Re: [Tinycc-devel] [PATCH] arm: Handle __builtin_frame_address(1) correctly

[Kirill Smelkov]

On Wed, Dec 05, 2012 at 11:26:32AM +0100, Thomas Preud'homme wrote:
> Le vendredi 30 novembre 2012 00:43:07, Daniel Glöckner a écrit :
> > 
> > You should not look at a leaf function to derive the GCC stack frame.
> > It is probably different from the generic stack frame because GCC
> > knows this function will never be part of a stack trace done by
> > another function of the final program.
> > 
> > Take a look at arch/arm/kernel/stacktrace.c inside the Linux kernel.
> > They assume that every function pushes {fp, sp, lr, pc} and that
> > fp points to the address where pc is stored. I don't know why
> > GCC pushes pc. I can only imagine this being done to keep the
> > stack aligned to 8 bytes. Or maybe it is for exception handling
> > or association of stack frame debug info..
> 
> > I have once seen a page in MSDN describing the ARM stack frame of
> > Windows CE. I don't know if ARM specified how stack frames should look
> > like.
> 
> Hi Kirill,
> 
> did you make any progress on the issue since Daniel's comments? Could you let 
> me know when you push a patch so that I can test it and bump master to equal 
> the mob branch?

Hi Thomas,

Unfortunately not that much. I was trying to investigate
bounds-checking failures for `tcc -b -run tcc ...` tests first and it looks
I'm close to understand what is going on (old bcheck code is not adapted
to /proc/sys/kernel/randomize_va_space = 2). The plan was to go to arm
issues then, but I'm going very slow, because for tcc I have only 15-20
minutes per day and it is all in underground. Sorry. Could this issue
please wait me for some time? (probably week or so...)

As to testing the patch, you don't need me to push it to try - just pass
the mail with the patch to `git am --scissors` and apply it locally and
play with it.


> Grischka, what do you think of releasing a rc1 after that? We got plenty of 
> features merged since last release and 3 years start to be quite a long time. 
> I know there is still many things to improve, including some I care about, 
> but 
> it might attract new contributions or new users to have a more recent release.

Grischka, if my currently-guilty-for-regressions voice counts, I'd like
to once again ask for release too.


Thanks,
Kirill