Messages in this thread | | | Date | Wed, 8 Jan 2014 13:05:26 -0500 | From | Santosh Shilimkar <> | Subject | Re: [PATCH] x86/mm: memblock: switch to use NUMA_NO_NODE |
| |
On Wednesday 08 January 2014 12:27 PM, Grygorii Strashko wrote: > Hi, > > On 01/08/2014 06:23 PM, Grygorii Strashko wrote: >> Update X86 code to use NUMA_NO_NODE instead of MAX_NUMNODES while >> calling memblock APIs, because memblock API is changed to use NUMA_NO_NODE and >> will produce warning during boot otherwise. >> >> See: >> https://lkml.org/lkml/2013/12/9/898 >> > [...] > > or, there are other 3 patches from Sergey Senozhatsky, which actually fix the same warnings: > https://lkml.org/lkml/2014/1/6/277 - [PATCH -next] x86 memtest: use NUMA_NO_NODE in do_one_pass() > https://lkml.org/lkml/2014/1/6/280 - [PATCH -next] e820: use NUMA_NO_NODE in memblock_find_dma_reserve() > http://comments.gmane.org/gmane.linux.kernel/1623429 - [PATCH -next] check: use NUMA_NO_NODE in setup_bios_corruption_check() > Either one should be fine though $subject patch would be my personal preference.
Andrew, This should kill at least 3 known memblock users with MAX_NUMNODES. Feel free to pick the patch(s) as per your preference.
Regards, Santosh
| |