I pretty much need the same thing as Programmatically determine NUMA node or PCI bus, device, function number of Direct3D9Ex device I’m looking for a way to programmatically determine which NUMA node a particular PCI device is connected to so that I can allocate memory for host to device transfers on that node. So, my code opens a device file
Tag: numa
How does linux calculate numa distance between two numa nodes ( internal logic of any numa library such as libnuma..etc )?
I tried to go through the numa_distance() and other related functions (From the 1st link ), But couldn’t understand. I am just trying to understand how linux calculates the NUMA distance between two nodes when this distance is said to vary based on architecture and NUMA interlink. I referred following links: https://github.com/jmesmon/numactl/blob/0df3f720e606a3706700e0487ba19d720f50c4b8/distance.c https://github.com/jmesmon/numactl/blob/0df3f720e606a3706700e0487ba19d720f50c4b8/numa.h https://github.com/jmesmon/numactl/blob/0df3f720e606a3706700e0487ba19d720f50c4b8/libnuma.c Answer Inside (recent versions of) the
Linux kernel: get information of page cache distribution over NUMA nodes
When Linux kernel runs on NUMA, each NUMA node has partially separate memory management. There is echo ‘?’ > /proc/sysrq-trigger function “Will dump current memory info to your console.” of SysRq (implemented as sysrq_handle_showmem and show_mem) to get basic memory statistics for every NUMA node to system console, dmesg and system kernel log. As I understand, there is data printed
How many NUMA nodes on a Power8 processor
I am using Ubuntu 15.04 on a two sockets Power8 machine, each socket has 10 cores. “numactl -H” outputs: The problem is, are there two NUMA nodes on each Power8 processor? Any why one has memory but the other one has nothing. I can’t find any document about this. Any information would be appreciated. A further question, if there are
How does the memory behind statically allocated huge pages get distributed across NUMA nodes?
In my /etc/default/grub file I have explicitly set aside N huge pages with “hugepages=N”. If I’m running on a box with 2 NUMA nodes, does N/2 huge pages get set aside for each node, or do they all go to node 0, or ….? Also is there a way on the command line to query how they’re split across nodes?
How do I know if my server has NUMA?
Hopping from Java Garbage Collection, I came across JVM settings for NUMA. Curiously I wanted to check if my CentOS server has NUMA capabilities or not. Is there a *ix command or utility that could grab this info? Answer I’m no expert here, but here’s something: Box 1, no NUMA: Box 2, some NUMA: