[PATCH v8 0/7] crash: Kernel handling of CPU and memory hot un/plug

[Eric DeVolder]

On 5/31/22 08:18, David Hildenbrand wrote:
> On 26.05.22 15:39, Sourabh Jain wrote:
>> Hello Eric,
>>
>> On 26/05/22 18:46, Eric DeVolder wrote:
>>>
>>>
>>> On 5/25/22 10:13, Sourabh Jain wrote:
>>>> Hello Eric,
>>>>
>>>> On 06/05/22 00:15, Eric DeVolder wrote:
>>>>> When the kdump service is loaded, if a CPU or memory is hot
>>>>> un/plugged, the crash elfcorehdr (for x86), which describes the CPUs
>>>>> and memory in the system, must also be updated, else the resulting
>>>>> vmcore is inaccurate (eg. missing either CPU context or memory
>>>>> regions).
>>>>>
>>>>> The current solution utilizes udev to initiate an unload-then-reload
>>>>> of the kdump image (e. kernel, initrd, boot_params, puratory and
>>>>> elfcorehdr) by the userspace kexec utility. In previous posts I have
>>>>> outlined the significant performance problems related to offloading
>>>>> this activity to userspace.
>>>>>
>>>>> This patchset introduces a generic crash hot un/plug handler that
>>>>> registers with the CPU and memory notifiers. Upon CPU or memory
>>>>> changes, this generic handler is invoked and performs important
>>>>> housekeeping, for example obtaining the appropriate lock, and then
>>>>> invokes an architecture specific handler to do the appropriate
>>>>> updates.
>>>>>
>>>>> In the case of x86_64, the arch specific handler generates a new
>>>>> elfcorehdr, and overwrites the old one in memory. No involvement
>>>>> with userspace needed.
>>>>>
>>>>> To realize the benefits/test this patchset, one must make a couple
>>>>> of minor changes to userspace:
>>>>>
>>>>>    - Disable the udev rule for updating kdump on hot un/plug changes.
>>>>>      Add the following as the first two lines to the udev rule file
>>>>>      /usr/lib/udev/rules.d/98-kexec.rules:
>>>>
>>>> If we can have a sysfs attribute to advertise this feature then
>>>> userspace
>>>> utilities (kexec tool/udev rules) can take action accordingly. In
>>>> short, it will
>>>> help us maintain backward compatibility.
>>>>
>>>> kexec tool can use the new sysfs attribute and allocate additional
>>>> buffer space
>>>> for elfcorehdr accordingly. Similarly, the checksum-related changes
>>>> can come
>>>> under this check.
>>>>
>>>> Udev rule can use this sysfs file to decide kdump service reload is
>>>> required or not.
>>>
>>> Great idea. I've been working on the corresponding udev and
>>> kexec-tools changes and your input/idea here is quite timely.
>>>
>>> I have boolean "crash_hotplug" as a core_param(), so it will show up as:
>>>
>>> # cat /sys/module/kernel/parameters/crash_hotplug
>>> N
>>
>> How about using 0-1 instead Y/N?
>> 0 = crash hotplug not supported
>> 1 = crash hotplug supported
>>
>> Also how about keeping sysfs here instead?
>> /sys/kernel/kexec_crash_hotplug
> 
> It's not only about hotplug, though. And actually we care about
> onlining/offlining. Hmm, I wonder if there is a better name for this
> automatic handling of cpu and memory devices.
> 
In the upcoming v9, there is no /sys/kernel/crash/kexec_crash_hotplug; I have sysfs attributes for 
memory blocks and CPUs named 'crash_hotplug' that can be utilized directly in udev rule as 
ATTR{crash_hotplug} to determine if the kernel is handling this for crash kernel update purposes.

Here's the current commit message for that change:

====
crash: memory and CPU hotplug sysfs attributes

This introduces the crash_hotplug attribute for memory and CPUs
for use by userspace.  This change directly facilitates the udev
rule for managing userspace re-loading of the crash kernel.

For memory, this changeset introduces the crash_hotplug attribute
to the /sys/devices/system/memory directory. For example:

  # udevadm info --attribute-walk /sys/devices/system/memory/memory81
   looking at device '/devices/system/memory/memory81':
     KERNEL=="memory81"
     SUBSYSTEM=="memory"
     DRIVER==""
     ATTR{online}=="1"
     ATTR{phys_device}=="0"
     ATTR{phys_index}=="00000051"
     ATTR{removable}=="1"
     ATTR{state}=="online"
     ATTR{valid_zones}=="Movable"

   looking at parent device '/devices/system/memory':
     KERNELS=="memory"
     SUBSYSTEMS==""
     DRIVERS==""
     ATTRS{auto_online_blocks}=="offline"
     ATTRS{block_size_bytes}=="8000000"
     ATTRS{crash_hotplug}=="1"

For CPUs, this changeset introduces the crash_hotplug attribute
to the /sys/devices/system/cpu directory. For example:

  # udevadm info --attribute-walk /sys/devices/system/cpu/cpu0
   looking at device '/devices/system/cpu/cpu0':
     KERNEL=="cpu0"
     SUBSYSTEM=="cpu"
     DRIVER=="processor"
     ATTR{crash_notes}=="277c38600"
     ATTR{crash_notes_size}=="368"
     ATTR{online}=="1"

   looking at parent device '/devices/system/cpu':
     KERNELS=="cpu"
     SUBSYSTEMS==""
     DRIVERS==""
     ATTRS{crash_hotplug}=="1"
     ATTRS{isolated}==""
     ATTRS{kernel_max}=="8191"
     ATTRS{nohz_full}=="  (null)"
     ATTRS{offline}=="4-7"
     ATTRS{online}=="0-3"
     ATTRS{possible}=="0-7"
     ATTRS{present}=="0-3"

With these changes in place, and by using the same attribute
crash_hotplug name, it is possible to efficiently instruct the
udev rule to skip crash kernel reloading.

For example, the following is the proposed udev rule change for RHEL
system 98-kexec.rules (as the first two lines of the rule file):

  # The kernel handles updates to crash elfcorehdr
  ATTRS{crash_hotplug}=="1", GOTO="kdump_reload_end"

When examined in the context of 98-kexec.rules, the above change
tests if crash_hotplug is set, and if so, it skips the userspace
initiated unload-then-reload of the crash kernel.
=====

Does that work for you?
Eric