Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-4990399

JDWP: JFrame construction performance regression over 1.4.2 (method entry err?)

    XMLWordPrintable

Details

    Description

      CAP bug report was originally 4969969. Since we were able to make some
      basic improvements on the debugger side, we have kept that bug number,
      this bug is being filed to track the method call increase
      in JFrame construction.

      It appears that the initial JFrame constructor is causing many more
      methods to be executed than in 1.4.2, from 15,000 methods to 377,000
      methods. Since the debugger 'step' functionality monitors the method
      entry events, it experienced a slowdown.

      This increase in method calls may be worth some investigation.

      =======================================================================


      Original debugger bug filed:

      ###@###.### 2003-12-17

      J2SE Version (please include all output from java -version flag):
          java version "1.5.0-beta"
          Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0-beta-b30)
          Java HotSpot(TM) Client VM (build 1.5.0-beta-b30, mixed mode)

      Does this problem occur on J2SE 1.3 or 1.4.x? Yes / No (pick one)
          No, 1.4.2 only takes 1 sec. b30 takes 18 sec.

      Operating System Configuration Information (be specific):
          Windows XP (version 5.1 Service Pack 1 (Build 2600))

      Hardware Configuration Information (be specific):

          Platform configuration:

          CPU: Intel 32bit Single Processor
          Processor Speed: 1893 MHz

          Operating System:
          Windows XP (version 5.1 Service Pack 1 (Build 2600))

          DirectX version: DirectX 8.1 or greater

          DirectX display devices:
          Device 0:
            Resolution (width x height x depth): 1024 x 768 x 32
            Device driver nv4_disp.dll version 6.13.10.3638

          GDI display devices:
          Primary: Name = \\.\DISPLAY1, Description = NVIDIA GeForce4 460 Go (Toshiba)

      Bug Description:

         Testing the VisualMust debugger with B30, there is a problem :
         when running the application step by step and entering into a "super()"
         to create a frame for example, the step takes 18 seconds and 100 % CPU
         (with a Pentium 4, 1900 MHz). With the same machine, with Java 1.4.2.
         the step takes less than 1 second.

      Steps to Reproduce (be specific):
        - unzip VisualMust.zip
        - run java -Xmx128m -classpath ./VisualMust.jar;./VisualMustDocs.jar;./JbcBeansEvaluation.jar;c:/jdk/lib/tools.jar VisualMust

      Attachments

        Issue Links

          Activity

            People

              serb Sergey Bylokhov
              ohair Kelly Ohair (Inactive)
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:
                Imported:
                Indexed: