Re: pcc build failure last couple days?

[bch%methodlogic.net@localhost]

On Tue, Aug 09, 2011 at 07:47:40PM +0100, Iain Hibbert wrote:
> On Tue, 9 Aug 2011, bch%methodlogic.net@localhost wrote:
> 
> > --- dependall-pcc ---
> > cc1: warnings being treated as errors
> > /usr/src/external/bsd/pcc/libexec/ccom/../../dist/pcc/cc/ccom/pftn.c: In 
> > function 'imop':
> > /usr/src/external/bsd/pcc/libexec/ccom/../../dist/pcc/cc/ccom/pftn.c:3059:14:
> >  error: operation on 'p->n_type' may be undefined
> > /usr/src/external/bsd/pcc/libexec/ccom/../../dist/pcc/cc/ccom/pftn.c:3076:14:
> >  error: operation on 'p->n_type' may be undefined
> > /usr/src/external/bsd/pcc/libexec/ccom/../../dist/pcc/cc/ccom/pftn.c:3100:14:
> >  error: operation on 'p->n_type' may be undefined
> > /usr/src/external/bsd/pcc/libexec/ccom/../../dist/pcc/cc/ccom/pftn.c:3108:14:
> >  error: operation on 'p->n_type' may be undefined
> 
> I guess this is related to the gcc-4.5 switchover
> 
> I have not updated sources for a couple of months so still using gcc 4.1.3
> (and, with a newer version of pcc) but line 3059 certainly looks gross..
> 
> does the patch attached fix it?

it changes the error msg... I'll dig into it, but for the record:

cc1: warnings being treated as errors
/usr/src/external/bsd/pcc/libexec/ccom/../../dist/pcc/mip/common.c:544:12: 
error: variably modified 'elm' at file scope
*** [common.o] Error code 1
nbmake: stopped in /usr/src/external/bsd/pcc/libexec/ccom
1 error
nbmake: stopped in /usr/src/external/bsd/pcc/libexec/ccom
*** [dependall] Error code 2
nbmake: stopped in /usr/src/external/bsd/pcc/libexec/ccom
1 error


> iain
> Index: pftn.c
> ===================================================================
> RCS file: /cvsroot/src/external/bsd/pcc/dist/pcc/cc/ccom/pftn.c,v
> retrieving revision 1.5
> diff -u -p -r1.5 pftn.c
> --- pftn.c    3 Jun 2010 19:07:59 -0000       1.5
> +++ pftn.c    9 Aug 2011 18:41:17 -0000
> @@ -3056,7 +3056,7 @@ imop(int op, NODE *l, NODE *r)
>       case PLUS:
>               if (li && ri) {
>                       p = buildtree(PLUS, l, r);
> -                     p->n_type = p->n_type += (FIMAG-FLOAT);
> +                     p->n_type += (FIMAG-FLOAT);
>               } else {
>                       /* If one is imaginary and one is real, make complex */
>                       if (li)
> @@ -3073,7 +3073,7 @@ imop(int op, NODE *l, NODE *r)
>       case MINUS:
>               if (li && ri) {
>                       p = buildtree(MINUS, l, r);
> -                     p->n_type = p->n_type += (FIMAG-FLOAT);
> +                     p->n_type += (FIMAG-FLOAT);
>               } else if (li) {
>                       q = cxstore(mxtyp);
>                       p = buildtree(ASSIGN, structref(ccopy(q), DOT, real),
> @@ -3097,7 +3097,7 @@ imop(int op, NODE *l, NODE *r)
>               if (li && ri)
>                       p = buildtree(UMINUS, p, NIL);
>               if (li ^ ri)
> -                     p->n_type = p->n_type += (FIMAG-FLOAT);
> +                     p->n_type += (FIMAG-FLOAT);
>               break;
>  
>       case DIV:
> @@ -3105,7 +3105,7 @@ imop(int op, NODE *l, NODE *r)
>               if (ri && !li)
>                       p = buildtree(UMINUS, p, NIL);
>               if (li ^ ri)
> -                     p->n_type = p->n_type += (FIMAG-FLOAT);
> +                     p->n_type += (FIMAG-FLOAT);
>               break;
>       default:
>               cerror("imop");


-- 
Brad Harder
Method Logic Digital Consulting
http://methodlogic.net/
http://twitter.com/bcharder