Hello community,
here is the log from the commit of package kdump for openSUSE:Factory checked in at 2017-06-30 18:39:25
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/kdump (Old)
and /work/SRC/openSUSE:Factory/.kdump.new (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "kdump"
Fri Jun 30 18:39:25 2017 rev:89 rq:507135 version:0.8.16
Changes:
--------
--- /work/SRC/openSUSE:Factory/kdump/kdump.changes 2017-06-20 10:59:20.308487979 +0200
+++ /work/SRC/openSUSE:Factory/.kdump.new/kdump.changes 2017-06-30 18:39:36.849321892 +0200
@@ -1,0 +2,60 @@
+Thu Jun 29 16:31:26 UTC 2017 - ptesarik@suse.com
+
+- kdump-activate-QETH-devices.patch: Activate QETH network devices
+ (bsc#1038669).
+
+-------------------------------------------------------------------
+Tue Jun 27 06:06:54 UTC 2017 - ptesarik@suse.com
+
+- Drop kdump-do-not-reload-FADUMP-on-CPU-memory-hotplug.patch:
+ Reload is needed even with fadmp (bsc#1040567).
+
+-------------------------------------------------------------------
+Mon Jun 26 11:24:47 UTC 2017 - ptesarik@suse.com
+
+- kdump-fail-if-fadump-cannot-be-registered.patch: Fail at startup
+ if fadump cannot be registered (bsc#1040567).
+- kdump-do-not-reload-FADUMP-on-CPU-memory-hotplug.patch: Do not
+ reload on CPU/memory hotplug when using FADUMP (bsc#1040567).
+
+-------------------------------------------------------------------
+Fri Jun 23 12:56:45 UTC 2017 - ptesarik@suse.com
+
+- kdump-explicitly-request-zFCP-devices.patch: Explicitly request
+ zFCP devices in kdump initrd (bsc#1008352).
+
+-------------------------------------------------------------------
+Fri Jun 23 11:51:11 UTC 2017 - ptesarik@suse.com
+
+- kdump-move-class-SystemCPU-to-header-file.patch: Move class
+ SystemCPU to a header file (bsc#1036223).
+- kdump-treat-KDUMP_CPUS-0-as-all-available-CPUs.patch: Treat
+ KDUMP_CPUS=0 as "all available CPUs" (bsc#1036223).
+- kdump-multithreading-by-default.patch: Use multithreading by
+ default (bsc#1036223).
+
+-------------------------------------------------------------------
+Thu Jun 22 12:42:50 UTC 2017 - ptesarik@suse.com
+
+- kdump-fix-save_dump-to-NFS.patch: Fix save_dump to NFS targets
+ (bsc#1045541).
+- kdump-invoke-subcommand-destructors-on-exit.patch: Invoke
+ subcommand destructors on exit (bsc#1045541).
+
+-------------------------------------------------------------------
+Fri Jun 16 12:04:14 UTC 2017 - msuchanek@suse.com
+
+- kdump-do-not-free-fadump-memory-when-immediate-reboot-is-requested.patch
+ Releasing fadump memory can take a long time so skip it when
+ rebooting anyway (bsc#1040610).
+
+-------------------------------------------------------------------
+Tue Jun 15 12:55:34 UTC 2017 - msuchanek@suse.com
+
+- kdump-do-not-check-bind-mount.patch: Do not request filesystem
+ check on bind mounts (bsc#1034169).
+- kdump-remount-sysroot-readwrite.patch: Also remount writable
+ any mounts that were already mounted readonly by systemd
+ (bsc#1034169).
+
+-------------------------------------------------------------------
@@ -28,0 +89,8 @@
+Thu Jun 1 14:50:25 UTC 2017 - msuchanek@suse.com
+
+- kdump-source-save_dump.patch: save_dump.sh is designed to be
+ sourced and has numerous toplevel return statements. Source it
+ from the service definition as well to prevent bash complaints.
+ (bcs#1034169).
+
+-------------------------------------------------------------------
@@ -157 +225 @@
-------------------------------------------------------------------
+-------------------------------------------------------------------
New:
----
kdump-activate-QETH-devices.patch
kdump-do-not-check-bind-mount.patch
kdump-do-not-free-fadump-memory-when-immediate-reboot-is-requested.patch
kdump-explicitly-request-zFCP-devices.patch
kdump-fail-if-fadump-cannot-be-registered.patch
kdump-fix-save_dump-to-NFS.patch
kdump-invoke-subcommand-destructors-on-exit.patch
kdump-move-class-SystemCPU-to-header-file.patch
kdump-multithreading-by-default.patch
kdump-remount-sysroot-readwrite.patch
kdump-source-save_dump.patch
kdump-treat-KDUMP_CPUS-0-as-all-available-CPUs.patch
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Other differences:
------------------
++++++ kdump.spec ++++++
--- /var/tmp/diff_new_pack.v6iJSV/_old 2017-06-30 18:39:37.661207696 +0200
+++ /var/tmp/diff_new_pack.v6iJSV/_new 2017-06-30 18:39:37.665207133 +0200
@@ -55,6 +55,18 @@
Patch11: %{name}-Routable-preferred-source-address.patch
Patch12: %{name}-URLTransfer-complete-target.patch
Patch13: %{name}-prepend-IP-address.patch
+Patch14: %{name}-do-not-free-fadump-memory-when-immediate-reboot-is-requested.patch
+Patch15: %{name}-do-not-check-bind-mount.patch
+Patch16: %{name}-source-save_dump.patch
+Patch17: %{name}-remount-sysroot-readwrite.patch
+Patch18: %{name}-fix-save_dump-to-NFS.patch
+Patch19: %{name}-invoke-subcommand-destructors-on-exit.patch
+Patch20: %{name}-move-class-SystemCPU-to-header-file.patch
+Patch21: %{name}-treat-KDUMP_CPUS-0-as-all-available-CPUs.patch
+Patch22: %{name}-multithreading-by-default.patch
+Patch23: %{name}-explicitly-request-zFCP-devices.patch
+Patch24: %{name}-fail-if-fadump-cannot-be-registered.patch
+Patch25: %{name}-activate-QETH-devices.patch
BuildRequires: asciidoc
BuildRequires: cmake
BuildRequires: gcc-c++
@@ -128,6 +140,18 @@
%patch11 -p1
%patch12 -p1
%patch13 -p1
+%patch14 -p1
+%patch15 -p1
+%patch16 -p1
+%patch17 -p1
+%patch18 -p1
+%patch19 -p1
+%patch20 -p1
+%patch21 -p1
+%patch22 -p1
+%patch23 -p1
+%patch24 -p1
+%patch25 -p1
%build
export CFLAGS="%{optflags}"
++++++ kdump-activate-QETH-devices.patch ++++++
From: Petr Tesarik
Date: Thu Jun 29 18:11:52 2017 +0200
Subject: Activate QETH network devices
References: bsc#1038669
Upstream: v0.8.17
Git-commit: 2288f2b56fec0298fdf21ff7a5dd03157d12242e
Take care of activating QETH devices in the kdump environment, because
dracut does not do it automatically.
Signed-off-by: Petr Tesarik
---
init/setup-kdump.functions | 72 +++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 72 insertions(+)
--- a/init/setup-kdump.functions
+++ b/init/setup-kdump.functions
@@ -457,6 +457,7 @@ function kdump_vlan_config() # {
# kdump_netif corresponding ifname= initrd parameter added
# kdump_iface device name in initrd
# kdump_kmods additional kernel modules updated
+# kdump_hwif hardware interfaces updated
function kdump_ifname_config() # {{{
{
kdump_iface="$1"
@@ -464,6 +465,8 @@ function kdump_ifname_config() #
if [ -z "$ifkind" ]
then
+ kdump_hwif="$kdump_hwif $kdump_iface"
+
local hwaddr
local addrtype=$(<"/sys/class/net/$kdump_iface/addr_assign_type")
if [ "$addrtype" -eq 0 ]
@@ -1038,6 +1041,69 @@ function kdump_filter_sysctl() #
} # }}}
#
+# Set up a QETH network interface
+# Parameters:
+# 1) _root: initrd temporary root
+# 2) _iface: interface name
+# Input variables:
+# kdump_hwif hardware network interfaces
+function kdump_setup_qeth() # {{{
+{
+ local _root="$1"
+ local _iface="$2"
+ local _dev=$( readlink "/sys/class/net/$_iface/device" )
+ _dev="${_dev##*/}"
+ local _cdev0=$( readlink "/sys/class/net/$_iface/device/cdev0" )
+ _cdev0="${_cdev0##*/}"
+ local _cdev1=$( readlink "/sys/class/net/$_iface/device/cdev1" )
+ _cdev1="${_cdev1##*/}"
+ local _cdev2=$( readlink "/sys/class/net/$_iface/device/cdev2" )
+ _cdev2="${_cdev2##*/}"
+ local _layer2=$("${_root}/etc/udev/rules.d/51-qeth-${_dev}.rules" <
Date: Wed Jun 21 15:31:21 2017 +0200
Subject: Do not request filesystem check on bind mounts
References: bsc#1034169
Upstream: v0.8.17
Git-commit: 395a14377816a19678da481cafdf2332e0b0c932
When creating fstab for kdump bind mounts the dump and fsck fields are not
specified. These should be 0 because the directory which is bind-mounted cannot
be checked nor dumped.
