Another Corrupt ROM

[shahij]

It is MakaraPlus
what corruption do you see on that?

Attachments

ZC1-30086.zip

(447.37 KiB) Downloaded 55 times

[fzabkar]

I don't see any obvious errors. However, the NvCache section is populated, so there appears to have been a power loss data protection event.

The UDSBFW segment is full and contains several 0xBEEF areas.

CELOG appears to have 13 entries.

Is there anything in terminal?

[shahij]

There is nothing in Terminal.
LED Registers Greyed

[fzabkar]

Did you dump the ROM via the terminal port or via a reader?

Am I correct in assuming that the drive doesn't spin?

Are you sure the PCB is OK?

[shahij]

Read by Programmer - Wrote to DONOR PCB.
Also attached DONOR ROM+ Imported RAP+SAP+CAP did not allow to SPIN.
with DONOR ROM-without imported adaptives, it spins, the DONOR PCB is fine

Attachments

DONOR ROM ROM_ZC1AVBAN_47A9DE18.zip

(442.49 KiB) Downloaded 46 times

[fzabkar]

The CRCs for the CAP, RAP and SAP all test OK, both with F3ROMExplorer and my own tool.

The SAP has a full and partial CRC. Both test OK with my tool.

The RAP is compressed with LZMA compression. I decompressed it with F3ROMExplorer, and the result looked OK and had the correct size, as per the LZMA header.

The CAP indicates that this a HP OEM model, MB4000GVYZK:

https://www.harddrivesdirect.com/product_info.php?products_id=493089_MB4000GVYZK

Seagate's warranty checker identifies the serial number, ZC1AW3PG, as belonging to an ST4000NM0035 Exos 7E8 4TB 512n SATA.

I don't see anything in either of the two IAP segments that would point to PUIS being enabled.

The donor has the same BOOTFW and SFW code as the patient.

Perhaps you could import the 3 adaptive segments one at a time to determine which one, if any, is causing the no-spin problem?

[pepe]

I don't see anything in either of the two IAP segments that would point to PUIS being enabled.

registers would show valid value as well as the interface state would be UP if it was PUIS...

[fzabkar]

This is most likely a red herring, but I found these two blocks of data in the donor's active IAP segment. They appear to consist of 24-byte and 20-byte records for 7 heads. The patient does not have these blocks.

Code:

Offset(h) 00   02   04   06   08   0A   0C   0E   10   12   14   16

000002B8            BB00 A100 E6FE F902 D5FD 4C02 3EFE 8A01 63FB CF00
000002D0  0100 0100 F400 3500 8100 BC01 32FF F901 00FE EC01 4BFB 3301
000002E8  0100 0100 65FF 5AFF 6402 3B00 FF02 F0FF 2402 1FFE 5805 9BFF
00000300  0100 0100 7AFF 2BFF EE01 8AFE E302 E2FE 9B02 B2FE 3B05 2F00
00000318  0100 0100 2E00 C400 D401 27FF 8300 DDFE C0FE 4500 FDFE 1EFF
00000330  0100 0100 2000 8F00 E5FF A7FE 63FF AFFE 84FF 6500 41FF 67FF
00000348  0100 0100 D4FF 18FF 7BFE 84FE 2100 4A00 8103 1701 7D03 2203
00000360  0100 0100

Code:

Offset(h) 00   02   04   06   08   0A   0C   0E   10   12

000003D4                      6AFC FCFE 0DFE FDF6 E8FE 0100
000003E8  9C00 1300 0100 0000 AEFC D6FF BBFF 83F9 06FF 0100
000003FC  4300 FAFF FFFF 0000 BAFC A8FE 38FD 01F5 7FFE 0100
00000410  D8FF FFFF FFFF 0000 F7FD 4501 BE01 02FC FBFE 0100
00000424  DCFF FCFF FFFF 0000 B5FB 1BFF E8FC FDF3 FFFD 0100
00000438  F0FF FAFF FFFF 0000 A4FD 6802 A404 DC00 6FFF 0100
0000044C  A9FF FAFF FFFF 0000 5DFC C100 0501 31FB B1FE 0100
00000460  DEFF FDFF 0000 0000

[fzabkar]

Your ROM is twice as big as it needs to be. That is, the second half is empty. Have you selected the correct flash memory chip?

Is it possible that your programmer is correctly programming the ROM with the first half of the dump and then rolling back to address 0 and reprogramming it again with 0xFF from the second half of the dump??

[shahij]

I read it by Programmer from Patient - Wrote by PC3K to donor
I soldered the ROM to DONOR today, and the behaviour is same, read the PATIENT ROM by PC3K, and compared Read by Programmer and Read by PC3K is same

An interesting findings, I wrote 2 DONOR ROM , the PCB behaves same as patient.

However, if I wrote back the ORIGINAL ROM to the PCB (the DONOR ROM the PCB came with), PCB initializes perfectly

[shahij]

Update# we fixed the original PCB, worked