If you want to emulate a ConnecCore 91 on a ConnectCore 93, download and install ConnectCore 93 images and follow the instructions in Emulate a ConnectCore 91 SoM on a ConnectCore 93. |
The ConnectCore 93 SOM variant included in the development kit has the following specifications:
Number of cores |
2 |
---|---|
RAM size |
1GB |
Digi offers the ConnectCore 93 SOM in different configurations, typically with different number of cores and RAM memory size. This topic explains how you can simulate a variant with fewer cores and/or less RAM memory and verify how well it performs for your application.
The following instructions and excerpts are generic. The information displayed may vary depending on the variant you have. |
Limit the number of operating cores
You can check the number of operating cores by looking at the kernel boot messages:
# dmesg | grep -i cpu
[ 0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=4, Nodes=1
[ 0.000000] rcu: RCU restricting CPUs from NR_CPUS=256 to nr_cpu_ids=4.
[ 0.006195] smp: Bringing up secondary CPUs ...
[ 0.008550] smp: Brought up 1 node, 4 CPUs
[ 1.926581] ledtrig-cpu: registered to indicate activity on CPUs
Deactivate cores at run-time
To deactivate a certain core at run-time:
# echo 0 > /sys/devices/system/cpu/cpuX/online
where X is the index of the core that you want to deactivate.
To deactivate a core from your application code, you can use the APIX library. See Manage CPU cores.
Deactivate cores at boot time
You can use the kernel command line argument maxcpus=N
to limit the number of cores that will operate in Linux.
For example, to limit the number of cores to 2, append the following to the environment variable extra_bootargs
:
=> env edit extra_bootargs
edit: maxcpus=2
If you want to keep this setting across reboots, save the environment with saveenv .
|
Boot Linux with this configuration and verify the number of active cores:
# dmesg | grep -i cpus
[ 0.000000] Kernel command line: console=ttyLP2,115200 root=PARTUUID=1c606ef5-f1ac-43b9-9bb5-d5c578580b6b rootwait rw fbcon=logo-pos:center maxcpus=2
[ 0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=4, Nodes=1
[ 0.000000] rcu: RCU restricting CPUs from NR_CPUS=256 to nr_cpu_ids=4.
[ 0.006210] smp: Bringing up secondary CPUs ...
[ 0.007099] smp: Brought up 1 node, 2 CPUs
[ 1.896259] ledtrig-cpu: registered to indicate activity on CPUs
Check the status of cores with:
# grep "" /sys/devices/system/cpu/cpu*/online
/sys/devices/system/cpu/cpu0/online:1
/sys/devices/system/cpu/cpu1/online:1
/sys/devices/system/cpu/cpu2/online:0
/sys/devices/system/cpu/cpu3/online:0
In the status log above, only the first two cores are online.
Simulate a smaller RAM size
You can check the current RAM size by looking at the kernel boot messages:
# dmesg | grep -i memory
[ 0.000000] Reserved memory: created CMA memory pool at 0x0000000096000000, size 640 MiB
[ 0.000000] Early memory node ranges
[ 0.000000] Memory: 1188060K/2015232K available (14652K kernel code, 1050K rwdata, 4584K rodata, 2496K init, 1035K bss, 171812K reserved, 655360K cma-reserved)
[ 0.140943] No USDPAA memory, no 'fsl,usdpaa-mem' in device-tree
[ 3.668498] Freeing unused kernel memory: 2496K
You can use the kernel command line argument mem=size
to limit the amount of RAM memory available to Linux.
For example, to limit the RAM memory to 512 MB, append the following to the environment variable extra_bootargs
:
=> env edit extra_bootargs
edit: mem=512M
If you want to keep this setting across reboots, save the environment with saveenv .
|
Boot Linux with this configuration and verify the available memory:
# dmesg | grep -i memory
[ 0.000000] Memory limited to 512MB
[ 0.000000] Reserved memory: created CMA memory pool at 0x000000009a000000, size 96 MiB
[ 0.000000] Early memory node ranges
[ 0.000000] Memory: 196060K/444416K available (14652K kernel code, 1050K rwdata, 4584K rodata, 2496K init, 1035K bss, 150052K reserved,98304K cma-reserved)
[ 0.123944] No USDPAA memory, no 'fsl,usdpaa-mem' in device-tree
[ 3.608580] Freeing unused kernel memory: 2496K