[PATCH v4 02/10] crash hp: Introduce CRASH_HOTPLUG configuration options

[David Hildenbrand]

On 03.03.22 11:22, Baoquan He wrote:
> On 03/02/22 at 10:20am, David Hildenbrand wrote:
>> On 01.03.22 21:04, Eric DeVolder wrote:
>>>
>>>
>>> On 2/22/22 21:25, Baoquan He wrote:
>>>> On 02/09/22 at 02:56pm, Eric DeVolder wrote:
>>>>> Support for CPU and memory hotplug for crash is controlled by the
>>>>> CRASH_HOTPLUG configuration option, introduced by this patch.
>>>>>
>>>>> The CRASH_HOTPLUG_ELFCOREHDR_SZ related configuration option is
>>>>> also introduced with this patch.
>>>>>
>>>>> Signed-off-by: Eric DeVolder <eric.devolder at oracle.com>
>>>>> ---
>>>>>   arch/x86/Kconfig | 26 ++++++++++++++++++++++++++
>>>>>   1 file changed, 26 insertions(+)
>>>>>
>>>>> diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
>>>>> index ebe8fc76949a..4e3374edab02 100644
>>>>> --- a/arch/x86/Kconfig
>>>>> +++ b/arch/x86/Kconfig
>>>>> @@ -2060,6 +2060,32 @@ config CRASH_DUMP
>>>>>   	  (CONFIG_RELOCATABLE=y).
>>>>>   	  For more details see Documentation/admin-guide/kdump/kdump.rst
>>>>>   
>>>>> +config CRASH_HOTPLUG
>>>>> +	bool "kernel updates of crash elfcorehdr"
>>>>> +	depends on CRASH_DUMP && (HOTPLUG_CPU || MEMORY_HOTPLUG) && KEXEC_FILE
>>>>> +	help
>>>>> +	  Enable the kernel to update the crash elfcorehdr (which contains
>>>>> +	  the list of CPUs and memory regions) directly when hot plug/unplug
>>>>> +	  of CPUs or memory. Otherwise userspace must monitor these hot
>>>>> +	  plug/unplug change notifications via udev in order to
>>>>> +	  unload-then-reload the crash kernel so that the list of CPUs and
>>>>> +	  memory regions is kept up-to-date. Note that the udev CPU and
>>>>> +	  memory change notifications still occur (however, userspace is not
>>>>> +	  required to monitor for crash dump purposes).
>>>>> +
>>>>> +config CRASH_HOTPLUG_ELFCOREHDR_SZ
>>>>> +	depends on CRASH_HOTPLUG
>>>>> +	int
>>>>> +	default 131072
>>>>> +	help
>>>>> +	  Specify the maximum size of the elfcorehdr buffer/segment.
>>>>> +	  The 128KiB default is sized so that it can accommodate 2048
>>>>> +	  Elf64_Phdr, where each Phdr represents either a CPU or a
>>>>> +	  region of memory.
>>>>> +	  For example, this size can accommodate hotplugging a machine
>>>>> +	  with up to 1024 CPUs and up to 1024 memory regions (e.g. 1TiB
>>>>> +	  with 1024 1GiB memory DIMMs).
>>>>
>>>> This example of memory could be a little misleading. The memory regions
>>>> may not be related to memory DIMMs. System could split them into many
>>>> smaller regions during bootup.
>>>
>>> I changed "with 1024 1GiB memory DIMMs" to "with 1024 1GiB hotplug memories".
>>> eric
>>
>> It's still not quite precise. Essentially it's the individual "System
>> RAM" entries in /proc/iomem
>>
>> Boot memory (i.e., a single DIMM) might be represented by multiple
>> entries due to rearranged holes (by the BIOS).
>>
>> While hoplugged DIMMs (under virt!) are usually represented using a
>> single range, it can be different on physical machines. Last but not
>> least, dax/kmem and virtio-mem behave in a different way.
> 
> Right. How about only mentioning the 'System RAM' entries in /proc/iomem
> as below? It's just giving an example, talking about the details of
> memory regions from each type may not be necessry here. People
> interested can refer to code or document related to get it.
> 
> 
> + default 131072
> + help
> +   Specify the maximum size of the elfcorehdr buffer/segment.
> +   The 128KiB default is sized so that it can accommodate 2048
> +   Elf64_Phdr, where each Phdr represents either a CPU or a
> +   region of memory.
> +   For example, this size can accommodate hotplugging a machine
> +   with up to 1024 CPUs and up to 1024 memory regions which are
>     represented by 'System RAM' entries in /proc/iomem.

Maybe changing the last paragraph to:

"For example, this size can accommodate a machine with up to 1024 CPUs
and up to 1024 memory regions, for example, as represented by 'System
RAM' entries in /proc/iomem."


-- 
Thanks,

David / dhildenb