Backport important fixes from upstream HEAD; also includes important zvol initialization fixes (by me) that were sent upstream

(Portage version: 2.2.0_alpha176/cvs/Linux x86_64, signed Manifest commit with key 0xBEE84C64)

7 files changed, 457 insertions, 1 deletions

diff --git a/sys-fs/zfs-kmod/ChangeLog b/sys-fs/zfs-kmod/ChangeLog
index ca52e7680a87..191258a8d4e8 100644
--- a/sys-fs/zfs-kmod/ChangeLog
+++ b/sys-fs/zfs-kmod/ChangeLog
@@ -1,6 +1,18 @@
 # ChangeLog for sys-fs/zfs-kmod
 # Copyright 1999-2013 Gentoo Foundation; Distributed under the GPL v2
-# $Header: /var/cvsroot/gentoo-x86/sys-fs/zfs-kmod/ChangeLog,v 1.29 2013/04/17 14:30:18 ryao Exp $
+# $Header: /var/cvsroot/gentoo-x86/sys-fs/zfs-kmod/ChangeLog,v 1.30 2013/05/28 13:08:52 ryao Exp $
+
+*zfs-kmod-0.6.1-r1 (28 May 2013)
+
+  28 May 2013; Richard Yao <ryao@gentoo.org>
+  +files/zfs-kmod-0.6.1-fix-getdents.patch,
+  +files/zfs-kmod-0.6.1-fix-txg_quiesce-deadlock.patch,
+  +files/zfs-kmod-0.6.1-fix-xattr-behavior-1.patch,
+  +files/zfs-kmod-0.6.1-fix-xattr-behavior-2.patch,
+  +files/zfs-kmod-0.6.1-fix-zvol-initialization.patch,
+  +zfs-kmod-0.6.1-r1.ebuild:
+  Backport important fixes from upstream HEAD; also includes important zvol
+  initialization fixes (by me) that were sent upstream
 
   17 Apr 2013; Richard Yao <ryao@gentoo.org> zfs-kmod-9999.ebuild:
   Remove /usr/src/zfs symlink from 9999 ebuild; sys-cluster/lustre (science
diff --git a/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-getdents.patch b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-getdents.patch
new file mode 100644
index 000000000000..ce7d5c0494d9
--- /dev/null
+++ b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-getdents.patch
@@ -0,0 +1,32 @@
+From 8f1e11b6105bf46a4258958eb3dcae2ab21ed8b2 Mon Sep 17 00:00:00 2001
+From: Caleb James DeLisle <calebdelisle@lavabit.com>
+Date: Tue, 16 Apr 2013 05:23:39 -0400
+Subject: [PATCH] Remove .readdir from zpl_file_operations table
+
+The zpl_readdir() function shouldn't be registered as part of
+the zpl_file_operations table, it must only be part of the
+zpl_dir_file_operations table.  By removing this callback
+the VFS will now correctly return ENOTDIR when calling
+getdents() on a file.
+
+Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
+Closes #1404
+---
+ module/zfs/zpl_file.c | 1 -
+ 1 file changed, 1 deletion(-)
+
+diff --git a/module/zfs/zpl_file.c b/module/zfs/zpl_file.c
+index 9c27b7f..db6a72c 100644
+--- a/module/zfs/zpl_file.c
++++ b/module/zfs/zpl_file.c
+@@ -446,7 +446,6 @@
+ 	.llseek		= generic_file_llseek,
+ 	.read		= zpl_read,
+ 	.write		= zpl_write,
+-	.readdir	= zpl_readdir,
+ 	.mmap		= zpl_mmap,
+ 	.fsync		= zpl_fsync,
+ #ifdef HAVE_FILE_FALLOCATE
+-- 
+1.8.1.6
+
diff --git a/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-txg_quiesce-deadlock.patch b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-txg_quiesce-deadlock.patch
new file mode 100644
index 000000000000..8513e78882b1
--- /dev/null
+++ b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-txg_quiesce-deadlock.patch
@@ -0,0 +1,73 @@
+From 57f5a2008e2e6acf58934cf43c5fdca0faffa73e Mon Sep 17 00:00:00 2001
+From: Brian Behlendorf <behlendorf1@llnl.gov>
+Date: Thu, 25 Apr 2013 16:29:22 -0700
+Subject: [PATCH] Fix txg_quiesce thread deadlock
+
+A deadlock was accidentally introduced by commit e95853a which
+can occur when the system is under memory pressure.  What happens
+is that while the txg_quiesce thread is holding the tx->tx_cpu
+locks it enters memory reclaim.  In the context of this memory
+reclaim it then issues synchronous I/O to a ZVOL swap device.
+Because the txg_quiesce thread is holding the tx->tx_cpu locks
+a new txg cannot be opened to handle the I/O.  Deadlock.
+
+The fix is straight forward.  Move the memory allocation outside
+the critical region where the tx->tx_cpu locks are held.  And for
+good measure change the offending allocation to KM_PUSHPAGE to
+ensure it never attempts to issue I/O during reclaim.
+
+Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
+Issue #1274
+---
+ module/zfs/dsl_pool.c |  2 +-
+ module/zfs/txg.c      | 14 +++++++-------
+ 2 files changed, 8 insertions(+), 8 deletions(-)
+
+diff --git a/module/zfs/dsl_pool.c b/module/zfs/dsl_pool.c
+index 704f034..771b265 100644
+--- a/module/zfs/dsl_pool.c
++++ b/module/zfs/dsl_pool.c
+@@ -143,7 +143,7 @@
+ {
+ 	txg_history_t *th, *rm;
+ 
+-	th = kmem_zalloc(sizeof(txg_history_t), KM_SLEEP);
++	th = kmem_zalloc(sizeof(txg_history_t), KM_PUSHPAGE);
+ 	mutex_init(&th->th_lock, NULL, MUTEX_DEFAULT, NULL);
+ 	th->th_kstat.txg = txg;
+ 	th->th_kstat.state = TXG_STATE_OPEN;
+diff --git a/module/zfs/txg.c b/module/zfs/txg.c
+index c7c3df3..7c820af 100644
+--- a/module/zfs/txg.c
++++ b/module/zfs/txg.c
+@@ -367,6 +367,13 @@
+ 	tx->tx_open_txg++;
+ 
+ 	/*
++	 * Now that we've incremented tx_open_txg, we can let threads
++	 * enter the next transaction group.
++	 */
++	for (c = 0; c < max_ncpus; c++)
++		mutex_exit(&tx->tx_cpu[c].tc_lock);
++
++	/*
+ 	 * Measure how long the txg was open and replace the kstat.
+ 	 */
+ 	th = dsl_pool_txg_history_get(dp, txg);
+@@ -376,13 +383,6 @@
+ 	dsl_pool_txg_history_add(dp, tx->tx_open_txg);
+ 
+ 	/*
+-	 * Now that we've incremented tx_open_txg, we can let threads
+-	 * enter the next transaction group.
+-	 */
+-	for (c = 0; c < max_ncpus; c++)
+-		mutex_exit(&tx->tx_cpu[c].tc_lock);
+-
+-	/*
+ 	 * Quiesce the transaction group by waiting for everyone to txg_exit().
+ 	 */
+ 	start = gethrtime();
+-- 
+1.8.1.6
+
diff --git a/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-xattr-behavior-1.patch b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-xattr-behavior-1.patch
new file mode 100644
index 000000000000..b4801708ae35
--- /dev/null
+++ b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-xattr-behavior-1.patch
@@ -0,0 +1,50 @@
+From f706421173c571371afff5e2a2ee0784c5e3f95d Mon Sep 17 00:00:00 2001
+From: Brian Behlendorf <behlendorf1@llnl.gov>
+Date: Wed, 17 Apr 2013 13:07:36 -0700
+Subject: [PATCH] Correctly return ERANGE in getxattr(2)
+
+According to the getxattr(2) man page the ERANGE errno should be
+returned when the size of the value buffer is to small to hold the
+result.  Prior to this patch the implementation would just truncate
+the value to size bytes.
