[Simh] Simh and OpenVMS Integrity V8.4 Network Support issues
Mark Pizzolato - Info Comm
Mark at infocomm.com
Fri Mar 30 14:38:49 EDT 2012
Hi Maurizio,
So, no matter what you do, networking isn't working for you. You no longer have crashes, rght?
Several things:
1) You are running everything (the build) from an account with full privileges enabled (needed to load the PCAP execlet), right?
2) please provide the simulator command file you are using for these tests.
3) what does 'show xq eth' display?
4) can you: 'attach xq eth0' or 'attach xq eth1'?
5) exactly what did you do when you 'tried to assign a mac address on the device ETH0/1 without success'?
6) how are you observing 'performance' to say that one version is good and one is unacceptable?
7) did you reboot the system between each of the below changes of pcap-vms software?
The difference between the vms-pcap.zip code from Matt and from the github vms-pcap.zip should only be a few things which may remove compiler & link warnings on some OS and/or compiler versions. I believe that both should work even when such warnings are produced. If you're not getting compile/link warnings then there should be no functional difference.
I'm quite surprised to hear that you conclude that V3.8-1 runs acceptably fast and the latest versions don't run equally as fast. There shouldn't have been any changes which caused this to happen. Please answer the above questions and maybe something obvious will appear. In any case once I see how you're measuring performance I'll repeat similar measurements here and look into what may cause it.
I look forward to your feedback.
- Mark
On Friday, March 30, 2012 at 11:10 AM, Maurizio De Tommaso wrote:
> Hi Mark,
>
> I give you some update about my tests of the SIMH with network support
> under OpenVMS I64 V8.4 UPD600.
>
> 1) SIMH code V3.8-1 from simh site (http://simh.trailing-edge.com/) + PCAP
> form Mark site (https://github.com/markpizz/simh/downloads). I tested also
> the PCAP from Matt site (http://www.9track.net/simh/pcap. I tried also to
> assign a mac address on the devices ETH0/1 without success. The network
> doesn't work but the performance of the VM is very good.
>
> 2) Simh code V3.8-2rc2-170-g9449c1f + PCAP form Mark site
> (https://github.com/markpizz/simh/downloads). I tested also the PCAP from
> Matt site (http://www.9track.net/simh/pcap). The network doesn't work
> and the performance of the VM is very poor, not acceptable.
>
> 3) Simh code v3.9-0-rc1+ + PCAP form Mark site
> (https://github.com/markpizz/simh/downloads). I tested also the PCAP from
> Matt site (http://www.9track.net/simh/pcap). The network doesn't work
> and the performance of the VM is very poor, not acceptable.
>
> I repeat, It's very strange because with the rx2600,as you reported before,
> you are able to work with the emulated network adapter without error
> (...but what about the SIM performace with the new code ?)
>
> As reported by Matt, perhaps it is something to do with the brand of the
> network cards ?
>
> Any hint will be appreciated.
>
> Best regards,
> /Maurizio
>
> ----- Original Message -----
> From: "Mark Pizzolato - Info Comm" <Mark at infocomm.com>
> To: "Maurizio De Tommaso (Gmail)" <maurizio.detommaso at gmail.com>;
> "Matt"
> <scope.matthew at btinternet.com>; "Mark Pizzolato - Info Comm"
> <Mark at infocomm.com>
> Cc: "Hittner, David T (IS)" <david.hittner at ngc.com>; "Simh Trailing-Edge
> Mailing List" <simh at trailing-edge.com>
> Sent: Saturday, March 24, 2012 11:06 PM
> Subject: RE: [Simh] Simh and OpenVMS Integrity V8.4 Network Support
> issues
>
>
> Hi Maurizio,
>
> On Saturday, March 24, 2012 at 11:10 AM, Maurizio De Tommaso wrote:
> > Thank you so much Mark and Matt for the update.
> >
> > I downloaded your two kits and I tested them (one at a time) with SIMH
> > V3.8-1 under OpenVMS V8.4 Update600 environment. I confirm you that
> > with these new two kits the CPUSPINWAIT crash seems to be gone.
> >
> > Unfortunally under the SIMH environment the network still doesn't work
> > (protocols used : SCS, Decnet, LAT and TCP/IP - all correctly
> > configured on
> > XQA: device). Basically, the same problems observed three weeks ago,
> > when I started my tests on this environment.
> >
> > It's very strange because with his rx2660, Mark is able to work with
> > the emulated network adapter without error. What am I doing wrong ?
>
> Actually my system is an rx2600.
>
> My testing covered all protocols.
>
> In my environment, I have a simh VAX instance running continuously. This
> node is configured as a cluster boot server and I have a SIMH node defined in
> the cluster with MAC address 08:00:2b:11:22:44.
>
> The host system (rx2600) is running OpenVMS 8.4 and is NOT a member of
> the cluster.
>
> The host system has 2 network interfaces, with both connected to the same
> switched network.
>
> sim> sh xq eth
> ETH devices:
> 0 IE0 (VMS Device: _EIA0:)
> 1 WE0 (VMS Device: _EWA0:)
>
> Using the configuration file you've got, with the MAC address on XQ set to
> 08:00:2b:11:22:44 I can attach to EITHER interface (eth0 or eth1) and the
> following will boot the system:
> sim> b cpu
> KA655-B V5.3, VMB 2.7
> Performing normal system tests.
> 40..39..38..37..36..35..34..33..32..31..30..29..28..27..26..25..
> 24..23..22..21..20..19..18..17..16..15..14..13..12..11..10..09..
> 08..07..06..05..04..03..
> Tests completed.
> >>>b xqa0
> (BOOT/R5:0 XQA0
>
>
>
> 2..
> -XQA0
> 1..0..
>
>
> %VAXcluster, system loaded from node VS4000 (AA-00-04-00-18-04)
> [...]
>
> The cluster boot works, the booted system has DECnet, LAT and IP
> configured and all protocols work (except when I use the interface the host is
> using for DECnet, LAT and IP I can't talk directly to the host, but I can talk to
> other systems on the LAN)
>
> Things should work with your locally booted system as well. You should be
> able to attach to eth0 (vs eth1) and talk to other systems on your network.
> You may want to try using the same interface which the host system is using
> (you would then be sure to be using a network interface which you are
> absolutely sure is working).
>
> There have been very significant changes to the simh networking code since
> simh v3.8-1, I'm running the new stuff.
>
> The simh vax code I'm running with can be found at
> https://github.com/markpizz/simh/zipball/master (Take what that URL
> returns and save it as a zip file).
> Move it to your VMS host machine in binary mode.
> Unpack it with these commands
> $ unzip {zipfilename.zip}
> $ set default [.directory-containing scp.c]
> $ set file/attri=RFM:STM makefile,*.mms,[...]*.c,[...]*.h,[...]*.txt
>
> Build it with:
> $ mms vax
>
> Let me know where you get.
>
> - Mark
>
>
> > From Integrity side :
> >
> > SYSTEM_YODA > run SYS$SYSDEVICE:[SIMH]VAX-I64.EXE
> > sim> show xq eth
> > ETH devices:
> > 0 LL0 (VMS Device: _LLA0:)
> > 1 IE1 (VMS Device: _EIB0:) <-- I reserve this network adapter for
> > SIMH emulator !!!
> >
> > sim> set cpu 512m
> > sim> set cpu conhalt
> > sim> load -r ka655x.bin
> > sim> attach NVR ka655.nvr
> > sim> set telnet 5305
> > sim> set TQ enable
> > sim> set TS disable
> > sim> set LPT enable
> > sim> set RQB disable
> > sim> set RQC disable
> > sim> set RQD disable
> > sim> set RL disable
> > sim> set RY disable
> > sim> set rq2 disable
> > sim> set rq3 enable
> > sim> set tq1 enable
> > sim> set tq2 disable
> > sim> set tq3 disable
> > sim> set RL disable
> > sim> set LPT disable
> > sim> set TQ TK70
> > sim> set rq0 ra92
> > sim> attach rq0 ovmsvaxsys.dsk
> > sim> set rq1 ra92
> > sim> attach rq1 private.dsk
> > sim> set xq type=delqa
> > sim> attach xq eth1
> > sim> boot cpu
> >
> >
> >
> >
> > From SIMH side :
> >
> >
> > LAN Data Structures
> > -------------------
> > LSB address = A1C07580
> > Device state = 0013 Inited,Run,Timer
> >
> > -- XQA Unit Summary 24-MAR-2012 18:50:23 --
> >
> > UCB UCB Addr Fmt Value Client State
> > --- -------- --- ----- ------ -----------
> > XQA0 A1BAFF00
> > XQA16 A1D3CE40 Eth 60-03 DECNET 0017 Strtn,Len,Uniq,Strtd
> > XQA17 A1BB0180 Eth 60-01 MOPDL 001F
> > Strtn,Len,Uniq,Share,Strtd
> > XQA18 A1D37DC0 Eth 90-00 LOOP 001D
> > Strtn,Uniq,Share,Strtd
> > XQA19 A1C7D880 Eth 60-04 LAT 0015 Strtn,Uniq,Strtd
> > XQA20 A1C80300 Eth 08-00 IP 0015 Strtn,Uniq,Strtd
> > XQA21 A1C7F900 Eth 08-06 ARP 0015 Strtn,Uniq,Strtd
> > XQA22 A1C7EF00 Eth 86-DD 0015 Strtn,Uniq,Strtd
> > XQA23 A1BBDF00 Eth 00-00 0000
> > XQA24 A1C80D00 Eth 00-00 0000
> >
> >
> > SYSTEM_SIMH> ifconfig -a
> > LO0:
> > flags=100c89<UP,LOOPBACK,NOARP,MULTICAST,SIMPLEX,NOCHECKSUM>
> > inet 127.0.0.1 netmask ff000000 ipmtu 4096
> >
> > QE0: flags=c43<UP,BROADCAST,RUNNING,MULTICAST,SIMPLEX>
> > inet 192.168.1.8 netmask ffffffc0 broadcast 192.168.1.63 ipmtu
> > 1500
> >
> >
> > SYSTEM_SIMH> ping 192.168.1.1 ! (the router) PING 192.168.1.1
> > (192.168.1.1): 56 data bytes
> >
> >
> > ----192.168.1.1 PING Statistics----
> > 4 packets transmitted, 0 packets received, 100% packet loss
> > %SYSTEM-F-TIMEOUT, device timeout
> >
> > SYSTEM_SIMH> ping 192.168.1.5 ! (the Integrity Server) PING
> > 192.168.1.5 (192.168.1.5): 56 data bytes
> >
> >
> > ----192.168.1.5 PING Statistics----
> > 4 packets transmitted, 0 packets received, 100% packet loss
> > %SYSTEM-F-TIMEOUT, device timeout
> >
> >
> >
> >
> > From Integrity side :
> >
> > LAN Data Structures
> > -------------------
> >
> > -- EIB Device Summary 24-MAR-2012 18:51:45 --
> >
> > LSB address = 9E81A000
> > Device state = 00003103 RUN,INITED,RING_AVAIL,FP_CPUAFF,AUTO_ENA
> >
> > UCB UCB Addr Fmt Value Client State
> > --- -------- --- ----- ------ -----------
> > EIB0 8E266D00
> > EIB6 8ECFF640 Eth 00-00 0000
> > EIB12 8E924B80 Eth 60-06 USER 0015 STRTN,UNIQ,STRTD
> >
> >
> > SYSTEM_YODA >ifconfig -a
> > LE0: flags=c43<UP,BROADCAST,RUNNING,MULTICAST,SIMPLEX>
> > *inet 192.168.1.5 netmask ffffffc0 broadcast 192.168.1.63 ipmtu
> > 1500
> >
> > LO0:
> > flags=100c89<UP,LOOPBACK,NOARP,MULTICAST,SIMPLEX,NOCHECKSUM>
> > inet 127.0.0.1 netmask ff000000 ipmtu 4096
> >
> >
> > SYSTEM_YODA >ping 192.168.1.8 ! (the simh Ip address) PING 192.168.1.8
> > (192.168.1.8): 56 data bytes
> >
> >
> > ----192.168.1.8 PING Statistics----
> > 4 packets transmitted, 0 packets received, 100% packet loss
> > %SYSTEM-F-TIMEOUT, device timeout
> >
> >
> >
> > It is very interesting to observe that the LAT protocol behaves
> > differently (Incoming and Outcoming connection are enabled on both
> > nodes - rx1600 and SIMH ) :
> >
> > SYSTEM_SIMH> mc latcp sho service
> >
> > Service Name Status Identification
> > ---------------- ----------- -------------------------------------------------
> > SIMH Available @SYS$MANAGER:ANNOUNCE.COM
> > VAX_SIMULATOR Available SIMH-Vax running on rx1620 & OpenVMS
> > V8.4
> >
> >
> > SYSTEM_YODA >mc latcp sho service
> >
> > Service Name Status Identification
> > ---------------- ----------- -------------------------------------------------
> > JEDY Available OpenVMS Alias Name Cluster
> > SIMH Available @SYS$MANAGER:ANNOUNCE.COM
> > VAX_SIMULATOR Available SIMH-Vax running on rx1620 & OpenVMS
> > V8.4
> > YODA Available @SYS$MANAGER:ANNOUNCE.COM
> >
> >
> > SYSTEM_YODA >set host/lat VAX_SIMULATOR %LAT-F-PROTERR,
> communication
> > protocol error -LAT-I-END, control returned to node YODA
> >
> > SYSTEM_YODA >set host/lat simh
> > %LAT-F-PROTERR, communication protocol error -LAT-I-END, control
> > returned to node YODA
> >
> > It seems that LAT protocol under SIMH is able to transmit but it is
> > not able to receive :
> >
> > EIB12 8E9279C0 Eth 60-06 USER 0015 STRTN,UNIQ,STRTD
> > (Integrity Server)
> > XQA11 A1D50080 Eth 60-04 LAT 0015 Strtn,Uniq,Strtd (Simh)
> >
> >
> > Any hint will be appreciated.
> >
> > Best regards,
> > /Maurizio
> >
> >
> > ----- Original Message -----
> > From: "Matt" <scope.matthew at btinternet.com>
> > To: "Mark Pizzolato - Info Comm" <Mark at infocomm.com>
> > Cc: "Maurizio De Tommaso (Gmail)" <maurizio.detommaso at gmail.com>;
> > "Hittner, David T (IS)" <david.hittner at ngc.com>; "Simh Trailing-Edge
> > Mailing List"
> > <simh at trailing-edge.com>
> > Sent: Thursday, March 22, 2012 10:37 PM
> > Subject: Re: [Simh] Simh and OpenVMS Integrity V8.4 Network Support
> > issues
> >
> >
> > > On 22/03/2012 20:42, Mark Pizzolato - Info Comm wrote:
> > >>
> > >> Well, I looked into this. I've got an rx2600 recently acquired
> > >> cheap on eBay. I updated the firmware on this system to the latest
> > >> before installing VMS and doing any testing.
> > >>
> > >> I started with the vms-pcap.zip downloaded from:
> > >> https://github.com/markpizz/simh/downloads
> > >>
> > >> Everything worked as expected. No crashes and full Ethernet behavior.
> > >>
> > >> I then took the vms-pcap.zip from http://www.9track.net/simh/pcap
> > >>
> > >> I expected these to be the same since I worked with Matt last year
> > >> when I picked up his fixes and merged them into the code on github.
> > >>
> > >> They were not the same. The code from
> > http://www.9track.net/simh/pcap
> > >> produced the crashes which Maurizio described.
> > >>
> > >> Looking for what was different, I compared the source. The code on
> > >> github.com had a version of pcapvcm.c which had comments describing
> > >> it
> > as
> > >> Version X-3. The code from 9track.net had comments describing it
> > >> as Version X-2. There was also a difference in snprintf.c which
> > >> cleaned up compiler warnings on some platforms. Additionally,
> > >> there were an extra file in the zip file from github.com called
> > >> 0readme_pcap-VMS.txt which describes the relative directory
> > >> structure the pcap-vms code should have with respect to the simh
> > >> source which references it (via simh's descrip.mms file).
> > >>
> > >> I'm not sure why Matt's website has the older code in its zip file
> > >> since I got the code on github from Matt. The right answer seems
> > >> to use the code from github.
> > >>
> > >> Meanwhile, I also looked at the libpcap provided with OpenVMS 8.4
> > TCP/IP
> > >> 5.7 ECO3. This libpcap implementation does NOT have
> > >> pcap_sendpacket,
> > so
> > >> it won't be useful for simh networking.
> > >>
> > >> - Mark Pizzolato
> > >>
> > >>
> > > I know what's happened here. I recently had to move to a new server,
> > > due to disk problems and the website content was restored from the
> > > copy on my local system, which should have been up to date but
> > > clearly wasn't. I think I've uploaded the correct version now (it
> > > identifies itself as X-3). Sorry for any inconvenience this has caused.
> > >
> > > Maurizio, does this fix the problems you were seeing?
> > >
> > > Matt
> >
>
>
More information about the Simh
mailing list