On Fri, Jan 03, 2020 at 11:23:47AM +0100, Stephan Gerhold wrote:
On Thu, Jan 02, 2020 at 11:05:37PM +0100, Greg Kroah-Hartman wrote:
From: Lee Jones lee.jones@linaro.org
[ Upstream commit 6b5c350648b857047b47acf74a57087ad27d6183 ]
Until now, MFD has assumed all child devices passed to it (via mfd_cells) are to be registered. It does not take into account requests from Device Tree and the like to disable child devices on a per-platform basis.
Well now it does.
Link: https://www.spinics.net/lists/arm-kernel/msg366309.html Link: https://lkml.org/lkml/2019/8/22/1350
As far as I understand it, this commit is not suitable for backporting. The link above explains the problem for a similar patch:
But I believe this would introduce a rather ugly bug in mfd_remove_devices() if the first sub-device is set to disabled: It iterates over the children devices to find the base address of the allocated "usage count" array, which is then used to free it. If the first sub-device is missing, it would free the wrong address.
This problem does not exist in mainline because the MFD usage counting was removed entirely as part of the larger "Simplify MFD Core" series [1].
None of the device trees in the stable kernels should depend on disabling MFD devices from the device tree. (They were written at a time when the MFD core ignored that request entirely...) Therefore, I would suggest to drop this patch entirely.
Thanks for catching this, now dropped from all stable queues.
greg k-h