At 09:34 AM 8/11/00 +0000, Gregor Okorn wrote:
>To: [EMAIL PROTECTED]
>Subject:
>--text follows this line--
>
>Please enter the details of your bug report here
>
>Hello Paul,
>
>Regularly and frequently, I observe hanging with JDebug. Typically, I
>can get it to stop at the first breakpoint I set, but usually, any
>attempts to step into or over just hang the application. Sometimes it
>succeeds in executing one step before hanging, and other times it may
>go as far as five steps or so. I haven't noticed a pattern yet, and
>was hoping to notice some pattern before actually sending this
>report, but since I'm too impatient to wait any longer, I'm hoping you
>might suggest some change to my debugging process to ferret out this
>problem. I really really want to use JDebug for my Java debugger,
>since it appears to be quite useful and much more efficient than
>polluting
>my code with System.out.println() debugging statements.
>
>Coincidentally, I saw the email from Lauri Tynor on 8/8/2000 regarding
>a similar observation with JDebug, and was kind of hoping to see some
>response to her issue that would address my issue at the same time -
>but so far haven't.
>
>Thanks for any insight into this problem,
>
>JDE> 8 51 step over 1
>
>
>(jde-dbo-event-set
>8 "all"
>(list "Thread" 1 "main" "runnable" "suspended by debugger"
>(list
>(list 0 "com.basystems.netman.cli.impl.CLICommand" "CLICommand.java"
>1533 "displayData")
>(list 1 "com.basystems.netman.cli.impl.CLIBasSonetLineCommand"
>"CLIBasSonetLineCommand.java" 268 "showInterface")
>(list 2 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 27376
>"ShowInterfacePosCmd")
>(list 3 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 10609
>"ShowPosInterfaceCmd")
>(list 4 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 10670
>"ShowInterfaceCommand")
>(list 5 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 363
>"ShowCmds")
>(list 6 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 159
>"Expression")
>(list 7 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 105
>"TopLevelExpression")
>(list 8 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 37
>"Start")
>(list 9 "com.basystems.netman.cli.parser.BasCliBase" "BasCliBase.java"
>1107 "processline")
>(list 10 "com.basystems.netman.cli.parser.BasCliBase" "BasCliBase.java"
>1081 "processInput")
>(list 11 "com.basystems.netman.cli.parser.BasCli" "BasCli.java" 22
>"main"))
>(list)
>nil)
>(list 'jde-dbo-step-event (list
>"com.basystems.netman.cli.impl.CLICommand" "CLICommand.java" 1533)))
>
>
>(jde-dbo-command-result 51)
>
>JDE> 8 52 get_locals 1 0
>
>JDE> 8 53 get_this 1 0
>
>
The debugger is hanging because JPDA is not responding to its request to
return the local variables and the fields of the "this" object of the
current stack frame. This could be because the class was not compiled with
all debug information or because there is a bug in JDEbug or JPDA or there
is some special problem with getting local variable and this object
information when debugging an attached process. Clearly this problem needs
further investigation. It would help me if you and other users could run a
few tests and see if the hangup regularly occurs after get_this. Meanwhile,
in the next release, which I plan to make tonight, as a workaround, I will
add a customization variable, that will allow you to enable/disable
automatic locals and/or this display while stepping plus commands to do
this manually.
- Paul
