[PATCH 1/3] mtd: nand: Add on-die ECC support

[Richard Weinberger]

Am 28.04.2015 um 00:36 schrieb Ben Shelton:
> On 04/28, Richard Weinberger wrote:
>> Am 27.04.2015 um 23:35 schrieb Ben Shelton:
>>> I tested this against the latest version of the PL353 NAND driver that Punnaiah
>>> has been working to upstream (copying her on this message).  With a few changes
>>> to that driver, I got it most of the way through initialization with on-die ECC
>>> enabled, but it segfaults here with a null pointer dereference because the
>>> PL353 driver does not implement chip->cmd_ctrl.  Instead, it implements a
>>> custom override of cmd->cmdfunc that does not call cmd_ctrl.  Looking through
>>> the other in-tree NAND drivers, it looks like most of them do implement
>>> cmd_ctrl, but quite a few of them do not (e.g. au1550nd, denali, docg4).
>>>
>>> What do you think would be the best way to handle this?  It seems like this gap
>>> could be bridged from either side -- either the PL353 driver could implement
>>> cmd_ctrl, at least as a stub version that provides the expected behavior in
>>> this case; or the on-die framework could break this out into a callback
>>> function with a default implementation that the driver could override to
>>> perform this behavior in the manner of its choosing.
>>
>> Oh, I thought every driver has to implement that function. ;-\
>> But you're right there is a corner case.
>>
>> What we could do is just using chip->cmdfunc() with a custom NAND command.
>> i.e. chip->cmdfunc(mtd, NAND_CMD_READMODE, -1, -1);
>>
>> Gerhard Sittig tried to introduce such a command some time ago:
>> http://lists.infradead.org/pipermail/linux-mtd/2014-April/053115.html
> 
> That sounds reasonable to me.  That's similar to how we're checking the
> NAND status after reads in our current out-of-tree PL353 driver.  We
> added the extra command:
> 
> + /*
> +  * READ0 command only, for checking read status. Note that the real command
> +  * here is 0x00, but we can't differentiate between READ0 where we need to
> +  * send a READSTART after the address bytes, or a READ0 by itself, after
> +  * a read status command to check the on-die ECC status. The high bit is
> +  * written into the unused end_cmd field, so we don't need to mask it off.
> +  */
> +#define NAND_CMD_READ0_ONLY 0x100
> 
> and then added it to the struct pl353_nand_command_format of the driver:
> 
>  static const struct pl353_nand_command_format pl353_nand_commands[] = {
>         {NAND_CMD_READ0, NAND_CMD_READSTART, 5, PL353_NAND_CMD_PHASE},
> +       {NAND_CMD_READ0_ONLY, NAND_CMD_NONE, 0, NAND_CMD_NONE},
>         {NAND_CMD_RNDOUT, NAND_CMD_RNDOUTSTART, 2, PL353_NAND_CMD_PHASE},
>         {NAND_CMD_READID, NAND_CMD_NONE, 1, NAND_CMD_NONE},
>         {NAND_CMD_STATUS, NAND_CMD_NONE, 0, NAND_CMD_NONE},

Yep. All you need to do in check_read_status_on_die() is switching back to reading
mode.

>>
>> Maybe Brian can bring some light into that too...
>>
>>> When I build this without CONFIG_MTD_NAND_ECC_ON_DIE enabled, I get the
>>> following warning here:
>>>
>>> In file included from drivers/mtd/nand/nand_base.c:46:0:
>>> include/linux/mtd/nand_ondie.h: In function 'nand_read_subpage_on_die':
>>> include/linux/mtd/nand_ondie.h:28:1: warning: no return statement in function returning non-void [-Wreturn-type]
>>> include/linux/mtd/nand_ondie.h: In function 'nand_read_page_on_die':
>>> include/linux/mtd/nand_ondie.h:34:1: warning: no return statement in function returning non-void [-Wreturn-type]
>>>
>>> Perhaps return an error code here, even though you'll never get past the BUG()?
>>
>> What gcc is this?
>> gcc 4.8 here does not warn, I thought it is smart enough that this function does never
>> return. Can it be that your .config has CONFIG_BUG=n?
>> Anyway, this functions clearly needs a return statement. :)
> 
> gcc 4.7.2, and you are correct that I had CONFIG_BUG off.  :)

Yeah, just noticed that BUG() with CONFIG_BUG=n does not have
a nonreturn attribute. So, gcc cannot know...

Thanks,
//richard