RE: [PATCH V2 05/12] ARM: SAMSUNG: Add common DMA operations

From: Boojin Kim <boojin.kim@samsung.com>
To: 'Grant Likely' <grant.likely@secretlab.ca>,
	'Kukjin Kim' <kgene.kim@samsung.com>
Cc: linux-arm-kernel@lists.infradead.org,
	linux-samsung-soc@vger.kernel.org,
	'Vinod Koul' <vinod.koul@intel.com>,
	'Dan Williams' <dan.j.williams@intel.com>,
	'Jassi Brar' <jassisinghbrar@gmail.com>,
	'Liam Girdwood' <lrg@ti.com>,
	'Mark Brown' <broonie@opensource.wolfsonmicro.com>,
	'Linus Walleij' <linus.walleij@linaro.org>
Subject: RE: [PATCH V2 05/12] ARM: SAMSUNG: Add common DMA operations
Date: Sat, 16 Jul 2011 09:39:31 +0900	[thread overview]
Message-ID: <001301cc4350$d4c09fa0$7e41dee0$%kim@samsung.com> (raw)
In-Reply-To: <20110715025347.GU2927@ponder.secretlab.ca>

Grant Likely wrote:

> On Wed, Jul 13, 2011 at 05:47:30PM +0900, Kukjin Kim wrote:
> > From: Boojin Kim <boojin.kim@samsung.com>
> >
> > This patch adds common DMA operations which are used for Samsung DMA
> > drivers. Currently there are two types of DMA driver for Samsung SoCs.
> > The one is S3C-DMA for S3C SoCs and the other is PL330-DMA for S5P SoCs.
> > This patch provides funcion pointers for common DMA operations to DMA
> > client driver like SPI and Audio. It makes DMA client drivers support
> > multi-platform.
> > In addition, this common DMA operations implement the shared actions
> > that are needed for DMA client driver. For example shared actions are
> > filter() function for dma_request_channel() and parameter passing for
> > device_prep_slave_sg().
> >
> > Signed-off-by: Boojin Kim <boojin.kim@samsung.com>
> > Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
> > ---
> >  arch/arm/mach-s3c2410/include/mach/dma.h       |    3 +-
> >  arch/arm/plat-samsung/Makefile                 |    6 +-
> >  arch/arm/plat-samsung/dma-ops.c                |  131
> +++++++++++++++++++++++
> >  arch/arm/plat-samsung/include/plat/dma-ops.h   |   47 +++++++++
> >  arch/arm/plat-samsung/include/plat/dma-pl330.h |    3 +
> >  arch/arm/plat-samsung/include/plat/dma.h       |    2 +-
> >  arch/arm/plat-samsung/s3c-dma-ops.c            |  132
> ++++++++++++++++++++++++
> >  7 files changed, 320 insertions(+), 4 deletions(-)
> >  create mode 100644 arch/arm/plat-samsung/dma-ops.c
> >  create mode 100644 arch/arm/plat-samsung/include/plat/dma-ops.h
> >  create mode 100644 arch/arm/plat-samsung/s3c-dma-ops.c
> >
> > diff --git a/arch/arm/mach-s3c2410/include/mach/dma.h b/arch/arm/mach-
> s3c2410/include/mach/dma.h
> > index b2b2a5b..71a662d 100644
> > --- a/arch/arm/mach-s3c2410/include/mach/dma.h
> > +++ b/arch/arm/mach-s3c2410/include/mach/dma.h
> > @@ -13,7 +13,6 @@
> >  #ifndef __ASM_ARCH_DMA_H
> >  #define __ASM_ARCH_DMA_H __FILE__
> >
> > -#include <plat/dma.h>
> >  #include <linux/sysdev.h>
> >
> >  #define MAX_DMA_TRANSFER_SIZE   0x100000 /* Data Unit is half word  */
> > @@ -51,6 +50,8 @@ enum dma_ch {
> >  	DMACH_MAX,		/* the end entry */
> >  };
> >
> > +#include <plat/dma.h>
> > +
> >  #define DMACH_LOW_LEVEL	(1<<28)	/* use this to specifiy hardware
> ch no */
> >
> >  /* we have 4 dma channels */
> > diff --git a/arch/arm/plat-samsung/Makefile b/arch/arm/plat-
> samsung/Makefile
> > index 53eb15b..9ecf2aa 100644
> > --- a/arch/arm/plat-samsung/Makefile
> > +++ b/arch/arm/plat-samsung/Makefile
> > @@ -62,9 +62,11 @@ obj-$(CONFIG_SAMSUNG_DEV_PWM)	+= dev-pwm.o
> >
> >  # DMA support
> >
> > -obj-$(CONFIG_S3C_DMA)		+= dma.o
> > +obj-$(CONFIG_S3C_DMA)		+= dma.o s3c-dma-ops.o
> >
> > -obj-$(CONFIG_S3C_PL330_DMA)	+= s3c-pl330.o
> > +obj-$(CONFIG_DMADEV_PL330)	+= dma-ops.o
> > +
> > +obj-$(CONFIG_S3C_PL330_DMA)	+= s3c-pl330.o s3c-dma-ops.o
> >
> >  # PM support
> >
> > diff --git a/arch/arm/plat-samsung/dma-ops.c b/arch/arm/plat-
> samsung/dma-ops.c
> > new file mode 100644
> > index 0000000..94a9d33
> > --- /dev/null
> > +++ b/arch/arm/plat-samsung/dma-ops.c
> > @@ -0,0 +1,131 @@
> > +/* linux/arch/arm/plat-samsung/dma-ops.c
> > + *
> > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > + *		http://www.samsung.com
> > + *
> > + * Samsung DMA Operations
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License version 2 as
> > + * published by the Free Software Foundation.
> > +*/
> > +
> > +#include <linux/kernel.h>
> > +#include <linux/errno.h>
> > +#include <linux/dmaengine.h>
> > +#include <linux/amba/pl330.h>
> > +
> > +#include <mach/dma.h>
> > +
> > +static bool filter(struct dma_chan *chan, void *param)
> > +{
> > +	struct dma_pl330_peri *peri = (struct dma_pl330_peri *)chan-
> >private;
> 
> Is this code pl330 specific?  If so, 'pl330' should probably be
> referenced in the function name.

I will address your comment.

> 
> > +	unsigned dma_ch = (unsigned)param;
> > +
> > +	if (peri->peri_id != dma_ch)
> > +		return false;
> > +
> > +	return true;
> > +}
> > +
> > +static unsigned dma_request(enum dma_ch dma_ch, struct samsung_dma_info
> *info)
> > +{
> > +	struct dma_chan *chan;
> > +	dma_cap_mask_t mask;
> > +	struct dma_slave_config slave_config;
> > +
> > +	dma_cap_zero(mask);
> > +	dma_cap_set(info->cap, mask);
> > +
> > +	chan = dma_request_channel(mask, filter, (void *)dma_ch);
> > +
> > +	if (info->direction == DMA_FROM_DEVICE) {
> > +		memset(&slave_config, 0, sizeof(struct dma_slave_config));
> > +		slave_config.direction = info->direction;
> > +		slave_config.src_addr = info->fifo;
> > +		slave_config.src_addr_width = info->width;
> > +		dmaengine_slave_config(chan, &slave_config);
> > +	} else if (info->direction == DMA_TO_DEVICE) {
> > +		memset(&slave_config, 0, sizeof(struct dma_slave_config));
> > +		slave_config.direction = info->direction;
> > +		slave_config.dst_addr = info->fifo;
> > +		slave_config.dst_addr_width = info->width;
> > +		dmaengine_slave_config(chan, &slave_config);
> > +	}
> > +
> > +	return (unsigned)chan;
> > +}
> > +
> > +static int dma_release(unsigned ch, struct s3c2410_dma_client *client)
> > +{
> > +	dma_release_channel((struct dma_chan *)ch);
> > +
> > +	return 0;
> > +}
> > +
> > +static int dma_prepare(unsigned ch, struct samsung_dma_prep_info *info)
> > +{
> > +	struct scatterlist sg;
> > +	struct dma_chan *chan = (struct dma_chan *)ch;
> > +	struct dma_async_tx_descriptor *desc;
> > +
> > +	switch (info->cap) {
> > +	case DMA_SLAVE:
> > +		sg_init_table(&sg, 1);
> > +		sg_dma_len(&sg) = info->len;
> > +		sg_set_page(&sg, pfn_to_page(PFN_DOWN(info->buf)),
> > +			    info->len, offset_in_page(info->buf));
> > +		sg_dma_address(&sg) = info->buf;
> > +
> > +		desc = chan->device->device_prep_slave_sg(chan,
> > +			&sg, 1, info->direction, DMA_PREP_INTERRUPT);
> > +		break;
> > +	case DMA_CYCLIC:
> > +		desc = chan->device->device_prep_dma_cyclic(chan,
> > +			info->buf, info->len, info->period,
info->direction);
> > +		break;
> > +	default:
> > +		dev_err(&chan->dev->device, "unsupported format\n");
> > +		return -EFAULT;
> > +	}
> > +
> > +	if (!desc) {
> > +		dev_err(&chan->dev->device, "cannot prepare cyclic dma\n");
> > +		return -EFAULT;
> > +	}
> > +
> > +	desc->callback = info->fp;
> > +	desc->callback_param = info->fp_param;
> > +
> > +	dmaengine_submit((struct dma_async_tx_descriptor *)desc);
> > +
> > +	return 0;
> > +}
> > +
> > +static inline int dma_trigger(unsigned ch)
> > +{
> > +	dma_async_issue_pending((struct dma_chan *)ch);
> > +
> > +	return 0;
> > +}
> > +
> > +static inline int dma_flush(unsigned ch)
> > +{
> > +	return dmaengine_terminate_all((struct dma_chan *)ch);
> > +}
> > +
> > +static struct samsung_dma_ops dmaeng_ops = {
> > +	.request	= dma_request,
> > +	.release	= dma_release,
> > +	.prepare	= dma_prepare,
> > +	.trigger	= dma_trigger,
> > +	.started	= NULL,
> > +	.flush		= dma_flush,
> > +	.stop		= dma_flush,
> > +};
> 
> Even though these function are all local statics, you should use a
> samsung prefix to avoid namespace collisions.  So, they should be
> named something like: samsung_dmaeng_ops, samsung_dmaeng_request(),
> samsung_dmaeng_release(), etc.  The ops structure and the functions
> should have the same prefix.

I will address your comment.

> 
> > +
> > +void *samsung_dma_get_ops(void)
> > +{
> > +	return (void *)&dmaeng_ops;
> > +}
> > +EXPORT_SYMBOL(samsung_dma_get_ops);
> 
> If all that is needed is a reference to the dma ops, then you could
> simply export samsung_dmaeng_ops() without a separate function.

Grant, Thanks for your comments. I can't understand this comment well.
Do you mean to change function name from 'samsung_dma_get_ops()' to
'samsung_dmaeng_ops()' ?
Or export 'dmaeng_ops' variable instead of making this
'samsung_dma_get_ops()' function

> 
> > diff --git a/arch/arm/plat-samsung/include/plat/dma-ops.h
> b/arch/arm/plat-samsung/include/plat/dma-ops.h
> > new file mode 100644
> > index 0000000..eea4130
> > --- /dev/null
> > +++ b/arch/arm/plat-samsung/include/plat/dma-ops.h
> > @@ -0,0 +1,47 @@
> > +/* arch/arm/plat-samsung/include/plat/dma-ops.h
> > + *
> > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > + *		http://www.samsung.com
> > + *
> > + * Samsung DMA support
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License version 2 as
> > + * published by the Free Software Foundation.
> > +*/
> > +
> > +#include <linux/dmaengine.h>
> > +
> > +struct samsung_dma_prep_info {
> > +	enum dma_transaction_type cap;
> > +	enum dma_data_direction direction;
> > +	unsigned buf;
> > +	unsigned long period;
> > +	unsigned long len;
> > +	void (*fp)(void *data);
> > +	void *fp_param;
> > +};
> > +
> > +struct samsung_dma_info {
> > +	enum dma_transaction_type cap;
> > +	enum dma_data_direction direction;
> > +	enum dma_slave_buswidth width;
> > +	unsigned fifo;
> > +	struct s3c2410_dma_client *client;
> > +};
> > +
> > +struct samsung_dma_ops {
> > +	unsigned (*request)(enum dma_ch ch, struct samsung_dma_info *info);
> > +	int (*release)(unsigned ch, struct s3c2410_dma_client *client);
> > +	int (*prepare)(unsigned ch, struct samsung_dma_prep_info *info);
> > +	int (*trigger)(unsigned ch);
> > +	int (*started)(unsigned ch);
> > +	int (*flush)(unsigned ch);
> > +	int (*stop)(unsigned ch);
> > +};
> > +
> > +/*
> > + * samsung_dma_get_ops
> > + * get the set of dma operations
> > + */
> > +extern void *samsung_dma_get_ops(void);
> > diff --git a/arch/arm/plat-samsung/include/plat/dma-pl330.h
> b/arch/arm/plat-samsung/include/plat/dma-pl330.h
> > index c402719..be84bec 100644
> > --- a/arch/arm/plat-samsung/include/plat/dma-pl330.h
> > +++ b/arch/arm/plat-samsung/include/plat/dma-pl330.h
> > @@ -1,4 +1,7 @@
> >  /*
> > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > + *		http://www.samsung.com
> > + *
> >   * Copyright (C) 2010 Samsung Electronics Co. Ltd.
> >   *	Jaswinder Singh <jassi.brar@samsung.com>
> >   *
> 
> 
> Heh, this patch doesn't make any changes to this file, and samsung
> already has a copyright notice on it anyway.  You should probably drop
> this hunk.

I will address your comment.

> 
> > diff --git a/arch/arm/plat-samsung/include/plat/dma.h b/arch/arm/plat-
> samsung/include/plat/dma.h
> > index 2e8f8c6..90da162 100644
> > --- a/arch/arm/plat-samsung/include/plat/dma.h
> > +++ b/arch/arm/plat-samsung/include/plat/dma.h
> > @@ -124,4 +124,4 @@ extern int s3c2410_dma_getposition(unsigned int
> channel,
> >  extern int s3c2410_dma_set_opfn(unsigned int, s3c2410_dma_opfn_t rtn);
> >  extern int s3c2410_dma_set_buffdone_fn(unsigned int, s3c2410_dma_cbfn_t
> rtn);
> >
> > -
> 
> nitpick: Unrelated whitespace change.  One blank line of whitespace is
> sufficient anyway.

I will address your comment.

> > +#include <plat/dma-ops.h>
> > diff --git a/arch/arm/plat-samsung/s3c-dma-ops.c b/arch/arm/plat-
> samsung/s3c-dma-ops.c
> > new file mode 100644
> > index 0000000..17b1be0
> > --- /dev/null
> > +++ b/arch/arm/plat-samsung/s3c-dma-ops.c
> > @@ -0,0 +1,132 @@
> > +/* linux/arch/arm/plat-samsung/s3c-dma-ops.c
> > + *
> > + * Copyright (c) 2011 Samsung Electronics Co., Ltd.
> > + *		http://www.samsung.com
> > + *
> > + * Samsung S3C-DMA Operations
> > + *
> > + * This program is free software; you can redistribute it and/or modify
> > + * it under the terms of the GNU General Public License version 2 as
> > + * published by the Free Software Foundation.
> > +*/
> > +
> > +#include <linux/kernel.h>
> > +#include <linux/errno.h>
> > +#include <linux/slab.h>
> > +#include <linux/types.h>
> > +
> > +#include <mach/dma.h>
> > +
> > +struct cb_data {
> 
> struct s3c2410_cb_data {
> 
> > +	void (*fp) (void *);
> > +	void *fp_param;
> > +	unsigned ch;
> > +	struct list_head node;
> > +};
> > +
> > +static LIST_HEAD(dma_list);
> > +
> > +static void dma_cb(struct s3c2410_dma_chan *channel,
> > +		   void *param, int size, enum s3c2410_dma_buffresult res)
> > +{
> > +	struct cb_data *data = (struct cb_data *)param;
> 
> param is a void*.  The (struct cb_data*) cast is not needed.
> 
> > +
> > +	data->fp(data->fp_param);
> > +}
> > +
> > +static unsigned dma_request(enum dma_ch dma_ch, struct samsung_dma_info
> *info)
> 
> These functions should *definitely* be named differently from the
> dma_* ops in the other file so that you can differentiate between
> them, and to make them grep-friendly.  This goes for *all* file-scope
> symbols.

I will address your comment.

> 
> > +{
> > +	struct cb_data *data;
> > +
> > +	if (s3c2410_dma_request(dma_ch, info->client, NULL) < 0) {
> > +		s3c2410_dma_free(dma_ch, info->client);
> > +		return 0;
> > +	}
> > +
> > +	data = kmalloc(sizeof(struct cb_data), GFP_KERNEL);
> 
> kzalloc()
> 
> > +	data->fp = NULL;
> 
> Drop this line after converting to kzalloc()

I will address your comment.

> 
> > +	data->ch = (unsigned)dma_ch;
> 
> Is the cast necessary?
> 
> > +	list_add_tail(&data->node, &dma_list);
> > +
> > +	if (info->direction == DMA_FROM_DEVICE)
> > +		s3c2410_dma_devconfig(dma_ch, S3C2410_DMASRC_HW,
info->fifo);
> > +	else
> > +		s3c2410_dma_devconfig(dma_ch, S3C2410_DMASRC_MEM, info-
> >fifo);
> 
> From arch/arm/plat-samsung/include/plat/dma.h:
> 	enum s3c2410_dmasrc {
> 		S3C2410_DMASRC_HW,	/* source is memory */
> 		S3C2410_DMASRC_MEM	/* source is hardware */
> 	};
> 
> from dma-mapping.h:
> 	enum dma_data_direction {
> 		DMA_BIDIRECTIONAL = 0,
> 		DMA_TO_DEVICE = 1,
> 		DMA_FROM_DEVICE = 2,
> 		DMA_NONE = 3,
> 	};
> 
> /me thinks it would all look nicer if s3c2410 dma just replaced
> s3c2410_dmasrc to dma_data_direction, and from what I can tell it
> would just be a simple search and replace.  :-)
> 

I will address your comment.

> > +
> > +	if (info->cap == DMA_CYCLIC)
> > +		s3c2410_dma_setflags(dma_ch, S3C2410_DMAF_CIRCULAR);
> > +
> > +	s3c2410_dma_config(dma_ch, info->width);
> > +
> > +	return (unsigned)dma_ch;
> > +}
> > +
> > +static int dma_release(unsigned ch, struct s3c2410_dma_client *client)
> 
> unsigned int
> 
> > +{
> > +	struct cb_data *data;
> > +
> > +	list_for_each_entry(data, &dma_list, node)
> > +	    if (data->ch == ch)
> > +		break;
> 
> nit: incorrect indentation.  Use tab characters instead of spaces.
> I've got "set list" and "set listchars=tab:\|-,trail:-" in my ~/.vimrc
> so I can see the difference between tabs and spaces.
> 

I will address your comment.

> > +	list_del(&data->node);
> 
> What happens if the channel isn't found in the list?  Can that
> situation happen?
> 
> What happens if two drivers call dma_release simultaneously?  It
> looks like a mutex is needed to protext the dma_list.
> 
> > +
> > +	s3c2410_dma_free((enum dma_ch)ch, client);
> 
> All the casting in this patch makes me nervous.  When a lot of casting
> is required, I wonder if the API needs to be changed.

I will remove casting to "enum dma_ch"

> 
> > +	kfree(data);
> > +
> > +	return 0;
> > +}
> > +
> > +static int dma_prepare(unsigned ch, struct samsung_dma_prep_info *info)
> > +{
> > +	struct cb_data *data;
> > +
> > +	list_for_each_entry(data, &dma_list, node)
> > +	    if (data->ch == ch)
> > +		break;
> > +
> > +	if (!data->fp) {
> > +		s3c2410_dma_set_buffdone_fn((enum dma_ch)ch, dma_cb);
> > +		data->fp = info->fp;
> > +		data->fp_param = info->fp_param;
> > +	}
> > +	s3c2410_dma_enqueue((enum dma_ch)ch, (void *)data, info->buf,
> > +			    info->period);
> > +
> > +	return 0;
> > +}
> > +
> > +static inline int dma_trigger(unsigned ch)
> > +{
> > +	return s3c2410_dma_ctrl(ch, S3C2410_DMAOP_START);
> > +}
> > +
> > +static inline int dma_started(unsigned ch)
> > +{
> > +	return s3c2410_dma_ctrl(ch, S3C2410_DMAOP_STARTED);
> > +}
> > +
> > +static inline int dma_flush(unsigned ch)
> > +{
> > +	return s3c2410_dma_ctrl(ch, S3C2410_DMAOP_FLUSH);
> > +}
> > +
> > +static inline int dma_stop(unsigned ch)
> > +{
> > +	return s3c2410_dma_ctrl(ch, S3C2410_DMAOP_STOP);
> > +}
> > +
> > +static struct samsung_dma_ops s3c_dma_ops = {
> > +	.request	= dma_request,
> > +	.release	= dma_release,
> > +	.prepare	= dma_prepare,
> > +	.trigger	= dma_trigger,
> > +	.started	= dma_started,
> > +	.flush		= dma_flush,
> > +	.stop		= dma_stop,
> > +};
> > +
> > +void *samsung_dma_get_ops(void)
> > +{
> > +	return (void *)&s3c_dma_ops;
> > +}
> > +EXPORT_SYMBOL(samsung_dma_get_ops);
> 
> This is a problem.  This patch adds two implementations of
> samsung_dma_get_ops(). New code needs to be multiplatform friendly.
> That means that the code nees to handle both dma-ops.c and
> s3c-dma-ops.c compiled into the kernel at the same time and select the
> correct one at runtime.
> 

I will address your comment.

> g.
> 
> > --
> > 1.7.1
> >