Testing: Old IDE drives that are not USB IDE adapter compatible

[rare_j]

Yes the old one is the one that comes up as Myson Century.
I'm not sure what's up with USBview, run as administrator, it works fine to view every other usb device on the system, including the 'easy ide' adapter, but crashes when viewing the startech.

[rare_j]

I came across another very old hard drive, Conner CP2084.

BusType: 0x07
Vendor: 'Conner P'
Product: 'eripherals 84MB'
Revision: '1.62'
Serial: ''
Size: 0

INQUIRY:
000000011F000000436F6E6E6572205065726970686572616C732038344D4220 ........Conner Peripherals 84MB
312E3632 1.62

READ CAPACITY:
FFFFFFFF00000200 ........

INQUIRY 1F:
000000011F000000436F6E6E6572205065726970686572616C732038344D4220 ........Conner Peripherals 84MB
312E3632 1.62

MODE SENSE:
00000000 ....

MODE SENSE 86:
860B00020000FFFFFFFF000100 .............

Very unlikely this one has an Amiga partition on it. But I thought it might be interesting.

[Toni Wilen]

Originally Posted by rare_j

OK, this is the old working ide adapter with the old amiga drive.

It works! SPT_READ_10.BIN contains valid RDB blocks.

SPT READ(10) 0 failed
IOCTL_SCSI_PASS_THROUGH_DIRECT: 28.00.00.00.00.21.00.00.01.00.
SENSE: 70.00.05.00.00.00.00.0A.00.00.00.00.20.00.00.00.00.00.
SPT READ(10) 33 failed
IOCTL_SCSI_PASS_THROUGH_DIRECT: 28.00.00.00.00.22.00.00.01.00.
SENSE: 70.00.05.00.00.00.00.0A.00.00.00.00.20.00.00.00.00.00.
SPT READ(10) 34 failed
<snip>

So it really looks like drive takes the LBA as CHS because 0 is rejected (Sectors starts from one) and drive apparently has 32 sectors. (But spec says 38? Or perhaps adapter executed initialize drive parameters command? Not that it matters, it is logical geometry only) Fortunately IDE uses exact same register and register bits for both LBA and CHS.

This confirms that at least some adapters don't work because they don't understand CHS-only IDE drives.

Anyway, it looks like it is relatively easy to detect correct head and sector values just by checking specific "LBA" values (if they read or fail) and then whole drive can be imaged easily.

[mark_k]

That's a lucky coincidence! The bridge firmware doesn't specifically handle the case where the number of LBA sectors in IDENTIFY DEVICE data is 0, so it subtracts 1 and puts that ($FFFFFFFF) in the READ CAPACITY data for the maximum LBA. And presumably its internal READ/WRITE command validation then allows anything in the READ/WRITE (10) CDB LBA bytes.

[rare_j]

So which hardware is at fault? The drives or the adapters?

[mark_k]

The adapters. The drives are most likely fine.

Originally IDE drives were accessed by specifying a cylinder, head & sector number. Whereas all modern IDE drives also support LBA access (specify a block number). The firmware in the adapters only works with drives that support LBA access. They don't work with older drives which only support C/H/S, not LBA access.

Since it's optional for drives to support LBA access, the adapter firmware is at fault for not handling that situation.

[rare_j]

I found what seems to be some relevent info here:
http://www.linuxquestions.org/questi...5/#post5143646

Is this similar to what Toni has done?

[demolition]

I got one 341MB drive here. It seems to be found by Windows which wants to initialize it (it should have an RDB on it so that would be normal I guess). I tried using WinImage to create an image but it just comes up with the error 'Incorrect function'.

WinUAE can see the drive and marks it as '[OS] [0K,RW]' however it crashes when I click the 'Read Identity' button:
Unhandled exception at 0x0114DA25 in winuae.exe: 0xC000041D: An unhandled exception was encountered during a user callback.

USBview output:

[Port12]  :  USB Mass Storage Device


Is Port User Connectable:         yes
Is Port Debug Capable:            no
Companion Port Number:            0
Companion Hub Symbolic Link Name: 
Protocols Supported:
 USB 1.1:                         yes
 USB 2.0:                         yes
 USB 3.0:                         no

Device Power State:               PowerDeviceD0

       ---===>Device Information<===---
English product name: "ST9385AG        "

ConnectionStatus:                  
Current Config Value:              0x01  -> Device Bus Speed: High (is not SuperSpeed or higher capable)
Device Address:                    0x0C
Open Pipes:                           2

          ===>Device Descriptor<===
bLength:                           0x12
bDescriptorType:                   0x01
bcdUSB:                          0x0200
bDeviceClass:                      0x00  -> This is an Interface Class Defined Device
bDeviceSubClass:                   0x00
bDeviceProtocol:                   0x00
bMaxPacketSize0:                   0x40 = (64) Bytes
idVendor:                        0x13FD = Initio Corporation
idProduct:                       0x0540
bcdDevice:                       0x0000
iManufacturer:                     0x01
     English (United States)  "Initio  "
iProduct:                          0x02
     English (United States)  "ST9385AG        "
iSerialNumber:                     0x03
     English (United States)  "00101005000000000"
bNumConfigurations:                0x01

          ---===>Open Pipes<===---

          ===>Endpoint Descriptor<===
bLength:                           0x07
bDescriptorType:                   0x05
bEndpointAddress:                  0x81  -> Direction: IN - EndpointID: 1
bmAttributes:                      0x02  -> Bulk Transfer Type
wMaxPacketSize:                  0x0200 = 0x200 max bytes
bInterval:                         0x00

          ===>Endpoint Descriptor<===
bLength:                           0x07
bDescriptorType:                   0x05
bEndpointAddress:                  0x02  -> Direction: OUT - EndpointID: 2
bmAttributes:                      0x02  -> Bulk Transfer Type
wMaxPacketSize:                  0x0200 = 0x200 max bytes
bInterval:                         0x01

       ---===>Full Configuration Descriptor<===---

          ===>Configuration Descriptor<===
bLength:                           0x09
bDescriptorType:                   0x02
wTotalLength:                    0x0020  -> Validated
bNumInterfaces:                    0x01
bConfigurationValue:               0x01
iConfiguration:                    0x00
bmAttributes:                      0xC0  -> Self Powered
MaxPower:                          0x01 =   2 mA

          ===>Interface Descriptor<===
bLength:                           0x09
bDescriptorType:                   0x04
bInterfaceNumber:                  0x00
bAlternateSetting:                 0x00
bNumEndpoints:                     0x02
bInterfaceClass:                   0x08  -> This is a Mass Storage USB Device Interface Class
bInterfaceSubClass:                0x06
bInterfaceProtocol:                0x50
iInterface:                        0x00

          ===>Endpoint Descriptor<===
bLength:                           0x07
bDescriptorType:                   0x05
bEndpointAddress:                  0x81  -> Direction: IN - EndpointID: 1
bmAttributes:                      0x02  -> Bulk Transfer Type
wMaxPacketSize:                  0x0200 = 0x200 max bytes
bInterval:                         0x00

          ===>Endpoint Descriptor<===
bLength:                           0x07
bDescriptorType:                   0x05
bEndpointAddress:                  0x02  -> Direction: OUT - EndpointID: 2
bmAttributes:                      0x02  -> Bulk Transfer Type
wMaxPacketSize:                  0x0200 = 0x200 max bytes
bInterval:                         0x01

[Toni Wilen]

Use winuae.7z, -log -logflush command line parameters (so that crash won't truncate the log). Attach winuaebootlog.txt.

[demolition]

Here's the log.

[Toni Wilen]

Nothing useful in the log. This is something that shouldn't happen..

Does it create dump file? (that dbghelp message may mean you have disabled some program crash related features in Windows). Use 2.6.0 b2 to generate the dump file, others are useless.

And if dump is not available: redownload winuae.7z and attach new log.

[demolition]

Don't know if this is what you need, but here's a crash dump saved from VS:
http://62.135.246.180/winuae_crashdump.zip

Where do I get WinUAE 2.6.0b2?

Edit: Also attached a new log.

[Toni Wilen]

Argh. Of course it was supposed to be 3.6.0 beta 2. But latest log helped, wait for next reply

[Toni Wilen]

This is the next reply

(Redownload, log, the usual. May not even crash anymore.)

[demolition]

Well it still crashes, but now the GUI just hangs instead of invoking the debugger.

[Toni Wilen]

Originally Posted by demolition

Well it still crashes, but now the GUI just hangs instead of invoking the debugger.

IOCTL_SCSI_PASS_THROUGH_DIRECT: 5A.00.BF.00.00.00.00.FF.00.00. OK (13 bytes)
MODE SENSE LEN 34315
MODE SENSE DBD 65535

Looks like your adapter returns bogus data for MODE SENSE. Returns 13 bytes but length field in returned data is 34315 bytes! DBD size is also totally invalid.

More validation added to detect invalid data..

[demolition]

New log

[Toni Wilen]

Originally Posted by demolition

New log

Now it hung when it attempted to guess number of sectors. Blocks 0 to 4 read find but 5th (chs=0,0,6) hung.

I don't think there are drives with only 5 sectors per track so perhaps the drive is damaged? If you keep log window open, does it time out and continue? (Timeout is 2mins, I changed it to 10s now)

Does the drive label mention any geometry values?

[demolition]

After two timeouts it finished and returned this information:

BusType: 0x07
Vendor: 'Initio'
Product: 'ST9385AG'
Revision: '2.35'
Serial: '00101005000000000'
Size: 0

INQUIRY:
000000021F000000496E6974696F202053543933383541472020202020202020 ........Initio  ST9385AG        
322E3335                                                         2.35                            

READ CAPACITY:
FFFFFFFF00000200                                                 ........                        

MODE SENSE:
00020000

Also attached new log.

I can't rule out that the drive is faulty somehow but it was working when I pulled it from an A600, although that is perhaps 5 years ago.. I could try hooking it up to a computer with on-board IDE to see if it works better there. This is how I managed to make an image of the drive after pulling it from the Amiga.

Edit: Oh, and drive label says 934 cyls, 14 heads, 51 sects, 341.4 MB which I guess you could also find if you look up the model name ST9385AG.

[demolition]

The drive seems to work with an on-board IDE controller. In the BIOS it finds the correct geometry and in Ubuntu (this PC does not have Windows) it finds the drive like this (from dmesg):

[    0.993945] ata1: PATA max UDMA/133 cmd 0x1f0 ctl 0x3f6 bmdma 0xfc00 irq 14
[    1.196284] ata1.00: ATA-0: ST9385AG, 24.24.01, max MWDMA1
[    1.196289] ata1.00: 666876 sectors, multi 0, CHS 934/14/51
[    1.196295] ata1: nv_mode_filter: 0x18f&0x739f->0x18f, BIOS=0x0 (0x0) ACPI=0x8f (180:600:0x12)
[    1.228215] ata1.00: configured for MWDMA1
[    1.228367] scsi 0:0:0:0: Direct-Access     ATA      ST9385AG         4.01 PQ: 0 ANSI: 5
[    1.228625] sd 0:0:0:0: [sda] 666876 512-byte logical blocks: (341 MB/326 MiB)
[    1.228674] sd 0:0:0:0: [sda] Write Protect is off
[    1.228677] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
[    1.228696] sd 0:0:0:0: Attached scsi generic sg0 type 0

Edit: Ok, the drive may not be 100% ok. Just tried making an image with dd and it stopped @127MiB with an IO-error (error log says 'media error').