All of lore.kernel.org
 help / color / mirror / Atom feed
From: Marek Szyprowski <m.szyprowski@samsung.com>
To: 'KyongHo Cho' <pullip.cho@samsung.com>
Cc: linux-arm-kernel@lists.infradead.org,
	linaro-mm-sig@lists.linaro.org, linux-mm@kvack.org,
	linux-arch@vger.kernel.org,
	'Kyungmin Park' <kyungmin.park@samsung.com>,
	'Russell King - ARM Linux' <linux@arm.linux.org.uk>,
	'Joerg Roedel' <joro@8bytes.org>, 'Arnd Bergmann' <arnd@arndb.de>
Subject: RE: [Linaro-mm-sig] [PATCH 3/8] ARM: dma-mapping: use asm-generic/dma-mapping-common.h
Date: Tue, 21 Jun 2011 13:47:03 +0200	[thread overview]
Message-ID: <002501cc3008$f000d600$d0028200$%szyprowski@samsung.com> (raw)
In-Reply-To: <BANLkTimHE2jzQAav465WaG3iWVeHPyNRNQ@mail.gmail.com>

Hello,

On Monday, June 20, 2011 4:33 PM KyongHo Cho wrote:

> On Mon, Jun 20, 2011 at 4:50 PM, Marek Szyprowski
> <m.szyprowski@samsung.com> wrote:
> > +static inline void set_dma_ops(struct device *dev, struct dma_map_ops
> *ops)
> > +{
> > +       dev->archdata.dma_ops = ops;
> > +}
> > +
> 
> Who calls set_dma_ops()?
> In the mach. initialization part?

Yes, some board, machine or device bus initialization code is supposed to
call this function. Just 'git grep dma_set_ops' and you will see. In my
patch series one of the clients of set_dma_ops function is dmabounce 
framework (it is called in dmabounce_register_dev() function).

> What if a device driver does not want to use arch's dma_map_ops
> when machine init procedure set a dma_map_ops?

Could you elaborate on this case? The whole point of dma-mapping framework
is to hide the implementation of DMA mapping operation from the driver. 
The driver should never fiddle with dma map ops directly.

> Even though, may arch defiens their dma_map_ops in archdata of device
> structure,
> I think it is not a good idea that is device structure contains a
> pointer to dma_map_ops
> that may not be common to all devices in a board.

It is up to the board/bus startup code to set dma ops correctly.

> I also think that it is better to attach and to detach dma_map_ops
> dynamically.

What's the point of such operations? Why do you want to change dma
mapping methods in runtime?

> Moreover, a mapping is not permanent in our Exynos platform
> because a System MMU may be turned off while runtime.

This is theoretically possible. The System MMU (Samsung IOMMU
controller) driver can change dma_map_ops back to NULL on remove moving
back the client device to generic ARM dma mapping implementation.

> DMA API must come with IOMMU API to initialize IOMMU in runtime.

I don't understand what's the problem here. 

Best regards
-- 
Marek Szyprowski
Samsung Poland R&D Center



--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

WARNING: multiple messages have this Message-ID (diff)
From: Marek Szyprowski <m.szyprowski@samsung.com>
To: 'KyongHo Cho' <pullip.cho@samsung.com>
Cc: linux-arm-kernel@lists.infradead.org,
	linaro-mm-sig@lists.linaro.org, linux-mm@kvack.org,
	linux-arch@vger.kernel.org,
	'Kyungmin Park' <kyungmin.park@samsung.com>,
	'Russell King - ARM Linux' <linux@arm.linux.org.uk>,
	'Joerg Roedel' <joro@8bytes.org>, 'Arnd Bergmann' <arnd@arndb.de>
Subject: RE: [Linaro-mm-sig] [PATCH 3/8] ARM: dma-mapping: use asm-generic/dma-mapping-common.h
Date: Tue, 21 Jun 2011 13:47:03 +0200	[thread overview]
Message-ID: <002501cc3008$f000d600$d0028200$%szyprowski@samsung.com> (raw)
Message-ID: <20110621114703.sZv3g6IUBDI5etnSyRxVXKYmmov4ci76W0fJDYnGRco@z> (raw)
In-Reply-To: <BANLkTimHE2jzQAav465WaG3iWVeHPyNRNQ@mail.gmail.com>

Hello,

On Monday, June 20, 2011 4:33 PM KyongHo Cho wrote:

> On Mon, Jun 20, 2011 at 4:50 PM, Marek Szyprowski
> <m.szyprowski@samsung.com> wrote:
> > +static inline void set_dma_ops(struct device *dev, struct dma_map_ops
> *ops)
> > +{
> > +       dev->archdata.dma_ops = ops;
> > +}
> > +
> 
> Who calls set_dma_ops()?
> In the mach. initialization part?

Yes, some board, machine or device bus initialization code is supposed to
call this function. Just 'git grep dma_set_ops' and you will see. In my
patch series one of the clients of set_dma_ops function is dmabounce 
framework (it is called in dmabounce_register_dev() function).

> What if a device driver does not want to use arch's dma_map_ops
> when machine init procedure set a dma_map_ops?

Could you elaborate on this case? The whole point of dma-mapping framework
is to hide the implementation of DMA mapping operation from the driver. 
The driver should never fiddle with dma map ops directly.

> Even though, may arch defiens their dma_map_ops in archdata of device
> structure,
> I think it is not a good idea that is device structure contains a
> pointer to dma_map_ops
> that may not be common to all devices in a board.

It is up to the board/bus startup code to set dma ops correctly.

> I also think that it is better to attach and to detach dma_map_ops
> dynamically.

What's the point of such operations? Why do you want to change dma
mapping methods in runtime?

> Moreover, a mapping is not permanent in our Exynos platform
> because a System MMU may be turned off while runtime.

This is theoretically possible. The System MMU (Samsung IOMMU
controller) driver can change dma_map_ops back to NULL on remove moving
back the client device to generic ARM dma mapping implementation.

> DMA API must come with IOMMU API to initialize IOMMU in runtime.

I don't understand what's the problem here. 

Best regards
-- 
Marek Szyprowski
Samsung Poland R&D Center




WARNING: multiple messages have this Message-ID (diff)
From: m.szyprowski@samsung.com (Marek Szyprowski)
To: linux-arm-kernel@lists.infradead.org
Subject: [Linaro-mm-sig] [PATCH 3/8] ARM: dma-mapping: use asm-generic/dma-mapping-common.h
Date: Tue, 21 Jun 2011 13:47:03 +0200	[thread overview]
Message-ID: <002501cc3008$f000d600$d0028200$%szyprowski@samsung.com> (raw)
In-Reply-To: <BANLkTimHE2jzQAav465WaG3iWVeHPyNRNQ@mail.gmail.com>

Hello,

On Monday, June 20, 2011 4:33 PM KyongHo Cho wrote:

> On Mon, Jun 20, 2011 at 4:50 PM, Marek Szyprowski
> <m.szyprowski@samsung.com> wrote:
> > +static inline void set_dma_ops(struct device *dev, struct dma_map_ops
> *ops)
> > +{
> > + ? ? ? dev->archdata.dma_ops = ops;
> > +}
> > +
> 
> Who calls set_dma_ops()?
> In the mach. initialization part?

Yes, some board, machine or device bus initialization code is supposed to
call this function. Just 'git grep dma_set_ops' and you will see. In my
patch series one of the clients of set_dma_ops function is dmabounce 
framework (it is called in dmabounce_register_dev() function).

> What if a device driver does not want to use arch's dma_map_ops
> when machine init procedure set a dma_map_ops?

Could you elaborate on this case? The whole point of dma-mapping framework
is to hide the implementation of DMA mapping operation from the driver. 
The driver should never fiddle with dma map ops directly.

> Even though, may arch defiens their dma_map_ops in archdata of device
> structure,
> I think it is not a good idea that is device structure contains a
> pointer to dma_map_ops
> that may not be common to all devices in a board.

It is up to the board/bus startup code to set dma ops correctly.

> I also think that it is better to attach and to detach dma_map_ops
> dynamically.

What's the point of such operations? Why do you want to change dma
mapping methods in runtime?

> Moreover, a mapping is not permanent in our Exynos platform
> because a System MMU may be turned off while runtime.

