Enhance the ZFS vdev layer to maintain both a logical and a physical

minimum allocation size for devices.  Use this information to
automatically increase ZFS's minimum allocation size for new top-level
vdevs to a value that more closely matches the optimum device
allocation size.

Use GEOM's stripesize attribute, if set, as the physical sector
size of the GEOM.

Calculate the minimum blocksize of each metaslab class.  Use the
calculated value instead of SPA_MINBLOCKSIZE (512b) when determining
the likelyhood of compression yeilding a reduction in physical space
usage.

Report devices with sub-optimal block size configuration in "zpool
status".  Also properly fail attempts to attach devices with a
logical block size greater than 8kB, since this will cause corruption
to ZFS's label area.

Sponsored by:	Spectra Logic Corporaion
MFC after:	2 weeks

Background
==========
Many modern devices use physical allocation units that are much
larger than the minimum logical allocation size accessible by
external commands.  Two prevalent examples of this are 512e disk
drives (512b logical sector, 4K physical sector) and flash devices
(512b logical sector, 4K or larger allocation block size, and 128k
or larger erase block size).  Operations that modify less than the
physical sector size result in a costly read-modify-write or garbage
collection sequence on these devices.

Simply exporting the true physical sector of the device to ZFS would
yield optimal performance, but has two serious drawbacks:

1) Existing pools created with devices that have different logical
   and physical block sizes, but were configured to use the logical
   block size (e.g. because the OS version used for pool construction
   reported the logical block size instead of the physical block
   size) will suddenly find that the vdev allocation size has
   increased.  This can be easily tolerated for active members of
   the array, but ZFS would prevent replacement of a vdev with
   another identical device because it now appears that the smaller
   allocation size required by the pool is not supported by the new
   device.

2) The device's physical block size may be too large to be supported
   by ZFS.  The optimal allocation size for the vdev may be quite
   large.  For example, a RAID controller may export a vdev that
   requires read-modify-write cycles unless accessed using 64k
   aligned/sized requests.  ZFS currently has an 8k minimum block
   size limit.

Reporting both the logical and physical allocation sizes for vdevs
solves these problems.  A device may be used so long as the logical
block size is compatible with the configuration.  By comparing the
logical and physical block sizes, new configurations can be optimized
and administrators can be notified of any existing pools that are
sub-optimal.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/spa.h:
	Add the SPA_ASHIFT constant.  ZFS currently has a hard upper
	limit of 13 (8k) for ashift and this constant is used to
	both document and enforce this limit.

sys/cddl/contrib/opensolaris/uts/common/sys/fs/zfs.h:
	Add the VDEV_AUX_ASHIFT_TOO_BIG error code.

	Add fields for exporting the configured, logical, and
	physical ashift to the vdev_stat_t structure.

	Add VDEV_STAT_VALID() macro which can be used to verify the
	presence of required vdev_stat_t fields in nvlist data.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev.c:
	Provide a SYSCTL_PROC handler for "max_auto_ashift".  Since
	the limit is only referenced long after boot when a create
	operation occurs, there's no compelling need for it to be
	a boot time configurable tunable.  This also allows the
	validation code for the max_auto_ashift value to be contained
	within the sysctl handler.

	Populate the new fields in the vdev_stat_t structure.

	Fail vdev opens if the vdev reports an ashift larger than
	SPA_MAXASHIFT.

	Propogate vdev_logical_ashift and vdev_physical_ashift between
	child and parent vdevs as is done for vdev_ashift.

	In vdev_open(), restore code that fails opens for devices
	where vdev_ashift grows.  This can only happen now if the
	device's logical ashift grows, which means it really isn't
	safe to use the device.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/vdev_impl.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_file.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_geom.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_mirror.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_missing.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_raidz.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev_root.c:
	Update the vdev_open() API so that both logical (what was
	just ashift before) and physical ashift are reported.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/vdev_impl.h:
	Add two new fields, vdev_physical_ashift and vdev_logical_ashift,
	to vdev_t.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/vdev.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/spa_config.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/spa.c:
	Add vdev_ashift_optimize().  Call it anytime a new top-level
	vdev is allocated.

cddl/contrib/opensolaris/cmd/zpool/zpool_main.c:
	Add text for the VDEV_AUX_ASHIFT_TOO_BIG error.

	For each sub-optimally configured leaf vdev, report configured
	and native block sizes.

cddl/contrib/opensolaris/cmd/zpool/zpool_main.c:
cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h:
cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c:
	Introduce a new zpool status: ZPOOL_STATUS_NON_NATIVE_ASHIFT.
	This status is reported on healthy pools containing vdevs
	configured to use a block size smaller than their reported
	physical block size.

cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c:
	Update find_vdev_problem() and supporting functions to
	provide the full vdev_stat_t structure to problem checking
	routines, and to allow decent into replacing vdevs.

	Add a vdev_non_native_ashift() validator which is used on
	the full vdev tree to check for ZPOOL_STATUS_NON_NATIVE_ASHIFT.

cddl/contrib/opensolaris/lib/libzpool/common/kernel.c:
cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h:
	Enhance sysctl userland stubs now that a SYSCTL_PROC handler
	is used in vdev.c.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/metaslab.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/metaslab_impl.h:
	When the group membership of a metaslab class changes (i.e.
	when a vdev is added or removed from a pool), walk the group
	list to determine the smallest block size currently available
	and record this in the metaslab class.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/metaslab.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/metaslab.c:
	Add the metaslab_class_get_minblocksize() accessor.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/sys/zio_compress.h:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zio_compress.c:
sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:
	In zio_compress_data(), take the minimum blocksize as an
	input parameter instead of assuming SPA_MINBLOCKSIZE.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:
	In l2arc_compress_buf(), pass SPA_MINBLOCKSIZE as the minimum
	blocksize of the device.  The l2arc code performs has it's own
	code for deciding if compression is worth while, so this
	effectively disables zio_compress_data() from second guessing
	the original decision.

sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zio.c:
	In zio_write_bp_init(), use the minimum blocksize of the
	normal metaslab class when compressing data.

5 files changed, 127 insertions, 37 deletions

diff --git a/cddl/contrib/opensolaris/cmd/zpool/zpool_main.c b/cddl/contrib/opensolaris/cmd/zpool/zpool_main.c
index bfb7114..1686467 100644
--- a/cddl/contrib/opensolaris/cmd/zpool/zpool_main.c
+++ b/cddl/contrib/opensolaris/cmd/zpool/zpool_main.c
@@ -1295,12 +1295,13 @@ print_status_config(zpool_handle_t *zhp, const char *name, nvlist_t *nv,
     int namewidth, int depth, boolean_t isspare)
 {
 	nvlist_t **child;
-	uint_t c, children;
+	uint_t c, vsc, children;
 	pool_scan_stat_t *ps = NULL;
 	vdev_stat_t *vs;
 	char rbuf[6], wbuf[6], cbuf[6];
 	char *vname;
 	uint64_t notpresent;
+	uint64_t ashift;
 	spare_cbdata_t cb;
 	const char *state;
 
@@ -1309,7 +1310,7 @@ print_status_config(zpool_handle_t *zhp, const char *name, nvlist_t *nv,
 		children = 0;
 
 	verify(nvlist_lookup_uint64_array(nv, ZPOOL_CONFIG_VDEV_STATS,
-	    (uint64_t **)&vs, &c) == 0);
+	    (uint64_t **)&vs, &vsc) == 0);
 
 	state = zpool_state_to_name(vs->vs_state, vs->vs_aux);
 	if (isspare) {
@@ -1363,6 +1364,10 @@ print_status_config(zpool_handle_t *zhp, const char *name, nvlist_t *nv,
 			(void) printf(gettext("unsupported feature(s)"));
 			break;
 
+		case VDEV_AUX_ASHIFT_TOO_BIG:
+			(void) printf(gettext("unsupported minimum blocksize"));
+			break;
+
 		case VDEV_AUX_SPARED:
 			verify(nvlist_lookup_uint64(nv, ZPOOL_CONFIG_GUID,
 			    &cb.cb_guid) == 0);
@@ -1405,6 +1410,12 @@ print_status_config(zpool_handle_t *zhp, const char *name, nvlist_t *nv,
 			(void) printf(gettext("corrupted data"));
 			break;
 		}
+	} else if (children == 0 && !isspare &&
+	    VDEV_STAT_VALID(vs_physical_ashift, vsc) &&
+	    vs->vs_configured_ashift < vs->vs_physical_ashift) {
+		(void) printf(
+		    gettext("  block size: %dB configured, %dB native"),
+		    1 << vs->vs_configured_ashift, 1 << vs->vs_physical_ashift);
 	}
 
 	(void) nvlist_lookup_uint64_array(nv, ZPOOL_CONFIG_SCAN_STATS,
@@ -4268,6 +4279,15 @@ status_callback(zpool_handle_t *zhp, void *data)
 		    "'zpool clear'.\n"));
 		break;
 
+	case ZPOOL_STATUS_NON_NATIVE_ASHIFT:
+		(void) printf(gettext("status: One or more devices are "
+		    "configured to use a non-native block size.\n"
+		    "\tExpect reduced performance.\n"));
+		(void) printf(gettext("action: Replace affected devices with "
+		    "devices that support the\n\tconfigured block size, or "
+		    "migrate data to a properly configured\n\tpool.\n"));
+		break;
+
 	default:
 		/*
 		 * The remaining errors can't actually be generated, yet.
diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h
index 981e9bd..0a5ca82 100644
--- a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h
+++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h
@@ -326,6 +326,7 @@ typedef enum {
 	ZPOOL_STATUS_RESILVERING,	/* device being resilvered */
 	ZPOOL_STATUS_OFFLINE_DEV,	/* device online */
 	ZPOOL_STATUS_REMOVED_DEV,	/* removed device */
+	ZPOOL_STATUS_NON_NATIVE_ASHIFT,	/* (e.g. 512e dev with ashift of 9) */
 
 	/*
 	 * Finally, the following indicates a healthy pool.
diff --git a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c
index 71b27a1..906883c 100644
--- a/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c
+++ b/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_status.c
@@ -73,57 +73,66 @@ static char *zfs_msgid_table[] = {
 
 /* ARGSUSED */
 static int
-vdev_missing(uint64_t state, uint64_t aux, uint64_t errs)
+vdev_missing(vdev_stat_t *vs, uint_t vsc)
 {
-	return (state == VDEV_STATE_CANT_OPEN &&
-	    aux == VDEV_AUX_OPEN_FAILED);
+	return (vs->vs_state == VDEV_STATE_CANT_OPEN &&
+	    vs->vs_aux == VDEV_AUX_OPEN_FAILED);
 }
 
 /* ARGSUSED */
 static int
-vdev_faulted(uint64_t state, uint64_t aux, uint64_t errs)
+vdev_faulted(vdev_stat_t *vs, uint_t vsc)
 {
-	return (state == VDEV_STATE_FAULTED);
+	return (vs->vs_state == VDEV_STATE_FAULTED);
 }
 
 /* ARGSUSED */
 static int
-vdev_errors(uint64_t state, uint64_t aux, uint64_t errs)
+vdev_errors(vdev_stat_t *vs, uint_t vsc)
 {
-	return (state == VDEV_STATE_DEGRADED || errs != 0);
+	return (vs->vs_state == VDEV_STATE_DEGRADED ||
+	    vs->vs_read_errors != 0 || vs->vs_write_errors != 0 ||
+	    vs->vs_checksum_errors != 0);
 }
 
 /* ARGSUSED */
 static int
-vdev_broken(uint64_t state, uint64_t aux, uint64_t errs)
+vdev_broken(vdev_stat_t *vs, uint_t vsc)
 {
-	return (state == VDEV_STATE_CANT_OPEN);
+	return (vs->vs_state == VDEV_STATE_CANT_OPEN);
 }
 
 /* ARGSUSED */
 static int
-vdev_offlined(uint64_t state, uint64_t aux, uint64_t errs)
+vdev_offlined(vdev_stat_t *vs, uint_t vsc)
 {
-	return (state == VDEV_STATE_OFFLINE);
+	return (vs->vs_state == VDEV_STATE_OFFLINE);
 }
 
 /* ARGSUSED */
 static int
-vdev_removed(uint64_t state, uint64_t aux, uint64_t errs)
+vdev_removed(vdev_stat_t *vs, uint_t vsc)
 {
-	return (state == VDEV_STATE_REMOVED);
+	return (vs->vs_state == VDEV_STATE_REMOVED);
+}
+
+static int
+vdev_non_native_ashift(vdev_stat_t *vs, uint_t vsc)
+{
+	return (VDEV_STAT_VALID(vs_physical_ashift, vsc) &&
+	    vs->vs_configured_ashift < vs->vs_physical_ashift);
 }
 
 /*
  * Detect if any leaf devices that have seen errors or could not be opened.
  */
 static boolean_t
-find_vdev_problem(nvlist_t *vdev, int (*func)(uint64_t, uint64_t, uint64_t))
+find_vdev_problem(nvlist_t *vdev, int (*func)(vdev_stat_t *, uint_t),
+    boolean_t ignore_replacing)
 {
 	nvlist_t **child;
 	vdev_stat_t *vs;
-	uint_t c, children;
-	char *type;
+	uint_t c, vsc, children;
 
 	/*
 	 * Ignore problems within a 'replacing' vdev, since we're presumably in
@@ -131,23 +140,25 @@ find_vdev_problem(nvlist_t *vdev, int (*func)(uint64_t, uint64_t, uint64_t))
 	 * out again.  We'll pick up the fact that a resilver is happening
 	 * later.
 	 */
-	verify(nvlist_lookup_string(vdev, ZPOOL_CONFIG_TYPE, &type) == 0);
-	if (strcmp(type, VDEV_TYPE_REPLACING) == 0)
-		return (B_FALSE);
+	if (ignore_replacing == B_TRUE) {
+		char *type;
+
+		verify(nvlist_lookup_string(vdev, ZPOOL_CONFIG_TYPE,
+		    &type) == 0);
+		if (strcmp(type, VDEV_TYPE_REPLACING) == 0)
+			return (B_FALSE);
+	}
 
 	if (nvlist_lookup_nvlist_array(vdev, ZPOOL_CONFIG_CHILDREN, &child,
 	    &children) == 0) {
 		for (c = 0; c < children; c++)
-			if (find_vdev_problem(child[c], func))
+			if (find_vdev_problem(child[c], func, ignore_replacing))
 				return (B_TRUE);
 	} else {
 		verify(nvlist_lookup_uint64_array(vdev, ZPOOL_CONFIG_VDEV_STATS,
-		    (uint64_t **)&vs, &c) == 0);
+		    (uint64_t **)&vs, &vsc) == 0);
 
-		if (func(vs->vs_state, vs->vs_aux,
-		    vs->vs_read_errors +
-		    vs->vs_write_errors +
-		    vs->vs_checksum_errors))
+		if (func(vs, vsc) != 0)
 			return (B_TRUE);
 	}
 
@@ -157,7 +168,7 @@ find_vdev_problem(nvlist_t *vdev, int (*func)(uint64_t, uint64_t, uint64_t))
 	if (nvlist_lookup_nvlist_array(vdev, ZPOOL_CONFIG_L2CACHE, &child,
 	    &children) == 0) {
 		for (c = 0; c < children; c++)
-			if (find_vdev_problem(child[c], func))
+			if (find_vdev_problem(child[c], func, ignore_replacing))
 				return (B_TRUE);
 	}
 
@@ -270,15 +281,15 @@ check_status(nvlist_t *config, boolean_t isimport)
 	 * Bad devices in non-replicated config.
 	 */
 	if (vs->vs_state == VDEV_STATE_CANT_OPEN &&
-	    find_vdev_problem(nvroot, vdev_faulted))
+	    find_vdev_problem(nvroot, vdev_faulted, B_TRUE))
 		return (ZPOOL_STATUS_FAULTED_DEV_NR);
 
 	if (vs->vs_state == VDEV_STATE_CANT_OPEN &&
-	    find_vdev_problem(nvroot, vdev_missing))
+	    find_vdev_problem(nvroot, vdev_missing, B_TRUE))
 		return (ZPOOL_STATUS_MISSING_DEV_NR);
 
 	if (vs->vs_state == VDEV_STATE_CANT_OPEN &&
-	    find_vdev_problem(nvroot, vdev_broken))
+	    find_vdev_problem(nvroot, vdev_broken, B_TRUE))
 		return (ZPOOL_STATUS_CORRUPT_LABEL_NR);
 
 	/*
@@ -300,32 +311,38 @@ check_status(nvlist_t *config, boolean_t isimport)
 	/*
 	 * Missing devices in a replicated config.
 	 */
-	if (find_vdev_problem(nvroot, vdev_faulted))
+	if (find_vdev_problem(nvroot, vdev_faulted, B_TRUE))
 		return (ZPOOL_STATUS_FAULTED_DEV_R);
-	if (find_vdev_problem(nvroot, vdev_missing))
+	if (find_vdev_problem(nvroot, vdev_missing, B_TRUE))
 		return (ZPOOL_STATUS_MISSING_DEV_R);
-	if (find_vdev_problem(nvroot, vdev_broken))
+	if (find_vdev_problem(nvroot, vdev_broken, B_TRUE))
 		return (ZPOOL_STATUS_CORRUPT_LABEL_R);
 
 	/*
 	 * Devices with errors
 	 */
-	if (!isimport && find_vdev_problem(nvroot, vdev_errors))
+	if (!isimport && find_vdev_problem(nvroot, vdev_errors, B_TRUE))
 		return (ZPOOL_STATUS_FAILING_DEV);
 
 	/*
 	 * Offlined devices
 	 */
-	if (find_vdev_problem(nvroot, vdev_offlined))
+	if (find_vdev_problem(nvroot, vdev_offlined, B_TRUE))
 		return (ZPOOL_STATUS_OFFLINE_DEV);
 
 	/*
 	 * Removed device
 	 */
-	if (find_vdev_problem(nvroot, vdev_removed))
+	if (find_vdev_problem(nvroot, vdev_removed, B_TRUE))
 		return (ZPOOL_STATUS_REMOVED_DEV);
 
 	/*
+	 * Suboptimal, but usable, ashift configuration.
+	 */
+	if (find_vdev_problem(nvroot, vdev_non_native_ashift, B_FALSE))
+		return (ZPOOL_STATUS_NON_NATIVE_ASHIFT);
+
+	/*
 	 * Outdated, but usable, version
 	 */
 	if (SPA_VERSION_IS_SUPPORTED(version) && version != SPA_VERSION)
diff --git a/cddl/contrib/opensolaris/lib/libzpool/common/kernel.c b/cddl/contrib/opensolaris/lib/libzpool/common/kernel.c
index c5c7b66..71c3dce 100644
--- a/cddl/contrib/opensolaris/lib/libzpool/common/kernel.c
+++ b/cddl/contrib/opensolaris/lib/libzpool/common/kernel.c
@@ -591,6 +591,12 @@ dprintf_setup(int *argc, char **argv)
 		dprintf_print_all = 1;
 }
 
+int
+sysctl_handle_64(SYSCTL_HANDLER_ARGS)
+{
+	return (0);
+}
+
 /*
  * =========================================================================
  * debug printfs
diff --git a/cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h b/cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h
index 2107505..da153c7 100644
--- a/cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h
+++ b/cddl/contrib/opensolaris/lib/libzpool/common/sys/zfs_context.h
@@ -659,11 +659,55 @@ typedef	uint32_t	idmap_rid_t;
 
 #define	SX_SYSINIT(name, lock, desc)
 
+#define SYSCTL_HANDLER_ARGS struct sysctl_oid *oidp, void *arg1,	\
+	intptr_t arg2, struct sysctl_req *req
+
+/*
+ * This describes the access space for a sysctl request.  This is needed
+ * so that we can use the interface from the kernel or from user-space.
+ */
+struct sysctl_req {
+	struct thread	*td;		/* used for access checking */
+	int		lock;		/* wiring state */
+	void		*oldptr;
+	size_t		oldlen;
+	size_t		oldidx;
+	int		(*oldfunc)(struct sysctl_req *, const void *, size_t);
+	void		*newptr;
+	size_t		newlen;
+	size_t		newidx;
+	int		(*newfunc)(struct sysctl_req *, void *, size_t);
+	size_t		validlen;
+	int		flags;
+};
+
+SLIST_HEAD(sysctl_oid_list, sysctl_oid);
+
+/*
+ * This describes one "oid" in the MIB tree.  Potentially more nodes can
+ * be hidden behind it, expanded by the handler.
+ */
+struct sysctl_oid {
+	struct sysctl_oid_list *oid_parent;
+	SLIST_ENTRY(sysctl_oid) oid_link;
+	int		oid_number;
+	u_int		oid_kind;
+	void		*oid_arg1;
+	intptr_t	oid_arg2;
+	const char	*oid_name;
+	int 		(*oid_handler)(SYSCTL_HANDLER_ARGS);
+	const char	*oid_fmt;
+	int		oid_refcnt;
+	u_int		oid_running;
+	const char	*oid_descr;
+};
+
 #define	SYSCTL_DECL(...)
 #define	SYSCTL_NODE(...)
 #define	SYSCTL_INT(...)
 #define	SYSCTL_UINT(...)
 #define	SYSCTL_ULONG(...)
+#define	SYSCTL_PROC(...)
 #define	SYSCTL_QUAD(...)
 #define	SYSCTL_UQUAD(...)
 #ifdef TUNABLE_INT
@@ -675,6 +719,8 @@ typedef	uint32_t	idmap_rid_t;
 #define	TUNABLE_ULONG(...)
 #define	TUNABLE_QUAD(...)
 
+int sysctl_handle_64(SYSCTL_HANDLER_ARGS);
+
 /* Errors */
 
 #ifndef	ERESTART