Re: ARM/Thumb long calls stubs

[OMakarenko]

> > LD generates only one stub for ARM_func. That only stub may be not
> > accessible from some parts of thumb code. 
> > 
> > So how can this be done without linker support?
> 
> To workaround the problem you could provide ARM-to-ARM or
> THUMB-to-THUMB long call entry points.  eg:
> 
        yes, thank you for suggestion. That is exactly the long calls stub I
want the linker to generate automatically as it has all needed information.

> > It would be nice for the linker to add long-calls stubs instead. Is
> > that possible?
> 
> Anything is possible, so yes it could be done.
> 
        :) I mean is it difficult to do? Are there any processors that ld
can already generate long call (or branch) stubs for? Probably I could add
that support for long calls on ARM if only I had an example. I am porting a
large piece of code from ARM native compiler/linker to gcc-tools and the
only show stopper is the long calls support. As I told you gcc misbehaves
when I use long calls attributes/pragma/option. I am not sure why yet and
going to try CVS version. 

        ARM native tools linker solves all this problems. It generates
additional long calls stubs when needed and it is smart enough to generate
additional interwork stubs when it sees that one stub will not do.

        LD beats it in general but this only point.

> > It doesn't. It generates wrong offsets instead and hence broken
> > code.
> 
> Please could you submit a simple test case then ?
> 
        I will try to strip down my code and will send the example asap.

        Best,
        =oleg