| Message ID | 1303235057-4745-1-git-send-email-broonie@opensource.wolfsonmicro.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Tue, Apr 19, 2011 at 12:44 PM, Mark Brown <broonie@opensource.wolfsonmicro.com> wrote: > Commit 373e6a (mmc: sdhci: R1B command handling + MMC_CAP_ERASE) moved the > handling of SDHCI_QUIRK_TIMEOUT_USES_SDCLK from sdhci_calc_timeout() to > sdhci_add_host(). This causes division by zero errors on at least the S3C > SDHCI controller as the quirk implementation needs host->clock set to work > but host->clock has not been set when sdhci_add_host() is called. > > Fix this by backing out that portion of the change, the clock may vary at > runtime anyway. It does occur to me that we may want to move the quirk to > where we set the clock but this seems more invasive and I'm concerned > about undesirable side effects. > > Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> > --- > drivers/mmc/host/sdhci.c | 7 ++++--- > 1 files changed, 4 insertions(+), 3 deletions(-) > > diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c > index f79dead..da57bdb 100644 > --- a/drivers/mmc/host/sdhci.c > +++ b/drivers/mmc/host/sdhci.c > @@ -616,6 +616,9 @@ static u8 sdhci_calc_timeout(struct sdhci_host *host, struct mmc_command *cmd) > target_timeout = data->timeout_ns / 1000 + > data->timeout_clks / host->clock; > > + if (host->quirks & SDHCI_QUIRK_DATA_TIMEOUT_USES_SDCLK) > + host->timeout_clk = host->clock / 1000; > + > /* > * Figure out needed cycles. > * We do this in steps in order to fit inside a 32 bit int. > @@ -626,6 +629,7 @@ static u8 sdhci_calc_timeout(struct sdhci_host *host, struct mmc_command *cmd) > * => > * (1) / (2) > 2^6 > */ > + BUG_ON(!host->timeout_clk); > count = 0; > current_timeout = (1 << 13) * 1000 / host->timeout_clk; > while (current_timeout < target_timeout) { > @@ -1894,9 +1898,6 @@ int sdhci_add_host(struct sdhci_host *host) > if (caps & SDHCI_TIMEOUT_CLK_UNIT) > host->timeout_clk *= 1000; > > - if (host->quirks & SDHCI_QUIRK_DATA_TIMEOUT_USES_SDCLK) > - host->timeout_clk = host->clock / 1000; > - > /* > * Set host parameters. > */ > -- > 1.7.4.1 > > Sorry about that. I thought I had the right idea. I'll try to be more careful. A -- 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
Hi Mark, On Tue, Apr 19 2011, Mark Brown wrote: > Commit 373e6a (mmc: sdhci: R1B command handling + MMC_CAP_ERASE) moved the > handling of SDHCI_QUIRK_TIMEOUT_USES_SDCLK from sdhci_calc_timeout() to > sdhci_add_host(). This causes division by zero errors on at least the S3C > SDHCI controller as the quirk implementation needs host->clock set to work > but host->clock has not been set when sdhci_add_host() is called. > > Fix this by backing out that portion of the change, the clock may vary at > runtime anyway. It does occur to me that we may want to move the quirk to > where we set the clock but this seems more invasive and I'm concerned > about undesirable side effects. > > Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Thanks for catching this, pushed to mmc-next for .40. - Chris.
diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c index f79dead..da57bdb 100644 --- a/drivers/mmc/host/sdhci.c +++ b/drivers/mmc/host/sdhci.c @@ -616,6 +616,9 @@ static u8 sdhci_calc_timeout(struct sdhci_host *host, struct mmc_command *cmd) target_timeout = data->timeout_ns / 1000 + data->timeout_clks / host->clock; + if (host->quirks & SDHCI_QUIRK_DATA_TIMEOUT_USES_SDCLK) + host->timeout_clk = host->clock / 1000; + /* * Figure out needed cycles. * We do this in steps in order to fit inside a 32 bit int. @@ -626,6 +629,7 @@ static u8 sdhci_calc_timeout(struct sdhci_host *host, struct mmc_command *cmd) * => * (1) / (2) > 2^6 */ + BUG_ON(!host->timeout_clk); count = 0; current_timeout = (1 << 13) * 1000 / host->timeout_clk; while (current_timeout < target_timeout) { @@ -1894,9 +1898,6 @@ int sdhci_add_host(struct sdhci_host *host) if (caps & SDHCI_TIMEOUT_CLK_UNIT) host->timeout_clk *= 1000; - if (host->quirks & SDHCI_QUIRK_DATA_TIMEOUT_USES_SDCLK) - host->timeout_clk = host->clock / 1000; - /* * Set host parameters. */
Commit 373e6a (mmc: sdhci: R1B command handling + MMC_CAP_ERASE) moved the handling of SDHCI_QUIRK_TIMEOUT_USES_SDCLK from sdhci_calc_timeout() to sdhci_add_host(). This causes division by zero errors on at least the S3C SDHCI controller as the quirk implementation needs host->clock set to work but host->clock has not been set when sdhci_add_host() is called. Fix this by backing out that portion of the change, the clock may vary at runtime anyway. It does occur to me that we may want to move the quirk to where we set the clock but this seems more invasive and I'm concerned about undesirable side effects. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> --- drivers/mmc/host/sdhci.c | 7 ++++--- 1 files changed, 4 insertions(+), 3 deletions(-)