To: 386users@TWG.COM Subject: 80386 mailing list, vol 4 #15 Date: 6 Mar 89 09:42:40 EST (Mon) From: "Wm E. Davidsen" 80386 User's mailing list vol 4 #15 Mar 6, 1989 In this issue: PS 2/80 Dual Async Adapter/A problems multitasking/multiuser on 80386 MIPS Magazine Bell Tech serial card comment American Digicom motherboards - request for info 386 Motherboards, comments wanted Re: multitasking/multiuser on 80386 Re: 386 cpu detection Wyse/Amdek 386 compatibility The addresses for the list are now: 386users@TWG.COM - for contributions to the list or ...!uunet!TWG.COM!386users 386users-request@TWG.COM - for administrivia or ...!uunet!TWG.COM!386users-request P L E A S E N O T E If you want to get on or off the list, or change your address, please mail to the 386users-request address, or the message will be delayed by having to hand forward it (for your convenience, not mine). ---------------------------------------------------------------------- Date: Tue, 28 Feb 89 11:24:18 -0500 (EST) From: "Curtis P. Yeske" Subject: PS 2/80 Dual Async Adapter/A problems Here at GSIA we are experiencing a strange problem with our PS 2/80's. We use the IBM PS/2 Dual Async Adapter/A cards to print to a HP LaserJet II. At the end of the print we get either a double tab twelve or so spaces, (we can not tell which). This causes the first line of the next print to be shifted to the right. The behavior occurs only from the DOS PRINT and DOS "Copy xxx.xx prn:" commands, (version 3.30). Programs such as Kermit, WP, which print directly to the printer do not have this problem. Has anyone seen this problem before? Suggestions? Thank you for your time and consideration, Curt Yeske Graduate School of Industrial Administration Carnegie Mellon cy13@andrew.cmu.edu (412) 268-3091 ------------------------------ From: barkdoll@cattell.psych.upenn.edu (Ed Barkdoll) Subject: multitasking/multiuser on 80386 Date: 25 Feb 89 14:54:03 GMT I am looking for a multitasking and multiuser (maybe) operating system for the 80386. Some of the systems I'm considering are: VM/386, Unix, DESQview, MS windows/386, OS/2, PC-MOS, Concurrent DOS 386. If you either know about pros/cons of any of these systems (or others) or know of good info sources about them, I'd appreciate your e-mail response. edwin barkdoll [barkdoll@cattell.psych.upenn.edu] ------------------------------ From: tbetz@dasys1.UUCP (Tom Betz) Subject: MIPS Magazine Date: 28 Feb 89 01:01:52 GMT MIPS Magazine I just received the premiere issue of a new magazine, and I'd like to recommend it to all those who miss the old Byte and it's focus on small, state-of-the-art computing. MIPS was created 'to help readers evaluate new alternatives in high-performance personal systems, principally 80386 systems and their challengers, including engineering workstations and the Macintosh II.' It seems as though it might do a pretty good job of this. I particularly like their approach to benchmarking tests, which can be very revealing in simulation of real-world-type situations. The premiere issue features a look at the Everex Step/33 (a 33 MHz 80386 machine), Silicon Graphics' Personal IRIS, the Definicon SP-ARC1 AT SPARC board, the Macintosh SE/30, and a comparison of the various Unixen available for the 80386. Columnists include David Betz (no relation, author of XLISP) on programming; A.G.W. Cameron on Engineering and Scientific applications (with a first column discussing the Sun 386i); Peter Varhol on business applications; and David L. Peltz on graphics. It's a slim volume, but full of good information... IMHO, it's worth a look. I plan to subscribe. Next month: a hands-on evaluation of the NeXT, comparative reviews of >100MB hard disks and high-throughput interfaces, and a comparison of multitasking features in Unix, OS/2, extended DOS, Multifinder and VMS. Available from MIPS Publishing, PO Box 54024, Boulder, CO 80322-4204 - Subscriptions are $19.94/year. -- "Still I sing bonny boys, bonny mad boys, | Tom Betz, 114 Woodworth Bedlam boys are bonny, | Yonkers, NY 10701-2509 For they all go bare, and they live by the air, | (914) 375-1510 And they want nor drink nor money." - Steeleye Span | tbetz@dasys1.UUCP ------------------------------ From: Paul A Vixie Subject: Bell Tech serial card comment Date: Tue, 28 Feb 89 08:42:54 PST I have no experience with the ACE-8; it sounds like it'll do 38Kbaud per port, but I don't know how smart it is. I expect it to be a good card, though. I've used Bell Tech ICC-6 in a 386/24/12 (24MHz CPU, 12MHz IO) running 386/ix. The