Travis Downs
Travis Downs
Isn't the solution right in the quoted text in the `README.Debian`? Specifically this part: ``` Note, since fzf 0.29.0-1, the bash completion is installed for bash by default. Feel free...
Awesome, reading it now. What's the approximate cost of the `PFCSTART/PFCEND` calls? Do the make a kernel transition, or does the LKM enable user-space setting & reading of the PMC...
The 240 cycles is for reading all 8 counters, right? Is there an option to only read a subset?
Indeed, I don't suggest a general solution because I think "it's complicated". Personally I wouldn't embark on the path of trying to virtualize counters: this already happens with `perf_events` so...
> I suppose what's left is that if a counter gets programmed with an L1I_MISS event code, then there should be a conditional pfcWRMSR() invocation within the kernel module. Yes,...
Indeed, but creating the more specific versions shouldn't cause any additional variance, since you'd also generate the specific version of `pfcRemoveBias` in the same way. This all happening at compile-time,...
Yes, I think only the last of those can be done with the MSR. What I was alluding to though was to use the MSR to do the reads of...
Following up on "similar ways to do this" I just ran across Andi Kleen's [`jevents`](https://github.com/andikleen/pmu-tools/tree/master/jevents). It uses the perf infrastructure to do self-monitoring, and I think it even allows user...
To be clear, the `perf_events` userpage implementation mostly just involves a userland `rdpmc`. So that we are on the same page, I'm talking about the code in the _cap_user_rdpmc_ section...
> On my part I tried to see whether or not it is possible to enable RDPMC reads other than by the perf_events API or by a superuser-priv'ed echo 2...