Re: [PATCH v2] mmc: dw_mmc: add support for RK3288

From: Doug Anderson
Date: Tue Jul 29 2014 - 12:38:20 EST

Addy,

On Mon, Jul 28, 2014 at 9:52 PM, Doug Anderson <dianders@xxxxxxxxxxxx> wrote:
> Addy,
>
> On Wed, Jul 9, 2014 at 8:31 PM, Addy Ke <addy.ke@xxxxxxxxxxxxxx> wrote:
>> This patch focuses on clock setting for RK3288 mmc controller.
>>
>> In RK3288 mmc controller, CLKDIV register can only be set 0 or 1,
>> and if DDR 8bit mode, CLKDIV register must be set 1.
>>
>> Signed-off-by: Addy Ke <addy.ke@xxxxxxxxxxxxxx>
>> ---
>> changes since v1:
>> - dw_mci_rk3288_setup_clock: do not call clk_get_rate(), just use the
>> host->bus_hz which is already called by dw_mmc.c, suggested by Jaehoon Chung
>>
>> .../devicetree/bindings/mmc/rockchip-dw-mshc.txt | 4 +-
>> drivers/mmc/host/dw_mmc-pltfm.c | 50 +++++++++++++++++++++-
>> 2 files changed, 51 insertions(+), 3 deletions(-)
>>
>> diff --git a/Documentation/devicetree/bindings/mmc/rockchip-dw-mshc.txt b/Documentation/devicetree/bindings/mmc/rockchip-dw-mshc.txt
>> index c559f3f..e3f95cd 100644
>> --- a/Documentation/devicetree/bindings/mmc/rockchip-dw-mshc.txt
>> +++ b/Documentation/devicetree/bindings/mmc/rockchip-dw-mshc.txt
>> @@ -10,7 +10,9 @@ extensions to the Synopsys Designware Mobile Storage Host Controller.
>> Required Properties:
>>
>> * compatible: should be
>> - - "rockchip,rk2928-dw-mshc": for Rockchip RK2928 and following
>> + - "rockchip,rk2928-dw-mshc": for Rockchip RK2928 and following,
>> + before RK3288
>> + - "rockchip,rk3288-dw-mshc": for Rockchip RK3288
>>
>> Example:
>>
>> diff --git a/drivers/mmc/host/dw_mmc-pltfm.c b/drivers/mmc/host/dw_mmc-pltfm.c
>> index d4a47a9..809c28b 100644
>> --- a/drivers/mmc/host/dw_mmc-pltfm.c
>> +++ b/drivers/mmc/host/dw_mmc-pltfm.c
>> @@ -21,17 +21,61 @@
>> #include <linux/mmc/mmc.h>
>> #include <linux/mmc/dw_mmc.h>
>> #include <linux/of.h>
>> +#include <linux/clk.h>
>>
>> #include "dw_mmc.h"
>> #include "dw_mmc-pltfm.h"
>>
>> +#define RK3288_CLKGEN_DIV 2
>
> Yup, this matches what I see in the TRM. It will always divide by 2
> to allow for 4 phases (picking the phases not supported yet). Default
> phase looks to be 180 degrees which is why we've (currently) got
> USE_HOLD_REG hardcoded. :)
>
>> +
>> static void dw_mci_pltfm_prepare_command(struct dw_mci *host, u32 *cmdr)
>> {
>> *cmdr |= SDMMC_CMD_USE_HOLD_REG;
>> }
>>
>> -static const struct dw_mci_drv_data rockchip_drv_data = {
>> +static int dw_mci_rk3288_setup_clock(struct dw_mci *host)
>> +{
>> + host->bus_hz /= RK3288_CLKGEN_DIV;
>> +
>> + return 0;
>> +}
>> +
>> +static void dw_mci_rk3288_set_ios(struct dw_mci *host, struct mmc_ios *ios)
>> +{
>> + int ret;
>> + unsigned int cclkin;
>> +
>> + /*
>> + * cclkin: source clock of mmc controller.
>> + * bus_hz: card interface clock generated by CLKGEN.
>> + * bus_hz = cclkin / RK3288_CLKGEN_DIV;
>> + * ios->clock = (div == 0) ? bus_hz : (bus_hz / (2 * div))
>> + *
>> + * Note: div can only be 0 or 1
>> + * if DDR50 8bit mode, div must be set 1
>
> Makes sense. So this function is essentially reversing the logic in
> dw_mmc and making sure that we'll get the right DIV (0 or 1) in
> dw_mci_setup_bus().
>
>
>> + */
>> + if ((ios->bus_width == MMC_BUS_WIDTH_8) &&
>> + (ios->timing == MMC_TIMING_UHS_DDR50 ||
>
> Probably don't need UHS_DDR50 since (I think) you can't have an 8-bit
> SD card--only MMC, right? ...but it doesn't hurt.
>
>
>> + ios->timing == MMC_TIMING_MMC_DDR52))
>> + cclkin = 2 * ios->clock * RK3288_CLKGEN_DIV;
>> + else
>> + cclkin = ios->clock * RK3288_CLKGEN_DIV;
>> +
>> + ret = clk_set_rate(host->ciu_clk, cclkin);
>> + if (ret)
>> + dev_warn(host->dev, "failed to set rate %uHz\n", ios->clock);
>> +
>> + host->bus_hz = clk_get_rate(host->ciu_clk) / RK3288_CLKGEN_DIV;
>> +}
>> +
>> +static const struct dw_mci_drv_data rk2928_drv_data = {
>> + .prepare_command = dw_mci_pltfm_prepare_command,
>> +};
>> +
>> +static const struct dw_mci_drv_data rk3288_drv_data = {
>> .prepare_command = dw_mci_pltfm_prepare_command,
>> + .set_ios = dw_mci_rk3288_set_ios,
>> + .setup_clock = dw_mci_rk3288_setup_clock,
>> };
>>
>> static const struct dw_mci_drv_data socfpga_drv_data = {
>> @@ -95,7 +139,9 @@ EXPORT_SYMBOL_GPL(dw_mci_pltfm_pmops);
>> static const struct of_device_id dw_mci_pltfm_match[] = {
>> { .compatible = "snps,dw-mshc", },
>> { .compatible = "rockchip,rk2928-dw-mshc",
>> - .data = &rockchip_drv_data },
>> + .data = &rk2928_drv_data },
>> + { .compatible = "rockchip,rk3288-dw-mshc",
>> + .data = &rk3288_drv_data },
>> { .compatible = "altr,socfpga-dw-mshc",
>> .data = &socfpga_drv_data },
>> {},
>
> Reviewed-by: Doug Anderson <dianders@xxxxxxxxxxxx>
> Tested-by: Doug Anderson <dianders@xxxxxxxxxxxx>

It turns out that I spoke too soon. I realized that in my testing I
didn't have "DDR" mode enabled for my eMMC card. When I did that then
your patch didn't work.

The problem is that dw_mci_setup_bus() doesn't realize that you've
changed "bus_hz" so it needs to re-run. You can fix it like this:


diff --git a/drivers/mmc/host/dw_mmc-pltfm.c b/drivers/mmc/host/dw_mmc-pltfm.c
index 809c28b..bcc96d0 100644
--- a/drivers/mmc/host/dw_mmc-pltfm.c
+++ b/drivers/mmc/host/dw_mmc-pltfm.c
@@ -44,6 +44,7 @@ static void dw_mci_rk3288_set_ios(struct dw_mci
*host, struct mmc_ios *ios)
{
int ret;
unsigned int cclkin;
+ u32 bus_hz;

/*
* cclkin: source clock of mmc controller.
@@ -65,7 +66,11 @@ static void dw_mci_rk3288_set_ios(struct dw_mci
*host, struct mmc_ios *ios)
if (ret)
dev_warn(host->dev, "failed to set rate %uHz\n", ios->clock);

- host->bus_hz = clk_get_rate(host->ciu_clk) / RK3288_CLKGEN_DIV;
+ bus_hz = clk_get_rate(host->ciu_clk) / RK3288_CLKGEN_DIV;
+ if (bus_hz != host->bus_hz) {
+ host->bus_hz = bus_hz;
+ host->current_speed = 0; /* force dw_mci_setup_bus() */
+ }
}

static const struct dw_mci_drv_data rk2928_drv_data = {


Do you want to spin the patch for this?

-Doug
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/