On SteamOS 2.0 is better a kernel 4.4 than 4.1 ?

Cordialy

Is SteamOS using the default .config file from Debian or will you provide your own?

You can revert commit fledermaus@7d4732c in Brewmaster. The hack was needed for the LEDs and rumble to work via Bluetooth on kernel 3.10, but as of kernel 3.15 the HID system underwent some major work to add proper support for these devices and the workaround is now unnecessary.

I see that you have taken the updates from 4.1.30 in your brewmaster-4.1-proposed branch. This contains a fix for the oops when attached an unknown Xbox One gamepad. This fix is 07761df

However, 4.1.30 doesn't contain an earlier fix also marked for stable that fixes functionality on controllers that aren't in xpad's VID/PID table. Please consider porting: torvalds/linux@c7f1429

The side effect (in addition to the oops fix) is that xpad correctly derives the XTYPE constant from the interface protocol. Without that code, any unknown Xbox One controller is treated like a wired Xbox 360 controller.

It's not a big deal that c7f1429 wasn't merged into 4.1, since the upstream driver was missing the longer initialization packet that 2015 and later firmware requires. It's likely that the majority of unrecognized controllers wouldn't work anyway. SteamOS xpad could benefit though, since it was updated to support the newer controllers.

You can reproduce the condition that c7f1429 fixes by removing the new Xbox One S controller VID/PID entry from the device table in xpad in brewmaster-4.1-proposed and attaching the controller. Without the fix, the Xbox One S controller will not work (though the oops is fixed). With that commit (and in the mainline kernel), the controller works correctly without a VID/PID entry.

steamos_kernel Github releases has been outdated for a while... :(

Greetings,

Thank you for your amazing job. For real, I enjoy what you've done so far with Debian GNU/Linux.
I would enjoy the latest improvements from the kernel 4.15, and I hope Valve will update the SteamOS kernel anytime soon.

So here is my request/question, is there any ETA for the next kernel update ?

Regards

I work for Rivet networks whom develops and distributes the Killer Networking products. We are working to upstream this change to mainline linux Kernel, but in the meantime can we have this patched in SteamOS Kernel also?

Add Killer E2400 device ID in alx driver.

Signed-off-by: Owen Lin [email protected]

diff -uprN alx_orig/main.c alx/main.c
--- alx_orig/main.c Mon Sep 7 13:00:58 2015
+++ alx/main.c Mon Sep 7 14:47:03 2015
@@ -1537,6 +1537,7 @@ static const struct pci_device_id alx_pc
{ PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_AR8162),
.driver_data = ALX_DEV_QUIRK_MSI_INTX_DISABLE_BUG },
{ PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_AR8171) },

• { PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_E2400) },
  { PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_AR8172) },
  {}
  };
  diff -uprN alx_orig/reg.h alx/reg.h
  --- alx_orig/reg.h Mon Sep 7 13:00:58 2015
  +++ alx/reg.h Mon Sep 7 14:46:16 2015
  @@ -39,6 +39,7 @@

define ALX_DEV_ID_E2200 0xe091

define ALX_DEV_ID_AR8162 0x1090

define ALX_DEV_ID_AR8171 0x10A1

+#define ALX_DEV_ID_E2400 0xe0A1

define ALX_DEV_ID_AR8172 0x10A0

/* rev definition,

MSI are claiming out-of-the-box SteamOS Support (with a capital S) for their GAMING-series motherboards, e.g. http://www.msi.com/product/mb/X99A-GAMING-7.html#hero-overview

The OEM requirements at https://github.com/ValveSoftware/SteamOS/blob/master/OEM.md say hardware should use in-tree kernel modules only, for all hardware.

The hardware sensor chip used by MSI on their recent hardware is NOT supported by the SteamOS kernel - it requires a 3.12 kernel (or backports from post-Alchemist nct6775.c)

I have tested the brewmaster_beta 2.91 and found there is still otp failure.
There is a little difference between my patch and upstream patch.

My patch is as below.

• /* Some firmware revisions have an incomplete WoWLAN implementation
• * despite WMI service bit being advertised. This feature flag is used
• * to distinguish whether WoWLAN is really supported or not.
• */
• ATH10K_FW_FEATURE_WOWLAN_SUPPORT = 6,
• /* Don't trust error code from otp.bin */
• ATH10K_FW_FEATURE_IGNORE_OTP_RESULT,

Upstream patch doesn't inculde ATH10K_FW_FEATURE_WOWLAN_SUPPORT.
Please help to check this.

BR,
Owen

I noticed on line 635 in xpad.c that you call xpad360w_process_packet... however, this function allocates memory using GFP_KERNEL. This can cause a sleep which is an issue in interrupt context.

I did a fresh install of steam os and the bluetooth adapter is not found.

I did some research I and it looks like similar issues were found by other users, but after reviewing the code it looks like this should be patched.
https://bbs.archlinux.org/viewtopic.php?id=126487

Any help would be appreciated.

uname -a

desktop@steamos:~$ uname -a
Linux steamos 4.1.0-0.steamos.3-amd64 #1 SMP Debian 4.1.6-2+steamos3 (2015-10-22) x86_64 GNU/Linux

lsusb

desktop@steamos:~$ lsusb
Bus 009 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 008 Device 002: ID 28de:1142
Bus 008 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 001 Device 003: ID 0cf3:9271 Atheros Communications, Inc. AR9271 802.11n
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 007 Device 003: ID 046d:c52b Logitech, Inc. Unifying Receiver
Bus 007 Device 004: ID 0cf3:3002 Atheros Communications, Inc. AR3011 Bluetooth
Bus 007 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 006 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 005 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

dmesg | grep ath

desktop@steamos:~$ dmesg | grep ath
[    2.073278] usb 7-3: firmware: direct-loading firmware ath3k-1.fw
[    2.099890] usb 1-4: ath9k_htc: Firmware htc_9271.fw requested
[    2.100147] usbcore: registered new interface driver ath9k_htc
[    2.384093] usb 1-4: ath9k_htc: Transferred FW: htc_9271.fw, size: 50980
[    2.438854] usbcore: registered new interface driver ath3k
[    2.623930] ath9k_htc 1-4:1.0: ath9k_htc: HTC initialized with 33 credits
[    3.850998] ath9k_htc 1-4:1.0: ath9k_htc: FW Version: 1.3
[    3.851004] ath9k_htc 1-4:1.0: FW RMW support: Off
[    3.851008] ath: EEPROM regdomain: 0x60
[    3.851011] ath: EEPROM indicates we should expect a direct regpair map
[    3.851015] ath: Country alpha2 being used: 00
[    3.851016] ath: Regpair used: 0x60
[    4.078937] usb 7-3: firmware: direct-loading firmware ath3k-1.fw
[    9.073645] ath3k: probe of 7-3:1.0 failed with error -110

