On 04/25/2014 01:58 AM, Amos Jeffries wrote: > On 25/04/2014 12:46 p.m., Alex Rousskov wrote: >> Do not leak implicit ACLs during reconfigure. >> >> Many ACLs implicitly created by Squid when grouping multiple ACLs were >> not destroyed because those implicit ACLs where not covered by the >> global ACL registry used for ACL destruction. >> >> See also: r13210 which did not go far enough because it incorrectly >> assumed that all InnerNode() children are aclRegister()ed and, hence, >> will be centrally freed.
> -0. Is this a "negative" vote from "Squid3 voting" rules point of view? http://wiki.squid-cache.org/MergeProcedure#Squid3_Voting > I believe we should move to reference counting ACLs instead of > continuing to work around these edge cases. I agree that reference counting is an overall better design for ACLs, of course. However, since refcounting ACLs would be a large change that nobody has volunteered to implement in the foreseeable future (AFAIK), I suggest that this [significant] leak fix should go in now. Any other votes/opinions? Thank you, Alex.
