Re: buffer overflow with long file paths in psfgen in NAMD 2.11b2?

From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Tue Dec 08 2015 - 16:41:22 CST

Fixed by extending the errmsg buffer. Thanks.

Jim

On Tue, 8 Dec 2015, Brian Radak wrote:

> I'm hitting a weird error when running some old psfgen scripts with the
> updated psfgen shipped with NAMD 2.11b2.
>
> psfgen) reading structure from psf file aced.psf
> *** buffer overflow detected ***: psfgen terminated
> ======= Backtrace: =========
> <plus typical pointer garbage>
>
> It was hard to make a minimal working example but it appears to happen in
> readpsf when the PSF file was made with a topology file with a long path and
> this same file path (or I guess one of similar length?) is loaded again.
>
> I've attached something like a test case, although the file path for the
> "long" topology file might need to be fiddled with to make this work.
> Alternatively, maybe this has to do with the fact that I keep my topology
> files on a separate disk that is symlinked into my home directory? In any
> event, I've never seen this problem before today.
>
> Brian
>
> P.S. To run the test case, first make some kind of long path for
> "top_all36_prot.rtf" and modify build.pgn and alchpatch.pgn to match. Then
> run the scripts through psfgen in that order:
>
> psfgen < build.pgn && psfgen < alchpatch.pgn
>
> --
> Brian Radak
> Theta Early Science Program Postdoctoral Appointee
> Leadership Computing Facility
> Argonne National Laboratory
>
> 9700 South Cass Avenue
> Building 240, 1.D.16
> Lemont, IL 60439-4871
> Tel: 630/252-8643
> email: bradak_at_anl.gov
>

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:22:18 CST