[U-Boot] Problem with mounting a jffs2 partition on flash.

[Marek Vasut <marex@denx.de>]

On Tuesday, February 09, 2016 at 04:36:33 PM, Bakhvalov, Denis (Nokia - 
PL/Wroclaw) wrote:
> Hi,

Hi!

> Thank you for you responses!
> 
> @To Marek:
> > It would be very helpful to see the whole boot output, not just U-Boot
> > output. For example SPL output is missing. Are you using mainline U-Boot
> > SPL or not ?
> 
> Here is the output from SPL:
> 
> U-Boot SPL 2013.01.01-svn564 (May 27 2015 - 17:58:12)
> Board Type : XXXX // I changed it
> Preloader Ver.: 00.01.03.00
> QSPI Access Mode: Single Access
> BOARD : Altera SOCFPGA Arria V Board
> CLOCK: EOSC1 clock 25000 KHz
> CLOCK: EOSC2 clock 25000 KHz
> CLOCK: F2S_SDR_REF clock 0 KHz
> CLOCK: F2S_PER_REF clock 0 KHz
> CLOCK: MPU clock 800 MHz
> CLOCK: DDR clock 400 MHz
> CLOCK: UART clock 100000 KHz
> CLOCK: MMC clock 12500 KHz
> CLOCK: QSPI clock 400000 KHz
> RESET: COLD
> INFO : Watchdog enabled
> SDRAM: Initializing MMR registers
> SDRAM: Calibrating PHY
> SEQ.C: Preparing to start memory calibration
> SEQ.C: CALIBRATION PASSED
> SDRAM: 512 MiB
> SF: Read data capture delay calibrated to 3 (0 - 7)
> SF: Detected S25FL512S with page size 512, total: 67108864

That's not mainline SPL. Please use mainline SPL with mainline U-Boot.
Mixing and matching various versions of U-Boot and SPL is not supported.

Is there any reason why you are using non-mainline SPL ?

> Here is the output from U-Boot:
> 
> U-Boot 2013.01.01-svn578 (Jan 08 2016 - 16:02:29)
> VERSION : 00.02.04.00
> U-Boot code: 01000040 -> 01049080  BSS: -> 01089908
> IRQ Stack: 0badc0de
> FIQ Stack: 0badc0de
> CPU   : Altera SOCFPGA Platform
> BOARD : XXXX  // I changed it
> I2C:   ready
> monitor len: 000898C8
> ramsize: 1F000000
> TLB table from 1eff0000 to 1eff4000
> Top of RAM usable for U-Boot at: 1eff0000
> Reserving 550k for U-Boot at: 1ef66000
> Reserving 640k for malloc() at: 1eec6000
> Reserving 32 Bytes for Board Info at: 1eec5fe0
> Reserving 128 Bytes for Global Data at: 1eec5f60
> Reserving 2048 Bytes for IRQ stack at: 1eec5760
> New Stack Pointer is: 1eec5750
> DRAM:  496 MiB
> relocation Offset is: 1df65fc0
> dram_bank_mmu_setup: bank: 0
> monitor flash len: 00051840
> Now running in RAM - U-Boot at: 1ef66000
> spi_setup_slave: bus 0 cs 0 max_hz 50MHz mode 3
> spi_claim_bus: bus:0 cs:0
> cadence_qspi_apb_chipselect : chipselect 0 decode 0
> cadence_qspi_apb_config_baudrate_div: ref_clk 400000000Hz sclk 50000000Hz
> Div 0x3 cadence_qspi_apb_config_baudrate_div: ref_clk 400000000Hz sclk
> 1000000Hz Div 0xf cadence_qspi_apb_config_baudrate_div: ref_clk
> 400000000Hz sclk 50000000Hz Div 0x3 SF: Read data capture delay calibrated
> to 3 (0 - 7)
> SF: Configuration Register : READ : CMD : 0x35
> spi_claim_bus: bus:0 cs:0
> cadence_qspi_apb_chipselect : chipselect 0 decode 0
> cadence_qspi_apb_config_baudrate_div: ref_clk 400000000Hz sclk 50000000Hz
> Div 0x3 SF: Bank Register : WRITE : CMD : 0x17 : BANK : 0x80
> SF: Opt Area Data : READ : CMD : { 0x4b : 0x00000010 }
> spi_claim_bus: bus:0 cs:0
> cadence_qspi_apb_chipselect : chipselect 0 decode 0
> cadence_qspi_apb_config_baudrate_div: ref_clk 400000000Hz sclk 50000000Hz
> Div 0x3 SF: Detected S25FL512S with page size 512, total 67108864 bytes
> SF: Detected S25FL512S with page size 512, total: 67108864
> READ: 0x100000 => cmd = { 0x0b 0x0010000008 } len = 0x20000
> spi_claim_bus: bus:0 cs:0
> cadence_qspi_apb_chipselect : chipselect 0 decode 0
> cadence_qspi_apb_config_baudrate_div: ref_clk 400000000Hz sclk 50000000Hz
> Div 0x3 READ: 0x140000 => cmd = { 0x0b 0x0014000008 } len = 0x20000
> spi_claim_bus: bus:0 cs:0
> cadence_qspi_apb_chipselect : chipselect 0 decode 0
> cadence_qspi_apb_config_baudrate_div: ref_clk 400000000Hz sclk 50000000Hz
> Div 0x3 Destroy Hash Table: 1efae2a0 table = 00000000
> Create Hash Table: N=512
> ...
> // then a lot of prints setting env variables.

I thought you were using mainline U-Boot, but this output is from non-mainline
U-Boot now. And it also contains some odd debug prints which are not even 
present in the U-Boot Altera delivers. I have to admit, I am quite confused.

I suppose you have Arria V SoCDK and you're trying to run U-Boot 2016.01 on
it, right ? Can you load the mainline SPL and mainline U-Boot and see if
that works for you? If it does not, please share the entire boot log .

If you have a custom board, you will need to change the memory timing bits
based on the output from Quartus . This is slightly more involved in both
the altera setup and mainline setup.

> > How did you install the U-Boot on your board ?
> 
> U-boot is simply stored to flash using USB blaster from Altera.

I was more interested in the procedure you used. I usually install mainline
U-Boot onto a board which boots from QSPI this way:

$ quartus_hps -c 1 -o PV -a 0x0 u-boot-with-spl-dtb.sfp

> @To Wolfgang:
> > What do you mean by "current" here?  UBI/UBIFS support in U-Boot was
> > actually adapted from the Linux kernel, so Linux support for UBI/UBIFS
> > predates the U-Boot code by a long, long time.
> 
> Ok, I tried to make it work with UBI/UBIFS:
> 
> 1. From Linux env I erased partition:
> flasheraseall /dev/mtd5
> 2. I'm able to work with this partition is Linux env:
> ubiattach /dev/ubi_ctrl -m 5
> ubimkvol /dev/ubi0 -N myUbifs -s 35MiB
> mount -t ubifs ubi0_0 /mnt
> 3. But from U-Boot it is not working:
> All commands as before:
> U-BOOT # sf probe 0 0 0
> U-BOOT # mtdparts
> ...
> ---list_partitions---
> 
> device nor0 <snor0>, # parts = 4
>  #: name                size            offset          mask_flags
>  0: boot                0x00100000      0x00000000      0
>  1: bootenv             0x00080000      0x00100000      0
>  2: SomeInfo            0x01980000      0x00180000      0
>  3: jffs2spart          0x02500000      0x01b00000      0
> 
> U-BOOT # ubi part ubifspart
> 
> ---mtdparts_init---
> last_ids  : nor0=snor0
> env_ids   : nor0=snor0
> last_parts:
> mtdparts=snor0:1M(boot),512k(bootenv),26112k(SomeInfo),-(ubifspart)
> env_parts :
> mtdparts=snor0:1M(boot),512k(bootenv),26112k(SomeInfo),-(ubifspart)
> 
> last_partition : nor0,0
> env_partition  : nor0,0
> --- find_dev_and_part ---
> id = ubifspart
> --- find_dev_and_part ---
> id = ubifspart
> Creating 1 MTD partitions on "nor0":
> 0x000001b00000-0x000004000000 : "mtd=3"
> raise: Signal # 8 caught
> raise: Signal # 8 caught
> UBI: attaching mtd1 to ubi0
> raise: Signal # 8 caught
> UBI: physical eraseblock size:   0 bytes (0 KiB)
> UBI: logical eraseblock size:    -128 bytes
> UBI: smallest flash I/O unit:    1
> UBI: VID header offset:          64 (aligned 64)
> UBI: data offset:                128
> UBI: empty MTD device detected
> UBI: create volume table (copy #1)
> UBI error: ubi_scan_get_free_peb: no eraseblocks found
> UBI error: ubi_init: cannot attach mtd1
> UBI error: ubi_init: UBI error: cannot initialize UBI, error -28
> UBI init error 28
> Command failed, result=28U-BOOT #
> 
> I'm confused about those prints:
> raise: Signal # 8 caught
> There are just so much of them... Even when executing sf probe command.

Is this u-boot 2016.01 or the ancient version provided by Altera please ?

> > Do you mean you are using a prehistoric Linux kernel version?
> 
> No, Linux version is 3.10.31.

Is there any reason why you don't use Linux 4.4 ? There are customers who
are using this on socfpga just fine, including FPGA manager and DT overlays
to deal with the FPGA part.

> > It appears it is hanging in the DRAM initialization phase.  This is a
> > very common issue.  Please make sure your memory configuration is
> > correct.
> > You can deal with this later; first you need to have the system
> > running at all.  You should focus on this first.
> 
> This is what I'm going to do next.

I'm looking forward to your findings!