How to utilize Intel Turbo Boost Technology in QNX

My target is Xeon D-Family, it has Turbo Boost Technology. While QNX prefers operating in low frequency, even in higher load.

cpufreq command can tell Turbo Boost Support in this target, but it says the model is not supported. It seems that this model is not covered in cpufreq tool.

The boost state is preferred in my project. How to make QNX running in a higher frequency.


This sounds to me like the kind of feature that would have to be handled by the kernel. You can’t modify the kernel itself but you can provide call outs at boot time. Even so, if it can be done at all, this sounds like it would be a challenging feature to implement.

You could approach QNX about this but at the very least you will need a support contract. It they are willing at all, it is likely they would ask for a large cash payment. That seems to be the way QNX works these days, coin opperated.

But there is a reason to reconsider this entirely. I’m not sure if it is the same feature, but I have a gaming computer with an Intel processor, and it can boost one of the cores to a higher frequency at the expense of the other cores. What does that do to the scheduling algorithm in a real time OS? Would you have the OS always running the highest priority processes on the higher frequency core? That could cause process migration issues. If not, you might be letting a lower priority process use a higher value resource.

The Xeon-D is an older processor and googling around I see support for QNX for it.

I’m curious to know which version of QNX you are using that is returning this error.


QNX version is QNX7.1

You are using the latest version of QNX.

I’d make an account at Foundry27 which is where the QNX people hang out and ask your question there. They may know why it isn’t working or isn’t supported or they may know a way to make it work (not sure whether OSTech or OSMeta is the better place to ask but either should get you an answer).