All of lore.kernel.org
 help / color / mirror / Atom feed
From: Masahiro Yamada <yamada.masahiro@socionext.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH V2 2/2] common: add new boot media kconfig entry
Date: Wed, 29 Jun 2016 09:33:42 +0900	[thread overview]
Message-ID: <CAK7LNAQMjRwAiY6jPRVWCORjMoBAkNzE+KcZXuruEJZaNOp3eQ@mail.gmail.com> (raw)
In-Reply-To: <20160628063955.GB28340@linux-7smt.suse>

Hi.

2016-06-28 15:39 GMT+09:00 Peng Fan <van.freenix@gmail.com>:
> Hi Masahiro,
> On Tue, Jun 28, 2016 at 02:24:00PM +0900, Masahiro Yamada wrote:
>>Hi.
>>
>>
>>2016-06-28 14:02 GMT+09:00 Peng Fan <van.freenix@gmail.com>:
>>> Hi Tom,
>>>
>>> On Fri, Jun 24, 2016 at 06:57:44PM -0400, Tom Rini wrote:
>>>>On Sun, Jun 19, 2016 at 06:20:52PM +0800, Peng Fan wrote:
>>>>> Hi Masahiro,
>>>>>
>>>>> +Simon
>>>>> On Fri, Jun 17, 2016 at 07:08:23PM +0900, Masahiro Yamada wrote:
>>>>> >2016-06-17 18:39 GMT+09:00 Peng Fan <van.freenix@gmail.com>:
>>>>> >> Add CONFIG_{SD|NAND|ONENAND|SPI|QSPI|SATA}_BOOT kconfig entries.
>>>>> >>
>>>>> >> SoCs supports loading U-Boot from different medias to DRAM, such as
>>>>> >> i.MX6/7 supports loading U-Boot to DRAM from sd/emmc/nand/qspi/spi/sata
>>>>> >> and etc. For i.MX, imximage will generate different IVT headers according
>>>>> >> to boot medias.
>>>>> >>
>>>>> >> Signed-off-by: Peng Fan <peng.fan@nxp.com>
>>>>> >> Cc: Simon Glass <sjg@chromium.org>
>>>>> >> Cc: Heiko Schocher <hs@denx.de>
>>>>> >> Cc: Joe Hershberger <joe.hershberger@ni.com>
>>>>> >> Cc: Bin Meng <bmeng.cn@gmail.com>
>>>>> >> Cc: Christophe Ricard <christophe-h.ricard@st.com>
>>>>> >> Cc: Nikita Kiryanov <nikita@compulab.co.il>
>>>>> >> Cc: Francois Retief <fgretief@spaceteq.co.za>
>>>>> >> Cc: Tom Rini <trini@konsulko.com>
>>>>> >> ---
>>>>> >>
>>>>> >> V2:
>>>>> >>  Move NOR_BOOT to the patch 1/2.
>>>>> >>  The idea of this patch is for adding different boot media support for
>>>>> >>  i.MXes. And I'll post out following patches if this patch is accepted.
>>>>> >>  I ran moveconfig.py, but I did not include the results into a patch.
>>>>> >>  This patch does not break the boards which defined NAND_BOOT/SD_BOOT and
>>>>> >>  etc, and I prefer to let board owners to move to defconfig later.
>>>>> >>
>>>>> >>  common/Kconfig | 48 ++++++++++++++++++++++++++++++++++++++++++++++++
>>>>> >>  1 file changed, 48 insertions(+)
>>>>> >>
>>>>> >> diff --git a/common/Kconfig b/common/Kconfig
>>>>> >> index 04d092c..f0f6ee1 100644
>>>>> >> --- a/common/Kconfig
>>>>> >> +++ b/common/Kconfig
>>>>> >> @@ -108,6 +108,54 @@ config NOR_BOOT
>>>>> >>           as the ROM only partially sets up pinmux.  We also default to using
>>>>> >>           NOR for environment.
>>>>> >>
>>>>> >> +config NAND_BOOT
>>>>> >> +       bool "Support for booting from NAND flash"
>>>>> >> +       default n
>>>>> >> +       help
>>>>> >> +         Enabling this will make a U-Boot binary that is capable of being
>>>>> >> +         booted via NAND flash. This is not a must, some SoCs need this,
>>>>> >> +         somes not.
>>>>> >> +
>>>>> >> +config ONENAND_BOOT
>>>>> >> +       bool "Support for booting from ONENAND"
>>>>> >> +       default n
>>>>> >> +       help
>>>>> >> +         Enabling this will make a U-Boot binary that is capable of being
>>>>> >> +         booted via ONENAND. This is not a must, some SoCs need this,
>>>>> >> +         somes not.
>>>>> >> +
>>>>> >> +config QSPI_BOOT
>>>>> >> +       bool "Support for booting from QSPI flash"
>>>>> >> +       default n
>>>>> >> +       help
>>>>> >> +         Enabling this will make a U-Boot binary that is capable of being
>>>>> >> +         booted via QSPI flash. This is not a must, some SoCs need this,
>>>>> >> +         somes not.
>>>>> >> +
>>>>> >> +config SATA_BOOT
>>>>> >> +       bool "Support for booting from SATA"
>>>>> >> +       default n
>>>>> >> +       help
>>>>> >> +         Enabling this will make a U-Boot binary that is capable of being
>>>>> >> +         booted via SATA. This is not a must, some SoCs need this,
>>>>> >> +         somes not.
>>>>> >> +
>>>>> >> +config SD_BOOT
>>>>> >> +       bool "Support for booting from SD/EMMC"
>>>>> >> +       default n
>>>>> >> +       help
>>>>> >> +         Enabling this will make a U-Boot binary that is capable of being
>>>>> >> +         booted via SD/EMMC. This is not a must, some SoCs need this,
>>>>> >> +         somes not.
>>>>> >> +
>>>>> >> +config SPI_BOOT
>>>>> >> +       bool "Support for booting from SPI flash"
>>>>> >> +       default n
>>>>> >> +       help
>>>>> >> +         Enabling this will make a U-Boot binary that is capable of being
>>>>> >> +         booted via SPI flash. This is not a must, some SoCs need this,
>>>>> >> +         somes not.
>>>>> >> +
>>>>> >>  endmenu
>>>>> >
>>>>> >
>>>>> >Do you intend to replace
>>>>> >CONFIG_SPL_NOR_SUPPORT
>>>>> >CONFIG_SPL_NAND_SUPPORT
>>>>> >CONFIG_SPL_USB_SUPPORT
>>>>> >CONFIG_SPL_MMC_SUPPORT
>>>>> >etc. with these options?
>>>>> >
>>>>>
>>>>> I missed these.
>>>>>
>>>>> >
>>>>> >Currently, common/spl/spl.c uses CONFIG_SPL_*_SUPPORT
>>>>> >to enable/disable capable boot devices.
>>>>>
>>>>> I think we could use a common option to replace the ones used in SPL.
>>>>
>>>>I'm not sure that CONFIG_xxx_BOOT and CONFIG_SPL_xxx_SUPPORT are the
>>>>same thing.  For example, CONFIG_NOR_BOOT on am335x means "we are
>>>>building to support booting from NOR, so include all of that stuff
>>>>that's normally not included".  While CONFIG_SPL_NAND_SUPPORT means
>>>>include NAND support in SPL, which is not mutually exclusive with MMC,
>>>>etc.
>>>
>>> ok. Then, better keep them. For i.MX6, CONFIG_XX_BOOT have the same meaning
>>> with am335x as you said.
>>>
>>> Do you have other comments? If not touching SPL, I would like to see this
>>> patch set applied.
>>>
>>> Thanks,
>>> Peng.
>>>>
>>
>>
>>I am not familiar with TI SoCs,
>
> Me too.

I missed to read git-log, I meant:
I am not familiar with i.MX SoCs.
(and I am not familiar with TI SoC, either)


>>so please help me to understand this.
>>
>>
>>+config NOR_BOOT
>>+       bool "Boot from NOR"
>>+       default n
>>+       help
>>+         U-Boot image is stored in NOR flash.
>>
>>
>>In the statement "U-Boot image is stored in NOR flash",
>>what "U-Boot" stands for?
>>SPL? or U-Boot proper?
>>
>>
>>If it points to SPL, it makes sense.
>
> In my case,
> SPL + U-Boot + OS.   means SPL
> Only U-Boot  + OS.   means U-Boot.


So, this board disables SPL
for an in-place executable device, like NOR.


I did likewise for my boards before,
and I found this was a PITA.

For a boot mode with SPL,
lowlevel_init code must be compiled into SPL.

For a boot mode without SPL,
lowlevel_init code must go into U-Boot proper.


Then, I ended up sprinkling ifdef CONFIG_SPL and
ifdef CONFIG_SPL_BUILD here are there.

This also loaded me per boot-mode testing
because boot sequence is too different for with/without SPL.


In my personal opinion, SPL should be enabled all the time.
Even for NOR boot, we can split the boot image
into SPL and U-Boot proper by using common/spl/spl_nor.c

Now, my SoC "select SPL"
and I can use an identical boot image
regardless of boot mode.

This may not be always possible,
but I am wondering if we could do someting
to mitigate the pain from per boot-mode defconfig.





>>
>>Theoretically, we can store SPL and U-Boot proper in different devices.
>>
>>CONFIG_FOO_BOOT means that SPL is stored in a FOO device.
>>(In other words, the hard-wired Boot ROM loads the SPL from the device FOO.
>>
>>
>>CONFIG_SPL_FOO_SUPPORT means that SPL supports loading U-Boot proper
>>from FOO a device.
>>In other words, the U-Boot proper can be stored in FOO device.
>
> From doc/README.SPL
> To support generic U-Boot libraries and drivers in the SPL binary one can
> optionally define CONFIG_SPL_XXX_SUPPORT. Currently following options
> are supported:
>
> ....
> CONFIG_SPL_MMC_SUPPORT (drivers/mmc/libmmc.o)
> ....
> CONFIG_SPL_NAND_SUPPORT (drivers/mtd/nand/libnand.o)
>
>>
>>
>>Correct?
>>
>>
>>One more question,
>>what will happen if both CONFIG_NOR_BOOT and CONFIG_NAND_BOOT are enabled?
>
> In my case, we do not allow CONFIG_NOR_BOOT and CONFIG_NAND_BOOT both defined.
> If both defined, U-Boot(I do not use spl) may not boot up in my case.


Perhaps, we should use a "choice" menu
to choose CONFIG_*_BOOT exclusively,
but I am not sure if it is the case for other vendors.
It looks like CONFIG_*_BOOT is used by TI as well as Freescale.


At least for my SoCs, I do not need these options at all
because one single boot image can boot from any device.


This patch is adding
  # CONFIG_NOR_BOOT is not set
  # CONFIG_NAND_BOOT is not set
  # CONFIG_QSPI_BOOT is not set
  ...

to my .config, which does not make sense.


Perhaps, can we define these symbols in an SoC Kconfig,
or surround them with "if <SOC>"?



-- 
Best Regards
Masahiro Yamada

  reply	other threads:[~2016-06-29  0:33 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-06-17  9:39 [U-Boot] [PATCH V2 1/2] Kconfig: make NOR_BOOT a common option Peng Fan
2016-06-17  9:39 ` [U-Boot] [PATCH V2 2/2] common: add new boot media kconfig entry Peng Fan
2016-06-17 10:08   ` Masahiro Yamada
2016-06-19 10:20     ` Peng Fan
2016-06-24 22:57       ` Tom Rini
2016-06-28  5:02         ` Peng Fan
2016-06-28  5:24           ` Masahiro Yamada
2016-06-28  6:39             ` Peng Fan
2016-06-29  0:33               ` Masahiro Yamada [this message]
2016-06-29  2:08             ` Tom Rini
2016-06-28 19:53   ` [U-Boot] [PATCH] configs: Re-sync after boot menu changes Tom Rini
2016-06-29  1:59   ` [U-Boot] [U-Boot, V2, 2/2] common: add new boot media kconfig entry Tom Rini
2016-06-17 22:59 ` [U-Boot] [PATCH V2 1/2] Kconfig: make NOR_BOOT a common option Simon Glass
2016-06-29  1:59 ` [U-Boot] [U-Boot, V2, " Tom Rini

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAK7LNAQMjRwAiY6jPRVWCORjMoBAkNzE+KcZXuruEJZaNOp3eQ@mail.gmail.com \
    --to=yamada.masahiro@socionext.com \
    --cc=u-boot@lists.denx.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.