>but chud branch predictor, what if- >it won't
>the loop... has ended...
best thread on 4chan right now and it's not even close
i686 was a mistake
Billions must Die
>>108599744Billions must execute instruction #23784
you maybe have read about zen amp
>>108599109>turn out it doesTHE NORTH BRIDGE HAS FALLEN BILLIONS MUST BUY BULLDOZER
I will make you cache miss and refill your pipeline
>>108599109>context switchBillion cycles must die
The voltage has fallen, billions must shut down
>s[i] == '\0'
>>108599810I miss my caches so much, bros
>hey chud branch predictor, what is at this memory address I don't have access to?>billions must read
>the pipeline has fallen>billions must flush
>but chud branch predictor, you're accessing memory at an address that's a function of privileged data without bound checking!>If something happens, you'll load a specific byte to cache which the current thread can extract via a timing attack, inferring the contents of the privileged memory!
kek
>>108599109The Linux kernel code is full of theseif (unlikely(ptr)) return 0;
if (unlikely(ptr)) return 0;
>>108599109That's literally the ryzen platform.There's a mini neural network chud that decides it
>>108601534Good self-modifying code doesn’t need a fucking branch predictor.Do likewise, gents.
>>108599109nothing ever jumps forward
mitigations=auto nosmt=force memtest=4
>>108601550>self modifying code>*incessantly invalidates your instruction cache*what do?
>>108601622What? just run as it is on the cache, it's roughly the same shit so it's no problem!
mutts won, humans lost
>>108601730If you use self modifying code to skip branches, then I assume you mean removing branches from the machine code in memory, right? The problem is that this invalidates the instruction cache line for the branch, which will cause a far worse stall than a mispredicted branch.If you determine at runtime that your code always or never takes a branch, then it's probably marginally better to pay the stall penalty once, rather than use up space in the branch predictor.If your branch is sometimes taken, then it's basically never worth it to modify the instruction on that condition. The branch predictor will essentially do that for you in a much faster way.
>>108601897I wasnt making a that smart post, just a "stupid CPU execute old version of code" deal that fucks with everything.
>>108601622You backpatch at least 64 bytes behind if you’re worried about your precious cache line.> my code fits in a cache lineGood, but unroll the first iteration of it. The first iteration patches the next before the prefetch, and the subsequent iteration(s) patch the first or previous unrolled iteration.
>>108600155Just finished modding some extra L1 onto my rig, taking my girl out to celebrate. Now my shit is SO cache
Where is the optimistic predictor that does a IT'S HAPPENING every time
>>108604460>The Altman module
>>108606171kek
>>108599109>Branch predictor chad trying to create certainty in an uncertain future, succeeding 99% of the time
>>108599810The pipeline has stalled...
>>108608859>0xA: IRQL_not_less_or_equal
>>108599109Incel Chud Ultra 1488K
>>108602115Lol
good thread
>>108602115kek. this thread is gold thanks OP
>>108607644He just has great pattern recognition.
>>108602115Your girl is a pornstar
>>108607644Cache is getting warm bros
>>108599109i laugh every time i see this
>>108599109kek good thread
>>108599109Keyed.
open the port. now.
>>108599109op from original thread here. nice to see that this joke is still funny to some people
>>108599603on god we don't care about this boomer type of humor unc
BILLIONS of instructions must execute.Per second. Total Instruction Execution
>>108616251go back to moping about the zoomettes drunk on unc cock, zoomer
Megabytes?
>>1085991090x90 ever happens
