https://issues.apache.org/bugzilla/show_bug.cgi?id=50651
--- Comment #8 from Konstantin Kolinko <knst.koli...@gmail.com> 2011-01-26 19:45:54 EST --- Created an attachment (id=26554) --> (https://issues.apache.org/bugzilla/attachment.cgi?id=26554) catalina.2011-01-27.log (In reply to comment #7) > Whether the recycle method should be idempotent is a separate discussion. > I agree that it should, but it doesn't mean it should be called multiple > times. > Cause the bugs that come up from coding like that, is way harder to track > down. I am sure that if we skip recycling the processor before its reuse in some case, the consequences will be harder than if we recycle it twice. The code is already hard to follow at some places. > So, why not fix the real problem ? I think the bug in recycle() is the real problem. Though I do not mind patching CoyoteAdapter.log() as well. Part of the necessary changes is in r1059548 [2] [2] http://svn.apache.org/viewvc/tomcat/trunk/java/org/apache/catalina/connector/CoyoteAdapter.java?r1=1059548&r2=1059547&pathrev=1059548 Here is one more occurrence of the issue, just using the browser (Firefox 3.6.13): 1. Using Tomcat 6.0.31. Configure Tomcat with Http11NioProtocol connector and AccessLogValve in <Host> 2. Clear cache of the browser 3. Access http://localhost:8080/examples/jsp/plugin/plugin.jsp 4. Wait for more than 10 seconds. 5. The pair of exceptions occur: 27.01.2011 3:09:37 org.apache.coyote.http11.Http11NioProtocol$Http11ConnectionHandler process SEVERE: Error reading request, ignored java.lang.NullPointerException at org.apache.coyote.http11.InternalNioOutputBuffer.recycle(InternalNioOutputBuffer.java:351) at org.apache.coyote.http11.Http11NioProcessor.recycle(Http11NioProcessor.java:1007) at org.apache.coyote.http11.Http11NioProtocol$Http11ConnectionHandler.release(Http11NioProtocol.java:655) at org.apache.coyote.http11.Http11NioProtocol$Http11ConnectionHandler.process(Http11NioProtocol.java:751) at org.apache.tomcat.util.net.NioEndpoint$SocketProcessor.run(NioEndpoint.java:2262) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:662) 27.01.2011 3:09:37 org.apache.tomcat.util.net.NioEndpoint$SocketProcessor run SEVERE: java.lang.NullPointerException at org.apache.coyote.http11.InternalNioOutputBuffer.recycle(InternalNioOutputBuffer.java:351) at org.apache.coyote.http11.Http11NioProcessor.recycle(Http11NioProcessor.java:1007) at org.apache.coyote.http11.Http11NioProtocol$Http11ConnectionHandler.release(Http11NioProtocol.java:655) at org.apache.coyote.http11.Http11NioProtocol$Http11ConnectionHandler.process(Http11NioProtocol.java:776) at org.apache.tomcat.util.net.NioEndpoint$SocketProcessor.run(NioEndpoint.java:2262) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:662) -- see attachment. Access log: 127.0.0.1 - - [27/Jan/2011:03:09:26 +0300] "GET /examples/jsp/plugin/plugin.jsp HTTP/1.1" 200 1655 127.0.0.1 - - [27/Jan/2011:03:09:26 +0300] "GET /favicon.ico HTTP/1.1" 200 21630 127.0.0.1 - - [27/Jan/2011:03:09:27 +0300] "GET /examples/jsp/plugin/applet/Clock2.class HTTP/1.1" 304 - I am not able to reproduce this with usual static or JSP pages. I suspect that some cause is that Http11NioProtocol#process(NioChannel) there are calls to release(socket, processor) in "if (state == SocketState.OPEN)" and in "else" branches -- Http11NioProtocol.java:751 is where the exception occurs, but Http11NioProcessor#process(NioChannel) called from there already calls recycle() by itself. I think this comes from r1030613. There were later changes in Http11NioProcessor to add to this. The second exception in the pair, at (Http11NioProtocol.java:776) occurs after a catch block. This comes from r1033341. In this place it is hard to guess whether recycle() call already happened or not. The reproduction recipe in comment 5 is different from the both of mine ones. -- Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
