Announcement

Collapse
No announcement yet.

VirtIO Bluetooth Driver On The Way To The Linux Kernel

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

  • geearf
    replied
    Originally posted by czz0 View Post
    I don't think this is niche relative to VM desktop use

    Many people who game in a gpu passthrough VM setup also passthrough a bluetooth controller to connect game controllers.
    Why won't you just pass the USB instead and let the guest OS deal with bluetooth?

    Leave a comment:


  • edwaleni
    replied
    Originally posted by torsionbar28 View Post
    Because the use cases for USB are clear. What purpose is served by paravirtual BT devices? Bluetooth is consumer technology that has near-zero deployment in the datacenter. Not saying this new driver is a bad thing, it's not, but I'm not really sure what the point is.

    Edit: I'm thinking it must be for bored sysadmins who want to use a PS4 controller in a VM for some on-the-job gaming.
    Not all paravirtualization is done in the datacenter. With Covid some firms have rolled out secure VM's to their remote employees.

    Rather than have the AC98 emulated through the sound driver, the virt-bt interface allows BT headsets to be used directly and not get the dropouts, crackles or clipping (the goal I am assuming).

    I will have to see if there are any benefits for HVD users like in Citrix, or in the VMWare Horizon client where USB devices can be exposed, but audio devices are hooked.

    Leave a comment:


  • xen0n
    replied
    If anything this certainly can help testing the Bluetooth stack. Bluetooth depends on the network stack but uses its own address family, parallel to the IP ones (AF_INET and AF_INET6), so the original virtio network support probably wasn't enough. Being able to simulate Bluetooth chatter without mocking whole hardware devices seems nice, anyone who has written more than a few mocks for just about any unit test can attest to that...

    Leave a comment:


  • foolishgrunt
    replied
    Originally posted by torsionbar28 View Post
    I'm more interested in what real-world practical use cases there are. Things people are actually doing, or intend to do, with virtualized BT. That's where I'm just not seeing anything.
    Well, Intel sponsored this, and I doubt they would if they didn't plan on using it themselves.

    Leave a comment:


  • czz0
    replied
    I don't think this is niche relative to VM desktop use

    Many people who game in a gpu passthrough VM setup also passthrough a bluetooth controller to connect game controllers.

    Leave a comment:


  • microcode
    replied
    Sweet, this is actually useful to me.

    Leave a comment:


  • torsionbar28
    replied
    Originally posted by skeevy420 View Post
    Wanting to use cheap BT keyboard/mouse setups instead of KVM switches
    Who wants to do this? And why? How would it even work, you'd have to manually pair/unpair the devices from each and every server? Seems like a PITA vs using a traditional KVM switch.

    Originally posted by skeevy420 View Post
    and/or running lots of wires at a multi-VM workstation; simply not enough (free) USB ports
    A $15 USB hub solves for port count. Besides, Linux on the desktop is barely a thing. Multi-seat Linux on the desktop is a niche within a niche. I've read articles about it, but not once have I seen it in use anywhere. PC's are cheap these days, and individual tiny modern form factors make more sense than a single shared tower.

    Originally posted by skeevy420 View Post
    used as a makeshift audio device with headphones.
    Who is listening to audio, inside of a VM, with headphones, where the wearer of said headphones is in the vicinity of the hypervisor server?

    Originally posted by skeevy420 View Post
    I'm sure other people have more ideas than PS4 controllers (which is also my use-case).
    I'm sure people have creative ideas. After all, you can implement RAID-5 on a stack of USB floppy drives. I'm more interested in what real-world practical use cases there are. Things people are actually doing, or intend to do, with virtualized BT. That's where I'm just not seeing anything.

    Leave a comment:


  • skeevy420
    replied
    Originally posted by torsionbar28 View Post
    Because the use cases for USB are clear. What purpose is served by paravirtual BT devices? Bluetooth is consumer technology that has near-zero deployment in the datacenter. Not saying this new driver is a bad thing, it's not, but I'm not really sure what the point is.

    Edit: I'm thinking it must be for bored sysadmins who want to use a PS4 controller in a VM for some on-the-job gaming.
    Wanting to use cheap BT keyboard/mouse setups instead of KVM switches and/or running lots of wires at a multi-VM workstation; simply not enough (free) USB ports; used as a makeshift audio device with headphones. I'm sure other people have more ideas than PS4 controllers (which is also my use-case).

    Leave a comment:


  • torsionbar28
    replied
    Originally posted by edwaleni View Post
    Niche perhaps, but desirable. If can one can paravirtualize access to USB devices, why not bluetooth?
    Because the use cases for USB are clear. What purpose is served by paravirtual BT devices? Bluetooth is consumer technology that has near-zero deployment in the datacenter. Not saying this new driver is a bad thing, it's not, but I'm not really sure what the point is.

    Edit: I'm thinking it must be for bored sysadmins who want to use a PS4 controller in a VM for some on-the-job gaming.
    Last edited by torsionbar28; 07 April 2021, 09:51 AM.

    Leave a comment:


  • edwaleni
    replied
    Niche perhaps, but desirable. If can one can paravirtualize access to USB devices, why not bluetooth?

    v4 of Bluetooth may be marginal, but this might work better with BT v5 with its range and performance enhancements.

    Leave a comment:

Working...
X