In the Linux kernel, the following vulnerability has been resolved:

gpio: zynq: fix runtime PM leak on request error path

pm runtime get sync() leaves the usage counter incremented even when it fails, and zynq gpio request() returns the error without dropping it.

gpiolib does not call ->free() when ->request() fails, so zynq gpio free(), which holds the only matching pm runtime put(), never runs. The reference is leaked and the controller can no longer runtime-suspend, so its clock stays enabled.

Switch to pm runtime resume and get(), which only increments the usage counter on success.