All of lore.kernel.org
 help / color / mirror / Atom feed
From: Anshuman Khandual <khandual@linux.vnet.ibm.com>
To: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: linuxppc-dev@ozlabs.org,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: how to implement platform specific per process parameter?
Date: Mon, 06 Aug 2012 13:27:56 +0530	[thread overview]
Message-ID: <501F7904.4050206@linux.vnet.ibm.com> (raw)
In-Reply-To: <501F514C.2090909@ozlabs.ru>

> There are 4 options:

> 1. [not a kernel interface] use ptrace to execute the register changing
> command inside the specified pid. The next context switch saves the new
> value in the thread_struct. Dirty hack.
> 
> 2. Add a new syscall which would receive pid + register value and do the
> job. A bit too much.
> 
> 3. Add some hook in /proc filesystem but so far there were no platform
> specific bits, right?
> 
> 4. Implement a static node "/sys/devices/system/cpu/dscr_control".
> write() would parse the input stream, call scanf("%d %x", &pid, &dscr)
> and do the job.
> 


/sys/ interface would be appropriate I believe. But in this way we can take a new
(pid, dscr) and update thread_struct. But there should be a way to enlist all
(pid, dscr) values which are explicitly set by the user and different than that
of /sys/devices/system/cpu/dscr_default. So that we can know which process is holding
to what value of DSCR at any point of time.

 

> What is the correct approach? Thanks.
> 
> 


Regards
Anshuman

WARNING: multiple messages have this Message-ID (diff)
From: Anshuman Khandual <khandual@linux.vnet.ibm.com>
To: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	linuxppc-dev@ozlabs.org
Subject: Re: how to implement platform specific per process parameter?
Date: Mon, 06 Aug 2012 13:27:56 +0530	[thread overview]
Message-ID: <501F7904.4050206@linux.vnet.ibm.com> (raw)
In-Reply-To: <501F514C.2090909@ozlabs.ru>

> There are 4 options:

> 1. [not a kernel interface] use ptrace to execute the register changing
> command inside the specified pid. The next context switch saves the new
> value in the thread_struct. Dirty hack.
> 
> 2. Add a new syscall which would receive pid + register value and do the
> job. A bit too much.
> 
> 3. Add some hook in /proc filesystem but so far there were no platform
> specific bits, right?
> 
> 4. Implement a static node "/sys/devices/system/cpu/dscr_control".
> write() would parse the input stream, call scanf("%d %x", &pid, &dscr)
> and do the job.
> 


/sys/ interface would be appropriate I believe. But in this way we can take a new
(pid, dscr) and update thread_struct. But there should be a way to enlist all
(pid, dscr) values which are explicitly set by the user and different than that
of /sys/devices/system/cpu/dscr_default. So that we can know which process is holding
to what value of DSCR at any point of time.

 

> What is the correct approach? Thanks.
> 
> 


Regards
Anshuman


  reply	other threads:[~2012-08-06  7:58 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-08-06  5:08 how to implement platform specific per process parameter? Alexey Kardashevskiy
2012-08-06  5:08 ` Alexey Kardashevskiy
2012-08-06  7:57 ` Anshuman Khandual [this message]
2012-08-06  7:57   ` Anshuman Khandual

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=501F7904.4050206@linux.vnet.ibm.com \
    --to=khandual@linux.vnet.ibm.com \
    --cc=aik@ozlabs.ru \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linuxppc-dev@ozlabs.org \
    /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.