diff mbox

[v2,4/5] acpi_nfit, libnvdimm: Add support for clear poison list and bad blocks

Message ID 148892081379.46486.4526071392091548993.stgit@djiang5-desk3.ch.intel.com (mailing list archive)
State New, archived
Headers show

Commit Message

Dave Jiang March 7, 2017, 9:06 p.m. UTC
Providing mechanism to clear poison list via the ndctl ND_CMD_CLEAR_ERROR
call. We will update the poison list and also the badblocks at region level
if the region is in dax mode or in pmem mode and not active.

Signed-off-by: Dave Jiang <dave.jiang@intel.com>
---
 drivers/acpi/nfit/core.c         |   63 ++++++++++++++++++++++++++++++++++++++
 drivers/acpi/nfit/nfit.h         |    2 +
 drivers/nvdimm/bus.c             |   55 +++++++++++++++++++++++++++++----
 drivers/nvdimm/core.c            |   17 ++++++++--
 drivers/nvdimm/region.c          |    9 +++++
 include/linux/libnvdimm.h        |    6 +++-
 tools/testing/nvdimm/test/nfit.c |   23 ++++++++------
 7 files changed, 154 insertions(+), 21 deletions(-)

Comments

Johannes Thumshirn March 8, 2017, 8 a.m. UTC | #1
On 03/07/2017 10:06 PM, Dave Jiang wrote:
> Providing mechanism to clear poison list via the ndctl ND_CMD_CLEAR_ERROR
> call. We will update the poison list and also the badblocks at region level
> if the region is in dax mode or in pmem mode and not active.
> 
> Signed-off-by: Dave Jiang <dave.jiang@intel.com>
> ---

Thanks,
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Dan Williams March 15, 2017, 7:09 a.m. UTC | #2
On Tue, Mar 7, 2017 at 1:06 PM, Dave Jiang <dave.jiang@intel.com> wrote:
> Providing mechanism to clear poison list via the ndctl ND_CMD_CLEAR_ERROR
> call. We will update the poison list and also the badblocks at region level
> if the region is in dax mode or in pmem mode and not active.
>
> Signed-off-by: Dave Jiang <dave.jiang@intel.com>
> ---
>  drivers/acpi/nfit/core.c         |   63 ++++++++++++++++++++++++++++++++++++++
>  drivers/acpi/nfit/nfit.h         |    2 +
>  drivers/nvdimm/bus.c             |   55 +++++++++++++++++++++++++++++----
>  drivers/nvdimm/core.c            |   17 ++++++++--
>  drivers/nvdimm/region.c          |    9 +++++
>  include/linux/libnvdimm.h        |    6 +++-
>  tools/testing/nvdimm/test/nfit.c |   23 ++++++++------
>  7 files changed, 154 insertions(+), 21 deletions(-)
>
> diff --git a/drivers/acpi/nfit/core.c b/drivers/acpi/nfit/core.c
> index e7b05df..97bc2bf 100644
> --- a/drivers/acpi/nfit/core.c
> +++ b/drivers/acpi/nfit/core.c
> @@ -94,6 +94,63 @@ static struct acpi_device *to_acpi_dev(struct acpi_nfit_desc *acpi_desc)
>         return to_acpi_device(acpi_desc->dev);
>  }
>
> +int acpi_nfit_forget_poison(struct nvdimm_bus_descriptor *nd_desc,
> +               unsigned int cmd, void *buf)
> +{
> +       struct acpi_nfit_desc *acpi_desc = to_acpi_nfit_desc(nd_desc);
> +       struct nvdimm_bus *nvdimm_bus = acpi_desc->nvdimm_bus;
> +       struct nfit_spa *nfit_spa;
> +       struct nd_cmd_clear_error *clear_err = buf;
> +       int found_match = 0;
> +
> +       if ((cmd != ND_CMD_CLEAR_ERROR) || !nvdimm_bus || !clear_err->cleared)
> +               return 0;

I think we need to move part of this check out to the caller because
you don't have enough information here to differentiate bus commands
from dimm commands. That "cmd != ND_CMD_CLEAR_ERROR" check fails if
the command is ND_CMD_GET_CONFIG_SIZE on a DIMM device.

