>>>> "AP" == Andy Piper <andy(a)beasys.com>
writes:
AP> Jeff Miller wrote:
> xemacs compiled ok, but it core dumps upon startup. What is
odd,
> though, xemacs -vanilla & xemacs -q will start up.
AP> Hmmn, interesting. Well x_resize_window is trying to resize a tabs
AP> window by the looks of things. Can you see what xwininfo reveals about
AP> the subwindow structure?
is this the type of info you are looking for?
xwininfo -tree
xwininfo: Please select the window about which you
would like information by clicking the
mouse in that window.
xwininfo: Window id: 0x3c0002d "emacs: *scratch*"
Root window id: 0x26 (the root window) (has no name)
Parent window id: 0x8005df (has no name)
1 child:
0x3c0002e (has no name): () 584x606+0+0 +564+254
5 children:
0x3c000ad (has no name): () 15x484+573+88 +1137+342
0x3c0003f (has no name): () 128x21+0+0 +564+254
1 child:
0x3c00040 (has no name): () 128x21+0+0 +564+254
1 child:
0x3c00041 (has no name): () 126x65530+1+26 +565+280
0x3c00031 (has no name): () 128x21+4+67 +568+321
1 child:
0x3c00036 (has no name): () 128x21+0+0 +568+321
1 child:
0x3c0003e (has no name): () 126x65530+1+26 +569+347
0x3c00030 (has no name): () 584x580+0+26 +564+280
0x3c0002f (has no name): () 582x24+0+0 +564+254
AP> Can you also find out what in your environment is triggering the
AP> crash?
it happens when flyspell is loaded.
AP> Also can you try changing x_finalize_image_instance so that it
AP> writes something like 0xdeadbeef to the clip and subwindow slots -
if you can tell me what to do...... or give me an example.... i am not
sure what these "slots" are or how to set them.
AP> I'm trying to determine whether these are null because of GC or
AP> because they never got created in the first place. If you can
AP> also pop up the stack to update_frame_subwindows and see what the
AP> subwindow_cachel length is (and what it contains).
is it ok to just add a print statement or something in this
function..... I'm not much of a gdb wizard......
i have discovered after that fact that the xemacs -vanilla with your
patch will core dump.... if i start doing things. It's start up ok.
But a C-x C-f, M-x will core dump. I can send the backtrace if you
think it will be anymore useful than the previous ones.
i will say though that things seem to be improving. I probably have
not had an actually crash (prior to your informational patch) for
about a week.
Jeff