NetBSD-Bugs archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: kern/58508: experimental wg(4) queues LIFO, not FIFO, pending first handshake



The following reply was made to PR kern/58508; it has been noted by GNATS.

From: Taylor R Campbell <riastradh%NetBSD.org@localhost>
To: gnats-bugs%NetBSD.org@localhost, netbsd-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: kern/58508: experimental wg(4) queues LIFO, not FIFO, pending first handshake
Date: Mon, 29 Jul 2024 14:47:59 +0000

 On reflection, I realize that LIFO and FIFO are not the right terms
 here.
 
 A single-entry queue is always both LIFO (last-in/first-out) and
 FIFO (first-in/first-out).
 
 What changed is the choice of which packets to drop when we are asked
 to transmit new ones and the (single-entry) queue has filled up.  Do
 we drop the oldest packet, or the newest packet (or something else)?
 
 Previously I had chosen drop-newest (but not for any particular reason
 I can remember, and certainly not for any reason I wrote down).  Now
 the choice is drop-oldest, like arp/ndp does.
 
 Too late to fix the commit message, oh well.
 


Home | Main Index | Thread Index | Old Index