RE: [PATCH net-next] net: mana: Add support for variable page sizes of ARM64

From: Haiyang Zhang Sent: Tuesday, June 11, 2024 10:44 AM > > > -Original Message- > > From: Michael Kelley > > Sent: Tuesday, June 11, 2024 12:35 PM > > To: Haiyang Zhang ; > > mailto:linux-hyperv@vger.kernel.org; > > mailto:

Re: [PATCH v1 1/3] mm: pass meminit_context to __free_pages_core()

On 11.06.24 21:41, Tim Chen wrote: On Fri, 2024-06-07 at 11:09 +0200, David Hildenbrand wrote: In preparation for further changes, let's teach __free_pages_core() about the differences of memory hotplug handling. Move the memory hotplug specific handling from generic_online_page() to __free_pag

Re: [PATCH v1 1/3] mm: pass meminit_context to __free_pages_core()

On Fri, 2024-06-07 at 11:09 +0200, David Hildenbrand wrote: > In preparation for further changes, let's teach __free_pages_core() > about the differences of memory hotplug handling. > > Move the memory hotplug specific handling from generic_online_page() to > __free_pages_core(), use adjust_manage

Re: [PATCH v1 2/3] mm/memory_hotplug: initialize memmap of !ZONE_DEVICE with PageOffline() instead of PageReserved()

On Tue, 11 Jun 2024 11:42:56 +0200 David Hildenbrand wrote: > > We'll leave the ZONE_DEVICE case alone for now. > > > > @Andrew, can we add here: > > "Note that self-hosted vmemmap pages will no longer be marked as > reserved. This matches ordinary vmemmap pages allocated from the buddy > du

Re: [PATCH v1 1/3] mm: pass meminit_context to __free_pages_core()

On 11.06.24 21:19, Andrew Morton wrote: On Tue, 11 Jun 2024 12:06:56 +0200 David Hildenbrand wrote: On 07.06.24 11:09, David Hildenbrand wrote: In preparation for further changes, let's teach __free_pages_core() about the differences of memory hotplug handling. Move the memory hotplug specif

Re: [PATCH v1 1/3] mm: pass meminit_context to __free_pages_core()

On Tue, 11 Jun 2024 12:06:56 +0200 David Hildenbrand wrote: > On 07.06.24 11:09, David Hildenbrand wrote: > > In preparation for further changes, let's teach __free_pages_core() > > about the differences of memory hotplug handling. > > > > Move the memory hotplug specific handling from generic_o

RE: [PATCH net-next] net: mana: Add support for variable page sizes of ARM64

> -Original Message- > From: Haiyang Zhang > Sent: Tuesday, June 11, 2024 1:44 PM > To: Michael Kelley ; linux-hyperv@vger.kernel.org; > net...@vger.kernel.org; Paul Rosswurm > Cc: Dexuan Cui ; step...@networkplumber.org; KY > Srinivasan ; o...@aepfle.de; vkuznets > ; da...@davemloft.n

RE: [PATCH net-next] net: mana: Add support for variable page sizes of ARM64

(resending in plain text) > -Original Message- > From: Michael Kelley > Sent: Tuesday, June 11, 2024 12:35 PM > To: Haiyang Zhang ; > mailto:linux-hyperv@vger.kernel.org; > mailto:net...@vger.kernel.org > Cc: Dexuan Cui

RE: [PATCH net-next] net: mana: Add support for variable page sizes of ARM64

From: LKML haiyangz On Behalf Of Haiyang Zhang Sent: Monday, June 10, 2024 2:23 PM > > As defined by the MANA Hardware spec, the queue size for DMA is 4KB > minimal, and power of 2. You say the hardware requires 4K "minimal". But the definitions in this patch hardcode to 4K, as if that's the on

Re: [PATCH v2 1/6] arm64/hyperv: Support DeviceTree

On 5/16/2024 8:27 AM, Roman Kisel wrote: On 5/15/2024 3:02 PM, Elliot Berman wrote: On Tue, May 14, 2024 at 03:43:48PM -0700, Roman Kisel wrote: The Virtual Trust Level platforms rely on DeviceTree, and the arm64/hyperv code supports ACPI only. Update the logic to support DeviceTree on boo

Re: [PATCH 1/1] x86/hyperv: Set X86_FEATURE_TSC_KNOWN_FREQ when Hyper-V provides frequency

On 6/5/2024 7:55 PM, mhkelle...@gmail.com wrote: From: Michael Kelley A Linux guest on Hyper-V gets the TSC frequency from a synthetic MSR, if available. In this case, set X86_FEATURE_TSC_KNOWN_FREQ so that Linux doesn't unnecessarily do refined TSC calibration when setting up the TSC clocks

Re: [PATCH v2 6/6] drivers/pci/hyperv/arm64: vPCI MSI IRQ domain from DT

On 6/7/2024 12:55 PM, Bjorn Helgaas wrote: On Wed, May 15, 2024 at 01:12:38PM -0500, Bjorn Helgaas wrote: On Wed, May 15, 2024 at 09:34:09AM -0700, Roman Kisel wrote: On 5/15/2024 2:48 AM, Saurabh Singh Sengar wrote: On Tue, May 14, 2024 at 03:43:53PM -0700, Roman Kisel wrote: The hyperv

Re: [PATCH v1 1/3] mm: pass meminit_context to __free_pages_core()

On 07.06.24 11:09, David Hildenbrand wrote: In preparation for further changes, let's teach __free_pages_core() about the differences of memory hotplug handling. Move the memory hotplug specific handling from generic_online_page() to __free_pages_core(), use adjust_managed_page_count() on the me

Re: [PATCH v1 2/3] mm/memory_hotplug: initialize memmap of !ZONE_DEVICE with PageOffline() instead of PageReserved()

On 07.06.24 11:09, David Hildenbrand wrote: We currently initialize the memmap such that PG_reserved is set and the refcount of the page is 1. In virtio-mem code, we have to manually clear that PG_reserved flag to make memory offlining with partially hotplugged memory blocks possible: has_unmovab

Re: [PATCH v1 2/3] mm/memory_hotplug: initialize memmap of !ZONE_DEVICE with PageOffline() instead of PageReserved()

On 11.06.24 09:45, Oscar Salvador wrote: On Mon, Jun 10, 2024 at 10:56:02AM +0200, David Hildenbrand wrote: There are fortunately not that many left. I'd even say marking them (vmemmap) reserved is more wrong than right: note that ordinary vmemmap pages after memory hotplug are not reserved! On

Re: [PATCH v1 2/3] mm/memory_hotplug: initialize memmap of !ZONE_DEVICE with PageOffline() instead of PageReserved()

On Fri, Jun 07, 2024 at 11:09:37AM +0200, David Hildenbrand wrote: > We currently initialize the memmap such that PG_reserved is set and the > refcount of the page is 1. In virtio-mem code, we have to manually clear > that PG_reserved flag to make memory offlining with partially hotplugged > memory

Re: [PATCH v1 2/3] mm/memory_hotplug: initialize memmap of !ZONE_DEVICE with PageOffline() instead of PageReserved()

On Mon, Jun 10, 2024 at 10:56:02AM +0200, David Hildenbrand wrote: > There are fortunately not that many left. > > I'd even say marking them (vmemmap) reserved is more wrong than right: note > that ordinary vmemmap pages after memory hotplug are not reserved! Only > bootmem should be reserved. Ok