latest driver in 386/ix (1.0.6) seems broken somehow since as of the last software upgrade the ICC gets wedged once in a while and needs to be rebooted. This is no fault of the ICC, though. Under 386/ix 1.0.5, it was flawless. The ICC-6 will go 19,200 pretty much flat out on all six ports if you can get your CPU to pump the data out fast enough. And it supports RTS/CTS hardware handshake in the standard non-standard full-duplex way, making it compatible with the Telebit Trailblazer (S58=2, S68=255) and letting you have out-of- band flow control (^S and ^Q make it all the way in to gnuemacs, how nice). The ICC has an 80186 and 512KB RAM on it; there's a daemon you start up in /etc/rc2.d that watches the heartbeat on the card -- when the heartbeat goes away, the daemon downloads the card's system image and starts it back up. There's an online tool called "idebug" that looks Just Like DOS DEBUG; with this you can play all kinds of games, including a RESET command that kills the card which causes the daemon to download, etc. (This is what my client does when the card wedges in 386/ix 1.0.6...) Design and construction of the card is well done. It "looks nice". I've also used this card in a 386/20/8 Xenix (pre-V.3-compatible) system. Worked the first time. You can put up to four ICC cards in a system. They need 16-bit slots. The later (current) version of the card has an on-board crystal so it doesn't care about the bus clock (though I imagine syncing to it for a transfer is Not Very Efficient). The back panel of the card has six RJ-45 jacks on it; these are pin-compatible with the 3B2 and you can buy RJ45->DB25 connectors for it just about anywhere. Works with MOD-TAP stuff, too. Nice as their hardware is, Bell Tech's tech support and sales people are too hard to reach. I always had the feeling that they needed more phone lines or maybe a better PBX. So the ACE, as I said, ought to be a pretty good card, based on the ICC's reputation. If anyone has specific information about (and, ideally, direct experience with) the ACE, I'd like to hear about it. Paul Vixie moderator, M/L ------------------------------ From: leech@zeta.cs.unc.edu Subject: American Digicom motherboards - request for info Date: 28 Feb 89 23:31:18 GMT I would be interested in hearing from anyone who is familiar with American Digicom Corp. 386 motherboards (or has other comments about the company). Their latest 25 MHz cached board looks attractive. Thanks! -- Jon Leech (leech@cs.unc.edu) __@/ "The satellite was one of seven launched yesterday by the Soviets." - BBC News ------------------------------ From: neal@polyslo.CalPoly.EDU (Neal Pollack) Subject: 386 Motherboards, comments wanted Date: 1 Mar 89 03:04:52 GMT I, and probably safe to say WE, need a comparision and comments about the quality and bang/buck of the 386 motherboards. I wish to build a 386 next month. I am sure a lot of other net readers are also interested. Who makes the good 386 motherboards and why. Cache versus interleave memory? size of cache? wait states? cost? DTK, AMI, Trillian, who else. Please post so all can learn and benefit. The magazines seem to have no interest in this, but I sure want to make the right choice. If I get mail, I will post also. Neal Pollack pyramid!polyslo!neal pyramid!polyslo!csun!smcnet!neal (213)-452-9351 ------------------------------ From: everett@hpcvlx.HP.COM (Everett Kaser) Subject: Re: multitasking/multiuser on 80386 Date: 27 Feb 89 16:43:36 GMT You might want to check out the second most recent issue of PC Magazine (I think it's volume 8, number 4). Their cover story was multitasking systems and they reviewed most of the ones you're asking about. Everett Kaser !hplabs!hp-pcd!everett everett%hpcvlx@hplabs.hp.com ------------------------------ From: martin@zeus.hf.intel.com (Martin Wilde) Subject: Re: 386 cpu detection Date: 27 Feb 89 17:04:18 GMT > Does anyone know a simple way of detecting whether code is executing >(in real mode) on a 386 vs non-386 (286/86) cpu? I'll summarize responses. >Thanks. >-- >Ray Berry KB7HT uucp: ...{uw-beaver|uiucuxc}tikal!ole!ray CS: 73407,3152 >Seattle Silicon Corp. 3075 112th Ave NE. Bellevue WA 98004 (206) 828 4422 Here is a small program that was presented in IBM PC Technical Journal a while back. It seems to work fine. I have tried it on both 286 Xenix and 386 Unix and it reports the cputype correctly. The source is in Microsoft Assembler and normal C code. I make to warranties or guarantees about the code, just passing it on. #!