Message ID | 1481689038-128781-2-git-send-email-shawn.lin@rock-chips.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On 14 December 2016 at 05:17, Shawn Lin <shawn.lin@rock-chips.com> wrote: > 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> > > --- > > drivers/mmc/core/sdio.c | 22 +++++++++++++++------- > 1 file changed, 15 insertions(+), 7 deletions(-) > > diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c > index 47f824b..1d736e4 100644 > --- a/drivers/mmc/core/sdio.c > +++ b/drivers/mmc/core/sdio.c > @@ -636,7 +636,11 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, > * 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, > @@ -647,9 +651,6 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, > } else if (err) { > ocr &= ~R4_18V_PRESENT; > } > - err = 0; > - } else { > - ocr &= ~R4_18V_PRESENT; > } > > /* > @@ -706,11 +707,18 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, > } > > /* > - * Read the common registers. > + * Read the common registers. Note that we should try to > + * validate whether ush would work or not. ush? I guess "UHS" is what you want? > */ > + retries = 1; > err = sdio_read_cccr(card, ocr); > - if (err) > - goto remove; > + if (err) { > + mmc_sdio_retry_init_card(host, card, &retries); > + if (ocr & R4_18V_PRESENT) > + goto try_again; Re-trying like this, would potentially cause us looping forever. Do you really need to retry in this case? If so, there need to be a limit for how many times. > + else > + goto remove; > + } > > /* > * Read the common CIS tuples. > -- > 1.9.1 > Kind regards Uffe -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 2016/12/29 23:31, Ulf Hansson wrote: > On 14 December 2016 at 05:17, Shawn Lin <shawn.lin@rock-chips.com> wrote: >> 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> >> >> --- >> >> drivers/mmc/core/sdio.c | 22 +++++++++++++++------- >> 1 file changed, 15 insertions(+), 7 deletions(-) >> >> diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c >> index 47f824b..1d736e4 100644 >> --- a/drivers/mmc/core/sdio.c >> +++ b/drivers/mmc/core/sdio.c >> @@ -636,7 +636,11 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, >> * 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, >> @@ -647,9 +651,6 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, >> } else if (err) { >> ocr &= ~R4_18V_PRESENT; >> } >> - err = 0; >> - } else { >> - ocr &= ~R4_18V_PRESENT; >> } >> >> /* >> @@ -706,11 +707,18 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, >> } >> >> /* >> - * Read the common registers. >> + * Read the common registers. Note that we should try to >> + * validate whether ush would work or not. > > ush? I guess "UHS" is what you want? yes, my typo. > >> */ >> + retries = 1; > > > >> err = sdio_read_cccr(card, ocr); >> - if (err) >> - goto remove; >> + if (err) { >> + mmc_sdio_retry_init_card(host, card, &retries); >> + if (ocr & R4_18V_PRESENT) >> + goto try_again; > > Re-trying like this, would potentially cause us looping forever. > > Do you really need to retry in this case? If so, there need to be a > limit for how many times. Actually I set the retries to 1, so it won't cause us looping forever. But the issue doesn't seems so easy to be fixed as... most of the controllers have its own bit to indicate the signaling voltage likes UHS_REG for dw_mmc and HISPD for sdhci. If the card was fixed to 1V8 and mmc core could not do anything to notify this info to the host drivers due to the missing vqmmc. So the situation would be the host drivers still use 3V3 internal logic settings to communicate with the 1V8 supplied cards when initializing the cards. This is a potential risk. But I didn't have clear idea about how to fix this either. New dt property like sdio-fixed-1v8 for mmc_of_parse()? > >> + else >> + goto remove; >> + } >> >> /* >> * Read the common CIS tuples. >> -- >> 1.9.1 >> > > Kind regards > Uffe > > >
diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c index 47f824b..1d736e4 100644 --- a/drivers/mmc/core/sdio.c +++ b/drivers/mmc/core/sdio.c @@ -636,7 +636,11 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, * 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, @@ -647,9 +651,6 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, } else if (err) { ocr &= ~R4_18V_PRESENT; } - err = 0; - } else { - ocr &= ~R4_18V_PRESENT; } /* @@ -706,11 +707,18 @@ static int mmc_sdio_init_card(struct mmc_host *host, u32 ocr, } /* - * Read the common registers. + * Read the common registers. Note that we should try to + * validate whether ush would work or not. */ + retries = 1; err = sdio_read_cccr(card, ocr); - if (err) - goto remove; + if (err) { + mmc_sdio_retry_init_card(host, card, &retries); + if (ocr & R4_18V_PRESENT) + goto try_again; + else + goto remove; + } /* * Read the common CIS tuples.
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> --- drivers/mmc/core/sdio.c | 22 +++++++++++++++------- 1 file changed, 15 insertions(+), 7 deletions(-)