cpu: moved all vendor headers to CPUNAME/include/vendor/#6700
cpu: moved all vendor headers to CPUNAME/include/vendor/#6700jnohlgard merged 7 commits intoRIOT-OS:masterfrom
Conversation
doc/doxygen/riot.doxyfile
Outdated
| */cpu/lpc11u34/include/LPC11Uxx.h \ | ||
| */cpu/lpc1768/include/LPC17xx.h \ | ||
| */cpu/lpc*/include/core_cm*.h \ | ||
| */cpu/*/include/vendor/* |
There was a problem hiding this comment.
good point. will adapt
|
Could you also adapt the cppcheck filters in |
3677255 to
44c2aa6
Compare
|
done |
733ceca to
75a7be7
Compare
|
fixed excludes for externc and some empty line at EOF issues. |
|
I'm globally +1 with the goal of this PR but as you have noticed, I have some opened PRs that will be impacted : #6651, #6025, #6553, #6594, #6615, #6625. |
|
Good idea, but what about having two separate trees, instead of polluting the include tree. Also, while I can see that having to prefix vendor/ on all vendor includes has its merits with regards to readability, it may also cause a hassle when vendor headers include other vendor headers (e.g. Atmel CPUs). |
|
@aabadie: I would prefer to merge this PR first (as there are probably another 20 PRs or so, that would need to be merged first...), as right now there is the effort for the 'code quality task force'. @gebart I think this PR is not |
doc/doxygen/riot.doxyfile
Outdated
| */cpu/lpc11u34/include/LPC11Uxx.h \ | ||
| */cpu/lpc1768/include/LPC17xx.h \ | ||
| */cpu/lpc*/include/core_cm*.h \ | ||
| */vendor/* \ |
There was a problem hiding this comment.
nice cleanup here! 👍
Could you move it to the correct place to keep the list sorted?
|
@haukepetersen After looking at this properly on my PC I am starting to agree that this is a good solution. @aabadie the rebase of other PRs touching these vendor headers should be quite uncomplicated. I agree with @haukepetersen that merging this is preferable compared to merging everything else first. There are new PRs opened all the time so it would be an infinite catching up to try to get this merged. |
kaspar030
left a comment
There was a problem hiding this comment.
ACK. (Needs more ACKs, though.)
OlegHahm
left a comment
There was a problem hiding this comment.
I don't think that splitting it up into so many commits is necessary, but I leave it to Hauke if he wants to squash or not. ACK for the change.
|
@gebart, indeed, just tested the rebase, things works fine (even if I have to modify files afterwards but that was expected anyways). |
75a7be7 to
19bf9f8
Compare
|
Addressed comment by @gebart above and squashed. Now waiting for the CI once more to make sure I didn't break anything and then we should be ready to merge this. |
|
Murdock2 currently chokes as cppcheck takes longer than 5 minutes, which is Murdock2's default job timeout. The other CI's agree, so I'd say merge! |
The reason for this is to simplify handling of vendor headers in the future, e.g. by easier excluding them from doxygen builds etc...