On Sat, 24 May 2014, Cory Smelosky wrote:
Damn.
boot [1,54]rsx11m.sys
SYSTEM FAULT DETECTED AT PC=044112 FACILITY=000300 ERROR CODE=000100
CRASH -- CONT WITH SCRATCH MEDIA ON MU000
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
At 11:18 PM +0200 24/5/14, Hans Vlems wrote:
The DIX spec says that hardware addresses ( the ones you put in the simh configuration file starting with 08-00-2B) must be unique .
I put no address in my simh conf file.
I tried to change it to -AA-BB-01 as you suggested above. I have now lost ssh access to my emulator and figure how to clear my ARP cache (arp -d -a wont 't do it under OS X).
--
Jean-Yves Bernier
On Sat, 24 May 2014, Dave McGuire wrote:
On 05/24/2014 05:14 PM, Cory Smelosky wrote:
Looks like progress!
TIME
17:09:44 24-MAY-14
;
TKB @RSXBLD
I don't recall it getting to this stage before!
Ah-HA, that's the biggie! Crossin' my fingers..
Yup!
It seems to have output less this time and jumped straight to the taskbuilding...it must've preallocated space last round.
-Dave
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On 05/24/2014 05:14 PM, Cory Smelosky wrote:
Looks like progress!
TIME
17:09:44 24-MAY-14
;
TKB @RSXBLD
I don't recall it getting to this stage before!
Ah-HA, that's the biggie! Crossin' my fingers..
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
At 10:26 PM +0200 24/5/14, Johnny Billquist wrote:
If I were to guess, there are packets dropped.
I can see the segment counter running backward at times, like this:
23:15:56.109519 10.2 > 10.1 586 data 45058>40193 seg 18
23:15:56.109855 10.2 > 10.1 378 data 45058>40193 seg 19
23:15:56.110118 10.2 > 10.1 586 data 45058>40193 seg 20
23:15:56.110277 10.2 > 10.1 586 data 45058>40193 seg 21
23:15:56.110548 10.2 > 10.1 586 data 45058>40193 seg 22
23:15:56.110736 10.2 > 10.1 364 data 45058>40193 seg 23
23:15:56.196612 10.2 > 10.1 586 data 45058>40193 seg 24
23:15:56.196816 10.2 > 10.1 586 data 45058>40193 seg 25
23:15:56.197113 10.2 > 10.1 586 data 45058>40193 seg 26
23:15:56.197316 10.2 > 10.1 394 data 45058>40193 seg 27
23:16:06.113164 10.2 > 10.1 586 data 45058>40193 seg 20
23:16:06.113458 10.2 > 10.1 586 data 45058>40193 seg 21
23:16:06.113664 10.2 > 10.1 586 data 45058>40193 seg 22
23:16:06.113994 10.2 > 10.1 364 data 45058>40193 seg 23
23:16:06.114152 10.2 > 10.1 586 data 45058>40193 seg 24
23:16:06.114304 10.2 > 10.1 586 data 45058>40193 seg 25
23:16:06.114657 10.2 > 10.1 586 data 45058>40193 seg 26
23:16:06.117048 10.2 > 10.1 394 data 45058>40193 seg 27
--
Jean-Yves Bernier
The DIX spec says that hardware addresses ( the ones you put in the simh configuration file starting with 08-00-2B) must be unique .
Verzonden vanaf mijn BlackBerry 10-smartphone.
Origineel bericht
Van: Jean-Yves Bernier
Verzonden: zaterdag 24 mei 2014 23:14
Aan: hecnet at Update.UU.SE
Beantwoorden: hecnet at Update.UU.SE
Onderwerp: Re: [HECnet] Same MAC address on different nodes
At 10:21 PM +0200 24/5/14, Johnny Billquist wrote:
So "Physical" is not "Hardware", and there is no need to SET XQ MAC.
I tried and it maked no change.
Right. Since your machine will set a different MAC address anyway,
at start, when you enable DECnet. The physical address is whatever
you set in simh before starting. It have no actual relevance once
DECnet starts. If you were to run something else, which did not use
DECnet (such as Unix), the address you set in simh will be the
actual MAC address used, since IP do not change the MAC address of
interfaces. No need.
You mean "the Hardware address is whatever you set in simh before
starting". CEX will set it to "Physical address" (in DECNET parlance)
which is AA-00-04-00-(1024 x area + node).
The MAC address is AA-00-04-00-01-28 for DECNET and 08-00-2B-AA-BB-CC
for Linux.
So, a NIC may have different MAC addresses depending of the network
stack running. Am I right here?
Really hard to give a good explanation on why without having good
access to all parts of the setup to be able to test and examine
things.
Well, my system is working 99%, that's not so bad :)
--
Jean-Yves Bernier
At 10:21 PM +0200 24/5/14, Johnny Billquist wrote:
So "Physical" is not "Hardware", and there is no need to SET XQ MAC.
I tried and it maked no change.
Right. Since your machine will set a different MAC address anyway, at start, when you enable DECnet. The physical address is whatever you set in simh before starting. It have no actual relevance once DECnet starts. If you were to run something else, which did not use DECnet (such as Unix), the address you set in simh will be the actual MAC address used, since IP do not change the MAC address of interfaces. No need.
You mean "the Hardware address is whatever you set in simh before starting". CEX will set it to "Physical address" (in DECNET parlance) which is AA-00-04-00-(1024 x area + node).
The MAC address is AA-00-04-00-01-28 for DECNET and 08-00-2B-AA-BB-CC for Linux.
So, a NIC may have different MAC addresses depending of the network stack running. Am I right here?
Really hard to give a good explanation on why without having good access to all parts of the setup to be able to test and examine things.
Well, my system is working 99%, that's not so bad :)
--
Jean-Yves Bernier
On Sat, 24 May 2014, Cory Smelosky wrote:
That was a mostly-joking comment. I'm not bothered by symbol errors I can add to a prefix file. Thank GOD I don't need to do a full executive gen every time!
Thank you!
-Dave
Looks like progress!
TIME
17:09:44 24-MAY-14
;
TKB @RSXBLD
I don't recall it getting to this stage before!
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On Sat, 24 May 2014, Dave McGuire wrote:
On 05/24/2014 05:00 PM, Cory Smelosky wrote:
No, IND is not picky about tabs, but why didn't you just add the line
to SGNPREFIX as I suggested instead?
I must have missed you mentioning putting it in SGNPREFIX. I was
reading rather quickly.
*grumble*
AT.T0 -- Undefined symbol $I$CBP
.IFT $I$CBP .AND .IFT $K$DAS .DATA #0 SET /PLCTL=ICB:600.:200.:51.
How many symbols need I manually define?! :P
Calm down, Cory. You've gotten farther with RSX in a few days than I
did in probably six months when I got my first PDP-11 in the 80s, and I
had a full (paper) manual set. You are doing friggin' FANTASTIC.
Seriously.
That was a mostly-joking comment. I'm not bothered by symbol errors I can add to a prefix file. Thank GOD I don't need to do a full executive gen every time!
Thank you!
-Dave
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On 05/24/2014 05:00 PM, Cory Smelosky wrote:
No, IND is not picky about tabs, but why didn't you just add the line
to SGNPREFIX as I suggested instead?
I must have missed you mentioning putting it in SGNPREFIX. I was
reading rather quickly.
*grumble*
AT.T0 -- Undefined symbol $I$CBP
.IFT $I$CBP .AND .IFT $K$DAS .DATA #0 SET /PLCTL=ICB:600.:200.:51.
How many symbols need I manually define?! :P
Calm down, Cory. You've gotten farther with RSX in a few days than I
did in probably six months when I got my first PDP-11 in the 80s, and I
had a full (paper) manual set. You are doing friggin' FANTASTIC.
Seriously.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
