Warren & Chris,
I think you were both right about the problem with exceptions. The problem
happens when inside the try/catch block, a method throws another exception
that is not caught. I was able to duplicate this behavior with gcc and
without the use of mysql. After I placed the method in question into another
nested try/catch, the problem resolved itself inside my simple application.
However, the problem is still in mysql++. I will move forward without the
use of exceptions. I still would be curious to see why this happens.
Sincerely,
Alex
-----Original Message-----
From: Warren Young [mailto:mysqlpp@stripped]
Sent: Monday, February 14, 2005 1:10 PM
To: MySQL++ Mailing List
Subject: Re: mysql++1.7.28 thread safety
Earl Miles wrote:
> I run MySQL++ in a multi-threaded application. I have to modify the
> Makefile to add the thread safety flag to gcc (on Linux it's -pthread;
> it varies from platform to platform) and be sure to link using
> -lmysqlclient_r.
Check this out:
http://ac-archive.sourceforge.net/Installed_Packages/acx_pthread.html
Integrating that free macro into the autoconf process is the right way
to do the first part of what you did. This is in the Wishlist.
But before doing this change, we should move to using a config
subdirectory, and moving all the configuration tests into that subdir,
cleaning up configure.in. This is also a Wishlist item.
Once that is done, some autoconf test for selecting libmysqlclient_r
instead of the regular library should be added.
Patches thoughtfully considered for all of these, as always.
I'm probably going to make a MySQL++ release fairly soon, but if someone
gets back to me saying that you're going to work on this, I'll hold the
release until you finish your work.
--
MySQL++ Mailing List
For list archives: http://lists.mysql.com/plusplus
To unsubscribe:
http://lists.mysql.com/plusplus?unsub=1
