[Cryptography] tail recursion in C [was Re: "NSA-linked Cisco exploit poses bigger threat than previously thought"]

Tom Mitchell mitch at niftyegg.com
Sun Aug 28 15:40:29 EDT 2016


On Sun, Aug 28, 2016 at 12:16 AM, Ron Garret <ron at flownet.com> wrote:
>
> On Aug 27, 2016, at 7:16 AM, Howard Chu <hyc at symas.com> wrote:
>
> > Ron Garret wrote:
> >> All of these are reasons why I dispute the claim that C is “a good
> portable assembler.”
>


> > We live in a world of finite resources. We are all always operating
> under very tight resource constraints. Wasting cycles or bytes is not doing
> anyone any favors.
>
> You can’t be serious.  Remember, we’re not talking about the efficiency of
> the compiled code here, we’re talking about the efficiency of the
> *compiler*.  Arguing that we should use unsafe languages in order to
> conserve electricity is just ridiculous.
>

I once thought that modern processors would remove a lot of the frugal
habits of many old timers (Clearly emacs is not frugal.).  The resources of
a Raspberry Pi are luxurious...

However the IOT is emerging on some rather sparse resources
and those tool chains and programmers are wrestling with the old
fast vs space battles.  Stacks so small that recursion of any kind
is a big risk and should be recast as looping structures.  And no operating
system...

For this and other reasons C is not going away quickly and the freedom
and ambiguous bits of C make optimization near impossible.
Instruction sets are growing in complexity so quality hand coding
has become very uncommon and a "portable" assembler is still
needed.

Compiler improvements and languages continue but seem to
demand a R-pi or bigger and do not target the tiny IOT devices.

Machine learning is making valuable discoveries. Next year
it could be improving compilers and compiler output.  ;-)

Machine learning could erase some cryptographic assumptions
faster than quantum computing.


-- 
  T o m    M i t c h e l l
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20160828/cdd35ebe/attachment.html>


More information about the cryptography mailing list