[Coco] Glenside IDE problem

Bill Pierce ooogalapasooo at aol.com
Wed Dec 10 14:42:00 EST 2014


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
-- 
"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

-- 
Coco mailing list
Coco at maltedmedia.com
https://pairlist5.pair.net/mailman/listinfo/coco


 


More information about the Coco mailing list