Re: Improving the kernel/i386 timecounter performance (GSoC proposal)

From: Tim Kientzle <kientzle_at_freebsd.org>
Date: Mon, 30 Mar 2009 12:29:03 -0700
Poul-Henning Kamp wrote:
> In message <20090329180745.GB38985_at_server.vk2pj.dyndns.org>, Peter Jeremy write
> s:
> 
>>> I'm assuming folks are still in love with the TSC because it still the
>>> cheapest as oppose ACPI-fast or HPET to even contemplate this?
>> That is its major advantage.  It might be feasible to export all the
>> data necessary to implement the complete CLOCK_*_FAST family.
> 
> The general attraction is that it can be read from userland by unpriviledged
> programs.
> 
> On systems where the ACPI or HPET hardware can be memory-mapped, I should
> be equally possible to map those read-only into userland processes.
> 
> Now _THAT_ would be interesting.

Which brings us back to having a page of code
provided by the kernel so that the kernel can
determine the appropriate implementation
(depending on the hardware availability) and so
that userland can invoke the functions without
going through a task switch.  Libc can then
either invoke these directly or, if the page is
unavailable for any reason, use the system calls.

Tim
Received on Mon Mar 30 2009 - 17:29:08 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:45 UTC