PFS3AIO v3.1 Final - Page 2

supaduper · 12 February 2018, 15:44

Quote:

Originally Posted by Sinphaltimus

I'm slightly confused about the larger than 5GB partition instructions Toni put here..

So just to ask for clarity on this.

If the partition is bigger than 5GB, follow the same instructions as you posted for smaller than 5GB partition then run the original PFSDoctor on it and if no errors I'm good to go? If errors then I need to reformat

OR

Do I run the original PFSDoctor BEFORE doing this change over?

In a word yes, but aways backup your partitions before changing any values or running software, safety First

Sinphaltimus · 12 February 2018, 17:11

Quote:

Originally Posted by supaduper

In a word yes, but aways backup your partitions before changing any values or running software, safety First

Yep absolutely, just wanted to ensure I was interpreting all this info correctly.

Thanks for the affirmation.

Romanujan · 12 February 2018, 17:16

FS-UAE, AmigaOS 4.1FE, HD image with PFS 'attached' to A4000 ID - there is an error visible during booting (see screenshot) - but the disk image seems to be functioning properly.

Otherwise - no problems found.

Toni Wilen · 12 February 2018, 17:30

Quote:

Originally Posted by Romanujan

FS-UAE, AmigaOS 4.1FE, HD image with PFS 'attached' to A4000 ID - there is an error visible during booting (see screenshot) - but the disk image seems to be functioning properly.

Otherwise - no problems found.

How large is the image? (Exactly, in bytes)

Fieldday · 12 February 2018, 18:00

Any chance for implantation the bug fixes to PFS 5.3? Sometimes got also strange errors during copy action with hundred files.
Thank you for your hard work

indigolemon · 12 February 2018, 18:37

Updated, all partitions check out fine (including a 20G and a 24G) - cheers Toni!

Romanujan · 12 February 2018, 19:37

Quote:

Originally Posted by Toni Wilen

How large is the image? (Exactly, in bytes)

3,758,096,896 bytes - but it's a VHD file (I use it to transfer files between my OS 3.x and 4.1 installations). Contains just one partition, which starts from cylinder 2 and ends on cylinder 3583.

Amiga1992 · 12 February 2018, 19:39

Quote:

Originally Posted by Toni Wilen

This version 100% fixes very old (has at least existed since commercial 18.x versions, possibly much longer) bug that causes wrong index block and similar errors when creating lots of new files/dirs quickly. PFSDoctor also reports errors, for example "x reserved blocks not used but allocated".

Trigger condition:
- >=5G partition size ("super index" partition mode)
- Lots of small files created quickly (for example unpacking archive with thousands of small files, for example EmeraldMines_v1.0_CD.lha)

This is exactly what happened to me! And PFSDoctor erased EVERYTHING when "fixing" it.

Will have to upgrade immediately. I guess I have to reformat the disk to be able to use it?

Toni Wilen · 12 February 2018, 19:42

Quote:

3,758,096,896 bytes - but it's a VHD file (I use it to transfer files between my OS 3.x and 4.1 installations). Contains just one partition, which starts from cylinder 2 and ends on cylinder 3583.

Then you need to use some utility (or check logs or use winuae which shows it) to get the actual size. I am quite sure PFS3 reported LBA is larger than max image LBA.

Romanujan · 12 February 2018, 20:53

The WinUAE 3.6.0 (run under Wine) just shows:

Code:

29127/4/63, 7340004/7340033 blocks, 3584.0MB/3584.0MB
RDSK...@........ [5244534B 00000040 A2B28AC9 00000007]

But I have created a hard drive image of 3,758,096,384 bytes (1024 * 1024 * 3584), created 1 partition (PFS3AIO 3.0 test1) filing the whole image (using OS 3.9) - and the result was exactly the same (error message during OS 4.1 FE startup, after booting the partition became fully useable).

Toni Wilen · 12 February 2018, 21:51

It won't be fully usable, few blocks at the end are not available and will cause problems when disk fills (any block level check program probably also complains). Most likely cause is some geometry calculation difference between programs, or another using CHS addressing and another not.

Does the message appear if you mount it in UAE or some SCSI controller?

This safety check hasn't changed and since first pfs3aio.

Romanujan · 12 February 2018, 22:59

On OS 3.9 - I can mount it via uaehf.device, via internal A1200 IDE, via BlizzardPPC SCSI - works all the time, without warnings. Attaching the image to IDE and using scsi.device 52.3 (from OS4 installation CD) also works properly.

On OS 4.1FE - I was using it attached to IDE, and I had this message during the startup. Attached to CyberstormPPC SCSI - and system starts normally, without problems, the disk image seems to be working correctly (so I have the workaround

quite likely I was using earlier PFS3AIO with Cyberstorm SCSI, I just don't remember right now, I don't play with OS 4 much).

It seems this is the IDE+OS4 which causes the problem - when I have some time I will try to test this using WinUAE (maybe the problem is FS-UAE specific).

[EDIT] Reducing partition size within the image also solved the IDE+OS4 error during the startup.

supaduper · 12 February 2018, 23:17

Like Toni said always leave a few blocks off the end then you wont get these problems

Toni Wilen · 13 February 2018, 08:39

It is caused by geometry differences between ROM KS driver and OS4 driver. I guess ROM driver uses CHS geometry to calculate max size (C*H*S = total blocks) which is practically always smaller than actual total LBA (which is used by OS3.5+ or at least 3.9+). Only affects IDE controllers. SCSI always uses real last block.

Easiest option is to initialize the drive under WB 3.1. Or leave few unused megabytes at the end of drive.

This check is one of the main points of PFS3AIO: if block that filesystem thinks is the last block can't be read (error returned or block is outside of 4G "barrier" and TD64/NSD or Direct SCSI is not available and/or returns error): Error is shown and partition is not mounted to protect the data.

thomas · 18 February 2018, 14:12

Quote:

Originally Posted by Toni Wilen

PFS3AIO v3.0 Test #1 (Based on v19 PFS3)

Could you please change the version string to indicate aio somewhere?

Toni Wilen · 18 February 2018, 15:51

Quote:

Originally Posted by thomas

Could you please change the version string to indicate aio somewhere?

Do you mean $VER: string inside binary?

thomas · 18 February 2018, 16:08

Yes.

Toni Wilen · 19 February 2018, 19:56

Test #2 released. First post updated.

kgc210 · 19 February 2018, 20:12

Do you still have to select DOS type ie:
PFS\03 = 0x50465303 or PDS\03 = 0x50445303

Or is it fully automatic now?

kolla · 21 February 2018, 03:27

When using FFS, one nice feature is that one can run "version" against the device...

Code:

version full df0:
filesystem 45.16 (11.Apr.2010)

Would be cool if this was more commonly implemented

12 February 2018, 17:16	#23
Romanujan Registered User Join Date: Dec 2007 Location: Szczecin/Poland Posts: 424	FS-UAE, AmigaOS 4.1FE, HD image with PFS 'attached' to A4000 ID - there is an error visible during booting (see screenshot) - but the disk image seems to be functioning properly. Otherwise - no problems found. Attached Thumbnails

12 February 2018, 20:53	#30
Romanujan Registered User Join Date: Dec 2007 Location: Szczecin/Poland Posts: 424	The WinUAE 3.6.0 (run under Wine) just shows: Code: 29127/4/63, 7340004/7340033 blocks, 3584.0MB/3584.0MB RDSK...@........ [5244534B 00000040 A2B28AC9 00000007] But I have created a hard drive image of 3,758,096,384 bytes (1024 * 1024 * 3584), created 1 partition (PFS3AIO 3.0 test1) filing the whole image (using OS 3.9) - and the result was exactly the same (error message during OS 4.1 FE startup, after booting the partition became fully useable).

