Direct IO support

Direct IO via the O_DIRECT flag was originally introduced in XFS by
IRIX for database workloads. Its purpose was to allow the database
to bypass the page and buffer caches to prevent unnecessary IO
operations (e.g.  readahead) while preventing contention for system
memory between the database and kernel caches.

On Illumos, there is a library function called directio(3C) that
allows user space to provide a hint to the file system that Direct IO
is useful, but the file system is free to ignore it. The semantics
are also entirely a file system decision. Those that do not
implement it return ENOTTY.

Since the semantics were never defined in any standard, O_DIRECT is
implemented such that it conforms to the behavior described in the
Linux open(2) man page as follows.

    1.  Minimize cache effects of the I/O.

    By design the ARC is already scan-resistant which helps mitigate
    the need for special O_DIRECT handling.  Data which is only
    accessed once will be the first to be evicted from the cache.
    This behavior is in consistent with Illumos and FreeBSD.

    Future performance work may wish to investigate the benefits of
    immediately evicting data from the cache which has been read or
    written with the O_DIRECT flag.  Functionally this behavior is
    very similar to applying the 'primarycache=metadata' property
    per open file.

    2. O_DIRECT _MAY_ impose restrictions on IO alignment and length.

    No additional alignment or length restrictions are imposed.

    3. O_DIRECT _MAY_ perform unbuffered IO operations directly
       between user memory and block device.

    No unbuffered IO operations are currently supported.  In order
    to support features such as transparent compression, encryption,
    and checksumming a copy must be made to transform the data.

    4. O_DIRECT _MAY_ imply O_DSYNC (XFS).

    O_DIRECT does not imply O_DSYNC for ZFS.  Callers must provide
    O_DSYNC to request synchronous semantics.

    5. O_DIRECT _MAY_ disable file locking that serializes IO
       operations.  Applications should avoid mixing O_DIRECT
       and normal IO or mmap(2) IO to the same file.  This is
       particularly true for overlapping regions.

    All I/O in ZFS is locked for correctness and this locking is not
    disabled by O_DIRECT.  However, concurrently mixing O_DIRECT,
    mmap(2), and normal I/O on the same file is not recommended.

This change is implemented by layering the aops->direct_IO operations
on the existing AIO operations.  Code already existed in ZFS on Linux
for bypassing the page cache when O_DIRECT is specified.

References:
  * http://xfs.org/docs/xfsdocs-xml-dev/XFS_User_Guide/tmp/en-US/html/ch02s09.html
  * https://blogs.oracle.com/roch/entry/zfs_and_directio
  * https://ext4.wiki.kernel.org/index.php/Clarifying_Direct_IO's_Semantics
  * https://illumos.org/man/3c/directio

Signed-off-by: Richard Yao <ryao@gentoo.org>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes openzfs#224

config/kernel-vfs-direct_IO.m4

@@ -0,0 +1,130 @@
+dnl #
+dnl # Linux 4.6.x API change
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_VFS_DIRECT_IO_ITER], [
+	AC_MSG_CHECKING([whether aops->direct_IO() uses iov_iter])
+	ZFS_LINUX_TRY_COMPILE([
+		#include <linux/fs.h>
+
+		ssize_t test_direct_IO(struct kiocb *kiocb,
+		    struct iov_iter *iter) { return 0; }
+
+		static const struct address_space_operations
+		    aops __attribute__ ((unused)) = {
+			.direct_IO = test_direct_IO,
+		};
+	],[
+	],[
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_DIRECT_IO_ITER, 1,
+		    [aops->direct_IO() uses iov_iter without rw])
+		zfs_ac_direct_io="yes"
+	],[
+		AC_MSG_RESULT([no])
+	])
+])
+
+dnl #
+dnl # Linux 4.1.x API change
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_VFS_DIRECT_IO_ITER_OFFSET], [
+	AC_MSG_CHECKING(
+	    [whether aops->direct_IO() uses iov_iter with offset])
+	ZFS_LINUX_TRY_COMPILE([
+		#include <linux/fs.h>
+
+		ssize_t test_direct_IO(struct kiocb *kiocb,
+		    struct iov_iter *iter, loff_t offset) { return 0; }
+
+		static const struct address_space_operations
+		    aops __attribute__ ((unused)) = {
+			.direct_IO = test_direct_IO,
+		};
+	],[
+	],[
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_DIRECT_IO_ITER_OFFSET, 1,
+		    [aops->direct_IO() uses iov_iter with offset])
+		zfs_ac_direct_io="yes"
+	],[
+		AC_MSG_RESULT([no])
+	])
+])
+
+dnl #
+dnl # Linux 3.16.x API change
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_VFS_DIRECT_IO_ITER_RW_OFFSET], [
+	AC_MSG_CHECKING(
+	    [whether aops->direct_IO() uses iov_iter with rw and offset])
+	ZFS_LINUX_TRY_COMPILE([
+		#include <linux/fs.h>
+
+		ssize_t test_direct_IO(int rw, struct kiocb *kiocb,
+		    struct iov_iter *iter, loff_t offset) { return 0; }
+
+		static const struct address_space_operations
+		    aops __attribute__ ((unused)) = {
+		    .direct_IO = test_direct_IO,
+		};
+	],[
+	],[
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_DIRECT_IO_ITER_RW_OFFSET, 1,
+		    [aops->direct_IO() uses iov_iter with rw and offset])
+		zfs_ac_direct_io="yes"
+	],[
+		AC_MSG_RESULT([no])
+	])
+])
+
+dnl #
+dnl # Ancient Linux API (predates git)
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_VFS_DIRECT_IO_IOVEC], [
+	AC_MSG_CHECKING([whether aops->direct_IO() uses iovec])
+	ZFS_LINUX_TRY_COMPILE([
+		#include <linux/fs.h>
+
+		ssize_t test_direct_IO(int rw, struct kiocb *kiocb,
+		    const struct iovec *iov, loff_t offset,
+		    unsigned long nr_segs) { return 0; }
+
+		static const struct address_space_operations
+		    aops __attribute__ ((unused)) = {
+		    .direct_IO = test_direct_IO,
+		};
+	],[
+	],[
+		AC_MSG_RESULT([yes])
+		AC_DEFINE(HAVE_VFS_DIRECT_IO_IOVEC, 1,
+		    [aops->direct_IO() uses iovec])
+		zfs_ac_direct_io="yes"
+	],[
+		AC_MSG_RESULT([no])
+	])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_VFS_DIRECT_IO], [
+	zfs_ac_direct_io="no"
+
+	if test "$zfs_ac_direct_io" = "no"; then
+		ZFS_AC_KERNEL_VFS_DIRECT_IO_ITER
+	fi
+
+	if test "$zfs_ac_direct_io" = "no"; then
+		ZFS_AC_KERNEL_VFS_DIRECT_IO_ITER_OFFSET
+	fi
+
+	if test "$zfs_ac_direct_io" = "no"; then
+		ZFS_AC_KERNEL_VFS_DIRECT_IO_ITER_RW_OFFSET
+	fi
+
+	if test "$zfs_ac_direct_io" = "no"; then
+		ZFS_AC_KERNEL_VFS_DIRECT_IO_IOVEC
+	fi
+
+	if test "$zfs_ac_direct_io" = "no"; then
+		AC_MSG_ERROR([no; unknown direct IO interface])
+	fi
+])

config/kernel.m4

@@ -144,6 +144,7 @@ AC_DEFUN([ZFS_AC_CONFIG_KERNEL], [
 	ZFS_AC_KERNEL_LSEEK_EXECUTE
 	ZFS_AC_KERNEL_VFS_ITERATE
 	ZFS_AC_KERNEL_VFS_RW_ITERATE
+	ZFS_AC_KERNEL_VFS_DIRECT_IO
 	ZFS_AC_KERNEL_GENERIC_WRITE_CHECKS
 	ZFS_AC_KERNEL_KMAP_ATOMIC_ARGS
 	ZFS_AC_KERNEL_FOLLOW_DOWN_ONE

configure.ac

@@ -280,6 +280,7 @@ AC_CONFIG_FILES([
 	tests/zfs-tests/tests/functional/hkdf/Makefile
 	tests/zfs-tests/tests/functional/inheritance/Makefile
 	tests/zfs-tests/tests/functional/inuse/Makefile
+	tests/zfs-tests/tests/functional/io/Makefile
 	tests/zfs-tests/tests/functional/kstat/Makefile
 	tests/zfs-tests/tests/functional/large_files/Makefile
 	tests/zfs-tests/tests/functional/largest_pool/Makefile

module/zfs/zpl_file.c

@@ -438,6 +438,58 @@ zpl_aio_write(struct kiocb *kiocb, const struct iovec *iovp,
 }
 #endif /* HAVE_VFS_RW_ITERATE */
 
