linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: laurent.pinchart@ideasonboard.com (Laurent Pinchart)
To: linux-arm-kernel@lists.infradead.org
Subject: [RFC PATCH 4/7] iommu: provide helper function to configure an IOMMU for an of master
Date: Tue, 02 Sep 2014 13:51:54 +0300	[thread overview]
Message-ID: <1485866.XbpZvVaUr8@avalon> (raw)
In-Reply-To: <1409327670-3495-5-git-send-email-will.deacon@arm.com>

Hi Will,

Thank you for the patch.

On Friday 29 August 2014 16:54:27 Will Deacon wrote:
> The generic IOMMU device-tree bindings can be used to add arbitrary OF
> masters to an IOMMU with a compliant binding.
> 
> This patch introduces of_iommu_configure, which does exactly that.
> 
> Signed-off-by: Will Deacon <will.deacon@arm.com>
> ---
>  drivers/iommu/Kconfig    |  2 +-
>  drivers/iommu/of_iommu.c | 36 ++++++++++++++++++++++++++++++++++++
>  include/linux/of_iommu.h |  6 ++++++
>  3 files changed, 43 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig
> index dd5112265cc9..6d13f962f156 100644
> --- a/drivers/iommu/Kconfig
> +++ b/drivers/iommu/Kconfig
> @@ -15,7 +15,7 @@ if IOMMU_SUPPORT
> 
>  config OF_IOMMU
>         def_bool y
> -       depends on OF
> +       depends on OF && IOMMU_API
> 
>  config FSL_PAMU
>  	bool "Freescale IOMMU support"
> diff --git a/drivers/iommu/of_iommu.c b/drivers/iommu/of_iommu.c
> index f9209666157c..a7d3b3a13b83 100644
> --- a/drivers/iommu/of_iommu.c
> +++ b/drivers/iommu/of_iommu.c
> @@ -19,6 +19,7 @@
> 
>  #include <linux/export.h>
>  #include <linux/limits.h>
> +#include <linux/iommu.h>

Pet peeve of mine, how about keeping the headers alphabetically sorted ?

>  #include <linux/of.h>
>  #include <linux/of_iommu.h>
> 
> @@ -90,6 +91,41 @@ int of_get_dma_window(struct device_node *dn, const char
> *prefix, int index, }
>  EXPORT_SYMBOL_GPL(of_get_dma_window);
> 
> +int of_iommu_configure(struct device *dev)
> +{
> +	struct of_phandle_args iommu_spec;
> +	struct bus_type *bus = dev->bus;
> +	const struct iommu_ops *ops = bus->iommu_ops;
> +	int ret = -EINVAL, idx = 0;
> +
> +	if (!iommu_present(bus))
> +		return -ENODEV;
> +
> +	/*
> +	 * We don't currently walk up the tree looking for a parent IOMMU.
> +	 * See the `Notes:' section of
> +	 * Documentation/devicetree/bindings/iommu/iommu.txt
> +	 */
> +	while (!of_parse_phandle_with_args(dev->of_node, "iommus",
> +					   "#iommu-cells", idx,
> +					   &iommu_spec)) {
> +		void *data = of_iommu_get_data(iommu_spec.np);
> +
> +		of_node_put(iommu_spec.np);
> +		if (!ops->add_device_master_ids)
> +			return -ENODEV;

Can't you move this outside of the loop ?

> +
> +		ret = ops->add_device_master_ids(dev, iommu_spec.args_count,
> +						 iommu_spec.args, data);

I'm curious, how do you envision support for devices connected to multiple 
IOMMUs ? IOMMU drivers usually allocate a per-device archdata structure in the 
add_device operation and store it in the dev->archdata.iommu field. How would 
that work with multiple invocations of add_device or add_device_master_ids ?

> +		if (ret)
> +			break;
> +
> +		idx++;
> +	}
> +
> +	return ret;
> +}
> +
>  void __init of_iommu_init(void)
>  {
>  	struct device_node *np;
> diff --git a/include/linux/of_iommu.h b/include/linux/of_iommu.h
> index 29f2f3f88d6a..85c6d1152624 100644
> --- a/include/linux/of_iommu.h
> +++ b/include/linux/of_iommu.h
> @@ -1,6 +1,7 @@
>  #ifndef __OF_IOMMU_H
>  #define __OF_IOMMU_H
> 
> +#include <linux/device.h>
>  #include <linux/of.h>
> 
>  #ifdef CONFIG_OF_IOMMU
> @@ -10,6 +11,7 @@ extern int of_get_dma_window(struct device_node *dn, const
> char *prefix, size_t *size);
> 
>  extern void of_iommu_init(void);
> +extern int of_iommu_configure(struct device *dev);
> 
>  #else
> 
> @@ -21,6 +23,10 @@ static inline int of_get_dma_window(struct device_node
> *dn, const char *prefix, }
> 
>  static inline void of_iommu_init(void) { }
> +static inline int of_iommu_configure(struct device *dev)
> +{
> +	return -ENODEV;
> +}
> 
>  #endif	/* CONFIG_OF_IOMMU */

-- 
Regards,

Laurent Pinchart

  parent reply	other threads:[~2014-09-02 10:51 UTC|newest]

Thread overview: 51+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-08-29 15:54 [RFC PATCH 0/7] Introduce automatic DMA configuration for IOMMU masters Will Deacon
2014-08-29 15:54 ` [RFC PATCH 1/7] iommu: provide early initialisation hook for IOMMU drivers Will Deacon
2014-09-01  7:52   ` Thierry Reding
2014-09-01 14:31     ` Arnd Bergmann
2014-09-01 16:36       ` Will Deacon
2014-09-02  6:56     ` Laurent Pinchart
2014-09-02 14:47   ` Varun Sethi
2014-09-02 15:04     ` Arnd Bergmann
2014-08-29 15:54 ` [RFC PATCH 2/7] dma-mapping: replace set_arch_dma_coherent_ops with arch_setup_dma_ops Will Deacon
2014-09-01 14:27   ` Arnd Bergmann
2014-09-01 16:20     ` Will Deacon
2014-08-29 15:54 ` [RFC PATCH 3/7] iommu: add new iommu_ops callback for adding a device with a set of IDs Will Deacon
2014-09-01  8:13   ` Thierry Reding
2014-09-01 14:39     ` Arnd Bergmann
2014-09-01 16:34       ` Will Deacon
2014-09-01 17:18         ` Arnd Bergmann
2014-08-29 15:54 ` [RFC PATCH 4/7] iommu: provide helper function to configure an IOMMU for an of master Will Deacon
2014-09-01  8:29   ` Thierry Reding
2014-09-01 14:46     ` Arnd Bergmann
2014-09-01 16:40       ` Will Deacon
2014-09-01 20:18         ` Arnd Bergmann
2014-09-02 10:03           ` Will Deacon
2014-09-02 12:15             ` Arnd Bergmann
2014-09-02 13:05               ` Will Deacon
2014-09-02 14:01                 ` Arnd Bergmann
2014-09-02 20:59                   ` jroedel at suse.de
2014-09-03  9:45                     ` Will Deacon
2014-09-02 15:03               ` Varun Sethi
2014-09-02 15:08                 ` Arnd Bergmann
2014-09-02 10:23       ` Laurent Pinchart
2014-09-02 10:51   ` Laurent Pinchart [this message]
2014-09-02 11:03     ` Will Deacon
2014-09-02 19:08       ` Laurent Pinchart
2014-09-02 14:55   ` Varun Sethi
2014-08-29 15:54 ` [RFC PATCH 5/7] dma-mapping: detect and configure IOMMU in of_dma_configure Will Deacon
2014-09-01 14:53   ` Arnd Bergmann
2014-08-29 15:54 ` [RFC PATCH 6/7] arm: call iommu_init before of_platform_populate Will Deacon
2014-08-29 15:54 ` [RFC PATCH 7/7] arm: dma-mapping: plumb our iommu mapping ops into arch_setup_dma_ops Will Deacon
2014-09-02  6:26 ` [RFC PATCH 0/7] Introduce automatic DMA configuration for IOMMU masters Marek Szyprowski
2014-09-02  8:31   ` Will Deacon
2014-09-02  8:48     ` Marek Szyprowski
2014-09-02  8:56       ` Arnd Bergmann
2014-09-02 10:42         ` Marek Szyprowski
2014-09-02 10:57           ` Will Deacon
2014-09-02 12:24             ` Marek Szyprowski
2014-09-02 12:43               ` Arnd Bergmann
2014-09-02 21:50                 ` Laurent Pinchart
2014-09-02 12:22           ` Arnd Bergmann
2014-09-02 12:30             ` Marek Szyprowski
2014-09-02 12:46               ` Arnd Bergmann
2014-09-02 13:11                 ` Marek Szyprowski

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1485866.XbpZvVaUr8@avalon \
    --to=laurent.pinchart@ideasonboard.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).