Illumos #5244 zio pipeline callers should explicitly invoke next stage

References:
https://www.illumos.org/projects/illumos-gate//issues/5244
https://reviews.csiden.org/r/119/diff/#

Porting Notes:
1. The unported "2932 support crash dumps to raidz, etc. pools" caused a
merge conflict due to a copyright difference in module/zfs/vdev_raidz.c.
2. The unported "4128 disks in zpools never go away when pulled" and
additional Linux-specific changes caused merge conflicts in
module/zfs/vdev_disk.c.
3. Our changes to use the TQ_PUSHPAGE extension and vdev_file_taskq
caused merge conflicts in module/zfs/vdev_file.c. The taskq import that
I plan to send in the future would have prevented this particular
conflict.

Reviewed by: Matthew Ahrens mahrens@delphix.com
Reviewed by: Adam Leventhal ahl@delphix.com
Reviewed by: Alex Reece alex.reece@delphix.com
Reviewed by: Christopher Siden christopher.siden@delphix.com
Ported-by: Richard Yao <ryao@gentoo.org>

include/sys/vdev_impl.h

@@ -61,7 +61,7 @@ typedef int	vdev_open_func_t(vdev_t *vd, uint64_t *size, uint64_t *max_size,
     uint64_t *ashift);
 typedef void	vdev_close_func_t(vdev_t *vd);
 typedef uint64_t vdev_asize_func_t(vdev_t *vd, uint64_t psize);
-typedef int	vdev_io_start_func_t(zio_t *zio);
+typedef void	vdev_io_start_func_t(zio_t *zio);
 typedef void	vdev_io_done_func_t(zio_t *zio);
 typedef void	vdev_state_change_func_t(vdev_t *vd, int, int);
 typedef void	vdev_hold_func_t(vdev_t *vd);

include/sys/zio.h

@@ -153,9 +153,6 @@ typedef enum zio_priority {
 	ZIO_PRIORITY_NOW		/* non-queued i/os (e.g. free) */
 } zio_priority_t;
 
-#define	ZIO_PIPELINE_CONTINUE		0x100
-#define	ZIO_PIPELINE_STOP		0x101
-
 enum zio_flag {
 	/*
 	 * Flags inherited by gang, ddt, and vdev children,

lib/libzpool/taskq.c

@@ -25,6 +25,7 @@
 /*
  * Copyright 2011 Nexenta Systems, Inc. All rights reserved.
  * Copyright 2012 Garrett D'Amore <garrett@damore.org>.  All rights reserved.
+ * Copyright (c) 2014 by Delphix. All rights reserved.
  */
 
 #include <sys/zfs_context.h>
@@ -33,8 +34,10 @@ int taskq_now;
 taskq_t *system_taskq;
 
 #define	TASKQ_ACTIVE	0x00010000
+#define	TASKQ_NAMELEN	31
 
 struct taskq {
+	char		tq_name[TASKQ_NAMELEN + 1];
 	kmutex_t	tq_lock;
 	krwlock_t	tq_threadlock;
 	kcondvar_t	tq_dispatch_cv;
@@ -280,6 +283,7 @@ taskq_create(const char *name, int nthreads, pri_t pri,
 	cv_init(&tq->tq_dispatch_cv, NULL, CV_DEFAULT, NULL);
 	cv_init(&tq->tq_wait_cv, NULL, CV_DEFAULT, NULL);
 	cv_init(&tq->tq_maxalloc_cv, NULL, CV_DEFAULT, NULL);
+	(void) strncpy(tq->tq_name, name, TASKQ_NAMELEN + 1);
 	tq->tq_flags = flags | TASKQ_ACTIVE;
 	tq->tq_active = nthreads;
 	tq->tq_nthreads = nthreads;

module/zfs/vdev_disk.c

@@ -655,7 +655,7 @@ vdev_disk_io_flush(struct block_device *bdev, zio_t *zio)
 	return (0);
 }
 
-static int
+static void
 vdev_disk_io_start(zio_t *zio)
 {
 	vdev_t *v = zio->io_vd;
@@ -667,7 +667,9 @@ vdev_disk_io_start(zio_t *zio)
 
 		if (!vdev_readable(v)) {
 			zio->io_error = SET_ERROR(ENXIO);
-			return (ZIO_PIPELINE_CONTINUE);
+			zio_interrupt(zio);
+			return;
+
 		}
 
 		switch (zio->io_cmd) {
@@ -683,7 +685,7 @@ vdev_disk_io_start(zio_t *zio)
 
 			error = vdev_disk_io_flush(vd->vd_bdev, zio);
 			if (error == 0)
-				return (ZIO_PIPELINE_STOP);
+				return;
 
 			zio->io_error = error;
 			if (error == ENOTSUP)
@@ -695,8 +697,8 @@ vdev_disk_io_start(zio_t *zio)
 			zio->io_error = SET_ERROR(ENOTSUP);
 		}
 
-		return (ZIO_PIPELINE_CONTINUE);
-
+		zio_execute(zio);
+		return;
 	case ZIO_TYPE_WRITE:
 		flags = WRITE;
 		break;
@@ -707,17 +709,17 @@ vdev_disk_io_start(zio_t *zio)
 
 	default:
 		zio->io_error = SET_ERROR(ENOTSUP);
-		return (ZIO_PIPELINE_CONTINUE);
+		zio_interrupt(zio);
+		return;
 	}
 
 	error = __vdev_disk_physio(vd->vd_bdev, zio, zio->io_data,
 	    zio->io_size, zio->io_offset, flags);
 	if (error) {
 		zio->io_error = error;
-		return (ZIO_PIPELINE_CONTINUE);
+		zio_interrupt(zio);
+		return;
 	}
-
-	return (ZIO_PIPELINE_STOP);
 }
 
 static void

module/zfs/vdev_file.c

@@ -20,7 +20,7 @@
  */
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright (c) 2013 by Delphix. All rights reserved.
+ * Copyright (c) 2011, 2014 by Delphix. All rights reserved.
  */
 
 #include <sys/zfs_context.h>
@@ -161,7 +161,7 @@ vdev_file_io_strategy(void *arg)
 	zio_interrupt(zio);
 }
 
-static int
+static void
 vdev_file_io_start(zio_t *zio)
 {
 	vdev_t *vd = zio->io_vd;
@@ -171,7 +171,8 @@ vdev_file_io_start(zio_t *zio)
 		/* XXPOLICY */
 		if (!vdev_readable(vd)) {
 			zio->io_error = SET_ERROR(ENXIO);
-			return (ZIO_PIPELINE_CONTINUE);
+			zio_interrupt(zio);
+			return;
 		}
 
 		switch (zio->io_cmd) {
@@ -187,13 +188,12 @@ vdev_file_io_start(zio_t *zio)
 			zio->io_error = SET_ERROR(ENOTSUP);
 		}
 
-		return (ZIO_PIPELINE_CONTINUE);
+		zio_execute(zio);
+		return;
 	}
 
 	VERIFY3U(taskq_dispatch(vdev_file_taskq, vdev_file_io_strategy, zio,
 	    TQ_PUSHPAGE), !=, 0);
-
-	return (ZIO_PIPELINE_STOP);
 }
 
 /* ARGSUSED */

module/zfs/vdev_mirror.c

@@ -24,7 +24,7 @@
  */
 
 /*
- * Copyright (c) 2013 by Delphix. All rights reserved.
+ * Copyright (c) 2012, 2014 by Delphix. All rights reserved.
  */
 
 #include <sys/zfs_context.h>
@@ -327,7 +327,7 @@ vdev_mirror_child_select(zio_t *zio)
 	return (-1);
 }
 
-static int
+static void
 vdev_mirror_io_start(zio_t *zio)
 {
 	mirror_map_t *mm;
@@ -352,7 +352,8 @@ vdev_mirror_io_start(zio_t *zio)
 				    zio->io_type, zio->io_priority, 0,
 				    vdev_mirror_scrub_done, mc));
 			}
-			return (ZIO_PIPELINE_CONTINUE);
+			zio_execute(zio);
+			return;
 		}
 		/*
 		 * For normal reads just pick one child.
@@ -378,7 +379,7 @@ vdev_mirror_io_start(zio_t *zio)
 		c++;
 	}
 
-	return (ZIO_PIPELINE_CONTINUE);
+	zio_execute(zio);
 }
 
 static int

module/zfs/vdev_missing.c

@@ -24,7 +24,7 @@
  */
 
 /*
- * Copyright (c) 2013 by Delphix. All rights reserved.
+ * Copyright (c) 2012, 2014 by Delphix. All rights reserved.
  */
 
