lordzahl Wrote: ------------------------------------------------------- > bodhi Wrote: > ------------------------------------------------------- > > All, > > > > Have you tried with the HDD inserted in the > slot? > No without HDD in the slot. I don't have a HDD to > insert for the next few days. It would be good to see if the HDD was powered upby bodhi - Debian
Here is DTB version 6 tarball for testing. I've modified the NAND definition to use armada-nand, similar to the NAS326 and WD MyCloud boxes. And modified bootargs. Copy the attached DTB to your USB rootfs /boot/dts/ folder. Power up, interrupt serial console and set the envs (note that we are not loading uInitrd). We want to ignore that for now and see if the USB drive is poweredby bodhi - Debian
Manuel, QuoteAre you sure "num-cs = <1>;" is correct? Not sure.by bodhi - Debian
All, Have you tried with the HDD inserted in the slot?by bodhi - Debian
Here is DTB version 5 tarball for testing. I've modified the NAND definition to use armada-nand, similar to the NAS326 and WD MyCloud boxes. And modified bootargs. Copy the attached DTB to your USB rootfs /boot/dts/ folder. Power up, interrupt serial console and set the envs (note that we are not loading uInitrd). We want to ignore that for now and see if the USB drive is poweredby bodhi - Debian
QuoteOk, went the kwboot way again. This took 2h (thats not really normal, is it? 4MB at 115200 baud should be faster?! Notice the "-q 0 -s 1". This makes the timing easier in my experience. $ ../kwboot-x86_64 -f -t -B 115200 /dev/ttyUSB0 -b u-boot-a38x-spi-uart.bin -q 0 -s 1 Sending boot message. Please reboot the target.../ Sending boot image... 0 % [.........................by bodhi - Debian
echowarrior108, With the HDD and the USB rootfs attached to the Pogo. Power up, interrupt serial console at count down, and setenv usb_set_bootargs 'setenv bootargs console=ttyS0,115200 root=LABEL=rootfs rootdelay=10' setenv uinitrd_addr 0x60e00000 setenv uimage_addr 0x60500000 setenv dtb_addr 0x62c00000 setenv dt_load_dtb 'ext2load usb 0:1 $dtb_addr /boot/dts/ox820-pogopby bodhi - Debian
lordzahl, > You mean information in GPL source? You may be > right that the stock dts nand configuration is > actually not valid. Right, now that I looked in my NSA326 notes. The NFC NAND is actually supported in mainline as armada-nand. So the NSA326 had a simple NAND flash defintion. We just have to specify that in the bootargs.by bodhi - Debian
There was something about the NFC NAND but I forgot where the information is. The NAS326 box has a similar stock DTS as this box, but I ended up using + flash@d0000 { + status = "okay"; + num-cs = <1>; + marvell,nand-keep-config; + marvell,nand-enby bodhi - Debian
removed.by bodhi - Debian
echowarrior108, > I have no trouble getting into the pogoplug with > the serial but past that I am not sure what to do, > kinda feeling stupid asking because I am sure that > its been asked on other posts:) No, there is alway the first time for everything :) so far you did great! Quote> OX820 # usb tree > USB device tree: > 1 Hub (480 Mb/s, 0mA) > |by bodhi - Debian
@Manuel, Could you ask for u-boot GPL from Thecus? if it is available I'm thinking I will acquire this box to build new u-boot for Armada.by bodhi - Debian
lordzahl Wrote: ------------------------------------------------------- > bodhi Wrote: > > > Please try normal kwboot with > > u-boot-a38x-spi-uart.bin (without -p option). > > Sure, can do. After several tries: > > $ ../kwboot-x86_64 -f -t -B 115200 > /dev/ttyUSB0 -b u-boot-a38x-spi-uart.bin > Sending boot message. Please reboot the > targetby bodhi - Debian
bodhi Wrote: ------------------------------------------------------- > Here is a modified DTB version 3 for testing. This > time we have the correct SPI flash type, and the > SPI driver (in mtdparts). Sorry, looks like I upload a wrong version. This was the same, no change from before.by bodhi - Debian
habibie, Apparently, during u-boot installation, you have missed step 8f. Quote8. Flashing default u-boot envs image (if you are upgrading from 2016.05-tld-1 u-boot, you can skip this step 8). As described in step 1, u-boot envs must be defined in /etc/fw_env.config as # MTD device name Device offset Env. size Flash sector size Number of sectors /dev/mtd0 0xc0000 0x20000 0x20000by bodhi - uBoot
