1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
|
To install, enter just enter the command make.
You will be prompted for the installation locations.
If the binary doesn't work, delete it and re-compile.
NOTE: The rexec daemon (in.rexecd) installed with the
Red Hat 4.1 distribution doesn't seem to work all that well--
it does not shutdown the socket when the remote process exits,
so the client hangs indefinitely.
Sometimes, problems may be due to the daemon, not the client.
Send me (sadd@msc.cornell.edu) any problems with
installation or the code.
Mike Sadd
Change Log:
Jul. 6, 1997: (Version 1.4)
No new features added. Many people have been using
this rexec for non-Linux machines, so I cleaned up
the code to make it more portable. I have verified
that it now works correctly under AIX 4.1, but not
SunOS 4.1.4.
Unnecessary code to set up the auxillary channel
has been removed, since this is done already by rexec(3).
The signal handling routine has been re-written to
improve portability and reliability.
Users are warned in the man page that their password
will be transmitted clear text accross the network.
Feb. 24, 1997: (Version 1.3)
Mostly bug fixes for features introduced in 1.2.
Fortunately, 1.2 has not been distributed widely.
Enabled signal handling only after the rexec
subroutine returns. Once again, changed signal handling
to deal with buggy daemons--all signals are echoed,
but remain raised locally. This means that the local
rexec process will die with a CNTRL-C, even if the
remote wouldn't. This is intended to give the user an
out if the remote daemon refuses to shut down its end
of the socket.
Added the -b option if the user wishes for the command
to behave similarly to rsh(1).
Fixed a bug in echo_fd(), which lost data if an interrupt was
received while writing.
Rexec now allows the usage
rexec hostname -l username
as used with some rsh(1) commands.
Feb. 14, 1997: (Version 1.2)
Fixed much of the code related to signal handling.
Except for SIGTSTP, signals are simply echoed; (rexec
waits for the remote command to drop its standard out.)-changed
2/24/97
SIGTSTP is echoed to the remote command, then suspends
local execution. Remote execution should also be
suspended if it uses the same decimal codes for
signals.
Orange Gopher (2/10/97) and Johannes Plass
(plass@dipmza.physik.uni-mainz.de, Oct. 17 1996),
both provided me with modified codes which implemented
new command line switches. At their suggestion I
added the -l and -p switch, and the environmental
variables REXEC_USER and REXEC_PASS. These may be
considered a security risk, but no one is being forced
to use them :) By own practice is to use
no command line switches, then set up a .netrc
file on my single user machine, or be happy with
being prompted by the rexec(3) routine for the info.
Johannes also pointed out how to make getopt conform
to POSIX.
Cleaned up the man page to conform to Linux standards.
Sept. 16, 1996: Added code to allow rexec to pass signals to
the remote process. (Version 1.1)
Sept. 12, 1996: Original code (Version 1)
|
