kexec-tools/dracut-module-setup.sh

1073 lines
32 KiB
Bash
Raw Normal View History

#!/bin/bash
_DRACUT_KDUMP_NM_TMP_DIR="$DRACUT_TMPDIR/$$-DRACUT_KDUMP_NM"
_save_kdump_netifs() {
unique_netifs[$1]=1
}
_get_kdump_netifs() {
echo -n "${!unique_netifs[@]}"
}
kdump_module_init() {
if ! [[ -d "${initdir}/tmp" ]]; then
mkdir -p "${initdir}/tmp"
fi
mkdir -p "$_DRACUT_KDUMP_NM_TMP_DIR"
. /lib/kdump/kdump-lib.sh
}
check() {
[[ $debug ]] && set -x
#kdumpctl sets this explicitly
if [[ -z $IN_KDUMP ]] || [[ ! -f /etc/kdump.conf ]]; then
return 1
fi
return 0
}
depends() {
local _dep="base shutdown"
kdump_module_init
add_opt_module() {
[[ " $omit_dracutmodules " != *\ $1\ * ]] && _dep="$_dep $1"
}
if is_squash_available; then
add_opt_module squash
else
dwarning "Required modules to build a squashed kdump image is missing!"
fi
if is_wdt_active; then
add_opt_module watchdog
fi
if is_ssh_dump_target; then
_dep="$_dep ssh-client"
fi
if is_lvm2_thinp_dump_target; then
if dracut --list-modules | grep -q lvmthinpool-monitor; then
add_opt_module lvmthinpool-monitor
else
dwarning "Required lvmthinpool-monitor modules is missing! Please upgrade dracut >= 057."
fi
fi
if [[ "$(uname -m)" == "s390x" ]]; then
_dep="$_dep znet"
fi
if [[ -n "$(ls -A /sys/class/drm 2> /dev/null)" ]] || [[ -d /sys/module/hyperv_fb ]]; then
add_opt_module drm
fi
if is_generic_fence_kdump || is_pcs_fence_kdump; then
_dep="$_dep network"
fi
echo "$_dep"
}
kdump_is_bridge() {
[[ -d /sys/class/net/"$1"/bridge ]]
}
kdump_is_bond() {
[[ -d /sys/class/net/"$1"/bonding ]]
}
kdump_is_team() {
[[ -f /usr/bin/teamnl ]] && teamnl "$1" ports &> /dev/null
}
kdump_is_vlan() {
[[ -f /proc/net/vlan/"$1" ]]
}
# $1: netdev name
source_ifcfg_file() {
local ifcfg_file
dwarning "Network Scripts are deprecated. You are encouraged to set up network by NetworkManager."
ifcfg_file=$(get_ifcfg_filename "$1")
if [[ -f ${ifcfg_file} ]]; then
. "${ifcfg_file}"
else
dwarning "The ifcfg file of $1 is not found!"
fi
}
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
# $1: repeat times
# $2: string to be repeated
# $3: separator
repeatedly_join_str() {
local _count="$1"
local _str="$2"
local _separator="$3"
local i _res
if [[ $_count -le 0 ]]; then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
echo -n ""
return
fi
i=0
_res="$_str"
((_count--))
while [[ $i -lt $_count ]]; do
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
((i++))
_res="${_res}${_separator}${_str}"
done
echo -n "$_res"
}
# $1: prefix
# $2: ipv6_flag="-6" indicates it's IPv6
# Given a prefix, calculate the netmask (equivalent of "ipcalc -m")
# by concatenating three parts,
# 1) the groups with all bits set 1
# 2) a group with partial bits set to 0
# 3) the groups with all bits set to 0
cal_netmask_by_prefix() {
local _prefix="$1"
local _ipv6_flag="$2" _ipv6
local _bits_per_octet=8
local _count _res _octets_per_group _octets_total _seperator _total_groups
local _max_group_value _max_group_value_repr _bits_per_group _tmp _zero_bits
if [[ $_ipv6_flag == "-6" ]]; then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
_ipv6=1
else
_ipv6=0
fi
if [[ $_prefix -lt 0 || $_prefix -gt 128 ]] \
|| ( ((!_ipv6)) && [[ $_prefix -gt 32 ]]); then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
derror "Bad prefix:$_prefix for calculating netmask"
exit 1
fi
if ((_ipv6)); then
_octets_per_group=2
_octets_total=16
_seperator=":"
else
_octets_per_group=1
_octets_total=4
_seperator="."
fi
_total_groups=$((_octets_total / _octets_per_group))
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
_bits_per_group=$((_octets_per_group * _bits_per_octet))
_max_group_value=$(((1 << _bits_per_group) - 1))
if ((_ipv6)); then
_max_group_value_repr=$(printf "%x" $_max_group_value)
else
_max_group_value_repr="$_max_group_value"
fi
_count=$((_prefix / _octets_per_group / _bits_per_octet))
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
_first_part=$(repeatedly_join_str "$_count" "$_max_group_value_repr" "$_seperator")
_res="$_first_part"
_tmp=$((_octets_total * _bits_per_octet - _prefix))
_zero_bits=$((_tmp % _bits_per_group))
if [[ $_zero_bits -ne 0 ]]; then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
_second_part=$((_max_group_value >> _zero_bits << _zero_bits))
if ((_ipv6)); then
_second_part=$(printf "%x" $_second_part)
fi
((_count++))
if [[ -z $_first_part ]]; then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
_res="$_second_part"
else
_res="${_first_part}${_seperator}${_second_part}"
fi
fi
_count=$((_total_groups - _count))
if [[ $_count -eq 0 ]]; then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
echo -n "$_res"
return
fi
if ((_ipv6)) && [[ $_count -gt 1 ]]; then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
# use condensed notion for IPv6
_third_part=":"
else
_third_part=$(repeatedly_join_str "$_count" "0" "$_seperator")
fi
if [[ -z $_res ]] && ((!_ipv6)); then
Implement IP netmask calculation to replace "ipcalc -m" Recently, dracut-network drops depedency on dhcp-client which requires ipcalc. Thus the dependency chain "kexec-tools -> dracut-network -> dhcp-client -> ipcalc" is broken. When NIC is configured to a static IP, kexec-tools depended on "ipcalc -m" to get netmask. This commit implements the shell equivalent of "ipcalc -m". The following test code shows cal_netmask_by_prefix is consistent with "ipcalc -m", #!/bin/bash . dracut-module-setup.sh for i in {0..128}; do mask_expected=$(ipcalc -m fe::/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "-6") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv6 tests passed" for i in {0..32}; do mask_expected=$(ipcalc -m 8.8.8.8/$i| cut -d"=" -f2) mask_actual=$(cal_netmask_by_prefix $i "") if [[ "$mask_expected" != "$mask_actual" ]]; then echo "prefix="$i, "expected="$mask_expected, "acutal="$mask_actual exit fi done echo "IPv4 tests passed" i=-2 res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi res=$(cal_netmask_by_prefix "$i" "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix<0" exit fi i=33 $(cal_netmask_by_prefix $i "") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>32 for IPv4" exit fi i=129 $(cal_netmask_by_prefix $i "-6") if [[ $? -ne 1 ]]; then echo "cal_netmask_by_prefix should exit when prefix>128 for IPv4" exit fi echo "Bad prefixes tests passed" echo "All tests passed" Reported-by: Jie Li <jieli@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-04-08 03:44:26 +00:00
echo -n "${_third_part}"
else
echo -n "${_res}${_seperator}${_third_part}"
fi
}
kdump_get_mac_addr() {
cat "/sys/class/net/$1/address"
}
#Bonding or team master modifies the mac address
#of its slaves, we should use perm address
kdump_get_perm_addr() {
local addr
addr=$(ethtool -P "$1" | sed -e 's/Permanent address: //')
if [[ -z $addr ]] || [[ $addr == "00:00:00:00:00:00" ]]; then
derror "Can't get the permanent address of $1"
else
echo "$addr"
fi
}
apply_nm_initrd_generator_timeouts() {
local _timeout_conf
_timeout_conf=$_DRACUT_KDUMP_NM_TMP_DIR/timeout_conf
cat << EOF > "$_timeout_conf"
[device-95-kdump]
carrier-wait-timeout=30000
[connection-95-kdump]
ipv4.dhcp-timeout=90
ipv6.dhcp-timeout=90
EOF
inst "$_timeout_conf" "/etc/NetworkManager/conf.d/95-kdump-timeouts.conf"
}
use_ipv4_or_ipv6() {
local _netif=$1 _uuid=$2
if [[ -v "ipv4_usage[$_netif]" ]]; then
nmcli connection modify --temporary "$_uuid" ipv4.may-fail no &> >(ddebug)
fi
if [[ -v "ipv6_usage[$_netif]" ]]; then
nmcli connection modify --temporary "$_uuid" ipv6.may-fail no &> >(ddebug)
fi
if [[ -v "ipv4_usage[$_netif]" ]] && [[ ! -v "ipv6_usage[$_netif]" ]]; then
nmcli connection modify --temporary "$_uuid" ipv6.method disabled &> >(ddebug)
elif [[ ! -v "ipv4_usage[$_netif]" ]] && [[ -v "ipv6_usage[$_netif]" ]]; then
nmcli connection modify --temporary "$_uuid" ipv4.method disabled &> >(ddebug)
fi
}
_clone_nmconnection() {
local _clone_output _name _unique_id
_unique_id=$1
_name=$(nmcli --get-values connection.id connection show "$_unique_id")
if _clone_output=$(nmcli connection clone --temporary uuid "$_unique_id" "$_name"); then
sed -E -n "s/.* \(.*\) cloned as.*\((.*)\)\.$/\1/p" <<< "$_clone_output"
return 0
fi
return 1
}
Address the cases where a NIC has a different name in kdump kernel A NIC may get a different name in the kdump kernel from 1st kernel in cases like, - kernel assigned network interface names are not persistent e.g. [1] - there is an udev rule to rename the NIC in the 1st kernel but the kdump initrd may not have that rule e.g. [2] If NM tries to match a NIC with a connection profile based on NIC name i.e. connection.interface-name, it will fail the above bases. A simple solution is to ask NM to match a connection profile by MAC address. Note we don't need to do this for user-created NICs like vlan, bridge and bond. An remaining issue is passing the name of a NIC via the kdumpnic dracut command line parameter which requires passing ifname=<interface>:<MAC> to have fixed NIC name. But we can simply drop this requirement. kdumpnic is needed because kdump needs to get the IP by NIC name and use the IP to created a dumping folder named "{IP}-{DATE}". We can simply pass the IP to the kdump kernel directly via a new dracut command line parameter kdumpip instead. In addition to the benefit of simplifying the code, there are other three benefits brought by this approach, - make use of whatever network to transfer the vmcore. Because as long as we have the network to we don't care which NIC is active. - if obtained IP in the kdump kernel is different from the one in the 1st kernel. "{IP}-{DATE}" would better tell where the dumped vmcore comes from. - without passing ifname=<interface>:<MAC> to kdump initrd, the issue of there are two interfaces with the same MAC address for Azure Hyper-V NIC SR-IOV [3] is resolved automatically. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1121778 [2] https://bugzilla.redhat.com/show_bug.cgi?id=810107 [3] https://bugzilla.redhat.com/show_bug.cgi?id=1962421 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 06:25:01 +00:00
_match_nmconnection_by_mac() {
local _unique_id _dev _mac _mac_field
_unique_id=$1
_dev=$2
_mac=$(kdump_get_perm_addr "$_dev")
[[ $_mac != 'not set' ]] || return
_mac_field=$(nmcli --get-values connection.type connection show "$_unique_id").mac-address
nmcli connection modify --temporary "$_unique_id" "$_mac_field" "$_mac" &> >(ddebug)
nmcli connection modify --temporary "$_unique_id" "connection.interface-name" "" &> >(ddebug)
}
# Clone and modify NM connection profiles
#
# This function makes use of "nmcli clone" to automatically convert ifcfg-*
# files to Networkmanager .nmconnection connection profiles and also modify the
# properties of .nmconnection if necessary.
clone_and_modify_nmconnection() {
local _dev _cloned_nmconnection_file_path _tmp_nmconnection_file_path _old_uuid _uuid
_dev=$1
_nmconnection_file_path=$2
_old_uuid=$(nmcli --get-values connection.uuid connection show filename "$_nmconnection_file_path")
if ! _uuid=$(_clone_nmconnection "$_old_uuid"); then
derror "Failed to clone $_old_uuid"
exit 1
fi
use_ipv4_or_ipv6 "$_dev" "$_uuid"
nmcli connection modify --temporary uuid "$_uuid" connection.wait-device-timeout 60000 &> >(ddebug)
Address the cases where a NIC has a different name in kdump kernel A NIC may get a different name in the kdump kernel from 1st kernel in cases like, - kernel assigned network interface names are not persistent e.g. [1] - there is an udev rule to rename the NIC in the 1st kernel but the kdump initrd may not have that rule e.g. [2] If NM tries to match a NIC with a connection profile based on NIC name i.e. connection.interface-name, it will fail the above bases. A simple solution is to ask NM to match a connection profile by MAC address. Note we don't need to do this for user-created NICs like vlan, bridge and bond. An remaining issue is passing the name of a NIC via the kdumpnic dracut command line parameter which requires passing ifname=<interface>:<MAC> to have fixed NIC name. But we can simply drop this requirement. kdumpnic is needed because kdump needs to get the IP by NIC name and use the IP to created a dumping folder named "{IP}-{DATE}". We can simply pass the IP to the kdump kernel directly via a new dracut command line parameter kdumpip instead. In addition to the benefit of simplifying the code, there are other three benefits brought by this approach, - make use of whatever network to transfer the vmcore. Because as long as we have the network to we don't care which NIC is active. - if obtained IP in the kdump kernel is different from the one in the 1st kernel. "{IP}-{DATE}" would better tell where the dumped vmcore comes from. - without passing ifname=<interface>:<MAC> to kdump initrd, the issue of there are two interfaces with the same MAC address for Azure Hyper-V NIC SR-IOV [3] is resolved automatically. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1121778 [2] https://bugzilla.redhat.com/show_bug.cgi?id=810107 [3] https://bugzilla.redhat.com/show_bug.cgi?id=1962421 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 06:25:01 +00:00
# For physical NIC i.e. non-user created NIC, ask NM to match a
# connection profile based on MAC address
_match_nmconnection_by_mac "$_uuid" "$_dev"
_cloned_nmconnection_file_path=$(nmcli --get-values UUID,FILENAME connection show | sed -n "s/^${_uuid}://p")
_tmp_nmconnection_file_path=$_DRACUT_KDUMP_NM_TMP_DIR/$(basename "$_nmconnection_file_path")
cp "$_cloned_nmconnection_file_path" "$_tmp_nmconnection_file_path"
# change uuid back to old value in case it's refered by other connection
# profile e.g. connection.master could be interface name of the master
# device or UUID of the master connection.
sed -i -E "s/(^uuid=).*$/\1${_old_uuid}/g" "$_tmp_nmconnection_file_path"
nmcli connection del "$_uuid" &> >(ddebug)
echo -n "$_tmp_nmconnection_file_path"
}
_install_nmconnection() {
local _src _nmconnection_name _dst
_src=$1
_nmconnection_name=$(basename "$_src")
_dst="/etc/NetworkManager/system-connections/$_nmconnection_name"
inst "$_src" "$_dst"
}
kdump_install_nmconnections() {
local _netif _nm_conn_path _cloned_nm_path
while IFS=: read -r _netif _nm_conn_path; do
[[ -v "unique_netifs[$_netif]" ]] || continue
if _cloned_nm_path=$(clone_and_modify_nmconnection "$_netif" "$_nm_conn_path"); then
_install_nmconnection "$_cloned_nm_path"
else
derror "Failed to install the .nmconnection for $_netif"
exit 1
fi
done <<< "$(nmcli -t -f device,filename connection show --active)"
# Stop dracut 35network-manger to calling nm-initrd-generator.
# Note this line of code can be removed after NetworkManager >= 1.35.2
# gets released.
echo > "${initdir}/usr/libexec/nm-initrd-generator"
}
kdump_install_nm_netif_allowlist() {
local _netif _except_netif _netif_allowlist _netif_allowlist_nm_conf
for _netif in $1; do
_per_mac=$(kdump_get_perm_addr "$_netif")
if [[ "$_per_mac" != 'not set' ]]; then
_except_netif="mac:$_per_mac"
else
_except_netif="interface-name:$_netif"
fi
_netif_allowlist="${_netif_allowlist}except:${_except_netif};"
done
_netif_allowlist_nm_conf=$_DRACUT_KDUMP_NM_TMP_DIR/netif_allowlist_nm_conf
cat << EOF > "$_netif_allowlist_nm_conf"
[device-others]
match-device=${_netif_allowlist}
managed=false
EOF
inst "$_netif_allowlist_nm_conf" "/etc/NetworkManager/conf.d/10-kdump-netif_allowlist.conf"
}
Reduce kdump memory consumption by only installing needed NIC drivers Even after having asked NM to stop managing a unneeded NIC, a NIC driver may still waste memory. For example, mlx5_core uses a substantial amount of memory during driver initialization, ======== Report format module_summary: ======== Module mlx5_core using 350.2MB (89650 pages), peak allocation 367.4MB (94056 pages) Module squashfs using 13.1MB (3360 pages), peak allocation 13.1MB (3360 pages) Module overlay using 2.1MB (550 pages), peak allocation 2.2MB (555 pages) Module dns_resolver using 0.9MB (219 pages), peak allocation 5.2MB (1338 pages) Module mlxfw using 0.7MB (172 pages), peak allocation 5.3MB (1349 pages) ======== Report format module_summary END ======== ======== Report format module_top: ======== Top stack usage of module mlx5_core: (null) Pages: 89650 (peak: 94056) ret_from_fork (0xffffda088b4165f8) Pages: 60007 (peak: 60007) kthread (0xffffda088b4bd7e4) Pages: 60007 (peak: 60007) worker_thread (0xffffda088b4b48d0) Pages: 60007 (peak: 60007) process_one_work (0xffffda088b4b3f40) Pages: 60007 (peak: 60007) work_for_cpu_fn (0xffffda088b4aef00) Pages: 53906 (peak: 53906) local_pci_probe (0xffffda088b9e1e44) Pages: 53906 (peak: 53906) probe_one mlx5_core (0xffffda084f899cc8) Pages: 53518 (peak: 53518) mlx5_init_one mlx5_core (0xffffda084f8994ac) Pages: 49756 (peak: 49756) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899100) Pages: 44434 (eak: 44434) mlx5_satisfy_startup_pages mlx5_core (0xffffda084f8a4f24) Pages: 44434 (peak: 44434) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899078) Pages: 5285 (peak: 5285) mlx5_cmd_init mlx5_core (0xffffda084f89e414) Pages: 4818 (peak: 4818) mlx5_alloc_cmd_msg mlx5_core (0xffffda084f89aaa0) Pages: 4403 (peak: 4403) This memory consumption is completely unnecessary when kdump doesn't need this NIC. Only install needed NIC drivers to prevent this kind of waste. Note 1. this patch depends on [1] to ask dracut to not install NIC drivers. 2. "ethtool -i" somehow fails to get the vlan driver 3. team.ko doesn't depend on the team mode drivers so we need to install the team mode drivers manually. [1] https://github.com/dracutdevs/dracut/pull/1789 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2022-05-19 03:39:25 +00:00
_get_nic_driver() {
ethtool -i "$1" | sed -n -E "s/driver: (.*)/\1/p"
}
_get_hpyerv_physical_driver() {
local _physical_nic
_physical_nic=$(find /sys/class/net/"$1"/ -name 'lower_*' | sed -En "s/\/.*lower_(.*)/\1/p")
[[ -n $_physical_nic ]] || return
_get_nic_driver "$_physical_nic"
}
Reduce kdump memory consumption by only installing needed NIC drivers Even after having asked NM to stop managing a unneeded NIC, a NIC driver may still waste memory. For example, mlx5_core uses a substantial amount of memory during driver initialization, ======== Report format module_summary: ======== Module mlx5_core using 350.2MB (89650 pages), peak allocation 367.4MB (94056 pages) Module squashfs using 13.1MB (3360 pages), peak allocation 13.1MB (3360 pages) Module overlay using 2.1MB (550 pages), peak allocation 2.2MB (555 pages) Module dns_resolver using 0.9MB (219 pages), peak allocation 5.2MB (1338 pages) Module mlxfw using 0.7MB (172 pages), peak allocation 5.3MB (1349 pages) ======== Report format module_summary END ======== ======== Report format module_top: ======== Top stack usage of module mlx5_core: (null) Pages: 89650 (peak: 94056) ret_from_fork (0xffffda088b4165f8) Pages: 60007 (peak: 60007) kthread (0xffffda088b4bd7e4) Pages: 60007 (peak: 60007) worker_thread (0xffffda088b4b48d0) Pages: 60007 (peak: 60007) process_one_work (0xffffda088b4b3f40) Pages: 60007 (peak: 60007) work_for_cpu_fn (0xffffda088b4aef00) Pages: 53906 (peak: 53906) local_pci_probe (0xffffda088b9e1e44) Pages: 53906 (peak: 53906) probe_one mlx5_core (0xffffda084f899cc8) Pages: 53518 (peak: 53518) mlx5_init_one mlx5_core (0xffffda084f8994ac) Pages: 49756 (peak: 49756) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899100) Pages: 44434 (eak: 44434) mlx5_satisfy_startup_pages mlx5_core (0xffffda084f8a4f24) Pages: 44434 (peak: 44434) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899078) Pages: 5285 (peak: 5285) mlx5_cmd_init mlx5_core (0xffffda084f89e414) Pages: 4818 (peak: 4818) mlx5_alloc_cmd_msg mlx5_core (0xffffda084f89aaa0) Pages: 4403 (peak: 4403) This memory consumption is completely unnecessary when kdump doesn't need this NIC. Only install needed NIC drivers to prevent this kind of waste. Note 1. this patch depends on [1] to ask dracut to not install NIC drivers. 2. "ethtool -i" somehow fails to get the vlan driver 3. team.ko doesn't depend on the team mode drivers so we need to install the team mode drivers manually. [1] https://github.com/dracutdevs/dracut/pull/1789 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2022-05-19 03:39:25 +00:00
kdump_install_nic_driver() {
local _netif _driver _drivers
_drivers=()
for _netif in $1; do
dracut-module-setup.sh: skip installing driver for the loopback interface Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2151500 Currently, kdump initrd fails to be built when dumping vmcore to localhost via ssh or nfs, kdumpctl[3331]: Cannot get driver information: Operation not supported kdumpctl[1991]: dracut: Failed to get the driver of lo dracut[2020]: Failed to get the driver of lo kdumpctl[1775]: kdump: mkdumprd: failed to make kdump initrd kdumpctl[1775]: kdump: Starting kdump: [FAILED] systemd[1]: kdump.service: Main process exited, code=exited, status=1/FAILURE systemd[1]: kdump.service: Failed with result 'exit-code'. systemd[1]: Failed to start Crash recovery kernel arming. systemd[1]: kdump.service: Consumed 1.710s CPU time. This is because the loopback interface is used for transferring vmcore and ethtool can't get the driver of the loopback interface. In fact, once COFNIG_NET is enabled, the loopback device is enabled and there is no driver for the loopback device. So skip installing driver for the loopback device. The loopback interface is implemented in linux/drivers/net/loopback.c and always has the name "lo". So we can safely tell if a network interface is the loopback interface by its name. Fixes: a65dde2d ("Reduce kdump memory consumption by only installing needed NIC drivers") Reported-by: Martin Pitt <mpitt@redhat.com> Reported-by: Rich Megginson <rmeggins@redhat.com> Reviewed-by: Lichen Liu <lichliu@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com>
2022-12-14 02:12:17 +00:00
[[ $_netif == lo ]] && continue
Reduce kdump memory consumption by only installing needed NIC drivers Even after having asked NM to stop managing a unneeded NIC, a NIC driver may still waste memory. For example, mlx5_core uses a substantial amount of memory during driver initialization, ======== Report format module_summary: ======== Module mlx5_core using 350.2MB (89650 pages), peak allocation 367.4MB (94056 pages) Module squashfs using 13.1MB (3360 pages), peak allocation 13.1MB (3360 pages) Module overlay using 2.1MB (550 pages), peak allocation 2.2MB (555 pages) Module dns_resolver using 0.9MB (219 pages), peak allocation 5.2MB (1338 pages) Module mlxfw using 0.7MB (172 pages), peak allocation 5.3MB (1349 pages) ======== Report format module_summary END ======== ======== Report format module_top: ======== Top stack usage of module mlx5_core: (null) Pages: 89650 (peak: 94056) ret_from_fork (0xffffda088b4165f8) Pages: 60007 (peak: 60007) kthread (0xffffda088b4bd7e4) Pages: 60007 (peak: 60007) worker_thread (0xffffda088b4b48d0) Pages: 60007 (peak: 60007) process_one_work (0xffffda088b4b3f40) Pages: 60007 (peak: 60007) work_for_cpu_fn (0xffffda088b4aef00) Pages: 53906 (peak: 53906) local_pci_probe (0xffffda088b9e1e44) Pages: 53906 (peak: 53906) probe_one mlx5_core (0xffffda084f899cc8) Pages: 53518 (peak: 53518) mlx5_init_one mlx5_core (0xffffda084f8994ac) Pages: 49756 (peak: 49756) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899100) Pages: 44434 (eak: 44434) mlx5_satisfy_startup_pages mlx5_core (0xffffda084f8a4f24) Pages: 44434 (peak: 44434) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899078) Pages: 5285 (peak: 5285) mlx5_cmd_init mlx5_core (0xffffda084f89e414) Pages: 4818 (peak: 4818) mlx5_alloc_cmd_msg mlx5_core (0xffffda084f89aaa0) Pages: 4403 (peak: 4403) This memory consumption is completely unnecessary when kdump doesn't need this NIC. Only install needed NIC drivers to prevent this kind of waste. Note 1. this patch depends on [1] to ask dracut to not install NIC drivers. 2. "ethtool -i" somehow fails to get the vlan driver 3. team.ko doesn't depend on the team mode drivers so we need to install the team mode drivers manually. [1] https://github.com/dracutdevs/dracut/pull/1789 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2022-05-19 03:39:25 +00:00
_driver=$(_get_nic_driver "$_netif")
if [[ -z $_driver ]]; then
derror "Failed to get the driver of $_netif"
exit 1
fi
if [[ $_driver == "802.1Q VLAN Support" ]]; then
# ethtool somehow doesn't return the driver name for a VLAN NIC
_driver=8021q
elif [[ $_driver == "team" ]]; then
# install the team mode drivers like team_mode_roundrobin.ko as well
_driver='=drivers/net/team'
elif [[ $_driver == "hv_netvsc" ]]; then
# A Hyper-V VM may have accelerated networking
# https://learn.microsoft.com/en-us/azure/virtual-network/accelerated-networking-overview
# Install the driver of physical NIC as well
_drivers+=("$(_get_hpyerv_physical_driver "$_netif")")
Reduce kdump memory consumption by only installing needed NIC drivers Even after having asked NM to stop managing a unneeded NIC, a NIC driver may still waste memory. For example, mlx5_core uses a substantial amount of memory during driver initialization, ======== Report format module_summary: ======== Module mlx5_core using 350.2MB (89650 pages), peak allocation 367.4MB (94056 pages) Module squashfs using 13.1MB (3360 pages), peak allocation 13.1MB (3360 pages) Module overlay using 2.1MB (550 pages), peak allocation 2.2MB (555 pages) Module dns_resolver using 0.9MB (219 pages), peak allocation 5.2MB (1338 pages) Module mlxfw using 0.7MB (172 pages), peak allocation 5.3MB (1349 pages) ======== Report format module_summary END ======== ======== Report format module_top: ======== Top stack usage of module mlx5_core: (null) Pages: 89650 (peak: 94056) ret_from_fork (0xffffda088b4165f8) Pages: 60007 (peak: 60007) kthread (0xffffda088b4bd7e4) Pages: 60007 (peak: 60007) worker_thread (0xffffda088b4b48d0) Pages: 60007 (peak: 60007) process_one_work (0xffffda088b4b3f40) Pages: 60007 (peak: 60007) work_for_cpu_fn (0xffffda088b4aef00) Pages: 53906 (peak: 53906) local_pci_probe (0xffffda088b9e1e44) Pages: 53906 (peak: 53906) probe_one mlx5_core (0xffffda084f899cc8) Pages: 53518 (peak: 53518) mlx5_init_one mlx5_core (0xffffda084f8994ac) Pages: 49756 (peak: 49756) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899100) Pages: 44434 (eak: 44434) mlx5_satisfy_startup_pages mlx5_core (0xffffda084f8a4f24) Pages: 44434 (peak: 44434) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899078) Pages: 5285 (peak: 5285) mlx5_cmd_init mlx5_core (0xffffda084f89e414) Pages: 4818 (peak: 4818) mlx5_alloc_cmd_msg mlx5_core (0xffffda084f89aaa0) Pages: 4403 (peak: 4403) This memory consumption is completely unnecessary when kdump doesn't need this NIC. Only install needed NIC drivers to prevent this kind of waste. Note 1. this patch depends on [1] to ask dracut to not install NIC drivers. 2. "ethtool -i" somehow fails to get the vlan driver 3. team.ko doesn't depend on the team mode drivers so we need to install the team mode drivers manually. [1] https://github.com/dracutdevs/dracut/pull/1789 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2022-05-19 03:39:25 +00:00
fi
_drivers+=("$_driver")
done
dracut-module-setup.sh: skip installing driver for the loopback interface Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2151500 Currently, kdump initrd fails to be built when dumping vmcore to localhost via ssh or nfs, kdumpctl[3331]: Cannot get driver information: Operation not supported kdumpctl[1991]: dracut: Failed to get the driver of lo dracut[2020]: Failed to get the driver of lo kdumpctl[1775]: kdump: mkdumprd: failed to make kdump initrd kdumpctl[1775]: kdump: Starting kdump: [FAILED] systemd[1]: kdump.service: Main process exited, code=exited, status=1/FAILURE systemd[1]: kdump.service: Failed with result 'exit-code'. systemd[1]: Failed to start Crash recovery kernel arming. systemd[1]: kdump.service: Consumed 1.710s CPU time. This is because the loopback interface is used for transferring vmcore and ethtool can't get the driver of the loopback interface. In fact, once COFNIG_NET is enabled, the loopback device is enabled and there is no driver for the loopback device. So skip installing driver for the loopback device. The loopback interface is implemented in linux/drivers/net/loopback.c and always has the name "lo". So we can safely tell if a network interface is the loopback interface by its name. Fixes: a65dde2d ("Reduce kdump memory consumption by only installing needed NIC drivers") Reported-by: Martin Pitt <mpitt@redhat.com> Reported-by: Rich Megginson <rmeggins@redhat.com> Reviewed-by: Lichen Liu <lichliu@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com> Signed-off-by: Coiby Xu <coxu@redhat.com>
2022-12-14 02:12:17 +00:00
[[ -n ${_drivers[*]} ]] || return
Reduce kdump memory consumption by only installing needed NIC drivers Even after having asked NM to stop managing a unneeded NIC, a NIC driver may still waste memory. For example, mlx5_core uses a substantial amount of memory during driver initialization, ======== Report format module_summary: ======== Module mlx5_core using 350.2MB (89650 pages), peak allocation 367.4MB (94056 pages) Module squashfs using 13.1MB (3360 pages), peak allocation 13.1MB (3360 pages) Module overlay using 2.1MB (550 pages), peak allocation 2.2MB (555 pages) Module dns_resolver using 0.9MB (219 pages), peak allocation 5.2MB (1338 pages) Module mlxfw using 0.7MB (172 pages), peak allocation 5.3MB (1349 pages) ======== Report format module_summary END ======== ======== Report format module_top: ======== Top stack usage of module mlx5_core: (null) Pages: 89650 (peak: 94056) ret_from_fork (0xffffda088b4165f8) Pages: 60007 (peak: 60007) kthread (0xffffda088b4bd7e4) Pages: 60007 (peak: 60007) worker_thread (0xffffda088b4b48d0) Pages: 60007 (peak: 60007) process_one_work (0xffffda088b4b3f40) Pages: 60007 (peak: 60007) work_for_cpu_fn (0xffffda088b4aef00) Pages: 53906 (peak: 53906) local_pci_probe (0xffffda088b9e1e44) Pages: 53906 (peak: 53906) probe_one mlx5_core (0xffffda084f899cc8) Pages: 53518 (peak: 53518) mlx5_init_one mlx5_core (0xffffda084f8994ac) Pages: 49756 (peak: 49756) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899100) Pages: 44434 (eak: 44434) mlx5_satisfy_startup_pages mlx5_core (0xffffda084f8a4f24) Pages: 44434 (peak: 44434) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899078) Pages: 5285 (peak: 5285) mlx5_cmd_init mlx5_core (0xffffda084f89e414) Pages: 4818 (peak: 4818) mlx5_alloc_cmd_msg mlx5_core (0xffffda084f89aaa0) Pages: 4403 (peak: 4403) This memory consumption is completely unnecessary when kdump doesn't need this NIC. Only install needed NIC drivers to prevent this kind of waste. Note 1. this patch depends on [1] to ask dracut to not install NIC drivers. 2. "ethtool -i" somehow fails to get the vlan driver 3. team.ko doesn't depend on the team mode drivers so we need to install the team mode drivers manually. [1] https://github.com/dracutdevs/dracut/pull/1789 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2022-05-19 03:39:25 +00:00
instmods "${_drivers[@]}"
}
kdump_setup_bridge() {
local _netdev=$1
local _dev
for _dev in "/sys/class/net/$_netdev/brif/"*; do
[[ -e $_dev ]] || continue
_dev=${_dev##*/}
if kdump_is_bond "$_dev"; then
kdump_setup_bond "$_dev" || return 1
elif kdump_is_team "$_dev"; then
kdump_setup_team "$_dev"
elif kdump_is_vlan "$_dev"; then
kdump_setup_vlan "$_dev"
fi
_save_kdump_netifs "$_dev"
done
}
kdump_setup_bond() {
local _netdev="$1"
local _dev
for _dev in $(< "/sys/class/net/$_netdev/bonding/slaves"); do
_save_kdump_netifs "$_dev"
done
}
kdump_setup_team() {
local _netdev=$1
local _dev
for _dev in $(teamnl "$_netdev" ports | awk -F':' '{print $2}'); do
_save_kdump_netifs "$_dev"
done
}
kdump_setup_vlan() {
local _netdev=$1
local _parent_netif
_parent_netif="$(awk '/^Device:/{print $2}' /proc/net/vlan/"$_netdev")"
#Just support vlan over bond and team
if kdump_is_bridge "$_parent_netif"; then
derror "Vlan over bridge is not supported!"
exit 1
elif kdump_is_bond "$_parent_netif"; then
kdump_setup_bond "$_parent_netif" || return 1
elif kdump_is_team "$_parent_netif"; then
kdump_setup_team "$_parent_netif" || return 1
fi
_save_kdump_netifs "$_parent_netif"
}
Simplify setup_znet by copying connection profile to initrd /usr/lib/udev/ccw_init [1] shipped by s390utils extracts the values of SUBCHANNELS, NETTYPE and LAYER2 from /etc/sysconfig/network-scripts/ifcfg-* or /etc/NetworkManager/system-connections/*.nmconnection to activate znet network device. If the connection profile is copied to initrd, there is no need to set up the "rd.znet" dracut cmdline parameter. There are two cases addressed by this commit, 1. znet network interface is a slave of bonding/teaming/vlan/bridging network. The connection profile has been copied to initrd by kdump_copy_nmconnection_file and it contains the info needed by ccw_init. 2. znet network interface is a slave of bonding/teaming/vlan/bridging network. The corresponding ifcfg-*/*.nmconnection file may not contain info like SUBCHANNELS [2]. In this case, copy the ifcfg-*/*.nmconnection file that has this info to the kdump initrd. Also to prevent the copied connection profile from being chosen by NM, set connection.autoconnect=false for this connection profile. With this implementation, there is also no need to check if znet is used beforehand. Note 1. ccw_init doesn't care if SUBCHANNELS, NETTYPE and LAYER2 comes from an active NM profile or not. If an inactive NM profile contains this info, it needs to be copied to the kdump initrd as well. 2. "rd.znet_ifname=$_netdev:${SUBCHANNELS}" is no longer needed needed because now there is no renaming of s390x network interfaces when reusing NetworkManager profiles. rd.znet_ifname was introduced in commit ce0305d ("Add a new option 'rd.znet_ifname' in order to use it in udev rules") to address the special case of non-persistent MAC address by renaming a network interface by SUBCHANNELS. [1] https://src.fedoraproject.org/rpms/s390utils/blob/rawhide/f/ccw_init [2] https://bugzilla.redhat.com/show_bug.cgi?id=2064708 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 07:26:00 +00:00
_find_znet_nmconnection() {
LANG=C grep -s -E -i -l \
"^s390-subchannels=([0-9]\.[0-9]\.[a-f0-9]+;){0,2}" \
"$1"/*.nmconnection | LC_ALL=C sed -e "$2"
Iterate /sys/bus/ccwgroup/devices to tell if we should set up rd.znet This patch fixes bz1941106 and bz1941905 which passed empty rd.znet to the kernel command line in the following cases, - The IBM (Z15) KVM guest uses virtio for all devices including network device, so there is no znet device for IBM KVM guest. So we can't assume a s390x machine always has a znet device. - When a bridged network is used, kexec-tools tries to obtain the znet configuration from the ifcfg script of the bridged network rather than from the ifcfg script of znet device. We can iterate /sys/bus/ccwgroup/devices to tell if there if there is a znet network device. By getting an ifname from znet, we can also avoid mistaking the slave netdev as a znet network device in a bridged network or bonded network. Note: This patch also assumes there is only one znet device as commit 7148c0a30dfc48221eadf255e8a89619f98a8752 ("add s390x netdev setup") which greatly simplifies the code. According to IBM [1], there could be more than znet devices for a z/VM system and a z/VM system may have a non-znet network device like ConnectX. Since kdump_setup_znet was introduced in 2012 and so far there is no known customer complaint that invalidates this assumption I think it's safe to assume an IBM z/VM system only has one znet device. Besides, there is no z/VM system found on beaker to test the alternative scenarios. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1941905#c13 Signed-off-by: Coiby Xu <coxu@redhat.com> Acked-by: Kairui Song <kasong@redhat.com>
2021-06-06 23:26:03 +00:00
}
Simplify setup_znet by copying connection profile to initrd /usr/lib/udev/ccw_init [1] shipped by s390utils extracts the values of SUBCHANNELS, NETTYPE and LAYER2 from /etc/sysconfig/network-scripts/ifcfg-* or /etc/NetworkManager/system-connections/*.nmconnection to activate znet network device. If the connection profile is copied to initrd, there is no need to set up the "rd.znet" dracut cmdline parameter. There are two cases addressed by this commit, 1. znet network interface is a slave of bonding/teaming/vlan/bridging network. The connection profile has been copied to initrd by kdump_copy_nmconnection_file and it contains the info needed by ccw_init. 2. znet network interface is a slave of bonding/teaming/vlan/bridging network. The corresponding ifcfg-*/*.nmconnection file may not contain info like SUBCHANNELS [2]. In this case, copy the ifcfg-*/*.nmconnection file that has this info to the kdump initrd. Also to prevent the copied connection profile from being chosen by NM, set connection.autoconnect=false for this connection profile. With this implementation, there is also no need to check if znet is used beforehand. Note 1. ccw_init doesn't care if SUBCHANNELS, NETTYPE and LAYER2 comes from an active NM profile or not. If an inactive NM profile contains this info, it needs to be copied to the kdump initrd as well. 2. "rd.znet_ifname=$_netdev:${SUBCHANNELS}" is no longer needed needed because now there is no renaming of s390x network interfaces when reusing NetworkManager profiles. rd.znet_ifname was introduced in commit ce0305d ("Add a new option 'rd.znet_ifname' in order to use it in udev rules") to address the special case of non-persistent MAC address by renaming a network interface by SUBCHANNELS. [1] https://src.fedoraproject.org/rpms/s390utils/blob/rawhide/f/ccw_init [2] https://bugzilla.redhat.com/show_bug.cgi?id=2064708 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 07:26:00 +00:00
# setup s390 znet
#
# Note part of code is extracted from ccw_init provided by s390utils
kdump_setup_znet() {
Simplify setup_znet by copying connection profile to initrd /usr/lib/udev/ccw_init [1] shipped by s390utils extracts the values of SUBCHANNELS, NETTYPE and LAYER2 from /etc/sysconfig/network-scripts/ifcfg-* or /etc/NetworkManager/system-connections/*.nmconnection to activate znet network device. If the connection profile is copied to initrd, there is no need to set up the "rd.znet" dracut cmdline parameter. There are two cases addressed by this commit, 1. znet network interface is a slave of bonding/teaming/vlan/bridging network. The connection profile has been copied to initrd by kdump_copy_nmconnection_file and it contains the info needed by ccw_init. 2. znet network interface is a slave of bonding/teaming/vlan/bridging network. The corresponding ifcfg-*/*.nmconnection file may not contain info like SUBCHANNELS [2]. In this case, copy the ifcfg-*/*.nmconnection file that has this info to the kdump initrd. Also to prevent the copied connection profile from being chosen by NM, set connection.autoconnect=false for this connection profile. With this implementation, there is also no need to check if znet is used beforehand. Note 1. ccw_init doesn't care if SUBCHANNELS, NETTYPE and LAYER2 comes from an active NM profile or not. If an inactive NM profile contains this info, it needs to be copied to the kdump initrd as well. 2. "rd.znet_ifname=$_netdev:${SUBCHANNELS}" is no longer needed needed because now there is no renaming of s390x network interfaces when reusing NetworkManager profiles. rd.znet_ifname was introduced in commit ce0305d ("Add a new option 'rd.znet_ifname' in order to use it in udev rules") to address the special case of non-persistent MAC address by renaming a network interface by SUBCHANNELS. [1] https://src.fedoraproject.org/rpms/s390utils/blob/rawhide/f/ccw_init [2] https://bugzilla.redhat.com/show_bug.cgi?id=2064708 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 07:26:00 +00:00
local _config_file _unique_name _NM_conf_dir
local __sed_discard_ignored_files='/\(~\|\.bak\|\.old\|\.orig\|\.rpmnew\|\.rpmorig\|\.rpmsave\)$/d'
if [[ "$(uname -m)" != "s390x" ]]; then
return
fi
Simplify setup_znet by copying connection profile to initrd /usr/lib/udev/ccw_init [1] shipped by s390utils extracts the values of SUBCHANNELS, NETTYPE and LAYER2 from /etc/sysconfig/network-scripts/ifcfg-* or /etc/NetworkManager/system-connections/*.nmconnection to activate znet network device. If the connection profile is copied to initrd, there is no need to set up the "rd.znet" dracut cmdline parameter. There are two cases addressed by this commit, 1. znet network interface is a slave of bonding/teaming/vlan/bridging network. The connection profile has been copied to initrd by kdump_copy_nmconnection_file and it contains the info needed by ccw_init. 2. znet network interface is a slave of bonding/teaming/vlan/bridging network. The corresponding ifcfg-*/*.nmconnection file may not contain info like SUBCHANNELS [2]. In this case, copy the ifcfg-*/*.nmconnection file that has this info to the kdump initrd. Also to prevent the copied connection profile from being chosen by NM, set connection.autoconnect=false for this connection profile. With this implementation, there is also no need to check if znet is used beforehand. Note 1. ccw_init doesn't care if SUBCHANNELS, NETTYPE and LAYER2 comes from an active NM profile or not. If an inactive NM profile contains this info, it needs to be copied to the kdump initrd as well. 2. "rd.znet_ifname=$_netdev:${SUBCHANNELS}" is no longer needed needed because now there is no renaming of s390x network interfaces when reusing NetworkManager profiles. rd.znet_ifname was introduced in commit ce0305d ("Add a new option 'rd.znet_ifname' in order to use it in udev rules") to address the special case of non-persistent MAC address by renaming a network interface by SUBCHANNELS. [1] https://src.fedoraproject.org/rpms/s390utils/blob/rawhide/f/ccw_init [2] https://bugzilla.redhat.com/show_bug.cgi?id=2064708 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 07:26:00 +00:00
_NM_conf_dir="/etc/NetworkManager/system-connections"
_config_file=$(_find_znet_nmconnection "$initdir/$_NM_conf_dir" "$__sed_discard_ignored_files")
if [[ -n "$_config_file" ]]; then
ddebug "$_config_file has already contained the znet config"
return
fi
_config_file=$(LANG=C grep -s -E -i -l \
"^[[:space:]]*SUBCHANNELS=['\"]?([0-9]\.[0-9]\.[a-f0-9]+,){0,2}" \
/etc/sysconfig/network-scripts/ifcfg-* \
| LC_ALL=C sed -e "$__sed_discard_ignored_files")
if [[ -z "$_config_file" ]]; then
_config_file=$(_find_znet_nmconnection "$_NM_conf_dir" "$__sed_discard_ignored_files")
fi
if [[ -n "$_config_file" ]]; then
_unique_name=$(cat /proc/sys/kernel/random/uuid)
nmcli connection clone --temporary "$_config_file" "$_unique_name" &> >(ddebug)
nmcli connection modify --temporary "$_unique_name" connection.autoconnect false
inst "/run/NetworkManager/system-connections/${_unique_name}.nmconnection" "${_NM_conf_dir}/${_unique_name}.nmconnection"
nmcli connection del "$_unique_name" &> >(ddebug)
fi
}
kdump_get_remote_ip() {
local _remote _remote_temp
_remote=$(get_remote_host "$1")
if is_hostname "$_remote"; then
_remote_temp=$(getent ahosts "$_remote" | grep -v : | head -n 1)
if [[ -z $_remote_temp ]]; then
_remote_temp=$(getent ahosts "$_remote" | head -n 1)
fi
_remote=$(echo "$_remote_temp" | awk '{print $1}')
fi
echo "$_remote"
}
# Collect netifs needed by kdump
# $1: destination host
kdump_collect_netif_usage() {
Address the cases where a NIC has a different name in kdump kernel A NIC may get a different name in the kdump kernel from 1st kernel in cases like, - kernel assigned network interface names are not persistent e.g. [1] - there is an udev rule to rename the NIC in the 1st kernel but the kdump initrd may not have that rule e.g. [2] If NM tries to match a NIC with a connection profile based on NIC name i.e. connection.interface-name, it will fail the above bases. A simple solution is to ask NM to match a connection profile by MAC address. Note we don't need to do this for user-created NICs like vlan, bridge and bond. An remaining issue is passing the name of a NIC via the kdumpnic dracut command line parameter which requires passing ifname=<interface>:<MAC> to have fixed NIC name. But we can simply drop this requirement. kdumpnic is needed because kdump needs to get the IP by NIC name and use the IP to created a dumping folder named "{IP}-{DATE}". We can simply pass the IP to the kdump kernel directly via a new dracut command line parameter kdumpip instead. In addition to the benefit of simplifying the code, there are other three benefits brought by this approach, - make use of whatever network to transfer the vmcore. Because as long as we have the network to we don't care which NIC is active. - if obtained IP in the kdump kernel is different from the one in the 1st kernel. "{IP}-{DATE}" would better tell where the dumped vmcore comes from. - without passing ifname=<interface>:<MAC> to kdump initrd, the issue of there are two interfaces with the same MAC address for Azure Hyper-V NIC SR-IOV [3] is resolved automatically. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1121778 [2] https://bugzilla.redhat.com/show_bug.cgi?id=810107 [3] https://bugzilla.redhat.com/show_bug.cgi?id=1962421 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 06:25:01 +00:00
local _destaddr _srcaddr _route _netdev
_destaddr=$(kdump_get_remote_ip "$1")
_route=$(kdump_get_ip_route "$_destaddr")
_srcaddr=$(kdump_get_ip_route_field "$_route" "src")
_netdev=$(kdump_get_ip_route_field "$_route" "dev")
if kdump_is_bridge "$_netdev"; then
kdump_setup_bridge "$_netdev"
elif kdump_is_bond "$_netdev"; then
kdump_setup_bond "$_netdev" || return 1
elif kdump_is_team "$_netdev"; then
kdump_setup_team "$_netdev"
elif kdump_is_vlan "$_netdev"; then
kdump_setup_vlan "$_netdev"
fi
_save_kdump_netifs "$_netdev"
if [[ ! -f ${initdir}/etc/cmdline.d/50neednet.conf ]]; then
# network-manager module needs this parameter
echo "rd.neednet" >> "${initdir}/etc/cmdline.d/50neednet.conf"
fi
Address the cases where a NIC has a different name in kdump kernel A NIC may get a different name in the kdump kernel from 1st kernel in cases like, - kernel assigned network interface names are not persistent e.g. [1] - there is an udev rule to rename the NIC in the 1st kernel but the kdump initrd may not have that rule e.g. [2] If NM tries to match a NIC with a connection profile based on NIC name i.e. connection.interface-name, it will fail the above bases. A simple solution is to ask NM to match a connection profile by MAC address. Note we don't need to do this for user-created NICs like vlan, bridge and bond. An remaining issue is passing the name of a NIC via the kdumpnic dracut command line parameter which requires passing ifname=<interface>:<MAC> to have fixed NIC name. But we can simply drop this requirement. kdumpnic is needed because kdump needs to get the IP by NIC name and use the IP to created a dumping folder named "{IP}-{DATE}". We can simply pass the IP to the kdump kernel directly via a new dracut command line parameter kdumpip instead. In addition to the benefit of simplifying the code, there are other three benefits brought by this approach, - make use of whatever network to transfer the vmcore. Because as long as we have the network to we don't care which NIC is active. - if obtained IP in the kdump kernel is different from the one in the 1st kernel. "{IP}-{DATE}" would better tell where the dumped vmcore comes from. - without passing ifname=<interface>:<MAC> to kdump initrd, the issue of there are two interfaces with the same MAC address for Azure Hyper-V NIC SR-IOV [3] is resolved automatically. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1121778 [2] https://bugzilla.redhat.com/show_bug.cgi?id=810107 [3] https://bugzilla.redhat.com/show_bug.cgi?id=1962421 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 06:25:01 +00:00
if [[ ! -f ${initdir}/etc/cmdline.d/60kdumpip.conf ]]; then
echo "kdump_remote_ip=$_destaddr" > "${initdir}/etc/cmdline.d/60kdumpip.conf"
fi
if is_ipv6_address "$_srcaddr"; then
ipv6_usage[$_netdev]=1
else
ipv4_usage[$_netdev]=1
fi
}
# Setup dracut to bring up network interface that enable
# initramfs accessing giving destination
kdump_install_net() {
local _netifs
_netifs=$(_get_kdump_netifs)
if [[ -n "$_netifs" ]]; then
kdump_install_nmconnections
apply_nm_initrd_generator_timeouts
Simplify setup_znet by copying connection profile to initrd /usr/lib/udev/ccw_init [1] shipped by s390utils extracts the values of SUBCHANNELS, NETTYPE and LAYER2 from /etc/sysconfig/network-scripts/ifcfg-* or /etc/NetworkManager/system-connections/*.nmconnection to activate znet network device. If the connection profile is copied to initrd, there is no need to set up the "rd.znet" dracut cmdline parameter. There are two cases addressed by this commit, 1. znet network interface is a slave of bonding/teaming/vlan/bridging network. The connection profile has been copied to initrd by kdump_copy_nmconnection_file and it contains the info needed by ccw_init. 2. znet network interface is a slave of bonding/teaming/vlan/bridging network. The corresponding ifcfg-*/*.nmconnection file may not contain info like SUBCHANNELS [2]. In this case, copy the ifcfg-*/*.nmconnection file that has this info to the kdump initrd. Also to prevent the copied connection profile from being chosen by NM, set connection.autoconnect=false for this connection profile. With this implementation, there is also no need to check if znet is used beforehand. Note 1. ccw_init doesn't care if SUBCHANNELS, NETTYPE and LAYER2 comes from an active NM profile or not. If an inactive NM profile contains this info, it needs to be copied to the kdump initrd as well. 2. "rd.znet_ifname=$_netdev:${SUBCHANNELS}" is no longer needed needed because now there is no renaming of s390x network interfaces when reusing NetworkManager profiles. rd.znet_ifname was introduced in commit ce0305d ("Add a new option 'rd.znet_ifname' in order to use it in udev rules") to address the special case of non-persistent MAC address by renaming a network interface by SUBCHANNELS. [1] https://src.fedoraproject.org/rpms/s390utils/blob/rawhide/f/ccw_init [2] https://bugzilla.redhat.com/show_bug.cgi?id=2064708 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2021-09-23 07:26:00 +00:00
kdump_setup_znet
kdump_install_nm_netif_allowlist "$_netifs"
Reduce kdump memory consumption by only installing needed NIC drivers Even after having asked NM to stop managing a unneeded NIC, a NIC driver may still waste memory. For example, mlx5_core uses a substantial amount of memory during driver initialization, ======== Report format module_summary: ======== Module mlx5_core using 350.2MB (89650 pages), peak allocation 367.4MB (94056 pages) Module squashfs using 13.1MB (3360 pages), peak allocation 13.1MB (3360 pages) Module overlay using 2.1MB (550 pages), peak allocation 2.2MB (555 pages) Module dns_resolver using 0.9MB (219 pages), peak allocation 5.2MB (1338 pages) Module mlxfw using 0.7MB (172 pages), peak allocation 5.3MB (1349 pages) ======== Report format module_summary END ======== ======== Report format module_top: ======== Top stack usage of module mlx5_core: (null) Pages: 89650 (peak: 94056) ret_from_fork (0xffffda088b4165f8) Pages: 60007 (peak: 60007) kthread (0xffffda088b4bd7e4) Pages: 60007 (peak: 60007) worker_thread (0xffffda088b4b48d0) Pages: 60007 (peak: 60007) process_one_work (0xffffda088b4b3f40) Pages: 60007 (peak: 60007) work_for_cpu_fn (0xffffda088b4aef00) Pages: 53906 (peak: 53906) local_pci_probe (0xffffda088b9e1e44) Pages: 53906 (peak: 53906) probe_one mlx5_core (0xffffda084f899cc8) Pages: 53518 (peak: 53518) mlx5_init_one mlx5_core (0xffffda084f8994ac) Pages: 49756 (peak: 49756) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899100) Pages: 44434 (eak: 44434) mlx5_satisfy_startup_pages mlx5_core (0xffffda084f8a4f24) Pages: 44434 (peak: 44434) mlx5_function_setup.constprop.0 mlx5_core (0xffffda084f899078) Pages: 5285 (peak: 5285) mlx5_cmd_init mlx5_core (0xffffda084f89e414) Pages: 4818 (peak: 4818) mlx5_alloc_cmd_msg mlx5_core (0xffffda084f89aaa0) Pages: 4403 (peak: 4403) This memory consumption is completely unnecessary when kdump doesn't need this NIC. Only install needed NIC drivers to prevent this kind of waste. Note 1. this patch depends on [1] to ask dracut to not install NIC drivers. 2. "ethtool -i" somehow fails to get the vlan driver 3. team.ko doesn't depend on the team mode drivers so we need to install the team mode drivers manually. [1] https://github.com/dracutdevs/dracut/pull/1789 Signed-off-by: Coiby Xu <coxu@redhat.com> Reviewed-by: Thomas Haller <thaller@redhat.com> Reviewed-by: Philipp Rudo <prudo@redhat.com>
2022-05-19 03:39:25 +00:00
kdump_install_nic_driver "$_netifs"
fi
}
# install etc/kdump/pre.d and /etc/kdump/post.d
kdump_install_pre_post_conf() {
if [[ -d /etc/kdump/pre.d ]]; then
for file in /etc/kdump/pre.d/*; do
if [[ -x $file ]]; then
dracut_install "$file"
elif [[ $file != "/etc/kdump/pre.d/*" ]]; then
echo "$file is not executable"
fi
done
fi
if [[ -d /etc/kdump/post.d ]]; then
for file in /etc/kdump/post.d/*; do
if [[ -x $file ]]; then
dracut_install "$file"
elif [[ $file != "/etc/kdump/post.d/*" ]]; then
echo "$file is not executable"
fi
done
fi
}
default_dump_target_install_conf() {
local _target _fstype
local _mntpoint _save_path
is_user_configured_dump_target && return
_save_path=$(get_bind_mount_source "$(get_save_path)")
_target=$(get_target_from_path "$_save_path")
_mntpoint=$(get_mntpoint_from_target "$_target")
_fstype=$(get_fs_type_from_target "$_target")
if is_fs_type_nfs "$_fstype"; then
kdump_collect_netif_usage "$_target"
_fstype="nfs"
else
_target=$(kdump_get_persistent_dev "$_target")
fi
echo "$_fstype $_target" >> "${initdir}/tmp/$$-kdump.conf"
# don't touch the path under root mount
if [[ $_mntpoint != "/" ]]; then
_save_path=${_save_path##"$_mntpoint"}
fi
#erase the old path line, then insert the parsed path
sed -i "/^path/d" "${initdir}/tmp/$$-kdump.conf"
echo "path $_save_path" >> "${initdir}/tmp/$$-kdump.conf"
}
#install kdump.conf and what user specifies in kdump.conf
kdump_install_conf() {
local _opt _val _pdev
kdump_read_conf > "${initdir}/tmp/$$-kdump.conf"
while read -r _opt _val; do
# remove inline comments after the end of a directive.
case "$_opt" in
raw)
_pdev=$(persistent_policy="by-id" kdump_get_persistent_dev "$_val")
sed -i -e "s#^${_opt}[[:space:]]\+$_val#$_opt $_pdev#" "${initdir}/tmp/$$-kdump.conf"
;;
ext[234] | xfs | btrfs | minix | virtiofs)
_pdev=$(kdump_get_persistent_dev "$_val")
sed -i -e "s#^${_opt}[[:space:]]\+$_val#$_opt $_pdev#" "${initdir}/tmp/$$-kdump.conf"
;;
ssh | nfs)
kdump_collect_netif_usage "$_val"
;;
dracut_args)
if [[ $(get_dracut_args_fstype "$_val") == nfs* ]]; then
kdump_collect_netif_usage "$(get_dracut_args_target "$_val")"
fi
;;
kdump_pre | kdump_post | extra_bins)
# shellcheck disable=SC2086
dracut_install $_val
;;
core_collector)
dracut_install "${_val%%[[:blank:]]*}"
;;
esac
done <<< "$(kdump_read_conf)"
kdump_install_pre_post_conf
default_dump_target_install_conf
kdump_configure_fence_kdump "${initdir}/tmp/$$-kdump.conf"
inst "${initdir}/tmp/$$-kdump.conf" "/etc/kdump.conf"
rm -f "${initdir}/tmp/$$-kdump.conf"
}
# Default sysctl parameters should suffice for kdump kernel.
# Remove custom configurations sysctl.conf & sysctl.d/*
remove_sysctl_conf() {
# As custom configurations like vm.min_free_kbytes can lead
# to OOM issues in kdump kernel, avoid them
rm -f "${initdir}/etc/sysctl.conf"
rm -rf "${initdir}/etc/sysctl.d"
rm -rf "${initdir}/run/sysctl.d"
rm -rf "${initdir}/usr/lib/sysctl.d"
}
kdump_iscsi_get_rec_val() {
local result
# The open-iscsi 742 release changed to using flat files in
# /var/lib/iscsi.
result=$(/sbin/iscsiadm --show -m session -r "$1" | grep "^${2} = ")
result=${result##* = }
echo "$result"
}
kdump_get_iscsi_initiator() {
local _initiator
local initiator_conf="/etc/iscsi/initiatorname.iscsi"
[[ -f $initiator_conf ]] || return 1
while read -r _initiator; do
[[ -z ${_initiator%%#*} ]] && continue # Skip comment lines
case $_initiator in
InitiatorName=*)
initiator=${_initiator#InitiatorName=}
echo "rd.iscsi.initiator=${initiator}"
return 0
;;
*) ;;
esac
done < ${initiator_conf}
return 1
}
# Figure out iBFT session according to session type
is_ibft() {
[[ "$(kdump_iscsi_get_rec_val "$1" "node.discovery_type")" == fw ]]
}
kdump_setup_iscsi_device() {
local path=$1
local tgt_name
local tgt_ipaddr
local username
local password
local userpwd_str
local username_in
local password_in
local userpwd_in_str
local netroot_str
local initiator_str
local netroot_conf="${initdir}/etc/cmdline.d/50iscsi.conf"
local initiator_conf="/etc/iscsi/initiatorname.iscsi"
dinfo "Found iscsi component $1"
module-setup: suppress the early iscsi error messages Currently, we throw the error message at the very beginning, as a result on a pure-hardware(all-offload) iscsi machine with many iscsi partitions, we suffered from too much noise as follows: iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.1/host0/session1 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.3/host1/session2 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.1/host0/session1 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.3/host1/session2 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.1/host0/session1 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.3/host1/session2 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.1/host0/session1 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.3/host1/session2 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.1/host0/session1 iscsiadm: No records found Unable to find iscsi record for /sys/devices/pci0000:00/0000:00:06.0/0000:07:00.0/0000:08:01.3/host1/session2 kexec: loaded kdump kernel Starting kdump: [OK] There's no need to know the very early error messages, we can remove the error output which is actually normal for the pure hardware iscsi. As for unexpected errors, we kept the error outputs in the succeeding kdump_iscsi_get_rec_val() calls by not appending "2>/dev/null". Signed-off-by: Xunlei Pang <xlpang@redhat.com> Acked-by: Dave Young <dyoung@redhat.com>
2017-07-24 13:31:39 +00:00
# Check once before getting explicit values, so we can bail out early,
# e.g. in case of pure-hardware(all-offload) iscsi.
if ! /sbin/iscsiadm -m session -r "$path" &> /dev/null; then
return 1
fi
if is_ibft "$path"; then
return
fi
# Remove software iscsi cmdline generated by 95iscsi,
# and let kdump regenerate here.
rm -f "${initdir}/etc/cmdline.d/95iscsi.conf"
tgt_name=$(kdump_iscsi_get_rec_val "$path" "node.name")
tgt_ipaddr=$(kdump_iscsi_get_rec_val "$path" "node.conn\[0\].address")
# get and set username and password details
username=$(kdump_iscsi_get_rec_val "$path" "node.session.auth.username")
[[ $username == "<empty>" ]] && username=""
password=$(kdump_iscsi_get_rec_val "$path" "node.session.auth.password")
[[ $password == "<empty>" ]] && password=""
username_in=$(kdump_iscsi_get_rec_val "$path" "node.session.auth.username_in")
[[ -n $username ]] && userpwd_str="$username:$password"
# get and set incoming username and password details
[[ $username_in == "<empty>" ]] && username_in=""
password_in=$(kdump_iscsi_get_rec_val "$path" "node.session.auth.password_in")
[[ $password_in == "<empty>" ]] && password_in=""
[[ -n $username_in ]] && userpwd_in_str=":$username_in:$password_in"
kdump_collect_netif_usage "$tgt_ipaddr"
# prepare netroot= command line
# FIXME: Do we need to parse and set other parameters like protocol, port
# iscsi_iface_name, netdev_name, LUN etc.
if is_ipv6_address "$tgt_ipaddr"; then
tgt_ipaddr="[$tgt_ipaddr]"
fi
netroot_str="netroot=iscsi:${userpwd_str}${userpwd_in_str}@$tgt_ipaddr::::$tgt_name"
[[ -f $netroot_conf ]] || touch "$netroot_conf"
# If netroot target does not exist already, append.
if ! grep -q "$netroot_str" "$netroot_conf"; then
echo "$netroot_str" >> "$netroot_conf"
dinfo "Appended $netroot_str to $netroot_conf"
fi
# Setup initator
if ! initiator_str=$(kdump_get_iscsi_initiator); then
derror "Failed to get initiator name"
return 1
fi
# If initiator details do not exist already, append.
if ! grep -q "$initiator_str" "$netroot_conf"; then
echo "$initiator_str" >> "$netroot_conf"
dinfo "Appended $initiator_str to $netroot_conf"
fi
}
kdump_check_iscsi_targets() {
# If our prerequisites are not met, fail anyways.
type -P iscsistart > /dev/null || return 1
kdump_check_setup_iscsi() {
local _dev
_dev=$1
[[ -L /sys/dev/block/$_dev ]] || return
cd "$(readlink -f "/sys/dev/block/$_dev")" || return 1
until [[ -d sys || -d iscsi_session ]]; do
cd ..
done
[[ -d iscsi_session ]] && kdump_setup_iscsi_device "$PWD"
}
[[ $hostonly ]] || [[ $mount_needs ]] && {
for_each_host_dev_and_slaves_all kdump_check_setup_iscsi
}
}
# hostname -a is deprecated, do it by ourself
get_alias() {
local ips
local entries
local alias_set
ips=$(hostname -I)
for ip in $ips; do
# in /etc/hosts, alias can come at the 2nd column
if entries=$(grep "$ip" /etc/hosts | awk '{ $1=""; print $0 }'); then
alias_set="$alias_set $entries"
fi
done
echo "$alias_set"
}
is_localhost() {
local hostnames
local shortnames
local aliasname
local nodename=$1
hostnames=$(hostname -A)
shortnames=$(hostname -A -s)
aliasname=$(get_alias)
hostnames="$hostnames $shortnames $aliasname"
for name in ${hostnames}; do
if [[ $name == "$nodename" ]]; then
return 0
fi
done
return 1
}
# retrieves fence_kdump nodes from Pacemaker cluster configuration
get_pcs_fence_kdump_nodes() {
local nodes
pcs cluster sync > /dev/null 2>&1 && pcs cluster cib-upgrade > /dev/null 2>&1
# get cluster nodes from cluster cib, get interface and ip address
nodelist=$(pcs cluster cib | xmllint --xpath "/cib/status/node_state/@uname" -)
# nodelist is formed as 'uname="node1" uname="node2" ... uname="nodeX"'
# we need to convert each to node1, node2 ... nodeX in each iteration
for node in ${nodelist}; do
# convert $node from 'uname="nodeX"' to 'nodeX'
eval "$node"
nodename="$uname"
# Skip its own node name
if is_localhost "$nodename"; then
continue
fi
nodes="$nodes $nodename"
done
echo "$nodes"
}
# retrieves fence_kdump args from config file
get_pcs_fence_kdump_args() {
if [[ -f $FENCE_KDUMP_CONFIG_FILE ]]; then
. "$FENCE_KDUMP_CONFIG_FILE"
echo "$FENCE_KDUMP_OPTS"
fi
}
get_generic_fence_kdump_nodes() {
local filtered
local nodes
kdump-lib.sh: add a config value retrive helper Add a helper kdump_get_conf_val to replace get_option_value. It can help cover more corner cases in the code, like when there are multiple spaces in config file, config value separated by a tab, heading spaces, or trailing comments. And this uses "sed group command" and "sed hold buffer", make it much faster than previous `grep <config> | tail -1`. This helper is supposed to provide a universal way for kexec-tools scripts to read in config value. Currently, different scripts are reading the config in many different fragile ways. For example, following codes are found in kexec-tools script code base: 1. grep ^force_rebuild $KDUMP_CONFIG_FILE echo $_force_rebuild | cut -d' ' -f2 2. grep ^kdump_post $KDUMP_CONFIG_FILE | cut -d\ -f2 3. awk '/^sshkey/ {print $2}' $conf_file 4. grep ^path $KDUMP_CONFIG_FILE | cut -d' ' -f2- 1, 2, and 4 will fail if the space is replaced by, e.g. a tab 1 and 2 might fail if there are multiple spaces between config name and config value: "kdump_post /var/crash/scripts/kdump-post.sh" A space will be read instead of config value. 1, 2, 3 will fail if there are space in file path, like: "kdump_post /var/crash/scripts dir/kdump-post.sh" 4 will fail if there are trailing comments: "path /var/crash # some comment here" And all will fail if there are heading space, " path /var/crash" And all will most likely cause problems if the config file contains the same option more than once. And all of them are slower than the new sed call. Old get_option_value is also very slow and doesn't handle heading space. Although we never claim to support heading space or tailing comments before, it's harmless to be more robust on config reading, and many conf files in /etc support heading spaces. And have a faster and safer config reading helper makes it easier to clean up the code. Signed-off-by: Kairui Song <kasong@redhat.com> Acked-by: Philipp Rudo <prudo@redhat.com>
2021-08-16 15:25:14 +00:00
nodes=$(kdump_get_conf_val "fence_kdump_nodes")
for node in ${nodes}; do
# Skip its own node name
if is_localhost "$node"; then
continue
fi
filtered="$filtered $node"
done
echo "$filtered"
}
# setup fence_kdump in cluster
# setup proper network and install needed files
kdump_configure_fence_kdump() {
local kdump_cfg_file=$1
local nodes
local args
if is_generic_fence_kdump; then
nodes=$(get_generic_fence_kdump_nodes)
elif is_pcs_fence_kdump; then
nodes=$(get_pcs_fence_kdump_nodes)
# set appropriate options in kdump.conf
echo "fence_kdump_nodes $nodes" >> "${kdump_cfg_file}"
args=$(get_pcs_fence_kdump_args)
if [[ -n $args ]]; then
echo "fence_kdump_args $args" >> "${kdump_cfg_file}"
fi
else
# fence_kdump not configured
return 1
fi
# setup network for each node
for node in ${nodes}; do
kdump_collect_netif_usage "$node"
done
dracut_install /etc/hosts
dracut_install /etc/nsswitch.conf
dracut_install "$FENCE_KDUMP_SEND"
}
# Install a random seed used to feed /dev/urandom
# By the time kdump service starts, /dev/uramdom is already fed by systemd
kdump_install_random_seed() {
local poolsize
poolsize=$(< /proc/sys/kernel/random/poolsize)
if [[ ! -d "${initdir}/var/lib/" ]]; then
mkdir -p "${initdir}/var/lib/"
fi
dd if=/dev/urandom of="${initdir}/var/lib/random-seed" \
bs="$poolsize" count=1 2> /dev/null
}
kdump_install_systemd_conf() {
# Kdump turns out to require longer default systemd mount timeout
# than 1st kernel(90s by default), we use default 300s for kdump.
if ! grep -q -r "^[[:space:]]*DefaultTimeoutStartSec=" "${initdir}/etc/systemd/system.conf"*; then
mkdir -p "${initdir}/etc/systemd/system.conf.d"
echo "[Manager]" > "${initdir}/etc/systemd/system.conf.d/kdump.conf"
echo "DefaultTimeoutStartSec=300s" >> "${initdir}/etc/systemd/system.conf.d/kdump.conf"
fi
# Forward logs to console directly, and don't read Kmsg, this avoids
# unneccessary memory consumption and make console output more useful.
# Only do so for non fadump image.
mkdir -p "${initdir}/etc/systemd/journald.conf.d"
echo "[Journal]" > "${initdir}/etc/systemd/journald.conf.d/kdump.conf"
echo "Storage=volatile" >> "${initdir}/etc/systemd/journald.conf.d/kdump.conf"
echo "ReadKMsg=no" >> "${initdir}/etc/systemd/journald.conf.d/kdump.conf"
echo "ForwardToConsole=yes" >> "${initdir}/etc/systemd/journald.conf.d/kdump.conf"
}
remove_cpu_online_rule() {
local file=${initdir}/usr/lib/udev/rules.d/40-redhat.rules
if [[ -f $file ]]; then
sed -i '/SUBSYSTEM=="cpu"/d' "$file"
fi
}
install() {
declare -A unique_netifs ipv4_usage ipv6_usage
local arch
kdump_module_init
kdump_install_conf
remove_sysctl_conf
# Onlining secondary cpus breaks kdump completely on KVM on Power hosts
# Though we use maxcpus=1 by default but 40-redhat.rules will bring up all
# possible cpus by default. (rhbz1270174 rhbz1266322)
# Thus before we get the kernel fix and the systemd rule fix let's remove
# the cpu online rule in kdump initramfs.
arch=$(uname -m)
if [[ "$arch" = "ppc64le" ]] || [[ "$arch" = "ppc64" ]]; then
remove_cpu_online_rule
fi
if is_ssh_dump_target; then
kdump_install_random_seed
fi
dracut_install -o /etc/adjtime /etc/localtime
inst "$moddir/monitor_dd_progress" "/kdumpscripts/monitor_dd_progress"
chmod +x "${initdir}/kdumpscripts/monitor_dd_progress"
inst "/bin/dd" "/bin/dd"
inst "/bin/tail" "/bin/tail"
inst "/bin/date" "/bin/date"
inst "/bin/sync" "/bin/sync"
inst "/bin/cut" "/bin/cut"
inst "/bin/head" "/bin/head"
inst "/bin/awk" "/bin/awk"
inst "/bin/sed" "/bin/sed"
inst "/bin/stat" "/bin/stat"
inst "/sbin/makedumpfile" "/sbin/makedumpfile"
inst "/sbin/vmcore-dmesg" "/sbin/vmcore-dmesg"
inst "/usr/bin/printf" "/sbin/printf"
inst "/usr/bin/logger" "/sbin/logger"
inst "/usr/bin/chmod" "/sbin/chmod"
inst "/lib/kdump/kdump-lib-initramfs.sh" "/lib/kdump-lib-initramfs.sh"
inst "/lib/kdump/kdump-logger.sh" "/lib/kdump-logger.sh"
inst "$moddir/kdump.sh" "/usr/bin/kdump.sh"
inst "$moddir/kdump-capture.service" "$systemdsystemunitdir/kdump-capture.service"
systemctl -q --root "$initdir" add-wants initrd.target kdump-capture.service
kdump-emergency: fix "Transaction is destructive" emergency failure We met a problem that the kdump emergency service failed to start when the target dump timeout(we passed "rd.timeout=30" to kdump), it reported "Transaction is destructive" messages: [ TIME ] Timed out waiting for device dev-mapper-fedora\x2droot.device. [DEPEND] Dependency failed for Initrd Root Device. [ SKIP ] Ordering cycle found, skipping System Initialization [DEPEND] Dependency failed for /sysroot. [DEPEND] Dependency failed for Initrd Root File System. [DEPEND] Dependency failed for Reload Configuration from the Real Root. [ SKIP ] Ordering cycle found, skipping System Initialization [ SKIP ] Ordering cycle found, skipping Initrd Default Target [DEPEND] Dependency failed for File System Check on /dev/mapper/fedora-root. [ OK ] Reached target Initrd File Systems. [ OK ] Stopped dracut pre-udev hook. [ OK ] Stopped dracut cmdline hook. Starting Setup Virtual Console... Starting Kdump Emergency... [ OK ] Reached target Initrd Default Target. [ OK ] Stopped dracut initqueue hook. Failed to start kdump-error-handler.service: Transaction is destructive. See system logs and 'systemctl status kdump-error-handler.service' for details. [FAILED] Failed to start Kdump Emergency. See 'systemctl status emergency.service' for details. [DEPEND] Dependency failed for Emergency Mode. This is because in case of root failure, initrd-root-fs.target will trigger systemd emergency target which requires the systemd emergency service actually is kdump-emergency.service, then our kdump-emergency.service starts kdump-error-handler.service with "systemctl isolate"(see 99kdumpbase/kdump-emergency.service, we replace systemd's with this one under kdump). This will lead to systemd two contradictable jobs queued as an atomic transaction: job 1) the emergency service gets started by initrd-root-fs.target job 2) the emergency service gets stopped due to "systemctl isolate" thereby throwing "Transaction is destructive". In order to solve it, we can utilize "IgnoreOnIsolate=yes" for both kdump-emergency.service and kdump-emergency.target. Unit with attribute "IgnoreOnIsolate=yes" won't be stopped when isolating another unit, they can keep going as expected in case be triggered by any failure. We add kdump-emergency.target dedicated to kdump the similar way as did for kdump-emergency.service(i.e. will replace systemd's emergency.target with kdump-emergency.target under kdump), and adds "IgnoreOnIsolate=yes" into both of them. Signed-off-by: Xunlei Pang <xlpang@redhat.com> Acked-by: Dave Young <dyoung@redhat.com> Acked-by: Pratyush Anand <panand@redhat.com> [bhe: improve the patch log about IgnoreOnIsolate="]
2017-03-27 04:07:32 +00:00
# Replace existing emergency service and emergency target
Introduce kdump error handling service Now upon failure kdump script might not be called at all and it might not be able to execute default action. It results in a hang. Because we disable emergency shell and rely on kdump.sh being invoked through dracut-pre-pivot hook. But it might happen that we never call into dracut-pre-pivot hook because certain systemd targets could not reach due to failure in their dependencies. In those cases error handling code does not run and system hangs. For example: sysroot-var-crash.mount --> initrd-root-fs.target --> initrd.target \ --> dracut-pre-pivot.service --> kdump.sh If /sysroot/var/crash mount fails, initrd-root-fs.target will not be reached. And then initrd.target will not be reached, dracut-pre-pivot.service wouldn't run. Finally kdump.sh wouldn't run. To solve this problem, we need to separate the error handling code from dracut-pre-pivot hook, and every time when a failure shows up, the separated code can be called by the emergency service. By default systemd provides an emergency service which will drop us into shell every time upon a critical failure. It's very convenient for us to re-use the framework of systemd emergency, because we don't have to touch the other parts of systemd. We can use our own script instead of the default one. This new scheme will overwrite emergency shell and replace with kdump error handling code. And this code will do the error handling as needed. Now, we will not rely on dracut-pre-pivot hook running always. Instead whenever error happens and it is serious enough that emergency shell needed to run, now kdump error handler will run. dracut-emergency is also replaced by kdump error handler and it's enabled again all the way down. So all the failure (including systemd and dracut) in 2nd kernel could be captured, and trigger kdump error handler. dracut-initqueue is a special case, which calls "systemctl start emergency" directly, not via "OnFailure=emergency". In case of failure, emergency is started, but not in a isolation mode, which means dracut-initqueue is still running. On the other hand, emergency will call dracut-initqueue again when default action is dump_to_rootfs. systemd would block on the last dracut-initqueue, waiting for the first instance to exit, which leaves us hang. It looks like the following: dracut-initqueue (running) --> call dracut-emergency: --> dracut-emergency (running) --> kdump-error-handler.sh (running) --> call dracut-initqueue: --> blocking and waiting for the original instance to exit. To fix this, I'd like to introduce a wrapper emergency service. This emegency service will replace both the systemd and dracut emergency. And this service does nothing but to isolate to real kdump error handler service: dracut-initqueue (running) --> call dracut-emergency: --> dracut-emergency isolate to kdump-error-handler.service --> dracut-emergency and dracut-initqueue will both be stopped and kdump-error-handler.service will run kdump-error-handler.sh. In a normal failure case, this still works: foo.service fails --> trigger emergency.service --> emergency.service isolates to kdump-error-handler.service --> kdump-error-handler.service will run kdump-error-handler.sh Signed-off-by: WANG Chao <chaowang@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com> Acked-by: Dave Young <dyoung@redhat.com>
2014-05-08 11:37:15 +00:00
cp "$moddir/kdump-emergency.service" "$initdir/$systemdsystemunitdir/emergency.service"
kdump-emergency: fix "Transaction is destructive" emergency failure We met a problem that the kdump emergency service failed to start when the target dump timeout(we passed "rd.timeout=30" to kdump), it reported "Transaction is destructive" messages: [ TIME ] Timed out waiting for device dev-mapper-fedora\x2droot.device. [DEPEND] Dependency failed for Initrd Root Device. [ SKIP ] Ordering cycle found, skipping System Initialization [DEPEND] Dependency failed for /sysroot. [DEPEND] Dependency failed for Initrd Root File System. [DEPEND] Dependency failed for Reload Configuration from the Real Root. [ SKIP ] Ordering cycle found, skipping System Initialization [ SKIP ] Ordering cycle found, skipping Initrd Default Target [DEPEND] Dependency failed for File System Check on /dev/mapper/fedora-root. [ OK ] Reached target Initrd File Systems. [ OK ] Stopped dracut pre-udev hook. [ OK ] Stopped dracut cmdline hook. Starting Setup Virtual Console... Starting Kdump Emergency... [ OK ] Reached target Initrd Default Target. [ OK ] Stopped dracut initqueue hook. Failed to start kdump-error-handler.service: Transaction is destructive. See system logs and 'systemctl status kdump-error-handler.service' for details. [FAILED] Failed to start Kdump Emergency. See 'systemctl status emergency.service' for details. [DEPEND] Dependency failed for Emergency Mode. This is because in case of root failure, initrd-root-fs.target will trigger systemd emergency target which requires the systemd emergency service actually is kdump-emergency.service, then our kdump-emergency.service starts kdump-error-handler.service with "systemctl isolate"(see 99kdumpbase/kdump-emergency.service, we replace systemd's with this one under kdump). This will lead to systemd two contradictable jobs queued as an atomic transaction: job 1) the emergency service gets started by initrd-root-fs.target job 2) the emergency service gets stopped due to "systemctl isolate" thereby throwing "Transaction is destructive". In order to solve it, we can utilize "IgnoreOnIsolate=yes" for both kdump-emergency.service and kdump-emergency.target. Unit with attribute "IgnoreOnIsolate=yes" won't be stopped when isolating another unit, they can keep going as expected in case be triggered by any failure. We add kdump-emergency.target dedicated to kdump the similar way as did for kdump-emergency.service(i.e. will replace systemd's emergency.target with kdump-emergency.target under kdump), and adds "IgnoreOnIsolate=yes" into both of them. Signed-off-by: Xunlei Pang <xlpang@redhat.com> Acked-by: Dave Young <dyoung@redhat.com> Acked-by: Pratyush Anand <panand@redhat.com> [bhe: improve the patch log about IgnoreOnIsolate="]
2017-03-27 04:07:32 +00:00
cp "$moddir/kdump-emergency.target" "$initdir/$systemdsystemunitdir/emergency.target"
Introduce kdump error handling service Now upon failure kdump script might not be called at all and it might not be able to execute default action. It results in a hang. Because we disable emergency shell and rely on kdump.sh being invoked through dracut-pre-pivot hook. But it might happen that we never call into dracut-pre-pivot hook because certain systemd targets could not reach due to failure in their dependencies. In those cases error handling code does not run and system hangs. For example: sysroot-var-crash.mount --> initrd-root-fs.target --> initrd.target \ --> dracut-pre-pivot.service --> kdump.sh If /sysroot/var/crash mount fails, initrd-root-fs.target will not be reached. And then initrd.target will not be reached, dracut-pre-pivot.service wouldn't run. Finally kdump.sh wouldn't run. To solve this problem, we need to separate the error handling code from dracut-pre-pivot hook, and every time when a failure shows up, the separated code can be called by the emergency service. By default systemd provides an emergency service which will drop us into shell every time upon a critical failure. It's very convenient for us to re-use the framework of systemd emergency, because we don't have to touch the other parts of systemd. We can use our own script instead of the default one. This new scheme will overwrite emergency shell and replace with kdump error handling code. And this code will do the error handling as needed. Now, we will not rely on dracut-pre-pivot hook running always. Instead whenever error happens and it is serious enough that emergency shell needed to run, now kdump error handler will run. dracut-emergency is also replaced by kdump error handler and it's enabled again all the way down. So all the failure (including systemd and dracut) in 2nd kernel could be captured, and trigger kdump error handler. dracut-initqueue is a special case, which calls "systemctl start emergency" directly, not via "OnFailure=emergency". In case of failure, emergency is started, but not in a isolation mode, which means dracut-initqueue is still running. On the other hand, emergency will call dracut-initqueue again when default action is dump_to_rootfs. systemd would block on the last dracut-initqueue, waiting for the first instance to exit, which leaves us hang. It looks like the following: dracut-initqueue (running) --> call dracut-emergency: --> dracut-emergency (running) --> kdump-error-handler.sh (running) --> call dracut-initqueue: --> blocking and waiting for the original instance to exit. To fix this, I'd like to introduce a wrapper emergency service. This emegency service will replace both the systemd and dracut emergency. And this service does nothing but to isolate to real kdump error handler service: dracut-initqueue (running) --> call dracut-emergency: --> dracut-emergency isolate to kdump-error-handler.service --> dracut-emergency and dracut-initqueue will both be stopped and kdump-error-handler.service will run kdump-error-handler.sh. In a normal failure case, this still works: foo.service fails --> trigger emergency.service --> emergency.service isolates to kdump-error-handler.service --> kdump-error-handler.service will run kdump-error-handler.sh Signed-off-by: WANG Chao <chaowang@redhat.com> Acked-by: Vivek Goyal <vgoyal@redhat.com> Acked-by: Dave Young <dyoung@redhat.com>
2014-05-08 11:37:15 +00:00
# Also redirect dracut-emergency to kdump error handler
ln_r "$systemdsystemunitdir/emergency.service" "$systemdsystemunitdir/dracut-emergency.service"
# Check for all the devices and if any device is iscsi, bring up iscsi
# target. Ideally all this should be pushed into dracut iscsi module
# at some point of time.
kdump_check_iscsi_targets
kdump_install_systemd_conf
# nfs/ssh dump will need to get host ip in second kernel and need to call 'ip' tool, see get_host_ip for more detail
if is_nfs_dump_target || is_ssh_dump_target; then
inst "ip"
fi
kdump_install_net
# For the lvm type target under kdump, in /etc/lvm/lvm.conf we can
# safely replace "reserved_memory=XXXX"(default value is 8192) with
# "reserved_memory=1024" to lower memory pressure under kdump. We do
# it unconditionally here, if "/etc/lvm/lvm.conf" doesn't exist, it
# actually does nothing.
sed -i -e \
's/\(^[[:space:]]*reserved_memory[[:space:]]*=\)[[:space:]]*[[:digit:]]*/\1 1024/' \
"${initdir}/etc/lvm/lvm.conf" &> /dev/null
# Save more memory by dropping switch root capability
dracut_no_switch_root
}