[meta-xilinx] Help with boot failure
Nathan Rossi
nathan at nathanrossi.com
Tue Jun 20 06:27:20 PDT 2017
On 20 June 2017 at 03:29, Rich Wilson <richaw at gmail.com> wrote:
> Thank you for your response, Nathan.
> Top posting is supposed to be bad. So scroll way down to the bottom!
> And I may have discovered a big hint, see the end of my response.
>
> On Sun, Jun 18, 2017 at 11:54 PM, Nathan Rossi <nathan at nathanrossi.com>
> wrote:
>>
>> On 19 June 2017 at 15:59, Rich Wilson <richaw at gmail.com> wrote:
>> > I have tried to follow the directions in meta-xilinx to build
>> > microzed-zynq7, and the results aren't booting (from micro-SD).
>> > Boot results:
>> > =====================================================
>> > U-Boot SPL 2017.01 (Jun 17 2017 - 07:44:02)
>> > mmc boot
>> > Trying to boot from MMC1reading system.dtb
>> > spl_load_image_fat_os: error reading image system.dtb, err - -1
>> > reading u-boot.img
>> > reading u-boot.img
>> >
>> >
>> > U-Boot 2017.01 (Jun 17 2017 - 07:44:02 -0700)
>> >
>> > Model: Zynq MicroZED Board
>> > Board: Xilinx Zynq
>> > DRAM: ECC disabled 1 GiB
>> > MMC: sdhci at e0100000: 0
>> > SF: Detected s25fl128s_64k with page size 256 Bytes, erase size 64 KiB,
>> > total 1B
>> > *** Warning - bad CRC, using default environment
>> >
>> > In: serial at e0001000
>> > Out: serial at e0001000
>> > Err: serial at e0001000
>> > Model: Zynq MicroZED Board
>> > Board: Xilinx Zynq
>> > Net: ZYNQ GEM: e000b000, phyaddr 0, interface rgmii-id
>> >
>> > Warning: ethernet at e000b000 (eth0) using random MAC address -
>> > 92:62:5b:ad:43:d0
>> > eth0: ethernet at e000b000
>> > ** Bad device size - mmc 0 **
>> > Checking if uenvcmd is set ...
>> > Hit any key to stop autoboot: 0
>> > Copying FIT from SD to RAM...
>> > ** Bad device size - mmc 0 **
>>
>> Your issue appears to be here, for some reason U-Boot is failing to
>> access the SD card. Are you able to use the mmcinfo and fatls commands
>> in U-Boot to list and or manually load the images from your SD card?
>>
>> If not then there might be an issue with the card or with how you have
>> set up the partitions/etc.
>>
>> OE can generate an image with partition table/etc setup such that it
>> can be directly dd'd to the disk. To enable that up just add the
>> following to your local.conf:
>>
>> IMAGE_FSTYPES_append = " wic"
>> WKS_FILES = "sdimage-bootpart.wks"
>>
>> Re-run your image build and the output image will be
>> "core-image-minimal.wic".
>>
>> > Zynq>
>> > ====================================================
>> > Bitbake:
>> > rw at linux-pbup:~/poky/build> bitbake core-image-minimal
>> > Parsing recipes: 100% |##########################################| Time:
>> > 0:00:33
>> > Parsing of 871 .bb files complete (0 cached, 871 parsed). 1342 targets,
>> > 80
>> > skipped, 0 masked, 0 errors.
