Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
×
Operating Systems Privacy Linux Hardware

Dell Adding Hardware Privacy Driver For Linux (phoronix.com) 46

Posted by BeauHD from the hardware-kill-switches dept.
According to Phoronix, a Dell privacy driver is is being prepared for the Linux kernel. From the report: Beginning in Dell's 2021 laptop models they are providing hardware-based "privacy buttons" to disable microphone and camera support. These new Dell privacy buttons are basically hardware kill switches for the microphone and web camera video stream. The Dell privacy driver sent out on Tuesday for the Linux kernel is about manipulating the relevant LEDs and tracking the status of the hardware-based controls where as the actual toggling of the audio/video support is handled by the hardware.

The Dell privacy driver in its current form is talked about for the camera and microphone support but the patch does also note a "PRIVACY_SCREEN_STATUS" bit as well. Presumably they will be extending this privacy driver as well for privacy screen handling around reducing the horizontal/vertical viewing angles of the display. The dell-privacy Linux driver in its initial form can be found via the kernel mailing list. It's great seeing Dell working on this driver punctually for Linux ahead of their next-gen laptops.
This discussion has been archived. No new comments can be posted.

Dell Adding Hardware Privacy Driver For Linux More

Dell Adding Hardware Privacy Driver For Linux

Comments Filter:

  • If it is hardware kill switch... (Score:4, Informative)

    by sinij ( 911942 ) on Wednesday November 04, 2020 @09:32PM (#60685688)
    ... why does it need a driver?
    • Indeed... On something the scale of a laptop, electrical cutout or buzz off. Preferably a mechanical shutter for the camera, as they are still dust/smudge/scratch magnets. If your OS takes more than a few seconds to reinitialize your device, it (either, I guess) is probably broken.
      • Not to say that soft RF kill switch should be excluded - just that an electrical disconnect is the only way to be reasonably sure that it's off (excluding various sorts of tampering)
      • Lenovo laptops have these, there's a mechanical shutter that both covers the camera and disconnects it electrically (the USB device disappears). However even if some evil malwarez is somehow faking out the USB-disconnect, the fact that there's a shutter over the camera means it still can't be used for anything.

    • Re: (Score:2)

      by dohzer ( 867770 )

      Oh, the kill switch *is* a hardware kill switch. But the driver allows government actors to enable the software-controlled "revive switch". It's essentially another power switch in parallel with the switch you thought would totally disable the camera and mic.

      • Re: (Score:3, Interesting)

        by rtb61 ( 674572 )

        You need a driver to let you know when it is on not off. Off would be the default, on is the active. Simply place the switch as close to the camera and microphone, part of the housing preferably and red is on and nothing is off (a switch on either side). The software driver should let you know of change for a while, how long, your choice, permanently on screen whilst it is on or just for 30 seconds at change state and on boot and return from sleep. Generate a real awareness of camera and microphone off or o

    • Re:If it is hardware kill switch... (Score:5, Informative)

      by fahrbot-bot ( 874524 ) on Wednesday November 04, 2020 @09:57PM (#60685746)

      ... why does it need a driver?

      (sigh)

      From TFS: [emphasis mine]

      The Dell privacy driver sent out on Tuesday for the Linux kernel is about manipulating the relevant LEDs and tracking the status of the hardware-based controls where as the actual toggling of the audio/video support is handled by the hardware.

      • Comment removed (Score:5, Insightful)

        by account_deleted ( 4530225 ) on Wednesday November 04, 2020 @10:01PM (#60685756)
        Comment removed based on user account deletion

        • "If the driver can toggle indicator leds and it is button based, then the leds cannot be trusted, the leds should be controlled by hardware too."

          Sure they can.

          When the switch is in the position where the camera has power, the LEDs have no power. When the switch (or button, if that's what's hanging you up) is in the position where the camera has power, the LED has no power. No problem. If I have the camera off 99% of the time like so many people will do, I don't want the glowing light to be mandatory, and be

          • Comment removed based on user account deletion

            • "Button generally implies momentary switch. And again, if the driver can control the led then you have to trust software in addition to the hardware."

              Button can imply whatever, the device doesn't exist yet, and Dell didn't write the word button. It's really silly to speculate much on.

              But no, nothing fundamental about the problem requires you to trust the software if there's an LED that can be controlled by software.

              It's very simple:

              Switch, button, or what have you in positon 1:

              LED has power, webcam does not

              • Though actually, the source code says what the button is, and screw that, Purism still has the best design.

              • Re: (Score:2)

                by AmiMoJo ( 196126 )

                Looking at the source code it appears that they have a hardware switch for the camera and another for the microphone. There is some SMBUS device (presumably a little MCU) that generates keypress events when the switches are changed.

                The "switches" could be momentary buttons that the MCU reads, and then uses to control some FETs or whatever to connect and disconnect the mic and camera. But it's a shame that the LEDs are not connected to the MCU, or the physical switches if that's what they are. I suppose Dell

            • If it's a hardware kill switch, it MUST be a on-off toggle switch rather than momentary. The toggle mechanism may be mechanical (like those pop-out=off toggle buttons) or a hard-wired electrical toggle switch controlled by a mechanical momentary switch, but it can't be software.

