Re: [PATCH 1/8] reconfiguration leaks: implicit ACLs

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.
Received on Fri Jun 13 2014 - 19:57:58 MDT