libnvdimm, pfn: Pad pfn namespaces relative to other regions

Commit cfe30b872058 "libnvdimm, pmem: adjust for section collisions with
'System RAM'" enabled Linux to workaround occasions where platform
firmware arranges for "System RAM" and "Persistent Memory" to collide
within a single section boundary. Unfortunately, as reported in this
issue [1], platform firmware can inflict the same collision between
persistent memory regions.

The approach of interrogating iomem_resource does not work in this
case because platform firmware may merge multiple regions into a single
iomem_resource range. Instead provide a method to interrogate regions
that share the same parent bus.

This is a stop-gap until the core-MM can grow support for hotplug on
sub-section boundaries.

[1]: https://github.com/pmem/ndctl/issues/76

Fixes: cfe30b872058 ("libnvdimm, pmem: adjust for section collisions with...")
Cc: <stable@vger.kernel.org>
Reported-by: Patrick Geary <patrickg@supermicro.com>
Tested-by: Patrick Geary <patrickg@supermicro.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>

3 files changed, 80 insertions, 27 deletions

diff --git a/drivers/nvdimm/nd-core.h b/drivers/nvdimm/nd-core.h
index 182258f64417..d0c621b32f72 100644
--- a/drivers/nvdimm/nd-core.h
+++ b/drivers/nvdimm/nd-core.h
@@ -111,6 +111,8 @@ resource_size_t nd_pmem_available_dpa(struct nd_region *nd_region,
 		struct nd_mapping *nd_mapping, resource_size_t *overlap);
 resource_size_t nd_blk_available_dpa(struct nd_region *nd_region);
 resource_size_t nd_region_available_dpa(struct nd_region *nd_region);
+int nd_region_conflict(struct nd_region *nd_region, resource_size_t start,
+		resource_size_t size);
 resource_size_t nvdimm_allocated_dpa(struct nvdimm_drvdata *ndd,
 		struct nd_label_id *label_id);
 int alias_dpa_busy(struct device *dev, void *data);
diff --git a/drivers/nvdimm/pfn_devs.c b/drivers/nvdimm/pfn_devs.c
index 24c64090169e..6f22272e8d80 100644
--- a/drivers/nvdimm/pfn_devs.c
+++ b/drivers/nvdimm/pfn_devs.c
@@ -649,14 +649,47 @@ static u64 phys_pmem_align_down(struct nd_pfn *nd_pfn, u64 phys)
 			ALIGN_DOWN(phys, nd_pfn->align));
 }
 
+/*
+ * Check if pmem collides with 'System RAM', or other regions when
+ * section aligned.  Trim it accordingly.
+ */
+static void trim_pfn_device(struct nd_pfn *nd_pfn, u32 *start_pad, u32 *end_trunc)
+{
+	struct nd_namespace_common *ndns = nd_pfn->ndns;
+	struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
+	struct nd_region *nd_region = to_nd_region(nd_pfn->dev.parent);
+	const resource_size_t start = nsio->res.start;
+	const resource_size_t end = start + resource_size(&nsio->res);
+	resource_size_t adjust, size;
+
+	*start_pad = 0;
+	*end_trunc = 0;
+
+	adjust = start - PHYS_SECTION_ALIGN_DOWN(start);
+	size = resource_size(&nsio->res) + adjust;
+	if (region_intersects(start - adjust, size, IORESOURCE_SYSTEM_RAM,
+				IORES_DESC_NONE) == REGION_MIXED
+			|| nd_region_conflict(nd_region, start - adjust, size))
+		*start_pad = PHYS_SECTION_ALIGN_UP(start) - start;
+
+	/* Now check that end of the range does not collide. */
+	adjust = PHYS_SECTION_ALIGN_UP(end) - end;
+	size = resource_size(&nsio->res) + adjust;
+	if (region_intersects(start, size, IORESOURCE_SYSTEM_RAM,
+				IORES_DESC_NONE) == REGION_MIXED
+			|| !IS_ALIGNED(end, nd_pfn->align)
+			|| nd_region_conflict(nd_region, start, size + adjust))
+		*end_trunc = end - phys_pmem_align_down(nd_pfn, end);
+}
+
 static int nd_pfn_init(struct nd_pfn *nd_pfn)
 {
 	u32 dax_label_reserve = is_nd_dax(&nd_pfn->dev) ? SZ_128K : 0;
 	struct nd_namespace_common *ndns = nd_pfn->ndns;
-	u32 start_pad = 0, end_trunc = 0;
+	struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
 	resource_size_t start, size;
-	struct nd_namespace_io *nsio;
 	struct nd_region *nd_region;
+	u32 start_pad, end_trunc;
 	struct nd_pfn_sb *pfn_sb;
 	unsigned long npfns;
 	phys_addr_t offset;
@@ -688,30 +721,7 @@ static int nd_pfn_init(struct nd_pfn *nd_pfn)
 
