minutes 2006-09-19

[Stan Cox <scox at redhat dot com>]

Utrace (Chris <moller>)
 -ptrace on utrace behavior appears to be different than original utrace
 -If a kill fails then the next ptrace operation is ignored (#203385?)
 -Thread/Process status when a process clones and then later kills the
  clone (#?)
 -Action:  Get the C version of the test for 203385 written and
 committed

 Backtrace (Mike <mcvet>)
 -Most of Alex's patches committed
 -Source display for previous stack frames works
 -Action: Investigate the lost frames in backtraces (+cagney)

 Fstack (Nurdin <npremji>)
 -What is the correct way to handle stack backtracing?
  Perhaps use the cascading observer model which simplifies iterating  
  through stack frames.
 -Action: Work out multi-threaded fstack (+pmuldoon)
 -Action: Investigate the gdb pstack bug (+cagney)

 64 bit (Tim <timoore>)
 -32 on 64 bit is mostly done
 -Adding automatic build 32 on 64 testing
 -IBM supplied some of the 32 on 64 testing framework infrastructure
 -ppc executables default to 32 bit to save cache space due to the large
  register set.
 -Resolved modulo status of the testsuite
 
 Breakpoint (Mark <mjw>)
 -Three steps needed for high level single threaded breakpoints:
  1. single instruction stepping
  2. address information for file/line
  3. stack frame information
 -Multithreading breakpoint support more difficult (Cagney)
 -Action:  Con call high-level breakpoints (+mjw, +cagney as needed)

 Disassemble (Mike <mcvet>)
 -Single threaded case mostly done
 -Need to define semantics for multiple threaded case, which currently
  blocks the main thread
 -Mostly resolved.  This topic can probably transition to handling multi
  threaded displays

 Testing & Dogtail (Phil <pmuldoon>)
 -Demo by Len
 -Is current state okay for the testing framework?
 -Memory smashing bug has proved elusive.
  -Seems related to cleanup in java gnome
 -Tom Tromey's suggestion for two different objects to handle
  enter/exit may simplify handling complicated cases)
 -Status needed on dogtail, memory smashing bug, Tom's suggestion
 -Action: Is dogtail okay for the testing framework (+ldimaggio)
 -Action: Beat the Java-GNOME memory smash to death (+cagney)

 Source (Stan <stan>, Rick <rmoseley>)
 -Simple variables for current frame now handled. 
 -Looking at prior frames support with libunwind.
 -Looking at gui integration
 -Action: Con call on variables in source window 

 Coverage (Stan, Andrew)
 -Investigate infrastructure changes needed to support gcov

  Misc
 -Investigate vte bug which causes intermittent console window problem
   (Andrew?)
 -Come up with workflows for terminal window (Phil?)
 -Chris problems with demise of cipe
 -Action: Get closure on the console (cagney)
 -Action: cipe/vpn solution (moller, cagney)