[PATCH v3 0/5] davinci_mmc fixes

[PATCH v3 0/5] davinci_mmc fixes

[David Lechner]

I have replaced one of my patches with the patch from Peter Ujfalusi and fixed
a mistake in the "fix unwinding..." patch.

Tested working on LEGO MINDSTORMS EV3 (da850-ish).

David Lechner (4):
  mmc: davinci: remove matching string
  mmc: davinci: fix unwinding in probe
  mmc: davinci: prepare clock
  ARM: davinci: remove mmc dma resources

Peter Ujfalusi (1):
  mmc: davinci_mmc: Use dma_request_chan() to requesting DMA channel

 arch/arm/mach-davinci/devices-da8xx.c |  20 -----
 arch/arm/mach-davinci/devices.c       |  16 ----
 drivers/mmc/host/davinci_mmc.c        | 147 ++++++++++++----------------------
 3 files changed, 50 insertions(+), 133 deletions(-)

-- 
1.9.1

[PATCH v3 1/5] mmc: davinci_mmc: Use dma_request_chan() to requesting DMA channel

[David Lechner]

From: Peter Ujfalusi <peter.ujfalusi@ti.com>

With the new dma_request_chan() the client driver does not need to look for
the DMA resource and it does not need to pass filter_fn anymore.
By switching to the new API the driver can now support deferred probing
against DMA.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
---

v3 changes: This is a new patch in v3 (from Peter Ujfalusi) and replaces my
previous "mmc: davinci: don't use dma platform resources" patch.


 drivers/mmc/host/davinci_mmc.c | 53 +++++++++++-------------------------------
 1 file changed, 14 insertions(+), 39 deletions(-)

diff --git a/drivers/mmc/host/davinci_mmc.c b/drivers/mmc/host/davinci_mmc.c
index 693144e..eb3ca9e 100644
--- a/drivers/mmc/host/davinci_mmc.c
+++ b/drivers/mmc/host/davinci_mmc.c
@@ -32,12 +32,10 @@
 #include <linux/delay.h>
 #include <linux/dmaengine.h>
 #include <linux/dma-mapping.h>
-#include <linux/edma.h>
 #include <linux/mmc/mmc.h>
 #include <linux/of.h>
 #include <linux/of_device.h>
 
-#include <linux/platform_data/edma.h>
 #include <linux/platform_data/mmc-davinci.h>
 
 /*
@@ -202,7 +200,6 @@ struct mmc_davinci_host {
 	u32 buffer_bytes_left;
 	u32 bytes_left;
 
-	u32 rxdma, txdma;
 	struct dma_chan *dma_tx;
 	struct dma_chan *dma_rx;
 	bool use_dma;
@@ -513,35 +510,20 @@ davinci_release_dma_channels(struct mmc_davinci_host *host)
 
 static int __init davinci_acquire_dma_channels(struct mmc_davinci_host *host)
 {
-	int r;
-	dma_cap_mask_t mask;
-
-	dma_cap_zero(mask);
-	dma_cap_set(DMA_SLAVE, mask);
-
-	host->dma_tx =
-		dma_request_slave_channel_compat(mask, edma_filter_fn,
-				&host->txdma, mmc_dev(host->mmc), "tx");
-	if (!host->dma_tx) {
+	host->dma_tx = dma_request_chan(mmc_dev(host->mmc), "tx");
+	if (IS_ERR(host->dma_tx)) {
 		dev_err(mmc_dev(host->mmc), "Can't get dma_tx channel\n");
-		return -ENODEV;
+		return PTR_ERR(host->dma_tx);
 	}
 