 	memset(pfn_sb, 0, sizeof(*pfn_sb));
 
-	/*
-	 * Check if pmem collides with 'System RAM' when section aligned and
-	 * trim it accordingly
-	 */
-	nsio = to_nd_namespace_io(&ndns->dev);
-	start = PHYS_SECTION_ALIGN_DOWN(nsio->res.start);
-	size = resource_size(&nsio->res);
-	if (region_intersects(start, size, IORESOURCE_SYSTEM_RAM,
-				IORES_DESC_NONE) == REGION_MIXED) {
-		start = nsio->res.start;
-		start_pad = PHYS_SECTION_ALIGN_UP(start) - start;
-	}
-
-	start = nsio->res.start;
-	size = PHYS_SECTION_ALIGN_UP(start + size) - start;
-	if (region_intersects(start, size, IORESOURCE_SYSTEM_RAM,
-				IORES_DESC_NONE) == REGION_MIXED
-			|| !IS_ALIGNED(start + resource_size(&nsio->res),
-				nd_pfn->align)) {
-		size = resource_size(&nsio->res);
-		end_trunc = start + size - phys_pmem_align_down(nd_pfn,
-				start + size);
-	}
-
+	trim_pfn_device(nd_pfn, &start_pad, &end_trunc);
 	if (start_pad + end_trunc)
 		dev_info(&nd_pfn->dev, "%s alignment collision, truncate %d bytes\n",
 				dev_name(&ndns->dev), start_pad + end_trunc);
@@ -722,7 +732,7 @@ static int nd_pfn_init(struct nd_pfn *nd_pfn)
 	 * implementation will limit the pfns advertised through
 	 * ->direct_access() to those that are included in the memmap.
 	 */
-	start += start_pad;
+	start = nsio->res.start + start_pad;
 	size = resource_size(&nsio->res);
 	npfns = PFN_SECTION_ALIGN_UP((size - start_pad - end_trunc - SZ_8K)
 			/ PAGE_SIZE);
diff --git a/drivers/nvdimm/region_devs.c b/drivers/nvdimm/region_devs.c
index 174a418cb171..e7377f1028ef 100644
--- a/drivers/nvdimm/region_devs.c
+++ b/drivers/nvdimm/region_devs.c
@@ -1184,6 +1184,47 @@ int nvdimm_has_cache(struct nd_region *nd_region)
 }
 EXPORT_SYMBOL_GPL(nvdimm_has_cache);
 
+struct conflict_context {
+	struct nd_region *nd_region;
+	resource_size_t start, size;
+};
+
+static int region_conflict(struct device *dev, void *data)
+{
+	struct nd_region *nd_region;
+	struct conflict_context *ctx = data;
+	resource_size_t res_end, region_end, region_start;
+
+	if (!is_memory(dev))
+		return 0;
+
+	nd_region = to_nd_region(dev);
+	if (nd_region == ctx->nd_region)
+		return 0;
+
+	res_end = ctx->start + ctx->size;
+	region_start = nd_region->ndr_start;
+	region_end = region_start + nd_region->ndr_size;
+	if (ctx->start >= region_start && ctx->start < region_end)
+		return -EBUSY;
+	if (res_end > region_start && res_end <= region_end)
+		return -EBUSY;
+	return 0;
+}
+
+int nd_region_conflict(struct nd_region *nd_region, resource_size_t start,
+		resource_size_t size)
+{
+	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(&nd_region->dev);
+	struct conflict_context ctx = {
+		.nd_region = nd_region,
+		.start = start,
+		.size = size,
+	};
+
+	return device_for_each_child(&nvdimm_bus->dev, &ctx, region_conflict);
+}
+
 void __exit nd_region_devs_exit(void)
 {
 	ida_destroy(&region_ida);