21 February 2018, 03:27	#40
kolla Banned Join Date: Nov 2007 Location: Trondheim, Norway Posts: 1,893	When using FFS, one nice feature is that one can run "version" against the device... Code: version full df0: filesystem 45.16 (11.Apr.2010) Would be cool if this was more commonly implemented

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
PFS3aio vs SFS/02	White	support.WinUAE	10	13 April 2018 21:44
PFS3AIO Probs ?	Foul	support.FS-UAE	0	21 February 2018 22:49
PFS3AIO Questions	manic23	support.Hardware	26	22 April 2014 13:49
Pfs3aio ????	stingray76	support.Hardware	2	19 February 2014 09:15
FinalWriter 97; FinalCalc 1.04; Final Data 2.00 & Final Copy	aebrown-u	request.Apps	20	12 July 2009 11:19

12 February 2018, 18:00	#25
Fieldday Registered User Join Date: May 2009 Location: Germany Posts: 87	Any chance for implantation the bug fixes to PFS 5.3? Sometimes got also strange errors during copy action with hundred files. Thank you for your hard work

12 February 2018, 18:37	#26
indigolemon Bit Copying Bard Join Date: Jan 2017 Location: Kelty, Fife, Scotland Age: 41 Posts: 1,293	Updated, all partitions check out fine (including a 20G and a 24G) - cheers Toni!

12 February 2018, 21:51	#31
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,505	It won't be fully usable, few blocks at the end are not available and will cause problems when disk fills (any block level check program probably also complains). Most likely cause is some geometry calculation difference between programs, or another using CHS addressing and another not. Does the message appear if you mount it in UAE or some SCSI controller? This safety check hasn't changed and since first pfs3aio.

12 February 2018, 22:59	#32
Romanujan Registered User Join Date: Dec 2007 Location: Szczecin/Poland Posts: 424	On OS 3.9 - I can mount it via uaehf.device, via internal A1200 IDE, via BlizzardPPC SCSI - works all the time, without warnings. Attaching the image to IDE and using scsi.device 52.3 (from OS4 installation CD) also works properly. On OS 4.1FE - I was using it attached to IDE, and I had this message during the startup. Attached to CyberstormPPC SCSI - and system starts normally, without problems, the disk image seems to be working correctly (so I have the workaround quite likely I was using earlier PFS3AIO with Cyberstorm SCSI, I just don't remember right now, I don't play with OS 4 much). It seems this is the IDE+OS4 which causes the problem - when I have some time I will try to test this using WinUAE (maybe the problem is FS-UAE specific). [EDIT] Reducing partition size within the image also solved the IDE+OS4 error during the startup.

12 February 2018, 23:17	#33
supaduper Registered User Join Date: Oct 2012 Location: Krypton Posts: 1,210	Like Toni said always leave a few blocks off the end then you wont get these problems

13 February 2018, 08:39	#34
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,505	It is caused by geometry differences between ROM KS driver and OS4 driver. I guess ROM driver uses CHS geometry to calculate max size (CHS = total blocks) which is practically always smaller than actual total LBA (which is used by OS3.5+ or at least 3.9+). Only affects IDE controllers. SCSI always uses real last block. Easiest option is to initialize the drive under WB 3.1. Or leave few unused megabytes at the end of drive. This check is one of the main points of PFS3AIO: if block that filesystem thinks is the last block can't be read (error returned or block is outside of 4G "barrier" and TD64/NSD or Direct SCSI is not available and/or returns error): Error is shown and partition is not mounted to protect the data.

18 February 2018, 16:08	#37
thomas Registered User Join Date: Jan 2002 Location: Germany Posts: 6,985	Yes.

19 February 2018, 19:56	#38
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,505	Test #2 released. First post updated.

19 February 2018, 20:12	#39
kgc210 Registered User Join Date: Jun 2016 Location: Stoke-On-Trent, England Posts: 450	Do you still have to select DOS type ie: PFS\03 = 0x50465303 or PDS\03 = 0x50445303 Or is it fully automatic now?

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