* [PATCH v3 1/4] dmaengine: rcar-dmac: Fix DMACHCLR handling if iommu is mapped
2019-09-09 6:34 [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support Yoshihiro Shimoda
@ 2019-09-09 6:34 ` Yoshihiro Shimoda
2019-09-09 6:34 ` [PATCH v3 2/4] dmaengine: rcar-dmac: Use of_data values instead of a macro Yoshihiro Shimoda
` (3 subsequent siblings)
4 siblings, 0 replies; 8+ messages in thread
From: Yoshihiro Shimoda @ 2019-09-09 6:34 UTC (permalink / raw)
To: vinod.koul; +Cc: dmaengine, linux-renesas-soc, Yoshihiro Shimoda
The commit 20c169aceb45 ("dmaengine: rcar-dmac: clear pertinence
number of channels") forgets to clear the last channel by
DMACHCLR in rcar_dmac_init() (and doesn't need to clear the first
channel) if iommu is mapped to the device. So, this patch fixes it
by using "channels_mask" bitfield.
Note that the hardware and driver don't support more than 32 bits
in DMACHCLR register anyway, so this patch should reject more than
32 channels in rcar_dmac_parse_of().
Fixes: 20c169aceb459575 ("dmaengine: rcar-dmac: clear pertinence number of channels")
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/1567424643-26629-1-git-send-email-yoshihiro.shimoda.uh@renesas.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
(cherry picked from commit cf24aac38698bfa1d021afd3883df3c4c65143a4)
[just cherry-picked it for slave-dma/next]
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
---
drivers/dma/sh/rcar-dmac.c | 28 +++++++++++++++++++---------
1 file changed, 19 insertions(+), 9 deletions(-)
diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
index eb6f231..3993ab6 100644
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -192,6 +192,7 @@ struct rcar_dmac_chan {
* @iomem: remapped I/O memory base
* @n_channels: number of available channels
* @channels: array of DMAC channels
+ * @channels_mask: bitfield of which DMA channels are managed by this driver
* @modules: bitmask of client modules in use
*/
struct rcar_dmac {
@@ -202,6 +203,7 @@ struct rcar_dmac {
unsigned int n_channels;
struct rcar_dmac_chan *channels;
+ unsigned int channels_mask;
DECLARE_BITMAP(modules, 256);
};
@@ -438,7 +440,7 @@ static int rcar_dmac_init(struct rcar_dmac *dmac)
u16 dmaor;
/* Clear all channels and enable the DMAC globally. */
- rcar_dmac_write(dmac, RCAR_DMACHCLR, GENMASK(dmac->n_channels - 1, 0));
+ rcar_dmac_write(dmac, RCAR_DMACHCLR, dmac->channels_mask);
rcar_dmac_write(dmac, RCAR_DMAOR,
RCAR_DMAOR_PRI_FIXED | RCAR_DMAOR_DME);
@@ -814,6 +816,9 @@ static void rcar_dmac_stop_all_chan(struct rcar_dmac *dmac)
for (i = 0; i < dmac->n_channels; ++i) {
struct rcar_dmac_chan *chan = &dmac->channels[i];
+ if (!(dmac->channels_mask & BIT(i)))
+ continue;
+
/* Stop and reinitialize the channel. */
spin_lock_irq(&chan->lock);
rcar_dmac_chan_halt(chan);
@@ -1774,6 +1779,8 @@ static int rcar_dmac_chan_probe(struct rcar_dmac *dmac,
return 0;
}
+#define RCAR_DMAC_MAX_CHANNELS 32
+
static int rcar_dmac_parse_of(struct device *dev, struct rcar_dmac *dmac)
{
struct device_node *np = dev->of_node;
@@ -1785,12 +1792,16 @@ static int rcar_dmac_parse_of(struct device *dev, struct rcar_dmac *dmac)
return ret;
}
- if (dmac->n_channels <= 0 || dmac->n_channels >= 100) {
+ /* The hardware and driver don't support more than 32 bits in CHCLR */
+ if (dmac->n_channels <= 0 ||
+ dmac->n_channels >= RCAR_DMAC_MAX_CHANNELS) {
dev_err(dev, "invalid number of channels %u\n",
dmac->n_channels);
return -EINVAL;
}
+ dmac->channels_mask = GENMASK(dmac->n_channels - 1, 0);
+
return 0;
}
@@ -1800,7 +1811,6 @@ static int rcar_dmac_probe(struct platform_device *pdev)
DMA_SLAVE_BUSWIDTH_2_BYTES | DMA_SLAVE_BUSWIDTH_4_BYTES |
DMA_SLAVE_BUSWIDTH_8_BYTES | DMA_SLAVE_BUSWIDTH_16_BYTES |
DMA_SLAVE_BUSWIDTH_32_BYTES | DMA_SLAVE_BUSWIDTH_64_BYTES;
- unsigned int channels_offset = 0;
struct dma_device *engine;
struct rcar_dmac *dmac;
struct resource *mem;
@@ -1829,10 +1839,8 @@ static int rcar_dmac_probe(struct platform_device *pdev)
* level we can't disable it selectively, so ignore channel 0 for now if
* the device is part of an IOMMU group.
