<samueldr>
there was one thing that was clear from the whole "speculative execution is flawed" thing: no one was going to be spared
<gchristensen>
yup
<samueldr>
and in a sick twisted way... intel might be the safer bet since their market share means they're the juiciest target
<samueldr>
they're getting a ton of security audits for free!*
<Ashy>
it seems to be the same old story overall, the safest computer is one that's switched off and locked in a safe, preferably without any data on it
<samueldr>
make sure there's no batteries in there
<samueldr>
:/ some details are lacking...
<samueldr>
>> Nailgun attack requires that the debug authentication signals are enabled
<samueldr>
as someone that is not intimately involved in ARM things, I do not even know what this means for the end-user
<samueldr>
is this something that every CPU has that must be turned of via software? is it microcode/fuse backed and left on accidentally?
<samueldr>
is this something that can be checked?
<samueldr>
(looks like this is not spec ex related at all at a first glance)
orivej has quit [Ping timeout: 248 seconds]
ryantrinkle has quit [Ping timeout: 244 seconds]
ryantrinkle has joined #nixos-aarch64
ryantrinkle has quit [Ping timeout: 268 seconds]
orivej has joined #nixos-aarch64
jackdk has quit [Ping timeout: 268 seconds]
FRidh has joined #nixos-aarch64
<andi->
This is the issues class where you have such great facilities within the CPU to measure execution latency so you can retrieve data from other parts of the system. At least that is how I understood this and a few of the recent Intel bugs.
* DigitalKiwi
wonders nailgun, row hammer...when is coffin nail going to come along and finish it off?