WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-devel

RE: [Xen-devel] [patch] new version of find_domain_by_id() without refer

To: "Keir Fraser" <keir@xxxxxxxxxxxxx>, <xen-devel@xxxxxxxxxxxxxxxxxxx>
Subject: RE: [Xen-devel] [patch] new version of find_domain_by_id() without reference count [0/6]
From: "Santos, Jose Renato G" <joserenato.santos@xxxxxx>
Date: Sat, 9 Dec 2006 15:57:50 -0600
Cc: "Turner, Yoshio" <yoshio_turner@xxxxxx>, G John Janakiraman <john@xxxxxxxxxxxxxxxxxxx>
Delivery-date: Sat, 09 Dec 2006 13:57:55 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
In-reply-to: <C1A0326E.5962%keir@xxxxxxxxxxxxx>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
Thread-index: AccbR+BB71+g4tAvS++Fs5r59z+B5gALJ+tsABlzeOA=
Thread-topic: [Xen-devel] [patch] new version of find_domain_by_id() without reference count [0/6]
 

> -----Original Message-----
> From: Keir Fraser [mailto:keir@xxxxxxxxxxxxx] 
> Sent: Saturday, December 09, 2006 1:29 AM
> To: Santos, Jose Renato G; xen-devel@xxxxxxxxxxxxxxxxxxx
> Cc: Turner, Yoshio; G John Janakiraman
> Subject: Re: [Xen-devel] [patch] new version of 
> find_domain_by_id() without reference count [0/6]
> 
> On 9/12/06 4:09 am, "Santos, Jose Renato G" <joserenato.santos@xxxxxx>
> wrote:
> 
> > I only did minor tests on x86-32. Xen and dom0 boots fine and I can 
> > create and destroy domains. But, no more exaustive tests 
> were done. I 
> > carefully checked if I removed all put_domain() associated 
> with each 
> > modified invocation of find_domain_by_id but mistakes are always 
> > possible. It would be good to put this to some more 
> exhaustive tests 
> > before pushing it to the main tree. Waiting for post 3.0.4 
> release is 
> > strongly suggested.
> 
> It can't go in until RCU is merged and that won't happen for 
> 3.0.4 anyway.
> 
> When you post these again after 3.0.4 (which I'm sure you'll 
> have to do) I'd actually like the names as follows:
>  1. Rename find_domain_by_id() to get_domain_by_id().
>  2. Call the new function find_domain_by_id().
> 
> This has the obvious disadvantage of an old/new name clash so 
> we effectively change semantics of an existing function. 
> *But* the new names are better
> imo: get_domain_by_id() more obviously pairs with put_domain(), and
> find_domain_by_id_noref() is ugly and this rename allows us 
> to drop the _noref suffix unambiguously.
> 

  Yes, I also thought we needed better names but my thinking was that
  we could change the names in separate patches after we were sure
  no code was still using find_domain_by_id() with the old semantics.
  Since I can only run and compile the x86 version I was afraid that
  I might have missed some code using find_domain_by_id() in the
  other archs. But I guess I was being too conservative. It is 
  probably fine to change the names as you suggested in the next
  patch. 

> I'd be interested to know what this does for 
> find/get_domain_by_id() percentages in your network profiles.
> 

  For TCP TX find_domain_by_id() cost is reduced from 3.05% of the
  total CPU cycles to 0.19% in my machine using the new version. 
  I have not run a RX benchmark yet, but I can get those numbers
  Monday.

  Regards

  Renato

>  -- Keir
> 
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel

<Prev in Thread] Current Thread [Next in Thread>