Re: [PATCH V3 2/4] ACPI/scan: Clean up acpi_check_dma

From: Bjorn Helgaas
Date: Mon Sep 14 2015 - 12:34:23 EST

Next message: Grant Grundler: "Re: [PATCH V3] mmc: block: Add new ioctl to send multi commands"
Previous message: Scott Wood: "Re: [PATCH v2] powerpc32: memcpy/memset: only use dcbz once cache is enabled"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Wed, Aug 26, 2015 at 08:54:37PM +0700, Suravee Suthikulpanit wrote:
> The original name of acpi_check_dma() function does not clearly tell what
> exactly it is checking. Also, returning two boolean values (one to indicate
> device is DMA capability, and the other to inidicate device coherency

s/inidicate/indicate/

> attribute) can be confusing.
>
> So, in order to simplify the function, this patch renames acpi_check_dma()
> to acpi_check_dma_coherency() to clearly indicate the purpose of this
> function, and only returns an integer where -1 means DMA not supported,
> 1 means coherent DMA, and 0 means non-coherent DMA.

I think acpi_check_dma_coherency() is better, but only slightly. It
still doesn't give a hint about the *sense* of the return value. I
think it'd be easier to read if there were two functions, e.g.,

int acpi_dma_supported(struct acpi_device *adev)
{
if (!adev)
return 0;

if (adev->flags.cca_seen)
return 1;

/*
* Per ACPI 6.0 sec 6.2.17, assume devices can do cache-coherent
* DMA on "Intel platforms". Presumably that includes all x86 and
* ia64, and other arches will set CONFIG_ACPI_CCA_REQUIRED=y.
*/
if (!IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED))
return 1;

return 0;
}

int acpi_dma_is_coherent(struct acpi_device *adev)
{
if (!acpi_dma_supported(adev))
return 0;

return adev->flags.coherent_dma;
}

struct platform_device *acpi_create_platform_device(...)
{
...
if (acpi_dma_supported(adev))
pdevinfo.dma_mask = DMA_BIT_MASK(32);

int acpi_bind_one(...)
{
...
if (acpi_dma_supported(acpi_dev))
arch_setup_dma_ops(dev, 0, 0, NULL, retval);

bool device_dma_is_coherent(...)
{
...
return acpi_dma_is_coherent(ACPI_COMPANION(dev)):

> Also, this patch moves the function into drivers/acpi/scan.c since
> it is easier to follow the logic in the acpi_init_coherency() in the
> same file here.
>
> Signed-off-by: Suravee Suthikulpanit <Suravee.Suthikulpanit@xxxxxxx>
> Suggested-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
> Suggested-by: Rafael J. Wysocki <rjw@xxxxxxxxxxxxx>
> CC: Catalin Marinas <catalin.marinas@xxxxxxx>
> CC: Rob Herring <robh+dt@xxxxxxxxxx>
> CC: Will Deacon <will.deacon@xxxxxxx>
> ---
> drivers/acpi/acpi_platform.c | 7 ++++++-
> drivers/acpi/glue.c | 6 +++---
> drivers/acpi/scan.c | 39 +++++++++++++++++++++++++++++++++++++++
> drivers/base/property.c | 10 +++++++---
> include/acpi/acpi_bus.h | 36 ++----------------------------------
> include/linux/acpi.h | 4 ++--
> 6 files changed, 59 insertions(+), 43 deletions(-)
>
> diff --git a/drivers/acpi/acpi_platform.c b/drivers/acpi/acpi_platform.c
> index 06a67d5..0b17a78 100644
> --- a/drivers/acpi/acpi_platform.c
> +++ b/drivers/acpi/acpi_platform.c
> @@ -103,7 +103,12 @@ struct platform_device *acpi_create_platform_device(struct acpi_device *adev)
> pdevinfo.res = resources;
> pdevinfo.num_res = count;
> pdevinfo.fwnode = acpi_fwnode_handle(adev);
> - pdevinfo.dma_mask = acpi_check_dma(adev, NULL) ? DMA_BIT_MASK(32) : 0;
> +
> + if (acpi_check_dma_coherency(adev) < 0)
> + pdevinfo.dma_mask = 0;
> + else
> + pdevinfo.dma_mask = DMA_BIT_MASK(32);
> +
> pdev = platform_device_register_full(&pdevinfo);
> if (IS_ERR(pdev))
> dev_err(&adev->dev, "platform device creation failed: %ld\n",
> diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c
> index b9657af..9be7f0f 100644
> --- a/drivers/acpi/glue.c
> +++ b/drivers/acpi/glue.c
> @@ -168,7 +168,6 @@ int acpi_bind_one(struct device *dev, struct acpi_device *acpi_dev)
> struct list_head *physnode_list;
> unsigned int node_id;
> int retval = -EINVAL;
> - bool coherent;
>
> if (has_acpi_companion(dev)) {
> if (acpi_dev) {
> @@ -225,8 +224,9 @@ int acpi_bind_one(struct device *dev, struct acpi_device *acpi_dev)
> if (!has_acpi_companion(dev))
> ACPI_COMPANION_SET(dev, acpi_dev);
>
> - if (acpi_check_dma(acpi_dev, &coherent))
> - arch_setup_dma_ops(dev, 0, 0, NULL, coherent);
> + retval = acpi_check_dma_coherency(acpi_dev);
> + if (retval >= 0)
> + arch_setup_dma_ops(dev, 0, 0, NULL, retval);
>
> acpi_physnode_link_name(physical_node_name, node_id);
> retval = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
> diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
> index ec25635..1d68289 100644
> --- a/drivers/acpi/scan.c
> +++ b/drivers/acpi/scan.c
> @@ -2182,6 +2182,45 @@ void acpi_free_pnp_ids(struct acpi_device_pnp *pnp)
> kfree(pnp->unique_id);
> }
>
> +/**
> + * acpi_check_dma_coherency - Check DMA coherency for the specified device.
> + * @adev: The pointer to acpi device to check coherency attribute
> + *
> + * Return -ENOTSUPP if DMA is not supported. Otherwise, return 1 if the
> + * device support coherent DMA, or 0 for non-coherent DMA.
> + */
> +int acpi_check_dma_coherency(struct acpi_device *adev)
> +{
> + int ret = -ENOTSUPP;
> +
> + if (!adev)
> + return ret;

There's no need for the local "ret". You can "return -ENOTSUPP" here and
below, which is easier to read than looking back to see where "ret" was
defined and whether it was modified after being initialized.

> +
> + /**
> + * Currently, we only support _CCA=1 (i.e. coherent_dma=1)
> + * This should be equivalent to specifying dma-coherent for
> + * a device in OF.
> + *
> + * For the case when _CCA=0 (i.e. coherent_dma=0 && cca_seen=1),
> + * we have two choices:
> + * 1. Do not support and disable DMA.

I know you didn't write this comment, but do we actually *disable* DMA in
the sense of turning off PCI bus mastering or calling an ACPI method that
disables DMA by this device? I suspect we just don't set up DMA ops and
masks for this device.

> + * 2. Support but rely on arch-specific cache maintenance for
> + * non-coherenje DMA operations.