[..]
> @@ -353,6 +410,12 @@ int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, struct nvdimm *nvdimm,
>         }
>
>         xlat_rc = xlat_status(nvdimm, buf, cmd, fw_status);
> +       if (xlat_rc >= 0) {

I think we should make this:

    if (nvdimm == NULL && cmd == ND_CMD_CLEAR_ERROR && xlat_rc >= 0)

> +               xlat_rc = acpi_nfit_forget_poison(nd_desc, cmd, buf);

We shouldn't be messing with the status return for the clear error
command that needs to go unmodified back to the caller.

> +               if (xlat_rc)
> +                       dev_err(dev, "%s:%s %s unable to forget poison\n",
> +                                       __func__, dimm_name, cmd_name);

I think this can just be ignored. The acpi_nfit_forget_poison()
routine should just be marked void since the only error it knows how
to return was "no error was found to clear". That's only worth a debug
statement not an error return.
diff mbox

Patch

diff --git a/drivers/acpi/nfit/core.c b/drivers/acpi/nfit/core.c
index e7b05df..97bc2bf 100644
--- a/drivers/acpi/nfit/core.c
+++ b/drivers/acpi/nfit/core.c
@@ -94,6 +94,63 @@  static struct acpi_device *to_acpi_dev(struct acpi_nfit_desc *acpi_desc)
 	return to_acpi_device(acpi_desc->dev);
 }
 
+int acpi_nfit_forget_poison(struct nvdimm_bus_descriptor *nd_desc,
+		unsigned int cmd, void *buf)
+{
+	struct acpi_nfit_desc *acpi_desc = to_acpi_nfit_desc(nd_desc);
+	struct nvdimm_bus *nvdimm_bus = acpi_desc->nvdimm_bus;
+	struct nfit_spa *nfit_spa;
+	struct nd_cmd_clear_error *clear_err = buf;
+	int found_match = 0;
+
+	if ((cmd != ND_CMD_CLEAR_ERROR) || !nvdimm_bus || !clear_err->cleared)
+		return 0;
+
+	/* clearing the poison list we keep track of */
+	__nvdimm_forget_poison(nvdimm_bus, clear_err->address,
+			clear_err->cleared);
+
+	mutex_lock(&acpi_desc->init_mutex);
+	list_for_each_entry(nfit_spa, &acpi_desc->spas, list) {
+		struct acpi_nfit_system_address *spa = nfit_spa->spa;
+		resource_size_t spa_begin, spa_end, clear_begin, clear_end;
+
+		if (nfit_spa_type(spa) != NFIT_SPA_PM)
+			continue;
+
+		spa_begin = spa->address;
+		spa_end = spa->address + spa->length - 1;
+		clear_begin = clear_err->address;
+		clear_end = clear_err->address + clear_err->cleared - 1;
+
+		/* make sure clear_err range is within a SPA range */
+		if ((clear_begin >= spa_begin &&
+				clear_begin < spa_end) &&
+				(clear_end > spa_begin &&
+				 clear_end <= spa_end)) {
+			found_match = 1;
+			break;
+		}
+	}
+	mutex_unlock(&acpi_desc->init_mutex);
+
+	/* now sync the badblocks lists from the poison list */
+	if (found_match) {
+		struct resource res;
+
+		res.start = clear_err->address;
+		res.end = clear_err->address + clear_err->cleared - 1;
+		nvdimm_region_badblocks_clear(nfit_spa->nd_region, &res);
+	} else {
+		dev_dbg(acpi_desc->dev,
+			"Unable to find NFIT SPA that matches error addr\n");
+		return -ENODEV;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(acpi_nfit_forget_poison);
+
 static int xlat_bus_status(void *buf, unsigned int cmd, u32 status)
 {
 	struct nd_cmd_clear_error *clear_err;
@@ -353,6 +410,12 @@  int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, struct nvdimm *nvdimm,
 	}
 