-	host->dma_rx =
-		dma_request_slave_channel_compat(mask, edma_filter_fn,
-				&host->rxdma, mmc_dev(host->mmc), "rx");
-	if (!host->dma_rx) {
+	host->dma_rx = dma_request_chan(mmc_dev(host->mmc), "rx");
+	if (IS_ERR(host->dma_rx)) {
 		dev_err(mmc_dev(host->mmc), "Can't get dma_rx channel\n");
-		r = -ENODEV;
-		goto free_master_write;
+		dma_release_channel(host->dma_tx);
+		return PTR_ERR(host->dma_rx);
 	}
 
 	return 0;
-
-free_master_write:
-	dma_release_channel(host->dma_tx);
-
-	return r;
 }
 
 /*----------------------------------------------------------------------*/
@@ -1253,18 +1235,6 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 	host = mmc_priv(mmc);
 	host->mmc = mmc;	/* Important */
 
-	r = platform_get_resource(pdev, IORESOURCE_DMA, 0);
-	if (!r)
-		dev_warn(&pdev->dev, "RX DMA resource not specified\n");
-	else
-		host->rxdma = r->start;
-
-	r = platform_get_resource(pdev, IORESOURCE_DMA, 1);
-	if (!r)
-		dev_warn(&pdev->dev, "TX DMA resource not specified\n");
-	else
-		host->txdma = r->start;
-
 	host->mem_res = mem;
 	host->base = ioremap(mem->start, mem_size);
 	if (!host->base)
@@ -1291,8 +1261,13 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 	host->mmc_irq = irq;
 	host->sdio_irq = platform_get_irq(pdev, 1);
 
-	if (host->use_dma && davinci_acquire_dma_channels(host) != 0)
-		host->use_dma = 0;
+	if (host->use_dma) {
+		ret = davinci_acquire_dma_channels(host);
+		if (ret == -EPROBE_DEFER)
+			goto out;
+		else if (ret)
+			host->use_dma = 0;
+	}
 
 	/* REVISIT:  someday, support IRQ-driven card detection.  */
 	mmc->caps |= MMC_CAP_NEEDS_POLL;
-- 
1.9.1

[PATCH v3 2/5] mmc: davinci: remove matching string

[David Lechner]

The string "MMCSDCLK" is not actually used for clock lookup, so can be removed.

Signed-off-by: David Lechner <david@lechnology.com>
Acked-by: Sekhar Nori <nsekhar@ti.com>
---

v3 changes: none.

 drivers/mmc/host/davinci_mmc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/mmc/host/davinci_mmc.c b/drivers/mmc/host/davinci_mmc.c
index eb3ca9e..498c42d 100644
--- a/drivers/mmc/host/davinci_mmc.c
+++ b/drivers/mmc/host/davinci_mmc.c
@@ -1241,7 +1241,7 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 		goto out;
 
 	ret = -ENXIO;
-	host->clk = clk_get(&pdev->dev, "MMCSDCLK");
+	host->clk = clk_get(&pdev->dev, NULL);
 	if (IS_ERR(host->clk)) {
 		ret = PTR_ERR(host->clk);
 		goto out;
-- 
1.9.1

[PATCH v3 3/5] mmc: davinci: fix unwinding in probe

[David Lechner]

Unwiding from an error in davinci_mmcsd_probe was a mess. Some errors were
not handled and not all paths unwound correctly. Also using devm_ where
possible to simplify things.

Signed-off-by: David Lechner <david@lechnology.com>
---

v3 changes: fixed ret = -ENODEV mistake. Updated to apply on top of patch from
Peter Ujfalusi.


 drivers/mmc/host/davinci_mmc.c | 96 ++++++++++++++++--------------------------
 1 file changed, 37 insertions(+), 59 deletions(-)

diff --git a/drivers/mmc/host/davinci_mmc.c b/drivers/mmc/host/davinci_mmc.c
index 498c42d..e75407d 100644
--- a/drivers/mmc/host/davinci_mmc.c
+++ b/drivers/mmc/host/davinci_mmc.c
@@ -1205,7 +1205,7 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 	struct mmc_davinci_host *host = NULL;
 	struct mmc_host *mmc = NULL;
 	struct resource *r, *mem = NULL;
-	int ret = 0, irq = 0;
+	int ret, irq;
 	size_t mem_size;
 	const struct platform_device_id *id_entry;
 
