RSX11 Tape file Structure?
Holm Tiffe
I want to install RSX11 just for fun on my PDP11/53 thats more a stack of
PCBs than a computer...
This site here has some different software packages:
http://pdp-11.org.ru/files.pl
Ther is a file named rsx11m42.zip that contains a image from a bootable tape.
It runs with simh, already testet this. Now I want to install it with a real
Tape on the PDP11. I have an Tandberg TDC4222 5GB Tape Drive connected to an
TMSCP simulating DILOG SC703 that works fine with 2.11BSD.
My ssems to be that the file m42kit.tap from the zip archive is an image
from an entire tape w/o any blocking information. I've tried blockszizes of
512 na 1024Bytes already but I get always this:
KDJ11-D/S E.03 Media not bootable
What could I do now?
Kind Regards,
Holm
Roger Ivie
> Ther is a file named rsx11m42.zip that contains a image from a bootable tape.
> It runs with simh, already testet this.
> My ssems to be that the file m42kit.tap from the zip archive is an image
> from an entire tape w/o any blocking information.
It's been a while since I looked at the SIMH tape image format, but as I
recall each block consists of a block size followed by the data for the
block. I'm pretty certain the SIMH manuals have documentation.
I believe SIMH used the same tape format as rmt; you might try looking
for rmt documentation as well.
--
roger ivie
ri...@ridgenet.net
Johnny Billquist
As Roger Ivie suggests - if it works with simh, there must already be
blocking information in the file, or else simh would not be able to use
the file either.
However, the most common format for tapes on a file for simh is actually
4 bytes of blocksize, then data, then the same four bytes of block size
again. The exception being if the block is a tape mark, which is just a
single four bytes of zero. Not repeated.
An alternative format is just two bytes of block size, followed by the
data. And then next block.
By the way, I hope you also understand how to actually write a tape from
a Unix system, with correct block sizes. Don't try to use dd, since that
will most definitely not be useful.
As for the actual block sizes of the tape, for a bootable tape on a
PDP-11 in general, the first two blocks are usually identical, and is
boot blocks. 512 bytes each. After that, it's up to the boot block (and
thus the system).
For RSX, a popular block size for tapes is 4144 bytes. But since RSX
uses ANSI format tapes, you'll have some 80 byte blocks in there with
volume headers and file headers, and so on, before you actually get to
any data blocks. And also, for backups and the like, the last block of
any save set can be just about any size. And tape marks are also used,
which look like 0-byte blocks when you read them. But writing tape marks
is a special operation.
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: b...@softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
Holm Tiffe
> As Roger Ivie suggests - if it works with simh, there must already be
> blocking information in the file, or else simh would not be able to use
> the file either.
>
> However, the most common format for tapes on a file for simh is actually
> 4 bytes of blocksize, then data, then the same four bytes of block size
> again. The exception being if the block is a tape mark, which is just a
> single four bytes of zero. Not repeated.
>
> An alternative format is just two bytes of block size, followed by the
> data. And then next block.
>
> By the way, I hope you also understand how to actually write a tape from
> a Unix system, with correct block sizes. Don't try to use dd, since that
> will most definitely not be useful.
>
> As for the actual block sizes of the tape, for a bootable tape on a
> PDP-11 in general, the first two blocks are usually identical, and is
> boot blocks. 512 bytes each. After that, it's up to the boot block (and
> thus the system).
>
> For RSX, a popular block size for tapes is 4144 bytes. But since RSX
> uses ANSI format tapes, you'll have some 80 byte blocks in there with
> volume headers and file headers, and so on, before you actually get to
> any data blocks. And also, for backups and the like, the last block of
> any save set can be just about any size. And tape marks are also used,
> which look like 0-byte blocks when you read them. But writing tape marks
> is a special operation.
>
> Johnny
:-) Yes, ioctls..
Ok, thanks guys for the hint in the right direction.
I've already read simh's documentation about tape structures and
hacked together a simple program that wrote me a tape.
It worked at the first try! :-)
Still wondering if there is a "raw" device in simh just to exchange data
without filtering. I'e wondered about hte ugly characters in a uuencoded
tar file that I wrote out in simh just to get the 2.11BSD Kernel Configuration
to the real machine. Filtered with tail and col -b to get it clean...
Now I kow better what happened here.
I've never seen block sizes of 4144 before, tought this was a bug
in my program first..
[snip]
foffs 0006a664, blen bl 581 = 4096:1000
foffs 0006b66c, blen bl 582 = 4096:1000
foffs 0006c674, blen bl 583 = 2048:0800
foffs 0006ce7c, blen bl 584 = 80:0050
foffs 0006ced4, blen bl 585 = 4096:1000
foffs 0006dedc, blen bl 586 = 4096:1000
foffs 0006eee4, blen bl 587 = 4096:1000
foffs 0006feec, blen bl 588 = 4096:1000
foffs 00070ef4, blen bl 589 = 4096:1000
foffs 00071efc, blen bl 590 = 4096:1000
foffs 00072f04, blen bl 591 = 4096:1000
foffs 00073f0c, blen bl 592 = 4096:1000
foffs 00074f14, blen bl 593 = 3584:0e00
foffs 00075d1c, blen bl 594 = 80:0050
foffs 00075d74, blen bl 595 = 4096:1000
foffs 00076d7c, blen bl 596 = 4096:1000
foffs 00077d84, blen bl 597 = 4096:1000
foffs 00078d8c, blen bl 598 = 4096:1000
foffs 00079d94, blen bl 599 = 4096:1000
foffs 0007ad9c, blen bl 600 = 4096:1000
foffs 0007bda4, blen bl 601 = 4096:1000
foffs 0007cdac, blen bl 602 = 4096:1000
foffs 0007ddb4, blen bl 603 = 2048:0800
foffs 0007e5bc, blen bl 604 = 80:0050
foffs 0007e614, blen bl 605 = 4144:1030
foffs 0007f64c, blen bl 606 = 4144:1030
foffs 00080684, blen bl 607 = 4144:1030
foffs 000816bc, blen bl 608 = 4144:1030
[snap]
The tape has a rather complex structure... hmmm
that are ~200k of those logs above.
Now I have to decrypt what SYSGEN want's to know from me :-~
BTW: In Sysgens device listing has no DEQNA, can I set this to DEUNA?
What means non mapped System? It thinks I have 1n 11/34, typing in
11/53 doesn't work....
Regards,
Holm
Bob Eager
> Yes, ioctls..
>
> Ok, thanks guys for the hint in the right direction. I've already read
> simh's documentation about tape structures and hacked together a simple
> program that wrote me a tape. It worked at the first try!
> Still wondering if there is a "raw" device in simh just to exchange data
> without filtering. I'e wondered about hte ugly characters in a uuencoded
> tar file that I wrote out in simh just to get the 2.11BSD Kernel
> Configuration to the real machine. Filtered with tail and col -b to get
> it clean... Now I kow better what happened here.
Tried using the paper tape punch?
--
Use the BIG mirror service in the UK:
http://www.mirrorservice.org
*lightning protection* - a w_tom conductor
Holm Tiffe
Bob Eager <rd...@spamcop.net> writes:
> On Mon, 21 Feb 2011 11:53:57 +0000, Holm Tiffe wrote:
>
>> In article <ijsdn9$nh7$1...@iltempo.update.uu.se>,
>> Johnny Billquist <b...@softjar.se> writes:
>>> On 2011-02-20 20:24, Holm Tiffe wrote:
>>>> Hi,
>
> Tried using the paper tape punch?
>
No, not jet. Thanks for the Idea, I'll try this.
Regards,
Holm
Bob Eager
It's worked for me in the past. Let us know how you get on.
Johnny Billquist
> Ok, thanks guys for the hint in the right direction.
No problems.
> I've already read simh's documentation about tape structures and
> hacked together a simple program that wrote me a tape.
> It worked at the first try! :-)
Excellent!
> Still wondering if there is a "raw" device in simh just to exchange data
> without filtering. I'e wondered about hte ugly characters in a uuencoded
> tar file that I wrote out in simh just to get the 2.11BSD Kernel Configuration
> to the real machine. Filtered with tail and col -b to get it clean...
> Now I kow better what happened here.
What do you mean by a "raw" device? Send just random bytes? How about a
terminal line?
A raw chunk of "data"? What about a disk?
It is in fact only tapes that are special, since you need to simulate
the physical tape marks, and Unix and other "modern" OSes don't really
have the concept of records.
> I've never seen block sizes of 4144 before, tought this was a bug
> in my program first..
:-)
> The tape has a rather complex structure... hmmm
> that are ~200k of those logs above.
If you just read the specification for ANSI format tapes, you'll have it
all. It's mot really that complex.
What makaes it more complex is the fact that those 4144 byte records are
actually a part of a BRU saveset, so in fact, that is another level of
archiving, in which you then have directories, files, and so on...
> Now I have to decrypt what SYSGEN want's to know from me :-~
Ah... SYSGEN for 11M. That is a complex beast... :-)
However, the manual is available on bitsavers.
Also, if you are running 11M+ instead of 11M, the sysgen is much more
user friendly.
> BTW: In Sysgens device listing has no DEQNA, can I set this to DEUNA?
No.
The DEUNA and DEQNA are very different. Unfortunately there is no
separate device driver for a DEQNA from DEC. There is a driver module if
you install DECnet, but otherwise you are semi-out of luck.
However, I think there was/is a SEQNA device driver written by someone
and submitted to DECUS. If you search through
ftp://ftp.update.uu.se/pub/pdp11/rsx it should be there somewhere.
> What means non mapped System? It thinks I have 1n 11/34, typing in
> 11/53 doesn't work....
Non-mapped system is a system without an MMU...
As for CPU model. It might be that SYSGEN isn't really trying to guess
what CPU you have, and is just presenting some kind of a default value.
You should be able to list what models it knows about.
What version are you trying to install, btw? The 11/34 is a bit older
than an 11/53, so maybe you are trying a version that was released
before the 11/53 existed?
Holm Tiffe
Bob Eager <rd...@spamcop.net> writes:
> On Mon, 21 Feb 2011 16:27:24 +0000, Holm Tiffe wrote:
>
>> In article <8sf6ds...@mid.individual.net>,
>> Bob Eager <rd...@spamcop.net> writes:
>>> On Mon, 21 Feb 2011 11:53:57 +0000, Holm Tiffe wrote:
>>>
>>>> In article <ijsdn9$nh7$1...@iltempo.update.uu.se>,
>>>> Johnny Billquist <b...@softjar.se> writes:
>>>>> On 2011-02-20 20:24, Holm Tiffe wrote:
>>>>>> Hi,
>> [..]>> it clean... Now I kow better what happened here.
>>>
>>> Tried using the paper tape punch?
>>>
>>>
>> No, not jet. Thanks for the Idea, I'll try this.
>
> It's worked for me in the past. Let us know how you get on.
>
Hmm, it doesn't look that 2.11BSD has a device or driver for punch....
Regards,
Holm
Bob Eager
It should have raw disk devices. dd from a file to /dev/rrd... or
whatever (I forget the device names after the decades), after attaching a
file to that disk.
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-21 12:53, Holm Tiffe wrote:
>> Ok, thanks guys for the hint in the right direction.
>
> No problems.
>
>> I've already read simh's documentation about tape structures and
>> hacked together a simple program that wrote me a tape.
>> It worked at the first try! :-)
>
> Excellent!
>
>> Still wondering if there is a "raw" device in simh just to exchange data
>> without filtering. I'e wondered about hte ugly characters in a uuencoded
>> tar file that I wrote out in simh just to get the 2.11BSD Kernel Configuration
>> to the real machine. Filtered with tail and col -b to get it clean...
>> Now I kow better what happened here.
>
> What do you mean by a "raw" device? Send just random bytes? How about a
> terminal line?
> A raw chunk of "data"? What about a disk?
I meant a way out of simhs 2.11BSD system to the Host system (here FreeBSD 8).
I tought forst that writing a tar file to one of simhs tape device files is
a easy way to do that.
With a disk file this is possible too, I must write to the raw (character)
device then. I've first thinkered about a tape, since a tape was what I've
wanted (to put onme in the 11/53 drive).
Now I have the network up and running so this isn't a problem anymore, but
the GENERIC Kernel had no networking.
>
>
> It is in fact only tapes that are special, since you need to simulate
> the physical tape marks, and Unix and other "modern" OSes don't really
> have the concept of records.
Ok,ok...
>
>> I've never seen block sizes of 4144 before, tought this was a bug
>> in my program first..
>
>:-)
>
>> The tape has a rather complex structure... hmmm
>> that are ~200k of those logs above.
>
> If you just read the specification for ANSI format tapes, you'll have it
> all. It's mot really that complex.
...but not really that interesting too. I think I'll never get a tape from an
other source than me...
>
> What makaes it more complex is the fact that those 4144 byte records are
> actually a part of a BRU saveset, so in fact, that is another level of
> archiving, in which you then have directories, files, and so on...
BRU... aha. :-|
I've read this while trying to install RSX11. Don't know what this means.
>
>> Now I have to decrypt what SYSGEN want's to know from me :-~
>
> Ah... SYSGEN for 11M. That is a complex beast... :-)
> However, the manual is available on bitsavers.
>
> Also, if you are running 11M+ instead of 11M, the sysgen is much more
> user friendly.
Where to get one?
There is a RSX-11M-PLUS V2.1 Baseline (Disk image RD52) and a
RSX-11M V3.2 for 11/45, 124KW, 50Hz (Disk image RK07) on the website
http://pdp-11.org.ru/files.pl. Sould I thake one from them?
Installing from a tape was relatively easy, but I don't have an RD52 nor
a RK07. But I have an RQDX3 and old 20MB disks....
I don't know of this 11/45 OS will run on my 11/53.
BTW: we have 50Hz here.
>
>> BTW: In Sysgens device listing has no DEQNA, can I set this to DEUNA?
>
> No.
> The DEUNA and DEQNA are very different. Unfortunately there is no
> separate device driver for a DEQNA from DEC. There is a driver module if
> you install DECnet, but otherwise you are semi-out of luck.
> However, I think there was/is a SEQNA device driver written by someone
> and submitted to DECUS. If you search through
> ftp://ftp.update.uu.se/pub/pdp11/rsx it should be there somewhere.
I'll look for it.
>
>> What means non mapped System? It thinks I have 1n 11/34, typing in
>> 11/53 doesn't work....
>
> Non-mapped system is a system without an MMU...
Ok, got it right also.
>
> As for CPU model. It might be that SYSGEN isn't really trying to guess
> what CPU you have, and is just presenting some kind of a default value.
> You should be able to list what models it knows about.
> What version are you trying to install, btw? The 11/34 is a bit older
> than an 11/53, so maybe you are trying a version that was released
> before the 11/53 existed?
>
> Johnny
>
The rsx11 came from here:
http://pdp-11.org.ru/files.pl
Caution, thats a russian site. look for a file "rsx11m42.zip" with google and
let google translate the site for you.
9 8 7 6 5 4 3 2 1
Commands are Help, Boot, List, Map, Test and Wrap.
Type a command then press the RETURN key: B DU1
DU1
RSX-11M V4.2 BL38 124.K MAPPED
>RED DU1:=SY:
>RED DU1:=LB:
>MOU DU1:11MKITV42
>@DU1:[1,2]STARTUP
>* PLEASE ENTER TIME AND DATE (HR:MN DD-MMM-YY) [S]: 19:05 22-FEB-11
>TIM 19:05 22-FEB-11
>* ENTER LINE WIDTH OF THIS TERMINAL [D D:132.]:
>SET /BUF=TI:132.
>ACS SY:/BLKS=1024.
>@ <EOF>
>
I've already let sysgen run and it rattelet some time on the disk, but..
It seems, that I have an chicken and egg problem:
I don't know anything about RSX11, thats all "§$)$&/$&/&" !.
There is no documentation on the net for Sysadmins that know nothing
about installing RSX11 right for the first time.
The beginners docs always presuppose a running and fine set up system
to begin with.
Because of this something like this is a big problem:
;
>; Create Executive build files
>;
>PIP RSXMC.MAC=SY0:RSXMC0.MAC/AP
>;
>* 1. Do you wish to edit any of the Executive files? [Y/N]: N
>;
.... yes, I've made a typo and wanted to edit the file TTDRVBLD.CMD to
add a controller. But how the f***k can I edit this file?
I've failed at this position to produce a directory listing with pip, typed
something in that put almost any assembler source file to TI:. Could stop it
with CTRL-C, but the next typed in char resumed this until reset. :-(
Is anyone here that says UNIX-OSes are cryptic and user unfriendly?
WHO SAYS THIS? (raised eyebrow)
Ok, next problem is the language, I've a natural german speaker, never had
english in the school or something like this (its only from Unix and
communicating with People like you on the net, yes I know swedish is
different to english also...), so I don't fully understood this setup process.
I'm suspecting now, that my install to the DU1 was a mistake.
From that what I've read in the meantime I suppose that my DU1 is now
something like a scratch disk from where the installion to the real
system disk has to take place. If this is the case I have to remove the
entire "\xa7$)$&/$&/&", add a new "scratch disk" to the UC07 and start
from the beginning..
Is this the case?
Regards,
Holm
Bob Eager
> On Tue, 22 Feb 2011 17:43:39 +0000, Holm Tiffe wrote:
>
>> In article <8sfm43...@mid.individual.net>,
>> Bob Eager <rd...@spamcop.net> writes:
>>> On Mon, 21 Feb 2011 16:27:24 +0000, Holm Tiffe wrote:
>>>
>>>> In article <8sf6ds...@mid.individual.net>,
>>>> Bob Eager <rd...@spamcop.net> writes:
>>>>> On Mon, 21 Feb 2011 11:53:57 +0000, Holm Tiffe wrote:
>>>>>
>>>>>> In article <ijsdn9$nh7$1...@iltempo.update.uu.se>,
>>>>>> Johnny Billquist <b...@softjar.se> writes:
>>>>>>> On 2011-02-20 20:24, Holm Tiffe wrote:
>>>>>>>> Hi,
>>>> [..]>> it clean... Now I kow better what happened here.
>>>>>
>>>>> Tried using the paper tape punch?
>>>>>
>>>>>
>>>> No, not jet. Thanks for the Idea, I'll try this.
>>>
>>> It's worked for me in the past. Let us know how you get on.
>>>
>>>
>> Hmm, it doesn't look that 2.11BSD has a device or driver for punch....
>
> It should have raw disk devices. dd from a file to /dev/rrd... or
> whatever (I forget the device names after the decades), after attaching
> a file to that disk.
2.11BSD should have files in /dev that look like the normal disk names
with an 'r' on the front. Write to those, and you're writing to the
'disk' from its very first block. That should give you a file (attached
to the disk in simh) with exactly what you wrote. You write it using 'dd'.
Rich Alderson
> In article <ijudu1$q62$1...@iltempo.update.uu.se>,
> Johnny Billquist <b...@softjar.se> writes:
>> What makaes it more complex is the fact that those 4144 byte records are
>> actually a part of a BRU saveset, so in fact, that is another level of
>> archiving, in which you then have directories, files, and so on...
> BRU... aha. :-|
> I've read this while trying to install RSX11. Don't know what this means.
Backup/Restore Utility.
[snip]
> I've already let sysgen run and it rattelet some time on the disk, but..
> It seems, that I have an chicken and egg problem:
> I don't know anything about RSX11, thats all "§$)$&/$&/&" !.
> There is no documentation on the net for Sysadmins that know nothing
> about installing RSX11 right for the first time.
> The beginners docs always presuppose a running and fine set up system
> to begin with.
I find the following at Bitsavers.org.
http://bitsavers.org/pdf/dec/pdp11/rsx11/RSX11A/DEC-11-IRSAA-A-D_pgmrRef_Mar73.pdf
http://bitsavers.org/pdf/dec/pdp11/rsx11/RSX11D_V6.2_Feb77/DEC-11-OXDIA-E-sysgen_Feb77.pdf
http://bitsavers.org/pdf/dec/pdp11/rsx11/RSX11M_V4.1_Apr83/AA-H625C-TC_sysgen_Apr83.pdf
Don't let the name of the first one scare you off. There is a chapter on
system generation in this manual. (RSX-11A was the earliest version of
the OS, and the documentation was more compact.)
I'm recommending both 11A and 11D as well as 11M, so that you get a feel
for how DEC wrote their documentation, and something of a feel for why
the later versions look the way they do.
> Because of this something like this is a big problem:
OK, we've solved, or at least we've begun to solve, this big problem.
> Ok, next problem is the language, I've a natural german speaker, never
> had english in the school or something like this (its only from Unix
> and communicating with People like you on the net, yes I know swedish
> is different to english also...), so I don't fully understood this
> setup process.
OK, now *that's* a big problem. Ich studierte Deutsch, and I'd still
hesitate to try to install software I didn't know using German docs,
much less Swedish (to use your examples).
I applaud your courage.
> I'm suspecting now, that my install to the DU1 was a mistake.
> From that what I've read in the meantime I suppose that my DU1 is now
> something like a scratch disk from where the installion to the real
> system disk has to take place. If this is the case I have to remove the
> entire "\xa7$)$&/$&/&", add a new "scratch disk" to the UC07 and start
> from the beginning..
It depends on what disk(s) you will use on your system. See chapter 2 of
the 11M manual above for a table of how many disks and how many tapes are
involved depending on what kind of disk.
--
Rich Alderson ne...@alderson.users.panix.com
the russet leaves of an autumn oak/inspire once again the failed poet/
to take up his pen/and essay to place his meagre words upon the page...
Johnny Billquist
Well, you know that you can run tar against disks as well?
(Or atleast you used to be able to do that way back in ancient history...)
Anyway, since the problem is solved, I think we can leave this.
>> What makaes it more complex is the fact that those 4144 byte records are
>> actually a part of a BRU saveset, so in fact, that is another level of
>> archiving, in which you then have directories, files, and so on...
>
> BRU... aha. :-|
> I've read this while trying to install RSX11. Don't know what this means.
Rich Alderson already answered this one. To clarify a bit more - BRU is
the RSX equivalent to dump/restore (or tar) in Unix, more or less. But
they are not compatible, and don't totally overlap as functionality goes.
>>> Now I have to decrypt what SYSGEN want's to know from me :-~
>>
>> Ah... SYSGEN for 11M. That is a complex beast... :-)
>> However, the manual is available on bitsavers.
>>
>> Also, if you are running 11M+ instead of 11M, the sysgen is much more
>> user friendly.
>
> Where to get one?
How about looking at ftp://www.trailing-edge.com/pub/rsx_dists/ ?
> There is a RSX-11M-PLUS V2.1 Baseline (Disk image RD52) and a
> RSX-11M V3.2 for 11/45, 124KW, 50Hz (Disk image RK07) on the website
> http://pdp-11.org.ru/files.pl. Sould I thake one from them?
I think not. See above... ;-)
> Installing from a tape was relatively easy, but I don't have an RD52 nor
> a RK07. But I have an RQDX3 and old 20MB disks....
> I don't know of this 11/45 OS will run on my 11/53.
>
> BTW: we have 50Hz here.
The frequency is specified when you run a SYSGEN. If you get it wrong,
it only means that the clock in the OS will be running a bit fast, or
slow, depending on which way it is wrong. :-)
If you want to install on a real 11/53, and only have a 20MB disk, you
are going to be very short on space, and you might not be able to
install 11M+ correctly at all...
Start playing with simh to get a feel for the whole thing.
> I've already let sysgen run and it rattelet some time on the disk, but..
>
> It seems, that I have an chicken and egg problem:
>
> I don't know anything about RSX11, thats all "§$)$&/$&/&" !.
:-)
But we're here to help you.
> There is no documentation on the net for Sysadmins that know nothing
> about installing RSX11 right for the first time.
Sure there is:
http://www.bitsavers.org/pdf/dec/pdp11/rsx11/RSX11M_V4.1_Apr83/
or
http://www.bitsavers.org/pdf/dec/pdp11/rsx11/RSX11Mplus_V4.x/
will get you pretty recent versions of manuals. Be prepared, though.
These are big manuals, with lots of information. Not something your read
through in one hour. But there you will find whole manuals dedicated to
only SYSGEN, as well as manuals for system administration, program
development, just use the system, and whatnot.
> The beginners docs always presuppose a running and fine set up system
> to begin with.
What documentation are you thinking of here?
> Because of this something like this is a big problem:
>
> ;
>> ; Create Executive build files
>> ;
>> PIP RSXMC.MAC=SY0:RSXMC0.MAC/AP
>> ;
>> * 1. Do you wish to edit any of the Executive files? [Y/N]: N
>> ;
>
>
> .... yes, I've made a typo and wanted to edit the file TTDRVBLD.CMD to
> add a controller. But how the f***k can I edit this file?
You do not want to edit the file to add a controller. At this point, you
have already passed by the point in the SYSGEN, where you tell which
controllers you have. Editing TTDRVBLD.CMD is many times not enough to
add additional controllers, so don't even try that unless you are an
expert on RSX systems. SYSGEN is your friend here, and you should use it.
> I've failed at this position to produce a directory listing with pip, typed
> something in that put almost any assembler source file to TI:. Could stop it
> with CTRL-C, but the next typed in char resumed this until reset. :-(
> Is anyone here that says UNIX-OSes are cryptic and user unfriendly?
>
> WHO SAYS THIS? (raised eyebrow)
Me? :-)
Anyway, there are some basic concepts you really need to know to get the
hang of things. RSX is sometimes, and in some ways, not a trivial
experience. And since you have not even got a system setup properly yet,
it becomes even more "exciting". :-)
This is like running a Unix system in single user mode, with no file
systems mounted yet, and no /etc/fstab, or just about any other config
file set up. How easy do you think such a scenario would be for someone
who have never seen unix before?
Anyway, to point out a few very basic things first of all:
^C don't abort any program in RSX. Any command you type in RSX is
essentially like doing anything in unix, and terminating it with "&".
^C in RSX just brings up a prompt from your command line interpreter. At
that point, you can kill the command, if you want to. Pressing enter
will just finish the CLI part, and the program will continue its output,
if that is what it was doing.
Getting a directory listing: PIP /LI
In your baseline system, you are using MCR, which is a very weird CLI
for "young" people. Once you have the system set up, you can have DCL
instead, which is way more user friendly.
> Ok, next problem is the language, I've a natural german speaker, never had
> english in the school or something like this (its only from Unix and
> communicating with People like you on the net, yes I know swedish is
> different to english also...), so I don't fully understood this setup process.
Can't really help much with the language. While I do speak basic german
(I live in Zürich), I'm not at all proficient enough to use it at this
level.
However, the setup process as such is that when you install the system
from tape, it restores a very basic, rudimentary RSX system for you. The
main purpose of this system is really just to allow you to run through a
SYSGEN, in order to compile a system which actually match what hardware
you have, and what options you want.
> I'm suspecting now, that my install to the DU1 was a mistake.
No, it was not (really).
> From that what I've read in the meantime I suppose that my DU1 is now
> something like a scratch disk from where the installion to the real
> system disk has to take place. If this is the case I have to remove the
> entire "\xa7$)$&/$&/&", add a new "scratch disk" to the UC07 and start
> from the beginning..
No. You can install to any disk. At boot time, RSX sees what disk it was
booted from, and sets up pointers that this disk is the system disk. The
end result of this is that any disk is as good as another as your system
disk. It makes no difference. Running a SYSGEN means you will compile a
whole kernel, and lots of system programs and device drivers, which will
be used to build a new system image. This image normally will live on
the same disk. You then bootstrap the new image, and then you use that
one instead. You can (of course) move to another disk if you want, but
there is no requirement to do this.
Also, you can change the unit number of a disk and reboot, and it will
still work fine. There is no hard connections inside a built system to a
unit number. What unit the system disk have is detected at boot time (as
mentioned above), and everything adapts to this. Unlike Unix, which
sometimes have weird configs at build time which ties the kernel to a
specific disk for the root file system and swap area, and in most cases
point to specific disks in /etc/fstab which must be changed if you move
disks around.
So, with RSX, your DU1: can just be changed to DU0:, and off you go.
Johnny Billquist
> ho...@freibergnet.de (Holm Tiffe) writes:
>
>> In article<ijudu1$q62$1...@iltempo.update.uu.se>,
>> Johnny Billquist<b...@softjar.se> writes:
>
>> It seems, that I have an chicken and egg problem:
>
>> I don't know anything about RSX11, thats all "§$)$&/$&/&" !.
>> There is no documentation on the net for Sysadmins that know nothing
>> about installing RSX11 right for the first time.
>> The beginners docs always presuppose a running and fine set up system
>> to begin with.
>
> I find the following at Bitsavers.org.
>
> http://bitsavers.org/pdf/dec/pdp11/rsx11/RSX11A/DEC-11-IRSAA-A-D_pgmrRef_Mar73.pdf
>
> http://bitsavers.org/pdf/dec/pdp11/rsx11/RSX11D_V6.2_Feb77/DEC-11-OXDIA-E-sysgen_Feb77.pdf
>
> http://bitsavers.org/pdf/dec/pdp11/rsx11/RSX11M_V4.1_Apr83/AA-H625C-TC_sysgen_Apr83.pdf
>
> Don't let the name of the first one scare you off. There is a chapter on
> system generation in this manual. (RSX-11A was the earliest version of
> the OS, and the documentation was more compact.)
>
> I'm recommending both 11A and 11D as well as 11M, so that you get a feel
> for how DEC wrote their documentation, and something of a feel for why
> the later versions look the way they do.
Nice with pointers, and the recommendations are well meant. But for a
novice, these are hefty manuals already. Having to plow through several
of them might scare him off forever, Rich. :-)
I would just start directly with the 11M SYSGEN manual, and at a later
time, when everything is working fine, and you start asking yourself
"why", then you can start with the full story. :-)
>> Ok, next problem is the language, I've a natural german speaker, never
>> had english in the school or something like this (its only from Unix
>> and communicating with People like you on the net, yes I know swedish
>> is different to english also...), so I don't fully understood this
>> setup process.
>
> OK, now *that's* a big problem. Ich studierte Deutsch, and I'd still
> hesitate to try to install software I didn't know using German docs,
> much less Swedish (to use your examples).
>
> I applaud your courage.
Me too.. (And Swedish is simple... ;-) )
>> I'm suspecting now, that my install to the DU1 was a mistake.
>> From that what I've read in the meantime I suppose that my DU1 is now
>> something like a scratch disk from where the installion to the real
>> system disk has to take place. If this is the case I have to remove the
>> entire "\xa7$)$&/$&/&", add a new "scratch disk" to the UC07 and start
>> from the beginning..
>
> It depends on what disk(s) you will use on your system. See chapter 2 of
> the 11M manual above for a table of how many disks and how many tapes are
> involved depending on what kind of disk.
Unless you install to RL01, RL02 or RK06, you normally get by with a
single disk. (Not so fond memories of installing to RL02 disks pop into
my head... I think 5 are required, although only 2 drives.)
DU implies MSCP disk. Hopefully enough to live and breathe...
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-22 19:41, Holm Tiffe wrote:
>> In article<ijudu1$q62$1...@iltempo.update.uu.se>,
>> Johnny Billquist<b...@softjar.se> writes:
>>> On 2011-02-21 12:53, Holm Tiffe wrote:
>>>> Ok, thanks guys for the hint in the right direction.
>>>
>> wanted (to put onme in the 11/53 drive).
>> Now I have the network up and running so this isn't a problem anymore, but
>> the GENERIC Kernel had no networking.
>
> Well, you know that you can run tar against disks as well?
Yes, I know.
> (Or atleast you used to be able to do that way back in ancient history...)
> Anyway, since the problem is solved, I think we can leave this.
>
>>> What makaes it more complex is the fact that those 4144 byte records are
>>> actually a part of a BRU saveset, so in fact, that is another level of
>>> archiving, in which you then have directories, files, and so on...
>>
>> BRU... aha. :-|
>> I've read this while trying to install RSX11. Don't know what this means.
>
> Rich Alderson already answered this one. To clarify a bit more - BRU is
> the RSX equivalent to dump/restore (or tar) in Unix, more or less. But
> they are not compatible, and don't totally overlap as functionality goes.
Ahh, ok.
>
>>>> Now I have to decrypt what SYSGEN want's to know from me :-~
>>>
>>> Ah... SYSGEN for 11M. That is a complex beast... :-)
>>> However, the manual is available on bitsavers.
>>>
>>> Also, if you are running 11M+ instead of 11M, the sysgen is much more
>>> user friendly.
>>
>> Where to get one?
>
> How about looking at ftp://www.trailing-edge.com/pub/rsx_dists/ ?
>
>> There is a RSX-11M-PLUS V2.1 Baseline (Disk image RD52) and a
>> RSX-11M V3.2 for 11/45, 124KW, 50Hz (Disk image RK07) on the website
>> http://pdp-11.org.ru/files.pl. Sould I thake one from them?
>
> I think not. See above... ;-)
Ok, also I' restarting with this: ftp://www.trailing-edge.com/pub/rsx_dists/rsx11mpbl87.dsk.bz2 ?
>
>> Installing from a tape was relatively easy, but I don't have an RD52 nor
>> a RK07. But I have an RQDX3 and old 20MB disks....
>> I don't know of this 11/45 OS will run on my 11/53.
