I forgot to mention a critical point. Any PDP-11 runing Unix (except very early versions and Miniunix), and certainly from Edition 6 onwards, MUST have EIS. Even the Unix bootstraps used EIS. So binaries built for Unix, will run on a 11/34/35/40/44/45/50/53/55/60/70/73/83/84/93/94 !! Floating point is a bit more problematic. The kernel (see crevat latter) didn't require it, but had to save FP status and registers on context switches. Quite a few processors had the FPP as an option, and so there was FPP emulation build into the kernel (conditionally). There were versions of the C compiler that had code tables for the FIS, to suit the 11/35/40. I have a distant memory, that I have seen FPP instructions used for some integer arithmetic for speed. I cannot recall if it was in the kernel or C libraries. It was conditional, and may have been in the latter BSD versions, but I don't have the source code online. Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id LAA75986 for pups-liszt; Wed, 10 May 2000 11:50:48 +1000 (EST) Received: from henry.cs.adfa.edu.au (henry.cs.adfa.edu.au [131.236.21.158]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id LAA75982 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 11:50:42 +1000 (EST) Received: (from wkt@localhost) by henry.cs.adfa.edu.au (8.9.2/8.9.3) id LAA90764; Wed, 10 May 2000 11:50:31 +1000 (EST) (envelope-from wkt) From: Warren Toomey <wkt(a)cs.adfa.edu.au> Message-Id: <200005100150.LAA90764(a)henry.cs.adfa.edu.au> Subject: Re: Help: PDP-11 instruction classification In-Reply-To: <85og6g18xo.fsf(a)junk.nocrew.org> from lars brinkhoff at "May 9, 2000 9:25:55 am" To: lars(a)nocrew.org (lars brinkhoff) Date: Wed, 10 May 2000 11:50:31 +1000 (EST) Cc: pups(a)minnie.cs.adfa.edu.au (Unix Heritage Society) Reply-To: wkt(a)cs.adfa.edu.au X-Mailer: ELM [version 2.4ME+ PL43 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk In article by lars brinkhoff: I've got some details on FADD, FSUB, FMUL, and FDIV. I could fax you the relevant details from the processor handbook, and/or give you source code from my Apout simulator. FADD is instruction 07500x, ADDF is 1720xx. They are different. I don't know enough about the CIS extensions. Can you name some of the opcode names, so that I can look in the handbooks I have here: 04, 05, 10, 15, 20, 34, 35, 40, 45, 55, 60, 70. Warren Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id NAA76328 for pups-liszt; Wed, 10 May 2000 13:00:15 +1000 (EST) Received: from moe.2bsd.com (0(a)MOE.2BSD.COM [206.139.202.200]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id NAA76324 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 13:00:06 +1000 (EST) Received: (from sms@localhost) by moe.2bsd.com (8.9.3/8.9.3) id TAA18914 for pups(a)minnie.cs.adfa.edu.au; Tue, 9 May 2000 19:58:25 -0700 (PDT) Date: Tue, 9 May 2000 19:58:25 -0700 (PDT) From: "Steven M. Schultz" <sms(a)moe.2bsd.com> Message-Id: <200005100258.TAA18914(a)moe.2bsd.com> To: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk The CIS was a quite expensive option that was only available (as I recall) on the 11/44 and the later KDJ-11 based systems (11/83 and up where it occupied another socket on the cpu board). The CIS instructions are: L2D L3D ADDP ADDN ADDNI ADDPI ASHN ASHP ASHNI ASHPI CMPC CMPCI CMPN CMPP CMPNI CMPPI CVTLN CVTLP CVTLNI CVTLPI CVTNL CVTPL CVTLNI CVTPLI CVTNP CVTPN CVTNPI CVTPNI DIVP DIVPI LOCC LOCCI MATC MATCI MOVC MOVCI MOVRC MOVRCI MOVTC MOVTCI MULP MULPI SCANC SCANCI SKPC SKPCI SPANC SPANCI SUBN SUBP SUBNI SUBPI Try the processor handbook that covers the 11/44+11/70 - I think that's where the CIS is described. At one time it looked like where I worked was going to get a CIS board for the 11/44 to help with the COBOL runtime we'd written/ported. It never came to be ;( But I did add the opcodes to the assembler and did some initial coding