Configured Memory access latency and bandwidth not taking effect
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
QEMU |
Invalid
|
Undecided
|
Unassigned |
Bug Description
I was trying to configure latencies and bandwidths between nodes in a NUMA emulation using QEMU 5.0.0.
Host : Ubuntu 20.04 64 bit
Guest : Ubuntu 18.04 64 bit
The machine configured has 2 nodes. Each node has 2 CPUs and has been allocated 3GB of memory. The memory access latencies and bandwidths for a local access (i.e from initiator 0 to target 0, and from initiator 1 to target 1) are set as 40ns and 10GB/s respectively. The memory access latencies and bandwidths for a remote access (i.e from initiator 1 to target 0, and from initiator 0 to target 1) are set as 80ns and 5GB/s respectively.
The command line launch is as follows.
sudo x86_64-
-machine hmat=on \
-boot c \
-enable-kvm \
-m 6G,slots=
-object memory-
-object memory-
-numa node,nodeid=
-numa node,nodeid=
-smp 4,sockets=
-numa cpu,node-
-numa cpu,node-
-numa cpu,node-
-numa cpu,node-
-numa dist,src=
-net nic \
-net user \
-hda testing.img \
-numa hmat-lb,
-numa hmat-lb,
-numa hmat-lb,
-numa hmat-lb,
-numa hmat-lb,
-numa hmat-lb,
-numa hmat-lb,
-numa hmat-lb,
Then the latencies and bandwidths between the nodes were tested using the Intel Memory Latency Checker v3.9 (https:/
Latency_matrix with idle latencies (in ns)
Numa Node
. .0. . .1.
0 36.2 36.4
1 34.9 35.4
Bandwidth_matrix with memory bandwidths (in MB/s)
Numa Node
. . .0. . . .1.
0 15167.1 15308.9
1 15226.0 15234.0
A test was also conducted with the tool “lat_mem_rd” from lmbench to measure the memory read latencies. This also gave results which did not match the config.
Any information on why the config latency and bandwidth values are not applied, would be appreciated.
description: | updated |
description: | updated |
description: | updated |
Changed in qemu: | |
status: | New → Invalid |
On Sat, 25 Jul 2020 07:26:38 -0000
Vishnu Dixit <email address hidden> wrote:
> Public bug reported: softmmu/ qemu-system- x86_64 \ 2,maxmem= 7G \ backend- ram,size= 3G,id=m0 \ backend- ram,size= 3G,id=m1 \ 0,memdev= m0 \ 1,memdev= m1 \ 4,maxcpus= 4 \ id=0,socket- id=0 \ id=0,socket- id=1 \ id=1,socket- id=2 \ id=1,socket- id=3 \ 0,dst=1, val=20 \ initiator= 0,target= 0,hierarchy= memory, data-type= access- latency, latency= 40 \ initiator= 0,target= 0,hierarchy= memory, data-type= access- bandwidth, bandwidth= 10G \ initiator= 0,target= 1,hierarchy= memory, data-type= access- latency, latency= 80 \ initiator= 0,target= 1,hierarchy= memory, data-type= access- bandwidth, bandwidth= 5G \ initiator= 1,target= 0,hierarchy= memory, data-type= access- latency, latency= 80 \ initiator= 1,target= 0,hierarchy= memory, data-type= access- bandwidth, bandwidth= 5G \ initiator= 1,target= 1,hierarchy= memory, data-type= access- latency, latency= 40 \ initiator= 1,target= 1,hierarchy= memory, data-type= access- bandwidth, bandwidth= 10G \ /software. intel.com/ content/ www/us/ en/develop/ articles/ intelr- latency- checker. html). But the obtained results did not match the
>
> I was trying to configure latencies and bandwidths between nodes in a
> NUMA emulation using QEMU 5.0.0.
>
> Host : Ubuntu 20.04 64 bit
> Guest : Ubuntu 18.04 64 bit
>
> The machine configured has 2 nodes. Each node has 2 CPUs and has been
> allocated 3GB of memory. The memory access latencies and bandwidths for
> a local access (i.e from initiator 0 to target 0, and from initiator 1
> to target 1) are set as 40ns and 10GB/s respectively. The memory access
> latencies and bandwidths for a remote access (i.e from initiator 1 to
> target 0, and from initiator 0 to target 1) are set as 80ns and 5GB/s
> respectively.
>
> The command line launch is as follows.
>
> sudo x86_64-
> -machine hmat=on \
> -boot c \
> -enable-kvm \
> -m 6G,slots=
> -object memory-
> -object memory-
> -numa node,nodeid=
> -numa node,nodeid=
> -smp 4,sockets=
> -numa cpu,node-
> -numa cpu,node-
> -numa cpu,node-
> -numa cpu,node-
> -numa dist,src=
> -net nic \
> -net user \
> -hda testing.img \
> -numa hmat-lb,
> -numa hmat-lb,
> -numa hmat-lb,
> -numa hmat-lb,
> -numa hmat-lb,
> -numa hmat-lb,
> -numa hmat-lb,
> -numa hmat-lb,
>
> Then the latencies and bandwidths between the nodes were tested using
> the Intel Memory Latency Checker v3.9
> (https:/
> memory-
> configuration. The following are the results obtained.
>
> Latency_matrix with idle latencies (in ns)
>
> Numa Node
> . .0. . .1.
> 0 36.2 36.4
> 1 34.9 35.4
>
> Bandwidth_matrix with memory bandwidths (in MB/s)
>
> Numa Node
> . . .0. . . .1.
> 0 15167.1 15308.9
> 1 15226.0 15234.0
>
> A test was also conducted with the tool “lat_mem_rd” from lmbench to
> measure the memory read latencies. This also gave results which did not
> match the config.
>
> Any information on why the config latency and bandwidth values are not
> applied, would be appreciated.
There is no information about host hardware, so I'd onldy hazard a guess
that host is non NUMA machine so all guest RAM and CPUs are in the same
latency domain, so that's why you are seeing pretty much the same timings.
QEMU nor KMV do not simullate HW latencies at all, all that is configured...