lkml.org 
[lkml]   [2009]   [Jun]   [23]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    SubjectRe: I.2 - Grouping
    From
    On Tue, Jun 23, 2009 at 7:16 AM, Paul Mackerras<paulus@samba.org> wrote:
    > Ingo Molnar writes:
    >
    >> > 2/ Grouping
    >> >
    >> > By design, an event can only be part of one group at a time.
    >
    > To clarify this statement of Stephane's, a _counter_ can only be in
    > one group.  You can have multiple counters counting the same _event_
    > and those counters can (obviously) be in different groups.
    >
    Okay.

    What happens if I do:
    fd0 = perf_counter_open(&hwc1, getpid(), -1, -1, 0);
    fd1 = perf_counter_open(&hwc2, getpid(), -1, fd0, 0);

    And then:
    fd2 = perf_counter_open(&hwc2, getpid(), -1, fd1, 0);

    >> > Events in a group are guaranteed to be active on the PMU at the
    >> > same time. That means a group cannot have more events than there
    >> > are available counters on the PMU. Tools may want to know the
    >> > number of counters available in order to group their events
    >> > accordingly, such that reliable ratios could be computed. It seems
    >> > the only way to know this is by trial and error. This is not
    >> > practical.
    >>
    >> Groups are there to support heavily constrained PMUs, and for them
    >> this is the only way, as there is no simple linear expression for
    >> how many counters one can load on the PMU.
    >
    > That's not the only reason for having groups, or even the main reason
    > IMO.  The main reason for having groups is to provide a way to ask the
    > scheduler to ensure that two or more counters are always scheduled
    > together, so that you can meaningfully do arithmetic operations on the
    > counter values that would be sensitive to the statistical noise
    > introduced by the scheduling, such as ratios and differences.
    >
    > In other words, grouping is there because we don't guarantee to have
    > all counters scheduled onto the PMU whenever possible.  Heavily
    > constrained PMUs increase the need for scheduling, but even if
    > counters are completely orthogonal there are only a fixed number of
    > them so we still need to schedule counters at some point.
    >
    I agree completely.
    --
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

    \
     
     \ /
      Last update: 2009-06-23 09:39    [W:3.930 / U:0.268 seconds]
    ©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site