Re: egcs-1.1.2 ping bug also causes miscompilation of pcbit isdn drive

Horst von Brand (vonbrand@inf.utfsm.cl)
Thu, 17 Jun 1999 15:59:48 -0400


<hagopiar@vuser.vu.union.edu> said:
> On Thu, 17 Jun 1999, Horst von Brand wrote:
> > It just so happens that the kernel is _the_ code that most streches the
> > compiler, AFAIK. Breakage due to ill-understood features, usage of outright
> > compiler bugs and reliance on optimizations done a certain way (or not
> > done) does happen.

> I believe that the compiler itself is a larger code base, but it likely
> doesn't have as much hack work in it. :-)

Hard to say. First gcc I used was something like 15 years ago...

Sure, gcc et al were sleeping for years (so that gcc-2.7.2 became "de
facto" gcc standard), but with the egcs project the speed has picked up
quite a bit, usually there are 300Kb or so patches weekly (gzipped). Not
recently, as the effort is going into a rock-stable gcc-2.95 release right
now.

Besides, compilers aren't "harder" or "easier" than kernels, just _very_
different problems. Have you ever tried to write a program that writes a
program? It is surprisingly hard to do, more so if the target language is
low-level. Now add in the requirement that the result will be munged by
several optimization passes (which you'll have to consider, and maintain
too!) and then translated into a incredible variety of target languages,
sometimes very broken ones, and comply with strange ABIs. Oh, BTW the
compiler has to be able to be compiled correctly with the wildest variety
of broken compilers, has to work with weird files in /usr/include, and all
that on quite different operating systems (DOS, Win32, VMS, Unixy, et al).

-- 
Dr. Horst H. von Brand                       mailto:vonbrand@inf.utfsm.cl
Departamento de Informatica                     Fono: +56 32 654431
Universidad Tecnica Federico Santa Maria              +56 32 654239
Casilla 110-V, Valparaiso, Chile                Fax:  +56 32 797513

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/