[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <878qxiowmy.fsf@yhuang6-desk2.ccr.corp.intel.com>
Date: Wed, 31 Jul 2024 15:20:37 +0800
From: "Huang, Ying" <ying.huang@...el.com>
To: Gregory Price <gourry@...rry.net>
Cc: linux-mm@...ck.org, akpm@...ux-foundation.org, dave.jiang@...el.com,
Jonathan.Cameron@...wei.com, horenchuang@...edance.com,
linux-kernel@...r.kernel.org, linux-acpi@...r.kernel.org,
dan.j.williams@...el.com, lenb@...nel.org, "Aneesh Kumar K.V"
<aneesh.kumar@...ux.ibm.com>
Subject: Re: [PATCH] acpi/hmat,mm/memtier: always register hmat adist
calculation callback
Gregory Price <gourry@...rry.net> writes:
> On Wed, Jul 31, 2024 at 09:22:32AM +0800, Huang, Ying wrote:
>> Gregory Price <gourry@...rry.net> writes:
>>
>> > This presumes driver configured devices, which is not always the case.
>> >
>> > kmem.c will set MEMTIER_DEFAULT_DAX_ADISTANCE
>> >
>> > but if BIOS/EFI has set up the node instead, you get the default of
>> > MEMTIER_ADISTANCE_DRAM if HMAT is not present or otherwise not sane.
>>
>> "efi_fake_mem=" kernel parameter can be used to add "EFI_MEMORY_SP" flag
>> to the memory range, so that kmem.c can manage it.
>>
>
> In this case, the system is configured explicitly so that kmem does not
> manage it. In fact, some systems still cannot be managed with
> EFI_MEMORY_SP due to hpa!=spa issues that the driver cannot manage.
Sorry, I don't understand. IIUC, kmem.c can manage almost any memory
range via drivers/dax/hmem. Please check
drivers/dax/hmem/device.c
drivers/dax/hmem/hmem.c
Could you elaborate why kmem.c doesn't work for some memory range?
>> > Not everyone is going to have the ability to get a platform vendor to
>> > fix a BIOS bug, and I've seen this in production.
>>
>> So, some vendor build a machine with broken/missing HMAT/CDAT and wants
>> users to use CXL memory devices in it? Have the vendor tested whether
>> CXL memory devices work?
>>
>
> As I mentioned, the broken aspect is being fixed, however there are
> existing production hardware which do not have HMAT entries.
>
>> > But the first step here would be creating two modes. HMAT-is-sane and
>> > CPU/Non-CPU seems reasonable to me but open to opinions.
>>
>> IMHO, we should reduce user configurable knobs unless we can prove it
>> is really necessary.
>>
>
> That's fair and valid.
>
> But I think a feature that worked in 5.x should work in 6.x, and right
> now the change in node placement breaks hardware that worked with 5.x
> which happened to have broken or missing HMAT.
--
Best Regards,
Huang, Ying
Powered by blists - more mailing lists