Mail Archives: djgpp-workers/2002/06/13/18:44:08
Hello.
Charles Sandmann wrote:
[snip]
> Do this (important test). Launch cmd.exe. If you launch the task manager,
> sort by image name, check the number of NTVDM.EXE images. Run emacs,
> you should see a new NTVDM. Record it's PID. Do the CTRL thing, does it's
> PID go away when it exits? The cmd window will remain. Do you get any
> other info, or just dropped back to command line (when NTVDM aborts you
> typically just go back to the prompt with no message).
>
> This test determines if it happens without nesting (important fact), if
> there is any crash info (important fact), if NTVDM dies (important).
> Does it go away if you are not using unixy sbrk?
>
> If it happens standalone, and NTVDM dies, then it indicates something
> in the hardware interrupt routine isn't quite right.
Here's my testing method: Start up Emacs, holding down Ctrl. Then go to the
scratch buffer, select the text, copy and paste it lots of times (Ctrl+Y),
move around a bit using the cursors (pressing Ctrl fairly often to skip
words), try transposing a bit (Ctrl+T).
Here are the results with Emacs CVS built against DJGPP 2.03 refresh++
pre-release:
Emacs run from cmd.exe does not crash on start-up, when I hold down Ctrl. I
does bomb out after pressing Ctrl 20-30 times, say (maybe more, maybe less).
NTVDM disappears from the Task Manager. I don't get a traceback.
Emacs run directly from a shortcut seems pretty stable. I spent a couple
minutes doing operations involving Ctrl and it did not crash.
This is all with Unixy sbrk. If I disable Unixy sbrk, then Emacs works fine.
So, is this a nesting problem? What else can I do to help diagnose the
problem?
Bye, Rich =]
--
Richard Dawe [ http://www.phekda.freeserve.co.uk/richdawe/ ]
- Raw text -