]> git.ipfire.org Git - thirdparty/systemd.git/log
thirdparty/systemd.git
3 years agomount-util: use UMOUNT_NOFOLLOW in recursive umounter
Lennart Poettering [Sat, 27 Jun 2020 09:13:01 +0000 (11:13 +0200)] 
mount-util: use UMOUNT_NOFOLLOW in recursive umounter

When we only want to unmount mount points below some path then it is
against our interest to follow symlinks. Hence don't.

3 years agomount-util: add destructor helper that umounts + rmdirs a path
Lennart Poettering [Sat, 27 Jun 2020 08:38:07 +0000 (10:38 +0200)] 
mount-util: add destructor helper that umounts + rmdirs a path

3 years agonamespace-util: introduce helper for combining unshare() + MS_SLAVE remount
Lennart Poettering [Sun, 2 Feb 2020 17:56:12 +0000 (18:56 +0100)] 
namespace-util: introduce helper for combining unshare() + MS_SLAVE remount

We have multiple places we do these two non-trivial operations together,
let's introduce a unified helper for doing both at once.

3 years agocore: refresh unit cache when building a transaction if UNIT_NOT_FOUND
Luca Boccassi [Fri, 3 Jul 2020 17:45:19 +0000 (18:45 +0100)] 
core: refresh unit cache when building a transaction if UNIT_NOT_FOUND

When a command asks to load a unit directly and it is in state
UNIT_NOT_FOUND, and the cache is outdated, we refresh it and
attempto to load again.
Use the same logic when building up a transaction and a dependency in
UNIT_NOT_FOUND state is encountered.
Update the unit test to exercise this code path.

3 years agoMerge pull request #16380 from yuwata/network-dhcp6-update-prefix-route
Zbigniew Jędrzejewski-Szmek [Tue, 7 Jul 2020 07:05:08 +0000 (09:05 +0200)] 
Merge pull request #16380 from yuwata/network-dhcp6-update-prefix-route

Network: update acquired dhcp6 prefix routes

3 years agoMerge pull request #16381 from yuwata/dhcp4-renewing
Zbigniew Jędrzejewski-Szmek [Tue, 7 Jul 2020 06:59:35 +0000 (08:59 +0200)] 
Merge pull request #16381 from yuwata/dhcp4-renewing

dhcp4: fixes around renewing address

3 years agoMerge pull request #16383 from yuwata/network-dhcp6-radv-update-prefix
Zbigniew Jędrzejewski-Szmek [Tue, 7 Jul 2020 06:53:31 +0000 (08:53 +0200)] 
Merge pull request #16383 from yuwata/network-dhcp6-radv-update-prefix

network: do not restart radv engine when adding prefix

3 years agohwdb: Add Google Pixel Slate (nocturne) (#16377)
satmandu [Tue, 7 Jul 2020 06:45:31 +0000 (02:45 -0400)] 
hwdb: Add Google Pixel Slate (nocturne) (#16377)

3 years agonetwork: fix memleak
Yu Watanabe [Tue, 7 Jul 2020 03:17:46 +0000 (12:17 +0900)] 
network: fix memleak

If init() is failed, the allocated memory was leaked.

3 years agonetwork: do not restart radv engine when adding prefix 16383/head
Yu Watanabe [Tue, 7 Jul 2020 02:19:49 +0000 (11:19 +0900)] 
network: do not restart radv engine when adding prefix

Replaces d469cea3bde53bc39317c8b433c825bb4790cbe5.

3 years agonetwork: fix indentation
Yu Watanabe [Tue, 7 Jul 2020 02:18:28 +0000 (11:18 +0900)] 
network: fix indentation

3 years agoRevert "network: Don't send RA with zero router lifetime when restarting radv"
Yu Watanabe [Tue, 7 Jul 2020 02:06:08 +0000 (11:06 +0900)] 
Revert "network: Don't send RA with zero router lifetime when restarting radv"

This reverts commit d469cea3bde53bc39317c8b433c825bb4790cbe5.

3 years agofix typo
Tomer Shechner [Tue, 7 Jul 2020 01:35:35 +0000 (04:35 +0300)] 
fix typo

I was thoroughly reading your nice coding style page and found out that you guys missed an 's'.

😁

3 years agonetwork: Don't send RA with zero router lifetime when restarting radv
Michael Marley [Sun, 5 Jul 2020 10:46:27 +0000 (06:46 -0400)] 
network: Don't send RA with zero router lifetime when restarting radv
While investigating https://github.com/systemd/systemd/issues/16356, I
discovered that networkd stops the radv service before adding or updating
prefixes and then starts it again.  This causes networkd to send an RA with
a router lifetime of zero, causing the routes to flap on systems receiving
the RA for a fraction of a second before radv is started again and proper
RAs are sent.  That has the potential to cause issues with latency-sensitive
traffic like gaming or VoIP.  This patch adds a boolean argument to the
sd_radv_stop() function to control this behavior.  The zero lifetime RA is
still sent whenever radv is actually being stopped, but when it is being
restarted for a prefix update (from networkd-dhcp6.c), the final RA is no
longer sent to avoid the route flapping.

3 years agodhcp4: do not try to renew address when client has no lease 16381/head
Yu Watanabe [Tue, 7 Jul 2020 01:04:42 +0000 (10:04 +0900)] 
dhcp4: do not try to renew address when client has no lease

3 years agodhcp4: do not renew address if client is not running yet
Yu Watanabe [Tue, 7 Jul 2020 00:43:10 +0000 (09:43 +0900)] 
dhcp4: do not renew address if client is not running yet

3 years agonetwork: always update acquired prefix route 16380/head
Yu Watanabe [Tue, 7 Jul 2020 00:20:39 +0000 (09:20 +0900)] 
network: always update acquired prefix route

Otherwise, routes become lifetime 0.

Fixes #16356.

3 years agonetwork: decrease indentation level
Yu Watanabe [Tue, 7 Jul 2020 00:14:21 +0000 (09:14 +0900)] 
network: decrease indentation level

3 years agonetwork: fix indentation
Yu Watanabe [Mon, 6 Jul 2020 23:47:38 +0000 (08:47 +0900)] 
network: fix indentation

3 years agoMerge pull request #16374 from keszybz/docs-and-networkd
Yu Watanabe [Mon, 6 Jul 2020 23:29:56 +0000 (08:29 +0900)] 
Merge pull request #16374 from keszybz/docs-and-networkd

Docs and networkd

3 years agohwdb: update again for v246
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 15:53:46 +0000 (17:53 +0200)] 
hwdb: update again for v246

Just a few minor changes.

3 years agoDHCPv4: client add support for DHCP server allow list
Susant Sahani [Fri, 26 Jun 2020 03:28:52 +0000 (03:28 +0000)] 
DHCPv4: client add support for DHCP server allow list

3 years agoMerge pull request #16344 from keszybz/update-utmp-erofs
Lennart Poettering [Mon, 6 Jul 2020 15:52:17 +0000 (17:52 +0200)] 
Merge pull request #16344 from keszybz/update-utmp-erofs

Make update-utmp not fail if it only fails to write wtmp records

3 years agohwdb: add another Logitech M705 mouse variant
Maciej S. Szmigiero [Sun, 5 Jul 2020 16:16:52 +0000 (18:16 +0200)] 
hwdb: add another Logitech M705 mouse variant

There seems to be a newer version of Logitech M705 mouse in the wild.
Let's add it to the hwdb.

Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
3 years agoMerge pull request #16373 from JackFangXN/master
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 15:14:07 +0000 (17:14 +0200)] 
Merge pull request #16373 from JackFangXN/master

userdbctl homectl use table_log_add_error()

3 years agounit-name: fix a potential memory leak
Frantisek Sumsal [Mon, 6 Jul 2020 13:22:28 +0000 (15:22 +0200)] 
unit-name: fix a potential memory leak

Also, add a test which verifies the issue is indeed gone.

Fixes: CID#1429014
Followup to: ab19db01ae1826efb3cbdf6dcb6a14412f8844d4

3 years agonetworkd: we should not loop when extract_first_word() fails 16374/head
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 14:21:34 +0000 (16:21 +0200)] 
networkd: we should not loop when extract_first_word() fails

While at it, define iterator in the loop to reduce the indentation a bit.

3 years agoman: do not say that tasks are threads and processes
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 14:31:57 +0000 (16:31 +0200)] 
man: do not say that tasks are threads and processes

This is confusing because the reader might think that processes and threads are
counted separately. Another issue pointed out in #16363.

3 years agoman: one more typo
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 14:11:27 +0000 (16:11 +0200)] 
man: one more typo

Also from #16363.

3 years agoMerge pull request #16370 from keszybz/tree-wide-spelling
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 13:57:47 +0000 (15:57 +0200)] 
Merge pull request #16370 from keszybz/tree-wide-spelling

Assorted spelling, markup, and grammar fixes

3 years agonetwork: tc: fix n_prio bounds check
Vito Caputo [Mon, 6 Jul 2020 07:36:51 +0000 (00:36 -0700)] 
network: tc: fix n_prio bounds check

Looks like a simple copy and paste error

Fixes https://github.com/systemd/systemd/issues/16367

3 years agouserdbctl homectl use table_log_add_error() 16373/head
fangxiuning [Mon, 6 Jul 2020 13:32:22 +0000 (21:32 +0800)] 
userdbctl homectl use table_log_add_error()

Signed-off-by: fangxiuning <fangxiuning123@126.com>
3 years agoFix two typos found by codespell 16370/head
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 13:09:23 +0000 (15:09 +0200)] 
Fix two typos found by codespell

3 years agotree-wide: drop quotes from around [section]
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 09:00:06 +0000 (11:00 +0200)] 
tree-wide: drop quotes from around [section]

For users, the square brackets already serve as markup and clearly delineate
the section name from surrounding text. Putting additional markup around that
only adds clutter. Also, we were very inconsistent in using the quotes. Let's
just drop them altogether.

3 years agotree-wide: fixes for assorted grammar and spelling issues
Zbigniew Jędrzejewski-Szmek [Mon, 6 Jul 2020 08:49:59 +0000 (10:49 +0200)] 
tree-wide: fixes for assorted grammar and spelling issues

Fixes #16363. Also includes some changes where I generalized the pattern.

3 years agotravis: install missing deps for the Coverity run
Frantisek Sumsal [Mon, 6 Jul 2020 08:51:02 +0000 (10:51 +0200)] 
travis: install missing deps for the Coverity run

Let's install libzstd & libfido2 to cover two recently added features.
In case of libfido2 this should also get rid of the 'dead code' issues
found by Coverity, like CID#1430168, CID#1430167, CID#1430166, or
CID#1430165.

3 years agoMerge pull request #16287 from yuwata/udevadm-info-attribute-walk
Yu Watanabe [Mon, 6 Jul 2020 05:47:51 +0000 (14:47 +0900)] 
Merge pull request #16287 from yuwata/udevadm-info-attribute-walk

udevadm: make --attribute-walk show subdir attributes

3 years agocatalog: update Polish translation
Piotr Drąg [Sun, 5 Jul 2020 10:32:05 +0000 (12:32 +0200)] 
catalog: update Polish translation

3 years agoupdate-utmp: do not fail on EROFS 16344/head
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 14:35:52 +0000 (16:35 +0200)] 
update-utmp: do not fail on EROFS

Right now systemd-update-utmp.service would fail on read-only /var because
it was not able to write the wtmp record. But it still writes the utmp
record just fine, so runtime information is OK. I don't think we need to
make too much fuss about not being able to save wtmp info.

3 years agoDefine _cleanup_ helper for setutxent()+endutxent()
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 14:24:04 +0000 (16:24 +0200)] 
Define _cleanup_ helper for setutxent()+endutxent()

3 years agotest: use KILL instead of SIGKILL in TEST-52-HONORFIRSTSHUTDOWN
Frantisek Sumsal [Sat, 4 Jul 2020 18:18:02 +0000 (20:18 +0200)] 
test: use KILL instead of SIGKILL in TEST-52-HONORFIRSTSHUTDOWN

SIG-prefixed signals for `kill` are not POSIX compliant, so on Ubuntu CI
(which defaults to dash instead of bash) the TEST-52 contains following
error:

[ 9693.549638] sh[51]: + systemctl poweroff --no-block
[ 9693.553130] systemd-logind[26]: System is powering down.
[ 9693.608911] sh[54]: /bin/sh: 1: kill: Illegal option -S

This can be reproduced manually as well, either by running dash, or bash
in POSIX mode:

$ dash -c 'kill -SIGKILL 123'
dash: 1: kill: Illegal option -S

$ bash --posix -c 'kill -SIGKILL 123'
bash: line 0: kill: SIGKILL: invalid signal specification

3 years agosd-device: be more defensive in FOREACH_DEVICE_SYSATTR 16287/head
Zbigniew Jędrzejewski-Szmek [Sun, 5 Jul 2020 08:55:18 +0000 (10:55 +0200)] 
sd-device: be more defensive in FOREACH_DEVICE_SYSATTR

Follow-up for fadcc1222949ed57ca2ce143f2eb9b93ea0dab1b.

3 years agocore: set private section name for automount units
Dan Callaghan [Fri, 3 Jul 2020 09:13:08 +0000 (19:13 +1000)] 
core: set private section name for automount units

Because this was left unset, the unit_write_setting() function was
refusing to write out the automount-specific TimeoutIdleSec= and
DirectoryMode= settings when creating transient automount units.
Set it to the proper value in line with other unit types.

3 years agohwdb: Pavilion 13 x360: Tablet mode and SYSRQ key
MaxVerevkin [Sat, 4 Jul 2020 06:43:34 +0000 (09:43 +0300)] 
hwdb: Pavilion 13 x360: Tablet mode and SYSRQ key

3 years agoMerge pull request #15442 from poettering/fido2
Zbigniew Jędrzejewski-Szmek [Fri, 3 Jul 2020 15:27:15 +0000 (17:27 +0200)] 
Merge pull request #15442 from poettering/fido2

add fido2 authentication support to homed

3 years agoudevadm: sort entries in `udevadm info -a` by attribute name
Yu Watanabe [Fri, 26 Jun 2020 06:44:41 +0000 (15:44 +0900)] 
udevadm: sort entries in `udevadm info -a` by attribute name

3 years agosd-device: make FOREACH_DEVICE_SYSATTR() list attributes in subdirectories
Yu Watanabe [Fri, 26 Jun 2020 06:29:43 +0000 (15:29 +0900)] 
sd-device: make FOREACH_DEVICE_SYSATTR() list attributes in subdirectories

Then, `udevadm info -a` can show subdir attributes.

Closes #12191.

3 years agoinitialize arg_clock_usec
gzjsgdsb [Fri, 3 Jul 2020 09:51:37 +0000 (17:51 +0800)] 
initialize arg_clock_usec

3 years agoMerge pull request #16348 from yuwata/news-sriov
Zbigniew Jędrzejewski-Szmek [Fri, 3 Jul 2020 10:13:14 +0000 (12:13 +0200)] 
Merge pull request #16348 from yuwata/news-sriov

NEWS: two updates about networkd

3 years agoNEWS: mention .network [SR-IOV] section 16348/head
Yu Watanabe [Fri, 3 Jul 2020 06:33:06 +0000 (15:33 +0900)] 
NEWS: mention .network [SR-IOV] section

3 years agoNEWS: WithoutRA= is not a boolean option anymore
Yu Watanabe [Fri, 3 Jul 2020 06:25:28 +0000 (15:25 +0900)] 
NEWS: WithoutRA= is not a boolean option anymore

3 years agoNEWS: tweak WithoutRA= description a bit more
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 16:12:03 +0000 (18:12 +0200)] 
NEWS: tweak WithoutRA= description a bit more

Follow-up for cbdf02952e831aa12c34f15d2466995c4ca4fc5b.

3 years agovarious daemons: emit Stopping... notification before destructing the manager object
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 12:56:34 +0000 (14:56 +0200)] 
various daemons: emit Stopping... notification before destructing the manager object

This is mostly cosmetic, but let's reorder the destructors so that
we do the final sd_notify() call before we run the destructor for
the manager object.

3 years agoinitctl: use _cleanup_ and run()
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 12:50:29 +0000 (14:50 +0200)] 
initctl: use _cleanup_ and run()

3 years agoinitctl,update-utmp: define iterator variable in loop
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 12:25:06 +0000 (14:25 +0200)] 
initctl,update-utmp: define iterator variable in loop

3 years agoupdate-utmp,initctl: drop ppid check
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 11:44:43 +0000 (13:44 +0200)] 
update-utmp,initctl: drop ppid check

Such checks make debugging harder but serve no useful purpose otherwise.
We got rid of all the checks for root, let's kill those too.

3 years agonetwork: tc: rename several settings which take size in bytes
Yu Watanabe [Wed, 1 Jul 2020 07:05:32 +0000 (16:05 +0900)] 
network: tc: rename several settings which take size in bytes

3 years agoMerge pull request #16343 from jirislaby/master
hadess [Thu, 2 Jul 2020 13:19:33 +0000 (15:19 +0200)] 
Merge pull request #16343 from jirislaby/master

hwdb: Add accel orientation quirk for UMAX VisionBook 10Wi Pro

3 years agoMerge pull request #16338 from keszybz/spelling2
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 13:10:27 +0000 (15:10 +0200)] 
Merge pull request #16338 from keszybz/spelling2

Tree-wide spelling fixes

3 years agoMerge pull request #16336 from yuwata/ifindex-cleanups
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 13:08:41 +0000 (15:08 +0200)] 
Merge pull request #16336 from yuwata/ifindex-cleanups

tiny cleanups related to ifindex

3 years agoNEWS: update document about WithoutRA=
Yu Watanabe [Thu, 2 Jul 2020 07:22:39 +0000 (16:22 +0900)] 
NEWS: update document about WithoutRA=

Follow-up for 838d39af93e16e749ef90acf915a634b380abbf6.

3 years agoMerge pull request #16219 from ssahani/network-sr-iov
Yu Watanabe [Thu, 2 Jul 2020 11:24:13 +0000 (20:24 +0900)] 
Merge pull request #16219 from ssahani/network-sr-iov

network: Introduce SR-IOV

3 years agohwdb: Add accel orientation quirk for UMAX VisionBook 10Wi Pro 16343/head
Jiri Slaby [Thu, 2 Jul 2020 11:06:04 +0000 (13:06 +0200)] 
hwdb: Add accel orientation quirk for UMAX VisionBook 10Wi Pro

3 years agotree-wide: "unparseable" → "unparsable" 16338/head
Zbigniew Jędrzejewski-Szmek [Thu, 2 Jul 2020 07:58:23 +0000 (09:58 +0200)] 
tree-wide: "unparseable" → "unparsable"

"unparsable" is the more common spelling. We already pick "movable" over
"moveable". Let's do the same with this pair.

3 years agoresolve: use appropriate type 16336/head
Yu Watanabe [Thu, 2 Jul 2020 07:12:06 +0000 (16:12 +0900)] 
resolve: use appropriate type

3 years agoutil: use setsockopt_int() at one more place
Yu Watanabe [Thu, 2 Jul 2020 07:03:28 +0000 (16:03 +0900)] 
util: use setsockopt_int() at one more place

3 years agolibsystemd-network: rename index -> ifindex
Yu Watanabe [Thu, 2 Jul 2020 06:49:19 +0000 (15:49 +0900)] 
libsystemd-network: rename index -> ifindex

3 years agodhcp6, radv: only accept valid ifindex
Yu Watanabe [Thu, 2 Jul 2020 06:45:33 +0000 (15:45 +0900)] 
dhcp6, radv: only accept valid ifindex

3 years agotree-wide: spell "lifecycle" without hyphen everywhere
Zbigniew Jędrzejewski-Szmek [Thu, 21 Feb 2019 06:45:45 +0000 (07:45 +0100)] 
tree-wide: spell "lifecycle" without hyphen everywhere

We had 2 more instances of unhyphentated spelling.

3 years agotree: wide "the the" and other trivial grammar fixes
Zbigniew Jędrzejewski-Szmek [Tue, 30 Jun 2020 13:24:57 +0000 (15:24 +0200)] 
tree: wide "the the" and other trivial grammar fixes

3 years agonetwork: Allow DHCPv6 client to start solicit mode
Susant Sahani [Wed, 1 Jul 2020 23:03:17 +0000 (23:03 +0000)] 
network: Allow DHCPv6 client  to start solicit mode

Now DHCPv6 client start only DHCPv6 INFORMATION-REQUEST mode.
Let's enable Solicit mode too.

3 years agol10n: update Czech Translation
AsciiWolf [Wed, 1 Jul 2020 14:40:52 +0000 (16:40 +0200)] 
l10n: update Czech Translation

3 years agoupdate TODO 15442/head
Lennart Poettering [Thu, 16 Apr 2020 09:12:39 +0000 (11:12 +0200)] 
update TODO

3 years agoman: update homectl man page with documentation for new features
Lennart Poettering [Thu, 16 Apr 2020 08:51:54 +0000 (10:51 +0200)] 
man: update homectl man page with documentation for new features

3 years agodocs: document new FIDO2 user record fields
Lennart Poettering [Thu, 16 Apr 2020 08:03:44 +0000 (10:03 +0200)] 
docs: document new FIDO2 user record fields

3 years agohomectl: add --pkcs11-uri=auto and --pkcs-11-uri=list support
Lennart Poettering [Thu, 16 Apr 2020 07:44:55 +0000 (09:44 +0200)] 
homectl: add --pkcs11-uri=auto and --pkcs-11-uri=list support

We have the same for FIDO2 devices, for listing suitable devices, or
picking the right one automatically, let's add that for PKCS11 too.

3 years agohomectl: add acquired fido2 PIN to user record
Lennart Poettering [Wed, 15 Apr 2020 19:56:31 +0000 (21:56 +0200)] 
homectl: add acquired fido2 PIN to user record

If we successfully acquired the PIN for the fido2 key, let's add it to
our user record, so that we can pass it to homed, which will need it
too.

3 years agohomectl: split out pkcs#11 related code bits into own .c/.h file
Lennart Poettering [Wed, 15 Apr 2020 19:53:03 +0000 (21:53 +0200)] 
homectl: split out pkcs#11 related code bits into own .c/.h file

There's some highly specific PKCS#11 code in homectl.c. Let's split that
out, since it is easily isolatable, to make homectl.c a bit more
readable.

No funcional changes, just some moving around and renaming two functions
to make them more suitably named when exported.

3 years agohomectl: rework how we log when doing a home directory update
Lennart Poettering [Wed, 15 Apr 2020 17:35:55 +0000 (19:35 +0200)] 
homectl: rework how we log when doing a home directory update

When updating a home directory we might update the record first, then
resize the image and finally synchronize the passwords to the storage
layers. These are three individually authenticated operations. Since
each might require touching a FIDO2 or PKCS#11 key we should say what we
are doing. Hence do so.

Usually we are pretty quiet with what we do, and let's stick to that.
Hence show this information only if we actually do more than one thing.
If we only update (and do not resize/sync passwords) then let's be quiet
as usual, as the command line then sufficiently clarifies what we are
doing.

3 years agohomectl: do generic error handling/retry also when creating a home directory
Lennart Poettering [Wed, 15 Apr 2020 17:35:07 +0000 (19:35 +0200)] 
homectl: do generic error handling/retry also when creating a home directory

After all, when creating we might need interaction with the security
token too, and our initial attempt to create the user will fail, since
we do not allow interactive auth on the security token, so that we then
can print a log message and retry with interactive auth then enabled.

3 years agohomectl: show touch emoji when asking for PKCS#11 protected auth path
Lennart Poettering [Wed, 15 Apr 2020 17:34:41 +0000 (19:34 +0200)] 
homectl: show touch emoji when asking for PKCS#11 protected auth path

3 years agopam-systemd-home: print helpful message when token's PIN is locked
Lennart Poettering [Tue, 14 Apr 2020 17:26:48 +0000 (19:26 +0200)] 
pam-systemd-home: print helpful message when token's PIN is locked

3 years agohomed: add support for authenticating with fido2 hmac-secret tokens
Lennart Poettering [Tue, 14 Apr 2020 17:18:09 +0000 (19:18 +0200)] 
homed: add support for authenticating with fido2 hmac-secret tokens

3 years agohomectl: add support for enrolling FIDO2 HMAC-SECRET tokens
Lennart Poettering [Tue, 14 Apr 2020 13:54:12 +0000 (15:54 +0200)] 
homectl: add support for enrolling FIDO2 HMAC-SECRET tokens

3 years agouser-record: add fields for FIDO2 HMAC authentication options
Lennart Poettering [Tue, 14 Apr 2020 13:53:05 +0000 (15:53 +0200)] 
user-record: add fields for FIDO2 HMAC authentication options

3 years agobuild-sys: add libfido2 as optional dependency
Lennart Poettering [Tue, 14 Apr 2020 13:54:52 +0000 (15:54 +0200)] 
build-sys: add libfido2 as optional dependency

3 years agouser-record: rename JSON field "pkcs11Pin" to "tokenPin"
Lennart Poettering [Tue, 14 Apr 2020 17:05:09 +0000 (19:05 +0200)] 
user-record: rename JSON field "pkcs11Pin" to "tokenPin"

We'd like to use it for FIDO2 tokens too, and the concept is entirely
generic, hence let's just reuse the field, but rename it. Read the old
name for compatibility, and treat the old name and the new name as
identical for most purposes.

3 years agouser-record: securely erase pkcs#11 when assigned NULL too
Lennart Poettering [Tue, 14 Apr 2020 13:43:01 +0000 (15:43 +0200)] 
user-record: securely erase pkcs#11 when assigned NULL too

3 years agohomectl: add missing log messages when json_variant_format() fails
Lennart Poettering [Wed, 15 Apr 2020 17:31:33 +0000 (19:31 +0200)] 
homectl: add missing log messages when json_variant_format() fails

3 years agopkcs11-util: reduce scope of a variable
Lennart Poettering [Tue, 14 Apr 2020 12:12:09 +0000 (14:12 +0200)] 
pkcs11-util: reduce scope of a variable

3 years agomacro: add new FOREACH_POINTER() macro magic
Lennart Poettering [Tue, 14 Apr 2020 16:52:24 +0000 (18:52 +0200)] 
macro: add new FOREACH_POINTER() macro magic

This allows us to iterate through a series of specified pointers. It's a
bit like FOREACH_STRING(), but for all kinds of pointers.

3 years agolocale-util: add support for touch emoji
Lennart Poettering [Wed, 15 Apr 2020 16:45:16 +0000 (18:45 +0200)] 
locale-util: add support for touch emoji

We can use this to highlight when users are supposed to touch their
security tokens.

3 years agodoc: add recentry introduced transient settings
Yu Watanabe [Wed, 1 Jul 2020 06:05:14 +0000 (15:05 +0900)] 
doc: add recentry introduced transient settings

Also sort entries for service settings.

3 years agoMerge pull request #16303 from poettering/dbus-util-split
Yu Watanabe [Wed, 1 Jul 2020 05:15:40 +0000 (14:15 +0900)] 
Merge pull request #16303 from poettering/dbus-util-split

shared: split src/shared/bus-util.c into multiple files

3 years agotest-network: add tests for SR-IOV 16219/head
Yu Watanabe [Mon, 22 Jun 2020 10:48:34 +0000 (19:48 +0900)] 
test-network: add tests for SR-IOV

This also adjusts test_sysctl_disable_ipv6(), as rt_trap flag may be
added to routes on some environments.

3 years agonetwork: add MACAddress= setting in [SR-IOV] section
Yu Watanabe [Mon, 22 Jun 2020 10:48:33 +0000 (19:48 +0900)] 
network: add MACAddress= setting in [SR-IOV] section

3 years agonetwork: add VLANProtocol= setting in [SR-IOV] section
Yu Watanabe [Mon, 22 Jun 2020 10:26:31 +0000 (19:26 +0900)] 
network: add VLANProtocol= setting in [SR-IOV] section

3 years agonetwork: Introduce SR-IOV
Susant Sahani [Sun, 21 Jun 2020 11:17:34 +0000 (11:17 +0000)] 
network: Introduce SR-IOV

SR-IOV provides the ability to partition a single physical PCI
resource into virtual PCI functions which can then be injected in
to a VM. In the case of network VFs, SR-IOV improves north-south n
etwork performance (that is, traffic with endpoints outside the
host machine) by allowing traffic to bypass the host machine’s network stack.

3 years agosd-netlink: Add SR-IOV netlink properties
Susant Sahani [Sun, 21 Jun 2020 11:16:53 +0000 (11:16 +0000)] 
sd-netlink: Add SR-IOV netlink properties

3 years agoudev: spi: include chip select number in ID_PATH
Marc Kleine-Budde [Sun, 28 Jun 2020 19:43:34 +0000 (21:43 +0200)] 
udev: spi: include chip select number in ID_PATH

All devices behind a SPI controller have the same udev ID_PATH property.
This is a problem for predicable network names for CAN controllers.

CAN controllers, in contrast to Ethernet controllers, don't have a MAC
Address, so there's no way to tell two CAN controllers on the same SPI
host controller apart:

$ udevadm info /sys/class/net/can0
P: /devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.1/net/can0
L: 0
E: DEVPATH=/devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.1/net/can0
E: INTERFACE=can0
E: IFINDEX=3
E: SUBSYSTEM=net
E: USEC_INITIALIZED=11187199
E: ID_PATH=platform-fe204000.spi
E: ID_PATH_TAG=platform-fe204000_spi
E: SYSTEMD_ALIAS=/sys/subsystem/net/devices/can0
E: TAGS=:systemd:

$ udevadm info /sys/class/net/can1
P: /devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.0/net/can1
L: 0
E: DEVPATH=/devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.0/net/can1
E: INTERFACE=can1
E: IFINDEX=4
E: SUBSYSTEM=net
E: USEC_INITIALIZED=11192211
E: ID_PATH=platform-fe204000.spi
E: ID_PATH_TAG=platform-fe204000_spi
E: SYSTEMD_ALIAS=/sys/subsystem/net/devices/can1
E: TAGS=:systemd:

With this the chip select number is added to the ID_PATH, to make
predictable network names possible.

$ sudo udevadm info /sys/class/net/can0
P: /devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.1/net/can0
L: 0
E: DEVPATH=/devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.1/net/can0
E: INTERFACE=can0
E: IFINDEX=3
E: SUBSYSTEM=net
E: USEC_INITIALIZED=11187199
E: ID_PATH=platform-fe204000.spi-cs-1
E: ID_PATH_TAG=platform-fe204000_spi-cs-1
E: SYSTEMD_ALIAS=/sys/subsystem/net/devices/can0
E: TAGS=:systemd:

$ sudo udevadm info /sys/class/net/can1
P: /devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.0/net/can1
L: 0
E: DEVPATH=/devices/platform/soc/fe204000.spi/spi_master/spi0/spi0.0/net/can1
E: INTERFACE=can1
E: IFINDEX=4
E: SUBSYSTEM=net
E: USEC_INITIALIZED=11192211
E: ID_PATH=platform-fe204000.spi-cs-0
E: ID_PATH_TAG=platform-fe204000_spi-cs-0
E: SYSTEMD_ALIAS=/sys/subsystem/net/devices/can1
E: TAGS=:systemd: