[U-Boot] [RFC PATCH] kconfig: introduce kconfig for UBI

From: Andrew F. Davis <afd@ti.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [RFC PATCH] kconfig: introduce kconfig for UBI
Date: Tue, 20 Sep 2016 09:21:17 -0500	[thread overview]
Message-ID: <bc3a71bd-bf0b-7526-2c06-fc24958b6933@ti.com> (raw)
In-Reply-To: <57E0C87A.7030905@denx.de>

On 09/20/2016 12:26 AM, Heiko Schocher wrote:
> Hello Andrew,
> 
> Am 19.09.2016 um 17:38 schrieb Andrew F. Davis:
>> On 09/14/2016 12:03 AM, Heiko Schocher wrote:
>>> move the UBI config options into Kconfig.
>>>
>>> Signed-off-by: Heiko Schocher <hs@denx.de>
>>> ---
>>> Tested with tbot:
>>> http://lists.denx.de/pipermail/u-boot/2016-June/258119.html
>>>
>>> result:
>>> Boards      : 1196
>>> compile err : 36
>>> not checked : 0
>>> U-Boot good : 1157 bad 3
>>> SPL good    : 427 bad 0
>>> ('compile err :', [...]
>>> ('bad         :', ['sandbox', 'sandbox_noblk', 'sandbox_spl'])
>>>
>>> I have 36 boards, which do not compile, at least with my
>>> toolchains used, see:
>>> https://github.com/hsdenx/tbot/blob/master/config/tbot_uboot_kconfig_check.cfg#L68
>>>
>>> but none of them had UBI support, so this should be OK.
>>>
>>> One change which should be discussed in microblaze-generic:
>>> (I preceded the patchsnipset with a 'D' so I hope it does
>>>   not confuse patchwork)
>>> Ddiff --git a/include/configs/microblaze-generic.h
>>> b/include/configs/microblaze-generic.h
>>> Dindex e5bf700..d8e65f8 100644
>>> D--- a/include/configs/microblaze-generic.h
>>> D+++ b/include/configs/microblaze-generic.h
>>> D@@ -177,7 +177,6 @@
>>> D
>>> D #if defined(FLASH)
>>> D # define CONFIG_CMD_JFFS2
>>> D-# define CONFIG_CMD_UBI
>>> D # undef CONFIG_CMD_UBIFS
>>> D
>>> D # if !defined(RAMENV)
>>> D@@ -192,7 +191,6 @@
>>> D # endif
>>> D #else
>>> D # undef CONFIG_CMD_JFFS2
>>> D-# undef CONFIG_CMD_UBI
>>> D # undef CONFIG_CMD_UBIFS
>>> D #endif
>>> D #endif
>>> D@@ -202,7 +200,6 @@
>>> D #endif
>>> D
>>> D #if defined(CONFIG_CMD_UBIFS)
>>> D-# define CONFIG_CMD_UBI
>>> D # define CONFIG_LZO
>>> D #endif
>>>
>>> Hmm... how to handle such a move?
>>> Just ignore it, as my test does not detect different
>>> binaries?
>>
>> This kind of thing has shown up for me a bit, some of this logic should
>> be moved over to Kconfig, others like these maybe not. These could be
>> replaced by defaulting CMD_UBI when a FLASH or UBI enabled defconfig is
>> selected.
>>
>> IMHO all of this can be added later when the Kconfig conversion is
>> finished, for now if the binary doesn't change I see no harm.
> 
> Yep.
> 
>> [snip]
>>
>>> +config CMD_UBI
>>> +    tristate "Enable UBI - Unsorted block images commands"
>>> +    select CRC32
>>> +    select MTD_UBI
>>> +    help
>>> +      UBI is a software layer above MTD layer which admits of LVM-like
>>
>>                                                              ^^ use of?
> 
> I have this text from linux:/work/hs/linux/drivers/mtd/ubi/Kconfig
> but you are right, changed.
> 
> Do you want to send a fix for this to the MTD mailinglist?
> 

I'll add it to my small fix patch stack and send it to the trivial patch
monkey next merge cycle.

>> Otherwise:
>>
>> Reviewed-by: Andrew F. Davis <afd@ti.com>
> 
> Thanks!
> 
> bye,
> Heiko
>>
>>> +      logical volumes on top of MTD devices, hides some complexities of
>>> +      flash chips like wear and bad blocks and provides some other
>>> useful
>>> +      capabilities. Please, consult the MTD web site for more details
>>> +      (www.linux-mtd.infradead.org). Activate this option if you want
>>> +      to use U-Boot UBI commands.
>>> +
>>
> 