Need help in figuring out the flow of RTEMS codes

[Richi Dubey]

Dear Dr. Bloom,

Thank you for your suggestions. Yes, I realized after your mail that
different architecture would have different assembly code, which I was
not aware of and assumed every .S file has the same rules.

I will set up an appropriate environment as you both suggested.

Thanks.





On Fri, May 29, 2020 at 8:34 PM Gedare Bloom <gedare at rtems.org> wrote:
>
> On Fri, May 29, 2020 at 1:59 AM Richi Dubey <richidubey at gmail.com> wrote:
> >
> > This makes sense. Thanks a lot!
> >
> >
> > On Fri, May 29, 2020 at 1:18 PM Sebastian Huber
> > <sebastian.huber at embedded-brains.de> wrote:
> > >
> > > Hello Richi,
> > >
> > > On 29/05/2020 09:36, Richi Dubey wrote:
> > > > Hii,
> > > >
> > > > I had a small doubt about all the source code/procedures that a code
> > > > goes through. Where can I read more about it? For instance, when I am
> > > > debugging smpschededf02.exe with gdb, It starts  with
> > > > bsps/sparc/shared/start/start.S, line 107(Why did it start with
> > > > start.s?) and the last line of this file is:
> > > >
> > > >   mov     %g0, %o0                ! command line
> > > >          call    SYM(boot_card)          ! does not return
> > > >           sub     %sp, 0x60, %sp         ! room for boot_card to save args
> > > >
> > > > After this it starts executing bsps/shared/start/bootcard.c: line 53 .
> > > > How did this happen? Is it because of call? If yes, why was sub
> > > > executed(in gdb) before executing code from bootcard.c?
> > > >
> > > > I hope I am not trying to going too deep into this. I just want to
> > > > have a overall picture of how a code gets executed.
> > >
> > > if you want to learn how code executes at processor level, then I would
> > > recommend to work with an ARM (Qemu) or RISC-V (SIS, Qemu) simulator.
> > > These are the currently most relevant architectures. SPARC is a dead
> > > architecture and it has some peculiarities which were a nice idea in the
> > > 1980s.
> > >
> > > > This is necessary
> > > > for writing your own code in RTEMS, right?
> > > It helps, but it is not necessary.
> +1
>
> This is quirky assembly stuff. Your project should not require writing
> assembly, but it is helpful if you can start to learn how to read it
> for debugging.
>
> If you know a different assembly language, you might try to set up an
> environment with that target and use it for low-level debugging
> purposes. Mostly though, gdb should take care of translating the
> executing lines to source linkage, so you would only face trouble when
> debugging assembly source like this. As Sebastian mentioned, if you
> don't know an assembly language, your time is better spent to learn a
> modern widely used one.
>
> > _______________________________________________
> > devel mailing list
> > devel at rtems.org
> > http://lists.rtems.org/mailman/listinfo/devel