Bug #1697 for Net-Ping: 450

Mon Oct 21 12:20:57 2002 Guest - Ticket created

Subject:

450_service.t

This on FreeBSD 4.7 running perl 5.8 and Net-Ping-2.23 PERL_DL_NONLAZY=1 /usr/local/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/100_load...........ok t/110_icmp_inst......ok t/120_udp_inst.......ok t/130_tcp_inst.......ok t/140_stream_inst....ok t/150_syn_inst.......ok t/200_ping_tcp.......ok t/250_ping_hires.....ok t/300_ping_stream....ok 11/12 skipped: Local echo port is off t/400_ping_syn.......ok t/450_service........NOK 2# Failed test 2 in t/450_service.t at line 40 # t/450_service.t line 40 is: ok !!$sock1; t/450_service........NOK 3# Failed test 3 in t/450_service.t at line 54 # t/450_service.t line 54 is: ok !!$sock2; Can't call method "sockport" on an undefined value at t/450_service.t line 56. t/450_service........dubious Test returned status 49 (wstat 12544, 0x3100) DIED. FAILED tests 2-37 Failed 36/37 tests, 2.70% okay Failed Test Stat Wstat Total Fail Failed List of Failed ------------------------------------------------------------------------------- t/450_service.t 49 12544 37 36 97.30% 2-37 11 subtests skipped. Failed 1/11 test scripts, 90.91% okay. 36/98 subtests failed, 63.27% okay. *** Error code 2 Stop in /root/.cpan/build/Net-Ping-2.23. /usr/bin/make test -- NOT OK Running make install make test had returned bad status, won't install without force

Mon Oct 21 12:55:17 2002 rob [...] roobik.com - Correspondence added

Date:	Mon, 21 Oct 2002 10:54:44 -0600 (MDT)
From:	Rob Brown <rob [...] roobik.com>
To:	Guest via RT <bug-Net-Ping [...] rt.cpan.org>
Cc:	bbb [...] cpan.org
Subject:	Re: [cpan #1697] 450_service.t
RT-Send-Cc:

Version 2.24 hopefully fixes this. It's in cvs: cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping cd Net-Ping perl Makefile.PL make perl -Iblib t/450_service.t make test Let me know if this still fails and what the error is. Rob On Mon, 21 Oct 2002, Guest via RT wrote: Show quoted text

> > This message about Net-Ping was sent to you by guest via rt.cpan.org > > Full context and any attached attachments can be found at: > <URL: https://rt.cpan.org/Ticket/Display.html?id=1697 > > > This on FreeBSD 4.7 running perl 5.8 and Net-Ping-2.23 > > PERL_DL_NONLAZY=1 /usr/local/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t > t/100_load...........ok > t/110_icmp_inst......ok > t/120_udp_inst.......ok > t/130_tcp_inst.......ok > t/140_stream_inst....ok > t/150_syn_inst.......ok > t/200_ping_tcp.......ok > t/250_ping_hires.....ok > t/300_ping_stream....ok > 11/12 skipped: Local echo port is off > t/400_ping_syn.......ok > t/450_service........NOK 2# Failed test 2 in t/450_service.t at line 40 > # t/450_service.t line 40 is: ok !!$sock1; > t/450_service........NOK 3# Failed test 3 in t/450_service.t at line 54 > # t/450_service.t line 54 is: ok !!$sock2; > Can't call method "sockport" on an undefined value at t/450_service.t line 56. > t/450_service........dubious > Test returned status 49 (wstat 12544, 0x3100) > DIED. FAILED tests 2-37 > Failed 36/37 tests, 2.70% okay > Failed Test Stat Wstat Total Fail Failed List of Failed > ------------------------------------------------------------------------------- > t/450_service.t 49 12544 37 36 97.30% 2-37 > 11 subtests skipped. > Failed 1/11 test scripts, 90.91% okay. 36/98 subtests failed, 63.27% okay. > *** Error code 2 > > Stop in /root/.cpan/build/Net-Ping-2.23. > /usr/bin/make test -- NOT OK > Running make install > make test had returned bad status, won't install without force > >

Mon Oct 21 17:13:24 2002 Guest - Correspondence added

From:

larrysh [...] cpan.org

It appears to run fine now... I do get this running the 450_service.t manualy Can't locate object method "tcp_service_check" via package "Net::Ping" at t/450_service.t line 74. And when running perl Makefile.PL for the first time: perl Makefile.PL Checking if your kit is complete... Warning: the following files are missing in your kit: README perl-Net-Ping.spec Please inform the author. Writing Makefile for Net::Ping But I assume that is because you don't store those two files in CVS, as they are autogenerated, and go away after repeated runs of the Makefile.PL

Mon Oct 21 17:29:05 2002 rob [...] roobik.com - Correspondence added

Date:	Mon, 21 Oct 2002 15:28:57 -0600 (MDT)
From:	Rob Brown <rob [...] roobik.com>
To:	Dan Kogai <dankogai [...] dan.co.jp>
Cc:	easmith [...] beatrice.rutgers.edu, <hv [...] crypt.org>, <h.m.brand [...] hccnet.nl>, Rob Brown <bbb [...] cpan.org>, Blair Zajac <blair [...] orcaware.com>, Perl 5 Porters <perl5-porters [...] perl.org>, <cpan-testers [...] perl.org>, <Jost.Krieger+smokeback [...] ruhr-uni-bochum.de>, <bug-Net-Ping [...] rt.cpan.org>
Subject:	[cpan #1697] 450_service.t
RT-Send-Cc:

Wow, this seems nastier than I was hoping. I didn't think about the dual CPU boxes and such that could cause a non-blocking connect to really connect or return the error code immediately. I also still think that it COULD be a matter of Fcntl because when I comment out this section: $ grep -2 fcntl lib/Net/Ping.pm # Set O_NONBLOCK property on filehandle my $flags = 0; if (fcntl($fh, F_GETFL, $flags)) { if (!fcntl($fh, F_SETFL, $flags | O_NONBLOCK)) { croak("fcntl F_SETFL: $!"); } } else { croak("fcntl F_GETFL: $!"); } everything happens and fails exactly like you say occurs on freebsd and solaris boxes even on linux. And this is exactly the behavior on Win32 platform whether I comment it out or not. NOTE: If this is not the most correct way to do a nonblocking tcp connect, then tell me and I'll make the repairs. How does &IO::Handle::blocking or &IO::Socket::INET::connect implement it? Similar, right? Okay, I'm even handling this case too. Are there any errors now? Do a whole "make test": cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping cd Net-Ping perl Makefile.PL make make test Dang, I wish I had a Solaris or FreeBSD box to run this test on. I hate bothering y'all (not to mention bloating my cvs server with cruffy revisions). This is getting frustrating. -- Rob On Tue, 22 Oct 2002, Dan Kogai wrote: Show quoted text

> On Tuesday, Oct 22, 2002, at 03:32 Asia/Tokyo, Blair Zajac wrote:

> > This worked for me on Irix on perl 5.6.1 and 5.8.0.

> > Still not good'uff on FreeBSD; >

> > > sudo perl -Mblib t/450_service.t Password:1..26

> > # Running under perl version 5.008 for freebsd > > # Current time local: Mon Oct 21 18:55:24 2002 > > # Current time GMT: Tue Oct 22 03:55:24 2002 > > # Using Test.pm version 1.22 > > ok 1 > > ok 2 > > ok 3 > > ok 4 > > ok 5 > > ok 6 > > ok 7 > > ok 8 > > ok 9 > > ok 10 > > ok 11 > > ok 12 > > WARNING: Nonblocking connect connected anyway? (freebsd) at > > /usr/home/dankogai/work/Net-Ping/blib/lib/Net/Ping.pm line 779. > > ok 13 > > ok 14 > > ok 15 > > ok 16 > > ok 17 > > not ok 18 > > # Failed test 18 in t/450_service.t at line 145 > > # t/450_service.t line 145 is: ok $p -> ack(); > > ok 19 > > ok 20 > > WARNING: Nonblocking connect connected anyway? (freebsd) at > > /usr/home/dankogai/work/Net-Ping/blib/lib/Net/Ping.pm line 779. > > ok 21 > > ok 22 > > ok 23 > > ok 24 > > not ok 25 > > # Failed test 25 in t/450_service.t at line 186 > > # t/450_service.t line 186 is: if (!ok $p -> ping("127.0.0.1")) {warn > > "ERRNO: $!";} > > ERRNO: Connection refused at t/450_service.t line 186. > > ok 26

> > It's not a matter of Fcntl. There are many (and more and more) systems > that refuse SYN. Rob, I strongly suggest you change your default > platform to OpenBSD or FreeBSD if OpenBSD is too difficult to install. > Linux is way too loose on default network configuration. > > Dan the Man with Too Many "make test" For a Day > >

Mon Oct 21 17:35:55 2002 blair [...] orcaware.com - Correspondence added

Date:	Mon, 21 Oct 2002 14:34:08 -0700
From:	Blair Zajac <blair [...] orcaware.com>
To:	Rob Brown <rob [...] roobik.com>
Cc:	Dan Kogai <dankogai [...] dan.co.jp>, easmith [...] beatrice.rutgers.edu, hv [...] crypt.org, h.m.brand [...] hccnet.nl, Rob Brown <bbb [...] cpan.org>, Perl 5 Porters <perl5-porters [...] perl.org>, cpan-testers [...] perl.org, Jost.Krieger+smokeback [...] ruhr-uni-bochum.de, bug-Net-Ping [...] rt.cpan.org
Subject:	Re: [cpan #1697] 450_service.t
RT-Send-Cc:

Rob Brown wrote: Show quoted text

> > Wow, this seems nastier than I was hoping. I didn't > think about the dual CPU boxes and such that could > cause a non-blocking connect to really connect > or return the error code immediately. I also still > think that it COULD be a matter of Fcntl because when > I comment out this section: > > $ grep -2 fcntl lib/Net/Ping.pm > # Set O_NONBLOCK property on filehandle > my $flags = 0; > if (fcntl($fh, F_GETFL, $flags)) { > if (!fcntl($fh, F_SETFL, $flags | O_NONBLOCK)) { > croak("fcntl F_SETFL: $!"); > } > } else { > croak("fcntl F_GETFL: $!"); > } > > everything happens and fails exactly like you say > occurs on freebsd and solaris boxes even on linux. > And this is exactly the behavior on Win32 platform > whether I comment it out or not. > > NOTE: If this is not the most correct way to do a > nonblocking tcp connect, then tell me and I'll make > the repairs. How does &IO::Handle::blocking or > &IO::Socket::INET::connect implement it? Similar, > right? > > Okay, I'm even handling this case too. Are there > any errors now? Do a whole "make test": > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping > cd Net-Ping > perl Makefile.PL > make > make test

Works fine on Irix and RedHat 8.0. Best, Blair -- Blair Zajac <blair@orcaware.com> Web and OS performance plots - http://www.orcaware.com/orca/

Mon Oct 21 18:05:36 2002 easmith [...] beatrice.rutgers.edu - Correspondence added

From:	Ed Allen Smith <easmith [...] beatrice.rutgers.edu>
Date:	Mon, 21 Oct 2002 18:07:33 -0400
To:	rob [...] roobik.com
Cc:	bbb [...] cpan.org, blair [...] orcaware.com, bug-Net-Ping [...] rt.cpan.org
Subject:	Re: [cpan #1697] 450_service.t
RT-Send-Cc:

In message <Pine.LNX.4.44.0210211458310.27092-100000@rox.roobik.com> (on 21 October 2002 15:28:57 -0600), rob@roobik.com (Rob Brown) wrote: Show quoted text

