I just re-found a quote about Unix processes that I'd "lost". It's by
Steve Johnson:
Dennis Ritchie encouraged modularity by telling all and sundry that
function calls were really, really cheap in C. Everybody started
writing small functions and modularizing. Years later we found out
that function calls were still expensive on the PDP-11, and VAX code
was often spending 50% of its time in the CALLS instruction. Dennis
had lied to us! But it was too late; we were all hooked...
http://www.catb.org/esr/writings/taoup/html/modularitychapter.html
Steve, can you recollect when you said this, was it just a quote for
Eric's book or did it come from elsewhere?
Does anybodu have a measure of the expense of function calls under Unix
on either platform?
Cheers, Warren
The comments in the rp06 walking across the floor reminds me of a time when I was installing netnews at a very new company and as the data transferred from the tape (we didn’t have a modem yet, that was to happen in a week or so) to the disk the disk got into a walking state. I was standing with my foot against the front of the drive to keep it from moving when a friend walked into the machine room and asked what I was doing. I said ‘keeping it all together’ and he asked about my foot. I moved it, the disk walked out a small amount and I replace my foot. He laughed and walked away. About 15 or 20 minutes later it was all back to ‘normal’ and I never had another problem with it.
David
> On Jan 2, 2016, at 1:31 PM, tuhs-request(a)minnie.tuhs.org wrote:
>
> several
> RP06 (200MB) removable disks (for a picture and description, see
>
> http://www.columbia.edu/cu/computinghistory/rp06.html
As late as 1990, every UNIX I knew of still used the
expensive calls/ret instructions for subroutine calls.
I vaguely remember a consensus (and I certainly shared
the feeling) that in hindsight it would have been better
to use jsb/rsb, but changing everything would have been
so much work that nobody wanted to do it.
1990 was already past peak VAX in the UNIX world, so
I can't imagine anyone bothering to make such a change
to an existing system after then. Especially a system
that already had many existing installations who would
have to deal with the resulting compatibility problem.
During the latter part of the 1990s, I was actively
supporting a private UNIX system just for myself on
a few MicroVAXes at home. One of the things I did
was to write a VAX code generator for the then-current
version of lcc (the one around which the book was
written), so as to have an ISO-compatible compiler
and convert all of /usr/src (not so big even in those
days) to ISO. It was an interesting exercise and I
learned a lot, but even then, I wasn't brave enough to
adopt an incompatible subroutine-calling convention.
Another big time waste in the original VAX UNIX was
the system-call interface: arguments were left on the
stack (where they had been put before calling the
syscall stub routine in libc); the kernel then had
to do a full-fledged copyin to get them. It occurred
to me more than once to change the convention and have
the syscall stubs copy the arguments into registers
before executing the chmk (syscall) instruction.
That instruction didn't touch the registers; the
kernel saved them early in the chmk trap routine,
in its own address space, so no copying or access
checking would have been required to fetch their
call-time contents.
That would still have been a messy change to make,
because I'd have to be sure every program had been
relinked with the new-style libc before changing the
kernel. (This was a system without shared libraries.)
But on a personal system it would have been doable.
I never did.
It's possible that current UNIX-descended/cloned systems
that have VAX ports, like Linux or Open/Free/NetBSD,
have had a chance to start over and do better on
subroutine calls and system calls. Does anyone know?
Norman Wilson
Toronto ON
> that's 28+13 = 41 memory cycles.
> ...
> purely in overhead (counting putting the args on the stack as overhead).
Oh, I missed an instruction for de-stacking the arguments, which was typically
something like 'add #N, sp', so another two instruction word fetches, or 43
cycles.
Ironically, if N=4, the compiler used to emit a 'cmp (sp)+, (sp)+', which is
more efficient space-wise (one word instead of two), but less time-wise
(3 cycles instead of 2).
Noel
> From: Warren Toomey
> I just re-found a quote about Unix processes
> ..
