Blame SPECS/kmod-kvdo.spec

0e75ee
%global commit                  1ea1bde8da0b27d675fc34ceef35d375434055b6
0e75ee
%global gittag                  6.2.6.3
bd4e69
%global shortcommit             %(c=%{commit}; echo ${c:0:7})
f1426a
%define spec_release            83
bd4e69
bd4e69
%define kmod_name		kvdo
bd4e69
%define kmod_driver_version	%{gittag}
613e0b
%define kmod_rpm_release	%{spec_release}
f1426a
%define kmod_kernel_version	4.18.0-363.el8
bd4e69
%define kmod_headers_version	%(rpm -qa kernel-devel | sed 's/^kernel-devel-//')
bd4e69
%define kmod_kbuild_dir		.
bd4e69
%define kmod_dependencies       %{nil}
bd4e69
%define kmod_build_dependencies	%{nil}
bd4e69
%define kmod_devel_package	0
bd4e69
bd4e69
Source0:	https://github.com/dm-vdo/%{kmod_name}/archive/%{commit}/%{kmod_name}-%{shortcommit}.tar.gz
bd4e69
bd4e69
%define findpat %( echo "%""P" )
bd4e69
bd4e69
Name:		kmod-kvdo
bd4e69
Version:	%{kmod_driver_version}
bd4e69
Release:	%{kmod_rpm_release}%{?dist}
bd4e69
Summary:	Kernel Modules for Virtual Data Optimizer
bd4e69
License:	GPLv2+
bd4e69
URL:		http://github.com/dm-vdo/kvdo
bd4e69
BuildRoot:	%(mktemp -ud %{_tmppath}/%{name}-%{version}-%{release}-XXXXXX)
bd4e69
BuildRequires:  elfutils-libelf-devel
bd4e69
BuildRequires:	glibc
bd4e69
BuildRequires:	kernel-devel >= %{kmod_kernel_version}
bd4e69
bd4e69
# Disable the kernel-debug requirement for now
bd4e69
%if 0
bd4e69
# kernel-debug appears to not be necessary at the moment. 
bd4e69
BuildRequires:  kernel-debug >= %{kmod_kernel_version}
bd4e69
%endif
bd4e69
bd4e69
BuildRequires:  libuuid-devel
bd4e69
BuildRequires:  redhat-rpm-config
bd4e69
ExcludeArch:    i686
bd4e69
ExcludeArch:    ppc
bd4e69
ExcludeArch:    ppc64
bd4e69
ExcludeArch:    s390
bd4e69
bd4e69
%global kernel_source() /usr/src/kernels/%{kmod_headers_version}
bd4e69
bd4e69
%global _use_internal_dependency_generator 0
e707d8
Provides:         kernel-modules = %{kmod_kernel_version}.%{_target_cpu}
e707d8
Provides:         kmod-%{kmod_name} = %{?epoch:%{epoch}:}%{version}-%{release}
e707d8
Requires(post):   %{_sbindir}/weak-modules
e707d8
Requires(postun): %{_sbindir}/weak-modules
f1426a
Requires:         kernel-core-uname-r    >= %{kmod_kernel_version}
f1426a
Requires:         kernel-modules-uname-r >= %{kmod_kernel_version}
bd4e69
bd4e69
%if "%{kmod_build_dependencies}" != ""
bd4e69
BuildRequires:  %{kmod_build_dependencies}
bd4e69
%endif
bd4e69
%if "%{kmod_dependencies}" != ""
bd4e69
Requires:       %{kmod_dependencies}
bd4e69
%endif
bd4e69
bd4e69
%description
bd4e69
Virtual Data Optimizer (VDO) is a device mapper target that delivers
bd4e69
block-level deduplication, compression, and thin provisioning.
bd4e69
bd4e69
This package provides the kernel modules for VDO.
bd4e69
bd4e69
%pre
bd4e69
# During the install, check whether kvdo or uds is loaded.  A warning here
bd4e69
# indicates that a previous install was not completely removed.  This message
bd4e69
# is purely informational to the user.
bd4e69
for module in kvdo uds; do
bd4e69
  if grep -q "^${module}" /proc/modules; then
bd4e69
    if [ "${module}" == "kvdo" ]; then
bd4e69
      echo "WARNING: Found ${module} module previously loaded (Version: $(cat /sys/kvdo/version 2>/dev/null || echo Unknown)).  A reboot is recommended before attempting to use the newly installed module."
bd4e69
    else
bd4e69
      echo "WARNING: Found ${module} module previously loaded.  A reboot is recommended before attempting to use the newly installed module."
bd4e69
    fi
bd4e69
  fi
bd4e69
done
bd4e69
bd4e69
%post
bd4e69
modules=( $(find /lib/modules/%{kmod_headers_version}/extra/kmod-%{kmod_name} | grep '\.ko$') )
bd4e69
printf '%s\n' "${modules[@]}" >> /var/lib/rpm-kmod-posttrans-weak-modules-add
bd4e69
bd4e69
%pretrans -p <lua>
bd4e69
posix.unlink("/var/lib/rpm-kmod-posttrans-weak-modules-add")
bd4e69
bd4e69
%posttrans
bd4e69
if [ -f "/var/lib/rpm-kmod-posttrans-weak-modules-add" ]; then
bd4e69
	modules=( $(cat /var/lib/rpm-kmod-posttrans-weak-modules-add) )
bd4e69
	rm -rf /var/lib/rpm-kmod-posttrans-weak-modules-add
bd4e69
	printf '%s\n' "${modules[@]}" | %{_sbindir}/weak-modules --dracut=/usr/bin/dracut --add-modules
bd4e69
fi
bd4e69
bd4e69
%preun
bd4e69
rpm -ql kmod-kvdo-%{kmod_driver_version}-%{kmod_rpm_release}%{?dist}.$(arch) | grep '\.ko$' > /var/run/rpm-kmod-%{kmod_name}-modules
bd4e69
# Check whether kvdo or uds is loaded, and if so attempt to remove it.  A
bd4e69
# failure to unload means there is still something using the module.  To make
bd4e69
# sure the user is aware, we print a warning with recommended instructions.
bd4e69
for module in kvdo uds; do
bd4e69
  if grep -q "^${module}" /proc/modules; then
bd4e69
    warnMessage="WARNING: ${module} in use.  Changes will take effect after a reboot."
bd4e69
    modprobe -r ${module} 2>/dev/null || echo ${warnMessage} && /usr/bin/true
bd4e69
  fi
bd4e69
done
bd4e69
bd4e69
%postun
bd4e69
modules=( $(cat /var/run/rpm-kmod-%{kmod_name}-modules) )
bd4e69
rm /var/run/rpm-kmod-%{kmod_name}-modules
bd4e69
printf '%s\n' "${modules[@]}" | %{_sbindir}/weak-modules --dracut=/usr/bin/dracut --remove-modules
bd4e69
bd4e69
%files
bd4e69
%defattr(644,root,root,755)
bd4e69
/lib/modules/%{kmod_headers_version}
bd4e69
/etc/depmod.d/%{kmod_name}.conf
bd4e69
/usr/share/doc/kmod-%{kmod_name}/greylist.txt
bd4e69
bd4e69
%prep
bd4e69
%setup -n %{kmod_name}-%{commit}
bd4e69
%{nil}
bd4e69
set -- *
bd4e69
mkdir source
bd4e69
mv "$@" source/
bd4e69
mkdir obj
bd4e69
bd4e69
%build
bd4e69
rm -rf obj
bd4e69
cp -r source obj
bd4e69
make -C %{kernel_source} M=$PWD/obj/%{kmod_kbuild_dir} V=1 \
bd4e69
	NOSTDINC_FLAGS="-I $PWD/obj/include -I $PWD/obj/include/uapi"
bd4e69
# mark modules executable so that strip-to-file can strip them
bd4e69
find obj/%{kmod_kbuild_dir} -name "*.ko" -type f -exec chmod u+x '{}' +
bd4e69
bd4e69
whitelist="/lib/modules/kabi-current/kabi_whitelist_%{_target_cpu}"
bd4e69
bd4e69
for modules in $( find obj/%{kmod_kbuild_dir} -name "*.ko" -type f -printf "%{findpat}\n" | sed 's|\.ko$||' | sort -u ) ; do
bd4e69
	# update depmod.conf
bd4e69
	module_weak_path=$(echo $modules | sed 's/[\/]*[^\/]*$//')
bd4e69
	if [ -z "$module_weak_path" ]; then
bd4e69
		module_weak_path=%{name}
bd4e69
	else
bd4e69
		module_weak_path=%{name}/$module_weak_path
bd4e69
	fi
bd4e69
	echo "override $(echo $modules | sed 's/.*\///') $(echo %{kmod_headers_version} | sed 's/\.[^\.]*$//').* weak-updates/$module_weak_path" >> source/depmod.conf
bd4e69
bd4e69
	# update greylist
bd4e69
	nm -u obj/%{kmod_kbuild_dir}/$modules.ko | sed 's/.*U //' |  sed 's/^\.//' | sort -u | while read -r symbol; do
bd4e69
		grep -q "^\s*$symbol\$" $whitelist || echo "$symbol" >> source/greylist
bd4e69
	done
bd4e69
done
bd4e69
sort -u source/greylist | uniq > source/greylist.txt
bd4e69
bd4e69
%install
bd4e69
export INSTALL_MOD_PATH=$RPM_BUILD_ROOT
bd4e69
export INSTALL_MOD_DIR=extra/%{name}
bd4e69
make -C %{kernel_source} modules_install V=1 \
bd4e69
	M=$PWD/obj/%{kmod_kbuild_dir}
bd4e69
# Cleanup unnecessary kernel-generated module dependency files.
bd4e69
find $INSTALL_MOD_PATH/lib/modules -iname 'modules.*' -exec rm {} \;
bd4e69
bd4e69
install -m 644 -D source/depmod.conf $RPM_BUILD_ROOT/etc/depmod.d/%{kmod_name}.conf
bd4e69
install -m 644 -D source/greylist.txt $RPM_BUILD_ROOT/usr/share/doc/kmod-%{kmod_name}/greylist.txt
bd4e69
bd4e69
%clean
bd4e69
rm -rf $RPM_BUILD_ROOT
bd4e69
bd4e69
%changelog
f1426a
* Thu Feb 03 2022 - Andy Walsh <awalsh@redhat.com> - 6.2.6.3-83
f1426a
- Adjusted kernel dependencies to grab the right packages.
f1426a
- Resolves: rhbz#2011426
f1426a
- Rebuilt for latest kernel.
f1426a
- Relates: rhbz#2000909
f1426a
0e75ee
* Wed Nov 03 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.6.3-82
0e75ee
- Fixed a bug which prevented the resumption of a suspended read-only vdo.
0e75ee
- Resolves: rhbz#1996893
0e75ee
5dc392
* Mon Oct 11 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.72-81.8_6
5dc392
- Bumped NVR for new 4.18 kernel
5dc392
- Relates: rhbz#2000909
5dc392
449490
* Fri Aug 27 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.72-80
449490
- Bumped NVR for new 4.18 kernel
449490
- Related: rhbz#1939279
449490
1e77a3
* Tue Aug 10 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.72-79
1e77a3
- Fixed a bug which could result in the UDS index issuing I/O while
1e77a3
  suspended.
1e77a3
- Resolves: rhbz#1990180
1e77a3
1e77a3
* Thu Aug 05 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.71-79
1e77a3
- Reduced context switches when a vdo is idle.
1e77a3
- Resolves: rhbz#1886738
1e77a3
52da5e
* Thu Jul 22 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.65-79
52da5e
- Removed unneeded allocations from the previous fixes for rebuilding
52da5e
  a converted index.
52da5e
- Resolves: rhbz#1966824
52da5e
52da5e
* Thu Jul 15 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.62-79
52da5e
- Fixed chapter computation for a converted sparse index.
52da5e
- Resolves: rhbz#1965516
52da5e
- Fixed invalidation of converted chapters.
52da5e
- Resolves: rhbz#1966818
52da5e
- Removed extraneous fields from the super block of a converted index.
52da5e
- Resolves: rhbz#1965546
52da5e
- Fixed calculation of the number of expiring chapters in a converted
52da5e
  index.
52da5e
- Resolves: rhbz#1975546
52da5e
- Fixed bugs rebuilding a converted index.
52da5e
- Resolves: rhbz#1966824
52da5e
76af58
* Mon Jun 21 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.41-79
76af58
- Bumped NVR for new 4.18 kernel
76af58
- Relates: rhbz#1939279
76af58
adc15a
* Tue Jun 01 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.41-78
adc15a
- Fixed bugs in reading the UDS index of a VDO volume which was converted
adc15a
  to LVM.
adc15a
- Resolves: rhbz#1928284
adc15a
adc15a
* Thu May 20 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.21-78
adc15a
- Added support for VDO volumes which have been converted to LVM.
adc15a
- Related: rhbz#1928284
adc15a
613e0b
* Thu May 13 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.5.11-78
613e0b
- Introduced new memory size parameter values for UDS indexes which have
613e0b
  been converted from vdo script management to LVM.
613e0b
- Resolves: rhbz#1928284
613e0b
290ac9
* Tue Mar 16 2021 - Andy Walsh <awalsh@redhat.com> - 6.2.4.26-77.8_5
39bed5
- Bumped NVR for new 4.18 kernel
290ac9
- Relates: rhbz#1939279
39bed5
d4b601
* Sat Nov 28 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.4.26-76
d4b601
- Bumped NVR for new 4.18 kernel
d4b601
- Relates: rhbz#1894978
d4b601
e23b8d
* Thu Nov 05 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.4.26-75
e23b8d
- Bumped NVR for new 4.18 kernel
e23b8d
- Relates: rhbz#1894978
e23b8d
e23b8d
* Mon Nov 02 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.4.26-74
e23b8d
- Modified physical growth to validate the new VDO size against the size of
e23b8d
  the underlying storage.
e23b8d
- Resolves: rhbz#1732922
e23b8d
- Fixed issues which prevented lvrename from working on lvm managed
e23b8d
  VDO devices.
e23b8d
- Resolves: rhbz#1888419
e23b8d
8f8228
* Thu Oct 01 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.4.14-74
8f8228
- Fixed a bug which causes the UDS index to consume an excessive
8f8228
  number of CPU cycles when the VDO device is idle.
8f8228
- Resolves: rhbz#1870660
c5d37b
e707d8
* Thu Jul 16 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.3.107-74
e707d8
- Updated dependencies to prevent installing kernel-rt incorrectly.
e707d8
- Resolves: rhbz#1811923
e707d8
- Bumped requirement for new 4.18 kernel
e707d8
- Relates: rhbz#1812069
e707d8
5910ff
* Fri Jun 19 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.3.107-73
5910ff
- Fixed a bug which can cause a soft lockup if users interrupt processes
5910ff
  waiting on dm-setup commands.
5910ff
- Resolves: rhbz#1844651
5910ff
- Fixed a rare race which could cause a suspend of a VDO device to fail.
5910ff
- Resolves: rhbz#1847747
5910ff
4eb647
* Tue Jun 02 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.3.91-73
4eb647
- Removed unused UDS bio statistics.
4eb647
- Resolves: rhbz#1827762
4eb647
- Removed support for old kernels.
4eb647
- Resolves: rhbz#1827762
4eb647
- Fixed Coverity errors.
4eb647
- Resolves: rhbz#1827763
4eb647
- Improved the error message when slab journal locks overflow.
4eb647
- Resolves: rhbz#1827761
4eb647
- Fixed a bug which could result in VDO issuing I/O while suspended.
4eb647
- Resolves: rhbz#1824789
4eb647
- Fixed a rare double-enqueue bug in the recovery journal.
4eb647
- Resolves: rhbz#1824802
4eb647
- Modified VDO to not allocate an index if the VDO device is started
4eb647
  with deduplication disabled.
4eb647
- Resolves: rhbz#1755448
4eb647
- Nodified VDO to not log spurious journal lock warnings when cleaning up
4eb647
  write requests which failed due to the VDO going read-only.
4eb647
- Resolves: rhbz#1840455
4eb647
8fcd6f
* Mon Apr 27 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117_8.3-73
8fcd6f
- Bumped NVR for new 4.18 kernel
8fcd6f
- Relates: rhbz#1812069
8fcd6f
8fcd6f
* Mon Apr 27 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117_8.3-72
8fcd6f
- Bumped NVR for new 4.18 kernel
8fcd6f
- Relates: rhbz#1812069
8fcd6f
8fcd6f
* Fri Apr 17 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117_8.3-71
8fcd6f
- Bumped NVR for new 4.18 kernel
8fcd6f
- Relates: rhbz#1812069
8fcd6f
480291
* Fri Apr 17 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117_8.3-70
480291
- Bumped NVR for new 4.18 kernel
480291
- Relates: rhbz#1812069
480291
480291
* Wed Mar 25 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117_8.3-69
480291
- Bumped NVR for new 4.18 kernel
480291
- Relates: rhbz#1812069
480291
480291
* Wed Mar 11 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117-68
480291
- Bumped NVR for new 4.18 kernel
480291
- Relates: rhbz#1812069
480291
480291
* Tue Mar 10 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117-67
480291
- Bumped NVR for new 4.18 kernel
480291
- Relates: rhbz#1812069
480291
480291
* Tue Mar 10 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117-66
480291
- Bumped NVR for new 4.18 kernel
480291
- Relates: rhbz#1812069
480291
fac727
* Sat Mar 07 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117-65
fac727
- Bumped NVR for new 4.18 kernel
fac727
- Relates: rhbz#1764816
fac727
8be025
* Tue Feb 11 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.117-63
8be025
- Fixed a bug which would cause the UDS index to be perceived as corrupt
8be025
  when a VDO volume is moved to a system with a different endianness.
8be025
- Resolves: rhbz#1745211
8be025
- Modified UDS index rebuild to be interruptible so that shutting down a
8be025
  VDO whose index is rebuilding need not wait for the index rebuild to
8be025
  complete.
8be025
- Resolves: rhbz#1737639
8be025
- Prevented two VDO devices from being started on the same storage.
8be025
- Resolves: rhbz#1725052
8be025
- Fixed rare races which could result in VDO issuing I/O while suspended.
8be025
- Resolves: rhbz#1766358
8be025
- Fixed crashes when re-suspending a VDO after it had been resumed.
8be025
- Resolves: rhbz#1765253
8be025
- Made async mode ACID. Added async-unsafe mode to preserve the performance of
8be025
  the old implementation.
8be025
- Resolves: rhbz#1657301
8be025
8be025
* Tue Jan 14 2020 - Andy Walsh <awalsh@redhat.com> - 6.2.2.24-63
8be025
- Bumped NVR for new 4.18 kernel
8be025
- Relates: rhbz#1764816
8be025
8be025
* Tue Dec 03 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.2.24-62
8be025
- Bumped NVR for new 4.18 kernel
8be025
- Relates: rhbz#1764816
8be025
8be025
* Tue Nov 26 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.2.24-61
8be025
- Bumped NVR for new 4.18 kernel
8be025
- Relates: rhbz#1764816
8be025
8be025
* Mon Nov 11 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.2.24-60
8be025
- Bumped NVR for new 4.18 kernel
8be025
- Relates: rhbz#1764816
8be025
8be025
* Thu Oct 31 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.2.24-59
8be025
- Bumped NVR for new 4.18 kernel
8be025
- Relates: rhbz#1764816
8be025
8be025
* Wed Oct 30 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.2.24-58
8be025
- Incremented the dm target version to allow lvm to tell whether a VDO
8be025
- Resolves: rhbz#1752893
8be025
8be025
* Thu Oct 24 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.2.18-58
8be025
- Bumped NVR for new 4.18 kernel
8be025
- Relates: rhbz#1764816
8be025
8be025
* Thu Oct 17 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.2.18-57
8be025
- Reduced the default number of index zones as the previous default
8be025
  attempted to maximize index performance at the expense of all other
8be025
  processes.
8be025
- Resolves: rhbz#1703507
8be025
- Fixed an assertion when resuming a VDO device which was not suspended
8be025
  with the no-flush flag.
8be025
- Resolves: rhbz#1752893
8be025
bd4e69
* Fri Sep 13 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.138-57
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1695330
bd4e69
bd4e69
* Thu Aug 08 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.138-56
bd4e69
- Fixed a crash on allocation failure and a use-after-free race introduced by
bd4e69
  the changes to avoid issuing I/O while suspended.
bd4e69
- Resolves: rhbz#1659303
bd4e69
bd4e69
* Fri Aug 02 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.134-56
bd4e69
- Finished converting the VDO device to conform to the DM convention of not
bd4e69
  issuing I/O while suspended.
bd4e69
- Resolves: rhbz#1659303
bd4e69
- Fixed a bug where crash recovery could use the wrong threads for certain
bd4e69
  operations potentially resulting in memory corruption.
bd4e69
- Resolves: rhbz#1703621
bd4e69
- Fixed a bug which could cause segfaults when running the vdostatus
bd4e69
  command.
bd4e69
- Resolves: rhbz#1669960
bd4e69
- Eliminated a backtrace from the error logged when creating a VDO device
bd4e69
  with an erroneous physical size in the table line.
bd4e69
- Resolves: rhbz#1717435
bd4e69
- Fixed a possible use-after-free of the UDSConfiguration.
bd4e69
- Resolves: rhbz#1653802
bd4e69
- Made VDO a singleton device because multi-segment devices containing a
bd4e69
  VDO have a number of issues.
bd4e69
- Resolves: rhbz#1725077
bd4e69
- Converted the VDO device to conform to the DM convention of not issuing
bd4e69
  writes from the constructor.
bd4e69
- Resolves: rhbz#1669086
bd4e69
bd4e69
* Mon Jul 29 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.102-56
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1695330
bd4e69
bd4e69
* Fri Jul 12 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.102-55
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1695330
bd4e69
bd4e69
* Thu Jun 27 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.102-54
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1695330
bd4e69
bd4e69
* Fri Jun 14 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.102-53
bd4e69
- Continued converting the VDO device to conform to the DM convention of
bd4e69
  not issuing I/O while suspended.
bd4e69
  - Resolves: rhbz#1659303
bd4e69
- Added more rate limiting of error logging in both the kvdo and uds
bd4e69
  modules in order to avoid soft-lockups on newer kernels.
bd4e69
  - Resolves: rhbz#1703243
bd4e69
- Eliminated the passing of addresses of unaligned fields in packed
bd4e69
  - Resolves: rhbz#1718058
bd4e69
bd4e69
* Tue May 21 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.48-53
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1695330
bd4e69
bd4e69
* Tue May 21 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.48-52
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1695330
bd4e69
bd4e69
* Sun May 05 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.1.48-51
bd4e69
- Modified the setting of the dm target version for VDO devices to only
bd4e69
  change when the dm interface (i.e. table line) changes instead of tying
bd4e69
  it to the VDO version.
bd4e69
  - Resolves: rhbz#1665298
bd4e69
- Improved error handling when resizing VDO devices.
bd4e69
  - Resolves: rhbz#1659247
bd4e69
- Reduced, removed, and/or rate limited error logging to avoid
bd4e69
  soft-lockups.
bd4e69
  - Resolves: rhbz#1678785
bd4e69
  - Resolves: rhbz#1698664
bd4e69
- Began converting the VDO device to conform to the DM convention of not
bd4e69
  issuing I/O while suspended.
bd4e69
  - Relates: rhbz#1659303
bd4e69
- Added a dmsetup message to close the UDS index of a running VDO device.
bd4e69
  - Relates: rhbz#1643291
bd4e69
bd4e69
* Fri May 03 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-51
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1695330
bd4e69
bd4e69
* Mon Feb 25 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-50
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Feb 15 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-49
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Feb 13 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-48
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Sat Feb 09 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-47
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Feb 08 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-46
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Feb 06 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-45
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Jan 16 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-44
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Mon Jan 14 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-43
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Jan 11 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-42
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Mon Jan 07 2019 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-41
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Dec 19 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-40
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Mon Dec 17 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-39
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Dec 14 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.293-38
bd4e69
- Improved dmsetup error reporting of invalid thread counts.
bd4e69
- rhbz#1645324
bd4e69
- Allowed VDO backing devices to specified by device number.
bd4e69
- Resolves: rhbz#1594285
bd4e69
- Eliminated memory allocations when suspended.
bd4e69
- Resolves: rhbz#1658348
bd4e69
- Improved error handling during suspend.
bd4e69
- Resolves: rhbz#1658348
bd4e69
bd4e69
* Wed Dec 12 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-38
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Mon Dec 10 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-37
bd4e69
- Bumped NVR for driver signing
bd4e69
- Relates: rhbz#1589873
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Thu Nov 29 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-36
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Nov 16 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.273-35
bd4e69
- Fixed more error path memory leaks in the uds and kvdo modules.
bd4e69
- Resolves: rhbz#1639854
bd4e69
- Removed the read cache.
bd4e69
- Resolves: rhbz#1639512
bd4e69
- Fixed a bug which prevented parsing of version 0 table lines.
bd4e69
- Resolves: rhbz#1643639
bd4e69
- In order to properly handle version 0 table lines, made no-op physical
bd4e69
  growth not an error.
bd4e69
- Resolves: rhbz#1643639
bd4e69
- Limited the number of logical zones to 60.
bd4e69
- Resolves: rhbz#1645324
bd4e69
- Converted to use the kernel's bio zeroing method instead of a VDO
bd4e69
  specific one.
bd4e69
- Resolves: rhbz#1647446
bd4e69
- Added a missing call to flush_cache_page() after writing pages which may
bd4e69
  be owned by the page cache or a user as required by the kernel.
bd4e69
- Resolves: rhbz#1647446
bd4e69
- Added a version 2 table line which uses DM-style optional parameters.
bd4e69
- Resolves: rhbz#1648469
bd4e69
- Fixed a bug in the statistics tracking partial I/Os.
bd4e69
- Resolves: rhbz#1648496
bd4e69
- Added a maximum discard size table line parameter and removed the
bd4e69
  corresponding sysfs parameter which applied to all VDO devices.
bd4e69
- Resolves: rhbz#1648469
bd4e69
bd4e69
* Wed Nov 07 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-35
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Oct 24 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-34
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Oct 24 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-33
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Oct 19 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-32
bd4e69
- Depend on more of the NVR for the kernel package.
bd4e69
- Resolves: rhbz#1640699
bd4e69
bd4e69
* Tue Oct 16 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-31
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Tue Oct 16 2018 - Tomas Kopecek <tkopecek@redhat.com> - 6.2.0.239-30
bd4e69
- Bumped NVR for driver signing
bd4e69
- Relates: rhbz#1589873
bd4e69
bd4e69
* Mon Oct 15 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-29
bd4e69
- Bumped NVR for driver signing
bd4e69
- Relates: rhbz#1589873
bd4e69
bd4e69
* Thu Oct 11 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-28
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Oct 10 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-27
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Sun Oct 07 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.239-26
bd4e69
- Fixed error path memory leaks in the uds and kvdo modules.
bd4e69
- Resolves: rhbz#1609403
bd4e69
- Modified the physical and logical growth procedures to be consistent with
bd4e69
  other device mapper targets.
bd4e69
- Resolves: rhbz#1631868
bd4e69
bd4e69
* Fri Sep 28 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-26
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Thu Sep 27 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-25
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Tue Sep 25 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-24
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Tue Sep 25 2018 - Joseph Chapman <jochapma@redhat.com> - 6.2.0.219-23
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Mon Sep 24 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-22
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Sep 21 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-21
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Thu Sep 20 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-20
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Sep 19 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-19
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Tue Sep 18 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-18
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Mon Sep 17 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.219-17
bd4e69
- Fixed error path memory leaks in the uds and kvdo modules.
bd4e69
- Resolves: rhbz#1609403
bd4e69
- Fixed conflict errors when installing RPMs via dnf.
bd4e69
- Resolves: rhbz#1601103
bd4e69
- Fixed a hang when recovering a VDO volume with a physical size larger
bd4e69
  than 16TB.
bd4e69
- Resolves: rhbz#1628316
bd4e69
- Fixed some potential initialization issues in the UDS module.
bd4e69
- Resolves: rhbz#1609403
bd4e69
- Fixed a use-after-free bug in a UDS error path.
bd4e69
- Resolves: rhbz#1609403
bd4e69
- Added missing va_end() calls.
bd4e69
- Resolves: rhbz#1627953
bd4e69
- Modified Makefile to take build flags from rpmbuild.
bd4e69
- Resolves: rhbz#1624184
bd4e69
bd4e69
* Fri Sep 14 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-16
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Thu Sep 13 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-15
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Sep 12 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-14
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Aug 29 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-13
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Wed Aug 29 2018 - Joseph Chapman <jochapma@redhat.com> - 6.2.0.197-12
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Tue Aug 28 2018 - Josh Boyer <jwboyer@redhat.com> - 6.2.0.197-11
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1623006
bd4e69
bd4e69
* Fri Aug 24 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-10
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1611663
bd4e69
bd4e69
* Mon Aug 20 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-9
bd4e69
- Patched a new compiler warning out
bd4e69
- Relates: rhbz#1611663
bd4e69
bd4e69
* Mon Aug 20 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-8
bd4e69
- Bumped NVR for new 4.18 kernel
bd4e69
- Relates: rhbz#1611663
bd4e69
bd4e69
* Mon Aug 13 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-7
bd4e69
- Bumped NVR for 4.18 rebase
bd4e69
- Resolves: rhbz#1534087
bd4e69
bd4e69
* Wed Aug  8 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.197-6
bd4e69
- Enabled the setting of max_discard_sectors for VDO devices via sysfs.
bd4e69
  This allows users stacking dm-thin devices on top of VDO to set a value which
bd4e69
  is large enough that dm-thin will send discards to VDO.
bd4e69
- Resolves: rhbz#1612349
bd4e69
bd4e69
* Sat Jul 28 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.187-6
bd4e69
- No longer attempt to unload modules in %preun
bd4e69
- Resolves: rhbz#1553420
bd4e69
- Fixed a bug in %preun that was attempting to call 'dnf'
bd4e69
- Resolves: rhbz#1598924
bd4e69
- Fixed weak-modules calls to use proper location for dracut
bd4e69
- Resolves: rhbz#1609178
bd4e69
- Fixed a potential deadlock in the UDS index by using the kernel supplied
bd4e69
  struct callback instead of our own implementation of synchronous
bd4e69
  callbacks.
bd4e69
- Eliminated obsolete code and fields from UDS.
bd4e69
- Converted the VDO module to use numeric.h from the UDS module instead of
bd4e69
  having its own version.
bd4e69
- Fixed a bug which would cause incorrect encoding of VDO data structures
bd4e69
  on disk.
bd4e69
- Removed or modified logging which prints pointers from the kernel since
bd4e69
  newer kernels obfuscate the pointer values for security reasons.
bd4e69
- Eliminated confusing and spurious error messages when rebuilding a UDS
bd4e69
  index.
bd4e69
- Improved memory allocation by not using the incorrect __GFP_NORETRY flag
bd4e69
  and by using the memalloc_noio_save mechanism.
bd4e69
- Finished conversion of the encoding and decoding of the VDO's on-disk
bd4e69
  structures to be platform independent.
bd4e69
- Converted VDO to use the atomic API from the UDS module instead of its
bd4e69
  own.
bd4e69
- Fixed memory leaks in UDS error paths.
bd4e69
- Fixed a potential stack overflow when reaping the recovery journal.
bd4e69
bd4e69
* Fri Jul 06 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.132-5
bd4e69
- Rebuilt to work with 4.18 kernel.
bd4e69
bd4e69
* Thu Jun 28 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.132-4
bd4e69
Note: This is a pre-release version, future versions of VDO may not support
bd4e69
VDO devices created with this version.
bd4e69
- Removed obsolete code.
bd4e69
- Continued conversion of the encoding and decoding of the VDO's on-disk
bd4e69
  structures to be platform independent.
bd4e69
- Adopted use of gcc's built-in byte order macros.
bd4e69
- Converted the VDO module to use the platform independent version of the
bd4e69
  Murmur3 hash from the UDS module.
bd4e69
- Improved counting of dedupe timeouts by including in the count queries
bd4e69
  which are not made due to their being a lack of resources from previous
bd4e69
  queries taking too long.
bd4e69
- Improved checking that VDO does not allocate memory from its own threads
bd4e69
  during normal operation.
bd4e69
- Fixed a bug which caused crashes with VDO on top of RAID-50.
bd4e69
- Fixed a bug which caused VDO to ignore most flush requests on kernels
bd4e69
  later than 4.10
bd4e69
- Resolves: rhbz#1594062
bd4e69
bd4e69
* Thu Jun 21 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.109-4
bd4e69
Note: This is a pre-release version, future versions of VDO may not support
bd4e69
VDO devices created with this version.
bd4e69
- Removed obsolete code.
bd4e69
- Made uses of memory barriers and atomics more portable across platforms.
bd4e69
- Converted the encoding and decoding of many of VDO's on-disk structures
bd4e69
  to be platform independent.
bd4e69
- Made the implementation of VDO's in-memory data structures platform
bd4e69
  independent.
bd4e69
- Fixed a logging bug which resulted in single log message being split
bd4e69
  across multiple log lines on newer kernels.
bd4e69
- Fixed a bug which would cause attempts to grow the physical size of a VDO
bd4e69
  device to fail if the device below the VDO was resized while the VDO was
bd4e69
  offline.
bd4e69
- Converted to use GCC's built-in macros for determining endianness.
bd4e69
- Converted some non-performance critical atomics to be spinlock protected
bd4e69
  in order to avoid dealing with memory barrier portability issues.
bd4e69
- Fixed a bug which could cause data loss when discarding unused portions
bd4e69
  of a VDO's logical space.
bd4e69
- Reduced memory usage (slightly) by rearranging structures to pack better
bd4e69
  in memory.
bd4e69
- Modified grow physical to fail in the prepare step if the size isn't
bd4e69
  changing, avoiding a suspend-and-resume cycle.
bd4e69
- Added support for building with a 4.18 kernel.
bd4e69
bd4e69
* Mon Jun 04 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.71-4
bd4e69
Note: This is a pre-release version, future versions of VDO may not support
bd4e69
VDO devices created with this version.
bd4e69
- Updated to compile on aarch64, ppc64le, and s390x processor architectures
bd4e69
  in addition to x86.
bd4e69
- Updated atomics, memory barriers, and other synchronization mechanisms to
bd4e69
  work on aarch64, ppc64le, and s390x processor architectures in addition
bd4e69
  to x86.
bd4e69
- Fixed thread safety issues in the UDS page cache.
bd4e69
- Removed obsolete code and interfaces from the UDS module.
bd4e69
- Added /sys/kvdo/version which contains the currently loaded version of
bd4e69
  the kvdo module.
bd4e69
- Updated the UDS module to consistently generate and encode on-disk data
bd4e69
  regardless of the processor architecture.
bd4e69
- Began Updating the VDO module to consistently encode on-disk data
bd4e69
  regardless of the processor architecture.
bd4e69
- Added logging of normal operation when a VDO device starts normally.
bd4e69
- Fixed a potential use-after-free race when shutting down a VDO device.
bd4e69
- Modified allocations made from VDO index threads to use the correct flags.
bd4e69
- Exported the MurmurHash3 implementation from the UDS module rather than
bd4e69
  having a seperate copy in the VDO module.
bd4e69
- Fixed handling of I/O errors in 4.13 and later kernels.
bd4e69
- Exported functions for handling endian conversions from the UDS module
bd4e69
  for use by the VDO module.
bd4e69
bd4e69
* Tue May 01 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.35-3
bd4e69
- Enabled aarch64 builds
bd4e69
bd4e69
* Fri Apr 27 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.35-2
bd4e69
Note: This is a pre-release version, future versions of VDO may not support
bd4e69
VDO devices created with this version.
bd4e69
- Added validation that the release version numbers in the geometry and
bd4e69
  super block match on load.
bd4e69
- Fixed compilation problems on newer versions of GCC.
bd4e69
bd4e69
* Tue Apr 24 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.32-2
bd4e69
Note: This is a pre-release version, future versions of VDO may not support
bd4e69
VDO devices created with this version.
bd4e69
- Merged the funnel queue implementations in the uds and kvdo modules.
bd4e69
- Improved deduplication of concurrent requests containing the same data.
bd4e69
- Enabled loading of VDO devices created with version 6.0 or 6.1.
bd4e69
- Moved atomic.h from the UDS module to the VDO module since the UDS module
bd4e69
  doesn't use it.
bd4e69
- Removed spurious error messages when first creating the index for a new
bd4e69
  VDO.
bd4e69
- Added validation that the release version numbers in the geometry block
bd4e69
  and VDO super block match.
bd4e69
- Fixed bug in UDS on architectures with page sizes larger than 4K.
bd4e69
- Reflected kernel change of SECTOR_SHIFT and SECTOR_SIZE from enums to
bd4e69
  macros.
bd4e69
- Continued to remove obsolete functionality from the UDS module.
bd4e69
- Continued to add support for architectures other than x86.
bd4e69
- Fixed a thread-safety issue in UDS module's chapter cache.
bd4e69
bd4e69
* Tue Apr 17 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.4-2
bd4e69
- Fixed path to _sbindir for weak-mldules
bd4e69
- Resolves: rhbz#1566144
bd4e69
         
bd4e69
* Fri Apr 13 2018 - Andy Walsh <awalsh@redhat.com> - 6.2.0.4-1
bd4e69
- Updated to use github for Source0
bd4e69
- Removed unused sections
bd4e69
- Initial RHEL8 RPM rhbz#1534087
bd4e69
         
bd4e69
* Fri Apr 13 2018 - J. corwin Coburn <corwin@redhat.com> - 6.2.0.4-1
bd4e69
- Initial pre-release for RHEL 8.
bd4e69
  - Please be aware that this version is not compatible with previous versions
bd4e69
    of VDO. Support for loading or upgrading devices created with VDO version
bd4e69
    6.1 will be available soon.
bd4e69
- Management tools will work with both python 2 and python 3.
bd4e69
- Dedupe path improvements.
bd4e69
- Beginnings of support for non-x86 architectures.
bd4e69
- Removed obsolete code from UDS.
bd4e69
bd4e69
* Tue Feb 27 2018 - Andy Walsh <awalsh@redhat.com> - 6.1.0.153-15
bd4e69
- Fixed preun handling of loaded modules
bd4e69
- Resolves: rhbz#1549178
bd4e69
bd4e69
* Fri Feb 16 2018 - Joseph Chapman <jochapma@redhat.com> - 6.1.0.149-13
bd4e69
- Sync mode is safe if underlying storage changes to requiring flushes
bd4e69
- Resolves: rhbz#1540777
bd4e69
bd4e69
* Wed Feb 07 2018 - Joseph Chapman <jochapma@redhat.com> - 6.1.0.146-13
bd4e69
- Module target is now "vdo" instead of "dedupe"
bd4e69
- Fixed a bug where crash recovery led to failed kernel page request
bd4e69
- Improved modification log messages
bd4e69
- Improved package description and summary fields
bd4e69
- Resolves: rhbz#1535127
bd4e69
- Resolves: rhbz#1535597
bd4e69
- Resolves: rhbz#1540696
bd4e69
- Resolves: rhbz#1541409
bd4e69
bd4e69
* Tue Feb 06 2018 - Andy Walsh <awalsh@redhat.com> - 6.1.0.144-13
bd4e69
- Updated summary and descriptions
bd4e69
- Resolves: rhbz#1541409
bd4e69
bd4e69
* Thu Feb 01 2018 - Joseph Chapman <jochapma@redhat.com> - 6.1.0.130-12
bd4e69
- Fix General Protection Fault unlocking UDS callback mutex
bd4e69
- Removing kmod-kvdo package unloads kernel module
bd4e69
- Fix URL to point to GitHub tree
bd4e69
- Resolves: rhbz#1510176
bd4e69
- Resolves: rhbz#1533260
bd4e69
- Resolves: rhbz#1539061
bd4e69
bd4e69
* Fri Jan 19 2018 - Joseph Chapman <jochapma@redhat.com> - 6.1.0.124-11
bd4e69
- Fixed provisional referencing for dedupe.
bd4e69
- Only log a bio submission from a VDO to itself.
bd4e69
- vdoformat cleans up metadata properly after fail.
bd4e69
- Resolves: rhbz#1511587
bd4e69
- Resolves: rhbz#1520972
bd4e69
- Resolves: rhbz#1532481
bd4e69
bd4e69
* Wed Jan 10 2018 - Joseph Chapman <jochapma@redhat.com> - 6.1.0.114-11
bd4e69
- /sys/uds permissions now resticted to superuser only
bd4e69
- Remove /sys/uds files that should not be used in production
bd4e69
- Removing kvdo module reports version
bd4e69
- VDO automatically chooses the proper write policy by default
bd4e69
- Fixed a Coverity-detected error path leak
bd4e69
- Resolves: rhbz#1525305
bd4e69
- Resolves: rhbz#1527734
bd4e69
- Resolves: rhbz#1527737
bd4e69
- Resolves: rhbz#1527924
bd4e69
- Resolves: rhbz#1528399
bd4e69
bd4e69
* Thu Dec 21 2017 - Joseph Chapman <jochapma@redhat.com> - 6.1.0.106-11
bd4e69
- Detect journal overflow after 160E of writes
bd4e69
- Clean up UDS threads when removing last VDO
bd4e69
- Resolves: rhbz#1512968
bd4e69
- Resolves: rhbz#1523240
bd4e69
bd4e69
* Tue Dec 12 2017 Joe Chapman <jochapma@redhat.com> 6.1.0.97-11
bd4e69
- Default logical size is no longer over-provisioned
bd4e69
- Remove debug logging when verifying dedupe advice
bd4e69
- Resolves: rhbz#1519330
bd4e69
bd4e69
* Fri Dec 08 2017 Joe Chapman <jochapma@redhat.com> 6.1.0.89-11
bd4e69
- improve metadata cleanup after vdoformat failure
bd4e69
- log REQ_FLUSH & REQ_FUA at level INFO
bd4e69
- improve performance of cuncurrent write requests with the same data
bd4e69
- Resolves: rhbz#1520972
bd4e69
- Resolves: rhbz#1521200
bd4e69
bd4e69
* Fri Dec 01 2017 Joe Chapman <jochapma@redhat.com> 6.1.0.72-10
bd4e69
- clear VDO metadata on a vdo remove call
bd4e69
- fix create of new dedupe indices
bd4e69
- add magic number to VDO geometry block
bd4e69
- do less logging when stopping a VDO
bd4e69
- add a UUID
bd4e69
- Resolves: rhbz#1512127
bd4e69
- Resolves: rhbz#1516081
bd4e69
- Resolves: rhbz#1511109
bd4e69
- Resolves: rhbz#1515183
bd4e69
bd4e69
* Fri Nov 17 2017 Joe Chapman <jochapma@redhat.com> 6.1.0.55-9
bd4e69
- fail loading an uncreated index more gracefully
bd4e69
- remove spurious/unnecessary files from the distribution
bd4e69
- fix kernel module version
bd4e69
- make logging less chatty
bd4e69
- fix an integer overflow in makeVDOLayout
bd4e69
- Resolves: rhbz#1511034
bd4e69
- Resolves: rhbz#1511109
bd4e69
- Resolves: rhbz#1511096
bd4e69
bd4e69
* Fri Nov 10 2017 Joe Chapman <jochapma@redhat.com> 6.1.0.44-8
bd4e69
- fix readCacheSize handling large numbers
bd4e69
- vdoformat signals error when it finds a geometry block
bd4e69
- prevent kernel oops when loading an old geometry block
bd4e69
- vdoformat silently rounds down physical sizes to a block boundary
bd4e69
- UDS threads identify related VDO device
bd4e69
- clean up contents of source tarballs
bd4e69
- Resolves: rhbz#1505936
bd4e69
- Resolves: rhbz#1507996
bd4e69
- Resolves: rhbz#1509466
bd4e69
- Resolves: rhbz#1510558
bd4e69
- Resolves: rhbz#1510585
bd4e69
- Resolves: rhbz#1511201
bd4e69
- Resolves: rhbz#1511209
bd4e69
bd4e69
* Fri Nov 03 2017 Joe Chapman <jochapma@redhat.com> 6.1.0.34-7
bd4e69
- Bugfixes
bd4e69
- Resolves: rhbz#1491422
bd4e69
bd4e69
* Mon Oct 30 2017 Joe Chapman <jochapma@redhat.com> 6.1.0.13-6
bd4e69
- Fixed some scanning tool complaints
bd4e69
- Resolves: rhbz#1491422
bd4e69
bd4e69
* Tue Oct 24 2017 Andy Walsh <awalsh@redhat.com> 6.1.0.0-6
bd4e69
- Fixed kernel requirement to allow subsequent releases without updating spec
bd4e69
- Resolves: rhbz#1491422
bd4e69
bd4e69
* Fri Oct 20 2017 John Wiele <jwiele@redhat.com> 6.1.0.0-5
bd4e69
- Bumped kernel requirement to 3.10.0-741
bd4e69
- Resolves: rhbz#1491422
bd4e69
bd4e69
* Tue Oct 17 2017 John Wiele <jwiele@redhat.com> 6.1.0.0-4
bd4e69
- Resolved some missing symbols
bd4e69
- Resolves: rhbz#1491422
bd4e69
bd4e69
* Mon Oct 16 2017 John Wiele <jwiele@redhat.com> 6.1.0.0-3
bd4e69
- Updated to provide a useable package
bd4e69
- Resolves: rhbz#1491422
bd4e69
bd4e69
* Sat Oct 14 2017 Andy Walsh <awalsh@redhat.com> 6.1.0.0-2
bd4e69
- Removed invalid requirement and some unnecessary comments in spec
bd4e69
- Resolves: rhbz#1491422
bd4e69
bd4e69
* Wed Oct 11 2017 John Wiele <jwiele@redhat.com> 6.1.0.0-1
bd4e69
- Initial vdo module for Driver Update Program
bd4e69
- Resolves: rhbz#1491422