remove bogus GFP_DMA32 flags for dma allocations

remove bogus GFP_DMA32 flags for dma allocations

[Christoph Hellwig]

dma_alloc_* internally selects the zone to allocate from based on the
DMA mask.  Remove all explicit uses of GFP_DMA32 with dma_alloc_*.

[PATCH 1/8] cpufreq: tegra186: don't pass GFP_DMA32 to dma_alloc_coherent

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/cpufreq/tegra186-cpufreq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/cpufreq/tegra186-cpufreq.c b/drivers/cpufreq/tegra186-cpufreq.c
index 1f59966573aa..f1e09022b819 100644
--- a/drivers/cpufreq/tegra186-cpufreq.c
+++ b/drivers/cpufreq/tegra186-cpufreq.c
@@ -121,7 +121,7 @@ static struct cpufreq_frequency_table *init_vhint_table(
 	void *virt;
 
 	virt = dma_alloc_coherent(bpmp->dev, sizeof(*data), &phys,
-				  GFP_KERNEL | GFP_DMA32);
+				  GFP_KERNEL);
 	if (!virt)
 		return ERR_PTR(-ENOMEM);
 
-- 
2.19.1

Re: [PATCH 1/8] cpufreq: tegra186: don't pass GFP_DMA32 to dma_alloc_coherent

[Rafael J. Wysocki]

On Sat, Oct 13, 2018 at 5:17 PM Christoph Hellwig <hch@lst.de> wrote:
>
> The DMA API does its own zone decisions based on the coherent_dma_mask.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

> ---
>  drivers/cpufreq/tegra186-cpufreq.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/cpufreq/tegra186-cpufreq.c b/drivers/cpufreq/tegra186-cpufreq.c
> index 1f59966573aa..f1e09022b819 100644
> --- a/drivers/cpufreq/tegra186-cpufreq.c
> +++ b/drivers/cpufreq/tegra186-cpufreq.c
> @@ -121,7 +121,7 @@ static struct cpufreq_frequency_table *init_vhint_table(
>         void *virt;
>
>         virt = dma_alloc_coherent(bpmp->dev, sizeof(*data), &phys,
> -                                 GFP_KERNEL | GFP_DMA32);
> +                                 GFP_KERNEL);
>         if (!virt)
>                 return ERR_PTR(-ENOMEM);
>
> --
> 2.19.1
>

Re: [PATCH 1/8] cpufreq: tegra186: don't pass GFP_DMA32 to dma_alloc_coherent

[Christoph Hellwig]

On Mon, Oct 15, 2018 at 09:43:04AM +0200, Rafael J. Wysocki wrote:
> On Sat, Oct 13, 2018 at 5:17 PM Christoph Hellwig <hch@lst.de> wrote:
> >
> > The DMA API does its own zone decisions based on the coherent_dma_mask.
> >
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> 
> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

Can you pick it up through the cpufreq tree?

Re: [PATCH 1/8] cpufreq: tegra186: don't pass GFP_DMA32 to dma_alloc_coherent

[Rafael J. Wysocki]

On Wed, Oct 17, 2018 at 9:19 AM Christoph Hellwig <hch@lst.de> wrote:
>
> On Mon, Oct 15, 2018 at 09:43:04AM +0200, Rafael J. Wysocki wrote:
> > On Sat, Oct 13, 2018 at 5:17 PM Christoph Hellwig <hch@lst.de> wrote:
> > >
> > > The DMA API does its own zone decisions based on the coherent_dma_mask.
> > >
> > > Signed-off-by: Christoph Hellwig <hch@lst.de>
> >
> > Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
>
> Can you pick it up through the cpufreq tree?

Sure, I'll do that, thanks!

[PATCH 2/8] firmware: tegra: don't pass GFP_DMA32 to dma_alloc_coherent

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/firmware/tegra/bpmp-debugfs.c | 11 +++++------
 drivers/firmware/tegra/bpmp.c         |  2 +-
 2 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/firmware/tegra/bpmp-debugfs.c b/drivers/firmware/tegra/bpmp-debugfs.c
index f7f6a0a5cb07..567160897bac 100644
--- a/drivers/firmware/tegra/bpmp-debugfs.c
+++ b/drivers/firmware/tegra/bpmp-debugfs.c
@@ -218,12 +218,12 @@ static int debugfs_show(struct seq_file *m, void *p)
 		return -ENOENT;
 
 	namevirt = dma_alloc_coherent(bpmp->dev, namesize, &namephys,
-				      GFP_KERNEL | GFP_DMA32);
+				      GFP_KERNEL);
 	if (!namevirt)
 		return -ENOMEM;
 
 	datavirt = dma_alloc_coherent(bpmp->dev, datasize, &dataphys,
-				      GFP_KERNEL | GFP_DMA32);
+				      GFP_KERNEL);
 	if (!datavirt) {
 		ret = -ENOMEM;
 		goto free_namebuf;
@@ -269,12 +269,12 @@ static ssize_t debugfs_store(struct file *file, const char __user *buf,
 		return -ENOENT;
 
 	namevirt = dma_alloc_coherent(bpmp->dev, namesize, &namephys,
-				      GFP_KERNEL | GFP_DMA32);
+				      GFP_KERNEL);
 	if (!namevirt)
 		return -ENOMEM;
 
 	datavirt = dma_alloc_coherent(bpmp->dev, datasize, &dataphys,
-				      GFP_KERNEL | GFP_DMA32);
+				      GFP_KERNEL);
 	if (!datavirt) {
 		ret = -ENOMEM;
 		goto free_namebuf;
@@ -422,8 +422,7 @@ int tegra_bpmp_init_debugfs(struct tegra_bpmp *bpmp)
 	if (!root)
 		return -ENOMEM;
 
-	virt = dma_alloc_coherent(bpmp->dev, sz, &phys,
-				  GFP_KERNEL | GFP_DMA32);
+	virt = dma_alloc_coherent(bpmp->dev, sz, &phys, GFP_KERNEL);
 	if (!virt) {
 		ret = -ENOMEM;
 		goto out;
diff --git a/drivers/firmware/tegra/bpmp.c b/drivers/firmware/tegra/bpmp.c
index 14a456afa379..e6d2356ccec3 100644
--- a/drivers/firmware/tegra/bpmp.c
+++ b/drivers/firmware/tegra/bpmp.c
@@ -531,7 +531,7 @@ static int tegra_bpmp_get_firmware_tag(struct tegra_bpmp *bpmp, char *tag,
 	int err;
 
 	virt = dma_alloc_coherent(bpmp->dev, MSG_DATA_MIN_SZ, &phys,
-				  GFP_KERNEL | GFP_DMA32);
+				  GFP_KERNEL);
 	if (!virt)
 		return -ENOMEM;
 
-- 
2.19.1

[PATCH 3/8] spi: pic32-sqi: don't pass GFP_DMA32 to dma_alloc_coherent

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/spi/spi-pic32-sqi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/spi/spi-pic32-sqi.c b/drivers/spi/spi-pic32-sqi.c
index bd1c6b53283f..ab53e461d80f 100644
--- a/drivers/spi/spi-pic32-sqi.c
+++ b/drivers/spi/spi-pic32-sqi.c
@@ -468,7 +468,7 @@ static int ring_desc_ring_alloc(struct pic32_sqi *sqi)
 	/* allocate coherent DMAable memory for hardware buffer descriptors. */
 	sqi->bd = dma_zalloc_coherent(&sqi->master->dev,
 				      sizeof(*bd) * PESQI_BD_COUNT,
-				      &sqi->bd_dma, GFP_DMA32);
+				      &sqi->bd_dma, GFP_KERNEL);
 	if (!sqi->bd) {
 		dev_err(&sqi->master->dev, "failed allocating dma buffer\n");
 		return -ENOMEM;
-- 
2.19.1

Applied "spi: pic32-sqi: don't pass GFP_DMA32 to dma_alloc_coherent" to the spi tree

[Mark Brown]

The patch

   spi: pic32-sqi: don't pass GFP_DMA32 to dma_alloc_coherent

has been applied to the spi tree at

   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi.git 

All being well this means that it will be integrated into the linux-next
tree (usually sometime in the next 24 hours) and sent to Linus during
the next merge window (or sooner if it is a bug fix), however if
problems are discovered then the patch may be dropped or reverted.  

