On Wed, Oct 21, 2009 at 9:46 AM, Kenneth Hoste <kenneth.ho...@ugent.be> wrote: > Hi Stephane, > > Thanks for your quick reply. Some clarification and further questions > below... > > On Oct 21, 2009, at 8:41 AM, stephane eranian wrote: > >> Ken, >> >> On Tue, Oct 20, 2009 at 4:06 PM, Kenneth Hoste <kenneth.ho...@ugent.be> >> wrote: >>> >>> The thing we are unable to explain is that the micro-ops per instruction >>> rate rises significantly when comparing Core i7 (Nehalem architecture) >>> to Core 2 (Core architecture). And that while micro-op fusion is reported >>> to be improved in the more recent Core i7 processors. >>> >> For Nehalem, things are a bit more complicated. Here is >> what the documentation says: >> >> C2H 01H UOPS_RETIRED.ANY >> Counts the number of micro-ops >> retired, (macro-fused=1, micro- >> fused=2, others=1; maximum count >> of 8 per cycle). Most instructions >> are composed of one or two micro- >> ops. Some instructions are decoded >> into longer sequences such as >> repeat instructions, floating point >> transcendental instructions, and >> assists. >> >> You need to subtract the number of uops micro-fused. I think >> there is another event for this. > > It's unclear to me why we would need to substract the number of uops > micro-fused... > Could you elaborate on this?
My mistake, I think the count is correct in that it give you the number of uops that would have retired without fusion. 2 fused micro-ops = increment of 2. As for Core 2, there are some errata but that's for instructions retired. Did you try breaking down ops_retired:any, into its various components to see how they add up? > >> >> Although, would you have a micro-benchmark that demonstrate >> this behavior? That would help figure out what is going on. >> >> I am not aware of an erratum for this event on Nehalem. > > The numbers seems to suggest that the Nehalem numbers might be what we > expect, but the Core 2 numbers probably are not (i.e. they are too low). > > Are you aware of a problem with counting uops on Core 2? > > greetings, > > Kenneth > ------------------------------------------------------------------------------ Come build with us! The BlackBerry(R) Developer Conference in SF, CA is the only developer event you need to attend this year. Jumpstart your developing skills, take BlackBerry mobile applications to market and stay ahead of the curve. Join us from November 9 - 12, 2009. Register now! http://p.sf.net/sfu/devconference _______________________________________________ perfmon2-devel mailing list perfmon2-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
