* [PATCH 01/11] dmaengine: pl08x: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
@ 2014-10-16 9:00 ` Maxime Ripard
2014-10-16 9:00 ` [PATCH 02/11] dmaengine: bcm2835: " Maxime Ripard
` (11 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:00 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/amba-pl08x.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c
index e34024b000a4..1364d00881dd 100644
--- a/drivers/dma/amba-pl08x.c
+++ b/drivers/dma/amba-pl08x.c
@@ -2164,7 +2164,6 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
__func__, ret);
goto out_no_memcpy;
}
- pl08x->memcpy.chancnt = ret;
/* Register slave channels */
ret = pl08x_dma_init_virtual_channels(pl08x, &pl08x->slave,
@@ -2175,7 +2174,6 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id)
__func__, ret);
goto out_no_slave;
}
- pl08x->slave.chancnt = ret;
ret = dma_async_device_register(&pl08x->memcpy);
if (ret) {
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 02/11] dmaengine: bcm2835: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
2014-10-16 9:00 ` [PATCH 01/11] dmaengine: pl08x: Remove chancnt affectations Maxime Ripard
@ 2014-10-16 9:00 ` Maxime Ripard
2014-10-16 9:00 ` [PATCH 03/11] dmaengine: jz4740: " Maxime Ripard
` (10 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:00 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/bcm2835-dma.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/dma/bcm2835-dma.c b/drivers/dma/bcm2835-dma.c
index 68007974961a..def66fbbb7f3 100644
--- a/drivers/dma/bcm2835-dma.c
+++ b/drivers/dma/bcm2835-dma.c
@@ -525,8 +525,6 @@ static int bcm2835_dma_chan_init(struct bcm2835_dmadev *d, int chan_id, int irq)
vchan_init(&c->vc, &d->ddev);
INIT_LIST_HEAD(&c->node);
- d->ddev.chancnt++;
-
c->chan_base = BCM2835_DMA_CHANIO(d->base, chan_id);
c->ch = chan_id;
c->irq_number = irq;
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 03/11] dmaengine: jz4740: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
2014-10-16 9:00 ` [PATCH 01/11] dmaengine: pl08x: Remove chancnt affectations Maxime Ripard
2014-10-16 9:00 ` [PATCH 02/11] dmaengine: bcm2835: " Maxime Ripard
@ 2014-10-16 9:00 ` Maxime Ripard
2014-10-16 9:00 ` [PATCH 04/11] dmaengine: fsldma: " Maxime Ripard
` (9 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:00 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/dma-jz4740.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/dma/dma-jz4740.c b/drivers/dma/dma-jz4740.c
index 6a9d89c93b1f..8b7d8f4cafaf 100644
--- a/drivers/dma/dma-jz4740.c
+++ b/drivers/dma/dma-jz4740.c
@@ -562,10 +562,9 @@ static int jz4740_dma_probe(struct platform_device *pdev)
dd->device_prep_dma_cyclic = jz4740_dma_prep_dma_cyclic;
dd->device_control = jz4740_dma_control;
dd->dev = &pdev->dev;
- dd->chancnt = JZ_DMA_NR_CHANS;
INIT_LIST_HEAD(&dd->channels);
- for (i = 0; i < dd->chancnt; i++) {
+ for (i = 0; i < JZ_DMA_NR_CHANS; i++) {
chan = &dmadev->chan[i];
chan->id = i;
chan->vchan.desc_free = jz4740_dma_desc_free;
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 04/11] dmaengine: fsldma: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (2 preceding siblings ...)
2014-10-16 9:00 ` [PATCH 03/11] dmaengine: jz4740: " Maxime Ripard
@ 2014-10-16 9:00 ` Maxime Ripard
2014-10-16 9:00 ` [PATCH 05/11] dmaengine: k3: " Maxime Ripard
` (8 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:00 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/fsldma.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c
index d5d6885ab341..75ae6405a5ee 100644
--- a/drivers/dma/fsldma.c
+++ b/drivers/dma/fsldma.c
@@ -1332,7 +1332,6 @@ static int fsl_dma_chan_probe(struct fsldma_device *fdev,
/* Add the channel to DMA device channel list */
list_add_tail(&chan->common.device_node, &fdev->common.channels);
- fdev->common.chancnt++;
dev_info(fdev->dev, "#%d (%s), irq %d\n", chan->id, compatible,
chan->irq != NO_IRQ ? chan->irq : fdev->irq);
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 05/11] dmaengine: k3: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (3 preceding siblings ...)
