Illumos 4970-4974 - extreme rewind enhancements

4970 need controls on i/o issued by zpool import -XF
4971 zpool import -T should accept hex values
4972 zpool import -T implies extreme rewind, and thus a scrub
4973 spa_load_retry retries the same txg
4974 spa_load_verify() reads all data twice
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: Dan McDonald <danmcd@omniti.com>
Reviewed by: George Wilson <george.wilson@delphix.com>
Approved by: Robert Mustacchi <rm@joyent.com>

References:
  https://www.illumos.org/issues/4970
  https://www.illumos.org/issues/4971
  https://www.illumos.org/issues/4972
  https://www.illumos.org/issues/4973
  https://www.illumos.org/issues/4974
  illumos/illumos-gate@e42d205

Notes:
    This set of patches adds a set of tunable parameters for the
    "extreme rewind" mode of pool import which allows control over
    the traversal performed during such an import.

Ported by: Tim Chase <tim@chase2k.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes openzfs#2598

cmd/zpool/zpool_main.c

@@ -22,7 +22,7 @@
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
  * Copyright 2011 Nexenta Systems, Inc. All rights reserved.
- * Copyright (c) 2013 by Delphix. All rights reserved.
+ * Copyright (c) 2011, 2014 by Delphix. All rights reserved.
  * Copyright (c) 2012 by Frederik Wessels. All rights reserved.
  * Copyright (c) 2012 by Cyril Plisko. All rights reserved.
  */
@@ -2082,7 +2082,7 @@ zpool_do_import(int argc, char **argv)
 
 		case 'T':
 			errno = 0;
-			txg = strtoull(optarg, &endptr, 10);
+			txg = strtoull(optarg, &endptr, 0);
 			if (errno != 0 || *endptr != '\0') {
 				(void) fprintf(stderr,
 				    gettext("invalid txg value\n"));

man/man5/zfs-module-parameters.5

@@ -230,6 +230,52 @@ they operate close to quota or capacity limits.
 Default value: 24
 .RE
 
+.sp
+.ne 2
+.na
+\fBspa_load_verify_data\fR (int)
+.ad
+.RS 12n
+Whether to traverse data blocks during an "extreme rewind" (\fB-X\fR)
+import.  Use 0 to disable and 1 to enable.
+
+An extreme rewind import normally performs a full traversal of all
+blocks in the pool for verification.  If this parameter is set to 0,
+the traversal skips non-metadata blocks.  It can be toggled once the
+import has started to stop or start the traversal of non-metadata blocks.
+.sp
+Default value: 1
+.RE
+
+.sp
+.ne 2
+.na
+\fBspa_load_verify_metadata\fR (int)
+.ad
+.RS 12n
+Whether to traverse blocks during an "extreme rewind" (\fB-X\fR)
+pool import.  Use 0 to disable and 1 to enable.
+
+An extreme rewind import normally performs a full traversal of all
+blocks in the pool for verification.  If this parameter is set to 1,
+the traversal is not performed.  It can be toggled once the import has
+started to stop or start the traversal.
+.sp
+Default value: 1
+.RE
+
+.sp
+.ne 2
+.na
+\fBspa_load_verify_maxinflight\fR (int)
+.ad
+.RS 12n
+Maximum concurrent I/Os during the traversal performed during an "extreme
+rewind" (\fB-X\fR) pool import.
+.sp
+Default value: 10000
+.RE
+
 .sp
 .ne 2
 .na

module/zfs/spa.c

@@ -1855,6 +1855,7 @@ spa_load_verify_done(zio_t *zio)
 	spa_load_error_t *sle = zio->io_private;
 	dmu_object_type_t type = BP_GET_TYPE(bp);
 	int error = zio->io_error;
+	spa_t *spa = zio->io_spa;
 
 	if (error) {
 		if ((BP_GET_LEVEL(bp) != 0 || DMU_OT_IS_METADATA(type)) &&
@@ -1864,23 +1865,56 @@ spa_load_verify_done(zio_t *zio)
 			atomic_add_64(&sle->sle_data_count, 1);
 	}
 	zio_data_buf_free(zio->io_data, zio->io_size);
+
+	mutex_enter(&spa->spa_scrub_lock);
+	spa->spa_scrub_inflight--;
+	cv_broadcast(&spa->spa_scrub_io_cv);
+	mutex_exit(&spa->spa_scrub_lock);
 }
 
+/*
+ * Maximum number of concurrent scrub i/os to create while verifying
+ * a pool while importing it.
+ */
+int spa_load_verify_maxinflight = 10000;
+int spa_load_verify_metadata = B_TRUE;
+int spa_load_verify_data = B_TRUE;
+
 /*ARGSUSED*/
 static int
 spa_load_verify_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
     const zbookmark_phys_t *zb, const dnode_phys_t *dnp, void *arg)
 {
-	if (!BP_IS_HOLE(bp) && !BP_IS_EMBEDDED(bp)) {
-		zio_t *rio = arg;
-		size_t size = BP_GET_PSIZE(bp);
-		void *data = zio_data_buf_alloc(size);
+	zio_t *rio;
+	size_t size;
+	void *data;
 
-		zio_nowait(zio_read(rio, spa, bp, data, size,
-		    spa_load_verify_done, rio->io_private, ZIO_PRIORITY_SCRUB,
-		    ZIO_FLAG_SPECULATIVE | ZIO_FLAG_CANFAIL |
-		    ZIO_FLAG_SCRUB | ZIO_FLAG_RAW, zb));
-	}
+	if (BP_IS_HOLE(bp) || BP_IS_EMBEDDED(bp))
+		return (0);
+	/*
+	 * Note: normally this routine will not be called if
+	 * spa_load_verify_metadata is not set.  However, it may be useful
+	 * to manually set the flag after the traversal has begun.
+	 */
+	if (!spa_load_verify_metadata)
+		return (0);
+	if (BP_GET_BUFC_TYPE(bp) == ARC_BUFC_DATA && !spa_load_verify_data)
+		return (0);
+
+	rio = arg;
+	size = BP_GET_PSIZE(bp);
+	data = zio_data_buf_alloc(size);
+
+	mutex_enter(&spa->spa_scrub_lock);
+	while (spa->spa_scrub_inflight >= spa_load_verify_maxinflight)
+		cv_wait(&spa->spa_scrub_io_cv, &spa->spa_scrub_lock);
+	spa->spa_scrub_inflight++;
+	mutex_exit(&spa->spa_scrub_lock);
+
+	zio_nowait(zio_read(rio, spa, bp, data, size,
+	    spa_load_verify_done, rio->io_private, ZIO_PRIORITY_SCRUB,
+	    ZIO_FLAG_SPECULATIVE | ZIO_FLAG_CANFAIL |
+	    ZIO_FLAG_SCRUB | ZIO_FLAG_RAW, zb));
 	return (0);
 }
 
@@ -1891,7 +1925,7 @@ spa_load_verify(spa_t *spa)
 	spa_load_error_t sle = { 0 };
 	zpool_rewind_policy_t policy;
 	boolean_t verify_ok = B_FALSE;
-	int error;
+	int error = 0;
 
 	zpool_get_rewind_policy(spa->spa_config, &policy);
 
@@ -1901,8 +1935,11 @@ spa_load_verify(spa_t *spa)
 	rio = zio_root(spa, NULL, &sle,
 	    ZIO_FLAG_CANFAIL | ZIO_FLAG_SPECULATIVE);
 
-	error = traverse_pool(spa, spa->spa_verify_min_txg,
-	    TRAVERSE_PRE | TRAVERSE_PREFETCH, spa_load_verify_cb, rio);
+	if (spa_load_verify_metadata) {
+		error = traverse_pool(spa, spa->spa_verify_min_txg,
+		    TRAVERSE_PRE | TRAVERSE_PREFETCH_METADATA,
+		    spa_load_verify_cb, rio);
+	}
 
 	(void) zio_wait(rio);
 
@@ -2781,7 +2818,7 @@ spa_load_retry(spa_t *spa, spa_load_state_t state, int mosconfig)
 	spa_unload(spa);
 	spa_deactivate(spa);
 
-	spa->spa_load_max_txg--;
+	spa->spa_load_max_txg = spa->spa_uberblock.ub_txg - 1;
 
 	spa_activate(spa, mode);
 	spa_async_suspend(spa);
@@ -2811,6 +2848,8 @@ spa_load_best(spa_t *spa, spa_load_state_t state, int mosconfig,
 		spa_set_log_state(spa, SPA_LOG_CLEAR);
 	} else {
 		spa->spa_load_max_txg = max_request;
+		if (max_request != UINT64_MAX)
+			spa->spa_extreme_rewind = B_TRUE;
 	}
 
 	load_error = rewind_error = spa_load(spa, state, SPA_IMPORT_EXISTING,
@@ -6603,3 +6642,17 @@ EXPORT_SYMBOL(spa_prop_clear_bootfs);
 /* asynchronous event notification */
 EXPORT_SYMBOL(spa_event_notify);
 #endif
+
+#if defined(_KERNEL) && defined(HAVE_SPL)
+module_param(spa_load_verify_maxinflight, int, 0644);
+MODULE_PARM_DESC(spa_load_verify_maxinflight,
+	"Max concurrent traversal I/Os while verifying pool during import -X");
+
+module_param(spa_load_verify_metadata, int, 0644);
+MODULE_PARM_DESC(spa_load_verify_metadata,
+	"Set to traverse metadata on pool import");
+
+module_param(spa_load_verify_data, int, 0644);
+MODULE_PARM_DESC(spa_load_verify_data,
+	"Set to traverse data on pool import");
+#endif