You may get further e-mails resulting from automated or manual testing
and review of the tree, please engage with people reporting problems and
send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they
should be sent as incremental updates against current git, existing
patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying
to this mail.

Thanks,
Mark

From ec506e9246bf42795f1fa8a5cd00740e5686ba73 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Sat, 13 Oct 2018 17:17:02 +0200
Subject: [PATCH] spi: pic32-sqi: don't pass GFP_DMA32 to dma_alloc_coherent

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
---
 drivers/spi/spi-pic32-sqi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/spi/spi-pic32-sqi.c b/drivers/spi/spi-pic32-sqi.c
index 62e6bf1f50b1..d7e4e18ec3df 100644
--- a/drivers/spi/spi-pic32-sqi.c
+++ b/drivers/spi/spi-pic32-sqi.c
@@ -468,7 +468,7 @@ static int ring_desc_ring_alloc(struct pic32_sqi *sqi)
 	/* allocate coherent DMAable memory for hardware buffer descriptors. */
 	sqi->bd = dma_zalloc_coherent(&sqi->master->dev,
 				      sizeof(*bd) * PESQI_BD_COUNT,
-				      &sqi->bd_dma, GFP_DMA32);
+				      &sqi->bd_dma, GFP_KERNEL);
 	if (!sqi->bd) {
 		dev_err(&sqi->master->dev, "failed allocating dma buffer\n");
 		return -ENOMEM;
-- 
2.19.0.rc2

[PATCH 4/8] sound: hpios: don't pass GFP_DMA32 to dma_alloc_coherent

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 sound/pci/asihpi/hpios.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sound/pci/asihpi/hpios.c b/sound/pci/asihpi/hpios.c
index 5ef4fe964366..7c91330af719 100644
--- a/sound/pci/asihpi/hpios.c
+++ b/sound/pci/asihpi/hpios.c
@@ -49,7 +49,7 @@ u16 hpios_locked_mem_alloc(struct consistent_dma_area *p_mem_area, u32 size,
 	/*?? any benefit in using managed dmam_alloc_coherent? */
 	p_mem_area->vaddr =
 		dma_alloc_coherent(&pdev->dev, size, &p_mem_area->dma_handle,
-		GFP_DMA32 | GFP_KERNEL);
+		GFP_KERNEL);
 
 	if (p_mem_area->vaddr) {
 		HPI_DEBUG_LOG(DEBUG, "allocated %d bytes, dma 0x%x vma %p\n",
-- 
2.19.1

Re: [alsa-devel] [PATCH 4/8] sound: hpios: don't pass GFP_DMA32 to dma_alloc_coherent

[Takashi Iwai]

On Sat, 13 Oct 2018 17:17:03 +0200,
Christoph Hellwig wrote:
> 
> The DMA API does its own zone decisions based on the coherent_dma_mask.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Reviewed-by: Takashi Iwai <tiwai@suse.de>


Would you like to take this as a series, or shall I take individually
through sound tree?


