Re: [PATCH 06/44 take 2] [UBI] startup code

[Artem Bityutskiy <dedekind@infradead.org>]

On Mon, 2007-02-19 at 10:59 +0000, Christoph Hellwig wrote:
> On Sat, Feb 17, 2007 at 06:54:54PM +0200, Artem Bityutskiy wrote:
> > +/* UBI headers must take 64 bytes. The below is a hacky way to ensure this */
> > +static int __ubi_check_ec_hdr_size[(UBI_EC_HDR_SIZE == 64) - 1]
> > +    __attribute__ ((__unused__));
> > +static int __ubi_check_ec_hdr_size[(UBI_VID_HDR_SIZE == 64) - 1]
> > +    __attribute__ ((__unused__));
> 
> please use BUILD_BUG_ON instead.

Will be done, thanks.

> > +
> > +static int ubi_attach_mtd_dev(const char *mtd_dev, int vid_hdr_offset,
> > +			      int data_offset);
> > +static void ubi_destroy_dev(int ubi_num);
> 
> Can you reorder the code to avoid all these forward declarations please?

Could you please submit a CodingStyle patch that would contain a
requirement to use the "higher-level functions at the bottom,
lower-layer at top"? Because I just use the opposite.

> > +	/* Attach MTD devices */
> > +	for (i = 0; i < mtd_devs; i++) {
> > +		struct mtd_dev_param *p = &mtd_dev_param[i];
> > +
> > +		cond_resched();
> > +		err = -EINVAL;
> 
> This looks very odd.

What exactly?

> > +module_param_call(mtd, ubi_mtd_param_parse, NULL, NULL, 000);
> > +MODULE_PARM_DESC(mtd, "MTD devices to attach. Parameter format: "
> > +		      "mtd=<name|num>[,<vid_hdr_offs>,<data_offs>]. "
> > +		      "Multiple \"mtd\" parameters may be specified.\n"
> > +		      "MTD devices may be specified by their number or name. "
> > +		      "Optional \"vid_hdr_offs\" and \"data_offs\" parameters "
> > +		      "specify UBI VID header position and data starting "
> > +		      "position to be used by UBI.\n"
> > +		      "Example: mtd=content,1984,2048 mtd=4 - attach MTD device"
> > +		      "with name content using VID header offset 1984 and data "
> > +		      "start 2048, and MTD device number 4 using default "
> > +		      "offsets");
> 
> This is a very odd paramater interface.  We really don't want drivers to use
> module_param_call directly.  You probably want various module_param_array calls
> instead.

Why not? We tried to avoid this but found out that this is the most
decent interface. Specific advises are welcome.

Thank you,
Artem.

-- 
Best regards,
Artem Bityutskiy (Битюцкий Артём)