2014-10-16 9:00 ` [PATCH 04/11] dmaengine: fsldma: " Maxime Ripard
@ 2014-10-16 9:00 ` Maxime Ripard
2014-10-16 9:00 ` [PATCH 06/11] dmaengine: mpc512x: " Maxime Ripard
` (7 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:00 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/k3dma.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/dma/k3dma.c b/drivers/dma/k3dma.c
index a1f911aaf220..58219aa649fd 100644
--- a/drivers/dma/k3dma.c
+++ b/drivers/dma/k3dma.c
@@ -722,7 +722,6 @@ static int k3_dma_probe(struct platform_device *op)
d->slave.device_issue_pending = k3_dma_issue_pending;
d->slave.device_control = k3_dma_control;
d->slave.copy_align = DMA_ALIGN;
- d->slave.chancnt = d->dma_requests;
/* init virtual channel */
d->chans = devm_kzalloc(&op->dev,
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 06/11] dmaengine: mpc512x: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (4 preceding siblings ...)
2014-10-16 9:00 ` [PATCH 05/11] dmaengine: k3: " Maxime Ripard
@ 2014-10-16 9:00 ` Maxime Ripard
2014-10-16 9:00 ` [PATCH 07/11] dmaengine: omap: " Maxime Ripard
` (6 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:00 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/mpc512x_dma.c | 12 +++++++-----
1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c
index 881db2bcb48b..fc08bb3af440 100644
--- a/drivers/dma/mpc512x_dma.c
+++ b/drivers/dma/mpc512x_dma.c
@@ -885,6 +885,7 @@ static int mpc_dma_probe(struct platform_device *op)
struct resource res;
ulong regs_start, regs_size;
int retval, i;
+ u8 chancnt;
mdma = devm_kzalloc(dev, sizeof(struct mpc_dma), GFP_KERNEL);
if (!mdma) {
@@ -956,10 +957,6 @@ static int mpc_dma_probe(struct platform_device *op)
dma = &mdma->dma;
dma->dev = dev;
- if (mdma->is_mpc8308)
- dma->chancnt = MPC8308_DMACHAN_MAX;
- else
- dma->chancnt = MPC512x_DMACHAN_MAX;
dma->device_alloc_chan_resources = mpc_dma_alloc_chan_resources;
dma->device_free_chan_resources = mpc_dma_free_chan_resources;
dma->device_issue_pending = mpc_dma_issue_pending;
@@ -972,7 +969,12 @@ static int mpc_dma_probe(struct platform_device *op)
dma_cap_set(DMA_MEMCPY, dma->cap_mask);
dma_cap_set(DMA_SLAVE, dma->cap_mask);
- for (i = 0; i < dma->chancnt; i++) {
+ if (mdma->is_mpc8308)
+ chancnt = MPC8308_DMACHAN_MAX;
+ else
+ chancnt = MPC512x_DMACHAN_MAX;
+
+ for (i = 0; i < chancnt; i++) {
mchan = &mdma->channels[i];
mchan->chan.device = dma;
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 07/11] dmaengine: omap: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (5 preceding siblings ...)
2014-10-16 9:00 ` [PATCH 06/11] dmaengine: mpc512x: " Maxime Ripard
@ 2014-10-16 9:00 ` Maxime Ripard
2014-10-16 9:01 ` [PATCH 08/11] dmaengine: sa11x0: " Maxime Ripard
` (5 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:00 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/omap-dma.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/dma/omap-dma.c b/drivers/dma/omap-dma.c
index 4cf7d9a950d7..b1e8c7317c0c 100644
--- a/drivers/dma/omap-dma.c
+++ b/drivers/dma/omap-dma.c
@@ -1069,8 +1069,6 @@ static int omap_dma_chan_init(struct omap_dmadev *od, int dma_sig)
vchan_init(&c->vc, &od->ddev);
INIT_LIST_HEAD(&c->node);
- od->ddev.chancnt++;
-
return 0;
}
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 08/11] dmaengine: sa11x0: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (6 preceding siblings ...)