@@ -1215,38 +1215,38 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 		return -ENOENT;
 	}
 
-	ret = -ENODEV;
 	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	irq = platform_get_irq(pdev, 0);
 	if (!r || irq == NO_IRQ)
-		goto out;
+		return -ENODEV;
 
-	ret = -EBUSY;
 	mem_size = resource_size(r);
-	mem = request_mem_region(r->start, mem_size, pdev->name);
+	mem = devm_request_mem_region(&pdev->dev, r->start, mem_size,
+				      pdev->name);
 	if (!mem)
-		goto out;
+		return -EBUSY;
 
-	ret = -ENOMEM;
 	mmc = mmc_alloc_host(sizeof(struct mmc_davinci_host), &pdev->dev);
 	if (!mmc)
-		goto out;
+		return -ENOMEM;
 
 	host = mmc_priv(mmc);
 	host->mmc = mmc;	/* Important */
 
 	host->mem_res = mem;
-	host->base = ioremap(mem->start, mem_size);
+	host->base = devm_ioremap(&pdev->dev, mem->start, mem_size);
 	if (!host->base)
-		goto out;
+		goto ioremap_fail;
 
-	ret = -ENXIO;
-	host->clk = clk_get(&pdev->dev, NULL);
+	host->clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(host->clk)) {
 		ret = PTR_ERR(host->clk);
-		goto out;
+		goto clk_get_fail;
 	}
-	clk_enable(host->clk);
+	ret = clk_enable(host->clk);
+	if (ret)
+		goto clk_enable_fail;
+
 	host->mmc_input_clk = clk_get_rate(host->clk);
 
 	init_mmcsd_host(host);
@@ -1264,7 +1264,7 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 	if (host->use_dma) {
 		ret = davinci_acquire_dma_channels(host);
 		if (ret == -EPROBE_DEFER)
-			goto out;
+			goto dma_probe_defer;
 		else if (ret)
 			host->use_dma = 0;
 	}
@@ -1321,15 +1321,17 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 
 	ret = mmc_add_host(mmc);
 	if (ret < 0)
-		goto out;
+		goto mmc_add_host_fail;
 
-	ret = request_irq(irq, mmc_davinci_irq, 0, mmc_hostname(mmc), host);
+	ret = devm_request_irq(&pdev->dev, irq, mmc_davinci_irq, 0,
+			       mmc_hostname(mmc), host);
 	if (ret)
-		goto out;
+		goto request_irq_fail;
 
 	if (host->sdio_irq >= 0) {
-		ret = request_irq(host->sdio_irq, mmc_davinci_sdio_irq, 0,
-				  mmc_hostname(mmc), host);
+		ret = devm_request_irq(&pdev->dev, host->sdio_irq,
+				       mmc_davinci_sdio_irq, 0,
+				       mmc_hostname(mmc), host);
 		if (!ret)
 			mmc->caps |= MMC_CAP_SDIO_IRQ;
 	}
@@ -1342,28 +1344,18 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 
 	return 0;
 
-out:
+request_irq_fail:
+	mmc_remove_host(mmc);
+mmc_add_host_fail:
 	mmc_davinci_cpufreq_deregister(host);
 cpu_freq_fail:
-	if (host) {
-		davinci_release_dma_channels(host);
-
-		if (host->clk) {
-			clk_disable(host->clk);
-			clk_put(host->clk);
-		}
-
-		if (host->base)
-			iounmap(host->base);
-	}
-
-	if (mmc)
-		mmc_free_host(mmc);
-
-	if (mem)
-		release_resource(mem);
-
-	dev_dbg(&pdev->dev, "probe err %d\n", ret);
+	davinci_release_dma_channels(host);
+dma_probe_defer:
+	clk_disable(host->clk);
+clk_enable_fail:
+clk_get_fail:
+ioremap_fail:
+	mmc_free_host(mmc);
 
 	return ret;
 }
@@ -1372,25 +1364,11 @@ static int __exit davinci_mmcsd_remove(struct platform_device *pdev)
 {
 	struct mmc_davinci_host *host = platform_get_drvdata(pdev);
 