thanks,

Takashi

> ---
>  sound/pci/asihpi/hpios.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/sound/pci/asihpi/hpios.c b/sound/pci/asihpi/hpios.c
> index 5ef4fe964366..7c91330af719 100644
> --- a/sound/pci/asihpi/hpios.c
> +++ b/sound/pci/asihpi/hpios.c
> @@ -49,7 +49,7 @@ u16 hpios_locked_mem_alloc(struct consistent_dma_area *p_mem_area, u32 size,
>  	/*?? any benefit in using managed dmam_alloc_coherent? */
>  	p_mem_area->vaddr =
>  		dma_alloc_coherent(&pdev->dev, size, &p_mem_area->dma_handle,
> -		GFP_DMA32 | GFP_KERNEL);
> +		GFP_KERNEL);
>  
>  	if (p_mem_area->vaddr) {
>  		HPI_DEBUG_LOG(DEBUG, "allocated %d bytes, dma 0x%x vma %p\n",
> -- 
> 2.19.1
> 
> _______________________________________________
> Alsa-devel mailing list
> Alsa-devel@alsa-project.org
> http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
> 

Re: [alsa-devel] [PATCH 4/8] sound: hpios: don't pass GFP_DMA32 to dma_alloc_coherent

[Christoph Hellwig]

On Sat, Oct 13, 2018 at 06:18:28PM +0200, Takashi Iwai wrote:
> On Sat, 13 Oct 2018 17:17:03 +0200,
> Christoph Hellwig wrote:
> > 
> > The DMA API does its own zone decisions based on the coherent_dma_mask.
> > 
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> 
> Reviewed-by: Takashi Iwai <tiwai@suse.de>
> 
> 
> Would you like to take this as a series, or shall I take individually
> through sound tree?

There is nothing that depends on this, so feel free to apply the
two sound patches to your tree.

Re: [alsa-devel] [PATCH 4/8] sound: hpios: don't pass GFP_DMA32 to dma_alloc_coherent

[Takashi Iwai]

On Sat, 13 Oct 2018 18:35:40 +0200,
Christoph Hellwig wrote:
> 
> On Sat, Oct 13, 2018 at 06:18:28PM +0200, Takashi Iwai wrote:
> > On Sat, 13 Oct 2018 17:17:03 +0200,
> > Christoph Hellwig wrote:
> > > 
> > > The DMA API does its own zone decisions based on the coherent_dma_mask.
> > > 
> > > Signed-off-by: Christoph Hellwig <hch@lst.de>
> > 
> > Reviewed-by: Takashi Iwai <tiwai@suse.de>
> > 
> > 
> > Would you like to take this as a series, or shall I take individually
> > through sound tree?
> 
> There is nothing that depends on this, so feel free to apply the
> two sound patches to your tree.

OK, thanks.


Takashi

[PATCH 5/8] sound: intel/sst: don't pass GFP_DMA32 to dma_alloc_coherent

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 sound/soc/intel/common/sst-firmware.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sound/soc/intel/common/sst-firmware.c b/sound/soc/intel/common/sst-firmware.c
index 11041aedea31..1e067504b604 100644
--- a/sound/soc/intel/common/sst-firmware.c
+++ b/sound/soc/intel/common/sst-firmware.c
@@ -355,7 +355,7 @@ struct sst_fw *sst_fw_new(struct sst_dsp *dsp,
 
 	/* allocate DMA buffer to store FW data */
 	sst_fw->dma_buf = dma_alloc_coherent(dsp->dma_dev, sst_fw->size,
-				&sst_fw->dmable_fw_paddr, GFP_DMA | GFP_KERNEL);
+				&sst_fw->dmable_fw_paddr, GFP_KERNEL);
 	if (!sst_fw->dma_buf) {
 		dev_err(dsp->dev, "error: DMA alloc failed\n");
 		kfree(sst_fw);
-- 
2.19.1

Re: [alsa-devel] [PATCH 5/8] sound: intel/sst: don't pass GFP_DMA32 to dma_alloc_coherent

[Takashi Iwai]

On Sat, 13 Oct 2018 17:17:04 +0200,
Christoph Hellwig wrote:
> 
> The DMA API does its own zone decisions based on the coherent_dma_mask.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Reviewed-by: Takashi Iwai <tiwai@suse.de>


thanks,

Takashi

> ---
>  sound/soc/intel/common/sst-firmware.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/sound/soc/intel/common/sst-firmware.c b/sound/soc/intel/common/sst-firmware.c
> index 11041aedea31..1e067504b604 100644
> --- a/sound/soc/intel/common/sst-firmware.c
> +++ b/sound/soc/intel/common/sst-firmware.c
> @@ -355,7 +355,7 @@ struct sst_fw *sst_fw_new(struct sst_dsp *dsp,
>  
>  	/* allocate DMA buffer to store FW data */
>  	sst_fw->dma_buf = dma_alloc_coherent(dsp->dma_dev, sst_fw->size,
> -				&sst_fw->dmable_fw_paddr, GFP_DMA | GFP_KERNEL);
> +				&sst_fw->dmable_fw_paddr, GFP_KERNEL);
>  	if (!sst_fw->dma_buf) {
>  		dev_err(dsp->dev, "error: DMA alloc failed\n");
>  		kfree(sst_fw);
> -- 
> 2.19.1
> 
> _______________________________________________
> Alsa-devel mailing list
> Alsa-devel@alsa-project.org
> http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
> 

Applied "ASoC: intel: don't pass GFP_DMA32 to dma_alloc_coherent" to the asoc tree

[Mark Brown]

The patch

   ASoC: intel: don't pass GFP_DMA32 to dma_alloc_coherent

has been applied to the asoc tree at

   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git 

All being well this means that it will be integrated into the linux-next
tree (usually sometime in the next 24 hours) and sent to Linus during
the next merge window (or sooner if it is a bug fix), however if
problems are discovered then the patch may be dropped or reverted.  

You may get further e-mails resulting from automated or manual testing
and review of the tree, please engage with people reporting problems and
send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they
should be sent as incremental updates against current git, existing
patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying
to this mail.

Thanks,
Mark

From 3b991038498bc5011b063d6a804503c577a79434 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Sat, 13 Oct 2018 17:17:04 +0200
Subject: [PATCH] ASoC: intel: don't pass GFP_DMA32 to dma_alloc_coherent

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
---
 sound/soc/intel/common/sst-firmware.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sound/soc/intel/common/sst-firmware.c b/sound/soc/intel/common/sst-firmware.c
index 11041aedea31..1e067504b604 100644
--- a/sound/soc/intel/common/sst-firmware.c
+++ b/sound/soc/intel/common/sst-firmware.c
@@ -355,7 +355,7 @@ struct sst_fw *sst_fw_new(struct sst_dsp *dsp,
 
 	/* allocate DMA buffer to store FW data */
 	sst_fw->dma_buf = dma_alloc_coherent(dsp->dma_dev, sst_fw->size,
-				&sst_fw->dmable_fw_paddr, GFP_DMA | GFP_KERNEL);
+				&sst_fw->dmable_fw_paddr, GFP_KERNEL);
 	if (!sst_fw->dma_buf) {
 		dev_err(dsp->dev, "error: DMA alloc failed\n");
 		kfree(sst_fw);
-- 
2.19.0.rc2

[PATCH 6/8] drm: sti: don't pass GFP_DMA32 to dma_alloc_wc

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/gpu/drm/sti/sti_gdp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c
index c32de6cbf061..cdf0a1396e00 100644
--- a/drivers/gpu/drm/sti/sti_gdp.c
+++ b/drivers/gpu/drm/sti/sti_gdp.c
@@ -517,7 +517,7 @@ static void sti_gdp_init(struct sti_gdp *gdp)
 	/* Allocate all the nodes within a single memory page */
 	size = sizeof(struct sti_gdp_node) *
 	    GDP_NODE_PER_FIELD * GDP_NODE_NB_BANK;
-	base = dma_alloc_wc(gdp->dev, size, &dma_addr, GFP_KERNEL | GFP_DMA);
+	base = dma_alloc_wc(gdp->dev, size, &dma_addr, GFP_KERNEL);
 
 	if (!base) {
 		DRM_ERROR("Failed to allocate memory for GDP node\n");
-- 
2.19.1

