Remoting host policy for mac.

Remoting host policy for mac.

BUG=NONE
TEST=1) Start remoting session with Mac as host
	 2) Go to the preferences file for the browser (com.google.Chrome.plist)
	 3) Add a bool pref "RemoteAccessHostFirewallTraversal" set to NONE
	 4) Wait 15 minutes
	 5) You should not be able to do NAT traversal.

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=98546

[dmac, 2011-08-25 21:10:50]

PTAL (if you don't mind)... otherwise I'll find someone else.
Yes I kind of wimped out, but I didn't think the complexity was worth it. I'd
rather keep things simple and effective even if there is a slight delay before
the policy would be applied.

[awong, 2011-08-25 21:32:05]

http://codereview.chromium.org/7748027/diff/1001/remoting/host/plugin/policy_...
File remoting/host/plugin/policy_hack/nat_policy_mac.mm (right):

http://codereview.chromium.org/7748027/diff/1001/remoting/host/plugin/policy_...
remoting/host/plugin/policy_hack/nat_policy_mac.mm:59: for (size_t i = 0; i <
ARRAYSIZE_UNSAFE(policies); ++i) {
:-/

Would it be worth promoting this struct declaration to a more normal type
(perhaps inside an anonymous namespace) so we could use arraysize()?

http://codereview.chromium.org/7748027/diff/1001/remoting/host/plugin/policy_...
remoting/host/plugin/policy_hack/nat_policy_mac.mm:78: << 
static_cast<bool>(policies[i].is_allowed);
Should also explicitly list what state we actually went into.

[awong, 2011-08-25 23:50:58]

LGTM

Other than the 2 comments earlier, this looks good.  And since the 2 comments
earlier are basically nits, LGTM.  I'd appreciate fixing the second one at least
though.

On 2011/08/25 21:32:05, awong wrote:
>
http://codereview.chromium.org/7748027/diff/1001/remoting/host/plugin/policy_...
> File remoting/host/plugin/policy_hack/nat_policy_mac.mm (right):
> 
>
http://codereview.chromium.org/7748027/diff/1001/remoting/host/plugin/policy_...
> remoting/host/plugin/policy_hack/nat_policy_mac.mm:59: for (size_t i = 0; i <
> ARRAYSIZE_UNSAFE(policies); ++i) {
> :-/
> 
> Would it be worth promoting this struct declaration to a more normal type
> (perhaps inside an anonymous namespace) so we could use arraysize()?
> 
>
http://codereview.chromium.org/7748027/diff/1001/remoting/host/plugin/policy_...
> remoting/host/plugin/policy_hack/nat_policy_mac.mm:78: << 
> static_cast<bool>(policies[i].is_allowed);
> Should also explicitly list what state we actually went into.

[dmac, 2011-08-26 21:04:19]

PTAL again