Clean up dead code and fix a warning: unused-but-set-variable spotted by gcc4.9.

[ae]

I have not any strong objection about this. But I prefer the commented code should be left here. I added it to be compatible with current behavior, but in future we can change this.

[ngie]

In D4720#100209, @ae wrote:

I have not any strong objection about this. But I prefer the commented code should be left here. I added it to be compatible with current behavior, but in future we can change this.

I think removing the code's ok, but maybe a better XXX comment should be put in so someone could fill in the blanks later (I ran into some code in lib/libc/rpc recently like this with unimplemented paths)?

[araujo]

In D4720#100240, @ngie wrote:

In D4720#100209, @ae wrote:

I have not any strong objection about this. But I prefer the commented code should be left here. I added it to be compatible with current behavior, but in future we can change this.

I think removing the code's ok, but maybe a better XXX comment should be put in so someone could fill in the blanks later (I ran into some code in lib/libc/rpc recently like this with unimplemented paths)?

I agree with @ngie, @ae could you give me any input about how we could add some comments there? Maybe we can remove that commented code, and insert at the comments the revision that add that code and an explanation about what that should do.

what do you think @ae?

[ae]

sys/net/if_gre.c

718 ↗

(On Diff #11737)

XXX: The current implementation uses the key only for outgoing packets. But we can check the key value here, or even in the encapcheck function.

[araujo]

Address @ae comments, and for now we don't remove the code.
We will still have the gcc-4.9 warning, but for those that are cleaning it up, they will know by
the comment, it is something not implemented yet.

[ngie]

sys/net/if_gre.c

718 ↗

(On Diff #11787)

Please add a newline before the XXX per style(9).

[araujo]

Respect style(9) by @ngie.

[ngie]

Sidenote: I'm fine (personally) with removing the variables or commenting out the code under another #ifdef (e.g. KEY_SUPPORT_NOTYET), as long as the comments are there documenting the item that needs to be implemented (so two birds could be killed with one stone)

sys/net/if_gre.c

719 ↗

(On Diff #11788)

Please add a space between the XXX: and The.

[araujo]

In D4720#100407, @ngie wrote:

Sidenote: I'm fine (personally) with removing the variables or commenting out the code under another #ifdef (e.g. KEY_SUPPORT_NOTYET), as long as the comments are there documenting the item that needs to be implemented (so two birds could be killed with one stone)

I would prefer a #ifdef too, the meaning of this patch is to lower the compiler warnings. I will address @ngie suggestion together with the 'space'.

[araujo]

Wrap using #ifdef notyet those variables and statements not yet implemented.
Suggested by @ngie.

@ae what do you think about it?

[araujo]

Cosmetic fix, one tab missed.

[ngie]

LGTM. @ae should sign off too.