OpenSAML user discussion

["George Stanchev" <>]

I still haven't got to building on linux, but for solaris, here is what
I found so far:

acx_pthread.m4, line 108, if you switch from

"acx_pthread_flags="-pthread -pthreads pthread -mt $acx_pthread_flags"

to

acx_pthread_flags="-pthreads -pthread pthread -mt $acx_pthread_flags"

It fixes part of the problem. I googled around and found other
open source projects's CVS or SVN commits in which they fix this file
in this way. It basically forces the configuration scripts to look
for the right option first on solaris.

For some reason, there are two such options emitted at build time and
I cannot find which one generates the second one. Will look at it
Monday or over the weekend. I must admit I am unix configuration
newbie so it might take me a bit longer than someone who knows that
stuff.

George

-----Original Message-----
From: Scott Cantor 
[mailto:]
 
Sent: Thursday, May 10, 2007 6:23 PM
To: 

Subject: RE: building cpp-xmltooling on unix

> I think can look into that -pthread issue for a workaround and will 
> let you know if I find anything.

Are you saying that gcc returns an error correctly on Linux when the
configure test runs, but not on Solaris? That would explain a lot of
glitches people seem to have there, I imagine.

> at the top of that Threads.h file, it'd be great.

It's added, Windows has a signal header too.

> Its worth mentioning that even with the g++ option being wrong, the 
> library biulds and links fine with no other errors or warnings.

Alright, well, CC doesn't, it's not close to building yet for me. I may
well check in fixes that break gcc until I port and test on Linux again.

-- Scott




**********************************************************************
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. Any unauthorized review, use, disclosure or distribution is 
prohibited. If you are not the intended recipient, please contact the sender 
by reply e-mail and destroy all copies of the original message.