TI中文支持网自研omapl138板子，nand和官方omapl-lcdk一样，自己制作ubi文件系统烧写到nand上，总是起不来，内核启动信息：
ubi0: attaching mtd4
ubi0 warning: ubi_io_read: error -74 (ECC error) while reading 64 bytes from PEB 0:0, read only 64 bytes, retry
ubi0 warning: ubi_io_read: error -74 (ECC error) while reading 64 bytes from PEB 0:0, read only 64 bytes, retry
ubi0 warning: ubi_io_read: error -74 (ECC error) while reading 64 bytes from PEB 0:0, read only 64 bytes, retry
ubi0 error: ubi_io_read: error -74 (ECC error) while reading 64 bytes from PEB 0:0, read 64 bytes
CPU: 0 PID: 1 Comm: swapper Not tainted 4.19.59-g5f8c1c6121 #15
mmc0: host does not support reading read-only switch, assuming write-enable
Hardware name: Generic DA850/OMAP-L138/AM18x
Backtrace:[<c00147b8>] (dump_backtrace) from [<c0014a30>] (show_stack+0x18/0x1c)
r7:00000000 r6:00000040 r5:ffffffb6 r4:c6024000
mmc0: new high speed SDHC card at address aaaa
[<c0014a18>] (show_stack) from [<c06a4ab4>] (dump_stack+0x20/0x28)
mmcblk0: mmc0:aaaa SS16G 14.8 GiB[<c06a4a94>] (dump_stack) from [<c0434000>] (ubi_io_read+0x188/0x308)
[<c0433e78>] (ubi_io_read) from [<c04343b8>] (ubi_io_read_ec_hdr+0x4c/0x24c)
random: fast init done
mmcblk0:
r10:c63eb680 r9:00000000 r8:00000040 r7:00000000 r6:c6024000 r5:c63c4580
r4:c63d6c00
[<c043436c>] (ubi_io_read_ec_hdr) from [<c0439c84>] (ubi_attach+0x180/0x1634)
r10:c63eb680 r9:00000000 r8:00000000 r7:00000000 r6:c6024000 r5:c63c4580
r4:c63d6c00
[<c0439b04>] (ubi_attach) from [<c042d88c>] (ubi_attach_mtd_dev+0x484/0xbe0)
r10:00000978 r9:c63da400 r8:c6024040 r7:00000000 r6:c63da400 r5:c6024000
r4:0001f000
[<c042d408>] (ubi_attach_mtd_dev) from [<c0920838>] (ubi_init+0x18c/0x250)
r10:c09be788 r9:c098bd7c r8:c09be784 r7:c09be788 r6:c09be788 r5:c63da400
r4:00000000
[<c09206ac>] (ubi_init) from [<c000a540>] (do_one_initcall+0x74/0x1bc)
r10:00000000 r9:c090055c r8:ffffe000 r7:00000000 r6:c09206ac r5:c0947008
r4:c09a0c40
[<c000a4cc>] (do_one_initcall) from [<c0903ecc>] (kernel_init_freeable+0x120/0x1e0)
r8:c0932834 r7:c09a0c40 r6:c09a0c40 r5:c0932854 r4:c0942b7c
[<c0903dac>] (kernel_init_freeable) from [<c06b9ecc>] (kernel_init+0x10/0xf8)
r10:00000000 r9:00000000 r8:00000000 r7:00000000 r6:00000000 r5:c06b9ebc
r4:00000000
[<c06b9ebc>] (kernel_init) from [<c00090e0>] (ret_from_fork+0x14/0x34)
Exception stack(0xc602dfb0 to 0xc602dff8)
dfa0: 00000000 00000000 00000000 00000000
dfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
dfe0: 00000000 00000000 00000000 00000000 00000013 00000000
r5:c06b9ebc r4:00000000
我查看u-boot使用的是
#ifdef CONFIG_SYS_NAND_4BIT_HW_ECC_OOBFIRST
nand->ecc.mode = NAND_ECC_HW_OOB_FIRST;
nand->ecc.size = 512;
nand->ecc.bytes = 10;
nand->ecc.strength = 4;
nand->ecc.calculate = nand_davinci_4bit_calculate_ecc;
nand->ecc.correct = nand_davinci_4bit_correct_data;
nand->ecc.hwctl = nand_davinci_4bit_enable_hwecc;
nand->ecc.layout = &nand_davinci_4bit_layout_oobfirst;
linux kernel是
nand@2000000,0 {
795 compatible = "ti,davinci-nand";
796 #address-cells = <0x1>;
797 #size-cells = <0x1>;
798 reg = <0x0 0x2000000 0x2000000 0x1 0x0 0x8000>;
799 ti,davinci-chipselect = <0x1>;
800 ti,davinci-mask-ale = <0x0>;
801 ti,davinci-mask-cle = <0x0>;
802 ti,davinci-mask-chipsel = <0x0>;
803 ti,davinci-nand-buswidth = <0x10>;
804 ti,davinci-ecc-mode = "hw";
805 ti,davinci-ecc-bits = <0x4>;
806 ti,davinci-nand-use-bbt;
这种方式计算ecc，用nfsroot方式启动后，把某个nand 分区格式化为ubi格式可以挂载读写，操作没问题。所以总感觉是因为ecc计算方法不一致导致的。不知道官方板子是用那种文件系统，具体怎么做的，有没有文档之类的？我这个问题已经困扰好多天了，感觉要放弃的节奏。麻烦TI的巨佬帮忙。
李大闲：
补充一点，就是我的ubi.img文件系统，通过nfs起来之后，用ubifomat可以写入到nand，顺利起来，但是从u-boot 用nand write就死活不行。。
Nancy Wang：
回复 李大闲:
看一下这个培训视频是否有帮助。
training.ti.com/restoring-and-updating-u-boot-nand-omap-l138
李大闲：
回复 Nancy Wang:
视频和问题没多大关系，另外flashtoos也找不到。
我只是想确定下官方用的什么文件系统，怎么做的，这个按理来说应该有的！
目前查看uboot和linux对ecc的处理看起来是一致的，但就是起不来！
Nancy Wang：
回复 李大闲:
这个链接参考看一下。
https://software-dl.ti.com/processor-sdk-linux/esd/docs/06_03_00_106/linux/Foundational_Components/Kernel/Kernel_Drivers/Storage/NAND.html
Nancy Wang：
回复 李大闲:
昨天将您的问题发布在英文论坛，可以在后面跟帖补充，谢谢！
e2e.ti.com/…/3368097
