netdev - Re: [PATCH 2/5] e1000e: Add interrupt moderation run-time ethtool interface

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 linux-hardening linux-cve-announce 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:	Mon, 14 Apr 2008 12:31:45 -0700
From:	Rick Jones <rick.jones2@...com>
To:	Andi Kleen <andi@...stfloor.org>
CC:	Auke Kok <auke-jan.h.kok@...el.com>, jeff@...zik.org,
	netdev@...r.kernel.org, e1000-devel@...ts.sourceforge.net
Subject: Re: [PATCH 2/5] e1000e: Add interrupt moderation run-time ethtool
 interface

Andi Kleen wrote:
> Auke Kok <auke-jan.h.kok@...el.com> writes:
> 
> 
>>The ethtool -c / -C interface can now be used to modify the
>>irq moderation algorithm. This change does not require an
>>adapter reset and can thus be used at all times. The adapter
>>only supports changing/reading rx-usecs which has special
>>values for 0, 1 and 3:
>>
>>0 - no irq moderation whatsoever
>>1 - normal moderation favoring regular mixed traffic (default)
>>3 - best attempt at low latency possible at cost of CPU
> 
> 
> It would be nice if ethtool instead of having this as a obscure magic
> number had a higher level "--low-latency" or similar parameter
> (that could also potentially set similar settings on other NICs)
> 
> You think there would be a clean way to implement this? 

The biggest problems might be ones of interpretation as to whether a 
given setting was indeed:

--favor-latency
--favor-throutput
--favor-mixed

But it would indeed at one level be more "clear" to a random admin (IMO).

I use "--favor-mumble" above since what might be the --low-latency 
through one NIC might not be as low as the "--low-latency" though 
another NIC, and it also sort of implicitly suggests a trade-off is 
being made.

rick jones
--
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