Closed bohde closed 5 years ago
:pushpin: Commit fc8fdd5 has been approved by joshbohde
:hourglass: Testing commit fc8fdd5 with merge 37e4001...
:broken_heart: Test failed - status-jenkins
:hourglass: Testing commit fc8fdd5 with merge 3d7c1b8...
@reconfig-bot r+
:pushpin: Commit 7ae8162 has been approved by zynaptic
:hourglass: Testing commit 7ae8162 with merge e94488c...
:broken_heart: Test failed - status-jenkins
@reconfig-bot r+
:pushpin: Commit fdc3661 has been approved by pwaller
:hourglass: Testing commit fdc3661 with merge c1bd42b...
@reconfig-bot r-
:broken_heart: Test failed - status-jenkins
I've made the xclbin prerequisite to the clocking report an order only one rather than removing it completely. However, looking at the way the clock information is extracted I suspect that may not be the correct input file to be using - IIRC the xclbin file is a zipped file of all the build artifacts to be used for AFI generation, so I'm not sure how the extraction script is supposed to work with that.
I believe it takes advantage of the fact that the line of interest appears in the binary, and sed seems to be OK with it. My preferred method would have been to use the xcl splitting tooling, but for some reason we decided this was good enough for now. We can always revisit that decision, but as I understand it is meant to work unless another mistake has crept in.
OK - I think this finally fixes things. Going to reinstate #233 in the release notes and give it the r+. Should then be good for a public release next week.
@reconfig-bot r+
:pushpin: Commit bea9254 has been approved by zynaptic
:hourglass: Testing commit bea9254 with merge bc83c38...
:sunny: Test successful - status-jenkins Approved by: zynaptic Pushing bc83c382959f964c2391330c9764441989037650 to master...
@reconfig-bot r+