ned@pebbles.cad.mcc.com (CME Ned Nowotny) (07/07/89)
I have recently grabbed a copy of the uuencoded and compressed s4diag from the net (Thanks, Thad.) and have a few questions. First, I have succesfully extracted the executable to build a new diagnostics disk and feel pretty comfortable with using it to build a new system disk. However, I would feel more comfortable if other people who have formatted out disks with more than 1024 cylinders would post the executable size and checksums for their s4diag software. Best of all, can anyone tell me the origins of this software? Next, does anyone have a reliable set of instructions for using s4diag in expert mode? Can the original s4diag run a disk surface test on more that 1024 cylinders? In other words, is there any chance that the limit in the original s4diag only applied to formatting disks? (Pretty unlikely, huh?) Now with regard to another comment I have seen. Someone claims that the stock compress program shipped with the PC7300 is compiled to only compress (and uncompress?) using up to 12-bit compression. However, I have had no problem uncompressing anything I have extracted from the net or files that I have compressed at work (on Suns and Vaxen) and tranferred to my machine at home. Was this person mistaken? (For what its worth, it is a good idea to get the compress softwre in the archives anyway. For one thing, you get the source. For another, you get man pages. The only problem I have had with the stock distribution was the need to link zcat to compress. Those with the 3.51 or later version of the 3B1/7300 OS may have had a different experience.) Ned Nowotny, MCC CAD Program, Box 200195, Austin, TX 78720 Ph: (512) 338-3715 ARPA: ned@mcc.com UUCP: ...!cs.utexas.edu!milano!cadillac!ned ------------------------------------------------------------------------------- "We have ways to make you scream." - Intel advertisement in the June 1989 DDJ.
thad@cup.portal.com (Thad P Floryan) (07/09/89)
Ned Nowotny asks several questions about the UNIXPC's diagnostics ... I've used the s4diag I posted to format and test (and put online! :-) both a Miniscribe 3085 (7 heads and 1170 cylinders) and a Maxtor XT2190 (15 heads and 1224 cylinders). I know nothing more about the origins of that specific s4diag than what I originally posted; I suspect it's an unofficial release from AT&T much like the 3.51c and 3.51d "pseudo-Fixdisks." The instructions for s4diag expert mode are detailed on pages 3-24 through 3-30 in the AT&T UNIX PC REFERENCE MANUAL. If someone has a scanner that can read these pages (8-1/2" by 11"), this would be the safest way; I'm not going to trust my typing all the charts and tables. One aspect of this that makes me uncomfortable is that the MANUAL is copyrighted and I don't know the legal aspects of so copying. Any words of wisdom from the net? But NOTE THIS: the MANUAL *IS* out of print; we couldn't even buy one for our UNIXPC SIG ... a person at AT&T in NJ photocopied the entire manual and sent it to us for the SIG's library. Are copyrights valid on out-of-print books? re the question: " Can the original s4diag run a disk surface test on more that 1024 cylinders? In other words, is there any chance that the limit in the original s4diag only applied to formatting disks? (Pretty unlikely, huh?) " NO! I wasted days trying to figure out WTF was wrong with the system I was upgrading with the Miniscribe 3085. The "stock" s4diag permitted specifying 1170 cylinders and gave every indication of doing everything "right", passed the surface check, etc. My installation of the Foundation Set would then bomb when attempting to make a filesystem on the HD. ONLY with the "new" s4diag did everything succeed. I suspect the "stock" s4diag EITHER has the constant "1024" hardcoded somewhere OR is doing modulo 1024 arithmetic in a "for" loop. My "guess" is that after 1024 cylinders, the format wrapped back to cylinder zero and overwrote the sector and cylinder information there. After using the "new" s4diag and reformatting, everything went smoothly (BTW: this assumes one has the WD2010 chip for > 1024 cylinders (and the P5.1 mod for > 8 heads)). And in reply to several others' questions: I believe the "new" s4diag I posted is the same one available from several other sources (such as ICUS and John Milton); if someone would compare the checksums of what I posted with what is available from those two sources, then we'd know! Thad Floryan [ thad@cup.portal.com (OR) ..!sun!portal!cup.portal.com!thad ]
lenny@icus.islp.ny.us (Lenny Tropiano) (07/10/89)
In article <20250@cup.portal.com> thad@cup.portal.com (Thad P Floryan) writes: ... |>I've used the s4diag I posted to format and test (and put online! :-) both a |>Miniscribe 3085 (7 heads and 1170 cylinders) and a Maxtor XT2190 (15 heads |>and 1224 cylinders). I know nothing more about the origins of that specific |>s4diag than what I originally posted; I've used this s4diag (the same one that Thad posted) to format and so surface tests on the 2nd hard drive installed on my UNIX pc. It also allows one to partition it into smaller slices, rather than using the default partitioning which is the entire disk, minus 4000 1K blocks (non- multi-user) and 5000 1K blocks (multi-user) for swap. During the format phase you are prompted for each of the slices, from slice 2 to slice f (/dev/rfp0[01][2-f]). |>I suspect it's an unofficial release from AT&T much like the 3.51c |>and 3.51d "pseudo-Fixdisks." 3.51b and 3.51c were kernel versions never released because of the bugs that existed in them. Those bugs were ironed out by myself and a couple of other *beta* testers. In the forthcoming FIXDISK, the kernel versions and the fixes from them, from 3.51a all the way up to what I have now 3.51dE1 should be included in the kernel that they distribute. As soon as *I* know, or one of the other testers know, what *EXACTLY* will be on the FIXDISKS, and when they will be available, and how to get them, etc, etc, etc... We'll be sure to let the net know! |>The instructions for s4diag expert mode are detailed on pages 3-24 through |>3-30 in the AT&T UNIX PC REFERENCE MANUAL. ... There are some other problems with the s4test/expert mode that are currently being cleaned up in the deep caverns within AT&T. Hopefully these fixes (more specifically cleaning up the problems with the expert mode) will make it to the next FIXDISK releases. Specifically there will be code in there to do formatting, testing, bad block entering of the 2nd drive (if you have one installed), the partitioning code, the 2048 cylinder limit, and the fix for the ambiguity of the "DR" command in the 31-"i>" interactive device test. BTW: if you look in the reference manual DR could mean select which "DRive" or "DataRegister" ... ... |>And in reply to several others' questions: I believe the "new" s4diag I posted |>is the same one available from several other sources (such as ICUS and John |>Milton); if someone would compare the checksums of what I posted with what is |>available from those two sources, then we'd know! |> Yes, they are one in the same... -Lenny -- Lenny Tropiano ICUS Software Systems [w] +1 (516) 589-7930 lenny@icus.islp.ny.us Telex; 154232428 ICUS [h] +1 (516) 968-8576 {ames,talcott,decuac,hombre,pacbell,sbcs}!icus!lenny attmail!icus!lenny ICUS Software Systems -- PO Box 1; Islip Terrace, NY 11752