Closed PiJoules closed 4 years ago
My wild guess is that it might be related to this fix #38 that I haven't merged yet -- there's a bug in writing out data file with very long symbol names etc. I will try to actually test that case myself soon and merge the fix.
Can you try with master branch now? If my guess was correct, then the merge of #38 should have fixed this. If it still does not work, then getting your json files somehow would be useful for fixing the issue.
Hi, I see the same error message (building https://gitlab.cern.ch/lhcb/LHCb) with today's master of ClangBuildAnalyzer 858cc42908337b385a495bf70cc90363d71e60c9.
You can download my json files (for a few days) here if you want https://cernbox.cern.ch/index.php/s/JVLy1bITaVOoZ7f
EDIT: I should add I tried the same json files on a different machine and here things work …
EDIT2: my bad, I exceeded the quota in my home directory which rightfully causes the corruption. After solving my filesystem issues I compared 5b2550eb6b84de6ea884511bc1eaed9bb75c6777 = v1.2.0 against 280e5c5d48d1a9d8807030845e4fc6ceb84b410a = master now. With the old version I get the checksum mismatch error. with the new version I don't get the error. So if @PiJoules's error is the same as the one I see with v1.2.0 then this should be fixed in the current master
I'll assume this is fixed. Comment on the issue if it's still happening!
I'm attempting to profile fuchsia compilations and am running into:
Do you have any ways of going about debugging this? This is from a fresh ClangBuildAnalyzer checkout and there didn't seem to be any issues with building or running it. I'm able to assert that
-ftime-trace
is working as intended since I can see the.json
files in my build. The capture file I also made is about 470 MB.