On 05/12/2014 22:56, Mark Thomas wrote: > On 05/12/2014 22:42, Rémy Maucherat wrote: >> 2014-12-05 23:22 GMT+01:00 Mark Thomas <ma...@apache.org>: >> >>>> My Tomcat trunk is quite broken right now (and the ci run isn't looking >>>> very good). It could be related to this. >>> >>> Hmm. I guess I must have missed something. Sorry about that. When I >>> looked through the code it looked surprisingly harmless. >>> >>> There are other fixes (AsynStateMachine, error handling) that I would >>> have thought would have been more likely to cause problems. >>> >>> I'll take a look over the weekend. If this gets in the way before I fix >>> it and you figure out which commit triggered the issue feel free to >>> revert it. >>> >> I have an apparent recycling issue (the first request to a webapp is fine, >> but then all others would fail), so I suspected this one. I switched to the >> revision before and it is fixed. >> The related bug doesn't sound like it really needs to be fixed. If it >> causes any impact, it would be better to say it is invalid. > > I think your point re Servlet 6.2.3 is sufficient to revert this and > resolve the bug as invalid. I'll do that shortly although I do want to > see if I can figure out what I missed.
I've found the problem. I've reverted my fix. One question I do have is that I wonder if this was an intended consequence of adding that language to the spec or whether it was only meant to apply to non-async mode. Mark --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