 	xlat_rc = xlat_status(nvdimm, buf, cmd, fw_status);
+	if (xlat_rc >= 0) {
+		xlat_rc = acpi_nfit_forget_poison(nd_desc, cmd, buf);
+		if (xlat_rc)
+			dev_err(dev, "%s:%s %s unable to forget poison\n",
+					__func__, dimm_name, cmd_name);
+	}
 
 	if (cmd_rc)
 		*cmd_rc = xlat_rc;
diff --git a/drivers/acpi/nfit/nfit.h b/drivers/acpi/nfit/nfit.h
index fc29c2e..e881780 100644
--- a/drivers/acpi/nfit/nfit.h
+++ b/drivers/acpi/nfit/nfit.h
@@ -243,4 +243,6 @@  void __acpi_nvdimm_notify(struct device *dev, u32 event);
 int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, struct nvdimm *nvdimm,
 		unsigned int cmd, void *buf, unsigned int buf_len, int *cmd_rc);
 void acpi_nfit_desc_init(struct acpi_nfit_desc *acpi_desc, struct device *dev);
+int acpi_nfit_forget_poison(struct nvdimm_bus_descriptor *nd_desc,
+		unsigned int cmd, void *buf);
 #endif /* __NFIT_H__ */
diff --git a/drivers/nvdimm/bus.c b/drivers/nvdimm/bus.c
index 23d4a17..9e76143 100644
--- a/drivers/nvdimm/bus.c
+++ b/drivers/nvdimm/bus.c
@@ -27,6 +27,7 @@ 
 #include <linux/nd.h>
 #include "nd-core.h"
 #include "nd.h"
+#include "pfn.h"
 
 int nvdimm_major;
 static int nvdimm_bus_major;
@@ -218,7 +219,8 @@  long nvdimm_clear_poison(struct device *dev, phys_addr_t phys,
 	if (cmd_rc < 0)
 		return cmd_rc;
 
-	nvdimm_clear_from_poison_list(nvdimm_bus, phys, len);
+	nvdimm_forget_poison(nvdimm_bus, phys, len);
+
 	return clear_err.cleared;
 }
 EXPORT_SYMBOL_GPL(nvdimm_clear_poison);
@@ -769,16 +771,55 @@  void wait_nvdimm_bus_probe_idle(struct device *dev)
 	} while (true);
 }
 
-static int pmem_active(struct device *dev, void *data)
+static int nd_pmem_forget_poison_check(struct device *dev, void *data)
 {
-	if (is_nd_pmem(dev) && dev->driver)
+	struct nd_cmd_clear_error *clear_err =
+		(struct nd_cmd_clear_error *)data;
+	struct nd_btt *nd_btt = is_nd_btt(dev) ? to_nd_btt(dev) : NULL;
+	struct nd_pfn *nd_pfn = is_nd_pfn(dev) ? to_nd_pfn(dev) : NULL;
+	struct nd_dax *nd_dax = is_nd_dax(dev) ? to_nd_dax(dev) : NULL;
+	struct nd_namespace_common *ndns = NULL;
+	struct nd_namespace_io *nsio;
+	resource_size_t offset = 0, end_trunc = 0, start, end, pstart, pend;
+
+	if (nd_dax || !dev->driver)
+		return 0;
+
+	start = clear_err->address;
+	end = clear_err->address + clear_err->cleared - 1;
+
+	if (nd_btt || nd_pfn || nd_dax) {
+		if (nd_btt)
+			ndns = nd_btt->ndns;
+		else if (nd_pfn)
+			ndns = nd_pfn->ndns;
+		else if (nd_dax)
+			ndns = nd_dax->nd_pfn.ndns;
+
+		if (!ndns)
+			return 0;
+	} else
+		ndns = to_ndns(dev);
+
+	nsio = to_nd_namespace_io(&ndns->dev);
+	pstart = nsio->res.start + offset;
+	pend = nsio->res.end - end_trunc;
+
+	if ((pstart >= start) && (pend <= end))
 		return -EBUSY;
+
 	return 0;
+
+}
+
+static int nd_ns_forget_poison_check(struct device *dev, void *data)
+{
+	return device_for_each_child(dev, data, nd_pmem_forget_poison_check);
 }
 
