[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <afcf367f-5b33-5249-ad85-2addede18dc5@linux.com>
Date: Thu, 29 Feb 2024 08:56:20 -0800 (PST)
From: "Christoph Lameter (Ampere)" <cl@...ux.com>
To: Huang Shijie <shijie@...amperecomputing.com>
cc: kuba@...nel.org, patches@...erecomputing.com, davem@...emloft.net,
horms@...nel.org, edumazet@...gle.com, ast@...nel.org, dhowells@...hat.com,
linyunsheng@...wei.com, aleksander.lobakin@...el.com,
linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
cl@...amperecomputing.com
Subject: Re: [PATCH v2] net: skbuff: set FLAG_SKB_NO_MERGE for
skbuff_fclone_cache
On Tue, 27 Feb 2024, Huang Shijie wrote:
> Since we do not set FLAG_SKB_NO_MERGE for skbuff_fclone_cache,
> the current skbuff_fclone_cache maybe not really allocated, it maybe
> used an exist old kmem_cache. In NUMA, the fclone allocated by
> alloc_skb_fclone() maybe in remote node.
This is not the right approach. If you want to force a local allocation
you need to use GFP_THISNODE. Merging has nothing to do with locality.
Powered by blists - more mailing lists