On 3/18/26 09:15, Eric Biggers wrote: > Doesn't DOITM still need to be enabled by default for now, given that > data-dependent timing is a backwards compatibility break and code hasn't > been updated to account for it yet?
Please read: https://lore.kernel.org/all/[email protected]/ Especially: > The execution latency of the DOIT instructions[1] does not depend on the > value of data operands on all currently-supported Intel processors. I believe that's a long winded way of saying there's no "backwards compatibility break".
