mboehmer

Members
  • Posts

    138
  • Joined

  • Last visited

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

mboehmer's Achievements

  1. New information: the "last byte read fixed to 0x00" happens only on slow SPI speeds, to my best knowledge it starts with SPI frequencies slower than 300kHz. SPI transmissions always work, correct data is sent out and received (according to logic analyzer). If you want to test it on your own, connect a jumper between pins 19 and 21 on the 40pin expansion header, and use a simple SPI transmission to see if things work. You need to open the SPI device first. SPI transmissions in my code are limited to 20bytes usually (one MachXO2 FlashROM page write/read). uint8_t test_spi( int fd, uint8_t size, uint32_t speed ) { uint8_t tx[size]; uint8_t rx[size]; memset(tx, 0xaa, size); memset(rx, 0x55, size); struct spi_ioc_transfer tr = { .tx_buf = (unsigned long)tx, .rx_buf = (unsigned long)rx, .len = ARRAY_SIZE(tx), .delay_usecs = 0, .speed_hz = speed, .bits_per_word = 8, }; if( ioctl(fd, SPI_IOC_MESSAGE(1), &tr) == -1) { return OP_ERROR; } /* print received data */ for( uint8_t i = 0; i < size; i++ ) { printf("%02x ", rx[i]); } printf("\n"); /* print transmitted data */ for( uint8_t i = 0; i < size; i++ ) { printf("%02x ", tx[i]); } printf("\n"); return OP_OK; }
  2. mboehmer

    Odroid C4

    I just came across a bug on C4 SPI - it trashes any transfer with 16++ bytes by overwriting the last received byte in the RX buffer by 0x00. Does it make sense at all to file a bug? Bug can be reproduced by simple means on the board I think (using a MOSI-MISO jumper). I can provide source code of "broken" SPI call, and logic analyzer shots if needed. Greetings, Michael
  3. I can confirm now a bug in the SPI kernel driver (AFAIK) for Odroid C4. SPI transfers on Odroid C4 work only reliably with a transfer size of 15 bytes or less. As soon as a transfer on SPI has more than 15 bytes (i.e. 16++), the last byte of the RX field of the transfer is overwritten by 0x00. I can see that correct data is being returned by the connected SPI slave on my logic analyzer. Same user code works without problems on C2 systems. I can file a bug with extended documentation here.
  4. Hi all, I have upgraded an Odroid C2 setup to C4, and encountered some strange issues. The SPI interface of C2 is bit banging, while the C4 offers "real" hardware support for SPI. I use a current kernel (5.10.102) with Bullseye. I use two SPI /CS lines, while one SPI controller takes care of "user" interaction with the LatticeSemi MachXO2 FPGA on the addon board. The other SPI controller is used to reflash the MachXO2 by the SPI. On the C2, everything works fine (while slow, due to bitbanging). On the C4, the same code on Odroid and FPGA fails - but only on the reflashing interface, not the user interface. The user interface uses "short" SPI accesses with usually 7 bytes in a command. The programming interface has long accesses with usually 16bytes++ for writing one 128bit FlashROM pages. As first result, the last byte of FlashROM page (16bytes) is usually fixed to zero. It seems to happen on readback. Has anyone encountered similar problems? Thanks in advance, Michael
  5. mboehmer

    Odroid C4

    After "apt update" I made a normal reboot. That was already end of game. Power cycle followed afterwards, with no change in behaviour. I set up now a new eMMC with normal Focal system (no desktop), and excluded the following files from upgrade: linux-dtb-current-meson64/focal 20.11 arm64 [upgradable from: 20.08.22] linux-focal-root-current-odroidc4/focal 20.11 arm64 [upgradable from: 20.08.22] linux-image-current-meson64/focal 20.11 arm64 [upgradable from: 20.08.22] linux-u-boot-odroidc4-current/focal 20.11 arm64 [upgradable from: 20.08.22] One of them (or the way it is updated) should be the culprit.
  6. mboehmer

    Odroid C4

    Anyone had problems with "apt update / apt upgrade"? Today I updated the packets on my C4, and for some reasons the system didn't boot anymore. I will try to reproduce with a freshly etched eMMC tonight.
  7. mboehmer

    Odroid C4

    Small update on the official image (Armbian 20.08.22 Focal with Linux 5.9.6-meson64): when I boot from eMMC, everything is fine. Inserting a Samsung uSD card leads to repeated "mmc1: error -84 whilst initialization". Trying to boot from uSD card (same image, working on eMMC) fails. System is stuck with"Starting kernel..." on console. Attaching the same uSD card by USB card reader works. So in case you find troubles with uSD card, check brand and product.
  8. mboehmer

    Odroid C4

    Hi, have been off for a few days, too many ongoing projects. I'm really happy to see that C4 is now "supported", and want to express my thanks to all people involved! Aynway, a small issue: the download image is reporting to be officially supported, while both selfcompiled images (buster/focal) report themselfs as "trunk" images without official support. Welcome to Armbian 20.11.0-trunk Buster with Linux 5.9.9-meson64 No end-user support: built from trunk I used a freshly installed build system from yesterday night. Is this expected behaviour? The network controller problem (dwmac reset issue) of previous builts is gone, btw. uSD card issues have not been verified, I'm using eMMC on my systems. So far, Michael
  9. HI all, first of all, I found thi sissue adopting a new board to the CSC tree, done by @balbes150. As it seems to be a problem of RK3288 in general, I post here, in case please move the topic to the approriate place. I have a JLD076 board, and started adjusting the DTS to the schematic. According to the schematic, and the TRM of RK3289, there is a maximum number of 160 GPIOs. This is also the number I get from my schematics. So I expected th get suitable number of pins from gpioinfo (bank:number 0:19, 1:4, 2:18, 3:28, 4:26, 5:12, 6:19, 7:24, 8:10). Instead, I get listed (0:24, 1...7:32, 8:16). I tried now to set ngpios for the &gpioX entries in the DTS, both in the rk3288.dtsi and the rk3288-ev-act8846.dts, and verified by dtc that the ngpios entry is in the DTB file, but gpioinfo still shows the wrong number of pins. This seems to be hardcoded somewhere... any ideas on what I do wrong here? Best regards, any help is appreciated. Michael
  10. We are down... and modules seem to be in good shape. Let's hope the powerup will work as expected.
  11. To get things more structured, I start over now, and try to get things done more systematically. (1) The image found here can't be started from uSD card. I press the maskrom button shortly, following console output, and see that the RK3288 tries to boot from uSD card, but fails. (2) The images proposed by @balbes150 can be booted from uSD card, by shortly pressing the maskrom button. (3) A shortly pressed maskrom button forces uSD card booting, a long pressed maskrom button forces the board to go to RKUSB mode. Edir: @balbes150 is right, it starts also directly from uSD card. (4) I changed now to Armbian_20.09_Arm-32_bionic_current_5.9.0-rc6.img image from @balbes150. It boots, but lacks ethernet. Changing the /boot/extlinux/extlinux.conf file to rk3288-evb-act8846.dtb I get ethernet. As the genuine Linaro reports itself as "evb" and carries the ACT8846 I think this is a good starting point. EDIT: got the versions mixed up. Sorry. (5) Uboot in this image seems to use random MACs, so accessing the board is a bit nasty. is the ethaddr env supported in this Uboot? EDIT: works like a charm. No more random MACs anymore I will extend this message with new informations as they arise.
  12. I recompiled the DTS and applied it in /boot/extlinux/extlinux.conf and get ERROR: reserving fdt memory region failed (addr=7df00000 size=d4638) Loading Device Tree to 0f8a1000, end 0f8b77e4 ... OK I will try thw Q8 image now, but it seems to lack ethernet support. Can you add it to the DTB? Otherwise I have to work all over USB serial which is a pain
  13. The schematic is not under NDA; but under gentlemen's agreement. I have to ask first.
  14. I have extracted the device tree from the eMMC based linux now, and attached it here. Does this help us? I have no access to the eMMC partitions for some reasons, as the eMMC partitions are Android type dts_orig.txt
  15. Thank you for your good explanation. I'm in contact with @balbes150 already, he helped me a lot. I used one of his images, and could successfully boot the board from uSD card. My main rror was to keep the maskrom switch pressed too long. If pressed shortly, it forces the RK3288 to skip the eMMC card and use the uSD card instead - it fails to load the kernel directly (as you pointed out) and loads the included 2020 Uboot, which does the rest. If pressed too long, it goes to real MaskROM loader... AFAIK it is possible to reflash the single sectors of eMMC later with newer Uboot and rootfs? Some help on how to backup and restore the current image would be great (I did some "backup" by rkdeveltool ( rkdeveloptool rl 0x0 $((7456 * 2048)) backup.data rkdeveloptool rl 0x0 $((7456 * 2048)) backup.data but I'm not sure if this is really the way to go. For DTB work, I will need some help, but I have a complete schematics of the board, documentation on how to compile for that specific card, and I already asked the manufacturer for his DTS file. I think - apart from DTS syntax - there is no problem to get all information. I would be happy to have some kind of official Armbian build, even if I have to get maintainer I have started now with rk3288-evb-act8846.dtb, and get ethernet working, other stuff not tried yet. Maybe you can send me your DTS file to check about ACT8846 stuff first?