Ram initialization and small c.

[Eric W. Biederman]

Steve Gehlbach <steve at nexpath.com> writes:

> Eric W. Biederman wrote:
> 
> > The requirements of these small tweaks when building are why I am convinced
> > that in the long run we need actual compiler support.  Either in gcc or
> > in one we write ourselves.  For now we have an easier way to generate the
> > assembly code.
> >
> 
> Actually I don't mind the inline tweaking approach and I think it is
> easier to write and debug than assy, but others will disagree.

The only part I don't like is that you have to keep the .S files
around...

> My concern is that a register-only compiler may be an intractable
> problem, and any way you go at it (custom or .md file for gcc) you
> will have to tweak and craft the C code just as much. 

It should have problems only when you run out of registers,
which I think is a rather different problem.

> If in fact this
> is true, then the benefit over inlining gcc seems minimal.  Wish there
> were a way to answer this question before you invest a month of time.

Oh.  I think I can have some answers in less than a month.  That is
just what I have allocated for investigation.  But it will take a week
or so to get something working well enough to compare with the inline
tweaking approach.

Plus write now I am enjoying the change of pace.

Eric