Where to discuss BSD2.11 bugs / patches?

[Heinz-Bernd Eggenstein]

Where is the right place to discuss BSD2.11 bugs and fixes (independent from use under SIMH)?

("at" and "atq" are in need of some Y2K fixes, but maybe this is already known. Is there a bug tracker?).

HBE

[Warner Losh]

On Fri, Sep 17, 2021, 6:33 PM 'Heinz-Bernd Eggenstein' via [PiDP-11] <pid...@googlegroups.com> wrote:

Where is the right place to discuss BSD2.11 bugs and fixes (independent from use under SIMH)?

("at" and "atq" are in need of some Y2K fixes, but maybe this is already known. Is there a bug tracker?).

Make sure you have the latest...

Don't think that there is one.

Warner 

HBE

--
You received this message because you are subscribed to the Google Groups "[PiDP-11]" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-11/eaeebfa8-be22-4da5-bcc4-89f1bb004d3fn%40googlegroups.com.

[Johnny Billquist]

No bug tracker exists, and we don't really have some official list or
anything.

But you can mail Steven or me if you have some patch. You're also
welcome to just report things you've observed, and I'll try to keep
track of it.

So far it's been low enough volume, and simple enough that it hasn't
really been any need for something more advanced.

Johnny

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com

> <mailto:pidp-11+u...@googlegroups.com>.

> To view this discussion on the web visit
> https://groups.google.com/d/msgid/pidp-11/eaeebfa8-be22-4da5-bcc4-89f1bb004d3fn%40googlegroups.com

> <https://groups.google.com/d/msgid/pidp-11/eaeebfa8-be22-4da5-bcc4-89f1bb004d3fn%40googlegroups.com?utm_medium=email&utm_source=footer>.

--
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

[Heinz-Bernd Eggenstein]

Ah ok. Well it's a bit unexpected to still find Y2K bugs, but I have a system at patch level 469 and "at" is not touched in 470 and 471 so I guess it's a current problem.

Simply try

at 2am somefile

followed by atq

Execution time is supposed to be encoded in the spool-file name in /usr/spool/at with a leading 2-digit yy field,  but will actually be written with a three digit 121 currently (2021 -1900)  which confuses e.g. atq which expects the yy.ddd.hhmm.* filenames.

I briefly looked into the code and there are several places that need fixing, something like what was done for 4.3BSD-Quasijarus (using 4 digit yyyy names for the spool files).

Cheers

HB

[Johnny Billquist]

Thanks.

Yes, there are probably still a whole bunch of Y2K issues around.

I'll poke at at and friend.

Johnny

> <https://groups.google.com/d/msgid/pidp-11/eaeebfa8-be22-4da5-bcc4-89f1bb004d3fn%40googlegroups.com?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/pidp-11/eaeebfa8-be22-4da5-bcc4-89f1bb004d3fn%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
>
> --
> 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
>

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com
> <mailto:pidp-11+u...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/pidp-11/ba0028e4-3d33-4e06-a62a-7d2d3915da7cn%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/ba0028e4-3d33-4e06-a62a-7d2d3915da7cn%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Johnny Billquist]

Yuck. There are actually several bugs in there. I think I got them all
fixed now. I guess this will be scheduled for the eventual patch 472...

Johnny

[Heinz-Bernd Eggenstein]

Excellent, thanks a ton!

> actually several bugs in there

From a quick look at the diffs with a 4.3BSD version I noticed problems with "long" (> 8 or 9 character" user names and leap year calculation ..at least.

Looking forward to 472

Thanks again

HBE

[jay.logue]

On a related topic, I looked for distribution tape images containing the more recent patch levels and couldn't find any.   So I created my own (although I need to update it to 471, and 472 as soon as that's available).

Any interest in putting these online somewhere?  Are there versions somewhere that I just didn't find?

--Jay

[Johnny Billquist]

I've been considering cutting a newer dist, but haven't come around to
it. Others have asked, but I don't think I've seen anyone actually put
something up. I bet there are people who would appreciate it.

