Linux CXL
 help / color / mirror / Atom feed
From: Ben Widawsky <ben.widawsky@intel.com>
To: linux-cxl@vger.kernel.org, nvdimm@lists.linux.dev,
	linux-pci@vger.kernel.org
Cc: patches@lists.linux.dev, Bjorn Helgaas <helgaas@kernel.org>,
	Ben Widawsky <ben.widawsky@intel.com>,
	Alison Schofield <alison.schofield@intel.com>,
	Dan Williams <dan.j.williams@intel.com>,
	Ira Weiny <ira.weiny@intel.com>,
	Jonathan Cameron <Jonathan.Cameron@Huawei.com>,
	Vishal Verma <vishal.l.verma@intel.com>
Subject: [PATCH v2 15/15] cxl/region: Create an nd_region
Date: Wed, 12 Jan 2022 15:47:49 -0800	[thread overview]
Message-ID: <20220112234749.1965960-16-ben.widawsky@intel.com> (raw)
In-Reply-To: <20220112234749.1965960-1-ben.widawsky@intel.com>

LIBNVDIMM supports the creation of regions for both persistent and
volatile memory ranges. The cxl_region driver is capable of handling the
CXL side of region creation but will reuse LIBVDIMM for interfacing with
the rest of the kernel.

TODO: CXL regions can go away. As a result the nd_region must also be
torn down.

TODO2: Handle mappings. LIBNVDIMM is capable of being informed about
which parts of devices contribute to a region and validating whether or
not the region is configured properly. To do this properly requires
tracking allocations per device.

Signed-off-by: Ben Widawsky <ben.widawsky@intel.com>
---
 drivers/cxl/core/pmem.c | 16 +++++++++++++
 drivers/cxl/cxl.h       |  1 +
 drivers/cxl/region.c    | 52 +++++++++++++++++++++++++++++++++++++++++
 3 files changed, 69 insertions(+)

diff --git a/drivers/cxl/core/pmem.c b/drivers/cxl/core/pmem.c
index bfcf51fbda5d..762a08c6f073 100644
--- a/drivers/cxl/core/pmem.c
+++ b/drivers/cxl/core/pmem.c
@@ -213,6 +213,22 @@ struct cxl_nvdimm *to_cxl_nvdimm(struct device *dev)
 }
 EXPORT_SYMBOL_NS_GPL(to_cxl_nvdimm, CXL);
 
+static int match_cxl_nvdimm(struct device *dev, void *data)
+{
+	return is_cxl_nvdimm(dev);
+}
+
+struct cxl_nvdimm *cxl_find_nvdimm(struct cxl_memdev *cxlmd)
+{
+	struct device *dev;
+
+	dev = device_find_child(&cxlmd->dev, NULL, match_cxl_nvdimm);
+	if (!dev)
+		return NULL;
+	return to_cxl_nvdimm(dev);
+}
+EXPORT_SYMBOL_NS_GPL(cxl_find_nvdimm, CXL);
+
 static struct cxl_nvdimm *cxl_nvdimm_alloc(struct cxl_memdev *cxlmd)
 {
 	struct cxl_nvdimm *cxl_nvd;
diff --git a/drivers/cxl/cxl.h b/drivers/cxl/cxl.h
index 6f9cabb77c08..a7b90356914d 100644
--- a/drivers/cxl/cxl.h
+++ b/drivers/cxl/cxl.h
@@ -434,6 +434,7 @@ bool is_cxl_nvdimm(struct device *dev);
 bool is_cxl_nvdimm_bridge(struct device *dev);
 int devm_cxl_add_nvdimm(struct device *host, struct cxl_memdev *cxlmd);
 struct cxl_nvdimm_bridge *cxl_find_nvdimm_bridge(struct cxl_memdev *cxlmd);
+struct cxl_nvdimm *cxl_find_nvdimm(struct cxl_memdev *cxlmd);
 
 /*
  * Unit test builds overrides this to __weak, find the 'strong' version
diff --git a/drivers/cxl/region.c b/drivers/cxl/region.c
index d00305655f5a..d4a7e8d47c11 100644
--- a/drivers/cxl/region.c
+++ b/drivers/cxl/region.c
@@ -623,6 +623,52 @@ static int bind_region(struct cxl_region *region)
 	return rc;
 }
 
