VC8-compatible debugging in the works!

Vikas Kumar walburn at gmail.com
Tue Mar 21 06:17:12 PST 2006


this is great work Peter and Michael. Is this debugging support only
for VC8 ? Does it already exist for Linux ? Do you really create
DWARF2 symbols when I assemble a code using yasm  ? Maybe i should do
a dwarfdump and check...

-Vikas

On 3/21/06, Michael Urman <mu at tortall.net> wrote:
> On Tue, Mar 21, 2006 at 12:56:05AM -0800, Peter Johnson wrote:
> > Obviously the basic mechanism would be something along the lines of a
> > [DEBUG] directive, but I'm looking at the parameters of said directive
> > (label to hook it into the code, plus some flags?), or even multiple
> > directives?  And of course there's the age-old question: would these
> > directives and parameters be debugging-format specific (advantage to
> > getting fuller access to all features of a particular debug format), or a
> > common set (advantage to porting same code to different debug formats).
> > Maybe a combination of both?  Discussion welcome.
>
> To answer this I think we need to figure out what information can be
> represented in all formats, what can be represented only in some
> formats, and what is actually useful for an assembly programmer.  Once
> we do that the decision should be relatively obvious.
>
> Things that seem likely to be useful to specify are:
>  * variable types - is this dd an integer or a pointer; if a pointer, what
>    does it point to?
>  * parameters to and local variables of a function
>  * calling convention of a function
>  * whether this code be stepped into - possibly most useful for macros if
>    that's not already implied by other means like the linemgr.
>  * ...
>
> The use cases for all of the above are to try to make a C debugger more
> capable of showing the sorts of information they can show for a C
> program.  Being able to show variables is paramount.  Being able to show
> callstacks would be great to have.  Being able to avoid useless stepping
> when not explicitly requested by a breakpoint would be nice.
>
> What of these and other useful items are representable in the supported
> debug formats?  They should give us a decent cross section of what
> support we can expect, so it would be reasonable to discard information
> (with warning) in debug formats without such capabilities if we come
> across them later.
>
> Michael
> --
> Michael Urman  [- mu at tortall.net -]
> _______________________________________________
> yasm-devel mailing list
> yasm-devel at tortall.net
> http://cvs.tortall.net/mailman/listinfo/yasm-devel
>


--
Capitalization is the difference between "I had to help my uncle Jack
off a horse.." and "I had to help my uncle jack off a horse.."


More information about the yasm-devel mailing list