+
+Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
+Closes #1408
+---
+ module/zfs/zpl_xattr.c | 12 ++++++++++--
+ 1 file changed, 10 insertions(+), 2 deletions(-)
+
+diff --git a/module/zfs/zpl_xattr.c b/module/zfs/zpl_xattr.c
+index c03764f..a7e38e6 100644
+--- a/module/zfs/zpl_xattr.c
++++ b/module/zfs/zpl_xattr.c
+@@ -225,6 +225,11 @@
+ 		goto out;
+ 	}
+ 
++	if (size < i_size_read(xip)) {
++		error = -ERANGE;
++		goto out;
++	}
++
+ 	error = zpl_read_common(xip, value, size, 0, UIO_SYSSPACE, 0, cr);
+ out:
+ 	if (xip)
+@@ -263,9 +268,12 @@
+ 	if (!size)
+ 		return (nv_size);
+ 
+-	memcpy(value, nv_value, MIN(size, nv_size));
++	if (size < nv_size)
++		return (-ERANGE);
++
++	memcpy(value, nv_value, nv_size);
+ 
+-	return (MIN(size, nv_size));
++	return (nv_size);
+ }
+ 
+ static int
+-- 
+1.8.1.6
+
diff --git a/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-xattr-behavior-2.patch b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-xattr-behavior-2.patch
new file mode 100644
index 000000000000..bb4b1c4a3a34
--- /dev/null
+++ b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-xattr-behavior-2.patch
@@ -0,0 +1,37 @@
+From 0377189b884fab7db02a95088e05712c7cf336f7 Mon Sep 17 00:00:00 2001
+From: Brian Behlendorf <behlendorf1@llnl.gov>
+Date: Wed, 8 May 2013 09:20:04 -0700
+Subject: [PATCH] Only check directory xattr on ENOENT
+
+When SA xattrs are enabled only fallback to checking the directory
+xattrs when the name is not found as a SA xattr.  Otherwise, the SA
+error which should be returned to the caller is overwritten by the
+directory xattr errors.  Positive return values indicating success
+will also be immediately returned.
+
+In the case of #1437 the ERANGE error was being correctly returned
+by zpl_xattr_get_sa() only to be overridden with ENOENT which was
+returned by the subsequent unnessisary call to zpl_xattr_get_dir().
+
+Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
+Closes #1437
+---
+ module/zfs/zpl_xattr.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/module/zfs/zpl_xattr.c b/module/zfs/zpl_xattr.c
+index a7e38e6..eb2c00d 100644
+--- a/module/zfs/zpl_xattr.c
++++ b/module/zfs/zpl_xattr.c
+@@ -288,7 +288,7 @@
+ 
+ 	if (zsb->z_use_sa && zp->z_is_sa) {
+ 		error = zpl_xattr_get_sa(ip, name, value, size);
+-		if (error >= 0)
++		if (error != -ENOENT)
+ 			goto out;
+ 	}
+ 
+-- 
+1.8.1.6
+
diff --git a/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-zvol-initialization.patch b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-zvol-initialization.patch
new file mode 100644
index 000000000000..660d76c82ccf
--- /dev/null
+++ b/sys-fs/zfs-kmod/files/zfs-kmod-0.6.1-fix-zvol-initialization.patch
@@ -0,0 +1,113 @@
+diff --git a/module/zfs/spa.c b/module/zfs/spa.c
+index 82ee445..eaf05b9 100644
+--- a/module/zfs/spa.c
++++ b/module/zfs/spa.c
+@@ -64,6 +64,7 @@
+ #include <sys/zfs_ioctl.h>
+ #include <sys/dsl_scan.h>
+ #include <sys/zfeature.h>
++#include <sys/zvol.h>
+ 
+ #ifdef	_KERNEL
+ #include <sys/bootprops.h>
+@@ -2856,6 +2857,7 @@ spa_open_common(const char *pool, spa_t **spapp, void *tag, nvlist_t *nvpolicy,
+ 	spa_load_state_t state = SPA_LOAD_OPEN;
+ 	int error;
+ 	int locked = B_FALSE;
++	int firstopen = B_FALSE;
+ 
+ 	*spapp = NULL;
+ 
+@@ -2879,6 +2881,8 @@ spa_open_common(const char *pool, spa_t **spapp, void *tag, nvlist_t *nvpolicy,
