This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Omniscient Debugging for C...
- From: Bil Lewis <bil at bayarea dot net>
- To: themadrasi <themadrasi at hotpop dot com>
- Cc: "Joseph S. Myers" <jsm at polyomino dot org dot uk>, gcc at gcc dot gnu dot org, replaydebugger at yahoo dot co dot in
- Date: Wed, 9 Jun 2004 05:33:42 -0700 (PDT)
- Subject: Re: Omniscient Debugging for C...
Joseph,
I see why you were unimpressed.
The project below does indeed have a vague relation to omniscient
debugging, but it's *very* vague. As a programmer using the two
tools, they would appear completely unrelated. Hence, trying to
the GCC tool in place of the ODB would convince you that there's
not much there.
Let me assure you that omniscient debugging *IS* incredible.
(Don't take my word for it. Try it yourself. It's on my site.
Should take you ten minutes.)
Then we can look at the question. Is there anyone interested
in creating history?
-Bil
(I am not knocking the GCC project. It's not supposed to be an
ODB. I suspect it's quite good for what it does.
On Fri, 4 Jun 2004, themadrasi wrote:
> Joseph S. Myers wrote:
>
> >On Thu, 3 Jun 2004, Bil Lewis wrote:
> >
> >
> >
> >>I've written a new kind of debugger -- it records state changes
> >>and allows the programmer to essentially "go backwards in time."
> >>
> >>
> >
> >Cf. <http://www.uwsg.iu.edu/hypermail/linux/kernel/9901.2/0531.html>:
> >
> ># I have enough information available in the proxy ptrace filter to
> ># implement PTRACE_SINGLESTEP_BACKWARDS. How would you like to have that
> ># capability in gdb? "Execute backwards until this data watchpoint
> ># changes." Imagine a graphical debugger with a scrollbar for time,
> ># where the top is "beginning of execution" and the bottom is "end of
> ># execution."
> >
> >(very specific to Linux 1.3.42, and with other limitations).
> >
> >
> >
> http://opensource.codito.com/lizard is a more recent project for a linux
> 2.4 series kernel using a later gdb version and gcc 3.3 . works
> currently only for single threaded apps.
>
> cheers
> ramana
>
>
>