[Tcl-bugs] [ tcl-Bugs-1639656 ] unixInit.test unixInit-1.2 fail Solaris 9

[SourceForge.net <no...@so...>]

Bugs item #1639656, was opened at 2007-01-20 02:48
Message generated for change (Comment added) made by das
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=110894&aid=1639656&group_id=10894

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: 27. Channel Types
Group: current: 8.4.14
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Larry W. Virden (lvirden)
Assigned to: Andreas Kupries (andreas_kupries)
Summary: unixInit.test unixInit-1.2 fail Solaris 9

Initial Comment:
built on sparc solaris 9, sun c compiler.

compiled with
 ./configure  --prefix=/tmp/tcl8414 --enable-shared --enable-symbols --enable-stubs --enable-64bit --enable-threads

unixInit.test

==== unixInit-1.2 initialisation: standard channel type deduction FAILED
==== Contents of test case:

    # pipe1 is a connection to a server that reports what port it
    # starts on, and delivers a constant string to the first client to
    # connect to that port before exiting.
    set pipe1 [open "|[list [interpreter]]" r+]
    puts $pipe1 {
        proc accept {channel host port} {
            puts $channel {puts [fconfigure stdin -peername]; exit}
            close $channel
            exit
        }
        puts [fconfigure [socket -server accept 0] -sockname]
        vwait forever  }
    # Note the backslash above; this is important to make sure that the
    # whole string is read before an [exit] can happen...
    flush $pipe1
    set port [lindex [gets $pipe1] 2]
    set sock [socket localhost $port]
    # pipe2 is a connection to a Tcl interpreter that takes its orders
    # from the socket we hand it (i.e. the server we create above.)
    # These orders will tell it to print out the details about the
    # socket it is taking instructions from, hopefully identifying it
    # as a socket.  Which is what this test is all about.
    set pipe2 [open "|[list [interpreter] <@$sock]" r]
    set result [gets $pipe2]

    # Clear any pending data; stops certain kinds of (non-important) errors
    fconfigure $pipe1 -blocking 0; gets $pipe1
    fconfigure $pipe2 -blocking 0; gets $pipe2

    # Close the pipes and the socket.
    close $pipe2
    close $pipe1
    catch {close $sock}

    # Can't use normal comparison, as hostname varies due to some
    # installations having a messed up /etc/hosts file.
    if {
        [string equal 127.0.0.1 [lindex $result 0]] &&
        [string equal $port     [lindex $result 2]]
    } then {
        subst "OK"
    } else {
        subst "Expected: `[list 127.0.0.1 localhost $port]', Got `$result'"
    }

---- Result was:
Expected: `127.0.0.1 localhost 41536', Got `0.0.0.0 localhost 41536'
---- Result should have been (exact matching):
OK
==== unixInit-1.2 FAILED



----------------------------------------------------------------------

>Comment By: Daniel A. Steffen (das)
Date: 2007-01-24 03:07

Message:
Logged In: YES 
user_id=90580
Originator: NO

that same bug existed on the HEAD and was fixed there at the same time, so
if you don't see similar errors with 8.5, that change probably fixed this
issue as well.

to get the tip of core-8-4-branch from CVS, use
	cvs -d:pserver:ano...@tc...:/cvsroot/tcl co -r
core-8-4-branch tcl

or you could manually apply the patch in question to 8.4.14, it is e.g.
available at
	http://rutherglen.ics.mq.edu.au/fisheye/rdiff/Tcl?csid=core-8-4-branch:das:20061219041410&u&N


----------------------------------------------------------------------

Comment By: Larry W. Virden (lvirden)
Date: 2007-01-24 02:18

Message:
Logged In: YES 
user_id=15949
Originator: YES

I'm not really certain how to get that code - right now, I get my cvs head
snapshots from the activestate tcl ftp site, and their tcl snapshots are of
the tcl 8.5 branch.

----------------------------------------------------------------------

Comment By: Daniel A. Steffen (das)
Date: 2007-01-20 12:23

Message:
Logged In: YES 
user_id=90580
Originator: NO

I fixed a 64 bit bug in TclpInetNtoa that caused 0.0.0.0 test failures on
a ppc64 mac, please verify if the tip of core-8-4-branch still exhibits
your problem.

2006-12-19  Daniel Steffen  <da...@us...>
	* unix/tclUnixThrd.c (TclpInetNtoa): fix for 64 bit.


----------------------------------------------------------------------

Comment By: Daniel A. Steffen (das)
Date: 2007-01-20 12:23

Message:
Logged In: YES 
user_id=90580
Originator: NO

I fixed a 64 bit bug in TclpInetNtoa that caused 0.0.0.0 test failures on
a ppc64 mac, please verify if the tip of core-8-4-branch still exhibits
your problem.

2006-12-19  Daniel Steffen  <da...@us...>
	* unix/tclUnixThrd.c (TclpInetNtoa): fix for 64 bit.


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=110894&aid=1639656&group_id=10894