Steelseries SRW-S1 support?

[mungewell]

[Nic KERS]

[mungewell]

Woo Hoo!

Very glad it's working for you, and hopefully we can get this into the mainline kernel so others will get the benefit too... I'm willing to do the leg work on that, but it's normal to credit testers with a 'tested-by' comment. If you'd like to be mentioned can you please PM me your email address?

The deadzone might be a generic problem with Linux - I think that it puts a default value on joysticks unless told otherwise. Check the reported values in jstest verses the values within the HID stream to confirm that it's not the wheel doing it. There is the jscal app which (is supposed) to fix the problem. I also found this util (G25 Manage) which can set the deadzone too:


Regarding Axis3/4 - hats are reported as 2D axis under Linux. If you want to used these for buttons use the 'ATOB' mode in the SD configuration page.

Also it would be nice to confirm all the buttons/dials report OK (in-case there are other changes to the HID descriptor needed) - in helping you out this week I noticed my G27 shifter also contains a X/Y joystick function that is not reported to OS!!


Once you get SD running I'd be interested in your opinions on how well this device works in game, and if the calibration menu needs any additional controls in order to get this set up nicely.

Simon

PS Does the SRW-S1 support any form of rumble under Windows and do the RPM LEDs have variable brightness?

[mungewell]

With looking at the pictures on:


The 'assist' dials don't seem to be reported (probably part of that 'vendor-defined' bit), can you check the HID stream for how these actually report. We might be able to get them working too.

Simon

[Nic KERS]

[mungewell]

[Nic KERS]

That video's mad! It's a different model wheel too. Mine hasn't got a gear number display or whatever he's flicking with his thumbs - unless he's just quick on the buttons. The cockpit driver presses buttons!

I'll do those things you mentioned to try, tomorrow. Thanks again.

(Ah. Got it now. Just reading the link.)

[mungewell]

I think the guy has stuck another LED module on the front of it.... he's obviously very into sim racing!! (not that there's anything wrong with that).

My 3.6rt kernel built fine and appears to work, so I suggest you double check for slip ups.
Simon.

[Nic KERS]

[Nic KERS]

Right. First difference is in the first line of the entries in syslog when it's plugged in.

3.8 (working):
Jan 17 14:47:02 z68 kernel: [ 90.042745] usb 1-1.5: new full-speed USB device number 5 using ehci-pci

3.6 (not working):
Jan 17 19:19:01 z68 kernel: [ 6778.815362] usb 3-1.5: new full-speed USB device number 5 using ehci_hcd

So 3.6 wants to use ehci_hcd instead of ehci-pci

Probably not right.

Then the mtp-probe goes for bus 1, device 5 on 3.8 but bus 3 device 5 on 3.6 The different bus designation is reflected in the rest of the syslog entry. There are no other differences (between the syslog entries).

I'll add more as I find it.

(I didn't get far with make oldconfig and menuconfig. The latter is too hard for me. Too many questions...)