PFS3 or PFS3 SCSI Direct - Page 5

daxb · 31 October 2012, 10:04

Please no new name. Just enhance version number, please (e.g. PFS3.1 V18.6). Plus autor named in version string.

Toni Wilen · 31 October 2012, 11:11

It needs new name, current name is PFS3 for KS1.x (because that was the only point originally, to run on 1.x) and it does not make much sense anymore, it is much more now.

Officially PFS3 = non-direct version. PFS3DS = DS version. My version does both, it must be clear for user which version he/she has installed! (by checking hdtoolbox original filesystem path/name in filesystem page).

This is not a port or recompile of official version, it gets quite bad when/if there is official "18.6" or later version. Just asking for version or date does not mean anything. What if someone else compiles official PFS3 with different compiler options and also increases version number?

From support/maintenance point of view it _must_ have different name.

Bamiga2002 · 31 October 2012, 12:57

How about just PFS3-U as in "Unified"?

Zack · 31 October 2012, 17:53

Another suggestion could be PFS3AIO (As in "All-in-one")?

Toni Wilen · 08 November 2012, 16:45

Ok, I decided to use my original name idea, all-in-one. It is short and boring enough. Ultimate is a bit too final

Uploaded to Aminet.

Toni Wilen · 18 November 2012, 17:23

Quote:

Originally Posted by fgh

Can you test if real drives accept bad CHS values without actually building an adapter?

Very ugly test program attached. Supports A600/A1200 IDE only. Probably good idea to run it from ram disk. Disables interrupts temporarily and pokes IDE registers directly. Better reboot after testing. Source also included.

Output:

Identify device command output

Total LBA, CHS values (Identify words 1, 3 and 6), Another set of CHS values (Identify words 54, 55, 56)

Block zero hex output (Only to confirm that IDE read command works)

Reads CHS=17219/0/1 hex output or error if drive rejected it. This is not yet tested, don't have >8G CF cards. (SD to IDE adapter and 16G SDHC should arrive next week)

Toni Wilen · 20 November 2012, 16:28

SDHC-IDE adapter and 16G SHDC card arrived. Looks exactly same as fgh's except sticker that says "Delock".

idequery output:

Code:

848A78C00000001008000200003F01DB ..x..........?..
74000000202020202020202020202020 t...            
20202020202030200001000200043637       0 ......67
3238313330364D656D6F727920436172 281306Memory Car
64204164617074657220202020202020 d Adapter       
20202020202020202020202020200001               ..
00000300000002000000000778C00010 ............x...
003F740001DB0101740001DB00000007 .?t.....t.......
00030078007800F00078000000000000 ...x.x...x......
00000000000000000000000000000000 ................
007E0600000000000000000000000000 .~..............
001F0000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000001 ................

LBA:  0031159296 
C:  0000030912  H:  0000000016  S:  0000000063 
C:  0000030912  H:  0000000016  S:  0000000063 

C:  0000000000  H:  0000000000  S:  0000000001 
ST:  0000000000 
5244534B00000040E1F5AF2B00000007 RDSK...@...+....
0000020000000007FFFFFFFF00000001 ................
00000002FFFFFFFFFFFFFFFFFFFFFFFF ................
FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF ................
000043EC000001000000000700000001 ..C.............
000043EC000000000000000000000000 ..C.............
000043EC000043EC0000000300000000 ..C...C.........
00000000000000000000000000000000 ................
0000000000000DFF00000002000043EB ..............C.
00000700000000000000003800000000 ...........8....
47656E657269632053442F4D4D432020 Generic SD/MMC  
2020202020202020312E303000000000         1.00....
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

C:  0000017219  H:  0000000000  S:  0000000001 
ST:  0000000000  ERR:  0000000000 
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

Accepts >16383 cylinders.

Still need to test 32G SHDC and >8G non-SD drives.

fgh · 20 November 2012, 17:33

Good work, here you go..

16GB CF-card (a600 with €1 IDE-CF adapter):

Code:

044A3FFF000000107E000200003F01DC .J?.....~....?..
7FC00000202020203542313831453741 ....    5B181E7A
30303032453938430002000200045665 0002E98C......Ve
72362E30324B554C54494D4154452043 r6.02KULTIMATE C
46204341524420313647422020202020 F CARD 16GB
20202020202020202020202020208001               ..
00002B0040000200000000073FFF0010 ..+.@.......?...
003FFC1000FB01017FC001DC00000407 .?..............
00030078007800780078000000000000 ...x.x.x.x......
00000000000000000000000000000000 ................
00300000702B50044000700010044000 .0..p+P.@.p...@.
007F001E001E0000FFFE000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

LBA: 0031227840

C: 0000016383 H: 0000000016 S: 0000000063
C: 0000016383 H: 0000000016 S: 0000000063
C: 0000000000 H: 0000000000 S: 0000000001

ST: 0000000000

5244534B000000406F5E23FA00000007 RDSK...@o^#.....
0000020000000017FFFFFFFF00000001 ................
00000005FFFFFFFFFFFFFFFFFFFFFFFF ................
FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF ................
0000079700003EC20000000100000001 ......>.........
00000797000000000000000000000000 ................
00000797000007970000000300000000 ................
00000000000000000000000000000000 ................
0000000000007D830000000200000796 ......}.........
00003EC2000000000000007A00000000 ..>........z....
47656E657269632D436F6D7061637420 Generic-Compact 
466C617368202020312E303000000000 Flash   1.00....
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

C: 0000017219 H: 0000000000 S: 0000000001
ST: 0000000081 ERR: 0000000016

32GB SD-card (a600 with IDE-SD adapter):

Code:

848AF5B40000001008000200003F03C7 .............?..
78000000202020202020202020202020 x...            
20202020202030200001000200043637       0 ......67
3238313330364D656D6F727920436172 281306Memory Car
64204164617074657220202020202020 d Adapter       
20202020202020202020202020200001               ..
000003000000020000000007F5B40010 ................
003F74C003C70101780003C700000007 .?t.....x.......
00030078007800F00078000000000000 ...x.x...x......
00000000000000000000000000000000 ................
007E0600000000000000000000000000 .~..............
001F0000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000001 ................

LBA: 0063404032

C: 0000062900 H: 0000000016 S: 0000000063
C: 0000062900 H: 0000000016 S: 0000000063
C: 0000000000 H: 0000000000 S: 0000000001

ST: 0000000000

5244534B0000004021203B9E00000007 RDSK...@! ;.....
0000020000000007FFFFFFFF00000001 ................
00000004FFFFFFFFFFFFFFFFFFFFFFFF ................
FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF ................
0000F1DE000001000000000400000001 ................
0000F1DE000000000000000000000000 ................
0000F1DE0000F1DE0000000300000000 ................
00000000000000000000000000000000 ................
00000000000007FF000000020000F1DD ................
0000040000000000000000CA00000000 ................
5472616E73636E643332474220202020 Transcnd32GB
2020202020202020312E303000000000         1.00....
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

C: 0000017219 H: 0000000000 S: 0000000001
ST: 0000000000 ERR: 0000000000

00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

Toni Wilen · 20 November 2012, 17:41

Quote:

Originally Posted by fgh

Code:

C: 0000017219 H: 0000000000 S: 0000000001
ST: 0000000081 ERR: 0000000016

Read command failed

Quote:

Code:

C: 0000017219 H: 0000000000 S: 0000000001
ST: 0000000000 ERR: 0000000000

Works.

Looks like SD adapter is better choice than CF.

fgh · 20 November 2012, 17:54

Ok, so faking CHS values does not work..
It's still cool that you can use a 32GB SD card without patching scsi.device

One last little thing: You said IDE heads register max value is 15, but heads are counted from 0, so this still gives 16 heads in total, doesn't it?
(Meaning a600/1200 scsi.device supports up to ~128GB using 'too large' CHS)

Bamiga2002 · 24 November 2012, 11:45

Just informing that I've used PFS3-AIO for some time now and it works

. And I don't really see a big difference to the original 060-version speedwise.

Turran · 14 December 2012, 18:03

Are we still supposed to use 0x50465303 for identifier with PFS3-AIO? meaning I have to remove the old PFS3 first to be able to add it?

Toni Wilen · 14 December 2012, 19:51

Quote:

Originally Posted by Turran

Are we still supposed to use 0x50465303 for identifier with PFS3-AIO? meaning I have to remove the old PFS3 first to be able to add it?

Use hdtoolbox filesystem update option to replace existing PFS3 with AIO version. Everything else can stay the same.

Turran · 14 December 2012, 21:14

Clever, thanks =)

Although, on a new drive, pointing to the l:PFS3-AIO (yes, renamed it), I can still specify 0x50465303, correct? I dont even need the PFS 18.5 package anymore?

Thats what I did anyway. Works fine so far =) Comes up as version 18.0

Alexco · 10 March 2013, 16:28

Sorry to revive this old thread, but I face some strange problem using any version of PFS3DS (020,040,060) or Tonis AOI.
My Setup is an A1200 with custom Kick3.9 ROM (somewhat older with scsi.device 43.43), a BPPC 060 with 128MB RAM and a Disk on Module 16GB.
The 16GB contains 4 partitions: 900MB FFS (Boot), 100MB FFS (emergency boot), 7GB Data PFS3DS and 6GB Tools PFS3DS.
If I add any direct SCSI version of PFS3 into the RDB, the system won't boot.
The HDD LED blinks shortly followed by Power LED blinking 8x, followed by reset. The only solution is then to de-activate the PFS3DS partitions in early startup (and then remove the filesystem from RDB).

The same thing happens if I try to mount the partitions via "mount" and a mountfile.

Can someone explain where my error is??

RDB looks like this (readRDB Tool output)

Code:

Rigid Disk Block:
   Vendor          : KingSpec
   Product         :  KDM-44VS.2-016G
   Revision        : 1030
   Cylinders       :   20128
   Heads           :       6
   Sectors         :     256
   Bytes per block :     512
   Last disk
   Last LUN
   Last ID
Partitions:
   KDH0: bootable
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 2
      High Cyl    = 1112
      Buffers     = 200
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x1FE00
      Boot prio   = 0
      DOS type    = 0x444F5303
   KDH1: bootable
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 1113
      High Cyl    = 1321
      Buffers     = 80
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x1FE00
      Boot prio   = -5
      DOS type    = 0x444F5303
   KDH2:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 1322
      High Cyl    = 10888
      Buffers     = 80
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x1FE00
      Boot prio   = 0
      DOS type    = 0x50445303
   KDH3:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 10889
      High Cyl    = 20127
      Buffers     = 80
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x1FE00
      Boot prio   = 0
      DOS type    = 0x50445303
File systems:
   0x444F5303  45.15  L:FastFileSystem
   0x50445303  18.5   l:PFS3-060ds

Mountfile created with HDToolBox

Code:

/* RigidDiskBlock. */
KingSpec KDM-44VS.2-016G1030: disk = scsi.device
      Unit = 0
      BytesperBlock = 512
      Cylinders = 20128
      Heads = 6
      BlocksPerTrack = 256
      CylinderBlocks = 1536
      RDBlow = 0 ; RDBhi = 3071
      MinCyl = 2 ; MaxCyl = 20127
      Interleave = 1
      HiLun = TRUE
      HiID = TRUE
      HiDrive = TRUE
      Reselect = TRUE
      Synch = FALSE
      AutoParkSeconds = 0
#

/* Partition. */
KDH2: device = scsi.device
      Unit = 0
      Flags = 0
      /*! ReadOnly = FALSE */
      FileSystem = "l:PFS3_aio-handler"

      GlobVec = -1
      Surfaces = 6
      BlockSize = 512
      BlocksPerTrack = 256
      Reserved = 2
      Interleave = 0
      SectorsPerBlock = 1
      LowCyl = 1322 ; HighCyl = 10888
      PreAlloc = 0
      Buffers = 80
      BufMemType = 0
      DOSType = 0x50445303
      MaxTransfer = 130560
      Mask = 0x7ffffffe
      BootPri = -128
      Mount = 1
      /*! Bootable = FALSE */
      /*# BootBlocks = 0 */
#
/* Partition. */
KDH3: device = scsi.device
      Unit = 0
      Flags = 0
      /*! ReadOnly = FALSE */
      FileSystem = "l:PFS3-60ds"
      Surfaces = 6
      BlockSize = 512
      BlocksPerTrack = 256
      Reserved = 2
      Interleave = 0
      SectorsPerBlock = 1
      LowCyl = 10889 ; HighCyl = 20127
      PreAlloc = 0
      Buffers = 80
      BufMemType = 0
      DOSType = 0x50445303
      MaxTransfer = 130560
      Mask = 0x7ffffffe
      BootPri = -128
      Mount = 1
      /*! Bootable = FALSE */
      /*# BootBlocks = 0 */
#

Toni Wilen · 10 March 2013, 17:06

Quote:

Originally Posted by Alexco

Sorry to revive this old thread, but I face some strange problem using any version of PFS3DS (020,040,060) or Tonis AOI.

Does AIO work if you use PFS3 dostype instead of PDS3? (=NSD will be used, not DirectSCSI)

What does query program attached to this post: http://eab.abime.net/showpost.php?p=851462&postcount=86 output? (may not work with a 68060 with caches enabled)

Alexco · 10 March 2013, 17:35

No, the AOI version always reboots. I tested PFS03,PDS03 and PDS00.

Here is the output of your tool:

Code:

044A3FFF0000000F00000240003F01D7 .J?........@.?..
C0000000565332343431323132313430 ....VS2441212140
30303039202020200002000200003230 0009    ......20
3132313033304B696E6753706563204B 121030KingSpec K
444D2D343456532E322D303136474D53 DM-44VS.2-016GMS
20202020202020202020202020208001               ..
00000F0040000200000000073FFF000F ....@.......?...
003F3C4F00EC0101C00001D700000207 .?<O............
00030078007800780078000000000000 ...x.x.x.x......
00000000000000000000000000000000 ................
03FC0000702974004001700904004001 ....p)t.@.p...@.
007F0000000000000000604F00000000 ..........`O....
0000000000000000C00001D700000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

LBA: 0030916608
C: 0000016383 H: 0000000015 S: 0000000063
C: 0000016383 H: 0000000015 S: 0000000063

C: 0000000000 H: 0000000000 S: 0000000001
ST: 0000000000
5244534B000000400CD125F300000007 RDSK...@..%.....
0000020000000007FFFFFFFF00000001 ................
00000005FFFFFFFFFFFFFFFFFFFFFFFF ................
FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF ................
00004EA0000001000000000600000001 ..N.............
00004EA0000000000000000000000000 ..N.............
00004EA000004EA00000000300000000 ..N...N.........
00000000000000000000000000000000 ................
0000000000000BFF0000000200004E9F ..............N.
00000600000000000000003B00000000 ...........;....
4B696E6753706563204B444D2D343456 KingSpec KDM-44V
532E322D303136473130333000000000 S.2-016G1030....
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................
00000000000000000000000000000000 ................

C: 0000017219 H: 0000000000 S: 0000000001
ST: 0000000081 ERR: 0000000016

Toni Wilen · 11 March 2013, 14:37

Weird. Does it work if you disable PFS3 partitions in early startup, boot without startup-sequence, load later version, for example scsi.device 43.45p (temporarily replacing the rom built-in version), reboot.

Log output looks fine.

Alexco · 11 March 2013, 16:13

Hmm.
Small success. If I load scsi.device 43.45p using PSDLoadModule I can at least mount the partitions using the mountfile and the "mount" command.
Tried that with the AIO and the 60ds handler.
Will try the RDB thing later...

Edit:
Putting the AIO in RDB also works with 43.45p.

So what's the difference between 43.43 and 43.45p? And what about 44.2?

Toni Wilen · 14 March 2013, 12:27

Could you attach or email me RDB dump of your harddrive? (The one that does not work with PFS3DS) At least later hdtoolbox versions have RDB backup option that can be used to dump the RDB. RDB dump does not contain any user data.

I'd like to debug this problem, it probably is scsi.device bug but perhaps it can be worked around in PFS3.

31 October 2012, 17:53	#84
Zack Registered User Join Date: Feb 2004 Location: Valby, Denmark Age: 48 Posts: 90	Another suggestion could be PFS3AIO (As in "All-in-one")? Last edited by Zack; 31 October 2012 at 17:54. Reason: spelling

11 March 2013, 16:13	#99
Alexco Registered User Join Date: Jun 2008 Location: outer space Posts: 60	Hmm. Small success. If I load scsi.device 43.45p using PSDLoadModule I can at least mount the partitions using the mountfile and the "mount" command. Tried that with the AIO and the 60ds handler. Will try the RDB thing later... Edit: Putting the AIO in RDB also works with 43.45p. So what's the difference between 43.43 and 43.45p? And what about 44.2? Last edited by Alexco; 11 March 2013 at 20:36. Reason: Added some infos, typo

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Cant Install PFS3	videofx	support.Hardware	1	07 March 2013 18:14
PFS3 why cant i...	zharn	support.Apps	9	27 January 2013 06:27
PFS3 error: INVALID PFS3 COPY !!! WTF?	keropi	support.Apps	10	18 March 2008 22:30
Pfs3	Hewitson	request.Apps	3	22 December 2007 14:32
Installing PFS3 on 8.5GB SCSI HD	lopos2000	support.Apps	26	27 March 2007 19:31

31 October 2012, 10:04	#81
daxb Registered User Join Date: Oct 2009 Location: Germany Posts: 3,307	Please no new name. Just enhance version number, please (e.g. PFS3.1 V18.6). Plus autor named in version string.

31 October 2012, 11:11	#82
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,545	It needs new name, current name is PFS3 for KS1.x (because that was the only point originally, to run on 1.x) and it does not make much sense anymore, it is much more now. Officially PFS3 = non-direct version. PFS3DS = DS version. My version does both, it must be clear for user which version he/she has installed! (by checking hdtoolbox original filesystem path/name in filesystem page). This is not a port or recompile of official version, it gets quite bad when/if there is official "18.6" or later version. Just asking for version or date does not mean anything. What if someone else compiles official PFS3 with different compiler options and also increases version number? From support/maintenance point of view it _must_ have different name.

31 October 2012, 12:57	#83
Bamiga2002 BlizzardPPC'less Join Date: May 2004 Location: Finland Age: 46 Posts: 3,210	How about just PFS3-U as in "Unified"?

08 November 2012, 16:45	#85
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,545	Ok, I decided to use my original name idea, all-in-one. It is short and boring enough. Ultimate is a bit too final Uploaded to Aminet.

20 November 2012, 17:54	#90
fgh Registered User Join Date: Dec 2010 Location: Norway Posts: 824	Ok, so faking CHS values does not work.. It's still cool that you can use a 32GB SD card without patching scsi.device One last little thing: You said IDE heads register max value is 15, but heads are counted from 0, so this still gives 16 heads in total, doesn't it? (Meaning a600/1200 scsi.device supports up to ~128GB using 'too large' CHS)

24 November 2012, 11:45	#91
Bamiga2002 BlizzardPPC'less Join Date: May 2004 Location: Finland Age: 46 Posts: 3,210	Just informing that I've used PFS3-AIO for some time now and it works . And I don't really see a big difference to the original 060-version speedwise.

14 December 2012, 18:03	#92
Turran Moderator Join Date: May 2012 Location: Stockholm / Sweden Age: 49 Posts: 1,575	Are we still supposed to use 0x50465303 for identifier with PFS3-AIO? meaning I have to remove the old PFS3 first to be able to add it?

14 December 2012, 21:14	#94
Turran Moderator Join Date: May 2012 Location: Stockholm / Sweden Age: 49 Posts: 1,575	Clever, thanks =) Although, on a new drive, pointing to the l:PFS3-AIO (yes, renamed it), I can still specify 0x50465303, correct? I dont even need the PFS 18.5 package anymore? Thats what I did anyway. Works fine so far =) Comes up as version 18.0

11 March 2013, 14:37	#98
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,545	Weird. Does it work if you disable PFS3 partitions in early startup, boot without startup-sequence, load later version, for example scsi.device 43.45p (temporarily replacing the rom built-in version), reboot. Log output looks fine.

14 March 2013, 12:27	#100
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,545	Could you attach or email me RDB dump of your harddrive? (The one that does not work with PFS3DS) At least later hdtoolbox versions have RDB backup option that can be used to dump the RDB. RDB dump does not contain any user data. I'd like to debug this problem, it probably is scsi.device bug but perhaps it can be worked around in PFS3.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)