+static int connect_to_libnvdimm(struct cxl_region *region)
+{
+	struct nd_region_desc ndr_desc;
+	struct cxl_nvdimm_bridge *nvb;
+	struct nd_region *ndr;
+	int rc = 0;
+
+	nvb = cxl_find_nvdimm_bridge(region->config.targets[0]);
+	device_lock(&nvb->dev);
+	if (!nvb->nvdimm_bus) {
+		rc = -ENXIO;
+		goto out;
+	}
+
+	memset(&ndr_desc, 0, sizeof(ndr_desc));
+
+	ndr_desc.res = region->res;
+
+	ndr_desc.numa_node = memory_add_physaddr_to_nid(region->res->start);
+	ndr_desc.target_node = phys_to_target_node(region->res->start);
+	if (ndr_desc.numa_node == NUMA_NO_NODE) {
+		ndr_desc.numa_node =
+			memory_add_physaddr_to_nid(region->res->start);
+		dev_info(&region->dev,
+			 "changing numa node from %d to %d for CXL region %pR",
+			 NUMA_NO_NODE, ndr_desc.numa_node, region->res);
+	}
+	if (ndr_desc.target_node == NUMA_NO_NODE) {
+		ndr_desc.target_node = ndr_desc.numa_node;
+		dev_info(&region->dev,
+			 "changing target node from %d to %d for CXL region %pR",
+			 NUMA_NO_NODE, ndr_desc.target_node, region->res);
+	}
+
+	ndr = nvdimm_pmem_region_create(nvb->nvdimm_bus, &ndr_desc);
+	if (IS_ERR(ndr))
+		rc = PTR_ERR(ndr);
+	else
+		dev_set_drvdata(&region->dev, ndr);
+
+out:
+	device_unlock(&nvb->dev);
+	put_device(&nvb->dev);
+	return rc;
+}
+
 static void region_unregister(void *dev)
 {
 	struct cxl_region *region = to_cxl_region(dev);
@@ -704,6 +750,12 @@ static int cxl_region_probe(struct device *dev)
 		return ret;
 	}
 
+	ret = connect_to_libnvdimm(region);
+	if (ret) {
+		region_unregister(dev);
+		return ret;
+	}
+
 
 	region->active = true;
 	dev_info(dev, "Bound");
-- 
2.34.1


  parent reply	other threads:[~2022-01-12 23:48 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-01-12 23:47 [PATCH v2 00/15] CXL Region driver Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 01/15] cxl/core: Rename find_cxl_port Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 02/15] cxl/core: Track port depth Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 03/15] cxl/region: Add region creation ABI Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 04/15] cxl/region: Introduce concept of region configuration Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 05/15] cxl/mem: Cache port created by the mem dev Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 06/15] cxl/region: Introduce a cxl_region driver Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 07/15] cxl/acpi: Handle address space allocation Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 08/15] cxl/region: Address " Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 09/15] cxl/region: Implement XHB verification Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 10/15] cxl/region: HB port config verification Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 11/15] cxl/region: Add infrastructure for decoder programming Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 12/15] cxl/region: Collect host bridge decoders Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 13/15] cxl: Program decoders for regions Ben Widawsky
2022-01-12 23:47 ` [PATCH v2 14/15] cxl/pmem: Convert nvdimm bridge API to use memdev Ben Widawsky
2022-01-12 23:47 ` Ben Widawsky [this message]
2022-01-15 19:00 ` [PATCH v2 00/15] CXL Region driver Ben Widawsky

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220112234749.1965960-16-ben.widawsky@intel.com \
    --to=ben.widawsky@intel.com \
    --cc=Jonathan.Cameron@Huawei.com \
    --cc=alison.schofield@intel.com \
    --cc=dan.j.williams@intel.com \
    --cc=helgaas@kernel.org \
    --cc=ira.weiny@intel.com \
    --cc=linux-cxl@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=nvdimm@lists.linux.dev \
    --cc=patches@lists.linux.dev \
    --cc=vishal.l.verma@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox