Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Lawrie Griffiths
    @lawrie
    @emard I don't think @Dolu1990 is implying that the upstream version has his changes, just that he has merged the latest upstream into his version, so that it builds on recent Linux systems, like Ubuntu 20.04.
    e2kgh
    @e2kgh
    @Dolu1990 : "opencod", on my boards, which I hope to try in the near future, I have an FT4232 to program the FPGA & SPI Flash ...
    Lawrie Griffiths
    @lawrie
    @emard Have you looked at driving the OLED/LCD displays from SaxonSoc yet? It needs an ST7789 driver.
    emard
    @emard
    @lawrie yes I know, he pulled the openocd and applied required changes for softcore CPU. I currently have ST7789 plugged, are examples for this also besides ssd1331 (I haved it too but boxed around)
    How about some challinging stuff like getting USB linux host to accept a small USB1.1 hub at US2 port so kbd, mouse and something else to be plugged :)?
    Lawrie Griffiths
    @lawrie
    I think you might find that @Dolu1990 is working on that :)
    I did not attempt ST7789. I just wanted to show that the /dev/spidev stuff worked and the ssd1331 display seemed a simpler one to start with.
    emard
    @emard
    ahaaaaaa great! I have simple and reliable US2 PHY clock recovery core that runs well at 48MHz for USB1.1, its stuff for usbmouse and usbkbd in ulx3s-misc. OK I should unpack some SSD1331 and try this
    Dolu1990
    @Dolu1990
    So basicaly, i'm working on implement an USB OHCI controller, but that's quite some work
    Lawrie Griffiths
    @lawrie
    I am not sure what the best way to distribute optional software for SaxonSoc Linux is. We probably don't want to add it to the buildroot stuff as it isn't in buildroot packages.
    Dolu1990
    @Dolu1990
    i don't realy know, maybe some kind of out of tree packages ?
    emard
    @emard
    Yes something like risc32 tar extraction :)
    Lawrie Griffiths
    @lawrie
    The current software that we have is the riscv32_lcc compiler and library, @emard's mcpclock utility, @emard's tetris and my saxonsoc-lib, which currently has libraries for gpio, spidev and ssd1331. There was also some stuff that @kost built for the previous version.
    Lawrie Griffiths
    @lawrie
    We don't have a script for building the sd card image for this (Smp) version The one for the previous version used /dev/loop devices and I am bit dubious about that as which such devices are currently in use is system-dependent. That script added /etc/network/interfaces, create_keys.sh etc., as well.
    Dolu1990
    @Dolu1990
    right that's missing
    emard
    @emard
    @lawrie tetris early color support :) gamefield is bitarray so its scroll will currently reset colors
    Lawrie Griffiths
    @lawrie
    Looks much more like the real thing, in colour :)
    emard
    @emard
    :))) seems better, still I'm trying to use only basic vt100 sequences that most should support
    Paul Ruiz
    @pnru_gitlab

    There are now lots of system calls implemented, but they don't currently all have the wrappers that they need. For example, there is pselect6 and ppoll, but not select and poll wrappers.

    @lawrie the wrapper for select() seems to be fairly straightforward:
    https://github.com/brgl/uclibc-ng/blob/master/libc/sysdeps/linux/common/select.c
    It is a matter of normalising the timeout & adding a zero at the end of the parameter list.

    emard
    @emard
    TETRIS full color support (with vt100 scroll contols)... I hope it doesn't have bugs
    Lawrie Griffiths
    @lawrie
    A platform isn't complete until Tetris is running on it.
    emard
    @emard
    To support full color it's scrolling at the screen and bitmap is only internall kept for game logic.It had bugs at scrolling (looks like empty line where blocks should "fall" from above when lines are complete). I fixed and played 1 game and see no bugs, but more testing is needed :)
    I tried, bug is still there :(
    Screen first scrolls content down when lines fall, but accidentaly scrolls up and spoils the content... I don't know why. It's the same on PC linux, not just saxonsoc related
    Lawrie Griffiths
    @lawrie
    I see the score line repeated.
    emard
    @emard
    That is "ok", comes from hardware scrolling of complete screen. Should be fixed easly later
    bug looks like vt100 region-scroll doesn't work sometimes... hmmm
    Lawrie Griffiths
    @lawrie
    Also I can't rotate sometimes when at the side of the screen.
    emard
    @emard
    This is problem because rotation center is set wrong. They use some bitwise black magic to rotate piece bitmap in 4 bytes of RAM
    emard
    @emard
    @lawrie I fixed rotation center and added bot directions with keys 'k'-ccw and i-cw
    Erik Piehl
    @Speccery
    Got today my 2nd ULX3S from crowdsource campaign! Have to love that packing :)
    Goran Mahovlic
    @goran-mahovlic
    @Speccery I must say it was not so easy to pack 1000 boxes, but it was done with special care :)
    Erik Piehl
    @Speccery
    @goran-mahovlic thank you very much for the effort! 🙏
    Goran Mahovlic
    @goran-mahovlic
    It was my pleasure!
    cutting and pasting each stripe :)
    Erik Piehl
    @Speccery
    :) Yes and the result is just a beauty!
    By the way, I recently designed a built a bunch of ESP32 boards. I needed an UART for those designs, and I went with the FT231XS thanks to the ULX3S. Works like a charm. Lot of hand soldering :) in that project for me.
    kost
    @kost
    emard
    @emard
    @Speccery glad you liked ft231x design! I have long time not checked your ti99/4a, did you get sdram working?
    Erik Piehl
    @Speccery
    @emard well not that I got the 2nd ULX3S I am trying to get back to the project. So unfortunately I haven't been working on this recently, but will work on it now.
    emard
    @emard
    @kost thanx for the fresh tools! fujprog -t works but fujprog -t -P/dev/ttyUSB0 for me doesnt work. ujprog works with -t and with -t -P..
    @Speccery have some fun with new board, it has 16MB flash in ESP32.
    Erik Piehl
    @Speccery
    @emard thanks I will for sure. Was it so that the SDRAM here is different from the earlier board? Will it work the same way?
    emard
    @emard
    @Speccery SDRAM is ISSI speed grade 7 (spec'd as 133 MHz 2-2-2 ras/cas ws). On new boards ti is different than other boards but should be pin and protocol compatible. Small differences in timings are normally possible so we must test.
    This SDRAM has passed self-test and booting saxonsoc as Goran/Dobricas standard test procedure for all boards before shipping
    kost
    @kost
    @emard will investigate. thanks for reporting.
    opened issue here: kost/fujprog#9
    if you have anything else to add to help debug it - please do!
    Paul Ruiz
    @pnru_gitlab
    @emard I read that as: (i) clock speed can be 133MHz; and (ii) tRCD = 2 clocks, tCAS = 2 clocks and write latency is 2 clocks. If I read that correctly (i.e. tRCD = 15ns, etc.) it is a seriously fast grade 7 chip.
    emard
    @emard
    @kost nothing special with old compiled ujprog -t -p/dev/ttyUSB0 it works as terminal for e.g. esp32 or linux. Same ujprog newly compiled doesn't work, also fujprog doesn't work I guess someting in build env changed