[U-Boot] [PATCH] mmc: fix OCR Polling

If in mmc_send_op_cond, OCR_BUSY is set in CMD1's response, then state is transfered to Ready state, and there is no need to send CMD1 again. Otherwise following CMD1 will recieve no response, or timeour error from driver such as fsl_esdhc.c.
If not into Ready state in previous CMD1, then continue CMD1 command.
In mmc_complete_op_cond, we use the value mmc->op_cond_response from mmc_send_op_cond, since there should be no CMD1 command between mmc_send_op_cond and mmc_complete_op_cond
Before fixing this, uboot log shows: " CMD_SEND:0 ARG 0x00000000 MMC_RSP_NONE CMD_SEND:8 ARG 0x000001AA MMC_RSP_R1,5,6,7 0x18EC1504 CMD_SEND:55 ARG 0x00000000 MMC_RSP_R1,5,6,7 0x18EC1504 CMD_SEND:0 ARG 0x00000000 MMC_RSP_NONE CMD_SEND:1 ARG 0x00000000 MMC_RSP_R3,4 0x00FF8080 CMD_SEND:1 ARG 0x40300000 MMC_RSP_R3,4 0xC0FF8080 --> Already OCR_BUSY set CMD_SEND:1 ARG 0x40300000 MMC_RSP_R3,4 0x0096850A --> Failed CMD1 MMC init failed "
Using this patch, this issue is fixed, emmc can be detected correctly.
Signed-off-by: Peng Fan Peng.Fan@freescale.com --- drivers/mmc/mmc.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/drivers/mmc/mmc.c b/drivers/mmc/mmc.c index a13769e..43a9a8a 100644 --- a/drivers/mmc/mmc.c +++ b/drivers/mmc/mmc.c @@ -406,14 +406,23 @@ static int mmc_complete_op_cond(struct mmc *mmc)
mmc->op_cond_pending = 0; start = get_timer(0); - do { + /* + * If in mmc_send_op_cond, OCR_BUSY is set in CMD1's response, then + * state is transfered to Ready state, and there is no need to + * send CMD1 again. Otherwise following CMD1 will recieve no response, + * or timeour error from driver such as fsl_esdhc.c. + * + * If not into Ready state in previous CMD1, then continue CMD1 + * command. + */ + while (!(mmc->op_cond_response & OCR_BUSY)) { err = mmc_send_op_cond_iter(mmc, &cmd, 1); if (err) return err; if (get_timer(start) > timeout) return UNUSABLE_ERR; udelay(100); - } while (!(mmc->op_cond_response & OCR_BUSY)); + }
if (mmc_host_is_spi(mmc)) { /* read OCR for spi */ cmd.cmdidx = MMC_CMD_SPI_READ_OCR;

Hi Peng,
On Mar 19, 2015, at 10:22 , Peng Fan Peng.Fan@freescale.com wrote:
If in mmc_send_op_cond, OCR_BUSY is set in CMD1's response, then state is transfered to Ready state, and there is no need to send CMD1 again. Otherwise following CMD1 will recieve no response, or timeour error from driver such as fsl_esdhc.c.
If not into Ready state in previous CMD1, then continue CMD1 command.
In mmc_complete_op_cond, we use the value mmc->op_cond_response from mmc_send_op_cond, since there should be no CMD1 command between mmc_send_op_cond and mmc_complete_op_cond
Before fixing this, uboot log shows: " CMD_SEND:0 ARG 0x00000000 MMC_RSP_NONE CMD_SEND:8 ARG 0x000001AA MMC_RSP_R1,5,6,7 0x18EC1504 CMD_SEND:55 ARG 0x00000000 MMC_RSP_R1,5,6,7 0x18EC1504 CMD_SEND:0 ARG 0x00000000 MMC_RSP_NONE CMD_SEND:1 ARG 0x00000000 MMC_RSP_R3,4 0x00FF8080 CMD_SEND:1 ARG 0x40300000 MMC_RSP_R3,4 0xC0FF8080 --> Already OCR_BUSY set CMD_SEND:1 ARG 0x40300000 MMC_RSP_R3,4 0x0096850A --> Failed CMD1 MMC init failed "
Using this patch, this issue is fixed, emmc can be detected correctly.
Signed-off-by: Peng Fan Peng.Fan@freescale.com
drivers/mmc/mmc.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-)
Thanks for the patch, but Andrew’s patchset is more complete.
I’ll pick up his fix instead.
Regards
— Pantelis
participants (2)
-
Pantelis Antoniou
-
Peng Fan