 /* set_config requires an idle interleave set */
 static int nd_cmd_clear_to_send(struct nvdimm_bus *nvdimm_bus,
-		struct nvdimm *nvdimm, unsigned int cmd)
+		struct nvdimm *nvdimm, unsigned int cmd, void *data)
 {
 	struct nvdimm_bus_descriptor *nd_desc = nvdimm_bus->nd_desc;
 
@@ -792,8 +833,8 @@  static int nd_cmd_clear_to_send(struct nvdimm_bus *nvdimm_bus,
 
 	/* require clear error to go through the pmem driver */
 	if (!nvdimm && cmd == ND_CMD_CLEAR_ERROR)
-		return device_for_each_child(&nvdimm_bus->dev, NULL,
-				pmem_active);
+		return device_for_each_child(&nvdimm_bus->dev, data,
+				nd_ns_forget_poison_check);
 
 	if (!nvdimm || cmd != ND_CMD_SET_CONFIG_DATA)
 		return 0;
@@ -927,7 +968,7 @@  static int __nd_ioctl(struct nvdimm_bus *nvdimm_bus, struct nvdimm *nvdimm,
 	}
 
 	nvdimm_bus_lock(&nvdimm_bus->dev);
-	rc = nd_cmd_clear_to_send(nvdimm_bus, nvdimm, cmd);
+	rc = nd_cmd_clear_to_send(nvdimm_bus, nvdimm, cmd, buf);
 	if (rc)
 		goto out_unlock;
 
diff --git a/drivers/nvdimm/core.c b/drivers/nvdimm/core.c
index 9303cfe..40a3da0 100644
--- a/drivers/nvdimm/core.c
+++ b/drivers/nvdimm/core.c
@@ -574,14 +574,15 @@  int nvdimm_bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length)
 }
 EXPORT_SYMBOL_GPL(nvdimm_bus_add_poison);
 
-void nvdimm_clear_from_poison_list(struct nvdimm_bus *nvdimm_bus,
-		phys_addr_t start, unsigned int len)
+void __nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus, phys_addr_t start,
+		unsigned int len)
 {
 	struct list_head *poison_list = &nvdimm_bus->poison_list;
 	u64 clr_end = start + len - 1;
 	struct nd_poison *pl, *next;
 
-	nvdimm_bus_lock(&nvdimm_bus->dev);
+	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
+
 	WARN_ON_ONCE(list_empty(poison_list));
 
 	/*
@@ -634,9 +635,17 @@  void nvdimm_clear_from_poison_list(struct nvdimm_bus *nvdimm_bus,
 			continue;
 		}
 	}
+}
+EXPORT_SYMBOL_GPL(__nvdimm_forget_poison);
+
+void nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus,
+		phys_addr_t start, unsigned int len)
+{
+	nvdimm_bus_lock(&nvdimm_bus->dev);
+	__nvdimm_forget_poison(nvdimm_bus, start, len);
 	nvdimm_bus_unlock(&nvdimm_bus->dev);
 }
-EXPORT_SYMBOL_GPL(nvdimm_clear_from_poison_list);
+EXPORT_SYMBOL_GPL(nvdimm_forget_poison);
 
