-
Notifications
You must be signed in to change notification settings - Fork 2.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix xcodebuild to actually work on Darwin even without ZAP installed. #24539
Fix xcodebuild to actually work on Darwin even without ZAP installed. #24539
Conversation
PR #24539: Size comparison from 6648ef2 to 6e7ce8f Increases (6 builds for nrfconnect, psoc6, qpg, telink)
Decreases (4 builds for psoc6, telink)
Full report (33 builds for cyw30739, linux, mbed, nrfconnect, psoc6, qpg, telink)
|
6e7ce8f
to
7e77004
Compare
7e77004
to
c56a2e3
Compare
PR #24539: Size comparison from 6648ef2 to c56a2e3 Increases (5 builds for nrfconnect, qpg, telink)
Decreases (1 build for esp32)
Full report (35 builds for cyw30739, esp32, linux, mbed, nrfconnect, psoc6, qpg, telink)
|
c56a2e3
to
a103473
Compare
PR #24539: Size comparison from cbc8fc4 to a103473 Increases (7 builds for bl702, psoc6, telink)
Decreases (6 builds for bl602, nrfconnect, psoc6, telink)
Full report (54 builds for bl602, bl702, cc13x2_26x2, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
|
a103473
to
d131906
Compare
PR #24539: Size comparison from f683392 to d131906 Increases (11 builds for bl602, bl702, esp32, psoc6, telink)
Decreases (4 builds for bl602, esp32, psoc6, qpg)
Full report (43 builds for bl602, bl702, cc13x2_26x2, cyw30739, efr32, esp32, k32w, linux, mbed, nrfconnect, psoc6, qpg, telink)
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please update summary to explain what was wrong and the fix.
I am trying to understand the code by review, however there seem to be several changes like env location updates, codegen logic updates and xcode updates. I would like to understand so if we ever change xcode again, this kind of error does not occur anymore.
I added a bit of a description here, however I am not sure about the cpp-app codegen. Why do we codegen those, it was not needed before. We are going down the wrong path if we start adding more codegen pregen. |
I tried to make it clearer a bit what the problem was. Sorry; I was dealing with this very late last night. To be clear, I view this as a temp fix to unblock us for now, until we fix the "libCHIP.a depends on codegen" bug in the build system. Thank you for filling in the summary! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Approving with side discussion on followup to clean up parameters on codepregen at least.
The approach is as sane as we can get for now and want this to unblock basically now.
…ose are not needed anymore.
* Remove dependency of libChipController on app-specific codegen. * Remove the unused cluster id member from ClusterBase. * Make ClusterBase constructor public. * Use ClusterBase instead of generated cluster structs in src/controller. * Remove the CHIPClusters.h includes and codegen dependency. * Revert the non-workflow changes from PR ##24539, since those are not needed anymore. * Regenerate generated files.
…ect-chip#24555) * Remove dependency of libChipController on app-specific codegen. * Remove the unused cluster id member from ClusterBase. * Make ClusterBase constructor public. * Use ClusterBase instead of generated cluster structs in src/controller. * Remove the CHIPClusters.h includes and codegen dependency. * Revert the non-workflow changes from PR #project-chip#24539, since those are not needed anymore. * Regenerate generated files.
Darwin builds without ZAP installation fails and this failure is not detected by CI. This is because when building libCHIP.a we insist on running codegen even though nothing that depends on codegen should be present in libCHIP.a.
The fix for now is to change
zzz_generated/darwin
to be the output of pregenerate and set chip_code_pre_generated_directory to that directory when building libCHIP.a.Changes: