[Cbe-oss-dev] Fwd: adding virtual SPU timers

[Dean Burdick]

Yeah, you're right.  I had the same thought.  This option is no good.

Dean Burdick
Cell Software Development
IBM Austin, TX
512-838-0264 T/L: 678-0264




Arnd Bergmann <arnd at arndb.de> 
10/22/2007 10:19 AM

To
cbe-oss-dev at ozlabs.org
cc
Dean Burdick/Austin/IBM at IBMUS, Uzi Shvadron <SHVADRON at il.ibm.com>
Subject
Re: [Cbe-oss-dev] Fwd: adding virtual SPU timers






On Monday 22 October 2007, Dean Burdick wrote:
> I like the idea of maintaining both the wall time and a delta.  This 
would 
> allow us to provide both real time (for trace timestamps) and virtual 
time 
> (for measurements/timers) at the same time with the library.

Any ideas on how the SPU code should get at the offset? If we require
a syscall for every timer operation, that will significantly limit the
accuracy and add latency to the library functions.

Also, I'm still not sure what times should be included in the offset
between virtual time by the kernel (see the list from my original
mail).

                 Arnd <><

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ozlabs.org/pipermail/cbe-oss-dev/attachments/20071024/4368f2bf/attachment.htm>