lsmod

desktop@steamos:~$ lsmod
Module                  Size  Used by
btusb                  45056  0
btbcm                  16384  1 btusb
btintel                16384  1 btusb
ctr                    16384  2
ccm                    20480  2
arc4                   16384  2
bnep                   20480  2
snd_hda_codec_hdmi     53248  1
joydev                 20480  0
hid_logitech_hidpp     20480  0
nls_utf8               16384  1
nls_cp437              20480  1
vfat                   20480  1
fat                    65536  1 vfat
ath9k_htc              61440  0
ath9k_common           28672  1 ath9k_htc
ath9k_hw              425984  2 ath9k_common,ath9k_htc
ath                    28672  3 ath9k_common,ath9k_htc,ath9k_hw
mac80211              569344  1 ath9k_htc
hid_logitech_dj        20480  0
cfg80211              458752  4 ath,ath9k_common,mac80211,ath9k_htc
ath3k                  20480  0
bluetooth             442368  9 bnep,ath3k,btbcm,btusb,btintel
kvm_amd                61440  0
rfkill                 20480  5 cfg80211,bluetooth
kvm                   425984  1 kvm_amd
crct10dif_pclmul       16384  0
crc32_pclmul           16384  0
snd_hda_codec_realtek    73728  1
snd_hda_codec_generic    65536  1 snd_hda_codec_realtek
ghash_clmulni_intel    16384  0
snd_hda_intel          28672  5
aesni_intel           172032  4
efi_pstore             16384  1
snd_hda_controller     28672  1 snd_hda_intel
aes_x86_64             20480  1 aesni_intel
lrw                    16384  1 aesni_intel
snd_hda_codec          94208  5 snd_hda_codec_realtek,snd_hda_codec_hdmi,snd_hda_codec_generic,snd_hda_intel,snd_hda_controller
gf128mul               16384  1 lrw
snd_hda_core           24576  5 snd_hda_codec_realtek,snd_hda_codec_hdmi,snd_hda_codec_generic,snd_hda_codec,snd_hda_controller
glue_helper            16384  1 aesni_intel
ablk_helper            16384  1 aesni_intel
snd_hwdep              16384  1 snd_hda_codec
cryptd                 20480  3 ghash_clmulni_intel,aesni_intel,ablk_helper
psmouse               114688  0
snd_pcm                90112  4 snd_hda_codec_hdmi,snd_hda_codec,snd_hda_intel,snd_hda_controller
pcspkr                 16384  0
snd_timer              28672  1 snd_pcm
serio_raw              16384  0
efivars                20480  1 efi_pstore
k10temp                16384  0
i2c_piix4              24576  0
snd                    69632  18 snd_hda_codec_realtek,snd_hwdep,snd_timer,snd_hda_codec_hdmi,snd_pcm,snd_hda_codec_generic,snd_hda_codec,snd_hda_intel
soundcore              16384  1 snd
8250_fintek            16384  0
tpm_infineon           20480  0
tpm_tis                20480  0
shpchp                 32768  0
acpi_cpufreq           20480  0
tpm                    32768  2 tpm_tis,tpm_infineon
processor              28672  5 acpi_cpufreq
evdev                  24576  16
thermal_sys            36864  1 processor
nvidia               8593408  100
drm                   278528  4 nvidia
fuse                   90112  3
autofs4                36864  2
hid_generic            16384  0
usbhid                 49152  0
hid                   110592  4 hid_generic,usbhid,hid_logitech_dj,hid_logitech_hidpp
ext4                  499712  3
crc16                  16384  2 ext4,bluetooth
mbcache                20480  1 ext4
jbd2                   90112  1 ext4
sg                     32768  0
sd_mod                 40960  6
ohci_pci               16384  0
mxm_wmi                16384  0
ahci                   36864  5
libahci                28672  1 ahci
crc32c_intel           24576  0
libata                200704  2 ahci,libahci
scsi_mod              200704  3 sg,libata,sd_mod
xhci_pci               16384  0
ohci_hcd               45056  1 ohci_pci
ehci_pci               16384  0
xhci_hcd              155648  1 xhci_pci
ehci_hcd               77824  1 ehci_pci
usbcore               200704  14 ath3k,btusb,ohci_hcd,ohci_pci,ehci_hcd,ehci_pci,usbhid,ath9k_htc,xhci_hcd,xhci_pci
r8169                  77824  0
mii                    16384  1 r8169
usb_common             16384  1 usbcore
wmi                    20480  1 mxm_wmi
button                 16384  0

Hey guys,

I got strange freezes while playing DiRT: Showdown with my wireless XBox 360 controller.

I posted this first here at Virtual Programming:
https://github.com/virtual-programming/dirtshowdown-linux/issues/5#issuecomment-139026856

Is it true that this could be a Kernel bug? Who would be able to fix this?

when will steamos3 come?

We have tried 2.93 and Killer1535 bluetooth can't work.
We found ther is no qca Bluetooth firmware in /lib/firmware folder.
Please refer to below link.
https://git.kernel.org/cgit/linux/kernel/git/firmware/linux-firmware.git/tree/qca

BR,
Owen

Can you make the SteamOS 3.0 Kernel open-source on Github?

Ath10k driver is present in the current steamOS kernel, but the necessary firmware files are missing. We have verified these steps work to have a Killer 1535 Wireless module work with a clean install of Brewmaster 2.70

Download:
http://www.killernetworking.com/support/K1535_Debian/board.bin
and put it in the /lib/firmware/ath10k/QCA6174/hw3.0/ folder (create the folder if it doesn’t exist)

Download:
http://www.killernetworking.com/support/K1535_Debian/firmware-4.bin
and put it in the /lib/firmware/ath10k/QCA6174/hw3.0/ folder as firmware-4.bin.

