Conversation
|
Coverage does not work, so maybe the last rebase from 3.x is missing? |
Either this or the commit that fixes that in 3.x has not yet been merged to trunk |
|
Well, it came as a surprise : there were some wrong LCOV terminators (LCOV_EXCL_END instead of LCOV_EXCL_STOP), and even a missing LCOV_EXCL_START). |
|
@GitMensch Is it okay if I commit to SVN the non-CI part of this PR ? |
| - name: Adjust testsuite | ||
| run: | | ||
| cd tests/testsuite.src | ||
| sed -i '' '/AT_SETUP(\[runtime check: write to internal storage (1)\])/a \ |
There was a problem hiding this comment.
Best is to always leave a note why we exclude something.
There was a problem hiding this comment.
Please explain why this is needed, especially if it isn't in 3.x
There was a problem hiding this comment.
As stated in my previous comment:
Then there was also an unused exclusion (statment.def) - which is weird since it appears to be used in 3.x. So I prefered to just request the error to be ignored.
And I aso add to pass an extra option to ignore the negative count.
So it's just because there are more errors/warnings in 4.x - don't know why.
|
What about MinGW? We found the whole BDB part to be broken and the runner using a different bash and buildtools in 3.x CI |
Haven't touched MinGW in this PR. That would need to be solved in 3.x first. |
ddeclerck
force-pushed
the
fix_gc4_ci
branch
2 times, most recently
from
b7cc45f to
073c07f
Compare
|
three lcov markers fixed upstream (actually one was wrong and was removed) |
ddeclerck
force-pushed
the
fix_gc4_ci
branch
6 times, most recently
from
13de892 to
dc2d2c2
Compare
|
@GitMensch Is this one okay to merge too ? |
|
feel free to push the workspace definitions as those are this mirror's branch only - but I'm hesitate to have a difference beetween GC3 and GC4 in that Makefile - why should we need that? |
2 participants
No description provided.