+ 	if (spa->spa_state == POOL_STATE_UNINITIALIZED) {
+ 		zpool_rewind_policy_t policy;
+ 
++		firstopen = B_TRUE;
++
+ 		zpool_get_rewind_policy(nvpolicy ? nvpolicy : spa->spa_config,
+ 		    &policy);
+ 		if (policy.zrp_request & ZPOOL_DO_REWIND)
+@@ -2953,6 +2957,11 @@ spa_open_common(const char *pool, spa_t **spapp, void *tag, nvlist_t *nvpolicy,
+ 		mutex_exit(&spa_namespace_lock);
+ 	}
+ 
++#ifdef _KERNEL
++	if (firstopen)
++		zvol_create_minors(spa->spa_name);
++#endif
++
+ 	*spapp = spa;
+ 
+ 	return (0);
+diff --git a/module/zfs/zvol.c b/module/zfs/zvol.c
+index b41eeb2..f9387a7 100644
+--- a/module/zfs/zvol.c
++++ b/module/zfs/zvol.c
+@@ -1217,6 +1217,8 @@ zvol_alloc(dev_t dev, const char *name)
+ 	if (zv == NULL)
+ 		goto out;
+ 
++	spin_lock_init(&zv->zv_lock);
++
+ 	zv->zv_queue = blk_init_queue(zvol_request, &zv->zv_lock);
+ 	if (zv->zv_queue == NULL)
+ 		goto out_kmem;
+@@ -1250,7 +1252,6 @@ zvol_alloc(dev_t dev, const char *name)
+ 	    sizeof (rl_t), offsetof(rl_t, r_node));
+ 	zv->zv_znode.z_is_zvol = TRUE;
+ 
+-	spin_lock_init(&zv->zv_lock);
+ 	list_link_init(&zv->zv_next);
+ 
+ 	zv->zv_disk->major = zvol_major;
+@@ -1563,36 +1564,40 @@ zvol_init(void)
+ {
+ 	int error;
+ 
++	list_create(&zvol_state_list, sizeof (zvol_state_t),
++	            offsetof(zvol_state_t, zv_next));
++	mutex_init(&zvol_state_lock, NULL, MUTEX_DEFAULT, NULL);
++
+ 	zvol_taskq = taskq_create(ZVOL_DRIVER, zvol_threads, maxclsyspri,
+ 		                  zvol_threads, INT_MAX, TASKQ_PREPOPULATE);
+ 	if (zvol_taskq == NULL) {
+ 		printk(KERN_INFO "ZFS: taskq_create() failed\n");
+-		return (-ENOMEM);
++		error = -ENOMEM;
++		goto out1;
+ 	}
+ 
+ 	error = register_blkdev(zvol_major, ZVOL_DRIVER);
+ 	if (error) {
+ 		printk(KERN_INFO "ZFS: register_blkdev() failed %d\n", error);
+-		taskq_destroy(zvol_taskq);
+-		return (error);
++		goto out2;
+ 	}
+ 
+ 	blk_register_region(MKDEV(zvol_major, 0), 1UL << MINORBITS,
+ 	                    THIS_MODULE, zvol_probe, NULL, NULL);
+ 
+-	mutex_init(&zvol_state_lock, NULL, MUTEX_DEFAULT, NULL);
+-	list_create(&zvol_state_list, sizeof (zvol_state_t),
+-	            offsetof(zvol_state_t, zv_next));
+-
+-	(void) zvol_create_minors(NULL);
+-
+ 	return (0);
++
++out2:
++	taskq_destroy(zvol_taskq);
++out1:
++	mutex_destroy(&zvol_state_lock);
++	list_destroy(&zvol_state_list);
++	return (error);
+ }
+ 
+ void
+ zvol_fini(void)
+ {
+-	zvol_remove_minors(NULL);
+ 	blk_unregister_region(MKDEV(zvol_major, 0), 1UL << MINORBITS);
+ 	unregister_blkdev(zvol_major, ZVOL_DRIVER);
+ 	taskq_destroy(zvol_taskq);
diff --git a/sys-fs/zfs-kmod/zfs-kmod-0.6.1-r1.ebuild b/sys-fs/zfs-kmod/zfs-kmod-0.6.1-r1.ebuild
new file mode 100644
index 000000000000..47a64f2cb1ef
--- /dev/null
+++ b/sys-fs/zfs-kmod/zfs-kmod-0.6.1-r1.ebuild
@@ -0,0 +1,139 @@
+# Copyright 1999-2013 Gentoo Foundation
+# Distributed under the terms of the GNU General Public License v2
+# $Header: /var/cvsroot/gentoo-x86/sys-fs/zfs-kmod/zfs-kmod-0.6.1-r1.ebuild,v 1.1 2013/05/28 13:08:52 ryao Exp $
+
+EAPI="4"
+
+AT_M4DIR="config"
+AUTOTOOLS_AUTORECONF="1"
+AUTOTOOLS_IN_SOURCE_BUILD="1"
+
+inherit bash-completion-r1 flag-o-matic linux-info linux-mod toolchain-funcs autotools-utils
+
+if [ ${PV} == "9999" ] ; then
+	inherit git-2
+	MY_PV=9999
+	EGIT_REPO_URI="git://github.com/zfsonlinux/zfs.git"
+else
+	inherit eutils versionator
+	MY_PV=$(replace_version_separator 3 '-')
+	SRC_URI="https://github.com/zfsonlinux/zfs/archive/zfs-${MY_PV}.tar.gz"
+	S="${WORKDIR}/zfs-zfs-${MY_PV}"
+	KEYWORDS="~amd64"
+fi
+
+DESCRIPTION="Linux ZFS kernel module for sys-fs/zfs"
+HOMEPAGE="http://zfsonlinux.org/"
+
+LICENSE="CDDL debug? ( GPL-2+ )"
+SLOT="0"
+IUSE="custom-cflags debug +rootfs"
+RESTRICT="test"
+
+DEPEND="
+	=sys-kernel/spl-${PV}*
+	dev-lang/perl
+	virtual/awk
+"
+
+RDEPEND="${DEPEND}
+	!sys-fs/zfs-fuse
+"
+
+pkg_setup() {
+	linux-info_pkg_setup
+	CONFIG_CHECK="!DEBUG_LOCK_ALLOC
+		BLK_DEV_LOOP
+		EFI_PARTITION
+		IOSCHED_NOOP
+		MODULES
+		!PAX_KERNEXEC_PLUGIN_METHOD_OR
+		ZLIB_DEFLATE
+		ZLIB_INFLATE
+	"
+
+	use rootfs && \
+		CONFIG_CHECK="${CONFIG_CHECK} BLK_DEV_INITRD
+			DEVTMPFS"
+
+	kernel_is ge 2 6 26 || die "Linux 2.6.26 or newer required"
+
+	[ ${PV} != "9999" ] && \
+		{ kernel_is le 3 9 || die "Linux 3.9 is the latest supported version."; }
+
+	check_extra_config
+}
+
+src_prepare() {
+	if [ ${PV} != "9999" ]
+	then
+		# Correctness fix for getdents
+		epatch "${FILESDIR}/${P}-fix-getdents.patch"
+
+		# Prevent possible deadlock regression
+		epatch "${FILESDIR}/${P}-fix-txg_quiesce-deadlock.patch"
+
+		# Correctness fixes for xattr
+		epatch "${FILESDIR}/${P}-fix-xattr-behavior-1.patch"
+		epatch "${FILESDIR}/${P}-fix-xattr-behavior-2.patch"
+
+		# Make certain that zvols always appear
+		epatch "${FILESDIR}/${P}-fix-zvol-initialization.patch"
+	fi
+
+	# Remove GPLv2-licensed ZPIOS unless we are debugging
+	use debug || sed -e 's/^subdir-m += zpios$//' -i "${S}/module/Makefile.in"
+
+	autotools-utils_src_prepare
+}
+
+src_configure() {
+	use custom-cflags || strip-flags
+	filter-ldflags -Wl,*
+
+	set_arch_to_kernel
+	local myeconfargs=(
+		--bindir="${EPREFIX}/bin"
+		--sbindir="${EPREFIX}/sbin"
+		--with-config=kernel
+		--with-linux="${KV_DIR}"
+		--with-linux-obj="${KV_OUT_DIR}"
+		$(use_enable debug)
+	)
+	autotools-utils_src_configure
+}
+
+src_install() {
+	autotools-utils_src_install
+	dodoc AUTHORS COPYRIGHT DISCLAIMER README.markdown
+
+	# Provide /usr/src/zfs symlink for lustre
+	dosym "$(basename $(echo "${ED}/usr/src/zfs-"*))/${KV_FULL}" /usr/src/zfs
+}
+
+pkg_postinst() {
+	linux-mod_pkg_postinst
+
+	# Remove old modules
+	if [ -d "${EROOT}lib/modules/${KV_FULL}/addon/zfs" ]
+	then
+		ewarn "${PN} now installs modules in ${EROOT}lib/modules/${KV_FULL}/extra/zfs"
+		ewarn "Old modules were detected in ${EROOT}lib/modules/${KV_FULL}/addon/zfs"
+		ewarn "Automatically removing old modules to avoid problems."
+		rm -r "${EROOT}lib/modules/${KV_FULL}/addon/zfs" || die "Cannot remove modules"
+		rmdir --ignore-fail-on-non-empty "${EROOT}lib/modules/${KV_FULL}/addon"
+	fi
+
+	if use x86 || use arm
+	then
+		ewarn "32-bit kernels will likely require increasing vmalloc to"
+		ewarn "at least 256M and decreasing zfs_arc_max to some value less than that."
+	fi
+
+	ewarn "This version of ZFSOnLinux includes support for features flags."
+	ewarn "If you upgrade your pools to make use of feature flags, you will lose"
+	ewarn "the ability to import them using older versions of ZFSOnLinux."
+	ewarn "Any new pools will be created with feature flag support and will"
+	ewarn "not be compatible with older versions of ZFSOnLinux. To create a new"
+	ewarn "pool that is backward compatible, use zpool create -o version=28 ..."
+}