All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dan Williams <dan.j.williams@intel.com>
To: linux-nvdimm@lists.01.org
Cc: david@fromorbit.com, linux-kernel@vger.kernel.org, hch@lst.de
Subject: [PATCH 03/13] libnvdimm: introduce nvdimm_flush()
Date: Sat, 04 Jun 2016 13:52:48 -0700	[thread overview]
Message-ID: <146507356876.8347.4735880992775196026.stgit@dwillia2-desk3.amr.corp.intel.com> (raw)
In-Reply-To: <146507355220.8347.12117020810872172684.stgit@dwillia2-desk3.amr.corp.intel.com>

nvdimm_flush() is an alternative to the x86 pcommit instruction.  It is
an optional write flushing mechanism that an nvdimm bus can provide for
the pmem driver to consume.  In the case of the NFIT nvdimm-bus-provider
nvdimm_flush() is implemented as a series of flush-hint-address [1]
writes to each dimm in the interleave set that backs the namespace.  For
now this implementation is just a simple replacement of wmb_pmem() /
arch_has_wmb_pmem() with nvdimm_flush() / nvdimm_has_flush().

We defer the full implementation of nvdimm_flush() until the
implementation is prepared to also handle the blk-region case.

Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 drivers/nvdimm/pmem.c        |   27 +++++++++++++++++++++------
 drivers/nvdimm/region_devs.c |   35 +++++++++++++++++++++++++++++++++++
 include/linux/libnvdimm.h    |    2 ++
 3 files changed, 58 insertions(+), 6 deletions(-)

diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
index 608fc4464574..248a89736bbf 100644
--- a/drivers/nvdimm/pmem.c
+++ b/drivers/nvdimm/pmem.c
@@ -46,10 +46,24 @@ struct pmem_device {
 	struct badblocks	bb;
 };
 
+static struct device *to_dev(struct pmem_device *pmem)
+{
+	/*
+	 * nvdimm bus services need a 'dev' parameter, and we record the device
+	 * at init in bb.dev.
+	 */
+	return pmem->bb.dev;
+}
+
+static struct nd_region *to_region(struct pmem_device *pmem)
+{
+	return to_nd_region(to_dev(pmem)->parent);
+}
+
 static void pmem_clear_poison(struct pmem_device *pmem, phys_addr_t offset,
 		unsigned int len)
 {
-	struct device *dev = pmem->bb.dev;
+	struct device *dev = to_dev(pmem);
 	sector_t sector;
 	long cleared;
 
@@ -135,7 +149,7 @@ static blk_qc_t pmem_make_request(struct request_queue *q, struct bio *bio)
 		nd_iostat_end(bio, start);
 
 	if (bio_data_dir(bio))
-		wmb_pmem();
+		nvdimm_flush(to_region(pmem));
 
 	bio_endio(bio);
 	return BLK_QC_T_NONE;
@@ -149,7 +163,7 @@ static int pmem_rw_page(struct block_device *bdev, sector_t sector,
 
 	rc = pmem_do_bvec(pmem, page, PAGE_SIZE, 0, rw, sector);
 	if (rw & WRITE)
-		wmb_pmem();
+		nvdimm_flush(to_region(pmem));
 
 	/*
 	 * The ->rw_page interface is subtle and tricky.  The core
@@ -205,6 +219,7 @@ static int pmem_attach_disk(struct device *dev,
 		struct nd_namespace_common *ndns)
 {
 	struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
+	struct nd_region *nd_region = to_nd_region(dev->parent);
 	struct vmem_altmap __altmap, *altmap = NULL;
 	struct resource *res = &nsio->res;
 	struct nd_pfn *nd_pfn = NULL;
@@ -234,7 +249,7 @@ static int pmem_attach_disk(struct device *dev,
 	dev_set_drvdata(dev, pmem);
 	pmem->phys_addr = res->start;
 	pmem->size = resource_size(res);
-	if (!arch_has_wmb_pmem())
+	if (nvdimm_has_flush(nd_region) < 0)
 		dev_warn(dev, "unable to guarantee persistence of writes\n");
 
