Re: [PATCH 2/3] trace: allow HVMOP_xentrace to set trace record subclass

[George Dunlap <george.dunlap@eu.citrix.com>]

On 25/07/13 14:23, David Vrabel wrote:
> From: David Vrabel <david.vrabel@citrix.com>
>
> Allow guests adding trace records with HVMOP_xentrace to set the
> sub-class.  This allows different guest generated traces to be
> filtered by xentrace.
>
> Signed-off-by: David Vrabel <david.vrabel@citrix.com>
> ---
>   xen/arch/x86/hvm/hvm.c     |    4 ++--
>   xen/include/public/trace.h |    3 +++
>   2 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
> index e2701b6..b0d8094 100644
> --- a/xen/arch/x86/hvm/hvm.c
> +++ b/xen/arch/x86/hvm/hvm.c
> @@ -4388,10 +4388,10 @@ long do_hvm_op(unsigned long op, XEN_GUEST_HANDLE_PARAM(void) arg)
>               return -EFAULT;
>   
>           if ( tr.extra_bytes > sizeof(tr.extra)
> -             || (tr.event & ~((1u<<TRC_SUBCLS_SHIFT)-1)) )
> +             || (tr.event & ~((1u<<TRC_CLS_SHIFT)-1)) )
>               return -EINVAL;
>   
> -        trace_var(tr.event | TRC_GUEST, 1 /*cycles*/,
> +        trace_var(TRC_GUEST_EVENT(tr.event), 1 /*cycles*/,
>                     tr.extra_bytes, tr.extra);
>           break;
>       }
> diff --git a/xen/include/public/trace.h b/xen/include/public/trace.h
> index e2f60a6..bbbd38b 100644
> --- a/xen/include/public/trace.h
> +++ b/xen/include/public/trace.h
> @@ -244,6 +244,9 @@
>   #define TRC_HW_IRQ_UNMAPPED_VECTOR    (TRC_HW_IRQ + 0x7)
>   #define TRC_HW_IRQ_HANDLED            (TRC_HW_IRQ + 0x8)
>   
> +/* Guest event with guest-specified sub-class and number. */
> +#define TRC_GUEST_EVENT(e) (0x08000000 | (e & 0xffff))

Minor nit -- should this be ((1u<<TRC_CLS_SHIFT)-1)?

Alternately, would it make sense to add the following, and use it both 
places in this patch?

#define TRC_CLS_MASK ((1u<<TRC_CLS_SHIFT)-1)

  -George