[dm-devel] [PATCH] vmalloc: introduce vmap_pfn for persistent memory
Christoph Hellwig
hch at infradead.org
Wed Nov 8 15:04:47 UTC 2017
On Wed, Nov 08, 2017 at 07:33:09AM -0500, Mikulas Patocka wrote:
> We could use the function clwb() (or arch-independent wrapper dax_flush())
> - that uses the clflushopt instruction on Broadwell or clwb on Skylake -
> but it is very slow, write performance on Broadwell is only 350MB/s.
>
> So in practice I use the movnti instruction that bypasses cache. The
> write-combining buffer is flushed with sfence.
And what do you do for an architecture with virtuall indexed caches?
More information about the dm-devel
mailing list