2014-10-16 9:00 ` [PATCH 07/11] dmaengine: omap: " Maxime Ripard
@ 2014-10-16 9:01 ` Maxime Ripard
2014-10-16 9:01 ` [PATCH 09/11] dmaengine: sh: " Maxime Ripard
` (4 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:01 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/sa11x0-dma.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/dma/sa11x0-dma.c b/drivers/dma/sa11x0-dma.c
index 4b0ef043729a..2329d295efb5 100644
--- a/drivers/dma/sa11x0-dma.c
+++ b/drivers/dma/sa11x0-dma.c
@@ -829,7 +829,6 @@ static int sa11x0_dma_init_dmadev(struct dma_device *dmadev,
{
unsigned i;
- dmadev->chancnt = ARRAY_SIZE(chan_desc);
INIT_LIST_HEAD(&dmadev->channels);
dmadev->dev = dev;
dmadev->device_alloc_chan_resources = sa11x0_dma_alloc_chan_resources;
@@ -838,7 +837,7 @@ static int sa11x0_dma_init_dmadev(struct dma_device *dmadev,
dmadev->device_tx_status = sa11x0_dma_tx_status;
dmadev->device_issue_pending = sa11x0_dma_issue_pending;
- for (i = 0; i < dmadev->chancnt; i++) {
+ for (i = 0; i < ARRAY_SIZE(chan_desc); i++) {
struct sa11x0_dma_chan *c;
c = kzalloc(sizeof(*c), GFP_KERNEL);
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 09/11] dmaengine: sh: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (7 preceding siblings ...)
2014-10-16 9:01 ` [PATCH 08/11] dmaengine: sa11x0: " Maxime Ripard
@ 2014-10-16 9:01 ` Maxime Ripard
2014-10-26 16:31 ` Laurent Pinchart
2014-10-16 9:01 ` [PATCH 10/11] dmaengine: sirf: " Maxime Ripard
` (3 subsequent siblings)
12 siblings, 1 reply; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:01 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/sh/rcar-audmapp.c | 2 --
drivers/dma/sh/rcar-hpbdma.c | 2 --
drivers/dma/sh/shdma-base.c | 2 +-
drivers/dma/sh/shdmac.c | 2 --
drivers/dma/sh/sudmac.c | 2 --
5 files changed, 1 insertion(+), 9 deletions(-)
diff --git a/drivers/dma/sh/rcar-audmapp.c b/drivers/dma/sh/rcar-audmapp.c
index dabbf0aba2e9..fd1005e58551 100644
--- a/drivers/dma/sh/rcar-audmapp.c
+++ b/drivers/dma/sh/rcar-audmapp.c
@@ -248,7 +248,6 @@ static int audmapp_chan_probe(struct platform_device *pdev,
static void audmapp_chan_remove(struct audmapp_device *audev)
{
- struct dma_device *dma_dev = &audev->shdma_dev.dma_dev;
struct shdma_chan *schan;
int i;
@@ -256,7 +255,6 @@ static void audmapp_chan_remove(struct audmapp_device *audev)
BUG_ON(!schan);
shdma_chan_remove(schan);
}
- dma_dev->chancnt = 0;
}
static struct dma_chan *audmapp_of_xlate(struct of_phandle_args *dma_spec,
diff --git a/drivers/dma/sh/rcar-hpbdma.c b/drivers/dma/sh/rcar-hpbdma.c
index b212d9471ab5..5ab921865ffe 100644
--- a/drivers/dma/sh/rcar-hpbdma.c
+++ b/drivers/dma/sh/rcar-hpbdma.c
@@ -619,7 +619,6 @@ error:
static void hpb_dmae_chan_remove(struct hpb_dmae_device *hpbdev)
{
- struct dma_device *dma_dev = &hpbdev->shdma_dev.dma_dev;
struct shdma_chan *schan;
int i;
@@ -628,7 +627,6 @@ static void hpb_dmae_chan_remove(struct hpb_dmae_device *hpbdev)
shdma_chan_remove(schan);
}
- dma_dev->chancnt = 0;
}
static int hpb_dmae_remove(struct platform_device *pdev)
diff --git a/drivers/dma/sh/shdma-base.c b/drivers/dma/sh/shdma-base.c
index 42d497416196..ed5b0993edfa 100644
--- a/drivers/dma/sh/shdma-base.c
+++ b/drivers/dma/sh/shdma-base.c
@@ -951,7 +951,7 @@ void shdma_chan_probe(struct shdma_dev *sdev,
/* Add the channel to DMA device channel list */
list_add_tail(&schan->dma_chan.device_node,
&sdev->dma_dev.channels);
- sdev->schan[sdev->dma_dev.chancnt++] = schan;
+ sdev->schan[id] = schan;
}
EXPORT_SYMBOL(shdma_chan_probe);
diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c
index 58eb85770eba..b65317c6ea4e 100644
--- a/drivers/dma/sh/shdmac.c
+++ b/drivers/dma/sh/shdmac.c
@@ -572,7 +572,6 @@ err_no_irq:
static void sh_dmae_chan_remove(struct sh_dmae_device *shdev)
{
- struct dma_device *dma_dev = &shdev->shdma_dev.dma_dev;
struct shdma_chan *schan;
int i;
@@ -581,7 +580,6 @@ static void sh_dmae_chan_remove(struct sh_dmae_device *shdev)
shdma_chan_remove(schan);
}
- dma_dev->chancnt = 0;
}
static void sh_dmae_shutdown(struct platform_device *pdev)
diff --git a/drivers/dma/sh/sudmac.c b/drivers/dma/sh/sudmac.c
index 3ce103909896..69a5c7b87ee0 100644
--- a/drivers/dma/sh/sudmac.c
+++ b/drivers/dma/sh/sudmac.c
@@ -295,7 +295,6 @@ err_no_irq:
static void sudmac_chan_remove(struct sudmac_device *su_dev)
{
- struct dma_device *dma_dev = &su_dev->shdma_dev.dma_dev;
struct shdma_chan *schan;
int i;
@@ -304,7 +303,6 @@ static void sudmac_chan_remove(struct sudmac_device *su_dev)
shdma_chan_remove(schan);
}
- dma_dev->chancnt = 0;
}
static dma_addr_t sudmac_slave_addr(struct shdma_chan *schan)
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 09/11] dmaengine: sh: Remove chancnt affectations
2014-10-16 9:01 ` [PATCH 09/11] dmaengine: sh: " Maxime Ripard
@ 2014-10-26 16:31 ` Laurent Pinchart
0 siblings, 0 replies; 17+ messages in thread
From: Laurent Pinchart @ 2014-10-26 16:31 UTC (permalink / raw)
To: linux-arm-kernel
Hi Maxime,
Thank you for the patch.
On Thursday 16 October 2014 11:01:01 Maxime Ripard wrote:
> chanctnt is already filled by dma_async_device_register, which uses the
> channel list to know how much channels there is.
>
> Since it's already filled, we can safely remove it from the drivers' probe
> function.
>
> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> ---
> drivers/dma/sh/rcar-audmapp.c | 2 --
> drivers/dma/sh/rcar-hpbdma.c | 2 --
> drivers/dma/sh/shdma-base.c | 2 +-
> drivers/dma/sh/shdmac.c | 2 --
> drivers/dma/sh/sudmac.c | 2 --
> 5 files changed, 1 insertion(+), 9 deletions(-)
>
> diff --git a/drivers/dma/sh/rcar-audmapp.c b/drivers/dma/sh/rcar-audmapp.c
> index dabbf0aba2e9..fd1005e58551 100644
> --- a/drivers/dma/sh/rcar-audmapp.c
> +++ b/drivers/dma/sh/rcar-audmapp.c
> @@ -248,7 +248,6 @@ static int audmapp_chan_probe(struct platform_device
> *pdev,
>
> static void audmapp_chan_remove(struct audmapp_device *audev)
> {
> - struct dma_device *dma_dev = &audev->shdma_dev.dma_dev;
> struct shdma_chan *schan;
> int i;
>
> @@ -256,7 +255,6 @@ static void audmapp_chan_remove(struct audmapp_device
> *audev) BUG_ON(!schan);
> shdma_chan_remove(schan);
> }
> - dma_dev->chancnt = 0;
> }
>
> static struct dma_chan *audmapp_of_xlate(struct of_phandle_args *dma_spec,
> diff --git a/drivers/dma/sh/rcar-hpbdma.c b/drivers/dma/sh/rcar-hpbdma.c
> index b212d9471ab5..5ab921865ffe 100644
> --- a/drivers/dma/sh/rcar-hpbdma.c
> +++ b/drivers/dma/sh/rcar-hpbdma.c
> @@ -619,7 +619,6 @@ error:
>
> static void hpb_dmae_chan_remove(struct hpb_dmae_device *hpbdev)
> {
> - struct dma_device *dma_dev = &hpbdev->shdma_dev.dma_dev;
> struct shdma_chan *schan;
> int i;
>
> @@ -628,7 +627,6 @@ static void hpb_dmae_chan_remove(struct hpb_dmae_device
> *hpbdev)
>
> shdma_chan_remove(schan);
> }
> - dma_dev->chancnt = 0;
> }
>
> static int hpb_dmae_remove(struct platform_device *pdev)
> diff --git a/drivers/dma/sh/shdma-base.c b/drivers/dma/sh/shdma-base.c
> index 42d497416196..ed5b0993edfa 100644
> --- a/drivers/dma/sh/shdma-base.c
> +++ b/drivers/dma/sh/shdma-base.c
> @@ -951,7 +951,7 @@ void shdma_chan_probe(struct shdma_dev *sdev,
> /* Add the channel to DMA device channel list */
> list_add_tail(&schan->dma_chan.device_node,
> &sdev->dma_dev.channels);
> - sdev->schan[sdev->dma_dev.chancnt++] = schan;
> + sdev->schan[id] = schan;
> }
> EXPORT_SYMBOL(shdma_chan_probe);
>
> diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c
> index 58eb85770eba..b65317c6ea4e 100644
> --- a/drivers/dma/sh/shdmac.c
> +++ b/drivers/dma/sh/shdmac.c
> @@ -572,7 +572,6 @@ err_no_irq:
>
> static void sh_dmae_chan_remove(struct sh_dmae_device *shdev)
> {
> - struct dma_device *dma_dev = &shdev->shdma_dev.dma_dev;
> struct shdma_chan *schan;
> int i;
>
> @@ -581,7 +580,6 @@ static void sh_dmae_chan_remove(struct sh_dmae_device
> *shdev)
>
> shdma_chan_remove(schan);
> }
> - dma_dev->chancnt = 0;
> }
>
> static void sh_dmae_shutdown(struct platform_device *pdev)
> diff --git a/drivers/dma/sh/sudmac.c b/drivers/dma/sh/sudmac.c
> index 3ce103909896..69a5c7b87ee0 100644
> --- a/drivers/dma/sh/sudmac.c
> +++ b/drivers/dma/sh/sudmac.c
> @@ -295,7 +295,6 @@ err_no_irq:
>
> static void sudmac_chan_remove(struct sudmac_device *su_dev)
> {
> - struct dma_device *dma_dev = &su_dev->shdma_dev.dma_dev;
> struct shdma_chan *schan;
> int i;
>
> @@ -304,7 +303,6 @@ static void sudmac_chan_remove(struct sudmac_device
> *su_dev)
>
> shdma_chan_remove(schan);
> }
> - dma_dev->chancnt = 0;
> }
>
> static dma_addr_t sudmac_slave_addr(struct shdma_chan *schan)
--
Regards,
Laurent Pinchart
^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 10/11] dmaengine: sirf: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (8 preceding siblings ...)
