I can't reproduce that. I also can't easily wire in an MCP23017 to 44&45 at present as mine is on a HAT, but with "dtoverlay=mcp23017,noints,i2c_csi_dsi":I saw that, thanks for handling so quickly. I edited and recompiled the dtbo on my own device to test it out and the pinctrl error line does indeed disappear from dmesg. Unfortunately I won't be able to validate against an MCP23017 for a day or two as I accidentally damaged the chip last night and am waiting for a replacement. One thing to note with the update overlay: It seems that the i2c_csi_dsi parameter does not utilize I2C0 on pins 44 and 45 as described in the parameter documentation - instead, when that parameter is set SDA1 & SCL1 are enabled on pins 2 and 3:Code:
2: a0 pu | hi // GPIO2 = SDA13: a0 pu | hi // GPIO3 = SCL1
Code:
pi@bookworm32:~ $ pinctrl get 2,3 2: ip pu | hi // GPIO2 = input 3: ip pu | hi // GPIO3 = input pi@bookworm32:~ $ ls /dev/i2c*/dev/i2c-0 /dev/i2c-10 /dev/i2c-20 /dev/i2c-21 /dev/i2c-22Does /dev/i2c-1 exist / is listed by "i2cdetect -l"? If it exists, then it is correct that 2&3 will be set to a0 for I2C.
Statistics: Posted by 6by9 — Thu May 16, 2024 6:34 pm