> Years later we found out that function calls were still expensive
> on the PDP-11
> ..
> Does anybodu have a measure of the expense of function calls under Unix
> on either platform?
Procedure calls were not cheap on the PDP-11 with the V6/V6 C compiler (which
admittedly was not the most efficient with small routines, since it always
saved all three non-temporary registers, no matter whether the called routine
used them or not).
This was especially true when compared to the code produced by the compiler
with the optimizer turned on, if the programmer was careful about allocating
'register' variables, which was pretty good.
On most PDP-11's, the speed was basically linearly related to the number of
memory references (both instruction fetch, and data), since most -11 CPU's
were memory-bound for most instructions. So for that compiler, a subroutine
call had a fair amount of overhead:
inst data
call 4 1
2 0 if any automatic variables
1 1 minimum per single-word argument
csv 9 5
cret 9 5
(In V7, someone managed to bum one cycle out of csv, taking it down to 8+5.)
So assume a pair of arguments which were not register variables (i.e.
automatics, or in structures pointed to by register variables), and some
automatics in the called routine, and that's 4+2 for the arguments, plus 6+1,
a subtotal of 10+3; add in csv and cret, that's 28+13 = 41 memory cycles.
On a typical machine like an 11/40 or 11/23, which had roughly 1 megacycle
memory throughput, that meant 40 usec (on a 1 MIP machine) to do a procedure
call, purely in overhead (counting putting the args on the stack as overhead).
We found that, even with the limited memory on the -11, it made sense to run
the time/space tradeoff the other way for short things like queue
insertion/removal, and do them as macros.
A routine had to be pretty lengthy before it was worth paying the overhead, in
order to amortize the calling cost across a fair amount of work (although of
course, getting access to another three register variables could make the
compiled output for the routine somewhat shorter).
Noel
Folks remember, VAX was not designed with UNIX in mind. It had two primary
influences, assembly programmers (Cutler et al) and FORTRAN compiler
writers. The truth is, the Vax was incredibly successful in both UNIX and
its intended OS (VMS) sites, even if a number of the instructions it has
were ignored by the C compiler writers. The fact the C did not map to it
as well as it would for other architectures later is not surprising given
the design constraints - C and UNIX were not part of the design. But it was
good enough (actually pretty darned good for the time) and was very, very
successful - I certainly stopped running a PDP11 when Vaxen were generally
available. I would not stop doing that until the 68000 based workstations
came along.
>From my own experience, when Dave (Patterson) was writing the RISC papers
in the early 1980s, a number of us ex-industry grad student types @ USB
were taking his architecture course having just come off some very
successful systems from the Vax, DG Eagle, Pr1me 750, etc.. [I'll leave the
names of said parties off to protect the innocent]. But what I will say is
that the four of used sit in the back of his calls and chuckle. We used to
remind Dave that a lot of the choices that were made on those machines, we
not for "CS" style reasons. IMO: Dave really did not "get it" -- all of
those system designers did make architectural choices, but the drivers were
the code base from the customer sites not how how well spell or grep
worked. And those commercial systems generally did mapped well at what
the designers considered and >>why<< those engineers considered what they
did [years later a HBS professor Clay Christensen's book explained why].
I've said this in other forums, but I contend that when we used pure CS to
design world's greatest pure computer architecture (Alpha) we ultimately
failed in the market. The computer architecture was extremely successful
and many of miss it. Hey, I now work for a company with one of the worst
instruction sets/ISA from a Computer Science standpoint - INTEL*64 (C),
and like the Vax, it's easy to throw darts at the architecture from a
purity standpoint. Alpha was great, C and other languages map to it well,
and the designers followed all of the CS knowledge at the time. But as a
>>system<< it could not compete with the disruption caused by the 386 and
later it's child, INTEL*64. And like Vaxen, INTEL*64 is ugly, but it
continues to win because of the economics.
At Intel we look at very specific codes and how they map and the choices of
what new things to add, how the system morphs are directly defined by what
we see from customers and in the case of scientific codes, how well the
FORTRAN compiler can exploit it -- because it is the same places (the
national labs and very large end users like weather, automotive, oil/gas or
life sciences) that have the same Fortran code that still need to run ;-)
This is just want DEC did years ago with the VAX (and Alpha).
As an interesting footnote, the DNA from the old DEC Fortran compiler lives
on "ifort" (and icc). Some of the same folks are still working on the
code generator, although they are leaving us fairly rapidly as they
approach and pass their 70s. But that's a different story ;-)
So the question is not a particular calling sequence or set of instructions
is good, you need to look at the entire economics of the system - which to
me begs the question of if the smartphone/tablet and ARM be the disruptor
to INTEL*64 - time will tell.
Clem
On Sun, Jan 3, 2016 at 7:42 PM, <scj(a)yaccman.com
<https://mail.google.com/mail/?view=cm&fs=1&tf=1&to=scj@yaccman.com>> wrote:
> Well, I certainly said this on several occasions, and the fact that it is
> recorded more or less exactly as I remember saying it suggests that I may
> have even written it somewhere, but if so, I don't recall where...
>
> As part of the PCC work, I wrote a technical report on how to design a C
> calling sequence, but that was before the VAX. Early calling sequences
> had both a stack pointer and a frame pointer, but for most machines it
> was possible to get by with just one, so calling sequences got better as
> time went on. Also, RISC machines with many more registers than the
> PDP-11 also led to more efficient calls by putting some arguments in
> registers. Later standardizations like varargs were painful on some
> architectures (especially those which had different registers for pointers
> and integers).
>
> The CALLS instruction was indeed a pig -- a space-time tradeoff in the
> wrong direction! For languages like FORTRAN it might have been justified,
> but for C it was awful. It is my memory too that CALLS was abandoned,
> perhaps first at UCB. But I actually had little hands-on experience with
> the VAX C compiler...
>
> Steve
>
>
>
>
> > I just re-found a quote about Unix processes that I'd "lost". It's by
> > Steve Johnson:
> >
> > Dennis Ritchie encouraged modularity by telling all and sundry that
> > function calls were really, really cheap in C. Everybody started
> > writing small functions and modularizing. Years later we found out
> > that function calls were still expensive on the PDP-11, and VAX code
> > was often spending 50% of its time in the CALLS instruction. Dennis
> > had lied to us! But it was too late; we were all hooked...
> > http://www.catb.org/esr/writings/taoup/html/modularitychapter.html
> >
> > Steve, can you recollect when you said this, was it just a quote for
> > Eric's book or did it come from elsewhere?
> >
> > Does anybodu have a measure of the expense of function calls under Unix
> > on either platform?
> >
> > Cheers, Warren
> >
>
>
>
On 2016-01-04 00:53, Tim Bradshaw <tfb(a)tfeb.org> wrote:
>> >On 3 Jan 2016, at 23:35, Warren Toomey<wkt(a)tuhs.org> wrote:
>> >
>> >Does anybodu have a measure of the expense of function calls under Unix
>> >on either platform?
>> >
> I don't have the reference to hand, but one of the things Lisp implementations (probably Franz Lisp in particular) did on the VAX was not to use CALLS: they could do this because they didn't need to interoperate with C except at known points (where they would use the C calling conventions). This change made a really significant difference to function call performance and meant that on call-heavy code Lisp was often very competitive with C.
>
> I can look up the reference (or, really, ask someone who remembers).
>
> The VAX architecture and its performance horrors must have killed DEC, I guess.
I don't know if that is a really honest description of the VAX in
general, nor DEC. DEC thrived in the age of the VAX.
However, the CALLS/CALLG and RET instructions were really horrid for
performance. Any clever programmer started using JSB and RSB instead. as
they give you the plain straight forward call and return semantics
without all the extra stuff that the CALL instructions give.
