public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Robert Richter <robert.richter@amd.com>
To: "graalfs@linux.vnet.ibm.com" <graalfs@linux.vnet.ibm.com>
Cc: "mingo@elte.hu" <mingo@elte.hu>,
	"oprofile-list@lists.sf.net" <oprofile-list@lists.sf.net>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"linux-s390@vger.kernel.org" <linux-s390@vger.kernel.org>,
	"borntraeger@de.ibm.com" <borntraeger@de.ibm.com>,
	"schwidefsky@de.ibm.com" <schwidefsky@de.ibm.com>,
	"heiko.carstens@de.ibm.com" <heiko.carstens@de.ibm.com>,
	Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>,
	Maran Pakkirisamy <maranp@linux.vnet.ibm.com>
Subject: Re: [patch 4/4] Handle memory unmap while hardware sampling is running
Date: Mon, 3 Jan 2011 21:39:09 +0100	[thread overview]
Message-ID: <20110103203909.GT4739@erda.amd.com> (raw)
In-Reply-To: <20101220130629.974963910@linux.vnet.ibm.com>

On 20.12.10 08:05:45, graalfs@linux.vnet.ibm.com wrote:
> From: graalfs@linux.vnet.ibm.com
> 
> During memory unmap hardware sampling is deactivated.
> After all samples have been collected hardware sampling is reactivated again.

Do you mean you switch off the profiler for each buffer sync? This
happens per default every 100ms. Why is this needed? Which memory is
unmapped?

This hooks in the sync_buffer() are ugly. We must try to find a better
solution here.

There is oprofile_put_buff() that allows to write directly to the the
event buffer which is used in powerpc/cell.

-Robert

>  
> Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
> Signed-off-by: Maran Pakkirisamy <maranp@linux.vnet.ibm.com>
> Signed-off-by: Heinz Graalfs <graalfs@linux.vnet.ibm.com>
> ---
>  arch/s390/oprofile/hwsampler_files.c |    2 +-
>  drivers/oprofile/buffer_sync.c       |   13 +++++++++++++
>  2 files changed, 14 insertions(+), 1 deletion(-)
> 
> Index: linux-2.6/drivers/oprofile/buffer_sync.c
> ===================================================================
> --- linux-2.6.orig/drivers/oprofile/buffer_sync.c
> +++ linux-2.6/drivers/oprofile/buffer_sync.c
> @@ -32,6 +32,11 @@
>  #include <linux/sched.h>
>  #include <linux/gfp.h>
>  
> +#ifdef CONFIG_OPROFILE_HWSAMPLING_MODE
> +#include <asm/hwsampler.h>
> +extern unsigned long oprofile_hwsampler;
> +#endif
> +
>  #include "oprofile_stats.h"
>  #include "event_buffer.h"
>  #include "cpu_buffer.h"
> @@ -513,6 +518,10 @@ void sync_buffer(int cpu)
>  
>  	mutex_lock(&buffer_mutex);
>  
> +#ifdef CONFIG_OPROFILE_HWSAMPLING_MODE
> +	if (oprofile_hwsampler)
> +		hwsampler_deactivate(cpu);
> +#endif
>  	add_cpu_switch(cpu);
>  
>  	op_cpu_buffer_reset(cpu);
> @@ -569,6 +578,10 @@ void sync_buffer(int cpu)
>  
>  	mark_done(cpu);
>  
> +#ifdef CONFIG_OPROFILE_HWSAMPLING_MODE
> +	if (oprofile_hwsampler)
> +		hwsampler_activate(cpu);
> +#endif
>  	mutex_unlock(&buffer_mutex);
>  }
>  
> Index: linux-2.6/arch/s390/oprofile/hwsampler_files.c
> ===================================================================
> --- linux-2.6.orig/arch/s390/oprofile/hwsampler_files.c
> +++ linux-2.6/arch/s390/oprofile/hwsampler_files.c
> @@ -22,7 +22,7 @@ unsigned long oprofile_max_interval;
>  static unsigned long oprofile_sdbt_blocks = DEFAULT_SDBT_BLOCKS;
>  static unsigned long oprofile_sdb_blocks = DEFAULT_SDB_BLOCKS;
>  
> -static unsigned long oprofile_hwsampler;
> +unsigned long oprofile_hwsampler;
>  
>  static int oprofile_hwsampler_start(void)
>  {
> 
> 

-- 
Advanced Micro Devices, Inc.
Operating System Research Center


      reply	other threads:[~2011-01-03 20:39 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-12-20 13:05 [patch 0/4] OProfile support for System z's hardware sampling graalfs
2010-12-20 13:05 ` [patch 1/4] This patch adds support for hardware based sampling on System z processors (models z10 and up) graalfs
2011-01-03 17:06   ` Robert Richter
2011-01-19 16:54     ` Heinz Graalfs
2010-12-20 13:05 ` [patch 2/4] This patch enhances OProfile to support System zs hardware sampling feature graalfs
2011-01-03 19:02   ` Robert Richter
2011-01-19 16:55     ` Heinz Graalfs
2010-12-20 13:05 ` [patch 3/4] This patch introduces a new oprofile sample add function (oprofile_add_ext_hw_sample) graalfs
2011-01-04 15:34   ` Robert Richter
2011-01-07 16:31     ` Heinz Graalfs
2011-01-19 16:56     ` Heinz Graalfs
2010-12-20 13:05 ` [patch 4/4] Handle memory unmap while hardware sampling is running graalfs
2011-01-03 20:39   ` Robert Richter [this message]

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=20110103203909.GT4739@erda.amd.com \
    --to=robert.richter@amd.com \
    --cc=borntraeger@de.ibm.com \
    --cc=graalfs@linux.vnet.ibm.com \
    --cc=heiko.carstens@de.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=mahesh@linux.vnet.ibm.com \
    --cc=maranp@linux.vnet.ibm.com \
    --cc=mingo@elte.hu \
    --cc=oprofile-list@lists.sf.net \
    --cc=schwidefsky@de.ibm.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox