Skip to contentSkip to conversation

MMC: DWMMC: Fix FIFO_DEPTH calculation

PATCH New Rajeshwari Shinde 5 participants 5 messages May 24, 2013 – Jun 17, 2013 MMC, SD and SDIO

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 MMC: DWMMC: Fix FIFO_DEPTH calculation

Current DWMMC driver used to give FIFO underrun/overrun error every 3rd time for mmc rescan command.
In current code FIFO_DEPTH is getting calculated after reading the default FIFOTH register and extracting the RX_WMARK bits from it i.e (RX_WMARK = FIFO_DEPTH/2 -1). Instead of storing the correct value, we were recalculating the FIFO_DEPT each time which is not correct.

Based on "[PATCH V9 3/9] DWMMC: Initialise dwmci and resolve EMMC read write issues"
http://permalink.gmane.org/gmane.comp.boot-loaders.u-boot/160247

---
Changes in V2:
- remove fifoth_val local variable
- Rebased on top of "[PATCH V9 3/9] DWMMC: Initialise dwmci and resolve EMMC read write issues".

 drivers/mmc/dw_mmc.c |   14 ++++++--------
 1 files changed, 6 insertions(+), 8 deletions(-)
drivers/mmc/dw_mmc.c +6 -8
index 1307b12..5da20ed 100644
@@ -312,7 +312,7 @@ static void dwmci_set_ios(struct mmc *mmc)
 static int dwmci_init(struct mmc *mmc)
 {
 	struct dwmci_host *host = (struct dwmci_host *)mmc->priv;
-	u32 fifo_size, fifoth_val;
+	u32 fifo_size;
 
 	dwmci_writel(host, DWMCI_PWREN, 1);
 
@@ -332,15 +332,13 @@ static int dwmci_init(struct mmc *mmc)
 	dwmci_writel(host, DWMCI_IDINTEN, 0);
 	dwmci_writel(host, DWMCI_BMOD, 1);
 
-	fifo_size = dwmci_readl(host, DWMCI_FIFOTH);
-	fifo_size = ((fifo_size & RX_WMARK_MASK) >> RX_WMARK_SHIFT) + 1;
-	if (host->fifoth_val) {
-		fifoth_val = host->fifoth_val;
-	} else {
-		fifoth_val = MSIZE(0x2) | RX_WMARK(fifo_size / 2 - 1) |
+	if (!host->fifoth_val) {
+		fifo_size = dwmci_readl(host, DWMCI_FIFOTH);
+		fifo_size = ((fifo_size & RX_WMARK_MASK) >> RX_WMARK_SHIFT) + 1;
+		host->fifoth_val = MSIZE(0x2) | RX_WMARK(fifo_size / 2 - 1) |
 			TX_WMARK(fifo_size / 2);
 	}
-	dwmci_writel(host, DWMCI_FIFOTH, fifoth_val);
+	dwmci_writel(host, DWMCI_FIFOTH, host->fifoth_val);
 
 	dwmci_writel(host, DWMCI_CLKENA, 0);
 	dwmci_writel(host, DWMCI_CLKSRC, 0);
Signature · 1 line
1.7.4.4
#1 lore raw

PATCH v2 1/1 MMC: DWMMC: Fix FIFO_DEPTH calculation

On Fri, May 24, 2013 at 5:45 AM, Rajeshwari Shinde <[email protected]

Quoted 12 lines

wrote:

Current DWMMC driver used to give FIFO underrun/overrun error every 3rd time
for mmc rescan command.
In current code FIFO_DEPTH is getting calculated after reading the default FIFOTH
register and extracting the RX_WMARK bits from it i.e (RX_WMARK = FIFO_DEPTH/2 -1).
Instead of storing the correct value, we were recalculating the FIFO_DEPT each

DEPTH

Quoted 9 lines

time which is not correct.

Based on "[PATCH V9 3/9] DWMMC: Initialise dwmci and resolve EMMC read write issues"
http://permalink.gmane.org/gmane.comp.boot-loaders.u-boot/160247

Tested on pit using 'mmc rescan' and saw no problems now.

Quoted 49 lines

---
Changes in V2:
- remove fifoth_val local variable
- Rebased on top of "[PATCH V9 3/9] DWMMC: Initialise dwmci and
resolve
EMMC read write issues".

 drivers/mmc/dw_mmc.c |   14 ++++++--------
 1 files changed, 6 insertions(+), 8 deletions(-)
drivers/mmc/dw_mmc.c +5 -7
index 1307b12..5da20ed 100644
@@ -312,7 +312,7 @@ static void dwmci_set_ios(struct mmc *mmc)
 static int dwmci_init(struct mmc *mmc)
 {
        struct dwmci_host *host = (struct dwmci_host *)mmc->priv;
-       u32 fifo_size, fifoth_val;
+       u32 fifo_size;
 
        dwmci_writel(host, DWMCI_PWREN, 1);
 
@@ -332,15 +332,13 @@ static int dwmci_init(struct mmc *mmc)
        dwmci_writel(host, DWMCI_IDINTEN, 0);
        dwmci_writel(host, DWMCI_BMOD, 1);
 
-       fifo_size = dwmci_readl(host, DWMCI_FIFOTH);
-       fifo_size = ((fifo_size & RX_WMARK_MASK) >> RX_WMARK_SHIFT) + 1;
-       if (host->fifoth_val) {
-               fifoth_val = host->fifoth_val;
-       } else {
-               fifoth_val = MSIZE(0x2) | RX_WMARK(fifo_size / 2 - 1) |
+       if (!host->fifoth_val) {
+               fifo_size = dwmci_readl(host, DWMCI_FIFOTH);
+               fifo_size = ((fifo_size & RX_WMARK_MASK) >>
 RX_WMARK_SHIFT) + 1;
+               host->fifoth_val = MSIZE(0x2) | RX_WMARK(fifo_size / 2 -

1) |

                        TX_WMARK(fifo_size / 2);
        }
drivers/mmc/dw_mmc.c +1 -2
-       dwmci_writel(host, DWMCI_FIFOTH, fifoth_val);
+       dwmci_writel(host, DWMCI_FIFOTH, host->fifoth_val);
 
        dwmci_writel(host, DWMCI_CLKENA, 0);
        dwmci_writel(host, DWMCI_CLKSRC, 0);
--

1.7.4.4

#2 lore raw

PATCH v2 1/1 MMC: DWMMC: Fix FIFO_DEPTH calculation

Thanks,
Jaehoon Chung

Quoted 74 lines · Simon Glass

On Fri, May 24, 2013 at 5:45 AM, Rajeshwari Shinde <rajeshwari.s at samsung.com <mailto:[email protected]>> wrote:

Current DWMMC driver used to give FIFO underrun/overrun error every 3rd time for mmc rescan command.
In current code FIFO_DEPTH is getting calculated after reading the default FIFOTH register and extracting the RX_WMARK bits from it i.e (RX_WMARK = FIFO_DEPTH/2 -1). Instead of storing the correct value, we were recalculating the FIFO_DEPT each

DEPTH

time which is not correct.

Based on "[PATCH V9 3/9] DWMMC: Initialise dwmci and resolve EMMC read write issues"
http://permalink.gmane.org/gmane.comp.boot-loaders.u-boot/160247

Signed-off-by Hatim Ali <hatim.rv at samsung.com <mailto:[email protected]>>
Signed-off-by Rajeshwari Shinde <rajeshwari.s at samsung.com <mailto:[email protected]>>
Acked-by Simon Glass <sjg at chromium.org <mailto:[email protected]>>

Tested on pit using 'mmc rescan' and saw no problems now.

Tested-by Simon Glass <sjg at chromium.org <mailto:[email protected]>>

---
Changes in V2:
- remove fifoth_val local variable
- Rebased on top of "[PATCH V9 3/9] DWMMC: Initialise dwmci and resolve EMMC read write issues".
drivers/mmc/dw_mmc.c | 14 ++++++--------
1 files changed, 6 insertions(+), 8 deletions(-)

diff --git a/drivers/mmc/dw_mmc.c b/drivers/mmc/dw_mmc.c index 1307b12..5da20ed 100644
--- a/drivers/mmc/dw_mmc.c
+++ b/drivers/mmc/dw_mmc.c
@@ -312,7 +312,7 @@ static void dwmci_set_ios(struct mmc *mmc)
static int dwmci_init(struct mmc *mmc)
{
struct dwmci_host *host = (struct dwmci_host *)mmc->priv;
- u32 fifo_size, fifoth_val;
+ u32 fifo_size;

dwmci_writel(host, DWMCI_PWREN, 1);

@@ -332,15 +332,13 @@ static int dwmci_init(struct mmc *mmc)
dwmci_writel(host, DWMCI_IDINTEN, 0); dwmci_writel(host, DWMCI_BMOD, 1);

- fifo_size = dwmci_readl(host, DWMCI_FIFOTH);
- fifo_size = ((fifo_size & RX_WMARK_MASK) >> RX_WMARK_SHIFT) + 1;
- if (host->fifoth_val) {
- fifoth_val = host->fifoth_val;
- } else {
- fifoth_val = MSIZE(0x2) | RX_WMARK(fifo_size / 2 - 1) |
+ if (!host->fifoth_val) {
+ fifo_size = dwmci_readl(host, DWMCI_FIFOTH);
+ fifo_size = ((fifo_size & RX_WMARK_MASK) >> RX_WMARK_SHIFT) + 1;
+ host->fifoth_val = MSIZE(0x2) | RX_WMARK(fifo_size / 2 - 1) |
TX_WMARK(fifo_size / 2);
}
- dwmci_writel(host, DWMCI_FIFOTH, fifoth_val);
+ dwmci_writel(host, DWMCI_FIFOTH, host->fifoth_val);

dwmci_writel(host, DWMCI_CLKENA, 0);
dwmci_writel(host, DWMCI_CLKSRC, 0);
--
1.7.4.4

#3 lore raw

PATCH v2 1/1 MMC: DWMMC: Fix FIFO_DEPTH calculation

In reply to Rajeshwari Shinde
Quoted 7 lines · Rajeshwari Shinde

Current DWMMC driver used to give FIFO underrun/overrun error every 3rd time for mmc rescan command.
In current code FIFO_DEPTH is getting calculated after reading the default FIFOTH register and extracting the RX_WMARK bits from it i.e (RX_WMARK = FIFO_DEPTH/2 -1). Instead of storing the correct value, we were recalculating the FIFO_DEPT each time which is not correct.

Quoted 8 lines

Based on "[PATCH V9 3/9] DWMMC: Initialise dwmci and resolve EMMC read write issues"
http://permalink.gmane.org/gmane.comp.boot-loaders.u-boot/160247

Looks like the patch this is based on is already accepted, but not yet upstream, so I'm going to delegate the patch to Minkyu Kang

#4 lore raw

PATCH v2 1/1 MMC: DWMMC: Fix FIFO_DEPTH calculation

Quoted 17 lines · Andy Fleming
Quoted 7 lines · Rajeshwari Shinde

Current DWMMC driver used to give FIFO underrun/overrun error every 3rd time for mmc rescan command.
In current code FIFO_DEPTH is getting calculated after reading the default FIFOTH register and extracting the RX_WMARK bits from it i.e (RX_WMARK = FIFO_DEPTH/2 -1). Instead of storing the correct value, we were recalculating the FIFO_DEPT each time which is not correct.

Quoted 8 lines

Based on "[PATCH V9 3/9] DWMMC: Initialise dwmci and resolve EMMC read write issues"
http://permalink.gmane.org/gmane.comp.boot-loaders.u-boot/160247

Quoted 7 lines

Looks like the patch this is based on is already accepted, but not yet upstream, so I'm going to delegate the patch to Minkyu Kang

applied to u-boot-samsung.

Thanks,
Minkyu Kang.