diff options
author | Shawn Lin <shawn.lin@rock-chips.com> | 2017-01-09 16:56:20 +0800 |
---|---|---|
committer | Ulf Hansson <ulf.hansson@linaro.org> | 2017-02-13 13:20:13 +0100 |
commit | 5fc3d80ef496c3bc3e8c09f056e3586fb681de56 (patch) | |
tree | ee4887229f65ec02220617a7777e7936f044c031 /drivers/mmc/core/sdio.c | |
parent | 0eb51a58ad823b2081c69ec1cf165833dd11c450 (diff) | |
download | linux-5fc3d80ef496c3bc3e8c09f056e3586fb681de56.tar.bz2 |
mmc: sdio: don't use rocr to check if the card could support UHS mode
Per SDIO Simplified Specification V3, section 3.1.2, A host that
supports UHS-I sets S18R to 1 in the argument of CMD5 to request a
change of the signal voltage to 1.8V. If the card supports UHS-I and
the current signal voltage is 3.3V, S18A is set to 1 in the R4 response.
If the signal voltage is already 1.8V, the card sets S18A to 0 so that
host maintains the current signal voltage. UHS-I is supported in SD mode
and S18A is always 0 in SPI mode.
For the current code, if the signaling voltage is fixed 1.8v, so
the card will set S18A to 0 for rocr and thus we would clear the
R4_18V_PRESENT from ocr, which make core won't try to use uhs mode.
To fix it, we expect sdio_read_cccr would fail if the uhs mode won't
work at all. Note that it's interesting that some sdio cards still
response S18A even the voltage is fixed to 1.8v and the CMD11 will
also accepted and finish enabling UHS mode successfully. I guess this
is why folks didn't notice this problem. Anyway, fix it.
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Diffstat (limited to 'drivers/mmc/core/sdio.c')
-rw-r--r-- | drivers/mmc/core/sdio.c | 24 |
1 files changed, 17 insertions, 7 deletions
diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c index f417adc9feb1..6810b3a8e130 100644 --- a/drivers/mmc/core/sdio.c +++ b/drivers/mmc/core/sdio.c @@ -633,7 +633,11 @@ try_again: * to switch to 1.8V signaling level. No 1.8v signalling if * UHS mode is not enabled to maintain compatibility and some * systems that claim 1.8v signalling in fact do not support - * it. + * it. Per SDIO spec v3, section 3.1.2, if the voltage is already + * 1.8v, the card sets S18A to 0 in the R4 response. So it will + * fails to check rocr & R4_18V_PRESENT, but we still need to + * try to init uhs card. sdio_read_cccr will take over this task + * to make sure which speed mode should work. */ if (!powered_resume && (rocr & ocr & R4_18V_PRESENT)) { err = mmc_set_signal_voltage(host, MMC_SIGNAL_VOLTAGE_180, @@ -645,9 +649,6 @@ try_again: } else if (err) { ocr &= ~R4_18V_PRESENT; } - err = 0; - } else { - ocr &= ~R4_18V_PRESENT; } /* @@ -704,11 +705,20 @@ try_again: } /* - * Read the common registers. + * Read the common registers. Note that we should try to + * validate whether UHS would work or not. */ err = sdio_read_cccr(card, ocr); - if (err) - goto remove; + if (err) { + mmc_sdio_resend_if_cond(host, card); + if (ocr & R4_18V_PRESENT) { + /* Retry init sequence, but without R4_18V_PRESENT. */ + retries = 0; + goto try_again; + } else { + goto remove; + } + } /* * Read the common CIS tuples. |