Thoughts about kenv emulating sysctl

[Aleksandr Rybalko]

On Wed, 9 May 2012 09:05:47 -0700
Garrett Cooper <yanegomi at gmail.com> wrote:

> Hi Hackers,
>     I've been asked to write up a script to analyze tunables via kenv
> for archival purposes an to establish a baseline set of static
> variables. In order to make life easier (and be able to do all the
> grunt work in a shell one-liner instead of introducing a bug prone
> tunable parser) I have written up a patch which would make kenv
> function a bit more like sysctl, wrt the fact that sysctl -n
> suppresses suffixing a value with the variable name when executed
> like so:
> 
> # kenv LINES
> LINES="24"
> # kenv -n LINES
> 24
> 
>     I've also considered keeping the functional defaults and instead
> do the following...
> 
> # kenv -v LINES
> LINES="24"
> # kenv LINES
> 24
> 
>     Pro of the first form is that it matches sysctl, pro of the
> second form is that it doesn't break backwards 'compatibility'. I
> know kenv isn't a widely used utility (albeit, I have seen it used in
> a few spots outside of FreeBSD proper), but I was wondering if anyone
> could see any potential pitfalls or would have a large degree of
> heartburn over changing the default to match sysctl. Thanks!
> -Garrett_______________________________________________
> freebsd-hackers at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To
> unsubscribe, send any mail to
> "freebsd-hackers-unsubscribe at freebsd.org"

Hi Garret,

I use it for embedded, kenv is good transport shared by loader, kernel
and userland (since there is no RW storages).

IMO, kenv != sysctl, so we not need to match sysctl. But backwards
'compatibility' is good reason to select second way.

Thanks.

WWW
-- 
Aleksandr Rybalko <ray at ddteam.net>