[U-Boot] [PATCH v1 3/7] nand: add nand mtd concat support

Scott Wood scott.wood at nxp.com
Wed May 25 16:51:26 CEST 2016


On 05/25/2016 01:10 AM, Heiko Schocher wrote:
> Hello Scott,
> 
> Am 25.05.2016 um 07:30 schrieb Scott Wood:
>> On 05/25/2016 12:07 AM, Heiko Schocher wrote:
>>> add for nand devices mtd concat support. Generic MTD concat
>>> support is already ported to mainline, and used in the cfi_mtd
>>> driver. This patch adds it similiar for nand devices.
>>>
>>> Signed-off-by: Heiko Schocher <hs at denx.de>
>>> ---
>>>
>>>   drivers/mtd/nand/nand.c | 29 +++++++++++++++++++++++++++++
>>>   1 file changed, 29 insertions(+)
>>>
>>> diff --git a/drivers/mtd/nand/nand.c b/drivers/mtd/nand/nand.c
>>> index 8f0a921..5413123 100644
>>> --- a/drivers/mtd/nand/nand.c
>>> +++ b/drivers/mtd/nand/nand.c
>>> @@ -9,6 +9,7 @@
>>>   #include <common.h>
>>>   #include <nand.h>
>>>   #include <errno.h>
>>> +#include <linux/mtd/concat.h>
>>>
>>>   #ifndef CONFIG_SYS_NAND_BASE_LIST
>>>   #define CONFIG_SYS_NAND_BASE_LIST { CONFIG_SYS_NAND_BASE }
>>> @@ -30,6 +31,12 @@ static char dev_name[CONFIG_SYS_MAX_NAND_DEVICE][8];
>>>
>>>   static unsigned long total_nand_size; /* in kiB */
>>>
>>> +#ifdef CONFIG_MTD_CONCAT
>>> +static int nand_devices_found;
>>> +static struct mtd_info *mtd_nand_list[CONFIG_SYS_MAX_NAND_DEVICE];
>>> +static char c_mtd_name[16];
>>> +#endif
>>
>> Why is c_mtd_name file-scope rather than local?
> 
> fixed.
> 
>>> +
>>>   /* Register an initialized NAND mtd device with the U-Boot NAND command. */
>>>   int nand_register(int devnum)
>>>   {
>>> @@ -49,6 +56,9 @@ int nand_register(int devnum)
>>>   	 * via the mtdcore infrastructure (e.g. ubi).
>>>   	 */
>>>   	add_mtd_device(mtd);
>>> +#ifdef CONFIG_MTD_CONCAT
>>> +	mtd_nand_list[nand_devices_found++] = mtd;
>>> +#endif
>>
>> Why do we need another list of NAND devices?  Use nand_info[].
> 
> Ah, correct ... struct nand_info_t == struct mtd_info ... Hmm...
> why we need a "struct nand_info_t" ?

It's a typedef, not a struct...  and we don't need it.  I already have a
patch in the set I'm working on for the Linux NAND sync that removes the
typedef, and another that converts nand_info[] into an array of pointers
(which should make calling mtd_concat_create easier).

> 
>>>   #endif
>>>
>>>   	total_nand_size += mtd->size / 1024;
>>> @@ -102,4 +112,23 @@ void nand_init(void)
>>>   	 */
>>>   	board_nand_select_device(nand_info[nand_curr_device].priv, nand_curr_device);
>>>   #endif
>>> +
>>> +#ifdef CONFIG_MTD_CONCAT
>>> +	if (nand_devices_found > 1) {
>>> +		struct mtd_info *mtd;
>>> +
>>> +		/*
>>> +		 * We detected multiple devices. Concatenate them together.
>>> +		 */
>>> +		sprintf(c_mtd_name, "nand%d", nand_devices_found);
>>> +		mtd = mtd_concat_create(mtd_nand_list, nand_devices_found,
>>> +					c_mtd_name);
>>> +
>>> +		if (mtd == NULL)
>>> +			return;
>>> +
>>> +		if (add_mtd_device(mtd))
>>> +			return;
>>> +	}
>>> +#endif /* CONFIG_MTD_CONCAT */
>>>   }
>>
>> Please don't add new features that depend on the old-style NAND init.
> 
> Hmm... my add works for the "#ifdef CONFIG_SYS_NAND_SELF_INIT" case
> and the "else" case ... where would be the appropriate place?

You're right...  For some reason I was thinking that none of this was
called with self-init.

-Scott



More information about the U-Boot mailing list