So, this is HeadRush Pedalboard, right?
In simple terms, the device is an embedded Linux appliance: a small single-board computer using RockChip RK3288 System-On-Chip, running some variant of Linux, and some userspace software providing the user interface.
at drivers/misc/usb5537.c:69/usb5537_write_config()
ERROR: board_usb_init: failed to probe usb5537@2d: ret=-121
USB5537 refers to an USB 3.0/2.0 7-port hub. In the image, you can see at least three of them exposed in pin headers (but without any pins); some of the unpopulated connectors near the top of the image are also USB ports.
However, I cannot find the kernel driver,
drivers/misc/usb5537.c, anywhere. It is obviously an out-of-tree driver for thus USB hub, not developed by open-source developers (definitely not kernel.org) and it is not in RockChip kernel trees either. It is the first point I'd start looking for a problem, but since the sources are nowhere to be found, it's a no go. Apparently, copyright law is not something HeadRush is interested in following.
None of the HeadRush firmware files or user guides mention Linux. The license they are breaking by not providing sources for the
modified Linux kernel they are using, is the only thing that permits them to use the Linux kernel in their product. This means that software-wise, none of the Linux kernel developers like myself can help you with this.
If I were you, I'd raise a fuckstink at HeadRush for breaking copyright law, and not providing the sources for the GPL-licensed software they are selling as a key part of this device.
![Rant :rant:](https://www.eevblog.com/forum/Smileys/default/rant.gif.pagespeed.ce.koOB_HE-VX.gif)
If you at some point do get the exact sources for at least the kernel used in the product –– and that means the exact ones, including the
usb5537.c file, and not just
"it's a 4.4 kernel" because the fuck it is; the 4.4 kernels do not have that file nor are able to produce that error output you are getting, and what is used on that thing is not a normal Linux kernel, but a modified one ––, then it would be a quick matter of checking out the usb5537 driver if it could be the source of the issues. Who knows, it might have an obfuscated check to stop operating after some (randomized) interval, just to make sure you guitarists keep buying new pedalboards.
(Really, it's more like how fragile the driver is wrt. timeouts, response intervals and such, that change as the hardware ages and things like capacitors degrade a little bit. It is easy to write a driver that works with perfect hardware, but a
proper driver that is robust against that kind of variation, requires common sense and a lot of experience. Which cheap developers that copy-paste code from the internet –– which is what the names shown in the log indicates; I can see similar names in microcontroller examples for controlling the USB5537B chip, which indicates that the "driver" is basically a copy-paste hack-job to get the hub chip "working". And because it is a low-grade hack, it often doesn't, causing a lot of lost money and wasted pedalboards. HeadRush doesn't mind, because you customers keep buying new devices to replace the "broken" ones.)
The reason I'm so angry is that customers don't care, and tend to just blame us open source developers for making "a crappy Linux that failed in my pedalboard". Because
we didn't, HeadRush did, by fucking with the kernel. The results you have in your hands, with HeadRush laughing at us all the way to the bank. And RockChip (the designers of the RK3x88 chips) happens to be one of the rare Chinese designer firms that do publish the changes needed for the Linux kernel to support their chips and even push changes to the upstream Linux kernel developers; companies like Allwinner just give a middle finger to everybody, because they cannot be sued in any Western country for copyright violations, they don't need to care. So, this is Annoying.
Besides, even if one would get full sources to the kernel, HeadRush will never ever publish enough sources for you to generate a replacement firmware fixing the software issue –– assuming it is one ––, so you're fucked with this thing anyway. (Even if somebody like me finds the bug in the driver, and provides a fix, what is the likelihood that HeadRush accepts the fix and produces a new firmware fixing the issues? They'd only be reducing their future sales, because their old products would be usable for longer.)
Unless, of course, it is a pure hardware issue. Which I don't think it is, because it works well enough to boot into the Linux kernel.
![Angry >:(](https://www.eevblog.com/forum/Smileys/default/xangry.gif.pagespeed.ic.0w4zhwjvDw.png)