[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH 1 of 4] xentrace: reduce trace buffer size to something mfn_offset can reach


  • To: Olaf Hering <olaf@xxxxxxxxx>
  • From: George Dunlap <dunlapg@xxxxxxxxx>
  • Date: Thu, 26 May 2011 11:05:14 +0100
  • Cc: xen-devel@xxxxxxxxxxxxxxxxxxx, George Dunlap <george.dunlap@xxxxxxxxxx>
  • Delivery-date: Thu, 26 May 2011 03:06:44 -0700
  • Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:cc:content-type :content-transfer-encoding; b=iooUWHgNSL7s2iHiA3Ad6TDFQrGdZFPbtK/SbzMVALpVCmssu1oDiIxk/LQDmp63kD r99ORF3PBgVXmOKaZJOZDWcBgShRHOIsc7DHaLTju2cBxbQaXDPEciY3DJmARkB4xEqd C9f4m6ptDNoE9wdY/YO7B3uUSPRKd4n1y2Qcg=
  • List-id: Xen developer discussion <xen-devel.lists.xensource.com>

Acked-by: George Dunlap <george.dunlap@xxxxxxxxxxxxx>

Sorry for the delay!
 -George

On Tue, May 10, 2011 at 3:32 PM, Olaf Hering <olaf@xxxxxxxxx> wrote:
> # HG changeset patch
> # User Olaf Hering <olaf@xxxxxxxxx>
> # Date 1305037380 -7200
> # Node ID 8ac937fa527b28243227193bf4749feb3a234c2c
> # Parent  19452acd23045f40c4e18437f0a60f016757e5bd
> xentrace: reduce trace buffer size to something mfn_offset can reach
>
> The start of the array which holds the list of mfns for each cpus
> tracebuffer is stored in an unsigned short. This limits the total amount
> of pages for each cpu as the number of active cpus increases.
>
> Update the math in calculate_tbuf_size() to apply also this rule to the
> max number of trace pages. Without this change the index can overflow.
>
> Signed-off-by: Olaf Hering <olaf@xxxxxxxxx>
>
> diff -r 19452acd2304 -r 8ac937fa527b xen/common/trace.c
> --- a/xen/common/trace.c        Fri May 06 11:15:35 2011 +0100
> +++ b/xen/common/trace.c        Tue May 10 16:23:00 2011 +0200
> @@ -112,11 +112,14 @@ static int calculate_tbuf_size(unsigned
>     typeof(dummy_size.prod) max_size;
>     struct t_info dummy_pages;
>     typeof(dummy_pages.tbuf_size) max_pages;
> +    typeof(dummy_pages.mfn_offset[0]) max_mfn_offset;
> +    unsigned int max_cpus = num_online_cpus();
>     unsigned int t_info_words;
>
>     /* force maximum value for an unsigned type */
>     max_size = -1;
>     max_pages = -1;
> +    max_mfn_offset = -1;
>
>     /* max size holds up to n pages */
>     max_size /= PAGE_SIZE;
> @@ -124,6 +127,18 @@ static int calculate_tbuf_size(unsigned
>     if ( max_size < max_pages )
>         max_pages = max_size;
>
> +    /*
> +     * max mfn_offset holds up to n pages per cpu
> +     * The array of mfns for the highest cpu can start at the maximum value
> +     * mfn_offset can hold. So reduce the number of cpus and also the 
> mfn_offset.
> +     */
> +    max_mfn_offset -= t_info_first_offset - 1;
> +    max_cpus--;
> +    if ( max_cpus )
> +        max_mfn_offset /= max_cpus;
> +    if ( max_mfn_offset < max_pages )
> +        max_pages = max_mfn_offset;
> +
>     if ( pages > max_pages )
>     {
>         printk(XENLOG_INFO "xentrace: requested number of %u pages "
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-devel
>

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


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.