/bin/sh # shar: Shell Archiver (v1.22) # # Run the following text with /bin/sh to create: # Makefile # cpu.asm # cputype.c # sed 's/^X//' << 'SHAR_EOF' > Makefile && XCC = /bin/cc XASM=/bin/masm XCFLAGS = -O -Ml XASMFLAGS = -Mx X Xall: cpu.o X $(CC) $(CFLAGS) -o cputype cputype.c cpu.o X Xcpu.o: X $(ASM) $(ASMFLAGS) cpu.asm X Xclean: X rm -f cpu.o cputype cputype.o SHAR_EOF chmod 0644 Makefile || echo "restore of Makefile fails" sed 's/^X//' << 'SHAR_EOF' > cpu.asm && X; Function cputype(), for real OR protected mode. Returns (in AX) X; the value 0 (86), 20 (186), 30 (286), 40 (386); negative if protected mode. X X .286p ;enable protected-mode instructions X_text segment byte public 'code' ; use Microsoft C names X assume cs:_text X public _cputype X_cputype proc far X push bp X push sp ;86/186 will push SP-2, X pop ax ;286/386 will push SP X cmp ax,sp X jz not86 ;if equal, SP was pushed X mov ax,10 ;is it 86 or 186 X mov cl,32 ;186 uses count mod32 = 0 X shl ax,cl ;86 shifts 32 so ax = 0 X jnz exit ;non-zero: no shift, so 186 X mov ax,20 ;zero: shifted out all bits X jmp exit Xnot86: pushf ;Test 16 or 23 operand size: X mov ax,sp ;pushed 2 or 4 bytes flags? X popf ;restore SP X inc ax ;restore AX by 2 bytes X inc ax X cmp ax,sp ;did pushf change SP by 2? X jnz is32bit ;if not, then 4 bytes of flags X Xis16bit: sub sp,6 ;Is it 286 or 386 in 16-bit mode? X mov bp,sp ;allocate space for GDT ptr X sgdt qword ptr [bp] ;(use PWORD PTR for MASM5) X add sp,4 ;discard 2 words of GDT pointer X pop ax ;get third word X inc ah ;286 stores -1, 386 0 or 1 X jnz is386 X Xis286: mov ax,30 ;set return value X jmp testprot Xis32bit: db 66H ;16-bit override in 32-bit mode Xis386: mov ax,40 Xtestprot: smsw cx ;Protected? Machine status -> CX X ror cx,1 ;protection bit -> carry flag X jnc exit ;real mode if no carry X neg ax ;protected: return neg value Xexit: pop bp X ret X_cputype endp X_text ends X end SHAR_EOF chmod 0644 cpu.asm || echo "restore of cpu.asm fails" sed 's/^X//' << 'SHAR_EOF' > cputype.c && X/* X * This program determines the cputype. X * The function returns the type. X */ X Xextern short cputype(); X Xmain() X{ X short type; X X type = cputype(); X X switch (type) { X X case 0: printf("cpu is 8086\n"); X break; X X case 20: printf("cpu is 80186\n"); X break; X X case 30: printf("cpu is 80286 in real mode\n"); X break; X X case 40: printf("cpu is 80386 in real mode\n"); X break; X X case -30: printf("cpu is 80286 in protected mode\n"); X break; X X case -40: printf("cpu is 80386 in protected mode\n"); X break; X X default: printf("unknown cpu type!\n"); X break; X } X} SHAR_EOF chmod 0644 cputype.c || echo "restore of cputype.c fails" exit 0 _/_ | ______ __. __ / o ____ | textronix!reed!littlei!zeus:martin / / / <_(_/|_/ (_<__<_/ / < | ------------------------------ From: aland@infmx.UUCP (Dr. Scump) Subject: Wyse/Amdek 386 compatibility Date: 25 Feb 89 06:18:11 GMT I found a local outlet which is selling the Wyse 3216 (386/16 box) for super cheap, and I was wondering: does it *work* for the various flavors of UNIX? I'm most interested in Microport Sys V/386 or Interactive 386/ix, but would also like to hear about SCO XENIX Sys V/386 and whatever other flavors of 386 UNIX are out there. The basic question: is there anybody out there who is running 386 UNIX on one of these boxes? The Wyse is listed in the SCO 2.2.3 manuals, but Wyse says they have their own custom SCO port. I called Microport Tech Support about the machine, and they say it depends on the motherboard version. (BTW, when I called Microport I got right through to an engineer who had a list of "good" & "bad" motherboard versions for the Wyse; I was quite impressed!) I'm a bit wary of relying solely on motherboard version though; I heard they had some major QC problems for awhile, and I'd hate to cross a whole version off my list if they just had 1 bad machine... Anyway, I would appreciate VERY MUCH if anybody running 386 UNIX on one of the Wyse boxes could drop me a quick note with the details (what type & version of UNIX, what motherboard rev, etc.) Please email to the below address and will post a summary. Thanks much! -- Alan S. Denney @ Informix Software, Inc. {pyramid|uunet}!infmx!aland "I want to live! -------------------------------------------- as an honest man, Disclaimer: These opinions are mine alone. to get all I deserve If I am caught or killed, the secretary and to give all I can." will disavow any knowledge of my actions. - S. Vega ------------------------------ End of 80386 M/L ****************