@bodhi A bit of delay before I can check download-serial.sh due to returning this morning to find the rs816 we have already converted running in degraded mode after a disk failure, I do keep a spare disk for each array here but I had wanted to upgrade the disks to 4TB at some point anyway so decided to do a full re-install instead. In my rush I just went ahead and formatted to gpt and it wasnby richjn - Debian
@bodhi Bit of a holdup compiling send-stop-pattern $ gcc send-stop-pattern.c -o send-stop-pattern send-stop-pattern.c: In function ‘main’: send-stop-pattern.c:25:2: warning: implicit declaration of function ‘sleep’ [-Wimplicit-function-declaration] 25 | sleep(2); | ^~~~~ /usr/bin/ld: /tmp/ccLYyHWG.o: in function `main': send-stop-pattern.c:(.text+0x9): undefiby richjn - Debian
@bodhi Thanks for taking the time to explain, I understand now but am still bewildered why it accepts some files 100% of the time and others not at all but I guess if that's the reality we have to live with it. Would we be able to use the download_serial script brute force method to send our back up stock u-boot ?by richjn - Debian
@bodhi I'm very happy to try anything to unbrick the box so please don't read the following as me wanting to abandon other methods. I have been wondering if we are looking in the wrong direction. We already have a backup of a functioning u-boot and even if there was a procedure issue in the way it was collected we have another working box to try again. At first we thought it wasby richjn - Debian
@bodhi Looks like send-stop-pattern was compiled on something other than x86 ? or with a specific processor specified. $ ./download-serial.sh /dev/ttyUSB0 u-boot-a38x-2015T1_p18_Thecus-tld-4-spi-uart.bin Now reset or power cycle your Armada 38x board and then press enter ./download-serial.sh: line 14: ./send-stop-pattern: cannot execute binary file: Exec format error I did try to compby richjn - Debian
@bodhi >Try this close cousin u-boot that I rebbuilt for Thecus N2350: Its loading, went to 9% quickly now slowed down to a crawl but it is still going, fingers are crossed! Will update when its done which at current rate will be 25 mins unless it speeds up again. UPDATE: After an eternity it was finally all loaded but unless I'm meant to do something at the end point it a no gby richjn - Debian
@bodhi THIS ONE This is u-boot-armada-2011.12 which contains a Synology perl script (assumes you are cross compiling) to define your target; the target in our case is the rs815 which is a close relative of the rs816 but after my disastrous attempts maybe its not close enough. I think the main change was a bump in cpu speed from 1.33 to 1.8 EDIT Just re-downloaded myself and notice "by richjn - Debian
@bodhi My vm is provided by Qemu which supports arm* but it didn't like the armhf Wheezy installer, rather than waste time I decided to download the i386 Wheezy installer with a view to cross compiling and because there are no active repos for this grabbed the entire 13GB 3 dvd set burnt them and carried out the install no problem, I actually did need some packages from dvd 2 and 3 so itby richjn - Debian
@bodhi I didn't spend long looking but I can't immediately see a clean way of installing gcc-6 on Buster so rather than mess with an in-production server I will setup a Wheezy armhf vm which should be about the right vintage and see how that goes.by richjn - Debian
@bodhi I had tried to rebuild stock u-boot before which after a couple of simple things to fix eventually got stuck with gcc-8 based errors all over the place. At the time I wasn't motivated enough to downgrade to gcc-6 partially because I wasn't convinced that would be far enough. Marvel SDK for this board uses gcc-4.6. Very much appreciate you having a go at compiling this yourselby richjn - Debian
@bodhi >So far, that was the correct behavior Understood, it was just that the db-88f6820-amc got that far whereas db-88f6820-gp errors straight out so was wondering if tweaking the config of db-88f6820-amc might get us somewhere...... but I do know I'm just stabbing away in the dark.by richjn - Debian
@bodhi For no reason other than blind optimism I have tried to kwboot some of the targets under boards/Marvel with varying outcomes, I doubt any of this is useful info but thought I'd post it anyway. db-88f6820-amc $ kwboot -t -B 115200 /dev/ttyUSB0 -b u-boot-spl.kwb -p Sending boot message. Please reboot the target...- Sending boot image... 0 % [...............................by richjn - Debian
@bodhi Just for something to do I tried sending my dud u-boot-spl.kwb via kwboot and to my surprise it started to transfer but then borked out at around 25% progress. I had to control-c to show the progress below because it just clears when it fails (no scroll back) If it can send this kwb then maybe there was an issue with the mtd0 back up which affected previous attempts / something happeneby richjn - Debian
@bodhi > If you are brave and flash it blindly, you will end up spending a lot of time trying to recover with kwboot, or > JTAG. Curiosity had already go the better of me before I read your post and now I'm the proud owner of a shiny new brick :) I would have been happier if it had worked but I was quite prepared for this to be the outcome. I have been looking at a few armadaby richjn - Debian
@bodhi Whilst waiting for Debian Buster to be released I have spent a bit more time looking at u-boot again. On closer inspection it transpires (I believe at the request of kernel devs) rather than continue the previous mvebu_armada-3xxx naming convention they have shifted to a soc naming convention and in our case the board identification db-88f6820-gp which covers the armada-38x range (thereby richjn - Debian
@bodhi >I beleive I could build this easily (I have not tried though). If your vm is pre gcc-8 you probably can, I wasn't motivated enough to downgrade besides I wasn't sure how far I would have to go. I'm only going from memory but I think Synology are building with gcc-4.3 > Usually it will not work to use Armada XP on the Armada 38x. Bummer, I guess that killsby richjn - Debian
@bodhi Warning understood, I have no desire to brick one of my RS816's on a futile attempt but I am ok with a measured risk. I had already taken a quick look at the u-boot-armada-2011.12 (latest Synology GPL on Sourceforge) which contains a build config for the RS815 the predecessor but very close relation to the RS816. The customization was fairly minimal beyond defining memory and SPby richjn - Debian
@bodhi As suggested I tried a few times without hitting the 'magic' numbers I will park that for now. The u-boot implementation on this box sucks, I could spend a few hours seeing if I can rebuild stock u-boot but although I'm up for the risks of bricking my own box I guess a guide that starts with the following may prove an unpopular one :) QuoteStep 1: In order to have tby richjn - Debian
@bodhi Before I push this rs816 back to production for a second time did you want to explore trying to understand what's going on with kwboot or shall we park that one for now ? we will get a another chance when I do my other rs816.by richjn - Debian
Final armada-385-synology-rs816 DTS v14 Actually working and properly tested eth0 & eth1by richjn - Debian
@bodhi & @1000001101000 Thank you both for help ip link root@debian:~# ip link 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1024 link/ether 00:50:by richjn - Debian
KWBOOT Just tried kwboot so I could send you the log and now it's back to xmodem error. $ kwboot -a -t -B 115200 /dev/ttyUSB0 -b mtd0.stock.rs816 -p Sending boot message. Please reboot the target...- Sending boot image... BootROM - 1.73 (Boot) Booting from SPI flash General initialization - Version: 1.0.0 AVS selection from EFUSE disabled (Skip reading EFUSE values) Deteby richjn - Debian
@bodhi Quotemdio@72004 { pinctrl-names = "default"; pinctrl-0 = <&mdio_pins>; phy0: ethernet-phy@0 { reg = <0>; }; phy1: ethernet-phy@1 {by richjn - Debian
Thanks both for helping I don't have time to try much this morning but do have time for a couple of test; The dts also contains the following section do I need to expand that as well ? ethernet@70000 { pinctrl-names = "default"; phy = <&phy0>; phy-mode = "sgmii"; buffer-manager = <&bm>; bm,pool-long = <0>; sby richjn - Debian
@bodhi Regret to report some poor testing on my part which has only come to light now whilst trying to bring my second rs816 off-line for conversion. In order to take the second box out I needed to configure the second nic of the box which had already been converted but the second nic is not being "seen" in Debian. I'm afraid my testing had been as simple as swapping the wire frby richjn - Debian
@bodhi The box is back in the rack and in production, the issue is I have no serial access to it short of taking it back out of the rack and placing it next to my PC. As soon as I take the other one out of production I can do that plus any other re-runs you may require to complete the documentation of this conversion. I agree it would be good to give users that extra confidence so I will take myby richjn - Debian
@bodhi QuoteSorry, I'm only interested in ARM boxes :) I may have the same view too after the Ds1517+ leaves me with a bloody nose :) but ever the mug I will give it a try one day. First rs816 is back in the rack and running well. I will take the other down shortly and convert it over to native Debian but I may pause to see if I can get stock u-boot to build then, if so, update it / adby richjn - Debian
@bodhi Results below all good! So I will make a start pushing this one back to production, I agree with your initrd comments but I will leave it as is flash > zImage load raid1 but the next one with 8TB drives I suspect I will go with our just tested raw sector procedure. I'm not a very good proof reader as I tend to see what I think is there not what is actually there but if you wby richjn - Debian
@bodhi More winning = more shiraz Marvell>> <INTERRUPT> Marvell>> <INTERRUPT> Marvell>> scsi init AHCI init for unit0 Target spinup took 0 ms. SATA link 1 timeout. SATA link 0 timeout. SATA link 1 timeout. SATA link 2 timeout. SATA link 3 timeout. AHCI 0001.0000 32 slots 4 ports 6 Gbps 0xf impl SATA mode flags: 64bit ncq led only pmp fbss pio sluby richjn - Debian