tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jason Hall <>
Subject Re: jasper
Date Tue, 08 Nov 2016 13:35:18 GMT
Would like to share that it broke in version 8.0.37.

Just ran my app in Tomcat 8.0.36 and it ran fine.  I have included the zipped generated source
files from both.  File size is about the same, but did not count chars in the individual method(s)
- unless 8.0.36 ignores the counts.

Tomcat 8.0.37 gave the error:
08-Nov-2016 08:07:48.540 SEVERE [21] org.apache.catalina.core.ApplicationDispatcher.invoke
Servlet.service() for servlet action threw exception
 javax.servlet.jsp.JspException: ServletException in '/WEB-INF/jsp/rm_requisition_small_parcel_maintain.jsp':
Unable to compile class for JSP: 

An error occurred at line: [308] in the generated java file: [C:\Users\jason.hall\Documents\software\apache-tomcat-8.0.37\work\Catalina\localhost\FACTS2\org\apache\jsp\WEB_002dINF\jsp\]
The code of method _jspService(HttpServletRequest, HttpServletResponse) is exceeding the 65535
bytes limit

while 8.0.36 did not

the output from 8.0.36 from my flow goes:

08-Nov-2016 08:01:32.446 INFO [http-apr-9092-exec-28] Exiting, going to >>>
08-Nov-2016 08:01:32.446 INFO [http-apr-9092-exec-28] This action took 0ms to execute
08-Nov-2016 08:01:32.647 INFO [http-apr-9092-exec-28]
>>> Initializing
08-Nov-2016 08:01:35.681 INFO [http-apr-9092-exec-28] Exiting, going to >>>
08-Nov-2016 08:01:35.682 INFO [http-apr-9092-exec-28] This action took 3034ms
to execute
08-Nov-2016 08:01:47.453 INFO [http-apr-9092-exec-32] Exiting, going to >>>
08-Nov-2016 08:01:47.453 INFO [http-apr-9092-exec-32] This action took 2ms to
08-Nov-2016 08:01:47.494 INFO [http-apr-9092-exec-24]
>>> Initializing

You can see that the flow normally happened as it should in 8.0.36.  It allowed me to go to
"RmRequisitionSmallParcelMaintainForm" after behind the scenes compilation.

I did not attach the .class file, but that was obviously also generated in 8.0.36....


----- Original Message -----
From: "Jason Hall" <>
To: "Tomcat Users List" <>
Sent: Tuesday, November 8, 2016 7:27:23 AM
Subject: Re: jasper

BTW - this application uses the struts 1.2.7 framework -as you might be able to tell with
the tags in the JSP if you are familiar with struts 1.

This might take a little time, but I can attempt to pick different versions of tomcat to determine
which version generates the "too big" java code to compile if you wish, unless you know what
versions of tomcat with the associated jasper changes that may have caused this ?


----- Original Message -----
From: "Jason Hall" <>
To: "Tomcat Users List" <>
Sent: Tuesday, November 8, 2016 7:22:57 AM
Subject: Re: jasper

Problem appears to be in the _jspService method from first "try".
I have included both tomcat's and glassfish's generated java from the JSP.  This JSP's java
is generated from an initial selection from a previous JSP page.  I selected the EXACT selection
from the previous JSP on both glassfish and tomcat deployments to get the generated java code.

I also included the raw JSP.

If the text attachments do not show up chris, let me know.


----- Original Message -----
From: "Christopher Schultz" <>
To: "Tomcat Users List" <>
Sent: Monday, November 7, 2016 10:29:14 AM
Subject: Re: jasper

Hash: SHA256


On 11/4/16 2:47 PM, Violeta Georgieva wrote:
> Hi,
> 2016-11-04 19:40 GMT+02:00 Jason Hall <>:
>> Possible issue with generated source from jsp(s).....
>> I current installed Tomcat 8.5.6 and use it with netbeans.
>> On one of my JSP files, it is creating a source file with a
>> try/catch
> block that is MUCH greater than the 64K limit set by the JVM.   On 
> glassfish and websphere, this does not occur.
>> I haven't checked the size of the method in websphere, but did
>> save the
> generated source on glassfish and the method is 57K (still under
> the 64K limit).  The tomcat one was almost 800K I think.
>> The older netbeans 8.0 version of tomcat that is bundeled with it
>> does
> not have this problem (version
> Check this

This example seems a little outrageous, though: the difference is by
an order of magnitude, not by a few percent like in the bug report you

Jason, would it be possible to provide a sample JSP that balloons in
this way when compiled with Tomcat's JSP compiler?

- -chris
Comment: GPGTools -
Comment: Using GnuPG with Thunderbird -


To unsubscribe, e-mail:
For additional commands, e-mail:

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message