Johnny

> <https://groups.google.com/d/msgid/pidp-11/ba0028e4-3d33-4e06-a62a-7d2d3915da7cn%40googlegroups.com?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/pidp-11/ba0028e4-3d33-4e06-a62a-7d2d3915da7cn%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
> >>
> >
>
> --
> 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
>

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com
> <mailto:pidp-11+u...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/pidp-11/a27872c5-2791-4315-af50-d95259c3ef92n%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/a27872c5-2791-4315-af50-d95259c3ef92n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[jay.logue]

Cool.  I'd love to get some eyes on my process for preping the tape image, to make sure doing everything right.  I'll start another thread on the topic sometime this weekend.

--Jay

[Warner Losh]

My 2.11BSD pl 0 project prepares tapes for pl 0 automatically using newer host to bootstrap.

I'm sure they could be adopted to go the other way (install a newer system from the last

tape and then create a tape there). Older 2.11BSDs are a bit weirder than the 46x or 47x

ones, so a fair amount of hacking on the expect scripts might be needed.

https://github.com/bsdimp/mk211bsd

https://www.tuhs.org/Archive/Distributions/UCB/2.11BSD_patch457/

is the most up-to-date tape I can find. Maybe I'll take some time and try

to hack these scripts to do that update...

Warner

To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-11/589177a0-662e-49b9-909f-c75bf20e8029n%40googlegroups.com.

[Heinz-Bernd Eggenstein]

> Are there versions somewhere that I just didn't find?

The most up-to-date tape that I'm aware of is this one: https://github.com/chasecovello/211bsd-pidp11 which includes patches up to and including 469.

It's intended for use with the PiDP-11 and aside from regular 2.11BSD It also contains the "bonus stuff" from the image that is used if you follow Oscar's instructions :

- the code for the weather station "hack"

- an HTTP server that will be up and running by default (!!!!) when you use this tape.

Rather than create yet another tape image, maybe you want to contact Chase Covello and nudge him towards going two patchlevels up. Or maybe you want to fork his archive and do that and then he can merge it back if/when he likes.

Cheers

HB

[jay.logue]

Thanks for the pointer.  That looks to be a prebuilt disk image, rather than a distribution tape image.  Useful for my PiDP-11, but not quite what I'm looking for.

Lately I've been reliving the experience of installing 2.11BSD onto a bare system, both in simh and on my ragtag 11/73 system (equipped with a UC07 and an HP 97548S).  Hence my desire for an as-close-to-original-as-possible dist tape image, but with bug fixes.

--Jay

[jay.logue]

>My 2.11BSD pl 0 project prepares tapes for pl 0 automatically using newer host to bootstrap.

> I'm sure they could be adopted to go the other way (install a newer system from the last

> tape and then create a tape there).

That makes a lot of sense.  I'm thinking it wouldn't be hard to set up a github repo with an automated build that released a new dist tape and preinstalled disk image whenever a new patch dropped.

--Jay

[Warner Losh]

On Sat, Sep 18, 2021 at 11:23 AM jay.logue <jay....@gmail.com> wrote:

>My 2.11BSD pl 0 project prepares tapes for pl 0 automatically using newer host to bootstrap.

> I'm sure they could be adopted to go the other way (install a newer system from the last

> tape and then create a tape there).

That makes a lot of sense.  I'm thinking it wouldn't be hard to set up a github repo with an automated build that released a new dist tape and preinstalled disk image whenever a new patch dropped.

I need to learn how to make CI pipeline for github/gitlab repos... Maybe I'll start there with this project.

Warner

 

To unsubscribe from this group and stop receiving emails from it, send an email to pidp-11+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-11/ae75f376-fd0b-4b14-acbf-12c0d0f60549n%40googlegroups.com.

[Johnny Billquist]

There is a script in the system that is used to create the dist tapes.

cd /usr/src/sys/pdpstand
make distribution

Johnny

> <https://groups.google.com/d/msgid/pidp-11/a27872c5-2791-4315-af50-d95259c3ef92n%40googlegroups.com?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/pidp-11/a27872c5-2791-4315-af50-d95259c3ef92n%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
>
> --
> 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
>

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com
> <mailto:pidp-11+u...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/pidp-11/589177a0-662e-49b9-909f-c75bf20e8029n%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/589177a0-662e-49b9-909f-c75bf20e8029n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Johnny Billquist]

Well, a Makefile isn't a script, but I hope you get it.
Also, there are more targets:

tmscptape is there in case you have large tape capacity, and the whole
thing fits on a single tape.

