🚀 go-pugleaf

In article <4i6fo9$p28@solutions.solon.com>
schwarz@mips.complang.tuwien.ac.at (Konrad Schwarz) writes:

|> James Kanze US/ESC 60/3/141 #40763 writes:
|> |> Konrad Schwarz writes:
|> |> 
|> |> |> J. Kanze writes:
|> |> |> |>     [Concerning alloca...]
|> |> |> I: As far as I know, implementing alloca via the stack is trivial on
|> |> |> I: any general-purpose CPU, but I am interested in hearing of counter examples.
|> |> |> |> 
|> |> |> |> So what do you do on machines that don't have a stack?
|> |> 
|> |> |> General-purpose CPU's have stacks, at least in my book.
|> |> 
|> |> IBM mainframes don't have stacks.  Would you not consider them general
|> |> purpose CPU's?  (While most modern architectures have stacks, I
|> |> suspect that if you looked at all of the general-purpose CPU's in
|> |> existance, you'd find more architectures without a stack than with.)

|> You're telling me that a C implementation on an IBM mainframe doesn't
|> use a register as a stack pointer or doesn't reserve a continuous
|> region of memory as a stack?

I don't know about how IBM does their implementations, but...

|> Note that you'd be hard pressed to find a RISC machine with a dedicated
|> stack pointer register.

|> |> |> How do you
|> |> |> implement C on machines that don't have a stack?
|> |> 
|> |> The same way you implement any other language, I suppose.  On the one
|> |> machine I have any experience with, local frames were allocated and

|> Is this the afore mentioned IBM mainframe?  Was the language C---or
|> SIMULA :)?

The afore mentioned is an IBM mainframe compatible, and the language was
C.  To be frank, it was a disasterous implementation of C, but that is
beside the question.

|> |> freed using a mechanism similar to malloc/free.  Once allocated, the
|> |> size of the local frame could not be changed.

|> See my previous postings for how to implement alloca in this case.

I have heard that such implementations exist.  I am simply pointing out
that the reason alloca is not in the standard is because such crazy
implementations asthe one I just described exist.  (Even with a more
reasonable implementation of C on a 360's architecture, I can imagine
problems with alloca's of very large arrays, and perverse use of the
alloca function.)
-- 
James Kanze           (+33) 88 14 49 00          email: kanze@gabi-soft.fr
GABI Software, Sarl., 8 rue des Francs Bourgeois, 67000 Strasbourg, France
Conseils, études et réalisations en logiciel orienté objet --
              -- A la recherche d'une activité dans une region francophone