lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Sun, 20 Jul 2008 10:58:14 -0400
From:	jamal <hadi@...erus.ca>
To:	Roland Dreier <rdreier@...co.com>
Cc:	Patrick McHardy <kaber@...sh.net>,
	David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
	johannes@...solutions.net, linux-wireless@...r.kernel.org
Subject: Re: [PATCH 20/31]: pkt_sched: Perform bulk of qdisc destruction in
	RCU.

On Fri, 2008-18-07 at 10:10 -0700, Roland Dreier wrote:

> This is definitely true, but it is good to keep in mind that in the near
> future we will start to see things look a little like multiple "virtual
> wires."  This is because of new ethernet standards like per-priority
> pause, which makes it possible that one hardware ring on a NIC can
> transmit while another ring is paused (possibly because of congestion
> far off in the network).

Thats essentially what i am arguing for.
[I think some, not all, of the wireless qos schemes also have similar
scheduling].

My understanding of these wired "datacentre/virtualization" schemes you
describe is they are strict prio based. When the low prio "virtual wire"
is contending for the "physical wire" with a higher prio "virtual wire",
the high prio always wins. 
We just need to make sure this behavior is also maintained whatever
buffering scheme is used within or above the driver(qdisc level).

cheers,
jamal

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists