This is the mail archive of the mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Can't spawn shell for authenticated user in SSHD


I want merely the ability to do limited remote WINNT command prompt sys administration tasks without exposing myself to the plain text and implementation vulnerabilities of telent server; i.e., I want to launch /winnt/system32/cmd.exe upon logging in via a SSH client to SSHD on an NT 4.0, SP 6 system.

To that end I used the cygwin setup program to download and install from the Internet the openssh package. This process apparently worked correctly, as I can get components, including sshd, running. In fact, launching sshd in debug mode, I can use a ssh client on another machine and successfully (apparently, from the debug trace) authenticate to sshd. However, it looks as though when sshd proceeds to spawn my shell, there is some sort of a process launch/communication failure, with the following debug text emitted from sshd:

Accepted password for Administrator from port 1540 ssh2
debug1: Entering interactive session for SSH2.
debug1: server_init_dispatch_20
debug1: server_input_channel_open: ctype session rchan 0 win 8192 max 8192
debug1: input_session_request
debug1: channel 0: new [server-session]
debug1: session_new: init
debug1: session_new: session 0
debug1: session_open: channel 0
debug1: session_open: session 0: link with channel 0
debug1: server_input_channel_open: confirm session
debug1: session_by_channel: session 0 channel 0
debug1: session_input_channel_req: session 0 channel 0 request pty-req reply 0
debug1: Allocating pty.
debug1: session_pty_req: session 0 alloc /dev/tty0
debug1: session_by_channel: session 0 channel 0
debug1: session_input_channel_req: session 0 channel 0 request shell reply 0
      0 [main] sshd 290 sync_with_child: child 300(0x148) died before initialization with status code 0x80
    304 [main] sshd 290 sync_with_child: *** child state waiting for longjmp
Disconnecting: fork failed: Resource temporarily unavailable
debug1: Calling cleanup 0x414188(0x43e794)
debug1: session_pty_cleanup: session 0 release /dev/tty0
debug1: Calling cleanup 0x41eb24(0x0)
debug1: channel_free: channel 0: server-session, nchannels 1
debug1: Calling cleanup 0x4169cc(0x0)

The /etc/passwd entry for the user I'm logging in as looks like this:


and my mount environment looks like:

  C:\cygwin\bin on /usr/bin type system (textmode)
  C:\cygwin\lib on /usr/lib type system (textmode)
  C:\cygwin on / type system (textmode)
  c: on /c type system (textmode)

It appears to me from the debug output that the cmd.exe process does get launched, but then something goes awry, and I have been unsuccessful in resolving this.

Can anyone help me with this? I did try to search the archives for references to this problem, but found nothing that seemed to be an exact match to this scenario.

John Fishbeck

Unsubscribe info:
Bug reporting:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]