On 24.04.2019 23:40, Yi-Hung Wei wrote: > On Wed, Apr 24, 2019 at 6:00 AM Ilya Maximets <[email protected]> wrote: >> >> Recent commit in "sparse" broke checking the OVS sources, because >> 'make' uses '-MD' flag to generate dependencies as a side effect >> within compilation commands, but "sparse" skips all the build commands >> that contains '-MD' and friends. >> Let's revert the bad commit as a workaround before installing "sparse" >> in TravisCI. >> >> Additionally fixed a false-positive: >> ./lib/bitmap.h:64:29: error: shift too big (64) for type unsigned long >> >> CC: Yi-Hung Wei <[email protected]> >> Fixes: 879e8238dfdf ("travis: Update sparse git repo") >> Signed-off-by: Ilya Maximets <[email protected]> >> --- >> .travis/linux-prepare.sh | 9 ++++++++- >> lib/bitmap.h | 2 +- >> 2 files changed, 9 insertions(+), 2 deletions(-) >> >> diff --git a/.travis/linux-prepare.sh b/.travis/linux-prepare.sh >> index 50189c421..d41799e3e 100755 >> --- a/.travis/linux-prepare.sh >> +++ b/.travis/linux-prepare.sh >> @@ -8,7 +8,14 @@ set -ev >> # environments claim to have LLVM (llvm-config exists and works) but >> # linking against it fails. >> git clone git://git.kernel.org/pub/scm/devel/sparse/sparse.git >> -cd sparse && make -j4 HAVE_LLVM= install && cd .. >> +cd sparse >> +# Commit bb1bf748580d ("cgcc: gendeps for -MM, -MD & -MMD too") makes >> +# sparse ignore almost all source files, because 'make' uses '-MD' to >> +# generate dependencies as a side effect within compilation commands. >> +git revert bb1bf748580d --no-commit >> +git diff HEAD >> +make -j4 HAVE_LLVM= install >> +cd .. >> >> pip install --disable-pip-version-check --user six flake8 hacking >> pip install --user --upgrade docutils >> diff --git a/lib/bitmap.h b/lib/bitmap.h >> index eb469339e..0b62066c6 100644 > Hi Ilya, > > Thanks for figuring out the issue. Instead of reverting that faulty > commit, how about let's use the latest release of sparse? In this > way, we do not need to worry too much about the latest spare > development, and we will use the next sparse release when we have > tested it. > > That faulty commit is in spare v0.6.1-rc1 and afterwards. So, maybe > we can checkout v0.6.0 tag and use it for Jarvis? How do you think? >
Hi. I thought about this before sending the patch. The concern I have is that "sparse" is not that actively developed project without specified release process. It could take years for them to release a new minor version. At the same time development is contiguous. i.e. we could get a big number of new patches in "sparse" without even a minor release version change. For example, between v0.5.0 and v0.5.1 passed 3.5 years with ~400 commits. In general, I hope that issue will be fixed in a near future and we'll be able to just drop this workaround. Best regards, Ilya Maximets. _______________________________________________ dev mailing list [email protected] https://mail.openvswitch.org/mailman/listinfo/ovs-dev
