Enable Linux read-ahead for a single page on ZVOLs

Linux has read-ahead logic designed to accelerate sequential workloads.
ZFS has its own read-ahead logic called zprefetch that operates on both
ZVOLs and datasets. Having two prefetchers active at the same time can
cause overprefetching, which unnecessarily reduces IOPS performance on
CoW filesystems like ZFS.

Testing shows that entirely disabling the Linux prefetch results in
a significant performance penalty for reads while commensurate benefits
are seen in random writes. It appears that read-ahead benefits are
inversely proportional to random write benefits, and so a single page
of Linux-layer read-ahead appears to offer the middle ground for both
workloads.

Reviewed-by: Chunwei Chen <david.chen@osnexus.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <ryao@gentoo.org>
Issue #5902

config/kernel-blk-queue-bdi.m4

@@ -0,0 +1,20 @@
+dnl #
+dnl # 2.6.32 - 4.11, statically allocated bdi in request_queue
+dnl # 4.12 - x.y, dynamically allocated bdi in request_queue
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_BLK_QUEUE_BDI], [
+	AC_MSG_CHECKING([whether blk_queue bdi is dynamic])
+	ZFS_LINUX_TRY_COMPILE([
+		#include <linux/blkdev.h>
+	],[
+		struct request_queue q;
+		struct backing_dev_info bdi;
+		q.backing_dev_info = &bdi;
+	],[
+		AC_MSG_RESULT(yes)
+		AC_DEFINE(HAVE_BLK_QUEUE_BDI_DYNAMIC, 1,
+		    [blk queue backing_dev_info is dynamic])
+	],[
+		AC_MSG_RESULT(no)
+	])
+])

config/kernel.m4

@@ -32,6 +32,7 @@ AC_DEFUN([ZFS_AC_CONFIG_KERNEL], [
 	ZFS_AC_KERNEL_BIO_END_IO_T_ARGS
 	ZFS_AC_KERNEL_BIO_RW_BARRIER
 	ZFS_AC_KERNEL_BIO_RW_DISCARD
+	ZFS_AC_KERNEL_BLK_QUEUE_BDI
 	ZFS_AC_KERNEL_BLK_QUEUE_FLUSH
 	ZFS_AC_KERNEL_BLK_QUEUE_MAX_HW_SECTORS
 	ZFS_AC_KERNEL_BLK_QUEUE_MAX_SEGMENTS

include/linux/blkdev_compat.h

@@ -31,6 +31,7 @@
 
 #include <linux/blkdev.h>
 #include <linux/elevator.h>
+#include <linux/backing-dev.h>
 
 #ifndef HAVE_FMODE_T
 typedef unsigned __bitwise__ fmode_t;
@@ -128,6 +129,16 @@ __blk_queue_max_segments(struct request_queue *q, unsigned short max_segments)
 }
 #endif
 
+static inline void
+blk_queue_set_read_ahead(struct request_queue *q, unsigned long ra_pages)
+{
+#ifdef HAVE_BLK_QUEUE_BDI_DYNAMIC
+	q->backing_dev_info->ra_pages = ra_pages;
+#else
+	q->backing_dev_info.ra_pages = ra_pages;
+#endif
+}
+
 #ifndef HAVE_GET_DISK_RO
 static inline int
 get_disk_ro(struct gendisk *disk)

module/zfs/zvol.c

@@ -1468,6 +1468,9 @@ zvol_alloc(dev_t dev, const char *name)
 	blk_queue_make_request(zv->zv_queue, zvol_request);
 	blk_queue_set_write_cache(zv->zv_queue, B_TRUE, B_TRUE);
 
+	/* Limit read-ahead to a single page to prevent over-prefetching. */
+	blk_queue_set_read_ahead(zv->zv_queue, 1);
+
 	/* Disable write merging in favor of the ZIO pipeline. */
 	queue_flag_set(QUEUE_FLAG_NOMERGES, zv->zv_queue);