External HDD failure in SNEEK+DI

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Set up an external HDD for SNEEK+DI.
2. Boot SNEEK+DI.

What is the expected output? What do you see instead?
Expected output: System Menu. Instead, I get no signal, not even a black
screen.

What version of the product are you using? On what operating system?
SNEEK+DI r85, compiled on Windows XP.

Please provide any additional information below.
I attempted this build of SNEEK+DI with four different USB drives. Two of
them are 4GB thumb drives, one by Sandisk, one by Kingston. Both drives
worked with an identical file structure.

The other two USB drives are external hard drives. One is a generic
IDE-to-USB hard drive enclosure containing a Quantum Fireball 10.2 GB PATA
hard drive, the other is a 60 GB external drive by ACOM. Both of these
drives failed, using the same file structure as the thumb drives. I've
attached two USB Gecko debug dumps, labeled "Sneekerr-ide.txt" and
"Sneekerr-acom.txt". 

All four drives were formatted FAT32. Cluster sizes were: 4K for the
Kingston and Sandisk, 32K for the ACOM drive (formatted to 32 GB), and 8K
for the Quantum Fireball (formatted to 10 GB).

Original issue reported on code.google.com by kspams...@gmail.com on 7 Apr 2010 at 7:48

Attachments:

• Sneekerr-acom.txt
• Sneekerr-ide.txt

[GoogleCodeExporter]

I experienced this issue as well using 3 USB devices. All three with identical 
file/folder structures. All 3 work with the di.bin from R79. One (WD 60GB HDD) 
does 
not work with the di.bin from R80 (no signal). Using R80 boot2.bin and R79 
di.bin 
the same HD works. Must be the di.bin changes causing the issue.

Original comment by cranium9...@yahoo.com on 9 Apr 2010 at 2:07

[GoogleCodeExporter]

Thanks for the suggestion... r79's di.bin did not work with my USB drives, 
though. :(

Original comment by kspams...@gmail.com on 9 Apr 2010 at 7:21

[GoogleCodeExporter]

does this problem still exist in the latest build?

Original comment by urt...@gmail.com on 17 Jul 2010 at 10:50

[GoogleCodeExporter]

I tested today and still have the same problem.

Original comment by darkunde...@gmail.com on 25 Jul 2010 at 12:11

[GoogleCodeExporter]

Note the section
"OH1:configured USB device at port 0" (Wii's own USB code)
followed by
"port 0 reseted status:1005" (SNEEK's USB code)
in the log - the Wii's own USB stack claims the device before SNEEK does.

This is exactly what my fix in issue 54 fixes... I suggest you try those 
changes (my last comment in it is the most robust solution) - there's a good 
chance it will work, as this is exactly the same issue my harddrive had before.

I now, afterwards, realize issue 54 is a duplicate of this one - sorry for 
posting a duplicate! :)

Original comment by thuli...@gmail.com on 25 Jul 2010 at 12:30

[GoogleCodeExporter]

Tested r106 on the problem drives, both without and with the patch suggested by 
thulinma -- no luck. It seems SNEEK is actually mounting the drive this time, 
but it still throws qtd errors with both drives.

Original comment by kspams...@gmail.com on 25 Jul 2010 at 4:11

Attachments:

• sneekpatch-err-acom.txt
• sneekpatch-err-ide.txt

[GoogleCodeExporter]

I've been investigating this issue - especially since I noticed that one of my 
USB drives exhibits the exact same behavior as yours do (unlike my other drive, 
which was not working due to slow startup) - they also fail (nearly) every 
single time with qtd errors and handshake timeouts.

I've made an interesting discovery: I opened up the external HD that also has 
this problem and replaced the power to the drive with an external power source 
- and then it worked!
My theory is that what happens is the enclosure puts the harddrive in power 
saving mode, and on first access (mounting the FAT file system) puts the drive 
in normal powered mode. This seems to confuse the EHCI driver in SNEEK, causing 
all these errors. Using external power instead of the enclosure power "forces" 
the drive to remain in normal power mode all the time, thus negating this issue.

Weird, but it seems to work. kspamsink - do you have a way of testing this on 
your drives? Perhaps it is easiest put your drives in a different enclosure 
that does not support power saving / sleeping.

Original comment by thuli...@gmail.com on 31 Jul 2010 at 9:19

[GoogleCodeExporter]

I workaround this (even with r114) by turning on the wii WITHOUT the SD card.
Wait for the health screen -> put in the SD card and press reset
Now sneeks loads without problems (UNEEK and UNEEK+DI) (WD Essentials 1TB FAT32 
32kb)
Otherwise it will flash 3 times (cannot find kernel.bin) and no screen output.
This has to do with slow usb devices like thulinma stated.
Waiting for the health screen, you can hear the drive spinning up.

Original comment by digd...@zonnet.nl on 1 Aug 2010 at 12:21