2014-10-16 9:01 ` [PATCH 09/11] dmaengine: sh: " Maxime Ripard
@ 2014-10-16 9:01 ` Maxime Ripard
2014-10-16 9:01 ` [PATCH 11/11] dmaengine: sun6i: " Maxime Ripard
` (2 subsequent siblings)
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:01 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/sirf-dma.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/dma/sirf-dma.c b/drivers/dma/sirf-dma.c
index aac03ab10c54..fbe291584e8a 100644
--- a/drivers/dma/sirf-dma.c
+++ b/drivers/dma/sirf-dma.c
@@ -735,7 +735,6 @@ static int sirfsoc_dma_probe(struct platform_device *op)
dma = &sdma->dma;
dma->dev = dev;
- dma->chancnt = SIRFSOC_DMA_CHANNELS;
dma->device_alloc_chan_resources = sirfsoc_dma_alloc_chan_resources;
dma->device_free_chan_resources = sirfsoc_dma_free_chan_resources;
@@ -752,7 +751,7 @@ static int sirfsoc_dma_probe(struct platform_device *op)
dma_cap_set(DMA_INTERLEAVE, dma->cap_mask);
dma_cap_set(DMA_PRIVATE, dma->cap_mask);
- for (i = 0; i < dma->chancnt; i++) {
+ for (i = 0; i < SIRFSOC_DMA_CHANNELS; i++) {
schan = &sdma->channels[i];
schan->chan.device = dma;
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 11/11] dmaengine: sun6i: Remove chancnt affectations
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (9 preceding siblings ...)
2014-10-16 9:01 ` [PATCH 10/11] dmaengine: sirf: " Maxime Ripard
@ 2014-10-16 9:01 ` Maxime Ripard
2014-10-16 9:03 ` [PATCH 00/11] dmaengine: Cleanup chancnt Vinod Koul
2014-11-06 6:06 ` Vinod Koul
12 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-10-16 9:01 UTC (permalink / raw)
To: linux-arm-kernel
chanctnt is already filled by dma_async_device_register, which uses the channel
list to know how much channels there is.
Since it's already filled, we can safely remove it from the drivers' probe
function.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
---
drivers/dma/sun6i-dma.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/dma/sun6i-dma.c b/drivers/dma/sun6i-dma.c
index 1f92a56fd2b6..61f22fc1c631 100644
--- a/drivers/dma/sun6i-dma.c
+++ b/drivers/dma/sun6i-dma.c
@@ -937,7 +937,6 @@ static int sun6i_dma_probe(struct platform_device *pdev)
sdc->slave.device_prep_slave_sg = sun6i_dma_prep_slave_sg;
sdc->slave.device_prep_dma_memcpy = sun6i_dma_prep_dma_memcpy;
sdc->slave.device_control = sun6i_dma_control;
- sdc->slave.chancnt = NR_MAX_VCHANS;
sdc->slave.dev = &pdev->dev;
--
2.1.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH 00/11] dmaengine: Cleanup chancnt
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (10 preceding siblings ...)
2014-10-16 9:01 ` [PATCH 11/11] dmaengine: sun6i: " Maxime Ripard
@ 2014-10-16 9:03 ` Vinod Koul
2014-11-05 13:51 ` Maxime Ripard
2014-11-06 6:06 ` Vinod Koul
12 siblings, 1 reply; 17+ messages in thread
From: Vinod Koul @ 2014-10-16 9:03 UTC (permalink / raw)
To: linux-arm-kernel
On Thu, Oct 16, 2014 at 11:00:52AM +0200, Maxime Ripard wrote:
> Hi everyone,
>
> As pointed out in the documentation thread [1], the chancnt variable
> is filled during the driver registration in dma_async_device_register,
> from the liste of channels that has been set.
>
> This serie removes the occurences of the drivers that set it in their
> probe function.
Barring the typos in the cover letter, series looks fine.
Thanks
--
~Vinod
>
> Maxime
>
>
> 1: http://lists.infradead.org/pipermail/linux-arm-kernel/2014-October/292941.html
>
> Maxime Ripard (11):
> dmaengine: pl08x: Remove chancnt affectations
> dmaengine: bcm2835: Remove chancnt affectations
> dmaengine: jz4740: Remove chancnt affectations
> dmaengine: fsldma: Remove chancnt affectations
> dmaengine: k3: Remove chancnt affectations
> dmaengine: mpc512x: Remove chancnt affectations
> dmaengine: omap: Remove chancnt affectations
> dmaengine: sa11x0: Remove chancnt affectations
> dmaengine: sh: Remove chancnt affectations
> dmaengine: sirf: Remove chancnt affectations
> dmaengine: sun6i: Remove chancnt affectations
>
> drivers/dma/amba-pl08x.c | 2 --
> drivers/dma/bcm2835-dma.c | 2 --
> drivers/dma/dma-jz4740.c | 3 +--
> drivers/dma/fsldma.c | 1 -
> drivers/dma/k3dma.c | 1 -
> drivers/dma/mpc512x_dma.c | 12 +++++++-----
> drivers/dma/omap-dma.c | 2 --
> drivers/dma/sa11x0-dma.c | 3 +--
> drivers/dma/sh/rcar-audmapp.c | 2 --
> drivers/dma/sh/rcar-hpbdma.c | 2 --
> drivers/dma/sh/shdma-base.c | 2 +-
> drivers/dma/sh/shdmac.c | 2 --
> drivers/dma/sh/sudmac.c | 2 --
> drivers/dma/sirf-dma.c | 3 +--
> drivers/dma/sun6i-dma.c | 1 -
> 15 files changed, 11 insertions(+), 29 deletions(-)
>
> --
> 2.1.1
>
--
^ permalink raw reply [flat|nested] 17+ messages in thread* [PATCH 00/11] dmaengine: Cleanup chancnt
2014-10-16 9:03 ` [PATCH 00/11] dmaengine: Cleanup chancnt Vinod Koul
@ 2014-11-05 13:51 ` Maxime Ripard
0 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-11-05 13:51 UTC (permalink / raw)
To: linux-arm-kernel
Hi Vinod,
On Thu, Oct 16, 2014 at 02:33:46PM +0530, Vinod Koul wrote:
> On Thu, Oct 16, 2014 at 11:00:52AM +0200, Maxime Ripard wrote:
> > Hi everyone,
> >
> > As pointed out in the documentation thread [1], the chancnt variable
> > is filled during the driver registration in dma_async_device_register,
> > from the liste of channels that has been set.
> >
> > This serie removes the occurences of the drivers that set it in their
> > probe function.
> Barring the typos in the cover letter, series looks fine.
It's been 3 weeks already, and there's been no reviews justifying to
send a second version of the patch set.
Can we merge this in?
Thanks,
Maxime
--
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20141105/665c1e30/attachment.sig>
^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 00/11] dmaengine: Cleanup chancnt
2014-10-16 9:00 [PATCH 00/11] dmaengine: Cleanup chancnt Maxime Ripard
` (11 preceding siblings ...)
2014-10-16 9:03 ` [PATCH 00/11] dmaengine: Cleanup chancnt Vinod Koul
@ 2014-11-06 6:06 ` Vinod Koul
2014-11-06 9:58 ` Maxime Ripard
12 siblings, 1 reply; 17+ messages in thread
From: Vinod Koul @ 2014-11-06 6:06 UTC (permalink / raw)
To: linux-arm-kernel
On Thu, Oct 16, 2014 at 11:00:52AM +0200, Maxime Ripard wrote:
> Hi everyone,
>
> As pointed out in the documentation thread [1], the chancnt variable
> is filled during the driver registration in dma_async_device_register,
> from the liste of channels that has been set.
>
> This serie removes the occurences of the drivers that set it in their
> probe function.
Applied all, thanks
--
~Vinod
^ permalink raw reply [flat|nested] 17+ messages in thread* [PATCH 00/11] dmaengine: Cleanup chancnt
2014-11-06 6:06 ` Vinod Koul
@ 2014-11-06 9:58 ` Maxime Ripard
0 siblings, 0 replies; 17+ messages in thread
From: Maxime Ripard @ 2014-11-06 9:58 UTC (permalink / raw)
To: linux-arm-kernel
On Thu, Nov 06, 2014 at 11:36:53AM +0530, Vinod Koul wrote:
> On Thu, Oct 16, 2014 at 11:00:52AM +0200, Maxime Ripard wrote:
> > Hi everyone,
> >
> > As pointed out in the documentation thread [1], the chancnt variable
> > is filled during the driver registration in dma_async_device_register,
> > from the liste of channels that has been set.
> >
> > This serie removes the occurences of the drivers that set it in their
> > probe function.
>
> Applied all, thanks
Great, thanks!
Maxime
--
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20141106/5cfa82fc/attachment.sig>
^ permalink raw reply [flat|nested] 17+ messages in thread