svn commit: r353732 - head/sys/net

[Enji Cooper]

> On Oct 22, 2019, at 08:26, Gleb Smirnoff <glebius at freebsd.org> wrote:
> 
> On Tue, Oct 22, 2019 at 07:53:31AM -0700, Enji Cooper wrote:
> E> > Author: glebius
> E> > Date: Fri Oct 18 15:20:24 2019
> E> > New Revision: 353732
> E> > URL: https://svnweb.freebsd.org/changeset/base/353732
> E> > 
> E> > Log:
> E> >  Make rt_getifa_fib() static.
> E> 
> E> Why should it be made static? The answer to this question is missing from the commit message.
> 
> Because a function used only in a scope of a single file and not
> intended to be used outside should be made static.

Gleb,

Yep :)! I appreciate your doing this because it limits scopes and calls, allowing compilers to optimize away unnecessary calls and external symbol linkage.

My concern is coming back to the time old issue of providing sufficient context behind commits when making changes. Looking at the code change you made, I understand what you did... I just lack the context as to why the changes was made.

This issue is true for the recent KPI rototilling related to interface property enumeration. If I hadn’t seen your mailing list RFC, I would lack all context as to why the changes are being made. Even a quick reference to a one page write up on the wiki, a proposal, or release notes would be wonderful moves towards documenting why interface changes are being made.

I want to thank you so very much for all of the awesome work you have done and are continuing to do in the network space — I really appreciate it a lot! Again, I just wish the reasoning behind the work was made more apparent to onlookers as to why changes are being made.

Cheers :)!
-Enji