Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP
From: Dave Hansen <hidden>
Date: 2020-04-30 16:04:28
Also in:
linux-acpi, linux-hyperv, linux-mm, linux-s390, lkml, nvdimm, virtualization, xen-devel
Possibly related (same subject, not in this thread)
- 2020-05-02 · Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP · Dan Williams <hidden>
- 2020-05-02 · Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP · David Hildenbrand <hidden>
- 2020-05-01 · Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP · Dan Williams <hidden>
- 2020-05-01 · Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP · David Hildenbrand <hidden>
- 2020-05-01 · Re: [PATCH v2 2/3] mm/memory_hotplug: Introduce MHP_NO_FIRMWARE_MEMMAP · Dan Williams <hidden>
On 4/30/20 8:52 AM, David Hildenbrand wrote:
quoted
Justifying behavior by documentation that does not consider memory hotplug is bad thinking.Are you maybe confusing this patch series with the arm64 approach? This is not about ordinary hotplugged DIMMs. I'd love to get Dan's, Dave's and Michal's opinion.
The impact on kexec from the DAX "kmem" driver's use of hotplug was inadvertent and unfortunate. The problem statement and solution seem pretty sane to me.