Is there any special logic within tcl, when stdin & -out have
been closed? maybe, if next command is "exec" ?

I've got a tcl-script that gets started with stdin and stdout
connected to a socket.

Under some circumstances, these sockets get closed, and at that
point the script is only supposed to [exec ...] a (shortly
running) command, and then exit.

But instead, it keeps "hanging" after the exec.

Running it with "truss" shows closing of stdin & -out,
and immediately afterwards a pipe is created, and some
flag is set for them:

55572: 33847: close(0) = 0
55572: 33847: close(1) = 0

not sure if next 3 lines are aftermath of close, or part of [exec]

55572: 33847: pipe(0x2FF21FD8) = 0
55572: 33847: kfcntl(0, F_SETFD, 0x00000001) = 0
55572: 33847: kfcntl(1, F_SETFD, 0x00000001) = 0

... [exec ...]: kfork, childprocess does: ..., execve, ..., _exit(0)

and then the main process goes on to:

55572: 33847: kread(0, 0x205D2324, 4096) (sleeping...)

Reading from a pipe, whose write-side is kept by the same process.

I don't know where the read on stdin comes from. According to
the script it should do an "exit" right after the "exec".


So far, I cannot yet tell, if this is a "heisen-" or "bohr-" bug.
I see it happens a couple of times, but I don't get to see the
cases where it doesn't happen, if those exist.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)