 	if (!devm_request_mem_region(dev, res->start, resource_size(res),
@@ -302,7 +317,7 @@ static int pmem_attach_disk(struct device *dev,
 			/ 512);
 	if (devm_init_badblocks(dev, &pmem->bb))
 		return -ENOMEM;
-	nvdimm_badblocks_populate(to_nd_region(dev->parent), &pmem->bb, res);
+	nvdimm_badblocks_populate(nd_region, &pmem->bb, res);
 	disk->bb = &pmem->bb;
 	add_disk(disk);
 	revalidate_disk(disk);
@@ -345,8 +360,8 @@ static int nd_pmem_remove(struct device *dev)
 
 static void nd_pmem_notify(struct device *dev, enum nvdimm_event event)
 {
-	struct nd_region *nd_region = to_nd_region(dev->parent);
 	struct pmem_device *pmem = dev_get_drvdata(dev);
+	struct nd_region *nd_region = to_region(pmem);
 	resource_size_t offset = 0, end_trunc = 0;
 	struct nd_namespace_common *ndns;
 	struct nd_namespace_io *nsio;
diff --git a/drivers/nvdimm/region_devs.c b/drivers/nvdimm/region_devs.c
index 320f0f3ea367..420e1a5e2250 100644
--- a/drivers/nvdimm/region_devs.c
+++ b/drivers/nvdimm/region_devs.c
@@ -14,6 +14,7 @@
 #include <linux/highmem.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
+#include <linux/pmem.h>
 #include <linux/sort.h>
 #include <linux/io.h>
 #include <linux/nd.h>
@@ -796,6 +797,40 @@ struct nd_region *nvdimm_volatile_region_create(struct nvdimm_bus *nvdimm_bus,
 }
 EXPORT_SYMBOL_GPL(nvdimm_volatile_region_create);
 
+/**
+ * nvdimm_flush - flush any posted write queues between the cpu and pmem media
+ * @nd_region: blk or interleaved pmem region
+ */
+void nvdimm_flush(struct nd_region *nd_region)
+{
+	/*
+	 * TODO: replace wmb_pmem() usage with flush hint writes where
+	 * available.
+	 */
+	wmb_pmem();
+}
+EXPORT_SYMBOL_GPL(nvdimm_flush);
+
+/**
+ * nvdimm_has_flush - determine write flushing requirements
+ * @nd_region: blk or interleaved pmem region
+ *
+ * Returns 1 if writes require flushing
+ * Returns 0 if writes do not require flushing
+ * Returns -ENXIO if flushing capability can not be determined
+ */
+int nvdimm_has_flush(struct nd_region *nd_region)
+{
+	/*
+	 * TODO: return 0 / 1 for NFIT regions depending on presence of
+	 * flush hint tables
+	 */
+	if (arch_has_wmb_pmem())
+		return 1;
+	return -ENXIO;
+}
+EXPORT_SYMBOL_GPL(nvdimm_has_flush);
+
 void __exit nd_region_devs_exit(void)
 {
 	ida_destroy(&region_ida);
diff --git a/include/linux/libnvdimm.h b/include/linux/libnvdimm.h
index 0c3c30cbbea5..90eb3119c3ce 100644
--- a/include/linux/libnvdimm.h
+++ b/include/linux/libnvdimm.h
@@ -156,4 +156,6 @@ struct nvdimm *nd_blk_region_to_dimm(struct nd_blk_region *ndbr);
 unsigned int nd_region_acquire_lane(struct nd_region *nd_region);
 void nd_region_release_lane(struct nd_region *nd_region, unsigned int lane);
 u64 nd_fletcher64(void *addr, size_t len, bool le);
+void nvdimm_flush(struct nd_region *nd_region);
+int nvdimm_has_flush(struct nd_region *nd_region);
 #endif /* __LIBNVDIMM_H__ */

