Skip to contentSkip to conversation

mmc: dw_mmc: Allow lower TMOUT value than maximum

PATCH Superseded Mårten Lindahl 5 participants 6 messages Nov 10, 2021 – Nov 17, 2021

Subsystems (2)

v3 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 mmc: dw_mmc: Allow lower TMOUT value than maximum

The TMOUT register is always set with a full value for every transfer, which (with a 200MHz clock) will give a full DRTO of ~84 milliseconds. Since the software dto_timer acts as a backup in cases when this timeout is not long enough, it is normally not a problem. But setting a full value makes it impossible to test shorter timeouts, when for example testing data read times on different SD cards.

Add a function to set any value smaller than the maximum of 0xFFFFFF.

---

v2:
- Calculate new value before checking boundaries
- Include CLKDIV register to get proper value

v3:
- Use 'if-else' instead of 'goto'
- Don't touch response field when maximize data field

 drivers/mmc/host/dw_mmc.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)
drivers/mmc/host/dw_mmc.c +27 -1
index 6578cc64ae9e..637ae2aea9fa 100644
@@ -1283,6 +1283,32 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
 	mci_writel(host, CTYPE, (slot->ctype << slot->id));
 }
 
+static void dw_mci_set_data_timeout(struct dw_mci *host,
+				    unsigned int timeout_ns)
+{
+	unsigned int clk_div, tmp, tmout;
+
+	clk_div = (mci_readl(host, CLKDIV) & 0xFF) * 2;
+	if (clk_div == 0)
+		clk_div = 1;
+
+	tmp = DIV_ROUND_UP_ULL((u64)timeout_ns * host->bus_hz,
+			       NSEC_PER_SEC * clk_div);
+
+	/* TMOUT[7:0] (RESPONSE_TIMEOUT) */
+	tmout = 0xFF; /* Set maximum */
+
+	/* TMOUT[31:8] (DATA_TIMEOUT) */
+	if (!tmp || tmp > 0xFFFFFF)
+		tmout |= (0xFFFFFF << 8);
+	else
+		tmout |= (tmp & 0xFFFFFF) << 8;
+
+	mci_writel(host, TMOUT, tmout);
+	dev_dbg(host->dev, "timeout_ns: %u => TMOUT[31:8]: 0x%06x",
+		timeout_ns, tmout >> 8);
+}
+
 static void __dw_mci_start_request(struct dw_mci *host,
 				   struct dw_mci_slot *slot,
 				   struct mmc_command *cmd)
@@ -1303,7 +1329,7 @@ static void __dw_mci_start_request(struct dw_mci *host,
 
 	data = cmd->data;
 	if (data) {
-		mci_writel(host, TMOUT, 0xFFFFFFFF);
+		dw_mci_set_data_timeout(host, data->timeout_ns);
 		mci_writel(host, BYTCNT, data->blksz*data->blocks);
 		mci_writel(host, BLKSIZ, data->blksz);
 	}
Signature · 1 line
2.20.1
Reviewed-by Douglas Anderson on v3 1/1
#1 lore raw

Hi,

Quoted 5 lines · Mårten Lindahl

The TMOUT register is always set with a full value for every transfer, which (with a 200MHz clock) will give a full DRTO of ~84 milliseconds. Since the software dto_timer acts as a backup in cases when this timeout is not long enough, it is normally not a problem.

I'm not 100% sure about what the part about "the software dto_timer acts as a backup in cases when this timeout is not long enough" means. At the moment the software fallback should be set to the same as the hardware value, right? So the software backup doesn't help make the timeout longer...

Quoted 42 lines

But setting a full
value makes it impossible to test shorter timeouts, when for example testing data read times on different SD cards.

Add a function to set any value smaller than the maximum of 0xFFFFFF.

---

v2:
- Calculate new value before checking boundaries
- Include CLKDIV register to get proper value

v3:
- Use 'if-else' instead of 'goto'
- Don't touch response field when maximize data field

 drivers/mmc/host/dw_mmc.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)
drivers/mmc/host/dw_mmc.c +14 -0
index 6578cc64ae9e..637ae2aea9fa 100644
@@ -1283,6 +1283,32 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
        mci_writel(host, CTYPE, (slot->ctype << slot->id));
 }
 
+static void dw_mci_set_data_timeout(struct dw_mci *host,
+                                   unsigned int timeout_ns)
+{
+       unsigned int clk_div, tmp, tmout;
+
+       clk_div = (mci_readl(host, CLKDIV) & 0xFF) * 2;
+       if (clk_div == 0)
+               clk_div = 1;
+
+       tmp = DIV_ROUND_UP_ULL((u64)timeout_ns * host->bus_hz,
+                              NSEC_PER_SEC * clk_div);
+
+       /* TMOUT[7:0] (RESPONSE_TIMEOUT) */
+       tmout = 0xFF; /* Set maximum */

drivers/mmc/host/dw_mmc.c static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit) · line 1309

Quoted 10 lines
drivers/mmc/host/dw_mmc.c +10 -0
+
+       /* TMOUT[31:8] (DATA_TIMEOUT) */
+       if (!tmp || tmp > 0xFFFFFF)
+               tmout |= (0xFFFFFF << 8);
+       else
+               tmout |= (tmp & 0xFFFFFF) << 8;
+
+       mci_writel(host, TMOUT, tmout);
+       dev_dbg(host->dev, "timeout_ns: %u => TMOUT[31:8]: 0x%06x",
+               timeout_ns, tmout >> 8);

nit: In theory 0x%06x is slightly better written as %#08x. I wouldn't spin the patch just for that, though.

In any case, I think this is fine to the best of my knowledge of how this timeout is supposed to work, so I'm happy with:

maintainer Reviewed-by on v3 1/1
#2 lore raw
In reply to Mårten Lindahl

Hi Marten,

Quoted 10 lines · Mårten Lindahl

The TMOUT register is always set with a full value for every transfer, which (with a 200MHz clock) will give a full DRTO of ~84 milliseconds. Since the software dto_timer acts as a backup in cases when this timeout is not long enough, it is normally not a problem. But setting a full value makes it impossible to test shorter timeouts, when for example testing data read times on different SD cards.

Add a function to set any value smaller than the maximum of 0xFFFFFF.

Best Regards,
Jaehoon Chung

Quoted 60 lines

---

v2:
- Calculate new value before checking boundaries
- Include CLKDIV register to get proper value

v3:
- Use 'if-else' instead of 'goto'
- Don't touch response field when maximize data field

 drivers/mmc/host/dw_mmc.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)
drivers/mmc/host/dw_mmc.c +27 -1
index 6578cc64ae9e..637ae2aea9fa 100644
@@ -1283,6 +1283,32 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
 	mci_writel(host, CTYPE, (slot->ctype << slot->id));
 }
 
+static void dw_mci_set_data_timeout(struct dw_mci *host,
+				    unsigned int timeout_ns)
+{
+	unsigned int clk_div, tmp, tmout;
+
+	clk_div = (mci_readl(host, CLKDIV) & 0xFF) * 2;
+	if (clk_div == 0)
+		clk_div = 1;
+
+	tmp = DIV_ROUND_UP_ULL((u64)timeout_ns * host->bus_hz,
+			       NSEC_PER_SEC * clk_div);
+
+	/* TMOUT[7:0] (RESPONSE_TIMEOUT) */
+	tmout = 0xFF; /* Set maximum */
+
+	/* TMOUT[31:8] (DATA_TIMEOUT) */
+	if (!tmp || tmp > 0xFFFFFF)
+		tmout |= (0xFFFFFF << 8);
+	else
+		tmout |= (tmp & 0xFFFFFF) << 8;
+
+	mci_writel(host, TMOUT, tmout);
+	dev_dbg(host->dev, "timeout_ns: %u => TMOUT[31:8]: 0x%06x",
+		timeout_ns, tmout >> 8);
+}
+
 static void __dw_mci_start_request(struct dw_mci *host,
 				   struct dw_mci_slot *slot,
 				   struct mmc_command *cmd)
@@ -1303,7 +1329,7 @@ static void __dw_mci_start_request(struct dw_mci *host,
 
 	data = cmd->data;
 	if (data) {
-		mci_writel(host, TMOUT, 0xFFFFFFFF);
+		dw_mci_set_data_timeout(host, data->timeout_ns);
 		mci_writel(host, BYTCNT, data->blksz*data->blocks);
 		mci_writel(host, BLKSIZ, data->blksz);
 	}
#3 lore raw
In reply to Mårten Lindahl
Quoted 11 lines · Mårten Lindahl

The TMOUT register is always set with a full value for every transfer, which (with a 200MHz clock) will give a full DRTO of ~84 milliseconds. Since the software dto_timer acts as a backup in cases when this timeout is not long enough, it is normally not a problem. But setting a full value makes it impossible to test shorter timeouts, when for example testing data read times on different SD cards.

Add a function to set any value smaller than the maximum of 0xFFFFFF.

Applied for next, thanks!

Kind regards
Uffe

Quoted 62 lines

---

v2:
- Calculate new value before checking boundaries
- Include CLKDIV register to get proper value

v3:
- Use 'if-else' instead of 'goto'
- Don't touch response field when maximize data field

 drivers/mmc/host/dw_mmc.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)
drivers/mmc/host/dw_mmc.c +27 -2
index 6578cc64ae9e..637ae2aea9fa 100644
@@ -1283,6 +1283,32 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
        mci_writel(host, CTYPE, (slot->ctype << slot->id));
 }
 
+static void dw_mci_set_data_timeout(struct dw_mci *host,
+                                   unsigned int timeout_ns)
+{
+       unsigned int clk_div, tmp, tmout;
+
+       clk_div = (mci_readl(host, CLKDIV) & 0xFF) * 2;
+       if (clk_div == 0)
+               clk_div = 1;
+
+       tmp = DIV_ROUND_UP_ULL((u64)timeout_ns * host->bus_hz,
+                              NSEC_PER_SEC * clk_div);
+
+       /* TMOUT[7:0] (RESPONSE_TIMEOUT) */
+       tmout = 0xFF; /* Set maximum */
+
+       /* TMOUT[31:8] (DATA_TIMEOUT) */
+       if (!tmp || tmp > 0xFFFFFF)
+               tmout |= (0xFFFFFF << 8);
+       else
+               tmout |= (tmp & 0xFFFFFF) << 8;
+
+       mci_writel(host, TMOUT, tmout);
+       dev_dbg(host->dev, "timeout_ns: %u => TMOUT[31:8]: 0x%06x",
+               timeout_ns, tmout >> 8);
+}
+
 static void __dw_mci_start_request(struct dw_mci *host,
                                   struct dw_mci_slot *slot,
                                   struct mmc_command *cmd)
@@ -1303,7 +1329,7 @@ static void __dw_mci_start_request(struct dw_mci *host,
 
        data = cmd->data;
        if (data) {
-               mci_writel(host, TMOUT, 0xFFFFFFFF);
+               dw_mci_set_data_timeout(host, data->timeout_ns);
                mci_writel(host, BYTCNT, data->blksz*data->blocks);
                mci_writel(host, BLKSIZ, data->blksz);
        }
--

2.20.1

#4 lore raw
Quoted 9 lines · Ulf Hansson
Quoted 8 lines · Mårten Lindahl

The TMOUT register is always set with a full value for every transfer, which (with a 200MHz clock) will give a full DRTO of ~84 milliseconds. Since the software dto_timer acts as a backup in cases when this timeout is not long enough, it is normally not a problem. But setting a full value makes it impossible to test shorter timeouts, when for example testing data read times on different SD cards.

Quoted 9 lines
Quoted 3 lines

Add a function to set any value smaller than the maximum of 0xFFFFFF.

Applied for next, thanks!

Kind regards
Uffe

Hi Ulf!

During tests on some units I found that there is a potential overflow of the 32bit divider in this patch (if clk_div > 4).

Is it possible to stop this patch, until I created a new version of it? I guess it's better to update this patch, in favour of sending a new correction patch for it?

Kind regards
Mårten

Quoted 63 lines
Quoted 63 lines

---

v2:
- Calculate new value before checking boundaries
- Include CLKDIV register to get proper value

v3:
- Use 'if-else' instead of 'goto'
- Don't touch response field when maximize data field

 drivers/mmc/host/dw_mmc.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)
drivers/mmc/host/dw_mmc.c +27 -2
index 6578cc64ae9e..637ae2aea9fa 100644
@@ -1283,6 +1283,32 @@ static void dw_mci_setup_bus(struct dw_mci_slot *slot, bool force_clkinit)
        mci_writel(host, CTYPE, (slot->ctype << slot->id));
 }
 
+static void dw_mci_set_data_timeout(struct dw_mci *host,
+                                   unsigned int timeout_ns)
+{
+       unsigned int clk_div, tmp, tmout;
+
+       clk_div = (mci_readl(host, CLKDIV) & 0xFF) * 2;
+       if (clk_div == 0)
+               clk_div = 1;
+
+       tmp = DIV_ROUND_UP_ULL((u64)timeout_ns * host->bus_hz,
+                              NSEC_PER_SEC * clk_div);
+
+       /* TMOUT[7:0] (RESPONSE_TIMEOUT) */
+       tmout = 0xFF; /* Set maximum */
+
+       /* TMOUT[31:8] (DATA_TIMEOUT) */
+       if (!tmp || tmp > 0xFFFFFF)
+               tmout |= (0xFFFFFF << 8);
+       else
+               tmout |= (tmp & 0xFFFFFF) << 8;
+
+       mci_writel(host, TMOUT, tmout);
+       dev_dbg(host->dev, "timeout_ns: %u => TMOUT[31:8]: 0x%06x",
+               timeout_ns, tmout >> 8);
+}
+
 static void __dw_mci_start_request(struct dw_mci *host,
                                   struct dw_mci_slot *slot,
                                   struct mmc_command *cmd)
@@ -1303,7 +1329,7 @@ static void __dw_mci_start_request(struct dw_mci *host,
 
        data = cmd->data;
        if (data) {
-               mci_writel(host, TMOUT, 0xFFFFFFFF);
+               dw_mci_set_data_timeout(host, data->timeout_ns);
                mci_writel(host, BYTCNT, data->blksz*data->blocks);
                mci_writel(host, BLKSIZ, data->blksz);
        }
--

2.20.1

#5 lore raw
Quoted 21 lines · Marten Lindahl
Quoted 9 lines · Ulf Hansson
Quoted 8 lines · Mårten Lindahl

The TMOUT register is always set with a full value for every transfer, which (with a 200MHz clock) will give a full DRTO of ~84 milliseconds. Since the software dto_timer acts as a backup in cases when this timeout is not long enough, it is normally not a problem. But setting a full value makes it impossible to test shorter timeouts, when for example testing data read times on different SD cards.

Quoted 9 lines
Quoted 3 lines

Add a function to set any value smaller than the maximum of 0xFFFFFF.

Applied for next, thanks!

Kind regards
Uffe

Quoted 8 lines

Hi Ulf!

During tests on some units I found that there is a potential overflow of the 32bit divider in this patch (if clk_div > 4).

Is it possible to stop this patch, until I created a new version of it? I guess it's better to update this patch, in favour of sending a new correction patch for it?

I have dropped the patch from my next branch. Please send a v4.

[...]

Kind regards
Uffe