nixos/tests/phosh: init #173108
nixos/tests/phosh: init #173108
Conversation
|
@ofborg test phosh |
| phone.screenshot("01lockinfo") | ||
|
|
||
| with subtest("Check that we can unlock the screen"): | ||
| phone.send_chars("${pin}", delay=0.2) |
There was a problem hiding this comment.
Consider leaving a comment to explain why a longer delay is needed.
There was a problem hiding this comment.
Sure thing. I'm not entirely sure what the cause of the jankiness is, but I suspect it has something to do with the animations.
There was a problem hiding this comment.
What happens if the delay kwarg is not used?
There was a problem hiding this comment.
Seems to me like it defaults to 0.01s by default
|
@zhaofengli LGTM. can you please rebase? |
|
@ofborg test phosh |
|
The test doesn't work now before #183913 is merged. |
|
ok didn't get that before. So let's get that merged in first. |
|
#183913 has now been merged. |
|
@ofborg test phosh |
|
It seems like the test ran through on arm but didn't build on x86. can you please have a look @zhaofengli ? |
|
Hmm, looking at the logs there seem to be a huge number of rebuilds which I didn't encounter locally, and it failed building an unrelated package: |
|
@ofborg test phosh |
|
This is passing:
This is failing on aarch64 ofborg: https://logs.nix.ci/?key=nixos/nixpkgs.173108&attempt_id=455cc027-e508-480b-9d56-786702b17dd7 Screenshots in `$out`
|
Ok, this time the ofborg error is related to this actual test... as long as this is supposed to work on aarch64, we gonna get this test fixed, regardless if it works on some other machine. In the future it will has to be solid in the CI, too. Any chance that there might be some big-noisy-machine-multicore-flakiness? |
|
I ran the test on my aarch64-linux pinebook-pro. The first hurdle that qemu seems to fail on big.LITTLE architecture. To workaround that, disable the little cores: Then when I run the test, I'm able to reproduce a failure, but it looks different to ofBorg's: Screenshots in `$out`
The test log showed phoc segfaulted. I'll keep looking. |
|
If I remove Screenshots in `$out`
I will re-run this a few times to check it's consistent. |
|
|
||
| systemd.services.phosh = { | ||
| environment = { | ||
| # Accelerated graphics fail on phoc 0.20 (wlroots 0.15) |
There was a problem hiding this comment.
Ah, it sounds like you've tested this against #183907 (or similar) and found that the test would fail?
There was a problem hiding this comment.
Yeah, I was testing with #183907 and without using the software renderer it failed saying "Could not find needed globals" with outputs being NULL:
phone # [ 118.480569] systemd[838]: Starting Phosh, a shell for mobile phones...
phone # [ 118.746481] .phosh-wrapped[4390]: Could not find needed globals
phone # outputs: 0, layer_shell: 0x118d100, idle_manager: 0x1196510, inhibit: 0x1196490, xdg_wm: 0x118d000, xdg_output: 0x118d080, wlr_output_manager: 0x1142400, wlr_foreign_toplevel_manager: 0x114f740[ 118.816159] traps: .phosh-wrapped[4390] trap int3 ip:4a044f sp:7fffda8b42c0 error:0 in .phosh-wrapped[419000+c1000]
It wasn't consistent, and per "Accelerated graphics fail on phoc 0.20 (wlroots 0.15)" it sounds like this test is sensitive to phoc 0.17 (status quo) vs phoc 0.20 (upcoming: #183907). In which case, I expect it's preferable to do the phoc 0.20 upgrade (with manual testing), then add this test. |
zhaofengli
force-pushed
the
phosh-test
branch
from
b75c5f8
to
25cb2fd
Compare
|
@ofborg test phosh |
|
I like this! Great job adding this :) |
Description of changes
Makes Phosh updates easier to test.
Screenshots in
$outThings done
sandbox = trueset innix.conf? (See Nix manual)nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage./result/bin/)nixos/doc/manual/md-to-db.shto update generated release notes