Reported-by: Neil Brown
Signed-off-by: Michal Suchanek
--- kdump-0.8.16/init/module-setup.sh~ 2017-06-01 16:56:11.765609113 +0200
+++ kdump-0.8.16/init/module-setup.sh 2017-06-01 16:58:31.295723256 +0200
@@ -168,7 +168,7 @@
do
line=( ${fstab_lines[i]} )
if [ "${line[1]%/*}" = "/kdump" ] ; then
- fstab_lines[i]="/sysroot ${line[1]} none bind"
+ fstab_lines[i]="/sysroot ${line[1]} none bind 0 0"
fi
done
fi
++++++ kdump-do-not-free-fadump-memory-when-immediate-reboot-is-requested.patch ++++++
From: Michal Suchanek
Date: Wed Jun 21 15:25:49 2017 +0200
Subject: Do not free fadump memory when reboot is requested
References: bsc#1040610
Upstream: v0.8.17
Git-commit: 653f8c0931c638cbb0bd5c22fc0e1dfd7969b29a
Freeing fadump memory can take a long time and doing it when we are going to
reboot anyway is pointless.
In bsc#1034169 a kernel oops is triggered by freeing fadump memory killing
save_dump.sh and preventing the reboot requested by the user from happening
entirely.
Avoid the issue and speed up reboot by not freeing fadump memory when reboot is
requested.
Signed-off-by: Michal Suchanek
--- kdump-0.8.16/init/save_dump.sh~ 2017-06-16 13:59:02.017110055 +0200
+++ kdump-0.8.16/init/save_dump.sh 2017-06-16 13:59:35.629621303 +0200
@@ -49,7 +49,8 @@
if fadump_enabled; then
# release memory if possible
- test -f $FADUMP_RELEASE_MEM && echo 1 > $FADUMP_RELEASE_MEM
+ [ -f $FADUMP_RELEASE_MEM -a $KDUMP_IMMEDIATE_REBOOT != "yes" \
+ -a "$KDUMP_IMMEDIATE_REBOOT" != "YES" ] && echo 1 > $FADUMP_RELEASE_MEM
if [ "$KDUMP_FADUMP_SHELL" = "yes" \
-o "$KDUMP_FADUMP_SHELL" = "YES" ] ; then
echo
++++++ kdump-explicitly-request-zFCP-devices.patch ++++++
From: Petr Tesarik
Date: Fri Jun 23 14:53:50 2017 +0200
Subject: Explicitly request zFCP devices in kdump initrd
References: bsc#1008352
Upstream: v0.8.17
Git-commit: 81899709e4a12bea4819d4ccc10f960cb535b447
When using NPIV with LUN autoscan, dracut does not save the port and
LUN information. However, LUN autoscan is always disabled in kdump
environment, so the LUNs will never appear.
To fix this case, the full information must be saved explicitly in a
kdump initrd.
Signed-off-by: Petr Tesarik
---
init/module-setup.sh | 22 ++++++++++++++++++++++
1 file changed, 22 insertions(+)
--- a/init/module-setup.sh
+++ b/init/module-setup.sh
@@ -81,6 +81,25 @@ kdump_add_mpath_dev() {
fi
}
+kdump_cmdline_zfcp() {
+ is_zfcp() {
+ local _dev=$1
+ local _devpath=$(cd -P /sys/dev/block/$_dev ; echo $PWD)
+ local _sdev _lun _wwpn _ccw
+
+ [ "${_devpath#*/sd}" == "$_devpath" ] && return 1
+ _sdev="${_devpath%%/block/*}"
+ [ -e ${_sdev}/fcp_lun ] || return 1
+ _ccw=$(cat ${_sdev}/hba_id)
+ _lun=$(cat ${_sdev}/fcp_lun)
+ _wwpn=$(cat ${_sdev}/wwpn)
+ echo "rd.zfcp=${_ccw},${_wwpn},${_lun}"
+ }
+ [[ $hostonly ]] || [[ $mount_needs ]] && {
+ for_each_host_dev_and_slaves_all is_zfcp
+ } | sort -u
+}
+
kdump_cmdline_ip() {
[ "$kdump_neednet" = y ] || return 0
@@ -142,6 +161,9 @@ kdump_gen_mount_units() {
}
cmdline() {
+ local _arch=$(uname -m)
+ [ "$_arch" = "s390" -o "$_arch" = "s390x" ] && kdump_cmdline_zfcp
+
kdump_cmdline_ip
}
++++++ kdump-fail-if-fadump-cannot-be-registered.patch ++++++
From: Petr Tesarik
Date: Mon Jun 26 13:22:10 2017 +0200
Subject: Fail at startup if fadump cannot be registered
References: bsc#1040567
Upstream: v0.8.17
Git-commit: 99a38e490012da75d83efb83385d42edbea100e9
If writing "1" into /sys/kernel/fadump_registered fails for whatever
reason, do not return success.
Signed-off-by: Petr Tesarik
diff --git a/init/load.sh b/init/load.sh
index 1894551..4fc7a82 100755
--- a/init/load.sh
+++ b/init/load.sh
@@ -241,17 +241,19 @@ function load_kdump_fadump()
fi
local msg
+ local result=0
# The kernel fails with EINVAL if registered already
# (see bnc#814780)
- if [ $(cat "$FADUMP_REGISTERED") != "1" ] ; then
- local output=$( (echo 1 > "$FADUMP_REGISTERED") 2>&1)
- local result=$?
+ if [ "$(cat $FADUMP_REGISTERED)" != "1" ] ; then
+ local output
- if [ $result -eq 0 ] ; then
+ output=$( (echo 1 > "$FADUMP_REGISTERED") 2>&1)
+ if [ $? -eq 0 ] ; then
msg="Registered fadump"
else
msg="FAILED to register fadump: $output"
+ result=1
fi
else
msg="fadump is already registered"
++++++ kdump-fix-save_dump-to-NFS.patch ++++++
From: Petr Tesarik
Date: Thu Jun 22 14:23:02 2017 +0200
Subject: Fix save_dump to NFS targets
References: bsc#1045541
Upstream: v0.8.17
Git-commit: 11d36a645ab99c9805f9fa3ca9eabce3aa2a9418
For NFS, the parent directory of the target must be mounted, because
the target directory itself does not exist yet.
For this to work, split the target path to directory (i.e. parent
directory) and basename. Then construct a unique mount point from the
host name and the parent directory, and finally append the base name
again to form a subdirectory in the target (to be created by the file
transfer method).
Signed-off-by: Petr Tesarik
---
kdumptool/transfer.cc | 24 +++++++++++-------------
kdumptool/transfer.h | 4 +---
2 files changed, 12 insertions(+), 16 deletions(-)
--- a/kdumptool/transfer.cc
+++ b/kdumptool/transfer.cc
@@ -462,22 +462,20 @@ RootDirURL NFSTransfer::translate(const
if (!rt.check(config->KDUMP_NET_TIMEOUT.value()))
cerr << "WARNING: Dump target not reachable" << endl;
- string mountedDir = parser.getPath();
- FileUtil::nfsmount(parser.getHostname(), mountedDir,
- DEFAULT_MOUNTPOINT, options);
-
+ FilePath path = parser.getPath();
+ string mountedDir = path.dirName();
+ string rest = path.baseName();
m_mountpoint = DEFAULT_MOUNTPOINT;
- m_rest = parser.getPath();
- m_rest.replace(m_rest.begin(), m_rest.begin() + mountedDir.size(), "");
- m_rest.ltrim("/");
+ m_mountpoint.appendPath(parser.getHostname()).appendPath(mountedDir);
+ m_mountpoint.mkdir(true);
+ Debug::debug()->dbg("Path: %s, Mountpoint: %s, Rest: %s",
+ path.c_str(), m_mountpoint.c_str(), rest.c_str());
- (m_prefix = m_mountpoint).appendPath(m_rest);
-
- Debug::debug()->dbg("Mountpoint: %s, Rest: %s, Prefix: $s",
- m_mountpoint.c_str(), m_rest.c_str(), m_prefix.c_str());
+ FileUtil::nfsmount(parser.getHostname(), mountedDir,
+ m_mountpoint, options);
- return RootDirURL("file://" + m_prefix, "");
+ return RootDirURL("file://" + m_mountpoint + PATH_SEPARATOR + rest, "");
}
// -----------------------------------------------------------------------------
@@ -507,7 +505,7 @@ void NFSTransfer::close()
throw (KError)
{
Debug::debug()->trace("NFSTransfer::close()");
- if (m_mountpoint.size() > 0) {
+ if (!m_mountpoint.empty()) {
FileUtil::umount(m_mountpoint);
m_mountpoint.clear();
}
--- a/kdumptool/transfer.h
+++ b/kdumptool/transfer.h
@@ -271,9 +271,7 @@ class NFSTransfer : public URLTransfer {
private:
- std::string m_mountpoint;
- KString m_rest;
- FilePath m_prefix;
+ FilePath m_mountpoint;
FileTransfer *m_fileTransfer;
};
++++++ kdump-invoke-subcommand-destructors-on-exit.patch ++++++
From: Petr Tesarik
Date: Thu Jun 22 14:31:26 2017 +0200
Subject: Invoke subcommand destructors on exit
References: bsc#1045541
Upstream: v0.8.17
Git-commit: ff813cdc274caf7c006d438dbf30d83413bb896e
To execute the destructor for a subcommand, the object must be
explicitly freed in KdumpTool destructor. This fixes a bug that
NFS shares are not unmounted on exit.
Signed-off-by: Petr Tesarik
---
kdumptool/kdumptool.cc | 1 +
1 file changed, 1 insertion(+)
--- a/kdumptool/kdumptool.cc
+++ b/kdumptool/kdumptool.cc
@@ -67,6 +67,7 @@ KdumpTool::~KdumpTool()
throw ()
{
Debug::debug()->trace("KdumpTool::~KdumpTool()");
+ delete m_subcommand;
}
// -----------------------------------------------------------------------------
++++++ kdump-move-class-SystemCPU-to-header-file.patch ++++++
From: Petr Tesarik
Date: Fri Jun 23 13:19:41 2017 +0200
Subject: Move class SystemCPU to a header file
References: bsc#1036223
Upstream: v0.8.17
Git-commit: ded0d2117cce72fc98a7ca9ee577090889ef9b16
This allows to reuse the class outside of calibrate.cc.
Signed-off-by: Petr Tesarik
---
kdumptool/calibrate.cc | 47 ---------------------------------------------
kdumptool/calibrate.h | 51 +++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 51 insertions(+), 47 deletions(-)
--- a/kdumptool/calibrate.cc
+++ b/kdumptool/calibrate.cc
@@ -251,53 +251,6 @@ using std::ifstream;
//{{{ SystemCPU ----------------------------------------------------------------
-class SystemCPU {
-
- public:
- /**
- * Initialize a new SystemCPU object.
- *
- * @param[in] sysdir Mount point for sysfs
- */
- SystemCPU(const char *sysdir = "/sys")
- throw ()
- : m_cpudir(FilePath(sysdir).appendPath("devices/system/cpu"))
- {}
-
- protected:
- /**
- * Path to the cpu system devices base directory
- */
- const FilePath m_cpudir;
-
- /**
- * Count the number of CPUs in a cpuset
- *
- * @param[in] name Name of the cpuset ("possible", "present", "online")
- *
- * @exception KError if the file cannot be opened or parsed
- */
- unsigned long count(const char *name);
-
- public:
- /**
- * Count the number of online CPUs
- *
- * @exception KError see @c count()
- */
- unsigned long numOnline(void)
- { return count("online"); }
-
- /**
- * Count the number of offline CPUs
- *
- * @exception KError see @c count()
- */
- unsigned long numOffline(void)
- { return count("offline"); }
-
-};
-
// -----------------------------------------------------------------------------
unsigned long SystemCPU::count(const char *name)
{
--- a/kdumptool/calibrate.h
+++ b/kdumptool/calibrate.h
@@ -20,6 +20,7 @@
#define CALIBRATE_H
#include "subcommand.h"
+#include "fileutil.h"
//{{{ Calibrate ----------------------------------------------------------------
@@ -54,6 +55,56 @@ class Calibrate : public Subcommand {
};
//}}}
+//{{{ SystemCPU ----------------------------------------------------------------
+
+class SystemCPU {
+
+ public:
+ /**
+ * Initialize a new SystemCPU object.
+ *
+ * @param[in] sysdir Mount point for sysfs
+ */
+ SystemCPU(const char *sysdir = "/sys")
+ throw ()
+ : m_cpudir(FilePath(sysdir).appendPath("devices/system/cpu"))
+ {}
+
+ protected:
+ /**
+ * Path to the cpu system devices base directory
+ */
+ const FilePath m_cpudir;
+
+ /**
+ * Count the number of CPUs in a cpuset
+ *
+ * @param[in] name Name of the cpuset ("possible", "present", "online")
+ *
+ * @exception KError if the file cannot be opened or parsed
+ */
+ unsigned long count(const char *name);
+
+ public:
+ /**
+ * Count the number of online CPUs
+ *
+ * @exception KError see @c count()
+ */
+ unsigned long numOnline(void)
+ { return count("online"); }
+
+ /**
+ * Count the number of offline CPUs
+ *
+ * @exception KError see @c count()
+ */
+ unsigned long numOffline(void)
+ { return count("offline"); }
+
+};
+
+//}}}
#endif /* CALIBRATE_H */
++++++ kdump-multithreading-by-default.patch ++++++
From: Petr Tesarik
Date: Fri Jun 23 13:47:28 2017 +0200
Subject: Multithreading by default
References: bsc#1036223
Upstream: v0.8.17
Git-commit: 07534f44546e37c65a08fe7fb093255fbca6cd30
The "--num-threads" and "--split" makedumpfile options are mutually
exclusive. Before multithreading was implemented, "--split" was the
only way to use any additional CPUs. It makes more sense to use
multithreading by default.
To use "--split", the SPLIT flag can be specified.
Signed-off-by: Petr Tesarik
---
doc/man/kdump.5.txt.in | 16 +++++++++++-----
kdumptool/savedump.cc | 17 +++++++++++++++++
kdumptool/savedump.h | 1 +
sysconfig.kdump.in | 5 +++--
4 files changed, 32 insertions(+), 7 deletions(-)
--- a/doc/man/kdump.5.txt.in
+++ b/doc/man/kdump.5.txt.in
@@ -513,11 +513,17 @@ This is a space-separated list of flags
Because SFTP and FTP are not mounted, that option has no meaning when saving
the dump to SFTP and FTP.
-*NOSPLIT*::
- Disable the _--split_ option of *makedumpfile*(8). This option is normally
- added when KDUMP_CPUS>1, because otherwise the additional CPUs are idle.
- You can specify this flag to force the use of only one dumping process,
- regardless of the value of KDUMP_CPUS.
+*SPLIT*::
+ If KDUMP_CPUS>1, use the _--split_ option of *makedumpfile*(8) instead of
+ the default _--num-threads_.
+
+*SINGLE*::
+ Specify this flag to force the use of only one CPU for dumping, regardless
+ of the value of KDUMP_CPUS.
+ It disables the _--split_ and _--num-threads_ options of *makedumpfile*(8).
+
+ For compatibility with older versions, *NOSPLIT* is an alias for *SINGLE*.
+ Its use is deprecated.
*XENALLDOMAINS*::
When dumping a Xen virtualization host, *makedumpfile*(8) is normally
--- a/kdumptool/savedump.cc
+++ b/kdumptool/savedump.cc
@@ -291,11 +291,24 @@ void SaveDump::saveDump(const RootDirURL
cpus = syscpu.numOnline();
}
if (!config->kdumptoolContainsFlag("NOSPLIT") &&
+ !config->kdumptoolContainsFlag("SINGLE") &&
cpus > 1) {
if (!useElf)
m_split = cpus;
else
cerr << "Splitting ELF dumps is not supported." << endl;
+
+ if (config->kdumptoolContainsFlag("SPLIT")) {
+ if (!useElf)
+ m_split = cpus;
+ else
+ cerr << "Splitting ELF dumps is not supported." << endl;
+ } else {
+ if (!useElf)
+ m_threads = cpus - 1;
+ else
+ cerr << "Multithreading is unavailable for ELF dumps" << endl;
+ }
}
bool excludeDomU = false;
@@ -313,6 +326,10 @@ void SaveDump::saveDump(const RootDirURL
cmdline << "makedumpfile ";
if (m_split)
cmdline << "--split ";
+ if (m_threads) {
+ SystemCPU syscpu;
+ cmdline << "--num-threads " << m_threads << " ";
+ }
cmdline << config->MAKEDUMPFILE_OPTIONS.value() << " ";
cmdline << "-d " << config->KDUMP_DUMPLEVEL.value() << " ";
if (excludeDomU)
--- a/kdumptool/savedump.h
+++ b/kdumptool/savedump.h
@@ -113,6 +113,7 @@ class SaveDump : public Subcommand {
bool m_usedDirectSave;
bool m_useMakedumpfile;
unsigned long m_split;
+ unsigned long m_threads;
std::string m_crashtime;
std::string m_crashrelease;
std::string m_rootdir;
--- a/sysconfig.kdump.in
+++ b/sysconfig.kdump.in
@@ -271,14 +271,15 @@ KDUMP_POSTSCRIPT=""
#
KDUMP_COPY_KERNEL="yes"
-## Type: string(NOSPARSE,NOSPLIT,XENALLDOMAINS)
+## Type: string(NOSPARSE,SPLIT,SINGLE,XENALLDOMAINS)
## Default: ""
## ServiceRestart: kdump
#
# Space-separated list of flags to tweak the run-time behaviour of kdumptool:
#
# NOSPARSE disable creation of sparse files.
-# NOSPLIT do not pass "--split" to makedumpfile even if KDUMP_CPUS > 1
+# SPLIT split the dump file with "makedumpfile --split"
+# SINGLE use single CPU to save the dump
# XENALLDOMAINS do not filter out Xen DomU pages
#
# See also: kdump(5).
++++++ kdump-remount-sysroot-readwrite.patch ++++++
From: Michal Suchanek
Date: Wed Jun 21 15:23:30 2017 +0200
Subject: Remount /sysroot readwrite when used for kdump
References: bsc#1034169
Upstream: v0.8.17
Git-commit: b3e40e6cb5794e441ddabf1de88f3729ebd0cc6f
When kdump is saved to / (ie /var is not separate filesystem) /kdump/mnt0 is
bind mount of readonly /sysroot. Due to mount bug "mount /kdump/mnt0 -o
remount,rw" does nothing.
Remount the device by hand with "mount none /kdump/mnt0 -o remount,rw" which
avoids the bug.
Based on patch by Neil Brown
Signed-off-by: Michal Suchanek
--- kdump-0.8.16/init/save_dump.sh~ 2017-06-15 14:47:33.059006747 +0200
+++ kdump-0.8.16/init/save_dump.sh 2017-06-15 14:47:42.827155243 +0200
@@ -127,13 +127,38 @@
check_for_device "$@"
}
+function rw_fixup()
+{
+ # handle remounting existing readonly mounts readwrite
+ # mount -a works only for not yet mounted filesystems
+ # remounting bind mounts needs special incantation
+ while read dev mpt fs opt dummy ; do
+ case "$opt" in
+ *bind*)
+ if [ "$fs" = "none" ] && ! [ -w "$mpt" ]; then
+ mount none "$mpt" -o remount,rw
+ fi
+ ;;
+ ro,* | *,ro,* | *,ro) ;;
+ *)
+ if ! [ -w "$mpt" ]; then
+ mount "$mpt" -o remount,rw
+ fi
+ ;;
+ esac
+ done < /etc/fstab
+}
+
#
# Mounts all partitions listed in /etc/fstab.kdump
function mount_all()
{
local ret=0
- test -f /etc/fstab.kdump || return 0
+ if ! [ -f /etc/fstab.kdump ] ; then
+ rw_fixup
+ return 0
+ fi
if [ -f /etc/fstab ] ; then
mv /etc/fstab /etc/fstab.orig
@@ -143,6 +167,8 @@
mount -a
ret=$?
+ rw_fixup
+
if [ -f /etc/fstab.orig ] ; then
mv /etc/fstab.orig /etc/fstab
else
++++++ kdump-source-save_dump.patch ++++++
From: Michal Suchanek
Date: Wed Jun 21 15:33:01 2017 +0200
Subject: Source save_dump.sh in systemd service.
References: bcs#1034169
Upstream: v0.8.17
Git-commit: b053165dd633978b1a78f6f7f993d34a62ee97c3
save_dump.sh is designed to be sourced and contains a number of toplevel return
statements. Bash will complain when these are executed unless the script is
sourced so source it to prevent useless noise.
Reported-by: Neil Brown
Signed-off-by: Michal Suchanek
--- kdump-0.8.16/init/kdump-save.service.in~ 2016-10-05 09:37:12.000000000 +0200
+++ kdump-0.8.16/init/kdump-save.service.in 2017-06-01 17:04:00.428710624 +0200
@@ -17,7 +17,7 @@
[Service]
Type=oneshot
-ExecStart=/lib/kdump/save_dump.sh
+ExecStart=/bin/bash -c "source /lib/kdump/save_dump.sh"
StandardInput=tty
StandardOutput=tty
StandardError=tty
++++++ kdump-treat-KDUMP_CPUS-0-as-all-available-CPUs.patch ++++++
From: Petr Tesarik
Date: Fri Jun 23 13:43:15 2017 +0200
Subject: Treat KDUMP_CPUS=0 as "all available CPUs"
References: bsc#1036223
Upstream: v0.8.17
Git-commit: 34192abe9a0f5ddbcd580fa52b7fa4400533b9a4
Since SMP after crash is now considered stable, remove the warnings and
make the feature more useful by not requiring a fixed number of CPUs.
Signed-off-by: Petr Tesarik
---
doc/man/kdump.5.txt.in | 7 +++----
init/load.sh | 7 +++++--
kdumptool/calibrate.cc | 6 +++---
kdumptool/savedump.cc | 23 ++++++++++++++---------
kdumptool/savedump.h | 2 +-
sysconfig.kdump.in | 7 +------
6 files changed, 27 insertions(+), 25 deletions(-)
--- a/doc/man/kdump.5.txt.in
+++ b/doc/man/kdump.5.txt.in
@@ -136,13 +136,12 @@ the number of CPUs in your system.
This parameter modifies the _maxcpus_ parameter of the kdump kernel and
the number of *makedumpfile*(8) processes created.
+If the value is zero, all available CPUs are used, i.e. the _maxcpus_
+parameter is not added to the kdump kernel command line.
+
*Note:* This parameter does not work properly for the _ELF_ format,
because *makedumpfile*(8) does not support split _ELF_ dump files.
-*WARNING:* SETTING THIS NUMBER TO A NUMBER HIGHER THAN 1 IS EXPERIMENTAL!
-SOME MACHINES ARE KNOWN TO HANG OR REBOOT RANDOMLY DURING INITIALISATION
-OF THE DUMP KERNEL IF KDUMP_CPUS > 1.
-
Default is 1.
--- a/init/load.sh
+++ b/init/load.sh
@@ -63,14 +63,17 @@ function build_kdump_commandline()
local commandline="$KDUMP_COMMANDLINE"
if [ -z "$commandline" ] ; then
- local nr_cpus=$(cpus_param "$kdump_kernel")
+ local nr_cpus
commandline=$(
remove_from_commandline \
'root|resume|crashkernel|splash|mem|BOOT_IMAGE|showopts|zfcp\.allow_lun_scan|hugepages|acpi_no_memhotplug|cgroup_disable|unknown_nmi_panic|rd\.udev\.children-max' \
< /proc/cmdline)
+ if [ ${KDUMP_CPUS:-1} -ne 0 ] ; then
+ nr_cpus=$(cpus_param "$kdump_kernel")=${KDUMP_CPUS:-1}
+ fi
# Use deadline for saving the memory footprint
commandline="$commandline elevator=deadline sysrq=yes reset_devices acpi_no_memhotplug cgroup_disable=memory"
- commandline="$commandline irqpoll ${nr_cpus}=${KDUMP_CPUS:-1}"
+ commandline="$commandline irqpoll ${nr_cpus}"
commandline="$commandline root=kdump rootflags=bind rd.udev.children-max=8"
case $(uname -i) in
i?86|x86_64)
--- a/kdumptool/calibrate.cc
+++ b/kdumptool/calibrate.cc
@@ -905,10 +905,10 @@ void Calibrate::execute()
}
// Add memory based on CPU count
- unsigned long cpus;
- if (CAN_REDUCE_CPUS) {
+ unsigned long cpus = 0;
+ if (CAN_REDUCE_CPUS)
cpus = config->KDUMP_CPUS.value();
- } else {
+ if (!cpus) {
SystemCPU syscpu;
unsigned long online = syscpu.numOnline();
unsigned long offline = syscpu.numOffline();
--- a/kdumptool/savedump.cc
+++ b/kdumptool/savedump.cc
@@ -41,6 +41,7 @@
#include "identifykernel.h"
#include "email.h"
#include "routable.h"
+#include "calibrate.h"
using std::string;
using std::list;
@@ -60,7 +61,7 @@ using std::ifstream;
SaveDump::SaveDump()
throw ()
: m_dump(DEFAULT_DUMP), m_transfer(NULL), m_usedDirectSave(false),
- m_useMakedumpfile(false), m_useSplit(false), m_nomail(false)
+ m_useMakedumpfile(false), m_split(0), m_nomail(false)
{
Debug::debug()->trace("SaveDump::SaveDump()");
@@ -284,10 +285,15 @@ void SaveDump::saveDump(const RootDirURL
if (noDump)
return; // nothing to be done
+ unsigned long cpus = config->KDUMP_CPUS.value();
+ if (cpus) {
+ SystemCPU syscpu;
+ cpus = syscpu.numOnline();
+ }
if (!config->kdumptoolContainsFlag("NOSPLIT") &&
- config->KDUMP_CPUS.value() > 1) {
+ cpus > 1) {
if (!useElf)
- m_useSplit = true;
+ m_split = cpus;
else
cerr << "Splitting ELF dumps is not supported." << endl;
}
@@ -305,7 +311,7 @@ void SaveDump::saveDump(const RootDirURL
// use makedumpfile
ostringstream cmdline;
cmdline << "makedumpfile ";
- if (m_useSplit)
+ if (m_split)
cmdline << "--split ";
cmdline << config->MAKEDUMPFILE_OPTIONS.value() << " ";
cmdline << "-d " << config->KDUMP_DUMPLEVEL.value() << " ";
@@ -340,10 +346,9 @@ void SaveDump::saveDump(const RootDirURL
provider->setProgress(&progress);
else
cout << "Saving dump ..." << endl;
- if (m_useSplit) {
+ if (m_split) {
StringVector targets;
- const int cpus = config->KDUMP_CPUS.value();
- for (int i = 1; i <= cpus; ++i) {
+ for (unsigned long i = 1; i <= m_split; ++i) {
ostringstream ss;
ss << "vmcore" << i;
targets.push_back(ss.str());
@@ -482,8 +487,8 @@ void SaveDump::generateInfo()
ss << "Dump level : "
<< Stringutil::number2string(config->KDUMP_DUMPLEVEL.value()) << endl;
ss << "Dump format : " << config->KDUMP_DUMPFORMAT.value() << endl;
- if (m_useSplit && m_usedDirectSave)
- ss << "Split parts : " << config->KDUMP_CPUS.value() << endl;
+ if (m_split && m_usedDirectSave)
+ ss << "Split parts : " << m_split << endl;
ss << endl;
--- a/kdumptool/savedump.h
+++ b/kdumptool/savedump.h
@@ -112,7 +112,7 @@ class SaveDump : public Subcommand {
Transfer *m_transfer;
bool m_usedDirectSave;
bool m_useMakedumpfile;
- bool m_useSplit;
+ unsigned long m_split;
std::string m_crashtime;
std::string m_crashrelease;
std::string m_rootdir;
--- a/sysconfig.kdump.in
+++ b/sysconfig.kdump.in
@@ -18,13 +18,8 @@ KDUMP_KERNELVER=""
#
# Number of CPUs to be used in the kdump environment. You may want to
# increase the number if computing power is the bottleneck in your setup.
-# This value is used as the maxcpus parameter of the secondary kernel, so
-# DON'T CHANGE IT unless your system can reliably re-initialize SMP after
-# a kernel crash.
#
-# WARNING: SETTING THIS NUMBER TO A NUMBER HIGHER THAN 1 IS EXPERIMENTAL!
-# SOME MACHINES ARE KNOWN TO HANG OR REBOOT RANDOMLY DURING INITIALISATION
-# OF THE DUMP KERNEL IF KDUMP_CPUS > 1.
+# If the value is zero, use all available CPUs.
#
# See also: kdump(5).
#
