On 3/2/26 10:01, Huang, FangSheng (Jerry) wrote: > > > On 2/28/2026 4:34 AM, David Hildenbrand wrote: >> On 2/26/26 11:50, fanhuang wrote: >>> Add a 'memmap-type' option to NUMA node configuration that allows >>> specifying the memory type for a NUMA node. >>> >>> Supported values: >>> - normal: Regular system RAM (E820 type 1, default) >>> - spm: Specific Purpose Memory (E820 type 0xEFFFFFFF) >>> - reserved: Reserved memory (E820 type 2) >>> >>> The 'spm' type indicates Specific Purpose Memory - a hint to the guest >>> that this memory might be managed by device drivers based on guest >>> policy. >>> The 'reserved' type marks memory as not usable as RAM. >>> >>> Note: This option is only supported on x86 platforms. >>> >>> Usage: >>> -numa node,nodeid=1,memdev=m1,memmap-type=spm >>> >>> Signed-off-by: fanhuang <[email protected]> >>> --- >>> hw/core/numa.c | 24 ++++++++++++ >>> hw/i386/acpi-build.c | 8 ++++ >>> hw/i386/e820_memory_layout.c | 72 ++++++++++++++++++++++++++++++++++++ >>> hw/i386/e820_memory_layout.h | 12 +++--- >>> hw/i386/pc.c | 48 ++++++++++++++++++++++++ >>> include/system/numa.h | 7 ++++ >>> qapi/machine.json | 24 ++++++++++++ >>> qemu-options.hx | 14 ++++++- >>> 8 files changed, 202 insertions(+), 7 deletions(-) >> >> I didn't take a look at the x86 implementation bits. The high-level >> concept LGTM. >> >> In an ideal world, we'd only indicate the property if actually supported >> by the machine. Not sure if that is easy to achieve with the "-numa" >> option. So I guess this has to do :) >> > Hi David, > > Thanks for the review and the LGTM on the high-level concept! > > Regarding the per-machine property visibility — agreed, it would be > cleaner. Currently we handle it with a runtime error when memmap-type > is used on non-x86 machines, which seems like a reasonable compromise > given the "-numa" option structure. > > I was wondering if there's anything else you or the other maintainers > would like me to address in the current v6?
Not from my side, so Acked-by: David Hildenbrand <[email protected]> on the core bits. -- Cheers, David
