Hi Matthew,

On 11/6/2024 12:25 AM, Matthew Wilcox wrote:

On Tue, Nov 05, 2024 at 04:45:45PM +0000, Shivank Garg wrote:

quoted

This patch series introduces fbind() syscall to support NUMA memory
policies for KVM guest_memfd, allowing VMMs to configure memory placement
for guest memory. This addresses the current limitation where guest_memfd
allocations ignore NUMA policies, potentially impacting performance of
memory-locality-sensitive workloads.

Why does guest_memfd ignore numa policies?  The pagecache doesn't,
eg in vma_alloc_folio_noprof().

guest_memfd doesn't have VMAs and hence can't store policy information in
VMA and use vma_alloc_folio_noprof() that fetches mpol from VMA.

The folio allocation path from guest_memfd typically looks like this...

kvm_gmem_get_folio
  filemap_grab_folio
    __filemap_get_folio
      filemap_alloc_folio
        __folio_alloc_node_noprof
          -> goes to the buddy allocator

Hence, I am trying to have a version of filemap_alloc_folio() that takes an mpol.


Thanks,
Shivank