[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Xen-devel] [RFC PATCH v2 10/25] x86: NUMA: Move numa code and make it generic
- To: Vijay Kilari <vijay.kilari@xxxxxxxxx>
- From: Julien Grall <julien.grall@xxxxxxx>
- Date: Tue, 9 May 2017 09:26:31 +0100
- Cc: Tim Deegan <tim@xxxxxxx>, Stefano Stabellini <sstabellini@xxxxxxxxxx>, Wei Liu <wei.liu2@xxxxxxxxxx>, George Dunlap <George.Dunlap@xxxxxxxxxxxxx>, Andrew Cooper <andrew.cooper3@xxxxxxxxxx>, Ian Jackson <ian.jackson@xxxxxxxxxxxxx>, "xen-devel@xxxxxxxxxxxxx" <xen-devel@xxxxxxxxxxxxx>, Jan Beulich <jbeulich@xxxxxxxx>, Vijaya Kumar K <Vijaya.Kumar@xxxxxxxxxx>
- Delivery-date: Tue, 09 May 2017 08:26:53 +0000
- List-id: Xen developer discussion <xen-devel.lists.xen.org>
On 05/09/2017 08:39 AM, Vijay Kilari wrote:
On Mon, May 8, 2017 at 10:21 PM, Julien Grall <julien.grall@xxxxxxx> wrote:
On 28/03/17 16:53, vijay.kilari@xxxxxxxxx wrote:
+static int numa_setup(char *s);
+custom_param("numa", numa_setup);
+
+struct node_data node_data[MAX_NUMNODES];
+
+/* Mapping from pdx to node id */
+unsigned int memnode_shift;
+static typeof(*memnodemap) _memnodemap[64];
Also, you move the hardcoded 64 here. But have you checked it is valid for
ARM?
Regardless that, this sounds like something that should be turned into a
define and require a comment.
64 is good enough. This _memnodemap is used in case of NUMA failed or off,
in which case memnode_shift is 63 (BITS_PER_LONG -1).
If it based on BITS_PER_LONG, then you should use BITS_PER_LONG (via a
proper define) rather than hardcoding it.
--
Julien Grall
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel
|
