[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <SJ1PR11MB6083B9EC479B6B20684FC1C1FC2F2@SJ1PR11MB6083.namprd11.prod.outlook.com>
Date: Sat, 16 Mar 2024 00:08:12 +0000
From: "Luck, Tony" <tony.luck@...el.com>
To: Thomas Gleixner <tglx@...utronix.de>, "Chatre, Reinette"
<reinette.chatre@...el.com>, Rex Nie <rex.nie@...uarmicro.com>,
"james.morse@....com" <james.morse@....com>, "x86@...nel.org"
<x86@...nel.org>, Babu Moger <babu.moger@....com>, Peter Newman
<peternewman@...gle.com>, Borislav Petkov <bp@...en8.de>
CC: "Yu, Fenghua" <fenghua.yu@...el.com>, "ilpo.jarvinen@...ux.intel.com"
<ilpo.jarvinen@...ux.intel.com>, "linux-kernel@...r.kernel.org"
<linux-kernel@...r.kernel.org>
Subject: RE: 32bit resctrl? (was Re: [PATCH v2] fs/resctrl: fix domid loss
precision issue)
> Why? Making mon_data_bits::u larger in the way it has been done does not
> have any dependency on 32 or 64 builds unless I'm missing something.
That union is copied into the kernfs_node "priv" pointer field. So it has
to fit into whatever size the system is using as a pointer:
See:
mkdir_mondata_subdir()
->mon_addfile()
> Caring about 32biit resctrl on x86 is a pointless exercise.
Agree 100%
-Tony
Powered by blists - more mailing lists