Sun-Spots-Request@RICE.EDU (William LeFebvre) (07/20/88)
SUN-SPOTS DIGEST Tuesday, 19 July 1988 Volume 6 : Issue 146 Today's Topics: Re: Problems with SUN-Link X.25 Re: Problems with SPARC Assembler Directives & KCL Re: remote dumps Re: SUN3/50 SCSI Disks Re: 15-pin ethernet connectors bug fix for RCS under sparc fix for SCSI support under SunOS 4.0 fix for "find" under SunOS 4.0 NeWS for GNUs Questions about Kic and MAGIC for Sun Idledaemon for the Sun? SunOS systerm parameters? Graph Package? Scratchpad for SunTools? Send contributions to: sun-spots@rice.edu Send subscription add/delete requests to: sun-spots-request@rice.edu Bitnet readers can subscribe directly with the CMS command: TELL LISTSERV AT RICE SUBSCRIBE SUNSPOTS My Full Name Recent backissues are available via anonymous FTP from "titan.rice.edu". For volume X, issue Y, "get sun-spots/vXnY". They are also accessible through the archive server: mail the request "send sun-spots vXnY" to "archive-server@rice.edu" or mail the word "help" to the same address for more information. ---------------------------------------------------------------------- Date: Sun, 10 Jul 88 8:21 BST From: Piete Brooks <pb%computer-lab.cambridge.ac.uk@nss.cs.ucl.ac.uk> Subject: Re: Problems with SUN-Link X.25 Reference: v6n132 > [[ Well... How can we get a copy? --wnl ]] A UK version is available from the UK distributed info server (sorry -- only available to UK sites). An earlier version is on the UKSUG tape. There has been a request from John Gilmore for a copy for the next SUG tape and a copy to be deposited with Rick Adams. However, the sources are currently somewhat UK biased, in that: a) it conforms to the local guidelines for the use of X.29 (the "Green Book") [ plus one copy being used over TYMNET ] b) it has code to support X.29 over a transport service (TS29) c) spad has support for a totally brain damaged Q-bus interface (DEXPAND) which significantly impairs the low level polling (select isn't supported) It also could do with some tidying WRT the TELNET interface (hard as I do not really understand telnet and we do not have any "intelligent" interfaces to test it against) and the use of my "X.25 over TCP" encapsulation routines which need to be integrated into spad [it has inline code ! :-(]. I am seriously contemplating breaking one of my golden rules and providing TWO different sources [it's amazing what RCS can do :-)], the scrungy UK copy and a pure X.29 copy. So, if anyone is intersted in a PRE-RELEASE version on the conditions that: a) comments to pb@cl.cam.ac.uk (even if nothing not comment on!) b) flames to /dev/null c) know about other flavours of X.29 / know about telnet / are desparate d) will delete pre-release copy when main release is available. e) willing to forward/make available (on my request) to other testers feel free to email me for a copy [ I am away 15-25 Jul and 11-25 Aug ] PS: to get x29d to work in "message mode" (forward on completed line), it is necessary to have a mode wherby echoing is performed at the client PAD rather than by the tty or pty (i.e. half-duplex). To achive this I distribute a version of tty_pty which allows a Remote Echo mode (binaries & source patches for SUN 3.x and ultrix 1.2 -- others if wanted). I know that sun distribute a new tty_pty.o with SunLink, presumably to achieve the same goal -- does anyone know how they do it, i.e. what IOCTL call should I make and what info do I get ? ------------------------------ Date: Sun, 10 Jul 88 21:34:14 PDT From: Craig Leres <leres@helios.ee.lbl.gov> Subject: Re: Problems with SPARC Assembler Directives & KCL I received a "Sun-4 Assembly Language Reference Manual" (part number 800-2053-10, revision A, November 30, 1987) under a seperate cover. It seems to have the information you're looking for. Taking a stab at converting the code you posted: > .data > .even > . = . + 1024 > .globl _character_table > _character_table: > .space 1024 I came up with: .seg "data" .align 2 .skip 1024 .global _character_table _character_table: .skip 1024 The mnemonic .byte takes a list of 8 bit initializer values, so I don't think you want to use it in place of .space. There are also .half (16 bit) and .word (32 bit). I went to session on sparc at the SUG conference in San Jose last year. Several of us were promised sparc processor handbooks, but after waiting several months, I got the stupid session slides instead. If any one knows how to get a processor handbook, please let the rest of us in on the secret. As usual, adb is helpful in figuring out addressing modes, but I'd rather not have to do so much guessing. Craig ------------------------------ Date: Sun, 10 Jul 88 20:11:33 EDT From: vic@fine.Princeton.EDU Subject: Re: remote dumps The default blocking factor for dumps under SUN-OS is in fact 10k, the manual pages for dump were copied from BSD, and have NOT been kept up to date ( THIS IS ANNOYING ). Despite cliams to the contrary in the man page, the b option to dump specifies 512b blocks, and the default size is 20 of these i.e. 10 tape blocks. The magic 126 makes much more sense when looked at as 63k, the largest multiple of 1K that allows the byte count to fit into a short. Hope this helps. V. Duchovni <vic@fine.princeton.edu> Mathematics Dept. Princeton University. ------------------------------ Date: Mon, 11 Jul 88 10:33:50 BST From: Paul Hudson <mcvax!moncam!paul@uunet.uu.net> Subject: Re: SUN3/50 SCSI Disks We've successfully run WREN-IV (~300Mb) discs on a 3/50 - my main machine runs one now. Nice, quiet fast drives. Paul Hudson Snail mail: Monotype ADG Email: ...!ukc!acorn!moncam!paul Science Park, paul@moncam.co.uk Milton Road, Cambridge, CB4 4FQ ------------------------------ Date: 10 Jul 88 15:01:13 GMT From: phri!roy@philabs.philips.com (Roy Smith) Subject: Re: 15-pin ethernet connectors In v6n132, Bob Drzyzgula asks: > could someone please tell me *why* did they specify them in the first place I've often wondered about that. The best answer I can come up with is that some marketing bozo decided that they needed to be able to say, "it's so easy to install, you don't even need a screwdriver!". As for why they continue to use them, I guess it's a slavish need to follow standards, even when the standards suck. Given that attitude, however, I'm at a loss to explain the various and sundry flavors of serial ports which claim to be RS-232C. BTW, I happen to like the connector on a 3-Com 3C503 ethernet card for the IBM-PC. It's got the slide-lock gizmo to keep the standard-bearers happy, but also has an extra optional screw-on adaptor plate which you can add to increase the security of the connection. The only bad part is that while they supply the mounting holes for the extra plate, the don't supply the plate itself. I don't know why they leave it out; it probably only costs about $0.50 to make. Roy Smith, System Administrator Public Health Research Institute {allegra,philabs,cmcl2,rutgers}!phri!roy -or- phri!roy@uunet.uu.net ------------------------------ Date: Sun, 10 Jul 88 13:50:01 PDT From: Craig Leres <leres@helios.ee.lbl.gov> Subject: bug fix for RCS under sparc When I compiled my copy of RCS (version 3) on a Sun 4 (running 3.2), I discovered it has its own fprintf() and that it isn't portable to sparc (no big surprise). There were ifdefs for VFPRINTF but that code was broken. Appended are the diffs that make RCS use varargs(3) when VFPRINTF is defined. Note that this new code works just fine on a 68020 too. Craig RCS file: RCS/rcslex.c,v retrieving revision 3.4 diff -c -r3.4 rcslex.c *** /tmp/,RCSt1a06721 Sun Jul 10 13:18:56 1988 --- rcslex.c Sun Jul 10 13:18:14 1988 *************** *** 2,8 **** * RCS file input */ static char rcsid[]= ! "$Header: rcslex.c,v 3.4 86/09/23 04:09:11 van Exp $ Purdue CS"; /********************************************************************************* * Lexical Analysis. * Character mapping table, --- 2,8 ---- * RCS file input */ static char rcsid[]= ! "$Header: rcslex.c,v 3.5 88/07/10 13:17:46 leres Exp $ Purdue CS"; /********************************************************************************* * Lexical Analysis. * Character mapping table, *************** *** 25,30 **** --- 25,33 ---- */ /* $Log: rcslex.c,v $ + * Revision 3.5 88/07/10 13:17:46 leres + * Fix VFPRINTF code to use varargs(3). + * * Revision 3.4 86/09/23 04:09:11 van * portability / lint fixes from guy@sun, aug86 * *************** *** 57,64 **** --- 60,71 ---- #include "rcsbase.h" + #ifdef VFPRINTF + #include <varargs.h> + #endif + /* character mapping table */ enum tokens map[] = { EOFILE, /* this will end up at ctab[-1] */ *************** *** 617,631 **** fprintf(iop, fmt, args) FILE *iop; char *fmt; /* Function: formatted output. Same as fprintf in stdio, * but aborts program on error */ { #ifdef VFPRINTF ! VOID vfprintf(iop, fmt, &args); #else _doprnt(fmt, &args, iop); #endif VFPRINTF --- 624,652 ---- + #ifdef VFPRINTF + fprintf(va_alist) + va_dcl + #else fprintf(iop, fmt, args) FILE *iop; char *fmt; + #endif /* Function: formatted output. Same as fprintf in stdio, * but aborts program on error */ { #ifdef VFPRINTF ! FILE *iop; ! char *fmt; ! va_list ap; ! ! va_start(ap); ! iop = va_arg(ap, FILE *); ! fmt = va_arg(ap, char *); ! ! VOID vfprintf(iop, fmt, ap); ! va_end(ap); #else _doprnt(fmt, &args, iop); #endif VFPRINTF ------------------------------ Date: Sun, 10 Jul 88 20:42:37 PDT From: Craig Leres <leres@helios.ee.lbl.gov> Subject: fix for SCSI support under SunOS 4.0 As you'll remember, I put 4.0 up on a Sun 4 the other day. When it came time to configure a kernel with support for only the devices I had (a scsi disk and tape), I got undefined symbols. This particular problem is because the include file for the sw scsi controller is in the wrong place. As a result, kernels that are configured with sw and without sc, si, or se have some "important" routines ifdef'ed out. Diffs to the /sys/sundev/sc_conf.c are appended. Craig ------ *** /tmp/,RCSt1a03710 Sun Jul 10 20:41:45 1988 --- sc_conf.c Sun Jul 10 20:38:58 1988 *************** *** 11,16 **** --- 11,19 ---- #include "sc.h" #include "si.h" #include "se.h" + #ifdef REL4 + #include "sw.h" /* Sun 4/110 support */ + #endif #if ((NSC > 0) || (NSI > 0) || (NSE > 0) || (NSW > 0) || (NWDS > 0)) *************** *** 24,30 **** #include "../sundev/scsi.h" #else REL4 - #include "sw.h" /* Sun 4/110 support */ #include "wds.h" /* Sun ECD support */ #include <sys/types.h> --- 27,32 ---- ------------------------------ Date: Sun, 10 Jul 88 20:25:46 PDT From: Craig Leres <leres@helios.ee.lbl.gov> Subject: fix for "find" under SunOS 4.0 The bugs just seem to come out of the woodwork when you start using a new architecture, don't they? If I didn't know better, I'd think sparc was designed to find bugs in vax-developed code. After putting up 4.0 on a 4/110 and running /usr/lib/updatedb, I discovered that "find name" didn't work. Since I don't have 4.0 source (taunt, taunt) I compiled the 3.5 find source. It failed too. The problem turned out to be an uninitialized local variable in the routine fastfind(). Diffs to the 3.5 find are appended. Craig ------ RCS file: RCS/find.c,v retrieving revision 1.1 diff -c -r1.1 find.c *** /tmp/,RCSt1a08078 Sun Jul 10 20:20:36 1988 --- find.c Sun Jul 10 20:03:46 1988 *************** *** 1,4 **** --- 1,6 ---- #ifndef lint + static char rcsid[] = + "@(#) $Header: find.c,v 1.2 88/07/10 20:02:57 leres Exp $ (LBL)"; static char sccsid[] = "@(#)find.c 1.1 87/12/21 SMI"; /* from S5R2 4.7 */ #endif *************** *** 1203,1208 **** --- 1205,1212 ---- if ( strchr ( pathpart, '*' ) || strchr ( pathpart, '?' ) || strchr ( pathpart, '[' ) ) globflag = YES; + else + globflag = NO; patend = patprep ( pathpart ); c = getc ( fp ); ------------------------------ Date: Sun, 10 Jul 88 18:32:46 -0400 From: Henry B.J. Krempel <krempel@pacrat.npac.syr.edu> Subject: NeWS for GNUs There is a nice NeWS interface for GNU-emacs available from Chris Maio at columbia. It's available on the public ftp area at columbia.edu in a file called ps-emacs.tar. It has mouse support, and a configurable menu, amongst other things. It would be nice if someday the FSF would include this NeWS support in with its bag of supported proprietary software (VMS, Suntools), oh well. Henry B. J. Krempel <krempel@pacrat.npac.syr.edu> Northeast Parallel Architectures Center (NPAC) Syracuse University 250 Machinery Hall Syracuse, N.Y. 13244 ------------------------------ Date: Mon, 11 Jul 88 17:32:52 EDT From: Yoshihito Koya <yk@china.ctr.columbia.edu> Subject: Questions about Kic and MAGIC for Sun I am an undergraduate student working at Mixed Analog vlsi lab at Columbia university. I have two problems. First, I have an assignment to bring up an integrated circuit layout program called Kic, which is developed by UC Berkeley, on our SUN computer. But the trouble is that the program does not have the driver for the SUN terminal. I would appreciate if someone can tell me either how can I modify Kic terminal drivers so that it will work on Sun3/60, or if such a driver exists then how to aquire it. Second, I am also interested in having MAGIC, another layout editor, working on Sun 3/60 in color. Because Sun 3/60 has two layers of modes, B/W and color, MAGIC (it only has drivers for Sun2/120, by the way) chooses B/W mode instead of color mode which seems to have less priority. Again, I like to know how can I modify the driver to work on Sun3/60 or does such a driver exist. If you know the solution to any of the above problems, please reply through e-mail as soon as possible. I am in most desparate need for it. My account is: yk@greece.ctr.columbia.edu Thank you for your cooperation. Yoshi Koya Address: Att. Yoshihito Koya 1220 S.W. Mudd Building Columbia University / CTR New York, NY 10027 Telephone: (212) 280 - 4685 ------------------------------ Date: Mon, 11 Jul 88 08:23:51 edt From: mlijews@nswc-wo.arpa Subject: Idledaemon for the Sun? Does anyone have a version of idledaemon, or a similar program which will log off idle users of tty ports, for the Sun? I am running SunOS 3.3 and would much appreciate getting such a program if you have already gotten it runnning, before I start hacking idledaemon. Mike ------------------------------ Date: 10 Jul 88 14:59:56 GMT From: fortin@tut.cis.ohio-state.edu (Micheal Fortin) Subject: SunOS systerm parameters? Recently I have been trying to collect together a list of system parameters and their descriptions. As it turns out, looking thru the manuals is not the easiest way to do this. Does anyone out there know where I could get a rather complete list, or have a list that you could send to me??? thanks in advance mike ------------------------------ Date: Mon, 11 Jul 88 12:38:16 -0200 From: Yaakov Itzikowitz <yi@WISDOM.BITNET> Subject: Graph Package? Would appreciate information/experience about a graph package that run on SUN/3, 3-D for simple geometric shapes, can have database of shapes, combinations of shapes will form a specific set. Set shapes will produce necessary inventory. Input devices can be graph editor, ocr, digitizer. Ability to laser print a constructed set. Any idea of the availability of such a package ? Thanks. Yaakov ------------------------------ Date: Mon, 11 Jul 88 12:46:23 EDT From: bernhold@qtp.ufl.edu Subject: Scratchpad for SunTools? Before I try to reinvent the wheel, does anyone have a "scratchpad" tool which is similar to the Macintosh desk accessory? Just a window in which to write notes to yourself & such. We are running SunView/SunTools. [[ Why not just pop up another "textedit"? The joys of multitasking. --wnl ]] Thanks in advance... Dave Bernholdt bernhold@qtp.ufl.edu bernhold@ufpine.bitnet ------------------------------ End of SUN-Spots Digest ***********************