obache@netbsd.org writes:
> Synopsis: shells/tcsh segfaults on history substitutions
> 
> Responsible-Changed-From-To: pkg-manager->agc
> Responsible-Changed-By: obache@netbsd.org
> Responsible-Changed-When: Thu, 28 Jun 2007 12:52:03 +0000
> Responsible-Changed-Why:
> Assign static-tcsh maintainer.
> Works fine on my NetBSD 3.1 box. Current or GCC4 specific problem?
I'm fairly certain it is a tcsh problem.  I rebuilt it with -g,
and it seems to crash in a call to wcslen, with a NULL parameter.
(gdb) target core tcsh.core
Core was generated by `tcsh'.
Program terminated with signal 11, Segmentation fault.
#0  0x080a0182 in wcslen ()
(gdb) where
#0  0x080a0182 in wcslen ()
#1  0x0808f1de in Strbuf_append (buf=0x8102534, s=0x0) at tc.str.c:599
#2  0x0805ff7d in getsub (en=0x8141968) at sh.lex.c:854
#3  0x0805fa1f in getexcl (sc=0) at sh.lex.c:731
#4  0x0805ed8e in getC1 (flag=3) at sh.lex.c:469
#5  0x0805e626 in word (parsehtime=0) at sh.lex.c:294
#6  0x0805e33e in lex (hp=0x811a218) at sh.lex.c:185
#7  0x0804b1d7 in process (catch=1) at sh.c:1931
#8  0x0804a126 in main (argc=0, argv=0xbfbfe898) at sh.c:1304

Also, I compiled version 6.14.00 on the exact same system, and
that version does not segfault on !$:h history substitutions.
I'd say it's a tcsh 6.15.00 problem.
					-jarle