Re: [RFC][PATCH 0/4] new timeofday-based soft-timer subsystem

[Nishanth Aravamudan <nacc@us.ibm.com>]

On 17.05.2005 [16:33:00 -0700], Nishanth Aravamudan wrote:
> On 13.05.2005 [17:16:35 -0700], john stultz wrote:
> > All,
> > 	This patch implements the architecture independent portion of the new
> > time of day subsystem. For a brief description on the rework, see here:
> > http://lwn.net/Articles/120850/ (Many thanks to the LWN team for that
> > easy to understand writeup!)
> > 
> > 	I intend this to be the last RFC release and to submit this patch to
> > Andrew for for testing near the end of this month. So please, if you
> > have any complaints, suggestions, or blocking issues, let me know.
> 
> I have been working closely with John to re-work the soft-timer subsytem
> to use the new timeofday() subsystem. The following patches attempts to
> begin this process. I would greatly appreciate any comments.

<snip>

> I will try to get some current benchmark differentials posted tomorrow.
> The previous patch I released showed little difference between mainline,
> John's timeofday rework and my soft-timer rework in kernbench.

<snip>

Hrm, "tomorrow" became "two days from now," but here they are, kernbench
comparisons (in percent relative to mainline) for x86 and x86_64, 10
iterations each.

The x86_64 machine is a 2-way 2.0 GHz with 3.5 GB of RAM.
The x86 machine is 16-way (32 with HT) 1.4 GHz Xeon with 15 GB of RAM.

				x86

					Elapsed	User	System	CPU

2.6.12-rc4:				100%	100%	100%	100%

2.6.12-rc4 + John's patch:		100.3%	100%	99.8%	99.6%

2.6.12-rc4 + John's patch + my patch:	102.1%	101%	100%	98%

				x86_64

					Elapsed	User	System	CPU

2.6.12-rc4:				100%	100%	100%	100%

2.6.12-rc4 + John's patch:		99.5%	99.5%	99.1%	99.9%

2.6.12-rc4 + John's patch + my patch:	99.7%	99.7%	99.5%	100%

----

All in all, pretty consistent across the board.

Thanks,
Nish