All of lore.kernel.org
 help / color / mirror / Atom feed
* dma_declare_coherent_memory on main memory
@ 2018-12-07 15:34 ` Christoph Hellwig
  0 siblings, 0 replies; 13+ messages in thread
From: Christoph Hellwig @ 2018-12-07 15:34 UTC (permalink / raw)
  To: Shawn Guo, Sascha Hauer, Fabio Estevam
  Cc: iommu, linux-sh, linux-imx, linux-arm-kernel, linux-kernel

Hi all,

the ARM imx27/31 ports and various sh boards use
dma_declare_coherent_memory on main memory taken from the memblock
allocator.

Is there any good reason these couldn't be switched to CMA areas?
Getting rid of these magic dma_declare_coherent_memory area would
help making the dma allocator a lot simpler.

^ permalink raw reply	[flat|nested] 13+ messages in thread
* dma_declare_coherent_memory on main memory
@ 2018-12-07 15:34 ` Christoph Hellwig
  0 siblings, 0 replies; 13+ messages in thread
From: Christoph Hellwig @ 2018-12-07 15:34 UTC (permalink / raw)
  To: Shawn Guo, Sascha Hauer, Fabio Estevam
  Cc: linux-arm-kernel, iommu, linux-imx, linux-kernel, linux-sh

Hi all,

the ARM imx27/31 ports and various sh boards use
dma_declare_coherent_memory on main memory taken from the memblock
allocator.

Is there any good reason these couldn't be switched to CMA areas?
Getting rid of these magic dma_declare_coherent_memory area would
help making the dma allocator a lot simpler.

^ permalink raw reply	[flat|nested] 13+ messages in thread
* dma_declare_coherent_memory on main memory
@ 2018-12-07 15:34 ` Christoph Hellwig
  0 siblings, 0 replies; 13+ messages in thread
From: Christoph Hellwig @ 2018-12-07 15:34 UTC (permalink / raw)
  To: Shawn Guo, Sascha Hauer, Fabio Estevam
  Cc: iommu, linux-sh, linux-imx, linux-arm-kernel, linux-kernel

Hi all,

the ARM imx27/31 ports and various sh boards use
dma_declare_coherent_memory on main memory taken from the memblock
allocator.

Is there any good reason these couldn't be switched to CMA areas?
Getting rid of these magic dma_declare_coherent_memory area would
help making the dma allocator a lot simpler.

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
  2018-12-07 15:34 ` Christoph Hellwig
  (?)
  (?)
@ 2018-12-08 21:25   ` Rob Landley
  -1 siblings, 0 replies; 13+ messages in thread
From: Rob Landley @ 2018-12-08 21:25 UTC (permalink / raw)
  To: Christoph Hellwig, Shawn Guo, Sascha Hauer, Fabio Estevam
  Cc: iommu, linux-sh, linux-imx, linux-arm-kernel, linux-kernel

On 12/7/18 9:34 AM, Christoph Hellwig wrote:
> Hi all,
> 
> the ARM imx27/31 ports and various sh boards use
> dma_declare_coherent_memory on main memory taken from the memblock
> allocator.
> 
> Is there any good reason these couldn't be switched to CMA areas?
> Getting rid of these magic dma_declare_coherent_memory area would
> help making the dma allocator a lot simpler.

Not that I know of?

Rob

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
@ 2018-12-08 21:25   ` Rob Landley
  0 siblings, 0 replies; 13+ messages in thread
From: Rob Landley @ 2018-12-08 21:25 UTC (permalink / raw)
  To: Christoph Hellwig, Shawn Guo, Sascha Hauer, Fabio Estevam
  Cc: linux-arm-kernel, iommu, linux-imx, linux-kernel, linux-sh

On 12/7/18 9:34 AM, Christoph Hellwig wrote:
> Hi all,
> 
> the ARM imx27/31 ports and various sh boards use
> dma_declare_coherent_memory on main memory taken from the memblock
> allocator.
> 
> Is there any good reason these couldn't be switched to CMA areas?
> Getting rid of these magic dma_declare_coherent_memory area would
> help making the dma allocator a lot simpler.

Not that I know of?

Rob


^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
@ 2018-12-08 21:25   ` Rob Landley
  0 siblings, 0 replies; 13+ messages in thread
From: Rob Landley @ 2018-12-08 21:25 UTC (permalink / raw)
  To: Christoph Hellwig, Shawn Guo, Sascha Hauer, Fabio Estevam
  Cc: iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA,
	linux-sh-u79uwXL29TY76Z2rM5mHXA, linux-imx-3arQi8VN3Tc,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA

On 12/7/18 9:34 AM, Christoph Hellwig wrote:
> Hi all,
> 
> the ARM imx27/31 ports and various sh boards use
> dma_declare_coherent_memory on main memory taken from the memblock
> allocator.
> 
> Is there any good reason these couldn't be switched to CMA areas?
> Getting rid of these magic dma_declare_coherent_memory area would
> help making the dma allocator a lot simpler.

Not that I know of?

Rob

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
@ 2018-12-08 21:25   ` Rob Landley
  0 siblings, 0 replies; 13+ messages in thread
From: Rob Landley @ 2018-12-08 21:25 UTC (permalink / raw)
  To: Christoph Hellwig, Shawn Guo, Sascha Hauer, Fabio Estevam
  Cc: iommu, linux-sh, linux-imx, linux-arm-kernel, linux-kernel

On 12/7/18 9:34 AM, Christoph Hellwig wrote:
> Hi all,
> 
> the ARM imx27/31 ports and various sh boards use
> dma_declare_coherent_memory on main memory taken from the memblock
> allocator.
> 
> Is there any good reason these couldn't be switched to CMA areas?
> Getting rid of these magic dma_declare_coherent_memory area would
> help making the dma allocator a lot simpler.

Not that I know of?

Rob


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
  2018-12-07 15:34 ` Christoph Hellwig
  (?)
@ 2018-12-10  7:26   ` Sascha Hauer
  -1 siblings, 0 replies; 13+ messages in thread
From: Sascha Hauer @ 2018-12-10  7:26 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: linux-sh, linux-kernel, iommu, linux-imx, Fabio Estevam,
	Shawn Guo, linux-arm-kernel

On Fri, Dec 07, 2018 at 04:34:32PM +0100, Christoph Hellwig wrote:
> Hi all,
> 
> the ARM imx27/31 ports and various sh boards use
> dma_declare_coherent_memory on main memory taken from the memblock
> allocator.
> 
> Is there any good reason these couldn't be switched to CMA areas?
> Getting rid of these magic dma_declare_coherent_memory area would
> help making the dma allocator a lot simpler.

At least for i.MX27/31 I'd say this predates CMA support, so historical
reasons.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
@ 2018-12-10  7:26   ` Sascha Hauer
  0 siblings, 0 replies; 13+ messages in thread
From: Sascha Hauer @ 2018-12-10  7:26 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: Shawn Guo, Fabio Estevam, linux-arm-kernel, iommu, linux-imx,
	linux-kernel, linux-sh

On Fri, Dec 07, 2018 at 04:34:32PM +0100, Christoph Hellwig wrote:
> Hi all,
> 
> the ARM imx27/31 ports and various sh boards use
> dma_declare_coherent_memory on main memory taken from the memblock
> allocator.
> 
> Is there any good reason these couldn't be switched to CMA areas?
> Getting rid of these magic dma_declare_coherent_memory area would
> help making the dma allocator a lot simpler.

At least for i.MX27/31 I'd say this predates CMA support, so historical
reasons.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
@ 2018-12-10  7:26   ` Sascha Hauer
  0 siblings, 0 replies; 13+ messages in thread
From: Sascha Hauer @ 2018-12-10  7:26 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: linux-sh, linux-kernel, iommu, linux-imx, Fabio Estevam,
	Shawn Guo, linux-arm-kernel

On Fri, Dec 07, 2018 at 04:34:32PM +0100, Christoph Hellwig wrote:
> Hi all,
> 
> the ARM imx27/31 ports and various sh boards use
> dma_declare_coherent_memory on main memory taken from the memblock
> allocator.
> 
> Is there any good reason these couldn't be switched to CMA areas?
> Getting rid of these magic dma_declare_coherent_memory area would
> help making the dma allocator a lot simpler.

