Patch Set 1:

(1 comment)

https://gerrit.osmocom.org/#/c/4750/1/src/libfilter/bsc_msg_filter.c
File src/libfilter/bsc_msg_filter.c:

Line 208:       return 1;
> But that was "the global allow". So It might make sense to have an access-l
The above "allow" check does change the semantics significantly: In order to be 
able to deny anything, we need *some* deny rule. Without a deny rule, we always 
end up accepting everything; either we allow before, or we hit this 'return 1'. 
Now, *with* a deny rule in place, say a "deny .*", which acts as the default 
you propose, we would deny all IMSIs and never reach this 'return 1', unless we 
allow some IMSIs before that and exit early. So before this patch, you can 
either compose a deny regex that matches all IMSIs except the ones you want to 
allow (usually rather cumbersome), or you're stuck on allowing all or nothing. 
After this patch you can conveniently pick IMSI prefixes to allow, rather than 
having to negate a deny regex.


-- 
To view, visit https://gerrit.osmocom.org/4750
To unsubscribe, visit https://gerrit.osmocom.org/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: Idb6c8dd62aa90666ba6fcd213f59d79f5498da3f
Gerrit-PatchSet: 1
Gerrit-Project: osmo-bsc
Gerrit-Branch: master
Gerrit-Owner: Neels Hofmeyr <[email protected]>
Gerrit-Reviewer: Harald Welte <[email protected]>
Gerrit-Reviewer: Holger Freyther <[email protected]>
Gerrit-Reviewer: Jenkins Builder
Gerrit-Reviewer: Neels Hofmeyr <[email protected]>
Gerrit-HasComments: Yes

Reply via email to