[PATCH v4 2/4] dt-bindings: mmc: sdhci-of-dwcmhsc: Add Sophgo SG2042 support

Chen Wang unicorn_wang at outlook.com
Sun Jun 23 23:18:19 PDT 2024


On 2024/6/21 0:15, Jisheng Zhang wrote:
> On Tue, Jun 18, 2024 at 04:38:30PM +0800, Chen Wang wrote:
>> From: Chen Wang <unicorn_wang at outlook.com>
>>
>> SG2042 use Synopsys dwcnshc IP for SD/eMMC controllers.
>>
>> SG2042 defines 3 clocks for SD/eMMC controllers.
>> - AXI_EMMC/AXI_SD for aclk/hclk(Bus interface clocks in DWC_mshc)
>>    and blck(Core Base Clock in DWC_mshc), these 3 clocks share one
>>    source, so reuse existing "core".
> No, this seems not correct. This should be the "bus" clk, and your above
> sentence "aclk/hclk(Bus interface clocks in DWC_mshc)" implies this clk is
> for bus
>
>> - 100K_EMMC/100K_SD for cqetmclk(Timer clocks in DWC_mshc), so reuse
>>    existing "timer" which was added for rockchip specified.
>> - EMMC_100M/SD_100M for cclk(Card clocks in DWC_mshc), add new "card".
> I think this is "core" clk, no? Plz check which internal clks' clock
> source is the so called EMMC_100M/SD_100M.

hi, Jisheng,

Just want to double-confirm, corresponding to the definition of clock in 
the dwc-mshc specification, what's the "core" clock in the 
snps,dwcmshc-sdhci.yaml?

I get following clock definitions in user-guide of dwc-mshc 
specification, in section 1.8 "Speed and Clock Requirements"

- 1.8.1 Bus Interface Clocks,which are aclk and m_hclk/hclk
- 1.8.2 Timer Clocks,which are tmclk,cqetmclk
- 1.8.3 Card Clocks, whichi is cclk
- 1.8.4 Core Base Clock,which is bclk

I used to think "core" is "Core Base Clock".

[......]




More information about the linux-riscv mailing list