libnvdimm, pmem: Do not flush power-fail protected CPU caches

This commit:

5fdf8e5 ("libnvdimm: re-enable deep flush for pmem devices via fsync()")

intended to make sure that deep flush was always available even on
platforms which support a power-fail protected CPU cache.  An unintended
side effect of this change was that we also lost the ability to skip
flushing CPU caches on those power-fail protected CPU cache.

Fix this by skipping the low level cache flushing in dax_flush() if we have
CPU caches which are power-fail protected.  The user can still override this
behavior by manually setting the write_cache state of a namespace.  See
libndctl's ndctl_namespace_write_cache_is_enabled(),
ndctl_namespace_enable_write_cache() and
ndctl_namespace_disable_write_cache() functions.

Cc: <stable@vger.kernel.org>
Fixes: 5fdf8e5 ("libnvdimm: re-enable deep flush for pmem devices via fsync()")
Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>

drivers/nvdimm/region_devs.c

@@ -1132,7 +1132,8 @@ EXPORT_SYMBOL_GPL(nvdimm_has_flush);
 
 int nvdimm_has_cache(struct nd_region *nd_region)
 {
-	return is_nd_pmem(&nd_region->dev);
+	return is_nd_pmem(&nd_region->dev) &&
+		!test_bit(ND_REGION_PERSIST_CACHE, &nd_region->flags);
 }
 EXPORT_SYMBOL_GPL(nvdimm_has_cache);