krzk changed the topic of #linux-exynos to: Samsung S3C24xx and S3C64xx planned deprecation and removal (see LKML) | Linux Samsung SoC Exynos | https://exynos.wiki.kernel.org | This channel is logged: https://libera.irclog.whitequark.org/linux-exynos
GNUtoo has quit [Remote host closed the connection]
GNUtoo has joined #linux-exynos
<Leandro[m]1> Is this error related to my issues with the emmc? 🤔
<Leandro[m]1> Wait... I know 7885 and 7885 aren't exactly the same, what if that's my issue
<Leandro[m]1> 7884*
<krzk> Leandro[m]1: the exynos-clockout error? or what? Because I don't see there errors with emmc...
<krzk> Leandro[m]1: the clkout error comes because you do not have proper compatible for your PMU (clkout is modelled as a child)
<Leandro[m]1> krzk: Yes, and I posted the emmc error earlier, i don't quite know how irc works soo uhh
<krzk> Leandro[m]1: you can grep for the error and you will see the list of compatibles it looks for. It should not be critical, usually clkout is not used
<krzk> I did not see any error earlier. I saw only link to image starting with simple-fb
<krzk> ahh.. like earlier earlier :)
<Leandro[m]1> Yep
<krzk> Leandro[m]1: I thought you got answer to that... or you disappeared so I could not give it. Maybe I am mixing channels.. Anyway
<krzk> Leandro[m]1: anyway - git grep for it. This leads to "if (!host->bus_hz) {" - you miss "ciu
<krzk> "ciu" clock
<krzk> If you want a but more detailed help, link to your DTSI and DTS, because otherwise it is half guessing...
<Leandro[m]1> https://pastebin.com/Dc0pkBAz here's my dts
<Leandro[m]1> Dtsi is upstream exynos-7885.dtsi
<Leandro[m]1> Some of the things in the MMC node aren't a thing in mainline driver I heard so that is known to me
<krzk> Leandro[m]1: that's weird then. exynos7885.dtsi has ciu clock defined, so this should work. Maybe your cmu_fsys clock controller did not probe for some reason?
<krzk> Leandro[m]1: or the clock hierarchy or the clock registers are a bit different, thus `host->bus_hz = clk_get_rate(host->ciu_clk);` returned 0?
<krzk> You should compare vendor/downstream code around these clocks (CLK_GOUT_MMC_EMBD_SDCLKIN and CLK_DOUT_FSYS_MMC_EMBD)
<krzk> and add more debug commands to drivers/mmc/host/dw_mmc.c to know why bus_hz==0. Does the devm_clk_get() suceeds? If not, there will be only dev_dbg message, so maybe you do not see that failure
<Leandro[m]1> Hmm
<Leandro[m]1> I'll come back tomorrow when I need help lol