04:18 <alexeymin> what is the reason to use .suppress_bind_attrs = true in .driver (or not to use)?

04:22 <bamse> alexeymin: there's files named "unbind" sprinkled in /sys, which can be used to unbind a device from its driver...i.e. remove it

04:24 <bamse> alexeymin: so setting it to true means you don't need to care (although i haven't looked at the exact details in a while) about your device/driver having to do cleanup

04:28 <alexeymin> so basically if you're too lazy to write proper cleanup/remove handlers

04:36 <bamse> that's a typical reason for attempting it yes

04:36 <bamse> but there's cases where it's just too complicated, or there's no return if the driver goes away

04:38 <bamse> e.g. drivers/soc/qcom/smem.c has it set, because we hand out pointers to clients directly...so iounmap the memory would be bad

04:39 <bamse> and i initially wrote a version with an api where we could handle that, but it wasn't worth all the extra code in smem and the clients

11:01 <julian[m]12> bamse: Would it be possible to get it upstream to set usb-charge-current-limit to some large value for sony-castor?

20:44 <steev> sboyd: i'm playing around with 5.13 on the lenovo yoga c630, and i'm getting https://dpaste.com/EPEYEKGPJ#line-544 - not quite sure what is going on as this is definitely new with 5.13... but also... that code hasn't been touched since 2014? and the disp_cc_mdss_pclk0_clk_src code hasn't been touched since 2018?