But, for assembler programmers, the architecture was nice. For
compilers, it's more difficult to do things optimal, and of course, it
took quite a while before hardware designers had the tools, skill and
knowledge how to implement complex instruction sets fast in hardware.
But nowadays, that is definitely not a problem, and it was more or less
already solved by the time of the NVAX chip as well, which was actually
really fast compared to a lot of stuff when it came out.
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt(a)softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
Clem cole <clemc(a)ccc.com> writes on Thu, 31 Dec 2015 23:04:04 -0500
about SPICE:
>> ...
>> Anyway SPICE1 was actually started in the late 1960's by dop [Don
>> Pederson]. Ellis Cohen wrote SPICE2 for the CDC 6400 in the mid 70's,
>> added some new device models and created really novel bit of self
>> modifying Fortran the compiled the inner loop.
>>
>> You are correct it was really the first widely available FOSS code -
>> an idea that you correctly note dop created.
>> ...
SPICE wasn't the only such package, or even the earliest! Still, I'll
be grateful to list readers for pointers off-list (or on) to early
publications about SPICE that I can add to the bibliography archives.
The EISPACK system, which predated LINPACK, and both of which led to
the current LAPACK, and descendants like CLAPACK and ScaLAPACK, has an
older vintage. It began with Algol routines published in the
German/English journal Numerische Mathematik
http://www.math.utah.edu/pub/tex/bib/nummath.bibhttp://www.math.utah.edu/pub/tex/bib/nummath2000.bibhttp://www.math.utah.edu/pub/tex/bib/nummath2010.bib
[change .bib to .html for a similar view with live hyperlinks]
The first such routine may have been that in entry Martin:1965:SDPa in
nummath.bib, which appeared in Num. Math. 79(5) 355--361 (October
1965) doi:10.1007/BF01436248. That journal did not then record
"received" dates, so the best that I can do for now is to claim
"October 1965" as the start of published code for free and open source
software in the area of numerical analysis.
Publication of related algorithms continued for 6 years, and then they
were collected in the famous HACLA (Handbook for Automatic
Computation: Linear Algebra) volume in 1971 (ISBN 0-387-05414-6).
Because Algol was little used in the USA, a project was begun in that
country to translate the Algol code to Fortran. That project was
called NATS, which originally stood for the groups at (read their
names vertically)
Northwestern University
Argonne National Laboratory
Texas, University of (at Austin)
Stanford
but as more groups joined in the effort, and EISPACK begat LINPACK,
NATS was changed to mean National Activity to Test Software.
The EISPACK book appeared in two editions in 1976 (ISBN 0-387-06710-8)
and 1977 (0-387-08254-9), volumes 6 and 51, respectively of Springer's
Lecture Notes in Computer Science (now around 9000 published volumes).
The LINPACK book appeared in 1979 (ISBN 0-89871-172-X).
The LAPACK book has three editions, in 1992 (ISBN 0-89871-294-7), 1995
(ISBN 0-89871-345-5), and 1999 (ISBN 0-89871-447-8). In between them,
the ScaLAPACK book appeared in 1997 (ISBN 0-89871-400-1).
There were several other packages described in the 1984 book
Sources and Development of Mathematical Software
ISBN 0-13-823501-5
(entry Cowell:1984:SDM), including FUNPACK, MINPACK, IMSL, SLATEC,
Boeing, AT&T PORT, and NAG. Some are free, and others are commercial.
The Algol code from Numerische Mathematik, like the ACM Collected
Algorithms, the Applied Statistics algorithms, and the Transactions on
Mathematical Software algorithms, was intended to be freely available
to anyone for any purpose, and no license of any kind was claimed for
it. That tradition continues with all of its descendants in the *PACK
family.
I have old archives of source code for EISPACK and LINPACK, but
comment documentation in EISPACK does not include revision dates, just
references to page numbers in the HACLA volume from 1971, and rarely,
to journal articles from 1968, 1970 and 1973. My filesystem dates,
alas, only reflect the copying from distribution tape to disk, and my
oldest file date for EISPACK is 20-Apr-1981.
The LINPACK comments appear be almost entirely without dates: I found
only one:
snrm2.for:11:C C.L.LAWSON, 1978 JAN 08
The bibliography on the GNU Project at
http://www.math.utah.edu/pub/tex/bib/gnu.bib
records most of the books mentioned above, and it also contains as its
first entry, Galler:1960:LEC, a letter published in the April 1960
issue of Communications of the ACM from Bernie Galler, with this
field:
remark = "From the letter: ``\ldots{} it is clear that what is
being charged for is the development of the program,
and while I am particularly unhappy that it comes from
a university, I believe it is damaging to the whole
profession. There isn't a 704 installation that hasn't
directly benefited from the free exchange of programs
made possible by the distribution facilities of
SHARE. If we start to sell our programs, this will set
very undesirable precedents.''",
That is so far the earliest reference that I have found for the notion
that software should be free, long before Richard Stallman, Eric
Raymond, Linus Torvalds, and others became such well-known proponents
of that idea, and we had large and profitable companies like Red Hat
and SUSE devoted to supporting, for a fee, such software.
I was a graduate student in quantum chemistry at the Quantum Theory
Project (QTP) at the University of Florida in Gainesville in the late
1960s and early 1970s, and we had a general practice of sharing of
code among various university research groups, most notably through
the Quantum Chemistry Program Exchange (QCPE) hosted at the University
of Indiana in Bloomington, IN.
A search through my bibliography archives found my earliest recording,
a 6-Apr-1971 publication (by me), with mention of QCPE. Library
searches found a catalog entry for QCPE Catalog volume 19 (1987), so
perhaps volume 1 appeared in 1968. But no --- I just found in its
home institution's library catalog
http://www.iucat.iu.edu/?utf8=%E2%9C%93&search_field=all_fields&q=QCPE&high…
an entry dated 1963, with details
Publishing history: 1 (Apr. 1963)- Ceased with 71 (Nov. 1980).
Other widely-distributed programs of that time included Enrico
Clementi's IBM Research group's IBMOL (about 1965), and others named
MOLECULE (pre-1975), POLYATOM (1963), and Gaussian (1970).
The POLYATOM year appears to be the earliest of those: see the paper
by Michael Barnett at
http://dx.doi.org/10.1103/RevModPhys.35.571
It appears in a July 1963 journal issue, again without a "received"
date. It begins:
A system of programs is being written by Dr. Malcolm
C. Harrison, Dr. Jules W. Moskowitz, Dr. Brian T. Sutcliffe,
D. E. Ellis, and R. S. Pitzer, to perform nonempirical
calculations for small molecules.
I have met, or been in the same group as (Don Ellis), most of those,
and it is worth noting their affiliations to emphasize the broad
character of that work:
Malcolm Courant Institute of Mathematical Sciences, New York University, NY
Jules New York University, NY
Brian York University, York, UK
Don University of Florida (later, Northwestern University)
Russ Harvard, Cambridge, MA (later, Ohio State University)
Michael MIT, Cambridge, MA and various UK sites in academia and industry
(see https://en.wikipedia.org/wiki/Michael_P._Barnett)
On the subject of the Gaussian program, developed at Carnegie-Mellon
University, see the two sites
https://en.wikipedia.org/wiki/Gaussian_%28software%29http://www.bannedbygaussian.org/
The second decries the loss of openness of Gaussian, which remains a
widely-used commercial product.
