| --- Log opened Mon Feb 23 00:00:53 2015 | ||
| stekern | poke53281: nice with the jor1k rtems demo | 01:10 |
|---|---|---|
| stekern | cool, lk is starting to make some sort of noise with mmu's turned on now | 01:23 |
| stekern | proof that openrisc development is hard and sweaty work: http://1drv.ms/1vqZzS8 | 03:07 |
| stekern | I just saw this on linkedin, might be interesting for someone: https://github.com/LarsAsplund/vunit | 06:19 |
| olofk | stekern: Haha. I want your office | 07:08 |
| olofk | And I assume that lk kernel is making good noise | 07:08 |
| olofk | I looked at vunit and OSVVM when working for a client who was using VHDL. Never got to try it out though | 07:10 |
| Me1234 | Is this OK? | 07:52 |
| Me1234 | <hex_offset>720896</hex_offset> | 07:52 |
| Me1234 | In the orpsoc.cof file. | 07:52 |
| Me1234 | orpsoc.cof from here https://github.com/juliusbaxter/mor1kx-dev-env/blob/master/boards/altera/de0_nano/syn/quartus/bin/orpsoc.cof | 07:53 |
| Me1234 | with changed file names. | 07:54 |
| Me1234 | http://pastie.org/9974015 is the script for jic generation. | 07:56 |
| Me1234 | Does the FPGA configuration overlap the data to copy to RAM? | 07:57 |
| Me1234 | With orpsoc v2 it works. orpsoc v2 is 9781 LEs, orpsoc v3 is 11005 LEs | 07:58 |
| olofk | Me1234: Number of LEs doesn't matter here, but I guess the FPGA image must be smaller than 720896 bytes | 08:05 |
| poke53281 | stekern: Sweden is not supposed to have weather in Februrary, were you can sit outside. :) | 08:08 |
| poke53281 | stekern: I don't have a clue what I can do with rtems. "help" is the only command, that works. | 08:11 |
| stekern | poke53281: neither does finland, that's why we escape to thailand in the winter | 08:12 |
| olofk | poke53281: Yeah. I tried ls, then help. Ended up in the same position as you after that :) | 08:13 |
| poke53281 | Especially I would remlove the RT from RTEMS, when running in the browser. | 08:16 |
| olofk | :) | 08:17 |
| olofk | Perhaps EEMS as in Eventually | 08:18 |
| poke53281 | stekern: Never was in Tahiland. But it looks like you enjoy it. | 08:20 |
| poke53281 | s/Tahiland/Thailand/ | 08:20 |
| poke53281 | olofk: Also the internet is not really helpful. | 08:21 |
| poke53281 | https://github.com/RTEMS/rtems/blob/master/testsuites/samples/capture/capture.scn | 08:24 |
| poke53281 | olofk: This one seems to be a sample session | 08:24 |
| poke53281 | try "config" or "task" | 08:32 |
| poke53281 | I should finally add a scrollback buffer | 08:32 |
| poke53281 | But as long as I can't install busybox, it seems pretty useless | 08:33 |
| Me1234 | On real hardware (de0_nano) I see that mor1kx loops in bootrom (bootrom.S from orposc v2) for a long time, than jumps to 0x100, And then Illegal instruction exception occurs (0x700). | 08:57 |
| Me1234 | On simulator it loops there for time, less than 1ms and then jumps to 0x200. | 08:58 |
| Me1234 | I am uploading the vcd file to google drive. | 08:58 |
| Me1234 | Simulation vcd file: https://drive.google.com/file/d/0B5U7b-LVTCGtWXJ3ZDVhcExHTVE/view?usp=sharing | 09:11 |
| olofk | Me1234: Cool. I'll take a look at it. Hang in there. We'll get this sorted out eventually :) | 10:39 |
| stekern | Me1234: you get a ibus error on address 0xf0000194 | 12:28 |
| stekern | how large is the memory space for your ROM? | 12:28 |
| Me1234 | olofk, stekern:I understood it. Again wb_ram size 64 in wb_intercon.conf | 13:10 |
| Me1234 | olofk, stekern: Bootrom works with ORPSoC V3. Linux boots. | 14:07 |
| stekern | Me1234: \o/ | 14:13 |
| stekern | olofk: well, this kind of noise: http://oompa.chokladfabriken.org/openrisc/jor1k/demos/lk.html | 14:14 |
| olofk | Me1234: Awesome! | 14:17 |
| olofk | stekern: Cool! I love that we can fire up jor1k to show off new software quickly | 14:18 |
| stekern | I think I'll need the restart timer for lk to work properly though | 14:19 |
| olofk | Similar to the RTEMS issue? | 14:21 |
| stekern | well, I got from the rtems thread that jor1k only support continous mode, but I use the restart mode in lk | 14:22 |
| stekern | and the clock_tests doesn't work | 14:22 |
| stekern | neither does the thread_tests that uses the timer for timeouts | 14:23 |
| mor1kx | [mor1kx] bandvig pushed 1 new commit to withfpu: https://github.com/openrisc/mor1kx/commit/194f4bb690f5d80f8d69b51b144650137764e3ff | 14:37 |
| mor1kx | mor1kx/withfpu 194f4bb Andrey Bacherov: More pipelined division | 14:37 |
| Me1234 | In what units is wb_ram size? | 14:46 |
| Me1234 | I mean paremeter WB_RAM_MEM_DEPTH | 14:47 |
| Me1234 | I am confused about that | 14:48 |
| Me1234 | wb_ram_generic | 14:48 |
| Me1234 | #(.depth(depth/4), | 14:48 |
| Me1234 | What is current directory when cpecifying relative paths in orpsoc_top.v ? | 15:52 |
| Me1234 | I want to make orpsoc-cores with bootrom my computer's paths independent | 15:53 |
| Me1234 | . | 15:53 |
| Me1234 | Orpsoc with bootrom here now https://github.com/VladimirP1/orpsoc-cores. Help me with last asked question. | 15:58 |
| olofk | Me1234: Paths are relative to the tool working directory | 20:12 |
| olofk | e.g. build/de0_nano/bld-quartus | 20:13 |
| olofk | So generally it's ../src/<core>/path/inside/core | 20:13 |
| ysionneau | olofk: any news from matt about NetBSD port to openrisc ? | 20:20 |
| olofk | ysionneau: No. Haven't heard back | 20:26 |
| olofk | For GSoC, I think we already missed the deadline unfortunately | 20:26 |
| ysionneau | has openrisc ever been picked by google as a gsoc organization? | 20:36 |
| olofk | ysionneau: We've never applied | 20:37 |
| ysionneau | good, you never got refused :) | 20:37 |
| olofk | Exactly :) | 20:37 |
| olofk | Me1234: I pushed a basic bootloaders repo to https://github.com/olofk/or1k_bootloaders Plan is to add the SPI boot loader here as well | 20:38 |
| olofk | Maybe I should have put the bootloaders under the openrisc org. Well, can move it later | 21:03 |
| --- Log closed Tue Feb 24 00:00:54 2015 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!