On Thu, Sep 6, 2012 at 3:32 PM, Avi Kivity <a...@redhat.com> wrote: > On 09/02/2012 01:51 PM, Blue Swirl wrote: >> I've used the following snippet to check what happens with the last 100 >> commits: >> for i in '' 1 2 3 4 5 6 7 8 9; do for j in 0 1 2 3 4 5 6 7 8 9; do >> echo $i$j; git show HEAD~$i$j >/tmp/a; clear;head -20 /tmp/a; >> ./scripts/checkpatch.pl --root=. /tmp/a; read foo; done; done >> >> Sadly, it also shows how much stuff gets committed without checking, >> and on the other hand, the amount of false alarms. > > Maintainers should add a checkpatch invocation as a git hook (advisory > only), this could reduce the amount of violations getting into the tree.
I'd agree. But it looks like we have a few maintainers who actively oppose CODING_STYLE and HACKING and don't reject patches with violations or even inform the submitters about problems. The next level is that we have committers who commit pulls without checking, but I think the responsibility for checking patches for all possible aspects (especially since there are real, deep, technical and architectural issues to consider besides minor issues like style) should not be duplicated at every level. > > > -- > error compiling committee.c: too many arguments to function