Release 2.16.4 #3299

mattkae · 2024-03-27T18:27:57Z

Release Notes

mir (2.16.4) UNRELEASED; urgency=medium

Bugfix release:
- Ignoring zero length gamma curves for KMS outputs (Fixes: Crash on wake when outputs have the same geometry #3238)
- Add hardware cursor support (Fixes: Cursor from GDM Login remains on screen after logging into Miriway #3198)
- Screenshots now respect clipping areas and rotated outputs (Fixes: grimshot save output is using the exact opposite clip area on my machine #3236)
  (Fixes: Screenshots wrong when display rotated #3259)
- Fixed custom attributes propagation on outputs
  (Fixes: Not respecting snap-name ubuntu-frame#172)
- Handle the Ubuntu 64-bit time_t apocalypse (Fixes: PPA FTBFS armhf/noble #3285)

graphics/MultiplexingDisplay: Add HW cursor support.

mattkae · 2024-03-27T19:47:22Z

@RAOF : I implemented a test in WLCS that works in main but will not be cherry-picked into this PR, as it would introduce an ABI change. How can I ignore it here? Or would I need to disable it in WLCS?

Saviq · 2024-03-28T07:34:15Z

You'll need this as well:

tests/open_wrapper: Handle the Ubuntu 64-bit time_t apocalypse #3290

You can see the build status in:
https://launchpad.net/~mir-team/+archive/ubuntu/rc

Saviq · 2024-03-28T07:35:04Z

Oh and don't worry about the conflict, the way to deal with this is to merge locally and push to main. Don't want to push anything to release/* after tagging, as that will trigger another candidate build.

AlanGriffiths

There are 14 commits, but only 5 fixes mentioned in the changelog, that seems wrong! Obviously the changelog and build related changes (like #3257) do not need mention, but omitting 10 of 14 commits seems wrong.

On closer examination there fixes included than are mentioned in the changelog, e.g.: #3261, #3238, #3240 & #3278.

I need to be convinced that #3237 and #3220 justify inclusion in a bugfix release: they are extensive and change (admittedly deprecated) APIs. Ideally, bugfix releases should only include things that can be tested in isolation and with confidence the rest of the system is stable.

mattkae · 2024-03-28T13:07:40Z

I need to be convinced that #3237 and #3220 justify inclusion in a bugfix release: they are extensive and change (admittedly deprecated) APIs. Ideally, bugfix releases should only include things that can be tested in isolation and with confidence the rest of the system is stable.

I can part with #3237, but #3220 is dear to my heart as it fixes all of the problems that I'm having on Nvidia for miracle-wm. Without it, miracle-wm doesn't function out of the box on all of my systems. I understand the change is extensive, but it felt like a good time to test if that's working across different systems.

Ignore zero length gamma curves

@RAOF

fixes #3236 Am I right to think this @RAOF ? It looks like `CPUCopyOutputSurface` is trying to have a `GL` layout. I could be wrong, but this makes my screenshots work. ## Before ![2024-02-21T17:01:19,606591722-05:00](https://github.com/canonical/mir/assets/25062299/1835506f-6725-47d4-920a-f2fdcec4ce91) ## After ![2024-02-21T17:06:29,424484886-05:00](https://github.com/canonical/mir/assets/25062299/6dfbc1e2-8fcf-40e2-8913-7a610b9a5743)

AlanGriffiths · 2024-03-28T14:47:49Z

2024-03-28T13:52:46.3245631Z E: Tried to extract package, but file already exists. Exit...

WTF?

codecov · 2024-03-28T16:00:51Z

Codecov Report

Attention: Patch coverage is 75.00000% with 13 lines in your changes are missing coverage. Please review.

Project coverage is 77.40%. Comparing base (29d10d4) to head (994e25e).
Report is 247 commits behind head on main.

Files	Patch %	Lines
src/server/graphics/multiplexing_hw_cursor.cpp	44.44%	10 Missing ⚠️
src/server/frontend_wayland/wlr_screencopy_v1.cpp	0.00%	2 Missing ⚠️
src/miral/static_display_config.cpp	91.66%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3299      +/-   ##
==========================================
- Coverage   77.51%   77.40%   -0.11%     
==========================================
  Files        1066     1064       -2     
  Lines       75358    68072    -7286     
==========================================
- Hits        58410    52694    -5716     
+ Misses      16948    15378    -1570

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

AlanGriffiths · 2024-04-02T08:51:46Z

I'm running Miriway edge/mir-pr3299 on the same configuration I encountered #3238 (to test that fix) and only the external output shows content. (However, the cursor is visible on both outputs.)

[edit]

The above happened when editing the .display (from a "sidebyside" setup). Miriway shows both outputs after a restart.

[edit 1]

This also happens with edge and stable, so not a blocker for this release: #3308

AlanGriffiths · 2024-04-02T09:58:37Z

Testing ~rc13

Note: DO NOT use the beta channel for testing - that has an ~rc14 generated with additional commits before the latest "forced push".

Bugfix	Fix confirmed	Comment
#3238	@AlanGriffiths
#3198	@mattkae	I, @AlanGriffiths, don't see any difference on Intel - the cursor remains for ~10s
#3236	@mattkae
#3259	@mattkae
canonical/ubuntu-frame#172	@AlanGriffiths
#3278	@mattkae	I, @mattkae, am running into the "EGL claimed support for DMA-Buf import modifiers, but all queries failed" exception startup. This is because of a poor platform selection on my hybrid machine. This will be fixed by the "Better platform selection" work. `miral-app` selects properly and I do not see this error.

…hat the final image is correct (#3292) fixes #3259 # Screenshots ## Left ![20240318_15h54m23s_grim](https://github.com/canonical/mir/assets/25062299/7815dfb9-ed03-491f-8422-e417d5d39d3b) ## Right ![20240318_15h54m44s_grim](https://github.com/canonical/mir/assets/25062299/412bd5a8-4935-4c4f-8c9a-6c710bf232b4) ## Inverted (Scaled) ![20240318_15h55m27s_grim](https://github.com/canonical/mir/assets/25062299/4bd4a102-c666-431f-86f8-ec005d3c457a) ## Normal ![20240318_15h56m07s_grim](https://github.com/canonical/mir/assets/25062299/f50be354-d664-4589-a75a-63e818d8597f)

Fixes: canonical/ubuntu-frame#172

…ot_respect_other_exclusive_zones WLCS test

We're *mostly* insulated from this, but our tests need to intercept calls to `open`, which means that we're exposed to some glibc internals that change under `_TIME_BITS=64`. Fortunately umockdev in the archive has already hit this, so adapt that fix to our use. Fixes: #3285

… back into main

…ther_exclusive_zones WLCS tests needed to be excluded entirely as they spuriously fail

AlanGriffiths · 2024-04-03T11:29:08Z

Testing ~rc13 (mark II)

Note: Because I force-pushed after dropping a commit and editing the changelog we're still at ~rc13

Bugfix	Fix confirmed	Comment
#3238	@AlanGriffiths
#3198	@AlanGriffiths
#3236	@mattkae
#3259	@mattkae
canonical/ubuntu-frame#172	@AlanGriffiths

AlanGriffiths · 2024-04-04T11:37:20Z

I've had to force a rebuild of Miriway/beta with ppa:mir-team/rc, but that's now up to date with this PR. (It works as expected)

Saviq

@AlanGriffiths @mattkae #3300 is biting me on Pis, but I don't see any real problems, I'm +1 to tag v2.16.4

mattkae · 2024-04-04T14:40:56Z

Screenshot items have been confirmed 🎉

Merge pull request #3206 from canonical/fix-hw-cursor

a338ea9

graphics/MultiplexingDisplay: Add HW cursor support.

mattkae changed the title ~~Release/2.16~~ Release 2.16.4 Mar 27, 2024

mattkae requested a review from AlanGriffiths March 27, 2024 18:28

mattkae force-pushed the release/2.16 branch 2 times, most recently from eda00ca to a96f43c Compare March 27, 2024 19:14

