Rockbox

I'm seeing the same thing.

I've performed some checks by using a USB monitor program - http://www.hhdsoftware.com/usbmon.html which allows trial runs for free.
Unfortunately this shows nothing happening at all from the PC end when using the bootloader USB mode.

As a comparison, if I boot Rockbox first, the logged data starts with the following:
000001: Get Status Request (DOWN), 07.04.2006 12:59:41.2451792 +142.9856032
Get status from Device. Data length 0x2 bytes (00 00)
000002: Control Transfer (UP), 07.04.2006 12:59:41.2451792 +0.0
Pipe Handle: 0x85a0a9f4
03 00 ..
Setup Packet
80 00 00 00 00 00 02 00€.......
Recipient: Device
Request Type: Standard
Direction: Device->Host
Request: 0x0 (GET_STATUS)
Value: 0x0
Index: 0x0
Length: 0x2
...
(This then succeeds every time.)

I've now tried this on 3 different PCs - two running Windows XP SP2 and one running Windows 2000 SP4 - and all 3 machines give exactly the same results.
In the unlikely case it's relevant, my Iriver firmware is 1.28E.

I have the same problem, although the bootloader USB mode *sometimes* works for me. Unfortunately, I haven't been able to determine any difference in circumstances between the occasions when it works and those when it doesn't.

Strange - USB bootloader (with v5) works perfect for me (H340).

Can anybody give some more details???

Not much to it - method is as follows:
1) H340 is powered off and no charger is connecte.
2) Connect USB cable to USB 2.0 DEVICE port.
3) H340 powers up, displays "Rochbox Bootloader/Version 5/Batt: ?.??V" - then "Bootloader USB mode".
4) PC does not recognise anything happening.
If I boot into the IRiver firmware (1.28E), USB works fine every time.

Sorry, it works perfect here (with Win2000-SP4 and with WinXP-SP1).
Is somebody else able to reproduce this behaviour?

For me, the bootloader USB mode sometimes works and sometimes doesn't. When it doesn't, symptoms and steps to reproduce are exactly the same as above; unfortunately, i haven't been able to determine any difference in circumstances between the failures and the successes. Bootloader v5, firmware 1.29K.

Please add following detail: how are you connecting?
- USB direct on motherboard
- powered hub
- unpowered hub
- usb 1.1 or 2.0

I've seen it do strange things on disconnect when using the hub of a DELL monitor.

It is maybe also related to a setting in retailos about connecting on desktop or hub.
You can also try enabling/disabling usb charging in retailos

USB direct to motherboard on 3 different PC with different chipsets. No hubs involved at all. RetailOS works reliably on all 3 machines.
USB Charging in RetailOS was disabled, but enabling it makes no difference to Rockbox. USB Conn Mode in RetailOS was set to Desktop, tried Hub, but no change.

I used USB direct to motherboard, on Windows XP SP2, USB 2.0. Sometimes works, sometimes doesn't...any idea on how to debug it locally?

As noted above, the bootloader USB mode occasionally works for me but often fails. I have tried it with USB 2.0 direct to motherboard on several machines running Windows XP SP2 and also tried it with USB 1.1 direct to motherboard on a Linux kernel 2.6.14.3 system. Under Linux, nothing in lsusb -v changed and nothing was reported in the kernel logs.

I can confirm that I have seen more than 2 H300s not to be able to get recognised by Windows XP (and Ubunutu)(haven't tested other OSes).
It seems that this bootloader bug affects some H300 users. With or without a HUB I get the same results.
There is a whole thread for it in MisticRiver forums.

It would be good to know what Linus has to say about this, since he wrote the Bootloader.

There is also another issue, maybe related to this one:

When disconnecting from my dell monitor at work, the h340 stays in usb mode.

This depends on the usb settings of the original firmware (don't remember which one),
so I can solve it that way.

For what its worth, on my new computer with a nForce 4 chipset and USB 2.0 ports bootloader USB mode works perfectly every time.
On my old box with a nondescript VIA chipset and USB 1.1 ports it normally doesn't work, but given enough patience and restarts of the player it will work.
When Rockbox is up and running USB mode always works fine with both PCs. Both PCs run Windows XP and I have never touched any USB related settings in Windows.

Hi,

I've been experiencing problems with the USB bootloader too.
I'm running a recent Debian Linux, kernel 2.6.16.18. USB is built in and uses USB 2.0.
I own an iRiver H320 with a V5 bootloader and 1.30 EU firmware version.
When Rockbox is loaded, USB works very well.
Here is what's printed in the syslog.

