
We might want to use a special prefix like apr_pvt_ for symbols like
this which are only for APR to use internally. I dunno...
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
on the current
CROSS_PROCESS semantics. It won't work with LOCKALL semantics.
(This program shows me which lock mechanisms block out other threads
in the same process. I'd be lost without it :) )
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://ww
Dale Ghent <[EMAIL PROTECTED]> writes:
> In Solaris 8, it is preferable to use the IPv6-derived getaddrinfo() (as
> we already do)
FYI... On all platforms it is preferable to use getaddrinfo().
getipnodebyname() was in some earlier IPv6 basic API RFC but is not in
2553bis.
--
any).
I thought this was going to be replaced by something which was
provided a user id and password and attempted to valid it. At least
that is more generally available.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
r_pool_t *context;
apr_size_t nbytes;
apr_status_t rv;
char *buf;
char msgbuf[120];
if (apr_initialize() != APR_SUCCESS) {
exit(-1);
}
atexit(apr_terminate);
if (apr_pool_create(&context, NULL) != APR_SUCCESS) {
exit(-1);
}
if ((rv = apr_file_pipe_cre
if (hostname != NULL) {
> struct addrinfo hints, *ai;
> apr_sockaddr_t *cursa;
>
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
(*in)->cntxt = p;
-(*in)->fname = ""; // What was this??? : apr_pstrdup(p, "PIPE"); */
+(*in)->fname = NULL;
(*in)->pipe = 1;
(*in)->timeout = -1;
(*in)->ungetchar = -1;
@@ -174,7 +174,7 @@
(*out) = (apr_file_t *)apr_pcalloc(
(repeating the msg, Bill, since I inadvertently sent it direct to you
the first time)
"William A. Rowe, Jr." <[EMAIL PROTECTED]> writes:
> From: "Jeff Trawick" <[EMAIL PROTECTED]>
> Sent: Monday, July 23, 2001 11:23 AM
> ;dev
>
>
hat apr
> does not yet have an apr_gethostbyname function?
see apr_sockaddr_info_get()
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
should be a no-op plain cleanup or the close-pool code
should not call plain_cleanup if that ptr is NULL.
run_cleanups() unconditionally calls cleanup->plain_cleanup(), so
this tries to branch to zero if this is the inherit cleanup.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at
hints.ai_flags = AI_CANONNAME;
> +if (strcmp(hostname, "0.0.0.0") == 0) {
> +hints.ai_flags = AI_NUMERICHOST;
> +}
> +else {
> +hints.ai_flags = AI_CANONNAME;
> +}
> hints.ai_family = family;
> hin
== -1 && errno == EINTR);
> +
> +/* Solaris returns EAGAIN even though it sent bytes on a non-block
> sock */
> + if (rv == -1 && errno != EAGAIN) {
> +rv = errno;
> +return rv;
> +}
--
Jeff Trawick | [EMAIL PROTECTE
With this patch
1) we don't call wait_for_io_or_timeout() after successfully sending
data
2) we try again to send data after wait_for_io_or_timeout() finds that
the socket is now writable
3) the code looks more like the other APR network write operations
--
Jeff Trawick | [
Jeff Trawick <[EMAIL PROTECTED]> writes:
> With this patch
>
> 1) we don't call wait_for_io_or_timeout() after successfully sending
>data
>
> 2) we try again to send data after wait_for_io_or_timeout() finds that
>the socket is now writable
>
> 3)
(normal e-mail machine busted; hope this doesn't go out as HTML or otherwise
misformatted :( )
> arv = 0;
> do {
> /* Clear out the repeat */
> repeat = 0;
>
> /* socket, vecs, number of vecs, bytes written */
> rv = sendfilev(sock->socketdes, sfv, vecs, &
gt;
>
> > trawick 01/07/27 09:54:44
> >
> > Modified:memory/unix apr_sms_threads.c
> > Log:
> > gotta use apr_os_thread_equal() instead of comparing apr_os_thread_t
> > directly; the latter doesn't work on OS/390
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
OL can't be invoked more than once, so this
configure.in:83:AC_PROG_LIBTOOL
and the resulting configure (bogus since autoconf exits with an error)
creates a broken apu_select_dbm.h.
(platforms: Linux, HP, AIX, Tru64, etc. libtool version 1.3.5 on
Linux, 1.3beta on AIX, dunno what version elsewh
[EMAIL PROTECTED] writes:
> rbb 01/07/30 08:09:02
>
> Modified:.configure.in
> Log:
> Grab the libtool.m4 files from APR. This fixes the bug Jeff just reported.
> You will only see the bug if you have removed the old aclocal.m4 file.
c00l
f APR_HAS_THREADS
stuff
#endif
}
#endif
I zapped the uninteresting inner check for APR_HAS_THREADS.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Is anybody gonna be aggravated if I change test apps to exit with
status zero if they work as expected?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
the app.
An app is either going to have to code threads to always exit with a
result or its code to join can't do anything useful with the thread
status.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
dy been committed.
I'll try to find some time in the next 24hrs to look over the patch
and hopefully commit, assuming nobody voices any disagreements.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Aaron Bannert <[EMAIL PROTECTED]> writes:
> On Wed, Aug 01, 2001 at 05:21:52PM -0400, Jeff Trawick wrote:
> > Aaron Bannert <[EMAIL PROTECTED]> writes:
> >
> > > How about instead of assuming APR_SUCCESS we just leave it undefined?
