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

Re: [Xen-devel] [PATCH v6 03/14] x86: detect and initialize Intel CAT feature



>>> On 07.05.15 at 11:08, <chao.p.peng@xxxxxxxxxxxxxxx> wrote:
> On Fri, Apr 24, 2015 at 03:59:32PM +0100, Jan Beulich wrote:
>> >>> On 23.04.15 at 11:55, <chao.p.peng@xxxxxxxxxxxxxxx> wrote:
>> >  struct psr_cmt *__read_mostly psr_cmt;
>> > +
>> > +static unsigned long __read_mostly * cat_socket_init_bitmap;
>> > +static unsigned long __read_mostly * cat_socket_enable_bitmap;
>> > +static struct psr_cat_socket_info *__read_mostly cat_socket_info;
>> 
>> So the last line looks right, and the line of context left in place is
>> well formed too. Why did you not follow these good examples for
>> the two newly added lines?
>> 
>> And what is cat_socket_enable_bitmap needed for anyway?
> 
> Sorry for missing this. 
> 
> It indicates if CAT is enabled on the socket while
> cat_socket_init_bitmap only indicates the socket is initialized(but the
> socket may not support CAT or CAT is not enabled successfully by
> software).
> 
> cat_socket_init_bitmap is totally for performance optimization, to avoid
> initializing the same socket for more than one times.
> If you think one should removed, then cat_socket_init_bitmap will be
> the candidate.

If the two (by the end of the series) really serve different purposes,
then I'm fine with keeping them both. Otherwise please ditch one.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel


 


Rackspace

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