You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When running grcov, we see in our cobertura report that for certain lines, 18446744073709551615 (unsigned max int) number of hits are reported. There is no way we have this many hits in reality, and is probably a bug.
<line number="XXX" hits="18446744073709551615">
for example. I can't share the code, but I will note that it happens in a few seemingly random places, such as an if statement, a closure, and an empty line. What's more, is that the results are not deterministic -- sometimes, we see more lines report with this errant number of hits, sometimes less lines. Although, it appears that the lines that it reports on with this large value are always from the same set of lines, it just doesn't always pick the same number of them.
I also confirmed that this problem happens with lcov reports, so it's likely not isolated to cobertura. The html results do appear to be correct though, with the errant lines reporting somewhere between 0 and 4+ hits. I can't identify any patterns in the lines with this bad value.
When running grcov, we see in our cobertura report that for certain lines, 18446744073709551615 (unsigned max int) number of hits are reported. There is no way we have this many hits in reality, and is probably a bug.
for example. I can't share the code, but I will note that it happens in a few seemingly random places, such as an
if
statement, a closure, and an empty line. What's more, is that the results are not deterministic -- sometimes, we see more lines report with this errant number of hits, sometimes less lines. Although, it appears that the lines that it reports on with this large value are always from the same set of lines, it just doesn't always pick the same number of them.I also confirmed that this problem happens with
lcov
reports, so it's likely not isolated to cobertura. The html results do appear to be correct though, with the errant lines reporting somewhere between 0 and 4+ hits. I can't identify any patterns in the lines with this bad value.Here is the command we are running:
The text was updated successfully, but these errors were encountered: