Linus Torvalds writes: (Summary) :)
The thing is, even with the delayed TLB flushing, I don't think it should be *so* delayed that we should be seeing a TLB fill from garbage page tables.
garbage page tables.
But the part in Andy's patch that worries me the most is that But the part in Andy's patch that worries me the most is that + cpumask_clear_cpu(cpu, mm_cpumask(mm)); I'm not entirely convinced by that tlb_gen logic.
convinced by that tlb_gen logic.
I can't actually see anything *wrong* in the tlb_gen logic, but it worries me.
The thing is, even with the delayed TLB flushing, I don't think it should be *so* delayed that we should be seeing a TLB fill from garbage page tables.
garbage page tables.
But the part in Andy's patch that worries me the most is that But the part in Andy's patch that worries me the most is that + cpumask_clear_cpu(cpu, mm_cpumask(mm)); I'm not entirely convinced by that tlb_gen logic.
convinced by that tlb_gen logic.
I can't actually see anything *wrong* in the tlb_gen logic, but it worries me.