At least for i.MX27/31 I'd say this predates CMA support, so historical
reasons.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
  2018-12-10  7:26   ` Sascha Hauer
  (?)
@ 2018-12-10 19:02     ` Christoph Hellwig
  -1 siblings, 0 replies; 13+ messages in thread
From: Christoph Hellwig @ 2018-12-10 19:02 UTC (permalink / raw)
  To: Sascha Hauer
  Cc: linux-sh, linux-kernel, iommu, linux-imx, Fabio Estevam,
	Shawn Guo, Christoph Hellwig, linux-arm-kernel

On Mon, Dec 10, 2018 at 08:26:48AM +0100, Sascha Hauer wrote:
> > the ARM imx27/31 ports and various sh boards use
> > dma_declare_coherent_memory on main memory taken from the memblock
> > allocator.
> > 
> > Is there any good reason these couldn't be switched to CMA areas?
> > Getting rid of these magic dma_declare_coherent_memory area would
> > help making the dma allocator a lot simpler.
> 
> At least for i.MX27/31 I'd say this predates CMA support, so historical
> reasons.

Ok.  Do you still have test hardware for i.MX?  If so I'd love to see
if the patch below works.  Note that this is the brute force approach,
just having a global CMA pool might be a little more elegant.

diff --git a/arch/arm/configs/imx_v4_v5_defconfig b/arch/arm/configs/imx_v4_v5_defconfig
index 8661dd9b064a..88852e7e5e7e 100644
--- a/arch/arm/configs/imx_v4_v5_defconfig
+++ b/arch/arm/configs/imx_v4_v5_defconfig
@@ -185,3 +185,5 @@ CONFIG_NLS_ISO8859_1=y
 CONFIG_NLS_ISO8859_15=m
 CONFIG_FONTS=y
 CONFIG_FONT_8x8=y
+CONFIG_CMA=y
+CONFIG_DMA_CMA=y
diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
index 1ad5736c8fa6..16c8d717316c 100644
--- a/arch/arm/configs/imx_v6_v7_defconfig
+++ b/arch/arm/configs/imx_v6_v7_defconfig
@@ -444,3 +444,5 @@ CONFIG_PROVE_LOCKING=y
 # CONFIG_DEBUG_BUGVERBOSE is not set
 # CONFIG_FTRACE is not set
 # CONFIG_ARM_UNWIND is not set
+CONFIG_CMA=y
+CONFIG_DMA_CMA=y
diff --git a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
index 5169dfba9718..2339a50d5459 100644
--- a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
+++ b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
@@ -23,6 +23,7 @@
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
+#include <linux/cma.h>
 #include <linux/platform_device.h>
 #include <linux/mtd/physmap.h>
 #include <linux/i2c.h>
@@ -35,6 +36,7 @@
 #include <linux/platform_data/asoc-mx27vis.h>
 #include <media/soc_camera.h>
 #include <sound/tlv320aic32x4.h>
+#include <asm/dma-contiguous.h>
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/time.h>
@@ -245,6 +247,7 @@ static phys_addr_t mx2_camera_base __initdata;
 static void __init visstrim_analog_camera_init(void)
 {
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	gpio_set_value(TVP5150_PWDN, 1);
 	ndelay(1);
@@ -257,9 +260,10 @@ static void __init visstrim_analog_camera_init(void)
 	if (IS_ERR(pdev))
 		return;
 
-	dma_declare_coherent_memory(&pdev->dev, mx2_camera_base,
-				    mx2_camera_base, MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (cma_init_reserved_mem(mx2_camera_base, MX2_CAMERA_BUF_SIZE, 0,
+			"visstrim-cam", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base, MX2_CAMERA_BUF_SIZE);
 }
 
 static void __init visstrim_reserve(void)
@@ -440,13 +444,16 @@ static const struct imx_ssi_platform_data visstrim_m10_ssi_pdata __initconst = {
 static void __init visstrim_coda_init(void)
 {
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	pdev = imx27_add_coda();
-	dma_declare_coherent_memory(&pdev->dev,
-				    mx2_camera_base + MX2_CAMERA_BUF_SIZE,
-				    mx2_camera_base + MX2_CAMERA_BUF_SIZE,
-				    MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (IS_ERR(pdev))
+		return;
+	if (cma_init_reserved_mem(mx2_camera_base + MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE, 0, "visstrim-coda", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base + MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE);
 }
 
