Hi bodhi, bodhi Wrote: ------------------------------------------------------- > OK here is the DTB V7 for testing. Thank you very much. Here we go, nice progress! Using (from dts v6, but without the space mentioned a few posts before) setenv mtdparts 'mtdparts=armada-nand:-(ubifs);spi0.0:0x00100000(uboot),0x00010000@0x00100000(uboot_env)' setenv load_dtb_addr 0x1000000by lordzahl - Debian
Hi, Oh, sorry about that. I'll correct it, when i am home again, so you can check it. But the result was negative. The boot up stopped at the same position. - Manuelby lordzahl - Debian
Hi bodhi, > Please try again without the extra space! Here we go:by lordzahl - Debian
I build my own uboot image today from mainline u-boot: export CROSS_COMPILE=arm-none-eabi- make clearfog_defconfig Booting the resulting *kwb file shows exactly the same result as the last one from bodhi - i guess you build it the same way? I opened a new ticket at the thecus support to ask for uboot sources. Last time i answered the email about the GPL sources - hoping to directly reach tby lordzahl - Debian
bodhi Wrote: > > md5sum u-boot-spl-clearfog-tld-1.kwb > d9a8d444342fbf3fc1fcd74206460650 > u-boot-spl-clearfog-tld-1.kwb > The md5sum is fine, so something else is still wrong.by lordzahl - Debian
Hi, bodhi Wrote: > > kwboot-x86_64 -f -t -B 115200 /dev/ttyUSB0 -b > u-boot-spl-clearfog-tld-1.kwb -q 0 -s 1 -p > Ok, this one is interesting: The timing works quite well, about 50% success chance. The uploading only takes ~10s, but the last % i can see is ~25% and than the old output in the console is cleared and i get kc�ks�kh�d����ctkl�hs�by lordzahl - Debian
Hi bodhi, I did not take exact timings when trying to kwboot the first image. I think uploading the image took about 15min and i waited after that 30-60s for it to boot. To be clear about the the attempt with the third image: I actually restarted the box three times. The first two times i did not work and this is where the normal boot up messages are from. After the third insertion of the pby lordzahl - Debian
Hi bodhi, > Let's try kwboot with these u-boot SPL. Ok, here we go. First one (i'll update if the others are done.) $ ../../kwboot-x86_64 -f -t -B 115200 /dev/ttyUSB0 -b u-boot-spl-db-88f6820-amc-tld-1.bin -q 0 -s 1 Sending boot message. Please reboot the target...| Sending boot image... 0 % [......................................................................] 6 % [by lordzahl - Debian
bodhi Wrote: ------------------------------------------------------- > Wonder why do we need these GPIOs: > > > static const unsigned char HDD0_ATTACH = 26; > static const unsigned char HDD1_ATTACH = 27; > > > > Ah I think it is because these HDDs are > hot-swapable. > > This could cause some problems! Maybe the HDD ports need to be poweredby lordzahl - Debian
Hi, > OK then there is something we are not seeing. I > think I should add USB2 power in DTS, just in > case. Update: Nope. I think it must > have power always. There is no GPIO for USB2 > enabled! Yes, we are missing something. Just so you know: The usb2 is different than the usb3 ports: It is actually internal and normally not reachable, if the enclosure of the box is nby lordzahl - Debian
Hi, thanks. Version 6 result: edit: Ok, maybe nailman23 can test with HDD, i will not be home before monday and i have only taken the box with me, but no drives. BootROM - 1.73 Booting from SPI flash General initialization - Version: 1.0.0 AVS selection from EFUSE disabled (Skip reading EFUSE values) Overriding default AVS value to: 0x23 Detected Device ID 6820 High speed PHY -by lordzahl - Debian
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.by lordzahl - Debian
Hi bodhi, Thanks for the update. bodhi Wrote: ------------------------------------------------------- > Here is DTB version 5 tarball for testing. Here is the result, still no change. Are you sure "num-cs = <1>;" is correct? The dts from the gpl source states "0". This is the chip select number, right? I tried to change it myself, but i am missing the "by lordzahl - Debian
bodhi Wrote: ------------------------------------------------------- > 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 You mean information in GPL source? You may be right that the stock dts nand configuration is actually not valid. In a commentary it states: 92by lordzahl - Debian
Ok, 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 lordzahl - Debian
nailman23 Wrote: ------------------------------------------------------- > I've tried the DTB version 4 and it stops on > exactly the same step.Also tried with disks > attached but it did not bring any improvement. I can confirm that. edit: kwboot is running now. 30% upload after 30min. This takes forever. Btw i now use a cheap CP2012 USB to Serial Adapter.by lordzahl - Debian
> 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. Yes, i already did that last week as a reply to the GPL sources. Maybe i will open a new ticket as well.by lordzahl - Debian
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 target.../ Sending boot image... 0 % [......................................................................] 0 % [...............by lordzahl - Debian
Hi! I think i made progress with kwboot! After reading this post I downloaded from the solid-run page the download-serial.sh script. This scripts is similar to kwboot, interrupts the startup and then uploads a binary with the standard tool sx for the xmodem protocoll. I have nearly a 100% success rate with this script to get the NAK package of an interrupted startup sequence. But the uploaby lordzahl - Debian
Hi bodhi, > Here is a modified DTB version 3 for testing. This > time we have the correct SPI flash type, and the > SPI driver (in mtdparts). Still fails at the same point, see log. Next i will try kwboot again. edit: Still no luck with kwboot. Maybe it is because of my teensy as USB/Serial Converter. Maybe @nailman23 can try again. I had a little more luck with at least gettby lordzahl - Debian
Thanks for sharing the DTS. I will take a look this evening. I have no HDDs attached at the moment. Only the USB Stick. Couldn't wait and took a look into the dts. In the original boot log (with stock everything) the next lines are about nand flash. [ 3.550713] armada-nand f10d0000.nand: Initialize HAL based NFC in 8bit mode with DMA Disabled using BCH 4bit ECC [ 3.565269] NANDby lordzahl - Debian
Hi bodhi, > Here is a modified DTB version for testing. I tried it, but it hangs at the same position (Log at end of post). I double checked afterwards by diffing the dtb against the old one, that i actually tested your modified version. Just to be sure. I will give kwboot another try later today. Cheers, Manuel BootROM - 1.73 Booting from SPI flash General initializationby lordzahl - Debian
bodhi Wrote: ------------------------------------------------------- > Did it stop here? hang? > > > [ 0.920675] ata2: SATA max UDMA/133 mmio port 0x180 irq 43 > Yes, it just stops/hangs. Nothing happens after that. Inputs have no result. I tried a second time with the same result.by lordzahl - Debian
> 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 bad? Next up: Testing the DTB (thanks!!) Using the 4.14 kernel from you: BootROM - 1.73 Booting from SPI flashby lordzahl - Debian
Hi, > So I think the handshake did not occur. Sometime > this could be caused by the wire connections not > stable. I have seen reports here about some member > tried to hold the wires to the header during > kwbooting and it seems to work more reliable! It > could be that the xmodem protocol is more > sensitive than the other normal serial protocol. You might be rby lordzahl - Debian
Hi bodhi, Sure, can do: N2350:~# cat /proc/mtd dev: size erasesize name mtd0: 20000000 00020000 "ubifs" mtd1: 00400000 00010000 "U-Boot-img" mtd2: 00300000 00010000 "U-Boot-env" mtd3: 05885000 0001f000 "boot" mtd4: 008f6000 0001f000 "etc" mtd5: 1803b000 0001f000 "rom" mtd6: 00706000 0001f000 "version" N23by lordzahl - Debian
hi bodhi, Yes, J1 is a 4 hole pinout, which could be easily soldered and is used for the serial console by me at the moment. I just put some loose wires in these holes. I'll attach a picture of the backside of the pcb to this post.by lordzahl - Debian
Hi bodhi, Great that you found the GPIO definitions. bodhi Wrote: > Do you have a picture for this box that you can > upload here? Is there a serial pinouts header? I have linked an article with a picture in the first post. But if still needed i can get pictures myself and upload them here directly. It seems like J2 might be an unpopulated GPIO (20 pin)? Cheers, Manuelby lordzahl - Debian
@nailman23: Happy to hear that others are interested. Do you have a serial console yet? @bodhi: Okay, if i understand you correctly i should still use the dtb from your Debian/linux Images. We also do not need the kernel config. The only thing which is missing (at the moment) is the USB power. So i was thinking, the dts, which is used for this box does indeed not enable USB power on boot. Wby lordzahl - Debian
Hi! > Good to hear they are concientious with GPL. Last > time I browsed their website, GPL info seems old. They are indeed! Well the GPL info was actually old for most devices and i also started to loose hope, but today they answered and actually provided the GPL sources (so it seems). Seems like they prepared them just for me :) They can be found here: http://www.thecus.com/Dowby lordzahl - Debian