rsync - May 2002 - unexplained error (code 255)

Everything worked, files transferred fine however a -vvv reveals that
the process did not end clean.  The error is not seen with -v or even
-vv. Any insight? rsync version 2.5.5 protocol version 26, running on
Solaris 8.

Write failed flushing stdout buffer.
write stdout: Broken pipe
rsync error: unexplained error (code 255) at main.c(925)
_exit_cleanup(code=255, file=main.c, line=925): about to call exit(255)


Full -vvv output

mach9$/opt/rsync/bin/rsync -e ssh -vvv
admin@mach3:/home1/admin/rsyncdir/* 052002/mach3 opening connection
using ssh mach3 -l admin rsync --server --sender -vvv .
"/home1/admin/rsyncdir/*" 
server_sender starting pid=23230
make_file(1,datefile)
expand file_list to 4000 bytes, did move
recv_file_name(datefile)
make_file(1,unamefile)
recv_file_name(unamefile)
received 2 names
recv_file_list done
get_local_name count=2 052002/mach3
generator starting pid=27427 count=2
recv_files(2) starting
recv_generator(datefile,0)
send_file_list done
sending sums for 0
send_files starting
recv_generator(unamefile,1)
sending sums for 1
generate_files phase=1
send_files(0,/home1/admin/rsyncdir/datefile)
send_files mapped /home1/admin/rsyncdir/datefile of size 29
recv_files(datefile)
recv mapped datefile of size 29
datefile
calling match_sums /home1/admin/rsyncdir/datefile
built hash table
hash search b=700 len=29
match at 0 last_match=0 j=0 len=29 n=0
done hash search
sending file_sum
got file_sum
renaming .datefile.ThaWK1 to datefile
false_alarms=0 tag_hits=1 matches=1
sender finished /home1/admin/rsyncdir/datefile
send_files(1,/home1/admin/rsyncdir/unamefile)
send_files mapped /home1/admin/rsyncdir/unamefile of size 78
recv_files(unamefile)
recv mapped unamefile of size 78
unamefile
calling match_sums /home1/admin/rsyncdir/unamefile
built hash table
hash search b=700 len=78
match at 0 last_match=0 j=0 len=78 n=0
done hash search
sending file_sum
got file_sum
renaming .unamefile.UhaWK1 to unamefile
false_alarms=0 tag_hits=1 matches=1
sender finished /home1/admin/rsyncdir/unamefile
recv_files phase=1
send_files phase=1
generate_files phase=2
send files finished
total: matches=2  tag_hits=2  false_alarms=0 data=0
recv_files finished
wrote 60 bytes  read 1242 bytes  868.00 bytes/sec
total size is 107  speedup is 0.80
Write failed flushing stdout buffer.
write stdout: Broken pipe
rsync error: unexplained error (code 255) at main.c(925)
_exit_cleanup(code=255, file=main.c, line=925): about to call exit(255)

Brian D. Hamm, CISSP, CCNA
Network Design & Implementation
(o) 727-939-3080
(c) 727-424-4384
(f) 240-266-7185
(e) bdhamm@tampabay.rr.com

I have been using rsync 2.5.5 on Solaris 8 for quite a while w/ no
problems.  This week we upgraded our boxes to Solaris9 and we start
seeing (when -vvv is specified):
   recv_files finished
   wrote 16 bytes  read 335214 bytes  31926.67 bytes/sec
   total size is 107361593  speedup is 320.26
   rsync error: unexplained error (code 255) at main.c(1001)
   _exit_cleanup(code=255, file=main.c, line=1001): about to call
exit(255)

So it appears to have copied all of the files, but then died
afterwards.  I have searched this list's archives and have found this
error pop up quite a few times but have never seen a solution for it.  I
have tried the CVS version (as of Dec-04-2002 @ 11:00EST) and am still
seeing the problem.

I am using the following:
Rsync->2.5.6cvs  protocol version 26
OS->Solaris9
SSH->Sun_SSH_1.0, protocol versions 1.5/2.0

I am invoking it with: rsync -avz --progress -e ssh host:/path/to/file
.

I am sure that the problem is Sun's implementation of SSH on Solaris9. 
Does anyone have a fix/workaround?

Thanks,
Alex