 /* DMA deinterlace */
@@ -459,21 +466,22 @@ static void __init visstrim_deinterlace_init(void)
 {
 	int ret = -ENOMEM;
 	struct platform_device *pdev = &visstrim_deinterlace;
+	struct cma *cma;
 
 	ret = platform_device_register(pdev);
 
-	dma_declare_coherent_memory(&pdev->dev,
-				    mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
-				    mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
-				    MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (cma_init_reserved_mem(mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE, 0, "visstrim-deinterlace", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE);
 }
 
 /* Emma-PrP for format conversion */
 static void __init visstrim_emmaprp_init(void)
 {
 	struct platform_device *pdev;
-	int ret;
+	struct cma *cma;
 
 	pdev = imx27_add_mx2_emmaprp();
 	if (IS_ERR(pdev))
@@ -483,12 +491,12 @@ static void __init visstrim_emmaprp_init(void)
 	 * Use the same memory area as the analog camera since both
 	 * devices are, by nature, exclusive.
 	 */
-	ret = dma_declare_coherent_memory(&pdev->dev,
-				mx2_camera_base, mx2_camera_base,
-				MX2_CAMERA_BUF_SIZE,
-				DMA_MEMORY_EXCLUSIVE);
-	if (ret)
+	if (cma_init_reserved_mem(mx2_camera_base, MX2_CAMERA_BUF_SIZE, 0,
+			"visstrim-emmaprp", &cma)) {
 		pr_err("Failed to declare memory for emmaprp\n");
+		return;
+	}
+	dma_contiguous_early_fixup(mx2_camera_base, MX2_CAMERA_BUF_SIZE);
 }
 
 /* Audio */
diff --git a/arch/arm/mach-imx/mach-mx31moboard.c b/arch/arm/mach-imx/mach-mx31moboard.c
index 643a3d749703..1c0705ec0ccd 100644
--- a/arch/arm/mach-imx/mach-mx31moboard.c
+++ b/arch/arm/mach-imx/mach-mx31moboard.c
@@ -13,6 +13,7 @@
  */
 
 #include <linux/delay.h>
+#include <linux/cma.h>
 #include <linux/dma-mapping.h>
 #include <linux/gfp.h>
 #include <linux/gpio.h>
@@ -37,6 +38,7 @@
 #include <linux/usb/otg.h>
 #include <linux/usb/ulpi.h>
 
+#include <asm/dma-contiguous.h>
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/time.h>
@@ -466,6 +468,7 @@ static int __init mx31moboard_init_cam(void)
 {
 	int ret;
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	imx31_add_ipu_core();
 
@@ -473,18 +476,18 @@ static int __init mx31moboard_init_cam(void)
 	if (IS_ERR(pdev))
 		return PTR_ERR(pdev);
 
-	ret = dma_declare_coherent_memory(&pdev->dev,
-					  mx3_camera_base, mx3_camera_base,
-					  MX3_CAMERA_BUF_SIZE,
-					  DMA_MEMORY_EXCLUSIVE);
+	ret = cma_init_reserved_mem(mx3_camera_base, MX3_CAMERA_BUF_SIZE, 0,
+			"mx31cam", &cma);
 	if (ret)
-		goto err;
+		goto out_device_put;
+	dma_contiguous_early_fixup(mx3_camera_base, MX3_CAMERA_BUF_SIZE);
 
 	ret = platform_device_add(pdev);
 	if (ret)
-err:
-		platform_device_put(pdev);
-
+		goto out_device_put;
+	return 0;
+out_device_put:
+	platform_device_put(pdev);
 	return ret;
 
 }

^ permalink raw reply related	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
@ 2018-12-10 19:02     ` Christoph Hellwig
  0 siblings, 0 replies; 13+ messages in thread
From: Christoph Hellwig @ 2018-12-10 19:02 UTC (permalink / raw)
  To: Sascha Hauer
  Cc: Christoph Hellwig, Shawn Guo, Fabio Estevam, linux-arm-kernel,
	iommu, linux-imx, linux-kernel, linux-sh

On Mon, Dec 10, 2018 at 08:26:48AM +0100, Sascha Hauer wrote:
> > the ARM imx27/31 ports and various sh boards use
> > dma_declare_coherent_memory on main memory taken from the memblock
> > allocator.
> > 
> > Is there any good reason these couldn't be switched to CMA areas?
> > Getting rid of these magic dma_declare_coherent_memory area would
> > help making the dma allocator a lot simpler.
> 
> At least for i.MX27/31 I'd say this predates CMA support, so historical
> reasons.

Ok.  Do you still have test hardware for i.MX?  If so I'd love to see
if the patch below works.  Note that this is the brute force approach,
just having a global CMA pool might be a little more elegant.

diff --git a/arch/arm/configs/imx_v4_v5_defconfig b/arch/arm/configs/imx_v4_v5_defconfig
index 8661dd9b064a..88852e7e5e7e 100644
--- a/arch/arm/configs/imx_v4_v5_defconfig
+++ b/arch/arm/configs/imx_v4_v5_defconfig
@@ -185,3 +185,5 @@ CONFIG_NLS_ISO8859_1=y
 CONFIG_NLS_ISO8859_15=m
 CONFIG_FONTS=y
 CONFIG_FONT_8x8=y
+CONFIG_CMA=y
+CONFIG_DMA_CMA=y
diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
index 1ad5736c8fa6..16c8d717316c 100644
--- a/arch/arm/configs/imx_v6_v7_defconfig
+++ b/arch/arm/configs/imx_v6_v7_defconfig
@@ -444,3 +444,5 @@ CONFIG_PROVE_LOCKING=y
 # CONFIG_DEBUG_BUGVERBOSE is not set
 # CONFIG_FTRACE is not set
 # CONFIG_ARM_UNWIND is not set
+CONFIG_CMA=y
+CONFIG_DMA_CMA=y
diff --git a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
index 5169dfba9718..2339a50d5459 100644
--- a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
+++ b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
@@ -23,6 +23,7 @@
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
+#include <linux/cma.h>
 #include <linux/platform_device.h>
 #include <linux/mtd/physmap.h>
 #include <linux/i2c.h>
@@ -35,6 +36,7 @@
 #include <linux/platform_data/asoc-mx27vis.h>
 #include <media/soc_camera.h>
 #include <sound/tlv320aic32x4.h>
+#include <asm/dma-contiguous.h>
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/time.h>
@@ -245,6 +247,7 @@ static phys_addr_t mx2_camera_base __initdata;
 static void __init visstrim_analog_camera_init(void)
 {
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	gpio_set_value(TVP5150_PWDN, 1);
 	ndelay(1);
@@ -257,9 +260,10 @@ static void __init visstrim_analog_camera_init(void)
 	if (IS_ERR(pdev))
 		return;
 
-	dma_declare_coherent_memory(&pdev->dev, mx2_camera_base,
-				    mx2_camera_base, MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (cma_init_reserved_mem(mx2_camera_base, MX2_CAMERA_BUF_SIZE, 0,
+			"visstrim-cam", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base, MX2_CAMERA_BUF_SIZE);
 }
 
 static void __init visstrim_reserve(void)
@@ -440,13 +444,16 @@ static const struct imx_ssi_platform_data visstrim_m10_ssi_pdata __initconst = {
 static void __init visstrim_coda_init(void)
 {
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	pdev = imx27_add_coda();
-	dma_declare_coherent_memory(&pdev->dev,
-				    mx2_camera_base + MX2_CAMERA_BUF_SIZE,
-				    mx2_camera_base + MX2_CAMERA_BUF_SIZE,
-				    MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (IS_ERR(pdev))
+		return;
+	if (cma_init_reserved_mem(mx2_camera_base + MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE, 0, "visstrim-coda", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base + MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE);
 }
 
 /* DMA deinterlace */
@@ -459,21 +466,22 @@ static void __init visstrim_deinterlace_init(void)
 {
 	int ret = -ENOMEM;
 	struct platform_device *pdev = &visstrim_deinterlace;
+	struct cma *cma;
 
 	ret = platform_device_register(pdev);
 
-	dma_declare_coherent_memory(&pdev->dev,
-				    mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
-				    mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
-				    MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (cma_init_reserved_mem(mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE, 0, "visstrim-deinterlace", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE);
 }
 
 /* Emma-PrP for format conversion */
 static void __init visstrim_emmaprp_init(void)
 {
 	struct platform_device *pdev;
-	int ret;
+	struct cma *cma;
 
 	pdev = imx27_add_mx2_emmaprp();
 	if (IS_ERR(pdev))
@@ -483,12 +491,12 @@ static void __init visstrim_emmaprp_init(void)
 	 * Use the same memory area as the analog camera since both
 	 * devices are, by nature, exclusive.
 	 */
-	ret = dma_declare_coherent_memory(&pdev->dev,
-				mx2_camera_base, mx2_camera_base,
-				MX2_CAMERA_BUF_SIZE,
-				DMA_MEMORY_EXCLUSIVE);
-	if (ret)
+	if (cma_init_reserved_mem(mx2_camera_base, MX2_CAMERA_BUF_SIZE, 0,
+			"visstrim-emmaprp", &cma)) {
 		pr_err("Failed to declare memory for emmaprp\n");
+		return;
+	}
+	dma_contiguous_early_fixup(mx2_camera_base, MX2_CAMERA_BUF_SIZE);
 }
 
 /* Audio */
diff --git a/arch/arm/mach-imx/mach-mx31moboard.c b/arch/arm/mach-imx/mach-mx31moboard.c
index 643a3d749703..1c0705ec0ccd 100644
--- a/arch/arm/mach-imx/mach-mx31moboard.c
+++ b/arch/arm/mach-imx/mach-mx31moboard.c
@@ -13,6 +13,7 @@
  */
 
 #include <linux/delay.h>
+#include <linux/cma.h>
 #include <linux/dma-mapping.h>
 #include <linux/gfp.h>
 #include <linux/gpio.h>
@@ -37,6 +38,7 @@
 #include <linux/usb/otg.h>
 #include <linux/usb/ulpi.h>
 
+#include <asm/dma-contiguous.h>
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/time.h>
@@ -466,6 +468,7 @@ static int __init mx31moboard_init_cam(void)
 {
 	int ret;
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	imx31_add_ipu_core();
 
@@ -473,18 +476,18 @@ static int __init mx31moboard_init_cam(void)
 	if (IS_ERR(pdev))
 		return PTR_ERR(pdev);
 
-	ret = dma_declare_coherent_memory(&pdev->dev,
-					  mx3_camera_base, mx3_camera_base,
-					  MX3_CAMERA_BUF_SIZE,
-					  DMA_MEMORY_EXCLUSIVE);
+	ret = cma_init_reserved_mem(mx3_camera_base, MX3_CAMERA_BUF_SIZE, 0,
+			"mx31cam", &cma);
 	if (ret)
-		goto err;
+		goto out_device_put;
+	dma_contiguous_early_fixup(mx3_camera_base, MX3_CAMERA_BUF_SIZE);
 
 	ret = platform_device_add(pdev);
 	if (ret)
-err:
-		platform_device_put(pdev);
-
+		goto out_device_put;
+	return 0;
+out_device_put:
+	platform_device_put(pdev);
 	return ret;
 
 }

^ permalink raw reply related	[flat|nested] 13+ messages in thread
* Re: dma_declare_coherent_memory on main memory
@ 2018-12-10 19:02     ` Christoph Hellwig
  0 siblings, 0 replies; 13+ messages in thread
From: Christoph Hellwig @ 2018-12-10 19:02 UTC (permalink / raw)
  To: Sascha Hauer
  Cc: linux-sh, linux-kernel, iommu, linux-imx, Fabio Estevam,
	Shawn Guo, Christoph Hellwig, linux-arm-kernel

On Mon, Dec 10, 2018 at 08:26:48AM +0100, Sascha Hauer wrote:
> > the ARM imx27/31 ports and various sh boards use
> > dma_declare_coherent_memory on main memory taken from the memblock
> > allocator.
> > 
> > Is there any good reason these couldn't be switched to CMA areas?
> > Getting rid of these magic dma_declare_coherent_memory area would
> > help making the dma allocator a lot simpler.
> 
> At least for i.MX27/31 I'd say this predates CMA support, so historical
> reasons.

