[TUHS] /dev/drum

[Johnny Billquist]

On 2018-04-30 17:05, Noel Chiappa wrote:
>      > From: Johnny Billquist
> 
>      > This is where I disagree. The problem is that the chunks in the PDP-11
>      > do not describe things from a zero offset, while a segment do. Only
>      > chunk 0 is describing addresses from a 0 offset. And exactly which chunk
>      > is selected is based on the virtual address, and nothing else.
> 
> Well, you have something of a point, but... it depends on how you look at it.
> If you think of a PDP-11 address as holding two concatenated fields (3 bits of
> 'segment' and 13 bits of 'offset within segment'), not so much.

Err... That would be a pretty ugly way to look at the world. Not to 
mention that one segment silently slides over into the next one if it's 
more than 8K. No, I think I prefer to not try to look at the world that 
way. :-)

> IIRC there are other segmented machines that do things this way - I can't
> recall the details of any off the top of my head. (Well, there the KA10/KI10,
> with their writeable/write-protected 'chunks', but that's a bit of a
> degenerate case. I'm sure there is some segmented machine that works that way,
> but I can't recall it.)

The KA/KI without the Twenex pager pretty much only had two segments, 
unless I remember wrong. It was the low segment and the high segment. 
Well, ok, that also implied that it was based on the virtual address, 
but these two segments were definitely treated as two separate spaces, 
that were never concatenated.

> BTW, this reminds me of another key differentiator between paging and
> segments, which is that paging was originally _invisible_ to the user (except
> for setting the total size of the process), whereas segmentation is explicitly
> visible to the user.

Good point.
Well, wouldn't you say that the "chunks" on a PDP-11 are invisible to 
the user? Unless you are the kernel of course. Or run without protection.

> I think there is at least one PDP-11 OS which makes the 'chunks' visible to
> the user - MERT (and speaking of which, I need to get back on my project of
> trying to track down source/documentation for it).

Don't know that system. But you could just argue that RT-11 also makes 
it all visible as well. Any OS which don't give you proper protection 
will allow you to play with all hardware, no matter how it appears.

>      > Demand paging really is a separate thing from virtual memory. It's a
>      > very bad thing to try and conflate the two.
> 
> Really? I always worked on the basis that the two terms were synonyms - but
> I'm very open to the possibility that there is a use to having them have
> distinct meanings.

*Demand* paging is definitely a separate concept from virtual memory.

OSes like RSX, RSTS/E, or Unix, on PDP-11s do give you virtual memory. 
However, none of them implement demand paging. The PDP-11 can, as you 
noted, also support demand paging, but there is little need for such an 
advanced mechanism on that machine, for several reasons.

When I first read about demand paging, it was in the context of VMS, and 
the simplicitly, elegance, and efficiency of it really impressed me.

Starting execution of an image in VMS really just meant that the OS read 
in the executable image meta information, which gave things like what 
memory addressed were defined. VMS created the page table, and marked 
every page as invalid, and then it started executing in the process. The 
first thing that would happen would of course be that you got a page 
fault by trying to fetch the instruction at the first address of the 
program. VMS immediately suspends execution until it have read in that 
page from the binary, marks that page as now having a valid translation, 
and then resumes execution. Of course, most likely you'll have a whole 
storm of page faults almost immediately after a process starts 
executing, but after a while, you'll be in a sort of balance where you 
mostly run without any more page faults.
(I would assume Unix does it the same way on modern machines.)

Very different than in RSX, for example, where the OS first makes sure 
all required virtual memory have valid translations before the process 
is allowed to execute.

> I see a definition of 'virtual memory' below, but what would you use for
> 'paging'?

"Paging" is a bit of an unclear term for me here. It might be referring 
to the reading in of a page at a page fault, which would be a demand 
paging situation, the paging out of individual pages as a demand paged 
system requires pages to fulfill needs of a different process, or it 
might mean the reading in or writing out of all pages of a process, in 
which case it would be close to a synonym with "swapping". So, in which 
way do you mean "paging"?

> Now that I think about it, there are actually _three_ concepts: 'virtual
> memory', as you define it; what I will call 'non-residence' - i.e. a process
> can run without _all_ of its virtual memory being present in physical memory;
> and 'paging' - which I would define as 'use fixed-size blocks'. (The third is
> more of an engineering thing, rather than high-level architecture, since it
> means you never have to 'shuffle' core, as systems that used variable-sized
> things seem to.)
> 
> 'Non-residence' is actually orthogonal to 'paging'; I can imagine a paging
> system which didn't support non-residence, and vice versa (either swapping
> the entire virtual address space, or doing it a segment at a time if the
> system has segments).

Yeah. And non-residence, as you call it, is what demand paging is.
Pages are brought into physical memory on demand.

The definition you use for paging here is an interesting point. Not sure 
how much of a thing it would be, but it might be a thing sometimes as 
well. But then we get into the question I asked further down, how do you 
then deal with when you have different sized pages on a system?
But "paging" just meaning "fixed size pages" is certainly a distinction 
you can make. I think I wouldn't find much use for it, but that don't 
mean others might not, or that it wouldn't be a possible classification.

>      > each process would have to be aware of all the other processes that use
>      > memory, and make sure that no two processes try to use the same memory,
>      > or chaos ensues.
> 
> There's also the System 360 approach, where processes share a single address
> space (physical memory - no virtual memory on them!), but it uses protection
> keys on memory 'chunks' (not sure of the correct IBM term) to ensure that one
> process can't tromp on another's memory.

Oh. There is no real connection between virtual memory and memory 
protection. One can exist with or without the other.

>      >> a memory management device for the PDP-11 which provided 'real' paging,
>      >> the KT11-B?
> 
>      > have never read any technical details. Interesting read.
> 
> Yes, we were lucky to be able to retrieve detailed info on it! A PDP-11/20
> sold on eBay with a fairly complete set of KT11-B documentation, and allegedly
> a "KT11-B" as well, but alas, it turned out to 'only' be an RK11-C. Not that
> RK11-C's aren't cool, but on the 'cool scale' they are like 3, whereas a
> KT11-B would have been, like, 17! :-) Still, we managed to get the KT11-B
> 'manual' (such as it is) and prints online.

Really cool. Thanks for sharing.

> I'd love to find out equivalent detail for the KT11-A, but I've never seen
> anything on it. (And I've appealed before for the KS11, which an early PDP-11
> Unix apparently used, but no joy.)

Might have been just some internal early attempt that never got out of DEC?

>      > But how do you then view modern architectures which have different sized
>      > pages? Are they no longer pages then?
> 
> Actually, there is precedent for that. The original Multics hardware, the
> GE-645, supported two page sizes. That was dropped in later machines (the
> Honeywell 6000's) since it was decided that the extra complexity wasn't worth
> it.
> 
> I don't have any problem with several different page sizes, _if it makes
> engineering sense to support them_. (I assume that the rationale for their
> re-introduction is that in the age of 64-bit machines, page tables for very
> large 'chunks' can be very large if pages of ~1K or so are used, or something
> like.)
> 
> It does make real memory allocation (one of the advantages of paging) more
> difficult, since there would now be small and large page frames. Although I
> suppose it wouldn't be hard to coalesce them, if there are only two sizes, and
> one's a small power-of-2 multiple of the other - like 'fragments' in the
> Berkeley Fast File System for BSD4.2.

So, would you then say that such machines do not have pages, but have 
segments?
Or where do you draw the line? Is it some function of how many different 
sized pages there can be before you would call it segments? ;-)

But yes, it can make life more complex. But somewhere, this is all just 
math, and computers are usually pretty good with that.

> I have a query, though - how does a system with two page sizes know which to
> use? On Multics (and probably on the x86), it's a per-segment attribute. But
> on a system with a large, flat address space, how does the system know which
> parts of it are using small pages, and which large?

Beats me. I always just assumed it would be related to the virtual 
address. Some part of the address space use small pages, some other part 
uses large. But I have not ever cared to actually find out.
Just as some machines had I/O spaced mapped to several addresses, so 
that you could hit things cached or uncached, and possibly read/write 
different sized data efficiently without lots of fiddling.

   Johnny

-- 
Johnny Billquist                  || "I'm on a bus
                                   ||  on a psychedelic trip
email: bqt at softjar.se             ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol