c/zfsonlinux
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

update ZFS to 0.7.7

Browse Source 
by importing the upstream release as patches. replace user namespace
patch with version which has been applied usptream.
This commit is contained in:
Fabian Grünbichler 2018-04-03 14:25:21 +02:00
parent 8324054718
commit 75b07eca3e
55 changed files with 7614 additions and 69 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

67
zfs-patches/0003-Use-user-namespaces-for-FSETID-policy-check.patch
View File

@ -1,67 +0,0 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Wolfgang Bumiller <w.bumiller@proxmox.com>
Date: Tue, 31 Oct 2017 09:08:42 +0100
Subject: [PATCH] Use user namespaces for FSETID policy check.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
With this we also need to verify the group id of a file with
the setgid flag has a valid mapping in the current
namespace.
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/policy.c | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/module/zfs/policy.c b/module/zfs/policy.c
index 03e8f748b..dbbcfefa3 100644
--- a/module/zfs/policy.c
+++ b/module/zfs/policy.c
@@ -42,19 +42,26 @@
* all other cases this function must fail and return the passed err.
*/
static int
-priv_policy(const cred_t *cr, int capability, boolean_t all, int err)
+priv_policy_ns(const cred_t *cr, int capability, boolean_t all, int err,
+ struct user_namespace *ns)
{
ASSERT3S(all, ==, B_FALSE);
if (cr != CRED() && (cr != kcred))
return (err);
- if (!capable(capability))
+ if (!(ns ? ns_capable(ns, capability) : capable(capability)))
return (err);
return (0);
}
+static int
+priv_policy(const cred_t *cr, int capability, boolean_t all, int err)
+{
+ return priv_policy_ns(cr, capability, all, err, NULL);
+}
+
/*
* Checks for operations that are either client-only or are used by
* both clients and servers.
@@ -175,8 +182,11 @@ secpolicy_vnode_setid_retain(const cred_t *cr, boolean_t issuidroot)
int
secpolicy_vnode_setids_setgids(const cred_t *cr, gid_t gid)
{
+ if (!kgid_has_mapping(cr->user_ns, SGID_TO_KGID(gid)))
+ return (EPERM);
if (crgetfsgid(cr) != gid && !groupmember(gid, cr))
- return (priv_policy(cr, CAP_FSETID, B_FALSE, EPERM));
+ return (priv_policy_ns(cr, CAP_FSETID, B_FALSE, EPERM,
+ cr->user_ns));
return (0);
}
--
2.14.2

0
zfs-patches/0004-always-load-ZFS-module-on-boot.patch → zfs-patches/0003-always-load-ZFS-module-on-boot.patch
View File

89
zfs-patches/0004-Make-enable-debug-fail-when-given-bogus-args.patch Normal file
View File

@ -0,0 +1,89 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Richard Yao <richard.yao@clusterhq.com>
Date: Tue, 23 Sep 2014 13:31:33 -0400
Subject: [PATCH] Make --enable-debug fail when given bogus args
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Currently, bogus options to --enable-debug become --disable-debug. That
means that passing --enable-debug=true is analogous to --disable-debug,
but the result is counterintuitive. We switch to AS_CASE to allow us to
fail when given a bogus option.
Also, we modify the text printed to clarify that --enable-debug enables
assertions.
Reviewed-by: Chunwei Chen <tuxoko@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <richard.yao@clusterhq.com>
Closes #2734
(cherry picked from commit 0f1ff38476ba255eb1f066f675c8721168ff2295)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
config/zfs-build.m4 | 43 +++++++++++++++++++++++++------------------
1 file changed, 25 insertions(+), 18 deletions(-)
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index 7651dc2c1..78a87aef7 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -6,30 +6,37 @@ AC_DEFUN([ZFS_AC_LICENSE], [
AC_MSG_RESULT([$ZFS_META_LICENSE])
])
+AC_DEFUN([ZFS_AC_DEBUG_ENABLE], [
+ KERNELCPPFLAGS="${KERNELCPPFLAGS} -DDEBUG -Werror"
+ HOSTCFLAGS="${HOSTCFLAGS} -DDEBUG -Werror"
+ DEBUG_CFLAGS="-DDEBUG -Werror"
+ DEBUG_STACKFLAGS="-fstack-check"
+ DEBUG_ZFS="_with_debug"
+ AC_DEFINE(ZFS_DEBUG, 1, [zfs debugging enabled])
+])
+
+AC_DEFUN([ZFS_AC_DEBUG_DISABLE], [
+ KERNELCPPFLAGS="${KERNELCPPFLAGS} -DNDEBUG "
+ HOSTCFLAGS="${HOSTCFLAGS} -DNDEBUG "
+ DEBUG_CFLAGS="-DNDEBUG"
+ DEBUG_STACKFLAGS=""
+ DEBUG_ZFS="_without_debug"
+])
+
AC_DEFUN([ZFS_AC_DEBUG], [
- AC_MSG_CHECKING([whether debugging is enabled])
+ AC_MSG_CHECKING([whether assertion support will be enabled])
AC_ARG_ENABLE([debug],
[AS_HELP_STRING([--enable-debug],
- [Enable generic debug support @<:@default=no@:>@])],
+ [Enable assertion support @<:@default=no@:>@])],
[],
[enable_debug=no])
- AS_IF([test "x$enable_debug" = xyes],
- [
- KERNELCPPFLAGS="${KERNELCPPFLAGS} -DDEBUG -Werror"
- HOSTCFLAGS="${HOSTCFLAGS} -DDEBUG -Werror"
- DEBUG_CFLAGS="-DDEBUG -Werror"
- DEBUG_STACKFLAGS="-fstack-check"
- DEBUG_ZFS="_with_debug"
- AC_DEFINE(ZFS_DEBUG, 1, [zfs debugging enabled])
- ],
- [
- KERNELCPPFLAGS="${KERNELCPPFLAGS} -DNDEBUG "
- HOSTCFLAGS="${HOSTCFLAGS} -DNDEBUG "
- DEBUG_CFLAGS="-DNDEBUG"
- DEBUG_STACKFLAGS=""
- DEBUG_ZFS="_without_debug"
- ])
+ AS_CASE(["x$enable_debug"],
+ ["xyes"],
+ [ZFS_AC_DEBUG_ENABLE],
+ ["xno"],
+ [ZFS_AC_DEBUG_DISABLE],
+ [AC_MSG_ERROR([Unknown option $enable_debug])])
AC_SUBST(DEBUG_CFLAGS)
AC_SUBST(DEBUG_STACKFLAGS)
--
2.14.2

122
zfs-patches/0005-Implement-enable-debuginfo-to-force-debuginfo.patch Normal file
View File

@ -0,0 +1,122 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Richard Yao <richard.yao@clusterhq.com>
Date: Tue, 23 Sep 2014 14:29:30 -0400
Subject: [PATCH] Implement --enable-debuginfo to force debuginfo
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Inspection of a Ubuntu 14.04 x64 system revealed that the config file
used to build the kernel image differs from the config file used to
build kernel modules by the presence of CONFIG_DEBUG_INFO=y:
This in itself is insufficient to show that the kernel is built with
debuginfo, but a cursory analysis of the debuginfo provided and the
size of the kernel strongly suggests that it was built with
CONFIG_DEBUG_INFO=y while the modules were not. Installing
linux-image-$(uname -r)-dbgsym had no obvious effect on the debuginfo
provided by either the modules or the kernel.
The consequence is that issue reports from distributions such as Ubuntu
and its derivatives build kernel modules without debuginfo contain
nonsensical backtraces. It is therefore desireable to force generation
of debuginfo, so we implement --enable-debuginfo. Since the build system
can build both userspace components and kernel modules, the generic
--enable-debuginfo option will force debuginfo for both. However, it
also supports --enable-debuginfo=kernel and --enable-debuginfo=user for
finer grained control.
Enabling debuginfo for the kernel modules works by injecting
CONFIG_DEBUG_INFO=y into the make environment. This is enables
generation of debuginfo by the kernel build systems on all Linux
kernels, but the build environment is slightly different int hat
CONFIG_DEBUG_INFO has not been in the CPP. Adding -DCONFIG_DEBUG_INFO
would fix that, but it would also cause build failures on kernels where
CONFIG_DEBUG_INFO=y is already set. That would complicate its use in
DKMS environments that support a range of kernels and is therefore
undesireable. We could write a compatibility shim to enable
CONFIG_DEBUG_INFO only when it is explicitly disabled, but we forgo
doing that because it is unnecessary. Nothing in ZoL or the kernel uses
CONFIG_DEBUG_INFO in the CPP at this time and that is unlikely to
change.
Enabling debuginfo for the userspace components is done by injecting -g
into CPPFLAGS. This is not necessary because the build system honors the
environment's CPPFLAGS by appending them to the actual CPPFLAGS used,
but it is supported for consistency.
Reviewed-by: Chunwei Chen <tuxoko@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Richard Yao <richard.yao@clusterhq.com>
Closes #2734
(cherry picked from commit 834815e9f767c9c5e7220ff84f29b1f069822a4d)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
configure.ac | 1 +
config/zfs-build.m4 | 33 ++++++++++++++++++++++++++++++++-
2 files changed, 33 insertions(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac
index 0c7977ef8..e145aa370 100644
--- a/configure.ac
+++ b/configure.ac
@@ -55,6 +55,7 @@ ZFS_AC_LICENSE
ZFS_AC_PACKAGE
ZFS_AC_CONFIG
ZFS_AC_DEBUG
+ZFS_AC_DEBUGINFO
AC_CONFIG_FILES([
Makefile
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index 78a87aef7..a8609b829 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -38,12 +38,43 @@ AC_DEFUN([ZFS_AC_DEBUG], [
[ZFS_AC_DEBUG_DISABLE],
[AC_MSG_ERROR([Unknown option $enable_debug])])
- AC_SUBST(DEBUG_CFLAGS)
AC_SUBST(DEBUG_STACKFLAGS)
AC_SUBST(DEBUG_ZFS)
AC_MSG_RESULT([$enable_debug])
])
+AC_DEFUN([ZFS_AC_DEBUGINFO_KERNEL], [
+ KERNELMAKE_PARAMS="$KERNELMAKE_PARAMS CONFIG_DEBUG_INFO=y"
+])
+
+AC_DEFUN([ZFS_AC_DEBUGINFO_USER], [
+ DEBUG_CFLAGS="$DEBUG_CFLAGS -g"
+])
+
+AC_DEFUN([ZFS_AC_DEBUGINFO], [
+ AC_MSG_CHECKING([whether debuginfo support will be forced])
+ AC_ARG_ENABLE([debuginfo],
+ [AS_HELP_STRING([--enable-debuginfo],
+ [Force generation of debuginfo @<:@default=no@:>@])],
+ [],
+ [enable_debuginfo=no])
+
+ AS_CASE(["x$enable_debuginfo"],
+ ["xyes"],
+ [ZFS_AC_DEBUGINFO_KERNEL
+ ZFS_AC_DEBUGINFO_USER],
+ ["xkernel"],
+ [ZFS_AC_DEBUGINFO_KERNEL],
+ ["xuser"],
+ [ZFS_AC_DEBUGINFO_USER],
+ ["xno"],
+ [],
+ [AC_MSG_ERROR([Unknown option $enable_debug])])
+
+ AC_SUBST(DEBUG_CFLAGS)
+ AC_MSG_RESULT([$enable_debuginfo])
+])
+
AC_DEFUN([ZFS_AC_CONFIG_ALWAYS], [
ZFS_AC_CONFIG_ALWAYS_NO_UNUSED_BUT_SET_VARIABLE
ZFS_AC_CONFIG_ALWAYS_NO_BOOL_COMPARE
--
2.14.2

121
zfs-patches/0006-Add-configure-option-to-enable-gcov-analysis.patch Normal file
View File

@ -0,0 +1,121 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Fri, 15 Sep 2017 10:24:13 -0700
Subject: [PATCH] Add configure option to enable gcov analysis
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
* Add configure option to enable gcov analysis.
* Includes a few minor ctime fixes.
* Add codecov.yml configuration.
Reviewed-by: Prakash Surya <prakash.surya@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #6642
(cherry picked from commit 184087f822ef178d6f5c155681ee7c3d90f23321)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
configure.ac | 1 +
Makefile.am | 8 +++---
.github/codecov.yml | 9 +++++++
config/zfs-build.m4 | 31 ++++++++++++++++++++++
.../tests/functional/ctime/ctime_001_pos.ksh | 0
5 files changed, 45 insertions(+), 4 deletions(-)
create mode 100644 .github/codecov.yml
mode change 100644 => 100755 tests/zfs-tests/tests/functional/ctime/ctime_001_pos.ksh
diff --git a/configure.ac b/configure.ac
index e145aa370..ee754fd38 100644
--- a/configure.ac
+++ b/configure.ac
@@ -56,6 +56,7 @@ ZFS_AC_PACKAGE
ZFS_AC_CONFIG
ZFS_AC_DEBUG
ZFS_AC_DEBUGINFO
+ZFS_AC_GCOV
AC_CONFIG_FILES([
Makefile
diff --git a/Makefile.am b/Makefile.am
index 508d3f40e..732a373bd 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -29,10 +29,10 @@ distclean-local::
-o -name .pc -o -name .hg -o -name .git \) -prune -o \
\( -name '*.orig' -o -name '*.rej' -o -name '*~' \
-o -name '*.bak' -o -name '#*#' -o -name '.*.orig' \
- -o -name '.*.rej' -o -name '.script-config' -o -size 0 \
- -o -name '*%' -o -name '.*.cmd' -o -name 'core' \
- -o -name 'Makefile' -o -name 'Module.symvers' \
- -o -name '*.order' -o -name '*.markers' \) \
+ -o -name '.*.rej' -o -size 0 -o -name '*%' -o -name '.*.cmd' \
+ -o -name 'core' -o -name 'Makefile' -o -name 'Module.symvers' \
+ -o -name '*.order' -o -name '*.markers' -o -name '*.gcda' \
+ -o -name '*.gcno' \) \
-type f -print | xargs $(RM)
dist-hook:
diff --git a/.github/codecov.yml b/.github/codecov.yml
new file mode 100644
index 000000000..f36be39cb
--- /dev/null
+++ b/.github/codecov.yml
@@ -0,0 +1,9 @@
+codecov:
+ strict_yaml_branch: master # only use the latest copy on master branch
+
+comment: off
+
+coverage:
+ status:
+ project: off
+ patch: off
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index a8609b829..b84658a64 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -75,6 +75,37 @@ AC_DEFUN([ZFS_AC_DEBUGINFO], [
AC_MSG_RESULT([$enable_debuginfo])
])
+AC_DEFUN([ZFS_AC_GCOV_KERNEL], [
+])
+
+AC_DEFUN([ZFS_AC_GCOV_USER], [
+ DEBUG_CFLAGS="$DEBUG_CFLAGS -fprofile-arcs -ftest-coverage"
+])
+
+AC_DEFUN([ZFS_AC_GCOV], [
+ AC_MSG_CHECKING([whether gcov profiling will be enabled])
+ AC_ARG_ENABLE([gcov],
+ [AS_HELP_STRING([--enable-gcov],
+ [Enable gcov profiling @<:@default=no@:>@])],
+ [],
+ [enable_gcov=no])
+
+ AS_CASE(["x$enable_gcov"],
+ ["xyes"],
+ [ZFS_AC_GCOV_KERNEL
+ ZFS_AC_GCOV_USER],
+ ["xkernel"],
+ [ZFS_AC_GCOV_KERNEL],
+ ["xuser"],
+ [ZFS_AC_GCOV_USER],
+ ["xno"],
+ [],
+ [AC_MSG_ERROR([Unknown option $enable_gcov])])
+
+ AC_SUBST(DEBUG_CFLAGS)
+ AC_MSG_RESULT([$enable_gcov])
+])
+
AC_DEFUN([ZFS_AC_CONFIG_ALWAYS], [
ZFS_AC_CONFIG_ALWAYS_NO_UNUSED_BUT_SET_VARIABLE
ZFS_AC_CONFIG_ALWAYS_NO_BOOL_COMPARE
diff --git a/tests/zfs-tests/tests/functional/ctime/ctime_001_pos.ksh b/tests/zfs-tests/tests/functional/ctime/ctime_001_pos.ksh
old mode 100644
new mode 100755
--
2.14.2

43
zfs-patches/0007-Make-fno-inline-compile-option-more-accessible.patch Normal file
View File

@ -0,0 +1,43 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Prakash Surya <prakash.surya@delphix.com>
Date: Fri, 15 Sep 2017 11:47:11 -0700
Subject: [PATCH] Make "-fno-inline" compile option more accessible
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
When functions are inlined, it can make the system much more difficult
to instrument using tools such as ftrace, BPF, crash, etc. Thus, to aid
development and increase the system's observability, when the
"--enable-debuginfo" flag is specified, the "-fno-inline" compilation
option will be used for both userspace and kernel modules.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Prakash Surya <prakash.surya@delphix.com>
Closes #6605
(cherry picked from commit f1236ebf35d1531db75ac8a57b23e8fe48d6b780)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
config/zfs-build.m4 | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index b84658a64..c695a882e 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -45,10 +45,11 @@ AC_DEFUN([ZFS_AC_DEBUG], [
AC_DEFUN([ZFS_AC_DEBUGINFO_KERNEL], [
KERNELMAKE_PARAMS="$KERNELMAKE_PARAMS CONFIG_DEBUG_INFO=y"
+ KERNELCPPFLAGS="${KERNELCPPFLAGS} -fno-inline"
])
AC_DEFUN([ZFS_AC_DEBUGINFO_USER], [
- DEBUG_CFLAGS="$DEBUG_CFLAGS -g"
+ DEBUG_CFLAGS="${DEBUG_CFLAGS} -g -fno-inline"
])
AC_DEFUN([ZFS_AC_DEBUGINFO], [
--
2.14.2

399
zfs-patches/0008-Add-support-for-enable-code-coverage-option.patch Normal file
View File

@ -0,0 +1,399 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Prakash Surya <prakash.surya@delphix.com>
Date: Fri, 22 Sep 2017 18:49:57 -0700
Subject: [PATCH] Add support for "--enable-code-coverage" option
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
This change adds support for a new option that can be passed to the
configure script: "--enable-code-coverage". Further, the "--enable-gcov"
option has been removed, as this new option provides the same
functionality (plus more).
When using this new option the following make targets are available:
* check-code-coverage
* code-coverage-capture
* code-coverage-clean
Note: these make targets can only be run from the root of the project.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Prakash Surya <prakash.surya@delphix.com>
Closes #6670
(cherry picked from commit 6b278f3223c0322527836da2e11e33978e54a234)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
configure.ac | 2 +-
Makefile.am | 2 +
config/Rules.am | 2 +
config/ax_code_coverage.m4 | 264 +++++++++++++++++++++++++++++++++++++++++++++
config/zfs-build.m4 | 31 ------
5 files changed, 269 insertions(+), 32 deletions(-)
create mode 100644 config/ax_code_coverage.m4
diff --git a/configure.ac b/configure.ac
index ee754fd38..d71712e4c 100644
--- a/configure.ac
+++ b/configure.ac
@@ -50,13 +50,13 @@ AC_PROG_CC
AC_PROG_LIBTOOL
AM_PROG_AS
AM_PROG_CC_C_O
+AX_CODE_COVERAGE
ZFS_AC_LICENSE
ZFS_AC_PACKAGE
ZFS_AC_CONFIG
ZFS_AC_DEBUG
ZFS_AC_DEBUGINFO
-ZFS_AC_GCOV
AC_CONFIG_FILES([
Makefile
diff --git a/Makefile.am b/Makefile.am
index 732a373bd..b539ff30f 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -23,6 +23,8 @@ EXTRA_DIST = autogen.sh copy-builtin
EXTRA_DIST += config/config.awk config/rpm.am config/deb.am config/tgz.am
EXTRA_DIST += META DISCLAIMER COPYRIGHT README.markdown OPENSOLARIS.LICENSE
+@CODE_COVERAGE_RULES@
+
distclean-local::
-$(RM) -R autom4te*.cache
-find . \( -name SCCS -o -name BitKeeper -o -name .svn -o -name CVS \
diff --git a/config/Rules.am b/config/Rules.am
index 1d39e7779..215f09c34 100644
--- a/config/Rules.am
+++ b/config/Rules.am
@@ -6,6 +6,7 @@ AM_CFLAGS += ${NO_UNUSED_BUT_SET_VARIABLE}
AM_CFLAGS += ${NO_BOOL_COMPARE}
AM_CFLAGS += -fno-strict-aliasing
AM_CFLAGS += -std=gnu99
+AM_CFLAGS += $(CODE_COVERAGE_CFLAGS)
AM_CPPFLAGS = -D_GNU_SOURCE -D__EXTENSIONS__ -D_REENTRANT
AM_CPPFLAGS += -D_POSIX_PTHREAD_SEMANTICS -D_FILE_OFFSET_BITS=64
AM_CPPFLAGS += -D_LARGEFILE64_SOURCE -DHAVE_LARGE_STACKS=1
@@ -14,3 +15,4 @@ AM_CPPFLAGS += -DLIBEXECDIR=\"$(libexecdir)\"
AM_CPPFLAGS += -DRUNSTATEDIR=\"$(runstatedir)\"
AM_CPPFLAGS += -DSBINDIR=\"$(sbindir)\"
AM_CPPFLAGS += -DSYSCONFDIR=\"$(sysconfdir)\"
+AM_CPPFLAGS += $(CODE_COVERAGE_CPPFLAGS)
diff --git a/config/ax_code_coverage.m4 b/config/ax_code_coverage.m4
new file mode 100644
index 000000000..6484f0332
--- /dev/null
+++ b/config/ax_code_coverage.m4
@@ -0,0 +1,264 @@
+# ===========================================================================
+# https://www.gnu.org/software/autoconf-archive/ax_code_coverage.html
+# ===========================================================================
+#
+# SYNOPSIS
+#
+# AX_CODE_COVERAGE()
+#
+# DESCRIPTION
+#
+# Defines CODE_COVERAGE_CPPFLAGS, CODE_COVERAGE_CFLAGS,
+# CODE_COVERAGE_CXXFLAGS and CODE_COVERAGE_LIBS which should be included
+# in the CPPFLAGS, CFLAGS CXXFLAGS and LIBS/LIBADD variables of every
+# build target (program or library) which should be built with code
+# coverage support. Also defines CODE_COVERAGE_RULES which should be
+# substituted in your Makefile; and $enable_code_coverage which can be
+# used in subsequent configure output. CODE_COVERAGE_ENABLED is defined
+# and substituted, and corresponds to the value of the
+# --enable-code-coverage option, which defaults to being disabled.
+#
+# Test also for gcov program and create GCOV variable that could be
+# substituted.
+#
+# Note that all optimization flags in CFLAGS must be disabled when code
+# coverage is enabled.
+#
+# Usage example:
+#
+# configure.ac:
+#
+# AX_CODE_COVERAGE
+#
+# Makefile.am:
+#
+# @CODE_COVERAGE_RULES@
+# my_program_LIBS = ... $(CODE_COVERAGE_LIBS) ...
+# my_program_CPPFLAGS = ... $(CODE_COVERAGE_CPPFLAGS) ...
+# my_program_CFLAGS = ... $(CODE_COVERAGE_CFLAGS) ...
+# my_program_CXXFLAGS = ... $(CODE_COVERAGE_CXXFLAGS) ...
+#
+# This results in a "check-code-coverage" rule being added to any
+# Makefile.am which includes "@CODE_COVERAGE_RULES@" (assuming the module
+# has been configured with --enable-code-coverage). Running `make
+# check-code-coverage` in that directory will run the module's test suite
+# (`make check`) and build a code coverage report detailing the code which
+# was touched, then print the URI for the report.
+#
+# In earlier versions of this macro, CODE_COVERAGE_LDFLAGS was defined
+# instead of CODE_COVERAGE_LIBS. They are both still defined, but use of
+# CODE_COVERAGE_LIBS is preferred for clarity; CODE_COVERAGE_LDFLAGS is
+# deprecated. They have the same value.
+#
+# This code was derived from Makefile.decl in GLib, originally licenced
+# under LGPLv2.1+.
+#
+# LICENSE
+#
+# Copyright (c) 2012, 2016 Philip Withnall
+# Copyright (c) 2012 Xan Lopez
+# Copyright (c) 2012 Christian Persch
+# Copyright (c) 2012 Paolo Borelli
+# Copyright (c) 2012 Dan Winship
+# Copyright (c) 2015 Bastien ROUCARIES
+#
+# This library is free software; you can redistribute it and/or modify it
+# under the terms of the GNU Lesser General Public License as published by
+# the Free Software Foundation; either version 2.1 of the License, or (at
+# your option) any later version.
+#
+# This library is distributed in the hope that it will be useful, but
+# WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser
+# General Public License for more details.
+#
+# You should have received a copy of the GNU Lesser General Public License
+# along with this program. If not, see <https://www.gnu.org/licenses/>.
+
+#serial 25
+
+AC_DEFUN([AX_CODE_COVERAGE],[
+ dnl Check for --enable-code-coverage
+ AC_REQUIRE([AC_PROG_SED])
+
+ # allow to override gcov location
+ AC_ARG_WITH([gcov],
+ [AS_HELP_STRING([--with-gcov[=GCOV]], [use given GCOV for coverage (GCOV=gcov).])],
+ [_AX_CODE_COVERAGE_GCOV_PROG_WITH=$with_gcov],
+ [_AX_CODE_COVERAGE_GCOV_PROG_WITH=gcov])
+
+ AC_MSG_CHECKING([whether to build with code coverage support])
+ AC_ARG_ENABLE([code-coverage],
+ AS_HELP_STRING([--enable-code-coverage],
+ [Whether to enable code coverage support]),,
+ enable_code_coverage=no)
+
+ AM_CONDITIONAL([CODE_COVERAGE_ENABLED], [test x$enable_code_coverage = xyes])
+ AC_SUBST([CODE_COVERAGE_ENABLED], [$enable_code_coverage])
+ AC_MSG_RESULT($enable_code_coverage)
+
+ AS_IF([ test "$enable_code_coverage" = "yes" ], [
+ # check for gcov
+ AC_CHECK_TOOL([GCOV],
+ [$_AX_CODE_COVERAGE_GCOV_PROG_WITH],
+ [:])
+ AS_IF([test "X$GCOV" = "X:"],
+ [AC_MSG_ERROR([gcov is needed to do coverage])])
+ AC_SUBST([GCOV])
+
+ dnl Check if gcc is being used
+ AS_IF([ test "$GCC" = "no" ], [
+ AC_MSG_ERROR([not compiling with gcc, which is required for gcov code coverage])
+ ])
+
+ AC_CHECK_PROG([LCOV], [lcov], [lcov])
+ AC_CHECK_PROG([GENHTML], [genhtml], [genhtml])
+
+ AS_IF([ test -z "$LCOV" ], [
+ AC_MSG_ERROR([To enable code coverage reporting you must have lcov installed])
+ ])
+
+ AS_IF([ test -z "$GENHTML" ], [
+ AC_MSG_ERROR([Could not find genhtml from the lcov package])
+ ])
+
+ dnl Build the code coverage flags
+ dnl Define CODE_COVERAGE_LDFLAGS for backwards compatibility
+ CODE_COVERAGE_CPPFLAGS="-DNDEBUG"
+ CODE_COVERAGE_CFLAGS="-O0 -g -fprofile-arcs -ftest-coverage"
+ CODE_COVERAGE_CXXFLAGS="-O0 -g -fprofile-arcs -ftest-coverage"
+ CODE_COVERAGE_LIBS="-lgcov"
+ CODE_COVERAGE_LDFLAGS="$CODE_COVERAGE_LIBS"
+
+ AC_SUBST([CODE_COVERAGE_CPPFLAGS])
+ AC_SUBST([CODE_COVERAGE_CFLAGS])
+ AC_SUBST([CODE_COVERAGE_CXXFLAGS])
+ AC_SUBST([CODE_COVERAGE_LIBS])
+ AC_SUBST([CODE_COVERAGE_LDFLAGS])
+
+ [CODE_COVERAGE_RULES_CHECK='
+ -$(A''M_V_at)$(MAKE) $(AM_MAKEFLAGS) -k check
+ $(A''M_V_at)$(MAKE) $(AM_MAKEFLAGS) code-coverage-capture
+']
+ [CODE_COVERAGE_RULES_CAPTURE='
+ $(code_coverage_v_lcov_cap)$(LCOV) $(code_coverage_quiet) $(addprefix --directory ,$(CODE_COVERAGE_DIRECTORY)) --capture --output-file "$(CODE_COVERAGE_OUTPUT_FILE).tmp" --test-name "$(call code_coverage_sanitize,$(PACKAGE_NAME)-$(PACKAGE_VERSION))" --no-checksum --compat-libtool $(CODE_COVERAGE_LCOV_SHOPTS) $(CODE_COVERAGE_LCOV_OPTIONS)
+ $(code_coverage_v_lcov_ign)$(LCOV) $(code_coverage_quiet) $(addprefix --directory ,$(CODE_COVERAGE_DIRECTORY)) --remove "$(CODE_COVERAGE_OUTPUT_FILE).tmp" "/tmp/*" $(CODE_COVERAGE_IGNORE_PATTERN) --output-file "$(CODE_COVERAGE_OUTPUT_FILE)" $(CODE_COVERAGE_LCOV_SHOPTS) $(CODE_COVERAGE_LCOV_RMOPTS)
+ -@rm -f $(CODE_COVERAGE_OUTPUT_FILE).tmp
+ $(code_coverage_v_genhtml)LANG=C $(GENHTML) $(code_coverage_quiet) $(addprefix --prefix ,$(CODE_COVERAGE_DIRECTORY)) --output-directory "$(CODE_COVERAGE_OUTPUT_DIRECTORY)" --title "$(PACKAGE_NAME)-$(PACKAGE_VERSION) Code Coverage" --legend --show-details "$(CODE_COVERAGE_OUTPUT_FILE)" $(CODE_COVERAGE_GENHTML_OPTIONS)
+ @echo "file://$(abs_builddir)/$(CODE_COVERAGE_OUTPUT_DIRECTORY)/index.html"
+']
+ [CODE_COVERAGE_RULES_CLEAN='
+clean: code-coverage-clean
+distclean: code-coverage-clean
+code-coverage-clean:
+ -$(LCOV) --directory $(top_builddir) -z
+ -rm -rf $(CODE_COVERAGE_OUTPUT_FILE) $(CODE_COVERAGE_OUTPUT_FILE).tmp $(CODE_COVERAGE_OUTPUT_DIRECTORY)
+ -find . \( -name "*.gcda" -o -name "*.gcno" -o -name "*.gcov" \) -delete
+']
+ ], [
+ [CODE_COVERAGE_RULES_CHECK='
+ @echo "Need to reconfigure with --enable-code-coverage"
+']
+ CODE_COVERAGE_RULES_CAPTURE="$CODE_COVERAGE_RULES_CHECK"
+ CODE_COVERAGE_RULES_CLEAN=''
+ ])
+
+[CODE_COVERAGE_RULES='
+# Code coverage
+#
+# Optional:
+# - CODE_COVERAGE_DIRECTORY: Top-level directory for code coverage reporting.
+# Multiple directories may be specified, separated by whitespace.
+# (Default: $(top_builddir))
+# - CODE_COVERAGE_OUTPUT_FILE: Filename and path for the .info file generated
+# by lcov for code coverage. (Default:
+# $(PACKAGE_NAME)-$(PACKAGE_VERSION)-coverage.info)
+# - CODE_COVERAGE_OUTPUT_DIRECTORY: Directory for generated code coverage
+# reports to be created. (Default:
+# $(PACKAGE_NAME)-$(PACKAGE_VERSION)-coverage)
+# - CODE_COVERAGE_BRANCH_COVERAGE: Set to 1 to enforce branch coverage,
+# set to 0 to disable it and leave empty to stay with the default.
+# (Default: empty)
+# - CODE_COVERAGE_LCOV_SHOPTS_DEFAULT: Extra options shared between both lcov
+# instances. (Default: based on $CODE_COVERAGE_BRANCH_COVERAGE)
+# - CODE_COVERAGE_LCOV_SHOPTS: Extra options to shared between both lcov
+# instances. (Default: $CODE_COVERAGE_LCOV_SHOPTS_DEFAULT)
+# - CODE_COVERAGE_LCOV_OPTIONS_GCOVPATH: --gcov-tool pathtogcov
+# - CODE_COVERAGE_LCOV_OPTIONS_DEFAULT: Extra options to pass to the
+# collecting lcov instance. (Default: $CODE_COVERAGE_LCOV_OPTIONS_GCOVPATH)
+# - CODE_COVERAGE_LCOV_OPTIONS: Extra options to pass to the collecting lcov
+# instance. (Default: $CODE_COVERAGE_LCOV_OPTIONS_DEFAULT)
+# - CODE_COVERAGE_LCOV_RMOPTS_DEFAULT: Extra options to pass to the filtering
+# lcov instance. (Default: empty)
+# - CODE_COVERAGE_LCOV_RMOPTS: Extra options to pass to the filtering lcov
+# instance. (Default: $CODE_COVERAGE_LCOV_RMOPTS_DEFAULT)
+# - CODE_COVERAGE_GENHTML_OPTIONS_DEFAULT: Extra options to pass to the
+# genhtml instance. (Default: based on $CODE_COVERAGE_BRANCH_COVERAGE)
+# - CODE_COVERAGE_GENHTML_OPTIONS: Extra options to pass to the genhtml
+# instance. (Default: $CODE_COVERAGE_GENHTML_OPTIONS_DEFAULT)
+# - CODE_COVERAGE_IGNORE_PATTERN: Extra glob pattern of files to ignore
+#
+# The generated report will be titled using the $(PACKAGE_NAME) and
+# $(PACKAGE_VERSION). In order to add the current git hash to the title,
+# use the git-version-gen script, available online.
+
+# Optional variables
+CODE_COVERAGE_DIRECTORY ?= $(top_builddir)
+CODE_COVERAGE_OUTPUT_FILE ?= $(PACKAGE_NAME)-$(PACKAGE_VERSION)-coverage.info
+CODE_COVERAGE_OUTPUT_DIRECTORY ?= $(PACKAGE_NAME)-$(PACKAGE_VERSION)-coverage
+CODE_COVERAGE_BRANCH_COVERAGE ?=
+CODE_COVERAGE_LCOV_SHOPTS_DEFAULT ?= $(if $(CODE_COVERAGE_BRANCH_COVERAGE),\
+--rc lcov_branch_coverage=$(CODE_COVERAGE_BRANCH_COVERAGE))
+CODE_COVERAGE_LCOV_SHOPTS ?= $(CODE_COVERAGE_LCOV_SHOPTS_DEFAULT)
+CODE_COVERAGE_LCOV_OPTIONS_GCOVPATH ?= --gcov-tool "$(GCOV)"
+CODE_COVERAGE_LCOV_OPTIONS_DEFAULT ?= $(CODE_COVERAGE_LCOV_OPTIONS_GCOVPATH)
+CODE_COVERAGE_LCOV_OPTIONS ?= $(CODE_COVERAGE_LCOV_OPTIONS_DEFAULT)
+CODE_COVERAGE_LCOV_RMOPTS_DEFAULT ?=
+CODE_COVERAGE_LCOV_RMOPTS ?= $(CODE_COVERAGE_LCOV_RMOPTS_DEFAULT)
+CODE_COVERAGE_GENHTML_OPTIONS_DEFAULT ?=\
+$(if $(CODE_COVERAGE_BRANCH_COVERAGE),\
+--rc genhtml_branch_coverage=$(CODE_COVERAGE_BRANCH_COVERAGE))
+CODE_COVERAGE_GENHTML_OPTIONS ?= $(CODE_COVERAGE_GENHTML_OPTIONS_DEFAULT)
+CODE_COVERAGE_IGNORE_PATTERN ?=
+
+GITIGNOREFILES ?=
+GITIGNOREFILES += $(CODE_COVERAGE_OUTPUT_FILE) $(CODE_COVERAGE_OUTPUT_DIRECTORY)
+
+code_coverage_v_lcov_cap = $(code_coverage_v_lcov_cap_$(V))
+code_coverage_v_lcov_cap_ = $(code_coverage_v_lcov_cap_$(AM_DEFAULT_VERBOSITY))
+code_coverage_v_lcov_cap_0 = @echo " LCOV --capture"\
+ $(CODE_COVERAGE_OUTPUT_FILE);
+code_coverage_v_lcov_ign = $(code_coverage_v_lcov_ign_$(V))
+code_coverage_v_lcov_ign_ = $(code_coverage_v_lcov_ign_$(AM_DEFAULT_VERBOSITY))
+code_coverage_v_lcov_ign_0 = @echo " LCOV --remove /tmp/*"\
+ $(CODE_COVERAGE_IGNORE_PATTERN);
+code_coverage_v_genhtml = $(code_coverage_v_genhtml_$(V))
+code_coverage_v_genhtml_ = $(code_coverage_v_genhtml_$(AM_DEFAULT_VERBOSITY))
+code_coverage_v_genhtml_0 = @echo " GEN " $(CODE_COVERAGE_OUTPUT_DIRECTORY);
+code_coverage_quiet = $(code_coverage_quiet_$(V))
+code_coverage_quiet_ = $(code_coverage_quiet_$(AM_DEFAULT_VERBOSITY))
+code_coverage_quiet_0 = --quiet
+
+# sanitizes the test-name: replaces with underscores: dashes and dots
+code_coverage_sanitize = $(subst -,_,$(subst .,_,$(1)))
+
+# Use recursive makes in order to ignore errors during check
+check-code-coverage:'"$CODE_COVERAGE_RULES_CHECK"'
+
+# Capture code coverage data
+code-coverage-capture: code-coverage-capture-hook'"$CODE_COVERAGE_RULES_CAPTURE"'
+
+# Hook rule executed before code-coverage-capture, overridable by the user
+code-coverage-capture-hook:
+
+'"$CODE_COVERAGE_RULES_CLEAN"'
+
+A''M_DISTCHECK_CONFIGURE_FLAGS ?=
+A''M_DISTCHECK_CONFIGURE_FLAGS += --disable-code-coverage
+
+.PHONY: check-code-coverage code-coverage-capture code-coverage-capture-hook code-coverage-clean
+']
+
+ AC_SUBST([CODE_COVERAGE_RULES])
+ m4_ifdef([_AM_SUBST_NOTMAKE], [_AM_SUBST_NOTMAKE([CODE_COVERAGE_RULES])])
+])
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index c695a882e..17cc80462 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -76,37 +76,6 @@ AC_DEFUN([ZFS_AC_DEBUGINFO], [
AC_MSG_RESULT([$enable_debuginfo])
])
-AC_DEFUN([ZFS_AC_GCOV_KERNEL], [
-])
-
-AC_DEFUN([ZFS_AC_GCOV_USER], [
- DEBUG_CFLAGS="$DEBUG_CFLAGS -fprofile-arcs -ftest-coverage"
-])
-
-AC_DEFUN([ZFS_AC_GCOV], [
- AC_MSG_CHECKING([whether gcov profiling will be enabled])
- AC_ARG_ENABLE([gcov],
- [AS_HELP_STRING([--enable-gcov],
- [Enable gcov profiling @<:@default=no@:>@])],
- [],
- [enable_gcov=no])
-
- AS_CASE(["x$enable_gcov"],
- ["xyes"],
- [ZFS_AC_GCOV_KERNEL
- ZFS_AC_GCOV_USER],
- ["xkernel"],
- [ZFS_AC_GCOV_KERNEL],
- ["xuser"],
- [ZFS_AC_GCOV_USER],
- ["xno"],
- [],
- [AC_MSG_ERROR([Unknown option $enable_gcov])])
-
- AC_SUBST(DEBUG_CFLAGS)
- AC_MSG_RESULT([$enable_gcov])
-])
-
AC_DEFUN([ZFS_AC_CONFIG_ALWAYS], [
ZFS_AC_CONFIG_ALWAYS_NO_UNUSED_BUT_SET_VARIABLE
ZFS_AC_CONFIG_ALWAYS_NO_BOOL_COMPARE
--
2.14.2

97
zfs-patches/0009-Update-codecov.yml.patch Normal file
View File

@ -0,0 +1,97 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Fri, 22 Sep 2017 18:54:34 -0700
Subject: [PATCH] Update codecov.yml
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Update the codecov.yml to make the following functional changes.
* Do not require the CI testing to pass before posting results.
* Set red-yellow-green coverage percent from 50%-100%
* Allow a 1% drop in coverage to still be considered a pass.
* Reduce the size of the comment posted to the issue.
Additionally, the top level README.markdown has been updated
to include the codecov.io badge and the project summary reworded.
Reviewed-by: Prakash Surya <prakash.surya@delphix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #6669
(cherry picked from commit 3790bfa80f9b26fd750b4a554c5707b9d380aac4)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
README.markdown | 12 ++++++------
.github/codecov.yml | 31 ++++++++++++++++++++++++++-----
2 files changed, 32 insertions(+), 11 deletions(-)
diff --git a/README.markdown b/README.markdown
index fd2ca8c86..1c305b35d 100644
--- a/README.markdown
+++ b/README.markdown
@@ -1,9 +1,9 @@
-<p align="center"><img src="http://zfsonlinux.org/images/zfs-linux.png"/></p>
-ZFS is an advanced file system and volume manager which was originally
-developed for Solaris and is now maintained by the Illumos community.
+![img](http://zfsonlinux.org/images/zfs-linux.png)
-ZFS on Linux, which is also known as ZoL, is currently feature complete. It
-includes fully functional and stable SPA, DMU, ZVOL, and ZPL layers. And it's native!
+ZFS on Linux is an advanced file system and volume manager which was originally
+developed for Solaris and is now maintained by the OpenZFS community.
+
+[![codecov](https://codecov.io/gh/zfsonlinux/zfs/branch/master/graph/badge.svg)](https://codecov.io/gh/zfsonlinux/zfs)
# Official Resources
* [Site](http://zfsonlinux.org)
@@ -16,4 +16,4 @@ Full documentation for installing ZoL on your favorite Linux distribution can
be found at [our site](http://zfsonlinux.org/).
# Contribute & Develop
-We have a separate document with [contribution guidelines](./.github/CONTRIBUTING.md).
\ No newline at end of file
+We have a separate document with [contribution guidelines](./.github/CONTRIBUTING.md).
diff --git a/.github/codecov.yml b/.github/codecov.yml
index f36be39cb..e74c59a0f 100644
--- a/.github/codecov.yml
+++ b/.github/codecov.yml
@@ -1,9 +1,30 @@
codecov:
- strict_yaml_branch: master # only use the latest copy on master branch
-
-comment: off
+ notify:
+ require_ci_to_pass: no
coverage:
+ precision: 2
+ round: down
+ range: "50...100"
+
status:
- project: off
- patch: off
+ project:
+ default:
+ threshold: 1%
+
+ patch:
+ default:
+ threshold: 1%
+
+parsers:
+ gcov:
+ branch_detection:
+ conditional: yes
+ loop: yes
+ method: no
+ macro: no
+
+comment:
+ layout: "header, sunburst, diff"
+ behavior: default
+ require_changes: no
--
2.14.2

35
zfs-patches/0010-Fix-enable-code-coverage-debug-build.patch Normal file
View File

@ -0,0 +1,35 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Fri, 22 Sep 2017 22:16:18 -0700
Subject: [PATCH] Fix "--enable-code-coverage" debug build
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
When --enable-code-coverage is provided it should not result
in NDEBUG being defined. This is controlled by --enable-debug.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #6674
(cherry picked from commit 6897ea475fd0c82a74edacf374d4e339f9a9b86b)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
config/ax_code_coverage.m4 | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/config/ax_code_coverage.m4 b/config/ax_code_coverage.m4
index 6484f0332..4417d4444 100644
--- a/config/ax_code_coverage.m4
+++ b/config/ax_code_coverage.m4
@@ -124,7 +124,7 @@ AC_DEFUN([AX_CODE_COVERAGE],[
dnl Build the code coverage flags
dnl Define CODE_COVERAGE_LDFLAGS for backwards compatibility
- CODE_COVERAGE_CPPFLAGS="-DNDEBUG"
+ CODE_COVERAGE_CPPFLAGS=""
CODE_COVERAGE_CFLAGS="-O0 -g -fprofile-arcs -ftest-coverage"
CODE_COVERAGE_CXXFLAGS="-O0 -g -fprofile-arcs -ftest-coverage"
CODE_COVERAGE_LIBS="-lgcov"
--
2.14.2

153
zfs-patches/0011-Remove-vn_rename-and-vn_remove-dependency.patch Normal file
View File

@ -0,0 +1,153 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Thu, 19 Oct 2017 10:06:55 -0700
Subject: [PATCH] Remove vn_rename and vn_remove dependency
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
The only place vn_rename and vn_remove are used is when writing
out an updated pool configuration file. By truncating the file
instead of renaming and removing it we can avoid having to implement
these interfaces entirely. Functionally an empty cache file is
treated the same as a missing cache file. This is particularly
advantageous because the Linux kernel has never provided a way
to reliably implement vn_rename and vn_remove.
The cachefile_004_pos.ksh test case was updated to understand
that an empty cache file is the same as a missing one.
The zfs-import-* systemd service files were not updated to use
ConditionFileNotEmpty in place of ConditionPathExists. This
means that after exporting all pools and rebooting new pools
will not the scanned for on the next boot. This small change
should not impact normal usage since pools are not exported
as part of a normal shutdown.
Documentation was updated accordingly.
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Arkadiusz Bubała <arkadiusz.bubala@open-e.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes zfsonlinux/spl#648
Closes #6753
(cherry picked from commit 5d62588032aa1d13d7f789cf564a0d20c77a5762)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/spa_config.c | 34 ++++++++++++++++++----
man/man8/zpool.8 | 2 +-
.../functional/cachefile/cachefile_004_pos.ksh | 6 ++--
3 files changed, 32 insertions(+), 10 deletions(-)
diff --git a/module/zfs/spa_config.c b/module/zfs/spa_config.c
index 5b792b868..fea239014 100644
--- a/module/zfs/spa_config.c
+++ b/module/zfs/spa_config.c
@@ -147,6 +147,26 @@ out:
kobj_close_file(file);
}
+static int
+spa_config_remove(spa_config_dirent_t *dp)
+{
+#if defined(__linux__) && defined(_KERNEL)
+ int error, flags = FWRITE | FTRUNC;
+ uio_seg_t seg = UIO_SYSSPACE;
+ vnode_t *vp;
+
+ error = vn_open(dp->scd_path, seg, flags, 0644, &vp, 0, 0);
+ if (error == 0) {
+ (void) VOP_FSYNC(vp, FSYNC, kcred, NULL);
+ (void) VOP_CLOSE(vp, 0, 1, 0, kcred, NULL);
+ }
+
+ return (error);
+#else
+ return (vn_remove(dp->scd_path, UIO_SYSSPACE, RMFILE));
+#endif
+}
+
static int
spa_config_write(spa_config_dirent_t *dp, nvlist_t *nvl)
{
@@ -161,7 +181,10 @@ spa_config_write(spa_config_dirent_t *dp, nvlist_t *nvl)
* If the nvlist is empty (NULL), then remove the old cachefile.
*/
if (nvl == NULL) {
- err = vn_remove(dp->scd_path, UIO_SYSSPACE, RMFILE);
+ err = spa_config_remove(dp);
+ if (err == ENOENT)
+ err = 0;
+
return (err);
}
@@ -174,9 +197,9 @@ spa_config_write(spa_config_dirent_t *dp, nvlist_t *nvl)
#if defined(__linux__) && defined(_KERNEL)
/*
* Write the configuration to disk. Due to the complexity involved
- * in performing a rename from within the kernel the file is truncated
- * and overwritten in place. In the event of an error the file is
- * unlinked to make sure we always have a consistent view of the data.
+ * in performing a rename and remove from within the kernel the file
+ * is instead truncated and overwritten in place. This way we always
+ * have a consistent view of the data or a zero length file.
*/
err = vn_open(dp->scd_path, UIO_SYSSPACE, oflags, 0644, &vp, 0, 0);
if (err == 0) {
@@ -186,9 +209,8 @@ spa_config_write(spa_config_dirent_t *dp, nvlist_t *nvl)
err = VOP_FSYNC(vp, FSYNC, kcred, NULL);
(void) VOP_CLOSE(vp, oflags, 1, 0, kcred, NULL);
-
if (err)
- (void) vn_remove(dp->scd_path, UIO_SYSSPACE, RMFILE);
+ (void) spa_config_remove(dp);
}
#else
/*
diff --git a/man/man8/zpool.8 b/man/man8/zpool.8
index 328ba3dce..22579101a 100644
--- a/man/man8/zpool.8
+++ b/man/man8/zpool.8
@@ -655,7 +655,7 @@ Because the kernel destroys and recreates this file when pools are added and
removed, care should be taken when attempting to access this file.
When the last pool using a
.Sy cachefile
-is exported or destroyed, the file is removed.
+is exported or destroyed, the file will be empty.
.It Sy comment Ns = Ns Ar text
A text string consisting of printable ASCII characters that will be stored
such that it is available even if the pool becomes faulted.
diff --git a/tests/zfs-tests/tests/functional/cachefile/cachefile_004_pos.ksh b/tests/zfs-tests/tests/functional/cachefile/cachefile_004_pos.ksh
index ae54a9365..e0b81e166 100755
--- a/tests/zfs-tests/tests/functional/cachefile/cachefile_004_pos.ksh
+++ b/tests/zfs-tests/tests/functional/cachefile/cachefile_004_pos.ksh
@@ -98,13 +98,13 @@ log_must zpool set cachefile=$CPATH2 $TESTPOOL1
log_must pool_in_cache $TESTPOOL1 $CPATH2
log_must zpool set cachefile=$CPATH2 $TESTPOOL2
log_must pool_in_cache $TESTPOOL2 $CPATH2
-if [[ -f $CPATH1 ]]; then
+if [[ -s $CPATH1 ]]; then
log_fail "Verify set when cachefile is set on pool."
fi
log_must zpool export $TESTPOOL1
log_must zpool export $TESTPOOL2
-if [[ -f $CPATH2 ]]; then
+if [[ -s $CPATH2 ]]; then
log_fail "Verify export when cachefile is set on pool."
fi
@@ -117,7 +117,7 @@ log_must pool_in_cache $TESTPOOL2 $CPATH2
log_must zpool destroy $TESTPOOL1
log_must zpool destroy $TESTPOOL2
-if [[ -f $CPATH2 ]]; then
+if [[ -s $CPATH2 ]]; then
log_fail "Verify destroy when cachefile is set on pool."
fi
--
2.14.2

102
zfs-patches/0012-Fix-with-systemd-on-Debian-based-distributions-6963.patch Normal file
View File

@ -0,0 +1,102 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: LOLi <loli10K@users.noreply.github.com>
Date: Sun, 17 Dec 2017 23:08:48 +0100
Subject: [PATCH] Fix --with-systemd on Debian-based distributions (#6963)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
These changes propagate the "--with-systemd" configure option to the
RPM spec file, allowing Debian-based distributions to package
systemd-related files.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes #6591
Closes #6963
(cherry picked from commit 03658d5081c64e14898cc9be45da3305b27fac9e)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
rpm/generic/zfs.spec.in | 2 +-
config/user-systemd.m4 | 20 ++++++++++++++++----
config/zfs-build.m4 | 2 +-
3 files changed, 18 insertions(+), 6 deletions(-)
diff --git a/rpm/generic/zfs.spec.in b/rpm/generic/zfs.spec.in
index 8df57fa46..4a911b4c2 100644
--- a/rpm/generic/zfs.spec.in
+++ b/rpm/generic/zfs.spec.in
@@ -87,11 +87,11 @@ BuildRequires: libblkid-devel
BuildRequires: libudev-devel
BuildRequires: libattr-devel
%endif
+
%if 0%{?_systemd}
Requires(post): systemd
Requires(preun): systemd
Requires(postun): systemd
-BuildRequires: systemd
%endif
# The zpool iostat/status -c scripts call some utilities like lsblk and iostat
diff --git a/config/user-systemd.m4 b/config/user-systemd.m4
index c2105abce..de2a44f10 100644
--- a/config/user-systemd.m4
+++ b/config/user-systemd.m4
@@ -2,7 +2,8 @@ AC_DEFUN([ZFS_AC_CONFIG_USER_SYSTEMD], [
AC_ARG_ENABLE(systemd,
AC_HELP_STRING([--enable-systemd],
[install systemd unit/preset files [[default: yes]]]),
- [],enable_systemd=yes)
+ [enable_systemd=$enableval],
+ [enable_systemd=check])
AC_ARG_WITH(systemdunitdir,
AC_HELP_STRING([--with-systemdunitdir=DIR],
@@ -19,16 +20,27 @@ AC_DEFUN([ZFS_AC_CONFIG_USER_SYSTEMD], [
[install systemd module load files into dir [[/usr/lib/modules-load.d]]]),
systemdmoduleloaddir=$withval,systemdmodulesloaddir=/usr/lib/modules-load.d)
+ AS_IF([test "x$enable_systemd" = xcheck], [
+ AS_IF([systemctl --version >/dev/null 2>&1],
+ [enable_systemd=yes],
+ [enable_systemd=no])
+ ])
- AS_IF([test "x$enable_systemd" = xyes],
- [
+ AC_MSG_CHECKING(for systemd support)
+ AC_MSG_RESULT([$enable_systemd])
+
+ AS_IF([test "x$enable_systemd" = xyes], [
ZFS_INIT_SYSTEMD=systemd
ZFS_MODULE_LOAD=modules-load.d
+ DEFINE_SYSTEMD='--with systemd --define "_unitdir $(systemdunitdir)" --define "_presetdir $(systemdpresetdir)"'
modulesloaddir=$systemdmodulesloaddir
- ])
+ ],[
+ DEFINE_SYSTEMD='--without systemd'
+ ])
AC_SUBST(ZFS_INIT_SYSTEMD)
AC_SUBST(ZFS_MODULE_LOAD)
+ AC_SUBST(DEFINE_SYSTEMD)
AC_SUBST(systemdunitdir)
AC_SUBST(systemdpresetdir)
AC_SUBST(modulesloaddir)
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index 17cc80462..5eaa49c87 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -160,7 +160,7 @@ AC_DEFUN([ZFS_AC_RPM], [
])
RPM_DEFINE_COMMON='--define "$(DEBUG_ZFS) 1"'
- RPM_DEFINE_UTIL='--define "_dracutdir $(dracutdir)" --define "_udevdir $(udevdir)" --define "_udevruledir $(udevruledir)" --define "_initconfdir $(DEFAULT_INITCONF_DIR)" $(DEFINE_INITRAMFS)'
+ RPM_DEFINE_UTIL='--define "_dracutdir $(dracutdir)" --define "_udevdir $(udevdir)" --define "_udevruledir $(udevruledir)" --define "_initconfdir $(DEFAULT_INITCONF_DIR)" $(DEFINE_INITRAMFS) $(DEFINE_SYSTEMD)'
RPM_DEFINE_KMOD='--define "kernels $(LINUX_VERSION)" --define "require_spldir $(SPL)" --define "require_splobj $(SPL_OBJ)" --define "ksrc $(LINUX)" --define "kobj $(LINUX_OBJ)"'
RPM_DEFINE_DKMS=
--
2.14.2

137
zfs-patches/0013-ZTS-Fix-create-o_ashift-test-case.patch Normal file
View File

@ -0,0 +1,137 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: LOLi <loli10K@users.noreply.github.com>
Date: Tue, 19 Dec 2017 19:49:33 +0100
Subject: [PATCH] ZTS: Fix create-o_ashift test case
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
The function that fills the uberblock ring buffer on every device label
has been reworked to avoid occasional failures caused by a race
condition that prevents 'zpool sync' from writing some uberblock
sequentially: this happens when the pool sync ioctl dispatch code calls
txg_wait_synced() while we're already waiting for a TXG to sync.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes #6924
Closes #6977
(cherry picked from commit 6c891ade8bee9c54484d5cf9b939582b7a9b7eeb)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zdb/zdb.c | 8 ++--
.../cli_root/zpool_create/create-o_ashift.ksh | 47 ++++++++++------------
2 files changed, 26 insertions(+), 29 deletions(-)
diff --git a/cmd/zdb/zdb.c b/cmd/zdb/zdb.c
index 1097501e8..442685486 100644
--- a/cmd/zdb/zdb.c
+++ b/cmd/zdb/zdb.c
@@ -2716,10 +2716,6 @@ dump_label(const char *dev)
exit(1);
}
- if (ioctl(fd, BLKFLSBUF) != 0)
- (void) printf("failed to invalidate cache '%s' : %s\n", path,
- strerror(errno));
-
if (fstat64_blk(fd, &statbuf) != 0) {
(void) printf("failed to stat '%s': %s\n", path,
strerror(errno));
@@ -2727,6 +2723,10 @@ dump_label(const char *dev)
exit(1);
}
+ if (S_ISBLK(statbuf.st_mode) && ioctl(fd, BLKFLSBUF) != 0)
+ (void) printf("failed to invalidate cache '%s' : %s\n", path,
+ strerror(errno));
+
avl_create(&config_tree, cksum_record_compare,
sizeof (cksum_record_t), offsetof(cksum_record_t, link));
avl_create(&uberblock_tree, cksum_record_compare,
diff --git a/tests/zfs-tests/tests/functional/cli_root/zpool_create/create-o_ashift.ksh b/tests/zfs-tests/tests/functional/cli_root/zpool_create/create-o_ashift.ksh
index 6449c8a91..6a9c3e28c 100755
--- a/tests/zfs-tests/tests/functional/cli_root/zpool_create/create-o_ashift.ksh
+++ b/tests/zfs-tests/tests/functional/cli_root/zpool_create/create-o_ashift.ksh
@@ -44,47 +44,45 @@ verify_runnable "global"
function cleanup
{
- poolexists $TESTPOOL && destroy_pool $TESTPOOL
+ destroy_pool $TESTPOOL
log_must rm -f $disk
}
#
-# Commit the specified number of TXGs to the provided pool
-# We use 'zpool sync' here because we can't force it via sync(1) like on illumos
-# $1 pool name
-# $2 number of txg syncs
+# Fill the uberblock ring in every <device> label: we do this by committing
+# TXGs to the provided <pool> until every slot contains a valid uberblock.
+# NOTE: We use 'zpool sync' here because we can't force it via sync(1) like on
+# illumos
#
-function txg_sync
+function write_device_uberblocks # <device> <pool>
{
- typeset pool=$1
- typeset -i count=$2
- typeset -i i=0;
+ typeset device=$1
+ typeset pool=$2
- while [ $i -lt $count ]
+ while [ "$(zdb -quuul $device | grep -c 'invalid')" -ne 0 ]
do
- log_must sync_pool $pool true
- ((i = i + 1))
+ sync_pool $pool true
done
}
#
-# Verify device $1 labels contains $2 valid uberblocks in every label
-# $1 device
-# $2 uberblocks count
+# Verify every label on <device> contains <count> (valid) uberblocks
#
-function verify_device_uberblocks
+function verify_device_uberblocks # <device> <count>
{
typeset device=$1
typeset ubcount=$2
zdb -quuul $device | egrep '^(\s+)?Uberblock' |
- egrep -v 'invalid$' | awk \
- -v ubcount=$ubcount '{ uberblocks[$0]++; }
- END { for (i in uberblocks) {
- count++;
- if (uberblocks[i] != 4) { exit 1; }
- }
- if (count != ubcount) { exit 1; } }'
+ awk -v ubcount=$ubcount 'BEGIN { count=0 } { uberblocks[$0]++; }
+ END {
+ for (i in uberblocks) {
+ if (i ~ /invalid/) { continue; }
+ if (uberblocks[i] != 4) { exit 1; }
+ count++;
+ }
+ if (count != ubcount) { exit 1; }
+ }'
return $?
}
@@ -110,8 +108,7 @@ do
log_fail "Pool was created without setting ashift value to "\
"$ashift (current = $pprop)"
fi
- # force 128 txg sync to fill the uberblock ring
- txg_sync $TESTPOOL 128
+ write_device_uberblocks $disk $TESTPOOL
verify_device_uberblocks $disk ${ubcount[$i]}
if [[ $? -ne 0 ]]
then
--
2.14.2

49
zfs-patches/0014-Emit-an-error-message-before-MMP-suspends-pool.patch Normal file
View File

@ -0,0 +1,49 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: "John L. Hammond" <35266395+jhammond-intel@users.noreply.github.com>
Date: Wed, 17 Jan 2018 14:24:42 -0600
Subject: [PATCH] Emit an error message before MMP suspends pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
In mmp_thread(), emit an MMP specific error message before calling
zio_suspend() so that the administrator will understand why the pool
is being suspended.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: John L. Hammond <john.hammond@intel.com>
Closes #7048
(cherry picked from commit ecc972c7f009e1fa75900e276a4c1306c55b5722)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/mmp.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index 6f2aa3f59..e91ae628a 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -26,6 +26,7 @@
#include <sys/mmp.h>
#include <sys/spa.h>
#include <sys/spa_impl.h>
+#include <sys/time.h>
#include <sys/vdev.h>
#include <sys/vdev_impl.h>
#include <sys/zfs_context.h>
@@ -428,6 +429,10 @@ mmp_thread(spa_t *spa)
*/
if (!suspended && mmp_fail_intervals && multihost &&
(start - mmp->mmp_last_write) > max_fail_ns) {
+ cmn_err(CE_WARN, "MMP writes to pool '%s' have not "
+ "succeeded in over %llus; suspending pool",
+ spa_name(spa),
+ NSEC2SEC(start - mmp->mmp_last_write));
zio_suspend(spa, NULL);
}
--
2.14.2

85
zfs-patches/0015-Fix-systemd_-RPM-macros-usage-on-Debian-based-distri.patch Normal file
View File

@ -0,0 +1,85 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: LOLi <loli10K@users.noreply.github.com>
Date: Fri, 2 Feb 2018 22:50:42 +0100
Subject: [PATCH] Fix systemd_ RPM macros usage on Debian-based distributions
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Debian-based distributions do not seem to provide RPM macros for
dealing with systemd pre- and post- (un)install actions: this results
in errors when installing or upgrading .deb packages because the
resulting control scripts contain the following unresolved macros:
* %systemd_post
* %systemd_preun
* %systemd_postun
Fix this by providing default values for postinstall, preuninstall and
postuninstall scripts when these macros are not defined.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes #7074
Closes #7100
(cherry picked from commit 29b79dcfe90a67a919c2f42fbaa2e557ea484bd5)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
rpm/generic/zfs.spec.in | 21 ++++++++++++++++++++-
1 file changed, 20 insertions(+), 1 deletion(-)
diff --git a/rpm/generic/zfs.spec.in b/rpm/generic/zfs.spec.in
index 4a911b4c2..ce028794c 100644
--- a/rpm/generic/zfs.spec.in
+++ b/rpm/generic/zfs.spec.in
@@ -245,8 +245,15 @@ find %{?buildroot}%{_libdir} -name '*.la' -exec rm -f {} \;
%post
%if 0%{?_systemd}
+%if 0%{?systemd_post:1}
%systemd_post %{systemd_svcs}
%else
+if [ "$1" = "1" -o "$1" = "install" ] ; then
+ # Initial installation
+ systemctl preset %{systemd_svcs} >/dev/null || true
+fi
+%endif
+%else
if [ -x /sbin/chkconfig ]; then
/sbin/chkconfig --add zfs-import
/sbin/chkconfig --add zfs-mount
@@ -258,9 +265,17 @@ exit 0
%preun
%if 0%{?_systemd}
+%if 0%{?systemd_preun:1}
%systemd_preun %{systemd_svcs}
%else
-if [ "$1" = "0" ] && [ -x /sbin/chkconfig ]; then
+if [ "$1" = "0" -o "$1" = "remove" ] ; then
+ # Package removal, not upgrade
+ systemctl --no-reload disable %{systemd_svcs} >/dev/null || true
+ systemctl stop %{systemd_svcs} >/dev/null || true
+fi
+%endif
+%else
+if [ "$1" = "0" -o "$1" = "remove" ] && [ -x /sbin/chkconfig ]; then
/sbin/chkconfig --del zfs-import
/sbin/chkconfig --del zfs-mount
/sbin/chkconfig --del zfs-share
@@ -271,7 +286,11 @@ exit 0
%postun
%if 0%{?_systemd}
+%if 0%{?systemd_postun:1}
%systemd_postun %{systemd_svcs}
+%else
+systemctl --system daemon-reload >/dev/null || true
+%endif
%endif
%files
--
2.14.2

39
zfs-patches/0016-Bug-fix-in-qat_compress.c-for-vmalloc-addr-check.patch Normal file
View File

@ -0,0 +1,39 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: wli5 <weigang.li@intel.com>
Date: Tue, 6 Feb 2018 02:26:27 +0800
Subject: [PATCH] Bug fix in qat_compress.c for vmalloc addr check
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Remove the unused vmalloc address check, and function mem_to_page
will handle the non-vmalloc address when map it to a physical
address.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Weigang Li <weigang.li@intel.com>
Closes #7125
(cherry picked from commit 5f38142e7b3a8994afef60faa2e65d72e7993807)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/qat_compress.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/module/zfs/qat_compress.c b/module/zfs/qat_compress.c
index 4d17d7ac9..62655f56d 100644
--- a/module/zfs/qat_compress.c
+++ b/module/zfs/qat_compress.c
@@ -364,10 +364,6 @@ qat_compress(qat_compress_dir_t dir, char *src, int src_len,
Cpa32U dst_buffer_list_mem_size = sizeof (CpaBufferList) +
(num_dst_buf * sizeof (CpaFlatBuffer));
- if (!is_vmalloc_addr(src) || !is_vmalloc_addr(src + src_len - 1) ||
- !is_vmalloc_addr(dst) || !is_vmalloc_addr(dst + dst_len - 1))
- return (-1);
-
if (PHYS_CONTIG_ALLOC(&in_pages,
num_src_buf * sizeof (struct page *)) != CPA_STATUS_SUCCESS)
goto fail;
--
2.14.2

52
zfs-patches/0017-Fix-default-libdir-for-Debian-Ubuntu.patch Normal file
View File

@ -0,0 +1,52 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Mon, 5 Feb 2018 20:42:52 -0800
Subject: [PATCH] Fix default libdir for Debian/Ubuntu
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
The distribution provided architecture specific RPM macro files
for x86_64 and other architectures on Debian/Ubuntu specify the
wrong default libdir install location. When building deb packages
override _lib with the correct location.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #7083
Closes #7101
(cherry picked from commit f1dde3fb20eb27c05b57ad82e0905a587ced8ee8)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
config/zfs-build.m4 | 13 +++++++++++++
1 file changed, 13 insertions(+)
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index 5eaa49c87..7c19cecd6 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -164,6 +164,19 @@ AC_DEFUN([ZFS_AC_RPM], [
RPM_DEFINE_KMOD='--define "kernels $(LINUX_VERSION)" --define "require_spldir $(SPL)" --define "require_splobj $(SPL_OBJ)" --define "ksrc $(LINUX)" --define "kobj $(LINUX_OBJ)"'
RPM_DEFINE_DKMS=
+ dnl # Override default lib directory on Debian/Ubuntu systems. The provided
+ dnl # /usr/lib/rpm/platform/<arch>/macros files do not specify the correct
+ dnl # path for multiarch systems as described by the packaging guidelines.
+ dnl #
+ dnl # https://wiki.ubuntu.com/MultiarchSpec
+ dnl # https://wiki.debian.org/Multiarch/Implementation
+ dnl #
+ AS_IF([test "$DEFAULT_PACKAGE" = "deb"], [
+ MULTIARCH_LIBDIR="lib/$(dpkg-architecture -qDEB_HOST_MULTIARCH)"
+ RPM_DEFINE_UTIL+=' --define "_lib $(MULTIARCH_LIBDIR)"'
+ AC_SUBST(MULTIARCH_LIBDIR)
+ ])
+
SRPM_DEFINE_COMMON='--define "build_src_rpm 1"'
SRPM_DEFINE_UTIL=
SRPM_DEFINE_KMOD=
--
2.14.2

71
zfs-patches/0018-Remove-deprecated-zfs_arc_p_aggressive_disable.patch Normal file
View File

@ -0,0 +1,71 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Richard Elling <Richard.Elling@RichardElling.com>
Date: Wed, 7 Feb 2018 11:54:20 -0800
Subject: [PATCH] Remove deprecated zfs_arc_p_aggressive_disable
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
zfs_arc_p_aggressive_disable is no more. This PR removes docs
and module parameters for zfs_arc_p_aggressive_disable.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Richard Elling <Richard.Elling@RichardElling.com>
Closes #7135
(cherry picked from commit a58e1284d8998b0d7e409a6ec009d9d372aad41b)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/arc.c | 4 ----
man/man5/zfs-module-parameters.5 | 11 -----------
2 files changed, 15 deletions(-)
diff --git a/module/zfs/arc.c b/module/zfs/arc.c
index 264e67735..9d1d0db1d 100644
--- a/module/zfs/arc.c
+++ b/module/zfs/arc.c
@@ -391,7 +391,6 @@ unsigned long zfs_arc_dnode_limit_percent = 10;
*/
unsigned long zfs_arc_sys_free = 0;
int zfs_arc_min_prefetch_lifespan = 0;
-int zfs_arc_p_aggressive_disable = 1;
int zfs_arc_p_dampener_disable = 1;
int zfs_arc_meta_prune = 10000;
int zfs_arc_meta_strategy = ARC_STRATEGY_META_BALANCED;
@@ -7928,9 +7927,6 @@ MODULE_PARM_DESC(zfs_arc_meta_strategy, "Meta reclaim strategy");
module_param(zfs_arc_grow_retry, int, 0644);
MODULE_PARM_DESC(zfs_arc_grow_retry, "Seconds before growing arc size");
-module_param(zfs_arc_p_aggressive_disable, int, 0644);
-MODULE_PARM_DESC(zfs_arc_p_aggressive_disable, "disable aggressive arc_p grow");
-
module_param(zfs_arc_p_dampener_disable, int, 0644);
MODULE_PARM_DESC(zfs_arc_p_dampener_disable, "disable arc_p adapt dampener");
diff --git a/man/man5/zfs-module-parameters.5 b/man/man5/zfs-module-parameters.5
index 19e6becfd..d4daffde6 100644
--- a/man/man5/zfs-module-parameters.5
+++ b/man/man5/zfs-module-parameters.5
@@ -674,17 +674,6 @@ max arc_p
Default value: \fB0\fR.
.RE
-.sp
-.ne 2
-.na
-\fBzfs_arc_p_aggressive_disable\fR (int)
-.ad
-.RS 12n
-Disable aggressive arc_p growth
-.sp
-Use \fB1\fR for yes (default) and \fB0\fR to disable.
-.RE
-
.sp
.ne 2
.na
--
2.14.2

51
zfs-patches/0019-OpenZFS-8966-Source-file-zfs_acl.c-function-zfs_acls.patch Normal file
View File

@ -0,0 +1,51 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: WHR <msl0000023508@gmail.com>
Date: Sun, 14 Jan 2018 23:57:54 +0300
Subject: [PATCH] OpenZFS 8966 - Source file zfs_acl.c, function
zfs_aclset_common contains a use after end of the lifetime of a local
variable
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Authored by: WHR <msl0000023508@gmail.com>
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed by: Andriy Gapon <avg@FreeBSD.org>
Reviewed by: George Melikov <mail@gmelikov.ru>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Richard Lowe <richlowe@richlowe.net>
Ported-by: Giuseppe Di Natale <dinatale2@llnl.gov>
OpenZFS-issue: https://www.illumos.org/issues/8966
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/c95549fcdc
Closes #7141
(cherry picked from commit a196b3bc3dcc14bda32b2fbed40d5e436319aca6)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/zfs_acl.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/module/zfs/zfs_acl.c b/module/zfs/zfs_acl.c
index 7ddedeaaf..1fcfca0c7 100644
--- a/module/zfs/zfs_acl.c
+++ b/module/zfs/zfs_acl.c
@@ -1323,6 +1323,7 @@ zfs_aclset_common(znode_t *zp, zfs_acl_t *aclp, cred_t *cr, dmu_tx_t *tx)
sa_bulk_attr_t bulk[5];
uint64_t ctime[2];
int count = 0;
+ zfs_acl_phys_t acl_phys;
mode = zp->z_mode;
@@ -1369,7 +1370,6 @@ zfs_aclset_common(znode_t *zp, zfs_acl_t *aclp, cred_t *cr, dmu_tx_t *tx)
} else { /* Painful legacy way */
zfs_acl_node_t *aclnode;
uint64_t off = 0;
- zfs_acl_phys_t acl_phys;
uint64_t aoid;
if ((error = sa_lookup(zp->z_sa_hdl, SA_ZPL_ZNODE_ACL(zfsvfs),
--
2.14.2

145
zfs-patches/0020-Linux-4.16-compat-inode_set_iversion.patch Normal file
View File

@ -0,0 +1,145 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Thu, 8 Feb 2018 14:27:59 -0800
Subject: [PATCH] Linux 4.16 compat: inode_set_iversion()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
A new interface was added to manipulate the version field of an
inode. Add a inode_set_iversion() wrapper for older kernels and
use the new interface when available.
The i_version field was dropped from the trace point due to the
switch to an atomic64_t i_version type.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Chunwei Chen <david.chen@nutanix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #7148
(cherry picked from commit 310e63dfd18f59ad583631dfa2f55d40cedf1415)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/linux/vfs_compat.h | 14 ++++++++++++++
include/sys/trace_acl.h | 6 ++----
module/zfs/zpl_super.c | 2 +-
config/kernel-inode-set-iversion.m4 | 19 +++++++++++++++++++
config/kernel.m4 | 1 +
5 files changed, 37 insertions(+), 5 deletions(-)
create mode 100644 config/kernel-inode-set-iversion.m4
diff --git a/include/linux/vfs_compat.h b/include/linux/vfs_compat.h
index 6111f0afc..f51ff887d 100644
--- a/include/linux/vfs_compat.h
+++ b/include/linux/vfs_compat.h
@@ -578,4 +578,18 @@ current_time(struct inode *ip)
}
#endif
+/*
+ * 4.16 API change
+ * Added iversion interface for managing inode version field.
+ */
+#ifdef HAVE_INODE_SET_IVERSION
+#include <linux/iversion.h>
+#else
+static inline void
+inode_set_iversion(struct inode *ip, u64 val)
+{
+ ip->i_version = val;
+}
+#endif
+
#endif /* _ZFS_VFS_H */
diff --git a/include/sys/trace_acl.h b/include/sys/trace_acl.h
index 1057e560b..610bbe29c 100644
--- a/include/sys/trace_acl.h
+++ b/include/sys/trace_acl.h
@@ -68,7 +68,6 @@ DECLARE_EVENT_CLASS(zfs_ace_class,
__field(uint32_t, i_gid)
__field(unsigned long, i_ino)
__field(unsigned int, i_nlink)
- __field(u64, i_version)
__field(loff_t, i_size)
__field(unsigned int, i_blkbits)
__field(unsigned short, i_bytes)
@@ -103,7 +102,6 @@ DECLARE_EVENT_CLASS(zfs_ace_class,
__entry->i_gid = KGID_TO_SGID(ZTOI(zn)->i_gid);
__entry->i_ino = zn->z_inode.i_ino;
__entry->i_nlink = zn->z_inode.i_nlink;
- __entry->i_version = zn->z_inode.i_version;
__entry->i_size = zn->z_inode.i_size;
__entry->i_blkbits = zn->z_inode.i_blkbits;
__entry->i_bytes = zn->z_inode.i_bytes;
@@ -121,7 +119,7 @@ DECLARE_EVENT_CLASS(zfs_ace_class,
"mapcnt %llu size %llu pflags %llu "
"sync_cnt %u mode 0x%x is_sa %d "
"is_mapped %d is_ctldir %d is_stale %d inode { "
- "uid %u gid %u ino %lu nlink %u version %llu size %lli "
+ "uid %u gid %u ino %lu nlink %u size %lli "
"blkbits %u bytes %u mode 0x%x generation %x } } "
"ace { type %u flags %u access_mask %u } mask_matched %u",
__entry->z_id, __entry->z_unlinked, __entry->z_atime_dirty,
@@ -131,7 +129,7 @@ DECLARE_EVENT_CLASS(zfs_ace_class,
__entry->z_is_sa, __entry->z_is_mapped,
__entry->z_is_ctldir, __entry->z_is_stale, __entry->i_uid,
__entry->i_gid, __entry->i_ino, __entry->i_nlink,
- __entry->i_version, __entry->i_size, __entry->i_blkbits,
+ __entry->i_size, __entry->i_blkbits,
__entry->i_bytes, __entry->i_mode, __entry->i_generation,
__entry->z_type, __entry->z_flags, __entry->z_access_mask,
__entry->mask_matched)
diff --git a/module/zfs/zpl_super.c b/module/zfs/zpl_super.c
index b6ef60277..25e75a897 100644
--- a/module/zfs/zpl_super.c
+++ b/module/zfs/zpl_super.c
@@ -36,7 +36,7 @@ zpl_inode_alloc(struct super_block *sb)
struct inode *ip;
VERIFY3S(zfs_inode_alloc(sb, &ip), ==, 0);
- ip->i_version = 1;
+ inode_set_iversion(ip, 1);
return (ip);
}
diff --git a/config/kernel-inode-set-iversion.m4 b/config/kernel-inode-set-iversion.m4
new file mode 100644
index 000000000..9a7d7890e
--- /dev/null
+++ b/config/kernel-inode-set-iversion.m4
@@ -0,0 +1,19 @@
+dnl #
+dnl # 4.16 API change
+dnl # inode_set_iversion introduced to set i_version
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_INODE_SET_IVERSION], [
+ AC_MSG_CHECKING([whether inode_set_iversion() exists])
+ ZFS_LINUX_TRY_COMPILE([
+ #include <linux/iversion.h>
+ ],[
+ struct inode inode;
+ inode_set_iversion(&inode, 1);
+ ],[
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(HAVE_INODE_SET_IVERSION, 1,
+ [inode_set_iversion() exists])
+ ],[
+ AC_MSG_RESULT(no)
+ ])
+])
diff --git a/config/kernel.m4 b/config/kernel.m4
index b759ccd39..b83f021e8 100644
--- a/config/kernel.m4
+++ b/config/kernel.m4
@@ -65,6 +65,7 @@ AC_DEFUN([ZFS_AC_CONFIG_KERNEL], [
ZFS_AC_KERNEL_INODE_OPERATIONS_SET_ACL
ZFS_AC_KERNEL_INODE_OPERATIONS_GETATTR
ZFS_AC_KERNEL_INODE_SET_FLAGS
+ ZFS_AC_KERNEL_INODE_SET_IVERSION
ZFS_AC_KERNEL_GET_ACL_HANDLE_CACHE
ZFS_AC_KERNEL_SHOW_OPTIONS
ZFS_AC_KERNEL_FILE_INODE
--
2.14.2

53
zfs-patches/0021-Linux-4.11-compat-avoid-refcount_t-name-conflict.patch Normal file
View File

@ -0,0 +1,53 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Thu, 8 Feb 2018 14:31:08 -0800
Subject: [PATCH] Linux 4.11 compat: avoid refcount_t name conflict
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Related to commit 4859fe796, when directly using the kernel's
refcount functions in kernel compatibility code do not map
refcount_t to zfs_refcount_t. This leads to a type mismatch.
Longer term we should consider renaming refcount_t to
zfs_refcount_t in the zfs code base.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Chunwei Chen <david.chen@nutanix.com>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #7148
(cherry picked from commit 3713b7333532f221733f7b74189d996359ed4311)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/linux/vfs_compat.h | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/include/linux/vfs_compat.h b/include/linux/vfs_compat.h
index f51ff887d..7fcf3c055 100644
--- a/include/linux/vfs_compat.h
+++ b/include/linux/vfs_compat.h
@@ -272,6 +272,10 @@ lseek_execute(
* This is several orders of magnitude larger than expected grace period.
* At 60 seconds the kernel will also begin issuing RCU stall warnings.
*/
+#ifdef refcount_t
+#undef refcount_t
+#endif
+
#include <linux/posix_acl.h>
#if defined(HAVE_POSIX_ACL_RELEASE) && !defined(HAVE_POSIX_ACL_RELEASE_GPL_ONLY)
@@ -397,6 +401,8 @@ typedef mode_t zpl_equivmode_t;
#define zpl_posix_acl_valid(ip, acl) posix_acl_valid(acl)
#endif
+#define refcount_t zfs_refcount_t
+
#endif /* CONFIG_FS_POSIX_ACL */
/*
--
2.14.2

116
zfs-patches/0022-Fix-zdb-c-traverse-stop-on-damaged-objset-root.patch Normal file
View File

@ -0,0 +1,116 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Chunwei Chen <david.chen@nutanix.com>
Date: Tue, 30 Jan 2018 13:39:11 -0800
Subject: [PATCH] Fix zdb -c traverse stop on damaged objset root
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
If a corruption happens to be on a root block of an objset, zdb -c will
not correctly report the error, and it will not traverse the datasets
that come after. This is because traverse_visitbp, which does the
callback and reset error for TRAVERSE_HARD, is skipped when traversing
zil is failed in traverse_impl.
Here's example of what 'zdb -eLcc' command looks like on a pool with
damaged objset root:
== before patch:
Traversing all blocks to verify checksums ...
Error counts:
errno count
block traversal size 379392 != alloc 33987072 (unreachable 33607680)
bp count: 172
ganged count: 0
bp logical: 1678336 avg: 9757
bp physical: 130560 avg: 759 compression: 12.85
bp allocated: 379392 avg: 2205 compression: 4.42
bp deduped: 0 ref>1: 0 deduplication: 1.00
SPA allocated: 33987072 used: 0.80%
additional, non-pointer bps of type 0: 71
Dittoed blocks on same vdev: 101
== after patch:
Traversing all blocks to verify checksums ...
zdb_blkptr_cb: Got error 52 reading <54, 0, -1, 0> -- skipping
Error counts:
errno count
52 1
block traversal size 33963520 != alloc 33987072 (unreachable 23552)
bp count: 447
ganged count: 0
bp logical: 36093440 avg: 80745
bp physical: 33699840 avg: 75391 compression: 1.07
bp allocated: 33963520 avg: 75981 compression: 1.06
bp deduped: 0 ref>1: 0 deduplication: 1.00
SPA allocated: 33987072 used: 0.80%
additional, non-pointer bps of type 0: 76
Dittoed blocks on same vdev: 115
==
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes #7099
(cherry picked from commit 23227313a2016449176cbfcbae2d4fc463a2bc09)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/dmu_traverse.c | 26 +++++++++++++++++---------
1 file changed, 17 insertions(+), 9 deletions(-)
diff --git a/module/zfs/dmu_traverse.c b/module/zfs/dmu_traverse.c
index c78228d74..b494bef35 100644
--- a/module/zfs/dmu_traverse.c
+++ b/module/zfs/dmu_traverse.c
@@ -599,19 +599,27 @@ traverse_impl(spa_t *spa, dsl_dataset_t *ds, uint64_t objset, blkptr_t *rootbp,
/* See comment on ZIL traversal in dsl_scan_visitds. */
if (ds != NULL && !ds->ds_is_snapshot && !BP_IS_HOLE(rootbp)) {
+ enum zio_flag zio_flags = ZIO_FLAG_CANFAIL;
uint32_t flags = ARC_FLAG_WAIT;
objset_phys_t *osp;
arc_buf_t *buf;
- err = arc_read(NULL, td->td_spa, rootbp,
- arc_getbuf_func, &buf,
- ZIO_PRIORITY_ASYNC_READ, ZIO_FLAG_CANFAIL, &flags, czb);
- if (err != 0)
- return (err);
-
- osp = buf->b_data;
- traverse_zil(td, &osp->os_zil_header);
- arc_buf_destroy(buf, &buf);
+ err = arc_read(NULL, td->td_spa, rootbp, arc_getbuf_func,
+ &buf, ZIO_PRIORITY_ASYNC_READ, zio_flags, &flags, czb);
+ if (err != 0) {
+ /*
+ * If both TRAVERSE_HARD and TRAVERSE_PRE are set,
+ * continue to visitbp so that td_func can be called
+ * in pre stage, and err will reset to zero.
+ */
+ if (!(td->td_flags & TRAVERSE_HARD) ||
+ !(td->td_flags & TRAVERSE_PRE))
+ return (err);
+ } else {
+ osp = buf->b_data;
+ traverse_zil(td, &osp->os_zil_header);
+ arc_buf_destroy(buf, &buf);
+ }
}
if (!(flags & TRAVERSE_PREFETCH_DATA) ||
--
2.14.2

40
zfs-patches/0023-Fix-zle_decompress-out-of-bound-access.patch Normal file
View File

@ -0,0 +1,40 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Chunwei Chen <david.chen@nutanix.com>
Date: Thu, 1 Feb 2018 15:41:05 -0800
Subject: [PATCH] Fix zle_decompress out of bound access
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes #7099
(cherry picked from commit 5e566c57726226ceeca09b1eb19cb1c373622763)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/zle.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/module/zfs/zle.c b/module/zfs/zle.c
index 13c5673fb..613607faa 100644
--- a/module/zfs/zle.c
+++ b/module/zfs/zle.c
@@ -74,10 +74,14 @@ zle_decompress(void *s_start, void *d_start, size_t s_len, size_t d_len, int n)
while (src < s_end && dst < d_end) {
int len = 1 + *src++;
if (len <= n) {
+ if (src + len > s_end || dst + len > d_end)
+ return (-1);
while (len-- != 0)
*dst++ = *src++;
} else {
len -= n;
+ if (dst + len > d_end)
+ return (-1);
while (len-- != 0)
*dst++ = 0;
}
--
2.14.2

59
zfs-patches/0024-Fix-racy-assignment-of-zcb.zcb_haderrors.patch Normal file
View File

@ -0,0 +1,59 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Chunwei Chen <david.chen@nutanix.com>
Date: Thu, 1 Feb 2018 15:42:41 -0800
Subject: [PATCH] Fix racy assignment of zcb.zcb_haderrors
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
zcb_haderrors will be modified in zdb_blkptr_done, which is
asynchronous. So we must move this assignment after zio_wait.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes #7099
(cherry picked from commit c797f0898ec11669427e2fc481118bc2e2d8ef3f)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zdb/zdb.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/cmd/zdb/zdb.c b/cmd/zdb/zdb.c
index 442685486..0cc1656a8 100644
--- a/cmd/zdb/zdb.c
+++ b/cmd/zdb/zdb.c
@@ -3313,7 +3313,7 @@ dump_block_stats(spa_t *spa)
uint64_t norm_alloc, norm_space, total_alloc, total_found;
int flags = TRAVERSE_PRE | TRAVERSE_PREFETCH_METADATA | TRAVERSE_HARD;
boolean_t leaks = B_FALSE;
- int e, c;
+ int e, c, err;
bp_embedded_type_t i;
(void) printf("\nTraversing all blocks %s%s%s%s%s...\n\n",
@@ -3354,7 +3354,7 @@ dump_block_stats(spa_t *spa)
zcb.zcb_totalasize = metaslab_class_get_alloc(spa_normal_class(spa));
zcb.zcb_start = zcb.zcb_lastprint = gethrtime();
- zcb.zcb_haderrors |= traverse_pool(spa, 0, flags, zdb_blkptr_cb, &zcb);
+ err = traverse_pool(spa, 0, flags, zdb_blkptr_cb, &zcb);
/*
* If we've traversed the data blocks then we need to wait for those
@@ -3370,6 +3370,12 @@ dump_block_stats(spa_t *spa)
}
}
+ /*
+ * Done after zio_wait() since zcb_haderrors is modified in
+ * zdb_blkptr_done()
+ */
+ zcb.zcb_haderrors |= err;
+
if (zcb.zcb_haderrors) {
(void) printf("\nError counts:\n\n");
(void) printf("\t%5s %s\n", "errno", "count");
--
2.14.2

130
zfs-patches/0025-Fix-zdb-R-decompression.patch Normal file
View File

@ -0,0 +1,130 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Chunwei Chen <david.chen@nutanix.com>
Date: Thu, 1 Feb 2018 16:19:36 -0800
Subject: [PATCH] Fix zdb -R decompression
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
There are some issues in the zdb -R decompression implementation.
The first is that ZLE can easily decompress non-ZLE streams. So we add
ZDB_NO_ZLE env to make zdb skip ZLE.
The second is the random bytes appended to pabd, pbuf2 stuff. This serve
no purpose at all, those bytes shouldn't be read during decompression
anyway. Instead, we randomize lbuf2, so that we can make sure
decompression fill exactly to lsize by bcmp lbuf and lbuf2.
The last one is the condition to detect fail is wrong.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes #7099
Closes #4984
(cherry picked from commit 18c662b84566cd34e6f6fb982d6a01a415a4e3cd)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zdb/zdb.c | 38 ++++++++++++++++++--------------------
man/man8/zdb.8 | 4 +++-
2 files changed, 21 insertions(+), 21 deletions(-)
diff --git a/cmd/zdb/zdb.c b/cmd/zdb/zdb.c
index 0cc1656a8..2d80589ca 100644
--- a/cmd/zdb/zdb.c
+++ b/cmd/zdb/zdb.c
@@ -3895,13 +3895,6 @@ name:
return (NULL);
}
-/* ARGSUSED */
-static int
-random_get_pseudo_bytes_cb(void *buf, size_t len, void *unused)
-{
- return (random_get_pseudo_bytes(buf, len));
-}
-
/*
* Read a block from a pool and print it out. The syntax of the
* block descriptor is:
@@ -4064,17 +4057,8 @@ zdb_read_block(char *thing, spa_t *spa)
* every decompress function at every inflated blocksize.
*/
enum zio_compress c;
- void *pbuf2 = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL);
void *lbuf2 = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL);
- abd_copy_to_buf(pbuf2, pabd, psize);
-
- VERIFY0(abd_iterate_func(pabd, psize, SPA_MAXBLOCKSIZE - psize,
- random_get_pseudo_bytes_cb, NULL));
-
- VERIFY0(random_get_pseudo_bytes((uint8_t *)pbuf2 + psize,
- SPA_MAXBLOCKSIZE - psize));
-
/*
* XXX - On the one hand, with SPA_MAXBLOCKSIZE at 16MB,
* this could take a while and we should let the user know
@@ -4084,13 +4068,29 @@ zdb_read_block(char *thing, spa_t *spa)
for (lsize = psize + SPA_MINBLOCKSIZE;
lsize <= SPA_MAXBLOCKSIZE; lsize += SPA_MINBLOCKSIZE) {
for (c = 0; c < ZIO_COMPRESS_FUNCTIONS; c++) {
+ /*
+ * ZLE can easily decompress non zle stream.
+ * So have an option to disable it.
+ */
+ if (c == ZIO_COMPRESS_ZLE &&
+ getenv("ZDB_NO_ZLE"))
+ continue;
+
(void) fprintf(stderr,
"Trying %05llx -> %05llx (%s)\n",
(u_longlong_t)psize, (u_longlong_t)lsize,
zio_compress_table[c].ci_name);
+
+ /*
+ * We randomize lbuf2, and decompress to both
+ * lbuf and lbuf2. This way, we will know if
+ * decompression fill exactly to lsize.
+ */
+ VERIFY0(random_get_pseudo_bytes(lbuf2, lsize));
+
if (zio_decompress_data(c, pabd,
lbuf, psize, lsize) == 0 &&
- zio_decompress_data_buf(c, pbuf2,
+ zio_decompress_data(c, pabd,
lbuf2, psize, lsize) == 0 &&
bcmp(lbuf, lbuf2, lsize) == 0)
break;
@@ -4098,11 +4098,9 @@ zdb_read_block(char *thing, spa_t *spa)
if (c != ZIO_COMPRESS_FUNCTIONS)
break;
}
-
- umem_free(pbuf2, SPA_MAXBLOCKSIZE);
umem_free(lbuf2, SPA_MAXBLOCKSIZE);
- if (lsize <= psize) {
+ if (lsize > SPA_MAXBLOCKSIZE) {
(void) printf("Decompress of %s failed\n", thing);
goto out;
}
diff --git a/man/man8/zdb.8 b/man/man8/zdb.8
index 4e47de7be..d991aae4c 100644
--- a/man/man8/zdb.8
+++ b/man/man8/zdb.8
@@ -246,7 +246,9 @@ and, optionally,
.It Sy b Ar offset
Print block pointer
.It Sy d
-Decompress the block
+Decompress the block. Set environment variable
+.Nm ZBD_NO_ZLE
+to skip zle when guessing.
.It Sy e
Byte swap the block
.It Sy g
--
2.14.2

50
zfs-patches/0026-Fix-zdb-E-segfault.patch Normal file
View File

@ -0,0 +1,50 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Chunwei Chen <david.chen@nutanix.com>
Date: Thu, 1 Feb 2018 16:28:11 -0800
Subject: [PATCH] Fix zdb -E segfault
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
SPA_MAXBLOCKSIZE is too large for stack.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes #7099
(cherry picked from commit 31ff122aa2e20c7ed48617868085ddba7b4ad174)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zdb/zdb.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/cmd/zdb/zdb.c b/cmd/zdb/zdb.c
index 2d80589ca..90847d8d9 100644
--- a/cmd/zdb/zdb.c
+++ b/cmd/zdb/zdb.c
@@ -4139,11 +4139,12 @@ zdb_embedded_block(char *thing)
{
blkptr_t bp;
unsigned long long *words = (void *)&bp;
- char buf[SPA_MAXBLOCKSIZE];
+ char *buf;
int err;
- memset(&bp, 0, sizeof (blkptr_t));
+ buf = umem_alloc(SPA_MAXBLOCKSIZE, UMEM_NOFAIL);
+ bzero(&bp, sizeof (bp));
err = sscanf(thing, "%llx:%llx:%llx:%llx:%llx:%llx:%llx:%llx:"
"%llx:%llx:%llx:%llx:%llx:%llx:%llx:%llx",
words + 0, words + 1, words + 2, words + 3,
@@ -4161,6 +4162,7 @@ zdb_embedded_block(char *thing)
exit(1);
}
zdb_dump_block_raw(buf, BPE_GET_LSIZE(&bp), 0);
+ umem_free(buf, SPA_MAXBLOCKSIZE);
}
int
--
2.14.2

217
zfs-patches/0027-Fix-zdb-ed-on-objset-for-exported-pool.patch Normal file
View File

@ -0,0 +1,217 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Chunwei Chen <david.chen@nutanix.com>
Date: Thu, 1 Feb 2018 16:36:40 -0800
Subject: [PATCH] Fix zdb -ed on objset for exported pool
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
zdb -ed on objset for exported pool would failed with:
failed to own dataset 'qq/fs0': No such file or directory
The reason is that zdb pass objset name to spa_import, it uses that
name to create a spa. Later, when dmu_objset_own tries to lookup the spa
using real pool name, it can't find one.
We fix this by make sure we pass pool name rather than objset name to
spa_import.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Chunwei Chen <david.chen@nutanix.com>
Closes #7099
Closes #6464
(cherry picked from commit 478754a8f5ff0f3b9f6dfe4ce272efc1681d243e)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
.../tests/functional/cli_root/zdb/Makefile.am | 3 +-
cmd/zdb/zdb.c | 32 +++++++----
tests/runfiles/linux.run | 2 +-
.../tests/functional/clean_mirror/cleanup.ksh | 4 +-
.../tests/functional/cli_root/zdb/zdb_006_pos.ksh | 64 ++++++++++++++++++++++
5 files changed, 90 insertions(+), 15 deletions(-)
create mode 100755 tests/zfs-tests/tests/functional/cli_root/zdb/zdb_006_pos.ksh
diff --git a/tests/zfs-tests/tests/functional/cli_root/zdb/Makefile.am b/tests/zfs-tests/tests/functional/cli_root/zdb/Makefile.am
index 51170fbc8..d37bcf607 100644
--- a/tests/zfs-tests/tests/functional/cli_root/zdb/Makefile.am
+++ b/tests/zfs-tests/tests/functional/cli_root/zdb/Makefile.am
@@ -4,4 +4,5 @@ dist_pkgdata_SCRIPTS = \
zdb_002_pos.ksh \
zdb_003_pos.ksh \
zdb_004_pos.ksh \
- zdb_005_pos.ksh
+ zdb_005_pos.ksh \
+ zdb_006_pos.ksh
diff --git a/cmd/zdb/zdb.c b/cmd/zdb/zdb.c
index 90847d8d9..17a0ae251 100644
--- a/cmd/zdb/zdb.c
+++ b/cmd/zdb/zdb.c
@@ -4177,7 +4177,7 @@ main(int argc, char **argv)
int error = 0;
char **searchdirs = NULL;
int nsearch = 0;
- char *target;
+ char *target, *target_pool;
nvlist_t *policy = NULL;
uint64_t max_txg = UINT64_MAX;
int flags = ZFS_IMPORT_MISSING_LOG;
@@ -4380,6 +4380,20 @@ main(int argc, char **argv)
error = 0;
target = argv[0];
+ if (strpbrk(target, "/@") != NULL) {
+ size_t targetlen;
+
+ target_pool = strdup(target);
+ *strpbrk(target_pool, "/@") = '\0';
+
+ target_is_spa = B_FALSE;
+ targetlen = strlen(target);
+ if (targetlen && target[targetlen - 1] == '/')
+ target[targetlen - 1] = '\0';
+ } else {
+ target_pool = target;
+ }
+
if (dump_opt['e']) {
importargs_t args = { 0 };
nvlist_t *cfg = NULL;
@@ -4388,8 +4402,10 @@ main(int argc, char **argv)
args.path = searchdirs;
args.can_be_active = B_TRUE;
- error = zpool_tryimport(g_zfs, target, &cfg, &args);
+ error = zpool_tryimport(g_zfs, target_pool, &cfg, &args);
+
if (error == 0) {
+
if (nvlist_add_nvlist(cfg,
ZPOOL_REWIND_POLICY, policy) != 0) {
fatal("can't open '%s': %s",
@@ -4404,19 +4420,13 @@ main(int argc, char **argv)
(void) printf("\nConfiguration for import:\n");
dump_nvlist(cfg, 8);
}
- error = spa_import(target, cfg, NULL,
+ error = spa_import(target_pool, cfg, NULL,
flags | ZFS_IMPORT_SKIP_MMP);
}
}
- if (strpbrk(target, "/@") != NULL) {
- size_t targetlen;
-
- target_is_spa = B_FALSE;
- targetlen = strlen(target);
- if (targetlen && target[targetlen - 1] == '/')
- target[targetlen - 1] = '\0';
- }
+ if (target_pool != target)
+ free(target_pool);
if (error == 0) {
if (target_is_spa || dump_opt['R']) {
diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run
index 303c27529..ea2dbb282 100644
--- a/tests/runfiles/linux.run
+++ b/tests/runfiles/linux.run
@@ -73,7 +73,7 @@ tags = ['functional', 'clean_mirror']
[tests/functional/cli_root/zdb]
tests = ['zdb_001_neg', 'zdb_002_pos', 'zdb_003_pos', 'zdb_004_pos',
- 'zdb_005_pos']
+ 'zdb_005_pos', 'zdb_006_pos']
pre =
post =
tags = ['functional', 'cli_root', 'zdb']
diff --git a/tests/zfs-tests/tests/functional/clean_mirror/cleanup.ksh b/tests/zfs-tests/tests/functional/clean_mirror/cleanup.ksh
index ac3bfbca8..fb0db312e 100755
--- a/tests/zfs-tests/tests/functional/clean_mirror/cleanup.ksh
+++ b/tests/zfs-tests/tests/functional/clean_mirror/cleanup.ksh
@@ -38,10 +38,10 @@ df -F zfs -h | grep "$TESTFS " >/dev/null
[[ $? == 0 ]] && log_must zfs umount -f $TESTDIR
destroy_pool $TESTPOOL
-if is_mpath_device $MIRROR_PRIMARY; then
+if ( is_mpath_device $MIRROR_PRIMARY || is_loop_device $MIRROR_SECONDARY); then
parted $DEV_DSKDIR/$MIRROR_PRIMARY -s rm 1
fi
-if is_mpath_device $MIRROR_SECONDARY; then
+if ( is_mpath_device $MIRROR_SECONDARY || is_loop_device $MIRROR_SECONDARY); then
parted $DEV_DSKDIR/$MIRROR_SECONDARY -s rm 1
fi
# recreate and destroy a zpool over the disks to restore the partitions to
diff --git a/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_006_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_006_pos.ksh
new file mode 100755
index 000000000..97b00e9e1
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_006_pos.ksh
@@ -0,0 +1,64 @@
+#!/bin/ksh
+
+#
+# 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.
+#
+
+#
+# Copyright (c) 2018 by Nutanix. All rights reserved.
+#
+
+. $STF_SUITE/include/libtest.shlib
+
+#
+# Description:
+# zdb -d will work on imported/exported pool with pool/dataset argument
+#
+# Strategy:
+# 1. Create a pool
+# 2. Run zdb -d with pool and dataset arguments.
+# 3. Export the pool
+# 4. Run zdb -ed with pool and dataset arguments.
+#
+
+function cleanup
+{
+ datasetexists $TESTPOOL && destroy_pool $TESTPOOL
+ for DISK in $DISKS; do
+ zpool labelclear -f $DEV_RDSKDIR/$DISK
+ done
+}
+
+log_assert "Verify zdb -d works on imported/exported pool with pool/dataset argument"
+log_onexit cleanup
+
+verify_runnable "global"
+verify_disk_count "$DISKS" 2
+
+default_mirror_setup_noexit $DISKS
+log_must zfs snap $TESTPOOL/$TESTFS@snap
+
+log_must zdb -d $TESTPOOL
+log_must zdb -d $TESTPOOL/
+log_must zdb -d $TESTPOOL/$TESTFS
+log_must zdb -d $TESTPOOL/$TESTFS@snap
+
+log_must zpool export $TESTPOOL
+
+log_must zdb -ed $TESTPOOL
+log_must zdb -ed $TESTPOOL/
+log_must zdb -ed $TESTPOOL/$TESTFS
+log_must zdb -ed $TESTPOOL/$TESTFS@snap
+
+log_must zpool import $TESTPOOL
+
+cleanup
+
+log_pass "zdb -d works on imported/exported pool with pool/dataset argument"
--
2.14.2

553
zfs-patches/0028-Handle-zap_add-failures-in-mixed-case-mode.patch Normal file
View File

@ -0,0 +1,553 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: sanjeevbagewadi <sanjeev.bagewadi@gmail.com>
Date: Fri, 9 Feb 2018 23:45:53 +0530
Subject: [PATCH] Handle zap_add() failures in mixed case mode
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
With "casesensitivity=mixed", zap_add() could fail when the number of
files/directories with the same name (varying in case) exceed the
capacity of the leaf node of a Fatzap. This results in a ASSERT()
failure as zfs_link_create() does not expect zap_add() to fail. The fix
is to handle these failures and rollback the transactions.
Reviewed by: Matt Ahrens <mahrens@delphix.com>
Reviewed-by: Chunwei Chen <david.chen@nutanix.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Sanjeev Bagewadi <sanjeev.bagewadi@gmail.com>
Closes #7011
Closes #7054
(cherry picked from commit b3da003ebfad673bb4ada35f87a18a1ef175e95d)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
.../tests/functional/casenorm/Makefile.am | 1 +
include/sys/zap_leaf.h | 15 ++-
module/zfs/zap.c | 25 +++-
module/zfs/zap_leaf.c | 2 +-
module/zfs/zap_micro.c | 38 +++++-
module/zfs/zfs_dir.c | 29 ++++-
module/zfs/zfs_vnops.c | 73 ++++++++---
tests/runfiles/linux.run | 2 +-
.../functional/casenorm/mixed_create_failure.ksh | 136 +++++++++++++++++++++
9 files changed, 289 insertions(+), 32 deletions(-)
create mode 100755 tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh
diff --git a/tests/zfs-tests/tests/functional/casenorm/Makefile.am b/tests/zfs-tests/tests/functional/casenorm/Makefile.am
index 00a19c7ff..00cb59074 100644
--- a/tests/zfs-tests/tests/functional/casenorm/Makefile.am
+++ b/tests/zfs-tests/tests/functional/casenorm/Makefile.am
@@ -9,6 +9,7 @@ dist_pkgdata_SCRIPTS = \
insensitive_formd_lookup.ksh \
insensitive_none_delete.ksh \
insensitive_none_lookup.ksh \
+ mixed_create_failure.ksh \
mixed_formd_delete.ksh \
mixed_formd_lookup_ci.ksh \
mixed_formd_lookup.ksh \
diff --git a/include/sys/zap_leaf.h b/include/sys/zap_leaf.h
index e784c5963..a3da1036a 100644
--- a/include/sys/zap_leaf.h
+++ b/include/sys/zap_leaf.h
@@ -46,10 +46,15 @@ struct zap_stats;
* block size (1<<l->l_bs) - hash entry size (2) * number of hash
* entries - header space (2*chunksize)
*/
-#define ZAP_LEAF_NUMCHUNKS(l) \
- (((1<<(l)->l_bs) - 2*ZAP_LEAF_HASH_NUMENTRIES(l)) / \
+#define ZAP_LEAF_NUMCHUNKS_BS(bs) \
+ (((1<<(bs)) - 2*ZAP_LEAF_HASH_NUMENTRIES_BS(bs)) / \
ZAP_LEAF_CHUNKSIZE - 2)
+#define ZAP_LEAF_NUMCHUNKS(l) (ZAP_LEAF_NUMCHUNKS_BS(((l)->l_bs)))
+
+#define ZAP_LEAF_NUMCHUNKS_DEF \
+ (ZAP_LEAF_NUMCHUNKS_BS(fzap_default_block_shift))
+
/*
* The amount of space within the chunk available for the array is:
* chunk size - space for type (1) - space for next pointer (2)
@@ -74,8 +79,10 @@ struct zap_stats;
* which is less than block size / CHUNKSIZE (24) / minimum number of
* chunks per entry (3).
*/
-#define ZAP_LEAF_HASH_SHIFT(l) ((l)->l_bs - 5)
-#define ZAP_LEAF_HASH_NUMENTRIES(l) (1 << ZAP_LEAF_HASH_SHIFT(l))
+#define ZAP_LEAF_HASH_SHIFT_BS(bs) ((bs) - 5)
+#define ZAP_LEAF_HASH_NUMENTRIES_BS(bs) (1 << ZAP_LEAF_HASH_SHIFT_BS(bs))
+#define ZAP_LEAF_HASH_SHIFT(l) (ZAP_LEAF_HASH_SHIFT_BS(((l)->l_bs)))
+#define ZAP_LEAF_HASH_NUMENTRIES(l) (ZAP_LEAF_HASH_NUMENTRIES_BS(((l)->l_bs)))
/*
* The chunks start immediately after the hash table. The end of the
diff --git a/module/zfs/zap.c b/module/zfs/zap.c
index ee9962bff..9843d8c50 100644
--- a/module/zfs/zap.c
+++ b/module/zfs/zap.c
@@ -819,15 +819,19 @@ fzap_lookup(zap_name_t *zn,
return (err);
}
+#define MAX_EXPAND_RETRIES 2
+
int
fzap_add_cd(zap_name_t *zn,
uint64_t integer_size, uint64_t num_integers,
const void *val, uint32_t cd, void *tag, dmu_tx_t *tx)
{
zap_leaf_t *l;
+ zap_leaf_t *prev_l = NULL;
int err;
zap_entry_handle_t zeh;
zap_t *zap = zn->zn_zap;
+ int expand_retries = 0;
ASSERT(RW_LOCK_HELD(&zap->zap_rwlock));
ASSERT(!zap->zap_ismicro);
@@ -851,10 +855,29 @@ retry:
if (err == 0) {
zap_increment_num_entries(zap, 1, tx);
} else if (err == EAGAIN) {
+ /*
+ * If the last two expansions did not help, there is no point
+ * trying to expand again
+ */
+ if (expand_retries > MAX_EXPAND_RETRIES && prev_l == l) {
+ err = SET_ERROR(ENOSPC);
+ goto out;
+ }
+
err = zap_expand_leaf(zn, l, tag, tx, &l);
zap = zn->zn_zap; /* zap_expand_leaf() may change zap */
- if (err == 0)
+ if (err == 0) {
+ prev_l = l;
+ expand_retries++;
goto retry;
+ } else if (err == ENOSPC) {
+ /*
+ * If we failed to expand the leaf, then bailout
+ * as there is no point trying
+ * zap_put_leaf_maybe_grow_ptrtbl().
+ */
+ return (err);
+ }
}
out:
diff --git a/module/zfs/zap_leaf.c b/module/zfs/zap_leaf.c
index c342695c7..526e46606 100644
--- a/module/zfs/zap_leaf.c
+++ b/module/zfs/zap_leaf.c
@@ -53,7 +53,7 @@ static uint16_t *zap_leaf_rehash_entry(zap_leaf_t *l, uint16_t entry);
((h) >> \
(64 - ZAP_LEAF_HASH_SHIFT(l) - zap_leaf_phys(l)->l_hdr.lh_prefix_len)))
-#define LEAF_HASH_ENTPTR(l, h) (&zap_leaf_phys(l)->l_hash[LEAF_HASH(l, h)])
+#define LEAF_HASH_ENTPTR(l, h) (&zap_leaf_phys(l)->l_hash[LEAF_HASH(l, h)])
extern inline zap_leaf_phys_t *zap_leaf_phys(zap_leaf_t *l);
diff --git a/module/zfs/zap_micro.c b/module/zfs/zap_micro.c
index 3ebf995c6..34bef3e63 100644
--- a/module/zfs/zap_micro.c
+++ b/module/zfs/zap_micro.c
@@ -363,6 +363,41 @@ mze_find_unused_cd(zap_t *zap, uint64_t hash)
return (cd);
}
+/*
+ * Each mzap entry requires at max : 4 chunks
+ * 3 chunks for names + 1 chunk for value.
+ */
+#define MZAP_ENT_CHUNKS (1 + ZAP_LEAF_ARRAY_NCHUNKS(MZAP_NAME_LEN) + \
+ ZAP_LEAF_ARRAY_NCHUNKS(sizeof (uint64_t)))
+
+/*
+ * Check if the current entry keeps the colliding entries under the fatzap leaf
+ * size.
+ */
+static boolean_t
+mze_canfit_fzap_leaf(zap_name_t *zn, uint64_t hash)
+{
+ zap_t *zap = zn->zn_zap;
+ mzap_ent_t mze_tofind;
+ mzap_ent_t *mze;
+ avl_index_t idx;
+ avl_tree_t *avl = &zap->zap_m.zap_avl;
+ uint32_t mzap_ents = 0;
+
+ mze_tofind.mze_hash = hash;
+ mze_tofind.mze_cd = 0;
+
+ for (mze = avl_find(avl, &mze_tofind, &idx);
+ mze && mze->mze_hash == hash; mze = AVL_NEXT(avl, mze)) {
+ mzap_ents++;
+ }
+
+ /* Include the new entry being added */
+ mzap_ents++;
+
+ return (ZAP_LEAF_NUMCHUNKS_DEF > (mzap_ents * MZAP_ENT_CHUNKS));
+}
+
static void
mze_remove(zap_t *zap, mzap_ent_t *mze)
{
@@ -1191,7 +1226,8 @@ zap_add_impl(zap_t *zap, const char *key,
err = fzap_add(zn, integer_size, num_integers, val, tag, tx);
zap = zn->zn_zap; /* fzap_add() may change zap */
} else if (integer_size != 8 || num_integers != 1 ||
- strlen(key) >= MZAP_NAME_LEN) {
+ strlen(key) >= MZAP_NAME_LEN ||
+ !mze_canfit_fzap_leaf(zn, zn->zn_hash)) {
err = mzap_upgrade(&zn->zn_zap, tag, tx, 0);
if (err == 0) {
err = fzap_add(zn, integer_size, num_integers, val,
diff --git a/module/zfs/zfs_dir.c b/module/zfs/zfs_dir.c
index 9a8bbccd9..6398a1d15 100644
--- a/module/zfs/zfs_dir.c
+++ b/module/zfs/zfs_dir.c
@@ -742,7 +742,11 @@ zfs_dirent(znode_t *zp, uint64_t mode)
}
/*
- * Link zp into dl. Can only fail if zp has been unlinked.
+ * Link zp into dl. Can fail in the following cases :
+ * - if zp has been unlinked.
+ * - if the number of entries with the same hash (aka. colliding entries)
+ * exceed the capacity of a leaf-block of fatzap and splitting of the
+ * leaf-block does not help.
*/
int
zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag)
@@ -776,6 +780,24 @@ zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag)
NULL, &links, sizeof (links));
}
}
+
+ value = zfs_dirent(zp, zp->z_mode);
+ error = zap_add(ZTOZSB(zp)->z_os, dzp->z_id, dl->dl_name, 8, 1,
+ &value, tx);
+
+ /*
+ * zap_add could fail to add the entry if it exceeds the capacity of the
+ * leaf-block and zap_leaf_split() failed to help.
+ * The caller of this routine is responsible for failing the transaction
+ * which will rollback the SA updates done above.
+ */
+ if (error != 0) {
+ if (!(flag & ZRENAMING) && !(flag & ZNEW))
+ drop_nlink(ZTOI(zp));
+ mutex_exit(&zp->z_lock);
+ return (error);
+ }
+
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_PARENT(zfsvfs), NULL,
&dzp->z_id, sizeof (dzp->z_id));
SA_ADD_BULK_ATTR(bulk, count, SA_ZPL_FLAGS(zfsvfs), NULL,
@@ -813,11 +835,6 @@ zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag)
ASSERT(error == 0);
mutex_exit(&dzp->z_lock);
- value = zfs_dirent(zp, zp->z_mode);
- error = zap_add(ZTOZSB(zp)->z_os, dzp->z_id, dl->dl_name,
- 8, 1, &value, tx);
- ASSERT(error == 0);
-
return (0);
}
diff --git a/module/zfs/zfs_vnops.c b/module/zfs/zfs_vnops.c
index 6f6ce79db..8a7ad702c 100644
--- a/module/zfs/zfs_vnops.c
+++ b/module/zfs/zfs_vnops.c
@@ -1443,10 +1443,22 @@ top:
}
zfs_mknode(dzp, vap, tx, cr, 0, &zp, &acl_ids);
+ error = zfs_link_create(dl, zp, tx, ZNEW);
+ if (error != 0) {
+ /*
+ * Since, we failed to add the directory entry for it,
+ * delete the newly created dnode.
+ */
+ zfs_znode_delete(zp, tx);
+ remove_inode_hash(ZTOI(zp));
+ zfs_acl_ids_free(&acl_ids);
+ dmu_tx_commit(tx);
+ goto out;
+ }
+
if (fuid_dirtied)
zfs_fuid_sync(zfsvfs, tx);
- (void) zfs_link_create(dl, zp, tx, ZNEW);
txtype = zfs_log_create_txtype(Z_FILE, vsecp, vap);
if (flag & FIGNORECASE)
txtype |= TX_CI;
@@ -2037,13 +2049,18 @@ top:
*/
zfs_mknode(dzp, vap, tx, cr, 0, &zp, &acl_ids);
- if (fuid_dirtied)
- zfs_fuid_sync(zfsvfs, tx);
-
/*
* Now put new name in parent dir.
*/
- (void) zfs_link_create(dl, zp, tx, ZNEW);
+ error = zfs_link_create(dl, zp, tx, ZNEW);
+ if (error != 0) {
+ zfs_znode_delete(zp, tx);
+ remove_inode_hash(ZTOI(zp));
+ goto out;
+ }
+
+ if (fuid_dirtied)
+ zfs_fuid_sync(zfsvfs, tx);
*ipp = ZTOI(zp);
@@ -2053,6 +2070,7 @@ top:
zfs_log_create(zilog, tx, txtype, dzp, zp, dirname, vsecp,
acl_ids.z_fuidp, vap);
+out:
zfs_acl_ids_free(&acl_ids);
dmu_tx_commit(tx);
@@ -2062,10 +2080,14 @@ top:
if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS)
zil_commit(zilog, 0);
- zfs_inode_update(dzp);
- zfs_inode_update(zp);
+ if (error != 0) {
+ iput(ZTOI(zp));
+ } else {
+ zfs_inode_update(dzp);
+ zfs_inode_update(zp);
+ }
ZFS_EXIT(zfsvfs);
- return (0);
+ return (error);
}
/*
@@ -3683,6 +3705,13 @@ top:
VERIFY3U(zfs_link_destroy(tdl, szp, tx,
ZRENAMING, NULL), ==, 0);
}
+ } else {
+ /*
+ * If we had removed the existing target, subsequent
+ * call to zfs_link_create() to add back the same entry
+ * but, the new dnode (szp) should not fail.
+ */
+ ASSERT(tzp == NULL);
}
}
@@ -3853,14 +3882,18 @@ top:
/*
* Insert the new object into the directory.
*/
- (void) zfs_link_create(dl, zp, tx, ZNEW);
-
- if (flags & FIGNORECASE)
- txtype |= TX_CI;
- zfs_log_symlink(zilog, tx, txtype, dzp, zp, name, link);
+ error = zfs_link_create(dl, zp, tx, ZNEW);
+ if (error != 0) {
+ zfs_znode_delete(zp, tx);
+ remove_inode_hash(ZTOI(zp));
+ } else {
+ if (flags & FIGNORECASE)
+ txtype |= TX_CI;
+ zfs_log_symlink(zilog, tx, txtype, dzp, zp, name, link);
- zfs_inode_update(dzp);
- zfs_inode_update(zp);
+ zfs_inode_update(dzp);
+ zfs_inode_update(zp);
+ }
zfs_acl_ids_free(&acl_ids);
@@ -3868,10 +3901,14 @@ top:
zfs_dirent_unlock(dl);
- *ipp = ZTOI(zp);
+ if (error == 0) {
+ *ipp = ZTOI(zp);
- if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS)
- zil_commit(zilog, 0);
+ if (zfsvfs->z_os->os_sync == ZFS_SYNC_ALWAYS)
+ zil_commit(zilog, 0);
+ } else {
+ iput(ZTOI(zp));
+ }
ZFS_EXIT(zfsvfs);
return (error);
diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run
index ea2dbb282..8be3e1c62 100644
--- a/tests/runfiles/linux.run
+++ b/tests/runfiles/linux.run
@@ -55,7 +55,7 @@ tags = ['functional', 'cachefile']
# 'mixed_none_lookup', 'mixed_none_lookup_ci', 'mixed_none_delete',
# 'mixed_formd_lookup', 'mixed_formd_lookup_ci', 'mixed_formd_delete']
[tests/functional/casenorm]
-tests = ['case_all_values', 'norm_all_values']
+tests = ['case_all_values', 'norm_all_values', 'mixed_create_failure']
tags = ['functional', 'casenorm']
[tests/functional/chattr]
diff --git a/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh b/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh
new file mode 100755
index 000000000..51b5bb3f6
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/casenorm/mixed_create_failure.ksh
@@ -0,0 +1,136 @@
+#!/bin/ksh -p
+#
+#
+# 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.
+#
+#
+# Copyright 2018 Nutanix Inc. All rights reserved.
+#
+
+. $STF_SUITE/tests/functional/casenorm/casenorm.kshlib
+
+# DESCRIPTION:
+# For the filesystem with casesensitivity=mixed, normalization=none,
+# when multiple files with the same name (differing only in case) are created,
+# the number of files is limited to what can fit in a fatzap leaf-block.
+# And beyond that, it fails with ENOSPC.
+#
+# Ensure that the create/rename operations fail gracefully and not trigger an
+# ASSERT.
+#
+# STRATEGY:
+# Repeat the below steps for objects: files, directories, symlinks and hardlinks
+# 1. Create objects with same name but varying in case.
+# E.g. 'abcdefghijklmnop', 'Abcdefghijklmnop', 'ABcdefghijklmnop' etc.
+# The create should fail with ENOSPC.
+# 2. Create an object with name 'tmp_obj' and try to rename it to name that we
+# failed to add in step 1 above.
+# This should fail as well.
+
+verify_runnable "global"
+
+function cleanup
+{
+ destroy_testfs
+}
+
+log_onexit cleanup
+log_assert "With mixed mode: ensure create fails with ENOSPC beyond a certain limit"
+
+create_testfs "-o casesensitivity=mixed -o normalization=none"
+
+# Different object types
+obj_type=('file' 'dir' 'symlink' 'hardlink')
+
+# Commands to create different object types
+typeset -A ops
+ops['file']='touch'
+ops['dir']='mkdir'
+ops['symlink']='ln -s'
+ops['hardlink']='ln'
+
+# This function tests the following for a give object type :
+# - Create multiple objects with the same name (varying only in case).
+# Ensure that it eventually fails once the leaf-block limit is exceeded.
+# - Create another object with a different name. And attempt rename it to the
+# name (for which the create had failed in the previous step).
+# This should fail as well.
+# Args :
+# $1 - object type (file/dir/symlink/hardlink)
+# $2 - test directory
+#
+function test_ops
+{
+ typeset obj_type=$1
+ typeset testdir=$2
+
+ target_obj='target-file'
+
+ op="${ops[$obj_type]}"
+
+ log_note "The op : $op"
+ log_note "testdir=$testdir obj_type=$obj_type"
+
+ test_path="$testdir/$obj_type"
+ mkdir $test_path
+ log_note "Created test dir $test_path"
+
+ if [[ $obj_type = "symlink" || $obj_type = "hardlink" ]]; then
+ touch $test_path/$target_obj
+ log_note "Created target: $test_path/$target_obj"
+ op="$op $test_path/$target_obj"
+ fi
+
+ log_note "op : $op"
+ names='{a,A}{b,B}{c,C}{d,D}{e,E}{f,F}{g,G}{h,H}{i,I}{j,J}{k,K}{l,L}'
+ for name in $names; do
+ cmd="$op $test_path/$name"
+ out=$($cmd 2>&1)
+ ret=$?
+ log_note "cmd: $cmd ret: $ret out=$out"
+ if (($ret != 0)); then
+ if [[ $out = *@(No space left on device)* ]]; then
+ save_name="$test_path/$name"
+ break;
+ else
+ log_err "$cmd failed with unexpected error : $out"
+ fi
+ fi
+ done
+
+ log_note 'Test rename \"sample_name\" rename'
+ TMP_OBJ="$test_path/tmp_obj"
+ cmd="$op $TMP_OBJ"
+ out=$($cmd 2>&1)
+ ret=$?
+ if (($ret != 0)); then
+ log_err "cmd:$cmd failed out:$out"
+ fi
+
+ # Now, try to rename the tmp_obj to the name which we failed to add earlier.
+ # This should fail as well.
+ out=$(mv $TMP_OBJ $save_name 2>&1)
+ ret=$?
+ if (($ret != 0)); then
+ if [[ $out = *@(No space left on device)* ]]; then
+ log_note "$cmd failed as expected : $out"
+ else
+ log_err "$cmd failed with : $out"
+ fi
+ fi
+}
+
+for obj_type in ${obj_type[*]};
+do
+ log_note "Testing create of $obj_type"
+ test_ops $obj_type $TESTDIR
+done
+
+log_pass "Mixed mode FS: Ops on large number of colliding names fail gracefully"
--
2.14.2

56
zfs-patches/0029-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch Normal file
View File

@ -0,0 +1,56 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: sanjeevbagewadi <sanjeev.bagewadi@gmail.com>
Date: Tue, 13 Feb 2018 01:00:38 +0530
Subject: [PATCH] mmp should use a fixed tag for spa_config locks
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
mmp_write_uberblock() and mmp_write_done() should the same tag
for spa_config_locks.
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Sanjeev Bagewadi <sanjeev.bagewadi@gmail.com>
Closes #6530
Closes #7155
(cherry picked from commit d85011ed696b41f4de963f7932dfb53ea871278b)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/mmp.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index e91ae628a..1dfb7c05b 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -125,6 +125,7 @@ uint_t zfs_multihost_import_intervals = MMP_DEFAULT_IMPORT_INTERVALS;
uint_t zfs_multihost_fail_intervals = MMP_DEFAULT_FAIL_INTERVALS;
static void mmp_thread(spa_t *spa);
+char *mmp_tag = "mmp_write_uberblock";
void
mmp_init(spa_t *spa)
@@ -278,7 +279,7 @@ mmp_write_done(zio_t *zio)
unlock:
mutex_exit(&mts->mmp_io_lock);
- spa_config_exit(spa, SCL_STATE, FTAG);
+ spa_config_exit(spa, SCL_STATE, mmp_tag);
abd_free(zio->io_abd);
}
@@ -314,7 +315,7 @@ mmp_write_uberblock(spa_t *spa)
int label;
uint64_t offset;
- spa_config_enter(spa, SCL_STATE, FTAG, RW_READER);
+ spa_config_enter(spa, SCL_STATE, mmp_tag, RW_READER);
vd = mmp_random_leaf(spa->spa_root_vdev);
if (vd == NULL) {
spa_config_exit(spa, SCL_STATE, FTAG);
--
2.14.2

66
zfs-patches/0030-contrib-initramfs-add-missing-conf.d-zfs.patch Normal file
View File

@ -0,0 +1,66 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: LOLi <loli10K@users.noreply.github.com>
Date: Mon, 12 Feb 2018 20:40:00 +0100
Subject: [PATCH] contrib/initramfs: add missing conf.d/zfs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
When upgrading from the distribution-provided zfs-initramfs package on
root-on-zfs Ubuntu and Debian the system may fail to boot: this change
adds the missing initramfs configuration file.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Richard Laager <rlaager@wiktel.com>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes #7158
(cherry picked from commit a9ff89e05cd2f420e44b0e50c92d97c166772d2b)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
contrib/initramfs/Makefile.am | 6 ++++--
contrib/initramfs/conf.d/zfs | 8 ++++++++
2 files changed, 12 insertions(+), 2 deletions(-)
create mode 100644 contrib/initramfs/conf.d/zfs
diff --git a/contrib/initramfs/Makefile.am b/contrib/initramfs/Makefile.am
index 998e588ab..b22940821 100644
--- a/contrib/initramfs/Makefile.am
+++ b/contrib/initramfs/Makefile.am
@@ -1,8 +1,10 @@
initrddir = $(datarootdir)/initramfs-tools
-initrd_SCRIPTS = conf-hooks.d/zfs hooks/zfs scripts/zfs scripts/local-top/zfs
+initrd_SCRIPTS = \
+ conf.d/zfs conf-hooks.d/zfs hooks/zfs scripts/zfs scripts/local-top/zfs
EXTRA_DIST = \
+ $(top_srcdir)/contrib/initramfs/conf.d/zfs \
$(top_srcdir)/contrib/initramfs/conf-hooks.d/zfs \
$(top_srcdir)/contrib/initramfs/hooks/zfs \
$(top_srcdir)/contrib/initramfs/scripts/zfs \
@@ -10,7 +12,7 @@ EXTRA_DIST = \
$(top_srcdir)/contrib/initramfs/README.initramfs.markdown
install-initrdSCRIPTS: $(EXTRA_DIST)
- for d in conf-hooks.d hooks scripts scripts/local-top; do \
+ for d in conf.d conf-hooks.d hooks scripts scripts/local-top; do \
$(MKDIR_P) $(DESTDIR)$(initrddir)/$$d; \
cp $(top_srcdir)/contrib/initramfs/$$d/zfs \
$(DESTDIR)$(initrddir)/$$d/; \
diff --git a/contrib/initramfs/conf.d/zfs b/contrib/initramfs/conf.d/zfs
new file mode 100644
index 000000000..c67d75ba8
--- /dev/null
+++ b/contrib/initramfs/conf.d/zfs
@@ -0,0 +1,8 @@
+for x in $(cat /proc/cmdline)
+do
+ case $x in
+ root=ZFS=*|root=zfs:*)
+ BOOT=zfs
+ ;;
+ esac
+done
--
2.14.2

90
zfs-patches/0031-zfs-receive-fails-with-dataset-is-busy.patch Normal file
View File

@ -0,0 +1,90 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: LOLi <loli10K@users.noreply.github.com>
Date: Mon, 12 Feb 2018 21:28:59 +0100
Subject: [PATCH] 'zfs receive' fails with "dataset is busy"
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Receiving an incremental stream after an interrupted "zfs receive -s"
fails with the message "dataset is busy": this is because we still have
the hidden clone ../%recv from the resumable receive.
Improve the error message suggesting the existence of a partially
complete resumable stream from "zfs receive -s" which can be either
aborted ("zfs receive -A") or resumed ("zfs send -t").
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes #7129
Closes #7154
(cherry picked from commit 1d805a534b5e1768413a0242b8e92b97cb1a796c)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zfs/zfs_main.c | 2 +-
lib/libzfs/libzfs_sendrecv.c | 18 +++++++++++++++++-
2 files changed, 18 insertions(+), 2 deletions(-)
diff --git a/cmd/zfs/zfs_main.c b/cmd/zfs/zfs_main.c
index e8fe6a9fa..f57df8581 100644
--- a/cmd/zfs/zfs_main.c
+++ b/cmd/zfs/zfs_main.c
@@ -6072,7 +6072,7 @@ share_mount_one(zfs_handle_t *zhp, int op, int flags, char *protocol,
(void) fprintf(stderr, gettext("cannot %s '%s': "
"Contains partially-completed state from "
- "\"zfs receive -r\", which can be resumed with "
+ "\"zfs receive -s\", which can be resumed with "
"\"zfs send -t\"\n"),
cmdname, zfs_get_name(zhp));
return (1);
diff --git a/lib/libzfs/libzfs_sendrecv.c b/lib/libzfs/libzfs_sendrecv.c
index ec190022f..66d89067b 100644
--- a/lib/libzfs/libzfs_sendrecv.c
+++ b/lib/libzfs/libzfs_sendrecv.c
@@ -3254,6 +3254,7 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const char *tosnap,
zfs_type_t type;
boolean_t toplevel = B_FALSE;
boolean_t zoned = B_FALSE;
+ boolean_t hastoken = B_FALSE;
begin_time = time(NULL);
bzero(origin, MAXNAMELEN);
@@ -3535,6 +3536,11 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const char *tosnap,
/* we want to know if we're zoned when validating -o|-x props */
zoned = zfs_prop_get_int(zhp, ZFS_PROP_ZONED);
+ /* may need this info later, get it now we have zhp around */
+ if (zfs_prop_get(zhp, ZFS_PROP_RECEIVE_RESUME_TOKEN, NULL, 0,
+ NULL, NULL, 0, B_TRUE) == 0)
+ hastoken = B_TRUE;
+
/* gather existing properties on destination */
origprops = fnvlist_alloc();
fnvlist_merge(origprops, zhp->zfs_props);
@@ -3741,9 +3747,19 @@ zfs_receive_one(libzfs_handle_t *hdl, int infd, const char *tosnap,
break;
case EDQUOT:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
- "destination %s space quota exceeded"), name);
+ "destination %s space quota exceeded."), name);
(void) zfs_error(hdl, EZFS_NOSPC, errbuf);
break;
+ case EBUSY:
+ if (hastoken) {
+ zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
+ "destination %s contains "
+ "partially-complete state from "
+ "\"zfs receive -s\"."), name);
+ (void) zfs_error(hdl, EZFS_BUSY, errbuf);
+ break;
+ }
+ /* fallthru */
default:
(void) zfs_standard_error(hdl, ioctl_errno, errbuf);
}
--
2.14.2

247
zfs-patches/0032-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch Normal file
View File

@ -0,0 +1,247 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: George Wilson <george.wilson@delphix.com>
Date: Thu, 8 Feb 2018 15:04:14 -0500
Subject: [PATCH] OpenZFS 8857 - zio_remove_child() panic due to already
destroyed parent zio
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
PROBLEM
=======
It's possible for a parent zio to complete even though it has children
which have not completed. This can result in the following panic:
> $C
ffffff01809128c0 vpanic()
ffffff01809128e0 mutex_panic+0x58(fffffffffb94c904, ffffff597dde7f80)
ffffff0180912950 mutex_vector_enter+0x347(ffffff597dde7f80)
ffffff01809129b0 zio_remove_child+0x50(ffffff597dde7c58, ffffff32bd901ac0,
ffffff3373370908)
ffffff0180912a40 zio_done+0x390(ffffff32bd901ac0)
ffffff0180912a70 zio_execute+0x78(ffffff32bd901ac0)
ffffff0180912b30 taskq_thread+0x2d0(ffffff33bae44140)
ffffff0180912b40 thread_start+8()
> ::status
debugging crash dump vmcore.2 (64-bit) from batfs0390
operating system: 5.11 joyent_20170911T171900Z (i86pc)
image uuid: (not set)
panic message: mutex_enter: bad mutex, lp=ffffff597dde7f80
owner=ffffff3c59b39480 thread=ffffff0180912c40
dump content: kernel pages only
The problem is that dbuf_prefetch along with l2arc can create a zio tree
which confuses the parent zio and allows it to complete with while children
still exist. Here's the scenario:
zio tree:
pio
|--- lio
The parent zio, pio, has entered the zio_done stage and begins to check its
children to see there are still some that have not completed. In zio_done(),
the children are checked in the following order:
zio_wait_for_children(zio, ZIO_CHILD_VDEV, ZIO_WAIT_DONE)
zio_wait_for_children(zio, ZIO_CHILD_GANG, ZIO_WAIT_DONE)
zio_wait_for_children(zio, ZIO_CHILD_DDT, ZIO_WAIT_DONE)
zio_wait_for_children(zio, ZIO_CHILD_LOGICAL, ZIO_WAIT_DONE)
If pio, finds any child which has not completed then it stops executing and
goes to sleep. Each call to zio_wait_for_children() will grab the io_lock
while checking the particular child.
In this scenario, the pio has completed the first call to
zio_wait_for_children() to check for any ZIO_CHILD_VDEV children. Since
the only zio in the zio tree right now is the logical zio, lio, then it
completes that call and prepares to check the next child type.
In the meantime, the lio completes and in its callback creates a child vdev
zio, cio. The zio tree looks like this:
zio tree:
pio
|--- lio
|--- cio
The lio then grabs the parent's io_lock and removes itself.
zio tree:
pio
|--- cio
The pio continues to run but has already completed its check for ZIO_CHILD_VDEV
and will erroneously complete. When the child zio, cio, completes it will panic
the system trying to reference the parent zio which has been destroyed.
SOLUTION
========
The fix is to rework the zio_wait_for_children() logic to accept a bitfield
for all the children types that it's interested in checking. The
io_lock will is held the entire time we check all the children types. Since
the function now accepts a bitfield, a simple ZIO_CHILD_BIT() macro is provided
to allow for the conversion between a ZIO_CHILD type and the bitfield used by
the zio_wiat_for_children logic.
Authored by: George Wilson <george.wilson@delphix.com>
Reviewed by: Matthew Ahrens <mahrens@delphix.com>
Reviewed by: Andriy Gapon <avg@FreeBSD.org>
Reviewed by: Youzhong Yang <youzhong@gmail.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Approved by: Dan McDonald <danmcd@omniti.com>
Ported-by: Giuseppe Di Natale <dinatale2@llnl.gov>
OpenZFS-issue: https://www.illumos.org/issues/8857
OpenZFS-commit: https://github.com/openzfs/openzfs/commit/862ff6d99c
Issue #5918
Closes #7168
(cherry picked from commit 07ce5d739024cf9c638716c09f9934b4e629678c)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/sys/zio.h | 11 +++++++++++
module/zfs/zio.c | 49 +++++++++++++++++++++++++++++--------------------
2 files changed, 40 insertions(+), 20 deletions(-)
diff --git a/include/sys/zio.h b/include/sys/zio.h
index 4eaabc38c..0d741f8e2 100644
--- a/include/sys/zio.h
+++ b/include/sys/zio.h
@@ -215,6 +215,9 @@ enum zio_flag {
(((zio)->io_flags & ZIO_FLAG_VDEV_INHERIT) | \
ZIO_FLAG_CANFAIL)
+#define ZIO_CHILD_BIT(x) (1 << (x))
+#define ZIO_CHILD_BIT_IS_SET(val, x) ((val) & (1 << (x)))
+
enum zio_child {
ZIO_CHILD_VDEV = 0,
ZIO_CHILD_GANG,
@@ -223,6 +226,14 @@ enum zio_child {
ZIO_CHILD_TYPES
};
+#define ZIO_CHILD_VDEV_BIT ZIO_CHILD_BIT(ZIO_CHILD_VDEV)
+#define ZIO_CHILD_GANG_BIT ZIO_CHILD_BIT(ZIO_CHILD_GANG)
+#define ZIO_CHILD_DDT_BIT ZIO_CHILD_BIT(ZIO_CHILD_DDT)
+#define ZIO_CHILD_LOGICAL_BIT ZIO_CHILD_BIT(ZIO_CHILD_LOGICAL)
+#define ZIO_CHILD_ALL_BITS \
+ (ZIO_CHILD_VDEV_BIT | ZIO_CHILD_GANG_BIT | \
+ ZIO_CHILD_DDT_BIT | ZIO_CHILD_LOGICAL_BIT)
+
enum zio_wait_type {
ZIO_WAIT_READY = 0,
ZIO_WAIT_DONE,
diff --git a/module/zfs/zio.c b/module/zfs/zio.c
index 1d69d8d8d..cd0a473e0 100644
--- a/module/zfs/zio.c
+++ b/module/zfs/zio.c
@@ -491,21 +491,26 @@ zio_remove_child(zio_t *pio, zio_t *cio, zio_link_t *zl)
}
static boolean_t
-zio_wait_for_children(zio_t *zio, enum zio_child child, enum zio_wait_type wait)
+zio_wait_for_children(zio_t *zio, uint8_t childbits, enum zio_wait_type wait)
{
- uint64_t *countp = &zio->io_children[child][wait];
boolean_t waiting = B_FALSE;
mutex_enter(&zio->io_lock);
ASSERT(zio->io_stall == NULL);
- if (*countp != 0) {
- zio->io_stage >>= 1;
- ASSERT3U(zio->io_stage, !=, ZIO_STAGE_OPEN);
- zio->io_stall = countp;
- waiting = B_TRUE;
+ for (int c = 0; c < ZIO_CHILD_TYPES; c++) {
+ if (!(ZIO_CHILD_BIT_IS_SET(childbits, c)))
+ continue;
+
+ uint64_t *countp = &zio->io_children[c][wait];
+ if (*countp != 0) {
+ zio->io_stage >>= 1;
+ ASSERT3U(zio->io_stage, !=, ZIO_STAGE_OPEN);
+ zio->io_stall = countp;
+ waiting = B_TRUE;
+ break;
+ }
}
mutex_exit(&zio->io_lock);
-
return (waiting);
}
@@ -1296,9 +1301,10 @@ zio_write_compress(zio_t *zio)
* If our children haven't all reached the ready stage,
* wait for them and then repeat this pipeline stage.
*/
- if (zio_wait_for_children(zio, ZIO_CHILD_GANG, ZIO_WAIT_READY) ||
- zio_wait_for_children(zio, ZIO_CHILD_LOGICAL, ZIO_WAIT_READY))
+ if (zio_wait_for_children(zio, ZIO_CHILD_LOGICAL_BIT |
+ ZIO_CHILD_GANG_BIT, ZIO_WAIT_READY)) {
return (ZIO_PIPELINE_STOP);
+ }
if (!IO_IS_ALLOCATING(zio))
return (ZIO_PIPELINE_CONTINUE);
@@ -2229,8 +2235,9 @@ zio_gang_issue(zio_t *zio)
{
blkptr_t *bp = zio->io_bp;
- if (zio_wait_for_children(zio, ZIO_CHILD_GANG, ZIO_WAIT_DONE))
+ if (zio_wait_for_children(zio, ZIO_CHILD_GANG_BIT, ZIO_WAIT_DONE)) {
return (ZIO_PIPELINE_STOP);
+ }
ASSERT(BP_IS_GANG(bp) && zio->io_gang_leader == zio);
ASSERT(zio->io_child_type > ZIO_CHILD_GANG);
@@ -2561,8 +2568,9 @@ zio_ddt_read_done(zio_t *zio)
{
blkptr_t *bp = zio->io_bp;
- if (zio_wait_for_children(zio, ZIO_CHILD_DDT, ZIO_WAIT_DONE))
+ if (zio_wait_for_children(zio, ZIO_CHILD_DDT_BIT, ZIO_WAIT_DONE)) {
return (ZIO_PIPELINE_STOP);
+ }
ASSERT(BP_GET_DEDUP(bp));
ASSERT(BP_GET_PSIZE(bp) == zio->io_size);
@@ -3292,8 +3300,9 @@ zio_vdev_io_done(zio_t *zio)
vdev_ops_t *ops = vd ? vd->vdev_ops : &vdev_mirror_ops;
boolean_t unexpected_error = B_FALSE;
- if (zio_wait_for_children(zio, ZIO_CHILD_VDEV, ZIO_WAIT_DONE))
+ if (zio_wait_for_children(zio, ZIO_CHILD_VDEV_BIT, ZIO_WAIT_DONE)) {
return (ZIO_PIPELINE_STOP);
+ }
ASSERT(zio->io_type == ZIO_TYPE_READ || zio->io_type == ZIO_TYPE_WRITE);
@@ -3362,8 +3371,9 @@ zio_vdev_io_assess(zio_t *zio)
{
vdev_t *vd = zio->io_vd;
- if (zio_wait_for_children(zio, ZIO_CHILD_VDEV, ZIO_WAIT_DONE))
+ if (zio_wait_for_children(zio, ZIO_CHILD_VDEV_BIT, ZIO_WAIT_DONE)) {
return (ZIO_PIPELINE_STOP);
+ }
if (vd == NULL && !(zio->io_flags & ZIO_FLAG_CONFIG_WRITER))
spa_config_exit(zio->io_spa, SCL_ZIO, zio);
@@ -3578,9 +3588,10 @@ zio_ready(zio_t *zio)
zio_t *pio, *pio_next;
zio_link_t *zl = NULL;
- if (zio_wait_for_children(zio, ZIO_CHILD_GANG, ZIO_WAIT_READY) ||
- zio_wait_for_children(zio, ZIO_CHILD_DDT, ZIO_WAIT_READY))
+ if (zio_wait_for_children(zio, ZIO_CHILD_GANG_BIT | ZIO_CHILD_DDT_BIT,
+ ZIO_WAIT_READY)) {
return (ZIO_PIPELINE_STOP);
+ }
if (zio->io_ready) {
ASSERT(IO_IS_ALLOCATING(zio));
@@ -3721,11 +3732,9 @@ zio_done(zio_t *zio)
* If our children haven't all completed,
* wait for them and then repeat this pipeline stage.
*/
- if (zio_wait_for_children(zio, ZIO_CHILD_VDEV, ZIO_WAIT_DONE) ||
- zio_wait_for_children(zio, ZIO_CHILD_GANG, ZIO_WAIT_DONE) ||
- zio_wait_for_children(zio, ZIO_CHILD_DDT, ZIO_WAIT_DONE) ||
- zio_wait_for_children(zio, ZIO_CHILD_LOGICAL, ZIO_WAIT_DONE))
+ if (zio_wait_for_children(zio, ZIO_CHILD_ALL_BITS, ZIO_WAIT_DONE)) {
return (ZIO_PIPELINE_STOP);
+ }
/*
* If the allocation throttle is enabled, then update the accounting.
--
2.14.2

40
zfs-patches/0033-Clarify-zinject-8-explanation-of-e.patch Normal file
View File

@ -0,0 +1,40 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Olaf Faaland <faaland1@llnl.gov>
Date: Thu, 15 Feb 2018 09:50:06 -0800
Subject: [PATCH] Clarify zinject(8) explanation of -e
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Error injection of EIO or ENXIO simply sets the zio's io_error value,
rather than preventing the read or write from occurring. This is
important information as it affects how the probes must be used.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes #7172
(cherry picked from commit 26941ce90b8f056949a3bc29da606513af799781)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
man/man8/zinject.8 | 3 +++
1 file changed, 3 insertions(+)
diff --git a/man/man8/zinject.8 b/man/man8/zinject.8
index 50fecfb64..7f363974b 100644
--- a/man/man8/zinject.8
+++ b/man/man8/zinject.8
@@ -111,6 +111,9 @@ Specify
.BR "dtl" " for an ECHILD error,"
.BR "io" " for an EIO error where reopening the device will succeed, or"
.BR "nxio" " for an ENXIO error where reopening the device will fail."
+For EIO and ENXIO, the "failed" reads or writes still occur. The probe simply
+sets the error value reported by the I/O pipeline so it appears the read or
+write failed.
.TP
.BI "\-f" " frequency"
Only inject errors a fraction of the time. Expressed as a real number
--
2.14.2

199
zfs-patches/0034-Fix-config-issues-frame-size-and-headers.patch Normal file
View File

@ -0,0 +1,199 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: chrisrd <chris@onthe.net.au>
Date: Fri, 16 Feb 2018 07:58:23 +1100
Subject: [PATCH] Fix config issues: frame size and headers
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
1. With various (debug and/or tracing?) kernel options enabled it's
possible for 'struct inode' and 'struct super_block' to exceed the
default frame size, leaving errors like this in config.log:
build/conftest.c:116:1: error: the frame size of 1048 bytes is larger
than 1024 bytes [-Werror=frame-larger-than=]
Fix this by removing the frame size warning for config checks
2. Without the correct headers included, it's possible for declarations
to be missed, leaving errors like this in the config.log:
build/conftest.c:131:14: error: ‘struct nameidata’ declared inside
parameter list [-Werror]
Fix this by adding appropriate headers.
Note: Both these issues can result in silent config failures because
the compile failure is taken to mean "this option is not supported by
this kernel" rather than "there's something wrong with the config
test". This can lead to something merely annoying (compile failures) to
something potentially serious (miscompiled or misused kernel primitives
or functions). E.g. the fixes included here resulted in these
additional defines in zfs_config.h with linux v4.14.19:
Also, drive-by whitespace fixes in config/* files which don't mention
"GNU" (those ones look to be imported from elsewhere so leave them
alone).
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chris Dunlop <chris@onthe.net.au>
Closes #7169
(cherry picked from commit 5a84c60fb95ee7ca1393efe93d9d17c80497c6b6)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
config/deb.am | 8 ++++----
config/kernel-acl.m4 | 1 +
config/kernel-create-nameidata.m4 | 1 +
config/kernel-dentry-operations.m4 | 1 +
config/kernel-get-link.m4 | 2 +-
config/kernel-lookup-nameidata.m4 | 1 +
config/kernel-vm_node_stat.m4 | 2 +-
config/kernel.m4 | 2 +-
config/tgz.am | 4 ++--
config/user-libblkid.m4 | 2 +-
10 files changed, 14 insertions(+), 10 deletions(-)
diff --git a/config/deb.am b/config/deb.am
index 1b51f9316..58ab96e18 100644
--- a/config/deb.am
+++ b/config/deb.am
@@ -2,16 +2,16 @@ deb-local:
@(if test "${HAVE_DPKGBUILD}" = "no"; then \
echo -e "\n" \
"*** Required util ${DPKGBUILD} missing. Please install the\n" \
- "*** package for your distribution which provides ${DPKGBUILD},\n" \
+ "*** package for your distribution which provides ${DPKGBUILD},\n" \
"*** re-run configure, and try again.\n"; \
- exit 1; \
+ exit 1; \
fi; \
if test "${HAVE_ALIEN}" = "no"; then \
echo -e "\n" \
"*** Required util ${ALIEN} missing. Please install the\n" \
- "*** package for your distribution which provides ${ALIEN},\n" \
+ "*** package for your distribution which provides ${ALIEN},\n" \
"*** re-run configure, and try again.\n"; \
- exit 1; \
+ exit 1; \
fi)
deb-kmod: deb-local rpm-kmod
diff --git a/config/kernel-acl.m4 b/config/kernel-acl.m4
index 311484349..02cc020e5 100644
--- a/config/kernel-acl.m4
+++ b/config/kernel-acl.m4
@@ -184,6 +184,7 @@ AC_DEFUN([ZFS_AC_KERNEL_INODE_OPERATIONS_PERMISSION_WITH_NAMEIDATA], [
AC_MSG_CHECKING([whether iops->permission() wants nameidata])
ZFS_LINUX_TRY_COMPILE([
#include <linux/fs.h>
+ #include <linux/sched.h>
int permission_fn(struct inode *inode, int mask,
struct nameidata *nd) { return 0; }
diff --git a/config/kernel-create-nameidata.m4 b/config/kernel-create-nameidata.m4
index a71490a00..d4c155c57 100644
--- a/config/kernel-create-nameidata.m4
+++ b/config/kernel-create-nameidata.m4
@@ -5,6 +5,7 @@ AC_DEFUN([ZFS_AC_KERNEL_CREATE_NAMEIDATA], [
AC_MSG_CHECKING([whether iops->create() passes nameidata])
ZFS_LINUX_TRY_COMPILE([
#include <linux/fs.h>
+ #include <linux/sched.h>
#ifdef HAVE_MKDIR_UMODE_T
int inode_create(struct inode *inode ,struct dentry *dentry,
diff --git a/config/kernel-dentry-operations.m4 b/config/kernel-dentry-operations.m4
index 3182490c9..61f5a27af 100644
--- a/config/kernel-dentry-operations.m4
+++ b/config/kernel-dentry-operations.m4
@@ -5,6 +5,7 @@ AC_DEFUN([ZFS_AC_KERNEL_D_REVALIDATE_NAMEIDATA], [
AC_MSG_CHECKING([whether dops->d_revalidate() takes struct nameidata])
ZFS_LINUX_TRY_COMPILE([
#include <linux/dcache.h>
+ #include <linux/sched.h>
int revalidate (struct dentry *dentry,
struct nameidata *nidata) { return 0; }
diff --git a/config/kernel-get-link.m4 b/config/kernel-get-link.m4
index 022c49c54..3cda08c1b 100644
--- a/config/kernel-get-link.m4
+++ b/config/kernel-get-link.m4
@@ -41,7 +41,7 @@ AC_DEFUN([ZFS_AC_KERNEL_FOLLOW_LINK], [
AC_DEFINE(HAVE_FOLLOW_LINK_NAMEIDATA, 1,
[iops->follow_link() nameidata])
],[
- AC_MSG_ERROR(no; please file a bug report)
+ AC_MSG_ERROR(no; please file a bug report)
])
])
])
diff --git a/config/kernel-lookup-nameidata.m4 b/config/kernel-lookup-nameidata.m4
index 43f5fb4cb..5453be5e8 100644
--- a/config/kernel-lookup-nameidata.m4
+++ b/config/kernel-lookup-nameidata.m4
@@ -5,6 +5,7 @@ AC_DEFUN([ZFS_AC_KERNEL_LOOKUP_NAMEIDATA], [
AC_MSG_CHECKING([whether iops->lookup() passes nameidata])
ZFS_LINUX_TRY_COMPILE([
#include <linux/fs.h>
+ #include <linux/sched.h>
struct dentry *inode_lookup(struct inode *inode,
struct dentry *dentry, struct nameidata *nidata)
diff --git a/config/kernel-vm_node_stat.m4 b/config/kernel-vm_node_stat.m4
index e1c42f884..5dcd9d827 100644
--- a/config/kernel-vm_node_stat.m4
+++ b/config/kernel-vm_node_stat.m4
@@ -7,7 +7,7 @@ AC_DEFUN([ZFS_AC_KERNEL_VM_NODE_STAT], [
ZFS_LINUX_TRY_COMPILE([
#include <linux/mm.h>
#include <linux/vmstat.h>
- ],[
+ ],[
int a __attribute__ ((unused)) = NR_VM_NODE_STAT_ITEMS;
long x __attribute__ ((unused)) =
atomic_long_read(&vm_node_stat[0]);
diff --git a/config/kernel.m4 b/config/kernel.m4
index b83f021e8..d4a84b2b2 100644
--- a/config/kernel.m4
+++ b/config/kernel.m4
@@ -722,7 +722,7 @@ AC_DEFUN([ZFS_LINUX_COMPILE_IFELSE], [
modpost_flag=''
test "x$enable_linux_builtin" = xyes && modpost_flag='modpost=true' # fake modpost stage
AS_IF(
- [AC_TRY_COMMAND(cp conftest.c conftest.h build && make [$2] -C $LINUX_OBJ EXTRA_CFLAGS="-Werror $EXTRA_KCFLAGS" $ARCH_UM M=$PWD/build $modpost_flag) >/dev/null && AC_TRY_COMMAND([$3])],
+ [AC_TRY_COMMAND(cp conftest.c conftest.h build && make [$2] -C $LINUX_OBJ EXTRA_CFLAGS="-Werror $FRAME_LARGER_THAN $EXTRA_KCFLAGS" $ARCH_UM M=$PWD/build $modpost_flag) >/dev/null && AC_TRY_COMMAND([$3])],
[$4],
[_AC_MSG_LOG_CONFTEST m4_ifvaln([$5],[$5])]
)
diff --git a/config/tgz.am b/config/tgz.am
index 2997b1de2..0657d045d 100644
--- a/config/tgz.am
+++ b/config/tgz.am
@@ -2,9 +2,9 @@ tgz-local:
@(if test "${HAVE_ALIEN}" = "no"; then \
echo -e "\n" \
"*** Required util ${ALIEN} missing. Please install the\n" \
- "*** package for your distribution which provides ${ALIEN},\n" \
+ "*** package for your distribution which provides ${ALIEN},\n" \
"*** re-run configure, and try again.\n"; \
- exit 1; \
+ exit 1; \
fi)
tgz-kmod: tgz-local rpm-kmod
diff --git a/config/user-libblkid.m4 b/config/user-libblkid.m4
index 5bc7f466a..88e6f990b 100644
--- a/config/user-libblkid.m4
+++ b/config/user-libblkid.m4
@@ -6,7 +6,7 @@ AC_DEFUN([ZFS_AC_CONFIG_USER_LIBBLKID], [
LIBBLKID=
AC_CHECK_HEADER([blkid/blkid.h], [], [AC_MSG_FAILURE([
- *** blkid.h missing, libblkid-devel package required])])
+ *** blkid.h missing, libblkid-devel package required])])
AC_SUBST([LIBBLKID], ["-lblkid"])
AC_DEFINE([HAVE_LIBBLKID], 1, [Define if you have libblkid])
--
2.14.2

