[ndctl PATCH 1/2] ndctl/namespace: remove open coded is_namespace_active()

[ndctl PATCH 1/2] ndctl/namespace: remove open coded is_namespace_active()

[Vishal Verma]

Remove the open coded namespace active check with the one provided by
libndctl - ndctl_namespace_is_active().

is_namespace_active() additionally checked for a non-NULL 'ndns', which
the libndctl API doesn't do. However, all the callers either performed
that check themselves, or made prior assumptions about ndns being valid
by dereferencing it earlier.

Cc: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
---
 ndctl/namespace.c | 14 +++-----------
 1 file changed, 3 insertions(+), 11 deletions(-)

diff --git a/ndctl/namespace.c b/ndctl/namespace.c
index 7fb0007..ccd46d0 100644
--- a/ndctl/namespace.c
+++ b/ndctl/namespace.c
@@ -453,14 +453,6 @@ static int setup_namespace(struct ndctl_region *region,
 	return rc;
 }
 
-static int is_namespace_active(struct ndctl_namespace *ndns)
-{
-	return ndns && (ndctl_namespace_is_enabled(ndns)
-		|| ndctl_namespace_get_pfn(ndns)
-		|| ndctl_namespace_get_dax(ndns)
-		|| ndctl_namespace_get_btt(ndns));
-}
-
 /*
  * validate_namespace_options - init parameters for setup_namespace
  * @region: parent of the namespace to create / reconfigure
@@ -787,7 +779,7 @@ static struct ndctl_namespace *region_get_namespace(struct ndctl_region *region)
 	/* prefer the 0th namespace if it is idle */
 	ndctl_namespace_foreach(region, ndns)
 		if (ndctl_namespace_get_id(ndns) == 0
-				&& !is_namespace_active(ndns))
+				&& !ndctl_namespace_is_active(ndns))
 			return ndns;
 	return ndctl_region_get_namespace_seed(region);
 }
@@ -827,7 +819,7 @@ static int namespace_create(struct ndctl_region *region)
 		p.size = available;
 
 	ndns = region_get_namespace(region);
-	if (!ndns || is_namespace_active(ndns)) {
+	if (!ndns || ndctl_namespace_is_active(ndns)) {
 		debug("%s: no %s namespace seed\n", devname,
 				ndns ? "idle" : "available");
 		return -EAGAIN;
@@ -1074,7 +1066,7 @@ static int namespace_reconfig(struct ndctl_region *region,
 	}
 
 	ndns = region_get_namespace(region);
-	if (!ndns || is_namespace_active(ndns)) {
+	if (!ndns || !ndctl_namespace_is_active(ndns)) {
 		debug("%s: no %s namespace seed\n",
 				ndctl_region_get_devname(region),
 				ndns ? "idle" : "available");
-- 
2.20.1
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

[ndctl PATCH 2/2] ndctl/namespace: introduce ndctl_namespace_is_configurable()

[Vishal Verma]

The motivation for this change is that we want to refrain from
(re)configuring what appear to be partially configured namespaces.
Namespaces may end up in a state that looks partially configured when
the kernel isn't able to enable a namespace due to mismatched
PAGE_SIZE expectations. In such cases, ndctl should not treat those
as unconfigured 'seed' namespaces, and reuse them.

Add a new ndctl_namespace_is_configurable API, whish tests whether a
namespaces is active, and whether it is partially configured. If neither
are true, then it can be used for (re)configuration. Additionally, deal
with the corner case of ND_DEVICE_NAMESPACE_IO (legacy PMEM) namespaces
by testing whether the size attribute is read-only (which indicates such
a namespace). Legacy namespaces always appear as configured, since their
size cannot be changed, but they are also always re-configurable.

Use this API in namespace_reconfig() and namespace_create() to enable
this partial configuration detection.

Cc: Dan Williams <dan.j.williams@intel.com>
Reported-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
---

Hi Aneesh - could you please test these patches and see if they work for
the pfn_sb PAGE_SIZE mismatch scenario?

 ndctl/lib/libndctl.c   | 37 +++++++++++++++++++++++++++++++++++++
 ndctl/lib/libndctl.sym |  4 ++++
 ndctl/libndctl.h       |  1 +
 ndctl/namespace.c      |  6 +++---
 util/sysfs.c           | 18 ++++++++++++++++++
 util/sysfs.h           |  3 +++
 6 files changed, 66 insertions(+), 3 deletions(-)

diff --git a/ndctl/lib/libndctl.c b/ndctl/lib/libndctl.c
index 6596f94..1674cf9 100644
--- a/ndctl/lib/libndctl.c
+++ b/ndctl/lib/libndctl.c
@@ -3985,6 +3985,21 @@ static void region_refresh_children(struct ndctl_region *region)
 	daxs_init(region);
 }
 
+static bool namespace_size_is_writeable(struct ndctl_namespace *ndns)
+{
+	struct ndctl_ctx *ctx = ndctl_namespace_get_ctx(ndns);
+	char *path = ndns->ndns_buf;
+	int len = ndns->buf_len;
+
+	if (snprintf(path, len, "%s/size", ndns->ndns_path) >= len) {
+		err(ctx, "%s: buffer too small!\n",
+				ndctl_namespace_get_devname(ndns));
+		return false;
+	}
+
+	return sysfs_attr_writable(ctx, path);
+}
+
 NDCTL_EXPORT bool ndctl_namespace_is_active(struct ndctl_namespace *ndns)
 {
 	struct ndctl_btt *btt = ndctl_namespace_get_btt(ndns);
@@ -4182,6 +4197,28 @@ NDCTL_EXPORT int ndctl_namespace_is_configured(struct ndctl_namespace *ndns)
 	}
 }
 
+/*
+ * Check if a given 'seed' namespace is ok to configure.
+ * If a size or uuid is present, it is considered not configurable, except
+ * in the case of legacy (ND_DEVICE_NAMESPACE_IO) namespaces. In that case,
+ * the size is never zero, but the namespace can still be reconfigured.
+ * Detect this by testing whether the size is writable. If it is not writable,
+ * then it is ok to (re)configure. In other words, legacy namespaces are always
+ * 'configured', but they are also always 'configurable'.
+ */
+NDCTL_EXPORT int ndctl_namespace_is_configurable(struct ndctl_namespace *ndns)
+{
+	if (ndctl_namespace_is_active(ndns))
+		return 0;
+	if (ndctl_namespace_is_configured(ndns)) {
+		if (!namespace_size_is_writeable(ndns))
+			return 1;
+		return 0;
+	}
+	/* !active and !configured is configurable */
+	return 1;
+}
+
 NDCTL_EXPORT void ndctl_namespace_get_uuid(struct ndctl_namespace *ndns, uuid_t uu)
 {
 	memcpy(uu, ndns->uuid, sizeof(uuid_t));
diff --git a/ndctl/lib/libndctl.sym b/ndctl/lib/libndctl.sym
index c93c1ee..835d36b 100644
--- a/ndctl/lib/libndctl.sym
+++ b/ndctl/lib/libndctl.sym
@@ -423,3 +423,7 @@ LIBNDCTL_21 {
 LIBNDCTL_22 {
 	ndctl_dimm_security_is_frozen;
 } LIBNDCTL_21;
+
+LIBNDCTL_23 {
+	ndctl_namespace_is_configurable;
+} LIBNDCTL_22;
diff --git a/ndctl/libndctl.h b/ndctl/libndctl.h
index db398b3..e4dd22f 100644
--- a/ndctl/libndctl.h
+++ b/ndctl/libndctl.h
@@ -491,6 +491,7 @@ int ndctl_namespace_disable_safe(struct ndctl_namespace *ndns);
 bool ndctl_namespace_is_active(struct ndctl_namespace *ndns);
 int ndctl_namespace_is_valid(struct ndctl_namespace *ndns);
 int ndctl_namespace_is_configured(struct ndctl_namespace *ndns);
+int ndctl_namespace_is_configurable(struct ndctl_namespace *ndns);
 int ndctl_namespace_delete(struct ndctl_namespace *ndns);
 int ndctl_namespace_set_uuid(struct ndctl_namespace *ndns, uuid_t uu);
 void ndctl_namespace_get_uuid(struct ndctl_namespace *ndns, uuid_t uu);
diff --git a/ndctl/namespace.c b/ndctl/namespace.c
index ccd46d0..2979e78 100644
--- a/ndctl/namespace.c
+++ b/ndctl/namespace.c
@@ -779,7 +779,7 @@ static struct ndctl_namespace *region_get_namespace(struct ndctl_region *region)
 	/* prefer the 0th namespace if it is idle */
 	ndctl_namespace_foreach(region, ndns)
 		if (ndctl_namespace_get_id(ndns) == 0
-				&& !ndctl_namespace_is_active(ndns))
+				&& ndctl_namespace_is_configurable(ndns))
 			return ndns;
 	return ndctl_region_get_namespace_seed(region);
 }
@@ -819,7 +819,7 @@ static int namespace_create(struct ndctl_region *region)
 		p.size = available;
 
 	ndns = region_get_namespace(region);
-	if (!ndns || ndctl_namespace_is_active(ndns)) {
+	if (!ndns || !ndctl_namespace_is_configurable(ndns)) {
 		debug("%s: no %s namespace seed\n", devname,
 				ndns ? "idle" : "available");
 		return -EAGAIN;
@@ -1066,7 +1066,7 @@ static int namespace_reconfig(struct ndctl_region *region,
 	}
 
 	ndns = region_get_namespace(region);
-	if (!ndns || !ndctl_namespace_is_active(ndns)) {
+	if (!ndns || !ndctl_namespace_is_configurable(ndns)) {
 		debug("%s: no %s namespace seed\n",
 				ndctl_region_get_devname(region),
 				ndns ? "idle" : "available");
diff --git a/util/sysfs.c b/util/sysfs.c
index 9f7bc1f..773eeff 100644
--- a/util/sysfs.c
+++ b/util/sysfs.c
@@ -20,6 +20,7 @@
 #include <ctype.h>
 #include <fcntl.h>
 #include <dirent.h>
+#include <stdbool.h>
 #include <sys/stat.h>
 #include <sys/types.h>
 #include <sys/ioctl.h>
@@ -84,6 +85,23 @@ int __sysfs_write_attr_quiet(struct log_ctx *ctx, const char *path,
 	return write_attr(ctx, path, buf, 1);
 }
 
+bool __sysfs_attr_writable(struct log_ctx *ctx, const char *path)
+{
+	int fd = open(path, O_WRONLY|O_CLOEXEC);
+	bool rc = false;
+
+	if (fd < 0) {
+		if (errno != EACCES)
+			log_dbg(ctx, "failed to open %s: %s\n",
+				path, strerror(errno));
+	} else {
+		rc = true;
+	}
+
+	close(fd);
+	return rc;
+}
+
 int __sysfs_device_parse(struct log_ctx *ctx, const char *base_path,
 		const char *dev_name, void *parent, add_dev_fn add_dev)
 {
diff --git a/util/sysfs.h b/util/sysfs.h
index fb169c6..e487ed4 100644
--- a/util/sysfs.h
+++ b/util/sysfs.h
@@ -14,6 +14,7 @@
 #define __UTIL_SYSFS_H__
 
 #include <string.h>
+#include <stdbool.h>
 
 typedef void *(*add_dev_fn)(void *parent, int id, const char *dev_path);
 
@@ -24,12 +25,14 @@ int __sysfs_read_attr(struct log_ctx *ctx, const char *path, char *buf);
 int __sysfs_write_attr(struct log_ctx *ctx, const char *path, const char *buf);
 int __sysfs_write_attr_quiet(struct log_ctx *ctx, const char *path,
 		const char *buf);
+bool __sysfs_attr_writable(struct log_ctx *ctx, const char *path);
 int __sysfs_device_parse(struct log_ctx *ctx, const char *base_path,
 		const char *dev_name, void *parent, add_dev_fn add_dev);
 
 #define sysfs_read_attr(c, p, b) __sysfs_read_attr(&(c)->ctx, (p), (b))
 #define sysfs_write_attr(c, p, b) __sysfs_write_attr(&(c)->ctx, (p), (b))
 #define sysfs_write_attr_quiet(c, p, b) __sysfs_write_attr_quiet(&(c)->ctx, (p), (b))
+#define sysfs_attr_writable(c, p) __sysfs_attr_writable(&(c)->ctx, (p))
 #define sysfs_device_parse(c, b, d, p, fn) __sysfs_device_parse(&(c)->ctx, \
 		(b), (d), (p), (fn))
 
-- 
2.20.1
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

Re: [ndctl PATCH 2/2] ndctl/namespace: introduce ndctl_namespace_is_configurable()

[Dan Williams]

On Fri, Nov 1, 2019 at 1:27 PM Vishal Verma <vishal.l.verma@intel.com> wrote:
>
> The motivation for this change is that we want to refrain from
> (re)configuring what appear to be partially configured namespaces.
> Namespaces may end up in a state that looks partially configured when
> the kernel isn't able to enable a namespace due to mismatched
> PAGE_SIZE expectations. In such cases, ndctl should not treat those
> as unconfigured 'seed' namespaces, and reuse them.
>
> Add a new ndctl_namespace_is_configurable API, whish tests whether a
> namespaces is active, and whether it is partially configured. If neither
> are true, then it can be used for (re)configuration. Additionally, deal
> with the corner case of ND_DEVICE_NAMESPACE_IO (legacy PMEM) namespaces
> by testing whether the size attribute is read-only (which indicates such
> a namespace). Legacy namespaces always appear as configured, since their
> size cannot be changed, but they are also always re-configurable.
>
> Use this API in namespace_reconfig() and namespace_create() to enable
> this partial configuration detection.

A couple comments.... I think it's probably sufficient to just check
for ND_DEVICE_NAMESPACE_IO as I don't anticipate we'll ever have more
than one namespace type with a read-only size attribute. Also, how
about s/ndctl_namespace_is_configurable/ndctl_namespace_is_configuration_idle/?
Because to me all namespaces are always "configurable", but some may
have active non-default properties set.
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

Re: [ndctl PATCH 2/2] ndctl/namespace: introduce ndctl_namespace_is_configurable()

[Verma, Vishal L]

On Mon, 2019-11-04 at 08:35 -0800, Dan Williams wrote:
> On Fri, Nov 1, 2019 at 1:27 PM Vishal Verma <vishal.l.verma@intel.com> wrote:
> > The motivation for this change is that we want to refrain from
> > (re)configuring what appear to be partially configured namespaces.
> > Namespaces may end up in a state that looks partially configured when
> > the kernel isn't able to enable a namespace due to mismatched
> > PAGE_SIZE expectations. In such cases, ndctl should not treat those
> > as unconfigured 'seed' namespaces, and reuse them.
> > 
> > Add a new ndctl_namespace_is_configurable API, whish tests whether a
> > namespaces is active, and whether it is partially configured. If neither
> > are true, then it can be used for (re)configuration. Additionally, deal
> > with the corner case of ND_DEVICE_NAMESPACE_IO (legacy PMEM) namespaces
> > by testing whether the size attribute is read-only (which indicates such
> > a namespace). Legacy namespaces always appear as configured, since their
> > size cannot be changed, but they are also always re-configurable.
> > 
> > Use this API in namespace_reconfig() and namespace_create() to enable
> > this partial configuration detection.
> 
> A couple comments.... I think it's probably sufficient to just check
> for ND_DEVICE_NAMESPACE_IO as I don't anticipate we'll ever have more
> than one namespace type with a read-only size attribute.

Yep I did notice I could do that, but I decided to go to the source of
it instead of adding another ND_DEVICE_NAMESPACE_IO assumption. Also I
had already written sysfs_attr_writable() before I noticed that :)
But there are already assumptions around ND_DEVICE_NAMESPACE_IO, so
adding another one here is fine.

> Also, how about s/ndctl_namespace_is_configurable/ndctl_namespace_is_configuration_idle/?
> Because to me all namespaces are always "configurable", but some may
> have active non-default properties set.

Sounds reasonable, but how about ndctl_namespace_has_partial_config(),
which I think describes the situation better, and makes it
straightforward for a potential future --really-force (or -ff) option
where we might still want to blow away anything on this namespace and
reclaim it.
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

Re: [ndctl PATCH 2/2] ndctl/namespace: introduce ndctl_namespace_is_configurable()

[Dan Williams]

On Mon, Nov 4, 2019 at 10:48 AM Verma, Vishal L
<vishal.l.verma@intel.com> wrote:
>
> On Mon, 2019-11-04 at 08:35 -0800, Dan Williams wrote:
> > On Fri, Nov 1, 2019 at 1:27 PM Vishal Verma <vishal.l.verma@intel.com> wrote:
> > > The motivation for this change is that we want to refrain from
> > > (re)configuring what appear to be partially configured namespaces.
> > > Namespaces may end up in a state that looks partially configured when
> > > the kernel isn't able to enable a namespace due to mismatched
> > > PAGE_SIZE expectations. In such cases, ndctl should not treat those
> > > as unconfigured 'seed' namespaces, and reuse them.
> > >
> > > Add a new ndctl_namespace_is_configurable API, whish tests whether a
> > > namespaces is active, and whether it is partially configured. If neither
> > > are true, then it can be used for (re)configuration. Additionally, deal
> > > with the corner case of ND_DEVICE_NAMESPACE_IO (legacy PMEM) namespaces
> > > by testing whether the size attribute is read-only (which indicates such
> > > a namespace). Legacy namespaces always appear as configured, since their
> > > size cannot be changed, but they are also always re-configurable.
> > >
> > > Use this API in namespace_reconfig() and namespace_create() to enable
> > > this partial configuration detection.
> >
> > A couple comments.... I think it's probably sufficient to just check
> > for ND_DEVICE_NAMESPACE_IO as I don't anticipate we'll ever have more
> > than one namespace type with a read-only size attribute.
>
> Yep I did notice I could do that, but I decided to go to the source of
> it instead of adding another ND_DEVICE_NAMESPACE_IO assumption. Also I
> had already written sysfs_attr_writable() before I noticed that :)
> But there are already assumptions around ND_DEVICE_NAMESPACE_IO, so
> adding another one here is fine.
>
> > Also, how about s/ndctl_namespace_is_configurable/ndctl_namespace_is_configuration_idle/?
> > Because to me all namespaces are always "configurable", but some may
> > have active non-default properties set.
>
> Sounds reasonable, but how about ndctl_namespace_has_partial_config(),
> which I think describes the situation better, and makes it
> straightforward for a potential future --really-force (or -ff) option
> where we might still want to blow away anything on this namespace and
> reclaim it.

Does "_has_partial_config()" imply "_has_full_config()"? In other
words, what ndctl cares about are namespaces that can effectively be
used as seeds with no existing configuration parameters having been
set. So "_is_configuration_idle()" seems unambiguous where
"_has_partial_config()" does not.
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org

Re: [ndctl PATCH 2/2] ndctl/namespace: introduce ndctl_namespace_is_configurable()

[Verma, Vishal L]

On Mon, 2019-11-04 at 10:58 -0800, Dan Williams wrote:
> > 
> > > Also, how about s/ndctl_namespace_is_configurable/ndctl_namespace_is_configuration_idle/?
> > > Because to me all namespaces are always "configurable", but some may
> > > have active non-default properties set.
> > 
> > Sounds reasonable, but how about ndctl_namespace_has_partial_config(),
> > which I think describes the situation better, and makes it
> > straightforward for a potential future --really-force (or -ff) option
> > where we might still want to blow away anything on this namespace and
> > reclaim it.
> 
> Does "_has_partial_config()" imply "_has_full_config()"? In other
> words, what ndctl cares about are namespaces that can effectively be
> used as seeds with no existing configuration parameters having been
> set. So "_is_configuration_idle()" seems unambiguous where
> "_has_partial_config()" does not.

Good point, I'll go with _is_configuration_idle()
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org