Manuel, Good works :) I'm aware about this script too. > download-serial.sh > script. This is the one we should use: u-boot-a38x-spi-uart.bin. I also have one similar I built last year. > edit: Looking at the sources: download-serial.sh > spams the magic character sequence 6000 times. Yup. Because they don't know the timing either :) > The > mainby bodhi - Debian
Good works Tommy :)by bodhi - Debian
I think you both were most likely right! perhaps this box hang when we don't try initialize the NAND partitions. Will do a version 4 to add the correct NAND mtds back in. Overall, the kernel log looks very good.by bodhi - Debian
habibie, > I now only have a SATA HDD attached sans any USB > memory stick, and yet my Seagate GoFLEX Home still > loads the kirkwood-pogo_e02.dtb instead of > the kirkwood-goflexhome.dtb when booting > through a SATA HDD. This happens regardless I have > the USB memory stick attached or not. So, what you > said is not an issue, AFAICT. OK then post the netconsoby bodhi - uBoot
ShapeShifter499 Wrote: ------------------------------------------------------- > Hello Bodhi, > > The thread is too long to read through all of it. > I'm just wondering what your patches fix over the > mainline and why they haven't been added to the > mainline kernel. Yes the patches should be applied on top of the mainline source tree 4.4.117. There is aby bodhi - Debian
Here is a modified DTB version 3 for testing. This time we have the correct SPI flash type, and the SPI driver (in mtdparts). Copy the attached DTB to your USB rootfs /boot/dts/ folder. Power up, interrupt serial console and set the envs (note that we are not loading uInitrd). We want to ignore that for now and see if the USB drive is powered up OK first. setenv mtdparts 'mtdpartsby bodhi - Debian
But that's a good point, though. Here is the boot log for NAS326. [ 1.269893] ahci-mvebu f10a8000.sata: AHCI 0001.0000 32 slots 2 ports 6 Gbps 0x3 impl platform mode [ 1.279299] ahci-mvebu f10a8000.sata: flags: 64bit ncq sntf led only pmp fbs pio slum part sxs [ 1.299041] scsi host0: ahci-mvebu [ 1.312901] scsi host1: ahci-mvebu [ 1.316504] ata1: SATA max UDMA/133 mmby bodhi - Debian
I'd doubt that a wrong flash definition would cause the hang. However, on the subject of SPI flash. 1. You're correct. That should be MX25L3205D according to your boot log. The current compatible name is from the mainline DTS version. I was not worry about it. SF: Detected MX25L3205D with page size 64 KiB, total 4 MiB 2. I suspect this would be wrong in mtdparts. But agaiby bodhi - Debian
Manuel, Try removing the HDDs and boot with USB only. And I am eager to see if you can try kwboot again.by bodhi - Debian
Manuel Here is the DTS source. Perhaps you might be able to spot something I did was wrong.by bodhi - Debian
Here is a modified DTB version for testing. Copy the attached DTB to your USB rootfs /boot/dts/ folder. Power up, interrupt serial console and set the envs (note that I've changed the usb_bootcmd to not loading uInitrd). We want to ignore that for now and see if the USB drive is powered up OK first. setenv load_dtb_addr 0x1000000 setenv load_initrd_addr 0x2900000 setenv load_imaby bodhi - Debian
habibie, As I've said above, conflicting info from both USB and HDD rootfs https://forum.doozan.com/read.php?3,54589,54659#msg-54659 Use one for data and other for rootfs then problem will be gone.by bodhi - uBoot
No, that dirty hack is not reconmended :)by bodhi - uBoot
lordzahl Wrote: ------------------------------------------------------- > > This run is actually better. When this happen, > > immediately use the up arrow to recall the > command > > and execute it again. > > I did that and it ended in the exact same message. > So it seems it at least got into the right mode, > but the data connection is still pretty baby bodhi - Debian
> Yes, it just stops/hangs. Nothing happens after > that. Inputs have no result. > I tried a second time with the same result. OK! I did not have any SATA node in the DTS yet. Will upload another version in perhaps 8hrs or so.by bodhi - Debian