There is also a book on the subject of mathematics whose use is
encumbered by patents and copyrights:
Ben Klemens
Ma$+$h you can't use: patents, copyright, and software
ISBN 0-8157-4942-2
(entry Klemens:2006:MYC in http://www.math.utah.edu/pub/tex/bib/master.bib)
----------------------------------------
P.S. A final sad personal note on computing history:
When our DEC-20/60 (Arpanet node UTAH-SCIENCE, later science.utah.edu
and still later, math.utah.edu) was retired on 31-Oct-1990 (its
predecessor, a DEC-20/40 began operating in March 1978) we were faced
with several cabinets full of 9-track tapes (about 25MB each), several
RP06 (200MB) removable disks (for a picture and description, see
http://www.columbia.edu/cu/computinghistory/rp06.html
) and the contents of three washing-machine sized RP07 (600MB) disks,
and were moving to a new machine room in an adjacent building.
We were able to copy over the RP0[67] disk contents, and I still have
them online on my desktop, but the tapes were financially infeasible
for us to copy to disk on the new VAX 8600 server, and we were leaving
9-track tape technology behind. There were probably 500 to 1000 of
those tapes, and all that we could do was fill a dumpster with them,
because we had no place to store the physical volumes at the new site,
and no money for their bits. I have deeply regretted that loss of 25
years of my, and our, early computing history ever since.
Computers were for far too long crippled by too little memory and too
little permanent storage, and only post-2000 has that situation been
alleviated with radical reductions in storage costs per byte of data.
My new desktop 8TB drive is 3.6 million times cheaper per byte than an
RP06 drive was. Had we been able to foresee that dramatic growth in
capacity, we could have archived those tapes in an off-campus
warehouse for later (attempted) data retrieval.
------------------------------------------------------------------------
P.P.S. Besides VAX VMS, our migration path from TOPS-20 was primarily
to Unix, first on the Wollongong distribution of BSD (3.x, I think)
running on VAX 750 machines in the early 1980s, then on Sun 3
MC68000-based workstations in 1988 that ultimately evolved to an
eclectic mixture of CPUs and vendors. My software test lab now has
about 70 flavors of Unix on assorted physical and virtual machines,
with ARM, MIPS, PowerPC, SPARC, x86, and x86-64 processors. Our last
DEC Alpha CPU died with its power supply 16 months ago, and a
colleague still has a runnable MC68020 box (an old NeXT desktop).
-------------------------------------------------------------------------------
- Nelson H. F. Beebe Tel: +1 801 581 5254 -
- University of Utah FAX: +1 801 581 4148 -
- Department of Mathematics, 110 LCB Internet e-mail: beebe(a)math.utah.edu -
- 155 S 1400 E RM 233 beebe(a)acm.org beebe(a)computer.org -
- Salt Lake City, UT 84112-0090, USA URL: http://www.math.utah.edu/~beebe/ -
-------------------------------------------------------------------------------
> From: Jacob Goense <dugo(a)xs4all.nl>
> Mills's 1983 RFC889[2] calls the original PING Packet InterNet Groper.
I have a strong suspicion that Packet-etc is a 'backronym' from Dave Mills.
Note that the use of the term "echo" for a packet returned dates back quite a
while before that, see e.g. IEN 104, "Minutes of the Fault Isolation
Meeting", from March 1979:
"ability to echo packets off any gateway"
When ICMP was split from GGP (see IEN-109, RFC-777), the functionality
migrated from GGP to ICMP, and was generalized so that all hosts provided the
capability, not just routers.
Noel
Personally, I lean away from listing the nine billion debunked
names of cron. It's like adding a disclaimer to cat(1) to
explain that cat just copies data to standard output, it doesn't
transform it or compute how long it would take to send the data
over UUCP.
But it probably shows that I have been trying to write a couple
of manual pages lately (for some personal stuff, plus some docs
for work that are not technically manual pages but deserve the
same sort of conciseness).
Maybe Wikipedia-page format should admit an optional BUGS section.
Norman Wilson
Toronto ON
PS: seriously, though I wouldn't bother including the debunking
text myself, save perhaps on the Talk page to encourage editors
to delete any future attempts to revive the un-names, I have no
problem with Grog doing it. More power to him if he has the
energy!
