> Same question as for v1, are we sure that the partition block size is > the same as the MMC erase block size? Is this guaranteed by all > standards (all MMC standards, and all partition table standards)? > Otherwise I assume we should do some maths here to figure out the MMC > start block AND block size based on the partition start block AND block > size and to pass the correct information to blk_derase(). I hope they > also are guaranteed to be aligned somehow and that a partition block > size is a multiple of an MMC erase block size? This maths seems to be > done in mmc_berase() (at a quick glance, not sure it's appropriate).
Hi Quentin, I'm no expert on U-Boot internals, but if you look e.g. to part_get_info_whole_disk() then you can see that blksz is directly copied from struct blk_desc to struct disk_partition. Both the start block and block size in struct disk_partition are multiples of blksz, so I think that they can be passed directly to blk_derase(), which later calls mmc_berase(). Best regards Tomas
