bodhi, the USB3.0 port works well after I added these two lines. I can't wake up the devices with the IR remote after shutdown of libreELEC. With original Android OS it consumes ~3W after power off. I think it doesn't really shutdown. How it is usually solved that the IR sensor is still working while most components are powerd off? This means that the IR sensor must wake up all otherby shv - Debian
bodhi, I just found out that the CoolEME CM-MH1 is a 100% clone of the Beelink A1 because the name which is shown on the PCB is beelink a1. LibreELEC just uses the dts configuration of the mainline kernel: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/arch/arm64/boot/dts/rockchip?h=v5.16.7 If I compare rk3328-a1.dts with rk3328-roc-cc.dts I find the following differencby shv - Debian
Hello, I bought the CoolEME CM-MH1 RK3328 TV Box. This box seams to be a clone of Beelink A1 TV Box. Therefore I am using libreELEC and the DTB file which is made for Beelink A1 Box. My Box is running so far expect the USB3 port. Possibly some minor parts are different between these two boxes. dmesg output (Android@CoolEME) [ 1.558719] [0: swapper/0: 1] usbcore: registered newby shv - Debian
bodhi, I fetched the source code from your GitHub. -shvby shv - Debian
bodhi Wrote: ------------------------------------------------------- > But the thing that puzzles me is the SPL and u-boot currently in HDD was built for Pogo V3, > yet it booted without problem. So as fas as u-boot is > concerned, these 2 boxes are identical. > > Let me ping shv to see if he is still at his old email. I think he built the HDD u-boot images using my GitHub.by shv - Debian
bodhi, there is one defined maintainer of ox820 Openwrt code. I tried to contact him but didn't get an answer yet. I was already using stock OS as rescue system but wanted to see something new. If I need to rescue the SATA drive I would use another Linux system. Moreover I would need to restore kernel and rootfs in the flash. -shvby shv - Rescue System
bodhi, if I use a current snapshot initramfs PCI and networkinterface seam to work. root@OpenWrt:/# dmesg|grep pci [ 1.571965] oxnas-pcie 47c00000.pcie-controller: PCIe version/deviceID 0x82510b5 [ 1.579435] oxnas-pcie 47c00000.pcie-controller: link up [ 1.585080] oxnas-pcie 47c00000.pcie-controller: PCI host bridge to bus 0000:00 [ 1.592422] pci_bus 0000:00: root bus resourby shv - Rescue System
bodhi, the directory /etc/modules-boot.d contains just links to files which are stroed in the directory /etc/modules.d. There are text files which conain names or kernel modules which shall be loaded. The kernel modules are located in /lib/modules/4.14.131/<name>.ko. I will try to remove the definition of mtdparts. They are using no special/own u-boot but the original u-boot from cby shv - Rescue System
bodhi, my device is a Pogoplug Pro with RaLink 3390 PCI Wifi card. I compared the kernel files from https://downloads.openwrt.org/releases/18.06.4/targets/oxnas/ox820/ (Pro: cloudengines-pogoplug-pro-squashfs-sysupgrade.tar, V3: cloudengines-pogoplug-series-3-squashfs-sysupgrade.tar). The kernel files are different and the file for Pro contains the string Cloud Engines PogoPlug Pro while the kby shv - Rescue System
bodhi, can the dtb also be responsible that SATA, PCI and Wifi card don't work?by shv - Rescue System
bodhi, with /dev/mtd0 0x00100000 0x20000 0x20000 I get root@OpenWrt:/# fw_printenv Cannot read bad block mark: Invalid argument I think the right solution is /dev/mtd1 0xc0000 0x20000 0x20000 fw_printenv works with that. root@OpenWrt:/# fw_printenv autoload=no baudrate=115200 boot=bootm 60500000 boot_stock=run load_nand boot || run load_nand2 boot bootargs=console=ttyS0,11by shv - Rescue System
bodi, with the recommended change I get an error using fw_printenv. The ecc error about mtdblock1 still occur. root@OpenWrt:/# fw_printenv Warning: Bad CRC, using default environment bootcmd=bootp; setenv bootargs root=/dev/nfs nfsroot=${serverip}:${rootpath} ip=${ipaddr}:${serverip}:${gatewayip}:${netmask}:${hostname}::off; bootm bootdelay=5 baudrate=115200 My understanding is that fw_by shv - Rescue System
bodhi, this is the dmesg output of openwrt@pogoplug pro [ 0.000000] Booting Linux on physical CPU 0x0 [ 0.000000] Linux version 4.14.131 (buildbot@dbab1e23488d) (gcc version 7.3.0 (OpenWrt GCC 7.3.0 r7808-ef686b7292)) #0 SMP Thu Jun 27 12:18:52 2019 [ 0.000000] CPU: ARMv6-compatible processor [410fb025] revision 5 (ARMv7), cr=00c5787d [ 0.000000] CPU: PIPT / VIPT nonaliasinby shv - Rescue System
bodhi, I have the following uboot on the Pogoplug Pro: U-Boot 2013.10-tld-3 (May 15 2014 - 14:54:14) for OXNAS I used the following steps for the installation of openwrt with the file openwrt-18.06.4-oxnas-ox820-cloudengines-pogoplug-pro-initramfs-uImage on USB stick: 1. stop boot, adapt bootcmd for openwrt: 'nboot 60500000 0 440000; bootm' ## load openwrt from USB-Stick @uby shv - Rescue System
Hello, I used the following guideline to install OpenWrt on PogoPlug Pro: Link. The version 18.06.4 seams to work well. I didn't get the Wifi Interface working. SATA seams to be not supported. The memory layout is defined as follows: root@OpenWrt:~# cat /proc/mtd dev: size erasesize name mtd0: 00040000 00020000 "stage1" mtd1: 00380000 00020000 "u-boot" mby shv - Rescue System
bodhi, I found the solution within the following link: https://unix.stackexchange.com/questions/78535/how-to-get-rid-of-the-scanning-for-btrfs-file-systems-at-start-up. Maybe I had installed and tested some raid tools some years ago. Even btrfs-tools wasn't installed on my system anymore btrfs was still activated in initramfs. The problem was solved by execution of update-initramfs -ukby shv - Debian
Bodhi, I installed the new 89 kernel. Wifi and wap-supplicant are working fine. Some kernel modules are unintentionally loaded: btrfs 1003190 0 raid6_pq 87146 1 btrfs xor 4097 1 btrfs libcrc32c 845 0 What could be the reason? Where should I look at? It is running on my productive Squeeze system. Neither /etc/modules nby shv - Debian
bodhi, your rootfs works well. I upgraded from Squeeze to Stretch. Just samba has much worse throughput compared to Squeeze. I am using some performace options which are shown within the following link expect xfs - I am using ext4: http://falkhusemann.de/blog/2010/09/maximaldurchsatz-der-modifizierten-seagate-dockstar-optimieren But these options don't improve the performance of samby shv - Debian
bodhi, thanks for your hint. The problem seams to be caused by a different rootfs name. I am using "rootfs-ide". Therefore I must adapt fstab to this. - shvby shv - Debian
Hello bodhi, I want to recommend to add multicore zip pakages lbzip2 and pigz if you will update the rootfs. Additionally you might add some links to these executables into /usr/local/bin: lrwxrwxrwx 1 root staff 21 Sep 13 01:47 bunzip2 -> /usr//bin/lbzip2 lrwxrwxrwx 1 root staff 21 Sep 13 01:47 bzcat -> /usr/bin/lbzip2 lrwxrwxrwx 1 root staff 21 Sep 13 01:47 bzip2 ->by shv - Debian
@Bodhi, i am just testing your latest rootfs with kernel upgraded to 4.4.79 on two Pogoplug Pro devices, one with original RT3090 wifi card and one with AR9287 wifi card. With AR9287 the wifi works well. Both interfaces are working with your kernel. I am wondering that the device name is renamed from wlan0 to wlp0s0. dmesg [ 3.378900] rt2800pci 0000:00:00.0 wlp0s0: renamed from wlan0by shv - Debian
This is the script to deactivate automatic booting from SATA HDD: #!/bin/sh # uncomment line below and set to the correct disk #disk=/dev/sda if [ -z "${disk}" ] ; then echo "You must uncomment/set the 'disk' variable" exit -1 fi perl <<EOF | dd of="$disk" bs=512 print "\x00" x 0x1bc; EOF - shvby shv - Debian
Sorry, I posted in a wrong thread. -shvby shv - uBoot
@ youxiaojie, I want to recommend to start form a basic process and add additional description. ============================================================================ The device must support 512 Byte sector size either natively or by simulation. Step1: Attach the target drive to any linux device, e.g. PC or a working Pogoplug Step2: Prepare mbr partition table: sda1: rootby shv - Debian
bodhi Wrote: ------------------------------------------------------- > mtd2 (rootfs). I could upload mtd1 if shv does not > . The kernel mtd1 is pretty generic. I uploaded the mtd1 image of Seagate Goflex Net: Link. The mtd2 image of Seagate Goflex Net can be downloaded from this Link. @bodhi, I had to use the flash command without --noecc option to restore mtd2. nandwrite /devby shv - uBoot
bodhi Wrote: ------------------------------------------------------- > Awesome! pls do. In case somebody needs it. The mtd2 image of Seagate Goflex Net can be downloaded from this Link.by shv - uBoot
bodhi, I successfully recovered my GFN device after readout mtd2 image from a new virgin device. If you want I can upload the mtd2 image. -shvby shv - uBoot
bodhi, in contradiction to Pogoplug devices Goflex Net or Home provide Samba share functionality. Therefore I would prefere to flash an original Goflex rootfs. It would be nice if you could read out your Goflex Home rootfs if possible. With this rootfs I could also try if the recovery from http://support.goflexhome.hipserv.com works with Goflex Net. -shvby shv - uBoot
bodhi, I tried backup and restore and got the same results as before: ... [ 2.320000] sdhci: Secure Digital Host Controller Interface driver [ 2.320000] sdhci: Copyright(c) Pierre Ossman [ 2.330000] usbcore: registered new interface driver usbhid [ 2.330000] drivers/hid/usbhid/hid-core.c: v2.6:USB HID core driver [ 2.340000] TCP cubic registered [ 2.340000] NET: Regiby shv - uBoot
bodhi, before trying to flash uib file I want to try backup and restore of the current rootfs mtd2. Can you please tell me the commands with the right options to backup and restore mtd2(rootfs)? Are there some commands in the serial console to deal with bad sectors and try to recover the information form such sectors? Do you think that the ubi file form Goflex Home fits to Goflex Net? -by shv - uBoot