_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm

WARNING: multiple messages have this Message-ID (diff)
From: Dan Williams <dan.j.williams@intel.com>
To: linux-nvdimm@ml01.01.org
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>,
	david@fromorbit.com, linux-kernel@vger.kernel.org, hch@lst.de
Subject: [PATCH 03/13] libnvdimm: introduce nvdimm_flush()
Date: Sat, 04 Jun 2016 13:52:48 -0700	[thread overview]
Message-ID: <146507356876.8347.4735880992775196026.stgit@dwillia2-desk3.amr.corp.intel.com> (raw)
In-Reply-To: <146507355220.8347.12117020810872172684.stgit@dwillia2-desk3.amr.corp.intel.com>

nvdimm_flush() is an alternative to the x86 pcommit instruction.  It is
an optional write flushing mechanism that an nvdimm bus can provide for
the pmem driver to consume.  In the case of the NFIT nvdimm-bus-provider
nvdimm_flush() is implemented as a series of flush-hint-address [1]
writes to each dimm in the interleave set that backs the namespace.  For
now this implementation is just a simple replacement of wmb_pmem() /
arch_has_wmb_pmem() with nvdimm_flush() / nvdimm_has_flush().

We defer the full implementation of nvdimm_flush() until the
implementation is prepared to also handle the blk-region case.

Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 drivers/nvdimm/pmem.c        |   27 +++++++++++++++++++++------
 drivers/nvdimm/region_devs.c |   35 +++++++++++++++++++++++++++++++++++
 include/linux/libnvdimm.h    |    2 ++
 3 files changed, 58 insertions(+), 6 deletions(-)

diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
index 608fc4464574..248a89736bbf 100644
--- a/drivers/nvdimm/pmem.c
+++ b/drivers/nvdimm/pmem.c
@@ -46,10 +46,24 @@ struct pmem_device {
 	struct badblocks	bb;
 };
 
+static struct device *to_dev(struct pmem_device *pmem)
+{
+	/*
+	 * nvdimm bus services need a 'dev' parameter, and we record the device
+	 * at init in bb.dev.
+	 */
+	return pmem->bb.dev;
+}
+
+static struct nd_region *to_region(struct pmem_device *pmem)
+{
+	return to_nd_region(to_dev(pmem)->parent);
+}
+
 static void pmem_clear_poison(struct pmem_device *pmem, phys_addr_t offset,
 		unsigned int len)
 {
-	struct device *dev = pmem->bb.dev;
+	struct device *dev = to_dev(pmem);
 	sector_t sector;
 	long cleared;
 
@@ -135,7 +149,7 @@ static blk_qc_t pmem_make_request(struct request_queue *q, struct bio *bio)
 		nd_iostat_end(bio, start);
 
 	if (bio_data_dir(bio))
-		wmb_pmem();
+		nvdimm_flush(to_region(pmem));
 
 	bio_endio(bio);
 	return BLK_QC_T_NONE;
@@ -149,7 +163,7 @@ static int pmem_rw_page(struct block_device *bdev, sector_t sector,
 
 	rc = pmem_do_bvec(pmem, page, PAGE_SIZE, 0, rw, sector);
 	if (rw & WRITE)
-		wmb_pmem();
+		nvdimm_flush(to_region(pmem));
 
 	/*
 	 * The ->rw_page interface is subtle and tricky.  The core
@@ -205,6 +219,7 @@ static int pmem_attach_disk(struct device *dev,
 		struct nd_namespace_common *ndns)
 {
 	struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
+	struct nd_region *nd_region = to_nd_region(dev->parent);
 	struct vmem_altmap __altmap, *altmap = NULL;
 	struct resource *res = &nsio->res;
 	struct nd_pfn *nd_pfn = NULL;
@@ -234,7 +249,7 @@ static int pmem_attach_disk(struct device *dev,
 	dev_set_drvdata(dev, pmem);
 	pmem->phys_addr = res->start;
 	pmem->size = resource_size(res);
-	if (!arch_has_wmb_pmem())
+	if (nvdimm_has_flush(nd_region) < 0)
 		dev_warn(dev, "unable to guarantee persistence of writes\n");
 
 	if (!devm_request_mem_region(dev, res->start, resource_size(res),
@@ -302,7 +317,7 @@ static int pmem_attach_disk(struct device *dev,
 			/ 512);
 	if (devm_init_badblocks(dev, &pmem->bb))
 		return -ENOMEM;
-	nvdimm_badblocks_populate(to_nd_region(dev->parent), &pmem->bb, res);
+	nvdimm_badblocks_populate(nd_region, &pmem->bb, res);
 	disk->bb = &pmem->bb;
 	add_disk(disk);
 	revalidate_disk(disk);
@@ -345,8 +360,8 @@ static int nd_pmem_remove(struct device *dev)
 