AlanGriffiths requested changes Mar 28, 2024

View reviewed changes

Saviq and others added 3 commits March 28, 2024 09:39

Merge pull request #3248 from canonical/ignore-zero-length-gamma

d7fa473

Ignore zero length gamma curves

ci: add Alpine 3.19 (#3257)

a4a1351

mattkae force-pushed the release/2.16 branch 2 times, most recently from 567be30 to 14c0550 Compare March 28, 2024 13:49

AlanGriffiths approved these changes Mar 28, 2024

View reviewed changes

AlanGriffiths marked this pull request as ready for review April 2, 2024 10:13

AlanGriffiths requested a review from a team as a code owner April 2, 2024 10:13

AlanGriffiths and others added 8 commits April 3, 2024 11:41

Fix propagation of custom attributes on outputs (#3294)

3efb89b

Fixes: canonical/ubuntu-frame#172

version: update version to 2.16.4 and update changelog

03335a6

testing: ignore surfaces_with_exclusive_zone_set_to_negative_one_do_n…

967d872

…ot_respect_other_exclusive_zones WLCS test

debian: update changelog to better reflect changes for 2.16.4

b864aa9

ci: gtest filter should ignore all exclusive zones tests until merged…

932d874

… back into main

ci: surfaces_with_exclusive_zone_set_to_negative_one_do_not_respect_o…

fbad5e5

…ther_exclusive_zones WLCS tests needed to be excluded entirely as they spuriously fail

Drop dropped fix from changelog

994e25e

AlanGriffiths force-pushed the release/2.16 branch from 31782c0 to 994e25e Compare April 3, 2024 11:09

AlanGriffiths approved these changes Apr 4, 2024

View reviewed changes

Saviq approved these changes Apr 4, 2024

View reviewed changes

AlanGriffiths added this pull request to the merge queue Apr 8, 2024

Merged via the queue into main with commit ed32a17 Apr 8, 2024
43 of 48 checks passed

AlanGriffiths deleted the release/2.16 branch April 8, 2024 10:52

Saviq restored the release/2.16 branch May 6, 2024 09:00

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Release 2.16.4 #3299

Release 2.16.4 #3299

mattkae commented Mar 27, 2024 •

edited by Saviq

Loading

mattkae commented Mar 27, 2024

Saviq commented Mar 28, 2024

Saviq commented Mar 28, 2024

AlanGriffiths left a comment

mattkae commented Mar 28, 2024

AlanGriffiths commented Mar 28, 2024

codecov bot commented Mar 28, 2024 •

edited

Loading

AlanGriffiths commented Apr 2, 2024 •

edited

Loading

AlanGriffiths commented Apr 2, 2024 •

edited by mattkae

Loading

AlanGriffiths commented Apr 3, 2024 •

edited by mattkae

Loading

AlanGriffiths commented Apr 4, 2024

Saviq left a comment

mattkae commented Apr 4, 2024

Release 2.16.4 #3299

Release 2.16.4 #3299

Conversation

mattkae commented Mar 27, 2024 • edited by Saviq Loading

Release Notes

mattkae commented Mar 27, 2024

Saviq commented Mar 28, 2024

Saviq commented Mar 28, 2024

AlanGriffiths left a comment

Choose a reason for hiding this comment

mattkae commented Mar 28, 2024

AlanGriffiths commented Mar 28, 2024

codecov bot commented Mar 28, 2024 • edited Loading

Codecov Report

AlanGriffiths commented Apr 2, 2024 • edited Loading

AlanGriffiths commented Apr 2, 2024 • edited by mattkae Loading

Testing ~rc13

AlanGriffiths commented Apr 3, 2024 • edited by mattkae Loading

Testing ~rc13 (mark II)

AlanGriffiths commented Apr 4, 2024

Saviq left a comment

Choose a reason for hiding this comment

mattkae commented Apr 4, 2024

mattkae commented Mar 27, 2024 •

edited by Saviq

Loading

codecov bot commented Mar 28, 2024 •

edited

Loading

AlanGriffiths commented Apr 2, 2024 •

edited

Loading

AlanGriffiths commented Apr 2, 2024 •

edited by mattkae

Loading

AlanGriffiths commented Apr 3, 2024 •

edited by mattkae

Loading