              Though honestly, I don't trust what seems to be the the article's interpretation of "hardware-based privacy buttons" as "basically hardware kill switches". It sounds to me like it could easily be a dedicated hardware button that

              • Re: (Score:3, Insightful)

                by airport77 ( 7359656 )
                > camera on = Privacy Mode LED of This is an usability disaster, a sure recipe for making it useless for most users. A laptop's webcam is suposed to be off at all times, except the few times I need to use it. An indicator that is usually on and occasionally goes off is much more dificult to spot.

                • I absolutely agree - but it would work as TCW suggested.

                  It would work much better if it were
                  camera on = LED on
                  camera off = LED software controlled
                  The important feature is just that the software can't affect the indicator while the camera is on.

                  Presumably nobody would want to disable the privacy indicator by making the LED always on, but *blinking* the LED when a camera is (trying to) access the camera while it's in privacy mode could be very convenient reminder to users that their camera isn't working becau

            • Re: (Score:2)

              by hawk ( 1151 )

              No.

              Things like:

              "16MM ILLUMINATED PUSHBUTTON - BLUE LATCHING ON/OFF SWITCH"

              https://www.adafruit.com/produ... [adafruit.com]

              personally, though, I'd prefer a slide switch for a privacy circuit.

              For that matter, I'd prefer a spring-loaded switch with a magnet to keep it on, such that it automatically ends up off when power is lost . . .

              hawk

          • Comment removed based on user account deletion
    • Perhaps if you read the summary you would know the answer, as the purpose of the driver is explained quite thoroughly there.
    • Comment removed based on user account deletion

  • re basically hardware kill switches

    These are soiftware kill switches; they wouldn't need a fucking driver if they were done right - in hardware.

    • The kill switch is hardware based, the driver is only able to detect what state the switch is in, but can't change the state because it's not a software kill switch.

      Now the fact that the driver could be in charge of a status led means we might be able to fool a user into thinking the camera is on when it's off or vice versa. I'd really hope for an led that is tied into the camera hardware rather than separately controllable.

    • The speakers can be dynamically turned into microphones. And the light sensor can be handy. Dell missed it chance to turn off the CPU's management engine and hard wire a cut to the wifi and BT and camera V+ or power lines. My expectation would be jumpers or switches to switch off trusted chip crap. Then on mobiles the binary blob for the phones modem chip will respond to crafted SMS messages containing character codes not normally inputtable. A Dell rep once told me they throw out specs, and seek the lowest
  • on any of my Linux systems.

    • Because FBI. That's why. You are a patriot aren't you? Well then.

    • Status (Score:3)

      by JBMcB ( 73720 )

      So that when the hardware kill switch is engaged, your v4l driver knows that the reason it has stopped receiving data is because the hardware kills switch was engaged, and not because the hardware died, or was disconnected, or for some other reason.

  • Toobin? (Score:5, Funny)

    by h33t l4x0r ( 4107715 ) on Wednesday November 04, 2020 @10:01PM (#60685754)
    Do we all agree to call this a Toobin button?
  • Comment removed based on user account deletion

    • That's a light to indicate the camera is on, not a light to indicate the camera is off. The problem is thoroughly different.

      There's a physical switch. In one position, the camera and microphone has power, but the LED has no power. In the other position it's reversed.

      If when the camera is off, I don't want that LED on 100% of the time, turning it off in no way compromises my privacy -- I'm still relying on the position of the switch. Further, since the switch physically cuts power to the LED when the camera

      • Comment removed based on user account deletion

        • I'm not really sure what you're arguing. I mean, is a switch better than a button? Sure, and who the hell knows what the unlreleased hardware looks like. The point is, that video is about something quite different. It's just as different regardless of what the button, switch, toggle, widget, or whatever else you might call it looks like. There is no "but" here.

        • Re: (Score:2)

          by piojo ( 995934 )

          the summary specifically mentions a button which implies a momentary switch.

          I'm splitting hairs a little bit, but a momentary switch can be made to physically cut power, like by changing the state of a flip flop circuit. Ideally it would be designed so the button needs to be held for one second for the flip flop to be turned back on. This would not require software controls, but it would increase the complexity of the chip.

  • I have a laptop from 2006 with one of these (Score:4, Interesting)

    by spiritplumber ( 1944222 ) on Wednesday November 04, 2020 @11:23PM (#60685922) Homepage
    I have a laptop from 2006 with one of these switches. It's a little slide switch that turns off the 5V to the webcam. The only mildly annoying thing is that you have to detect the device again as if you plugged in an external webcam.
    • I disable the webcam on my laptop at startup by a script that poweris it off if it is on (this is on Linux, and it shows up as a USB camera).

      I can get it back by using the keyboard switch (Fn+F10) and since it's appearing on USB it's automatically detected and enabled.

  • Looks very promising for intruders!

  • The hardware LED is controlled separately from the hardware switch?? By software??

    Tbis is not a hardware kill switch. This is bullshit.

    That software or even firmware can control it, is utter bullshit.

    A hardware kill switch is something that *physically cuts the wires*.
    In this case to the camera and microphone.

    (It should also cut access to anything else that can vibrate, and whose vibrations can be measured in software/firmware, like possibly hard disk head.)

  • Their stickers will lose their computer jobs.

  • Um. Please, someone disassemble the driver.
  • No they have a Linux driver for Privacy hardware, it'd be nice to see one for the TPM and the fingerprint reader.

Slashdot Top Deals

"No matter where you go, there you are..." -- Buckaroo Banzai

Close