*/
- if (device_iommu_mapped(&pdev->dev)) {
- dmac->n_channels--;
- channels_offset = 1;
- }
+ if (device_iommu_mapped(&pdev->dev))
+ dmac->channels_mask &= ~BIT(0);
dmac->channels = devm_kcalloc(&pdev->dev, dmac->n_channels,
sizeof(*dmac->channels), GFP_KERNEL);
@@ -1890,8 +1898,10 @@ static int rcar_dmac_probe(struct platform_device *pdev)
INIT_LIST_HEAD(&engine->channels);
for (i = 0; i < dmac->n_channels; ++i) {
- ret = rcar_dmac_chan_probe(dmac, &dmac->channels[i],
- i + channels_offset);
+ if (!(dmac->channels_mask & BIT(i)))
+ continue;
+
+ ret = rcar_dmac_chan_probe(dmac, &dmac->channels[i], i);
if (ret < 0)
goto error;
}
--
2.7.4
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH v3 2/4] dmaengine: rcar-dmac: Use of_data values instead of a macro
2019-09-09 6:34 [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support Yoshihiro Shimoda
2019-09-09 6:34 ` [PATCH v3 1/4] dmaengine: rcar-dmac: Fix DMACHCLR handling if iommu is mapped Yoshihiro Shimoda
@ 2019-09-09 6:34 ` Yoshihiro Shimoda
2019-09-13 8:48 ` Simon Horman
2019-09-09 6:34 ` [PATCH v3 3/4] dmaengine: rcar-dmac: Use devm_platform_ioremap_resource() Yoshihiro Shimoda
` (2 subsequent siblings)
4 siblings, 1 reply; 8+ messages in thread
From: Yoshihiro Shimoda @ 2019-09-09 6:34 UTC (permalink / raw)
To: vinod.koul; +Cc: dmaengine, linux-renesas-soc, Yoshihiro Shimoda
Since we will have changed memory mapping of the DMAC in the future,
this patch uses of_data values instead of a macro to calculate
each channel's base offset.
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
drivers/dma/sh/rcar-dmac.c | 33 ++++++++++++++++++++++++++++-----
1 file changed, 28 insertions(+), 5 deletions(-)
diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
index 3993ab6..74996a0 100644
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -210,12 +210,20 @@ struct rcar_dmac {
#define to_rcar_dmac(d) container_of(d, struct rcar_dmac, engine)
+/*
+ * struct rcar_dmac_of_data - This driver's OF data
+ * @chan_offset_base: DMAC channels base offset
+ * @chan_offset_stride: DMAC channels offset stride
+ */
+struct rcar_dmac_of_data {
+ u32 chan_offset_base;
+ u32 chan_offset_stride;
+};
+
/* -----------------------------------------------------------------------------
* Registers
*/
-#define RCAR_DMAC_CHAN_OFFSET(i) (0x8000 + 0x80 * (i))
-
#define RCAR_DMAISTA 0x0020
#define RCAR_DMASEC 0x0030
#define RCAR_DMAOR 0x0060
@@ -1726,6 +1734,7 @@ static const struct dev_pm_ops rcar_dmac_pm = {
static int rcar_dmac_chan_probe(struct rcar_dmac *dmac,
struct rcar_dmac_chan *rchan,
+ const struct rcar_dmac_of_data *data,
unsigned int index)
{
struct platform_device *pdev = to_platform_device(dmac->dev);
@@ -1735,7 +1744,8 @@ static int rcar_dmac_chan_probe(struct rcar_dmac *dmac,
int ret;
rchan->index = index;
- rchan->iomem = dmac->iomem + RCAR_DMAC_CHAN_OFFSET(index);
+ rchan->iomem = dmac->iomem + data->chan_offset_base +
+ data->chan_offset_stride * index;
rchan->mid_rid = -EINVAL;
spin_lock_init(&rchan->lock);
@@ -1813,10 +1823,15 @@ static int rcar_dmac_probe(struct platform_device *pdev)
DMA_SLAVE_BUSWIDTH_32_BYTES | DMA_SLAVE_BUSWIDTH_64_BYTES;
struct dma_device *engine;
struct rcar_dmac *dmac;
+ const struct rcar_dmac_of_data *data;
struct resource *mem;
unsigned int i;
int ret;
+ data = of_device_get_match_data(&pdev->dev);
+ if (!data)
+ return -EINVAL;
+
dmac = devm_kzalloc(&pdev->dev, sizeof(*dmac), GFP_KERNEL);
if (!dmac)
return -ENOMEM;
@@ -1901,7 +1916,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
if (!(dmac->channels_mask & BIT(i)))
continue;
- ret = rcar_dmac_chan_probe(dmac, &dmac->channels[i], i);
+ ret = rcar_dmac_chan_probe(dmac, &dmac->channels[i], data, i);
if (ret < 0)
goto error;
}
@@ -1948,8 +1963,16 @@ static void rcar_dmac_shutdown(struct platform_device *pdev)
rcar_dmac_stop_all_chan(dmac);
}
+static const struct rcar_dmac_of_data rcar_dmac_data = {
+ .chan_offset_base = 0x8000,
+ .chan_offset_stride = 0x80,
+};
+
static const struct of_device_id rcar_dmac_of_ids[] = {
- { .compatible = "renesas,rcar-dmac", },
+ {
+ .compatible = "renesas,rcar-dmac",
+ .data = &rcar_dmac_data,
+ },
{ /* Sentinel */ }
};
MODULE_DEVICE_TABLE(of, rcar_dmac_of_ids);
--
2.7.4
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH v3 3/4] dmaengine: rcar-dmac: Use devm_platform_ioremap_resource()
2019-09-09 6:34 [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support Yoshihiro Shimoda
2019-09-09 6:34 ` [PATCH v3 1/4] dmaengine: rcar-dmac: Fix DMACHCLR handling if iommu is mapped Yoshihiro Shimoda
2019-09-09 6:34 ` [PATCH v3 2/4] dmaengine: rcar-dmac: Use of_data values instead of a macro Yoshihiro Shimoda
@ 2019-09-09 6:34 ` Yoshihiro Shimoda
2019-09-13 8:46 ` Simon Horman
2019-09-09 6:34 ` [PATCH v3 4/4] dmaengine: rcar-dmac: Add dma-channel-mask property support Yoshihiro Shimoda
2019-10-14 6:58 ` [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support Vinod Koul
4 siblings, 1 reply; 8+ messages in thread
From: Yoshihiro Shimoda @ 2019-09-09 6:34 UTC (permalink / raw)
To: vinod.koul; +Cc: dmaengine, linux-renesas-soc, Yoshihiro Shimoda
This patch uses devm_platform_ioremap_resource() instead of
using platform_get_resource() and devm_ioremap_resource() together
to simplify.
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
drivers/dma/sh/rcar-dmac.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
index 74996a0..542786d 100644
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -1824,7 +1824,6 @@ static int rcar_dmac_probe(struct platform_device *pdev)
struct dma_device *engine;
struct rcar_dmac *dmac;
const struct rcar_dmac_of_data *data;
- struct resource *mem;
unsigned int i;
int ret;
@@ -1863,8 +1862,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
return -ENOMEM;
/* Request resources. */
- mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
- dmac->iomem = devm_ioremap_resource(&pdev->dev, mem);
+ dmac->iomem = devm_platform_ioremap_resource(pdev, 0);
if (IS_ERR(dmac->iomem))
return PTR_ERR(dmac->iomem);
--
2.7.4
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH v3 3/4] dmaengine: rcar-dmac: Use devm_platform_ioremap_resource()
2019-09-09 6:34 ` [PATCH v3 3/4] dmaengine: rcar-dmac: Use devm_platform_ioremap_resource() Yoshihiro Shimoda
@ 2019-09-13 8:46 ` Simon Horman
0 siblings, 0 replies; 8+ messages in thread
From: Simon Horman @ 2019-09-13 8:46 UTC (permalink / raw)
To: Yoshihiro Shimoda; +Cc: vinod.koul, dmaengine, linux-renesas-soc
On Mon, Sep 09, 2019 at 03:34:51PM +0900, Yoshihiro Shimoda wrote:
> This patch uses devm_platform_ioremap_resource() instead of
> using platform_get_resource() and devm_ioremap_resource() together
> to simplify.
>
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
> ---
> drivers/dma/sh/rcar-dmac.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
> index 74996a0..542786d 100644
> --- a/drivers/dma/sh/rcar-dmac.c
> +++ b/drivers/dma/sh/rcar-dmac.c
> @@ -1824,7 +1824,6 @@ static int rcar_dmac_probe(struct platform_device *pdev)
> struct dma_device *engine;
> struct rcar_dmac *dmac;
> const struct rcar_dmac_of_data *data;
> - struct resource *mem;
> unsigned int i;
> int ret;
>
> @@ -1863,8 +1862,7 @@ static int rcar_dmac_probe(struct platform_device *pdev)
> return -ENOMEM;
>
> /* Request resources. */
> - mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> - dmac->iomem = devm_ioremap_resource(&pdev->dev, mem);
> + dmac->iomem = devm_platform_ioremap_resource(pdev, 0);
> if (IS_ERR(dmac->iomem))
> return PTR_ERR(dmac->iomem);
>
> --
> 2.7.4
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH v3 4/4] dmaengine: rcar-dmac: Add dma-channel-mask property support
2019-09-09 6:34 [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support Yoshihiro Shimoda
` (2 preceding siblings ...)
2019-09-09 6:34 ` [PATCH v3 3/4] dmaengine: rcar-dmac: Use devm_platform_ioremap_resource() Yoshihiro Shimoda
@ 2019-09-09 6:34 ` Yoshihiro Shimoda
2019-10-14 6:58 ` [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support Vinod Koul
4 siblings, 0 replies; 8+ messages in thread
From: Yoshihiro Shimoda @ 2019-09-09 6:34 UTC (permalink / raw)
To: vinod.koul; +Cc: dmaengine, linux-renesas-soc, Yoshihiro Shimoda
This patch adds dma-channel-mask property support not to reserve
some DMA channels for some reasons. (for example: a heterogeneous
CPU uses it.)
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
drivers/dma/sh/rcar-dmac.c | 10 +++++++++-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c
index 542786d..f06016d 100644
--- a/drivers/dma/sh/rcar-dmac.c
+++ b/drivers/dma/sh/rcar-dmac.c
@@ -203,7 +203,7 @@ struct rcar_dmac {
unsigned int n_channels;
struct rcar_dmac_chan *channels;
- unsigned int channels_mask;
+ u32 channels_mask;
DECLARE_BITMAP(modules, 256);
};
@@ -1810,7 +1810,15 @@ static int rcar_dmac_parse_of(struct device *dev, struct rcar_dmac *dmac)
return -EINVAL;
}
+ /*
+ * If the driver is unable to read dma-channel-mask property,
+ * the driver assumes that it can use all channels.
+ */
dmac->channels_mask = GENMASK(dmac->n_channels - 1, 0);
+ of_property_read_u32(np, "dma-channel-mask", &dmac->channels_mask);
+
+ /* If the property has out-of-channel mask, this driver clears it */
+ dmac->channels_mask &= GENMASK(dmac->n_channels - 1, 0);
return 0;
}
--
2.7.4
^ permalink raw reply related [flat|nested] 8+ messages in thread* Re: [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support
2019-09-09 6:34 [PATCH v3 0/4] dmaengine: rcar-dmac: use of_data and add dma-channel-mask support Yoshihiro Shimoda
` (3 preceding siblings ...)
2019-09-09 6:34 ` [PATCH v3 4/4] dmaengine: rcar-dmac: Add dma-channel-mask property support Yoshihiro Shimoda
@ 2019-10-14 6:58 ` Vinod Koul
4 siblings, 0 replies; 8+ messages in thread
From: Vinod Koul @ 2019-10-14 6:58 UTC (permalink / raw)
To: Yoshihiro Shimoda; +Cc: vinod.koul, dmaengine, linux-renesas-soc
On 09-09-19, 15:34, Yoshihiro Shimoda wrote:
> This patch series is based on the latest slave-dma.git / next branch.
>
> Changes from v2:
> - Rebase the latest slave-dma.git / next branch (In other words,
> this patch series doesn't depend any other branches.
> - Cherry-picked a patch which is contained in v5.3-rc8 to solve any
> dependency. (I'm not sure whether this is a right way or not...)
> https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=169317
Sorry for delay, I was on vacation + conference. Yeah ideally I would
merge fixes and this wouldn't be the case.
patch1 is no longer needed now, I have applied the rest
Thanks
--
~Vinod
^ permalink raw reply [flat|nested] 8+ messages in thread