Re: [PATCH 6/8] drm: sti: don't pass GFP_DMA32 to dma_alloc_wc

[Benjamin Gaignard]

Le sam. 13 oct. 2018 à 17:17, Christoph Hellwig <hch@lst.de> a écrit :
>
> The DMA API does its own zone decisions based on the coherent_dma_mask.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>

> ---
>  drivers/gpu/drm/sti/sti_gdp.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c
> index c32de6cbf061..cdf0a1396e00 100644
> --- a/drivers/gpu/drm/sti/sti_gdp.c
> +++ b/drivers/gpu/drm/sti/sti_gdp.c
> @@ -517,7 +517,7 @@ static void sti_gdp_init(struct sti_gdp *gdp)
>         /* Allocate all the nodes within a single memory page */
>         size = sizeof(struct sti_gdp_node) *
>             GDP_NODE_PER_FIELD * GDP_NODE_NB_BANK;
> -       base = dma_alloc_wc(gdp->dev, size, &dma_addr, GFP_KERNEL | GFP_DMA);
> +       base = dma_alloc_wc(gdp->dev, size, &dma_addr, GFP_KERNEL);
>
>         if (!base) {
>                 DRM_ERROR("Failed to allocate memory for GDP node\n");
> --
> 2.19.1
>
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel



-- 
Benjamin Gaignard

Graphic Study Group

Linaro.org │ Open source software for ARM SoCs

Follow Linaro: Facebook | Twitter | Blog

Re: [PATCH 6/8] drm: sti: don't pass GFP_DMA32 to dma_alloc_wc

[Benjamin Gaignard]

Le lun. 15 oct. 2018 à 11:12, Benjamin Gaignard
<benjamin.gaignard@linaro.org> a écrit :
>
> Le sam. 13 oct. 2018 à 17:17, Christoph Hellwig <hch@lst.de> a écrit :
> >
> > The DMA API does its own zone decisions based on the coherent_dma_mask.
> >
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
>
> Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>

Christoph do you plan to merge this patch on your own tree ? or would
like I put it directly in drm-misc-next branch ?

Regards,
Benjamin
>
> > ---
> >  drivers/gpu/drm/sti/sti_gdp.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c
> > index c32de6cbf061..cdf0a1396e00 100644
> > --- a/drivers/gpu/drm/sti/sti_gdp.c
> > +++ b/drivers/gpu/drm/sti/sti_gdp.c
> > @@ -517,7 +517,7 @@ static void sti_gdp_init(struct sti_gdp *gdp)
> >         /* Allocate all the nodes within a single memory page */
> >         size = sizeof(struct sti_gdp_node) *
> >             GDP_NODE_PER_FIELD * GDP_NODE_NB_BANK;
> > -       base = dma_alloc_wc(gdp->dev, size, &dma_addr, GFP_KERNEL | GFP_DMA);
> > +       base = dma_alloc_wc(gdp->dev, size, &dma_addr, GFP_KERNEL);
> >
> >         if (!base) {
> >                 DRM_ERROR("Failed to allocate memory for GDP node\n");
> > --
> > 2.19.1
> >
> > _______________________________________________
> > dri-devel mailing list
> > dri-devel@lists.freedesktop.org
> > https://lists.freedesktop.org/mailman/listinfo/dri-devel
>
>
>
> --
> Benjamin Gaignard
>
> Graphic Study Group
>
> Linaro.org │ Open source software for ARM SoCs
>
> Follow Linaro: Facebook | Twitter | Blog



-- 
Benjamin Gaignard

Graphic Study Group

Linaro.org │ Open source software for ARM SoCs

Follow Linaro: Facebook | Twitter | Blog

Re: [PATCH 6/8] drm: sti: don't pass GFP_DMA32 to dma_alloc_wc

[Christoph Hellwig]

On Tue, Oct 16, 2018 at 02:41:23PM +0200, Benjamin Gaignard wrote:
> Le lun. 15 oct. 2018 à 11:12, Benjamin Gaignard
> <benjamin.gaignard@linaro.org> a écrit :
> >
> > Le sam. 13 oct. 2018 à 17:17, Christoph Hellwig <hch@lst.de> a écrit :
> > >
> > > The DMA API does its own zone decisions based on the coherent_dma_mask.
> > >
> > > Signed-off-by: Christoph Hellwig <hch@lst.de>
> >
> > Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>
> 
> Christoph do you plan to merge this patch on your own tree ? or would
> like I put it directly in drm-misc-next branch ?

