From mboxrd@z Thu Jan 1 00:00:00 1970 From: jianxin.pan@amlogic.com (Jianxin Pan) Date: Sun, 4 Nov 2018 23:39:44 +0800 Subject: [PATCH v5 2/3] clk: meson: add DT documentation for emmc clock controller In-Reply-To: <154130064120.88331.5081431734592584205@swboyd.mtv.corp.google.com> References: <1539839245-13793-1-git-send-email-jianxin.pan@amlogic.com> <1539839245-13793-3-git-send-email-jianxin.pan@amlogic.com> <20181024145513.GA6647@ofant> <154130064120.88331.5081431734592584205@swboyd.mtv.corp.google.com> Message-ID: To: linus-amlogic@lists.infradead.org List-Id: linus-amlogic.lists.infradead.org Hi Stephen, Thank you for the review. On 2018/11/4 11:04, Stephen Boyd wrote: > Quoting Yixun Lan (2018-10-25 00:29:15) >> yes, I think the documentation need to be fixed >> >> for the final solution, we decide to make 'mmc-clkc' an independent node >> instead of being a sub-node of 'mmc', so both of them may exist in parallel.. >> >> the DT part may like this: >> >> sd_emmc_c_clkc: clock-controller at 7000 { >> compatible = "amlogic,axg-mmc-clkc", "syscon"; >> reg = <0x0 0x7000 0x0 0x4>; >> ... >> }; >> >> sd_emmc_c: mmc at 7000 { >> compatible = "amlogic,axg-mmc"; >> reg = <0x0 0x7000 0x0 0x800>; >> ... >> }; > > That's improper usage of DT. We don't want two devices at the same > register offset. sd_emmc_c_clkc is shared by nand and sd_emmc_c controller, and this clock is part of the MMC controller's register space. The idea of adding the clock-controller at 7000is introduced during the discussion in the NAND driver mainline effort: https://lkml.kernel.org/r/20180628090034.0637a062 at xps13 > > . >