Johnny

[Heinz-Bernd Eggenstein]

As for "bugs" in general: I guess there are quite a few programs in 2.11BSD that fall into the category:

"Why, oh why, is that code using 16 bit ints [here] and [there] instead of a 32 bit long :-( ??? "

My favorite example is "head" : 

   head -40000 some_long_file

will silently decide to output exactly zero lines, because the line count is parsed and processed as a (signed)  int ==> overflow to a negative number.

of course

  head -70000 some_long_file

will silently decide to output exactly 4464 lines ... you get the drift.

But is this a "bug" that should be corrected in the distribution????

OTOH, the source code is there for everyone who cares to roll their own enhanced version of head, put it in ~/bin and put that in the beginning of your $PATH.

Hmmmm...I honestly don't know. But there are cases where 32767 is really not expected to be a limit, maybe not even in 1975.

Cheers

HB

[Johnny Billquist]

Ah, yes. The limitations of 16-bit ints... That probably bites in a lot
of places. Although, if someone really wants to see more than the first
32000 lines of a file, I think they need a sanity check.

And it's not really that changing to a 32-bit integer removes the "bug".
It just moves the posts a little.

So no, I would probably argue that in this case, it's not really a bug,
but a limitation. And a reasonable one.

Johnny

> <https://groups.google.com/d/msgid/pidp-11/589177a0-662e-49b9-909f-c75bf20e8029n%40googlegroups.com?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/pidp-11/589177a0-662e-49b9-909f-c75bf20e8029n%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
> >>
> >
>
> --
> 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
>

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com
> <mailto:pidp-11+u...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/pidp-11/b0f33938-1e4d-4989-84f2-5adf8c0cc563n%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/b0f33938-1e4d-4989-84f2-5adf8c0cc563n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Heinz-Bernd Eggenstein]

> Although, if someone really wants to see more than the first
> 32000 lines of a file, I think they need a sanity check.

Ok, then I'll need one :-) No seriously, the output of head isn't always meant for human consumption, .e.g. I use it all the time to generate smaller testsets for stuff that crunches text files (small enough to finish in reasonable time, but large enough so the time-ing gives representative throughput estimates. ... that's how I noticed this one).

It should perhaps at least complain or mention the limitation in the man page.

Cheers

HB

[Johnny Billquist]

I could definitely agree that the limitation should be documented, or
even detected when hit at runtime.

On the other hand, then we should go around and do the same for a lot of
programs. At least I am not to keen on such work.

tail, for example, do use longs, but as observed, that just means the
posts are a little further out, but they are still there. So we should
document there as well, and everywhere else where integers are used as
arguments.

If you run some crunching on a real PDP-11, then 32000 lines will
already take quite a while... :-)

(Simulators just makes things miserably fast... :-) )

Johnny

> <https://groups.google.com/d/msgid/pidp-11/b0f33938-1e4d-4989-84f2-5adf8c0cc563n%40googlegroups.com?utm_medium=email&utm_source=footer

> <https://groups.google.com/d/msgid/pidp-11/b0f33938-1e4d-4989-84f2-5adf8c0cc563n%40googlegroups.com?utm_medium=email&utm_source=footer>>.
>
>
> --
> 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
>

> --
> You received this message because you are subscribed to the Google
> Groups "[PiDP-11]" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to pidp-11+u...@googlegroups.com
> <mailto:pidp-11+u...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/pidp-11/57ecaefc-07d2-494e-89c0-d2fa42c05f31n%40googlegroups.com
> <https://groups.google.com/d/msgid/pidp-11/57ecaefc-07d2-494e-89c0-d2fa42c05f31n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Heinz-Bernd Eggenstein]

> (Simulators just makes things miserably fast... :-) )

Valid point. However I use simh's THROTTLE to slow the sim down so that at least the available CPU benchmarks are in the right ballpark, and in that configuration I get about 50kB/s  I/O  throughput (disk and network) for code that processes stuff character-by-character like "wc" , with ca 80% of the time spent in "user" and 20% in "system" code according to "time", so maybe the underlying peak I/O throughput is more like 200kB/s. That still feels like the (say) 19870s or 80s (?) to me, but if there are ways to make the I/O throughput even more realistic compared to a real PDP-11, I'd be happy to do that. (I use the standard boot.ini for 2.11BSD in simh which as far as I understand it simulates a rather "modern" RQDX3 MSCP controller with a single  RA82, w/ fake 1GB capacity for simplicity).

Cheers

HB