[PATCH v1] mtd: rawnand: meson: initialize clock register
Liang Yang
liang.yang at amlogic.com
Sun Nov 12 22:47:29 PST 2023
Hi Arseniy,
I remember that you asked it in another email, but i forgot to reply. Sorry
On 2023/11/12 18:08, Arseniy Krasnov wrote:
> [ EXTERNAL EMAIL ]
>
> On 09.11.2023 08:40, Arseniy Krasnov wrote:
>> Clock register must be also initialized during controller probing. If
>> this is not performed (for example by bootloader before) - controller
>> will not work.
>>
>> Signed-off-by: Arseniy Krasnov <avkrasnov at salutedevices.com>
>> ---
>> drivers/mtd/nand/raw/meson_nand.c | 4 +++-
>> 1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/mtd/nand/raw/meson_nand.c b/drivers/mtd/nand/raw/meson_nand.c
>> index 0d4d358152d7..4e7fa943928c 100644
>> --- a/drivers/mtd/nand/raw/meson_nand.c
>> +++ b/drivers/mtd/nand/raw/meson_nand.c
>> @@ -91,6 +91,8 @@
>>
>> /* eMMC clock register, misc control */
>> #define CLK_SELECT_NAND BIT(31)
>> +#define CLK_ALWAYS_ON_NAND BIT(24)
>> +#define CLK_ENABLE_VALUE 0x245
> ^^^^^^
>
> Hi,
>
> @Liang, it will be great, if You'll give some details about this magic value. I get it
> from vendor's driver and it makes NAND controller alive, but I don't have any docs.
Bit 7:6 is used to select the clock source for NAND controller.
00 : more clock source
01 : fix pll is select, commonly 1GHZ in Amlogic soc chips.
The implementation in meson nfc driver should select '01', so you could
do like this :
#define CLK_SELECT_FIX_PLL2 BIT(6)
writel(CLK_ALWAYS_ON_NAND | CLK_SELECT_NAND | CLK_SELECT_FIX_PLL2,
nfc->reg_clk);
Bit 5:0 is the clock divider.
>
> Thanks, Arseniy
>
>>
>> #define NFC_CLK_CYCLE 6
>>
>> @@ -1152,7 +1154,7 @@ static int meson_nfc_clk_init(struct meson_nfc *nfc)
>> return PTR_ERR(nfc->nand_clk);
>>
>> /* init SD_EMMC_CLOCK to sane defaults w/min clock rate */
>> - writel(CLK_SELECT_NAND | readl(nfc->reg_clk),
>> + writel(CLK_ALWAYS_ON_NAND | CLK_SELECT_NAND | CLK_ENABLE_VALUE,
>> nfc->reg_clk);
>>
>> ret = clk_prepare_enable(nfc->core_clk);
--
Thanks,
Liang
More information about the linux-mtd
mailing list