Please pull it in through drm-misc-next, thanks!

Re: [PATCH 6/8] drm: sti: don't pass GFP_DMA32 to dma_alloc_wc

[Benjamin Gaignard]

Le mer. 17 oct. 2018 à 09:19, Christoph Hellwig <hch@lst.de> a écrit :
>
> On Tue, Oct 16, 2018 at 02:41:23PM +0200, Benjamin Gaignard wrote:
> > Le lun. 15 oct. 2018 à 11:12, Benjamin Gaignard
> > <benjamin.gaignard@linaro.org> a écrit :
> > >
> > > Le sam. 13 oct. 2018 à 17:17, Christoph Hellwig <hch@lst.de> a écrit :
> > > >
> > > > The DMA API does its own zone decisions based on the coherent_dma_mask.
> > > >
> > > > Signed-off-by: Christoph Hellwig <hch@lst.de>
> > >
> > > Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>
> >
> > Christoph do you plan to merge this patch on your own tree ? or would
> > like I put it directly in drm-misc-next branch ?
>
> Please pull it in through drm-misc-next, thanks!

Applied on drm-misc-next,
Thanks,
Benjamin

[PATCH 7/8] media: sti/bdisp: don't pass GFP_DMA32 to dma_alloc_attrs

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/media/platform/sti/bdisp/bdisp-hw.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/media/platform/sti/bdisp/bdisp-hw.c b/drivers/media/platform/sti/bdisp/bdisp-hw.c
index 26d9fa7aeb5f..4372abbb5950 100644
--- a/drivers/media/platform/sti/bdisp/bdisp-hw.c
+++ b/drivers/media/platform/sti/bdisp/bdisp-hw.c
@@ -510,7 +510,7 @@ int bdisp_hw_alloc_filters(struct device *dev)
 
 	/* Allocate all the filters within a single memory page */
 	size = (BDISP_HF_NB * NB_H_FILTER) + (BDISP_VF_NB * NB_V_FILTER);
-	base = dma_alloc_attrs(dev, size, &paddr, GFP_KERNEL | GFP_DMA,
+	base = dma_alloc_attrs(dev, size, &paddr, GFP_KERNEL,
 			       DMA_ATTR_WRITE_COMBINE);
 	if (!base)
 		return -ENOMEM;
-- 
2.19.1

Re: [PATCH 7/8] media: sti/bdisp: don't pass GFP_DMA32 to dma_alloc_attrs

[Benjamin Gaignard]

Le sam. 13 oct. 2018 à 17:18, Christoph Hellwig <hch@lst.de> a écrit :
>
> The DMA API does its own zone decisions based on the coherent_dma_mask.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>

> ---
>  drivers/media/platform/sti/bdisp/bdisp-hw.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/media/platform/sti/bdisp/bdisp-hw.c b/drivers/media/platform/sti/bdisp/bdisp-hw.c
> index 26d9fa7aeb5f..4372abbb5950 100644
> --- a/drivers/media/platform/sti/bdisp/bdisp-hw.c
> +++ b/drivers/media/platform/sti/bdisp/bdisp-hw.c
> @@ -510,7 +510,7 @@ int bdisp_hw_alloc_filters(struct device *dev)
>
>         /* Allocate all the filters within a single memory page */
>         size = (BDISP_HF_NB * NB_H_FILTER) + (BDISP_VF_NB * NB_V_FILTER);
> -       base = dma_alloc_attrs(dev, size, &paddr, GFP_KERNEL | GFP_DMA,
> +       base = dma_alloc_attrs(dev, size, &paddr, GFP_KERNEL,
>                                DMA_ATTR_WRITE_COMBINE);
>         if (!base)
>                 return -ENOMEM;
> --
> 2.19.1
>
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel



-- 
Benjamin Gaignard

Graphic Study Group

Linaro.org │ Open source software for ARM SoCs

Follow Linaro: Facebook | Twitter | Blog

Re: [PATCH 7/8] media: sti/bdisp: don't pass GFP_DMA32 to dma_alloc_attrs

[Christoph Hellwig]

On Mon, Oct 15, 2018 at 11:12:55AM +0200, Benjamin Gaignard wrote:
> Le sam. 13 oct. 2018 à 17:18, Christoph Hellwig <hch@lst.de> a écrit :
> >
> > The DMA API does its own zone decisions based on the coherent_dma_mask.
> >
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> 
> Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>

Can you pick it up through the media tree?

Re: [PATCH 7/8] media: sti/bdisp: don't pass GFP_DMA32 to dma_alloc_attrs

[Benjamin Gaignard]

Le mer. 17 oct. 2018 à 09:20, Christoph Hellwig <hch@lst.de> a écrit :
>
> On Mon, Oct 15, 2018 at 11:12:55AM +0200, Benjamin Gaignard wrote:
> > Le sam. 13 oct. 2018 à 17:18, Christoph Hellwig <hch@lst.de> a écrit :
> > >
> > > The DMA API does its own zone decisions based on the coherent_dma_mask.
> > >
> > > Signed-off-by: Christoph Hellwig <hch@lst.de>
> >
> > Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>
>
> Can you pick it up through the media tree?

No but Mauros or Hans (in CC) can add it.

Re: [PATCH 7/8] media: sti/bdisp: don't pass GFP_DMA32 to dma_alloc_attrs

[Mauro Carvalho Chehab]

Em Thu, 18 Oct 2018 14:00:40 +0200
Benjamin Gaignard <benjamin.gaignard@linaro.org> escreveu:

> Le mer. 17 oct. 2018 à 09:20, Christoph Hellwig <hch@lst.de> a écrit :
> >
> > On Mon, Oct 15, 2018 at 11:12:55AM +0200, Benjamin Gaignard wrote:  
> > > Le sam. 13 oct. 2018 à 17:18, Christoph Hellwig <hch@lst.de> a écrit :  
> > > >
> > > > The DMA API does its own zone decisions based on the coherent_dma_mask.
> > > >
> > > > Signed-off-by: Christoph Hellwig <hch@lst.de>  
> > >
> > > Reviewed-by: Benjamin Gaignard <benjamin.gaignard@linaro.org>  
> >
> > Can you pick it up through the media tree?  
> 
> No but Mauros or Hans (in CC) can add it.

I'm adding it. Sorry for the delay. All those trips for MS/KS made
harder to handle it earlier.

Thanks,
Mauro

[PATCH 8/8] fsl-diu-fb: don't pass GFP_DMA32 to dmam_alloc_coherent

[Christoph Hellwig]

The DMA API does its own zone decisions based on the coherent_dma_mask.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/video/fbdev/fsl-diu-fb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/video/fbdev/fsl-diu-fb.c b/drivers/video/fbdev/fsl-diu-fb.c
index bc9eb8afc313..6a49fe917bdb 100644
--- a/drivers/video/fbdev/fsl-diu-fb.c
+++ b/drivers/video/fbdev/fsl-diu-fb.c
@@ -1697,7 +1697,7 @@ static int fsl_diu_probe(struct platform_device *pdev)
 	int ret;
 
 	data = dmam_alloc_coherent(&pdev->dev, sizeof(struct fsl_diu_data),
-				   &dma_addr, GFP_DMA | __GFP_ZERO);
+				   &dma_addr, GFP_KERNEL | __GFP_ZERO);
 	if (!data)
 		return -ENOMEM;
 	data->dma_addr = dma_addr;
-- 
2.19.1