 static void nd_pmem_notify(struct device *dev, enum nvdimm_event event)
 {
-	struct nd_region *nd_region = to_nd_region(dev->parent);
 	struct pmem_device *pmem = dev_get_drvdata(dev);
+	struct nd_region *nd_region = to_region(pmem);
 	resource_size_t offset = 0, end_trunc = 0;
 	struct nd_namespace_common *ndns;
 	struct nd_namespace_io *nsio;
diff --git a/drivers/nvdimm/region_devs.c b/drivers/nvdimm/region_devs.c
index 320f0f3ea367..420e1a5e2250 100644
--- a/drivers/nvdimm/region_devs.c
+++ b/drivers/nvdimm/region_devs.c
@@ -14,6 +14,7 @@
 #include <linux/highmem.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
+#include <linux/pmem.h>
 #include <linux/sort.h>
 #include <linux/io.h>
 #include <linux/nd.h>
@@ -796,6 +797,40 @@ struct nd_region *nvdimm_volatile_region_create(struct nvdimm_bus *nvdimm_bus,
 }
 EXPORT_SYMBOL_GPL(nvdimm_volatile_region_create);
 
+/**
+ * nvdimm_flush - flush any posted write queues between the cpu and pmem media
+ * @nd_region: blk or interleaved pmem region
+ */
+void nvdimm_flush(struct nd_region *nd_region)
+{
+	/*
+	 * TODO: replace wmb_pmem() usage with flush hint writes where
+	 * available.
+	 */
+	wmb_pmem();
+}
+EXPORT_SYMBOL_GPL(nvdimm_flush);
+
+/**
+ * nvdimm_has_flush - determine write flushing requirements
+ * @nd_region: blk or interleaved pmem region
+ *
+ * Returns 1 if writes require flushing
+ * Returns 0 if writes do not require flushing
+ * Returns -ENXIO if flushing capability can not be determined
+ */
+int nvdimm_has_flush(struct nd_region *nd_region)
+{
+	/*
+	 * TODO: return 0 / 1 for NFIT regions depending on presence of
+	 * flush hint tables
+	 */
+	if (arch_has_wmb_pmem())
+		return 1;
+	return -ENXIO;
+}
+EXPORT_SYMBOL_GPL(nvdimm_has_flush);
+
 void __exit nd_region_devs_exit(void)
 {
 	ida_destroy(&region_ida);
diff --git a/include/linux/libnvdimm.h b/include/linux/libnvdimm.h
index 0c3c30cbbea5..90eb3119c3ce 100644
--- a/include/linux/libnvdimm.h
+++ b/include/linux/libnvdimm.h
@@ -156,4 +156,6 @@ struct nvdimm *nd_blk_region_to_dimm(struct nd_blk_region *ndbr);
 unsigned int nd_region_acquire_lane(struct nd_region *nd_region);
 void nd_region_release_lane(struct nd_region *nd_region, unsigned int lane);
 u64 nd_fletcher64(void *addr, size_t len, bool le);
+void nvdimm_flush(struct nd_region *nd_region);
+int nvdimm_has_flush(struct nd_region *nd_region);
 #endif /* __LIBNVDIMM_H__ */

  parent reply	other threads:[~2016-06-04 20:53 UTC|newest]

Thread overview: 73+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-06-04 20:52 [PATCH 00/13] deprecate pcommit Dan Williams
2016-06-04 20:52 ` Dan Williams
2016-06-04 20:52 ` [PATCH 01/13] driver core, libnvdimm: disable manual unbind of dimms while region active Dan Williams
2016-06-04 20:52   ` Dan Williams
2016-06-04 21:10   ` Greg Kroah-Hartman
2016-06-04 21:10     ` Greg Kroah-Hartman
2016-06-04 21:39     ` Dan Williams
2016-06-04 21:39       ` Dan Williams
2016-06-04 21:45       ` Greg Kroah-Hartman
2016-06-04 21:45         ` Greg Kroah-Hartman
2016-06-04 21:48         ` Dan Williams
2016-06-04 21:48           ` Dan Williams
2016-06-04 21:50   ` kbuild test robot
2016-06-04 21:50     ` kbuild test robot
2016-06-06 19:25   ` Linda Knippers
2016-06-06 19:25     ` Linda Knippers
2016-06-06 19:31     ` Dan Williams
2016-06-06 19:31       ` Dan Williams
2016-06-06 19:36       ` Dan Williams
2016-06-06 19:36         ` Dan Williams
2016-06-06 19:36       ` Linda Knippers
2016-06-06 19:36         ` Linda Knippers
2016-06-06 19:46         ` Dan Williams
2016-06-06 19:46           ` Dan Williams
2016-06-06 20:20           ` Linda Knippers
2016-06-06 20:20             ` Linda Knippers
2016-06-06 20:36             ` Dan Williams
2016-06-06 20:36               ` Dan Williams
2016-06-06 21:15               ` Linda Knippers
2016-06-06 21:15                 ` Linda Knippers
2016-06-04 20:52 ` [PATCH 02/13] nfit: always associate flush hints Dan Williams
2016-06-04 20:52   ` Dan Williams
2016-06-04 20:52 ` Dan Williams [this message]
2016-06-04 20:52   ` [PATCH 03/13] libnvdimm: introduce nvdimm_flush() Dan Williams
2016-06-06 17:45   ` Jeff Moyer
2016-06-04 20:52 ` [PATCH 04/13] libnvdimm, nfit: move flush hint mapping to dimm driver Dan Williams
2016-06-04 20:52   ` Dan Williams
2016-06-04 21:29   ` kbuild test robot
2016-06-04 21:29     ` kbuild test robot
2016-06-04 21:40   ` kbuild test robot
2016-06-04 21:40     ` kbuild test robot
2016-06-04 21:49   ` kbuild test robot
2016-06-04 21:49     ` kbuild test robot
2016-06-07 18:11   ` Kani, Toshimitsu
2016-06-07 18:11     ` Kani, Toshimitsu
2016-06-07 18:15     ` Dan Williams
2016-06-07 18:15       ` Dan Williams
2016-06-04 20:52 ` [PATCH 05/13] tools/testing/nvdimm: simulate multiple flush hints per-dimm Dan Williams
2016-06-04 20:52   ` Dan Williams
2016-06-04 20:53 ` [PATCH 06/13] libnvdimm: cycle flush hints per-cpu Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-04 20:53 ` [PATCH 07/13] libnvdimm, pmem: use REQ_FUA, REQ_FLUSH for nvdimm_flush() Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-04 20:53 ` [PATCH 08/13] fs/dax: remove wmb_pmem() Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-04 20:53 ` [PATCH 09/13] libnvdimm, pmem: use nvdimm_flush() for namespace I/O writes Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-04 20:53 ` [PATCH 10/13] pmem: kill wmb_pmem() Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-04 20:53 ` [PATCH 11/13] Revert "KVM: x86: add pcommit support" Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-06 15:14   ` Paolo Bonzini
2016-06-06 16:14     ` Dan Williams
2016-06-04 20:53 ` [PATCH 12/13] x86/insn: remove pcommit Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-04 20:53 ` [PATCH 13/13] pmem: kill __pmem address space Dan Williams
2016-06-04 20:53   ` Dan Williams
2016-06-04 22:18   ` kbuild test robot
2016-06-04 22:18     ` kbuild test robot
2016-06-05 17:41 ` [PATCH 00/13] deprecate pcommit Andy Lutomirski
2016-06-05 17:41   ` Andy Lutomirski
2016-06-05 18:48   ` Rudoff, Andy
2016-06-05 18:48     ` Rudoff, Andy

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=146507356876.8347.4735880992775196026.stgit@dwillia2-desk3.amr.corp.intel.com \
    --to=dan.j.williams@intel.com \
    --cc=david@fromorbit.com \
    --cc=hch@lst.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-nvdimm@lists.01.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.