 #ifdef CONFIG_BLK_DEV_INTEGRITY
 int nd_integrity_init(struct gendisk *disk, unsigned long meta_size)
diff --git a/drivers/nvdimm/region.c b/drivers/nvdimm/region.c
index 869a886..211435c 100644
--- a/drivers/nvdimm/region.c
+++ b/drivers/nvdimm/region.c
@@ -131,6 +131,15 @@  static void nd_region_notify(struct device *dev, enum nvdimm_event event)
 	device_for_each_child(dev, &event, child_notify);
 }
 
+void nvdimm_region_badblocks_clear(struct nd_region *nd_region,
+		struct resource *res)
+{
+	sector_t sector = (res->start - nd_region->ndr_start) >> 9;
+
+	badblocks_clear(&nd_region->bb, sector, resource_size(res) >> 9);
+}
+EXPORT_SYMBOL_GPL(nvdimm_region_badblocks_clear);
+
 static struct nd_device_driver nd_region_driver = {
 	.probe = nd_region_probe,
 	.remove = nd_region_remove,
diff --git a/include/linux/libnvdimm.h b/include/linux/libnvdimm.h
index 77e7af3..97b46ca 100644
--- a/include/linux/libnvdimm.h
+++ b/include/linux/libnvdimm.h
@@ -120,7 +120,9 @@  static inline struct nd_blk_region_desc *to_blk_region_desc(
 }
 
 int nvdimm_bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length);
-void nvdimm_clear_from_poison_list(struct nvdimm_bus *nvdimm_bus,
+void nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus,
+		phys_addr_t start, unsigned int len);
+void __nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus,
 		phys_addr_t start, unsigned int len);
 struct nvdimm_bus *nvdimm_bus_register(struct device *parent,
 		struct nvdimm_bus_descriptor *nfit_desc);
@@ -162,4 +164,6 @@  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);
+void nvdimm_region_badblocks_clear(struct nd_region *nd_region,
+		struct resource *res);
 #endif /* __LIBNVDIMM_H__ */
diff --git a/tools/testing/nvdimm/test/nfit.c b/tools/testing/nvdimm/test/nfit.c
index 798f176..bdd2a98 100644
--- a/tools/testing/nvdimm/test/nfit.c
+++ b/tools/testing/nvdimm/test/nfit.c
@@ -319,7 +319,8 @@  static int nfit_test_cmd_ars_status(struct ars_state *ars_state,
 	return 0;
 }
 
-static int nfit_test_cmd_clear_error(struct nd_cmd_clear_error *clear_err,
+static int nfit_test_cmd_setup_clear_error(
+		struct nd_cmd_clear_error *clear_err,
 		unsigned int buf_len, int *cmd_rc)
 {
 	const u64 mask = NFIT_TEST_CLEAR_ERR_UNIT - 1;
@@ -330,14 +331,11 @@  static int nfit_test_cmd_clear_error(struct nd_cmd_clear_error *clear_err,
 		return -EINVAL;
 
 	/*
-	 * Report 'all clear' success for all commands even though a new
-	 * scrub will find errors again.  This is enough to have the
-	 * error removed from the 'badblocks' tracking in the pmem
-	 * driver.
+	 * since this is a test and there's no actual _DSM command being
+	 * issued, we need to fake the result.
 	 */
-	clear_err->status = 0;
-	clear_err->cleared = clear_err->length;
-	*cmd_rc = 0;
+	*cmd_rc = clear_err->cleared = clear_err->length;
+
 	return 0;
 }
 
@@ -465,7 +463,14 @@  static int nfit_test_ctl(struct nvdimm_bus_descriptor *nd_desc,
 					cmd_rc);
 			break;
 		case ND_CMD_CLEAR_ERROR:
-			rc = nfit_test_cmd_clear_error(buf, buf_len, cmd_rc);
+			rc = nfit_test_cmd_setup_clear_error(buf,
+					buf_len, cmd_rc);
+			if (rc < 0)
+				return rc;
+
+			rc = acpi_nfit_forget_poison(nd_desc, cmd, buf);
+			if (rc < 0)
+				return rc;
 			break;
 		default:
 			return -ENOTTY;