Review sprint #1

[niemeyer]

##Summary

We have quite a few pending PRs open ranging from major features all the way to trivial tweaks. To ensure these get addressed timely, we’ll have a more focused effort over the next few days to work through them. If you are an author or a reviewer, let’s please collaborate towards that.

Below we have the PRs which were open when the sprint started this morning, and perhaps a few important ones that were opened later and then integrated into the sprint.

For any required updates unrelated to status changes, please comment below. Status changes will be updated regularly and automatically.

##Status of each PR:

• Pending review
• Closed without merging
• Review in progress
• Reviewed and accepted
• Merged
• Broken spread tests (click to see)

Implement snap-update-ns

• #3136: snap-confine: add code to ensure that / or /snap is mounted “shared”
• #3138: interfaces/mount: add Change.Perform
• #3209: interfaces/mount: add partial implementation of Change.Needed
• #3224: interfaces/mount: write current fstab files with mode 0644
• #3225: cmd/snap-update-ns: add actual implementation
• #3231: cmd/snap-discard-ns: remove current profile when cleaning up
• #3232: cmd/snap-confine: write current mount profile.
• #3233: packaging: remove current mount profiles when purging
• #3234: overlord/snapstate/backend,interfaces/mount: move ns management code.

Aliases v2

• #3137: overlord: switch to aliases v2 tasks for install/refresh etc ops plus transition
• #3191: many: implement snap alias <snap.app> (aliases v2)
• #3192: many: implement snap unalias (aliases v2)
• #3220: many: implement snap prefer (aliases v2)
• #3237: many: adjust /aliases and “snap aliases” to aliases v2, also some cleanup
• #3245: many: aliases v2 cleanups

Interface hooks

• #3212: api, ifacestate: resolve disconnect early
• #3119: interfaces: API additions for interface hooks
• #3120: interfaces/hooks: expose attrs to the interface API, snapctl enhancements (step #4)

##Desktop

• #2592: many: add support for session dbus services via snaps
• #3177: cmd/snap: make users Xauthority file available in snap environment
• #3244: many: fix review comments from PR #3177

##Repair capability

• #3211: assertions: add “repair” assertion
• #3221: snap-repair: add snap-repair run
• #3226: snap-repair: add snap-repair run

Interfaces

• #2749: interfaces/default: allow mknod for regular files, pipes and sockets (LP: #1636540)
• #2837: interfaces/apparmor: allow reading from ecryptfs
• #2869: interfaces/builtin: add online-accounts-service interface
• #2929: interfaces/builtin: add storage-framework-service interface
• #2941: interfaces/builtin: add network-status interface
• #2969: interfaces: mediate netlink sockets via seccomp
• #3045: interfaces: add random interface
• #3051: interfaces: add consoles interface
• #3205: tests: add openvswitch interface spread test
• #3208: interfaces: recover panics when sanitizing plugs and slots
• #3218: interfaces: allow writing to /run/systemd/journal/stdout by default
• #3219: interfaces/i2c: allow modifying device-specific sysfs entries

##Other

• #2833: many: allow core refresh.schedule setting
• #2895: client,cmd/snap: first pass at client messaging around modes
• #3018: tests: add empty initrd failover test
• #3026: cmd/snap-confine: use defensive argument parser
• #3111: snapd: initial implementation for systemd software watchdog for snapd
• #3117: tests: parameterize gadget snap channel
• #3141: many: show available “tracks” in snap info
• #3150: In-snap bash tab completion
• #3156: many: support debian in our CI
• #3171: snapstate: normalize gadget defaults
• #3197: store: retry on connection reset too
• #3214: cmd/snap: iterate interface tab completion
• #3223: many: mount squashfs as read-only
• #3217: snap: support for snap tasks --last=…
• #3222: many: fix test cases to work with different DistroLibExecDir
• #3230: tests: extend network-control spread test to cope with network namespaces
• #3240: snap: add snap refresh --time option
• #3242: tests: tweak time for econnreset test a bit more
• #3235: overlord/hooks: make sure only one hook for given snap is executed at a time.

##Stats

55 total → ( 2 waiting → 8 reviewing → 0 reviewed → 41 merged ) → 45 closed

[niemeyer]

Updates to the board since creation:

• Now broken spread tests are highlighted and linked to with
• Progress stats at the bottom of the board

[niemeyer]

[pedronis]

@chipaca @mvo in the morning I need 2nd reviews of:

#3192: many: implement snap unalias (aliases v2)
#3239 many: show alias changes on snap alias/unalias (aliases v2)

after that comes:

#3220 many: implement snap prefer (aliases v2)

but probably easier if the other two are merged first, also become it needs its own tweaks to show what changes happened…

[zyga]

Now broken spread tests are highlighted and linked to with

Those get stale as the tests are re-tried. Do I understand correctly that the board is now automatically updated by a script of some sort?

[pstolowski]

@niemeyer Could you please move #3119 and #3120 to the “Interface hooks” section?

[niemeyer]

@zyga Yes, there’s a script that updates it. Still manually run for now, but mostly automatic. There was indeed a bug in the logic checking for statuses, and it’s now fixed. Thanks for raising the issue.

@pstolowski Done!

[niemeyer]

@zyga Any chance we can get the error handling simplifications for snap-confine that we discussed back in early march and tersely described in #3026? This is waiting there for a while now… is there anything blocking it, or is it just waiting on the pipeline?

[zyga]

It’s just waiting in the pipe. I’ll send it today.

[niemeyer]

Apologies for the lack of updates on the board last week, and also the delays on acting on the reviewing itself. Some of us were in a planning sprint last week and that got everything delayed. Despite that, several PRs moved on thanks to the good folks that continue to push things forward meanwhile.

The board was just brought up-to-date, and we’re now picking this up again and driving to a conclusion in the next few days.

[niemeyer]

The first reviewing sprint is now finished!

We’ve run the sprint for 3 weeks, except we had an unrelated sprint in the middle which took the attention of some of us away from it for quite a while. Even then, we’ve managed to review and close 45 pull requests within that time frame, which is in fact more than we had on the board when the sprint started (after the first few days we added a few more). A few still remain, and will be sorted organically over the next days.

Thank you very much to everyone who joined ensuring PRs that were long open were properly sorted.