As a workaround, you can do something similar with GPF. You can create a subset which contains only the mask and then you can write the result as CSV format.
I have tried with this workaround, the problem is that the CSV file is huge (ca.2GB), probably containig all image pixels, even those which where not selected by my mask (whuich I created myself with an equation)…
The thing that I need is a list of only pixels satisfying a certain condition, their values and their coordinates…
You could set this equation as valid-pixel expression, then pixels which do not meet the expression will be NaN in the output. You could delete these pixels afterwards. But I agree it is still very large an not ideal.