Hello Rene, and Chris,
Thank you for reporting this problem. Problem
occurs at **c=3. I'll fix this problem in next
release.
Masaharu Goto
>Date: Mon, 25 Feb 2002 07:48:50 +0000
>From: Rene Brun <Rene.Brun@cern.ch>
>To: cint@pcroot.cern.ch
>Subject: [CINT] [Fwd: [ROOT] cint bug: a[i][j]*a[i][j]]
>
>Hi Masa,
>
>Another mail that was posted to roottalk 2 weeks ago
>
>Rene
>-----------------------------------------------------------------------------
-
--
>Date: Fri, 15 Feb 2002 15:25:28 -0500
>From: Chris Crawford <chris2@lns.mit.edu>
>To: roottalk@pcroot.cern.ch
>Cc: chris2@lns.mit.edu
>Subject: [ROOT] cint bug: a[i][j]*a[i][j]
>
>i would like to report a couple of eccentric bugs in cint 5.15.20
>illustrated by the following code "err.c":
>
>float sub(float** a) {return a[0][0]+a[0][0];}
>void err() { float **b = new float*; *b= new float(2.);
> for(int i=0;i==0;i++) { printf("%f\n",sub(b)); }}
>void er2() { float **c = new float*; *c= new float; **c=3.;
> for(int i=0;i==0;i++) { printf("%f\n",sub(c)); }}
>
>err() reports the error "Error: +" but still seems to work. it does so
>with either 'new' or 'malloc', only inside a 'for' or 'while' loop,
>which calls 'a[0][0]+a[0][0]' from a function. using '**a + **a' won't
>report an error.
>
>er2() crashes at **c=3, only when followed by the next line.
>
>any thoughts???
>--Chris Crawford
>
>
This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:50:43 MET