Drives in HDToolBox not detected full size (CHS settings?)

[LoadWB]

I just recalled that there may be another punch line here: the drive may report its size differently with respect to how it is asked.

The ATA scsi.device may receive an answer fit for a disk controller as present in an early 2000's PC. Back then the size of a sector was 512 bytes, and this is no longer the case with today's SSDs. Also, the number of sectors per track, etc. was much lower than what we use today.

Hence, the SSD may report its properties skewed towards "compatibility".

hrmmmmm maybe setting the sector size to 4096?

Yes, and with larger disks comes the unavoidable disk errors, bit-flips and so on, that needs to be handled somehow. And TRIM, UNMAP functionality for SSDs... soon enough drives will be replaced completely by various NVM solutions etc, interesting time ahead

TRIM support for modern SSDs would be nice.  In the beginning I recall some SSDs had what was essentially auto-TRIM to support OSes and filesystems which did not support TRIM, and I suspect that IDE SSDs, with their target being older systems, do as well.

[SnkBitten]

Have you tried configuring the drive under WinUAE on a fully updated/patched OS 3.9 setup?

Version reports HDToolBox 45.6 (2/13/2002) that I had used.  

Maybe install the drive under that and see if they show up correctly.

I also am using PFS_AIO instead of FFS.
http://aminet.net/package/disk/misc/pfs3aio

I did my drive setups that way, then used a "fresh" WinUAE config to install the OS's (3.1/3.9).

I had a updated and patched WinUAE build of OS 3.9, and just added the drives via an external USB cable with SATA and IDE connections + power.

[LoadWB]

Per my post in another thread, I am deliberately not using WinUAE for this project.  I have never used it, anyway, so I would have to work that up, too.  My intention is to be able to use the native tools to achieve my goals, illustrating short-falls and solutions as I go.  It's a challenge, and I'm already "cheating" by using a 3.9 boot set

[LoadWB]

Okay, I attached what I'm seeing with the 128GB SSD and what happens if I accept the default parameters.  If I change the block size to 4096 it only comes up to 88GB.

EDIT: If I set cylinders to 65536 and blocks per cylinder to 4096 I get 128GB exactly, but that is more than the detected size of the device.

[kolla]

128 GB or 128 GiB?

https://www.google.com/#q=convert+128+GB+GiB

[LoadWB]

Either way, the number in detection does not match the number in installation, and I would like to think the program is consistent.

[Thomas]

Either way, the number in detection does not match the number in installation, and I would like to think the program is consistent.

It is. Find out the true total number of blocks and increase the number of cylinders until the displayed number of blocks matches the true one as close as possible. In order to keep the number of cylinders small I would also change the number of heads to 16 so that you get 4096 blocks per cylinder.

You can use this program to find out the true total number of blocks: http://thomas-rapp.homepage.t-online.de/downloads/hddreport.lha

[LoadWB]

It is. Find out the true total number of blocks and increase the number of cylinders until the displayed number of blocks matches the true one as close as possible. In order to keep the number of cylinders small I would also change the number of heads to 16 so that you get 4096 blocks per cylinder.

You can use this program to find out the true total number of blocks: http://thomas-rapp.homepage.t-online.de/downloads/hddreport.lha

Nice.  And the work begins.

[kolla]

Well, it is after all an SSD, not a spinning disk :p

[SnkBitten]

Nice.  And the work begins.

Your photo looks like SCSI.device is 43.35 and not the latest patched 43.45

[Thomas]

30333 cylinders * 16 heads * 256 sectors = 124243968 sectors

Matches exactly, no wasted sectors at the end. And partitions aligned to 2-megabyte boundaries should meet all alignment requirements.

[LoadWB]

Your photo looks like SCSI.device is 43.35 and not the latest patched 43.45

Yeah... something got screwy there.  I'm going to have to try again.

30333 cylinders * 16 heads * 256 sectors = 124243968 sectors

Matches exactly, no wasted sectors at the end. And partitions aligned to 2-megabyte boundaries should meet all alignment requirements.

What's the blocks-per-track setting need to be?  I found some parameters change spontaneously as I'm entering.  Eh.  I'm done for the year.  Going home, leaving this at the office and I'll be back on Tuesday.  Going home to my working CSMKIII SCSI system  Thanks for the help, I'll get back on it after New Year.

[Thomas]

"heads" = tracks per cylinder
"sectors" = sectors per track
sectors per cylinder = heads * sectors

So:

Cylinders = 330333
Heads = 16
Sectors per Track = 256
Sectors per Cylinder = 4096

[nyteschayde]

For what it's worth, and for those using OS 4 on any of their machines, the boot floppy has a version of scsi.device that is >52. I use this on all my classic machines.

Version >NIL: scsi.device 52
If Warn
  LoadResident DEVS:A1200/scsi.device reboot quiet
EndIf

Suddenly, the second channel on the internal IDE adapter and other nice little benefits are usable on the system.

[LoadWB]

"heads" = tracks per cylinder
"sectors" = sectors per track
sectors per cylinder = heads * sectors

So:

Cylinders = 330333
Heads = 16
Sectors per Track = 256
Sectors per Cylinder = 4096

*forehead slap* Oh, duh.  I tried a little bit different values and am having problems.  My 3.9 installation is missing some stuff, and the 3.1 installation said (never did before) that my 2GB partition didn't have enough room for Fonts.  I'll try again Tuesday.

For what it's worth, and for those using OS 4 on any of their machines, the boot floppy has a version of scsi.device that is >52. I use this on all my classic machines.

Version >NIL: scsi.device 52
If Warn
  LoadResident DEVS:A1200/scsi.device reboot quiet
EndIf

Suddenly, the second channel on the internal IDE adapter and other nice little benefits are usable on the system.

Yeah, I completely forgot that I had done some customization to my boot floppy but then re-made it.  I had the patched scsi.device on there, but now I am curious about the OS4.1 scsi.device, and particularly surprised it is 68k. But then I guess the system has to boot 68k then hand over to PPC, right?

I'm assuming one of the benefits is less confusing size issues, but what others?

[Thomas]

My 3.9 installation is missing some stuff

More details needed.

the 3.1 installation said (never did before) that my 2GB partition didn't have enough room for Fonts.

Then the partition is bigger than 2GB. For 3.1 everything above 2 GB is negative. Put some big files on the partition so that free space shrinks below 2 GB. After installation these files can be deleted.

I'm assuming one of the benefits is less confusing size issues, but what others?

A new scsi.device does not fix bugs in HDToolbox and other programs, therefore size issues won't change.