> > > This seems
at wasn't vestigial. Now, if APR_HAS_THREADS isn't defined, this file
> won't compile, and
> make test will fail. :-(
Try it (or believe my explanation to Ian).
Have fun,
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
It fails for me on various Unix boxen at the point that we try to find
out how much shared memory is available. mm_available() gets called
but it fails because a semaphore descriptor is bogus (gets EBADF on a
lock request).
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site
35
> @@ -173,6 +173,11 @@
>}
>}
>
> +APR_DECLARE(void) apr_thread_yield()
Shouldn't this have another "void", as in
APR_DECLARE(void) apr_thread_yield(void)
With no parameter list declaration it looks like K&R.
--
Jeff Trawick | [EMAIL PROTECTED] | P
I'm happy to create a new APR_INCOMPLETE_CHAR status, hijack the
current doc for APR_INCOMPLETE, and change Apache/APR as appropriate.
Does APR_INCOMPLETE (for incomplete file status) need to be renamed as
well for clarity? APR_INCOMPLETE_STATUS may be sufficiently clear.
Comments?
--
"William A. Rowe, Jr." <[EMAIL PROTECTED]> writes:
> From: "Jeff Trawick" <[EMAIL PROTECTED]>
> Sent: Friday, August 03, 2001 7:28 AM
>
>
> > I'm happy to create a new APR_INCOMPLETE_CHAR status, hijack the
> > current doc for APR_IN
increment reg A
store reg A at x
x has been incremented twice and now has value 1
compare-and-swap can be used for this.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Bo
want to assume that any of them
are... certainly the customers I'm familiar with would not accept that
as an answer :)
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
ent? I don't recall any (in the implementation language
used on the product, a special keyword on a variable declaration
caused compare-and-swap logic to be generated; very hard to screw up)
other several instruction sequences? yep (one cause was that even
when the proper instruction sequence wa
result (as opposed to an error result). For
> example,
> where APR_OOPS means our buffer filled, and APR_UGH means we are still
> looking...
okay, okay, you're right of course :)
The simple fact is that I barely have time now to try to make truly
broken stuff work properly; I th
may be that a header file would then be missing.
I do know that if dsos are disabled we compile okay on Darwin now
(well, last week :) ).
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
ection. It is much better to find an error (e.g., ECONNREFUSED)
here than have a subsequent read or write fail. Applications don't
tend to have as friendly an error path when send or recv fail as they
do when connect fails since a connect failure is normal behavior (user
specified bad i
ust
readability and writability, then he calls getsockopt(,SOL_SOCKET,
SO_ERROR,) to retrieve the saved error from TCP.)
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
t; > + *Win32 way to rewind?
>
> No. By closing (and marking the handle as closed) the next apr_dir_read
> will reopen the directory walk. There is no such thing as an 'open dir'
> on win32, you are either reading the first, or the next entry.
thanks fo
ywhere
instead of just some places.
The text "a non-blocking connect has succeeded or has definitely
failed when the select pops successfully" is more appropriate than
what you mention above.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities
y "int" for SO_ERROR.
"int" for SO_ERROR would indicate the type of the 4th parameter to
getsockopt(), not the type of len, which is the same for all
levels/options
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
supports IPV6
I'm working on a different patch at the moment.
Thanks,
Jeff
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
s to.
> is anybody else seeing this?
I think so.
/----
a related note:
I suspect that on the platforms where IPv6 is disabled but
getaddrinfo() is used APR needs to change family=APR_UNSPEC on input
to family=AF_INET.
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
use ../apr/build/rules.mk has been wiped out.
What sort of problem are you hitting with apr-util coming first?
(welcome back)
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
el better if we had apr_initialize() call
apr_unix_setup_time() to do this?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
rv = errno;
> break;
> }
> +thefile->dataRead = bytesread;
> thefile->filePtr += thefile->dataRead;
> thefile->bufpos = 0;
> }
I just started playing with /error/HTTP_NOT_FOUND.html
Jeff Trawick <[EMAIL PROTECTED]> writes:
> I just started playing with /error/HTTP_NOT_FOUND.html.var at
> OtherBill's suggestion. Even with this patch, the seek back to the end
> of the de body is not going to the right place.
>
> gotta read some code...
I know the
, apr-util's makefiles
would remove it during *clean
Actually, #2 seems pretty simple.
Comments anyone?
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
l then apr should still be buildable.
as I understand it: my change didn't cause anything to be altered in
apr when you clean apr-util; my change didn't affect whether or not
apr is buildable when you rebuild apr-util
--
Jeff Trawick | [EMAIL PROTECTED] | PGP public key at w
ot; in
*bsdi*)
-INCLUDE_RULES=".include \"$top_builddir/../apr/build/rules.mk\""
+INCLUDE_RULES=".include \"$top_builddir/build/rules.mk\""
;;
*)
-INCLUDE_RULES="include $top_builddir/../apr/build/rules.mk"
+INCLUDE_RULES=&qu
801 - 900 of 1477 matches
Mail list logo