| Message ID | 20211124184603.3897245-1-john@metanate.com |
|---|---|
| Headers | show |
| Series | mmc: dw_mmc: start deprecating mshcN aliases | expand |
On Mittwoch, 24. November 2021 19:45:58 CET John Keeping wrote: > This series is prompted by discussion on a previous patch set [1] but is > a totally different approach and only a partial solution. > > With these patches, the dependency on the mshcN alias is totally removed > from dw_mmc-hi3798cv200 and dw_mmc-rockchip and dw_mmc-exynos moves > towards being able to consider the mshcN aliases deprecated. > > I haven't changed dw_mci_hi6220_caps here, although it looks like it's > possible to apply MMC_CAP_CMD23 to all controllers there with no change > in behaviour as the final entry is SDIO for which CMD23 is not > applicable IIUC. But I'm not familiar with that hardware and don't feel > confident making that change. > > [1] https://lore.kernel.org/all/20211116190244.1417591-1-john@metanate.com/ > > John Keeping (4): > mmc: dw_mmc: add common capabilities to replace caps > mmc: dw_mmc: hi3798cv200: use common_caps > mmc: dw_mmc: rockchip: use common_caps > mmc: dw_mmc: exynos: use common_caps > > drivers/mmc/host/dw_mmc-exynos.c | 9 +++++---- > drivers/mmc/host/dw_mmc-hi3798cv200.c | 9 +-------- > drivers/mmc/host/dw_mmc-rockchip.c | 11 +---------- > drivers/mmc/host/dw_mmc.c | 3 +++ > drivers/mmc/host/dw_mmc.h | 3 +++ > 5 files changed, 13 insertions(+), 22 deletions(-) > > For rockchip: Tested-by: Nicolas Frattaroli <frattaroli.nicolas@gmail.com> Tested on a rk3566 with no obvious issues arising. Regards, Nicolas Frattaroli
On 11/25/21 3:46 AM, John Keeping wrote: > The capabilities for all instances are the same, so use common_caps > instead of caps/num_caps to remove the dependency on the mshcN device > tree alias. > > Signed-off-by: John Keeping <john@metanate.com> Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com> Best Regards, Jaehoon Chung > --- > drivers/mmc/host/dw_mmc-hi3798cv200.c | 9 +-------- > 1 file changed, 1 insertion(+), 8 deletions(-) > > diff --git a/drivers/mmc/host/dw_mmc-hi3798cv200.c b/drivers/mmc/host/dw_mmc-hi3798cv200.c > index 39794f93826f..e9437ef8ef19 100644 > --- a/drivers/mmc/host/dw_mmc-hi3798cv200.c > +++ b/drivers/mmc/host/dw_mmc-hi3798cv200.c > @@ -23,12 +23,6 @@ struct hi3798cv200_priv { > struct clk *drive_clk; > }; > > -static unsigned long dw_mci_hi3798cv200_caps[] = { > - MMC_CAP_CMD23, > - MMC_CAP_CMD23, > - MMC_CAP_CMD23 > -}; > - > static void dw_mci_hi3798cv200_set_ios(struct dw_mci *host, struct mmc_ios *ios) > { > struct hi3798cv200_priv *priv = host->priv; > @@ -166,8 +160,7 @@ static int dw_mci_hi3798cv200_init(struct dw_mci *host) > } > > static const struct dw_mci_drv_data hi3798cv200_data = { > - .caps = dw_mci_hi3798cv200_caps, > - .num_caps = ARRAY_SIZE(dw_mci_hi3798cv200_caps), > + .common_caps = MMC_CAP_CMD23, > .init = dw_mci_hi3798cv200_init, > .set_ios = dw_mci_hi3798cv200_set_ios, > .execute_tuning = dw_mci_hi3798cv200_execute_tuning, >
On 11/25/21 3:46 AM, John Keeping wrote: > Move the common MMC_CAP_CMD23 capability to common_caps so that only the > special case of MMC_CAP_1_8V_DDR and MMC_CAP_8_BIT_DATA are set via > caps/num_caps. Both of those can, and should, be set via device tree > properties instead, so we can now say that exynos_dwmmc_caps is only > used for backwards compatibility. > > Signed-off-by: John Keeping <john@metanate.com> Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com> Added minor comment.. > --- > drivers/mmc/host/dw_mmc-exynos.c | 9 +++++---- > 1 file changed, 5 insertions(+), 4 deletions(-) > > diff --git a/drivers/mmc/host/dw_mmc-exynos.c b/drivers/mmc/host/dw_mmc-exynos.c > index c2dd29ef45c6..f76eeeb0cc53 100644 > --- a/drivers/mmc/host/dw_mmc-exynos.c > +++ b/drivers/mmc/host/dw_mmc-exynos.c > @@ -526,15 +526,16 @@ static int dw_mci_exynos_prepare_hs400_tuning(struct dw_mci *host, > > /* Common capabilities of Exynos4/Exynos5 SoC */ > static unsigned long exynos_dwmmc_caps[4] = { > - MMC_CAP_1_8V_DDR | MMC_CAP_8_BIT_DATA | MMC_CAP_CMD23, > - MMC_CAP_CMD23, > - MMC_CAP_CMD23, > - MMC_CAP_CMD23, > + MMC_CAP_1_8V_DDR | MMC_CAP_8_BIT_DATA, > + 0, > + 0, > + 0, > }; It can be removed all things. Best Regards, Jaehoon Chung > > static const struct dw_mci_drv_data exynos_drv_data = { > .caps = exynos_dwmmc_caps, > .num_caps = ARRAY_SIZE(exynos_dwmmc_caps), > + .common_caps = MMC_CAP_CMD23, > .init = dw_mci_exynos_priv_init, > .set_ios = dw_mci_exynos_set_ios, > .parse_dt = dw_mci_exynos_parse_dt, >
On Wed, 24 Nov 2021 at 20:24, Ulf Hansson <ulf.hansson@linaro.org> wrote: > > On Wed, 24 Nov 2021 at 19:46, John Keeping <john@metanate.com> wrote: > > > > This series is prompted by discussion on a previous patch set [1] but is > > a totally different approach and only a partial solution. > > > > With these patches, the dependency on the mshcN alias is totally removed > > from dw_mmc-hi3798cv200 and dw_mmc-rockchip and dw_mmc-exynos moves > > towards being able to consider the mshcN aliases deprecated. > > > > I haven't changed dw_mci_hi6220_caps here, although it looks like it's > > possible to apply MMC_CAP_CMD23 to all controllers there with no change > > in behaviour as the final entry is SDIO for which CMD23 is not > > applicable IIUC. But I'm not familiar with that hardware and don't feel > > confident making that change. > > > > [1] https://lore.kernel.org/all/20211116190244.1417591-1-john@metanate.com/ > > > > John Keeping (4): > > mmc: dw_mmc: add common capabilities to replace caps > > mmc: dw_mmc: hi3798cv200: use common_caps > > mmc: dw_mmc: rockchip: use common_caps > > mmc: dw_mmc: exynos: use common_caps > > > > drivers/mmc/host/dw_mmc-exynos.c | 9 +++++---- > > drivers/mmc/host/dw_mmc-hi3798cv200.c | 9 +-------- > > drivers/mmc/host/dw_mmc-rockchip.c | 11 +---------- > > drivers/mmc/host/dw_mmc.c | 3 +++ > > drivers/mmc/host/dw_mmc.h | 3 +++ > > 5 files changed, 13 insertions(+), 22 deletions(-) > > > > -- > > 2.34.0 > > > > This looks good to me, I intend to apply this later this week, unless > objections of course. > > In the meantime, I will continue to look at what we can do to resolve > the exynos/k3 issues around this. Let's consider additional changes to be on top of this, as this is certainly a nice step forward. So, applied for next, thanks! Kind regards Uffe