libnvdimm/pmem: Honor force_raw for legacy pmem regions

For recovery, where non-dax access is needed to a given physical address
range, and testing, allow the 'force_raw' attribute to override the
default establishment of a dev_pagemap.

Otherwise without this capability it is possible to end up with a
namespace that can not be activated due to corrupted info-block, and one
that can not be repaired due to a section collision.

Cc: <stable@vger.kernel.org>
Fixes: 004f1afbe1 ("libnvdimm, pmem: direct map legacy pmem by default")
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
This commit is contained in:
Dan Williams 2019-01-24 17:33:06 -08:00
parent 11a358109e
commit fa7d2e639c
1 changed files with 4 additions and 0 deletions

View File

@ -138,6 +138,7 @@ bool nd_is_uuid_unique(struct device *dev, u8 *uuid)
bool pmem_should_map_pages(struct device *dev)
{
struct nd_region *nd_region = to_nd_region(dev->parent);
struct nd_namespace_common *ndns = to_ndns(dev);
struct nd_namespace_io *nsio;
if (!IS_ENABLED(CONFIG_ZONE_DEVICE))
@ -149,6 +150,9 @@ bool pmem_should_map_pages(struct device *dev)
if (is_nd_pfn(dev) || is_nd_btt(dev))
return false;
if (ndns->force_raw)
return false;
nsio = to_nd_namespace_io(dev);
if (region_intersects(nsio->res.start, resource_size(&nsio->res),
IORESOURCE_SYSTEM_RAM,