> Wow, this seems nastier than I was hoping. I didn't > think about the dual CPU boxes and such that could > cause a non-blocking connect to really connect

I can test it out on a dual-processor box (with dual network interfaces at that) - IRIX 6.5.16m and Perl 5.8.0 - as soon as I stop getting the following error with CVS: cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping cvs checkout: cannot find password cvs checkout: use "cvs login" to log in first cvs [checkout aborted]: or set the CVS_PASSWORD environment variable (There may be a simple solution to this that I'm not thinking of because I just got up, of course....). -Allen Show quoted text

> Okay, I'm even handling this case too. Are there > any errors now? Do a whole "make test": > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping > cd Net-Ping > perl Makefile.PL > make > make test

-- Allen Smith http://cesario.rutgers.edu/easmith/ September 11, 2001 A Day That Shall Live In Infamy II "They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." - Benjamin Franklin

Mon Oct 21 19:21:30 2002 rob [...] roobik.com - Correspondence added

Date:	Mon, 21 Oct 2002 17:21:24 -0600 (MDT)
From:	Rob Brown <rob [...] roobik.com>
To:	Ed Allen Smith <easmith [...] beatrice.rutgers.edu>
Cc:	bbb [...] cpan.org, <bug-Net-Ping [...] rt.cpan.org>
Subject:	Re: [cpan #1697] 450_service.t
RT-Send-Cc:

Hmm, seems to be an old version of cvs. Guess you'll have to do this first: cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware login (Logging in to anonymous@cvs.roobik.com.) CVS password: (Just press enter, there is no password.) Then try the "checkout" line: cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping But I think I already got a report that it works on at least one Irix box. Not sure about dual CPU though. I mostly need Solaris and FreeBSD and OpenBSD to test. -- Rob On Mon, 21 Oct 2002, Ed Allen Smith wrote: Show quoted text

> In message <Pine.LNX.4.44.0210211458310.27092-100000@rox.roobik.com> (on 21 > October 2002 15:28:57 -0600), rob@roobik.com (Rob Brown) wrote:

> > Wow, this seems nastier than I was hoping. I didn't > > think about the dual CPU boxes and such that could > > cause a non-blocking connect to really connect

> > I can test it out on a dual-processor box (with dual network interfaces at > that) - IRIX 6.5.16m and Perl 5.8.0 - as soon as I stop getting the > following error with CVS: > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping > cvs checkout: cannot find password > cvs checkout: use "cvs login" to log in first > cvs [checkout aborted]: or set the CVS_PASSWORD environment variable > > (There may be a simple solution to this that I'm not thinking of because I > just got up, of course....). > > -Allen > >

> > Okay, I'm even handling this case too. Are there > > any errors now? Do a whole "make test": > > > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping > > cd Net-Ping > > perl Makefile.PL > > make > > make test

> > >

Mon Oct 21 19:38:56 2002 easmith [...] beatrice.rutgers.edu - Correspondence added

From:	Ed Allen Smith <easmith [...] beatrice.rutgers.edu>
Date:	Mon, 21 Oct 2002 19:42:55 -0400
To:	rob [...] roobik.com
Cc:	bbb [...] cpan.org, bug-Net-Ping [...] rt.cpan.org
Subject:	Re: [cpan #1697] 450_service.t
RT-Send-Cc:

In message <Pine.LNX.4.44.0210211717030.27092-100000@rox.roobik.com> (on 21 October 2002 17:21:24 -0600), rob@roobik.com (Rob Brown) wrote: Show quoted text

> Hmm, seems to be an old version of cvs.

Possible. The newer versions I've tried don't seem to compile properly with networking capabilities, IIRC. Show quoted text

> Guess you'll have to do this first: > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware login > (Logging in to anonymous@cvs.roobik.com.) > CVS password: (Just press enter, there is no password.) > > Then try the "checkout" line: > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping

Done, no problems. Show quoted text

> But I think I already got a report that it works on > at least one Irix box. Not sure about dual CPU though.

Looks OK, unless there's a system load dependence. Show quoted text

> I mostly need Solaris and FreeBSD and OpenBSD to test. > > -- > Rob > > On Mon, 21 Oct 2002, Ed Allen Smith wrote: >

> > In message <Pine.LNX.4.44.0210211458310.27092-100000@rox.roobik.com> (on 21 > > October 2002 15:28:57 -0600), rob@roobik.com (Rob Brown) wrote:

> > > Wow, this seems nastier than I was hoping. I didn't > > > think about the dual CPU boxes and such that could > > > cause a non-blocking connect to really connect

> > > > I can test it out on a dual-processor box (with dual network interfaces at > > that) - IRIX 6.5.16m and Perl 5.8.0 - as soon as I stop getting the > > following error with CVS: > > > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping > > cvs checkout: cannot find password > > cvs checkout: use "cvs login" to log in first > > cvs [checkout aborted]: or set the CVS_PASSWORD environment variable > > > > (There may be a simple solution to this that I'm not thinking of because I > > just got up, of course....). > > > > -Allen > > > >

> > > Okay, I'm even handling this case too. Are there > > > any errors now? Do a whole "make test": > > > > > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping > > > cd Net-Ping > > > perl Makefile.PL > > > make > > > make test

> > > > > >

>

-- Allen Smith http://cesario.rutgers.edu/easmith/ September 11, 2001 A Day That Shall Live In Infamy II "They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." - Benjamin Franklin

Tue Oct 22 11:37:20 2002 doughera [...] lafayette.edu - Correspondence added

Date:	Tue, 22 Oct 2002 11:37:22 -0400 (EDT)
From:	Andy Dougherty <doughera [...] lafayette.edu>
To:	Rob Brown <rob [...] roobik.com>
Cc:	Perl 5 Porters <perl5-porters [...] perl.org>, cpan-testers [...] perl.org, bug-Net-Ping [...] rt.cpan.org
Subject:	Re: [cpan #1697] 450_service.t
RT-Send-Cc:

On Mon, 21 Oct 2002, Rob Brown wrote: Show quoted text

> Okay, I'm even handling this case too. Are there > any errors now? Do a whole "make test": > > cvs -z3 -q -d :pserver:anonymous@cvs.roobik.com.:/usr/local/cvsroot/freeware checkout Net-Ping > cd Net-Ping > perl Makefile.PL > make > make test

Solaris 8 : PERL_DL_NONLAZY=1 /home/doughera/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/100_load...........ok t/110_icmp_inst......ok 1/2 skipped: icmp ping requires root privileges. t/120_udp_inst.......ok t/130_tcp_inst.......ok t/140_stream_inst....ok t/150_syn_inst.......ok t/200_ping_tcp.......ok t/250_ping_hires.....ok t/300_ping_stream....ok t/400_ping_syn.......skipped all skipped: no http port t/410_syn_host.......skipped all skipped: no http port t/450_service........# Failed test 18 in t/450_service.t at line 145 FAILED test 18 Failed 1/26 tests, 96.15% okay Failed 1/12 test scripts, 91.67% okay. 1/70 subtests failed, 98.57% okay. Failed Test Stat Wstat Total Fail Failed List of Failed ------------------------------------------------------------------------------- t/450_service.t 26 1 3.85% 18 2 tests and 1 subtest skipped. *** Error code 29 make: Fatal error: Command failed for target `test_dynamic' -- Andy Dougherty doughera@lafayette.edu

Thu Jan 30 00:12:45 2003 bbb [...] cpan.org - Taken

Thu Jan 30 00:23:44 2003 bbb [...] cpan.org - Correspondence added 60 min

Works on FreeBSD now.

Thu Jan 30 00:23:45 2003 bbb [...] cpan.org - Status changed from 'new' to 'resolved'

Bug #1697 for Net-Ping: 450_service.t