Messages in this thread | | | Date | Sat, 23 Apr 2011 00:15:22 +0200 | Subject | Re: perf_events: questions about cpu_has_ht_siblings() and offcore support | From | Stephane Eranian <> |
| |
On Fri, Apr 22, 2011 at 10:42 PM, Peter Zijlstra <peterz@infradead.org> wrote: > On Fri, 2011-04-22 at 15:46 +0200, Stephane Eranian wrote: >> >> perf stat -e offcore_response_0:dmd_data_rd,offcore_response_0:dmnd_rfo ...... >> >> This should still not be allowed. > > Aside from the problem you mention, I don't like the whole > offcore_response_[01] thing. The offcore things are fully symmetric so > the kernel should automatically select the other one if one is taken. > We could indeed do this. Though it would mean special casing event codes 0xbb and 0xb7. You'd have to lookup the per-core struct to decide whether it's worth it or not.
| |