All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: Lin Ming <ming.m.lin@intel.com>
Cc: Ingo Molnar <mingo@elte.hu>, Andi Kleen <andi@firstfloor.org>,
	Stephane Eranian <eranian@google.com>,
	Arnaldo Carvalho de Melo <acme@ghostprotocols.net>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	Robert Richter <robert.richter@amd.com>
Subject: Re: [PATCH 1/4] perf: Add memory load/store events generic code
Date: Wed, 06 Jul 2011 15:51:42 +0200	[thread overview]
Message-ID: <1309960302.3282.283.camel@twins> (raw)
In-Reply-To: <1309931621.18875.130.camel@minggr.sh.intel.com>

On Wed, 2011-07-06 at 13:53 +0800, Lin Ming wrote:

> Do you mean to use the "impossible combinations" to express the inverse?

Nah, impossible would be things like having neither LOAD nor STORE set.

> MEM_STORE_DCU_MISS as: store-l2-l3-dram
> MEM_STORE_STLB_MISS as: store-itlb-dtlb
> 
> How about below code?

Right, something like that. Robert can the IBS data source data be
mapped onto this as well?

> #define PERF_MEM_LOAD                   (1ULL << 0)
> #define PERF_MEM_STORE                  (1ULL << 1)
> #define PERF_MEM_ATOMIC                 (1ULL << 2)
> #define PERF_MEM_L1                     (1ULL << 3)
> #define PERF_MEM_L2                     (1ULL << 4)
> #define PERF_MEM_L3                     (1ULL << 5)
> #define PERF_MEM_RAM                    (1ULL << 6)
> #define PERF_MEM_UNKNOWN                (1ULL << 7)
> #define PERF_MEM_IO                     (1ULL << 8)
> #define PERF_MEM_UNCACHED               (1ULL << 9)
> #define PERF_MEM_RESERVED               (1ULL << 10)
> #define PERF_MEM_LOCAL                  (1ULL << 11)
> #define PERF_MEM_REMOTE                 (1ULL << 12)
> #define PERF_MEM_SNOOP                  (1ULL << 13)
> #define PERF_MEM_MODIFIED               (1ULL << 14)
> #define PERF_MEM_EXCLUSIVE              (1ULL << 15)
> #define PERF_MEM_SHARED                 (1ULL << 16)
> #define PERF_MEM_INVALID                (1ULL << 17)

> #define PERF_MEM_ITLB                   (1ULL << 18)
> #define PERF_MEM_DTLB                   (1ULL << 19)
> #define PERF_MEM_STLB                   (1ULL << 20)

Are these TLB hit or miss?

> #define PERF_MEM_STORE_L1D_HIT  \
>         (PERF_MEM_STORE | PERF_MEM_L1)
> 
> #define PERF_MEM_STORE_L1D_MISS \
>         (PERF_MEM_STORE | PERF_MEM_L2 | PERF_MEM_L3 | PERF_MEM_RAM)
> 
> #define PERF_MEM_STORE_STLB_HIT \
>         (PERF_MEM_STORE | PERF_MEM_STLB)
>       
> #define PERF_MEM_STORE_STLB_MISS \
>         (PERF_MEM_STORE | PERF_MEM_ITLB | PERF_MEM_DTLB)

Going by the definition in table 30-22 neither of these seem correct, a
STLB_HIT was defined as DTLB|STLB whereas a STLB_MISS was missing both
(resulting in a full page-table walk I presume).

> #define PERF_MEM_STORE_ATOMIC \
>         (PERF_MEM_STORE | PERF_MEM_ATOMIC)
> 
> #define PERF_MEM_LOAD_STLB_HIT  \
>         (PERF_MEM_LOAD | PERF_MEM_STLB)
>    
> #define PERF_MEM_LOAD_STLB_MISS \
>         (PERF_MEM_LOAD | PERF_MEM_ITLB | PERF_MEM_DTLB)

idem

> #define PERF_MEM_LOAD_ATOMIC \
>         (PERF_MEM_LOAD | PERF_MEM_ATOMIC)
> 


  reply	other threads:[~2011-07-06 13:52 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-04  8:02 [PATCH 0/4] perf: memory load/store events generalization Lin Ming
2011-07-04  8:02 ` [PATCH 1/4] perf: Add memory load/store events generic code Lin Ming
2011-07-04  8:33   ` Peter Zijlstra
2011-07-04  8:44     ` Peter Zijlstra
2011-07-05 12:03       ` Peter Zijlstra
2011-07-05 23:02         ` Paul Mackerras
2011-07-06 13:58           ` Peter Zijlstra
2011-07-08  7:18             ` Anton Blanchard
2011-07-08 15:18               ` Peter Zijlstra
2011-08-08 11:57                 ` Peter Zijlstra
2011-08-08 11:59                 ` Peter Zijlstra
2011-07-04 22:01     ` Andi Kleen
2011-07-05  8:43       ` Peter Zijlstra
2011-07-04 11:08   ` Peter Zijlstra
2011-07-04 11:16   ` Peter Zijlstra
2011-07-04 21:52     ` Andi Kleen
2011-07-05 11:54     ` Lin Ming
2011-07-05 14:17       ` Peter Zijlstra
2011-07-06  5:53         ` Lin Ming
2011-07-06 13:51           ` Peter Zijlstra [this message]
2011-07-07  2:01             ` Lin Ming
2011-07-04  8:02 ` [PATCH 2/4] perf, x86: Add Intel Nhm/Wsm/Snb load latency support Lin Ming
2011-07-05 13:17   ` Peter Zijlstra
2011-07-05 13:34     ` Lin Ming
2011-07-22 18:58   ` Stephane Eranian
2011-07-04  8:02 ` [PATCH 3/4] perf, x86: Add Intel SandyBridge pricise store support Lin Ming
2011-07-11  8:32   ` Peter Zijlstra
2011-07-11  8:57     ` Lin Ming
2011-07-11  8:52       ` Peter Zijlstra
2011-07-04  8:02 ` [PATCH 4/4] perf, tool: Add new command "perf mem" Lin Ming
2011-07-04 22:00   ` Andi Kleen
2011-07-05  1:35     ` Lin Ming
2011-07-22 18:55 ` [PATCH 0/4] perf: memory load/store events generalization Stephane Eranian
2011-07-22 21:01   ` Andi Kleen
2011-07-22 21:14     ` Stephane Eranian
2011-07-22 21:43       ` Andi Kleen
2011-07-22 21:59         ` Stephane Eranian

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1309960302.3282.283.camel@twins \
    --to=a.p.zijlstra@chello.nl \
    --cc=acme@ghostprotocols.net \
    --cc=andi@firstfloor.org \
    --cc=eranian@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=ming.m.lin@intel.com \
    --cc=mingo@elte.hu \
    --cc=robert.richter@amd.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.