[Coco] Glenside IDE problem
Gene Heskett
gheskett at wdtv.com
Wed Dec 10 19:53:31 EST 2014
On Wednesday 10 December 2014 14:42:00 Bill Pierce via Coco did opine
And Gene did reply:
> Gene, I "think" but I'm not sure, that the IDE descriptors use STP and
> DNS (they're not needed for IDE) as bitwise flags for several things.
> I think the "physical" drive number is bit 0 of DNS=. If you notice,
> i0 is dns=0 and i1 is dns=1. Why one would move it, I don't know, but
> it may be due to trying to increase the speed of the IDE by doing bit
> comparisons for flags that just need a bit. IDE can only have 2
> drives, so 1 bit would suffice, I want to say I read it in either one
> of the glenside docs or the super driver docs. I may well be wrong on
> this, but I think this is the case. Your scsi drivers would need more
> than a bit since you can access many drives, so you need the whole
> byte So the scsi descs would be set up different.
>
>
> once my drive finishes it's many hour format... I may check out my
> theory and see if that's what's going on with /i1. Oh, and yes, I plan
> to set up a couple of partitions. I had already read the docs on how
> this was done. Making new descriptors is no biggie. I aslo plan to do
> an hdbdos partition, maybe 2 or 3.... And yes, I know how to set the
> hdbdos offsets to get to more partitions. I have a VHD on the emulator
> that has 10 rsdos partitions on one drive. On bootup, a menu runs and
> you select the partition you want. Each time it changes partitions, it
> runs the menu again. This resets anychanges and allows you to go to
> another drive. If you want the menu after you run something in basic,
> just type "RUN"AUTOEXEC.BAS" and you can switch partitions again.
>
>
> This round was to just get the drive going as I've never had an IDE
> drive on the Coco and it's been a learning experience for sure :-) PS.
> It's up to $ABDF on the format :-)
>
>
> Bill Pierce
> "Today is a good day... I woke up" - Ritchie Havens
>
>
> My Music from the Tandy/Radio Shack Color Computer 2 & 3
> https://sites.google.com/site/dabarnstudio/
> Co-Contributor, Co-Editor for CocoPedia
> http://www.cocopedia.com/wiki/index.php/Main_Page
> E-Mail: ooogalapasooo at aol.com
>
>
>
>
> -----Original Message-----
> From: Gene Heskett <gheskett at wdtv.com>
> To: coco <coco at maltedmedia.com>
> Sent: Wed, Dec 10, 2014 2:09 pm
> Subject: Re: [Coco] Glenside IDE problem
>
>
> On Wednesday 10 December 2014 11:15:11 Bill Pierce via Coco did opine
>
> And Gene did reply:
> > I think I've found the problem. There is an error in the "i1_ide.dd"
> > provided on the Nitros9 distros. The "/i1" desc is internally named
> > as "i0" which when loaded in the boot with "i0" makes it clash with
> > i0.
>
> So it is, but I am still concerned about the drv number, I think the i1
> descriptor should be a drv=01.
>
> When its done formatting i0, the do a dir -e both. If you see the same
> . and .. dirs on i1 as you see on i0, then I think I've nailed it.
> OTOH, if you want to "partition" i0 with the descriptors offsets, you
> will want to reformat /i0 to the size you want then set i1 as the same
> drive, but with those offsets from i0's lsn0 in the wpc and ofs
> entries for i1. And if you want to put an HDBDOS partion on it, hold
> back on the size of i1, the roughly 90 megs it takes for 256 of the
> hdbdos vdisks then make an sh.dd that points to the starting sector of
> that partition in wpc & ofs. Then you'll have all 3 (why stop at 3 if
> the drive is big enough?)
> on one drive. If you have a slave drive too, configure it also like
> the above, and a cron driven backup can then keep the slave drive as a
> backup of the main drive. Good insurance IMO.
>
> Don't forget to keep dead tree printouts of the dmode output for each
> of those in case you have to reconstruct things later.
>
> There is one fly in this soup, dmode can change the nam to i00, but
> cannot then fix it back to i0 because the extra char over writes the R
> in RBF. Soooo, you'll have to use ded to restore it.
>
> Next question, whats the rbf.d definition for the length of the device
> name? No. In os9.d M$NAME reservation is only 2 bytes, but since that
> included alphabetical chars, one could us iA to iG, with suitable
> offsets filled in for 8 partitions. But at some point, the drive
> device table would overflow too. That FWIW, is 22 byte or more for
> the 1st DD.SIZ bytes of each entry in the device table, plus 18 more
> bytes for the rest of that drives table entry for $1E bytes per table
> entry.
>
> D.DevTbl is stored at $80 in the direct page, which on my L2 machine
> says it is located at $7500. That would be in the 4th dat slot and
> makes -0 sense. It had better be in block 0 not block 1($2000-3FFF)
> or block 2($4000-5FFF) but in block 3($6000-7FFF).
>
> But I have not found an entry limitation yet.
>
> But that right there blows what I know about Level 3 right out of the
> water. Bill Nobel, is my tracing correct?
>
> From a dmem dump of my L2 DAT record in the DP:
> {t2|08}/DD/NITROS9/3.3.0/6309L2/MODULES/RBF:dmem 0 90 10|dump
>
> Address 0 1 2 3 4 5 6 7 8 9 A B C D E F 0 2 4 6 8 A C E
> -------- ---- ---- ---- ---- ---- ---- ---- ---- ----------------
> 00000000 6C00 0900 0900 0000 0315 1203 00FC 0000 l............|..
>
> I yam confused. Can someone clarify?
>
> Besides, my wood arrived about an hour ago so I've other things to do
> the rest of the day.
>
> > Since I'm not yet ready to use that desc, (I plan to do partitions) I
> > just ommited it from the boot and the problem disappeared. So now
> > either my drive is formatting.... or the Coco is frozen :-) Actually,
> > it just started counting... and with a 4g format, I imagine I've got
> > a long time to wait. I've forgotten how slow format is on a real
> > Coco as I do all my VHD formatting in VCC (overclocked) so it only
> > takes a few seconds to format a 96meg drive :-)
> >
> >
> > Bill Pierce
> > "Today is a good day... I woke up" - Ritchie Havens
> >
> >
> > My Music from the Tandy/Radio Shack Color Computer 2 & 3
> > https://sites.google.com/site/dabarnstudio/
> > Co-Contributor, Co-Editor for CocoPedia
> > http://www.cocopedia.com/wiki/index.php/Main_Page
> > E-Mail: ooogalapasooo at aol.com
> >
> >
> >
> >
> > -----Original Message-----
> > From: Cocodemus <retrocanada76 at gmail.com>
> > To: Bill Pierce <ooogalapasooo at aol.com>; CoCoList for Color Computer
> > Enthusiasts <coco at maltedmedia.com> Sent: Wed, Dec 10, 2014 7:31 am
> > Subject: Re: [Coco] Glenside IDE problem
> >
> >
> > Is the jumper properly set? Ff50 or ff70? What does the descriptor
> > say?
> >
> > Sent from my iPhone
> >
> > > On Dec 9, 2014, at 11:45 PM, Bill Pierce via Coco
> > > <coco at maltedmedia.com>
> >
> > wrote:
> > > I just recently aquired a Glenside IDE board and I'm trying to set
> > > it up. From what I've read in the Super Driver manual, using
> > > rbsuper. llide, and the
> >
> > descriptors is "supposed" to be pretty much automatic.
> >
> > > I know the board and drive are connected properly as I can run
> > > Curtis's old
> >
> > "detect_ide21.b09" util and it idents my drive with all the proper
> > specs.
> >
> > > I have installed rbsuper, llide, i0, and ih in my boot, all from a
> > > fresh repo
> >
> > build.
> >
> > > Any access I try to make to /i0 gives me "Error 235: Bad Name"
> > > This includes format, dir, free ect. The only uitl that recognizes
> > > the drive
> >
> > is dmode which gives proper attr for /i0.
> >
> > > This seems to be a descriptor error and not a drive error.
> > > I have checked header parity and crc for all files and everything
> > > shows good. Does anybody know what I must do to make these drivers
> > > work with my Glenside
> >
> > board?
> >
> > > Is there something I need to change? I know the descriptor should
> > > be set
> >
> > specifically for my drive, but it should still give me a proper
> > device error (244, 214, etc) and not "Error 235: Bad Name".
> >
> > > Thanks :-)
> > >
> > >
> > > Bill Pierce
> > > "Today is a good day... I woke up" - Ritchie Havens
> > >
> > >
> > > My Music from the Tandy/Radio Shack Color Computer 2 & 3
> > > https://sites.google.com/site/dabarnstudio/
> > > Co-Contributor, Co-Editor for CocoPedia
> > > http://www.cocopedia.com/wiki/index.php/Main_Page
> > > E-Mail: ooogalapasooo at aol.com
>
> Cheers, Gene Heskett
Ack kcalc thats only 11,263,744 byte into the drive. That does seem a
little slow. How big is the drive?
Cheers, Gene Heskett
--
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>
US V Castleman, SCOTUS, Mar 2014 is grounds for Impeaching SCOTUS
More information about the Coco
mailing list