in preparation for getting the CIS instructions. You do a _lot_ of saving and restoring registers because the CIS instructions expect many of their operands to be in R2 thru R5 (R0 and R1 were used for returning results). That's a pretty high percentage (~100) of the available registers ;) Steven Schultz sms(a)moe.2bsd.com Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id NAA76464 for pups-liszt; Wed, 10 May 2000 13:30:12 +1000 (EST) Received: from moe.2bsd.com (0(a)MOE.2BSD.COM [206.139.202.200]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id NAA76452 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 13:30:00 +1000 (EST) Received: (from sms@localhost) by moe.2bsd.com (8.9.3/8.9.3) id UAA19095 for pups(a)minnie.cs.adfa.edu.au; Tue, 9 May 2000 20:22:57 -0700 (PDT) Date: Tue, 9 May 2000 20:22:57 -0700 (PDT) From: "Steven M. Schultz" <sms(a)moe.2bsd.com> Message-Id: <200005100322.UAA19095(a)moe.2bsd.com> To: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification (again!) Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk Hi -- Absolutely correct. By the time Unix was making its way out of the lab and becoming commercially available (V7 and later for certain) the 11/70 was the target machine and while you could order an 11/70 without floating point very few (that I saw) were bought that way. Even some of the DEC bootstraps used EIS - at least the later RSX-11D ones did. At one site we couldn't boot the 11/45 into RSX-11D, it would crash part way thru the boot process. The DEC folks would bring in their diagnostic disks and they would boot just fine. Finally I insisted they run full cpu diagnostics and voila - the 'div' (or was it the 'ash' - been a looong time) instruction was failing. The diagnostic packs were RT11 based (I believe) and were meant to run on all processors rather than the 45 and up. The compiler also checks if any FP is done and generates references to special symbols as needed. The linker can then link in dummy modules for parts of the 'printf' and int->floating conversion routines and save some space that way. A kluge but on a small machine every half kb counts. The kernel (at least V7 and later) has special code in it to catch and ignore the illegal instruction 'setd' if no hardware FP is present. This is because the crt0 routine (which receives control from the kernel when a program is run) forces double precision mode in all programs. Some floating point simulators (V7's for example) ran in user mode and caught the SIGILL signal). I guess running slow instead of crashing is a plus but wow did that bog down the system immensely. Of course one of the programs that was used a lot was a Fortran one and f77 generates a *lot* of floating point code. It wasn't just the emulation of the FP instructions (that would not have been too bad) but the overhead of the hundreds of signals and context switches was a system killer. 2.10BSD and later have an in-kernel FP emulator - alas it doesn't work. About the most I did was get it thru the assembler. All the systems I have come with builtin FP. At one time Tim Shoppa was going to take a stab at pulling the FP board out of his 11/44 and see about getting the emulator working - copious free time permitting of course ;) In the C library. The kernel itself can't (or at least shouldn't) do floating point stuff because that might happen at interrupt level and trash the current user process (if any) context. The C library code (at least in 2.11) uses FP to do the 32 bit divide and multiply. It's a lot faster (and much less code) to convert long to double, do the operations and then store&convert double back to int or long. There might be one or two other cases but that was the primary use. And thereby hangs a bit of a tale. The KDJ-11{A,B} handle faults during a stack push by a FP instruction differently than the 11/44 or 11/70. If you have access to the 2.11BSD bug archive (I know it's on 'ns.to.gd-es.com' and 'moe.2bsd.com' - not sure of who else is mirroring it) look for update #150 (Subject is "stack expansion bug on KDJ-11 cpus". Steven Schultz sms(a)moe.2bsd.com Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id NAA76504 for pups-liszt; Wed, 10 May 2000 13:45:07 +1000 (EST) Received: from moe.2bsd.com (0(a)MOE.2BSD.COM [206.139.202.200]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id NAA76494 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 13:44:59 +1000 (EST) Received: (from sms@localhost) by moe.2bsd.com (8.9.3/8.9.3) id UAA19137 for pups(a)minnie.cs.adfa.edu.au; Tue, 9 May 2000 20:32:53 -0700 (PDT) Date: Tue, 9 May 2000 20:32:53 -0700 (PDT) From: "Steven M. Schultz" <sms(a)moe.2bsd.com> Message-Id: <200005100332.UAA19137(a)moe.2bsd.com> To: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk Hi - Quite correct. Adding the EIS brings an 11/40 to the same level as an 11/45 without floating point. Yes - that's part of the EIS. Standard on the 11/45, 11/70 (and the later KDJ-11 systems such as the 11/53, 73, 83,etc) Not a popular option at all. At the time DEC was trying to make the 11 more of a COBOL machine but the CIS option was too little, too late and expensive (plus it wouldn't fit as I recall on a 11/70 - just the 11/44 and newer). ... Looks right to me. Two different machines. The FADD was part of the FIS option for the 11/40 only (I don't recall ever hearing of someone adding the FIS to a 11/45 or 70). The 11/45, 70 and later all had the FPU as an option or standard (the KDJ-11{A,B} had the instructions standard but you could buy (~$600 at the time) an accelerator chip to speed them up). I don't think there were any models after the 11/70 that used the FIS. There was also the CSM (Call Supervisor Mode) instruction which was on the 11/44 only (or did it make it into the later KDJ-11 line as well - I forget). Steven Schultz sms(a)moe.2bsd.com Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id QAA77102 for pups-liszt; Wed, 10 May 2000 16:56:39 +1000 (EST) Received: from Zeke.Update.UU.SE (IDENT:2026@Zeke.Update.uu.se [130.238.11.14]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id QAA77098 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 16:56:29 +1000 (EST) Received: from localhost (bqt@localhost) by Zeke.Update.UU.SE (8.8.8/8.8.8) with SMTP id IAA22713; Wed, 10 May 2000 08:56:05 +0200 Date: Wed, 10 May 2000 08:56:04 +0200 (MET DST) From: Johnny Billquist <bqt(a)Update.UU.SE> To: Warren Toomey <wkt(a)cs.adfa.edu.au> cc: lars brinkhoff <lars(a)nocrew.org>, Unix Heritage Society <pups(a)minnie.cs.adfa.edu.au> Subject: Re: Help: PDP-11 instruction classification In-Reply-To: <200005100150.LAA90764(a)henry.cs.adfa.edu.au> Message-ID: <Pine.VUL.3.93.1000510085424.22474D-100000(a)Zeke.Update.UU.SE> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk On Wed, 10 May 2000, Warren Toomey wrote: You don't have a handbook for any machine that could have CIS then Warren. It was only available for the 11/44 and F-11 based machines. (11/23 11/24) Johnny Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt(a)update.uu.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id RAA77158 for pups-liszt; Wed, 10 May 2000 17:02:53 +1000 (EST) Received: from junk.nocrew.org (mail(a)[212.73.17.42]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id RAA77154 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 17:02:45 +1000 (EST) Received: from lars by junk.nocrew.org with local (Exim 3.03 #1 (Debian)) id 12pQW0-0000VH-00; Wed, 10 May 2000 09:02:40 +0200 To: "Steven M. Schultz" <sms(a)moe.2bsd.com> Cc: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification References: <200005100332.UAA19137(a)moe.2bsd.com> From: lars brinkhoff <lars(a)nocrew.org> Date: 10 May 2000 09:02:40 +0200 In-Reply-To: "Steven M. Schultz"'s message of "Tue, 9 May 2000 20:32:53 -0700 (PDT)" Message-ID: <85g0rqzy3z.fsf(a)junk.nocrew.org> User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.5 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk "Steven M. Schultz" <sms(a)moe.2bsd.com> writes: What bit pattern does that instruction have? Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id RAA77208 for pups-liszt; Wed, 10 May 2000 17:11:51 +1000 (EST) Received: from Zeke.Update.UU.SE (IDENT:2026@Zeke.Update.uu.se [130.238.11.14]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id RAA77204 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 17:11:42 +1000 (EST) Received: from localhost (bqt@localhost) by Zeke.Update.UU.SE (8.8.8/8.8.8) with SMTP id JAA22854; Wed, 10 May 2000 09:11:25 +0200 Date: Wed, 10 May 2000 09:11:24 +0200 (MET DST) From: Johnny Billquist <bqt(a)Update.UU.SE> To: "Steven M. Schultz" <sms(a)moe.2bsd.com> cc: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification In-Reply-To: <200005100332.UAA19137(a)moe.2bsd.com> Message-ID: <Pine.VUL.3.93.1000510090352.22474E-100000(a)Zeke.Update.UU.SE> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk On Tue, 9 May 2000, Steven M. Schultz wrote: Hi, all. ... > > > Not a popular option at all. At the time DEC was trying to make the > 11 more of a COBOL machine but the CIS option was too little, too late > and expensive (plus it wouldn't fit as I recall on a 11/70 - just the > 11/44 and newer). Actually, not even those. It's the 11/44 and the KDF-11 based systems only, which means 11/23 and 11/24. ... FIS is only for 11/35, 11/40 and the LSI-11. And it's a real brain-damaged thing too. Lucky us FPU came along. Yes, CSM is in the J-11. There is also TSTSET in J11, but I wonder if anyone uses it. Speaking of CSM, by the way. Slighty off-topic perhaps, but in RSX-11M-PLUS they used to have a rather complex way of getting to supervisor mode in user programs, since the 11/70 didn't have CSM. Then came the J-11, and DEC changed things around on systems having CSM to use this instruction instead. It turned out that this improved code quite a lot, so they implemented the CSM instruction by emulation on the 11/70 as well. Johnny Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt(a)update.uu.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id RAA77339 for pups-liszt; Wed, 10 May 2000 17:41:47 +1000 (EST) Received: from junk.nocrew.org (mail(a)[212.73.17.42]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id RAA77335 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 17:41:39 +1000 (EST) Received: from lars by junk.nocrew.org with local (Exim 3.03 #1 (Debian)) id 12pR7b-0002L7-00; Wed, 10 May 2000 09:41:31 +0200 To: Johnny Billquist <bqt(a)Update.UU.SE> Cc: "Steven M. Schultz" <sms(a)moe.2bsd.com>, pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification References: <Pine.VUL.3.93.1000510090352.22474E-100000(a)Zeke.Update.UU.SE> From: lars brinkhoff <lars(a)nocrew.org> Date: 10 May 2000 09:41:31 +0200 In-Reply-To: Johnny Billquist's message of "Wed, 10 May 2000 09:11:24 +0200 (MET DST)" Message-ID: <85bt2ezwb8.fsf(a)junk.nocrew.org> User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.5 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk Johnny Billquist <bqt(a)Update.UU.SE> writes: Ok, so CSM is found in 11/44 and J11 machines? And TSTSET is found only in J11? Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id RAA77372 for pups-liszt; Wed, 10 May 2000 17:47:13 +1000 (EST) Received: from Zeke.Update.UU.SE (IDENT:2026@Zeke.Update.uu.se [130.238.11.14]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id RAA77368 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 17:47:02 +1000 (EST) Received: from localhost (bqt@localhost) by Zeke.Update.UU.SE (8.8.8/8.8.8) with SMTP id JAA23131; Wed, 10 May 2000 09:46:51 +0200 Date: Wed, 10 May 2000 09:46:50 +0200 (MET DST) From: Johnny Billquist <bqt(a)Update.UU.SE> To: lars brinkhoff <lars(a)nocrew.org> cc: "Steven M. Schultz" <sms(a)moe.2bsd.com>, pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification In-Reply-To: <85bt2ezwb8.fsf(a)junk.nocrew.org> Message-ID: <Pine.VUL.3.93.1000510094549.22474H-100000(a)Zeke.Update.UU.SE> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk On 10 May 2000, lars brinkhoff wrote: Yes. Yes. I think that the J11 had one or two other new instructions as well, but I'll have to look it up at home, unless someone beats me to it. Johnny Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt(a)update.uu.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id SAA77434 for pups-liszt; Wed, 10 May 2000 18:02:00 +1000 (EST) Received: from junk.nocrew.org (mail(a)[212.73.17.42]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id SAA77430 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 18:01:52 +1000 (EST) Received: from lars by junk.nocrew.org with local (Exim 3.03 #1 (Debian)) id 12pRRA-0003LQ-00; Wed, 10 May 2000 10:01:44 +0200 To: Johnny Billquist <bqt(a)Update.UU.SE> Cc: "Steven M. Schultz" <sms(a)moe.2bsd.com>, pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification References: <Pine.VUL.3.93.1000510094549.22474H-100000(a)Zeke.Update.UU.SE> From: lars brinkhoff <lars(a)nocrew.org> Date: 10 May 2000 10:01:44 +0200 In-Reply-To: Johnny Billquist's message of "Wed, 10 May 2000 09:46:50 +0200 (MET DST)" Message-ID: <857ld2zvdj.fsf(a)junk.nocrew.org> User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.5 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk Johnny Billquist <bqt(a)Update.UU.SE> writes: > > Yes. > I think that the J11 had one or two other new instructions as well, but > I'll have to look it up at home, unless someone beats me to it. WRTLCK perhaps? Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id UAA77937 for pups-liszt; Wed, 10 May 2000 20:55:53 +1000 (EST) Received: from Zeke.Update.UU.SE (IDENT:2026@Zeke.Update.uu.se [130.238.11.14]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id UAA77933 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 20:55:43 +1000 (EST) Received: from localhost (bqt@localhost) by Zeke.Update.UU.SE (8.8.8/8.8.8) with SMTP id MAA24782; Wed, 10 May 2000 12:55:33 +0200 Date: Wed, 10 May 2000 12:55:32 +0200 (MET DST) From: Johnny Billquist <bqt(a)Update.UU.SE> To: lars brinkhoff <lars(a)nocrew.org> cc: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification In-Reply-To: <857ld2zvdj.fsf(a)junk.nocrew.org> Message-ID: <Pine.VUL.3.93.1000510125423.22474I-100000(a)Zeke.Update.UU.SE> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk On 10 May 2000, lars brinkhoff wrote: > Johnny Billquist <bqt(a)Update.UU.SE> writes: > > > > Yes. > > I think that the J11 had one or two other new instructions as well, but > > I'll have to look it up at home, unless someone beats me to it. > > WRTLCK perhaps? Yup. That sounds familiar. The J11 have several design features aimed at multi-processor systems. None were produced, however. Johnny Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt(a)update.uu.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id XAA78420 for pups-liszt; Wed, 10 May 2000 23:02:28 +1000 (EST) Received: from junk.nocrew.org (mail(a)[212.73.17.42]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id XAA78416 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 23:02:18 +1000 (EST) Received: from lars by junk.nocrew.org with local (Exim 3.03 #1 (Debian)) for pups(a)minnie.cs.adfa.edu.au id 12pW7o-0000ZD-00; Wed, 10 May 2000 15:02:04 +0200 To: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification (again!) From: lars brinkhoff <lars(a)nocrew.org> Date: 10 May 2000 15:02:04 +0200 In-Reply-To: "Steven M. Schultz"'s message of "Tue, 