Partition Manager

PartMan 0.97 fails on Titanium with ADFS 4.04

I’ve uploaded a revised 0.97 if you’d like to try again.

The revised 0.97 version works. Thanks.
The FileCore formating is recognised, but the disc space is reported as ‘unallocated’
Both, the 256 GB and 2 TB discs.

add_partitions_MBR
==================
ADFS 4 checking for MBR partition table
 IDEUserOp_ADFS(0,1,&0,&0,..,&200)
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 00 00 00 00 00 00 40 20 00 05 BA 03 00   .........@ .....

  ADFS_IDEUserOp,&1000008
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 00 00 00 00 00 00 40 20 00 05 BA 03 00   .........@ .....

MBR partition table found

Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : EB 3C 90 43 61 73 74 6C 65 00 00 00 02 00 00 00   .<.Castle.......
&000010 : 00 00 00 00 00 00 00 00 BD 00 10 00 00 00 00 00   ................
&000020 : C3 1F EE 1E 00 00 00 00 00 60 32 4E 4F 20 4E 41   .........`2NO NA
&000030 : 4D 45 20 20 20 20 20 20 20 20 20 20 20 20 FA 31   ME            .1
&000040 : C0 8E D0 BC 00 7C FB 8E D8 E8 00 00 5E 83 C6 19   ............^...
&000050 : BB 07 00 FC AC 84 C0 74 06 B4 0E CD 10 EB F5 30   .......t.......0
&000060 : E4 CD 16 CD 19 0D 0A 4E 6F 6E 2D 73 79 73 74 65   .......Non-syste
&000070 : 6D 20 64 69 73 6B 0D 0A 50 72 65 73 73 20 61 6E   m disk..Press an
&000080 : 79 20 6B 65 79 20 74 6F 20 72 65 62 6F 6F 74 0D   y key to reboot.
&000090 : 0A 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0000A0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0000B0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0000C0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0000D0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0000E0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0000F0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000100 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000110 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000120 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000130 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000140 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000150 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000160 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000170 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000180 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000190 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0001A0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0001B0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0001C0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0001D0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0001E0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&0001F0 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 55 AA   ..............U.

No supported partitions found
MBR partition table not valid™
add_partitions_FileCore
=======================
 IDEUserOp_ADFS(0,1,&6,&0,..,&200)
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 06 00 00 00 00 00 40 20 00 05 BA 03 00   .........@ .....

  ADFS_IDEUserOp,&1000008
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 06 00 00 00 00 00 40 20 00 05 BA 03 00   .........@ .....


Boot DiscRec:
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 09 BD 10 00 13 0F 00 02 01 56 20 00 01 99 E8 02   .........V .....
&000010 : 00 E0 B2 38 00 00 00 00 00 00 00 00 00 00 00 00   ...8............
&000020 : 00 00 00 00 3A 00 00 00 03 01 07 00 01 00 00 00   ....:...........
&000030 : 00 08 00 00 00 00 00 00 00 00 00 00 00 00 00 50   ...............P

Full DiscRec is in Zone 939 at LBA &E8E3000
LBA size in bits=4096 Log2=9
Zone spare bits=32
DiscRec size in bits=480
Log2 bytes per map bit=15

 IDEUserOp_ADFS(0,1,&E8E3000,&0,..,&200)
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 00 30 8E 0E 00 00 40 20 00 05 BA 03 00   ....0....@ .....

  ADFS_IDEUserOp,&1000008
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 00 30 8E 0E 00 00 40 20 00 05 BA 03 00   ....0....@ .....


Full DiscRec:
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 04 00 00 00 53 42 50 72 02 00 00 00 00 08 00 00   ....SBPr........
&000010 : 00 00 00 00 00 00 00 00 01 80 1C 00 42 33 0D 00   ............B3..
&000020 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................
&000030 : 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00   ................

DiscRec probably invalid
Adding unallocated space
 Start of area=000000000
    Drive size=01D1C5970
Max accessible=020000000
Drive size below max

Unallocated block
-----------------
P_Start_LBA_Low%=&0
P_Start_LBA_High%=&0
P_Size_In_LBA_Low%=&1D1C5970
P_Size_In_LBA_High%=&0

Failed to find a valid FileCore structure
------
Checking for a full-disc FileCore
 IDEUserOp_ADFS(1,1,&0,&0,..,&1000)
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 00 00 00 00 00 00 50 20 00 05 BA 03 00   .........P .....

  ADFS_IDEUserOp,&1001008
Address :  0  1  2  3  4  5  6  7  8  9  A  B  C  D  E  F
&000000 : 00 01 00 00 00 00 00 00 00 50 20 00 05 BA 03 00   .........P .....
  

Thanks for retesting, I will however need to see the full debug log to figure out where it went wrong.

Sorry. Easter vacation. We are not at home. I can not test before weekend. :-(

Thanks for retesting, I will however need to see the full debug log to figure out where it went wrong.

see here

Thanks Anton, I’ll review and try to figure out what the issue is.

I’ve made a few tweaks to the 48-bit LBA code for ADFS4 and reuploaded if someone would like to try it on a test drive. I don’t think it will get it working, but there’s some additional debug output that might help.

Anton, your debug output is odd. Your primary drive has an MBR partition table, but no partitions and the FileCore middle zone didn’t seem to be where it was expected – I can’t explain why though. Your secondary drive, with the 4K sectors – all the discop’s appeared to fail. Finally, the SD card appears to only have an 8GB Linux partition – it that correct?

Jon, as far as I know there is nothing special with the primary drive. Yes, the secondary is a 4Kn 2 TB drive, FileCore formated. SD card 8GB Linux, correct.

With the actual PartMan I get the Error ‘Corruption of stack’
*where
Address &00000004 is system workspace

The debug output so far

With the actual PartMan I get the Error ‘Corruption of stack’

Up until that point it was working correctly. It found “ADFS::TIK” on the primary drive and had started reading the secondary 4K drive.

I can’t tell exactly where the stack corruption occurred, but it looks like it was caused shortly after reading LBA 1 from the 4K drive to check for an MBR partition table. I’ll recheck all the code for potential buffer-overruns when reading 4K sectors, as memory corruption is the logical explanation.

Anton, I’ve corrected a few potential buffer overruns and uploaded a revised build for testing.

Jon, there is something strange there.
Fresh extracted from zip, a empty PartMan window and the hourglas is shown, no response → reboot. Starting later fills window, but hangs mostly at closing window → reboot. I think it’s the 4Kn drive, as it work on the Pinebook Pro.

Apart from that, the FileCore at the primary drive is now recognised, from the 4Kn not.

new debug.txt

Thanks for the debug log, I can’t explain the odd behaviour but I can see why it failed to identify Filecore on the 4K drive. I’ll review all the code it called and double-check I’ve not missed something obvious.

I didn’t spot anything obvious that might have caused the locks (although I can guarantee its due to a bug in the way BASIC handles LOCAL ERROR ), so I’ve removed the error handler from the ADFS code. I’ll probably need to remove all the LOCAL ERROR handlers to prevent hard locks completely.

I have also modified the FileCore identify code to support 4K sectors in the latest build if you’d like to see if it identifies FileCore on the 4K drive.

Jon blamed

a bug in the way BASIC handles LOCAL ERROR

The much-missed Steve Drain asked you to clarify what you meant to no effect, so I’m going to do the same. There certainly is a long-standing bug concerning LOCAL, and there are potentially bugs regarding LOCAL arrays, but your description is sufficiently vague that certainly from this vantage point it could just as easily be your own code going wrong – is it possible that these things don’t do what you think they do?

DIM var LOCAL size – causes stack corruption if not preceded by an explicit LOCAL var

Nope. Not seeing that, and there’s no link between the ‘LOCALness’ of var and what DIM does. So I can’t guess what you’re seeing.

DIM var LOCAL errors when used after certain other variable declarations

Too vague to be helpful, but again I can’t see the link.

ON ERROR LOCAL can corrupt the error stack under certain conditions when the function exits. The workaround is to put a LOCAL ERROR statement before the ON ERROR LOCAL

ONERRORLOCAL and LOCALERROR do two different things. Were you expecting the former to do what the latter does?

Passing undimensioned arrays to FN RETURN parameters can corrupt the stack when the function exits.

Do you mean undimensioned array references e.g. a(), or something else? All array references are RETURN by definition – there’s no difference whether the PROC has a RETURN in its parameter list.

ON ERROR LOCAL appears to corrupt variables and causes hard-locks if an error is trapped and the procedure/function exited.

No it doesn’t. But your code can certainly corrupt variables and lock up. And beware of this construction:

DEFPROClooks_safe
ONERRORLOCAL:ENDPROC
PROCbut_no
ENDPROC

DEFPROCbut_no
LOCALfoobar
GOBBLEDEGOOK
ENDPROC

That will zero the global foobar variable, regardless of the ONERROR.

Is it possible that ONERRORLOCAL does not do what you have assumed it does?

Jon, with your latest version PartMgr (mostly) starts.
Click on close icon → machine hangs.
Quit from task manager (mostly) works, but subsequent program start fails.

New observation (not tested yet): closeing a filer window (4Kn disc) after PartMgr is running or (sucessfuly) quit → machine hangs.
It looks like PartMgr changed something (ADFS?) does not expect?

Tomorrow I will disconnect the 4Kn drive to check if this is realy 4Kn related.

New observation (not tested yet): closeing a filer window (4Kn disc) after PartMgr is running or (sucessfuly) quit → machine hangs.

That doesn’t sound good. Removing the 4K drive would certainly confirm if the issue is 4K specific or ADFS_IDEUserOp related. From your last debug log, all the reads succeeded so on the face of it, the code appears to be working. I suppose it’s possible that by using ADFS_IDEUserOp to read from the drive its somehow left either ADFS or SATADriver in an odd state.

Did it identify the FileCore partition on the 4K drive?

Jon blamed

Nemo, you posted your reply in the wrong thread.

The 4Kn drive removed, the machine does not hang anymore.
But there is still something wrong. At a subsequent run the DiscOp 3/IDENTIFY response: is empty (all zero) and nothing is recognised.
*RMReinit ADFS cure this.

I think, this should be solved before looking into the 4Kn thing.
You may contact me by PM.

Thanks, it sounds like my use of ADFS_IDEUserOp is somehow messing up ADFS. I’ll comment out the code in PM for the time being and will pull the source code for both ADFS4 and SATADriver and check through the relevant code line-by-line.

Once I have some tests, I’ll email you direct. I’ve made a note so you can remove your email from the post.

Have try 0.97.

Not sure what “Reserved” means in the Filecorepartition.

The debug file you find here.

Not sure what “Reserved” means in the Filecore partition

Reserved is a “Protective partition” that covers the FileCore portion of the disc.

Thanks for the debug log, it managed to identify SSD_FAT this time on ADFS 0:1 which means the 48-bit LBA reads worked. As it failed to identify ROS_Daten on ADFS 0:2, there’s still some issues with identifying FileCore on 4K drives that I need to resolve.

I’ve posted a build with further 4K sector tweaks, If you’d like to give it another go.

The new version works fine here. No side effects as previous noticed.
And the details from the 4Kn drive are shown now too.

Thanks for confirming.

Who wants to be the brave soul that tries to format drives under ADFS4?

Reserved is a “Protective partition” that covers the FileCore portion of the disc.

Not sure what it means. I can not see “handling differences” for me. Is usable like the other devices.

I’ve posted a build with further 4K sector tweaks, If you’d like to give it another go.

Who wants to be the brave soul that tries to format drives under ADFS4?

ADFS4 is not the Problem. My bootdevice is ADFS7…

Have do any write/read actions (round about 20GB). No problems.

Thanks a lot. Very nice work.

ADFS4 is not the Problem. My bootdevice is ADFS7…

ADFS4 as in “version 4 of ADFS”, not “ADFS drive number 4” :)