Speed up REPL Tests in CI by grouping apps in a single target #41501
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
PR #41501: Size comparison from a95f163 to 72a847b Full report (37 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, efr32, esp32, nrfconnect, nxp, psoc6, qpg, realtek, stm32, telink)
|
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #41501 +/- ##
==========================================
- Coverage 51.07% 51.06% -0.02%
==========================================
Files 1384 1385 +1
Lines 100886 100883 -3
Branches 13049 13054 +5
==========================================
- Hits 51531 51516 -15
- Misses 49355 49367 +12 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
PR #41501: Size comparison from a95f163 to 9a73a41 Full report (10 builds for cc13x4_26x4, cc32xx, nrfconnect, realtek, stm32)
|
soares-sergio
force-pushed
the
improve_ci
branch
from
9a73a41 to
72a847b
Compare
|
PR #41501: Size comparison from a95f163 to e16a5fc Full report (33 builds for bl602, bl702, bl702l, cc32xx, efr32, esp32, nrfconnect, nxp, psoc6, qpg, realtek, stm32, telink)
|
|
PR #41501: Size comparison from a95f163 to 7452467 Increases above 0.2%:
Full report (37 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, efr32, esp32, nrfconnect, nxp, psoc6, qpg, realtek, stm32, telink)
|
|
PR #41501: Size comparison from a95f163 to 8d86273 Increases above 0.2%:
Full report (34 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, efr32, esp32, nrfconnect, psoc6, qpg, realtek, stm32, telink)
|
|
PR #41501: Size comparison from a95f163 to 3a964c5 Full report (12 builds for cc13x4_26x4, cc32xx, nrfconnect, qpg, realtek, stm32)
|
|
PR #41501: Size comparison from a95f163 to b0c0fd3 Full report (12 builds for cc13x4_26x4, cc32xx, nrfconnect, qpg, realtek, stm32)
|
|
PR #41501: Size comparison from a95f163 to 257a1d6 Increases above 0.2%:
Full report (34 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, efr32, esp32, nrfconnect, psoc6, qpg, realtek, stm32, telink)
|
soares-sergio
changed the title
…es.py in separate calls and deleting out dir
|
PR #41501: Size comparison from 1f47e8d to fe60440 Full report (9 builds for cc13x4_26x4, cc32xx, realtek, stm32)
|
andy31415
reviewed
Oct 21, 2025
| # } | ||
|
|
||
| if (enable_linux_apps_for_ci_build) { | ||
| group("linux_apps_for_ci") { |
There was a problem hiding this comment.
I would not separate out a CI target - we want builds to still look like "ask to build closure/bridge/..." it should just happen faster.
That would allow us to change tests.yaml without needing to also change BUILD.gn (i.e. have less coupling)
There was a problem hiding this comment.
Looking at changes needed for build_examples, I think we can completely drop this part.
andy31415
reviewed
Oct 21, 2025
| # extra_build_deps += [ ":linux_camera_controller_app" ] | ||
| # } | ||
|
|
||
| if (enable_linux_apps_for_ci_build) { |
There was a problem hiding this comment.
can we just enable this by default?
|
PR #41501: Size comparison from 1f47e8d to d213e56 Full report (9 builds for cc13x4_26x4, cc32xx, realtek, stm32)
|
|
PR #41501: Size comparison from 1f47e8d to ceae248 Full report (5 builds for cc32xx, realtek, stm32)
|
|
PR #41501: Size comparison from 1f47e8d to 070ed97 Full report (9 builds for cc13x4_26x4, cc32xx, realtek, stm32)
|
andy31415
reviewed
Oct 22, 2025
andy31415
reviewed
Oct 22, 2025
soares-sergio
force-pushed
the
improve_ci
branch
from
070ed97 to
f0f7798
Compare
|
PR #41501: Size comparison from 9dda583 to 16645ca Full report (31 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, esp32, nrfconnect, psoc6, qpg, realtek, stm32, telink)
|
|
PR #41501: Size comparison from 9dda583 to 688dbf2 Full report (34 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, efr32, esp32, nrfconnect, psoc6, qpg, realtek, stm32, telink)
|
|
PR #41501: Size comparison from 9dda583 to 5c5b783 Full report (37 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, efr32, esp32, nrfconnect, nxp, psoc6, qpg, realtek, stm32, telink)
|
|
PR #41501: Size comparison from 9dda583 to c2d8104 Increases above 0.2%:
Full report (37 builds for bl602, bl702, bl702l, cc13x4_26x4, cc32xx, efr32, esp32, nrfconnect, nxp, psoc6, qpg, realtek, stm32, telink)
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
This PR significantly speeds up the build of example apps in the "REPL Linux Tests" CI job by introducing a unified build that compiles multiple applications within the same output directory in a single target, rather than building each example from scratch. This change allows the build system to reuse shared libraries and output directories, drastically reducing build times.
The build time for each individual app when building from scratch is about 6min.
We are building 28 apps "REPL Tests" workflow, which means a around 2h40min of build time.
From the 28 apps, I was able to aggregate 19 in a single target
linux_apps_for_ciin the rootBUILD.gnfile. This reduces the build time of those apps in theREPL Testsworkflow from ~2h to ~25min.You can build the
linux_apps_for_citarget locally by executing:Background
CI in the Matter SDK is way too slow. In one of my recent Pull Requests, it took 4h to run and affected my ability to move faster and land code upstream. So I'm investigating investigating ways to make CI faster.
I presented some stats in the last
swttstandup and I created slack channel (#swtt-ci) to group people interested in helping with this effort. Please join if you would like to help.Currently, the biggest CI offender is the REPL tests:
Limitations of this PR
We can still do better. I was not able yet to aggregate the following apps in the same target. Some of them require custom
args.gnior dependencies that need to be refactored. I left them alone for now and they are still building individually from scratch and taking a long time:Testing
Relying on CI to validate