Re: squid-2.4 release ?

Adrian Chadd wrote:

> I haven't thought of a cute way of solving this with those pointers.
> I've got cute ways of solving it WITHOUT those pointers, but that won't
> see the light of day in squid-2.4.

The simple way is to reserve space for the removal node in the store
entry, and then use the OFFSETOF macro from cbdata to go back from the
node to the StoreEntry.. And similary for memory entries..

The only criteria is that a fixed amount of space is set aside for the
removal policy, and that none of the policies attempts to use more than
that (a simple sizeof assert in the policy initialization should make
sure of that)

/Henrik
Received on Mon Jan 08 2001 - 17:41:41 MST