af: initial import of global AF_ definition header#3660
af: initial import of global AF_ definition header#3660miri64 merged 2 commits intoRIOT-OS:masterfrom
Conversation
e2d3712 to
ce51fbd
Compare
|
I removed most of the address families defined in I also removed the |
There was a problem hiding this comment.
OT: the net/ipv6 and net/ng_ipv6 will get merged to net/ipv6 soon-ish right?
There was a problem hiding this comment.
No net/ipv6 was taken out of GNRC and is independent of it (it contains general definitions like addresses, headers and related functions). net/ng_ipv6.h was moved to net/gnrc/ipv6.h yesterday.
AFAIU, |
|
@cgundogan ok, I misinterpreted it as part/step of the migration from the old stack. |
There was a problem hiding this comment.
ok, really nitpicking: too much newlines
The old stack is gone since about 2 weeks or so. The only things left are a socket_base skeleton, which will be hopefully removed as soon as the new connection API was removed, and |
There was a problem hiding this comment.
I didn't tested with oonf yet, but doesn't removing the both includes will harm using this package?
There was a problem hiding this comment.
I did not test it either, but judging from the diff it looked like those were only included for the AF_ macros.
There was a problem hiding this comment.
Seems to be broken in master.
There was a problem hiding this comment.
Seems to be broken in master.
No I'm just too stupid to use it ;-P
|
needs rebase :( |
|
Done. |
|
nice thx :) seems we need |
|
ok, since we have no |
|
I could have told you that (but failed to update the issue too). |
|
np :) |
|
I would say please squash and lets see what travis says |
Saw it, though that it was implementation dependent and shrugged it off :D |
|
Squashed |
|
nice, I would say ACK |
af: initial import of global AF_ definition header
There was a problem hiding this comment.
How did these additional white trailing spaces passed the CI?
There was a problem hiding this comment.
Some versions of git (my local e.g.) register patch files as binary files and thus ignore their content for patching, diff and whitespace checks.
There are a lot of redefinitions of the
AF_macros. This PR once and for all defines them all in a central place that is not dependent from POSIX. I chose to define them in a similar fashion as they are defined in Linux: as anenum, that is also exposed to the preprocessor. This way the arbitrary numbering is left to the compiler, while the predecessor also can access these constants (and as a bonus can simply convert them to strings with the#operand).