 /*
@@ -66,11 +66,11 @@ vdev_missing_close(vdev_t *vd)
 }
 
 /* ARGSUSED */
-static int
+static void
 vdev_missing_io_start(zio_t *zio)
 {
 	zio->io_error = SET_ERROR(ENOTSUP);
-	return (ZIO_PIPELINE_CONTINUE);
+	zio_execute(zio);
 }
 
 /* ARGSUSED */

module/zfs/vdev_raidz.c

@@ -21,7 +21,7 @@
 
 /*
  * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright (c) 2013 by Delphix. All rights reserved.
+ * Copyright (c) 2012, 2014 by Delphix. All rights reserved.
  */
 
 #include <sys/zfs_context.h>
@@ -1567,7 +1567,7 @@ vdev_raidz_child_done(zio_t *zio)
  *      vdevs have had errors, then create zio read operations to the parity
  *      columns' VDevs as well.
  */
-static int
+static void
 vdev_raidz_io_start(zio_t *zio)
 {
 	vdev_t *vd = zio->io_vd;
@@ -1611,7 +1611,8 @@ vdev_raidz_io_start(zio_t *zio)
 			    ZIO_FLAG_NODATA | ZIO_FLAG_OPTIONAL, NULL, NULL));
 		}
 
-		return (ZIO_PIPELINE_CONTINUE);
+		zio_execute(zio);
+		return;
 	}
 
 	ASSERT(zio->io_type == ZIO_TYPE_READ);
@@ -1651,7 +1652,7 @@ vdev_raidz_io_start(zio_t *zio)
 		}
 	}
 
-	return (ZIO_PIPELINE_CONTINUE);
+	zio_execute(zio);
 }
 
 

module/zfs/zio.c

@@ -61,6 +61,9 @@ kmem_cache_t *zio_data_buf_cache[SPA_MAXBLOCKSIZE >> SPA_MINBLOCKSHIFT];
 int zio_bulk_flags = 0;
 int zio_delay_max = ZIO_DELAY_MAX;
 
+#define	ZIO_PIPELINE_CONTINUE		0x100
+#define	ZIO_PIPELINE_STOP		0x101
+
 /*
  * The following actions directly effect the spa's sync-to-convergence logic.
  * The values below define the sync pass when we start performing the action.
@@ -2600,6 +2603,18 @@ zio_free_zil(spa_t *spa, uint64_t txg, blkptr_t *bp)
  * Read and write to physical devices
  * ==========================================================================
  */
+
+
+/*
+ * Issue an I/O to the underlying vdev. Typically the issue pipeline
+ * stops after this stage and will resume upon I/O completion.
+ * However, there are instances where the vdev layer may need to
+ * continue the pipeline when an I/O was not issued. Since the I/O
+ * that was sent to the vdev layer might be different than the one
+ * currently active in the pipeline (see vdev_queue_io()), we explicitly
+ * force the underlying vdev layers to call either zio_execute() or
+ * zio_interrupt() to ensure that the pipeline continues with the correct I/O.
+ */
 static int
 zio_vdev_io_start(zio_t *zio)
 {
@@ -2617,15 +2632,16 @@ zio_vdev_io_start(zio_t *zio)
 		/*
 		 * The mirror_ops handle multiple DVAs in a single BP.
 		 */
-		return (vdev_mirror_ops.vdev_op_io_start(zio));
+		vdev_mirror_ops.vdev_op_io_start(zio);
+		return (ZIO_PIPELINE_STOP);
 	}
 
 	/*
 	 * We keep track of time-sensitive I/Os so that the scan thread
 	 * can quickly react to certain workloads.  In particular, we care
 	 * about non-scrubbing, top-level reads and writes with the following
 	 * characteristics:
-	 * 	- synchronous writes of user data to non-slog devices
+	 *	- synchronous writes of user data to non-slog devices
 	 *	- any reads of user data
 	 * When these conditions are met, adjust the timestamp of spa_last_io
 	 * which allows the scan thread to adjust its workload accordingly.
@@ -2695,7 +2711,8 @@ zio_vdev_io_start(zio_t *zio)
 		}
 	}
 
-	return (vd->vdev_ops->vdev_op_io_start(zio));
+	vd->vdev_ops->vdev_op_io_start(zio);
+	return (ZIO_PIPELINE_STOP);
 }
 
 static int