On 4 April 2016 at 14:39, Ulf Hansson <ulf.hansson@xxxxxxxxxx> wrote:
> On 2 April 2016 at 01:04, Gwendal Grignou <gwendal@xxxxxxxxxxxx> wrote:
>> When the eMMC device does not support/allow sending RST_n signal,
>> try to do a brute force power cycle instead of returning EOPNOTSUPP.
>> If power cycle is not supported by the hardware, mmc_init_card still
>> send a CMD0, performing a Software Reset.
>>
>> Signed-off-by: Gwendal Grignou <gwendal@xxxxxxxxxxxx>
>> ---
>> Changes in v3:
>> - clock setting in already done in power_cycle.
>> - clarify behavior when power cycle is not implemented. .
>> drivers/mmc/core/core.c | 5 +++--
>> drivers/mmc/core/mmc.c | 24 +++++++++++-------------
>> 2 files changed, 14 insertions(+), 15 deletions(-)
>>
>> diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
>> index d8bbc78..d5bc2d8 100644
>> --- a/drivers/mmc/core/core.c
>> +++ b/drivers/mmc/core/core.c
>> @@ -2468,8 +2468,9 @@ int mmc_hw_reset(struct mmc_host *host)
>> ret = host->bus_ops->reset(host);
>> mmc_bus_put(host);
>>
>> - if (ret != -EOPNOTSUPP)
>> - pr_warn("%s: tried to reset card\n", mmc_hostname(host));
>> + if (ret)
>> + pr_warn("%s: tried to reset card, got error %d\n",
>> + mmc_hostname(host), ret);
>>
>> return ret;
>> }
>> diff --git a/drivers/mmc/core/mmc.c b/drivers/mmc/core/mmc.c
>> index 96cc7e2..92b7bea 100644
>> --- a/drivers/mmc/core/mmc.c
>> +++ b/drivers/mmc/core/mmc.c
>> @@ -1988,19 +1988,17 @@ static int mmc_reset(struct mmc_host *host)
>> {
>> struct mmc_card *card = host->card;
>>
>> - if (!(host->caps & MMC_CAP_HW_RESET) || !host->ops->hw_reset)
>> - return -EOPNOTSUPP;
>> -
>> - if (!mmc_can_reset(card))
>> - return -EOPNOTSUPP;
>> -
>> - mmc_set_clock(host, host->f_init);
>> -
>> - host->ops->hw_reset(host);
>> -
>> - /* Set initial state and call mmc_set_ios */
>> - mmc_set_initial_state(host);
>> -
>> + if ((host->caps & MMC_CAP_HW_RESET) && host->ops->hw_reset &&
>> + mmc_can_reset(card)) {
>> + /* If the card accept RST_n signal, send it. */
>> + mmc_set_clock(host, host->f_init);
>> + host->ops->hw_reset(host);
>> + /* Set initial state and call mmc_set_ios */
>> + mmc_set_initial_state(host);
>> + } else {
>> + /* Do a brute force power cycle */
>> + mmc_power_cycle(host, card->ocr);
>> + }
>> return mmc_init_card(host, card->ocr, card);
>> }
>>
>> --
>> 2.8.0.rc3.226.g39d4020
>>
>
> There are more two callers triggering the host->ops->hw_reset() to be
> called, which also have special treatment of the error code
> -EOPNOTSUPP. These are mmc_hw_reset() and mmc_blk_reset().
>
> Please make sure to update their error handling accordingly. Otherwise
> this looks okay to me.
After some more investigations, I realized that the error path in the
mmc block layer needed some additional cleanups.
Therefore I have decided to apply the v3 version of $subject patch for
next, then I am going to post some cleanups changes on top.
I also took the liberty to update the changelog as I think it needed
to be extended with some more information.
Thanks and kind regards
Uffe
--
To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
