Maelgrum wrote on 2023-10-05, 13:13:

sbcrack stage 2 for 4.12

For 4.12;

1Stage 2, variant 05 ...passed
2Stage 2, variant 07 ...passed
3Stage 2, variant 18 ...passed
4Stage 2, variant 44 ...passed

sbcrack stage 3 for 4.12

Maelgrum wrote on 2023-10-05, 14:04:

sbcrack stage 3 for 4.12

1SB Reset: done
2DSP version: 4.12
3MPU-401 init: done
4MPU-401 loopback check: done
5Stage 3, variant 44 ...passed
6Success!

Stack structure for variant 44:
C0 ACC
BF R0
BE DPH
BD DPL
BC ACC
BB PCH
BA PCL

looks good for attack ))

sbcrack stage 4 for 4.12

Maelgrum wrote on 2023-10-05, 14:27:

sbcrack stage 4 for 4.12

1SB Reset: done
2DSP version: 4.12
3MPU-401 init: done
4MPU-401 loopback check: done
5Stage 4, addr 1212 ...failed
6Stage 4, addr 1213 ...failed
7Stage 4, addr 1211 ...failed
8Stage 4, addr 1214 ...failed
9Stage 4, addr 1210 ...failed
10Stage 4, addr 1215 ...failed
11Stage 4, addr 120F ...Success!

S95Sedan wrote on 2023-10-05, 14:30:

Maelgrum wrote on 2023-10-05, 14:27:

sbcrack stage 4 for 4.12

Stage 4, addr 120F ...Success![/code]

Excellent !
Stage 5 not required - all checks are integrated in stage 4
sb16dump will be ready soon

sb16dump V0.14 experimental support for 4.12

Maelgrum wrote on 2023-10-05, 14:44:

sb16dump V0.14 experimental support for 4.12

Gives an error.
Added the mdump1 and 2.

1SB Reset: done
2DSP version: 4.12
3MPU-401 init: done
4Internal memory dump: done
5MPU-401 loopback check: done
6ERROR: Dump failed

some more sbcrack scan Stage4
scans full possible range

Maelgrum wrote on 2023-10-05, 14:44:

sb16dump V0.14 experimental support for 4.12

You had the same problem yesterday with 4.16. Keep in mind that considering resetting of DSP and returning of AA as a proper result is on my opinion very wrong. Resets can be triggered by faulty code too. Wrap around to address 0 past program memory would result in reset too. But 120F looks very probable for an address of E3 comparing it to 4.13. Instead I propose you to look for the proper functioning of E3, e.g. returning of copyright $.

Maelgrum wrote on 2023-10-05, 14:59:

some more sbcrack scan Stage4
scans full possible range

Same result all failed apart from the one that already passed. (120F )

sb16dump V0.15 experimental support for 4.12 (found error in previous version)

Maelgrum wrote on 2023-10-05, 15:08:

sb16dump V0.15 experimental support for 4.12 (found error in previous version)

Same as the other version.
Added Mdumps.

1SB Reset: done
2DSP version: 4.12
3MPU-401 init: done
4Internal memory dump: done
5MPU-401 loopback check: done
6ERROR: Dump failed

Maelgrum wrote on 2023-10-05, 15:08:

sb16dump V0.15 experimental support for 4.12 (found error in previous version)

This triggered an NMI for me on my AWE32

maxtherabbit wrote on 2023-10-05, 15:27:

Maelgrum wrote on 2023-10-05, 15:08:

sb16dump V0.15 experimental support for 4.12 (found error in previous version)

This triggered an NMI for me on my AWE32

I guess it jumps to a code which generates NMI for MIDI emulation with AWEUTIL.

maxtherabbit wrote on 2023-10-05, 15:27:

Maelgrum wrote on 2023-10-05, 15:08:

sb16dump V0.15 experimental support for 4.12 (found error in previous version)

This triggered an NMI for me on my AWE32

That happened to me when testing with SBCRACK and dumping the firmware for the AWE64 CT4520.

Just disable NMI to let it continue and do a cold boot afterwards.

I do not have the ability to disable NMIs on this board short of putting a piece of tape on the card edge or something

Apparently you can gate them by setting bit 7 at I/O port 0x70.

Now it's just ERROR: Dump failed

Perhaps toggling the MFBEN jumper setting helps?

maxtherabbit wrote on 2023-10-05, 15:51:

I do not have the ability to disable NMIs on this board short of putting a piece of tape on the card edge or something

Yes , tape would work perfectly. It is the rightmost pad on the XT edge connector but I guess you know that.