>> > NOTE: Resolving any missing task queue dependencies
>> >
>> > Build Configuration:
>> > BB_VERSION = "1.34.0"
>> > BUILD_SYS = "x86_64-linux"
>> > NATIVELSBSTRING = "universal-4.8"
>> > TARGET_SYS = "arm-poky-linux-gnueabi"
>> > MACHINE = "microzed-zynq7"
>> > DISTRO = "poky"
>> > DISTRO_VERSION = "2.3"
>> > TUNE_FEATURES = "arm armv7a vfp thumb neon callconvention-hard
>> > cortexa9"
>> > TARGET_FPU = "hard"
>> > meta
>> > meta-poky
>> > meta-yocto-bsp = "pyro:f01b909a266498853e6b3f10e6b39f2d95148129"
>> > meta-xilinx = "pyro:2371ecf0a0ad37cb775d4c09e89a1e70ab8f347a"
>> > meta-webserver = "pyro:5e82995148a2844c6f483ae5ddd1438d87ea9fb7"
>> > ====================================================
>> > Contents of boot partition (and you can see where I copied the files
>> > from):
>> >
>> > linux-pbup:/home/rw/poky/build/tmp/deploy/images/microzed-zynq7 # ls
>> > /mnt
>> > boot.bin microzed-zynq7.dtb
>> > uEnv.txt
>> > core-image-minimal-microzed-zynq7.cpio.gz.u-boot u-boot.img
>> > uImage
>> > =======================================================
>> > contents of uEnv.txt, which seemed to be equivalent to the directions
>> > (or
>> > do I need to replace the ${...} constructs? I don't think it's getting
>> > that
>> > far :) )
>> >
>> > kernel_image=uImage
>> > devicetree_image=microzed-zynq7.dtb
>> > bootargs=console=ttyPS0,115200 root=/dev/mmcblk0p2 rw rootwait
>> > earlyprintk
>> > loadkernel=fatload mmc 0 ${kernel_load_address} ${kernel_image}
>> > loaddtb=fatload mmc 0 ${devicetree_load_address} ${devicetree_image}
>> > bootkernel=run loadkernel && run loaddtb && bootm ${kernel_load_address}
>> > -
>> > ${devicetree_
>> > load_address}
>> > uenvcmd=run bootkernel
>> > ==============================================================
>> > Potential screwup by me: When I first bitbaked, meta-xilinx was at
>> > the master branch. After checking out pyro, I rebaked, but
>> > it didn't do anything. I removed /build/tmp, and the bitbake
>> > was pretty fast. Same result.
>>
>> It looks like your build is fine, and pryo/master point at the same
>> commit at the moment so there should be no differences. Also no need
>> to modify the generated uEnv.txt.
>>
>> Regards,
>> Nathan
>
>
> I added the requested lines to local.conf and bitbaked. The first time I
> tried this,
> it got stuck for over 10 minutes, guile do_compile IIRC, with over 99% user
> CPU time. That didn't seem right, so CTRL/C and tried again. Succeeded
> very quickly. Here's the evidence:
>
> Keyboard Interrupt, closing down...
>
>
> Second Keyboard Interrupt, stopping...
>
> Execution was interrupted, returning a non-zero exit code.
> NOTE: Sending SIGTERM to remaining 1 tasks
> Traceback (most recent call last):
> File "/home/rw/poky/bitbake/bin/bitbake", line 48, in <module>
> cookerdata.CookerConfiguration()))
> File "/home/rw/poky/bitbake/lib/bb/main.py", line 470, in bitbake_main
> server_connection.terminate()
> File "/home/rw/poky/bitbake/lib/bb/server/process.py", line 231, in
> terminate
> self.procserver.join(0.1)
> File "/usr/lib64/python3.4/multiprocessing/process.py", line 121, in join
> res = self._popen.wait(timeout)
> File "/usr/lib64/python3.4/multiprocessing/popen_fork.py", line 51, in
> wait
> if not wait([self.sentinel], timeout):
> File "/usr/lib64/python3.4/multiprocessing/connection.py", line 926, in
> wait
> ready = selector.select(timeout)
> File "/usr/lib64/python3.4/selectors.py", line 367, in select
> fd_event_list = self._poll.poll(timeout)
> KeyboardInterrupt
>
>
> And after the appropriate dd, here is the result:
>
> U-Boot SPL 2017.01 (Jun 19 2017 - 00:20:53)
> mmc boot
> Trying to boot from MMC1reading system.dtb
> spl_load_image_fat_os: error reading image system.dtb, err - -1
> reading u-boot.img
> reading u-boot.img
>
>
> U-Boot 2017.01 (Jun 19 2017 - 00:20:53 -0700)
>
> Model: Zynq MicroZED Board
> Board: Xilinx Zynq
> DRAM: ECC disabled 1 GiB
> MMC: sdhci at e0100000: 0
> SF: Detected s25fl128s_64k with page size 256 Bytes, erase size 64 KiB,
> total 1B
> *** Warning - bad CRC, using default environment
>
> In: serial at e0001000
> Out: serial at e0001000
> Err: serial at e0001000
> Model: Zynq MicroZED Board
> Board: Xilinx Zynq
> Net: ZYNQ GEM: e000b000, phyaddr 0, interface rgmii-id
>
> Warning: ethernet at e000b000 (eth0) using random MAC address -
> 92:62:5b:ad:43:d0
> eth0: ethernet at e000b000
> ** Bad device size - mmc 0 **
> Checking if uenvcmd is set ...
> Hit any key to stop autoboot: 0
> Copying FIT from SD to RAM...
> ** Bad device size - mmc 0 **
> Zynq> mmcinfo
> Device: sdhci at e0100000
> Manufacturer ID: 3
> OEM: 5344
> Name: SL16G
> Tran Speed: 50000000
> Rd Block Len: 512
> SD version 3.0
> High Capacity: Yes
> Capacity: 14.8 GiB
> Bus Width: 4-bit
> Erase Group Size: 512 Bytes
> Zynq> fatls mmc 0
> 69424 boot.bin
> 21929 microzed-zynq7.dtb
> 465013 u-boot.img
> 417 uenv.txt
> 3597520 uimage
>
> 5 file(s), 0 dir(s)
>
> Zynq>
>
>
> Now an interesting experiment. I typed three commands from
> uenvcmd=fatload mmc 0 0x3000000 ${kernel_image} && fatload mmc 0 0x2A00000
> ${devicetree_image} && bootm 0x3000000 - 0x2A00000
>
> and the kernel took off. But couldn't find root, no surprise.
>
> AHA! This might be a big hint what is going on:
> (after a reboot)
>
> Zynq> fatls mmc 0
> ** Bad device size - mmc 0 **
> Zynq> mmcinfo
> Device: sdhci at e0100000
> Manufacturer ID: 3
> OEM: 5344
> Name: SL16G
> Tran Speed: 50000000
> Rd Block Len: 512
> SD version 3.0
> High Capacity: Yes
> Capacity: 14.8 GiB
> Bus Width: 4-bit
> Erase Group Size: 512 Bytes
> Zynq> fatls mmc 0
> 69424 boot.bin
> 21929 microzed-zynq7.dtb
> 465013 u-boot.img
> 417 uenv.txt
> 3597520 uimage
>
> 5 file(s), 0 dir(s)
>
> Zynq>
I was able to reproduce (on a Zybo) the "Bad device size" that is
printed first (which is when u-boot tries to process preboot). However
I needed to use a different/slower? card to make it happen. And I was
only able to reproduce it in u-boot v2017.01 and not v2017.05 (which
is in oe-core master).
Would you be able to test u-boot v2017.05 and see if you can reproduce it there?
If you can still reproduce your issue there, the following patch might
be needed to force u-boot to rescan the mmc devices before attempting
to load the preboot config. The follow commit for meta-xilinx should
work for u-boot v2017.01 and v2017.05 (on pyro or master).
https://github.com/nathanrossi/meta-xilinx/commit/20ae90016aaacec9569c1cb34a60552231fe733d
You will probably also need this patch to get the autoboot working
properly, since mainline u-boot doesn't set all the variables used by
default.
https://github.com/nathanrossi/meta-xilinx/commit/6472328917c240d1ad03561f038a6fe2de8bb382
Regards,
Nathan
More information about the meta-xilinx
mailing list