[GRASS5] Re: v.digit segfault

Maris Nartiss maris.gis at gmail.com
Thu Apr 13 09:59:50 EDT 2006


Hi,

adding DEBUG didnt helped a lot. After drawing all vect features it
still segfaults without reporting any info.

Other ideas?

......
D2/5: node = 1056 NodeSymb = 12
D2/5: display_node(): node = 1056 color = 12
D2/5: set color to symb 12: 255 0 0
D2/5: display_icon()
D2/5: node = 1057 NodeSymb = 12
D2/5: display_node(): node = 1057 color = 12
D2/5: set color to symb 12: 255 0 0
D2/5: display_icon()
D5/5: driver_close()
D5/5:  -> closed
Segmentation fault

Maris.

On 4/11/06, Hamish <hamish_nospam at yahoo.com> wrote:
> > I know v.digit is broken by design, but it was workin.
>
> I for one think it is a lot better than its reputation.
>
> > Unfortunatly on my Gentoo box v.digit segfaults for both Grass 6.0.2
> > and grass-6.1.cvs_src_snapshot_2006_04_08.
> >
> > Symptoms: after comand run it displays vectors but no toolbar etc. is
> > shown. After that vect maps needs to be rebuilt with v.build.
>
> (I haven't seen this, works fine for me in Debian/stable tcltk 8.4)
>
>
> > I'm not an expert in gdb, but here's output:
> > GRASS 6.1.cvs (Agnese):~ > gdb v.digit
> > (gdb) run map=ttt
> > Starting program: /usr/local/grass-6.1.cvs/bin/v.digit map=ttt
> > [Thread debugging using libthread_db enabled]
> > [New Thread -1219950928 (LWP 22555)]
> >
> > Program received signal SIGSEGV, Segmentation fault.
> > [Switching to Thread -1219950928 (LWP 22555)]
> > 0xb78b47c3 in strlen () from /lib/libc.so.6
> > (gdb) bt
> > #0  0xb78b47c3 in strlen () from /lib/libc.so.6
> > #1  0xb79a9635 in Tcl_ExternalToUtfDString () from /usr/lib/libtcl.so
> > #2  0xb7a687c0 in Tk_MainEx () from /usr/lib/libtk.so
> > #3  0x080522d6 in main (argc=32, argv=0xbfab47d4) at main.c:199
> >
> > v.digit main.c 199 line is
> >     /* Open toolbox */
> >     Tk_Main(0, argv, Tcl_AppInit);
>
> Unfortunetly gdb isn't much use for mixed C and Tcl programs v.digit and
> NVIZ. It debugs the C to the line where it says "start main Tcl
> application" ... Which just tells you the bug is somewhere after the Tcl
> part begins..
>
> for more clues try:
>  g.gisenv set="DEBUG=5"
>
> (0 is off, 5 is most verbose)
>
>
>
>
> Hamish
>




More information about the grass-dev mailing list