Ok.  Do you still have test hardware for i.MX?  If so I'd love to see
if the patch below works.  Note that this is the brute force approach,
just having a global CMA pool might be a little more elegant.

diff --git a/arch/arm/configs/imx_v4_v5_defconfig b/arch/arm/configs/imx_v4_v5_defconfig
index 8661dd9b064a..88852e7e5e7e 100644
--- a/arch/arm/configs/imx_v4_v5_defconfig
+++ b/arch/arm/configs/imx_v4_v5_defconfig
@@ -185,3 +185,5 @@ CONFIG_NLS_ISO8859_1=y
 CONFIG_NLS_ISO8859_15=m
 CONFIG_FONTS=y
 CONFIG_FONT_8x8=y
+CONFIG_CMA=y
+CONFIG_DMA_CMA=y
diff --git a/arch/arm/configs/imx_v6_v7_defconfig b/arch/arm/configs/imx_v6_v7_defconfig
index 1ad5736c8fa6..16c8d717316c 100644
--- a/arch/arm/configs/imx_v6_v7_defconfig
+++ b/arch/arm/configs/imx_v6_v7_defconfig
@@ -444,3 +444,5 @@ CONFIG_PROVE_LOCKING=y
 # CONFIG_DEBUG_BUGVERBOSE is not set
 # CONFIG_FTRACE is not set
 # CONFIG_ARM_UNWIND is not set
+CONFIG_CMA=y
+CONFIG_DMA_CMA=y
diff --git a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
index 5169dfba9718..2339a50d5459 100644
--- a/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
+++ b/arch/arm/mach-imx/mach-imx27_visstrim_m10.c
@@ -23,6 +23,7 @@
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
+#include <linux/cma.h>
 #include <linux/platform_device.h>
 #include <linux/mtd/physmap.h>
 #include <linux/i2c.h>
@@ -35,6 +36,7 @@
 #include <linux/platform_data/asoc-mx27vis.h>
 #include <media/soc_camera.h>
 #include <sound/tlv320aic32x4.h>
+#include <asm/dma-contiguous.h>
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/time.h>
@@ -245,6 +247,7 @@ static phys_addr_t mx2_camera_base __initdata;
 static void __init visstrim_analog_camera_init(void)
 {
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	gpio_set_value(TVP5150_PWDN, 1);
 	ndelay(1);
@@ -257,9 +260,10 @@ static void __init visstrim_analog_camera_init(void)
 	if (IS_ERR(pdev))
 		return;
 
-	dma_declare_coherent_memory(&pdev->dev, mx2_camera_base,
-				    mx2_camera_base, MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (cma_init_reserved_mem(mx2_camera_base, MX2_CAMERA_BUF_SIZE, 0,
+			"visstrim-cam", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base, MX2_CAMERA_BUF_SIZE);
 }
 
 static void __init visstrim_reserve(void)
@@ -440,13 +444,16 @@ static const struct imx_ssi_platform_data visstrim_m10_ssi_pdata __initconst = {
 static void __init visstrim_coda_init(void)
 {
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	pdev = imx27_add_coda();
-	dma_declare_coherent_memory(&pdev->dev,
-				    mx2_camera_base + MX2_CAMERA_BUF_SIZE,
-				    mx2_camera_base + MX2_CAMERA_BUF_SIZE,
-				    MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (IS_ERR(pdev))
+		return;
+	if (cma_init_reserved_mem(mx2_camera_base + MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE, 0, "visstrim-coda", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base + MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE);
 }
 
 /* DMA deinterlace */
@@ -459,21 +466,22 @@ static void __init visstrim_deinterlace_init(void)
 {
 	int ret = -ENOMEM;
 	struct platform_device *pdev = &visstrim_deinterlace;
+	struct cma *cma;
 
 	ret = platform_device_register(pdev);
 
-	dma_declare_coherent_memory(&pdev->dev,
-				    mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
-				    mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
-				    MX2_CAMERA_BUF_SIZE,
-				    DMA_MEMORY_EXCLUSIVE);
+	if (cma_init_reserved_mem(mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE, 0, "visstrim-deinterlace", &cma))
+		return;
+	dma_contiguous_early_fixup(mx2_camera_base + 2 * MX2_CAMERA_BUF_SIZE,
+			MX2_CAMERA_BUF_SIZE);
 }
 
