APR broken on FreeBSD 4?

2004-10-15 Thread Craig Rodrigues
Hi, Is anyone seeing APR fail to build on FreeBSD 4, but the FreeBSD ports cluster is hitting this. I can't reproduce this. -- Craig Rodrigues http://crodrigues.org [EMAIL PROTECTED] - Forwarded message from Kris Kennaway [EMAIL PROTECTED] - Date: Thu, 14 Oct 2004 15:05:12

Can a thread write-lock a rwlock that it already has read-locked?

2004-10-15 Thread David Barrett
I'm experimenting with the reader/writer locks and have a question: Can a thread write-lock a thread on which it already holds the read lock? My tests seem to indicate that this is not allowed, but I want to confirm this is by design and not something I'm just doing wrong. Specifically,

UTF-8 hostnames?

2004-10-15 Thread David Barrett
Does apr_sockaddr_info_get (along with all the functions that take hostnames) accept UTF-8 hostnames?

Re: APR broken on FreeBSD 4?

2004-10-15 Thread Joe Orton
On Thu, Oct 14, 2004 at 07:14:40PM -0400, Craig Rodrigues wrote: checking for LIBS needed for pthreads... (cached) -lpthread adding -lpthread to LIBS checking for pthread.h... (cached) yes ... checking for sigwait... no checking for poll... no It looks like after -lpthread is added to LIBS

Re: UTF-8 hostnames?

2004-10-15 Thread Joe Orton
On Thu, Oct 14, 2004 at 08:04:35PM -0600, David Barrett wrote: Does apr_sockaddr_info_get (along with all the functions that take hostnames) accept UTF-8 hostnames? No, not really. To do non-ASCII DNS you need to use IDNA to map the hostnames to valid ASCII strings: there's an excellent

Re: UTF-8 hostnames?

2004-10-15 Thread David Barrett
Thanks for the detailed answer! On Fri, 15 Oct 2004 12:24 am, Joe Orton wrote: On Thu, Oct 14, 2004 at 08:04:35PM -0600, David Barrett wrote: Does apr_sockaddr_info_get (along with all the functions that take hostnames) accept UTF-8 hostnames? No, not really. To do non-ASCII DNS you need to use

Re: APR broken on FreeBSD 4?

2004-10-15 Thread Paul Querna
Joe Orton wrote: On Thu, Oct 14, 2004 at 07:14:40PM -0400, Craig Rodrigues wrote: checking for LIBS needed for pthreads... (cached) -lpthread adding -lpthread to LIBS checking for pthread.h... (cached) yes ... checking for sigwait... no checking for poll... no It looks like after -lpthread is

SOCEINPROGRESS returned on non-blocking, zero-timeout socket

2004-10-15 Thread David Barrett
I'm seeing unexpected behavior when I use apr_opt_set( s, APR_SO_NONBLOCK, 1 ) in conjunction with apr_socket_timeout( s, 0 ). Basically, the following code generates no error return values: 1# apr_socket_create( s, APR_INET, SOCK_STREAM, APR_PROTO_TCP, pool ); 2# apr_opt_set( s,

Re: Can a thread write-lock a rwlock that it already has read-locked?

2004-10-15 Thread Joe Orton
On Thu, Oct 14, 2004 at 06:17:39PM -0600, David Barrett wrote: I'm experimenting with the reader/writer locks and have a question: Can a thread write-lock a thread on which it already holds the read lock? Hey, I answered that one already :) apr_thread_rwlock.h in 1.0 says: * Note:

Re: APR 1.0.0 and CYGWIN

2004-10-15 Thread Max Bowsher
William A. Rowe, Jr. wrote: I should have added that the pseudo-fork and condition vars on cygwin would actually be better preserved by using the unix port. In fact, if the cygwin build simply used the win32 apr_file_io code I'd be quite happy with it relying on the other (unambiguous) features

Re: APR broken on FreeBSD 4?

2004-10-15 Thread Jun Kuriyama
At Fri, 15 Oct 2004 10:09:01 -0400, Craig Rodrigues wrote: Joe Orton wrote: On Thu, Oct 14, 2004 at 07:14:40PM -0400, Craig Rodrigues wrote: checking for LIBS needed for pthreads... (cached) -lpthread adding -lpthread to LIBS checking for pthread.h... (cached) yes Hmm, I cannot reproduce