[U-Boot] [PATCH v2 1/3] spl: mmc: support uboot image offset on main partition
sr at denx.de
Tue Aug 13 09:24:17 UTC 2019
On 12.08.19 16:25, Baruch Siach wrote:
> Hi Stefan,
> On Mon, Aug 12, 2019 at 03:13:58PM +0200, Stefan Roese wrote:
>> On 06.08.19 08:15, Stefan Roese wrote:
>>> On 29.07.19 09:12, Baruch Siach wrote:
>>>> On Armada 38x platforms the ROM code loads SPL from offset 0 of eMMC
>>>> hardware boot partitions. When there are no boot partitions (i.e. SD
>>>> card) the ROM skips the first sector that usually contains the (logical)
>>>> partition table. Since the generated .kwb image contains the main U-Boot
>>>> image in a fixed location (0x140 sectors by default), we end up with the
>>>> main U-Boot image in offset of 1 sector. The current workaround is to
>>>> manually set CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR to 0x141 to
>>>> compensate for that.
>>>> This patch uses the run-time detected boot partition to determine the
>>>> right offset of the main U-Boot partition. The generated .kwb image is
>>>> now compatible with both eMMC boot partition, and SD card main data
>>>> Signed-off-by: Baruch Siach <baruch at tkos.co.il>
>>>> v2: Rebase on top of current master
>>>> common/spl/Kconfig | 12 ++++++++++++
>>>> common/spl/spl_mmc.c | 15 +++++++++++++--
>>>> 2 files changed, 25 insertions(+), 2 deletions(-)
>>>> diff --git a/common/spl/Kconfig b/common/spl/Kconfig
>>>> index 5d6da5db89bc..e43eefe046bc 100644
>>>> --- a/common/spl/Kconfig
>>>> +++ b/common/spl/Kconfig
>>>> @@ -293,6 +293,18 @@ config SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR
>>>> Address on the MMC to load U-Boot from, when the MMC is being used
>>>> in raw mode. Units: MMC sectors (1 sector = 512 bytes).
>>>> +config SYS_MMCSD_RAW_MODE_U_BOOT_DATA_PART_OFFSET
>>>> + hex "U-Boot main hardware partition image offset"
>>>> + depends on SYS_MMCSD_RAW_MODE_U_BOOT_USE_SECTOR
>>>> + default 0x0
>>>> + help
>>>> + On some platforms SPL location depends on hardware partition. The ROM
>>>> + code skips the MBR sector when loading SPL from main hardware data
>>>> + partition. This adds offset to the main U-Boot image. Set this symbol
>>>> + to the number of skipped sectors.
>>>> + If unsure, leave the default.
>>>> config SYS_MMCSD_RAW_MODE_U_BOOT_USE_PARTITION
>>>> bool "MMC Raw mode: by partition"
>>>> diff --git a/common/spl/spl_mmc.c b/common/spl/spl_mmc.c
>>>> index b3619889f794..72439c029f4a 100644
>>>> --- a/common/spl/spl_mmc.c
>>>> +++ b/common/spl/spl_mmc.c
>>>> @@ -49,6 +49,16 @@ static ulong h_spl_load_read(struct spl_load_info *load, ulong sector,
>>>> return blk_dread(mmc_get_blk_desc(mmc), sector, count, buf);
>>>> +static __maybe_unused unsigned long spl_mmc_raw_uboot_offset(int part)
>>>> +#if IS_ENABLED(CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_USE_SECTOR)
>>> Use if (IS_ENABLED()) instead ?
> CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_DATA_PART_OFFSET is not defined when
> CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_USE_SECTOR is disabled. Using 'if
> (IS_ENABLED())' would break the build.
>>>> + if (part == 0)
>>> Why is this restricted to part == 0 ?
> As explained in the commit log, the ROM skips the first sector of the main
> data partition (where part == 0). Otherwise (where part != 0), the ROM load
> SPL from sector 0. That is the whole point of this patch set. Detect the SPL
> offset at run-time, instead of hard coding it at build time.
I see, thanks. Frankly, I don't really like this SoC specific extension
in the common code. Even though its "configurable" and won't "hurt" the
other users of this code. But I also don't see a better way to implement
this feature. So if nobody else objects, lets move forward with this
More information about the U-Boot