Re: Very odd 'bash' problem in RH7.1 - What!?!!

[Dennis McMurchy <denismcm@example.com>]

  This problem of executables that are there for 'ls', but can't be
found by 'bash' when I try to run them continues to puzzle me deeply.

  Tobias suggested, "Try ldd pppcosts".  Here's the result:

<denismcm@example.com:~>$ ldd ~/internet/pppcosts
/usr/bin/ldd: /home/denis/internet/pppcosts: No such file or directory

  Interesting, eh?  I should add here before I forget again, that the
only two files that have shown this behaviour (so far) are old
executables that I added myself to a fresh "Everything" installation of
RedHat 7.1 (I would have assumed that this had all the appropriate
older libraries that I might need, but it apparently doesn't include
the 'ld-linux.so.1' that 'pppcosts' probably needs).

  Now, I could try recompiling these things, but, surely, this new fact
that ldd, for example, reports that these files don't exist suggests that
the problem lies elsewhere, perhaps where Thomas O'Dowd suggests.  Also,
I just checked another old executable I've added in, and there seems to
be no problem running it, or getting e.g. ldd to see it:

	<denismcm@example.com:~>$ ldd fmtj
		libc.so.6 => /lib/i686/libc.so.6 (0x4002d000)
		/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)

On Mon, 28 May 2001, Thomas O'Dowd wrote:

> There is no rehash for bash. Do you have the same problem with
> a different shell, ie try it with tcsh...

  NO, appears to make no difference when I switch shells:

[denismcm@example.com ~]$ echo $SHELL
/bin/tcsh
[denismcm@example.com ~]$ ls ~/internet/pppcosts
/home/denis/internet/pppcosts
[denismcm@example.com ~]$ ~/internet/pppcosts
/home/denis/internet/pppcosts: Command not found.
denismcm@example.com ~]$ ls -l /usr/local/bin/xvscan
-rwxrwxr-x    1 bin      bin       1354434 May 28 01:46 /usr/local/bin/xvscan
[denismcm@example.com ~]$ /usr/local/bin/xvscan
/usr/local/bin/xvscan: Command not found.

On Mon, 28 May 2001, Thomas O'Dowd wrote:

> Could be something in
> the environment also, send us the output of the env command. I
> seem to remember having a wierd ksh problem something like this
> about 3 years ago and it was something in the environment. Can't
> remember if this was exactly it or not though but it sounds very
> similar.

  This is the output of the 'env' command:

PWD=/home/denis
TZ=KST-9
WINDOWID=18874381
HOSTNAME=tangrennorth
LD_LIBRARY_PATH=/home/denis/donegaldownloads/xforms/FORMS
DOOMWADDIR=/usr/games/doom
PVM_RSH=/usr/bin/rsh
QTDIR=/usr/lib/qt-2.3.0
LESSOPEN=|/usr/bin/lesspipe.sh %s
PS1=<\u@example.com:\w>$
XPVM_ROOT=/usr/share/pvm3/xpvm
KDEDIR=/usr
NNTPSERVER=nnrp.gol.com
SOCKS_NS=203.216.4.4
USER=denismcm
LS_COLORS=
SMTPSERVER=mail.gol.com
MACHTYPE=i386-redhat-linux-gnu
MAIL=/var/spool/mail/denismcm
INPUTRC=/etc/inputrc
XMODIFIERS=@example.com=none
LANG=en_US
DISPLAY=:0.0
LOGNAME=denismcm
SHLVL=4
XJDIC=/home/denisbin/xjdic22
TEXMF=/usr/local/share/texmf
WWW_HOME=/home/denis/internet/homepage.html
SESSION_MANAGER=local/tangrennorth:/tmp/.ICE-unix/1166
SHELL=/bin/bash
HOSTTYPE=i386
OSTYPE=linux-gnu
HISTSIZE=1000
LAMHELPFILE=/etc/lam/lam-helpfile
PVM_ROOT=/usr/share/pvm3
HOME=/home/denis
TERM=kterm
SSH_ASKPASS=/usr/libexec/openssh/gnome-ssh-askpass
XNLSPATH=/home/netscape/nls
PATH=/home/denis:/usr/bin:/usr/local/bin:/bin:/usr/X11R6/bin:.:/usr/games:/usr/games/doom:/home/denisbin/rvplayer5.0:
GNOME_DESKTOP_ICON=/usr/share/pixmaps/kterm.xpm
EXINIT=map q !}fmt 72 | map 
 !}fmtj 72 | map Z :.,$s/^//
_=/usr/bin/env
OLDPWD=/home/denis/internet