-	if (host) {
-		mmc_davinci_cpufreq_deregister(host);
-
-		mmc_remove_host(host->mmc);
-		free_irq(host->mmc_irq, host);
-		if (host->mmc->caps & MMC_CAP_SDIO_IRQ)
-			free_irq(host->sdio_irq, host);
-
-		davinci_release_dma_channels(host);
-
-		clk_disable(host->clk);
-		clk_put(host->clk);
-
-		iounmap(host->base);
-
-		release_resource(host->mem_res);
-
-		mmc_free_host(host->mmc);
-	}
+	mmc_remove_host(host->mmc);
+	mmc_davinci_cpufreq_deregister(host);
+	davinci_release_dma_channels(host);
+	clk_disable(host->clk);
+	mmc_free_host(host->mmc);
 
 	return 0;
 }
-- 
1.9.1

[PATCH v3 4/5] mmc: davinci: prepare clock

[David Lechner]

When trying to use this driver with the common clock framework, enabling
the clock fails because it was not prepared. This fixes the problem by
calling clk_prepare and clk_enable in a single function. Ditto for
clk_disable_unprepare.

Signed-off-by: David Lechner <david@lechnology.com>
Acked-by: Sekhar Nori <nsekhar@ti.com>
---

v3 changes: none.


 drivers/mmc/host/davinci_mmc.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/mmc/host/davinci_mmc.c b/drivers/mmc/host/davinci_mmc.c
index e75407d..d5d931b 100644
--- a/drivers/mmc/host/davinci_mmc.c
+++ b/drivers/mmc/host/davinci_mmc.c
@@ -1243,9 +1243,9 @@ static int __init davinci_mmcsd_probe(struct platform_device *pdev)
 		ret = PTR_ERR(host->clk);
 		goto clk_get_fail;
 	}
-	ret = clk_enable(host->clk);
+	ret = clk_prepare_enable(host->clk);
 	if (ret)
-		goto clk_enable_fail;
+		goto clk_prepare_enable_fail;
 
 	host->mmc_input_clk = clk_get_rate(host->clk);
 
@@ -1351,8 +1351,8 @@ mmc_add_host_fail:
 cpu_freq_fail:
 	davinci_release_dma_channels(host);
 dma_probe_defer:
-	clk_disable(host->clk);
-clk_enable_fail:
+	clk_disable_unprepare(host->clk);
+clk_prepare_enable_fail:
 clk_get_fail:
 ioremap_fail:
 	mmc_free_host(mmc);
@@ -1367,7 +1367,7 @@ static int __exit davinci_mmcsd_remove(struct platform_device *pdev)
 	mmc_remove_host(host->mmc);
 	mmc_davinci_cpufreq_deregister(host);
 	davinci_release_dma_channels(host);
-	clk_disable(host->clk);
+	clk_disable_unprepare(host->clk);
 	mmc_free_host(host->mmc);
 
 	return 0;
-- 
1.9.1

[PATCH v3 5/5] ARM: davinci: remove mmc dma resources

[David Lechner]

The davinci_mmc driver no longer uses platform resources for getting dma
channels. Instead lookup is now done using dma_slave_map.

Signed-off-by: David Lechner <david@lechnology.com>
Acked-by: Sekhar Nori <nsekhar@ti.com>
---

v3 changes: none.


 arch/arm/mach-davinci/devices-da8xx.c | 20 --------------------
 arch/arm/mach-davinci/devices.c       | 16 ----------------
 2 files changed, 36 deletions(-)

diff --git a/arch/arm/mach-davinci/devices-da8xx.c b/arch/arm/mach-davinci/devices-da8xx.c
index 725e693..add3771 100644
--- a/arch/arm/mach-davinci/devices-da8xx.c
+++ b/arch/arm/mach-davinci/devices-da8xx.c
@@ -751,16 +751,6 @@ static struct resource da8xx_mmcsd0_resources[] = {
 		.end	= IRQ_DA8XX_MMCSDINT0,
 		.flags	= IORESOURCE_IRQ,
 	},
-	{		/* DMA RX */
-		.start	= DA8XX_DMA_MMCSD0_RX,
-		.end	= DA8XX_DMA_MMCSD0_RX,
-		.flags	= IORESOURCE_DMA,
-	},
-	{		/* DMA TX */
-		.start	= DA8XX_DMA_MMCSD0_TX,
-		.end	= DA8XX_DMA_MMCSD0_TX,
-		.flags	= IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device da8xx_mmcsd0_device = {
@@ -788,16 +778,6 @@ static struct resource da850_mmcsd1_resources[] = {
 		.end	= IRQ_DA850_MMCSDINT0_1,
 		.flags	= IORESOURCE_IRQ,
 	},
-	{		/* DMA RX */
-		.start	= DA850_DMA_MMCSD1_RX,
-		.end	= DA850_DMA_MMCSD1_RX,
-		.flags	= IORESOURCE_DMA,
-	},
-	{		/* DMA TX */
-		.start	= DA850_DMA_MMCSD1_TX,
-		.end	= DA850_DMA_MMCSD1_TX,
-		.flags	= IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device da850_mmcsd1_device = {
diff --git a/arch/arm/mach-davinci/devices.c b/arch/arm/mach-davinci/devices.c
index 6257aa4..67d26c5 100644
--- a/arch/arm/mach-davinci/devices.c
+++ b/arch/arm/mach-davinci/devices.c
@@ -144,14 +144,6 @@ static struct resource mmcsd0_resources[] = {
 		.start = IRQ_SDIOINT,
 		.flags = IORESOURCE_IRQ,
 	},
-	/* DMA channels: RX, then TX */
-	{
-		.start = EDMA_CTLR_CHAN(0, DAVINCI_DMA_MMCRXEVT),
-		.flags = IORESOURCE_DMA,
-	}, {
-		.start = EDMA_CTLR_CHAN(0, DAVINCI_DMA_MMCTXEVT),
-		.flags = IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device davinci_mmcsd0_device = {
@@ -181,14 +173,6 @@ static struct resource mmcsd1_resources[] = {
 		.start = IRQ_DM355_SDIOINT1,
 		.flags = IORESOURCE_IRQ,
 	},
-	/* DMA channels: RX, then TX */
-	{
-		.start = EDMA_CTLR_CHAN(0, 30),	/* rx */
-		.flags = IORESOURCE_DMA,
-	}, {
-		.start = EDMA_CTLR_CHAN(0, 31),	/* tx */
-		.flags = IORESOURCE_DMA,
-	},
 };
 
 static struct platform_device davinci_mmcsd1_device = {
-- 
1.9.1

[PATCH v3 0/5] davinci_mmc fixes

[Ulf Hansson]

On 17 March 2016 at 04:45, David Lechner <david@lechnology.com> wrote:
> I have replaced one of my patches with the patch from Peter Ujfalusi and fixed
> a mistake in the "fix unwinding..." patch.
>
> Tested working on LEGO MINDSTORMS EV3 (da850-ish).
>
> David Lechner (4):
>   mmc: davinci: remove matching string
>   mmc: davinci: fix unwinding in probe
>   mmc: davinci: prepare clock
>   ARM: davinci: remove mmc dma resources
>
> Peter Ujfalusi (1):
>   mmc: davinci_mmc: Use dma_request_chan() to requesting DMA channel
>
>  arch/arm/mach-davinci/devices-da8xx.c |  20 -----
>  arch/arm/mach-davinci/devices.c       |  16 ----
>  drivers/mmc/host/davinci_mmc.c        | 147 ++++++++++++----------------------
>  3 files changed, 50 insertions(+), 133 deletions(-)
>
> --
> 1.9.1
>

Patch 3 triggers a compiler warning, please fix it.

../drivers/mmc/host/davinci_mmc.c: In function ?davinci_mmcsd_probe?:
../drivers/mmc/host/davinci_mmc.c:1208:6: warning: ?ret? may be used
uninitialized in this function [-Wmaybe-uninitialized]

I have applied patch 1->2 for next, so no need to resend these.

Kind regards
Uffe

[PATCH v3 0/5] davinci_mmc fixes

[David Lechner]

On 04/05/2016 06:12 AM, Ulf Hansson wrote:
>
> Patch 3 triggers a compiler warning, please fix it.
>
> ../drivers/mmc/host/davinci_mmc.c: In function ?davinci_mmcsd_probe?:
> ../drivers/mmc/host/davinci_mmc.c:1208:6: warning: ?ret? may be used
> uninitialized in this function [-Wmaybe-uninitialized]
>
> I have applied patch 1->2 for next, so no need to resend these.
>
> Kind regards
> Uffe
>

I am not getting a compiler warning with this commit. Can you please 
provide a git link so that I can be sure I am building using the same 
branch that you are using?

[PATCH v3 0/5] davinci_mmc fixes

[David Lechner]

On 04/05/2016 11:51 AM, David Lechner wrote:
> On 04/05/2016 06:12 AM, Ulf Hansson wrote:
>>
>> Patch 3 triggers a compiler warning, please fix it.
>>
>> ../drivers/mmc/host/davinci_mmc.c: In function ?davinci_mmcsd_probe?:
>> ../drivers/mmc/host/davinci_mmc.c:1208:6: warning: ?ret? may be used
>> uninitialized in this function [-Wmaybe-uninitialized]
>>
>> I have applied patch 1->2 for next, so no need to resend these.
>>
>> Kind regards
>> Uffe
>>
>
> I am not getting a compiler warning with this commit. Can you please
> provide a git link so that I can be sure I am building using the same
> branch that you are using?


I think I figured it out. I am applying the patches on top of the next 
branch from git://git.linaro.org/people/ulf.hansson/mmc.git. However, I 
am still not getting a compiler warning. Perhaps it has to do with the 
gcc version. I'm not exactly sure how to fix it since it is really not 
used uninitialized. Just set it to 0 in the variable declaration?

[PATCH v3 0/5] davinci_mmc fixes

[David Lechner]

On 04/05/2016 12:07 PM, David Lechner wrote:
> On 04/05/2016 11:51 AM, David Lechner wrote:
>> On 04/05/2016 06:12 AM, Ulf Hansson wrote:
>>>
>>> Patch 3 triggers a compiler warning, please fix it.
>>>
>>> ../drivers/mmc/host/davinci_mmc.c: In function ?davinci_mmcsd_probe?:
>>> ../drivers/mmc/host/davinci_mmc.c:1208:6: warning: ?ret? may be used
>>> uninitialized in this function [-Wmaybe-uninitialized]
>>>
>>> I have applied patch 1->2 for next, so no need to resend these.
>>>
>>> Kind regards
>>> Uffe
>>>
>>
>> I am not getting a compiler warning with this commit. Can you please
>> provide a git link so that I can be sure I am building using the same
>> branch that you are using?
>
>
> I think I figured it out. I am applying the patches on top of the next
> branch from git://git.linaro.org/people/ulf.hansson/mmc.git. However, I
> am still not getting a compiler warning. Perhaps it has to do with the
> gcc version. I'm not exactly sure how to fix it since it is really not
> used uninitialized. Just set it to 0 in the variable declaration?

OK. I guess it is my compiler that is broken. :-/
I stared at it long enough and I see the error now.