+#if defined(HAVE_VFS_RW_ITERATE)
+static ssize_t
+zpl_direct_IO_impl(struct kiocb *kiocb, struct iov_iter *iter)
+{
+	if (iov_iter_rw(iter) == WRITE)
+		return (zpl_iter_write(kiocb, iter));
+	else
+		return (zpl_iter_read(kiocb, iter));
+}
+#if defined(HAVE_VFS_DIRECT_IO_ITER)
+static ssize_t
+zpl_direct_IO(struct kiocb *kiocb, struct iov_iter *iter)
+{
+	return (zpl_direct_IO_impl(kiocb, iter));
+}
+#elif defined(HAVE_VFS_DIRECT_IO_ITER_OFFSET)
+static ssize_t
+zpl_direct_IO(struct kiocb *kiocb, struct iov_iter *iter, loff_t pos)
+{
+	ASSERT(pos, ==, kiocb->ki_pos);
+	return (zpl_direct_IO_impl(kiocb, iter));
+}
+#elif defined(HAVE_VFS_DIRECT_IO_ITER_RW_OFFSET)
+static ssize_t
+zpl_direct_IO(int rw, struct kiocb *kiocb, struct iov_iter *iter, loff_t pos)
+{
+	ASSERT(rw, ==, iov_iter_rw(iter));
+	ASSERT(pos, ==, kiocb->ki_pos);
+	return (zpl_direct_IO_impl(kiocb, iter));
+}
+#else
+#error "Unknown direct IO interface"
+#endif
+
+#else
+
+#if defined(HAVE_VFS_DIRECT_IO_IOVEC)
+static ssize_t
+zpl_direct_IO(int rw, struct kiocb *kiocb, const struct iovec *iovp,
+    loff_t pos, unsigned long nr_segs)
+{
+	if (rw == WRITE)
+		return (zpl_aio_write(kiocb, iovp, nr_segs, pos));
+	else
+		return (zpl_aio_read(kiocb, iovp, nr_segs, pos));
+}
+#else
+#error "Unknown direct IO interface"
+#endif
+
+#endif /* HAVE_VFS_RW_ITERATE */
+
 static loff_t
 zpl_llseek(struct file *filp, loff_t offset, int whence)
 {
@@ -929,6 +981,7 @@ const struct address_space_operations zpl_address_space_operations = {
 	.readpage	= zpl_readpage,
 	.writepage	= zpl_writepage,
 	.writepages	= zpl_writepages,
+	.direct_IO	= zpl_direct_IO,
 };
 
 const struct file_operations zpl_file_operations = {

scripts/commitcheck.sh

@@ -16,10 +16,11 @@ function test_url()
 }
 
 # test commit body for length
+# lines containing urls are exempt for the length limit.
 function test_commit_bodylength()
 {
     length="72"
-    body=$(git log -n 1 --pretty=%b "$REF" | grep -E -m 1 ".{$((length + 1))}")
+    body=$(git log -n 1 --pretty=%b "$REF" | grep -Ev "http(s)*://" | grep -E -m 1 ".{$((length + 1))}")
     if [ -n "$body" ]; then
         echo "error: commit message body contains line over ${length} characters"
         return 1

tests/runfiles/linux.run

@@ -558,6 +558,10 @@ tests = ['inherit_001_pos']
 pre =
 tags = ['functional', 'inheritance']
 
+[tests/functional/io]
+tests = ['sync', 'psync', 'libaio', 'posixaio', 'mmap']
+tags = ['functional', 'io']
+
 [tests/functional/inuse]
 tests = ['inuse_001_pos', 'inuse_003_pos', 'inuse_004_pos',
     'inuse_005_pos', 'inuse_006_pos', 'inuse_007_pos', 'inuse_008_pos',

tests/zfs-tests/tests/functional/Makefile.am

@@ -27,6 +27,7 @@ SUBDIRS = \
 	hkdf \
 	inheritance \
 	inuse \
+	io \
 	kstat \
 	large_files \
 	largest_pool \

tests/zfs-tests/tests/functional/io/Makefile.am

@@ -0,0 +1,12 @@
+pkgdatadir = $(datadir)/@PACKAGE@/zfs-tests/tests/functional/io
+dist_pkgdata_SCRIPTS = \
+	setup.ksh \
+	cleanup.ksh \
+	sync.ksh \
+	psync.ksh \
+	libaio.ksh \
+	posixaio.ksh \
+	mmap.ksh
+
+dist_pkgdata_DATA = \
+	io.cfg

tests/zfs-tests/tests/functional/io/cleanup.ksh

@@ -0,0 +1,31 @@
+#!/bin/ksh -p
+#
+# CDDL HEADER START
+#
+# The contents of this file are subject to the terms of the
+# Common Development and Distribution License (the "License").
+# You may not use this file except in compliance with the License.
+#
+# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+# or http://www.opensolaris.org/os/licensing.
+# See the License for the specific language governing permissions
+# and limitations under the License.
+#
+# When distributing Covered Code, include this CDDL HEADER in each
+# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+# If applicable, add the following below this CDDL HEADER, with the
+# fields enclosed by brackets "[]" replaced with your own identifying
+# information: Portions Copyright [yyyy] [name of copyright owner]
+#
+# CDDL HEADER END
+#
+
+#
+# Copyright (c) 2018 by Lawrence Livermore National Security, LLC.
+#
+
+. $STF_SUITE/include/libtest.shlib
+
+verify_runnable "global"
+
+default_cleanup

tests/zfs-tests/tests/functional/io/io.cfg

@@ -0,0 +1,25 @@
+#
+# CDDL HEADER START
+#
+# This file and its contents are supplied under the terms of the
+# Common Development and Distribution License ("CDDL"), version 1.0.
+# You may only use this file in accordance with the terms of version
+# 1.0 of the CDDL.
+#
+# A full copy of the text of the CDDL should have accompanied this
+# source.  A copy of the CDDL is also available via the Internet at
+# http://www.illumos.org/license/CDDL.
+#
+# CDDL HEADER END
+#
+
+#
+# Copyright (c) 2018 by Lawrence Livermore National Security, LLC.
+#
+
+FIO_COMMON_ARGS="--numjobs=8 --bs=128k --size=16M --fallocate=none --group_reporting --verify=sha1 --minimal"
+
+FIO_READ_ARGS="--name=rw --rw=read $FIO_COMMON_ARGS"
+FIO_WRITE_ARGS="--name=rw --rw=write $FIO_COMMON_ARGS"
+FIO_RANDREAD_ARGS="--name=rw --rw=randread $FIO_COMMON_ARGS"
+FIO_RANDWRITE_ARGS="--name=rw --rw=randwrite $FIO_COMMON_ARGS"

tests/zfs-tests/tests/functional/io/libaio.ksh

@@ -0,0 +1,65 @@
+#! /bin/ksh -p
+#
+# CDDL HEADER START
+#
+# The contents of this file are subject to the terms of the
+# Common Development and Distribution License (the "License").
+# You may not use this file except in compliance with the License.
+#
+# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+# or http://www.opensolaris.org/os/licensing.
+# See the License for the specific language governing permissions
+# and limitations under the License.
+#
+# When distributing Covered Code, include this CDDL HEADER in each
+# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+# If applicable, add the following below this CDDL HEADER, with the
+# fields enclosed by brackets "[]" replaced with your own identifying
+# information: Portions Copyright [yyyy] [name of copyright owner]
+#
+# CDDL HEADER END
+#
+
+#
+# Copyright (c) 2018 by Lawrence Livermore National Security, LLC.
+#
+
+. $STF_SUITE/include/libtest.shlib
+. $STF_SUITE/tests/functional/io/io.cfg
+
+#
+# DESCRIPTION:
+#	Verify Linux native asynchronous IO.
+#
+# STRATEGY:
+#	1. Use fio(1) in verify mode to perform write, read,
+#	   random read, and random write workloads.
+#	2. Repeat the test with additional fio(1) options.
+#
+
+verify_runnable "global"
+
+function cleanup
+{
+	log_must rm -f "$mntpnt/rw*"
+}
+
+log_assert "Verify Linux native asynchronous IO"
+
+log_onexit cleanup
+
+ioengine="--ioengine=libaio"
+mntpnt=$(get_prop mountpoint $TESTPOOL/$TESTFS)
+dir="--directory=$mntpnt"
+
+set -A fio_arg -- "--sync=0" "--sync=1" "--direct=0" "--direct=1"
+
+for arg in "${fio_arg[@]}"; do
+	log_must fio $dir $ioengine $arg $FIO_WRITE_ARGS
+	log_must fio $dir $ioengine $arg $FIO_READ_ARGS
+	log_must fio $dir $ioengine $arg $FIO_RANDWRITE_ARGS
+	log_must fio $dir $ioengine $arg $FIO_RANDREAD_ARGS
+	log_must rm -f "$mntpnt/rw*"
+done
+
+log_pass "Verified Linux native asynchronous IO"