s/Support but/Support DMA but/ ?
s/non-coherenje/non-coherent/

> + * Currently, we implement case 2 above.
> + *
> + * For the case when _CCA is missing (i.e. cca_seen=0) and
> + * platform specifies ACPI_CCA_REQUIRED, we do not support DMA,
> + * and fallback to arch-specific default handling.
> + *
> + * See acpi_init_coherency() for more info.
> + */
> + if (!adev->flags.cca_seen && IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED))
> + return ret;
> +
> + return adev->flags.coherent_dma;
> +}
> +EXPORT_SYMBOL_GPL(acpi_check_dma_coherency);
> +
> static void acpi_init_coherency(struct acpi_device *adev)
> {
> unsigned long long cca = 0;
> diff --git a/drivers/base/property.c b/drivers/base/property.c
> index f3f6d16..cd45dfc 100644
> --- a/drivers/base/property.c
> +++ b/drivers/base/property.c
> @@ -525,10 +525,14 @@ bool device_dma_is_coherent(struct device *dev)
> {
> bool coherent = false;
>
> - if (IS_ENABLED(CONFIG_OF) && dev->of_node)
> + if (IS_ENABLED(CONFIG_OF) && dev->of_node) {
> coherent = of_dma_is_coherent(dev->of_node);
> - else
> - acpi_check_dma(ACPI_COMPANION(dev), &coherent);
> + } else {
> + int ret = acpi_check_dma_coherency(ACPI_COMPANION(dev));
> +
> + if (ret >= 0)
> + coherent = ret;
> + }
>
> return coherent;
> }
> diff --git a/include/acpi/acpi_bus.h b/include/acpi/acpi_bus.h
> index 7ecb8e4..baf43ea 100644
> --- a/include/acpi/acpi_bus.h
> +++ b/include/acpi/acpi_bus.h
> @@ -382,40 +382,6 @@ struct acpi_device {
> void (*remove)(struct acpi_device *);
> };
>
> -static inline bool acpi_check_dma(struct acpi_device *adev, bool *coherent)
> -{
> - bool ret = false;
> -
> - if (!adev)
> - return ret;
> -
> - /**
> - * Currently, we only support _CCA=1 (i.e. coherent_dma=1)
> - * This should be equivalent to specifyig dma-coherent for
> - * a device in OF.
> - *
> - * For the case when _CCA=0 (i.e. coherent_dma=0 && cca_seen=1),
> - * There are two cases:
> - * case 1. Do not support and disable DMA.
> - * case 2. Support but rely on arch-specific cache maintenance for
> - * non-coherence DMA operations.
> - * Currently, we implement case 2 above.
> - *
> - * For the case when _CCA is missing (i.e. cca_seen=0) and
> - * platform specifies ACPI_CCA_REQUIRED, we do not support DMA,
> - * and fallback to arch-specific default handling.
> - *
> - * See acpi_init_coherency() for more info.
> - */
> - if (adev->flags.coherent_dma ||
> - (adev->flags.cca_seen && IS_ENABLED(CONFIG_ARM64))) {
> - ret = true;
> - if (coherent)
> - *coherent = adev->flags.coherent_dma;
> - }
> - return ret;
> -}
> -
> static inline bool is_acpi_node(struct fwnode_handle *fwnode)
> {
> return fwnode && fwnode->type == FWNODE_ACPI;
> @@ -640,6 +606,8 @@ static inline bool acpi_device_can_poweroff(struct acpi_device *adev)
> return adev->power.states[ACPI_STATE_D3_COLD].flags.valid;
> }
>
> +int acpi_check_dma_coherency(struct acpi_device *adev);
> +
> #else /* CONFIG_ACPI */
>
> static inline int register_acpi_bus_type(void *bus) { return 0; }
> diff --git a/include/linux/acpi.h b/include/linux/acpi.h
> index d2445fa..d350c9e 100644
> --- a/include/linux/acpi.h
> +++ b/include/linux/acpi.h
> @@ -562,9 +562,9 @@ static inline int acpi_device_modalias(struct device *dev,
> return -ENODEV;
> }
>
> -static inline bool acpi_check_dma(struct acpi_device *adev, bool *coherent)
> +static inline int acpi_check_dma_coherency(struct acpi_device *adev)
> {
> - return false;
> + return -ENOTSUPP;
> }
>
> #define ACPI_PTR(_ptr) (NULL)
> --
> 2.1.0
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Next message: Grant Grundler: "Re: [PATCH V3] mmc: block: Add new ioctl to send multi commands"
Previous message: Scott Wood: "Re: [PATCH v2] powerpc32: memcpy/memset: only use dcbz once cache is enabled"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]