Hi Mark
Thank you for your comment
I'm sorry but I'm having a hard time understanding what this means in the context of your patch. What is a "sleeping bug"?
Sorry for my Engilish. It mean "(Maybe) old fsi_pointer() has bug, but this bug doesn't run a mischief"
Ah, I see - so it probably was an existing bug but nobody reported it?
Yes
I suspect it's worth fixing in 3.3 then, there's often issues where userspace applications don't work that don't get reported well or at all, often these are due to issues in the pointer implementations. People end up coming up with configurations that work around them and never think that there might be a driver issue.
I'm not good at detail of user side applications, but aplay was working correctly on old/PIO FSI. So, I didn't notice it.
Best regards --- Kuninori Morimoto