Rough order of magnitude for errors on my (Celeron) box, my test: Dekker 10k ppm, Peterson 10 ppm.
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
@RichFelker That barriers are even needed is awful; my feeble mind just cannot visualize all the things that can go wrong w/ reordering. -
@RichFelker "reads can be speculative and return values from the future". Does this mean: mov eax, [ebx] mov [ebx], 0x01 eax could be 1?! -
@cr1901 No. On a single core there is a total order of all loads/stores. Also (from link) "Stores are not reordered with older loads".
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.