[illumos-Discuss] Little question about compilers and a future...

[Joerg Schilling]

"Garrett D'Amore" <garrett at damore.org> wrote:

> On Fri, 2010-09-17 at 19:08 +0200, Joerg Schilling wrote:

> > /usr/lib/cpp is more important than I thought before checking it's relevance.
> > It is used by rpcgen(1), dtrace(1), aw(1) and as(1) and it is called 1421 times 
> > when compiling Illumos. Today a full "nightly" succeeded with the OSS 
> > replacement I wrote.
>
>
> I'm really happy to hear this... I'm glad that you were able to run a
> full nightly -- this is in fact why I have such requirements -- so we
> don't get surprises. :-)

I did never make a public announcement for something being ready, before doing 
full tests. Last monday, I just mentioned that I reached a state where my cpp 
works for all rpcgen calls. Life is much easier if you try to avoid 
missunderstandings....


> Can you post a webrev somewhere?

I am currently using a only slightly modified Makefile from AT&T from 1978 and 
a webrev for a new program does not help anyway.

I am currently getting 156 warnings like:

../../intel/ia32/sys/trap.h: 74: T_AST redefined
../../i86pc/genassym/obj32/assym.h: 409: T_AST redefined
../../intel/ia32/sys/trap.h: 74: T_AST redefined
../../i86pc/genassym/obj32/assym.h: 409: T_AST redefined
../../i86pc/genassym/obj64/assym.h: 428: T_AST redefined
../../i86pc/genassym/obj64/assym.h: 462: FPU_EN redefined
../../i86pc/genassym/obj64/assym.h: 463: FPU_VALID redefined
../../i86pc/genassym/obj32/assym.h: 409: T_AST redefined
/home/schily/Solaris/schily147x/proto/root_i386/usr/include/ia32/sys/trap.h: 74: T_AST redefined
/home/schily/Solaris/schily147x/proto/root_i386/usr/include/ia32/sys/trap.h: 74: T_AST redefined
../../i86pc/genassym/obj64/assym.h: 428: T_AST redefined
../../i86pc/genassym/obj32/assym.h: 409: T_AST redefined
../../i86pc/genassym/obj64/assym.h: 428: T_AST redefined
../../i86pc/genassym/obj32/assym.h: 409: T_AST redefined
../../i86pc/genassym/obj64/assym.h: 428: T_AST redefined
../../i86pc/genassym/obj32/assym.h: 409: T_AST redefined

This is not a problem but the expected behavior of a cpp from 1978.
At that time, redefinitions with different whitespace in the #define line
have been flagged this way.

For me, to be able to have a self hosting SchilliX based on redistributable 
code, this is no problem. For publishing the source, I would like to find a way
to ignore different whitespace, but the parser is written like people did in 
the 1970s and so this is not as trivial as the other 10 changes I had to 
implement in order to be close enough to the Sun cpp.

P.S.: In general, I tend to finish tasks in chronological order and I am still 
waiting on a one month old webrev for star integration.

Jörg

-- 
 EMail:joerg at schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
       js at cs.tu-berlin.de                (uni)  
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily