In the Linux kernel, the following vulnerability has been resolved:
memblock: make memblocksetnode() also warn about use of MAX_NUMNODES
On an (old) x86 system with SRAT just covering space above 4Gb:
ACPI: SRAT: Node 0 PXM 0 [mem 0x100000000-0xfffffffff] hotplug
the commit referenced below leads to this NUMA configuration no longer being refused by a CONFIG_NUMA=y kernel (previously
NUMA: nodes only cover 6144MB of your 8185MB e820 RAM. Not used.
No NUMA configuration found
Faking a node at [mem 0x0000000000000000-0x000000027fffffff]
was seen in the log directly after the message quoted above), because of memblockvalidatenumacoverage() checking for NUMANONODE (only). This in turn led to memblockallocrangenid()'s warning about MAXNUMNODES triggering, followed by a NULL deref in memmapinit() when trying to access node 64's (NODE_SHIFT=6) node data.
To compensate said change, make memblocksetnode() warn on and adjust a passed in value of MAX_NUMNODES, just like various other functions already do.