VOLKER - Voice Of Logic Knowledge Experience & Responsibility's avatar
VOLKER - Voice Of Logic Knowledge Experience & Responsibility volker@nostrplebs.com 1 week ago
GM! Maybe I'm just stupid, or brainfogged from working through the night helping my friends copy with the CC bug, but isn't this logic wrong anyhow? If the board config defines HW_ENABLE_RNG as zero (logically: FALSE), and the SW later checks for it NOT being TRUE, then the hardware still isn't used, right? @NVK 🌞⚡️ @N Source: Kevin Kelbie's
coldcard.rip
A source-linked analysis of the COLDCARD random-number flaw, the affected firmware, and the Bitcoin sweeps that followed.
 image

Replies (2)

VOLKER - Voice Of Logic Knowledge Experience & Responsibility's avatar
VOLKER - Voice Of Logic Knowledge Experience & Responsibility volker@nostrplebs.com 1 week ago
Yes, I know about the #ifndef. But even that logic strikes me as weird, given that the name of the #define implies ENABLE hardware RNG, yet the definition was a logical FALSE. Which means, had the check been a proper if(!…) instead of #ifndef, then it still wouldn't use the HW RNG. And again, I may just be hung over, but this kind of logic inversion sometimes happens when programming, and in this particular case it's hard to detect, because the result is just loss of entropy. That's why I'm asking people who are more into the code, to review this part (I'm literally too old for that).
↑