Just remembered this one also exists. It's getting hard to keep track :P
I don't know if it's worth including, since I suppose it's not really a spectre/meltdown based vulnerability, but I figured it's something to consider. (maybe give a warning when 'paranoid' is specified?)
I doesn't seem to use use speculation, but does leak stuff from the L1 cache through side channel analysis when different threads share the same core. Because of the complexity of the problem, openBSD's solution was to disable hyperthreading. Some of the similarities with L1TF make me wonder if openBSD was actually working on mitigating both at the time. (see my comment on #229 from today)
Detection seems simple: are you an Intel processor with hyperthreading?
With the mitigation being: turn off hyperthreading
Though it may be possible to refine these criteria a bit.
Just remembered this one also exists. It's getting hard to keep track :P
I don't know if it's worth including, since I suppose it's not really a spectre/meltdown based vulnerability, but I figured it's something to consider. (maybe give a warning when 'paranoid' is specified?)
I doesn't seem to use use speculation, but does leak stuff from the L1 cache through side channel analysis when different threads share the same core. Because of the complexity of the problem, openBSD's solution was to disable hyperthreading. Some of the similarities with L1TF make me wonder if openBSD was actually working on mitigating both at the time. (see my comment on #229 from today)
Detection seems simple: are you an Intel processor with hyperthreading? With the mitigation being: turn off hyperthreading Though it may be possible to refine these criteria a bit.