Only one thing... but I guess this is mostly a question to Adrian:

+ *
+ * NOTE: COMM_ERR_CLOSING will NOT be called for CommReads' sitting in
a 
+ * DeferredReadManager.
  */


makes me further question the need and value of COMM_ERR_CLOSING...


What is the intention behind COMM_ERR_CLOSING? What is it supposed to
fix which cannot be fixed by not calling the callback and instead make
use of the close callback?

Regards
Henrik



m�n 2003-03-03 klockan 11.59 skrev Robert Collins:
> Ok, I'd like to apply this patch:
> http://www.squid-cache.org/~robertc/deferred.patch.bz2
> asap (Adrian has a couple of comms things to work on that depend on this
> to prevent collisions).
> 
> It
> * removes deferred reads across the board
> * converts nearly all unready reads to the new DeferredRead framework.
> * leaves ssl reading 1 byte per IO cycle at minimum.
> * Includes David Nicklay's epoll patch.
> 
> This should allow kqueue and epoll to properly support delay pools.
> 
> Cheers,
> Rob
-- 
Henrik Nordstrom <[EMAIL PROTECTED]>
MARA Systems AB, Sweden

Reply via email to