antwain.schneider writes via Kernel.org Bugzilla: rock64 exposes i2s1 on a secondary 22 pin header on the machine, this includes gpio2b7 since i don't care about i2s and those pins would be more valuable as gpios, i explicitly disabled i2s1 in the device tree i set gpio2b7 as a cs-gpio but it wasn't working, eventually i got bold and actually started looking at the grf in memory ma -k /dev/gpiomem dd 0x28 4 00000020 -------- -------- 00000001 no matter how many i2s1 things i disabled or removed from the dtb, GRF_GPIO2BH_IOMUX always read 00000001, meaning i2s1_mclk seems hardcoded eventually i found the description in commit 3818e4a7678ea70508b6b6e72a4cbf445e1e7dfa and realized gpio2b4 also gives me trouble i feel this is a bug, but probably not important for a 6 year old chip View: https://bugzilla.kernel.org/show_bug.cgi?id=217334#c0 You can reply to this message to join the discussion. -- Deet-doot-dot, I am a bot. Kernel.org Bugzilla (peebz 0.1)