This is the mail archive of the
gcc-help@gcc.gnu.org
mailing list for the GCC project.
Re: g77 size of variable too large
This one doesn't compile either, at least for me.
So, as an excuse to have one more cup of coffee before leaving
for work :), I distilled it down to a small example:
C parameter(it=164,jt=52,kt=100,nphit=8,nphito=8)
parameter(it=478,jt=303,kt=215,nphit=8,nphito=8)
common
1/vectr/phi(it,jt,kt,nphit),phio(it,jt,kt,nphito),prt(nphit),
2 resor(nphit),reref(nphit),urf(nphit),ustarv(it,jt,kt),
3 phic(it,jt,kt,6),cs(it,jt,kt),tracel(it,jt,kt),
4 csk(it,jt,kt),
6 conve(it,jt,kt),convn(it,jt,kt),convh(it,jt,kt),
7 smp(it,jt,kt),dtfals(nphit)
integer u,v,w,p,pp,m,te,ed,t,f1,f2,f3
logical solve,save,restrt,cycli,cyclk,steady,ortogo,echo,les
logical dynk,multig,expl,dyn
character plane*2,head*24,scheme*2,schtur*2,name(nphit)*2,linsol*2
end
$ g77 big.f
big.f:4: error: size of variable `vectr_' is too large
$
Fabian, is this the error you have been getting ?
regards,
bud davis
On Fri, Jan 09, 2004 at 12:46:34PM +0100, Fabian Braennstroem wrote:
> Hello,
>
> me again. I just saw on in the browser that you can't see the attached
> files. So here they are inline:
>
> *************************************************************************
> *** main.f *************************************************************
>
> program main
> include 'CASECOM'
> include 'COMMON'
> character*24 hedvis,hedden,hedx,hedy,hedz
> real*4 dtime, etime
> real*4 darray(2),dd,tarr(3)
> parameter (ittelaps=10)
> real*4 ttstepi,ttstep,ttstept,ttsteps,ttelaps(ittelaps)
> dimension ustart(jt)
> character*8 ttnames(ittelaps)
> data ttnames/' calcu ',' calcv ',' calcw ',' conv '
> &,' calcpe ',' calcte ',' vist ',' calced',' calcph'
> &,' SUM '/
>
> c------define integer pointers used in the vectors phi, urf, nsweep etc.
> u=1
> v=2
> w=3
> p=4
> pp=5
> te=6
> ed=7
> t=8
> f1=9
> f2=10
>
> end
>
>
>
> *************************************************************************
> *** COMMON *************************************************************
>
>
> c parameter(it=164,jt=52,kt=100,nphit=8,nphito=8)
> parameter(it=478,jt=303,kt=215,nphit=8,nphito=8)
> c it : i-dimension of arrays (must be >= ni)
> c jt : j-dimension of arrays (must be >= nj)
> c kt : k-dimension of arrays (must be >= nk)
> c nphit : dimension of last index in phi-vector (must be >= number of
> variables)
> c nphito: dimension of last index in phio-vector (if steady=.false., it
> must
> c be >= number of variables; if steady=.true. it can be set to 1
> in
> c order to save memory)
> c
> common
> 1/pres/ipref,jpref,kpref
> 1/prin1/plane
> 1/prin2/indmon,imon,jmon,kmon
> 1/vecti/nsweep(nphit)
> 1/vectl/solve(nphit)
> 1/vectc/head(nphit)
> 1/vectr/phi(it,jt,kt,nphit),phio(it,jt,kt,nphito),prt(nphit),
> 2 resor(nphit),reref(nphit),urf(nphit),ustarv(it,jt,kt),
> 3 phic(it,jt,kt,6),cs(it,jt,kt),tracel(it,jt,kt),
> 4 csk(it,jt,kt),
> 6 conve(it,jt,kt),convn(it,jt,kt),convh(it,jt,kt),
> 7 smp(it,jt,kt),dtfals(nphit)
> 1/alli/ni,nj,nk,nim1,njm1,nkm1,iter,maxit,nphmax,itstep,ntstep,
> 2 iturb,itime,jst1
> 1/allr/great,small,sormax,betap,time,dt(100000),tscoef,acrank
> 1/allc/scheme,schtur,name,linsol(nphit)
> 1/alll/save,restrt,cycli,cyclk,ortogo,steady,echo,les,expl,dyn,
> 2 dynk,multig
> common
> 1/geom/xc(it),yc(jt),zc(kt),vol(it,jt,kt),
> 2 areaex(jt,kt),areany(it,kt),areahz(it,jt),
> 5 fx(it),fy(jt),fz(kt)
> 1/flupr/urfvis,viscos,vismin,densit,vis(it,jt,kt)
> 1/turb/cmucd,cmu,cd,c1,c2,cappa,elog,pfun,prandl
> 1/coef/ap(it,jt,kt),an(it,jt,kt),as(it,jt,kt),ae(it,jt,kt),
> 2 aw(it,jt,kt),ah(it,jt,kt),al(it,jt,kt),sp(it,jt,kt),
> 3 su(it,jt,kt),suv(it,jt,kt),suw(it,jt,kt),spvw(it,jt,kt)
> 1/point/u,v,w,p,pp,m,te,ed,t,f1,f2,f3
> integer u,v,w,p,pp,m,te,ed,t,f1,f2,f3
> logical solve,save,restrt,cycli,cyclk,steady,ortogo,echo,les
> logical dynk,multig,expl,dyn
> character plane*2,head*24,scheme*2,schtur*2,name(nphit)*2,linsol*2
>
> Greetings!
> Fabian
>
> > > There's something missing in this example, as this doesn't compile.
> > > Please send us a complete, compilable example.
> > >
> >
> > Yes, the example is attached.
> >
> > Thanks!
> > Fabian
>