This is theoretically possible. The System MMU (Samsung IOMMU
controller) driver can change dma_map_ops back to NULL on remove moving
back the client device to generic ARM dma mapping implementation.

> DMA API must come with IOMMU API to initialize IOMMU in runtime.

I don't understand what's the problem here. 

Best regards
-- 
Marek Szyprowski
Samsung Poland R&D Center

  reply	other threads:[~2011-06-21 11:47 UTC|newest]

Thread overview: 201+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-06-20  7:50 [PATCH/RFC 0/8] ARM: DMA-mapping framework redesign Marek Szyprowski
2011-06-20  7:50 ` Marek Szyprowski
2011-06-20  7:50 ` Marek Szyprowski
2011-06-20  7:50 ` [PATCH 1/8] ARM: dma-mapping: remove offset parameter to prepare for generic dma_ops Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  8:35   ` Michal Nazarewicz
2011-06-20  8:35     ` Michal Nazarewicz
2011-06-20  8:35     ` Michal Nazarewicz
2011-06-20 10:46     ` Marek Szyprowski
2011-06-20 10:46       ` Marek Szyprowski
2011-06-20 10:46       ` Marek Szyprowski
2011-07-03 15:28   ` Russell King - ARM Linux
2011-07-03 15:28     ` Russell King - ARM Linux
2011-07-03 15:28     ` Russell King - ARM Linux
2011-07-26 12:56     ` Marek Szyprowski
2011-07-26 12:56       ` Marek Szyprowski
2011-07-26 12:56       ` Marek Szyprowski
2011-06-20  7:50 ` [PATCH 2/8] ARM: dma-mapping: implement dma_map_single on top of dma_map_page Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20 14:39   ` Russell King - ARM Linux
2011-06-20 14:39     ` Russell King - ARM Linux
2011-06-20 14:39     ` Russell King - ARM Linux
2011-06-20 15:15     ` Marek Szyprowski
2011-06-20 15:15       ` Marek Szyprowski
2011-06-20 15:15       ` Marek Szyprowski
2011-06-24 15:24       ` Arnd Bergmann
2011-06-24 15:24         ` Arnd Bergmann
2011-06-24 15:24         ` Arnd Bergmann
2011-06-27 14:29         ` Marek Szyprowski
2011-06-27 14:29           ` Marek Szyprowski
2011-06-27 14:29           ` Marek Szyprowski
2011-06-27 14:53           ` Arnd Bergmann
2011-06-27 14:53             ` Arnd Bergmann
2011-06-27 14:53             ` Arnd Bergmann
2011-06-27 15:06             ` Marek Szyprowski
2011-06-27 15:06               ` Marek Szyprowski
2011-06-27 15:06               ` Marek Szyprowski
2011-06-20  7:50 ` [PATCH 3/8] ARM: dma-mapping: use asm-generic/dma-mapping-common.h Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20 14:33   ` [Linaro-mm-sig] " KyongHo Cho
2011-06-20 14:33     ` KyongHo Cho
2011-06-20 14:33     ` KyongHo Cho
2011-06-21 11:47     ` Marek Szyprowski [this message]
2011-06-21 11:47       ` Marek Szyprowski
2011-06-21 11:47       ` Marek Szyprowski
2011-06-24  8:39       ` 'Joerg Roedel'
2011-06-24  8:39         ` 'Joerg Roedel'
2011-06-24  8:39         ` 'Joerg Roedel'
2011-06-24 15:36   ` Arnd Bergmann
2011-06-24 15:36     ` Arnd Bergmann
2011-06-24 15:36     ` Arnd Bergmann
2011-06-24 15:36     ` Arnd Bergmann
2011-06-27 12:18     ` Marek Szyprowski
2011-06-27 12:18       ` Marek Szyprowski
2011-06-27 12:18       ` Marek Szyprowski
2011-06-27 13:19       ` Arnd Bergmann
2011-06-27 13:19         ` Arnd Bergmann
2011-06-27 13:19         ` Arnd Bergmann
2011-07-07 12:09         ` Lennert Buytenhek
2011-07-07 12:09           ` Lennert Buytenhek
2011-07-07 12:09           ` Lennert Buytenhek
2011-07-07 12:38           ` Russell King - ARM Linux
2011-07-07 12:38             ` Russell King - ARM Linux
2011-07-07 12:38             ` Russell King - ARM Linux
2011-07-15  0:10             ` Lennert Buytenhek
2011-07-15  0:10               ` Lennert Buytenhek
2011-07-15  0:10               ` Lennert Buytenhek
2011-07-15  9:27               ` Russell King - ARM Linux
2011-07-15  9:27                 ` Russell King - ARM Linux
2011-07-15  9:27                 ` Russell King - ARM Linux
2011-07-15 21:53                 ` Lennert Buytenhek
2011-07-15 21:53                   ` Lennert Buytenhek
2011-07-15 21:53                   ` Lennert Buytenhek
2011-06-20  7:50 ` [PATCH 4/8] ARM: dma-mapping: implement dma sg methods on top of generic dma ops Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20 14:37   ` KyongHo Cho
2011-06-20 14:37     ` KyongHo Cho
2011-06-20 14:37     ` KyongHo Cho
2011-06-20 14:40   ` Russell King - ARM Linux
2011-06-20 14:40     ` Russell King - ARM Linux
2011-06-20 14:40     ` Russell King - ARM Linux
2011-06-20 15:23     ` Marek Szyprowski
2011-06-20 15:23       ` Marek Szyprowski
2011-06-20 15:23       ` Marek Szyprowski
2011-06-20  7:50 ` [PATCH 5/8] ARM: dma-mapping: move all dma bounce code to separate dma ops structure Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20 14:42   ` Russell King - ARM Linux
2011-06-20 14:42     ` Russell King - ARM Linux
2011-06-20 14:42     ` Russell King - ARM Linux
2011-06-20 15:31     ` Marek Szyprowski
2011-06-20 15:31       ` Marek Szyprowski
2011-06-20 15:31       ` Marek Szyprowski
2011-06-24 15:47       ` Arnd Bergmann
2011-06-24 15:47         ` Arnd Bergmann
2011-06-24 15:47         ` Arnd Bergmann
2011-06-27 14:20         ` Marek Szyprowski
2011-06-27 14:20           ` Marek Szyprowski
2011-06-27 14:20           ` Marek Szyprowski
2011-06-20  7:50 ` [PATCH 6/8] ARM: dma-mapping: remove redundant code and cleanup Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50 ` [PATCH 7/8] common: dma-mapping: change alloc/free_coherent method to more generic alloc/free_attrs Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20 14:45   ` KyongHo Cho
2011-06-20 14:45     ` KyongHo Cho
2011-06-20 14:45     ` KyongHo Cho
2011-06-20 15:06     ` Russell King - ARM Linux
2011-06-20 15:06       ` Russell King - ARM Linux
2011-06-20 15:06       ` Russell King - ARM Linux
2011-06-20 15:06       ` Russell King - ARM Linux
2011-06-20 15:14       ` [Linaro-mm-sig] " KyongHo Cho
2011-06-20 15:14         ` KyongHo Cho
2011-06-20 15:14         ` KyongHo Cho
2011-06-21 11:23     ` Marek Szyprowski
2011-06-21 11:23       ` Marek Szyprowski
2011-06-21 11:23       ` Marek Szyprowski
2011-06-22  0:00       ` [Linaro-mm-sig] " KyongHo Cho
2011-06-22  0:00         ` KyongHo Cho
2011-06-22  0:00         ` KyongHo Cho
2011-06-24  7:20         ` Marek Szyprowski
2011-06-24  7:20           ` Marek Szyprowski
2011-06-24  7:20           ` Marek Szyprowski
2011-06-24 15:51   ` Arnd Bergmann
2011-06-24 15:51     ` Arnd Bergmann
2011-06-24 15:51     ` Arnd Bergmann
2011-06-24 16:15     ` James Bottomley
2011-06-24 16:15       ` James Bottomley
2011-06-24 16:15       ` James Bottomley
2011-06-24 16:23       ` Arnd Bergmann
2011-06-24 16:23         ` Arnd Bergmann
2011-06-24 16:23         ` Arnd Bergmann
2011-06-27 12:23     ` Marek Szyprowski
2011-06-27 12:23       ` Marek Szyprowski
2011-06-27 12:23       ` Marek Szyprowski
2011-06-27 12:23       ` Marek Szyprowski
2011-06-27 13:22       ` Arnd Bergmann
2011-06-27 13:22         ` Arnd Bergmann
2011-06-27 13:22         ` Arnd Bergmann
2011-06-27 13:30         ` Marek Szyprowski
2011-06-27 13:30           ` Marek Szyprowski
2011-06-27 13:30           ` Marek Szyprowski
2011-06-24 15:53   ` Arnd Bergmann
2011-06-24 15:53     ` Arnd Bergmann
2011-06-24 15:53     ` Arnd Bergmann
2011-06-27 14:41     ` Marek Szyprowski
2011-06-27 14:41       ` Marek Szyprowski
2011-06-27 14:41       ` Marek Szyprowski
2011-06-20  7:50 ` [PATCH 8/8] ARM: dma-mapping: use alloc, mmap, free from dma_ops Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-20  7:50   ` Marek Szyprowski
2011-06-22  6:53   ` [Linaro-mm-sig] " KyongHo Cho
2011-06-22  6:53     ` KyongHo Cho
2011-06-22  6:53     ` KyongHo Cho
2011-06-22  4:53 ` [Linaro-mm-sig] [PATCH/RFC 0/8] ARM: DMA-mapping framework redesign Subash Patel
2011-06-22  4:53   ` Subash Patel
2011-06-22  4:53   ` Subash Patel
2011-06-22  6:59   ` Marek Szyprowski
2011-06-22  6:59     ` Marek Szyprowski
2011-06-22  6:59     ` Marek Szyprowski
2011-06-22  8:53     ` Subash Patel
2011-06-22  8:53       ` Subash Patel
2011-06-22  8:53       ` Subash Patel
2011-06-22  9:27       ` Marek Szyprowski
2011-06-22  9:27         ` Marek Szyprowski
2011-06-22  9:27         ` Marek Szyprowski
2011-06-22 16:00         ` Jordan Crouse
2011-06-22 16:00           ` Jordan Crouse
2011-06-22 16:00           ` Jordan Crouse
2011-06-23 13:09           ` Subash Patel
2011-06-23 13:09             ` Subash Patel
2011-06-23 13:09             ` Subash Patel
2011-06-23 16:24             ` Michael K. Edwards
2011-06-23 16:24               ` Michael K. Edwards
2011-06-23 16:24               ` Michael K. Edwards
2011-06-23 22:09               ` Michael K. Edwards
2011-06-23 22:09                 ` Michael K. Edwards
2011-06-23 22:09                 ` Michael K. Edwards
2011-06-25  5:23                 ` Jonathan Morton
2011-06-25  5:23                   ` Jonathan Morton
2011-06-25  5:23                   ` Jonathan Morton
2011-06-25  9:55                   ` Michael K. Edwards
2011-06-25  9:55                     ` Michael K. Edwards
2011-06-25  9:55                     ` Michael K. Edwards
2011-06-26  0:06                     ` Jonathan Morton
2011-06-26  0:06                       ` Jonathan Morton
2011-06-26  0:06                       ` Jonathan Morton
2011-06-24 15:20           ` Arnd Bergmann
2011-06-24 15:20             ` Arnd Bergmann
2011-06-24 15:20             ` Arnd Bergmann
2011-06-24  9:18 ` Joerg Roedel
2011-06-24  9:18   ` Joerg Roedel
2011-06-24  9:18   ` Joerg Roedel
2011-06-24 14:26   ` Marek Szyprowski
2011-06-24 14:26     ` Marek Szyprowski
2011-06-24 14:26     ` 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='002501cc3008$f000d600$d0028200$%szyprowski@samsung.com' \
    --to=m.szyprowski@samsung.com \
    --cc=arnd@arndb.de \
    --cc=joro@8bytes.org \
    --cc=kyungmin.park@samsung.com \
    --cc=linaro-mm-sig@lists.linaro.org \
    --cc=linux-arch@vger.kernel.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-mm@kvack.org \
    --cc=linux@arm.linux.org.uk \
    --cc=pullip.cho@samsung.com \
    /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 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.