On 19 September 2016 at 10:00, Wolfram Sang <wsa@xxxxxxxxxxxxx> wrote:
> On Thu, Sep 15, 2016 at 01:52:26PM +0200, Geert Uytterhoeven wrote:
>> Hi Wolfram,
>>
>> On Wed, Sep 14, 2016 at 7:05 PM, Wolfram Sang
>> <wsa+renesas@xxxxxxxxxxxxxxxxxxxx> wrote:
>> > The DTS changes for R-Car Gen3 will come via a seperate series. Note that
>> > 'non-removable' is not supported yet because of Runtime PM issues. It seems we
>> > need to overhaul Runtime PM handling for other reasons as well, so I suggest
>> > the basic support goes in like this and DTS do not use 'non-removable' for now.
>>
>> DT describes the hardware, not current limitations of the software.
>> Hence IMHO you should add the "non-removable" property to the DTS, and
>> work around its non-functioning in software.
>
> @Ulf: Would you accept such a workaround for a while?
With some comments about why, yes!
Br
Uffe
>
> diff --git a/drivers/mmc/host/tmio_mmc_pio.c b/drivers/mmc/host/tmio_mmc_pio.c
> index f21c92ec7121cc..5ab8af294f7c40 100644
> --- a/drivers/mmc/host/tmio_mmc_pio.c
> +++ b/drivers/mmc/host/tmio_mmc_pio.c
> @@ -1154,6 +1154,10 @@ int tmio_mmc_host_probe(struct tmio_mmc_host *_host,
> !mmc_card_is_removable(mmc) ||
> mmc->slot.cd_irq >= 0);
>
> + /* Workaround for NONREMOVABLE until we fix RuntimePM handling */
> + if (pdata->flags & TMIO_MMC_MIN_RCAR2)
> + _host->native_hotplug = true;
> +
> if (tmio_mmc_clk_enable(_host) < 0) {
> mmc->f_max = pdata->hclk;
> mmc->f_min = mmc->f_max / 512;
>
