Re: [ng-spice] Pole-zero analysis (and others)


To ng-spice@ieee.ing.uniroma1.it
From "Arno W. Peters" <a.w.peters@ieee.org>
Date Sat, 10 Jun 2000 21:33:21 +0200
Delivered-To mailing list ng-spice@ieee.ing.uniroma1.it
In-Reply-To <393E92C1.5EFCE33E@gmx.de >; from ManfredMetzger@gmx.de on Wed, Jun 07, 2000 at 08:21:53PM +0200
Mailing-List contact ng-spice-help@ieee.ing.uniroma1.it; run by ezmlm
References <393BFC93.8D1054D9@gmx.de > <20000605220809.A1043@cgmd80028.chello.nl > <393E92C1.5EFCE33E@gmx.de >
Reply-To ng-spice@ieee.ing.uniroma1.it
User-Agent Mutt/1.2i

On Wed, Jun 07, 2000 at 08:21:53PM +0200, Manfred Metzger wrote:

> Since moving AN_* only solves one part of the problem, this is not a
> proof that moving AN_* into dev.c corrects anything but you can take
> a debugger, set a breakpoint to VSRCpzSetup() and see that
> VSRCpzSetup() is not called before the crash. The program crashes in
> VSRCpzLoad() while accessing *(here->VSRCibrIbrptr). This pointer is
> only initialized in VSRCpzSetup().  If you now look into vsrcitf.h,
> which is included in dev.c, you see, that VSRCpzSetup() is not
> included in VSRCinfo, if AN_pz is not defined.

In my local tree, I have moved the affected code into dev.c.  It
solves the segfault for the circuit you have provided.  The results I
get are similar to the ones you report.

I have tried to reduce the complexity of the test case.  The program
now segfaults on the following circuit:

Multistage filter
v1 1 0 0 ac 1.0
r1 1 2 1k
c1 2 0 10p
.pz 1 0 2 0 vol pz
.print pz all
.end

Stack backtrace is as follows:

#0  0x808dd77 in alter ()
#1  0x808c4ef in CKTpzRunTrial ()
#2  0x808b196 in CKTpzFindZeros ()
#3  0x8089ad3 in PZan (ckt=0x80ff408, reset=1) at pzan.c:59
#4  0x807e6b5 in CKTdoJob (inCkt=0x80ff408, reset=1, inTask=0x80ff770)
    at cktdojob.c:130
#5  0x806e5f0 in if_run (t=0x80ff408 "", what=0x80b6889 "run", 
args=0x8100200, 
    tab=0x80ff390) at spiceif.c:188
#6  0x806badb in dosim (what=0x80b6889 "run", wl=0x0) at runcoms.c:237
#7  0x806bbc7 in ft_dorun (file=0x0) at runcoms.c:275
#8  0x8053843 in main (argc=1, argv=0xbffff974) at ./main.c:619

I have not yet been able to determine the cause of this segfault.

Regards,
-- 
Arno

PGP signature


Partial thread listing: