Last commit made on 2020-01-23
Get this branch:
git clone -b master

Branch merges

Branch information


Recent commits

2c7c168... by Maciej Borzecki 16 hours ago

Merge pull request #8037 from bboozzoo/bboozzoo/skip-nakedret

travis, tests, run-checks: skip nakedret

82c2943... by Maciej Borzecki 18 hours ago

travis, tests/unit/go: nakedret fix has landed

Drop the SKIP_FLAGS, but leave the knobs (and fixes).

Signed-off-by: Maciej Borzecki <email address hidden>

2d701c4... by Maciej Borzecki 22 hours ago

travis, tests/unit, run-checks: skip nakedret until upstream is fixed

Nakedret uptream broke and reports false positives for nested function literals
[1]. A PR with a fix is open already, but until it lands, we need to skip
nakedret errors.


Signed-off-by: Maciej Borzecki <email address hidden>

fc623b9... by Maciej Borzecki 22 hours ago

testutil: fix valid nakedret issues

Signed-off-by: Maciej Borzecki <email address hidden>

a1df748... by Michael Vogt on 2020-01-23

Merge pull request #8014 from mvo5/c20-snap-recovery-encrypt-20.04

tests: run `uc20-snap-recovery-encrypt` test on 20.04-64 as well

6405f9d... by Zygmunt Krynicki on 2020-01-22

Merge pull request #8031 from zyga/tweak/mount-ns-update

tests: update mount-ns test tables

793e9ca... by Claudio Matsuoka on 2020-01-22

tests: reuse existing mounts and fix key test error

We're now mounting newly created filesystems in snap-bootstrap, so
use these instead of remounting to check content deploy. Also unmount
the filesystem before closing the encrypted device.

Signed-off-by: Claudio Matsuoka <email address hidden>

ac3e5f5... by Ian Johnson on 2020-01-22

Merge pull request #8027 from mvo5/snap-disable-auto-imports-during-install-mode

snap: disable auto-import in uc20 install-mode

Do not try to auto-import block devices in install mode on UC20.
This will work-around the kernel crash LP:#1860231. We also do
not want to enable auto-import in this mode until we have a

This works around the kernel crash in LP 1860231.

d9e092b... by Zygmunt Krynicki on 2020-01-22

tests: update mount-ns test tables

Looking at the diff I see the following changes, on core systems:

 - addition of /system-data/etc/systemd/journald.conf.d
 - addition of /system-data/etc/environment
 - duplication of /system-data/etc/systemd/system (probably a bug)
 - lots of shifting of re-numbered numeric values

I think we could introduce yet another tweak, where certain numeric
values would be encoded only as delta from previous row, for instance
propagation values - this would reduce this diff down to a few lines

In addition I took a snapshot of core-20 mount table. Casual diff from
core18 reveals a number of alterations:

 - /boot/grub is mounted from /boot/EFI
 - /etc/machine-id mount from tmpfs (probably a bug, core18 was
   persistent from system-data)
 - /usr/lib/{firmware,modules} (core20) vs /lib/{firmware,modules}
   (core18) mounted from kernel snap - probably ok but unsure if

In addition I noticed we had /etc/fstab mounted from /image.fstab from
some tmpfs. Not sure what this is or how it got there.

As before, the tests are not enabled for core-* devices because of two
buggy tests leaving behind junk on core systems. I'm investigating if we
can fix those now.

Signed-off-by: Zygmunt Krynicki <email address hidden>

38ea16c... by Maciej Borzecki on 2020-01-22

Merge pull request #8030 from chipaca/command-chain-service

tests: add a command-chain service test

We didn't have a spread test for a service using command-chain.
This adds that.