On 11/03/2015 08:19, Mark Thomas wrote: > On 11/03/2015 08:01, Keiichi Fujino wrote: >> 2015-03-11 16:50 GMT+09:00 Mark Thomas <ma...@apache.org>: >> >>> On 11/03/2015 06:19, Keiichi Fujino wrote: >>>> 2015-03-11 7:35 GMT+09:00 <ma...@apache.org>: >>>> >>>>> Author: markt >>>>> Date: Tue Mar 10 22:35:19 2015 >>>>> New Revision: 1665736 >>>>> >>>>> URL: http://svn.apache.org/r1665736 >>>>> Log: >>>>> Stop re-using the SocketWrapper >>>>> With the introduction of upgrade and non-blocking, I/O can occur on >>>>> non-container threads. This makes it near impossible to track whether a >>>>> SocketWrapper (== connection) is still referenced or not, making re-use >>> a >>>>> risky proposition. >>> >>> <snip/> >>> >>>> I think this fix causes NPE. >>> >>> Quite possibly. >>> >>>> Do we need initialization of poller and socketBufferHandler in >>>> SocketWrapper? >>> >>> Again, quite possibly. >>> >>> I have been focussed on the crashing test case and this change didn't >>> break that (neither did it fix it). I haven't even done a basic smoke >>> test of these changes yet. What is the simplest way to reproduce this? >>> >>> >> I think this is reproduced by simple HTTP access. >> e.g. access to http://localhost:8080 by browser. > > Whoops :) > > Fixed.
I may have spoken too soon. Browser access works but I'm seeing lots of errors in the logs. They might be related to my local APR changes but maybe not. I'll take a closer look. Mark --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
