On 12/25/21 1:20 PM, Uwe Kleine-König wrote:
Hello,
some sidestory to this series: While working on fixing the struct dev
lifetime thing, I saw a few ugly things that I don't intend to fix
myself. Here is just a mail to let you know this with a list of issues.
Feel free to address or ignore:
- 104_QUAD_8 depends on X86, but compiles fine on ARCH=arm. Maybe
adding support for COMPILE_TEST would be a good idea.
- 104-quad-8.c uses devm_request_irq() and (now) devm_counter_add(). On
unbind an irq might be pending which results in quad8_irq_handler()
calling counter_push_event() for a counter that is already
unregistered. (The issue exists also without my changes.)
- I think intel-qep.c makes the counter unfunctional in
intel_qep_remove before the counter is unregistered.
- I wonder why counter is a bus and not a class device type. There is
no driver that would ever bind a counter device, is there? So
/sys/bus/counter/driver is always empty.
This last item has been brought up before. The conclusion was yes it
should have been a class but it is too late to change it now since it
would break userspace.
(I think bus was cargo-culted from the iio subsystem.)