9 May 2000 20:22:57 -0700 (PDT)" Message-ID: <85snvqy2wj.fsf(a)junk.nocrew.org> User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.5 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk Ok, I've tried to collect all information and make a handy table. I've also looked quit a bit at the processor feature table at http://www.pdp11.org/mirrors/www.psych.usyd.edu.au/pdp-11/table.html Could you please check that this table is correct? In particular, some features may be optional instead of standard (yes) or missing (no), or vice versa. Some models are listed as (see XXX), where XXX is a CPU model used in that machine. Are there more opportunities for doing that? It would be nice to have two tables: "Machine model 11/NN used CPU models X11 or Y11." "CPU model X11 had features A, B , and C." Or something like that. Model EIS EIS40 CSM TSTSET, FPP CIS FIS WRTLCK 03 (see LSI11) (LSI-11 or LSI-11/2) 04 no no no no no no no 05 no no no no no no no 10 no no no no no no no 15 no no no no no no no 20 no no no no no no no 21 (see T11) 23 (see F11) 24 (see F11) 34 yes yes no no yes no no 35 opt yes no no no? no opt 40 opt yes no no no no? opt 44 yes yes yes no opt opt no? 45 (see KB11) 50 (see KB11) 53 (see J11) (KDJ-11D?) 55 (see KB11+) (KB-11D) 60 yes yes no no yes[2] no no 70 (see KB11+) (KB-11B or KB-11C) 73 (see J11) (KDJ-11A or KDJ-11B) 83 (see J11) (KDJ-11B?C?) 84 (see J11) (KDJ-11B?C?) 93 (see J11) (KDJ-11D?) 94 (see J11) (KDJ-11D?) KB11 yes yes no no opt no no KB11+ yes yes no no yes[1] no no J11 yes yes yes yes yes opt no LSI11 opt yes no no no no opt T11 no no no no no no no F11 yes yes no no opt opt no [1] = really optional, but most shipped with fpp [2] = microcoded fpp standard, accelerated hardware fpp optional EIS = RTT, SPL, MARK, SXT, MUL, DIV, ASH, ASHC, XOR, SOB EIS40 = RTT, MARK, SXT, XOR, SOB FPP = CFCC, SETF, SETI, SETD, SETL, LDFPS, STFPS, STST, CLRF, TSTF, ABSF, NEGF, MULF, MODF, ADDF, LDF, SUBF, CMPF, STF, DIVF, STEXP, STCFI, STCFF', LDEXP, LDCIF, LDCFF' CIS = L2D, L3D, ADDP, ADDN, ADDNI, ADDPI, ASHN, ASHP, ASHNI, ASHPI, CMPC, CMPCI, CMPN, CMPP, CMPNI, CMPPI, CVTLN, CVTLP, CVT, CVTLNI, CVTLPI, CVTNL, CVTPL, CVLNI, CVTPLI, CVTNP, CVTPN, CVTNPI, CVTPNI, DIVP, DIVPI, LOCC, LOCCI, MATC, MATCI, MOVC, MOVCI, MOVRC, MOVRCI, MOVTC, MOVTCI, MULP, MULPI, SCANC, SCANCI, SKPC, SKPCI, SPANC, SPANCI, SUBN, SUBP, SUBNI, SUBPI FIS = FADD, FDIV, FMUL, FSUB KB11 = KB-11 KB11+ = KB-11B, KB-11C, or KB-11D J11 = KDJ-11A or KDJ-11B LSI11 = LSI-11 or LSI-11/2 T11 = ? F11 = ? Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id AAA78761 for pups-liszt; Thu, 11 May 2000 00:16:18 +1000 (EST) Received: from Zeke.Update.UU.SE (IDENT:2026@Zeke.Update.uu.se [130.238.11.14]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id AAA78757 for <pups(a)minnie.cs.adfa.edu.au>; Thu, 11 May 2000 00:16:08 +1000 (EST) Received: from localhost (bqt@localhost) by Zeke.Update.UU.SE (8.8.8/8.8.8) with SMTP id QAA26854; Wed, 10 May 2000 16:15:56 +0200 Date: Wed, 10 May 2000 16:15:55 +0200 (MET DST) From: Johnny Billquist <bqt(a)Update.UU.SE> To: lars brinkhoff <lars(a)nocrew.org> cc: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification (again!) In-Reply-To: <85snvqy2wj.fsf(a)junk.nocrew.org> Message-ID: <Pine.VUL.3.93.1000510154632.22474K-100000(a)Zeke.Update.UU.SE> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk On 10 May 2000, lars brinkhoff wrote: [table deleted...] First I think it's wrong of you to call the limited EIS stuff implemented by default on the 11/40 "EIS40". It's just that parts of the EIS was always available, with or without the optional EIS. Second, the FPU implemented in the 11/60 was a bit special, and not compatible with the normal FPU, which was an option. I'm not aware of anything that used the 11/60 internal FPU stuff. Specifically, it used the normal processor registers for