<firefrommoonligh>
That'd be nice. In the meanwhile, I'd manually keep track of which peripherals are dependent on clock speed (most?), and add a fn to your program that changes it for them via PAC. (Probably need unsafe raw pointers or `steal`.)
<adamgreig-m>
changelog for 1.11 is solely adding `<cpu>` if it doesn't exist, lol
<thalesfragoso[m]>
thanks, that should solve it :)
<adamgreig-m>
woohoo, i just placed the last airwire out of 500, finally i can sleep
<adamgreig-m>
350/500 in the first hour, but the final 2 in the last hour, typical
f_punk[m] has joined ##stm32-rs
<thalesfragoso[m]>
richardeoin on the sdmmc code (H7), when reading data we are checking the rxfifo half full flag to proceed to read from the fifo, shouldn't we be checking the rxfifo empty flag instead ?
<thalesfragoso[m]>
It seems like it will stall in the last 8 words on the 512 bytes block, probably someone didn't notice it yet because they didn't tried to read that much at a time ?
<thalesfragoso[m]>
* It seems like it will stall in the last 7 words on the 512 bytes block, probably someone didn't notice it yet because they didn't tried to read that much at a time ?
<thalesfragoso[m]>
Also, we are bailing out on the reading I'd `dataend` bit is set, but there might be data left on the fifo after that, no ?
<thalesfragoso[m]>
* Also, we are bailing out on the reading if `dataend` bit is set, but there might be data left on the fifo after that, no ?