[Bug 208911] Renesas USB controller - FW has invalid version :8224

[bugzilla-daemon@bugzilla.kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=208911

neoweb@hackspherelabs.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |neoweb@hackspherelabs.com

--- Comment #14 from neoweb@hackspherelabs.com ---
Hello,

I am not trying to waste anyone's time here.  This is what I had:

[    1.578294] xhci_hcd 0000:0a:00.0: FW has invalid version :8215
[    1.578331] xhci_hcd 0000:0a:00.0: Direct firmware load for
renesas_usb_fw.mem failed with error -2
[    1.578334] xhci_hcd 0000:0a:00.0: request_firmware failed: -2
[    1.578339] xhci_hcd: probe of 0000:0a:00.0 failed with error -2

I can find a ton of detail about this, I never knew that cards like this needed
firmware to be loaded on every boot.  I know FreeBSD will load firmware into
LSI Raid cards IIRC.  So it looks like from the comment of the final patch: 
"Test if ROM is present and loaded, if so we can skip everything" that my card
is going to use the firmware that is in the EEPROM.

I never knew you could load firmware into a USB 3.0 card like this.  It looks
like someone has even created a userspace util + SystemD setup for this:  

*https://mjott.de/blog/881-renesas-usb-3-0-controllers-vs-linux/
*https://github.com/markusj/upd72020x-load
*https://lore.kernel.org/patchwork/patch/686290/

So to get to my point.  I have noticed I have some issues sometimes with my
Renesas USB card manufactured by StarTech.  Just standard USB stuff, and issues
with pluggable USB 3.0 hubs...sometimes I have to reboot to get everything
working again, or reset the hub.

Am I doing myself a disservice not upgrading the firmware on it using the
Windows utils from Startech...or should I systemd with the github stuff above,
and upgrade the firmware every boot?

I don't know much about this stuff, but should linux always push a new firmware
to the card like CPU microcode, or FreeBSD.

Like I said, sorry if this is a waste of time or I am posting in the wrong
place, but it would be doing linux a disservice if all the Windows drivers
pushed newer better firmware to these devices, and worked better, because
firmware loading was just skipped.

Anyone have any idea?

-- 
You are receiving this mail because:
You are watching the assignee of the bug.