On Thu, 2012-01-12 at 09:34 +0100, John Johansen wrote: > On 01/11/2012 05:47 PM, Jamie Strandboge wrote:
> > So I started looking into some of my profiling requirements for this a > > bit more this week and while I think that there are use cases for > > matching, filtering and pinning, from a profiling perspective I think > > pinning and filtering (so long as you can filter in such a way as to > > clear the env var completely) are the most interesting. As an initial > > implementation, either would be quite useful. As an alternative, simply > > having the ability to clear/unset an environment variable would be quite > > useful. > > > So from an implementation pov > - matching is the easiest > - next by complete variable filtering/unsetting. > - Partial filtering of a variables values is a little harder, and needs a > good syntax that can distiguish between which parts of the variable > are being matched and what part of the value when matched should be > removed. > - setting a value would be next and will probably need a userspace helper > so it can manipulate the env memory > - and most difficult is pinning as it requires identifying and saving > off a value, and then setting it if it has changed. > > I think a first pass at this will probably get the first two. Cool, filtering (with the ability to clear) would be very useful and should allow me to not do weird workarounds like I am now. Would have to profile more with it, but it may even be good enough for the medium term. -- Jamie Strandboge | http://www.canonical.com
signature.asc
Description: This is a digitally signed message part
-- AppArmor mailing list [email protected] Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/apparmor
