non-display of assembly

non-display of assembly

[Tom Tromey]

Today, for a change, I'm debugging library code.  I'm trying to figure
out why a particular call to memmove() is failing.  Thinking that the
compiler might be inlining memmove(), I want to put the source window
into assembly mode.

This doesn't work.  The mouse pointer, when over the source window, is
a watch.  The source window itself is blank.  I don't get any sort of
error message.

Trying to switch the view back to source does not work.

I didn't try the "tk whatever" trick to reset it.

Any clues why this might happen?  The `disas' command works in the
console, but of course I'd rather not use that.

Tom

Re: non-display of assembly

[James Ingham]

Tom,

Some unexpected error was hit which messed up Insight.  The best first 
stab at debuging this sort of thing is to do:

setenv GDBTK_DEBUG 2

in the shell from which you launch gdb.  Then do whatever you normally do,
keeping an eye on the debugging window.  Errors will show up in an
orangish color.  This might give you a bit of a hint...

BTW, if you forget to set this env. var, you can also go to the
console, & type:

(gdb) tk ManagedWin::open DebugWin

Otherwise, I need a recipe for reproducing your bug before I can be
of much help.

Jim

 > Today, for a change, I'm debugging library code.  I'm trying to figure
 > out why a particular call to memmove() is failing.  Thinking that the
 > compiler might be inlining memmove(), I want to put the source window
 > into assembly mode.
 > 
 > This doesn't work.  The mouse pointer, when over the source window, is
 > a watch.  The source window itself is blank.  I don't get any sort of
 > error message.
 > 
 > Trying to switch the view back to source does not work.
 > 
 > I didn't try the "tk whatever" trick to reset it.
 > 
 > Any clues why this might happen?  The `disas' command works in the
 > console, but of course I'd rather not use that.
 > 
 > Tom
 >