Creating nightly hwloc snapshot SVN tarball was a success.
Snapshot: hwloc 1.1a1r2145
Start time: Tue May 25 21:01:02 EDT 2010
End time: Tue May 25 21:02:58 EDT 2010
Your friendly daemon,
Cyrador
On May 25, 2010, at 3:00 PM, Brice Goglin wrote:
> The attached patch implements a "linesize" cache attribute. You now get
> things like
> L3Cache #0 (8192KB line=64)
>L2Cache #0 (256KB line=64)
> L1Cache #0 (32KB line=64)
Cool!
>> Why is runtime icache information
Wheeler, Kyle Bruce, le Tue 25 May 2010 13:46:09 -0600, a écrit :
> I agree that, inherently, cache line size has nothing to do with topology.
> But on the other hand, it's particularly useful for parallel shared-memory
> applications (to avoid false-sharing),
The false-sharing part is where it
On May 25, 2010, at 1:00 PM, Brice Goglin wrote:
> Numerous ideas like this were proposed and we're not sure where to stop.
> If we start doing this, people will ask for the processor frequency, the
> number of floating point units per core, the associativity of the cache,
> the type of memory,
Hello!
I noticed that hwloc doesn't appear to have a way of reporting (minimum) cache
line size... which is obviously useful and important for avoiding false-sharing
issues. I've been hacking together a way to do it in what passes for a
cross-platform method. My code is currently in qthreads
On May 25, 2010, at 9:27 AM, Brice Goglin wrote:
> On 25/05/2010 15:23, Jeff Squyres wrote:
> > - 1.1 will be a "new feature" release. What features / branches do you
> > guys want to bring into the trunk and then create a 1.1 branch with?
>
> I think we have trac tickets for everything we
On 25/05/2010 15:23, Jeff Squyres wrote:
> - 1.1 will be a "new feature" release. What features / branches do you guys
> want to bring into the trunk and then create a 1.1 branch with?
I think we have trac tickets for everything we wanted in 1.1.
Brice
I made a 1.0.1 milestone; new tickets will be in that milestone by default.
I think we're all assuming the same thing, but let's put it in writing to make
sure:
- 1.0.1 will be a bug fix release, if necessary (I think OMPI, at least, would
like some kind of resolution to #35 -- I'm working on