38
zfs-patches/0035-Correct-count_uberblocks-in-mmp.kshlib.patch Normal file
View File

@ -0,0 +1,38 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Giuseppe Di Natale <dinatale2@users.noreply.github.com>
Date: Tue, 20 Feb 2018 16:28:52 -0800
Subject: [PATCH] Correct count_uberblocks in mmp.kshlib
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
A log_must call was causing count_uberblocks to return more
than just the uberblock count. Remove the log_must since it
was only logging a sleep.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Signed-off-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Closes #7191
(cherry picked from commit d5b10b3ef3b5e82b9e5a9bee122f1a0be007dae2)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
tests/zfs-tests/tests/functional/mmp/mmp.kshlib | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/zfs-tests/tests/functional/mmp/mmp.kshlib b/tests/zfs-tests/tests/functional/mmp/mmp.kshlib
index 4c46ae7a2..1f8d66fd9 100644
--- a/tests/zfs-tests/tests/functional/mmp/mmp.kshlib
+++ b/tests/zfs-tests/tests/functional/mmp/mmp.kshlib
@@ -198,6 +198,6 @@ function count_uberblocks # pool duration
typeset -i duration=$2
typeset hist_path="/proc/spl/kstat/zfs/$pool/multihost"
- log_must sleep $duration
+ sleep $duration
echo $(cat "$hist_path" | sed '1,2d' | wc -l)
}
--
2.14.2

131
zfs-patches/0036-Add-SMART-attributes-for-SSD-and-NVMe.patch Normal file
View File

@ -0,0 +1,131 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: bunder2015 <omfgbunder@gmail.com>
Date: Wed, 21 Feb 2018 16:52:47 -0500
Subject: [PATCH] Add SMART attributes for SSD and NVMe
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
This adds the SMART attributes required to probe Samsung SSD and NVMe
(and possibly others) disks when using the "zpool status -c" command.
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: bunder2015 <omfgbunder@gmail.com>
Closes #7183
Closes #7193
(cherry picked from commit c705d8386b21b08aefdc62b6b1a556aab6717316)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zpool/Makefile.am | 2 ++
cmd/zpool/zpool.d/nvme_err | 1 +
cmd/zpool/zpool.d/smart | 23 +++++++++++++++++++++--
3 files changed, 24 insertions(+), 2 deletions(-)
create mode 120000 cmd/zpool/zpool.d/nvme_err
diff --git a/cmd/zpool/Makefile.am b/cmd/zpool/Makefile.am
index 6eff1d143..c7b8b76e3 100644
--- a/cmd/zpool/Makefile.am
+++ b/cmd/zpool/Makefile.am
@@ -60,6 +60,7 @@ dist_zpoolexec_SCRIPTS = \
zpool.d/pend_sec \
zpool.d/off_ucor \
zpool.d/ata_err \
+ zpool.d/nvme_err \
zpool.d/pwr_cyc \
zpool.d/upath \
zpool.d/vendor
@@ -98,6 +99,7 @@ zpoolconfdefaults = \
pend_sec \
off_ucor \
ata_err \
+ nvme_err \
pwr_cyc \
upath \
vendor
diff --git a/cmd/zpool/zpool.d/nvme_err b/cmd/zpool/zpool.d/nvme_err
new file mode 120000
index 000000000..94f22861f
--- /dev/null
+++ b/cmd/zpool/zpool.d/nvme_err
@@ -0,0 +1 @@
+smart
\ No newline at end of file
diff --git a/cmd/zpool/zpool.d/smart b/cmd/zpool/zpool.d/smart
index 3721f30ed..4bc3af39d 100755
--- a/cmd/zpool/zpool.d/smart
+++ b/cmd/zpool/zpool.d/smart
@@ -23,6 +23,7 @@ off_ucor: Show SMART offline uncorrectable errors (ATA).
ata_err: Show SMART ATA errors (ATA).
pwr_cyc: Show SMART power cycle count (ATA).
serial: Show disk serial number.
+nvme_err: Show SMART NVMe errors (NVMe).
"
script=$(basename "$0")
@@ -37,7 +38,7 @@ smartctl_path=$(which smartctl)
if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
raw_out=$(eval "sudo $smartctl_path -a $VDEV_UPATH")
- # Are we a SAS or ATA drive? Look for the right line in smartctl:
+ # What kind of drive are we? Look for the right line in smartctl:
#
# SAS:
# Transport protocol: SAS
@@ -45,7 +46,10 @@ if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
# SATA:
# ATA Version is: 8
#
- type=$(echo "$raw_out" | grep -m 1 -Eo '^ATA|SAS$')
+ # NVMe:
+ # SMART/Health Information (NVMe Log 0xnn, NSID 0xnn)
+ #
+ type=$(echo "$raw_out" | grep -m 1 -Eo '^ATA|NVMe|SAS$')
out=$(echo "$raw_out" | awk '
# SAS specific
/read:/{print "rrd="$4"\nr_cor="$5"\nr_proc="$7"\nr_ucor="$8}
@@ -71,10 +75,21 @@ if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
# SATA common
/Temperature_Celsius/{print "temp="$10}
+/Airflow_Temperature_Cel/{print "temp="$10}
/SMART overall-health self-assessment test result:/{print "health="$6}
/Power_On_Hours/{print "hours_on="$10}
/Serial Number:/{print "serial="$3}
+# NVMe common
+/Temperature:/{print "temp="$2}
+/SMART overall-health self-assessment test result:/{print "health="$6}
+/Power On Hours:/{gsub("[^0-9]","",$4); print "hours_on="$4}
+/Serial Number:/{print "serial="$3}
+/Power Cycles:/{print "pwr_cyc="$3}
+
+# NVMe specific
+/Media and Data Integrity Errors:/{print "nvme_err="$6}
+
END {ORS="\n"; print ""}
');
fi
@@ -94,6 +109,8 @@ smart)
scripts="temp|health|r_ucor|w_ucor"
elif [ "$type" = "ATA" ] ; then
scripts="temp|health|ata_err|realloc|rep_ucor|cmd_to|pend_sec|off_ucor"
+ elif [ "$type" = "NVMe" ] ; then
+ scripts="temp|health|nvme_err"
fi
;;
smartx)
@@ -102,6 +119,8 @@ smartx)
scripts="hours_on|defect|nonmed|r_proc|w_proc"
elif [ "$type" = "ATA" ] ; then
scripts="hours_on|pwr_cyc"
+ elif [ "$type" = "NVMe" ] ; then
+ scripts="hours_on|pwr_cyc"
fi
;;
*)
--
2.14.2

52
zfs-patches/0037-Allow-modprobe-to-fail-when-called-within-systemd.patch Normal file
View File

@ -0,0 +1,52 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Matthew Thode <mthode@mthode.org>
Date: Wed, 21 Feb 2018 22:45:35 +0000
Subject: [PATCH] Allow modprobe to fail when called within systemd
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
This allows for systems with zfs built into the kernel manually to run
these services. Otherwise the service will fail to start.
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed-by: Kash Pande <kash@tripleback.net>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Matthew Thode <mthode@mthode.org>
Closes #7174
(cherry picked from commit 30ac8de48a2bb680e15dfe8879faaa7b461b771c)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
etc/systemd/system/zfs-import-cache.service.in | 2 +-
etc/systemd/system/zfs-import-scan.service.in | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/etc/systemd/system/zfs-import-cache.service.in b/etc/systemd/system/zfs-import-cache.service.in
index 9d677f8dd..726c468ca 100644
--- a/etc/systemd/system/zfs-import-cache.service.in
+++ b/etc/systemd/system/zfs-import-cache.service.in
@@ -12,7 +12,7 @@ ConditionPathExists=@sysconfdir@/zfs/zpool.cache
[Service]
Type=oneshot
RemainAfterExit=yes
-ExecStartPre=/sbin/modprobe zfs
+ExecStartPre=-/sbin/modprobe zfs
ExecStart=@sbindir@/zpool import -c @sysconfdir@/zfs/zpool.cache -aN
[Install]
diff --git a/etc/systemd/system/zfs-import-scan.service.in b/etc/systemd/system/zfs-import-scan.service.in
index cc428b0d0..8fe2c1072 100644
--- a/etc/systemd/system/zfs-import-scan.service.in
+++ b/etc/systemd/system/zfs-import-scan.service.in
@@ -11,7 +11,7 @@ ConditionPathExists=!@sysconfdir@/zfs/zpool.cache
[Service]
Type=oneshot
RemainAfterExit=yes
-ExecStartPre=/sbin/modprobe zfs
+ExecStartPre=-/sbin/modprobe zfs
ExecStart=@sbindir@/zpool import -aN -d /dev/disk/by-id -o cachefile=none
[Install]
--
2.14.2

85
zfs-patches/0038-Linux-4.16-compat-use-correct-_dec_and_test.patch Normal file
View File

@ -0,0 +1,85 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Tony Hutter <hutter2@llnl.gov>
Date: Thu, 22 Feb 2018 09:02:06 -0800
Subject: [PATCH] Linux 4.16 compat: use correct *_dec_and_test()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Use refcount_dec_and_test() on 4.16+ kernels, atomic_dec_and_test()
on older kernels. https://lwn.net/Articles/714974/
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes: #7179
Closes: #7211
(cherry picked from commit e5ba614d05bde1e98ba46d5b75cf6b8a0701c2ca)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/linux/vfs_compat.h | 6 +++++-
config/kernel-acl-refcount.m4 | 20 ++++++++++++++++++++
config/kernel.m4 | 1 +
3 files changed, 26 insertions(+), 1 deletion(-)
create mode 100644 config/kernel-acl-refcount.m4
diff --git a/include/linux/vfs_compat.h b/include/linux/vfs_compat.h
index 7fcf3c055..6347268af 100644
--- a/include/linux/vfs_compat.h
+++ b/include/linux/vfs_compat.h
@@ -288,9 +288,13 @@ zpl_posix_acl_release(struct posix_acl *acl)
{
if ((acl == NULL) || (acl == ACL_NOT_CACHED))
return;
-
+#ifdef HAVE_ACL_REFCOUNT
+ if (refcount_dec_and_test(&acl->a_refcount))
+ zpl_posix_acl_release_impl(acl);
+#else
if (atomic_dec_and_test(&acl->a_refcount))
zpl_posix_acl_release_impl(acl);
+#endif
}
#endif /* HAVE_POSIX_ACL_RELEASE */
diff --git a/config/kernel-acl-refcount.m4 b/config/kernel-acl-refcount.m4
new file mode 100644
index 000000000..43e3c442d
--- /dev/null
+++ b/config/kernel-acl-refcount.m4
@@ -0,0 +1,20 @@
+dnl #
+dnl # 4.16 kernel: check if struct posix_acl acl.a_refcount is a refcount_t.
+dnl # It's an atomic_t on older kernels.
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_ACL_HAS_REFCOUNT], [
+ AC_MSG_CHECKING([whether posix_acl has refcount_t])
+ ZFS_LINUX_TRY_COMPILE([
+ #include <linux/backing-dev.h>
+ #include <linux/refcount.h>
+ #include <linux/posix_acl.h>
+ ],[
+ struct posix_acl acl;
+ refcount_t *r __attribute__ ((unused)) = &acl.a_refcount;
+ ],[
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(HAVE_ACL_REFCOUNT, 1, [posix_acl has refcount_t])
+ ],[
+ AC_MSG_RESULT(no)
+ ])
+])
diff --git a/config/kernel.m4 b/config/kernel.m4
index d4a84b2b2..7bb86a96e 100644
--- a/config/kernel.m4
+++ b/config/kernel.m4
@@ -124,6 +124,7 @@ AC_DEFUN([ZFS_AC_CONFIG_KERNEL], [
ZFS_AC_KERNEL_HAVE_GENERIC_SETXATTR
ZFS_AC_KERNEL_CURRENT_TIME
ZFS_AC_KERNEL_VM_NODE_STAT
+ ZFS_AC_KERNEL_ACL_HAS_REFCOUNT
AS_IF([test "$LINUX_OBJ" != "$LINUX"], [
KERNELMAKE_PARAMS="$KERNELMAKE_PARAMS O=$LINUX_OBJ"
--
2.14.2

39
zfs-patches/0039-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch Normal file
View File

@ -0,0 +1,39 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Olaf Faaland <faaland1@llnl.gov>
Date: Thu, 22 Feb 2018 09:14:46 -0800
Subject: [PATCH] Do not initiate MMP writes while pool is suspended
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
While the pool is suspended on host A, it may be imported on host B.
If host A continued to write MMP blocks, it would be blindly
overwriting MMP blocks written by host B, and the blocks written by
host A would have outdated txg information.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes #7182
(cherry picked from commit b1f61f05b4b2edc77ec5f3d28d041d7b3c873f01)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/mmp.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index 1dfb7c05b..d003d79de 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -437,7 +437,7 @@ mmp_thread(spa_t *spa)
zio_suspend(spa, NULL);
}
- if (multihost)
+ if (multihost && !suspended)
mmp_write_uberblock(spa);
CALLB_CPR_SAFE_BEGIN(&cpr);
--
2.14.2

237
zfs-patches/0040-Report-duration-and-error-in-mmp_history-entries.patch Normal file
View File

@ -0,0 +1,237 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Olaf Faaland <faaland1@llnl.gov>
Date: Thu, 22 Feb 2018 15:34:34 -0800
Subject: [PATCH] Report duration and error in mmp_history entries
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
After an MMP write completes, update the relevant mmp_history entry
with the time between submission and completion, and the error
status of the write.
[faaland1@toss3a zfs]$ cat /proc/spl/kstat/zfs/pool/multihost
39 0 0x01 100 8800 69147946270893 72723903122926
id txg timestamp error duration mmp_delay vdev_guid
10607 1166 1518985089 0 138301 637785455 4882...
10608 1166 1518985089 0 136154 635407747 1151...
10609 1166 1518985089 0 803618560 633048078 9740...
10610 1166 1518985090 0 144826 633048078 4882...
10611 1166 1518985090 0 164527 666187671 1151...
Where duration = gethrtime_in_done_fn - gethrtime_at_submission, and
error = zio->io_error.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes #7190
(cherry picked from commit 2644784f49a6b6be063d54ca0e1662ee6bef7ebd)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/sys/mmp.h | 1 +
include/sys/spa.h | 4 +++-
include/sys/vdev_impl.h | 1 +
module/zfs/mmp.c | 13 ++++++++++--
module/zfs/spa_stats.c | 54 ++++++++++++++++++++++++++++++++++++++++---------
5 files changed, 60 insertions(+), 13 deletions(-)
diff --git a/include/sys/mmp.h b/include/sys/mmp.h
index 5b2fea1a6..1ce685f9c 100644
--- a/include/sys/mmp.h
+++ b/include/sys/mmp.h
@@ -42,6 +42,7 @@ typedef struct mmp_thread {
uint64_t mmp_delay; /* decaying avg ns between MMP writes */
uberblock_t mmp_ub; /* last ub written by sync */
zio_t *mmp_zio_root; /* root of mmp write zios */
+ uint64_t mmp_kstat_id; /* unique id for next MMP write kstat */
} mmp_thread_t;
diff --git a/include/sys/spa.h b/include/sys/spa.h
index 67235871f..53fa5514a 100644
--- a/include/sys/spa.h
+++ b/include/sys/spa.h
@@ -759,8 +759,10 @@ extern txg_stat_t *spa_txg_history_init_io(spa_t *, uint64_t,
struct dsl_pool *);
extern void spa_txg_history_fini_io(spa_t *, txg_stat_t *);
extern void spa_tx_assign_add_nsecs(spa_t *spa, uint64_t nsecs);
+extern int spa_mmp_history_set(spa_t *spa, uint64_t mmp_kstat_id, int io_error,
+ hrtime_t duration);
extern void spa_mmp_history_add(uint64_t txg, uint64_t timestamp,
- uint64_t mmp_delay, vdev_t *vd, int label);
+ uint64_t mmp_delay, vdev_t *vd, int label, uint64_t mmp_kstat_id);
/* Pool configuration locks */
extern int spa_config_tryenter(spa_t *spa, int locks, void *tag, krw_t rw);
diff --git a/include/sys/vdev_impl.h b/include/sys/vdev_impl.h
index 4c2e3cd2e..13c495822 100644
--- a/include/sys/vdev_impl.h
+++ b/include/sys/vdev_impl.h
@@ -238,6 +238,7 @@ struct vdev {
vdev_aux_t vdev_label_aux; /* on-disk aux state */
uint64_t vdev_leaf_zap;
hrtime_t vdev_mmp_pending; /* 0 if write finished */
+ uint64_t vdev_mmp_kstat_id; /* to find kstat entry */
/*
* For DTrace to work in userland (libzpool) context, these fields must
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index d003d79de..ee8e9201b 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -135,6 +135,7 @@ mmp_init(spa_t *spa)
mutex_init(&mmp->mmp_thread_lock, NULL, MUTEX_DEFAULT, NULL);
cv_init(&mmp->mmp_thread_cv, NULL, CV_DEFAULT, NULL);
mutex_init(&mmp->mmp_io_lock, NULL, MUTEX_DEFAULT, NULL);
+ mmp->mmp_kstat_id = 1;
}
void
@@ -244,7 +245,8 @@ mmp_write_done(zio_t *zio)
mmp_thread_t *mts = zio->io_private;
mutex_enter(&mts->mmp_io_lock);
- vd->vdev_mmp_pending = 0;
+ uint64_t mmp_kstat_id = vd->vdev_mmp_kstat_id;
+ hrtime_t mmp_write_duration = gethrtime() - vd->vdev_mmp_pending;
if (zio->io_error)
goto unlock;
@@ -278,9 +280,15 @@ mmp_write_done(zio_t *zio)
mts->mmp_last_write = gethrtime();
unlock:
+ vd->vdev_mmp_pending = 0;
+ vd->vdev_mmp_kstat_id = 0;
+
mutex_exit(&mts->mmp_io_lock);
spa_config_exit(spa, SCL_STATE, mmp_tag);
+ spa_mmp_history_set(spa, mmp_kstat_id, zio->io_error,
+ mmp_write_duration);
+
abd_free(zio->io_abd);
}
@@ -333,6 +341,7 @@ mmp_write_uberblock(spa_t *spa)
ub->ub_mmp_magic = MMP_MAGIC;
ub->ub_mmp_delay = mmp->mmp_delay;
vd->vdev_mmp_pending = gethrtime();
+ vd->vdev_mmp_kstat_id = mmp->mmp_kstat_id++;
zio_t *zio = zio_null(mmp->mmp_zio_root, spa, NULL, NULL, NULL, flags);
abd_t *ub_abd = abd_alloc_for_io(VDEV_UBERBLOCK_SIZE(vd), B_TRUE);
@@ -350,7 +359,7 @@ mmp_write_uberblock(spa_t *spa)
flags | ZIO_FLAG_DONT_PROPAGATE);
spa_mmp_history_add(ub->ub_txg, ub->ub_timestamp, ub->ub_mmp_delay, vd,
- label);
+ label, vd->vdev_mmp_kstat_id);
zio_nowait(zio);
}
diff --git a/module/zfs/spa_stats.c b/module/zfs/spa_stats.c
index 7ca359806..58967e9fc 100644
--- a/module/zfs/spa_stats.c
+++ b/module/zfs/spa_stats.c
@@ -718,21 +718,24 @@ spa_io_history_destroy(spa_t *spa)
*/
typedef struct spa_mmp_history {
+ uint64_t mmp_kstat_id; /* unique # for updates */
uint64_t txg; /* txg of last sync */
uint64_t timestamp; /* UTC time of of last sync */
uint64_t mmp_delay; /* nanosec since last MMP write */
uint64_t vdev_guid; /* unique ID of leaf vdev */
char *vdev_path;
uint64_t vdev_label; /* vdev label */
+ int io_error; /* error status of MMP write */
+ hrtime_t duration; /* time from submission to completion */
list_node_t smh_link;
} spa_mmp_history_t;
static int
spa_mmp_history_headers(char *buf, size_t size)
{
- (void) snprintf(buf, size, "%-10s %-10s %-12s %-24s %-10s %s\n",
- "txg", "timestamp", "mmp_delay", "vdev_guid", "vdev_label",
- "vdev_path");
+ (void) snprintf(buf, size, "%-10s %-10s %-10s %-6s %-10s %-12s %-24s "
+ "%-10s %s\n", "id", "txg", "timestamp", "error", "duration",
+ "mmp_delay", "vdev_guid", "vdev_label", "vdev_path");
return (0);
}
@@ -741,11 +744,12 @@ spa_mmp_history_data(char *buf, size_t size, void *data)
{
spa_mmp_history_t *smh = (spa_mmp_history_t *)data;
- (void) snprintf(buf, size, "%-10llu %-10llu %-12llu %-24llu %-10llu "
- "%s\n",
- (u_longlong_t)smh->txg, (u_longlong_t)smh->timestamp,
- (u_longlong_t)smh->mmp_delay, (u_longlong_t)smh->vdev_guid,
- (u_longlong_t)smh->vdev_label,
+ (void) snprintf(buf, size, "%-10llu %-10llu %-10llu %-6lld %-10lld "
+ "%-12llu %-24llu %-10llu %s\n",
+ (u_longlong_t)smh->mmp_kstat_id, (u_longlong_t)smh->txg,
+ (u_longlong_t)smh->timestamp, (longlong_t)smh->io_error,
+ (longlong_t)smh->duration, (u_longlong_t)smh->mmp_delay,
+ (u_longlong_t)smh->vdev_guid, (u_longlong_t)smh->vdev_label,
(smh->vdev_path ? smh->vdev_path : "-"));
return (0);
@@ -861,11 +865,40 @@ spa_mmp_history_destroy(spa_t *spa)
}
/*
- * Add a new MMP update to historical record.
+ * Set MMP write duration and error status in existing record.
+ */
+int
+spa_mmp_history_set(spa_t *spa, uint64_t mmp_kstat_id, int io_error,
+ hrtime_t duration)
+{
+ spa_stats_history_t *ssh = &spa->spa_stats.mmp_history;
+ spa_mmp_history_t *smh;
+ int error = ENOENT;
+
+ if (zfs_multihost_history == 0 && ssh->size == 0)
+ return (0);
+
+ mutex_enter(&ssh->lock);
+ for (smh = list_head(&ssh->list); smh != NULL;
+ smh = list_next(&ssh->list, smh)) {
+ if (smh->mmp_kstat_id == mmp_kstat_id) {
+ smh->io_error = io_error;
+ smh->duration = duration;
+ error = 0;
+ break;
+ }
+ }
+ mutex_exit(&ssh->lock);
+
+ return (error);
+}
+
+/*
+ * Add a new MMP write to historical record.
*/
void
spa_mmp_history_add(uint64_t txg, uint64_t timestamp, uint64_t mmp_delay,
- vdev_t *vd, int label)
+ vdev_t *vd, int label, uint64_t mmp_kstat_id)
{
spa_t *spa = vd->vdev_spa;
spa_stats_history_t *ssh = &spa->spa_stats.mmp_history;
@@ -882,6 +915,7 @@ spa_mmp_history_add(uint64_t txg, uint64_t timestamp, uint64_t mmp_delay,
if (vd->vdev_path)
smh->vdev_path = strdup(vd->vdev_path);
smh->vdev_label = label;
+ smh->mmp_kstat_id = mmp_kstat_id;
mutex_enter(&ssh->lock);
--
2.14.2

451
zfs-patches/0041-Fix-free-memory-calculation-on-v3.14.patch Normal file
View File

@ -0,0 +1,451 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: chrisrd <chris@onthe.net.au>
Date: Sat, 24 Feb 2018 03:50:06 +1100
Subject: [PATCH] Fix free memory calculation on v3.14+
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Provide infrastructure to auto-configure to enum and API changes in the
global page stats used for our free memory calculations.
arc_free_memory has been broken since an API change in Linux v3.14:
2016-07-28 v4.8 599d0c95 mm, vmscan: move LRU lists to node
2016-07-28 v4.8 75ef7184 mm, vmstat: add infrastructure for per-node
vmstats
These commits moved some of global_page_state() into
global_node_page_state(). The API change was particularly egregious as,
instead of breaking the old code, it silently did the wrong thing and we
continued using global_page_state() where we should have been using
global_node_page_state(), thus indexing into the wrong array via
NR_SLAB_RECLAIMABLE et al.
There have been further API changes along the way:
2017-07-06 v4.13 385386cf mm: vmstat: move slab statistics from zone to
node counters
2017-09-06 v4.14 c41f012a mm: rename global_page_state to
global_zone_page_state
...and various (incomplete, as it turns out) attempts to accomodate
these changes in ZoL:
2017-08-24 2209e409 Linux 4.8+ compatibility fix for vm stats
2017-09-16 787acae0 Linux 3.14 compat: IO acct, global_page_state, etc
2017-09-19 661907e6 Linux 4.14 compat: IO acct, global_page_state, etc
The config infrastructure provided here resolves these issues going back
to the original API change in v3.14 and is robust against further Linux
changes in this area.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Chris Dunlop <chris@onthe.net.au>
Closes #7170
(cherry picked from commit 338523dd6ec641cc4d552c3f67e1becfb9e22b0a)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/linux/Makefile.am | 3 +-
scripts/Makefile.am | 2 +
include/linux/page_compat.h | 78 ++++++++++++++++++++++++++
module/zfs/arc.c | 23 ++------
config/kernel-global_page_state.m4 | 109 +++++++++++++++++++++++++++++++++++++
config/kernel-vm_node_stat.m4 | 22 --------
config/kernel.m4 | 2 +-
scripts/enum-extract.pl | 58 ++++++++++++++++++++
8 files changed, 256 insertions(+), 41 deletions(-)
create mode 100644 include/linux/page_compat.h
create mode 100644 config/kernel-global_page_state.m4
delete mode 100644 config/kernel-vm_node_stat.m4
create mode 100755 scripts/enum-extract.pl
diff --git a/include/linux/Makefile.am b/include/linux/Makefile.am
index 9bb0b3493..89c2689f6 100644
--- a/include/linux/Makefile.am
+++ b/include/linux/Makefile.am
@@ -9,7 +9,8 @@ KERNEL_H = \
$(top_srcdir)/include/linux/kmap_compat.h \
$(top_srcdir)/include/linux/simd_x86.h \
$(top_srcdir)/include/linux/simd_aarch64.h \
- $(top_srcdir)/include/linux/mod_compat.h
+ $(top_srcdir)/include/linux/mod_compat.h \
+ $(top_srcdir)/include/linux/page_compat.h
USER_H =
diff --git a/scripts/Makefile.am b/scripts/Makefile.am
index 74b8b31a5..5a8abd135 100644
--- a/scripts/Makefile.am
+++ b/scripts/Makefile.am
@@ -5,6 +5,7 @@ EXTRA_DIST = dkms.mkconf dkms.postbuild kmodtool zfs2zol-patch.sed cstyle.pl
pkgdatadir = $(datadir)/@PACKAGE@
dist_pkgdata_SCRIPTS = \
$(top_builddir)/scripts/common.sh \
+ $(top_srcdir)/scripts/enum-extract.pl \
$(top_srcdir)/scripts/zimport.sh \
$(top_srcdir)/scripts/zfs.sh \
$(top_srcdir)/scripts/zfs-tests.sh \
@@ -15,3 +16,4 @@ dist_pkgdata_SCRIPTS = \
$(top_srcdir)/scripts/zpios-survey.sh \
$(top_srcdir)/scripts/smb.sh \
$(top_srcdir)/scripts/zfs-helpers.sh
+
diff --git a/include/linux/page_compat.h b/include/linux/page_compat.h
new file mode 100644
index 000000000..95acb7d53
--- /dev/null
+++ b/include/linux/page_compat.h
@@ -0,0 +1,78 @@
+#ifndef _ZFS_PAGE_COMPAT_H
+#define _ZFS_PAGE_COMPAT_H
+
+/*
+ * We have various enum members moving between two separate enum types,
+ * and accessed by different functions at various times. Centralise the
+ * insanity.
+ *
+ * < v4.8: all enums in zone_stat_item, via global_page_state()
+ * v4.8: some enums moved to node_stat_item, global_node_page_state() introduced
+ * v4.13: some enums moved from zone_stat_item to node_state_item
+ * v4.14: global_page_state() rename to global_zone_page_state()
+ *
+ * The defines used here are created by config/kernel-global_page_state.m4
+ */
+
+/*
+ * Create our own accessor functions to follow the Linux API changes
+ */
+#if defined(ZFS_GLOBAL_ZONE_PAGE_STATE)
+
+/* global_zone_page_state() introduced */
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_FILE_PAGES)
+#define nr_file_pages() global_node_page_state(NR_FILE_PAGES)
+#else
+#define nr_file_pages() global_zone_page_state(NR_FILE_PAGES)
+#endif
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_INACTIVE_ANON)
+#define nr_inactive_anon_pages() global_node_page_state(NR_INACTIVE_ANON)
+#else
+#define nr_inactive_anon_pages() global_zone_page_state(NR_INACTIVE_ANON)
+#endif
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_INACTIVE_FILE)
+#define nr_inactive_file_pages() global_node_page_state(NR_INACTIVE_FILE)
+#else
+#define nr_inactive_file_pages() global_zone_page_state(NR_INACTIVE_FILE)
+#endif
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_SLAB_RECLAIMABLE)
+#define nr_slab_reclaimable_pages() global_node_page_state(NR_SLAB_RECLAIMABLE)
+#else
+#define nr_slab_reclaimable_pages() global_zone_page_state(NR_SLAB_RECLAIMABLE)
+#endif
+
+#elif defined(ZFS_GLOBAL_NODE_PAGE_STATE)
+
+/* global_node_page_state() introduced */
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_FILE_PAGES)
+#define nr_file_pages() global_node_page_state(NR_FILE_PAGES)
+#else
+#define nr_file_pages() global_page_state(NR_FILE_PAGES)
+#endif
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_INACTIVE_ANON)
+#define nr_inactive_anon_pages() global_node_page_state(NR_INACTIVE_ANON)
+#else
+#define nr_inactive_anon_pages() global_page_state(NR_INACTIVE_ANON)
+#endif
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_INACTIVE_FILE)
+#define nr_inactive_file_pages() global_node_page_state(NR_INACTIVE_FILE)
+#else
+#define nr_inactive_file_pages() global_page_state(NR_INACTIVE_FILE)
+#endif
+#if defined(ZFS_ENUM_NODE_STAT_ITEM_NR_SLAB_RECLAIMABLE)
+#define nr_slab_reclaimable_pages() global_node_page_state(NR_SLAB_RECLAIMABLE)
+#else
+#define nr_slab_reclaimable_pages() global_page_state(NR_SLAB_RECLAIMABLE)
+#endif
+
+#else
+
+/* global_page_state() only */
+#define nr_file_pages() global_page_state(NR_FILE_PAGES)
+#define nr_inactive_anon_pages() global_page_state(NR_INACTIVE_ANON)
+#define nr_inactive_file_pages() global_page_state(NR_INACTIVE_FILE)
+#define nr_slab_reclaimable_pages() global_page_state(NR_SLAB_RECLAIMABLE)
+
+#endif /* ZFS_GLOBAL_ZONE_PAGE_STATE */
+
+#endif /* _ZFS_PAGE_COMPAT_H */
diff --git a/module/zfs/arc.c b/module/zfs/arc.c
index 9d1d0db1d..236794672 100644
--- a/module/zfs/arc.c
+++ b/module/zfs/arc.c
@@ -280,6 +280,7 @@
#include <sys/fs/swapnode.h>
#include <sys/zpl.h>
#include <linux/mm_compat.h>
+#include <linux/page_compat.h>
#endif
#include <sys/callb.h>
#include <sys/kstat.h>
@@ -4016,17 +4017,11 @@ arc_free_memory(void)
si_meminfo(&si);
return (ptob(si.freeram - si.freehigh));
#else
-#ifdef ZFS_GLOBAL_NODE_PAGE_STATE
return (ptob(nr_free_pages() +
- global_node_page_state(NR_INACTIVE_FILE) +
- global_node_page_state(NR_INACTIVE_ANON) +
- global_node_page_state(NR_SLAB_RECLAIMABLE)));
-#else
- return (ptob(nr_free_pages() +
- global_page_state(NR_INACTIVE_FILE) +
- global_page_state(NR_INACTIVE_ANON) +
- global_page_state(NR_SLAB_RECLAIMABLE)));
-#endif /* ZFS_GLOBAL_NODE_PAGE_STATE */
+ nr_inactive_file_pages() +
+ nr_inactive_anon_pages() +
+ nr_slab_reclaimable_pages()));
+
#endif /* CONFIG_HIGHMEM */
#else
return (spa_get_random(arc_all_memory() * 20 / 100));
@@ -4437,13 +4432,7 @@ arc_evictable_memory(void)
* Scale reported evictable memory in proportion to page cache, cap
* at specified min/max.
*/
-#ifdef ZFS_GLOBAL_NODE_PAGE_STATE
- uint64_t min = (ptob(global_node_page_state(NR_FILE_PAGES)) / 100) *
- zfs_arc_pc_percent;
-#else
- uint64_t min = (ptob(global_page_state(NR_FILE_PAGES)) / 100) *
- zfs_arc_pc_percent;
-#endif
+ uint64_t min = (ptob(nr_file_pages()) / 100) * zfs_arc_pc_percent;
min = MAX(arc_c_min, MIN(arc_c_max, min));
if (arc_dirty >= min)
diff --git a/config/kernel-global_page_state.m4 b/config/kernel-global_page_state.m4
new file mode 100644
index 000000000..f4a40011f
--- /dev/null
+++ b/config/kernel-global_page_state.m4
@@ -0,0 +1,109 @@
+dnl #
+dnl # 4.8 API change
+dnl #
+dnl # 75ef71840539 mm, vmstat: add infrastructure for per-node vmstats
+dnl # 599d0c954f91 mm, vmscan: move LRU lists to node
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_GLOBAL_NODE_PAGE_STATE], [
+ AC_MSG_CHECKING([whether global_node_page_state() exists])
+ ZFS_LINUX_TRY_COMPILE([
+ #include <linux/mm.h>
+ #include <linux/vmstat.h>
+ ],[
+ (void) global_node_page_state(0);
+ ],[
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(ZFS_GLOBAL_NODE_PAGE_STATE, 1, [global_node_page_state() exists])
+ ],[
+ AC_MSG_RESULT(no)
+ ])
+])
+
+dnl #
+dnl # 4.14 API change
+dnl #
+dnl # c41f012ade0b mm: rename global_page_state to global_zone_page_state
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_GLOBAL_ZONE_PAGE_STATE], [
+ AC_MSG_CHECKING([whether global_zone_page_state() exists])
+ ZFS_LINUX_TRY_COMPILE([
+ #include <linux/mm.h>
+ #include <linux/vmstat.h>
+ ],[
+ (void) global_zone_page_state(0);
+ ],[
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(ZFS_GLOBAL_ZONE_PAGE_STATE, 1, [global_zone_page_state() exists])
+ ],[
+ AC_MSG_RESULT(no)
+ ])
+])
+
+dnl #
+dnl # Create a define and autoconf variable for an enum member
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_ENUM_MEMBER], [
+ AC_MSG_CHECKING([whether enum $2 contains $1])
+ AS_IF([AC_TRY_COMMAND("${srcdir}/scripts/enum-extract.pl" "$2" "$3" | egrep -qx $1)],[
+ AC_MSG_RESULT([yes])
+ AC_DEFINE(m4_join([_], [ZFS_ENUM], m4_toupper($2), $1), 1, [enum $2 contains $1])
+ m4_join([_], [ZFS_ENUM], m4_toupper($2), $1)=1
+ ],[
+ AC_MSG_RESULT([no])
+ ])
+])
+
+dnl #
+dnl # Sanity check helpers
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_ERROR],[
+ AC_MSG_RESULT(no)
+ AC_MSG_RESULT([$1 in either node_stat_item or zone_stat_item: $2])
+ AC_MSG_RESULT([configure needs updating, see: config/kernel-global_page_state.m4])
+ AC_MSG_FAILURE([SHUT 'ER DOWN CLANCY, SHE'S PUMPIN' MUD!])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_CHECK], [
+ enum_check_a="m4_join([_], [$ZFS_ENUM_NODE_STAT_ITEM], $1)"
+ enum_check_b="m4_join([_], [$ZFS_ENUM_ZONE_STAT_ITEM], $1)"
+ AS_IF([test -n "$enum_check_a" -a -n "$enum_check_b"],[
+ ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_ERROR([$1], [DUPLICATE])
+ ])
+ AS_IF([test -z "$enum_check_a" -a -z "$enum_check_b"],[
+ ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_ERROR([$1], [NOT FOUND])
+ ])
+])
+
+dnl #
+dnl # Ensure the config tests are finding one and only one of each enum of interest
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_GLOBAL_ZONE_PAGE_STATE_SANITY], [
+ AC_MSG_CHECKING([global_page_state enums are sane])
+
+ ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_CHECK([NR_FILE_PAGES])
+ ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_CHECK([NR_INACTIVE_ANON])
+ ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_CHECK([NR_INACTIVE_FILE])
+ ZFS_AC_KERNEL_GLOBAL_PAGE_STATE_ENUM_CHECK([NR_SLAB_RECLAIMABLE])
+
+ AC_MSG_RESULT(yes)
+])
+
+dnl #
+dnl # enum members in which we're interested
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_GLOBAL_PAGE_STATE], [
+ ZFS_AC_KERNEL_GLOBAL_NODE_PAGE_STATE
+ ZFS_AC_KERNEL_GLOBAL_ZONE_PAGE_STATE
+
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_FILE_PAGES], [node_stat_item], [$LINUX/include/linux/mmzone.h])
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_INACTIVE_ANON], [node_stat_item], [$LINUX/include/linux/mmzone.h])
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_INACTIVE_FILE], [node_stat_item], [$LINUX/include/linux/mmzone.h])
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_SLAB_RECLAIMABLE], [node_stat_item], [$LINUX/include/linux/mmzone.h])
+
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_FILE_PAGES], [zone_stat_item], [$LINUX/include/linux/mmzone.h])
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_INACTIVE_ANON], [zone_stat_item], [$LINUX/include/linux/mmzone.h])
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_INACTIVE_FILE], [zone_stat_item], [$LINUX/include/linux/mmzone.h])
+ ZFS_AC_KERNEL_ENUM_MEMBER([NR_SLAB_RECLAIMABLE], [zone_stat_item], [$LINUX/include/linux/mmzone.h])
+
+ ZFS_AC_KERNEL_GLOBAL_ZONE_PAGE_STATE_SANITY
+])
diff --git a/config/kernel-vm_node_stat.m4 b/config/kernel-vm_node_stat.m4
deleted file mode 100644
index 5dcd9d827..000000000
--- a/config/kernel-vm_node_stat.m4
+++ /dev/null
@@ -1,22 +0,0 @@
-dnl #
-dnl # 4.8 API change
-dnl # kernel vm counters change
-dnl #
-AC_DEFUN([ZFS_AC_KERNEL_VM_NODE_STAT], [
- AC_MSG_CHECKING([whether to use vm_node_stat based fn's])
- ZFS_LINUX_TRY_COMPILE([
- #include <linux/mm.h>
- #include <linux/vmstat.h>
- ],[
- int a __attribute__ ((unused)) = NR_VM_NODE_STAT_ITEMS;
- long x __attribute__ ((unused)) =
- atomic_long_read(&vm_node_stat[0]);
- (void) global_node_page_state(0);
- ],[
- AC_MSG_RESULT(yes)
- AC_DEFINE(ZFS_GLOBAL_NODE_PAGE_STATE, 1,
- [using global_node_page_state()])
- ],[
- AC_MSG_RESULT(no)
- ])
-])
diff --git a/config/kernel.m4 b/config/kernel.m4
index 7bb86a96e..3e499e447 100644
--- a/config/kernel.m4
+++ b/config/kernel.m4
@@ -123,7 +123,7 @@ AC_DEFUN([ZFS_AC_CONFIG_KERNEL], [
ZFS_AC_KERNEL_RENAME_WANTS_FLAGS
ZFS_AC_KERNEL_HAVE_GENERIC_SETXATTR
ZFS_AC_KERNEL_CURRENT_TIME
- ZFS_AC_KERNEL_VM_NODE_STAT
+ ZFS_AC_KERNEL_GLOBAL_PAGE_STATE
ZFS_AC_KERNEL_ACL_HAS_REFCOUNT
AS_IF([test "$LINUX_OBJ" != "$LINUX"], [
diff --git a/scripts/enum-extract.pl b/scripts/enum-extract.pl
new file mode 100755
index 000000000..5112cc807
--- /dev/null
+++ b/scripts/enum-extract.pl
@@ -0,0 +1,58 @@
+#!/usr/bin/perl -w
+
+my $usage = <<EOT;
+usage: config-enum enum [file ...]
+
+Returns the elements from an enum declaration.
+
+"Best effort": we're not building an entire C interpreter here!
+EOT
+
+use warnings;
+use strict;
+use Getopt::Std;
+
+my %opts;
+
+if (!getopts("", \%opts) || @ARGV < 1) {
+ print $usage;
+ exit 2;
+}
+
+my $enum = shift;
+
+my $in_enum = 0;
+
+while (<>) {
+ # comments
+ s/\/\*.*\*\///;
+ if (m/\/\*/) {
+ while ($_ .= <>) {
+ last if s/\/\*.*\*\///s;
+ }
+ }
+
+ # preprocessor stuff
+ next if /^#/;
+
+ # find our enum
+ $in_enum = 1 if s/^\s*enum\s+${enum}(?:\s|$)//;
+ next unless $in_enum;
+
+ # remove explicit values
+ s/\s*=[^,]+,/,/g;
+
+ # extract each identifier
+ while (m/\b([a-z_][a-z0-9_]*)\b/ig) {
+ print $1, "\n";
+ }
+
+ #
+ # don't exit: there may be multiple versions of the same enum, e.g.
+ # inside different #ifdef blocks. Let's explicitly return all of
+ # them and let external tooling deal with it.
+ #
+ $in_enum = 0 if m/}\s*;/;
+}
+
+exit 0;
--
2.14.2

516
zfs-patches/0042-Add-scrub-after-resilver-zed-script.patch Normal file
View File

@ -0,0 +1,516 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Tony Hutter <hutter2@llnl.gov>
Date: Fri, 23 Feb 2018 11:38:05 -0800
Subject: [PATCH] Add scrub after resilver zed script
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
* Add a zed script to kick off a scrub after a resilver. The script is
disabled by default.
* Add a optional $PATH (-P) option to zed to allow it to use a custom
$PATH for its zedlets. This is needed when you're running zed under
the ZTS in a local workspace.
* Update test scripts to not copy in all-debug.sh and all-syslog.sh by
default. They can be optionally copied in as part of zed_setup().
These scripts slow down zed considerably under heavy events loads and
can cause events to be dropped or their delivery delayed. This was
causing some sporadic failures in the 'fault' tests.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Richard Laager <rlaager@wiktel.com>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes #4662
Closes #7086
(cherry picked from commit 99920d823e8e1510a0ad133c985bd2aa11a02834)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zed/Makefile.am | 6 +-
tests/zfs-tests/tests/functional/fault/Makefile.am | 3 +-
cmd/zed/zed_conf.h | 1 +
man/man8/zed.8.in | 10 +++-
cmd/zed/zed_conf.c | 7 ++-
cmd/zed/zed_event.c | 34 +++++++++--
cmd/zed/zed.d/resilver_finish-start-scrub.sh | 17 ++++++
cmd/zed/zed.d/zed.rc | 3 +
tests/runfiles/linux.run | 2 +-
tests/zfs-tests/include/commands.cfg | 1 +
tests/zfs-tests/include/libtest.shlib | 50 ++++++++++++++---
.../zfs-tests/tests/functional/events/cleanup.ksh | 2 +-
tests/zfs-tests/tests/functional/events/setup.ksh | 2 +-
tests/zfs-tests/tests/functional/fault/cleanup.ksh | 2 +-
.../functional/fault/scrub_after_resilver.ksh | 65 ++++++++++++++++++++++
tests/zfs-tests/tests/functional/fault/setup.ksh | 2 +-
16 files changed, 186 insertions(+), 21 deletions(-)
create mode 100755 cmd/zed/zed.d/resilver_finish-start-scrub.sh
create mode 100755 tests/zfs-tests/tests/functional/fault/scrub_after_resilver.ksh
diff --git a/cmd/zed/Makefile.am b/cmd/zed/Makefile.am
index 97733a512..ee44898cd 100644
--- a/cmd/zed/Makefile.am
+++ b/cmd/zed/Makefile.am
@@ -69,7 +69,8 @@ dist_zedexec_SCRIPTS = \
zed.d/statechange-notify.sh \
zed.d/vdev_clear-led.sh \
zed.d/vdev_attach-led.sh \
- zed.d/pool_import-led.sh
+ zed.d/pool_import-led.sh \
+ zed.d/resilver_finish-start-scrub.sh
zedconfdefaults = \
all-syslog.sh \
@@ -80,7 +81,8 @@ zedconfdefaults = \
statechange-notify.sh \
vdev_clear-led.sh \
vdev_attach-led.sh \
- pool_import-led.sh
+ pool_import-led.sh \
+ resilver_finish-start-scrub.sh
install-data-hook:
$(MKDIR_P) "$(DESTDIR)$(zedconfdir)"
diff --git a/tests/zfs-tests/tests/functional/fault/Makefile.am b/tests/zfs-tests/tests/functional/fault/Makefile.am
index eeff31261..abe28501d 100644
--- a/tests/zfs-tests/tests/functional/fault/Makefile.am
+++ b/tests/zfs-tests/tests/functional/fault/Makefile.am
@@ -4,4 +4,5 @@ dist_pkgdata_SCRIPTS = \
setup.ksh \
cleanup.ksh \
auto_online_001_pos.ksh \
- auto_replace_001_pos.ksh
+ auto_replace_001_pos.ksh \
+ scrub_after_resilver.ksh
diff --git a/cmd/zed/zed_conf.h b/cmd/zed/zed_conf.h
index 2bc634134..7d6b63b1d 100644
--- a/cmd/zed/zed_conf.h
+++ b/cmd/zed/zed_conf.h
@@ -37,6 +37,7 @@ struct zed_conf {
int state_fd; /* fd to state file */
libzfs_handle_t *zfs_hdl; /* handle to libzfs */
int zevent_fd; /* fd for access to zevents */
+ char *path; /* custom $PATH for zedlets to use */
};
struct zed_conf *zed_conf_create(void);
diff --git a/man/man8/zed.8.in b/man/man8/zed.8.in
index 2ab088d98..645e91795 100644
--- a/man/man8/zed.8.in
+++ b/man/man8/zed.8.in
@@ -27,6 +27,7 @@ ZED \- ZFS Event Daemon
[\fB\-L\fR]
[\fB\-M\fR]
[\fB\-p\fR \fIpidfile\fR]
+[\fB\-P\fR \fIpath\fR]
[\fB\-s\fR \fIstatefile\fR]
[\fB\-v\fR]
[\fB\-V\fR]
@@ -78,9 +79,16 @@ Read the enabled ZEDLETs from the specified directory.
.BI \-p\ pidfile
Write the daemon's process ID to the specified file.
.TP
+.BI \-P\ path
+Custom $PATH for zedlets to use. Normally zedlets run in a locked-down
+environment, with hardcoded paths to the ZFS commands ($ZFS, $ZPOOL, $ZED, ...),
+and a hardcoded $PATH. This is done for security reasons. However, the
+ZFS test suite uses a custom PATH for its ZFS commands, and passes it to zed
+with -P. In short, -P is only to be used by the ZFS test suite; never use
+it in production!
+.TP
.BI \-s\ statefile
Write the daemon's state to the specified file.
-
.SH ZEVENTS
.PP
A zevent is comprised of a list of nvpairs (name/value pairs). Each zevent
diff --git a/cmd/zed/zed_conf.c b/cmd/zed/zed_conf.c
index 5b27f1e4f..86671369c 100644
--- a/cmd/zed/zed_conf.c
+++ b/cmd/zed/zed_conf.c
@@ -155,6 +155,8 @@ _zed_conf_display_help(const char *prog, int got_err)
"Run daemon in the foreground.");
fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-M",
"Lock all pages in memory.");
+ fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-P",
+ "$PATH for ZED to use (only used by ZTS).");
fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-Z",
"Zero state file.");
fprintf(fp, "\n");
@@ -247,7 +249,7 @@ _zed_conf_parse_path(char **resultp, const char *path)
void
zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv)
{
- const char * const opts = ":hLVc:d:p:s:vfFMZ";
+ const char * const opts = ":hLVc:d:p:P:s:vfFMZ";
int opt;
if (!zcp || !argv || !argv[0])
@@ -275,6 +277,9 @@ zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv)
case 'p':
_zed_conf_parse_path(&zcp->pid_file, optarg);
break;
+ case 'P':
+ _zed_conf_parse_path(&zcp->path, optarg);
+ break;
case 's':
_zed_conf_parse_path(&zcp->state_file, optarg);
break;
diff --git a/cmd/zed/zed_event.c b/cmd/zed/zed_event.c
index 390235019..2a7ff16fd 100644
--- a/cmd/zed/zed_event.c
+++ b/cmd/zed/zed_event.c
@@ -733,12 +733,14 @@ _zed_event_add_nvpair(uint64_t eid, zed_strings_t *zsp, nvpair_t *nvp)
/*
* Restrict various environment variables to safe and sane values
- * when constructing the environment for the child process.
+ * when constructing the environment for the child process, unless
+ * we're running with a custom $PATH (like under the ZFS test suite).
*
* Reference: Secure Programming Cookbook by Viega & Messier, Section 1.1.
*/
static void
-_zed_event_add_env_restrict(uint64_t eid, zed_strings_t *zsp)
+_zed_event_add_env_restrict(uint64_t eid, zed_strings_t *zsp,
+ const char *path)
{
const char *env_restrict[][2] = {
{ "IFS", " \t\n" },
@@ -753,11 +755,35 @@ _zed_event_add_env_restrict(uint64_t eid, zed_strings_t *zsp)
{ "ZFS_RELEASE", ZFS_META_RELEASE },
{ NULL, NULL }
};
+
+ /*
+ * If we have a custom $PATH, use the default ZFS binary locations
+ * instead of the hard-coded ones.
+ */
+ const char *env_path[][2] = {
+ { "IFS", " \t\n" },
+ { "PATH", NULL }, /* $PATH copied in later on */
+ { "ZDB", "zdb" },
+ { "ZED", "zed" },
+ { "ZFS", "zfs" },
+ { "ZINJECT", "zinject" },
+ { "ZPOOL", "zpool" },
+ { "ZFS_ALIAS", ZFS_META_ALIAS },
+ { "ZFS_VERSION", ZFS_META_VERSION },
+ { "ZFS_RELEASE", ZFS_META_RELEASE },
+ { NULL, NULL }
+ };
const char *(*pa)[2];
assert(zsp != NULL);
- for (pa = env_restrict; *(*pa); pa++) {
+ pa = path != NULL ? env_path : env_restrict;
+
+ for (; *(*pa); pa++) {
+ /* Use our custom $PATH if we have one */
+ if (path != NULL && strcmp((*pa)[0], "PATH") == 0)
+ (*pa)[1] = path;
+
_zed_event_add_var(eid, zsp, NULL, (*pa)[0], "%s", (*pa)[1]);
}
}
@@ -902,7 +928,7 @@ zed_event_service(struct zed_conf *zcp)
while ((nvp = nvlist_next_nvpair(nvl, nvp)))
_zed_event_add_nvpair(eid, zsp, nvp);
- _zed_event_add_env_restrict(eid, zsp);
+ _zed_event_add_env_restrict(eid, zsp, zcp->path);
_zed_event_add_env_preserve(eid, zsp);
_zed_event_add_var(eid, zsp, ZED_VAR_PREFIX, "PID",
diff --git a/cmd/zed/zed.d/resilver_finish-start-scrub.sh b/cmd/zed/zed.d/resilver_finish-start-scrub.sh
new file mode 100755
index 000000000..6f9c0b309
--- /dev/null
+++ b/cmd/zed/zed.d/resilver_finish-start-scrub.sh
@@ -0,0 +1,17 @@
+#!/bin/sh
+# resilver_finish-start-scrub.sh
+# Run a scrub after a resilver
+#
+# Exit codes:
+# 1: Internal error
+# 2: Script wasn't enabled in zed.rc
+[ -f "${ZED_ZEDLET_DIR}/zed.rc" ] && . "${ZED_ZEDLET_DIR}/zed.rc"
+. "${ZED_ZEDLET_DIR}/zed-functions.sh"
+
+[ "${ZED_SCRUB_AFTER_RESILVER}" = "1" ] || exit 2
+[ -n "${ZEVENT_POOL}" ] || exit 1
+[ -n "${ZEVENT_SUBCLASS}" ] || exit 1
+zed_check_cmd "${ZPOOL}" || exit 1
+
+zed_log_msg "Starting scrub after resilver on ${ZEVENT_POOL}"
+"${ZPOOL}" scrub "${ZEVENT_POOL}"
diff --git a/cmd/zed/zed.d/zed.rc b/cmd/zed/zed.d/zed.rc
index a1dd33704..8b0e476d5 100644
--- a/cmd/zed/zed.d/zed.rc
+++ b/cmd/zed/zed.d/zed.rc
@@ -86,6 +86,9 @@
#
ZED_USE_ENCLOSURE_LEDS=1
+##
+# Run a scrub after every resilver
+#ZED_SCRUB_AFTER_RESILVER=1
##
# The syslog priority (e.g., specified as a "facility.level" pair).
diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run
index 8be3e1c62..89c923db1 100644
--- a/tests/runfiles/linux.run
+++ b/tests/runfiles/linux.run
@@ -421,7 +421,7 @@ tests = ['exec_001_pos', 'exec_002_neg']
tags = ['functional', 'exec']
[tests/functional/fault]
-tests = ['auto_online_001_pos', 'auto_replace_001_pos']
+tests = ['auto_online_001_pos', 'auto_replace_001_pos', 'scrub_after_resilver']
tags = ['functional', 'fault']
[tests/functional/features/async_destroy]
diff --git a/tests/zfs-tests/include/commands.cfg b/tests/zfs-tests/include/commands.cfg
index f6fd239de..936e54c1a 100644
--- a/tests/zfs-tests/include/commands.cfg
+++ b/tests/zfs-tests/include/commands.cfg
@@ -83,6 +83,7 @@ export SYSTEM_FILES='arp
pgrep
ping
pkill
+ printenv
printf
ps
pwd
diff --git a/tests/zfs-tests/include/libtest.shlib b/tests/zfs-tests/include/libtest.shlib
index 86f172a6d..48fb5e7c5 100644
--- a/tests/zfs-tests/include/libtest.shlib
+++ b/tests/zfs-tests/include/libtest.shlib
@@ -3339,9 +3339,32 @@ function wait_replacing #pool
done
}
+#
+# Wait for a pool to be scrubbed
+#
+# $1 pool name
+# $2 number of seconds to wait (optional)
+#
+# Returns true when pool has been scrubbed, or false if there's a timeout or if
+# no scrub was done.
+#
+function wait_scrubbed
+{
+ typeset pool=${1:-$TESTPOOL}
+ typeset iter=${2:-10}
+ for i in {1..$iter} ; do
+ if is_pool_scrubbed $pool ; then
+ return 0
+ fi
+ sleep 1
+ done
+ return 1
+}
+
#
# Setup custom environment for the ZED.
#
+# $@ Optional list of zedlets to run under zed.
function zed_setup
{
if ! is_linux; then
@@ -3359,6 +3382,7 @@ function zed_setup
if [[ -e $VDEVID_CONF_ETC ]]; then
log_fail "Must not have $VDEVID_CONF_ETC file present on system"
fi
+ EXTRA_ZEDLETS=$@
# Create a symlink for /etc/zfs/vdev_id.conf file.
log_must ln -s $VDEVID_CONF $VDEVID_CONF_ETC
@@ -3368,32 +3392,44 @@ function zed_setup
log_must cp ${ZEDLET_ETC_DIR}/zed.rc $ZEDLET_DIR
log_must cp ${ZEDLET_ETC_DIR}/zed-functions.sh $ZEDLET_DIR
+ # Scripts must only be user writable.
+ if [[ -n "$EXTRA_ZEDLETS" ]] ; then
+ saved_umask=$(umask)
+ log_must umask 0022
+ for i in $EXTRA_ZEDLETS ; do
+ log_must cp ${ZEDLET_LIBEXEC_DIR}/$i $ZEDLET_DIR
+ done
+ log_must umask $saved_umask
+ fi
+
# Customize the zed.rc file to enable the full debug log.
log_must sed -i '/\#ZED_DEBUG_LOG=.*/d' $ZEDLET_DIR/zed.rc
echo "ZED_DEBUG_LOG=$ZED_DEBUG_LOG" >>$ZEDLET_DIR/zed.rc
- # Scripts must only be user writable.
- saved_umask=$(umask)
- log_must umask 0022
- log_must cp ${ZEDLET_LIBEXEC_DIR}/all-syslog.sh $ZEDLET_DIR
- log_must cp ${ZEDLET_LIBEXEC_DIR}/all-debug.sh $ZEDLET_DIR
- log_must umask $saved_umask
}
#
# Cleanup custom ZED environment.
#
+# $@ Optional list of zedlets to remove from our test zed.d directory.
function zed_cleanup
{
if ! is_linux; then
return
fi
+ EXTRA_ZEDLETS=$@
log_must rm -f ${ZEDLET_DIR}/zed.rc
log_must rm -f ${ZEDLET_DIR}/zed-functions.sh
log_must rm -f ${ZEDLET_DIR}/all-syslog.sh
log_must rm -f ${ZEDLET_DIR}/all-debug.sh
log_must rm -f ${ZEDLET_DIR}/state
+
+ if [[ -n "$EXTRA_ZEDLETS" ]] ; then
+ for i in $EXTRA_ZEDLETS ; do
+ log_must rm -f ${ZEDLET_DIR}/$i
+ done
+ fi
log_must rm -f $ZED_LOG
log_must rm -f $ZED_DEBUG_LOG
log_must rm -f $VDEVID_CONF_ETC
@@ -3425,7 +3461,7 @@ function zed_start
# run ZED in the background and redirect foreground logging
# output to $ZED_LOG.
log_must truncate -s 0 $ZED_DEBUG_LOG
- log_must eval "zed -vF -d $ZEDLET_DIR -p $ZEDLET_DIR/zed.pid" \
+ log_must eval "zed -vF -d $ZEDLET_DIR -p $ZEDLET_DIR/zed.pid -P $PATH" \
"-s $ZEDLET_DIR/state 2>$ZED_LOG &"
return 0
diff --git a/tests/zfs-tests/tests/functional/events/cleanup.ksh b/tests/zfs-tests/tests/functional/events/cleanup.ksh
index bc536e260..4905342b7 100755
--- a/tests/zfs-tests/tests/functional/events/cleanup.ksh
+++ b/tests/zfs-tests/tests/functional/events/cleanup.ksh
@@ -26,6 +26,6 @@
. $STF_SUITE/include/libtest.shlib
-zed_cleanup
+zed_cleanup all-debug.sh all-syslog.sh
default_cleanup
diff --git a/tests/zfs-tests/tests/functional/events/setup.ksh b/tests/zfs-tests/tests/functional/events/setup.ksh
index 7113c1f39..2f81d16b1 100755
--- a/tests/zfs-tests/tests/functional/events/setup.ksh
+++ b/tests/zfs-tests/tests/functional/events/setup.ksh
@@ -28,6 +28,6 @@
DISK=${DISKS%% *}
-zed_setup
+zed_setup all-debug.sh all-syslog.sh
default_setup $DISK
diff --git a/tests/zfs-tests/tests/functional/fault/cleanup.ksh b/tests/zfs-tests/tests/functional/fault/cleanup.ksh
index f39f05d6f..d3de742b3 100755
--- a/tests/zfs-tests/tests/functional/fault/cleanup.ksh
+++ b/tests/zfs-tests/tests/functional/fault/cleanup.ksh
@@ -31,7 +31,7 @@ verify_runnable "global"
cleanup_devices $DISKS
zed_stop
-zed_cleanup
+zed_cleanup resilver_finish-start-scrub.sh
SD=$(lsscsi | nawk '/scsi_debug/ {print $6; exit}')
SDDEVICE=$(echo $SD | nawk -F / '{print $3}')
diff --git a/tests/zfs-tests/tests/functional/fault/scrub_after_resilver.ksh b/tests/zfs-tests/tests/functional/fault/scrub_after_resilver.ksh
new file mode 100755
index 000000000..558cb065f
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/fault/scrub_after_resilver.ksh
@@ -0,0 +1,65 @@
+#!/bin/ksh -p
+#
+# 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.
+#
+
+#
+# Copyright (c) 2018 by Lawrence Livermore National Security, LLC.
+# All rights reserved.
+#
+
+. $STF_SUITE/include/libtest.shlib
+. $STF_SUITE/tests/functional/fault/fault.cfg
+
+#
+# DESCRIPTION:
+# Test the scrub after resilver zedlet
+#
+# STRATEGY:
+# 1. Create a mirrored pool
+# 2. Fault a disk
+# 3. Replace the disk, starting a resilver
+# 4. Verify that a scrub happens after the resilver finishes
+#
+
+log_assert "Testing the scrub after resilver zedlet"
+
+# Backup our zed.rc
+zedrc_backup="$(mktemp)"
+log_must cp $ZEDLET_DIR/zed.rc $zedrc_backup
+
+# Enable ZED_SCRUB_AFTER_RESILVER
+eval "sed -i 's/\#ZED_SCRUB_AFTER_RESILVER/ZED_SCRUB_AFTER_RESILVER/g' $ZEDLET_DIR/zed.rc"
+
+function cleanup
+{
+ # Restore our zed.rc
+ log_must mv $zedrc_backup $ZEDLET_DIR/zed.rc
+ default_cleanup_noexit
+}
+
+log_onexit cleanup
+
+verify_disk_count "$DISKS" 3
+default_mirror_setup_noexit $DISK1 $DISK2
+
+log_must zpool offline -f $TESTPOOL $DISK1
+
+# Write to our degraded pool so we have some data to resilver
+log_must mkfile 16M $TESTDIR/file1
+
+# Replace the failed disks, forcing a resilver
+log_must zpool replace $TESTPOOL $DISK1 $DISK3
+
+# Wait for the resilver to finish, and then the subsequent scrub to finish.
+# Waiting for the scrub has the effect of waiting for both. Timeout after 10
+# seconds if nothing is happening.
+log_must wait_scrubbed $TESTPOOL 10
+log_pass "Successfully ran the scrub after resilver zedlet"
diff --git a/tests/zfs-tests/tests/functional/fault/setup.ksh b/tests/zfs-tests/tests/functional/fault/setup.ksh
index 3d54d4f21..484bc4587 100755
--- a/tests/zfs-tests/tests/functional/fault/setup.ksh
+++ b/tests/zfs-tests/tests/functional/fault/setup.ksh
@@ -28,7 +28,7 @@
verify_runnable "global"
-zed_setup
+zed_setup resilver_finish-start-scrub.sh
zed_start
# Create a scsi_debug device to be used with auto-online (if using loop devices)
--
2.14.2

322
zfs-patches/0043-Add-SMART-self-test-results-to-zpool-status-c.patch Normal file
View File

@ -0,0 +1,322 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Tony Hutter <hutter2@llnl.gov>
Date: Tue, 27 Feb 2018 09:31:27 -0800
Subject: [PATCH] Add SMART self-test results to zpool status -c
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Add in SMART self-test results to zpool status|iostat -c. This
works for both SAS and SATA drives.
Also, add plumbing to allow the 'smart' script to take smartctl
output from a directory of output text files instead of running
it against the vdevs.
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes #7178
(cherry picked from commit 5e3085e360161456fe2af697494c479de0ee2085)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/zpool/Makefile.am | 14 ++++-
cmd/zpool/zpool.d/smart | 132 +++++++++++++++++++++++++++++++++++-----
cmd/zpool/zpool.d/smart_test | 1 +
cmd/zpool/zpool.d/test_ended | 1 +
cmd/zpool/zpool.d/test_progress | 1 +
cmd/zpool/zpool.d/test_status | 1 +
cmd/zpool/zpool.d/test_type | 1 +
7 files changed, 133 insertions(+), 18 deletions(-)
create mode 120000 cmd/zpool/zpool.d/smart_test
create mode 120000 cmd/zpool/zpool.d/test_ended
create mode 120000 cmd/zpool/zpool.d/test_progress
create mode 120000 cmd/zpool/zpool.d/test_status
create mode 120000 cmd/zpool/zpool.d/test_type
diff --git a/cmd/zpool/Makefile.am b/cmd/zpool/Makefile.am
index c7b8b76e3..d07f8d616 100644
--- a/cmd/zpool/Makefile.am
+++ b/cmd/zpool/Makefile.am
@@ -63,7 +63,12 @@ dist_zpoolexec_SCRIPTS = \
zpool.d/nvme_err \
zpool.d/pwr_cyc \
zpool.d/upath \
- zpool.d/vendor
+ zpool.d/vendor \
+ zpool.d/smart_test \
+ zpool.d/test_type \
+ zpool.d/test_status \
+ zpool.d/test_progress \
+ zpool.d/test_ended
zpoolconfdefaults = \
enc \
@@ -102,7 +107,12 @@ zpoolconfdefaults = \
nvme_err \
pwr_cyc \
upath \
- vendor
+ vendor \
+ smart_test \
+ test_type \
+ test_status \
+ test_progress \
+ test_ended
install-data-hook:
$(MKDIR_P) "$(DESTDIR)$(zpoolconfdir)"
diff --git a/cmd/zpool/zpool.d/smart b/cmd/zpool/zpool.d/smart
index 4bc3af39d..64b5f6e4e 100755
--- a/cmd/zpool/zpool.d/smart
+++ b/cmd/zpool/zpool.d/smart
@@ -24,8 +24,44 @@ ata_err: Show SMART ATA errors (ATA).
pwr_cyc: Show SMART power cycle count (ATA).
serial: Show disk serial number.
nvme_err: Show SMART NVMe errors (NVMe).
+smart_test: Show SMART self-test results summary.
+test_type: Show SMART self-test type (short, long... ).
+test_status: Show SMART self-test status.
+test_progress: Show SMART self-test percentage done.
+test_ended: Show when the last SMART self-test ended (if supported).
"
+# Hack for developer testing
+#
+# If you set $samples to a directory containing smartctl output text files,
+# we will use them instead of running smartctl on the vdevs. This can be
+# useful if you want to test a bunch of different smartctl outputs. Also, if
+# $samples is set, and additional 'file' column is added to the zpool output
+# showing the filename.
+samples=
+
+# get_filename_from_dir DIR
+#
+# Look in directory DIR and return a filename from it. The filename returned
+# is chosen quasi-sequentially (based off our PID). This allows us to return
+# a different filename every time this script is invoked (which we do for each
+# vdev), without having to maintain state.
+get_filename_from_dir()
+{
+ dir=$1
+ pid="$$"
+ num_files=$(find "$dir" -maxdepth 1 -type f | wc -l)
+ mod=$((pid % num_files))
+ i=0
+ find "$dir" -type f -printf "%f\n" | while read -r file ; do
+ if [ "$mod" = "$i" ] ; then
+ echo "$file"
+ break
+ fi
+ i=$((i+1))
+ done
+}
+
script=$(basename "$0")
if [ "$1" = "-h" ] ; then
@@ -35,8 +71,16 @@ fi
smartctl_path=$(which smartctl)
-if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
- raw_out=$(eval "sudo $smartctl_path -a $VDEV_UPATH")
+if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ] || [ -n "$samples" ] ; then
+ if [ -n "$samples" ] ; then
+ # cat a smartctl output text file instead of running smartctl
+ # on a vdev (only used for developer testing).
+ file=$(get_filename_from_dir $samples)
+ echo "file=$file"
+ raw_out=$(cat "$samples/$file")
+ else
+ raw_out=$(eval "sudo $smartctl_path -a $VDEV_UPATH")
+ fi
# What kind of drive are we? Look for the right line in smartctl:
#
@@ -49,7 +93,6 @@ if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
# NVMe:
# SMART/Health Information (NVMe Log 0xnn, NSID 0xnn)
#
- type=$(echo "$raw_out" | grep -m 1 -Eo '^ATA|NVMe|SAS$')
out=$(echo "$raw_out" | awk '
# SAS specific
/read:/{print "rrd="$4"\nr_cor="$5"\nr_proc="$7"\nr_ucor="$8}
@@ -58,10 +101,11 @@ if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
/Elements in grown defect list/{print "defect="$6}
# SAS common
+/SAS/{type="sas"}
/Drive Temperature:/{print "temp="$4}
# Status can be a long string, substitute spaces for '_'
/SMART Health Status:/{printf "health="; for(i=4;i<=NF-1;i++){printf "%s_", $i}; printf "%s\n", $i}
-/number of hours powered up/{print "hours_on="$7}
+/number of hours powered up/{print "hours_on="$7; hours_on=int($7)}
/Serial number:/{print "serial="$3}
# SATA specific
@@ -74,13 +118,16 @@ if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
/Power_Cycle_Count/{print "pwr_cyc="$10}
# SATA common
+/SATA/{type="sata"}
/Temperature_Celsius/{print "temp="$10}
/Airflow_Temperature_Cel/{print "temp="$10}
+/Current Temperature:/{print "temp="$3}
/SMART overall-health self-assessment test result:/{print "health="$6}
-/Power_On_Hours/{print "hours_on="$10}
+/Power_On_Hours/{print "hours_on="$10; hours_on=int($10)}
/Serial Number:/{print "serial="$3}
# NVMe common
+/NVMe/{type="nvme"}
/Temperature:/{print "temp="$2}
/SMART overall-health self-assessment test result:/{print "health="$6}
/Power On Hours:/{gsub("[^0-9]","",$4); print "hours_on="$4}
@@ -90,39 +137,92 @@ if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ]; then
# NVMe specific
/Media and Data Integrity Errors:/{print "nvme_err="$6}
-END {ORS="\n"; print ""}
+# SMART self-test info
+/Self-test execution status:/{progress=tolower($4)} # SAS
+/SMART Self-test log/{test_seen=1} # SAS
+/SMART Extended Self-test Log/{test_seen=1} # SATA
+/# 1/{
+ test_type=tolower($3"_"$4);
+ # Status could be one word ("Completed") or multiple ("Completed: read
+ # failure"). Look for the ":" to see if we need to grab more words.
+
+ if ($5 ~ ":")
+ status=tolower($5""$6"_"$7)
+ else
+ status=tolower($5)
+ if (status=="self")
+ status="running";
+
+ if (type == "sas") {
+ hours=int($(NF-4))
+ } else {
+ hours=int($(NF-1))
+ # SATA reports percent remaining, rather than percent done
+ # Convert it to percent done.
+ progress=(100-int($(NF-2)))"%"
+ }
+ # When we int()-ify "hours", it converts stuff like "NOW" and "-" into
+ # 0. In those cases, set it to hours_on, so they will cancel out in
+ # the "hours_ago" calculation later on.
+ if (hours == 0)
+ hours=hours_on
+
+ if (test_seen) {
+ print "test="hours_on
+ print "test_type="test_type
+ print "test_status="status
+ print "test_progress="progress
+ }
+ # Not all drives report hours_on
+ if (hours_on && hours) {
+ total_hours_ago=(hours_on-hours)
+ days_ago=int(total_hours_ago/24)
+ hours_ago=(total_hours_ago % 24)
+ if (days_ago != 0)
+ ago_str=days_ago"d"
+ if (hours_ago !=0)
+ ago_str=ago_str""hours_ago"h"
+ print "test_ended="ago_str
+ }
+}
+
+END {print "type="type; ORS="\n"; print ""}
');
fi
+type=$(echo "$out" | grep '^type=' | cut -d '=' -f 2)
-# if type is not set by now, either we don't have a block device
-# or smartctl failed. Either way, default to ATA and set out to
-# nothing
+# If type is not set by now, either we don't have a block device
+# or smartctl failed. Either way, default to ATA and set $out to
+# nothing.
if [ -z "$type" ]; then
- type="ATA"
+ type="sata"
out=
fi
case $script in
smart)
# Print temperature plus common predictors of drive failure
- if [ "$type" = "SAS" ] ; then
+ if [ "$type" = "sas" ] ; then
scripts="temp|health|r_ucor|w_ucor"
- elif [ "$type" = "ATA" ] ; then
+ elif [ "$type" = "sata" ] ; then
scripts="temp|health|ata_err|realloc|rep_ucor|cmd_to|pend_sec|off_ucor"
- elif [ "$type" = "NVMe" ] ; then
+ elif [ "$type" = "nvme" ] ; then
scripts="temp|health|nvme_err"
fi
;;
smartx)
# Print some other interesting stats
- if [ "$type" = "SAS" ] ; then
+ if [ "$type" = "sas" ] ; then
scripts="hours_on|defect|nonmed|r_proc|w_proc"
- elif [ "$type" = "ATA" ] ; then
+ elif [ "$type" = "sata" ] ; then
scripts="hours_on|pwr_cyc"
- elif [ "$type" = "NVMe" ] ; then
+ elif [ "$type" = "nvme" ] ; then
scripts="hours_on|pwr_cyc"
fi
;;
+smart_test)
+ scripts="test_type|test_status|test_progress|test_ended"
+ ;;
*)
scripts="$script"
esac
diff --git a/cmd/zpool/zpool.d/smart_test b/cmd/zpool/zpool.d/smart_test
new file mode 120000
index 000000000..94f22861f
--- /dev/null
+++ b/cmd/zpool/zpool.d/smart_test
@@ -0,0 +1 @@
+smart
\ No newline at end of file
diff --git a/cmd/zpool/zpool.d/test_ended b/cmd/zpool/zpool.d/test_ended
new file mode 120000
index 000000000..94f22861f
--- /dev/null
+++ b/cmd/zpool/zpool.d/test_ended
@@ -0,0 +1 @@
+smart
\ No newline at end of file
diff --git a/cmd/zpool/zpool.d/test_progress b/cmd/zpool/zpool.d/test_progress
new file mode 120000
index 000000000..94f22861f
--- /dev/null
+++ b/cmd/zpool/zpool.d/test_progress
@@ -0,0 +1 @@
+smart
\ No newline at end of file
diff --git a/cmd/zpool/zpool.d/test_status b/cmd/zpool/zpool.d/test_status
new file mode 120000
index 000000000..94f22861f
--- /dev/null
+++ b/cmd/zpool/zpool.d/test_status
@@ -0,0 +1 @@
+smart
\ No newline at end of file
diff --git a/cmd/zpool/zpool.d/test_type b/cmd/zpool/zpool.d/test_type
new file mode 120000
index 000000000..94f22861f
--- /dev/null
+++ b/cmd/zpool/zpool.d/test_type
@@ -0,0 +1 @@
+smart
\ No newline at end of file
--
2.14.2

66
zfs-patches/0044-Fix-zpool-8-list-example-to-match-actual-format.patch Normal file
View File

@ -0,0 +1,66 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Tomohiro Kusumi <kusumi.tomohiro@gmail.com>
Date: Thu, 1 Mar 2018 01:54:53 +0900
Subject: [PATCH] Fix zpool(8) list example to match actual format
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
a05dfd00 (Illumos 5147) has swapped FRAG and EXPANDSZ,
so it's natural to modify these examples.
# zpool list | head -1
NAME SIZE ALLOC FREE EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT
^^^^^^^^^^^^^^^
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@osnexus.com>
Closes #7244
(cherry picked from commit bcaba38e428be8dc90564d15c7e13f154c535edd)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
man/man8/zpool.8 | 20 ++++++++++----------
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/man/man8/zpool.8 b/man/man8/zpool.8
index 22579101a..7c99f5868 100644
--- a/man/man8/zpool.8
+++ b/man/man8/zpool.8
@@ -2087,10 +2087,10 @@ is faulted due to a missing device.
The results from this command are similar to the following:
.Bd -literal
# zpool list
-NAME SIZE ALLOC FREE FRAG EXPANDSZ CAP DEDUP HEALTH ALTROOT
-rpool 19.9G 8.43G 11.4G 33% - 42% 1.00x ONLINE -
-tank 61.5G 20.0G 41.5G 48% - 32% 1.00x ONLINE -
-zion - - - - - - - FAULTED -
+NAME SIZE ALLOC FREE EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT
+rpool 19.9G 8.43G 11.4G - 33% 42% 1.00x ONLINE -
+tank 61.5G 20.0G 41.5G - 48% 32% 1.00x ONLINE -
+zion - - - - - - - FAULTED -
.Ed
.It Sy Example 7 No Destroying a ZFS Storage Pool
The following command destroys the pool
@@ -2215,12 +2215,12 @@ In this example, the pool will not be able to utilize this extra capacity until
all the devices under the raidz vdev have been expanded.
.Bd -literal
# zpool list -v data
-NAME SIZE ALLOC FREE FRAG EXPANDSZ CAP DEDUP HEALTH ALTROOT
-data 23.9G 14.6G 9.30G 48% - 61% 1.00x ONLINE -
- raidz1 23.9G 14.6G 9.30G 48% -
- sda - - - - -
- sdb - - - - 10G
- sdc - - - - -
+NAME SIZE ALLOC FREE EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT
+data 23.9G 14.6G 9.30G - 48% 61% 1.00x ONLINE -
+ raidz1 23.9G 14.6G 9.30G - 48%
+ sda - - - - -
+ sdb - - - 10G -
+ sdc - - - - -
.Ed
.It Sy Example 16 No Adding output columns
Additional columns can be added to the
--
2.14.2

275
zfs-patches/0045-Fix-some-typos.patch Normal file
View File

@ -0,0 +1,275 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: John Eismeier <32205350+jeis2497052@users.noreply.github.com>
Date: Wed, 28 Feb 2018 11:57:10 -0500
Subject: [PATCH] Fix some typos
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed by: George Melikov <mail@gmelikov.ru>
Signed-off-by: John Eismeier <john.eismeier@gmail.com>
Closes #7237
(cherry picked from commit 33bb1e82568a9734cc3f41d3e1d33003ebf0e123)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
cmd/arcstat/arcstat.py | 4 ++--
cmd/dbufstat/dbufstat.py | 4 ++--
contrib/initramfs/scripts/zfs | 2 +-
man/man8/zpool.8 | 2 +-
tests/zfs-tests/tests/functional/migration/migration_001_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_002_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_003_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_004_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_005_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_006_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_007_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_008_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_009_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_010_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_011_pos.ksh | 4 ++--
tests/zfs-tests/tests/functional/migration/migration_012_pos.ksh | 4 ++--
16 files changed, 30 insertions(+), 30 deletions(-)
diff --git a/cmd/arcstat/arcstat.py b/cmd/arcstat/arcstat.py
index aa54ee87a..85c83ccc4 100755
--- a/cmd/arcstat/arcstat.py
+++ b/cmd/arcstat/arcstat.py
@@ -280,7 +280,7 @@ def init():
"outfile",
"help",
"verbose",
- "seperator",
+ "separator",
"columns"
]
)
@@ -299,7 +299,7 @@ def init():
hflag = True
if opt in ('-v', '--verbose'):
vflag = True
- if opt in ('-s', '--seperator'):
+ if opt in ('-s', '--separator'):
sep = arg
i += 1
if opt in ('-f', '--columns'):
diff --git a/cmd/dbufstat/dbufstat.py b/cmd/dbufstat/dbufstat.py
index dda0a143f..42bb0c7c7 100755
--- a/cmd/dbufstat/dbufstat.py
+++ b/cmd/dbufstat/dbufstat.py
@@ -474,7 +474,7 @@ def main():
"help",
"infile",
"outfile",
- "seperator",
+ "separator",
"types",
"verbose",
"extended"
@@ -499,7 +499,7 @@ def main():
ofile = arg
if opt in ('-r', '--raw'):
raw += 1
- if opt in ('-s', '--seperator'):
+ if opt in ('-s', '--separator'):
sep = arg
if opt in ('-t', '--types'):
tflag = True
diff --git a/contrib/initramfs/scripts/zfs b/contrib/initramfs/scripts/zfs
index 8770a2e8e..86329e764 100644
--- a/contrib/initramfs/scripts/zfs
+++ b/contrib/initramfs/scripts/zfs
@@ -478,7 +478,7 @@ destroy_fs()
echo "Message: $ZFS_STDERR"
echo "Error: $ZFS_ERROR"
echo ""
- echo "Failed to destroy '$fs'. Please make sure that '$fs' is not availible."
+ echo "Failed to destroy '$fs'. Please make sure that '$fs' is not available."
echo "Hint: Try: zfs destroy -Rfn $fs"
echo "If this dryrun looks good, then remove the 'n' from '-Rfn' and try again."
/bin/sh
diff --git a/man/man8/zpool.8 b/man/man8/zpool.8
index 7c99f5868..6d7c2271c 100644
--- a/man/man8/zpool.8
+++ b/man/man8/zpool.8
@@ -733,7 +733,7 @@ man page. In order to enable this property each host must set a unique hostid.
See
.Xr genhostid 1
.Xr zgenhostid 8
-.Xr spl-module-paramters 5
+.Xr spl-module-parameters 5
for additional details. The default value is
.Sy off .
.It Sy version Ns = Ns Ar version
diff --git a/tests/zfs-tests/tests/functional/migration/migration_001_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_001_pos.ksh
index 4d5fbb9ff..875d2f7c7 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_001_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_001_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "tar cf $TESTDIR/tar$$.tar $BNAME"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $TESTDIR $SUMA $SUMB "tar xf $TESTDIR/tar$$.tar"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to ZFS fs"
-log_pass "Successully migrated test file from ZFS fs to ZFS fs".
+log_pass "Successfully migrated test file from ZFS fs to ZFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_002_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_002_pos.ksh
index e0655248d..6b97e2a40 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_002_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_002_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "tar cf $TESTDIR/tar$$.tar $BNAME"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $NONZFS_TESTDIR $SUMA $SUMB "tar xf $TESTDIR/tar$$.tar"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to UFS fs"
-log_pass "Successully migrated test file from ZFS fs to UFS fs".
+log_pass "Successfully migrated test file from ZFS fs to UFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_003_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_003_pos.ksh
index 904a2b1a0..dd0baeaa9 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_003_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_003_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "tar cf $NONZFS_TESTDIR/tar$$.tar $BNAME"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $TESTDIR $SUMA $SUMB "tar xvf $NONZFS_TESTDIR/tar$$.tar"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"UFS fs to ZFS fs"
-log_pass "Successully migrated test file from UFS fs to ZFS fs".
+log_pass "Successfully migrated test file from UFS fs to ZFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_004_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_004_pos.ksh
index 6d33dd5b3..00a6cc172 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_004_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_004_pos.ksh
@@ -67,7 +67,7 @@ cd $cwd
(( $? != 0 )) && log_untested "Could not change directory to $cwd"
migrate_cpio $TESTDIR "$TESTDIR/cpio$$.cpio" $SUMA $SUMB
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to ZFS fs"
-log_pass "Successully migrated test file from ZFS fs to ZFS fs".
+log_pass "Successfully migrated test file from ZFS fs to ZFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_005_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_005_pos.ksh
index a41b19b5f..4386596f7 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_005_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_005_pos.ksh
@@ -67,7 +67,7 @@ cd $cwd
(( $? != 0 )) && log_untested "Could not change directory to $cwd"
migrate_cpio $NONZFS_TESTDIR "$TESTDIR/cpio$$.cpio" $SUMA $SUMB
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to UFS fs"
-log_pass "Successully migrated test file from ZFS fs to UFS fs".
+log_pass "Successfully migrated test file from ZFS fs to UFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_006_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_006_pos.ksh
index 5b444421a..9b5c9166e 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_006_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_006_pos.ksh
@@ -67,7 +67,7 @@ cd $cwd
(( $? != 0 )) && log_untested "Could not change directory to $cwd"
migrate_cpio $TESTDIR "$NONZFS_TESTDIR/cpio$$.cpio" $SUMA $SUMB
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to ZFS fs"
-log_pass "Successully migrated test file from UFS fs to ZFS fs".
+log_pass "Successfully migrated test file from UFS fs to ZFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_007_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_007_pos.ksh
index c3197052c..0d136550f 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_007_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_007_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "dd if=$BNAME obs=128k of=$TESTDIR/dd$$.dd"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $TESTDIR $SUMA $SUMB "dd if=$TESTDIR/dd$$.dd obs=128k of=$BNAME"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to ZFS fs"
-log_pass "Successully migrated test file from ZFS fs to ZFS fs".
+log_pass "Successfully migrated test file from ZFS fs to ZFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_008_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_008_pos.ksh
index 2e51eef36..f62b1f33a 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_008_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_008_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "dd if=$BNAME obs=128k of=$TESTDIR/dd$$.dd"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $NONZFS_TESTDIR $SUMA $SUMB "dd if=$TESTDIR/dd$$.dd obs=128k of=$BNAME"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to ZFS fs"
-log_pass "Successully migrated test file from ZFS fs to UFS fs".
+log_pass "Successfully migrated test file from ZFS fs to UFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_009_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_009_pos.ksh
index 7749494e5..907be39eb 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_009_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_009_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "dd if=$BNAME obs=128k of=$NONZFS_TESTDIR/dd$$.dd"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $TESTDIR $SUMA $SUMB "dd if=$NONZFS_TESTDIR/dd$$.dd obs=128k of=$BNAME"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to ZFS fs"
-log_pass "Successully migrated test file from UFS fs to ZFS fs".
+log_pass "Successfully migrated test file from UFS fs to ZFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_010_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_010_pos.ksh
index a11ab72df..e80dd67cd 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_010_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_010_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "cp $BNAME $TESTDIR/cp$$.cp"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $TESTDIR $SUMA $SUMB "cp $TESTDIR/cp$$.cp $BNAME"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to ZFS fs"
-log_pass "Successully migrated test file from ZFS fs to ZFS fs".
+log_pass "Successfully migrated test file from ZFS fs to ZFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_011_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_011_pos.ksh
index 17e1c78f9..2d7ecb45e 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_011_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_011_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "cp $BNAME $TESTDIR/cp$$.cp"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $NONZFS_TESTDIR $SUMA $SUMB "cp $TESTDIR/cp$$.cp $BNAME"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"ZFS fs to UFS fs"
-log_pass "Successully migrated test file from ZFS fs to UFS fs".
+log_pass "Successfully migrated test file from ZFS fs to UFS fs".
diff --git a/tests/zfs-tests/tests/functional/migration/migration_012_pos.ksh b/tests/zfs-tests/tests/functional/migration/migration_012_pos.ksh
index 823dabeae..fd9c45491 100755
--- a/tests/zfs-tests/tests/functional/migration/migration_012_pos.ksh
+++ b/tests/zfs-tests/tests/functional/migration/migration_012_pos.ksh
@@ -60,7 +60,7 @@ prepare $DNAME "cp $BNAME $NONZFS_TESTDIR/cp$$.cp"
(( $? != 0 )) && log_fail "Unable to create src archive"
migrate $TESTDIR $SUMA $SUMB "cp $NONZFS_TESTDIR/cp$$.cp $BNAME"
-(( $? != 0 )) && log_fail "Uable to successfully migrate test file from" \
+(( $? != 0 )) && log_fail "Unable to successfully migrate test file from" \
"UFS fs to ZFS fs"
-log_pass "Successully migrated test file from UFS fs to ZFS fs".
+log_pass "Successfully migrated test file from UFS fs to ZFS fs".
--
2.14.2

46
zfs-patches/0046-Increment-zil_itx_needcopy_bytes-properly.patch Normal file
View File

@ -0,0 +1,46 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: chrisrd <chris@onthe.net.au>
Date: Sat, 3 Mar 2018 05:01:53 +1100
Subject: [PATCH] Increment zil_itx_needcopy_bytes properly
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
In zil_lwb_commit() with TX_WRITE, we copy the log write record (lrw)
into the log write block (lwb) and send it off using zil_lwb_add_txg().
If we also have WR_NEED_COPY, we additionally copy the lwr's data into
the lwb to be sent off. If the lwr + data doesn't fit into the lwb, we
send the lrw and as much data as will fit (dnow bytes), then go back
and do the same with the remaining data.
Each time through this loop we're sending dnow data bytes. I.e.
zil_itx_needcopy_bytes should be incremented by dnow.
Reviewed-by: Richard Elling <Richard.Elling@RichardElling.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Chris Dunlop <chris@onthe.net.au>
Closes #6988
Closes #7176
(cherry picked from commit 792f88131c647a70440c709c78d43210db6c6534)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/zil.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/module/zfs/zil.c b/module/zfs/zil.c
index 4d714cefc..645b1d4d8 100644
--- a/module/zfs/zil.c
+++ b/module/zfs/zil.c
@@ -1167,8 +1167,7 @@ cont:
lrw->lr_offset += dnow;
lrw->lr_length -= dnow;
ZIL_STAT_BUMP(zil_itx_needcopy_count);
- ZIL_STAT_INCR(zil_itx_needcopy_bytes,
- lrw->lr_length);
+ ZIL_STAT_INCR(zil_itx_needcopy_bytes, dnow);
} else {
ASSERT(itx->itx_wr_state == WR_INDIRECT);
dbuf = NULL;
--
2.14.2

201
zfs-patches/0047-Change-checksum-IO-delay-ratelimit-values.patch Normal file
View File

@ -0,0 +1,201 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Tony Hutter <hutter2@llnl.gov>
Date: Sun, 4 Mar 2018 17:34:51 -0800
Subject: [PATCH] Change checksum & IO delay ratelimit values
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Change checksum & IO delay ratelimit thresholds from 5/sec to 20/sec.
This allows zed to actually trigger if a bunch of these events arrive in
a short period of time (zed has a threshold of 10 events in 10 sec).
Previously, if you had, say, 100 checksum errors in 1 sec, it would get
ratelimited to 5/sec which wouldn't trigger zed to fault the drive.
Also, convert the checksum and IO delay thresholds to module params for
easy testing.
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
Closes #7252
(cherry picked from commit 6dc40e2ada2d0d008bd314ff3525f2b0acc2bb01)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/sys/vdev_impl.h | 2 --
include/sys/zfs_ratelimit.h | 12 +++++++++---
module/zcommon/zfs_comutil.c | 4 ++--
module/zfs/vdev.c | 23 +++++++++++++++++++++--
man/man5/zfs-module-parameters.5 | 39 +++++++++++++++++++++++++++++++++++++++
5 files changed, 71 insertions(+), 9 deletions(-)
diff --git a/include/sys/vdev_impl.h b/include/sys/vdev_impl.h
index 13c495822..4f9f1a903 100644
--- a/include/sys/vdev_impl.h
+++ b/include/sys/vdev_impl.h
@@ -255,8 +255,6 @@ struct vdev {
* We rate limit ZIO delay and ZIO checksum events, since they
* can flood ZED with tons of events when a drive is acting up.
*/
-#define DELAYS_PER_SECOND 5
-#define CHECKSUMS_PER_SECOND 5
zfs_ratelimit_t vdev_delay_rl;
zfs_ratelimit_t vdev_checksum_rl;
};
diff --git a/include/sys/zfs_ratelimit.h b/include/sys/zfs_ratelimit.h
index f36e07841..012825fad 100644
--- a/include/sys/zfs_ratelimit.h
+++ b/include/sys/zfs_ratelimit.h
@@ -25,13 +25,19 @@
typedef struct {
hrtime_t start;
unsigned int count;
- unsigned int burst; /* Number to allow per interval */
- unsigned int interval; /* Interval length in seconds */
+
+ /*
+ * Pointer to number of events per interval. We do this to
+ * allow the burst to be a (changeable) module parameter.
+ */
+ unsigned int *burst;
+
+ unsigned int interval; /* Interval length in seconds */
kmutex_t lock;
} zfs_ratelimit_t;
int zfs_ratelimit(zfs_ratelimit_t *rl);
-void zfs_ratelimit_init(zfs_ratelimit_t *rl, unsigned int burst,
+void zfs_ratelimit_init(zfs_ratelimit_t *rl, unsigned int *burst,
unsigned int interval);
void zfs_ratelimit_fini(zfs_ratelimit_t *rl);
diff --git a/module/zcommon/zfs_comutil.c b/module/zcommon/zfs_comutil.c
index 52cb7e365..44cdc8523 100644
--- a/module/zcommon/zfs_comutil.c
+++ b/module/zcommon/zfs_comutil.c
@@ -215,7 +215,7 @@ const char *zfs_history_event_names[ZFS_NUM_LEGACY_HISTORY_EVENTS] = {
* interval: Interval time in seconds
*/
void
-zfs_ratelimit_init(zfs_ratelimit_t *rl, unsigned int burst,
+zfs_ratelimit_init(zfs_ratelimit_t *rl, unsigned int *burst,
unsigned int interval)
{
rl->count = 0;
@@ -270,7 +270,7 @@ zfs_ratelimit(zfs_ratelimit_t *rl)
rl->start = now;
rl->count = 0;
} else {
- if (rl->count >= rl->burst) {
+ if (rl->count >= *rl->burst) {
rc = 0; /* We're ratelimiting */
}
}
diff --git a/module/zfs/vdev.c b/module/zfs/vdev.c
index df07d893d..0786fbb83 100644
--- a/module/zfs/vdev.c
+++ b/module/zfs/vdev.c
@@ -56,6 +56,16 @@
*/
int metaslabs_per_vdev = 200;
+/*
+ * Rate limit delay events to this many IO delays per second.
+ */
+unsigned int zfs_delays_per_second = 20;
+
+/*
+ * Rate limit checksum events after this many checksum errors per second.
+ */
+unsigned int zfs_checksums_per_second = 20;
+
/*
* Virtual device management.
*/
@@ -357,8 +367,8 @@ vdev_alloc_common(spa_t *spa, uint_t id, uint64_t guid, vdev_ops_t *ops)
* and checksum events so that we don't overwhelm ZED with thousands
* of events when a disk is acting up.
*/
- zfs_ratelimit_init(&vd->vdev_delay_rl, DELAYS_PER_SECOND, 1);
- zfs_ratelimit_init(&vd->vdev_checksum_rl, CHECKSUMS_PER_SECOND, 1);
+ zfs_ratelimit_init(&vd->vdev_delay_rl, &zfs_delays_per_second, 1);
+ zfs_ratelimit_init(&vd->vdev_checksum_rl, &zfs_checksums_per_second, 1);
list_link_init(&vd->vdev_config_dirty_node);
list_link_init(&vd->vdev_state_dirty_node);
@@ -3776,5 +3786,14 @@ module_param(metaslabs_per_vdev, int, 0644);
MODULE_PARM_DESC(metaslabs_per_vdev,
"Divide added vdev into approximately (but no more than) this number "
"of metaslabs");
+
+module_param(zfs_delays_per_second, uint, 0644);
+MODULE_PARM_DESC(zfs_delays_per_second, "Rate limit delay events to this many "
+ "IO delays per second");
+
+module_param(zfs_checksums_per_second, uint, 0644);
+ MODULE_PARM_DESC(zfs_checksums_per_second, "Rate limit checksum events "
+ "to this many checksum errors per second (do not set below zed"
+ "threshold).");
/* END CSTYLED */
#endif
diff --git a/man/man5/zfs-module-parameters.5 b/man/man5/zfs-module-parameters.5
index d4daffde6..8d5ac2576 100644
--- a/man/man5/zfs-module-parameters.5
+++ b/man/man5/zfs-module-parameters.5
@@ -739,6 +739,34 @@ Disable pool import at module load by ignoring the cache file (typically \fB/etc
Use \fB1\fR for yes (default) and \fB0\fR for no.
.RE
+.sp
+.ne 2
+.na
+\fBzfs_checksums_per_second\fR (int)
+.ad
+.RS 12n
+Rate limit checksum events to this many per second. Note that this should
+not be set below the zed thresholds (currently 10 checksums over 10 sec)
+or else zed may not trigger any action.
+.sp
+Default value: 20
+.RE
+
+.sp
+.ne 2
+.na
+\fBzfs_commit_timeout_pct\fR (int)
+.ad
+.RS 12n
+This controls the amount of time that a ZIL block (lwb) will remain "open"
+when it isn't "full", and it has a thread waiting for it to be committed to
+stable storage. The timeout is scaled based on a percentage of the last lwb
+latency to avoid significantly impacting the latency of each individual
+transaction record (itx).
+.sp
+Default value: \fB5\fR%.
+.RE
+
.sp
.ne 2
.na
@@ -866,6 +894,17 @@ Note: \fBzfs_delay_scale\fR * \fBzfs_dirty_data_max\fR must be < 2^64.
Default value: \fB500,000\fR.
.RE
+.sp
+.ne 2
+.na
+\fBzfs_delays_per_second\fR (int)
+.ad
+.RS 12n
+Rate limit IO delay events to this many per second.
+.sp
+Default value: 20
+.RE
+
.sp
.ne 2
.na
--
2.14.2

98
zfs-patches/0048-Linux-4.16-compat-get_disk_and_module.patch Normal file
View File

@ -0,0 +1,98 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Giuseppe Di Natale <dinatale2@users.noreply.github.com>
Date: Mon, 5 Mar 2018 12:44:35 -0800
Subject: [PATCH] Linux 4.16 compat: get_disk_and_module()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
As of https://github.com/torvalds/linux/commit/fb6d47a, get_disk()
is now get_disk_and_module(). Add a configure check to determine
if we need to use get_disk_and_module().
Reviewed-by: loli10K <ezomori.nozomu@gmail.com>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Closes #7264
(cherry picked from commit 8d7f17798d0faa0001bf7257759be0c688b91044)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
include/linux/blkdev_compat.h | 8 ++++++++
module/zfs/zvol.c | 2 +-
config/kernel-get-disk-and-module.m4 | 19 +++++++++++++++++++
config/kernel.m4 | 1 +
4 files changed, 29 insertions(+), 1 deletion(-)
create mode 100644 config/kernel-get-disk-and-module.m4
diff --git a/include/linux/blkdev_compat.h b/include/linux/blkdev_compat.h
index c8a8e856d..4406493e4 100644
--- a/include/linux/blkdev_compat.h
+++ b/include/linux/blkdev_compat.h
@@ -139,6 +139,14 @@ blk_queue_set_read_ahead(struct request_queue *q, unsigned long ra_pages)
#endif
}
+#ifndef HAVE_GET_DISK_AND_MODULE
+static inline struct kobject *
+get_disk_and_module(struct gendisk *disk)
+{
+ return (get_disk(disk));
+}
+#endif
+
#ifndef HAVE_GET_DISK_RO
static inline int
get_disk_ro(struct gendisk *disk)
diff --git a/module/zfs/zvol.c b/module/zfs/zvol.c
index aac494209..62176e1cf 100644
--- a/module/zfs/zvol.c
+++ b/module/zfs/zvol.c
@@ -1559,7 +1559,7 @@ zvol_probe(dev_t dev, int *part, void *arg)
struct kobject *kobj;
zv = zvol_find_by_dev(dev);
- kobj = zv ? get_disk(zv->zv_disk) : NULL;
+ kobj = zv ? get_disk_and_module(zv->zv_disk) : NULL;
ASSERT(zv == NULL || MUTEX_HELD(&zv->zv_state_lock));
if (zv)
mutex_exit(&zv->zv_state_lock);
diff --git a/config/kernel-get-disk-and-module.m4 b/config/kernel-get-disk-and-module.m4
new file mode 100644
index 000000000..2a51a5af7
--- /dev/null
+++ b/config/kernel-get-disk-and-module.m4
@@ -0,0 +1,19 @@
+dnl #
+dnl # 4.16 API change
+dnl # Verify if get_disk_and_module() symbol is available.
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_GET_DISK_AND_MODULE],
+ [AC_MSG_CHECKING([whether get_disk_and_module() is available])
+ ZFS_LINUX_TRY_COMPILE_SYMBOL([
+ #include <linux/genhd.h>
+ ], [
+ struct gendisk *disk = NULL;
+ (void) get_disk_and_module(disk);
+ ], [get_disk_and_module], [block/genhd.c], [
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(HAVE_GET_DISK_AND_MODULE,
+ 1, [get_disk_and_module() is available])
+ ], [
+ AC_MSG_RESULT(no)
+ ])
+])
diff --git a/config/kernel.m4 b/config/kernel.m4
index 3e499e447..419ed1a2c 100644
--- a/config/kernel.m4
+++ b/config/kernel.m4
@@ -40,6 +40,7 @@ AC_DEFUN([ZFS_AC_CONFIG_KERNEL], [
ZFS_AC_KERNEL_BLK_QUEUE_MAX_SEGMENTS
ZFS_AC_KERNEL_BLK_QUEUE_HAVE_BIO_RW_UNPLUG
ZFS_AC_KERNEL_BLK_QUEUE_HAVE_BLK_PLUG
+ ZFS_AC_KERNEL_GET_DISK_AND_MODULE
ZFS_AC_KERNEL_GET_DISK_RO
ZFS_AC_KERNEL_GET_GENDISK
ZFS_AC_KERNEL_HAVE_BIO_SET_OP_ATTRS
--
2.14.2

46
zfs-patches/0049-Detect-long-config-lock-acquisition-in-mmp.patch Normal file
View File

@ -0,0 +1,46 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Olaf Faaland <faaland1@llnl.gov>
Date: Tue, 20 Feb 2018 17:33:51 -0800
Subject: [PATCH] Detect long config lock acquisition in mmp
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
If something holds the config lock as a writer for too long, MMP will
fail to issue MMP writes in a timely manner. This will result either in
the pool being suspended, or in an extreme case, in the pool not being
protected.
If the time to acquire the config lock exceeds 1/10 of the minimum
zfs_multihost_interval, report it in the zfs debug log.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes #7212
(cherry picked from commit c17922b8a9db00ad7e7d59a5ff975b2a1edcf887)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/mmp.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index ee8e9201b..634e48287 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -323,7 +323,13 @@ mmp_write_uberblock(spa_t *spa)
int label;
uint64_t offset;
+ hrtime_t lock_acquire_time = gethrtime();
spa_config_enter(spa, SCL_STATE, mmp_tag, RW_READER);
+ lock_acquire_time = gethrtime() - lock_acquire_time;
+ if (lock_acquire_time > (MSEC2NSEC(MMP_MIN_INTERVAL) / 10))
+ zfs_dbgmsg("SCL_STATE acquisition took %llu ns\n",
+ (u_longlong_t)lock_acquire_time);
+
vd = mmp_random_leaf(spa->spa_root_vdev);
if (vd == NULL) {
spa_config_exit(spa, SCL_STATE, FTAG);
--
2.14.2

786
zfs-patches/0050-Take-user-namespaces-into-account-in-policy-checks.patch Normal file
View File

@ -0,0 +1,786 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Wolfgang Bumiller <Blub@users.noreply.github.com>
Date: Thu, 8 Mar 2018 00:40:42 +0100
Subject: [PATCH] Take user namespaces into account in policy checks
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Change file related checks to use user namespaces and make
sure involved uids/gids are mappable in the current
namespace.
Note that checks without file ownership information will
still not take user namespaces into account, as some of
these should be handled via 'zfs allow' (otherwise root in a
user namespace could issue commands such as `zpool export`).
This also adds an initial user namespace regression test
for the setgid bit loss, with a user_ns_exec helper usable
in further tests.
Additionally, configure checks for the required user
namespace related features are added for:
* ns_capable
* kuid/kgid_has_mapping()
* user_ns in cred_t
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Closes #6800
Closes #7270
(cherry picked from commit 3808006edfc46b18f0a40a2e9df54c6567bf52cc)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
configure.ac | 2 +
tests/zfs-tests/cmd/Makefile.am | 1 +
tests/zfs-tests/cmd/user_ns_exec/Makefile.am | 6 +
tests/zfs-tests/tests/functional/Makefile.am | 1 +
.../tests/functional/user_namespace/Makefile.am | 7 +
module/zfs/policy.c | 66 +++++++-
tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c | 179 +++++++++++++++++++++
config/kernel-userns-capabilities.m4 | 67 ++++++++
config/kernel.m4 | 1 +
tests/runfiles/linux.run | 4 +
tests/zfs-tests/cmd/user_ns_exec/.gitignore | 1 +
tests/zfs-tests/include/commands.cfg | 1 +
.../tests/functional/user_namespace/cleanup.ksh | 25 +++
.../tests/functional/user_namespace/setup.ksh | 32 ++++
.../functional/user_namespace/user_namespace.cfg | 23 +++
.../user_namespace/user_namespace_001.ksh | 89 ++++++++++
.../user_namespace/user_namespace_common.kshlib | 23 +++
17 files changed, 521 insertions(+), 7 deletions(-)
create mode 100644 tests/zfs-tests/cmd/user_ns_exec/Makefile.am
create mode 100644 tests/zfs-tests/tests/functional/user_namespace/Makefile.am
create mode 100644 tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c
create mode 100644 config/kernel-userns-capabilities.m4
create mode 100644 tests/zfs-tests/cmd/user_ns_exec/.gitignore
create mode 100755 tests/zfs-tests/tests/functional/user_namespace/cleanup.ksh
create mode 100755 tests/zfs-tests/tests/functional/user_namespace/setup.ksh
create mode 100644 tests/zfs-tests/tests/functional/user_namespace/user_namespace.cfg
create mode 100755 tests/zfs-tests/tests/functional/user_namespace/user_namespace_001.ksh
create mode 100644 tests/zfs-tests/tests/functional/user_namespace/user_namespace_common.kshlib
diff --git a/configure.ac b/configure.ac
index d71712e4c..77e5764fc 100644
--- a/configure.ac
+++ b/configure.ac
@@ -153,6 +153,7 @@ AC_CONFIG_FILES([
tests/zfs-tests/callbacks/Makefile
tests/zfs-tests/cmd/Makefile
tests/zfs-tests/cmd/chg_usr_exec/Makefile
+ tests/zfs-tests/cmd/user_ns_exec/Makefile
tests/zfs-tests/cmd/devname2devid/Makefile
tests/zfs-tests/cmd/dir_rd_update/Makefile
tests/zfs-tests/cmd/file_check/Makefile
@@ -284,6 +285,7 @@ AC_CONFIG_FILES([
tests/zfs-tests/tests/functional/threadsappend/Makefile
tests/zfs-tests/tests/functional/tmpfile/Makefile
tests/zfs-tests/tests/functional/truncate/Makefile
+ tests/zfs-tests/tests/functional/user_namespace/Makefile
tests/zfs-tests/tests/functional/userquota/Makefile
tests/zfs-tests/tests/functional/upgrade/Makefile
tests/zfs-tests/tests/functional/vdev_zaps/Makefile
diff --git a/tests/zfs-tests/cmd/Makefile.am b/tests/zfs-tests/cmd/Makefile.am
index f55ff8ce2..1cce6947b 100644
--- a/tests/zfs-tests/cmd/Makefile.am
+++ b/tests/zfs-tests/cmd/Makefile.am
@@ -2,6 +2,7 @@ EXTRA_DIST = file_common.h
SUBDIRS = \
chg_usr_exec \
+ user_ns_exec \
devname2devid \
dir_rd_update \
file_check \
diff --git a/tests/zfs-tests/cmd/user_ns_exec/Makefile.am b/tests/zfs-tests/cmd/user_ns_exec/Makefile.am
new file mode 100644
index 000000000..5b4bc9aaa
--- /dev/null
+++ b/tests/zfs-tests/cmd/user_ns_exec/Makefile.am
@@ -0,0 +1,6 @@
+include $(top_srcdir)/config/Rules.am
+
+pkgexecdir = $(datadir)/@PACKAGE@/zfs-tests/bin
+
+pkgexec_PROGRAMS = user_ns_exec
+user_ns_exec_SOURCES = user_ns_exec.c
diff --git a/tests/zfs-tests/tests/functional/Makefile.am b/tests/zfs-tests/tests/functional/Makefile.am
index d68f254ef..cd60324f3 100644
--- a/tests/zfs-tests/tests/functional/Makefile.am
+++ b/tests/zfs-tests/tests/functional/Makefile.am
@@ -58,6 +58,7 @@ SUBDIRS = \
tmpfile \
truncate \
upgrade \
+ user_namespace \
userquota \
vdev_zaps \
write_dirs \
diff --git a/tests/zfs-tests/tests/functional/user_namespace/Makefile.am b/tests/zfs-tests/tests/functional/user_namespace/Makefile.am
new file mode 100644
index 000000000..0c0f6887a
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/user_namespace/Makefile.am
@@ -0,0 +1,7 @@
+pkgdatadir = $(datadir)/@PACKAGE@/zfs-tests/tests/functional/user_namespace
+dist_pkgdata_SCRIPTS = \
+ setup.ksh \
+ cleanup.ksh \
+ user_namespace_common.kshlib \
+ user_namespace.cfg \
+ user_namespace_001.ksh
diff --git a/module/zfs/policy.c b/module/zfs/policy.c
index 03e8f748b..55c932747 100644
--- a/module/zfs/policy.c
+++ b/module/zfs/policy.c
@@ -42,19 +42,47 @@
* all other cases this function must fail and return the passed err.
*/
static int
-priv_policy(const cred_t *cr, int capability, boolean_t all, int err)
+priv_policy_ns(const cred_t *cr, int capability, boolean_t all, int err,
+ struct user_namespace *ns)
{
ASSERT3S(all, ==, B_FALSE);
if (cr != CRED() && (cr != kcred))
return (err);
+#if defined(CONFIG_USER_NS) && defined(HAVE_NS_CAPABLE)
+ if (!(ns ? ns_capable(ns, capability) : capable(capability)))
+#else
if (!capable(capability))
+#endif
return (err);
return (0);
}
+static int
+priv_policy(const cred_t *cr, int capability, boolean_t all, int err)
+{
+ return (priv_policy_ns(cr, capability, all, err, NULL));
+}
+
+static int
+priv_policy_user(const cred_t *cr, int capability, boolean_t all, int err)
+{
+ /*
+ * All priv_policy_user checks are preceeded by kuid/kgid_has_mapping()
+ * checks. If we cannot do them, we shouldn't be using ns_capable()
+ * since we don't know whether the affected files are valid in our
+ * namespace. Note that kuid_has_mapping() came after cred->user_ns, so
+ * we shouldn't need to re-check for HAVE_CRED_USER_NS
+ */
+#if defined(CONFIG_USER_NS) && defined(HAVE_KUID_HAS_MAPPING)
+ return (priv_policy_ns(cr, capability, all, err, cr->user_ns));
+#else
+ return (priv_policy_ns(cr, capability, all, err, NULL));
+#endif
+}
+
/*
* Checks for operations that are either client-only or are used by
* both clients and servers.
@@ -102,10 +130,15 @@ secpolicy_vnode_any_access(const cred_t *cr, struct inode *ip, uid_t owner)
if (zpl_inode_owner_or_capable(ip))
return (0);
- if (priv_policy(cr, CAP_DAC_OVERRIDE, B_FALSE, EPERM) == 0)
+#if defined(CONFIG_USER_NS) && defined(HAVE_KUID_HAS_MAPPING)
+ if (!kuid_has_mapping(cr->user_ns, SUID_TO_KUID(owner)))
+ return (EPERM);
+#endif
+
+ if (priv_policy_user(cr, CAP_DAC_OVERRIDE, B_FALSE, EPERM) == 0)
return (0);
- if (priv_policy(cr, CAP_DAC_READ_SEARCH, B_FALSE, EPERM) == 0)
+ if (priv_policy_user(cr, CAP_DAC_READ_SEARCH, B_FALSE, EPERM) == 0)
return (0);
return (EPERM);
@@ -120,7 +153,12 @@ secpolicy_vnode_chown(const cred_t *cr, uid_t owner)
if (crgetfsuid(cr) == owner)
return (0);
- return (priv_policy(cr, CAP_FOWNER, B_FALSE, EPERM));
+#if defined(CONFIG_USER_NS) && defined(HAVE_KUID_HAS_MAPPING)
+ if (!kuid_has_mapping(cr->user_ns, SUID_TO_KUID(owner)))
+ return (EPERM);
+#endif
+
+ return (priv_policy_user(cr, CAP_FOWNER, B_FALSE, EPERM));
}
/*
@@ -152,7 +190,12 @@ secpolicy_vnode_setdac(const cred_t *cr, uid_t owner)
if (crgetfsuid(cr) == owner)
return (0);
- return (priv_policy(cr, CAP_FOWNER, B_FALSE, EPERM));
+#if defined(CONFIG_USER_NS) && defined(HAVE_KUID_HAS_MAPPING)
+ if (!kuid_has_mapping(cr->user_ns, SUID_TO_KUID(owner)))
+ return (EPERM);
+#endif
+
+ return (priv_policy_user(cr, CAP_FOWNER, B_FALSE, EPERM));
}
/*
@@ -175,8 +218,12 @@ secpolicy_vnode_setid_retain(const cred_t *cr, boolean_t issuidroot)
int
secpolicy_vnode_setids_setgids(const cred_t *cr, gid_t gid)
{
+#if defined(CONFIG_USER_NS) && defined(HAVE_KUID_HAS_MAPPING)
+ if (!kgid_has_mapping(cr->user_ns, SGID_TO_KGID(gid)))
+ return (EPERM);
+#endif
if (crgetfsgid(cr) != gid && !groupmember(gid, cr))
- return (priv_policy(cr, CAP_FSETID, B_FALSE, EPERM));
+ return (priv_policy_user(cr, CAP_FSETID, B_FALSE, EPERM));
return (0);
}
@@ -222,7 +269,12 @@ secpolicy_vnode_setid_modify(const cred_t *cr, uid_t owner)
if (crgetfsuid(cr) == owner)
return (0);
- return (priv_policy(cr, CAP_FSETID, B_FALSE, EPERM));
+#if defined(CONFIG_USER_NS) && defined(HAVE_KUID_HAS_MAPPING)
+ if (!kuid_has_mapping(cr->user_ns, SUID_TO_KUID(owner)))
+ return (EPERM);
+#endif
+
+ return (priv_policy_user(cr, CAP_FSETID, B_FALSE, EPERM));
}
/*
diff --git a/tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c b/tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c
new file mode 100644
index 000000000..cd46738bd
--- /dev/null
+++ b/tests/zfs-tests/cmd/user_ns_exec/user_ns_exec.c
@@ -0,0 +1,179 @@
+/*
+ * 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
+ */
+
+#include <stdio.h>
+#include <unistd.h>
+#include <string.h>
+#include <limits.h>
+#include <sys/types.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/wait.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <signal.h>
+#include <sched.h>
+
+#define EXECSHELL "/bin/sh"
+#define UIDMAP "0 100000 65536"
+
+static int
+child_main(int argc, char *argv[], int sync_pipe)
+{
+ char sync_buf;
+ char cmds[BUFSIZ] = { 0 };
+ char sep[] = " ";
+ int i, len;
+
+ if (unshare(CLONE_NEWUSER | CLONE_NEWNS) != 0) {
+ perror("unshare");
+ return (1);
+ }
+
+ /* tell parent we entered the new namespace */
+ if (write(sync_pipe, "1", 1) != 1) {
+ perror("write");
+ return (1);
+ }
+
+ /* wait for parent to setup the uid mapping */
+ if (read(sync_pipe, &sync_buf, 1) != 1) {
+ (void) fprintf(stderr, "user namespace setup failed\n");
+ return (1);
+ }
+
+ close(sync_pipe);
+
+ if (setuid(0) != 0) {
+ perror("setuid");
+ return (1);
+ }
+ if (setgid(0) != 0) {
+ perror("setgid");
+ return (1);
+ }
+
+ len = 0;
+ for (i = 1; i < argc; i++) {
+ (void) snprintf(cmds+len, sizeof (cmds)-len,
+ "%s%s", argv[i], sep);
+ len += strlen(argv[i]) + strlen(sep);
+ }
+
+ if (execl(EXECSHELL, "sh", "-c", cmds, (char *)NULL) != 0) {
+ perror("execl: " EXECSHELL);
+ return (1);
+ }
+
+ return (0);
+}
+
+static int
+set_idmap(pid_t pid, const char *file)
+{
+ int result = 0;
+ int mapfd;
+ char path[PATH_MAX];
+
+ (void) snprintf(path, sizeof (path), "/proc/%d/%s", (int)pid, file);
+
+ mapfd = open(path, O_WRONLY);
+ if (mapfd < 0) {
+ result = errno;
+ perror("open");
+ return (errno);
+ }
+
+ if (write(mapfd, UIDMAP, sizeof (UIDMAP)-1) != sizeof (UIDMAP)-1) {
+ perror("write");
+ result = (errno);
+ }
+
+ close(mapfd);
+
+ return (result);
+}
+
+int
+main(int argc, char *argv[])
+{
+ char sync_buf;
+ int result, wstatus;
+ int syncfd[2];
+ pid_t child;
+
+ if (argc < 2 || strlen(argv[1]) == 0) {
+ (void) printf("\tUsage: %s <commands> ...\n", argv[0]);
+ return (1);
+ }
+
+ if (socketpair(AF_UNIX, SOCK_STREAM, 0, syncfd) != 0) {
+ perror("socketpair");
+ return (1);
+ }
+
+ child = fork();
+ if (child == (pid_t)-1) {
+ perror("fork");
+ return (1);
+ }
+
+ if (child == 0) {
+ close(syncfd[0]);
+ return (child_main(argc, argv, syncfd[1]));
+ }
+
+ close(syncfd[1]);
+
+ result = 0;
+ /* wait for the child to have unshared its namespaces */
+ if (read(syncfd[0], &sync_buf, 1) != 1) {
+ perror("read");
+ kill(child, SIGKILL);
+ result = 1;
+ goto reap;
+ }
+
+ /* write uid mapping */
+ if (set_idmap(child, "uid_map") != 0 ||
+ set_idmap(child, "gid_map") != 0) {
+ result = 1;
+ kill(child, SIGKILL);
+ goto reap;
+ }
+
+ /* tell the child to proceed */
+ if (write(syncfd[0], "1", 1) != 1) {
+ perror("write");
+ kill(child, SIGKILL);
+ result = 1;
+ goto reap;
+ }
+ close(syncfd[0]);
+
+reap:
+ while (waitpid(child, &wstatus, 0) != child)
+ kill(child, SIGKILL);
+ if (result == 0)
+ result = WEXITSTATUS(wstatus);
+
+ return (result);
+}
diff --git a/config/kernel-userns-capabilities.m4 b/config/kernel-userns-capabilities.m4
new file mode 100644
index 000000000..fa3381978
--- /dev/null
+++ b/config/kernel-userns-capabilities.m4
@@ -0,0 +1,67 @@
+dnl #
+dnl # 2.6.38 API change
+dnl # ns_capable() was introduced
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_NS_CAPABLE], [
+ AC_MSG_CHECKING([whether ns_capable exists])
+ ZFS_LINUX_TRY_COMPILE([
+ #include <linux/capability.h>
+ ],[
+ ns_capable((struct user_namespace *)NULL, CAP_SYS_ADMIN);
+ ],[
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(HAVE_NS_CAPABLE, 1,
+ [ns_capable exists])
+ ],[
+ AC_MSG_RESULT(no)
+ ])
+])
+
+dnl #
+dnl # 2.6.39 API change
+dnl # struct user_namespace was added to struct cred_t as
+dnl # cred->user_ns member
+dnl # Note that current_user_ns() was added in 2.6.28.
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_CRED_USER_NS], [
+ AC_MSG_CHECKING([whether cred_t->user_ns exists])
+ ZFS_LINUX_TRY_COMPILE([
+ #include <linux/cred.h>
+ ],[
+ struct cred cr;
+ cr.user_ns = (struct user_namespace *)NULL;
+ ],[
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(HAVE_CRED_USER_NS, 1,
+ [cred_t->user_ns exists])
+ ],[
+ AC_MSG_RESULT(no)
+ ])
+])
+
+dnl #
+dnl # 3.4 API change
+dnl # kuid_has_mapping() and kgid_has_mapping() were added to distinguish
+dnl # between internal kernel uids/gids and user namespace uids/gids.
+dnl #
+AC_DEFUN([ZFS_AC_KERNEL_KUID_HAS_MAPPING], [
+ AC_MSG_CHECKING([whether kuid_has_mapping/kgid_has_mapping exist])
+ ZFS_LINUX_TRY_COMPILE([
+ #include <linux/uidgid.h>
+ ],[
+ kuid_has_mapping((struct user_namespace *)NULL, KUIDT_INIT(0));
+ kgid_has_mapping((struct user_namespace *)NULL, KGIDT_INIT(0));
+ ],[
+ AC_MSG_RESULT(yes)
+ AC_DEFINE(HAVE_KUID_HAS_MAPPING, 1,
+ [kuid_has_mapping/kgid_has_mapping exist])
+ ],[
+ AC_MSG_RESULT(no)
+ ])
+])
+
+AC_DEFUN([ZFS_AC_KERNEL_USERNS_CAPABILITIES], [
+ ZFS_AC_KERNEL_NS_CAPABLE
+ ZFS_AC_KERNEL_CRED_USER_NS
+ ZFS_AC_KERNEL_KUID_HAS_MAPPING
+])
diff --git a/config/kernel.m4 b/config/kernel.m4
index 419ed1a2c..910d4ff25 100644
--- a/config/kernel.m4
+++ b/config/kernel.m4
@@ -126,6 +126,7 @@ AC_DEFUN([ZFS_AC_CONFIG_KERNEL], [
ZFS_AC_KERNEL_CURRENT_TIME
ZFS_AC_KERNEL_GLOBAL_PAGE_STATE
ZFS_AC_KERNEL_ACL_HAS_REFCOUNT
+ ZFS_AC_KERNEL_USERNS_CAPABILITIES
AS_IF([test "$LINUX_OBJ" != "$LINUX"], [
KERNELMAKE_PARAMS="$KERNELMAKE_PARAMS O=$LINUX_OBJ"
diff --git a/tests/runfiles/linux.run b/tests/runfiles/linux.run
index 89c923db1..25ae3fe5e 100644
--- a/tests/runfiles/linux.run
+++ b/tests/runfiles/linux.run
@@ -648,6 +648,10 @@ tags = ['functional', 'truncate']
tests = [ 'upgrade_userobj_001_pos' ]
tags = ['functional', 'upgrade']
+[tests/functional/user_namespace]
+tests = ['user_namespace_001']
+tags = ['functional', 'user_namespace']
+
[tests/functional/userquota]
tests = [
'userquota_001_pos', 'userquota_002_pos', 'userquota_003_pos',
diff --git a/tests/zfs-tests/cmd/user_ns_exec/.gitignore b/tests/zfs-tests/cmd/user_ns_exec/.gitignore
new file mode 100644
index 000000000..655867a64
--- /dev/null
+++ b/tests/zfs-tests/cmd/user_ns_exec/.gitignore
@@ -0,0 +1 @@
+/user_ns_exec
diff --git a/tests/zfs-tests/include/commands.cfg b/tests/zfs-tests/include/commands.cfg
index 936e54c1a..0d768a3cf 100644
--- a/tests/zfs-tests/include/commands.cfg
+++ b/tests/zfs-tests/include/commands.cfg
@@ -164,4 +164,5 @@ export ZFSTEST_FILES='chg_usr_exec
rename_dir
rm_lnkcnt_zero_file
threadsappend
+ user_ns_exec
xattrtest'
diff --git a/tests/zfs-tests/tests/functional/user_namespace/cleanup.ksh b/tests/zfs-tests/tests/functional/user_namespace/cleanup.ksh
new file mode 100755
index 000000000..61caf3910
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/user_namespace/cleanup.ksh
@@ -0,0 +1,25 @@
+#!/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
+#
+
+. $STF_SUITE/include/libtest.shlib
+
+default_cleanup
diff --git a/tests/zfs-tests/tests/functional/user_namespace/setup.ksh b/tests/zfs-tests/tests/functional/user_namespace/setup.ksh
new file mode 100755
index 000000000..354cc9a6b
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/user_namespace/setup.ksh
@@ -0,0 +1,32 @@
+#!/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
+#
+
+. $STF_SUITE/include/libtest.shlib
+
+if ! [ -f /proc/self/uid_map ]; then
+ log_unsupported "The kernel doesn't support user namespaces."
+fi
+
+verify_runnable "both"
+
+DISK=${DISKS%% *}
+default_setup $DISK
diff --git a/tests/zfs-tests/tests/functional/user_namespace/user_namespace.cfg b/tests/zfs-tests/tests/functional/user_namespace/user_namespace.cfg
new file mode 100644
index 000000000..9e55398e2
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/user_namespace/user_namespace.cfg
@@ -0,0 +1,23 @@
+#
+# 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
+#
+
+export ROOT_UID=100000
+export OTHER_UID=101000
diff --git a/tests/zfs-tests/tests/functional/user_namespace/user_namespace_001.ksh b/tests/zfs-tests/tests/functional/user_namespace/user_namespace_001.ksh
new file mode 100755
index 000000000..6be30ab4d
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/user_namespace/user_namespace_001.ksh
@@ -0,0 +1,89 @@
+#!/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
+#
+
+. $STF_SUITE/tests/functional/user_namespace/user_namespace_common.kshlib
+
+#
+#
+# DESCRIPTION:
+# Regression test for secpolicy_vnode_setids_setgids
+#
+#
+# STRATEGY:
+# 1. Create files with various owners.
+# 2. Try to set setgid bit.
+#
+
+verify_runnable "both"
+
+# rroot: real root,
+# uroot: root within user namespace
+# uother: other user within user namespace
+set -A files rroot_rroot uroot_uroot uroot_other uother_uroot uother_uother
+
+function cleanup
+{
+ for i in ${files[*]}; do
+ log_must rm -f $TESTDIR/$i
+ done
+}
+
+log_onexit cleanup
+
+log_assert "Check root in user namespaces"
+
+TOUCH=$(readlink -e $(which touch))
+CHMOD=$(readlink -e $(which chmod))
+
+for i in ${files[*]}; do
+ log_must $TOUCH $TESTDIR/$i
+ log_must $CHMOD 0644 $TESTDIR/$i
+done
+
+log_must chown 0:0 $TESTDIR/rroot_rroot
+log_must chown $ROOT_UID:$ROOT_UID $TESTDIR/uroot_uroot
+log_must chown $ROOT_UID:$OTHER_UID $TESTDIR/uroot_other
+log_must chown $OTHER_UID:$ROOT_UID $TESTDIR/uother_uroot
+log_must chown $OTHER_UID:$OTHER_UID $TESTDIR/uother_uother
+
+log_mustnot user_ns_exec $CHMOD 02755 $TESTDIR/rroot_rroot
+log_mustnot test -g $TESTDIR/rroot_rroot
+
+log_must user_ns_exec $CHMOD 02755 $TESTDIR/uroot_uroot
+log_must test -g $TESTDIR/uroot_uroot
+
+log_must user_ns_exec $CHMOD 02755 $TESTDIR/uroot_other
+log_must test -g $TESTDIR/uroot_other
+
+log_must user_ns_exec $CHMOD 02755 $TESTDIR/uother_uroot
+log_must test -g $TESTDIR/uother_uroot
+
+log_must user_ns_exec $CHMOD 02755 $TESTDIR/uother_uother
+log_must test -g $TESTDIR/uother_uother
+
+log_mustnot user_ns_exec $TOUCH $TESTDIR/rroot_rroot
+log_must $CHMOD 0666 $TESTDIR/rroot_rroot
+for i in ${files[*]}; do
+ log_must user_ns_exec $TOUCH $TESTDIR/$i
+done
+
+log_pass "Check root in user namespaces"
diff --git a/tests/zfs-tests/tests/functional/user_namespace/user_namespace_common.kshlib b/tests/zfs-tests/tests/functional/user_namespace/user_namespace_common.kshlib
new file mode 100644
index 000000000..8577294d0
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/user_namespace/user_namespace_common.kshlib
@@ -0,0 +1,23 @@
+#
+# 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
+#
+
+. $STF_SUITE/include/libtest.shlib
+. $STF_SUITE/tests/functional/user_namespace/user_namespace.cfg
--
2.14.2

232
zfs-patches/0051-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch Normal file
View File

@ -0,0 +1,232 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Paul Zuchowski <31706010+PaulZ-98@users.noreply.github.com>
Date: Wed, 7 Mar 2018 20:03:33 -0500
Subject: [PATCH] zdb and inuse tests don't pass with real disks
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Due to zpool create auto-partioning in Linux (i.e. sdb1),
certain utilities need to use the parition (sdb1) while
others use the whole disk name (sdb).
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Paul Zuchowski <pzuchowski@datto.com>
Closes #6939
Closes #7261
(cherry picked from commit 0a0af41bd926ca20a26d7476d792ce82bca9c077)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
tests/zfs-tests/include/libtest.shlib | 31 +++++++++++++++++++---
.../tests/functional/cli_root/zdb/zdb_003_pos.ksh | 5 ++++
.../tests/functional/cli_root/zdb/zdb_004_pos.ksh | 11 +++++---
.../tests/functional/cli_root/zdb/zdb_005_pos.ksh | 4 +++
.../tests/functional/inuse/inuse_005_pos.ksh | 8 +++---
.../tests/functional/inuse/inuse_008_pos.ksh | 5 ++--
.../tests/functional/inuse/inuse_009_pos.ksh | 10 +++++--
7 files changed, 59 insertions(+), 15 deletions(-)
diff --git a/tests/zfs-tests/include/libtest.shlib b/tests/zfs-tests/include/libtest.shlib
index 48fb5e7c5..d51d73e61 100644
--- a/tests/zfs-tests/include/libtest.shlib
+++ b/tests/zfs-tests/include/libtest.shlib
@@ -767,7 +767,11 @@ function zero_partitions #<whole_disk_name>
typeset i
if is_linux; then
- log_must parted $DEV_DSKDIR/$diskname -s -- mklabel gpt
+ DSK=$DEV_DSKDIR/$diskname
+ DSK=$(echo $DSK | sed -e "s|//|/|g")
+ log_must parted $DSK -s -- mklabel gpt
+ blockdev --rereadpt $DSK 2>/dev/null
+ block_device_wait
else
for i in 0 1 3 4 5 6 7
do
@@ -795,10 +799,11 @@ function set_partition #<slice_num> <slice_start> <size_plus_units> <whole_disk
typeset start=$2
typeset size=$3
typeset disk=$4
- [[ -z $slicenum || -z $size || -z $disk ]] && \
- log_fail "The slice, size or disk name is unspecified."
if is_linux; then
+ if [[ -z $size || -z $disk ]]; then
+ log_fail "The size or disk name is unspecified."
+ fi
typeset size_mb=${size%%[mMgG]}
size_mb=${size_mb%%[mMgG][bB]}
@@ -841,6 +846,10 @@ function set_partition #<slice_num> <slice_start> <size_plus_units> <whole_disk
blockdev --rereadpt $DEV_DSKDIR/$disk 2>/dev/null
block_device_wait
else
+ if [[ -z $slicenum || -z $size || -z $disk ]]; then
+ log_fail "The slice, size or disk name is unspecified."
+ fi
+
typeset format_file=/var/tmp/format_in.$$
echo "partition" >$format_file
@@ -3634,3 +3643,19 @@ function get_tunable_impl
return 1
}
+
+#
+# Get actual devices used by the pool (i.e. linux sdb1 not sdb).
+#
+function get_pool_devices #testpool #devdir
+{
+ typeset testpool=$1
+ typeset devdir=$2
+ typeset out=""
+
+ if is_linux; then
+ out=$(zpool status -P $testpool |grep ${devdir} | awk '{print $1}')
+ out=$(echo $out | sed -e "s|${devdir}/||g" | tr '\n' ' ')
+ fi
+ echo $out
+}
diff --git a/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_003_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_003_pos.ksh
index 4c57cb587..3c444ae98 100755
--- a/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_003_pos.ksh
+++ b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_003_pos.ksh
@@ -43,6 +43,11 @@ config_count=(1 2)
set -A DISK $DISKS
default_mirror_setup_noexit $DISKS
+
+DEVS=$(get_pool_devices ${TESTPOOL} ${DEV_RDSKDIR})
+log_note "$DEVS"
+[[ -n $DEVS ]] && set -A DISK $DEVS
+
log_must dd if=/dev/${DISK[0]} of=/dev/${DISK[1]} bs=1K count=256 conv=notrunc
for x in 0 1 ; do
diff --git a/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_004_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_004_pos.ksh
index 2b4057aa7..91a5c9799 100755
--- a/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_004_pos.ksh
+++ b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_004_pos.ksh
@@ -45,17 +45,22 @@ function cleanup
verify_runnable "global"
verify_disk_count "$DISKS" 2
set -A DISK $DISKS
+WHOLE_DISK=${DISK[0]}
default_mirror_setup_noexit $DISKS
-log_must zpool offline $TESTPOOL ${DISK[0]}
+DEVS=$(get_pool_devices ${TESTPOOL} ${DEV_RDSKDIR})
+[[ -n $DEVS ]] && set -A DISK $DEVS
+
+log_must zpool offline $TESTPOOL ${WHOLE_DISK}
log_must dd if=/dev/urandom of=$TESTDIR/testfile bs=1K count=2
log_must zpool export $TESTPOOL
+
log_must dd if=$DEV_RDSKDIR/${DISK[0]} of=$DEV_RDSKDIR/${DISK[1]} bs=1K count=256 conv=notrunc
-ubs=$(zdb -lu $DEV_RDSKDIR/${DISK[1]} | grep -e LABEL -e Uberblock -e 'labels = ')
+ubs=$(zdb -lu ${DISK[1]} | grep -e LABEL -e Uberblock -e 'labels = ')
log_note "vdev 1: ubs $ubs"
-ub_dump_counts=$(zdb -lu $DEV_RDSKDIR/${DISK[1]} | \
+ub_dump_counts=$(zdb -lu ${DISK[1]} | \
awk ' /LABEL/ {label=$NF; blocks[label]=0};
/Uberblock/ {blocks[label]++};
END {print blocks[0],blocks[1],blocks[2],blocks[3]}')
diff --git a/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_005_pos.ksh b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_005_pos.ksh
index 60bbb5615..f6730f611 100755
--- a/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_005_pos.ksh
+++ b/tests/zfs-tests/tests/functional/cli_root/zdb/zdb_005_pos.ksh
@@ -44,6 +44,10 @@ verify_disk_count "$DISKS" 2
set -A DISK $DISKS
default_mirror_setup_noexit $DISKS
+DEVS=$(get_pool_devices ${TESTPOOL} ${DEV_RDSKDIR})
+log_note "$DEVS"
+[[ -n $DEVS ]] && set -A DISK $DEVS
+
log_must dd if=/dev/zero of=$DEV_RDSKDIR/${DISK[1]} bs=1K count=256 conv=notrunc
log_must truncate -s 0 $TEMPFILE
diff --git a/tests/zfs-tests/tests/functional/inuse/inuse_005_pos.ksh b/tests/zfs-tests/tests/functional/inuse/inuse_005_pos.ksh
index 9dd1e25bf..6b0abf429 100755
--- a/tests/zfs-tests/tests/functional/inuse/inuse_005_pos.ksh
+++ b/tests/zfs-tests/tests/functional/inuse/inuse_005_pos.ksh
@@ -82,19 +82,18 @@ typeset -i i=0
unset NOINUSE_CHECK
while (( i < ${#vdevs[*]} )); do
-
for num in 0 1 2 3 ; do
eval typeset disk=\${FS_DISK$num}
zero_partitions $disk
done
-
typeset cyl=""
for num in 0 1 2 3 ; do
eval typeset slice=\${FS_SIDE$num}
disk=${slice%${SLICE_PREFIX}*}
- slice=${slice##*${SLICE_PREFIX}}
+ [[ -z $SLICE_PREFIX ]] && eval typeset disk=\${FS_DISK$num}
+ slice=$(echo $slice | awk '{ print substr($1,length($1),1) }')
log_must set_partition $slice "$cyl" $FS_SIZE $disk
- cyl=$(get_endslice $disk $slice)
+ [[ $num < 3 ]] && cyl=$(get_endslice $disk $slice)
done
if [[ -n $SINGLE_DISK && -n ${vdevs[i]} ]]; then
@@ -115,7 +114,6 @@ while (( i < ${#vdevs[*]} )); do
(( i = i + 1 ))
continue
fi
-
create_pool $TESTPOOL1 ${vdevs[i]} $vdisks spare $sdisks
verify_assertion "$rawtargets"
destroy_pool $TESTPOOL1
diff --git a/tests/zfs-tests/tests/functional/inuse/inuse_008_pos.ksh b/tests/zfs-tests/tests/functional/inuse/inuse_008_pos.ksh
index 95d39d958..ddc8fa7a4 100755
--- a/tests/zfs-tests/tests/functional/inuse/inuse_008_pos.ksh
+++ b/tests/zfs-tests/tests/functional/inuse/inuse_008_pos.ksh
@@ -92,9 +92,10 @@ done
for num in 0 1 2 3 ; do
eval typeset slice=\${FS_SIDE$num}
disk=${slice%${SLICE_PREFIX}*}
- slice=${slice##*${SLICE_PREFIX}}
+ [[ -z $SLICE_PREFIX ]] && eval typeset disk=\${FS_DISK$num}
+ slice=$(echo $slice | awk '{ print substr($1,length($1),1) }')
log_must set_partition $slice "$cyl" $FS_SIZE $disk
- cyl=$(get_endslice $disk $slice)
+ [[ $num < 3 ]] && cyl=$(get_endslice $disk $slice)
done
while (( i < ${#vdevs[*]} )); do
diff --git a/tests/zfs-tests/tests/functional/inuse/inuse_009_pos.ksh b/tests/zfs-tests/tests/functional/inuse/inuse_009_pos.ksh
index 6a9b9623c..a5e9fda59 100755
--- a/tests/zfs-tests/tests/functional/inuse/inuse_009_pos.ksh
+++ b/tests/zfs-tests/tests/functional/inuse/inuse_009_pos.ksh
@@ -82,13 +82,19 @@ typeset -i i=0
while (( i < ${#vdevs[*]} )); do
+ for num in 0 1 2 3 ; do
+ eval typeset disk=\${FS_DISK$num}
+ zero_partitions $disk
+ done
+
typeset cyl=""
for num in 0 1 2 3 ; do
eval typeset slice=\${FS_SIDE$num}
disk=${slice%${SLICE_PREFIX}*}
- slice=${slice##*${SLICE_PREFIX}}
+ [[ -z $SLICE_PREFIX ]] && eval typeset disk=\${FS_DISK$num}
+ slice=$(echo $slice | awk '{ print substr($1,length($1),1) }')
log_must set_partition $slice "$cyl" $FS_SIZE $disk
- cyl=$(get_endslice $disk $slice)
+ [[ $num < 3 ]] && cyl=$(get_endslice $disk $slice)
done
if [[ -n $SINGLE_DISK && -n ${vdevs[i]} ]]; then
--
2.14.2

39
zfs-patches/0052-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch Normal file
View File

@ -0,0 +1,39 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: LOLi <loli10K@users.noreply.github.com>
Date: Fri, 9 Mar 2018 22:52:37 +0100
Subject: [PATCH] Fix zfs-kmod builds when using rpm >= 4.14
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
With rpm-software-management/rpm@5e94633 a package version containing
invalid characters (most commonly a double '-') causes the kmod package
generation to terminate with an error. This change takes advantage of
the newly introduced rpm macro "_wrong_version_format_terminate_build"
to allow kmod packages to be built.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: loli10K <ezomori.nozomu@gmail.com>
Closes #7284
(cherry picked from commit dc0176eeec9719827610fc68742ea3c76796a81f)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
config/zfs-build.m4 | 2 ++
1 file changed, 2 insertions(+)
diff --git a/config/zfs-build.m4 b/config/zfs-build.m4
index 7c19cecd6..8fc3863f4 100644
--- a/config/zfs-build.m4
+++ b/config/zfs-build.m4
@@ -162,6 +162,8 @@ AC_DEFUN([ZFS_AC_RPM], [
RPM_DEFINE_COMMON='--define "$(DEBUG_ZFS) 1"'
RPM_DEFINE_UTIL='--define "_dracutdir $(dracutdir)" --define "_udevdir $(udevdir)" --define "_udevruledir $(udevruledir)" --define "_initconfdir $(DEFAULT_INITCONF_DIR)" $(DEFINE_INITRAMFS) $(DEFINE_SYSTEMD)'
RPM_DEFINE_KMOD='--define "kernels $(LINUX_VERSION)" --define "require_spldir $(SPL)" --define "require_splobj $(SPL_OBJ)" --define "ksrc $(LINUX)" --define "kobj $(LINUX_OBJ)"'
+ RPM_DEFINE_KMOD+=' --define "_wrong_version_format_terminate_build 0"'
+
RPM_DEFINE_DKMS=
dnl # Override default lib directory on Debian/Ubuntu systems. The provided
--
2.14.2

60
zfs-patches/0053-Handle-zio_resume-and-mmp-off.patch Normal file
View File

@ -0,0 +1,60 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Olaf Faaland <faaland1@llnl.gov>
Date: Thu, 8 Mar 2018 15:21:54 -0800
Subject: [PATCH] Handle zio_resume and mmp => off
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
When multihost is disabled on a pool, and the pool is resumed via zpool
clear, within a single cycle of the mmp thread's loop (e.g. while it's
in the cv_timedwait call), both mmp_last_write and mmp_delay should be
updated.
The original code mistakenly treated the two cases as if they could not
occur at the same time.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Olaf Faaland <faaland1@llnl.gov>
Closes #7286
(cherry picked from commit 267fd7b0f14251026c35d7ceab4fbbe2f06717e6)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
module/zfs/mmp.c | 14 ++++++++++----
1 file changed, 10 insertions(+), 4 deletions(-)
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index 634e48287..15eea41ab 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -409,16 +409,22 @@ mmp_thread(spa_t *spa)
}
/*
- * When MMP goes off => on, or spa goes suspended =>
- * !suspended, we know no writes occurred recently. We
- * update mmp_last_write to give us some time to try.
+ * MMP off => on, or suspended => !suspended:
+ * No writes occurred recently. Update mmp_last_write to give
+ * us some time to try.
*/
if ((!last_spa_multihost && multihost) ||
(last_spa_suspended && !suspended)) {
mutex_enter(&mmp->mmp_io_lock);
mmp->mmp_last_write = gethrtime();
mutex_exit(&mmp->mmp_io_lock);
- } else if (last_spa_multihost && !multihost) {
+ }
+
+ /*
+ * MMP on => off:
+ * mmp_delay == 0 tells importing node to skip activity check.
+ */
+ if (last_spa_multihost && !multihost) {
mutex_enter(&mmp->mmp_io_lock);
mmp->mmp_delay = 0;
mutex_exit(&mmp->mmp_io_lock);
--
2.14.2

72
zfs-patches/0054-Fix-MMP-write-frequency-for-large-pools.patch Normal file
View File

@ -0,0 +1,72 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Mon, 12 Mar 2018 11:26:05 -0700
Subject: [PATCH] Fix MMP write frequency for large pools
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
When a single pool contains more vdevs than the CONFIG_HZ for
for the kernel the mmp thread will not delay properly. Switch
to using cv_timedwait_sig_hires() to handle higher resolution
delays.
This issue was reported on Arch Linux where HZ defaults to only
100 and this could be fairly easily reproduced with a reasonably
large pool. Most distribution kernels set CONFIG_HZ=250 or
CONFIG_HZ=1000 and thus are unlikely to be impacted.
Reviewed-by: George Melikov <mail@gmelikov.ru>
Reviewed-by: Giuseppe Di Natale <dinatale2@llnl.gov>
Reviewed-by: Olaf Faaland <faaland1@llnl.gov>
Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #7205
Closes #7289
(cherry picked from commit c30e716c8149b7df7cf968373c22aa00b48d17b8)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
lib/libspl/include/sys/time.h | 6 +++++-
module/zfs/mmp.c | 6 +++---
2 files changed, 8 insertions(+), 4 deletions(-)
diff --git a/lib/libspl/include/sys/time.h b/lib/libspl/include/sys/time.h
index f05fcaa1c..dc645fa5c 100644
--- a/lib/libspl/include/sys/time.h
+++ b/lib/libspl/include/sys/time.h
@@ -54,10 +54,14 @@
#define MSEC2NSEC(m) ((hrtime_t)(m) * (NANOSEC / MILLISEC))
#endif
-#ifndef NSEC2MSEC
+#ifndef NSEC2MSEC
#define NSEC2MSEC(n) ((n) / (NANOSEC / MILLISEC))
#endif
+#ifndef USEC2NSEC
+#define USEC2NSEC(m) ((hrtime_t)(m) * (NANOSEC / MICROSEC))
+#endif
+
#ifndef NSEC2SEC
#define NSEC2SEC(n) ((n) / (NANOSEC / SEC))
#endif
diff --git a/module/zfs/mmp.c b/module/zfs/mmp.c
index 15eea41ab..dc5c6a74f 100644
--- a/module/zfs/mmp.c
+++ b/module/zfs/mmp.c
@@ -462,9 +462,9 @@ mmp_thread(spa_t *spa)
mmp_write_uberblock(spa);
CALLB_CPR_SAFE_BEGIN(&cpr);
- (void) cv_timedwait_sig(&mmp->mmp_thread_cv,
- &mmp->mmp_thread_lock, ddi_get_lbolt() +
- ((next_time - gethrtime()) / (NANOSEC / hz)));
+ (void) cv_timedwait_sig_hires(&mmp->mmp_thread_cv,
+ &mmp->mmp_thread_lock, next_time, USEC2NSEC(1),
+ CALLOUT_FLAG_ABSOLUTE);
CALLB_CPR_SAFE_END(&cpr, &mmp->mmp_thread_lock);
}
--
2.14.2

63
zfs-patches/0055-Tag-zfs-0.7.7.patch Normal file
View File

@ -0,0 +1,63 @@
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Tony Hutter <hutter2@llnl.gov>
Date: Wed, 14 Mar 2018 16:16:43 -0700
Subject: [PATCH] Tag zfs-0.7.7
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
META file and changelog updated.
Signed-off-by: Tony Hutter <hutter2@llnl.gov>
(cherry picked from commit 240ccfc13adb46faf3d6e5e17e25a9769dd06733)
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
---
rpm/generic/zfs-kmod.spec.in | 3 +++
rpm/generic/zfs.spec.in | 3 +++
META | 2 +-
3 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/rpm/generic/zfs-kmod.spec.in b/rpm/generic/zfs-kmod.spec.in
index cb23b0a8f..bf52fa2c4 100644
--- a/rpm/generic/zfs-kmod.spec.in
+++ b/rpm/generic/zfs-kmod.spec.in
@@ -191,6 +191,9 @@ chmod u+x ${RPM_BUILD_ROOT}%{kmodinstdir_prefix}/*/extra/*/*/*
rm -rf $RPM_BUILD_ROOT
%changelog
+* Wed Mar 14 2018 Tony Hutter <hutter2@llnl.gov> - 0.7.7-1
+- Released 0.7.7-1, detailed release notes are available at:
+- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.7
* Thu Feb 01 2018 Tony Hutter <hutter2@llnl.gov> - 0.7.6-1
- Released 0.7.6-1, detailed release notes are available at:
- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.6
diff --git a/rpm/generic/zfs.spec.in b/rpm/generic/zfs.spec.in
index ce028794c..7479e987e 100644
--- a/rpm/generic/zfs.spec.in
+++ b/rpm/generic/zfs.spec.in
@@ -352,6 +352,9 @@ systemctl --system daemon-reload >/dev/null || true
%endif
%changelog
+* Wed Mar 14 2018 Tony Hutter <hutter2@llnl.gov> - 0.7.7-1
+- Released 0.7.7-1, detailed release notes are available at:
+- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.7
* Thu Feb 01 2018 Tony Hutter <hutter2@llnl.gov> - 0.7.6-1
- Released 0.7.6-1, detailed release notes are available at:
- https://github.com/zfsonlinux/zfs/releases/tag/zfs-0.7.6
diff --git a/META b/META
index 1f4686df7..eefafb10d 100644
--- a/META
+++ b/META
@@ -1,7 +1,7 @@
Meta: 1
Name: zfs
Branch: 1.0
-Version: 0.7.6
+Version: 0.7.7
Release: 1
Release-Tags: relext
License: CDDL
--
2.14.2

55
zfs-patches/series
View File

@ -1,4 +1,55 @@
0001-remove-DKMS-modules-and-dracut-build.patch 0001-remove-DKMS-modules-and-dracut-build.patch
0002-import-with-d-dev-disk-by-id-in-scan-service.patch 0002-import-with-d-dev-disk-by-id-in-scan-service.patch
0003-Use-user-namespaces-for-FSETID-policy-check.patch 0003-always-load-ZFS-module-on-boot.patch
0004-always-load-ZFS-module-on-boot.patch 0004-Make-enable-debug-fail-when-given-bogus-args.patch
0005-Implement-enable-debuginfo-to-force-debuginfo.patch
0006-Add-configure-option-to-enable-gcov-analysis.patch
0007-Make-fno-inline-compile-option-more-accessible.patch
0008-Add-support-for-enable-code-coverage-option.patch
0009-Update-codecov.yml.patch
0010-Fix-enable-code-coverage-debug-build.patch
0011-Remove-vn_rename-and-vn_remove-dependency.patch
0012-Fix-with-systemd-on-Debian-based-distributions-6963.patch
0013-ZTS-Fix-create-o_ashift-test-case.patch
0014-Emit-an-error-message-before-MMP-suspends-pool.patch
0015-Fix-systemd_-RPM-macros-usage-on-Debian-based-distri.patch
0016-Bug-fix-in-qat_compress.c-for-vmalloc-addr-check.patch
0017-Fix-default-libdir-for-Debian-Ubuntu.patch
0018-Remove-deprecated-zfs_arc_p_aggressive_disable.patch
0019-OpenZFS-8966-Source-file-zfs_acl.c-function-zfs_acls.patch
0020-Linux-4.16-compat-inode_set_iversion.patch
0021-Linux-4.11-compat-avoid-refcount_t-name-conflict.patch
0022-Fix-zdb-c-traverse-stop-on-damaged-objset-root.patch
0023-Fix-zle_decompress-out-of-bound-access.patch
0024-Fix-racy-assignment-of-zcb.zcb_haderrors.patch
0025-Fix-zdb-R-decompression.patch
0026-Fix-zdb-E-segfault.patch
0027-Fix-zdb-ed-on-objset-for-exported-pool.patch
0028-Handle-zap_add-failures-in-mixed-case-mode.patch
0029-mmp-should-use-a-fixed-tag-for-spa_config-locks.patch
0030-contrib-initramfs-add-missing-conf.d-zfs.patch
0031-zfs-receive-fails-with-dataset-is-busy.patch
0032-OpenZFS-8857-zio_remove_child-panic-due-to-already-d.patch
0033-Clarify-zinject-8-explanation-of-e.patch
0034-Fix-config-issues-frame-size-and-headers.patch
0035-Correct-count_uberblocks-in-mmp.kshlib.patch
0036-Add-SMART-attributes-for-SSD-and-NVMe.patch
0037-Allow-modprobe-to-fail-when-called-within-systemd.patch
0038-Linux-4.16-compat-use-correct-_dec_and_test.patch
0039-Do-not-initiate-MMP-writes-while-pool-is-suspended.patch
0040-Report-duration-and-error-in-mmp_history-entries.patch
0041-Fix-free-memory-calculation-on-v3.14.patch
0042-Add-scrub-after-resilver-zed-script.patch
0043-Add-SMART-self-test-results-to-zpool-status-c.patch
0044-Fix-zpool-8-list-example-to-match-actual-format.patch
0045-Fix-some-typos.patch
0046-Increment-zil_itx_needcopy_bytes-properly.patch
0047-Change-checksum-IO-delay-ratelimit-values.patch
0048-Linux-4.16-compat-get_disk_and_module.patch
0049-Detect-long-config-lock-acquisition-in-mmp.patch
0050-Take-user-namespaces-into-account-in-policy-checks.patch
0051-zdb-and-inuse-tests-don-t-pass-with-real-disks.patch
0052-Fix-zfs-kmod-builds-when-using-rpm-4.14.patch
0053-Handle-zio_resume-and-mmp-off.patch
0054-Fix-MMP-write-frequency-for-large-pools.patch
0055-Tag-zfs-0.7.7.patch