  Many of these things are defined in my ~/.bash_profile, of course, but
quite a few aren't.  Is there anything suspicious-looking here?  I'm still
quite nervous after the cracker was in last week (via my
dynamically-assigned IP address, yet!).   I've done a fresh install,
but I have added in my (enormous) ~/ directory tree (though the only
compiled executable in there that I can think of was this 'pppcosts' -
I absolutely should recompile it to be on the safe side, I realize).
All of the many shell scripts that I use seem to work just fine.

PVM_RSH=/usr/bin/rsh
LESSOPEN=|/usr/bin/lesspipe.sh %s
INPUTRC=/etc/inputrc
LAMHELPFILE=/etc/lam/lam-helpfile
SSH_ASKPASS=/usr/libexec/openssh/gnome-ssh-askpass
_=/usr/bin/env

  This ASKPASS, in particular, worries me, but it may well be standard.

On Mon, 28 May 2001, Thomas O'Dowd wrote:

> Another trick I would try is running strace on your current shell
> process from another terminal. ie $ strace -p <PID_OF_OTHER_SHELL>
> then try the command and see if you can see what might be failing...
> you might want to use the -f option if it manages to fork and exec
> successfully hitting on the problem in the sub shell.

  I'm afraid that I couldn't understand the last two lines here, but
I did manage to try the 'strace -p <PID_OF_OTHER_SHELL>' that you
suggested (took a couple of tries to realize that I couldn't capture
the output without 2>&1 redirection).  This means nothing to me, but
here's the output for a run of the following command:

<denismcm@example.com:~>$ ~/internet/pppcosts
bash: /home/denis/internet/pppcosts: No such file or directory