Create config file with:

sudo nano /etc/modprobe.d/ath10k.conf

and put the following contents into the file: options ath10k_core skip_otp=y

Reboot.

If a different method is preferred/verified, we would be fine with that as well.

Thanks,
Thomas
Rivet Networks

Essentially, what I am seeing in this repo is the base Debian repository with the addition of new driver files (which is great). However, the base Debian repository and distribution therein has way too many old drivers that have nothing to do with Steam, nor will Steam execute on that hardware of that era. Also, you have chipsets under the Arch folder which is designed so that you can port the OS over to each of these different system. Problem is, some of these systems don't have enough memory to even run Steam, let alone games created for it.

My suggestion is to cut the old technology files and architectures you never plan on supporting, which will make your project smaller and even more customized. There will be a lot less "configure" flags to fiddle around with. By chopping out the old dead wood, it should make it a lot easier to maintain.

An example of what I am taking about is the EISA busline, it predates the PCI and PCI-E. As you've stated the recommended hardware, will have PCI-E slots in it, the least you should support are the original PCI slots, nothing older. That will also cut down on the amount of CPUs you need to support as well because you find certain families with the minimum of PCI v1.00 , Intel AGP and beyond.

Since this repository is not set as fork of torvalds/linux it makes it complicated to see what the actual changes are -> Requesting to add as fork or provide comparison (might be usefull for distro maintainers who allow kernel compilation).

Copypasting from the 2.117 announcement thread...

There appears to be a regression from the kernel's recent updates/changes to hid-sony.c. Connecting a DS4 via Bluetooth causes the system (and the controller) to lock up; here's the relevant dmesg output:

https://gist.github.com/flibitijibibo/1369c7da5b922983c7fbdcf336ea3921

It's reproducible with a USB connection, and I've duplicated this on a different PC running 2.117. AFAIK this is 100% reproducible with an original DualShock 4 controller.

MSI are claiming out-of-the-box SteamOS Support (with a capital S) for their GAMING-series motherboards, e.g. http://www.msi.com/product/mb/X99A-GAMING-7.html#hero-overview

The OEM requirements at https://github.com/ValveSoftware/SteamOS/blob/master/OEM.md say hardware should use in-tree kernel modules only, for all hardware.

The Atheros/Killer wifi chip used by MSI on some of their recent hardware is NOT supported by the SteamOS kernel - it requires a 4.0 kernel (or backports from post-Alchemist ath10k/). New firmware is also required for this hardware, which is not included in the firmware packages in SteamOS.

I work for Rivet networks whom develops and distributes the Killer Networking products. We are about to launch (in the coming weeks), a new E2500 product and we would like to ensure these systems can load SteamOS/Linux. We are working to upstream this change to mainline linux Kernel, but in the meantime can we have this patched in SteamOS Kernel also?

Add Killer E2500 device ID in alx driver.

Signed-off-by: Thomas Wooden [email protected]

diff -uprN alx_orig/main.c alx/main.c
--- alx_orig/main.c Mon Aug 22 12:05:12 2016
+++ alx/main.c Mon Aug 22 12:27:47 2016
@@ -1537,6 +1537,7 @@ static const struct pci_device_id alx_pc
{ PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_AR8162),
.driver_data = ALX_DEV_QUIRK_MSI_INTX_DISABLE_BUG },
{ PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_AR8171) },

{ PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_E2400) },
{ PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_E2500) },
{ PCI_VDEVICE(ATTANSIC, ALX_DEV_ID_AR8172) },
{}
};
diff -uprN alx_orig/reg.h alx/reg.h
--- alx_orig/reg.h Mon Aug 22 12:05:12 2016
+++ alx/reg.h Mon Aug 22 12:27:47 2016
@@ -39,6 +39,7 @@

define ALX_DEV_ID_E2200 0xe091

define ALX_DEV_ID_AR8162 0x1090

define ALX_DEV_ID_AR8171 0x10A1

+#define ALX_DEV_ID_E2400 0xe0A1
+#define ALX_DEV_ID_E2500 0xe0B1

define ALX_DEV_ID_AR8172 0x10A0

/* rev definition,