@lauriewired I do not understand, how exactly is it going to help to force 100% of hardware vendors to rush out RVA23 support or get left behind by Ubuntu? Aren't they going to just make *more* out of spec and seriously vulnerable processors with these new features?
URGENT, PLEASE RT:
on 14 June 2024, myself, @quantumcatgirl, and 49 others jointly signed a Plea for Prosecutorial Compassion for Singaporean trans woman Harvey. she now faces 9 strokes of the cane for resisting strip-searches by male prison officers. 1/
https://t.co/hXEPpR4ep9
i have once again fallen to my vices: trying to explain some small technical thing and having it spiral way out of control until i'm conflating release engineering with Doing A Gender and decide to announce another transition via my work's developer comms
https://t.co/FSB4xAW0rz
@whitequark this is probably not that cursed, but i think the pinmux names on one of the chips i've touched implies that this is the topology, and idk if i want to know for sure anymore
@__phantomderp ... landing pads. Staying in the realm of C, pthread cancellation (maybe even pthread_exit) will unwind. Is this intentionally excluded?
@kepe__@DrawsMiguel@StatsLime@mycoliza this simplifies hardware, and you can do the check in software anyway. if /0 is undefined behavior, you do nothing. if it is supposed to error, if (divisor == 0) error() is easier anyway, and an unlikely branch is practically free
@kepe__@DrawsMiguel@StatsLime@mycoliza riscv has no data dependent exceptions. [insert elaboration here] but basically they've tried very hard to never have exceptions based on an operand's value
if it comforts you, think "the result is unspecified, but for consistency's sake we'll specify one"
@mitsuhiko "safety" you say...
https://t.co/zoE4TZod6c
but yeah there's a reason when talking about safety it's about things a program "cannot do". there are always things you can do to make a program segfault (okay, bus error, tomato/tomato)
try this:
print("Please wiggle the RAM stick")