Queuing multiple command is implementation defined?

[matrem]

I just pay attention to this note of the OpenCL specification (1.0.48) :

@Note p50

It is possible that a device(s) becomes unavailable after a context and command-queues that use this device(s) have been created and commands have been queued to command-queues. In this case the behavior of OpenCL API calls that use this context (and command-queues) are considered to be implementation-defined. The user callback function, if specified when the context is created, can be used to record appropriate information in the errinfo, private_info arguments passed to the callback function when the device becomes unavailable.

Does it really means that second command queuing behavior is implementation dependent? :?

[affie]

The behavior of multiple command-queues is not implementation defined. What this text is saying is that the behavior of context and command-queues is implementation defined if the device associated with the context / queue becomes unavailable. An example of a device becoming unavailable would be the device getting powered off.

[matrem]

Ok I’m relieved, I understood with that sentence:

It is possible that a device(s) becomes unavailable after a context and command-queues that use this device(s) have been created and commands have been queued to command-queues.

that a device can become unavailable because of its usage by command(s) queued…