>>
>> BTW: we have 50Hz here.
>
> The frequency is specified when you run a SYSGEN. If you get it wrong,
> it only means that the clock in the OS will be running a bit fast, or
> slow, depending on which way it is wrong. :-)
Ok,ok, I'm an electronican and Unix Systems Admin...
>
> If you want to install on a real 11/53, and only have a 20MB disk, you
> are going to be very short on space, and you might not be able to
> install 11M+ correctly at all...
What you call real? No It isn't real, its a stack of PCBs in a H9278-A
Backplane w/o the card cage mechanics, stabilized with some matchboxes between
the cards and laying on my desk.. :-)
>
> Start playing with simh to get a feel for the whole thing.
>
>> I've already let sysgen run and it rattelet some time on the disk, but..
>>
>> It seems, that I have an chicken and egg problem:
>>
>> I don't know anything about RSX11, thats all "§$)$&/$&/&" !.
>
>:-)
> But we're here to help you.
Thats why I'm asking for help here :-)
>
>> There is no documentation on the net for Sysadmins that know nothing
>> about installing RSX11 right for the first time.
>
> Sure there is:
> http://www.bitsavers.org/pdf/dec/pdp11/rsx11/RSX11M_V4.1_Apr83/
> or
> http://www.bitsavers.org/pdf/dec/pdp11/rsx11/RSX11Mplus_V4.x/
Good, I'll get it and read something.
>
> will get you pretty recent versions of manuals. Be prepared, though.
> These are big manuals, with lots of information. Not something your read
> through in one hour. But there you will find whole manuals dedicated to
> only SYSGEN, as well as manuals for system administration, program
> development, just use the system, and whatnot.
Hmm, I'm trying to setup RSX11 just for fun, and "to have had a look at it".
Maybe later this is useful to get an old big eastern german PDP11-Clone running
again.
Take a look at this K1630:
http://www.robotrontechnik.de/index.htm?/html/computer/k1600.htm
This machine is standing in a museum and People from Robotrontechnik.de
would get it up aganin. Currently there are no system Software for that beast.
The Disks are bad (gummed up bearings it seems).
This is not a 1:1 clone of an Dec Machine, the PCBs and the Card connectors
are totally different. The Hardware should be somewhat compatible tough..
It has some I3000 based emulator Hardware for the Robotron R4000
(Honeywell DDP516 clone) too..
I think this Machine is some of the slowest existing PDP-11s..
In the nerxt time I'll get a complete card cadge from an russian
Elektronika-E60 too, this is a LSI11 with QBUS but... the russians made the
card connector in metric dimensions including the space between the cold
fingers...
Maybe I'll take the complete 19" CPU, its a question of space, lets see..
http://www.tiffe.de/Robotron/E60/DSCF0065.JPG
http://www.tiffe.de/Robotron/E60/DSCF0066.JPG
http://www.tiffe.de/Robotron/E60/DSCF0067.JPG
On the last picture at the right side you can see an card cage from the E60.
Thre are some images from the cards too:
http://www.tiffe.de/Robotron/E60/E60-01.jpg
http://www.tiffe.de/Robotron/E60/E60-02.jpg
http://www.tiffe.de/Robotron/E60/E60-03.jpg
http://www.tiffe.de/Robotron/E60/E60-04.jpg
http://www.tiffe.de/Robotron/E60/E60-05.jpg
http://www.tiffe.de/Robotron/E60/E60-06.jpg
http://www.tiffe.de/Robotron/E60/E60-07.jpg
The E60 had no hard disk.
There is plenty to play with..
>
>> The beginners docs always presuppose a running and fine set up system
>> to begin with.
>
> What documentation are you thinking of here?
Hmm, maybe something like the doc to for setting up 2.11BSD, like this:
http://minnie.tuhs.org/PUPS/Setup/2.11bsd_setup.html
>
>> Because of this something like this is a big problem:
>>
>> ;
>>> ; Create Executive build files
>>> ;
>>> PIP RSXMC.MAC=SY0:RSXMC0.MAC/AP
>>> ;
>>> * 1. Do you wish to edit any of the Executive files? [Y/N]: N
>>> ;
>>
>>
>> .... yes, I've made a typo and wanted to edit the file TTDRVBLD.CMD to
>> add a controller. But how the f***k can I edit this file?
>
> You do not want to edit the file to add a controller. At this point, you
> have already passed by the point in the SYSGEN, where you tell which
> controllers you have. Editing TTDRVBLD.CMD is many times not enough to
> add additional controllers, so don't even try that unless you are an
> expert on RSX systems. SYSGEN is your friend here, and you should use it.
Ok, I wanted to change the number of DL Disks frome one to two exactly,
I have an Dilog DQ614 here that emulates 2 RL disks.
>
>> I've failed at this position to produce a directory listing with pip, typed
>> something in that put almost any assembler source file to TI:. Could stop it
>> with CTRL-C, but the next typed in char resumed this until reset. :-(
>> Is anyone here that says UNIX-OSes are cryptic and user unfriendly?
>>
>> WHO SAYS THIS? (raised eyebrow)
>
> Me? :-)
Naaaa?!?
>
> Anyway, there are some basic concepts you really need to know to get the
> hang of things. RSX is sometimes, and in some ways, not a trivial
> experience. And since you have not even got a system setup properly yet,
> it becomes even more "exciting". :-)
> This is like running a Unix system in single user mode, with no file
> systems mounted yet, and no /etc/fstab, or just about any other config
> file set up. How easy do you think such a scenario would be for someone
> who have never seen unix before?
With the manual I've mentioned above this should be possible..
>
>
> Anyway, to point out a few very basic things first of all:
> ^C don't abort any program in RSX. Any command you type in RSX is
> essentially like doing anything in unix, and terminating it with "&".
> ^C in RSX just brings up a prompt from your command line interpreter. At
> that point, you can kill the command, if you want to. Pressing enter
> will just finish the CLI part, and the program will continue its output,
> if that is what it was doing.
>
> Getting a directory listing: PIP /LI
This wasn't that easy, tried this.
>
> In your baseline system, you are using MCR, which is a very weird CLI
> for "young" people. Once you have the system set up, you can have DCL
> instead, which is way more user friendly.
Hmm, I'm 48 years old, but no one asked me If I want an interface for
old People :-)
>
>> Ok, next problem is the language, I've a natural german speaker, never had
>> english in the school or something like this (its only from Unix and
>> communicating with People like you on the net, yes I know swedish is
>> different to english also...), so I don't fully understood this setup process.
>
> Can't really help much with the language. While I do speak basic german
> (I live in Zürich), I'm not at all proficient enough to use it at this
> level.
Normally I don't have problems with this, I've read the english documentation
for some Hardware first bevore I try to understand the german part.
The german part is almost translated at least twice .. you know what I mean?
>
> However, the setup process as such is that when you install the system
> from tape, it restores a very basic, rudimentary RSX system for you. The
> main purpose of this system is really just to allow you to run through a
> SYSGEN, in order to compile a system which actually match what hardware
> you have, and what options you want.
Ok, maybe I've invested not enough time for this task.
I have an Emulex UC07 currently in the machine, there is a 4GB IBM SCSI
Drive connected to the UC07. The Emulex is setup to split the drive in 2 parts
2GB each. (2GB is already to big for RT-11). I've collected some 1,2 and 4GB
SCSI Drives lately from some friends so disk space should not be a problem.
On the first half of the current disk sits happyly a 2.11BSD, I wanted to
install RSX-11 on the 2nd half.
In the card bin sits an UC08 to, (a KA630 and 18MB of memory too) and there
is a 2nd card cage for experiments in the future..
I have an RQDX3 too but used it only for the floppies jet.
If I try to install RSX-11M+ on simh then I have the problem to get the
data (or disk contents) to the real pdp11 again, that's why I've installed
this RSX11 from the tape (thats connected to an Dilog SQ703 wich emulates
TMSCP). The Tape drive is a Tandberg TDC4222 that was handy, it seems that
the hardware and at least the bootroms of the KDJ11-DS are happy with it.
Tieed to connect an DLT40 yesterday evening instead but it doesn't worked.
I think I have to setup it with the SQ703 first..
Regards and thanks to Rich too..
Holm
Holm Tiffe
ho...@freibergnet.de (Holm Tiffe) writes:
> In article <ik1cs2$105$1...@iltempo.update.uu.se>,
> Johnny Billquist <b...@softjar.se> writes:
>> On 2011-02-22 19:41, Holm Tiffe wrote:
>>> In article<ijudu1$q62$1...@iltempo.update.uu.se>,
>>> Johnny Billquist<b...@softjar.se> writes:
>>>> On 2011-02-21 12:53, Holm Tiffe wrote:
>>>>> Ok, thanks guys for the hint in the right direction.
>>>>
[..]
>>>
Both disk images there are approx. 160MB decompressed, to what device should I attach them in simh to boot?
Regards,
Holm
Holm Tiffe
Couldn't get to boot this, if I set rq0 to RD54 or RA92 I get this:
RSX-11M-PLUS V4.6 BL87 128.KW System:"RSXMPL"
SAV -- Task too big for partition, task removed - ...CON
>RED DU1:=SY:
17:14:03 *** GEN Checkpoint space allocation failure
17:14:03 *** GEN Checkpoint space allocation failure
17:14:04 *** GEN Checkpoint space allocation failure
17:14:04 *** GEN Checkpoint space allocation failure
17:14:05 *** GEN Checkpoint space allocation failure
17:14:05 *** GEN Checkpoint space allocation failure
MCR>
17:14:06 *** GEN Checkpoint space allocation failure
^R
MCR>
17:14:06 *** GEN Checkpoint space allocation failure
^R
MCR>
17:14:07 *** GEN Checkpoint space allocation failure
^R
MCR>
Simulation stopped, PC: 020432 (BR 20400)
sim>
So for which disk type are the images made?
Regards,
Holm
Rich Alderson
> Nice with pointers, and the recommendations are well meant. But for a
> novice, these are hefty manuals already. Having to plow through several
> of them might scare him off forever, Rich. :-)
The thought I had is that the earlier manuals assume less on the part of
the installer, and explain more for that reason. This gives the installer
some background for the 11M manual.
I speak from experience. I have yet to complete an 11M installation, and
I've been using the derivative RSX-20F for decades (as you know ;-).
Rich Alderson
> Couldn't get to boot this, if I set rq0 to RD54 or RA92 I get this:
> RSX-11M-PLUS V4.6 BL87 128.KW System:"RSXMPL"
> SAV -- Task too big for partition, task removed - ...CON
> >RED DU1:=SY:
> 17:14:03 *** GEN Checkpoint space allocation failure
> 17:14:03 *** GEN Checkpoint space allocation failure
> 17:14:04 *** GEN Checkpoint space allocation failure
> 17:14:04 *** GEN Checkpoint space allocation failure
> 17:14:05 *** GEN Checkpoint space allocation failure
> 17:14:05 *** GEN Checkpoint space allocation failure
> MCR>
> 17:14:06 *** GEN Checkpoint space allocation failure
> ^R
> MCR>
> 17:14:06 *** GEN Checkpoint space allocation failure
> ^R
> MCR>
> 17:14:07 *** GEN Checkpoint space allocation failure
> ^R
> MCR>
> Simulation stopped, PC: 020432 (BR 20400)
> sim>
> So for which disk type are the images made?
That's not a complaint about the disk. "Partition" is an operating system
workspace in RSX-11 (to a first approximation).
Before you boot the tape image, do a SHOW CONF at the simh prompt, and let
us see everything the emulated 11 has active. I feel like it's going to be
a memory sizing issue.
Johnny Billquist
> Johnny Billquist<b...@softjar.se> writes:
>
>> Nice with pointers, and the recommendations are well meant. But for a
>> novice, these are hefty manuals already. Having to plow through several
>> of them might scare him off forever, Rich. :-)
>
> The thought I had is that the earlier manuals assume less on the part of
> the installer, and explain more for that reason. This gives the installer
> some background for the 11M manual.
>
> I speak from experience. I have yet to complete an 11M installation, and
> I've been using the derivative RSX-20F for decades (as you know ;-).
Read through the 11M SYSGEN manual, and become dazed. :-)
Actually, the SYSGEN manual is one of the few that are actually totally
different even between 11M and 11M+. So I would suspect it differs quite
some from 11A and so on as well. So while it's good for knowledge and
understanding, I suspect that nothing but the 11M SYSGEN manual will
actually do when it comes to actually doing a SYSGEN.
Johnny Billquist
Heh! You are totally right, Rich.
Look at the first line printed out: RSX-11M-PLUS ... 128 KW.
Way too small. M-PLUS requires atleast 256 KW.
Fix that, and things will improve right away.
Yes, a partition in RSX refers to a chunk of memory.
Johnny Billquist
> In article<ik1cs2$105$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-22 19:41, Holm Tiffe wrote:
>>> In article<ijudu1$q62$1...@iltempo.update.uu.se>,
>>> Johnny Billquist<b...@softjar.se> writes:
>>>> On 2011-02-21 12:53, Holm Tiffe wrote:
>>>>> Now I have to decrypt what SYSGEN want's to know from me :-~
>>>>
>>>> Ah... SYSGEN for 11M. That is a complex beast... :-)
>>>> However, the manual is available on bitsavers.
>>>>
>>>> Also, if you are running 11M+ instead of 11M, the sysgen is much more
>>>> user friendly.
>>>
>>> Where to get one?
>>
>> How about looking at ftp://www.trailing-edge.com/pub/rsx_dists/ ?
>>
>>> There is a RSX-11M-PLUS V2.1 Baseline (Disk image RD52) and a
>>> RSX-11M V3.2 for 11/45, 124KW, 50Hz (Disk image RK07) on the website
>>> http://pdp-11.org.ru/files.pl. Sould I thake one from them?
>>
>> I think not. See above... ;-)
>
> Ok, also I' restarting with this: ftp://www.trailing-edge.com/pub/rsx_dists/rsx11mpbl87.dsk.bz2 ?
Yeah, that will give you a "modern" M+ system.
>>> Installing from a tape was relatively easy, but I don't have an RD52 nor
>>> a RK07. But I have an RQDX3 and old 20MB disks....
>>> I don't know of this 11/45 OS will run on my 11/53.
>>>
>>> BTW: we have 50Hz here.
>>
>> The frequency is specified when you run a SYSGEN. If you get it wrong,
>> it only means that the clock in the OS will be running a bit fast, or
>> slow, depending on which way it is wrong. :-)
>
> Ok,ok, I'm an electronican and Unix Systems Admin...
:-)
>> If you want to install on a real 11/53, and only have a 20MB disk, you
>> are going to be very short on space, and you might not be able to
>> install 11M+ correctly at all...
>
> What you call real? No It isn't real, its a stack of PCBs in a H9278-A
> Backplane w/o the card cage mechanics, stabilized with some matchboxes between
> the cards and laying on my desk.. :-)
Well, that is real enough for me. I meant real as opposed to something
simulated.
>>> There is no documentation on the net for Sysadmins that know nothing
>>> about installing RSX11 right for the first time.
>>
>> Sure there is:
>> http://www.bitsavers.org/pdf/dec/pdp11/rsx11/RSX11M_V4.1_Apr83/
>> or
>> http://www.bitsavers.org/pdf/dec/pdp11/rsx11/RSX11Mplus_V4.x/
>
> Good, I'll get it and read something.
If you go down the M+ path instead, it is in some ways more complex,
since it is a much more capable system, but also simpler, since there
are less possibilities to adopt to extreme setups, and more things are
always on. And since there are more resources, it can be more generous
with help and information and user friendly interfaces.
>> will get you pretty recent versions of manuals. Be prepared, though.
>> These are big manuals, with lots of information. Not something your read
>> through in one hour. But there you will find whole manuals dedicated to
>> only SYSGEN, as well as manuals for system administration, program
>> development, just use the system, and whatnot.
>
> Hmm, I'm trying to setup RSX11 just for fun, and "to have had a look at it".
11M+ will probably be easier to do this with than 11M. 11M is more
cryptic, and have the capability to run on more different setups of
hardware, resources and features, which makes the setting up process
much more complex.
> Maybe later this is useful to get an old big eastern german PDP11-Clone running
> again.
[...]
Cool.
>>> Because of this something like this is a big problem:
>>>
>>> ;
>>>> ; Create Executive build files
>>>> ;
>>>> PIP RSXMC.MAC=SY0:RSXMC0.MAC/AP
>>>> ;
>>>> * 1. Do you wish to edit any of the Executive files? [Y/N]: N
>>>> ;
>>>
>>>
>>> .... yes, I've made a typo and wanted to edit the file TTDRVBLD.CMD to
>>> add a controller. But how the f***k can I edit this file?
>>
>> You do not want to edit the file to add a controller. At this point, you
>> have already passed by the point in the SYSGEN, where you tell which
>> controllers you have. Editing TTDRVBLD.CMD is many times not enough to
>> add additional controllers, so don't even try that unless you are an
>> expert on RSX systems. SYSGEN is your friend here, and you should use it.
>
> Ok, I wanted to change the number of DL Disks frome one to two exactly,
> I have an Dilog DQ614 here that emulates 2 RL disks.
Ah, yes... No, editing TTDRVBLD.CMD will definitely not help you there.
Not only is TTDRVBLD.CMD a file that only contains pieces for assembling
the terminal driver, and nothing else. The number of units is not
defined in that file anyway. It is a case of some complex data
structures that needs to be set up correctly in Macro-11 source files.
>> In your baseline system, you are using MCR, which is a very weird CLI
>> for "young" people. Once you have the system set up, you can have DCL
>> instead, which is way more user friendly.
>
> Hmm, I'm 48 years old, but no one asked me If I want an interface for
> old People :-)
Heh. :-)
By "young", I mean people who don't have experience from old DEC
systems. If you've used RT-11, OS/8, Tops-10 or something similar, it
would probably be a little easier to understand some of the finer points
(such as commands usually having file arguments as <dst>=<src> )
>> However, the setup process as such is that when you install the system
>> from tape, it restores a very basic, rudimentary RSX system for you. The
>> main purpose of this system is really just to allow you to run through a
>> SYSGEN, in order to compile a system which actually match what hardware
>> you have, and what options you want.
>
> Ok, maybe I've invested not enough time for this task.
SYSGEN is an important thing to get right. But you can do it over and
over again. But it pays to run on a simulator, since running it on real,
old, hardware, might take a few hours for each round.
Cool. Sounds like a good idea. And yes, that seems like plenty of space.
> In the card bin sits an UC08 to, (a KA630 and 18MB of memory too) and there
> is a 2nd card cage for experiments in the future..
> I have an RQDX3 too but used it only for the floppies jet.
Stay away from the RQDX3 if you can. It's slow as hell, even though it
works. And only supports old, small disks...
> If I try to install RSX-11M+ on simh then I have the problem to get the
> data (or disk contents) to the real pdp11 again, that's why I've installed
> this RSX11 from the tape (thats connected to an Dilog SQ703 wich emulates
> TMSCP). The Tape drive is a Tandberg TDC4222 that was handy, it seems that
> the hardware and at least the bootroms of the KDJ11-DS are happy with it.
> Tieed to connect an DLT40 yesterday evening instead but it doesn't worked.
> I think I have to setup it with the SQ703 first..
You could be tricky, and let simh use a whole disk on its own, and then
just move that whole disk over to the real pdp-11... ;-)
But if you can write to tape from simh, then you could generate the
whole system on simh, back it up to tape, move the tape over to the real
machine, and restore the system there.
Holm Tiffe
Yes! I've set the cpu to 11/53 1M and this time that looked better :-)
I have now this:
sim> show tq
TQ, TU81 (180MB), address=17774500-17774503, vector=260, 4 units
TQ0, attached to tapefile, write enabled, SIMH format
TQ1, not attached, write enabled, SIMH format
TQ2, not attached, write enabled, SIMH format
TQ3, not attached, write enabled, SIMH format
sim> show rq
RQ, address=17772150-17772153*, vector=154, 4 units
RQ0, 159MB, attached to rq0file, write enabled, RD54
RQ1, 1505MB, attached to rsx11mpbl87.dsk, write enabled, RA92
RQ2, 159MB, attached to rq2file, write enabled, RD54
RQ3, 409KB, not attached, write enabled, RX50
sim>
... and played with sysgen (w/o to read previously the suggested manuals,
sorry, there waqs no time jet).
I've understand what it would know from me and it had build a system,
it sayed in [1,54]
Ok, thats clear, but now it gets different ways:
At the end it prints out this:
>; When you are finished with SYSGEN, software boot in your target
>; system. Set the date and time, and save the system without any
>; switches to verify that the system is working. Then save the
>; system again, this time with the /WB switch to make it hardware
>; bootable.
>;
>; The optional SAVE switches are:
>;
>; /WB writes the boot block
>; /SFILE="filename" uses "filename" as the startup file
>; /MOU="mount-switches" specifies mount switches to be used
>; when the system volume is mounted
>;
>; For example:
>;
>; >BOOT [1,54]RSX11M.SYS
>; XDT: 24
>;
>; XDT>G
>; RSX-11M-PLUS V3.0 BL24
>;
>; >
>; TIM 19-APR-85 12:00:00
>; >SAV
>;
>; RSX-11M-PLUS V3.0 BL24 256.K System:"MJTOAD"
>; >RED DB:=SY:
>; >RED DB:=LB:
>; >RED DB:=SP:
>; >MOU DB:"RSX11MPBL24"
>; >@DB:[1,2]STARTUP
>; .
>; .
>; .
>; >* Please enter time and date (HH:MM MM/DD/YY) [S]: ^Z
>; >@ <EOF>
>; >SAV /WB
>;
>; RSX-11M-PLUS V3.0 BL24 256.K System:"MJTOAD"
>; >RED DB:=SY:
>; >RED DB:=LB:
>; >RED DB:=SP:
>; >MOU DB:"RSX11MPBL24"
>; >@DB:[1,2]STARTUP
>; .
>; .
>; .
>; >* Please enter time and date (HH:MM MM/DD/YY) [S]: 12:01 4/19/85
>; >TIME 12:01 4/19/85
>; >ACS SY:/BLKS=1024.
>; >CON ONLINE ALL
>; .
>; .
>; .
>; >@ <EOF>
>; >SET /UIC=[1,54]
>; >PIP [2,54]*.*;*/DE
>; >
>;
>; You may wish to edit [1,2]STARTUP.CMD and [1,2]QMGSTART.CMD
>; to reflect your desired initialization operations.
>;
>;
>!REM MACT0
>!REM PIPT0
>!REM LBRT0
>!REM TKBT0
>!REM VMRT0
>;
>; End of PREPGEN
>;
>TIME
17:20:21 24-FEB-11
>;
>ASN =
>;
>@ <EOF>
>
Hmm, now I've entered "boot [1,54]RSX11M.SYS" as suggested.
It comes up:
RSX-11M-PLUS V4.6 BL87 512.KW System:"RSXMPL"
>RED DU1:=SY:
>RED DU1:=LB:
>RED DU1:=SP:
>MOU DU1:"RSX11MPBL87"
>@DU1:[1,2]STARTUP
>; PLEASE NOTE
>;
>; If you have not yet read.....
I've gave the System an other name as "RSXMPL", named it ELEVEN.
but it is insisting to be calles RSXMPL.
Even there is no XDP...
I've added another disk yesterday evening, there was an RZ26L-E under
the discs that came from friends. I've added it to the UC07, formated and
checked it so there is another GB of space.
BTW: Ive got the DLT functioning yesterday too, but removed it agian.
The Dilog SQ703 is behaving other then it sould be (i think).
The DLT is only working if I allow SCSI Arbitation on the SQ703 (Default off)
and if I set this to on, I can't boot the old (selfmade) RSX11 tape anymore.
The same happens if I change the defaul SCSI to MSCP Device mapping of the
SQ703 (nothing special, default is only one drive SCSI = to TMSCP 0, I've
only tried to add SCSI 1 to MSCP 1). In both cases the boot isnt workingu
anymore...
What todo next?
I think I can write a system to a disk with simh and dump it to one of the
free disks on the 11/53 using 2.11 BSD.
The other way is creating a simh tape, convert it to a realy tape using my
program mentioned before and booting from that tape.
Both is needing your help...
Kind Regards,
Holm
Johnny Billquist
> In article<4D6623BF...@softjar.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> Heh! You are totally right, Rich.
>>
>> Look at the first line printed out: RSX-11M-PLUS ... 128 KW.
>> Way too small. M-PLUS requires atleast 256 KW.
>>
>> Fix that, and things will improve right away.
>>
>> Yes, a partition in RSX refers to a chunk of memory.
>>
>> Johnny
>>
>
> Yes! I've set the cpu to 11/53 1M and this time that looked better :-)
Excellent!
[...]
> ... and played with sysgen (w/o to read previously the suggested manuals,
> sorry, there waqs no time jet).
SYSGEN for 11M+ is much nicer and more user friendly. That one can be
run without reading a manual, in general.
> I've understand what it would know from me and it had build a system,
> it sayed in [1,54]
Good.
> Ok, thats clear, but now it gets different ways:
>
> At the end it prints out this:
[...]
> Hmm, now I've entered "boot [1,54]RSX11M.SYS" as suggested.
>
> It comes up:
[...]
Well, you seem to not have understood what a PREPGEN is. :-)
A PREPGEN is a fake SYSGEN. It's like a test run. You go through the
motions, but without actually doing any of the steps.
The "good" thing is that you can rerun SYSGEN, and reuse the answers you
gave during the PREPGEN. But make a proper SYSGEN, which actually
creates the files that you expect, and then report back.
> BTW: Ive got the DLT functioning yesterday too, but removed it agian.
> The Dilog SQ703 is behaving other then it sould be (i think).
>
> The DLT is only working if I allow SCSI Arbitation on the SQ703 (Default off)
> and if I set this to on, I can't boot the old (selfmade) RSX11 tape anymore.
> The same happens if I change the defaul SCSI to MSCP Device mapping of the
> SQ703 (nothing special, default is only one drive SCSI = to TMSCP 0, I've
> only tried to add SCSI 1 to MSCP 1). In both cases the boot isnt workingu
> anymore...
Unfortunately I don't know enough about the SQ703 to help with that part.
> What todo next?
>
> I think I can write a system to a disk with simh and dump it to one of the
> free disks on the 11/53 using 2.11 BSD.
Hmm. Why not let simh write directly to a raw disk, and then you'll have
your disk set up already?
But sure, using 2.11BSD as an intermediate step would also work.
> The other way is creating a simh tape, convert it to a realy tape using my
> program mentioned before and booting from that tape.
>
> Both is needing your help...
I think you are ready for the next step... ;-)
Holm Tiffe
OMG!
Will try..
>
> A PREPGEN is a fake SYSGEN. It's like a test run. You go through the
> motions, but without actually doing any of the steps.
>
> The "good" thing is that you can rerun SYSGEN, and reuse the answers you
> gave during the PREPGEN. But make a proper SYSGEN, which actually
> creates the files that you expect, and then report back.
>
>> BTW: Ive got the DLT functioning yesterday too, but removed it agian.
>> The Dilog SQ703 is behaving other then it sould be (i think).
>>
>> The DLT is only working if I allow SCSI Arbitation on the SQ703 (Default off)
>> and if I set this to on, I can't boot the old (selfmade) RSX11 tape anymore.
>> The same happens if I change the defaul SCSI to MSCP Device mapping of the
>> SQ703 (nothing special, default is only one drive SCSI = to TMSCP 0, I've
>> only tried to add SCSI 1 to MSCP 1). In both cases the boot isnt workingu
>> anymore...
>
> Unfortunately I don't know enough about the SQ703 to help with that part.
>
>> What todo next?
>>
>> I think I can write a system to a disk with simh and dump it to one of the
>> free disks on the 11/53 using 2.11 BSD.
>
> Hmm. Why not let simh write directly to a raw disk, and then you'll have
> your disk set up already?
That's not an entirely good idea in this concrete case, since the Emulex UC07
is splitting the 4GB Drive in two 2GB halves. There is somew offset that
I don't know exactly. If I make a mistake here I toast the 2.11BSD Partition.
Maybe I can try this with the new added RZ26, but how about the geometry
information in simh vs. pdp11?
> But sure, using 2.11BSD as an intermediate step would also work.
>
>> The other way is creating a simh tape, convert it to a realy tape using my
>> program mentioned before and booting from that tape.
>>
>> Both is needing your help...
>
> I think you are ready for the next step... ;-)
>
> Johnny
>
How can I write a bootable (simh-)tape out of the newly generated system?
Regards,
Holm
Holm Tiffe
Ok, it looks now as I have the generated system booting from rq1 in simh:
sim> boot rq1
RSX-11M-PLUS V4.6 BL87 512.KW System:"ELEVEN"
>RED DU1:=SY:
>RED DU1:=LB:
>RED DU1:=SP:
>MOU DU1:"RSX11MPBL87"
>@DU1:[1,2]STARTUP
>; PLEASE NOTE
>;
>; If you have not yet read the system release notes, please do so
>; now before attempting to perform a SYSGEN or to utilize the new
>; features of this system.
>;
Z[c
SET -- Inquire cannot determine terminal type
>;
>; Please ignore any random characters that may have printed on your
>; terminal just now. They came from a SET /INQUIRE=TI: command.
>; Evidently your terminal does not recognize escape sequences.
>; This will not affect the running of this command file.
>;
>* Please enter time and date (HH:MM DD-MMM-YYYY) [S]: 08:33 25-feb-2011
>TIME 08:33 25-feb-2011
>ACS SY:/BLKS=1024.
>CON ONLINE ALL
>ELI /LOG/LIM
>CLI /INIT=DCL/CTRLC/DPR="<15><12>/$ /"
>INS LB:[1,1]RMSRESAB.TSK/RON=YES/PAR=GEN
>INS LB:[1,1]RMSLBL.TSK/RON=YES/PAR=GEN
>INS LB:[1,1]RMSLBM.TSK/RON=YES/PAR=GEN
>INS $QMGCLI
>INS $QMGCLI/TASK=...PRI
>INS $QMGCLI/TASK=...SUB
>QUE /START:QMG
>INS $QMGPRT/TASK=PRT.../SLV=NO
>QUE LP0:/CR/NM
>START/ACCOUNTING
>CON ESTAT LP0:
>QUE BAP0:/BATCH
>QUE BAP0:/AS:BATCH
>@ <EOF>
>
Regards,
Holm
Johnny Billquist
> In article<ik6q4t$q9d$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-24 17:52, Holm Tiffe wrote:
>>> What todo next?
>>>
>>> I think I can write a system to a disk with simh and dump it to one of the
>>> free disks on the 11/53 using 2.11 BSD.
>>
>> Hmm. Why not let simh write directly to a raw disk, and then you'll have
>> your disk set up already?
>
>
> That's not an entirely good idea in this concrete case, since the Emulex UC07
> is splitting the 4GB Drive in two 2GB halves. There is somew offset that
> I don't know exactly. If I make a mistake here I toast the 2.11BSD Partition.
Ah. Good point.
> Maybe I can try this with the new added RZ26, but how about the geometry
> information in simh vs. pdp11?
SCSI disks don't really have a geometry. So that is not an issue.
This will "just work".
>> But sure, using 2.11BSD as an intermediate step would also work.
>>
>>> The other way is creating a simh tape, convert it to a realy tape using my
>>> program mentioned before and booting from that tape.
>>>
>>> Both is needing your help...
>>
>> I think you are ready for the next step... ;-)
>>
>> Johnny
>>
>
> How can I write a bootable (simh-)tape out of the newly generated system?
You don't do it exactly like that.
What you do is a backup of the whole disk. This will be a simple BRU backup.
To bring this image up on a new system, you boot the normal installation
tape. That will bring a system called BRUSYS up on your system. This
system can only do a few things, one of them being to run BRU. Then you
replace the boot tape with your backup tape, and you restore from that
tape to your disk, followed by booting the disk.
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-25 08:02, Holm Tiffe wrote:
>> In article<ik6q4t$q9d$1...@iltempo.update.uu.se>,
>> Johnny Billquist<b...@softjar.se> writes:
>>> On 2011-02-24 17:52, Holm Tiffe wrote:
>>>> What todo next?
>>>>
>>>> I think I can write a system to a disk with simh and dump it to one of the
>>>> free disks on the 11/53 using 2.11 BSD.
>>>
>>> Hmm. Why not let simh write directly to a raw disk, and then you'll have
>>> your disk set up already?
>>
>>
>> That's not an entirely good idea in this concrete case, since the Emulex UC07
>> is splitting the 4GB Drive in two 2GB halves. There is somew offset that
>> I don't know exactly. If I make a mistake here I toast the 2.11BSD Partition.
>
> Ah. Good point.
>
>> Maybe I can try this with the new added RZ26, but how about the geometry
>> information in simh vs. pdp11?
>
> SCSI disks don't really have a geometry. So that is not an issue.
> This will "just work".
Hmmm, yes I know this.
I meant it more in sight of the filesystem sizes on that disk.
When I dump a backup of a 159MB Drive to this 1GB Disk, I'll have a
159MB drive later...
>
>>> But sure, using 2.11BSD as an intermediate step would also work.
>>>
>>>> The other way is creating a simh tape, convert it to a realy tape using my
>>>> program mentioned before and booting from that tape.
>>>>
>>>> Both is needing your help...
>>>
>>> I think you are ready for the next step... ;-)
>>>
>>> Johnny
>>>
>>
>> How can I write a bootable (simh-)tape out of the newly generated system?
>
> You don't do it exactly like that.
> What you do is a backup of the whole disk. This will be a simple BRU backup.
> To bring this image up on a new system, you boot the normal installation
> tape. That will bring a system called BRUSYS up on your system. This
> system can only do a few things, one of them being to run BRU. Then you
> replace the boot tape with your backup tape, and you restore from that
> tape to your disk, followed by booting the disk.
>
> Johnny
>
Aha, ok.
In the moment I have some problems with the newly added rz26 drive.
I can format and verify it with the UC07 firmware bur when I try to access
it from the BSDs side I get this:
# dd if=/dev/rra2a of=/dev/null count=2
ra2 st=3 sb=0 fl=0 en=9
/dev/rra2a: Input/output error
This is MSCP unit=2, Status Register=3 and Opcode=9..
I think I have to read the uda50 manual now to find out wwat the heck
this means.
Regards,
Holm
Holm Tiffe
Currently I'm copying the diskimage from simh with dd from 2.11BSD to
an empty disk partition (DU2). I had to disable a kernel protection mechanism
in 2.11BSDs ufs_disksubr.c which should prevent disklables from being over-
written. I could only wirte 2 blocks and got then an error "read only
filesystem":
}
/*
* Check for write to write-protected label area. This does not include
* sector 0 which is the boot block.
*/
if (bp->b_blkno + pi->p_offset <= LABELSECTOR &&
bp->b_blkno + pi->p_offset + sz > LABELSECTOR &&
!(bp->b_flags & B_READ) && !(dk->dk_flags & DKF_WLABEL))
{
bp->b_error = EROFS;
goto bad;
}
return(1); /* success */
bad:
I've commented the if(){} block out and build a new kernel, this kernel is
happyly copying the data now.. lets see if it boots..
I can set the flag DKF_WLABEL only if I build my own dd program, I'm to lazy
to make that now...
Maybe I run in the next problem, since the data are now put to DU2, RSX11
is build for DU1...
Regards,
Holm
Holm Tiffe
ho...@unicorn.tsht.lan (Holm Tiffe) writes:
> ...There is something curios with the RZ26L, an other disk does well.
> Currently I'm copying the diskimage from simh with dd from 2.11BSD to
It worked. Almost.
Type a command then press the RETURN key: B DU2
DU2
RSX-11M-PLUS V4.6 BL87 768.KW System:"ELEVEN"
>RED DU2:=SY:
>RED DU2:=LB:
>RED DU2:=SP:
>MOU DU2:"RSX11MPBL87"
>@DU2:[1,2]STARTUP
>; PLEASE NOTE
>;
>; If you have not yet read the system release notes, please do so
>; now before attempting to perform a SYSGEN or to utilize the new
>; features of this system.
>;
>;
SET -- Inquire cannot determine terminal type
>;
>; Please ignore any random characters that may have printed on your
>; terminal just now. They came from a SET /INQUIRE=TI: command.
>; Evidently your terminal does not recognize escape sequences.
>; This will not affect the running of this command file.
>;
>* Please enter time and date (HH:MM DD-MMM-YYYY) [S]: 21:17 25-feb-2011
>TIME 21:17 25-feb-2011
>ACS SY:/BLKS=1024.
>CON ONLINE ALL
>ELI /LOG/LIM
>CLI /INIT=DCL/CTRLC/DPR="<15><12>/$ /"
>INS LB:[1,1]RMSRESAB.TSK/RON=YES/PAR=GEN
>INS LB:[1,1]RMSLBL.TSK/RON=YES/PAR=GEN
>INS LB:[1,1]RMSLBM.TSK/RON=YES/PAR=GEN
>INS $QMGCLI
>INS $QMGCLI/TASK=...PRI
>INS $QMGCLI/TASK=...SUB
>QUE /START:QMG
>INS $QMGPRT/TASK=PRT.../SLV=NO
>QUE LP0:/CR/NM
>START/ACCOUNTING
21:17:07 STAT0 -- 73. *ERROR* on scan file open
21:17:07 STAT0 -- 4. FCS I/O error code = 334
file: DU2:[???,???]SYSSCAN.TMP;2
>CON ESTAT LP0:
>QUE BAP0:/BATCH
>QUE BAP0:/AS:BATCH
>@ <EOF>
>
... what are that lines:
21:17:07 STAT0 -- 73. *ERROR* on scan file open
21:17:07 STAT0 -- 4. FCS I/O error code = 334
file: DU2:[???,???]SYSSCAN.TMP;2
???
This is now from the "real" 11/53, not simulated anymore
Regards,
Holm
Johnny Billquist
> In article<ik7rsk$51v$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-25 08:02, Holm Tiffe wrote:
>>> Maybe I can try this with the new added RZ26, but how about the geometry
>>> information in simh vs. pdp11?
>>
>> SCSI disks don't really have a geometry. So that is not an issue.
>> This will "just work".
>
> Hmmm, yes I know this.
> I meant it more in sight of the filesystem sizes on that disk.
> When I dump a backup of a 159MB Drive to this 1GB Disk, I'll have a
> 159MB drive later...
Yes.
But why would you create a 159MB file system on a 1GB disk?
And why would you dump the file system out to the disk? Create the file
system on that disk to start with, and use that from simh. Then move the
disk over to the PDP-11, and use the disk on that system. No need to
"dump" anything.
>>> How can I write a bootable (simh-)tape out of the newly generated system?
>>
>> You don't do it exactly like that.
>> What you do is a backup of the whole disk. This will be a simple BRU backup.
>> To bring this image up on a new system, you boot the normal installation
>> tape. That will bring a system called BRUSYS up on your system. This
>> system can only do a few things, one of them being to run BRU. Then you
>> replace the boot tape with your backup tape, and you restore from that
>> tape to your disk, followed by booting the disk.
>>
>> Johnny
>>
>
> Aha, ok.
>
> In the moment I have some problems with the newly added rz26 drive.
> I can format and verify it with the UC07 firmware bur when I try to access
> it from the BSDs side I get this:
>
> # dd if=/dev/rra2a of=/dev/null count=2
> ra2 st=3 sb=0 fl=0 en=9
> /dev/rra2a: Input/output error
>
> This is MSCP unit=2, Status Register=3 and Opcode=9..
> I think I have to read the uda50 manual now to find out wwat the heck
> this means.
I have no idea myself, unfortunately...
Johnny Billquist
> ...There is something curios with the RZ26L, an other disk does well.
> Currently I'm copying the diskimage from simh with dd from 2.11BSD to
> an empty disk partition (DU2). I had to disable a kernel protection mechanism
> in 2.11BSDs ufs_disksubr.c which should prevent disklables from being over-
> written. I could only wirte 2 blocks and got then an error "read only
> filesystem":
You do know that the label area protection can be overridden? No need to
patch the code.
I think it's the disklabel program that you use to write-enable that part.
> Maybe I run in the next problem, since the data are now put to DU2, RSX11
> is build for DU1...
Like I said before. With RSX, this is not a problem. The system is not
built for a system disk on a specific unit. It uses whatever unit it
actually is booted from.
Johnny Billquist
.err 334
000334 (-36): %I/O-F-IE.SQC, file ID, sequence number check
Looks like a corrupted file system for the system accounting file.
Which is probably because the system was taken down without properly
being shut down, and the SYSSCAN.TMP file is a temporary file which gets
a bit mixed up.
If you look at LB:[1,6]SYSSCAN.TMP, you'll probably find that it's not a
valid file any more.
Delete it, and you should be back in business.
I'll happily explain what the File ID, sequence number check means some
other time.
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-25 21:20, Holm Tiffe wrote:
>> In article<ik8m1o$b9t$1...@news.freibergnet.de>,
>> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>>> ...There is something curios with the RZ26L, an other disk does well.
>>> Currently I'm copying the diskimage from simh with dd from 2.11BSD to
[..]
>>
>> 21:17:07 STAT0 -- 4. FCS I/O error code = 334
>> file: DU2:[???,???]SYSSCAN.TMP;2
>>
>> ???
>
> .err 334
> 000334 (-36): %I/O-F-IE.SQC, file ID, sequence number check
>
> Looks like a corrupted file system for the system accounting file.
> Which is probably because the system was taken down without properly
> being shut down, and the SYSSCAN.TMP file is a temporary file which gets
> a bit mixed up.
> If you look at LB:[1,6]SYSSCAN.TMP, you'll probably find that it's not a
> valid file any more.
> Delete it, and you should be back in business.
How to look, how to delete?
Shurely I've not made a proper shutdow., don't kow how.
>
> I'll happily explain what the File ID, sequence number check means some
> other time.
>
> Johnny
>
Yes, I have now the firs part from the docu that you menntioned on the sceen,
it's starting with login and some play with DCL.
Nothing about deleting privileged files or creating accounts...
Regards,
Holm
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-25 12:21, Holm Tiffe wrote:
>> In article<ik7rsk$51v$1...@iltempo.update.uu.se>,
>> Johnny Billquist<b...@softjar.se> writes:
>>> On 2011-02-25 08:02, Holm Tiffe wrote:
>>>> Maybe I can try this with the new added RZ26, but how about the geometry
>>>> information in simh vs. pdp11?
>>>
>>> SCSI disks don't really have a geometry. So that is not an issue.
>>> This will "just work".
>>
>> Hmmm, yes I know this.
>> I meant it more in sight of the filesystem sizes on that disk.
>> When I dump a backup of a 159MB Drive to this 1GB Disk, I'll have a
>> 159MB drive later...
>
> Yes.
> But why would you create a 159MB file system on a 1GB disk?
Because this is the only complete filesystem I have.
> And why would you dump the file system out to the disk? Create the file
> system on that disk to start with, and use that from simh. Then move the
> disk over to the PDP-11, and use the disk on that system. No need to
> "dump" anything.
This may be true if you know what todo.
Currently I don't know how to set up a filesystem with the size of the disks
I have on the PDP (ok, thats is the easyer task). Next I must use the 159MB
Distribution to format that disk on simh and populate the disk with RSX-11Mplus.
I've already mentioned that I kown nothing about RSX11.
I've found manuals describing those processes for RT-11 and for XXDP and
had setup such discs in the past. Now I have to krawl trough mountains of
paper to find the needed infos, since I know almost nothing.
[..]
Regards,
Holm
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-25 17:40, Holm Tiffe wrote:
>> ...There is something curios with the RZ26L, an other disk does well.
>> Currently I'm copying the diskimage from simh with dd from 2.11BSD to
>> an empty disk partition (DU2). I had to disable a kernel protection mechanism
>> in 2.11BSDs ufs_disksubr.c which should prevent disklables from being over-
>> written. I could only wirte 2 blocks and got then an error "read only
>> filesystem":
>
> You do know that the label area protection can be overridden? No need to
> patch the code.
> I think it's the disklabel program that you use to write-enable that part.
Yes, and it is the only one where this mechanism ever gets used.
I'm not shure if this feature is more a misfeature or a bug.
Modern BSDs don't have this feature anymore. For what is it good for?
Your can scrap easily the entire disk, only the label persists...
If I would use this mechanism I had to write a program which handles the
label area as special case and have to set the flag to allow overwriting
of the label.
dd is a program to dump disks, and this program doesnt work for dumping disks
b'cause of this ugly label preserving feature...
Regards,
Holm
>
>> Maybe I run in the next problem, since the data are now put to DU2, RSX11
>> is build for DU1...
>
> Like I said before. With RSX, this is not a problem. The system is not
> built for a system disk on a specific unit. It uses whatever unit it
> actually is booted from.
>
> Johnny
>
hmm, yesterday eavening I've checked if sysgen works, sysgen couldnt find
the 3 save files generatad at the previous runs on simh...
Regards,
Holm
Johnny Billquist
> In article<ik9kou$235$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-25 21:20, Holm Tiffe wrote:
>>> In article<ik8m1o$b9t$1...@news.freibergnet.de>,
>>> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>>>> ...There is something curios with the RZ26L, an other disk does well.
>>>> Currently I'm copying the diskimage from simh with dd from 2.11BSD to
> [..]
>>>
>>> 21:17:07 STAT0 -- 73. *ERROR* on scan file open
>>> 21:17:07 STAT0 -- 4. FCS I/O error code = 334
>>> file: DU2:[???,???]SYSSCAN.TMP;2
>>>
>>> ???
>>
>> .err 334
>> 000334 (-36): %I/O-F-IE.SQC, file ID, sequence number check
>>
>> Looks like a corrupted file system for the system accounting file.
>> Which is probably because the system was taken down without properly
>> being shut down, and the SYSSCAN.TMP file is a temporary file which gets
>> a bit mixed up.
>> If you look at LB:[1,6]SYSSCAN.TMP, you'll probably find that it's not a
>> valid file any more.
>> Delete it, and you should be back in business.
>
> How to look, how to delete?
In MCR:
PIP <filename>/DE
In DCL:
DELETE <filename>
> Shurely I've not made a proper shutdow., don't kow how.
RUN SHUTUP
>> I'll happily explain what the File ID, sequence number check means some
>> other time.
>>
>> Johnny
>>
>
> Yes, I have now the firs part from the docu that you menntioned on the sceen,
> it's starting with login and some play with DCL.
>
> Nothing about deleting privileged files or creating accounts...
If it's the "Introduction to RSX" manual, it will go through how you
delete files, but not how you create accounts.
That's in some system manager manual. But basically, you RUN ACNT
Johnny Billquist
> In article<ik9kh2$228$2...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-25 17:40, Holm Tiffe wrote:
>>> ...There is something curios with the RZ26L, an other disk does well.
>>> Currently I'm copying the diskimage from simh with dd from 2.11BSD to
>>> an empty disk partition (DU2). I had to disable a kernel protection mechanism
>>> in 2.11BSDs ufs_disksubr.c which should prevent disklables from being over-
>>> written. I could only wirte 2 blocks and got then an error "read only
>>> filesystem":
>>
>> You do know that the label area protection can be overridden? No need to
>> patch the code.
>> I think it's the disklabel program that you use to write-enable that part.
>
> Yes, and it is the only one where this mechanism ever gets used.
> I'm not shure if this feature is more a misfeature or a bug.
I think you are misunderstanding it...
> Modern BSDs don't have this feature anymore. For what is it good for?
Yes, they do. I'm using NetBSD a lot, and the protection of the label
area is definitely there.
> Your can scrap easily the entire disk, only the label persists...
Yes. But scrapping the label area is pretty much as bad as scrapping the
whole disk, but much faster. So you want to avoid people doing that
accidentally. Because if you dd to a partition starting at 0, the disk
label is the first part that gets overwritten, and then you are pretty
much lost. No way of recovering the disk.
But note: the protection is not lifted by disklabel only for disklabel
use. Disklabel can turn the protection off, so that you *then* can dd to
the label area. It's a protection that disklabel can turn on and off for
you, and you can then write to that area from any program.
> If I would use this mechanism I had to write a program which handles the
> label area as special case and have to set the flag to allow overwriting
> of the label.
> dd is a program to dump disks, and this program doesnt work for dumping disks
> b'cause of this ugly label preserving feature...
No, you are incorrect. dd can work perfectly fine. You just need to turn
off the protection of the area before running dd, and disklabel dies
that for you.
>>> Maybe I run in the next problem, since the data are now put to DU2, RSX11
>>> is build for DU1...
>>
>> Like I said before. With RSX, this is not a problem. The system is not
>> built for a system disk on a specific unit. It uses whatever unit it
>> actually is booted from.
>>
>> Johnny
>>
>
> hmm, yesterday eavening I've checked if sysgen works, sysgen couldnt find
> the 3 save files generatad at the previous runs on simh...
Well, it could be that you didn't place the answer files where SYSGEN
think they would be.
As a short point to remember, though. You should not refer to your
system disk as DUn: when you do things. You should always refer to it as
LB:, which will always be the system disk, no matter what unit, type, or
place it actually is at.
Johnny Billquist
> In article<ik9kck$228$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-25 12:21, Holm Tiffe wrote:
>>> In article<ik7rsk$51v$1...@iltempo.update.uu.se>,
>>> Johnny Billquist<b...@softjar.se> writes:
>>>> On 2011-02-25 08:02, Holm Tiffe wrote:
>>>>> Maybe I can try this with the new added RZ26, but how about the geometry
>>>>> information in simh vs. pdp11?
>>>>
>>>> SCSI disks don't really have a geometry. So that is not an issue.
>>>> This will "just work".
>>>
>>> Hmmm, yes I know this.
>>> I meant it more in sight of the filesystem sizes on that disk.
>>> When I dump a backup of a 159MB Drive to this 1GB Disk, I'll have a
>>> 159MB drive later...
>>
>> Yes.
>> But why would you create a 159MB file system on a 1GB disk?
>
> Because this is the only complete filesystem I have.
Ah! Good point.
>> And why would you dump the file system out to the disk? Create the file
>> system on that disk to start with, and use that from simh. Then move the
>> disk over to the PDP-11, and use the disk on that system. No need to
>> "dump" anything.
>
> This may be true if you know what todo.
> Currently I don't know how to set up a filesystem with the size of the disks
> I have on the PDP (ok, thats is the easyer task). Next I must use the 159MB
> Distribution to format that disk on simh and populate the disk with RSX-11Mplus.
Right.
So, what you do in such a case is that you backup the disk down to tape,
and then restore it to the new disk, which is larger. And thus you'll
have your system on a larger disk.
> I've already mentioned that I kown nothing about RSX11.
I know... I hope I don't come across as arrogant or stupid. It's just
that sometimes I'm not sure what you do, or why. So I ask stupid
questions, rather than making assumptions.
> I've found manuals describing those processes for RT-11 and for XXDP and
> had setup such discs in the past. Now I have to krawl trough mountains of
> paper to find the needed infos, since I know almost nothing.
Yes. The RSX manual set is pretty big.
However, it is atleast pretty well organized, so you should be able to
find pretty much which manual you need to read based on just the titles.
Holm Tiffe
Already tried that, ->privilege violation.
Currently uic[200,200]
it seems that I have to type something like SET /UIC=[1,6]
before? Will try this..
>
>> Shurely I've not made a proper shutdow., don't kow how.
>
> RUN SHUTUP
>
>>> I'll happily explain what the File ID, sequence number check means some
>>> other time.
>>>
>>> Johnny
>>>
>>
>> Yes, I have now the firs part from the docu that you menntioned on the sceen,
>> it's starting with login and some play with DCL.
>>
>> Nothing about deleting privileged files or creating accounts...
>
> If it's the "Introduction to RSX" manual, it will go through how you
> delete files, but not how you create accounts.
> That's in some system manager manual. But basically, you RUN ACNT
>
> Johnny
>
Don't get me wrong, I know, that I know nothing and there is losts of
documentation to read.
It seems to be very difficult to get a RSX-11 up and running properly
under this circumstances. Want I'm missing here is a short step by step
advice how to get such a system properly setup, the same as the installing
2.11BSD, in one document.
Next how to create users, and so on.
The fact not to have proper distribution media and hardware fitting to them
makes it surely not easyer so If I sum up what I have I have "new" hardware
and don't know much about it, an a OS from that I where know the philosphy
nor the essential commands and features that it has.
I have a mass of documentation and must cross wise jumping trough it to
get all the neccessary info I need. That's not really so easy..
Thats why I'm asking for primitive things like "how to delete".
Ti this time I'm already tried to delete this file and got this privilege
violation.
What I currently want, is a halfways clean installation of RSX11 to play
with..
Reading over the "system generation and installation guide" currently...
Regards,
Holm
Holm Tiffe
ho...@unicorn.tsht.lan (Holm Tiffe) writes:
> In article <ikafth$adv$1...@iltempo.update.uu.se>,
> Johnny Billquist <b...@softjar.se> writes:
>> On 2011-02-26 08:01, Holm Tiffe wrote:
>>> In article<ik9kou$235$1...@iltempo.update.uu.se>,
>>> Johnny Billquist<b...@softjar.se> writes:
>>>> On 2011-02-25 21:20, Holm Tiffe wrote:
>>>>> In article<ik8m1o$b9t$1...@news.freibergnet.de>,
>>>>> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>>>>>> ...There is something curios with the RZ26L, an other disk does well.
>>>>>> Currently I'm copying the diskimage from simh with dd from 2.11BSD to
[..]
>SET UIC=[1,6]
>DELETE DU1:[1,6]SYSSCAN.TMP
PIP -- Version must be explicit or "*"
DU1:[1,6]SYSSCAN.TMP
>DELETE DU1:[1,6]SYSSCAN.TMP;*
PIP -- Failed to mark file for delete
DU1:[1,6]SYSSCAN.TMP;1 -- File ID, sequence number check
PIP -- Failed to mark file for delete
DU1:[1,6]SYSSCAN.TMP;2 -- File ID, sequence number check
>
Hmpf...
Regards,
Holm
Johnny Billquist
> In article<ikafth$adv$1...@iltempo.update.uu.se>,
Ah! Ok, if that is the problem, then yes. SET /UIC=[1,54] will solve it.
To make a long story short: SET /UIC=[x,y] is the equivalent to su under
Unix.
Some important things to know, and differences:
A UIC under RSX is the User Identification Code. It's a combination of
gid and uid under Unix.
You always give it in numeric form. No way of giving a name, and get the
corresponding UIC in commands, normally.
All UICs use octal numbers.
All UICs with a group number of 10 or less is "privileged". In this case
it means, that it matches the SYSTEM field in the file protection field.
You can also change the protection of files even if you don't have
access to them.
SET /UIC changes who you are. You can enter any UIC, even those for
which there is no user.
SET /UIC can only change who you are if you are sitting on a privileged
terminal (so, normal users cannot do this trick).
[1,54] is the canonical system administration UIC.
>>>> I'll happily explain what the File ID, sequence number check means some
>>>> other time.
>>>>
>>>> Johnny
>>>>
>>>
>>> Yes, I have now the firs part from the docu that you menntioned on the sceen,
>>> it's starting with login and some play with DCL.
>>>
>>> Nothing about deleting privileged files or creating accounts...
>>
>> If it's the "Introduction to RSX" manual, it will go through how you
>> delete files, but not how you create accounts.
>> That's in some system manager manual. But basically, you RUN ACNT
>>
>> Johnny
>>
>
> Don't get me wrong, I know, that I know nothing and there is losts of
> documentation to read.
It's fine. Keep reading, and keep asking. I'll happily help, as soon as
I have time, and understand what the problems are.
> It seems to be very difficult to get a RSX-11 up and running properly
> under this circumstances. Want I'm missing here is a short step by step
> advice how to get such a system properly setup, the same as the installing
> 2.11BSD, in one document.
> Next how to create users, and so on.
There is no one short document that describes this, no.
DEC expected whoever installed RSX would read through a couple of
manuals before doing this.
However, it is all documented. It's just that there is a lot to read
through.
You should start with the SYSGEN manual, and then the system managers
manuals (I think it's two volumes).
> The fact not to have proper distribution media and hardware fitting to them
> makes it surely not easyer so If I sum up what I have I have "new" hardware
> and don't know much about it, an a OS from that I where know the philosphy
> nor the essential commands and features that it has.
Indeed. It's not easy, and especially not when you jump in at the deep
end. :-)
> I have a mass of documentation and must cross wise jumping trough it to
> get all the neccessary info I need. That's not really so easy..
>
> Thats why I'm asking for primitive things like "how to delete".
> Ti this time I'm already tried to delete this file and got this privilege
> violation.
And now that I know what problem you had with it, I hope the answer I
provided was a little more helpful.
> What I currently want, is a halfways clean installation of RSX11 to play
> with..
>
>
> Reading over the "system generation and installation guide" currently...
That's a good manual to read.
Johnny Billquist
> [..]
>
>> SET UIC=[1,6]
>> DELETE DU1:[1,6]SYSSCAN.TMP
> PIP -- Version must be explicit or "*"
> DU1:[1,6]SYSSCAN.TMP
Expected. And you obviously didn't have a problem solving that one.
>> DELETE DU1:[1,6]SYSSCAN.TMP;*
> PIP -- Failed to mark file for delete
> DU1:[1,6]SYSSCAN.TMP;1 -- File ID, sequence number check
> PIP -- Failed to mark file for delete
> DU1:[1,6]SYSSCAN.TMP;2 -- File ID, sequence number check
>>
>
> Hmpf...
I didn't mention this as much, but it was very much to be expected. It's
the same error you got as when you tried starting the accounting. Since
the files are broken, they are just as much broken for this purpose.
However, even though you got those errors, please check if the directory
entries didn't disappear now?
What this error basically meant is that the files themselves had already
been deleted, but the directory entry pointing to them was still around.
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-26 10:39, Holm Tiffe wrote:
>> [..]
>>
>>> SET UIC=[1,6]
>>> DELETE DU1:[1,6]SYSSCAN.TMP
>> PIP -- Version must be explicit or "*"
>> DU1:[1,6]SYSSCAN.TMP
>
> Expected. And you obviously didn't have a problem solving that one.
>
>>> DELETE DU1:[1,6]SYSSCAN.TMP;*
>> PIP -- Failed to mark file for delete
>> DU1:[1,6]SYSSCAN.TMP;1 -- File ID, sequence number check
>> PIP -- Failed to mark file for delete
>> DU1:[1,6]SYSSCAN.TMP;2 -- File ID, sequence number check
>>>
>>
>> Hmpf...
>
> I didn't mention this as much, but it was very much to be expected. It's
> the same error you got as when you tried starting the accounting. Since
> the files are broken, they are just as much broken for this purpose.
>
> However, even though you got those errors, please check if the directory
> entries didn't disappear now?
No, they are still there: (dir [1,6])
ACNTRN.SYS;36 10. 25-FEB-2011 08:33
SYSSCAN.TMP;1
Read attributes error - File ID, sequence number check
SYSSCAN.TMP;2
Read attributes error - File ID, sequence number check
ACNTRN.SYS;37 10. 25-FEB-2011 21:17
>
> What this error basically meant is that the files themselves had already
> been deleted, but the directory entry pointing to them was still around.
>
> Johnny
>
Next one:
>set uic=[1,54]
>run $FMT
FMT>du2:/verify
** WARNING - Data will be lost on DU2: **
Continue? [Y OR N]: y
Start formatting
FMT -- Privilege violation
FMT>
Regards,
Holm
Holm Tiffe
.. additiony question:
Has the console SLU from the KDJ11-D/S (or any other SLUs) support for
modem control signals? I'm asking this, since I've running the console at
38400 and using kermit on my FreeBSD System to connect to the pdp11.
If I do an drag&drop to the kermit, the receiver in the pdp11 ovverruns.
I you say yes, I'll warm up my soldering iron and will look what to connect..
Regards,
Holm
Johnny Billquist
> In article<ikaimi$b95$2...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-26 10:39, Holm Tiffe wrote:
>>> [..]
>>>
>>> PIP -- Failed to mark file for delete
>>> DU1:[1,6]SYSSCAN.TMP;1 -- File ID, sequence number check
>>> PIP -- Failed to mark file for delete
>>> DU1:[1,6]SYSSCAN.TMP;2 -- File ID, sequence number check
>>>>
>>>
>>> Hmpf...
>>
>> I didn't mention this as much, but it was very much to be expected. It's
>> the same error you got as when you tried starting the accounting. Since
>> the files are broken, they are just as much broken for this purpose.
>>
>> However, even though you got those errors, please check if the directory
>> entries didn't disappear now?
>
> No, they are still there: (dir [1,6])
>
> ACNTRN.SYS;36 10. 25-FEB-2011 08:33
> SYSSCAN.TMP;1
> Read attributes error - File ID, sequence number check
> SYSSCAN.TMP;2
> Read attributes error - File ID, sequence number check
> ACNTRN.SYS;37 10. 25-FEB-2011 21:17
Hmpf. I thought that the directory entries would be deleted even though
you got that error, but okay. Let's be simple, and just remove the
directory entries then.
PIP LB:[1,6]SYSSCAN.TMP;*/RM
(this is the same as doing unlink in Unix)
>> What this error basically meant is that the files themselves had already
>> been deleted, but the directory entry pointing to them was still around.
>>
>> Johnny
>>
>
> Next one:
>
>> set uic=[1,54]
>> run $FMT
> FMT>du2:/verify
>
> ** WARNING - Data will be lost on DU2: **
>
> Continue? [Y OR N]: y
>
> Start formatting
> FMT -- Privilege violation
>
> FMT>
Huh? That was weird. Could you tell me what the result of a "DEV TI:" tells?
Johnny Billquist
> .. additiony question:
>
> Has the console SLU from the KDJ11-D/S (or any other SLUs) support for
> modem control signals? I'm asking this, since I've running the console at
> 38400 and using kermit on my FreeBSD System to connect to the pdp11.
> If I do an drag&drop to the kermit, the receiver in the pdp11 ovverruns.
> I you say yes, I'll warm up my soldering iron and will look what to connect..
The answer is yes, but at the same time no. So no, don't bring out the
soldering iron.
DEC *never* used modem control signals for flow control. (To do that is
actually to abuse the RS-232 standard.) DEC always use XON/XOFF flow
control.
Modem control signals are used for what they are intended. So you can
for instance notice if someone gets disconnected, and the RSX will log
out the terminal at that point.
Holm Tiffe
Ok, this has worked now, no complaints from the system anymore.
>>>
>>
>> Next one:
>>
>>> set uic=[1,54]
>>> run $FMT
>> FMT>du2:/verify
>>
>> ** WARNING - Data will be lost on DU2: **
>>
>> Continue? [Y OR N]: y
>>
>> Start formatting
>> FMT -- Privilege violation
>>
>> FMT>
>
> Huh? That was weird. Could you tell me what the result of a "DEV TI:" tells?
>
> Johnny
>
Hmm.. whats the difference from FTM and initialize?
Some playing:
LI = MCR BUF = 80. HFILL = 0
LINES = 24. TERM = unknown OWNER = none
NOPRINTER_PORT NOPASTHRU NOSERIAL
LOWER PRIV NOHOLD NOSLAVE NOESC CRT NOFORM NOREMOTE
ECHO NOVFILL HHT NOFDX WRAP NORPA NOEBC TYPEAHEAD
NOCTRLC AVO ANSI DEC NOEDIT NOREGIS NOSOFT NOBLKMOD
HSYNC BRO NOABAUD TTSYNC
>SET UIC=[1,54]
MCR -- Task not in system
>SE
MCR -- Illegal function
>
>DEV TI:
TT0: [200,200] [200,200]
CLI = MCR BUF = 80. HFILL = 0
LINES = 24. TERM = unknown OWNER = none
NOPRINTER_PORT NOPASTHRU NOSERIAL
LOWER PRIV NOHOLD NOSLAVE NOESC CRT NOFORM NOREMOTE
ECHO NOVFILL HHT NOFDX WRAP NORPA NOEBC TYPEAHEAD
NOCTRLC AVO ANSI DEC NOEDIT NOREGIS NOSOFT NOBLKMOD
HSYNC BRO NOABAUD TTSYNC
>SET UIC=[1,54]
>RUN $FMT
FMT>DU2:/VERIFY
** WARNING - Data will be lost on DU2: **
Continue? [Y OR N]: Y
Start formatting
FMT -- Privilege violation
FMT>^Z
>DEV TI:
TT0: [1,54] [1,54]
CLI = MCR BUF = 80. HFILL = 0
LINES = 24. TERM = unknown OWNER = none
NOPRINTER_PORT NOPASTHRU NOSERIAL
LOWER PRIV NOHOLD NOSLAVE NOESC CRT NOFORM NOREMOTE
ECHO NOVFILL HHT NOFDX WRAP NORPA NOEBC TYPEAHEAD
NOCTRLC AVO ANSI DEC NOEDIT NOREGIS NOSOFT NOBLKMOD
HSYNC BRO NOABAUD TTSYNC
>MOUNT DU2:
>RUN $FTM
INS -- File not found
>RUN$FTM
MCR -- Syntax error
>RUN $FMT
FMT>DU2:/VERIFY
FMT -- Privilege violation
FMT>^Z
>INITIALIZE DU2:
INI -- Volume mounted Files-11
>DISMOUNT DU2:
MCR -- Task not in system
>R DISMOUNT
MCR -- Illegal function
>DEV DU2:
DU2: Mounted Loaded Label= Type=RA81
>HELP DMO
DMO ddnn:[["]volume-label["]][/keyword(s)]
/DEV
/TERM=term:
/LOCK=option
DMO /USER [/keyword(s)]
/DEV
/TERM=term:
/LOCK=option
The DISMOUNT command requests the file system to mark a volume for dismount and
release its control blocks. The DISMOUNT /USER command dismounts all volumes
that you have mounted.
If you specify a volume-label, it is checked against the label on the volume
to ensure that the proper volume is being dismounted. Privileged users can
dismount any volume.
For more information on the DISMOUNT keywords, type HELP DISMOUNT keyword.
>DMO DU2:
DMO -- TT0: dismounted from DU2: *** Final dismount initiated ***
>
11:49:08 *** DU2: -- Dismount complete
>INITIALIZE DU2:
INI -- Device is not mounted foreign
>MOUNT DU2:/FOR
>RUN $FMT
FMT>DU2:
** WARNING - Data will be lost on DU2: **
Continue? [Y OR N]: Y
Start formatting
FMT -- Invalid Operation
FMT>^Z
>INITIALIZE DU2:
INI -- Device not allocated to this terminal - DU2:
>DMO DU2:
11:50:18 *** DU2: -- Dismount complete
DMO -- TT0: dismounted from DU2: *** Final dismount initiated ***
>ALLOCATE DU2:
>MOUNT DU2:/FOR
>RUN $FMT
FMT>DU2:/VERIFY
** WARNING - Data will be lost on DU2: **
Continue? [Y OR N]: Y
Start formatting
FMT -- Invalid Operation
FMT>^Z
>INITIALIZE DU2:
WARNING -- This volume is not backwards compatible
Searching for bad block descriptor file
INI -- Failed to read bad block file
>
...have something read about BAD before..
It seems that a volume to be formated has to be mounted in some way,
allocated exclusively to a terminal..?
... always interresting to how to get some command started .. :-)
Will try to find out how big DU2 currently is at next...
>
>dev du2:
DU2: TT0: - Private Mounted Loaded Foreign Type=RA81
>bad du2:
MCR -- Task not in system
>BAD DU2:
MCR -- Task not in system
>run BAD DU2:
MCR -- Syntax error
>ins BAD
INS -- File not found
>run $BAD
BAD>du2:
Regards,
Holm
Johnny Billquist
> In article<ikakr4$bv3$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-26 11:15, Holm Tiffe wrote:
>>> In article<ikaimi$b95$2...@iltempo.update.uu.se>,
>>> Johnny Billquist<b...@softjar.se> writes:
>>>> On 2011-02-26 10:39, Holm Tiffe wrote:
>>>>> [..]
>>>>>
>> you got that error, but okay. Let's be simple, and just remove the
>> directory entries then.
>>
>> PIP LB:[1,6]SYSSCAN.TMP;*/RM
>>
>> (this is the same as doing unlink in Unix)
>>
>>>> What this error basically meant is that the files themselves had already
>>>> been deleted, but the directory entry pointing to them was still around.
>>>>
>>>> Johnny
>
> Ok, this has worked now, no complaints from the system anymore.
Good. If that had failed, it would have been very interesting.
However, at some point you might want to run the RSX equivalent to fsck,
which is VFY.
>>>>
>>>
>>> Next one:
>>>
>>>> set uic=[1,54]
>>>> run $FMT
>>> FMT>du2:/verify
>>>
>>> ** WARNING - Data will be lost on DU2: **
>>>
>>> Continue? [Y OR N]: y
>>>
>>> Start formatting
>>> FMT -- Privilege violation
>>>
>>> FMT>
>>
>> Huh? That was weird. Could you tell me what the result of a "DEV TI:" tells?
>>
>> Johnny
>>
>
> Hmm.. whats the difference from FTM and initialize?
FMT is formatting the disk. Ah! Right now I realized what your problem
is as well.
Anyway... Formatting is the actions of physically formatting a disk. It
has nothing to do with file systems and similar stuff.
INI is the creation of a file system on a disk.
By the way, notice how you get errors when you do SET UIC=[...]. It
should be SET /UIC=[...]
(notice the slash)
>> RUN $FMT
> FMT>DU2:/VERIFY
>
> ** WARNING - Data will be lost on DU2: **
>
> Continue? [Y OR N]: Y
>
> Start formatting
> FMT -- Privilege violation
Your problem is actually that RSX do not allow you to format this kind
of disks.
>> MOUNT DU2:
Yes. BAD is the program used to scan for bad blocks on a disk. However,
since DU disks guarantee to be free of bad blocks, BAD is not really
neccesary. I think there is a switch to INI to tell it to ignore bad
block information, and just initialize the file system.
> It seems that a volume to be formated has to be mounted in some way,
> allocated exclusively to a terminal..?
>
> ... always interresting to how to get some command started .. :-)
> Will try to find out how big DU2 currently is at next...
Yes. To create a file system on a disk, the disk must be allocated
exclusively to you, and it must be mounted foreign.
Have to go away for a while. Write more later...
Holm Tiffe
That's what I meant, you don't have a real chance to get that RSX-11 installed
when you know nothing about the os and don't have some Jonny hanging around...
>
> Have to go away for a while. Write more later...
>
> Johnny
>
Ok, I too have other things todo, we would try if it's possibler to roast
some sausage jet down in the garden :-)) The Sun is shining...
Ok, initialize is for the filesystem, fmt for low level format and it is ok,
that the DU disks can't be low level formated with fmt. (there are the
programs to the controllers for ..) understand.
VFY>du1:
Consistency check of index and bitmap on DU1:
File ID 004461,000002 . Owner [11,24]
File is marked for delete
File ID 005772,000002 VTDRV.OBJ;1 Owner [11,24]
Bad file header
Index indicates 167060. blocks free, 144240. blocks used out of 311300.
Bitmap indicates 167058. blocks free, 144242. blocks used out of 311300.
VFY>
???
I have now initialized the DU2 drive and it seems, that it has the capacity
taht it should have:
terminal are deallocated.
>dea du2:
>mou du2:
>run $VFY
VFY>du2:/FR
DU2: has 2044659. blocks free, 52236. blocks used out of 2096895.
VFY>^Z
>
I want to "move" the entire sysstem to du2 now (and back later, since du1
is that 159MB disk file from simh, written from tape with 2.11BSD to the
du2 first, moved to du1, bootet from there, initialized du2..)
I want to reinstall the BSD on du0 and clean du1, since the disk they live on
has 4 Gbye and is splitted into halves from the uc07. There must be something
gone wrong there, since the halves have only a little bit more than 1 Gbyte
each. The F.R.D program in the UC07 is a little bit of "$%&(/%&/(" too,
so this could easysly happen....
I do have a bootable tape from the BSD and I have a backup from the root
and the user fs (with all the patches applied), so this shouldn't be a
problem besides of the time that job is consuming.
The question is how I can move the RSX-11Mplus to the freshly initialized DU2
and get a clean system there?
An other way where to create a "Distribution" on tape, but it think that's
rather complex. Bootblocks, BRU and the FS...
Regards,
Holm
Holm Tiffe
Ahh.. using that BRU System?
>
>> I've already mentioned that I kown nothing about RSX11.
>
> I know... I hope I don't come across as arrogant or stupid. It's just
> that sometimes I'm not sure what you do, or why. So I ask stupid
> questions, rather than making assumptions.
No, you don't come across like that.
I do notice well that you are investing a lot of time for me and that you are
the only one with so much effort to get things cleared up for some kind
of an idiot..
I can understand that's not easy to describe the simplest commands to get
the thing going. But it is annoying to if you want to follow things like
"simplytdelete that file and you are back in business" when that simple
deleting a file is kicking you in the ass... :-)
I'm a unix guy and I make my living with some IT and electronc services
for other people and firms. (repairing devices, programing micros, faiting
trojans and viruses, managing an small ISP and so on)
Mostly that are things easy to understand like Unix (* evil gin *).
Yes I know, for other People Unix Systems are not easy to understand, mostly
they are understandable for them at all..
Unix IS user friendly, it's only a little bit picky where his freends are..
Think that's the same here.
Regards,
Holm
Holm Tiffe
...in the meantime I've managed to make a backup tape from du1: with bru
and restored it tu du2:.
The system there is booting fine, a vfy of du2 looks ok, the size of
du2 is 2096895 blocks and ober 2020000 are free.
If I make a sysgen here withe the answer files generated before, the filesystem gets full or something like this (directroy entries?).
18:54:45 26-FEB-2011
>;
>MAC @RSXASM
MAC -- I/O error on work file
OU:[11,24]DRGIN,LS:[11,34]DRGIN/-SP=IN:[1,1]EXEMC/ML,[11,10]RSXMC/PA:1,DRGIN
MAC -- Open failure on output file
OU:[11,24]DRGLI,LS:[11,34]DRGLI/-SP=IN:[1,1]EXEMC/ML,[11,10]RSXMC/PA:1,DRGLI
MAC -- Open failure on output file
OU:[11,24]DRGPP,LS:[11,34]DRGPP/-SP=IN:[1,1]EXEMC/ML,[11,10]RSXMC/PA:1,DRGPP
MAC -- Open failure on output file
OU:[11,24]DRGSS,LS:[11,34]DRGSS/-SP=IN:[1,1]EXEMC/ML,[11,10]RSXMC/PA:1,DRGSS
MAC -- Open failure on output file
[..]
>;
>; Now we build the concatenated object module file for the Executive.
>;
>PIP RSX11M.OBS=*.OBJ
PIP -- Open failure on output file
DU2:[11,24]RSX11M.OBS -- Index file full
>PIP TTDRV.OBS=*.TTY
PIP -- No such file(s)
SY0:[11,24]*.TTY
[..]
..Index file full. grmbl...
What is happening here?
vfy after with some rant from the logger again:
VFY>du2:
Consistency check of index and bitmap on DU2:
19:09:09 SYSLOG -- 70. *ERROR* On scan file open
19:09:09 SYSLOG -- 4. FCS I/O error code = 347
file: DU2:[???,???]SYSSCAN.TMP;1
19:14:09 SYSLOG -- 70. *ERROR* On scan file open
19:14:09 SYSLOG -- 4. FCS I/O error code = 347
file: DU2:[???,???]SYSSCAN.TMP;1
Index indicates 1952070. blocks free, 144825. blocks used out of 2096895.
Bitmap indicates 1952070. blocks free, 144825. blocks used out of 2096895.
VFY>
Regards,
Holm
Johnny Billquist
> In article<ikas66$2mr6$1...@news.freibergnet.de>,
> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>> In article<ikan8h$ckn$1...@iltempo.update.uu.se>,
>> Johnny Billquist<b...@softjar.se> writes:
>>> Yes. To create a file system on a disk, the disk must be allocated
>>> exclusively to you, and it must be mounted foreign.
>>
>> That's what I meant, you don't have a real chance to get that RSX-11 installed
>> when you know nothing about the os and don't have some Jonny hanging around...
I head what you say. But I'm actually having a lot of fun answering
this. And hopefully in the end perhaps there will be a new RSX fanatic
as the result of it all... :-D
>>>
>>> Have to go away for a while. Write more later...
>>>
>>> Johnny
>>>
>> Ok, I too have other things todo, we would try if it's possibler to roast
>> some sausage jet down in the garden :-)) The Sun is shining...
>>
>> Ok, initialize is for the filesystem, fmt for low level format and it is ok,
>> that the DU disks can't be low level formated with fmt. (there are the
>> programs to the controllers for ..) understand.
Right.
The FMT progrma can do formatting for quite a few things, but DU disks
are not among those, with one exception. FMT can format DU: flopppy disks.
I'm not totally clear on why DEC didn't allow FMT to format normal DU
disks. Instead you either need to run diagnostics programs to format the
disks, or else with controllers like yours, there is a formatting tool
built in to the controller.
>> VFY>du1:
>>
>> Consistency check of index and bitmap on DU1:
>>
>> File ID 004461,000002 . Owner [11,24]
>> File is marked for delete
>> File ID 005772,000002 VTDRV.OBJ;1 Owner [11,24]
>> Bad file header
You can repair this. Read the documenttion for VFY in the right manual.
It is slightly more tricky than it first might seem, since you need to
have write access to [0,0]INDEXF.SYS, which normally is protected from
direct manipulation.
So you need to remount the disk with that file unlocked, which there is
a special switch to MOU for. But dismounting and remounting the system
disk is a bit special, since you need to still have the system access
the disk when you have dismounted it.
And then, of course, you need to run VFY to repair the file structure,
using various switches, and maybe run several times doing different
operations. Like I said - read the manual for this.
>> I have now initialized the DU2 drive and it seems, that it has the capacity
>> taht it should have:
>>
>> terminal are deallocated.
>>
>>> dea du2:
>>> mou du2:
>>> run $VFY
>> VFY>du2:/FR
>>
>> DU2: has 2044659. blocks free, 52236. blocks used out of 2096895.
>> VFY>^Z
>>>
>>
>> I want to "move" the entire sysstem to du2 now (and back later, since du1
>> is that 159MB disk file from simh, written from tape with 2.11BSD to the
>> du2 first, moved to du1, bootet from there, initialized du2..)
>>
>> I want to reinstall the BSD on du0 and clean du1, since the disk they live on
>> has 4 Gbye and is splitted into halves from the uc07. There must be something
>> gone wrong there, since the halves have only a little bit more than 1 Gbyte
>> each. The F.R.D program in the UC07 is a little bit of "$%&(/%&/(" too,
>> so this could easysly happen....
Ah. Ok. But things are atleast progressing. I'll leave the BSD part out
of this, even though I've done quite a lot of things inside there as well.
>> The question is how I can move the RSX-11Mplus to the freshly initialized DU2
>> and get a clean system there?
>>
>> An other way where to create a "Distribution" on tape, but it think that's
>> rather complex. Bootblocks, BRU and the FS...
>>
>> Regards,
>>
>> Holm
>
> ...in the meantime I've managed to make a backup tape from du1: with bru
> and restored it tu du2:.
> The system there is booting fine, a vfy of du2 looks ok, the size of
> du2 is 2096895 blocks and ober 2020000 are free.
>
> If I make a sysgen here withe the answer files generated before, the filesystem gets full or something like this (directroy entries?).
[...]
Same kind of problem you can see happen in Unix.
You've run out of file headers, which are the same things as inodes in Unix.
Do a PIP /FR to see some more information on this.
As you've figured out how to move the files over, what more you need to
know is that INI have a switch or two that you should be interested in.
Among them, the number of file headers to reserve room for on the disk.
You should probably set that up to 65000., which is about the max. You
probably have a very low number right now.
Johnny Billquist
> In article<ikagi5$aoe$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-26 08:07, Holm Tiffe wrote:
>>> In article<ik9kck$228$1...@iltempo.update.uu.se>,
>>> Johnny Billquist<b...@softjar.se> writes:
>>>> And why would you dump the file system out to the disk? Create the file
>>>> system on that disk to start with, and use that from simh. Then move the
>>>> disk over to the PDP-11, and use the disk on that system. No need to
>>>> "dump" anything.
>>>
>>> This may be true if you know what todo.
>>> Currently I don't know how to set up a filesystem with the size of the disks
>>> I have on the PDP (ok, thats is the easyer task). Next I must use the 159MB
>>> Distribution to format that disk on simh and populate the disk with RSX-11Mplus.
>>
>> Right.
>> So, what you do in such a case is that you backup the disk down to tape,
>> and then restore it to the new disk, which is larger. And thus you'll
>> have your system on a larger disk.
>
> Ahh.. using that BRU System?
Right. And I think you already have figured out pretty much all of this
procedure.
>>> I've already mentioned that I kown nothing about RSX11.
>>
>> I know... I hope I don't come across as arrogant or stupid. It's just
>> that sometimes I'm not sure what you do, or why. So I ask stupid
>> questions, rather than making assumptions.
>
> No, you don't come across like that.
Thanks. :-)
> I do notice well that you are investing a lot of time for me and that you are
> the only one with so much effort to get things cleared up for some kind
> of an idiot..
Believe me, I'm having loads of fun trying to explain the finer points
of RSX. I know that RSX is a very strange beast for people who are used
to Unix. It might be even more so if you ever start programming on it
seriously. But it's rewarding, and after a while you start thinking "if
only I could do the same things in Unix", :-)
> I can understand that's not easy to describe the simplest commands to get
> the thing going. But it is annoying to if you want to follow things like
> "simplytdelete that file and you are back in business" when that simple
> deleting a file is kicking you in the ass... :-)
>
> I'm a unix guy and I make my living with some IT and electronc services
> for other people and firms. (repairing devices, programing micros, faiting
> trojans and viruses, managing an small ISP and so on)
>
> Mostly that are things easy to understand like Unix (* evil gin *).
The fact that you know Unix is helpful, and I try to make comparisons
and parallels to Unix when I can to help explain things a bit quicker.
But let me know if something still is unclear.
> Yes I know, for other People Unix Systems are not easy to understand, mostly
> they are understandable for them at all..
>
> Unix IS user friendly, it's only a little bit picky where his freends are..
> Think that's the same here.
Nah. I think that Unix just plain sucks. :-)
But that is what I mostly get paid to work on. Wish I had more serious
RSX work on my table... ;-)
(I actually don't think Unix is that horrible, it's just that I think
RSX in some ways is much nicer, but Unix do some neat things too.)
Holm Tiffe
Johnny Billquist <b...@softjar.se> writes:
> On 2011-02-26 19:32, Holm Tiffe wrote:
>> In article<ikas66$2mr6$1...@news.freibergnet.de>,
>> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>>> In article<ikan8h$ckn$1...@iltempo.update.uu.se>,
>>> Johnny Billquist<b...@softjar.se> writes:
> [...]
>
>>>> Yes. To create a file system on a disk, the disk must be allocated
>>>> exclusively to you, and it must be mounted foreign.
>>>
>>> That's what I meant, you don't have a real chance to get that RSX-11 installed
>>> when you know nothing about the os and don't have some Jonny hanging around...
>
> I head what you say. But I'm actually having a lot of fun answering
> this. And hopefully in the end perhaps there will be a new RSX fanatic
> as the result of it all... :-D
Hmm, lets see, currently that is all looking very cryptic..
>
>>> VFY>du1:
>>>
>>> Consistency check of index and bitmap on DU1:
>>>
>>> File ID 004461,000002 . Owner [11,24]
>>> File is marked for delete
>>> File ID 005772,000002 VTDRV.OBJ;1 Owner [11,24]
>>> Bad file header
>
> You can repair this. Read the documenttion for VFY in the right manual.
> It is slightly more tricky than it first might seem, since you need to
> have write access to [0,0]INDEXF.SYS, which normally is protected from
> direct manipulation.
> So you need to remount the disk with that file unlocked, which there is
> a special switch to MOU for. But dismounting and remounting the system
> disk is a bit special, since you need to still have the system access
> the disk when you have dismounted it.
>
> And then, of course, you need to run VFY to repair the file structure,
> using various switches, and maybe run several times doing different
> operations. Like I said - read the manual for this.
Currently it was much more easyer to copy the contents from DU1 again.
[..]
>
> Same kind of problem you can see happen in Unix.
> You've run out of file headers, which are the same things as inodes in Unix.
>
> Do a PIP /FR to see some more information on this.
DU2: has 1952805. blocks free, 144090. blocks used out of 2096895.
Largest contiguous space = 1044621. blocks
105. file headers are free, 3203. headers used out of 3308.
105. that aren't looking as to much....
>
> As you've figured out how to move the files over, what more you need to
> know is that INI have a switch or two that you should be interested in.
> Among them, the number of file headers to reserve room for on the disk.
> You should probably set that up to 65000., which is about the max. You
> probably have a very low number right now.
Yes, Initialized the tdisk previously as RA90 (or so) with that 65000.
But if I reinstall the data with bru, it always asking me if I want to
initialize the output Disk. If I answer with no, that thing complains that
there are files11 on the disk and aborts. I can only reteive the data if I
answer with yes and then it makes a default with a low number of file headers
it seems...
Regards,
Holm
>
> Johnny
>
Johnny Billquist
> In article<ikc3et$qql$1...@iltempo.update.uu.se>,
> Johnny Billquist<b...@softjar.se> writes:
>> On 2011-02-26 19:32, Holm Tiffe wrote:
>>> In article<ikas66$2mr6$1...@news.freibergnet.de>,
>>> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>>>> In article<ikan8h$ckn$1...@iltempo.update.uu.se>,
>>>> Johnny Billquist<b...@softjar.se> writes:
>>>> VFY>du1:
>>>>
>>>> Consistency check of index and bitmap on DU1:
>>>>
>>>> File ID 004461,000002 . Owner [11,24]
>>>> File is marked for delete
>>>> File ID 005772,000002 VTDRV.OBJ;1 Owner [11,24]
>>>> Bad file header
>>
>> You can repair this. Read the documenttion for VFY in the right manual.
>> It is slightly more tricky than it first might seem, since you need to
>> have write access to [0,0]INDEXF.SYS, which normally is protected from
>> direct manipulation.
>> So you need to remount the disk with that file unlocked, which there is
>> a special switch to MOU for. But dismounting and remounting the system
>> disk is a bit special, since you need to still have the system access
>> the disk when you have dismounted it.
>>
>> And then, of course, you need to run VFY to repair the file structure,
>> using various switches, and maybe run several times doing different
>> operations. Like I said - read the manual for this.
>
> Currently it was much more easyer to copy the contents from DU1 again.
Also, it is not critical to fix these things immediately. You will not
get any bad effects by leaving them be.
The marked for delete file is just staying around even though it should
be deleted. You can delete it at your own convenience, at any time.
The bad file header issues just mean that you have directory entries
that point to non-existing files. Not a real problem either. Just
slightly annoying sometimes.
>> Same kind of problem you can see happen in Unix.
>> You've run out of file headers, which are the same things as inodes in Unix.
>>
>> Do a PIP /FR to see some more information on this.
>
> DU2: has 1952805. blocks free, 144090. blocks used out of 2096895.
> Largest contiguous space = 1044621. blocks
> 105. file headers are free, 3203. headers used out of 3308.
>
> 105. that aren't looking as to much....
Nope. Definitely not fun... :-)
>> As you've figured out how to move the files over, what more you need to
>> know is that INI have a switch or two that you should be interested in.
>> Among them, the number of file headers to reserve room for on the disk.
>> You should probably set that up to 65000., which is about the max. You
>> probably have a very low number right now.
>
> Yes, Initialized the tdisk previously as RA90 (or so) with that 65000.
> But if I reinstall the data with bru, it always asking me if I want to
> initialize the output Disk. If I answer with no, that thing complains that
> there are files11 on the disk and aborts. I can only reteive the data if I
> answer with yes and then it makes a default with a low number of file headers
> it seems...
I'll start with a hint: check what switches BRU actually have... ;-)
In short - why not let BRU initialize the output disk with a different
number of file headers that it defaults to... ;-)
Holm Tiffe
Ok.
>
>>> Same kind of problem you can see happen in Unix.
>>> You've run out of file headers, which are the same things as inodes in Unix.
>>>
>>> Do a PIP /FR to see some more information on this.
>>
>> DU2: has 1952805. blocks free, 144090. blocks used out of 2096895.
>> Largest contiguous space = 1044621. blocks
>> 105. file headers are free, 3203. headers used out of 3308.
>>
>> 105. that aren't looking as to much....
>
> Nope. Definitely not fun... :-)
>
>>> As you've figured out how to move the files over, what more you need to
>>> know is that INI have a switch or two that you should be interested in.
>>> Among them, the number of file headers to reserve room for on the disk.
>>> You should probably set that up to 65000., which is about the max. You
>>> probably have a very low number right now.
>>
>> Yes, Initialized the tdisk previously as RA90 (or so) with that 65000.
>> But if I reinstall the data with bru, it always asking me if I want to
>> initialize the output Disk. If I answer with no, that thing complains that
>> there are files11 on the disk and aborts. I can only reteive the data if I
>> answer with yes and then it makes a default with a low number of file headers
>> it seems...
>
> I'll start with a hint: check what switches BRU actually have... ;-)
>
> In short - why not let BRU initialize the output disk with a different
> number of file headers that it defaults to... ;-)
>
> Johnny
>
:-) I was faster... in the meantime this is looking like this:
DU2: has 1889762. blocks free, 207133. blocks used out of 2096895.
Largest contiguous space = 982412. blocks
62311. file headers are free, 3189. headers used out of 65500.
BRU accepts the same flags as initialize itself regarding max and headers,
...yes...
Regards,
Holm
Holm Tiffe
While I'm trying a sasgen now that actually fits the system a quaetion about
the flow control again. I've set the kermit to make flow control with xon/xoff,
how I have to set ti: to make the same? I've still getting overruns..
Regards,
Holm
Johnny Billquist
> In article<ikdgf5$2dtu$1...@news.freibergnet.de>,
Cool! Much better. I notice that the disk seem to be about 1G in size. I
guess that means you have still not reconfigured the controller partitions.
>> BRU accepts the same flags as initialize itself regarding max and headers,
>> ...yes...
Indeed. Which is helpful. :-)
> While I'm trying a sasgen now that actually fits the system a quaetion about
> the flow control again. I've set the kermit to make flow control with xon/xoff,
> how I have to set ti: to make the same? I've still getting overruns..
DEV TI: will show the settings for the terminal from RSX point of view.
TTSYNC is the attribute telling the system to respect flow control from
a terminal, in case the system is outputting data faster than the
terminal can handle.
HSYNC is for the system doing flow control to the terminal, in case the
terminal is outputting data faster than the system can handle.
Holm Tiffe
No. That is was is coming next. You see the data from du2: wich is the 2nd
physical disc (that is 2GB in size and split into two from uc07) ind the first
"slice" of the disk. hopefully rsx cann see the du2 now, lets see.
The disc with the problem is du0/du1, un du0 lives the bsd, and on du1 should
finally reside rsx11.
Since I now know how to dump and restore the rsx11 it schouldn't be a problem
now to install it again onm du1 where currently that 159MB filesystem from
the simh resides in the 1,2Gb Slice. the disk du0/du1 is normally a 4gb disk.
>
>>> BRU accepts the same flags as initialize itself regarding max and headers,
>>> ...yes...
>
> Indeed. Which is helpful. :-)
>
>> While I'm trying a sasgen now that actually fits the system a quaetion about
>> the flow control again. I've set the kermit to make flow control with xon/xoff,
>> how I have to set ti: to make the same? I've still getting overruns..
>
> DEV TI: will show the settings for the terminal from RSX point of view.
>
> TTSYNC is the attribute telling the system to respect flow control from
> a terminal, in case the system is outputting data faster than the
> terminal can handle.
> HSYNC is for the system doing flow control to the terminal, in case the
> terminal is outputting data faster than the system can handle.
>
> Johnny
>
Ok, I'll look at them.
You will hear from me..
Regards,
Holm
Holm Tiffe
...answering my self seems to be a new hobby from me...
What do you think, are there still people that are interested to get
QBUS Prototype Boards and 2nd that want "empty" boards to make ther
own QBUS Memeory Cards out of ol 9 Chip PC-SIMMs ?
I have a friend that has made some memory boards for an old eastern
germany computer, the P8000 wich is some clone from the old Zilog S8000
System running Wega some sort of System III. (have two of them).
We mailed about that last time, the question is how many boards are needed
and what people are willin to pay for them.
I htink it where some sort of fun for me to build som Hardware for the PDP11,
DA and AD converter, parallel Interfaces and soon.
Last saw the rpice for an 512KB QBUS Memory Board on Ebay.. something like
US$ 800 or so... whow..
Regards,
Holm
Holm Tiffe
Maybe here comes the next problem. The UC07 is now properly set up to
make 2 2GB DUs from the 4GB IBM.
Currently I've restoring from tape the BRU Backup that I wrote from the
working system on DU2 to the tape, using the:
RSX-11M/RSX-11M-PLUS Standalone Copy System V03
RSX-11M/RSX-11M-PLUS Standalone Configuration and Disk Sizing Program
(that was on the older RSX-11 I've got from the russian site.)
Valid switches are:
/CSR=nnnnnn to change the default device CSR
/VEC=nnn to change the default device vector
/FOR=n to change the default magtape formatter number
/DEV to list all default device CSR and vectors
Enter first device: mu:
Enter second device: du1:
Hit RETURN and enter date and time as 'TIM HH:MM MM/DD/YY'
>20:05 27-feb-20\02\11
MCR -- 1
>run bru
>
BRU>/rew/initialize/max:65500/headers:65495
From: mu0:
To: du1:
BRU - Starting Tape 1 on MU0:
BRU -- *WARNING* -- No bad block data file found
BRU -- *WARNING* -- Allocation for system file exceeds volume limit
...BRU is restoring now, but is the disk to big now or what does it mean?
...vfy brings many bad blocks..
Regards,
Holm
Johnny Billquist
> In article<ikdo9m$2m4m$1...@news.freibergnet.de>,
> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>> In article<ikdk2u$a91$1...@iltempo.update.uu.se>,
>> Johnny Billquist<b...@softjar.se> writes:
>>> On 2011-02-27 13:43, Holm Tiffe wrote:
>>>> In article<ikdgf5$2dtu$1...@news.freibergnet.de>,
>>>> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>>>> While I'm trying a sasgen now that actually fits the system a quaetion about
>>>> the flow control again. I've set the kermit to make flow control with xon/xoff,
>>>> how I have to set ti: to make the same? I've still getting overruns..
>>>
>>> DEV TI: will show the settings for the terminal from RSX point of view.
>>>
>>> TTSYNC is the attribute telling the system to respect flow control from
>>> a terminal, in case the system is outputting data faster than the
>>> terminal can handle.
>>> HSYNC is for the system doing flow control to the terminal, in case the
>>> terminal is outputting data faster than the system can handle.
>>>
>>> Johnny
>>>
>>
>> Ok, I'll look at them.
>>
>> You will hear from me..
>>
>> Regards,
>>
>> Holm
>
> ...answering my self seems to be a new hobby from me...
:-)
> What do you think, are there still people that are interested to get
> QBUS Prototype Boards and 2nd that want "empty" boards to make ther
> own QBUS Memeory Cards out of ol 9 Chip PC-SIMMs ?
There is probably some hobbyist users who might be interested, but I
doubt you'd make it work financially.
> I have a friend that has made some memory boards for an old eastern
> germany computer, the P8000 wich is some clone from the old Zilog S8000
> System running Wega some sort of System III. (have two of them).
>
> We mailed about that last time, the question is how many boards are needed
> and what people are willin to pay for them.
> I htink it where some sort of fun for me to build som Hardware for the PDP11,
> DA and AD converter, parallel Interfaces and soon.
> Last saw the rpice for an 512KB QBUS Memory Board on Ebay.. something like
> US$ 800 or so... whow..
Doubt anyone would pay that price, but what do I know?
Personally, I think the one piece most people would be interested in is
a disc controller for modern disks, which looks like MSCP disks (ie. DU:
type).
But making such a controller is not totally trivial.
Johnny Billquist
> In article<ikdo9m$2m4m$1...@news.freibergnet.de>,
> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>> In article<ikdk2u$a91$1...@iltempo.update.uu.se>,
>> Johnny Billquist<b...@softjar.se> writes:
>>> On 2011-02-27 13:43, Holm Tiffe wrote:
>>>> In article<ikdgf5$2dtu$1...@news.freibergnet.de>,
>>>> ho...@unicorn.tsht.lan (Holm Tiffe) writes:
>>>>> BRU accepts the same flags as initialize itself regarding max and headers,
>>>>> ...yes...
>>>
>>> Indeed. Which is helpful. :-)
>
> Maybe here comes the next problem. The UC07 is now properly set up to
> make 2 2GB DUs from the 4GB IBM.
Good.
> Currently I've restoring from tape the BRU Backup that I wrote from the
> working system on DU2 to the tape, using the:
>
> RSX-11M/RSX-11M-PLUS Standalone Copy System V03
>
>
> RSX-11M/RSX-11M-PLUS Standalone Configuration and Disk Sizing Program
>
> (that was on the older RSX-11 I've got from the russian site.)
>
> Valid switches are:
> /CSR=nnnnnn to change the default device CSR
> /VEC=nnn to change the default device vector
> /FOR=n to change the default magtape formatter number
> /DEV to list all default device CSR and vectors
>
>
> Enter first device: mu:
>
> Enter second device: du1:
>
> Hit RETURN and enter date and time as 'TIM HH:MM MM/DD/YY'
>
>> 20:05 27-feb-20\02\11
> MCR -- 1
That means the command failed with an error. That old version of BRUSYS
is actually not Y2K compliant. :-)
2 digit years only. And '11' is not an expected, normal, year. :-)
It's probably expecting something between 70 and 99.
>> run bru
>>
> BRU>/rew/initialize/max:65500/headers:65495
> From: mu0:
> To: du1:
> BRU - Starting Tape 1 on MU0:
>
> BRU -- *WARNING* -- No bad block data file found
That is ok, and can be overridden.
> BRU -- *WARNING* -- Allocation for system file exceeds volume limit
Hum? That is weird. On newer versions of BRUSYS, this is a fatal error,
and not just a warning.
It could be that the version you are using is old enough to have some
problems with very large disks...
As a work-around, you could boot the most recent version of BRUSYS from
your running M+ 4.6 system.
When in there, just type:
BOOT LB:[6,54]BRUSYS
and you'll have a "modern" BRUSYS to work from.
> ...BRU is restoring now, but is the disk to big now or what does it mean?
It should basically mean that BRU could not create one of the primary
systems files on the disk because it was too big to fit on the disk.
But that should only happy if you actually are restoring to a rather
small disk.
> ...vfy brings many bad blocks..
That also sounds bad. :-/
Bob Koehler
> PIP -- Failed to mark file for delete
> DU1:[1,6]SYSSCAN.TMP;1 -- File ID, sequence number check
> PIP -- Failed to mark file for delete
> DU1:[1,6]SYSSCAN.TMP;2 -- File ID, sequence number check
>>
Try using VFY (or was it VFYx?) to fix the file system.