operations, and I don't remember exactly if the data format was compatible, but it didn't have double precision at all. I can probably dig up specific details if you are interested. I think I have a processor handbook for the 11/60 somewhere. About processor names: KA11 was the 11/15 and 11/20. KB11-A was the 11/45, 11/50. KB11-B was the 11/70. KB11-C was the 11/70. KB11-D was the 11/55. KB11-Cm and KB-11E was the never produced 11/74. KD11-A was the 11/35 and 11/40. KD11-B was the 11/05 and 11/10. KD11-D was the 11/04. KD11-E was the 11/34. KD11-EA was the 11/34a. KD11-[FHQ] was the 11/03. KD11-HA was the LSI-11/2. KD11-K was the 11/60. KD11-Z was the 11/44. KDF11 is F-11 based. KDJ11 is J-11 based. All as far as I can glean from various sources. KE11 is not a processor, but instead different processor options, such as EIS, FIS, EAE and such. Johnny Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt(a)update.uu.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id BAA79057 for pups-liszt; Thu, 11 May 2000 01:39:06 +1000 (EST) Received: from fw0.transarc.com (xfw.transarc.ibm.com [192.54.226.51]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id BAA79053 for <pups(a)minnie.cs.adfa.edu.au>; Thu, 11 May 2000 01:38:58 +1000 (EST) Received: from mailhost2.transarc.ibm.com (mailhost2.transarc.ibm.com [9.38.192.125]) by fw0.transarc.com (AIX4.2/UCB 8.7/8.7) with ESMTP id LAA09672 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 11:18:06 -0400 (EDT) Received: from smithfield.transarc.ibm.com (smithfield.transarc.ibm.com [9.38.192.92]) by mailhost2.transarc.ibm.com (8.8.0/8.8.0) with SMTP id LAA28911 for <pups(a)minnie.cs.adfa.edu.au>; Wed, 10 May 2000 11:37:56 -0400 (EDT) Date: Wed, 10 May 2000 11:37:51 -0400 (EDT) From: Pat Barron <pat(a)transarc.ibm.com> To: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification (again!) In-Reply-To: <200005100135.LAA02228(a)psychwarp.psych.usyd.edu.au> Message-ID: <Pine.GSO.3.96.1000510111921.6129B-100000(a)smithfield.transarc.ibm.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk On Wed, 10 May 2000 johnh(a)psych.usyd.edu.au wrote: > > Floating point is a bit more problematic. The kernel (see crevat latter) didn't > require it, but had to save FP status and registers on context switches. > Quite a few processors had the FPP as an option, and so there was FPP emulation > This discussion brings back memories ... I once found a bug in the kernel floating-point stuff on V7m on an 11/40 (without FIS). Turned out that, if you took a zero-length file, chmod'ed it to be executable, and then tried to run it, the kernel would take a path through the code that it didn't normally take, in which it tried to save the *real* FP registers - which were not there, and attempting to touch the missing registers would panic the machine. I found this while I was a lab assistant for a computer architecture course, taking care of this 11/40 that the department had just aquired (another department was about to discard it, so we picked it up ...), and which was being used by students learning assembly language programming. As you might imagine, these folks generated zero-length a.out files *all the time* (since that's what 'as' would sometimes output if your source code had errors in it), and sometimes they'd try to execute them. Therefore, the machine was crashing a couple of times a day until I found and fixed the bug ... --Pat. Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id CAA79425 for pups-liszt; Thu, 11 May 2000 02:30:10 +1000 (EST) Received: from moe.2bsd.com (0(a)MOE.2BSD.COM [206.139.202.200]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with ESMTP id CAA79421 for <pups(a)minnie.cs.adfa.edu.au>; Thu, 11 May 2000 02:30:02 +1000 (EST) Received: (from root@localhost) by moe.2bsd.com (8.9.3/8.9.3) id JAA28584 for pups(a)minnie.cs.adfa.edu.au; Wed, 10 May 2000 09:20:58 -0700 (PDT) Date: Wed, 10 May 2000 09:20:58 -0700 (PDT) From: "Steven M. Schultz" <sms(a)moe.2bsd.com> Message-Id: <200005101620.JAA28584(a)moe.2bsd.com> To: pups(a)minnie.cs.adfa.edu.au Subject: Re: Help: PDP-11 instruction classification Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk From: lars brinkhoff <lars(a)nocrew.org> 070DD It's a single operand instruction. Steven Schultz sms(a)moe.2bsd.com Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id DAA79549 for pups-liszt; Thu, 11 May 2000 03:06:26 +1000 (EST) Received: from timaxp.trailing-edge.com (timaxp.trailing-edge.com [63.73.218.130]) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) with SMTP id DAA79545 for <PUPS(a)MINNIE.CS.ADFA.EDU.AU>; Thu, 11 May 2000 03:06:18 +1000 (EST) Received: by timaxp.trailing-edge.com for PUPS(a)MINNIE.CS.ADFA.EDU.AU; Wed, 10 May 2000 13:06:12 -0400 Date: Wed, 10 May 2000 13:06:12 -0400 From: Tim Shoppa <SHOPPA(a)trailing-edge.com> To: PUPS(a)minnie.cs.adfa.edu.au Message-Id: <000510130612.202009f2(a)trailing-edge.com> Subject: Re: Help: PDP-11 instruction classification (again!) Sender: owner-pups(a)minnie.cs.adfa.edu.au Precedence: bulk It's complicated by the fact that often the same CPU module was used in differently DEC-labeled systems. For example, a late-rev KDJ11B with PMI memory in a Q-bus is an 11/83; the exact same CPU board with non-PMI memory in a Q-bus is an 11/73. And the exact same CPU board in a very different backplane is an 11/84. And an 11/73 can also have a KDJ11A in it, a very different module. See Micronote #39. At one point I began writing up a "KDJ11-x" FAQ, but never got it finished. There are many variations between different revs of the J11 CPU chip and the boards, especially with respect to whether FPJ11's work properly or not. Many (but not all) of the KDJ11 differences are well described in Micronote #39, _KDJ11-A and KDJ11-B Differences_. (Side hint: everyone should have a copy of the Micronotes. If you don't have a printed set, you can read them online at http://metalab.unc.edu/pub/academic/computer-science/history/pdp-11/ by clicking on "Micronotes"). Yes, the 11/53 is a KDJ11-D. 93's and 94's are KDJ11-E's. And lots of other systems. Some DEC peripherals (most noticably the early HSC storage controllers for VAXclusters) have J11's, several Xerox laserprinters used J11's, DEC PRO380's used J11's. Many third-party CPU boards use J11's, it's not unusual to see them scrounging the used market for HSC's to strip the J11 from, as the HSC's generally had late-rev J11's. (And Harris hasn't made the J11 chips for many years now.) Never sold as a "PDP-11" system, though the chip does implement the basic PDP-11 instruction set (and some of the add-ons.) It was sold by DEC in the KXT11-CA single board computer, which had a T11, 32K RAM, up to 32K of EPROM, 3 serial lines, some parallel I/O, and a Q-bus interface. For more information see Micronote #16, _KXT11-CA Development Tools_ Micronote #18, _KXT11-CA DMA Programming_ Micronote #32, _KXT11-CA Parallel I/O_ Micronote #34, _Programming KXT-11C Multi SLU_ You can find T11 chips in several Q-bus and Unibus peripherals, most notably the RQDX1, 2, and 3 (the chip labeled "27-17311-01"). aka "Fonz-11", the CPU chipset used in 11/23's, 11/24's, the lower-end DEC PRO's, etc. The CPU chipset used in the LSI-11/02 and /03 is a Western Digital chipset, and the same set was used (with different microcode) by other CPU makers. In particular, the Alpha Micro two-board S-100 set. -- Tim Shoppa Email: shoppa(a)trailing-edge.com Trailing Edge Technology WWW: http://www.trailing-edge.com/ 7328 Bradley Blvd Voice: 301-767-5917 Bethesda, MD, USA 20817 Fax: 301-767-5927 Received: (from major@localhost) by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id EAA79749 for pups-liszt; Thu, 11 May 2000 04:04:13 +1000 (EST)