rpms/NetworkManager
7
0
Fork 0
Code Issues Pull Requests Releases Activity

Merge branch 'c8' into a8

Browse Source
This commit is contained in:
eabdullin 2024-01-15 11:48:31 +03:00
commit 5ac7a12c8b
12 changed files with 4747 additions and 309 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

189
SOURCES/0004-Revert-infiniband-avoid-normalizing-the-p-key-when-r.patch
View File

@ -1,189 +0,0 @@
From a1b73d823f5ec30c240418137d62d183b6ff8ca7 Mon Sep 17 00:00:00 2001
From: Thomas Haller <thaller@redhat.com>
Date: Wed, 24 May 2023 09:44:59 +0200
Subject: [PATCH 1/2] Revert "infiniband: avoid normalizing the p-key when
reading from ifcfg"
Historically, initscripts' ifup-ib would set the highest bit of
PKEY_ID=. That changed and needs to be restored.
Note that it probably makes little sense to ever configure p-keys
without the highest bit set, because that flag indicates full membership
and kernel will automatically add it. At least, kernel will add the flag
for the p-key, but not for the automatically chosen interface name.
Meaning, writing 0x00f0 to create_child sysctl, results in an interface
"$parent.00f0", but `ip -d link` shows pkey 0x80f0.
As NetworkManager otherwise supports p-keys without the highest bit set,
and since that high bit is honored for the interface name, we cannot
just always add the high bit. NetworkManager always assuming the highest
bit is set, would change the interface names of existing configuration.
With this revert, when a user configures a small p-key and the profile
is stored in ifcfg-rh format, the settings backend will automatically
mangle the profile and set 0x8000. That is different from when the
profile is stored in keyfile format. Since using small p-keys is
probably an odd case, we don't try to workaround that any other way
(like that ifcfg format could represent the orignal value of the profile
and not doing such mangling, or to add the high bit throughout
NetworkManager to the p-key). It's an inconsistency, but given the
existing behaviors it seems best to stick (revert) to it.
This reverts commit a4fe16a426097eee263cb3ef831dcea468b1ca26.
Affected versions were 1.42.2+ and 1.40.2+.
See-also: https://src.fedoraproject.org/rpms/rdma/blob/05333c3602aa3c1d82a6363521bdd5a498eac6d0/f/rdma.ifup-ib#_75
https://bugzilla.redhat.com/show_bug.cgi?id=2209164
(cherry picked from commit f8e5e07355e23b6d59b1b1c9cd2387c6b40b214b)
---
.../plugins/ifcfg-rh/nms-ifcfg-rh-reader.c | 18 ++++++
.../plugins/ifcfg-rh/tests/test-ifcfg-rh.c | 57 ++++++++++++++-----
2 files changed, 60 insertions(+), 15 deletions(-)
diff --git a/src/core/settings/plugins/ifcfg-rh/nms-ifcfg-rh-reader.c b/src/core/settings/plugins/ifcfg-rh/nms-ifcfg-rh-reader.c
index 02ba843201..533379c678 100644
--- a/src/core/settings/plugins/ifcfg-rh/nms-ifcfg-rh-reader.c
+++ b/src/core/settings/plugins/ifcfg-rh/nms-ifcfg-rh-reader.c
@@ -5387,6 +5387,24 @@ parse_infiniband_p_key(shvarFile *ifcfg, int *out_p_key, char **out_parent, GErr
return FALSE;
}
+ /* The highest bit 0x8000 indicates full membership, which kernel always
+ * automatically sets.
+ *
+ * NetworkManager supports p-keys without the high bit set. That affects
+ * the interface name (nmp_utils_new_infiniband_name()) and is what
+ * we write to "create_child"/"delete_child" sysctl. Kernel will honor
+ * such p-keys for the interface name, but for other purposes it adds the
+ * highest bit. That makes using p-keys without the highest bit odd.
+ *
+ * Historically, /etc/sysconfig/network-scripts/ifup-ib would always add "|=0x8000".
+ * The reader does that too.
+ *
+ * Note that this means ifcfg cannot handle p-keys without the highest bit set,
+ * and when trying to store that to ifcfg format, the profile will be mangled/modified
+ * by the ifcg plugin (unlike keyfile backend, which preserves the original p-key value).
+ */
+ id |= 0x8000;
+
*out_p_key = id;
*out_parent = g_steal_pointer(&physdev);
return TRUE;
diff --git a/src/core/settings/plugins/ifcfg-rh/tests/test-ifcfg-rh.c b/src/core/settings/plugins/ifcfg-rh/tests/test-ifcfg-rh.c
index d2ac2b29db..01eb24216a 100644
--- a/src/core/settings/plugins/ifcfg-rh/tests/test-ifcfg-rh.c
+++ b/src/core/settings/plugins/ifcfg-rh/tests/test-ifcfg-rh.c
@@ -8383,21 +8383,21 @@ test_read_ipoib(void)
s_infiniband = nmtst_connection_assert_setting(connection, NM_TYPE_SETTING_INFINIBAND);
pkey = nm_setting_infiniband_get_p_key(s_infiniband);
- g_assert(pkey);
- g_assert_cmpint(pkey, ==, 12);
+ g_assert_cmpint(pkey, ==, 0x800c);
transport_mode = nm_setting_infiniband_get_transport_mode(s_infiniband);
- g_assert(transport_mode);
g_assert_cmpstr(transport_mode, ==, "connected");
}
static void
test_write_infiniband(gconstpointer test_data)
{
- const int TEST_IDX = GPOINTER_TO_INT(test_data);
- nmtst_auto_unlinkfile char *testfile = NULL;
- gs_unref_object NMConnection *connection = NULL;
- gs_unref_object NMConnection *reread = NULL;
+ const int TEST_IDX = GPOINTER_TO_INT(test_data);
+ nmtst_auto_unlinkfile char *testfile = NULL;
+ gs_unref_object NMConnection *connection = NULL;
+ gs_unref_object NMConnection *expected = NULL;
+ gs_unref_object NMConnection *reread = NULL;
+ gboolean reread_same = FALSE;
NMSettingConnection *s_con;
NMSettingInfiniband *s_infiniband;
NMSettingIPConfig *s_ip4;
@@ -8407,6 +8407,7 @@ test_write_infiniband(gconstpointer test_data)
NMIPAddress *addr;
GError *error = NULL;
const char *interface_name = NULL;
+ int p_key;
connection = nm_simple_connection_new();
@@ -8422,14 +8423,21 @@ test_write_infiniband(gconstpointer test_data)
NM_SETTING_INFINIBAND_SETTING_NAME,
NULL);
- if (NM_IN_SET(TEST_IDX, 1, 3))
- interface_name = "ib0.000c";
+ if (NM_IN_SET(TEST_IDX, 1, 2))
+ p_key = nmtst_get_rand_bool() ? 0x000c : 0x800c;
+ else
+ p_key = -1;
+
+ if (NM_IN_SET(TEST_IDX, 1, 3)) {
+ if (p_key >= 0x8000)
+ interface_name = "ib0.800c";
+ }
g_object_set(s_con, NM_SETTING_CONNECTION_INTERFACE_NAME, interface_name, NULL);
s_infiniband = _nm_connection_new_setting(connection, NM_TYPE_SETTING_INFINIBAND);
g_object_set(s_infiniband, NM_SETTING_INFINIBAND_TRANSPORT_MODE, "connected", NULL);
- if (NM_IN_SET(TEST_IDX, 1, 2)) {
+ if (p_key == -1) {
g_object_set(s_infiniband,
NM_SETTING_INFINIBAND_MAC_ADDRESS,
mac,
@@ -8439,7 +8447,7 @@ test_write_infiniband(gconstpointer test_data)
} else {
g_object_set(s_infiniband,
NM_SETTING_INFINIBAND_P_KEY,
- 12,
+ p_key,
NM_SETTING_INFINIBAND_PARENT,
"ib0",
NULL);
@@ -8468,13 +8476,32 @@ test_write_infiniband(gconstpointer test_data)
nmtst_assert_connection_verifies(connection);
- _writer_new_connection(connection, TEST_SCRATCH_DIR, &testfile);
-
- reread = _connection_from_file(testfile, NULL, TYPE_INFINIBAND, NULL);
+ if (p_key != -1 && p_key < 0x8000) {
+ expected = nm_simple_connection_new_clone(connection);
+ g_object_set(nm_connection_get_setting(expected, NM_TYPE_SETTING_INFINIBAND),
+ NM_SETTING_INFINIBAND_P_KEY,
+ (int) (p_key | 0x8000),
+ NULL);
+ } else
+ expected = g_object_ref(connection);
- nmtst_assert_connection_equals(connection, TRUE, reread, FALSE);
+ _writer_new_connection_reread(connection,
+ TEST_SCRATCH_DIR,
+ &testfile,
+ NO_EXPECTED,
+ &reread,
+ &reread_same);
+ _assert_reread_same(expected, reread);
+ if (p_key == -1 || p_key > 0x8000)
+ g_assert(reread_same);
+ else
+ g_assert(!reread_same);
g_assert_cmpstr(interface_name, ==, nm_connection_get_interface_name(reread));
+ g_assert_cmpint(nm_setting_infiniband_get_p_key(
+ _nm_connection_get_setting(reread, NM_TYPE_SETTING_INFINIBAND)),
+ ==,
+ p_key == -1 ? -1 : (p_key | 0x8000));
}
static void
--
2.39.3

78
SOURCES/0005-libnm-docs-clarify-behavior-of-infiniband.p-key-prop.patch
View File

@ -1,78 +0,0 @@
From 1e014d466a7008725e0b2c7cb41b1e00cb7868de Mon Sep 17 00:00:00 2001
From: Thomas Haller <thaller@redhat.com>
Date: Wed, 24 May 2023 10:44:58 +0200
Subject: [PATCH 2/2] libnm/docs: clarify behavior of infiniband.p-key property
(cherry picked from commit ea18e66ef657b55eca941dca3de4949b950e656b)
---
src/libnm-core-impl/nm-setting-infiniband.c | 19 ++++++++++++++++---
src/libnmc-setting/settings-docs.h.in | 2 +-
.../generate-docs-nm-settings-nmcli.xml.in | 2 +-
3 files changed, 18 insertions(+), 5 deletions(-)
diff --git a/src/libnm-core-impl/nm-setting-infiniband.c b/src/libnm-core-impl/nm-setting-infiniband.c
index 787b838b76..df296becba 100644
--- a/src/libnm-core-impl/nm-setting-infiniband.c
+++ b/src/libnm-core-impl/nm-setting-infiniband.c
@@ -448,9 +448,20 @@ nm_setting_infiniband_class_init(NMSettingInfinibandClass *klass)
* NMSettingInfiniband:p-key:
*
* The InfiniBand P_Key to use for this device. A value of -1 means to use
- * the default P_Key (aka "the P_Key at index 0"). Otherwise, it is a 16-bit
- * unsigned integer, whose high bit is set if it is a "full membership"
- * P_Key.
+ * the default P_Key (aka "the P_Key at index 0"). Otherwise, it is a
+ * 16-bit unsigned integer, whose high bit 0x8000 is set if it is a "full
+ * membership" P_Key. The values 0 and 0x8000 are not allowed.
+ *
+ * With the p-key set, the interface name is always "$parent.$p_key".
+ * Setting "connection.interface-name" to another name is not supported.
+ *
+ * Note that kernel will internally always set the full membership bit,
+ * although the interface name does not reflect that. Thus, not setting
+ * the high bit is probably not useful.
+ *
+ * If the profile is stored in ifcfg-rh format, then the full membership
+ * bit is automatically added. To get consistent behavior, it is
+ * best to only use p-key values with the full membership bit set.
**/
/* ---ifcfg-rh---
* property: p-key
@@ -459,6 +470,8 @@ nm_setting_infiniband_class_init(NMSettingInfinibandClass *klass)
* description: InfiniBand P_Key. The value can be a hex number prefixed with "0x"
* or a decimal number.
* When PKEY_ID is specified, PHYSDEV and DEVICE also must be specified.
+ * Note that ifcfg-rh format will always automatically set the full membership
+ * bit 0x8000. Other p-key cannot be stored.
* example: PKEY=yes PKEY_ID=2 PHYSDEV=mlx4_ib0 DEVICE=mlx4_ib0.8002
* ---end---
*/
diff --git a/src/libnmc-setting/settings-docs.h.in b/src/libnmc-setting/settings-docs.h.in
index 49bddb237f..01629a71f7 100644
--- a/src/libnmc-setting/settings-docs.h.in
+++ b/src/libnmc-setting/settings-docs.h.in
@@ -153,7 +153,7 @@
#define DESCRIBE_DOC_NM_SETTING_GSM_USERNAME N_("The username used to authenticate with the network, if required. Many providers do not require a username, or accept any username. But if a username is required, it is specified here.")
#define DESCRIBE_DOC_NM_SETTING_INFINIBAND_MAC_ADDRESS N_("If specified, this connection will only apply to the IPoIB device whose permanent MAC address matches. This property does not change the MAC address of the device (i.e. MAC spoofing).")
#define DESCRIBE_DOC_NM_SETTING_INFINIBAND_MTU N_("If non-zero, only transmit packets of the specified size or smaller, breaking larger packets up into multiple frames.")
-#define DESCRIBE_DOC_NM_SETTING_INFINIBAND_P_KEY N_("The InfiniBand P_Key to use for this device. A value of -1 means to use the default P_Key (aka \"the P_Key at index 0\"). Otherwise, it is a 16-bit unsigned integer, whose high bit is set if it is a \"full membership\" P_Key.")
+#define DESCRIBE_DOC_NM_SETTING_INFINIBAND_P_KEY N_("The InfiniBand P_Key to use for this device. A value of -1 means to use the default P_Key (aka \"the P_Key at index 0\"). Otherwise, it is a 16-bit unsigned integer, whose high bit 0x8000 is set if it is a \"full membership\" P_Key. The values 0 and 0x8000 are not allowed. With the p-key set, the interface name is always \"$parent.$p_key\". Setting \"connection.interface-name\" to another name is not supported. Note that kernel will internally always set the full membership bit, although the interface name does not reflect that. Thus, not setting the high bit is probably not useful. If the profile is stored in ifcfg-rh format, then the full membership bit is automatically added. To get consistent behavior, it is best to only use p-key values with the full membership bit set.")
#define DESCRIBE_DOC_NM_SETTING_INFINIBAND_PARENT N_("The interface name of the parent device of this device. Normally NULL, but if the \"p_key\" property is set, then you must specify the base device by setting either this property or \"mac-address\".")
#define DESCRIBE_DOC_NM_SETTING_INFINIBAND_TRANSPORT_MODE N_("The IP-over-InfiniBand transport mode. Either \"datagram\" or \"connected\".")
#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_ADDRESSES N_("A list of IPv4 addresses and their prefix length. Multiple addresses can be separated by comma. For example \"192.168.1.5/24, 10.1.0.5/24\". The addresses are listed in decreasing priority, meaning the first address will be the primary address.")
diff --git a/src/nmcli/generate-docs-nm-settings-nmcli.xml.in b/src/nmcli/generate-docs-nm-settings-nmcli.xml.in
index de20c60718..0a69c926e4 100644
--- a/src/nmcli/generate-docs-nm-settings-nmcli.xml.in
+++ b/src/nmcli/generate-docs-nm-settings-nmcli.xml.in
@@ -614,7 +614,7 @@
description="The IP-over-InfiniBand transport mode. Either &quot;datagram&quot; or &quot;connected&quot;." />
<property name="p-key"
alias="p-key"
- description="The InfiniBand P_Key to use for this device. A value of -1 means to use the default P_Key (aka &quot;the P_Key at index 0&quot;). Otherwise, it is a 16-bit unsigned integer, whose high bit is set if it is a &quot;full membership&quot; P_Key." />
+ description="The InfiniBand P_Key to use for this device. A value of -1 means to use the default P_Key (aka &quot;the P_Key at index 0&quot;). Otherwise, it is a 16-bit unsigned integer, whose high bit 0x8000 is set if it is a &quot;full membership&quot; P_Key. The values 0 and 0x8000 are not allowed. With the p-key set, the interface name is always &quot;$parent.$p_key&quot;. Setting &quot;connection.interface-name&quot; to another name is not supported. Note that kernel will internally always set the full membership bit, although the interface name does not reflect that. Thus, not setting the high bit is probably not useful. If the profile is stored in ifcfg-rh format, then the full membership bit is automatically added. To get consistent behavior, it is best to only use p-key values with the full membership bit set." />
<property name="parent"
alias="parent"
description="The interface name of the parent device of this device. Normally NULL, but if the &quot;p_key&quot; property is set, then you must specify the base device by setting either this property or &quot;mac-address&quot;." />
--
2.39.3

26
SOURCES/0007-unblock-autoconnect-upon-reapply-rh2217903.patch
View File

@ -1,26 +0,0 @@
diff -aruN NetworkManager-1.40.16/src/core/devices/nm-device.c NetworkManager-1.40.16.alma/src/core/devices/nm-device.c
--- NetworkManager-1.40.16/src/core/devices/nm-device.c 2023-02-23 11:42:52
+++ NetworkManager-1.40.16.alma/src/core/devices/nm-device.c 2023-08-08 17:55:53
@@ -12826,6 +12826,7 @@
NMConnection *con_old;
NMConnection *con_new;
GHashTableIter iter;
+ NMSettingsConnection *sett_conn;
if (priv->state < NM_DEVICE_STATE_PREPARE || priv->state > NM_DEVICE_STATE_ACTIVATED) {
g_set_error_literal(error,
@@ -12997,6 +12998,14 @@
if (priv->state >= NM_DEVICE_STATE_ACTIVATED)
nm_device_update_metered(self);
+
+ sett_conn = nm_device_get_settings_connection(self);
+ if (sett_conn) {
+ nm_settings_connection_autoconnect_blocked_reason_set(
+ sett_conn,
+ NM_SETTINGS_AUTO_CONNECT_BLOCKED_REASON_USER_REQUEST,
+ FALSE);
+ }
return TRUE;
}

1490
SOURCES/1001-cloud-setup-IMDSv2-rh2151987.patch Normal file
View File

File diff suppressed because it is too large Load Diff

104
SOURCES/1002-dns-add-support-to-no-aaaa-option-rh2144521.patch Normal file
View File

@ -0,0 +1,104 @@
From 53a9c6027f739daf8f49e2180e4ac51f73eae697 Mon Sep 17 00:00:00 2001
From: Fernando Fernandez Mancera <ffmancera@riseup.net>
Date: Tue, 21 Mar 2023 16:39:38 +0000
Subject: [PATCH] dns: add support to no-aaaa option
Users can set `no-aaaa` DNS option to suppress AAAA queries made by the
stub resolver, including AAAA lookups triggered by NSS-based interfaces
such as getaddrinfo. Only DNS lookups are affected.
(cherry picked from commit 9d4bbf78f0b3a80eec9115663bd9db2c6460b369)
(cherry picked from commit f71cd2eb72d97ee9119e812bba7bbd581c587114)
(cherry picked from commit 920ab658b259d940072c61ae43f7013bbb431440)
---
src/libnm-core-impl/nm-setting-ip-config.c | 8 +++++---
src/libnm-core-public/nm-setting-ip-config.h | 1 +
src/libnmc-setting/settings-docs.h.in | 4 ++--
src/nmcli/generate-docs-nm-settings-nmcli.xml.in | 4 ++--
4 files changed, 10 insertions(+), 7 deletions(-)
diff --git a/src/libnm-core-impl/nm-setting-ip-config.c b/src/libnm-core-impl/nm-setting-ip-config.c
index c8fc461396..0e163094f5 100644
--- a/src/libnm-core-impl/nm-setting-ip-config.c
+++ b/src/libnm-core-impl/nm-setting-ip-config.c
@@ -51,6 +51,7 @@ const NMUtilsDNSOptionDesc _nm_utils_dns_option_descs[] = {
{NM_SETTING_DNS_OPTION_USE_VC, FALSE, FALSE},
{NM_SETTING_DNS_OPTION_NO_RELOAD, FALSE, FALSE},
{NM_SETTING_DNS_OPTION_TRUST_AD, FALSE, FALSE},
+ {NM_SETTING_DNS_OPTION_NO_AAAA, FALSE, FALSE},
{NULL, FALSE, FALSE}};
static char *
@@ -6202,9 +6203,10 @@ nm_setting_ip_config_class_init(NMSettingIPConfigClass *klass)
* distinct from an empty list of properties.
*
* The currently supported options are "attempts", "debug", "edns0",
- * "inet6", "ip6-bytestring", "ip6-dotint", "ndots", "no-check-names",
- * "no-ip6-dotint", "no-reload", "no-tld-query", "rotate", "single-request",
- * "single-request-reopen", "timeout", "trust-ad", "use-vc".
+ * "inet6", "ip6-bytestring", "ip6-dotint", "ndots", "no-aaaa",
+ * "no-check-names", "no-ip6-dotint", "no-reload", "no-tld-query",
+ * "rotate", "single-request", "single-request-reopen", "timeout",
+ * "trust-ad", "use-vc".
*
* The "trust-ad" setting is only honored if the profile contributes
* name servers to resolv.conf, and if all contributing profiles have
diff --git a/src/libnm-core-public/nm-setting-ip-config.h b/src/libnm-core-public/nm-setting-ip-config.h
index acbdec0f61..b42c56a8f3 100644
--- a/src/libnm-core-public/nm-setting-ip-config.h
+++ b/src/libnm-core-public/nm-setting-ip-config.h
@@ -359,6 +359,7 @@ char *nm_ip_routing_rule_to_string(const NMIPRoutingRule *self,
#define NM_SETTING_DNS_OPTION_USE_VC "use-vc"
#define NM_SETTING_DNS_OPTION_NO_RELOAD "no-reload"
#define NM_SETTING_DNS_OPTION_TRUST_AD "trust-ad"
+#define NM_SETTING_DNS_OPTION_NO_AAAA "no-aaaa"
typedef struct _NMSettingIPConfigClass NMSettingIPConfigClass;
diff --git a/src/libnmc-setting/settings-docs.h.in b/src/libnmc-setting/settings-docs.h.in
index 35dfa49f00..62edc77f6b 100644
--- a/src/libnmc-setting/settings-docs.h.in
+++ b/src/libnmc-setting/settings-docs.h.in
@@ -168,7 +168,7 @@
#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_DHCP_TIMEOUT N_("A timeout for a DHCP transaction in seconds. If zero (the default), a globally configured default is used. If still unspecified, a device specific timeout is used (usually 45 seconds). Set to 2147483647 (MAXINT32) for infinity.")
#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_DHCP_VENDOR_CLASS_IDENTIFIER N_("The Vendor Class Identifier DHCP option (60). Special characters in the data string may be escaped using C-style escapes, nevertheless this property cannot contain nul bytes. If the per-profile value is unspecified (the default), a global connection default gets consulted. If still unspecified, the DHCP option is not sent to the server.")
#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_DNS N_("Array of IP addresses of DNS servers.")
-#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_DNS_OPTIONS N_("Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are \"attempts\", \"debug\", \"edns0\", \"inet6\", \"ip6-bytestring\", \"ip6-dotint\", \"ndots\", \"no-check-names\", \"no-ip6-dotint\", \"no-reload\", \"no-tld-query\", \"rotate\", \"single-request\", \"single-request-reopen\", \"timeout\", \"trust-ad\", \"use-vc\". The \"trust-ad\" setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have \"trust-ad\" enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then \"edns0\" and \"trust-ad\" are automatically added.")
+#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_DNS_OPTIONS N_("Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are \"attempts\", \"debug\", \"edns0\", \"inet6\", \"ip6-bytestring\", \"ip6-dotint\", \"ndots\", \"no-aaaa\", \"no-check-names\", \"no-ip6-dotint\", \"no-reload\", \"no-tld-query\", \"rotate\", \"single-request\", \"single-request-reopen\", \"timeout\", \"trust-ad\", \"use-vc\". The \"trust-ad\" setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have \"trust-ad\" enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then \"edns0\" and \"trust-ad\" are automatically added.")
#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_DNS_PRIORITY N_("DNS servers priority. The relative priority for DNS servers specified by this setting. A lower numerical value is better (higher priority). Negative values have the special effect of excluding other configurations with a greater numerical priority value; so in presence of at least one negative priority, only DNS servers from connections with the lowest priority value will be used. To avoid all DNS leaks, set the priority of the profile that should be used to the most negative value of all active connections profiles. Zero selects a globally configured default value. If the latter is missing or zero too, it defaults to 50 for VPNs (including WireGuard) and 100 for other connections. Note that the priority is to order DNS settings for multiple active connections. It does not disambiguate multiple DNS servers within the same connection profile. When multiple devices have configurations with the same priority, VPNs will be considered first, then devices with the best (lowest metric) default route and then all other devices. When using dns=default, servers with higher priority will be on top of resolv.conf. To prioritize a given server over another one within the same connection, just specify them in the desired order. Note that commonly the resolver tries name servers in /etc/resolv.conf in the order listed, proceeding with the next server in the list on failure. See for example the \"rotate\" option of the dns-options setting. If there are any negative DNS priorities, then only name servers from the devices with that lowest priority will be considered. When using a DNS resolver that supports Conditional Forwarding or Split DNS (with dns=dnsmasq or dns=systemd-resolved settings), each connection is used to query domains in its search list. The search domains determine which name servers to ask, and the DNS priority is used to prioritize name servers based on the domain. Queries for domains not present in any search list are routed through connections having the '~.' special wildcard domain, which is added automatically to connections with the default route (or can be added manually). When multiple connections specify the same domain, the one with the best priority (lowest numerical value) wins. If a sub domain is configured on another interface it will be accepted regardless the priority, unless parent domain on the other interface has a negative priority, which causes the sub domain to be shadowed. With Split DNS one can avoid undesired DNS leaks by properly configuring DNS priorities and the search domains, so that only name servers of the desired interface are configured.")
#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_DNS_SEARCH N_("List of DNS search domains. Domains starting with a tilde ('~') are considered 'routing' domains and are used only to decide the interface over which a query must be forwarded; they are not used to complete unqualified host names. When using a DNS plugin that supports Conditional Forwarding or Split DNS, then the search domains specify which name servers to query. This makes the behavior different from running with plain /etc/resolv.conf. For more information see also the dns-priority setting. When set on a profile that also enabled DHCP, the DNS search list received automatically (option 119 for DHCPv4 and option 24 for DHCPv6) gets merged with the manual list. This can be prevented by setting \"ignore-auto-dns\". Note that if no DNS searches are configured, the fallback will be derived from the domain from DHCP (option 15).")
#define DESCRIBE_DOC_NM_SETTING_IP4_CONFIG_GATEWAY N_("The gateway associated with this configuration. This is only meaningful if \"addresses\" is also set. Setting the gateway causes NetworkManager to configure a standard default route with the gateway as next hop. This is ignored if \"never-default\" is set. An alternative is to configure the default route explicitly with a manual route and /0 as prefix length. Note that the gateway usually conflicts with routing that NetworkManager configures for WireGuard interfaces, so usually it should not be set in that case. See \"ip4-auto-default-route\".")
@@ -194,7 +194,7 @@
#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_DHCP_SEND_HOSTNAME N_("If TRUE, a hostname is sent to the DHCP server when acquiring a lease. Some DHCP servers use this hostname to update DNS databases, essentially providing a static hostname for the computer. If the \"dhcp-hostname\" property is NULL and this property is TRUE, the current persistent hostname of the computer is sent.")
#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_DHCP_TIMEOUT N_("A timeout for a DHCP transaction in seconds. If zero (the default), a globally configured default is used. If still unspecified, a device specific timeout is used (usually 45 seconds). Set to 2147483647 (MAXINT32) for infinity.")
#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_DNS N_("Array of IP addresses of DNS servers.")
-#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_DNS_OPTIONS N_("Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are \"attempts\", \"debug\", \"edns0\", \"inet6\", \"ip6-bytestring\", \"ip6-dotint\", \"ndots\", \"no-check-names\", \"no-ip6-dotint\", \"no-reload\", \"no-tld-query\", \"rotate\", \"single-request\", \"single-request-reopen\", \"timeout\", \"trust-ad\", \"use-vc\". The \"trust-ad\" setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have \"trust-ad\" enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then \"edns0\" and \"trust-ad\" are automatically added.")
+#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_DNS_OPTIONS N_("Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are \"attempts\", \"debug\", \"edns0\", \"inet6\", \"ip6-bytestring\", \"ip6-dotint\", \"ndots\", \"no-aaaa\", \"no-check-names\", \"no-ip6-dotint\", \"no-reload\", \"no-tld-query\", \"rotate\", \"single-request\", \"single-request-reopen\", \"timeout\", \"trust-ad\", \"use-vc\". The \"trust-ad\" setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have \"trust-ad\" enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then \"edns0\" and \"trust-ad\" are automatically added.")
#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_DNS_PRIORITY N_("DNS servers priority. The relative priority for DNS servers specified by this setting. A lower numerical value is better (higher priority). Negative values have the special effect of excluding other configurations with a greater numerical priority value; so in presence of at least one negative priority, only DNS servers from connections with the lowest priority value will be used. To avoid all DNS leaks, set the priority of the profile that should be used to the most negative value of all active connections profiles. Zero selects a globally configured default value. If the latter is missing or zero too, it defaults to 50 for VPNs (including WireGuard) and 100 for other connections. Note that the priority is to order DNS settings for multiple active connections. It does not disambiguate multiple DNS servers within the same connection profile. When multiple devices have configurations with the same priority, VPNs will be considered first, then devices with the best (lowest metric) default route and then all other devices. When using dns=default, servers with higher priority will be on top of resolv.conf. To prioritize a given server over another one within the same connection, just specify them in the desired order. Note that commonly the resolver tries name servers in /etc/resolv.conf in the order listed, proceeding with the next server in the list on failure. See for example the \"rotate\" option of the dns-options setting. If there are any negative DNS priorities, then only name servers from the devices with that lowest priority will be considered. When using a DNS resolver that supports Conditional Forwarding or Split DNS (with dns=dnsmasq or dns=systemd-resolved settings), each connection is used to query domains in its search list. The search domains determine which name servers to ask, and the DNS priority is used to prioritize name servers based on the domain. Queries for domains not present in any search list are routed through connections having the '~.' special wildcard domain, which is added automatically to connections with the default route (or can be added manually). When multiple connections specify the same domain, the one with the best priority (lowest numerical value) wins. If a sub domain is configured on another interface it will be accepted regardless the priority, unless parent domain on the other interface has a negative priority, which causes the sub domain to be shadowed. With Split DNS one can avoid undesired DNS leaks by properly configuring DNS priorities and the search domains, so that only name servers of the desired interface are configured.")
#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_DNS_SEARCH N_("List of DNS search domains. Domains starting with a tilde ('~') are considered 'routing' domains and are used only to decide the interface over which a query must be forwarded; they are not used to complete unqualified host names. When using a DNS plugin that supports Conditional Forwarding or Split DNS, then the search domains specify which name servers to query. This makes the behavior different from running with plain /etc/resolv.conf. For more information see also the dns-priority setting. When set on a profile that also enabled DHCP, the DNS search list received automatically (option 119 for DHCPv4 and option 24 for DHCPv6) gets merged with the manual list. This can be prevented by setting \"ignore-auto-dns\". Note that if no DNS searches are configured, the fallback will be derived from the domain from DHCP (option 15).")
#define DESCRIBE_DOC_NM_SETTING_IP6_CONFIG_GATEWAY N_("The gateway associated with this configuration. This is only meaningful if \"addresses\" is also set. Setting the gateway causes NetworkManager to configure a standard default route with the gateway as next hop. This is ignored if \"never-default\" is set. An alternative is to configure the default route explicitly with a manual route and /0 as prefix length. Note that the gateway usually conflicts with routing that NetworkManager configures for WireGuard interfaces, so usually it should not be set in that case. See \"ip4-auto-default-route\".")
diff --git a/src/nmcli/generate-docs-nm-settings-nmcli.xml.in b/src/nmcli/generate-docs-nm-settings-nmcli.xml.in
index 9acb76481e..03e6c0b54b 100644
--- a/src/nmcli/generate-docs-nm-settings-nmcli.xml.in
+++ b/src/nmcli/generate-docs-nm-settings-nmcli.xml.in
@@ -656,7 +656,7 @@
<property name="dns-search"
description="List of DNS search domains. Domains starting with a tilde (&apos;~&apos;) are considered &apos;routing&apos; domains and are used only to decide the interface over which a query must be forwarded; they are not used to complete unqualified host names. When using a DNS plugin that supports Conditional Forwarding or Split DNS, then the search domains specify which name servers to query. This makes the behavior different from running with plain /etc/resolv.conf. For more information see also the dns-priority setting. When set on a profile that also enabled DHCP, the DNS search list received automatically (option 119 for DHCPv4 and option 24 for DHCPv6) gets merged with the manual list. This can be prevented by setting &quot;ignore-auto-dns&quot;. Note that if no DNS searches are configured, the fallback will be derived from the domain from DHCP (option 15)." />
<property name="dns-options"
- description="Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are &quot;attempts&quot;, &quot;debug&quot;, &quot;edns0&quot;, &quot;inet6&quot;, &quot;ip6-bytestring&quot;, &quot;ip6-dotint&quot;, &quot;ndots&quot;, &quot;no-check-names&quot;, &quot;no-ip6-dotint&quot;, &quot;no-reload&quot;, &quot;no-tld-query&quot;, &quot;rotate&quot;, &quot;single-request&quot;, &quot;single-request-reopen&quot;, &quot;timeout&quot;, &quot;trust-ad&quot;, &quot;use-vc&quot;. The &quot;trust-ad&quot; setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have &quot;trust-ad&quot; enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then &quot;edns0&quot; and &quot;trust-ad&quot; are automatically added." />
+ description="Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are &quot;attempts&quot;, &quot;debug&quot;, &quot;edns0&quot;, &quot;inet6&quot;, &quot;ip6-bytestring&quot;, &quot;ip6-dotint&quot;, &quot;ndots&quot;, &quot;no-aaaa&quot;, &quot;no-check-names&quot;, &quot;no-ip6-dotint&quot;, &quot;no-reload&quot;, &quot;no-tld-query&quot;, &quot;rotate&quot;, &quot;single-request&quot;, &quot;single-request-reopen&quot;, &quot;timeout&quot;, &quot;trust-ad&quot;, &quot;use-vc&quot;. The &quot;trust-ad&quot; setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have &quot;trust-ad&quot; enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then &quot;edns0&quot; and &quot;trust-ad&quot; are automatically added." />
<property name="dns-priority"
description="DNS servers priority. The relative priority for DNS servers specified by this setting. A lower numerical value is better (higher priority). Negative values have the special effect of excluding other configurations with a greater numerical priority value; so in presence of at least one negative priority, only DNS servers from connections with the lowest priority value will be used. To avoid all DNS leaks, set the priority of the profile that should be used to the most negative value of all active connections profiles. Zero selects a globally configured default value. If the latter is missing or zero too, it defaults to 50 for VPNs (including WireGuard) and 100 for other connections. Note that the priority is to order DNS settings for multiple active connections. It does not disambiguate multiple DNS servers within the same connection profile. When multiple devices have configurations with the same priority, VPNs will be considered first, then devices with the best (lowest metric) default route and then all other devices. When using dns=default, servers with higher priority will be on top of resolv.conf. To prioritize a given server over another one within the same connection, just specify them in the desired order. Note that commonly the resolver tries name servers in /etc/resolv.conf in the order listed, proceeding with the next server in the list on failure. See for example the &quot;rotate&quot; option of the dns-options setting. If there are any negative DNS priorities, then only name servers from the devices with that lowest priority will be considered. When using a DNS resolver that supports Conditional Forwarding or Split DNS (with dns=dnsmasq or dns=systemd-resolved settings), each connection is used to query domains in its search list. The search domains determine which name servers to ask, and the DNS priority is used to prioritize name servers based on the domain. Queries for domains not present in any search list are routed through connections having the &apos;~.&apos; special wildcard domain, which is added automatically to connections with the default route (or can be added manually). When multiple connections specify the same domain, the one with the best priority (lowest numerical value) wins. If a sub domain is configured on another interface it will be accepted regardless the priority, unless parent domain on the other interface has a negative priority, which causes the sub domain to be shadowed. With Split DNS one can avoid undesired DNS leaks by properly configuring DNS priorities and the search domains, so that only name servers of the desired interface are configured." />
<property name="addresses"
@@ -713,7 +713,7 @@
<property name="dns-search"
description="List of DNS search domains. Domains starting with a tilde (&apos;~&apos;) are considered &apos;routing&apos; domains and are used only to decide the interface over which a query must be forwarded; they are not used to complete unqualified host names. When using a DNS plugin that supports Conditional Forwarding or Split DNS, then the search domains specify which name servers to query. This makes the behavior different from running with plain /etc/resolv.conf. For more information see also the dns-priority setting. When set on a profile that also enabled DHCP, the DNS search list received automatically (option 119 for DHCPv4 and option 24 for DHCPv6) gets merged with the manual list. This can be prevented by setting &quot;ignore-auto-dns&quot;. Note that if no DNS searches are configured, the fallback will be derived from the domain from DHCP (option 15)." />
<property name="dns-options"
- description="Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are &quot;attempts&quot;, &quot;debug&quot;, &quot;edns0&quot;, &quot;inet6&quot;, &quot;ip6-bytestring&quot;, &quot;ip6-dotint&quot;, &quot;ndots&quot;, &quot;no-check-names&quot;, &quot;no-ip6-dotint&quot;, &quot;no-reload&quot;, &quot;no-tld-query&quot;, &quot;rotate&quot;, &quot;single-request&quot;, &quot;single-request-reopen&quot;, &quot;timeout&quot;, &quot;trust-ad&quot;, &quot;use-vc&quot;. The &quot;trust-ad&quot; setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have &quot;trust-ad&quot; enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then &quot;edns0&quot; and &quot;trust-ad&quot; are automatically added." />
+ description="Array of DNS options as described in man 5 resolv.conf. NULL means that the options are unset and left at the default. In this case NetworkManager will use default options. This is distinct from an empty list of properties. The currently supported options are &quot;attempts&quot;, &quot;debug&quot;, &quot;edns0&quot;, &quot;inet6&quot;, &quot;ip6-bytestring&quot;, &quot;ip6-dotint&quot;, &quot;ndots&quot;, &quot;no-aaaa&quot;, &quot;no-check-names&quot;, &quot;no-ip6-dotint&quot;, &quot;no-reload&quot;, &quot;no-tld-query&quot;, &quot;rotate&quot;, &quot;single-request&quot;, &quot;single-request-reopen&quot;, &quot;timeout&quot;, &quot;trust-ad&quot;, &quot;use-vc&quot;. The &quot;trust-ad&quot; setting is only honored if the profile contributes name servers to resolv.conf, and if all contributing profiles have &quot;trust-ad&quot; enabled. When using a caching DNS plugin (dnsmasq or systemd-resolved in NetworkManager.conf) then &quot;edns0&quot; and &quot;trust-ad&quot; are automatically added." />
<property name="dns-priority"
description="DNS servers priority. The relative priority for DNS servers specified by this setting. A lower numerical value is better (higher priority). Negative values have the special effect of excluding other configurations with a greater numerical priority value; so in presence of at least one negative priority, only DNS servers from connections with the lowest priority value will be used. To avoid all DNS leaks, set the priority of the profile that should be used to the most negative value of all active connections profiles. Zero selects a globally configured default value. If the latter is missing or zero too, it defaults to 50 for VPNs (including WireGuard) and 100 for other connections. Note that the priority is to order DNS settings for multiple active connections. It does not disambiguate multiple DNS servers within the same connection profile. When multiple devices have configurations with the same priority, VPNs will be considered first, then devices with the best (lowest metric) default route and then all other devices. When using dns=default, servers with higher priority will be on top of resolv.conf. To prioritize a given server over another one within the same connection, just specify them in the desired order. Note that commonly the resolver tries name servers in /etc/resolv.conf in the order listed, proceeding with the next server in the list on failure. See for example the &quot;rotate&quot; option of the dns-options setting. If there are any negative DNS priorities, then only name servers from the devices with that lowest priority will be considered. When using a DNS resolver that supports Conditional Forwarding or Split DNS (with dns=dnsmasq or dns=systemd-resolved settings), each connection is used to query domains in its search list. The search domains determine which name servers to ask, and the DNS priority is used to prioritize name servers based on the domain. Queries for domains not present in any search list are routed through connections having the &apos;~.&apos; special wildcard domain, which is added automatically to connections with the default route (or can be added manually). When multiple connections specify the same domain, the one with the best priority (lowest numerical value) wins. If a sub domain is configured on another interface it will be accepted regardless the priority, unless parent domain on the other interface has a negative priority, which causes the sub domain to be shadowed. With Split DNS one can avoid undesired DNS leaks by properly configuring DNS priorities and the search domains, so that only name servers of the desired interface are configured." />
<property name="addresses"
--
2.39.2

1531
SOURCES/1003-suppport-bond-port-prio-rh1920398.patch Normal file
View File

File diff suppressed because it is too large Load Diff

52
SOURCES/1004-team-don-t-try-to-connect-to-teamd-in-update_connect-rh2182029.patch Normal file
View File

@ -0,0 +1,52 @@
From 4a9d5b23ab513f0ee0b8f490e522f60d4ef3e4cd Mon Sep 17 00:00:00 2001
From: Beniamino Galvani <bgalvani@redhat.com>
Date: Thu, 4 May 2023 15:11:49 +0200
Subject: [PATCH] team: don't try to connect to teamd in update_connection()
In constructed(), NMDevice starts watching the D-Bus name owner or
monitoring the unix socket, and so it is always aware if teamd is
running. When it is, NMDevice connects to it and initializes
priv->tdc.
It is not useful to try to connect to teamd in update_connection()
because warnings will be generated by NM and by libteam if teamd is
not running. As explained above the connection is always initialized
when teamd is available, and so we can just check priv->tdc.
Fixes: ab586236e36b ('core: implement update_connection() for Team')
https://bugzilla.redhat.com/show_bug.cgi?id=2182029
https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/merge_requests/1631
(cherry picked from commit 93430627c245a0b33b873edca329fa716ccfb7d6)
(cherry picked from commit b60f0dd0a20db232c7edc01faa4562ce510ed107)
(cherry picked from commit f6f1a44559990765a5cbc940a74f54df5d8a30d0)
---
src/core/devices/team/nm-device-team.c | 9 +--------
1 file changed, 1 insertion(+), 8 deletions(-)
diff --git a/src/core/devices/team/nm-device-team.c b/src/core/devices/team/nm-device-team.c
index b745158ef8..1d2beb5e8a 100644
--- a/src/core/devices/team/nm-device-team.c
+++ b/src/core/devices/team/nm-device-team.c
@@ -228,17 +228,10 @@ update_connection(NMDevice *device, NMConnection *connection)
NMDeviceTeam *self = NM_DEVICE_TEAM(device);
NMSettingTeam *s_team = _nm_connection_ensure_setting(connection, NM_TYPE_SETTING_TEAM);
NMDeviceTeamPrivate *priv = NM_DEVICE_TEAM_GET_PRIVATE(self);
- struct teamdctl *tdc = priv->tdc;
/* Read the configuration only if not already set */
- if (!priv->config && ensure_teamd_connection(device))
+ if (!priv->config && priv->tdc) {
teamd_read_config(self);
-
- /* Restore previous tdc state */
- if (priv->tdc && !tdc) {
- teamdctl_disconnect(priv->tdc);
- teamdctl_free(priv->tdc);
- priv->tdc = NULL;
}
g_object_set(G_OBJECT(s_team), NM_SETTING_TEAM_CONFIG, _get_config(self), NULL);
--
2.40.1

0
SOURCES/0006-ipv6ll-don-t-regenerate-the-address-when-removed-rh2209353.patch → SOURCES/1005-ipv6ll-don-t-regenerate-the-address-when-removed-rh2209353.patch
View File

1299
SOURCES/1006-fix-read-infiniband-from-ifcfg-rh2209164.patch Normal file
View File

File diff suppressed because it is too large Load Diff

47
SOURCES/1007-unblock-autoconnect-on-reapply-rh2207690.patch Normal file
View File

@ -0,0 +1,47 @@
From 04c0fffdc3a24b66fcfd2e55714bc1308c219c24 Mon Sep 17 00:00:00 2001
From: Gris Ge <fge@redhat.com>
Date: Tue, 27 Jun 2023 15:02:54 +0800
Subject: [PATCH 1/1] setting-connection: Unblock autoconnect upon finish of
`Reapply`
The activation of a connection will clear the block of autoconnect,
we should do the same for reapply.
Signed-off-by: Gris Ge <fge@redhat.com>
(cherry picked from commit 0486efd3584c70179072f611e63b9c0ff6851b80)
(cherry picked from commit 18ce5f43bd16b3cc394424619652c782cb3795c3)
(cherry picked from commit 2695396939d2a867145f7db569aaf5cc6b0d742c)
---
src/core/devices/nm-device.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/src/core/devices/nm-device.c b/src/core/devices/nm-device.c
index 31acc1c1fe6b..2ac55fa83cbe 100644
--- a/src/core/devices/nm-device.c
+++ b/src/core/devices/nm-device.c
@@ -12826,6 +12826,7 @@ check_and_reapply_connection(NMDevice *self,
NMConnection *con_old;
NMConnection *con_new;
GHashTableIter iter;
+ NMSettingsConnection *sett_conn;
if (priv->state < NM_DEVICE_STATE_PREPARE || priv->state > NM_DEVICE_STATE_ACTIVATED) {
g_set_error_literal(error,
@@ -12998,6 +12999,14 @@ check_and_reapply_connection(NMDevice *self,
if (priv->state >= NM_DEVICE_STATE_ACTIVATED)
nm_device_update_metered(self);
+ sett_conn = nm_device_get_settings_connection(self);
+ if (sett_conn) {
+ nm_settings_connection_autoconnect_blocked_reason_set(
+ sett_conn,
+ NM_SETTINGS_AUTO_CONNECT_BLOCKED_REASON_USER_REQUEST,
+ FALSE);
+ }
+
return TRUE;
}
--
2.40.1

190
SOURCES/1008-cloud-setup-fix-terminating-in-reconfig-rh2221903.patch Normal file
View File

@ -0,0 +1,190 @@
From 1e74f06a69d0f01753d6f2f071202a41b92239bc Mon Sep 17 00:00:00 2001
From: Thomas Haller <thaller@redhat.com>
Date: Wed, 31 May 2023 13:06:22 +0200
Subject: [PATCH 1/2] cloud-setup: fix terminating in the middle of
reconfiguring the system
Once we start reconfiguring the system, we need to finish on all
interfaces. Otherwise, we might reconfigure some interfaces, abort
and leave the network broken. When that happens, a subsequent run
might also be unable to recover, because we are unable to reach the
HTTP meta data service.
https://bugzilla.redhat.com/show_bug.cgi?id=2207812
Fixes: 69f048bf0ca3 ('cloud-setup: add tool for automatic IP configuration in cloud')
(cherry picked from commit dab114f038f39e07080f71426d70e84449890088)
(cherry picked from commit 0a033798ac646c80669ab5d8a15362583f4d8ba4)
(cherry picked from commit fe243025e5751dda2e5a3694953f92c87372e008)
---
src/nm-cloud-setup/main.c | 49 ++++++++++++++++++++++++++-------------
1 file changed, 33 insertions(+), 16 deletions(-)
diff --git a/src/nm-cloud-setup/main.c b/src/nm-cloud-setup/main.c
index e1cbd1d4d8..01e41bd72e 100644
--- a/src/nm-cloud-setup/main.c
+++ b/src/nm-cloud-setup/main.c
@@ -15,6 +15,12 @@
/*****************************************************************************/
+typedef struct {
+ GCancellable *cancellable;
+ gboolean enabled;
+ gboolean signal_received;
+} SigTermData;
+
typedef struct {
GMainLoop *main_loop;
GCancellable *cancellable;
@@ -444,7 +450,7 @@ _nmc_mangle_connection(NMDevice *device,
/*****************************************************************************/
static gboolean
-_config_one(GCancellable *sigterm_cancellable,
+_config_one(SigTermData *sigterm_data,
NMClient *nmc,
const NMCSProviderGetConfigResult *result,
guint idx)
@@ -464,7 +470,7 @@ _config_one(GCancellable *sigterm_cancellable,
g_main_context_iteration(NULL, FALSE);
- if (g_cancellable_is_cancelled(sigterm_cancellable))
+ if (g_cancellable_is_cancelled(sigterm_data->cancellable))
return FALSE;
device = nm_g_object_ref(_nmc_get_device_by_hwaddr(nmc, hwaddr));
@@ -498,7 +504,7 @@ try_again:
g_clear_error(&error);
applied_connection = nmcs_device_get_applied_connection(device,
- sigterm_cancellable,
+ sigterm_data->cancellable,
&applied_version_id,
&error);
if (!applied_connection) {
@@ -560,8 +566,12 @@ try_again:
* during package upgrade. */
maybe_no_preserved_external_ip = TRUE;
+ /* Once we start reconfiguring the system, we cannot abort in the middle. From now on,
+ * any SIGTERM gets ignored until we are done. */
+ sigterm_data->enabled = FALSE;
+
if (!nmcs_device_reapply(device,
- sigterm_cancellable,
+ NULL,
applied_connection,
applied_version_id,
maybe_no_preserved_external_ip,
@@ -592,15 +602,13 @@ try_again:
}
static gboolean
-_config_all(GCancellable *sigterm_cancellable,
- NMClient *nmc,
- const NMCSProviderGetConfigResult *result)
+_config_all(SigTermData *sigterm_data, NMClient *nmc, const NMCSProviderGetConfigResult *result)
{
gboolean any_changes = FALSE;
guint i;
for (i = 0; i < result->n_iface_datas; i++) {
- if (_config_one(sigterm_cancellable, nmc, result, i))
+ if (_config_one(sigterm_data, nmc, result, i))
any_changes = TRUE;
}
@@ -612,13 +620,16 @@ _config_all(GCancellable *sigterm_cancellable,
static gboolean
sigterm_handler(gpointer user_data)
{
- GCancellable *sigterm_cancellable = user_data;
+ SigTermData *sigterm_data = user_data;
- if (!g_cancellable_is_cancelled(sigterm_cancellable)) {
- _LOGD("SIGTERM received");
- g_cancellable_cancel(user_data);
- } else
- _LOGD("SIGTERM received (again)");
+ _LOGD("SIGTERM received (%s) (%s)",
+ sigterm_data->signal_received ? "first time" : "again",
+ sigterm_data->enabled ? "cancel operation" : "ignore");
+
+ sigterm_data->signal_received = TRUE;
+
+ if (sigterm_data->enabled)
+ g_cancellable_cancel(sigterm_data->cancellable);
return G_SOURCE_CONTINUE;
}
@@ -633,6 +644,7 @@ main(int argc, const char *const *argv)
gs_unref_object NMClient *nmc = NULL;
nm_auto_free_nmcs_provider_get_config_result NMCSProviderGetConfigResult *result = NULL;
gs_free_error GError *error = NULL;
+ SigTermData sigterm_data;
_nm_logging_enabled_init(g_getenv(NMCS_ENV_VARIABLE("NM_CLOUD_SETUP_LOG")));
@@ -645,7 +657,12 @@ main(int argc, const char *const *argv)
sigterm_cancellable = g_cancellable_new();
- sigterm_source = nm_g_unix_signal_add_source(SIGTERM, sigterm_handler, sigterm_cancellable);
+ sigterm_data = (SigTermData){
+ .cancellable = sigterm_cancellable,
+ .enabled = TRUE,
+ .signal_received = FALSE,
+ };
+ sigterm_source = nm_g_unix_signal_add_source(SIGTERM, sigterm_handler, &sigterm_data);
provider = _provider_detect(sigterm_cancellable);
if (!provider)
@@ -676,7 +693,7 @@ main(int argc, const char *const *argv)
if (!result)
goto done;
- if (_config_all(sigterm_cancellable, nmc, result))
+ if (_config_all(&sigterm_data, nmc, result))
_LOGI("some changes were applied for provider %s", nmcs_provider_get_name(provider));
else
_LOGD("no changes were applied for provider %s", nmcs_provider_get_name(provider));
--
2.40.1
From 1d148ee9592e1292a62f1d932c85d4ca94e9d642 Mon Sep 17 00:00:00 2001
From: Thomas Haller <thaller@redhat.com>
Date: Mon, 5 Jun 2023 13:04:53 +0200
Subject: [PATCH 2/2] cloud-setup: clear error variable in
nmcs_device_reapply()
This is rather bad, because if we reach the "goto again" case,
the error variable is not cleared. Subsequently passing the
error location to nm_device_reapply_finish() will trigger a glib
warning.
Fixes: 29b0420be72f ('nm-cloud-setup: set preserve-external-ip flag during reapply')
(cherry picked from commit c70a5470be034c660b426ebdbef9e8e67609ece7)
(cherry picked from commit 98be3dd5acafa88e7477dcbb9d6420cb2e73ec01)
(cherry picked from commit 5cc38d1c6b1d76b1fa93cba021cf6a5472f12fa4)
---
src/nm-cloud-setup/nm-cloud-setup-utils.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/src/nm-cloud-setup/nm-cloud-setup-utils.c b/src/nm-cloud-setup/nm-cloud-setup-utils.c
index 7cf7959241..1410ecf7c1 100644
--- a/src/nm-cloud-setup/nm-cloud-setup-utils.c
+++ b/src/nm-cloud-setup/nm-cloud-setup-utils.c
@@ -833,6 +833,8 @@ nmcs_device_reapply(NMDevice *device,
NMDeviceReapplyFlags reapply_flags = NM_DEVICE_REAPPLY_FLAGS_PRESERVE_EXTERNAL_IP;
again:
+ g_clear_error(&data.error);
+
nm_device_reapply_async(device,
connection,
version_id,
--
2.40.1

50
SPECS/NetworkManager.spec
View File

@ -6,7 +6,7 @@
%global epoch_version 1 %global epoch_version 1
%global real_version 1.40.16 %global real_version 1.40.16
%global rpm_version %{real_version} %global rpm_version %{real_version}
%global release_version 4 %global release_version 9
%global snapshot %{nil} %global snapshot %{nil}
%global git_sha %{nil} %global git_sha %{nil}
%global bcond_default_debug 0 %global bcond_default_debug 0
@ -174,7 +174,7 @@ Name: NetworkManager
Summary: Network connection manager and user applications Summary: Network connection manager and user applications
Epoch: %{epoch_version} Epoch: %{epoch_version}
Version: %{rpm_version} Version: %{rpm_version}
Release: %{release_version}%{?snap}%{?dist}.alma Release: %{release_version}%{?snap}%{?dist}
Group: System Environment/Base Group: System Environment/Base
License: GPLv2+ and LGPLv2+ License: GPLv2+ and LGPLv2+
URL: https://networkmanager.dev/ URL: https://networkmanager.dev/
@ -193,16 +193,17 @@ Source7: readme-ifcfg-rh.txt
Patch1: 0001-cloud-setup-systemd-unit-rh1791758.patch Patch1: 0001-cloud-setup-systemd-unit-rh1791758.patch
Patch2: 0002-firewall-Default-to-iptables-backend-to-preserve-behavior.patch Patch2: 0002-firewall-Default-to-iptables-backend-to-preserve-behavior.patch
Patch3: 0003-order-ipv6-addresses.patch Patch3: 0003-order-ipv6-addresses.patch
# Patches were generated from NetworkManager github 1e014d466a7008725e0b2c7cb41b1e00cb7868de commit
Patch4: 0004-Revert-infiniband-avoid-normalizing-the-p-key-when-r.patch
Patch5: 0005-libnm-docs-clarify-behavior-of-infiniband.p-key-prop.patch
# This patch was taken from //gitlab.com/redhat/centos-stream/rpms/NetworkManager/-/raw/be40f1d84c6be08681327c3677962a3317f8aeb4/1005-ipv6ll-don-t-regenerate-the-address-when-removed-rh2209353.patch
Patch6: 0006-ipv6ll-don-t-regenerate-the-address-when-removed-rh2209353.patch
# Bugfixes that are only relevant until next rebase of the package. # Bugfixes that are only relevant until next rebase of the package.
# Patch1001: 1001-some.patch # Patch1001: 1001-some.patch
# https://git.almalinux.org/rpms/NetworkManager/commit/cc98a44c6dfde57ceeffa71654c0937a27b9ce43 Patch1001: 1001-cloud-setup-IMDSv2-rh2151987.patch
# and backported to 1.40.16 Patch1002: 1002-dns-add-support-to-no-aaaa-option-rh2144521.patch
Patch7: 0007-unblock-autoconnect-upon-reapply-rh2217903.patch Patch1003: 1003-suppport-bond-port-prio-rh1920398.patch
Patch1004: 1004-team-don-t-try-to-connect-to-teamd-in-update_connect-rh2182029.patch
Patch1005: 1005-ipv6ll-don-t-regenerate-the-address-when-removed-rh2209353.patch
Patch1006: 1006-fix-read-infiniband-from-ifcfg-rh2209164.patch
Patch1007: 1007-unblock-autoconnect-on-reapply-rh2207690.patch
Patch1008: 1008-cloud-setup-fix-terminating-in-reconfig-rh2221903.patch
Requires(post): systemd Requires(post): systemd
%if 0%{?fedora} || 0%{?rhel} >= 8 %if 0%{?fedora} || 0%{?rhel} >= 8
@ -1238,13 +1239,30 @@ fi
%changelog %changelog
* Tue Aug 08 2023 Eduard Abdullin <eabdullin@almalinux.org> - 1:1.40.16-4.alma * Wed Jul 19 2023 Beniamino Galvani <bgalvani@redhat.com> - 1:1.40.16-9
- Apply 0007-unblock-autoconnect-upon-reapply-rh2217903.patch - cloud-setup: fix terminating in the middle of reconfiguration (rh #2221903)
* Fri Jul 14 2023 Eduard Abdullin <eabdullin@almalinux.org> - 1:1.40.16-3.alma * Wed Jun 28 2023 Thomas Haller <thaller@redhat.com> - 1:1.40.16-8
- Apply 0004-Revert-infiniband-avoid-normalizing-the-p-key-when-r.patch - unblock autoconnect of profile on reapply (rh #2207690)
- Apply 0005-libnm-docs-clarify-behavior-of-infiniband.p-key-prop.patch
- Apply 0006-ipv6ll-don-t-regenerate-the-address-when-removed-rh2209353.patch * Mon Jun 5 2023 Thomas Haller <thaller@redhat.com> - 1:1.40.16-7
- fix reading infiniband p-key from ifcfg files (rh #2209164)
* Wed May 24 2023 Beniamino Galvani <bgalvani@redhat.com> - 1:1.40.16-6
- don't fail when the IPv6 link-local address is removed (rh #2209353)
* Wed May 17 2023 Fernando Fernandez Mancera <ferferna@redhat.com> - 1:1.40.16-5
- support bond port prio property (rh #1920398)
- team: don't try to connect to teamd in update_connection() (rh #2182029)
* Wed Apr 19 2023 Beniamino Galvani <bgalvani@redhat.com> - 1:1.40.16-4
- support the "no-aaaa" resolv.conf option (rh #2144521)
* Mon Mar 13 2023 Lubomir Rintel <lkundrak@v3.sk> - 1:1.40.16-3
- cloud-setup/ec2: fix regression with IMDSv2 support (rh #2181466)
* Mon Mar 13 2023 Lubomir Rintel <lkundrak@v3.sk> - 1:1.40.16-2
- cloud-setup/ec2: add IMDSv2 support (rh #2151986)
* Thu Feb 23 2023 Beniamino Galvani <bgalvani@redhat.com> - 1:1.40.16-1 * Thu Feb 23 2023 Beniamino Galvani <bgalvani@redhat.com> - 1:1.40.16-1
- Update to 1.40.16 release - Update to 1.40.16 release