On Sun, Nov 16 2025 at 16:16, Ma Ke wrote:
of_find_device_by_node() increments the reference count but it's never decremented, preventing proper device cleanup. Add put_device() properly to ensure references released before function return.
Found by code review.
By whom? You sent 7 patches today which touch random parts of the kernel:
[PATCH] irqchip: Fix error handling in qcom_mpm_init [PATCH] phy: HiSilicon: Fix error handling in hi3670_pcie_get_resources_from_pcie [PATCH] ASoC: codecs: wcd937x: Fix error handling in wcd937x codec driver [PATCH] ASoC: codecs: Fix error handling in pm4125 audio codec driver [PATCH] powerpc/warp: Fix error handling in pika_dtm_thread [PATCH] USB: Fix error handling in gadget driver [PATCH] USB: ohci-nxp: Fix error handling in ohci-hcd-nxp driver
and in all of them you claim to have found them by code review.
Why do I have doubts especially when I look at your email address?
Thanks,
tglx