[U-Boot] [PATCH 3/7] dfu: DFU backend implementation

[Lukasz Majewski]

Hi Marek,

> Dear Lukasz Majewski,
> 
> > Hi Marek,
> > 
> > > Dear Lukasz Majewski,
> > > 
> > > > New, separate driver at ./drivers/dfu has been added. It allows
> > > > platform and storage independent operation of DFU.
> > > > 
> > > > Signed-off-by: Lukasz Majewski <l.majewski at samsung.com>
> > > > Signed-off-by: Kyungmin Park <kyungmin.park at samsung.com>
> > > > Cc: Marek Vasut <marex at denx.de>
> > > > ---
> > > 
> > > [...]
> > > 
> > > > +#include <common.h>
> > > > +#include <malloc.h>
> > > > +#include <mmc.h>
> > > > +#include <fat.h>
> > > > +#include <dfu.h>
> > > > +#include <linux/list.h>
> > > > +#include <linux/compiler.h>
> > > > +
> > > > +static LIST_HEAD(dfu_list);
> > > > +static int dfu_alt_num;
> > > > +
> > > > +static int dfu_find_alt_num(char *s)
> > > > +{
> > > > +	int i = 0;
> > > > +
> > > > +	for (; *s; s++)
> > > > +		if (*s == ';')
> > > > +			i++;
> > > > +
> > > > +	return ++i;
> > > > +}
> > > > +
> > > > +static char *dfu_extract_entity(char** env)
> > > > +{
> > > > +	char *s = *env;
> > > > +
> > > > +	strsep(env, ";");
> > > > +	return s;
> > > > +}
> > > 
> > > Shall we not make these all generic? They seem to be quite helpful
> > > components.
> > 
> > It is a good topic for a discussion if those functions shall be
> > moved to ./lib/string.c.
> > I regarded them as a "little" helper functions for parsing DFU alt
> > setting env variable. Those are very short and build with methods
> > exported from string.c
> 
> Good point
> 
> > > > +
> > > > +char *dfu_extract_token(char** e, int *n)
> > > > +{
> > > > +	char *st = *e;
> > > > +
> > > > +	debug("%s: %s\n", __func__, st);
> > > > +
> > > > +	strsep(e, " ");
> > > > +	*n = *e - st;
> > > > +
> > > > +	return st;
> > > > +}
> > > > +
> > > > +static unsigned char __aligned(CONFIG_SYS_CACHELINE_SIZE)
> > > > +
> > > > dfu_buf[DFU_DATA_BUF_SIZE];
> > > 
> > > Can we not stack-allocate it with ALLOC_CACHE_ALIGN_BUFFER()?
> > 
> > The dfu_buf is 4 MiB (this is the size of DFU_DATA_BUF_SIZE). I
> > don't think, that allocating it on the stack (for stack allocation
> > the ALLOC_CACHE_ALIGN_BUFFER() is designed) is a good idea.
> 
> Heh, agreed :-)
> 
> > > [...]
> > > 
> > > > diff --git a/include/dfu.h b/include/dfu.h
> > > > new file mode 100644
> > > > index 0000000..f7d823b
> > > > --- /dev/null
> > > > +++ b/include/dfu.h
> > > 
> > > [...]
> > > 
> > > > +struct dfu_entity {
> > > > +	char			name[DFU_NAME_SIZE];
> > > > +	int                     alt;
> > > > +	void                    *dev_private;
> > > > +	int                     dev_num;
> > > > +	enum dfu_device_type    dev_type;
> > > > +	enum dfu_layout         layout;
> > > > +
> > > > +	union {
> > > > +		struct mmc_internal_data mmc;
> > > 
> > > This union seems redundant ;-)
> > 
> > Good point :-), but I predict, that DFU will be used to program
> > other memory types (OneNAND, or NAND). To support those, one needs
> > to extend the union with e.g struct onenand_internal_data onenand.
> > 
> > Since we don't have so many memory types, I think that union usage
> > is acceptable.
> 
> And will those pieces be implemented any soon ? :)

:-). Since I'm the OneNAND custodian I shall keep in the back of my
head, that support for this memory is important :-)

> 
> > > [...]
> 
> Best regards,
> Marek Vasut



-- 
Best regards,

Lukasz Majewski

Samsung Poland R&D Center | Linux Platform Group