diff options
| author | Andy Green <andy@openmoko.com> | 2009-01-04 17:06:03 +0000 |
|---|---|---|
| committer | Andy Green <agreen@pads.home.warmcat.com> | 2009-01-04 17:06:03 +0000 |
| commit | bcd4eff43a16ed339d4b17141d8828e9228b2d52 (patch) | |
| tree | 5d58d531a98d824a884d91fdb77406067c7cb39a | |
| parent | 8aaff459acdad4915bcdd8475d4314499980e0ac (diff) | |
fix-glamo-mci-ambiguous-timeout.patch
Sometimes we see failures with cards where the status is eg, 0x310.
This corresponds to a "no data" timeout and an assertion that the
data is present. This patch makes the data present status have
priority over the timeout status.
Signed-off-by: Andy Green <andy@openmoko.com>
| -rw-r--r-- | drivers/mfd/glamo/glamo-mci.c | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/drivers/mfd/glamo/glamo-mci.c b/drivers/mfd/glamo/glamo-mci.c index 3e3ffeaff71..778e47c7427 100644 --- a/drivers/mfd/glamo/glamo-mci.c +++ b/drivers/mfd/glamo/glamo-mci.c @@ -361,6 +361,10 @@ static void glamo_mci_irq(unsigned int irq, struct irq_desc *desc) writew(GLAMO_IRQ_MMC, glamo_mci_def_pdata.pglamo->base + GLAMO_REG_IRQ_CLEAR); + /* we ignore a data timeout report if we are also told the data came */ + if (status & GLAMO_STAT1_MMC_RB_DRDY) + status &= ~GLAMO_STAT1_MMC_DTOUT; + if (status & (GLAMO_STAT1_MMC_RTOUT | GLAMO_STAT1_MMC_DTOUT)) cmd->error = -ETIMEDOUT; |