From: Grant Lowe <>
Date: Mon Apr 01 2002 - 01:22:48 EST
Hi again gurus.
I have received about a dozen or so responses so far.  Some have said that
it doesn't run from the installed package.  Others said that it does work,
but you have to do a little tweaking of the default installs (for example
ssh now installs in /usr/local/ssh/bin, and before it installed in
/usr/local/bin, so I put in some symbolic links to make sure).  Others asked
if I had the correct packages in the first place (zlib, openssl, etc.). I
made sure I have all the packages.  A couple of people suggested I read the
instructions that are hyperlinked on the openssh-3.1p1 package area of
sunfreeware (one is a Microsnot Word document, the other is a hypertext
document, but are basically the same thing).  I followed those instructions
to the letter.  Somebody said that on the openssh mailing list the reason
for the segmentation fault is the versions of openssl and openssh are
incompatible.  Lastly, somebody suggested that I check my LD_LIBRARY_PATH
variable, especially to include /usr/local at the beginning of the path.
One thing I do remember from a couple of years ago when I built this from
scratch is I needed the libwrap.a and tcpd.h files from tcp wrappers.  Well,
the tcpd.h is still with tcp wrappers.  I couldn't find the libwrap.a
however.  Does this ring a bell with anyone?
It still doesn't work, though, so if anybody else has some more ideas, or
whatever, I'll add them to the final summary.  Thanks.
Original post:
Hi gurus.

I have a brand new, Solaris for Intel box.  It is Solaris 8, fully-patched,
with 128 MB RAM.  It is networked, and working just fine, except for SSH.
The problem I'm having is with openssh3.1p1.  I have installed gcc-3.0.3,
openssl-0.9.6c, and zlib-1.1.4.  These packages, as well as openssh-3.1p1,
came from  When I try to run ssh, it give me a
coredump.  Here's the truss of it:

# truss ssh
execve("/usr/local/ssh/bin/ssh", 0x080476DC,
0x080476E4)  argc = 1
open("/usr/lib/", O_RDONLY)              = 3
fxstat(2, 3, 0x08047574)                        = 0
mmap(0x00000000, 215720, PROT_READ|PROT_EXEC,
MAP_SHARED, 3, 0) = 0xDFB20000
mmap(0x00000000, 204800, PROT_READ|PROT_EXEC,
MAP_SHARED, 3, 0) = 0xDFAE0000
mmap(0xDFAE0000, 126060, PROT_READ|PROT_EXEC,
munmap(0xDFAFF000, 65536)                       = 0
MAP_PRIVATE|MAP_FIXED, 3, 126976) = 0xDFB0F000
open("/dev/zero", O_RDWR)                       = 4
close(3)                                        = 0
    Incurred fault #6, FLTBOUNDS  %pc = 0xDFAE9DEB
      siginfo: SIGSEGV SEGV_MAPERR addr=0x00000008
    Received signal #11, SIGSEGV [default]
      siginfo: SIGSEGV SEGV_MAPERR addr=0x00000008
        *** process killed ***

Any idea of what's going on?  Thanks for any and all help!

sunmanagers mailing list
Received on Mon Apr 1 00:19:16 2002

This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:42:39 EST