4 ms·
I wonder if this could theoretically happen for the main display in front of the steering wheel. Are there safety standards carmakers have to certify to when it
by zwieback 3y ago
I wonder if this could theoretically happen for the main display in front of the steering wheel. Are there safety standards carmakers have to certify to when it comes to speedometer, etc.?
- bri3d 3y agoYes. Extremely strict ones having to do with the visibility, reliability, and availability of certain UI elements as well as warning lights. In the US, this is all governed by the FMVSS (Federal Motor Vehicle Safety Standards). This is part of the reason why you'll still see separate warning lights or really weird/ugly overlays for some warnings on modern digital dashboards - in many situations, the safety critical / regulated information is actually being blitted to the screen by a _separate safety processor!_ running an RTOS, while the whiz-bang main display is powered by some commodity SoC running Linux or QNX.
- KMag 3y agoI really wish smart phone display chipsets had this blit overlay functionality for TOTP (including a per-device symmetric encryption key for importing/exporting TOTP secrets such that the main processor can deny itself future access to the secret, but the phone can still generate TOTP codes, without the TOTP processor needing to store all of the secrets itself). We've seen multiple secure enclave exploits. It's too bad some of this security functionality isn't baked into the display chipset. Yes, it's probably ugly (even if the chipset allows some forms of theming the TOTP code display), but it's better than having your brokerage account, your work login, etc. all exploited due to some 0day. It's also better than carrying around a handful of RSA physical tokens on your keychain (modulo phone breaking/being stolen/running out of battery).