read(0, "\33", 1)                       = 1
read(0, "[", 1)                         = 1
read(0, "A", 1)                         = 1
write(2, "~/internet/pppcosts", 19)     = 19
rt_sigprocmask(SIG_BLOCK, NULL, [], 8)  = 0
read(0, "\r", 1)                        = 1
write(2, "\n", 1)                       = 1
rt_sigprocmask(SIG_BLOCK, [INT], [], 8) = 0
ioctl(0, SNDCTL_TMR_STOP, {B38400 opost isig icanon echo ...}) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
rt_sigaction(SIGINT, {0x807c5d0, [], 0x4000000}, {0x80a1cc0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGTERM, {SIG_IGN}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGQUIT, {SIG_IGN}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGALRM, {0x807c540, [HUP INT ILL TRAP ABRT BUS FPE USR1 SEGV USR2 PIPE ALRM TERM XCPU XFSZ VTALRM UNUSED], 0x4000000}, {0x80a1cc0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGTSTP, {SIG_IGN}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGTTOU, {SIG_IGN}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGTTIN, {SIG_IGN}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGWINCH, {0x8070b10, [], 0x4000000}, {0x80a1d40, [], 0x4000000}, 8) = 0
rt_sigaction(SIGINT, {0x807c5d0, [], 0x4000000}, {0x807c5d0, [], 0x4000000}, 8) = 0
rt_sigprocmask(SIG_BLOCK, [INT CHLD], [], 8) = 0
fork()                                  = 2103
setpgid(2103, 2103)                     = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
--- SIGCHLD (Child exited) ---
wait4(-1, [WIFEXITED(s) && WEXITSTATUS(s) == 127], WNOHANG|WUNTRACED, NULL) = 2103
wait4(-1, 0xbffff86c, WNOHANG|WUNTRACED, NULL) = -1 ECHILD (No child processes)
sigreturn()                             = ? (mask now [])
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD], [CHLD], 8) = 0
rt_sigprocmask(SIG_SETMASK, [CHLD], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD TSTP TTIN TTOU], [CHLD], 8) = 0
ioctl(255, TIOCSPGRP, [2103])           = -1 EPERM (Operation not permitted)
rt_sigprocmask(SIG_SETMASK, [CHLD], NULL, 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD TSTP TTIN TTOU], [CHLD], 8) = 0
ioctl(255, TIOCSPGRP, [2013])           = 0
rt_sigprocmask(SIG_SETMASK, [CHLD], NULL, 8) = 0
ioctl(255, TCGETS, {B38400 opost isig icanon echo ...}) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD TTOU], [CHLD], 8) = 0
rt_sigprocmask(SIG_SETMASK, [CHLD], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD], [CHLD], 8) = 0
rt_sigprocmask(SIG_SETMASK, [CHLD], NULL, 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, NULL, [], 8)  = 0
rt_sigaction(SIGINT, {0x807c5d0, [], 0x4000000}, {0x807c5d0, [], 0x4000000}, 8) = 0
time(NULL)                              = 991107319
stat64("/var/spool/mail/denismcm", {st_mode=S_IFREG|0600, st_size=43985, ...}) = 0
time(NULL)                              = 991107319
rt_sigprocmask(SIG_BLOCK, [CHLD TTOU], [], 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, [CHLD TSTP TTIN TTOU], [], 8) = 0
ioctl(255, TIOCSPGRP, [2013])           = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
rt_sigaction(SIGINT, {0x807c5d0, [], 0x4000000}, {0x807c5d0, [], 0x4000000}, 8) = 0
rt_sigprocmask(SIG_BLOCK, [INT], [], 8) = 0
ioctl(0, TIOCGWINSZ, {ws_row=25, ws_col=80, ws_xpixel=0, ws_ypixel=0}) = 0
ioctl(0, TIOCSWINSZ, {ws_row=25, ws_col=80, ws_xpixel=0, ws_ypixel=0}) = 0
ioctl(0, TCGETS, {B38400 opost isig icanon echo ...}) = 0
ioctl(0, SNDCTL_TMR_STOP, {B38400 opost isig -icanon -echo ...}) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
rt_sigaction(SIGINT, {0x80a1cc0, [], 0x4000000}, {0x807c5d0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGTERM, {0x80a1cc0, [], 0x4000000}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGTERM, {SIG_IGN}, {0x80a1cc0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGQUIT, {0x80a1cc0, [], 0x4000000}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGQUIT, {SIG_IGN}, {0x80a1cc0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGALRM, {0x80a1cc0, [], 0x4000000}, {0x807c540, [HUP INT ILL TRAP ABRT BUS FPE USR1 SEGV USR2 PIPE ALRM TERM XCPU XFSZ VTALRM UNUSED], 0x4000000}, 8) = 0
rt_sigaction(SIGTSTP, {0x80a1cc0, [], 0x4000000}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGTSTP, {SIG_IGN}, {0x80a1cc0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGTTOU, {0x80a1cc0, [], 0x4000000}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGTTOU, {SIG_IGN}, {0x80a1cc0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGTTIN, {0x80a1cc0, [], 0x4000000}, {SIG_IGN}, 8) = 0
rt_sigaction(SIGTTIN, {SIG_IGN}, {0x80a1cc0, [], 0x4000000}, 8) = 0
rt_sigaction(SIGWINCH, {0x80a1d40, [], 0x4000000}, {0x8070b10, [], 0x4000000}, 8) = 0
write(2, "<denismcm@example.com:~>$ ", 27) = 27
rt_sigprocmask(SIG_BLOCK, NULL, [], 8)  = 0
read(0,

  Sure hope this makes sense to somebody out there.  I need to fix this
problem soon, or I'll have retreat to the "safety" and (relative)
sanity of RH6.2 again.

  Thanks for thinking about this.  Cheers, Dennis