{par script path}: {chld script path}: /bin/bash: bad interpreter: Permission denied

[nicebounce at trodman dot com]

<As always thanks for continually improving  cygwin; it's a wonderful set of tools.>

Here's an error message example:

  /adm/bin/ccm_build_scripts/run_large_nb.sh: s:/adm/bin/ccm_build_scripts/run_nb.sh: /bin/bash: bad interpreter: Permission denied

  "run_nb.sh" aborts immediately, not executing a single line as far as I know.

  Misc info:
    $ pwd
    /adm/bin/ccm_build_scripts

    $ head -1 run_large_nb.sh run_nb.sh
    ==> run_large_nb.sh <==
    #!/bin/bash -

    ==> run_nb.sh <==
    #!/bin/bash -

    $ ls -l run_large_nb.sh run_nb.sh
    -rwxr-xr-x    1 adm_tsr  SCM_ES_S    54021 Mar  3 11:15 run_large_nb.sh
    -rwxr-xr-x    1 adm_tsr  SCM_ES_S    13322 Sep  3  2002 run_nb.sh
    $ uname -a;echo $CYGWIN
    CYGWIN_NT-5.0 c7mkes123 1.3.20(0.73/3/2) 2003-02-08 12:10 i686 unknown unknown Cygwin
    binmode ntsec tty

The scripts involved have long been stable. The problem 
occurs sporatically. It happened early this week 3 times in 
one morning, then I rebooted the box and ran an intensive software 
build which exercised both scripts over a 22 hour period 
with no problems. 

Problem has been seen on both cygwin 17-1 and 20-1 (we skipped 18-1),
I'm not convinced it's a cygwin problem.  We're running Windows 2000
Server on all boxes involved; I believe our service packs are up to date.

The scripts reside on a network drive "s:" controlled by an HP VA7400 SAN;
perhaps the problem is SAN related.  The SAN is seen by our remote
w2k file server as a local SCSI block device.  The SAN hardware
is shared by many UNIX or Windows servers- each allocated
separate LUNs.  See cygwin mount table at end of e-mail.

I can not force the error msg to occur at will even when I make
the parent and/or child script nonexecutable.  Does anyone know
how to force the same error msg that we're getting?

I've searched my local cygwin mailing list archives for this error,
with no luck.

Any insights would be appreciated.  Since we run so long between
occurances of this problem, I'm not too concerned.  I wanted to
at least document this somewhat for others that might see it.

--
thanks/regards,
Tom Rodman

perl -e 'print unpack("u", "\.\=\$\!T\<F\]D\;6\%N\+F\-O\;0H\`");'

--v-v------------------C-U-T---H-E-R-E-------------------------v-v--
$ mount # drive S: is a network drive managed by a w2k box, the disk is on a SAN
c:\aut\cyg\bin on /usr/bin type system (binmode)
c:\aut\cyg\lib on /usr/lib type system (binmode)
s:\adm\bin\jcm on /sjcm type system (binmode)
c:\aut\cyg on / type system (binmode)
s:\public on /public type system (binmode,noexec)
s:\prep on /prep type system (binmode,noexec)
s:\user on /user type system (binmode,noexec)
c:\aut on /aut type system (binmode)
s:\adm on /adm type system (binmode)
a: on /drv/a type system (binmode)
c: on /drv/c type system (binmode)
s: on /drv/s type system (binmode,noexec)
y: on /drv/y type system (binmode)


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/