Re: core dump in column(1) with 33 columns

"Jeremy C. Reed" writes:
> On Sat, 23 Feb 2008, Greg Oster wrote:
> 
> > "Jeremy C. Reed" writes:
> > > tx:work$ jot 33 1 > J33           
> > > tx:work$ paste -s J33 | column -t 
> > > Memory fault (core dumped)
> > > 
> > > 
> > > Loaded symbols for /usr/libexec/ld.elf_so
> > > #0  0x0000000000401701 in maketbl () at 
> > > /usr/src/usr.bin/column/column.c:242
> > > 242                             t->len[coloff] = strlen(cols[coloff]);
> > 
> > What arch and OS version is this?  I can't replicate this on 4.99.54 
> > i386 or amd64, or 4.99.37 i386.  (I do see weird behaviour on 3.0 
> > i386.. it just prints lots and lots of spaces or something..)
> > 
> > Later...
> 
> Sorry ...
> 
> tx:rc.d$ uname -a 
> NetBSD tx.reedmedia.net 4.0_RC4 NetBSD 4.0_RC4 (GENERIC) #0: Thu Nov  8 
> 00:11:53 PST 2007  
> builds@wb28:/home/builds/ab/netbsd-4-0-RC4/amd64/200711080452Z-obj/home/build
> s/ab/netbsd-4-0-RC4/src/sys/arch/amd64/compile/GENERIC 
> amd64
> 
> I did update src/usr.bin/column (to HEAD) and built with -g.
> 
> Maybe try more columns?

Correct you are...   'jot 49 1 > foo ; paste -s foo | column -t' 
works for me.  'jot 50 1 > foo ; paste -s foo | column -t' segfaults.
Oddly,  'jot 50 1 | paste -s - | column -t' doesn't work either, but 
only prints a blank line instead of segfaulting?!!?  Yet there for:

 jot 50 1 > foo ; paste -s foo > t1
 jot 50 1 | paste -s - > t2

is no difference between t1 and t2!??!

Bizzare.

This is all on i386 with 4.99.54 -current from Feb 13.

Later...

Greg Oster