Jun 5 12:00:13 localhost kernel: usb 1-2: new high speed USB device using ehci_hcd and address 24
Jun 5 12:00:14 localhost kernel: usb 1-2: configuration #2 chosen from 1 choice
Jun 5 12:00:14 localhost kernel: scsi24 : SCSI emulation for USB Mass Storage devices
Jun 5 12:00:14 localhost kernel: usb-storage: device found at 24
Jun 5 12:00:14 localhost kernel: usb-storage: waiting for device to settle before scanning
Jun 5 12:00:19 localhost kernel: Vendor: TOSHIBA Model: MK2004GAL Rev: JC10
Jun 5 12:00:19 localhost kernel: Type: Direct-Access ANSI SCSI revision: 00
Jun 5 12:00:19 localhost kernel: SCSI device sdb: 39063024 512-byte hdwr sectors (20000 MB)
Jun 5 12:00:19 localhost kernel: sdb: Write Protect is off
Jun 5 12:00:19 localhost kernel: sdb: Mode Sense: 00 4a 00 00
Jun 5 12:00:19 localhost kernel: sdb: assuming drive cache: write through
Jun 5 12:00:19 localhost kernel: SCSI device sdb: 39063024 512-byte hdwr sectors (20000 MB)
Jun 5 12:00:19 localhost kernel: sdb: Write Protect is off
Jun 5 12:00:19 localhost kernel: sdb: Mode Sense: 00 4a 00 00
Jun 5 12:00:19 localhost kernel: sdb: assuming drive cache: write through
Jun 5 12:00:19 localhost kernel: sdb: sdb1
Jun 5 12:00:19 localhost kernel: sd 24:0:0:0: Attached scsi disk sdb
Jun 5 12:00:19 localhost kernel: usb-storage: device scan complete
Jun 5 12:00:21 localhost kernel: UDF-fs: No VRS found
Jun 5 12:00:21 localhost kernel: UDF-fs: No VRS found
Jun 5 12:00:21 localhost kernel: Unable to identify CD-ROM format.
Jun 5 12:00:21 localhost kernel: Unable to identify CD-ROM format.
Jun 5 12:00:21 localhost kernel: FAT: utf8 is not a recommended IO charset for FAT filesystems, filesystem will be case sensitive!

Then my H300 icon pops up on my KDE desktop and device is mounted perfectly.

Now, when using the USB bootloader (iRiver is off, no prower cable, then USB is plugged in) nothing gets printed.
USB chipset is an ICH5 and works perfectly under Linux.

Hope that helps.

I'm getting the same problems as mentioned above,
USB2 (motherboard), XP SP2.

Bootloader USB works for me if I select the "Hub" option in the original firmware, and not if "Desktop" is selected.

I wish that would be the case, but for me it will refuse to work even when in original firmware I select the Hub option.
Also I am not sure how this can have any relation to USB bootloader on rockbox, in any case more people should test these options just to be sure.

I've tried setting up the hub mode instead of desktop mode for USB. USB bootloader still doesn't work.
BTW I'm a bit sceptical about that 'hub mode' trick.... how could the iRiver original mode affect the Rockbox bootloader when there's a power off between the two things ?

Same problem here. The USB connection works perfectly in both the iriver firmware and Rockbox, but not in the bootloader. I use GNU/Linux, and nothing is printed in dmesg here either, just like in the above comment by Jerome. Motherboard USB 1.1.

It doesn't work at all for me either.

Using a h340 with v5 bootloader on usb2.0 connection (via chipset) to a gentoo box with kernels from 2.6.12ish to 2.6.19.

I've repeatedly checked the kernel logs from time to time and it never shows anything.

The updated H300 bootloader now in SVN resolves this issue for me.

Yes, the new bootloader fixes this problem for me too.

I will leave this task open until the final bootloader is released and more people would be able to verify it.

The latest h3x0 bootloader in svn doesn't boot my unit 50% of the time! The last thing it prints it the battery voltage. I get the feeling that disk spins up a bit earlier or later, so it may be related to the fact that I've got an 80GB disk installed (MK8007GAH)

Current SVN bootloader works for my H300+MK8007GAH.

However it fails to boot for petur's H300+MK8007GAH - crash appears to be in thread.c.

I'd guess that the change which broke it is r12514 or r12519 - maybe look at http://svn.rockbox.org/viewvc.cgi/trunk/bootloader/main.c?r1=12513;r2=12519;pathrev=12519