 /* Emma-PrP for format conversion */
 static void __init visstrim_emmaprp_init(void)
 {
 	struct platform_device *pdev;
-	int ret;
+	struct cma *cma;
 
 	pdev = imx27_add_mx2_emmaprp();
 	if (IS_ERR(pdev))
@@ -483,12 +491,12 @@ static void __init visstrim_emmaprp_init(void)
 	 * Use the same memory area as the analog camera since both
 	 * devices are, by nature, exclusive.
 	 */
-	ret = dma_declare_coherent_memory(&pdev->dev,
-				mx2_camera_base, mx2_camera_base,
-				MX2_CAMERA_BUF_SIZE,
-				DMA_MEMORY_EXCLUSIVE);
-	if (ret)
+	if (cma_init_reserved_mem(mx2_camera_base, MX2_CAMERA_BUF_SIZE, 0,
+			"visstrim-emmaprp", &cma)) {
 		pr_err("Failed to declare memory for emmaprp\n");
+		return;
+	}
+	dma_contiguous_early_fixup(mx2_camera_base, MX2_CAMERA_BUF_SIZE);
 }
 
 /* Audio */
diff --git a/arch/arm/mach-imx/mach-mx31moboard.c b/arch/arm/mach-imx/mach-mx31moboard.c
index 643a3d749703..1c0705ec0ccd 100644
--- a/arch/arm/mach-imx/mach-mx31moboard.c
+++ b/arch/arm/mach-imx/mach-mx31moboard.c
@@ -13,6 +13,7 @@
  */
 
 #include <linux/delay.h>
+#include <linux/cma.h>
 #include <linux/dma-mapping.h>
 #include <linux/gfp.h>
 #include <linux/gpio.h>
@@ -37,6 +38,7 @@
 #include <linux/usb/otg.h>
 #include <linux/usb/ulpi.h>
 
+#include <asm/dma-contiguous.h>
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/time.h>
@@ -466,6 +468,7 @@ static int __init mx31moboard_init_cam(void)
 {
 	int ret;
 	struct platform_device *pdev;
+	struct cma *cma;
 
 	imx31_add_ipu_core();
 
@@ -473,18 +476,18 @@ static int __init mx31moboard_init_cam(void)
 	if (IS_ERR(pdev))
 		return PTR_ERR(pdev);
 
-	ret = dma_declare_coherent_memory(&pdev->dev,
-					  mx3_camera_base, mx3_camera_base,
-					  MX3_CAMERA_BUF_SIZE,
-					  DMA_MEMORY_EXCLUSIVE);
+	ret = cma_init_reserved_mem(mx3_camera_base, MX3_CAMERA_BUF_SIZE, 0,
+			"mx31cam", &cma);
 	if (ret)
-		goto err;
+		goto out_device_put;
+	dma_contiguous_early_fixup(mx3_camera_base, MX3_CAMERA_BUF_SIZE);
 
 	ret = platform_device_add(pdev);
 	if (ret)
-err:
-		platform_device_put(pdev);
-
+		goto out_device_put;
+	return 0;
+out_device_put:
+	platform_device_put(pdev);
 	return ret;
 
 }

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply related	[flat|nested] 13+ messages in thread
end of thread, other threads:[~2018-12-10 19:02 UTC | newest]

Thread overview: 13+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-12-07 15:34 dma_declare_coherent_memory on main memory Christoph Hellwig
2018-12-07 15:34 ` Christoph Hellwig
2018-12-07 15:34 ` Christoph Hellwig
2018-12-08 21:25 ` Rob Landley
2018-12-08 21:25   ` Rob Landley
2018-12-08 21:25   ` Rob Landley
2018-12-08 21:25   ` Rob Landley
2018-12-10  7:26 ` Sascha Hauer
2018-12-10  7:26   ` Sascha Hauer
2018-12-10  7:26   ` Sascha Hauer
2018-12-10 19:02   ` Christoph Hellwig
2018-12-10 19:02     ` Christoph Hellwig
2018-12-10 19:02     ` Christoph Hellwig

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.