Announcement

Collapse
No announcement yet.

DualShock 4 Controller Support Being Dropped From HID-Sony In Favor Of New Driver

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • DualShock 4 Controller Support Being Dropped From HID-Sony In Favor Of New Driver

    Phoronix: DualShock 4 Controller Support Being Dropped From HID-Sony In Favor Of New Driver

    With the Linux 6.2 kernel Sony DualShock 4 controller support was added to the "hid-playstation" driver as the newer open-source driver maintained by Sony and started out originally in supporting the PlayStation 5 DualSense controllers. The DualShock 4 controller was long supported by the "hid-sony" driver while now in the Linux 6.3 kernel that support will be removed...

    https://www.phoronix.com/news/HID-Sony-Dropping-DS4

  • #2
    That DS4 support is removed from hid-sony doesn't surprise me, however that it is removed without any kind of depreciation period is interesting.
    Originally posted by phoronix View Post
    Many were wondering if this was going to happen since normally one piece of hardware isn't supported by two different drivers within the Linux kernel.
    That this is the norm is often claimed, but there are numerous historic and current examples where two different mainline drivers support the same hardware.
    • e100 vs. eepro100, one of the more well-known historic examples, where removal of eepro100 faced some resistance
    • b43 vs. brcmsmac
    • cciss vs. hpsa
    • fbdev vs. drm, though this was necessary due to different featuresets
    • radeon vs. amdgpu
    • thinkpad_acpi vs. various generic ACPI drivers
    • etc. (certainly there are more that I don't remember right now)

    Comment


    • #3
      Originally posted by chithanh View Post
      That DS4 support is removed from hid-sony doesn't surprise me, however that it is removed without any kind of depreciation period is interesting.
      That this is the norm is often claimed, but there are numerous historic and current examples where two different mainline drivers support the same hardware.
      • e100 vs. eepro100, one of the more well-known historic examples, where removal of eepro100 faced some resistance
      • b43 vs. brcmsmac
      • cciss vs. hpsa
      • fbdev vs. drm, though this was necessary due to different featuresets
      • radeon vs. amdgpu
      • thinkpad_acpi vs. various generic ACPI drivers
      • etc. (certainly there are more that I don't remember right now)
      It's more or less just a driver in-between evdev and the backend (usb or bluetooth).

      no visible changes in interface, nor would anyone actually need to specify the driver name. (Assuming there are no weirdly specific hacks included in either one).

      ​​​​

      Comment


      • #4
        Originally posted by chithanh View Post
        [*]radeon vs. amdgpu
        that's a side effect of newer architectures getting a new driver, itself being a new software architecture, old hw using radeon while new hw using amdgpu

        for the specific generation that happened to be able to overlap both, isnt/wasnt amdgpu more of a beta?

        Comment


        • #5
          kn00tcn

          But that is the same situation for Sony controllers:
          DualShock 3 old driver only
          DualShock 4 both drivers
          DualSense new driver only.

          Same btw. for cciss vs. hpsa

          In fact, radeon supports two GPU generations that amgdpu also supports, namely SI (Tahiti etc.) and CI (Hawaii etc.). SI took a long time with much deliberation to become first class member of amdgpu, but CI support was always fully there.

          Comment

          Working...
          X