[RFC PATCH v2] MTD: nand: add return value for write_page()/write_page_raw() functions in structure of nand_ecc_ctrl.
mikedunn at newsguy.com
Thu Jun 21 13:58:01 EDT 2012
On 06/20/2012 11:43 AM, Brian Norris wrote:
> I noticed, for instance, that docg4.c has some strange code involving
> a return in a void function (comment below). If that is the *only*
> existing return statement within an 'ecc_ctrl.write_page'
> implmentation, then this whole patch is unneeded; you can just remove
> the 'return' in docg4.c.
>>> diff --git a/drivers/mtd/nand/docg4.c b/drivers/mtd/nand/docg4.c
>>> index a225e49..0f2ffd7 100644
>>> --- a/drivers/mtd/nand/docg4.c
>>> +++ b/drivers/mtd/nand/docg4.c
>>> -static void docg4_write_page_raw(struct mtd_info *mtd, struct nand_chip
>>> +static int docg4_write_page_raw(struct mtd_info *mtd, struct nand_chip
>>> const uint8_t *buf, int oob_required)
>>> return write_page(mtd, nand, buf, false);
> Hmm, this used to be a void function, returning the result of another
> void function? I would think the compiler would have warned about
> these issues before. Anyway, this is a useless 'return' statement and
> can be killed with no real effect.
Oops, yes, this must have been a careless mistake when I combined the read_page
and read_page_raw functions into one, with wrappers for each case. Oddly, no
compiler warning. I'm trying to get around to putting together a patch with a
couple minor fixes for the docg4, and I'll add this to the list.
More information about the linux-arm-kernel