[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <c63dd8c4-a66a-4a97-ac94-70b3159ba3a8@suse.cz>
Date: Fri, 24 Jan 2025 09:42:07 +0100
From: Vlastimil Babka <vbabka@...e.cz>
To: Marco Elver <elver@...gle.com>
Cc: cl@...two.org, Alexander Potapenko <glider@...gle.com>,
Dmitry Vyukov <dvyukov@...gle.com>, Jonathan Corbet <corbet@....net>,
Andrew Morton <akpm@...ux-foundation.org>, Yang Shi <shy828301@...il.com>,
Huang Shijie <shijie@...amperecomputing.com>, kasan-dev@...glegroups.com,
workflows@...r.kernel.org, linux-doc@...r.kernel.org,
linux-kernel@...r.kernel.org, Christoph Lameter <cl@...ux.com>
Subject: Re: [PATCH] KFENCE: Clarify that sample allocations are not following
NUMA or memory policies
On 1/24/25 09:37, Marco Elver wrote:
> On Fri, 24 Jan 2025 at 09:13, Vlastimil Babka <vbabka@...e.cz> wrote:
>>
>> On 1/23/25 23:44, Christoph Lameter via B4 Relay wrote:
>> > From: Christoph Lameter <cl@...ux.com>
>> >
>> > KFENCE manages its own pools and redirects regular memory allocations
>> > to those pools in a sporadic way. The usual memory allocator features
>> > like NUMA, memory policies and pfmemalloc are not supported.
>>
>> Can it also violate __GFP_THISNODE constraint? That could be a problem, I
>> recall a problem in the past where it could have been not honoured by the
>> page allocator, leading to corruption of slab lists.
>
> KFENCE does not sample page allocator allocations. Is kmalloc()
> allowed to take __GFP_THISNODE?
Yeah and SLUB is honouring it.
Powered by blists - more mailing lists