Re: What if a TLB flush needed to sleep?

[Christoph Lameter]

On Tue, 25 Mar 2008, Luck, Tony wrote:

> 2) Is it feasible to rearrange the MM code so that we don't
> hold any locks while doing a TLB flush?  Or should I implement
> some sort of spin_only_semaphore?

The EMM notifier V2 patchset contains two patches that 
convert the immap_lock and the anon_vma lock to semaphores. After that
much of the TLB flushing is (tlb_finish_mmu, tlb_gather etc) is running 
without holding any spinlocks. There would need to be additional measures 
for flushing inherent in macros (like ptep_clear_flush). Currently the 
pte functions are called under pte lock.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/