Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1661624

Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake

Show all headers | View raw

Hi,

On Thu, Jun 8, 2017 at 1:15 AM, Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Wed, Jun 07, 2017 at 04:22:24PM -0700, Andi Kleen wrote:
>
> > diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h
> > index b1c0b187acfe..95daade294d7 100644
> > --- a/include/uapi/linux/perf_event.h
> > +++ b/include/uapi/linux/perf_event.h
> > @@ -931,14 +931,18 @@ union perf_mem_data_src {
> >                       mem_snoop:5,    /* snoop mode */
> >                       mem_lock:2,     /* lock instr */
> >                       mem_dtlb:7,     /* tlb access */
> > -                     mem_rsvd:31;
> > +                     mem_lvlx:8,     /* memory hierarchy level, ext */
> > +                     mem_snoopx:2,   /* snoop mode, ext */
> > +                     mem_rsvd:21;
> >       };
> >  };
> >  #elif defined(__BIG_ENDIAN_BITFIELD)
> >  union perf_mem_data_src {
> >       __u64 val;
> >       struct {
> > -             __u64   mem_rsvd:31,
> > +             __u64   mem_rsvd:21,
> > +                     mem_snoopx:2,   /* snoop mode, ext */
> > +                     mem_lvlx:8,     /* memory hierarchy level, ext */
> >                       mem_dtlb:7,     /* tlb access */
> >                       mem_lock:2,     /* lock instr */
> >                       mem_snoop:5,    /* snoop mode */
>
> So one thing we could do is add a mem_hops field and always set that,
> even for the old stuff. The old stuff will not know about that field and
> ignore the bits, but new stuff will then not need as many LVL bits.
>
That would be better than lvlx I think. I am guessing you're suggesting
an integer count here and not a bitmask. Right? Then I wonder why it
would need 8 bits or 255 possible levels!

> Of course, we then get into the problem of how many bits of hops we
> need.. Power guys ?
>
> > @@ -975,6 +979,16 @@ union perf_mem_data_src {
> >  #define PERF_MEM_LVL_UNC     0x2000 /* Uncached memory */
> >  #define PERF_MEM_LVL_SHIFT   5
> >
> > +#define PERF_MEM_LVLX_L4     0x01 /* L4 */
> > +#define PERF_MEM_LVLX_REM_L4    0x02 /* Remote L4 */
> > +#define PERF_MEM_LVLX_REM_RAM        0x04 /* Remote Ram, unknown hops */
> > +#define PERF_MEM_LVLX_PMEM   0x08 /* Persistent Memory */
> > +#define PERF_MEM_LVLX_REM_PMEM       0x10 /* Remote Persistent Memory */
> > +#define PERF_MEM_LVLX_REM_NA 0x20 /* Remote N/A level */
>
> Still wondering what the point of REM_NA is.. can you explain?
>
> > +/* 2 free */
> > +
> > +#define PERF_MEM_LVLX_SHIFT  33
> > +
> >  /* snoop mode */
> >  #define PERF_MEM_SNOOP_NA    0x01 /* not available */
> >  #define PERF_MEM_SNOOP_NONE  0x02 /* no snoop */
> > @@ -983,6 +997,10 @@ union perf_mem_data_src {
> >  #define PERF_MEM_SNOOP_HITM  0x10 /* snoop hit modified */
> >  #define PERF_MEM_SNOOP_SHIFT 19
> >
> > +#define PERF_MEM_SNOOPX_FWD  0x01 /* forward */
> > +/* 1 free */
> > +#define PERF_MEM_SNOOPX_SHIFT        41
> > +
> >  /* locked instruction */
> >  #define PERF_MEM_LOCK_NA     0x01 /* not available */
> >  #define PERF_MEM_LOCK_LOCKED 0x02 /* locked transaction */
> > --
> > 2.9.4
> >

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

Fix Skylake PEBS data source for perf Andi Kleen <andi@firstfloor.org> - 2017-06-08 01:30 +0200
  [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Andi Kleen <andi@firstfloor.org> - 2017-06-08 01:30 +0200
    Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Peter Zijlstra <peterz@infradead.org> - 2017-06-08 10:20 +0200
      Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Stephane Eranian <eranian@google.com> - 2017-06-08 21:50 +0200
        Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Peter Zijlstra <peterz@infradead.org> - 2017-06-08 22:10 +0200
          Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Andi Kleen <ak@linux.intel.com> - 2017-06-09 18:50 +0200
        Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Andi Kleen <andi@firstfloor.org> - 2017-06-08 22:20 +0200
          Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Peter Zijlstra <peterz@infradead.org> - 2017-06-08 22:30 +0200
      Re: [PATCH v2 2/4] perf/x86: Fix data source decoding for Skylake Madhavan Srinivasan <maddy@linux.vnet.ibm.com> - 2017-06-09 10:10 +0200

csiph-web