PFS3AIO and DirectSCSI and large IDE HD

[Doc Mindie]

Hmm... got a weird problem... at least I think it's weird....

Got the 40GB up and running with PFS\3 and a patched scsi.device. However, neither HDInstTools not HDToolBox seems to be able to access more than the first 8 GB... I'm stumped as to what is happening here, as both programmes seems to report to correct Head/Cylinder/Block per Cylinder numbers, as compared to what is on the actual HD itself.

Ideas/questions about what I'm doing (wrong)?

[thomas]

Originally Posted by Doc Mindie

Hmm... got a weird problem... at least I think it's weird....

Got the 40GB up and running with PFS\3 and a patched scsi.device. However, neither HDInstTools not HDToolBox seems to be able to access more than the first 8 GB... I'm stumped as to what is happening here, as both programmes seems to report to correct Head/Cylinder/Block per Cylinder numbers, as compared to what is on the actual HD itself.

Ideas/questions about what I'm doing (wrong)?

16383 * 16 * 63 * 512 = 8,455,200,768 bytes = ~7.87 GB

These values are only listed for backwards compatibility.

The only value which matters is the highest LBA a.k.a. the number of (user) sectors.

HDToolbox should estimate some nice values for cylinders / heads / blocks per track from the LBA. The actual values do not really matter. There are only two rules:

- a cylinder should be so big that the entire RDB area fits into two cylinders.

- And cylinders * heads * blocks per track should be as near as possible to the LBA but must not be higher.

The number of blocks per cylinder calculates as heads * blocks per track.

[Doc Mindie]

LBA 39.070.080

CBS is as you stated, 16383*16*63, 4 heads

I just noticed I was wrong in one regard, it's a 20GB HD, not a 40.... but still... not being able to access more than 8GB seems a bit... strange? Especially when I use the patched SCSI.device and Toni's PFS3AIO, I was positive I should be able to access all of it

[thomas]

It depends on the order in which you installed the hardware and software.

If you first connected the HDD, then initialized it with HDToolbox and scsi.device V40 and later installed scsi.device V43 or above, then HDToolbox created the partition table with what scsi.device V40 could see. Even though scsi.device V43 sees the entire drive, HDToolbox only allows partitions in the area the HDD was initialized with.

If you already have data stored on the drive and don't want to delete it, you can use this program to make the rest of the drive available to HDToolbox: http://aminet.net/package/disk/misc/fixhddsize

Otherwise I'd suggest to reinitialize the HDD while scsi.device V43 or above is running.

[Doc Mindie]

Thank, Thomas.

I've installed LoadModule and Don_Adan' scsi.device (from http://wt.exotica.org.uk/test.html, hacked/patched) on my 3.1 Install disk (obviously the backup-disk, I don't even touch my originals) and will now try to replace HDToolsBox with HDInstTools as well, to see if that helps anything.

As a sidenote... OS3.5/3.9... if I need to skip Setpatch patching scsi.device from there.... the Startup-sequence command should be something similar to: "Setpatch SKIPROMUPDATE scsi.device" is that correct? Or does my memory fail me (again)?

[fgh]

HDToolbox and HDInsttools will both do the job.
OS3.5/9 already patches HDToolbox to support 128GiB drives, but if you still want to try Don Adan's patch the syntax is correct, except you might have to quote 'scsi.device'.

Thomas: Does fixhddsize work specifically for the 7,87GiB scenario, or for any 'too small' RDB? (If I write an image of a full 8GB SD card to a 16GB card, can I use it to expand the RDB?)