The cyber news machine never stops cranking away, so check out today's Metacurity for the most critical infosec developments you should know, including
--The US, UK, and Australia sanction Russian bulletproof hosting providers,
--Alice Guo sentenced to life for running scam compound,
--Samourai Wallet operator sentenced to four years,
--SK Telecom rejects breach mediation proposal,
--USBP is massively surveilling Americans,
--Trump to preempt state AI …
OK so summary of findings on the A35 bringup attempts so far:
First, resetting the A35 subsystem via RCC.C1RSTCSETR does not seem to be the way to go. It disconnects the debugger and a bunch of bad things happen. Probably possible to recover from this, CPU2 is still running fine, but it's not getting me closer to a solution.
Alice Guo, Chinese national who ran huge scam centre while Philippines mayor, sentenced to life in prison
https://www.theguardian.com/world/2025/nov/20/alice-guo-chinese-nat…
As usual, STMicro makes silicon I mostly like, but I don't like their packages.
Here's the STM32MP257 TFBGA436 package with balls color coded (somewhat roughly) by power domain/function. I grouped some similar rails together e.g. all low voltage digital core balls are the same color (orange) even though there's multiple 0.8 / 0.82 / 0.9V domains there. All IO power (except dedicated PHY rails) is magenta regardless of which bank(s) it's for.
Things of note:
…
Jamming GPS signals is a favorite tactic of Putin. Just saying.
At least some of the U.S. warships that have deployed to the Caribbean in recent months have been jamming GPS signals in their vicinity
https://www.nytimes.com/2025/12/20/world/a
STM32MP2 bringup progress: I have a "FSBL" (really my whole firmware) image booting from QSPI flash, doing some very low level init, then hanging before it starts the UART.
But it's able to open up debug access without needing ST's wrapper. I hard code a few pointers for SFRs that I don't have structs for but it's enough that I can get gdb attached and the code isn't TOO ugly.
Now I need to figure out why Timer::Sleep() is being derpy in BSP_InitUART…