lkml.org 
[lkml]   [2023]   [Sep]   [18]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [RFC PATCH 3/9] Documentation/scrub-configure.rst: Add documentation for scrub driver
From
On 18.09.23 12:25, Shiju Jose wrote:
> Hi David,
>
> Thanks for looking into this.
>
>> -----Original Message-----
>> From: David Hildenbrand <david@redhat.com>
>> Sent: 18 September 2023 08:24
>> To: Shiju Jose <shiju.jose@huawei.com>; linux-acpi@vger.kernel.org; linux-
>> mm@kvack.org; linux-kernel@vger.kernel.org
>> Cc: rafael@kernel.org; lenb@kernel.org; naoya.horiguchi@nec.com;
>> tony.luck@intel.com; james.morse@arm.com; dave.hansen@linux.intel.com;
>> jiaqiyan@google.com; jthoughton@google.com; somasundaram.a@hpe.com;
>> erdemaktas@google.com; pgonda@google.com; rientjes@google.com;
>> duenwen@google.com; Vilas.Sridharan@amd.com; mike.malvestuto@intel.com;
>> gthelen@google.com; Linuxarm <linuxarm@huawei.com>; Jonathan Cameron
>> <jonathan.cameron@huawei.com>; tanxiaofei <tanxiaofei@huawei.com>;
>> Zengtao (B) <prime.zeng@hisilicon.com>
>> Subject: Re: [RFC PATCH 3/9] Documentation/scrub-configure.rst: Add
>> documentation for scrub driver
>>
>> On 15.09.23 19:28, shiju.jose@huawei.com wrote:
>>> From: Shiju Jose <shiju.jose@huawei.com>
>>>
>>> Add documentation for scrub driver, supports configure scrub
>>> parameters, in Documentation/scrub-configure.rst
>>>
>>> Signed-off-by: Shiju Jose <shiju.jose@huawei.com>
>>> ---
>>> Documentation/scrub-configure.rst | 55
>> +++++++++++++++++++++++++++++++
>>> 1 file changed, 55 insertions(+)
>>> create mode 100644 Documentation/scrub-configure.rst
>>>
>>> diff --git a/Documentation/scrub-configure.rst
>>> b/Documentation/scrub-configure.rst
>>> new file mode 100644
>>> index 000000000000..9f8581b88788
>>> --- /dev/null
>>> +++ b/Documentation/scrub-configure.rst
>>> @@ -0,0 +1,55 @@
>>> +==========================
>>> +Scrub subsystem driver
>>> +==========================
>>> +
>>> +Copyright (c) 2023 HiSilicon Limited.
>>> +
>>> +:Author: Shiju Jose <shiju.jose@huawei.com>
>>> +:License: The GNU Free Documentation License, Version 1.2
>>> + (dual licensed under the GPL v2) :Original Reviewers:
>>> +
>>> +- Written for: 6.7
>>> +- Updated for:
>>> +
>>> +Introduction
>>> +------------
>>> +The scrub subsystem driver provides the interface for configure the
>>
>> "... interface for configuring memory scrubbers in the system."
>>
>> are we only configuring firmware/hw-based memory scrubbing? I assume so.
> The scrub control could be used for the SW based memory scrubbing too.

Okay, looks like there is not too much hw/firmware specific in there
(besides these weird range changes).
[...]

>>> +-------
>>> +
>>> + The usage takes the form shown in this example::
>>> +
>>> + # echo 0x300000 > /sys/class/scrub/scrub0/region0/addr_base
>>> + # echo 0x100000 > /sys/class/scrub/scrub0/region0/addr_size
>>> + # cat /sys/class/scrub/scrub0/region0/speed_available
>>> + # 1-60
>>> + # echo 25 > /sys/class/scrub/scrub0/region0/speed
>>> + # echo 1 > /sys/class/scrub/scrub0/region0/enable
>>> +
>>> + # cat /sys/class/scrub/scrub0/region0/speed
>>> + # 0x19
>>
>> Is it reasonable to return the speed as hex? You set it as dec.
> Presently return speed as hex to reduce the number of callback function needed
> for reading the hex/dec data because the values for the address range
> need to be in hex.

If speed_available returns dec, speed better also return dec IMHO.

>
>>
>>> + # cat /sys/class/scrub/scrub0/region0/addr_base
>>> + # 0x100000
>>
>> But didn't we set it to 0x300000 ...
> This is an emulated example for testing the RASF/RAS2 definition.
> According to the RASF & RAS2 definition, the actual address range in the
> platform could vary from the requested address range for the patrol scrubbing.
> "The platform calculates the nearest patrol scrub boundary address
> from where it can start". The platform returns the actual address range
> in response to GET_PATROL_PARAMETERS command to the firmware.
> Please see section 5.2.21.2.1 Hardware-based Memory Scrubbing ,
> Table 5.87: Parameter Block Structure for PATROL_SCRUB in the
> ACPI 6.5 specification.
>

So you configure [0x300000 - 0x400000] and you get [0x100000 - 0x300000]

How does that make any sense? :)

Shouldn't we rather return an error when setting a range that is
impossible, instead of the hardware deciding to scrub something
completely different (as can be seen in the example)?

--
Cheers,

David / dhildenb

\
 
 \ /
  Last update: 2023-09-18 14:19    [W:0.103 / U:0.060 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site