SymbX/systemd
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
49,627 Commits 2 Branches 1 Tag
fd03a6a92bc2c4eae2165cbfcbcd93d82cd28ccf
Commit Graph
100 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Yu Watanabe 38ba3da0a1 network: rename UseFQDN= -> UseHostname= 
Prompted by
https://github.com/systemd/systemd/pull/18543#issuecomment-780798451.
 2021-02-18 21:47:37 +01:00
Yu Watanabe ccbd74f602 network: NHA_ID should be always set 2021-02-18 03:56:26 +09:00
Yu Watanabe c004cd2bbe network: constify arguments 2021-02-18 03:54:50 +09:00
Yu Watanabe 56223d926d network: introduce log_nexthop_debug() 2021-02-18 03:54:50 +09:00
Yu Watanabe c633628daf tree-wide: constify variables if possible 2021-02-18 03:48:07 +09:00
Yu Watanabe 94af46fc66 network: use temporary buffer for safety 2021-02-18 03:48:07 +09:00
Yu Watanabe 5380707aba network: use in_addr_prefix_to_string() 2021-02-18 03:48:07 +09:00
Yu Watanabe b1dea5cffa resolve: use sockaddr_in_addr() 2021-02-18 03:48:07 +09:00
Yu Watanabe bb3b08ad98 resolve: make manager_find_ifindex() or friends return earlier 2021-02-18 03:48:07 +09:00
Yu Watanabe 94876904bb tree-wide: use in_addr_is_set() or friends 2021-02-18 03:48:07 +09:00
Yu Watanabe 275468c033 network: assign values after all checks are passed 2021-02-18 03:48:07 +09:00
Yu Watanabe fa55043450 in-addr-util: introduce in6_addr_equal() 2021-02-18 03:48:06 +09:00
Yu Watanabe 1235befadd in-addr-util: introduce in6_addr_is_link_local() 2021-02-18 03:48:06 +09:00
Yu Watanabe 7653dcc328 in-addr-util: introduce in_addr_is_set() or friends 2021-02-18 03:48:06 +09:00
Yu Watanabe 0c0585ca25 network: enumerate nexthops before routes 
Preparation for the later commits.
Routes may have RTA_NH_ID attribute. To resolve the nexthop ID, all
nexthops must be enumerated earlier.
 2021-02-17 19:40:02 +01:00
Yu Watanabe 055e4909de network: nexthop: first create nexthops with ID 
Otherwise, an ID-less nexthop may conflict with a later nexthop with ID.
 2021-02-17 19:39:37 +01:00
Yu Watanabe 9997507421 firewall-util: replace nft_in6addr_to_range() with in_addr_prefix_range() 2021-02-17 22:57:37 +09:00
Yu Watanabe 1534c5791a in-addr-util: introduce in_addr_prefix_range() 
This will replace nft_in6addr_to_range() in later commit.
 2021-02-17 22:57:37 +09:00
Yu Watanabe 9164338b2e in-addr-util: make in_addr_prefix_nth() always return valid prefix 
Previously, e.g. in_addr_prefix_nth(2400::1, prefixlen=32, nth=1)
does not return 2400:1:: but does 2400:1::1.
 2021-02-17 22:57:37 +09:00
Yu Watanabe 7b6b05cff9 in-addr-util: make in_addr_prefix_nth() refuse prefixlen larger than maximum size 2021-02-17 22:57:37 +09:00
Yu Watanabe 518b6da5d3 in-addr-util: make in_addr_prefix_nth() returns 0 on success 2021-02-17 22:57:37 +09:00
Yu Watanabe 40785f53ba network: refuse IPv4 multipath route for IPv6 route 2021-02-17 22:08:15 +09:00
Yu Watanabe 35d39c94a9 network: Route::gw_family may be AF_UNSPEC 2021-02-17 21:11:13 +09:00
Yu Watanabe 6cd8f9b5f2 test-network: add tests for Family= in [NextHop] 2021-02-17 15:55:37 +09:00
Yu Watanabe 26ff450550 man: update explanations of settings in [NextHop] section 2021-02-17 15:55:37 +09:00
Yu Watanabe acfd8491fb network: nexthop: refuse 0 id 
We usually do not accept values which will be handled as unspecified.
Instead, this makes config_parse_nexthop_id() accept an empty string.
 2021-02-17 15:55:37 +09:00
Yu Watanabe f1923efccb network: nexthop: introduce Family= setting in [NextHop] section 
This is an alias of `Gateway=0.0.0.0` or `Gateway=::`.
 2021-02-17 15:55:37 +09:00
Yu Watanabe e9c4253d47 network: allow to configure nexthop with null address 
Closes #18446.
 2021-02-17 15:55:37 +09:00
Yu Watanabe 0008b5aee2 network: nexthop: unset gateway when an empty string is assigned 2021-02-17 15:55:37 +09:00
Yu Watanabe 3b4e3ebb51 netlink: introduce sd_netlink_message_has_flag() 2021-02-16 22:36:24 +01:00
Yu Watanabe 955b11ffd3 Merge pull request #18629 from yuwata/sd-netlink-nexthop-types 
sd-netlink: add nexthop related types
 2021-02-17 04:06:15 +09:00
Yu Watanabe 8900c05e6e netlink: drop sd_rtnl_message_{route,nexthop}_set_family() 
The family is already set when the message is created.
 2021-02-17 02:06:26 +09:00
Yu Watanabe fc80ed5f97 network: drop unnecessary family setting 
It is already set in sd_rtnl_message_new_nexthop().
 2021-02-17 02:06:26 +09:00
Yu Watanabe ccb4072e21 man: fix indentation 2021-02-16 17:59:21 +01:00
Yu Watanabe 735a3d73b7 netlink: fix assertions 2021-02-17 01:21:03 +09:00
Yu Watanabe 4684ec5e3d netlink: add nexthop related types 2021-02-17 01:16:42 +09:00
Yu Watanabe f5b7deb969 sd-netlink: add RTA_NH_ID attribute support 2021-02-17 01:16:10 +09:00
Yu Watanabe 0f1f933b6e tree-wide: fix typo 2021-02-15 10:15:04 +01:00
Yu Watanabe f24cbe56d7 Merge pull request #18588 from poettering/refuse-loops 
resolved: try hard to never enter packet loops between resolved's stub and resolved's client
 2021-02-15 11:10:32 +09:00
Yu Watanabe fcf99d0ec7 Merge pull request #18587 from poettering/rr-count-workaround 
two fixes for handling misformed packets more gracefully.
 2021-02-15 10:55:10 +09:00
Yu Watanabe 547f9b0dc5 Merge pull request #18563 from poettering/nss-resolve-no-valid 
timesyncd: for off dnssec in resolved
 2021-02-15 10:43:49 +09:00
Yu Watanabe 9ae4f96056 Merge pull request #18555 from yuwata/network-address-set-flag-on-remove 
network: address: also set IFA_FLAGS on remove
 2021-02-13 17:44:58 +09:00
Yu Watanabe b2af6d66fb Merge pull request #18455 from yuwata/network-change-link-state-only-when-new-address-or-route-will-be-assigned 
network: change link state only when new address or route will be assigned
 2021-02-13 17:43:27 +09:00
Yu Watanabe 4a70450104 test-network: merge test_address_static and test_address_preferred_lifetime_zero_ipv6 2021-02-12 23:09:21 +09:00
Yu Watanabe 53ae4762ef network: address: do not set IFA_F_PERMANENT flag 
The flag is automatically set by kernel when the valid lifetime is
infinite. Note that the flag in netlink message for IPv4 address is
ignored. See set_ifa_lifetime() in kernel's net/ipv4/devinet.c.
But the flag is honored for IPv6 address. And if the flag is set with
finite valid lifetime, the address will not removed automatically by
the kernel.
 2021-02-12 23:09:21 +09:00
Yu Watanabe a8481354f0 network: address: also set IFA_FLAGS on remove 
If an address is assigned with IFA_F_MANAGETEMPADDR, then the flag must
be also set on remove. Otherwise, temporary addresses will not be
removed. See also inet6_rtm_deladdr() in kernel's net/ipv6/addrconf.c.

Fixes #13218.
 2021-02-12 23:09:21 +09:00
Yu Watanabe 111ce98419 network: dhcp6: change link state into "configuring" only when a new address or route will be assigned 2021-02-12 22:44:05 +09:00
Yu Watanabe 24b445c2ce network: ndisc: change link state into "configuring" only when a new address or route will be assigned 2021-02-12 22:44:05 +09:00
Yu Watanabe b54aed29c6 network: dhcp6: fix condtion check 2021-02-12 22:43:39 +09:00
Yu Watanabe 0ef9f3c76f network: set return value at the end of the function 
The later netlink_call_async() call may fail. We should not touch the
return value when the function failed.
 2021-02-12 22:43:10 +09:00
Yu Watanabe d9eee312a7 network: make address_configure() or friends return 1 when the address is new 2021-02-12 22:43:10 +09:00
Yu Watanabe 0ef0c9caf7 Merge pull request #18554 from yuwata/network-address-set-NLM_F_REPLACE-flag-automatically 
network: automatically set NLM_F_REPLACE flag
 2021-02-12 13:47:58 +09:00
Yu Watanabe b3f9c17a6b tree-wide: use free_and_strdup_warn() 2021-02-12 03:05:08 +09:00
Yu Watanabe 37c0b60177 network: automatically set NLM_F_REPLACE flag 
The difference between sd_rtnl_message_new_addr() and
sd_rtnl_message_new_addr_update() is only whether NLM_F_REPLACE flag
will be set or not.

If an address is already assigned, then we need to set NLM_F_REPLACE
flag, otherwise, address_configure() will be fail.

This makes address_configure() judge whether we should use the flag or
not.
 2021-02-12 01:50:14 +09:00
Yu Watanabe fe44ce8dba network: address: also logs preferred lifetime 2021-02-12 01:49:59 +09:00
Yu Watanabe 7fb1d980af tree-wide: propagate error in xxx_from-string() 2021-02-12 01:19:18 +09:00
Yu Watanabe bde8467a0d network: dhcp: use string table 2021-02-12 01:19:18 +09:00
Yu Watanabe 58f99ed8a8 string-table: introduce DEFINE_STRING_TABLE_LOOKUP_FROM_STRING() 2021-02-12 01:19:18 +09:00
Yu Watanabe e4ffe10367 network: route: use _WITH_FALLBACK macros 2021-02-12 01:19:18 +09:00
Yu Watanabe 0fcc4e7484 networkd: lldp: use string table 2021-02-12 01:19:13 +09:00
Yu Watanabe 641bcbc56b string-table: introduce DEFINE_PRIVATE_STRING_TABLE_LOOKUP_FROM_STRING_WITH_BOOLEAN() 2021-02-12 01:17:13 +09:00
Yu Watanabe b98680b2b8 tree-wide: use error codes in xxx_from_string() 2021-02-12 01:17:13 +09:00
Yu Watanabe 0cbb768aba conf-parser: use return codes in xxx_from_string() 
Follow-up for #11484.
 2021-02-12 01:17:06 +09:00
Yu Watanabe 23c32ff830 network: use SD_BUS_METHOD_WITH_ARGS() macro 2021-02-10 21:34:47 +01:00
Yu Watanabe 01e3d04ee6 Merge pull request #18536 from poettering/uid-refs-simplify 
dynamic uid/gid reference handling simplifications
 2021-02-11 03:18:15 +09:00
Yu Watanabe fa7924db0b Merge pull request #11484 from keszybz/udevadm-error-logs 
Use real return codes in _from_string() functions
 2021-02-11 03:17:47 +09:00
Yu Watanabe aba1c9c85b bash-completion: loginctl: add -P option 2021-02-11 02:51:00 +09:00
Yu Watanabe e56aa17084 bash-completion: coredumpctl: add --json and -n options 2021-02-11 02:50:54 +09:00
Yu Watanabe 9a67c080b0 Merge pull request #18506 from keszybz/fuzz-systemctl-parse-argv 
Fuzzer for systemctl argline parsing
 2021-02-11 00:29:22 +09:00
Yu Watanabe 7bc38f59dc log: do not use uninitialized value 
Follow-up for 85cf96e3f5.
 2021-02-09 13:58:12 +01:00
Yu Watanabe beb75dd31d test-network: support protocol and linkdown flag for ff00::/8 route 
Fixes #18507.
 2021-02-09 12:25:27 +01:00
Yu Watanabe 8087644a13 tree-wide: replace strverscmp() and str_verscmp() with strverscmp_improved() 2021-02-09 14:25:03 +09:00
Yu Watanabe 87b7d9b6ff string-util: introduce strverscmp_improved() 
Unfortunately, strverscmp() from libc or str_verscmp() do not correctly
handle pre-release version, e.g. 247 vs 247~rc1.

This implement a new comparison function, which is based on the RPM's
rpmvercmp().
 2021-02-09 14:24:58 +09:00
Yu Watanabe e5bc5f1f5a fundamental: move several macros and functions into src/fundamental/ 
sd-boot has a copy of a subset of codes from libbasic. This makes
sd-boot share the code with libbasic, and dedup the code.

Note, startswith_no_case() is dropped from sd-boot, as
- it is not used,
- the previous implementation is not correct,
- gnu-efi does not have StrniCmp() or so.
 2021-02-09 14:22:54 +09:00
Yu Watanabe 822be62fb2 hostnamectl: use Table 2021-02-09 13:49:27 +09:00
Yu Watanabe f2d3ec7abf hostnamectl: show hint when user try to set transient hostname but static hostname is already used 2021-02-09 13:49:27 +09:00
Yu Watanabe e8acf09186 hostname: use free_and_strdup_and_warn() 2021-02-09 13:49:27 +09:00
Yu Watanabe ba12e41d05 hostname: re-read file later when failed to update file 
Previously, even when writing e.g. /etc/hostname fails, the static
hostname in Context is not restored. So, the subsequent call of the same
method succeeds:
```
$ sudo chattr +i /etc/hostname
$ sudo hostnamectl --static set-hostname aaa
Could not set static hostname: Access denied
$ echo $?
1
$ sudo hostnamectl --static set-hostname aaa
$ echo $?
0
```

This makes when updating file is failed, the saved stat is cleared. So,
the static hostname or machine information in the context are always
consistent to the corresponding files.
 2021-02-09 13:49:27 +09:00
Yu Watanabe c388618ec0 hostnamectl: unset pretty hostname only when no target is specified 2021-02-09 13:49:27 +09:00
Yu Watanabe 2d79e90a1f hostnamectl: try to set transient hostname even if updating static or pretty hostname failed 
If no target (--pretty, --static, or --transient) is specified, then
let's try to set transient hostname even if setting static or pretty
hostname failed. This may be useful for read-only filesystem.
 2021-02-09 13:49:15 +09:00
Yu Watanabe 957991b7ac hostname: introduce two bus errors for updating file 2021-02-09 13:37:27 +09:00
Yu Watanabe 0ce2a086da bus-error: align error definitions 2021-02-09 13:37:27 +09:00
Yu Watanabe 5c3d2e3f43 hostnamectl: improve log message on failure 2021-02-09 13:37:27 +09:00
Yu Watanabe 7bac23e8af man: fix typo 2021-02-04 19:17:50 +01:00
Yu Watanabe 436cde839f Merge pull request #18462 from poettering/copy-time 
timestamp fixes in copy.c
 2021-02-04 14:04:37 +09:00
Yu Watanabe 245f4b384d Merge pull request #18459 from poettering/discover-image 
shared: rename machine-image.[ch] → discover-image.[ch]
 2021-02-04 13:57:37 +09:00
Yu Watanabe 0fb613000d test-network: disable NDISC on veth-peer 
This should speed the tests up.
 2021-02-03 22:09:07 +00:00
Yu Watanabe 27897cecff Merge pull request #18319 from yuwata/udev-move-tx-rx-queues 
network,udev: move TransmitQueues=/ReceiveQueues= from .network to .link
 2021-02-03 18:18:57 +09:00
Yu Watanabe 89d2da287b Merge pull request #18423 from DaanDeMeyer/sd-boot-no-include 
More sd-boot cleanups
 2021-02-03 15:04:26 +09:00
Yu Watanabe face9fcc16 network,udev: move TransmitQueues=/ReceiveQueues= from .network to .link 
As the settings are mostly hardware setup, and merely see from network
layer.

See also discussions in
https://github.com/systemd/systemd/pull/18170#issuecomment-758807497
https://github.com/orgs/systemd/teams/systemd/discussions/1
 2021-02-03 14:44:51 +09:00
Yu Watanabe 1f8dc96c06 udev: rename TxQueueLength= -> TransmitQueueLength= 
As we usually (unfortunately not always though) do not use abbreviations.

Tx may be standard abbreviation, but we already have e.g.
TransmitChecksumOffload=. So, let's use Transmit instead of Tx.

Follow-up for ef4a91a7e8.
 2021-02-03 14:43:49 +09:00
Yu Watanabe d96d8a81ef test: add tests for RouteTable= setting 2021-02-03 14:39:25 +09:00
Yu Watanabe 552b90a2f2 network: use defined route table name in debug logs 
And rename route_table_from_string_full() to
manager_get_route_table_from_string().
 2021-02-03 14:36:50 +09:00
Yu Watanabe 310eff7274 network: make RouteTable= setting can take multiple name:number pairs in a line 
Follow-up for c038ce4606.
 2021-02-03 14:32:26 +09:00
Yu Watanabe ecf83c2429 libudev: add one more assertion 2021-02-02 03:23:31 +09:00
Yu Watanabe 65c637ad2c libudev: unset uptodate flag before free()ing entries 
udev_list_entry_free() also removes the entry from LIST if the flag is
set. This slightly optimizes the cleanup logic.
 2021-02-02 02:34:23 +09:00
Yu Watanabe 8e5ce38727 libudev: also drop the entry from LIST even if unique flag is set 
Otherwise, the list becomes dirty when an entry is freed.

This also remove the entry from the hashmap only when its name is set.
The name should be always set, so that does not change anything. But
just for safety.
 2021-02-02 02:28:33 +09:00
Yu Watanabe 140716a516 libudev: set entry->list after the entry is stored in the list 
This should not change anything. As hashmap_remove() is called before
hashmap_ensure_put(). So, even if hashmap_ensure_put() fails, a wrong
entry will not removed from the hashmap by udev_list_entry_free().
But anyway, just for safety.
 2021-02-02 02:22:59 +09:00
Yu Watanabe eaef130d3f libudev: use hashmap_ensure_put() 2021-02-02 02:22:35 +09:00
Yu Watanabe 8b9708d1d3 sd-event: retrieve more events when epoll_wait() returns number equivalent to the buffer size 
When epoll_wait() returns number equivalent to the buffer size, there
may exist remaining events which may have higher priority. To make priority
sorting correctly, let's retrieve all events.
 2021-02-02 01:14:39 +09:00