This is the mail archive of the
mailing list for the GDB project.
Re: GDB MI Reverse Commands added [1 of 3]
Jakob Engblom wrote:
Can someone familiar with the main reverse code illuminate me on how to call
into the reverse code in teh right way? It seems that the reverse.c code has
already changed from the code that was (bad style) copied into mi-main.c...
which is exactly what we want to avoid in the first place.
Or is the logical thing to do from MI to just callsubmitted has rotted since the patch was submitted...
reverse_continue (char *args, int from_tty)
exec_reverse_once ("continue", args, from_tty);
As the implementation of reverse continue? This does looks like the we
Just to clarify -- who's got the ball here? Calling 'reverse_continue' sounds
like reasonable thing to me
if one wants to do a reverse continue ;-)
I think the ball is mine.
But I need to understand how to put the two things together, and for that I need
some help by the people who did reverse.c.
That would be me, but I'm confused.
What changed? Reverse.c didn't change (I don't think...)
If MI used to work, why doesn't it work any more?
In particular, from MI, what should the "from_tty" argument be?
AFAIK, that should always be zero. "from_tty" really means "from_CLI".
> And the "args"?
MI args look different from the command-line arguments to me, so how can one
convert between the two?
That would depend entirely on the command.
But please let's start with my first question --
why are we having to do this? What is it that
changed, making these changes necessary?
I also have a gdb 7.0-compatible version of the current patch brewing, the one
submitted in August was really against 6.8, and is thus really a bit out of
date. We are currently shipping this patch with Simics, along with a patched
gdb that supports reverse over MI for the benefit of our Eclipse integration.
Jakob Engblom, PhD, Technical Marketing Manager
Virtutech Direct: +46 8 690 07 47
Drottningholmsvägen 22 Mobile: +46 709 242 646
11243 Stockholm Web: www.virtutech.com