IRC logs for #openrisc Wednesday, 2014-08-27

--- Log opened Wed Aug 27 00:00:25 2014
-!- masoud_ is now known as mpoloton00:07
olofkYes! port 333306:07
stekernI've just discovered uinput06:09
olofkuinput is cool. I used it on my rpi to control mplayer from a tac2 soldered to the pin header06:10
stekernI'm going to use it to get keyboard and mouse to the or1k core on the sockit board06:11
stekernspeaking of tac2, I built a 2-port tac2->usb-joystick device with an avr and v-usb to be able to play c64 games on emulators the way they were meant to06:14
olofkThat's the way :)06:14
stekernamiga games too06:14
olofkAnd Mega Drive too actually06:14
olofkIf you can get away with a single button :)06:15
stekernright, but I have megadrive controllers06:16
stekerncan't remember if I implemented support for more than one button though ;)06:16
olofkWe discovered that you could see a level map in Alien Breed on amiga if you hooked up a mega drive controller and pressed B06:17
olofk_franck_: What happened with the UART over JTAG stuff? Is there a fusesoc system available that uses it?06:22
stekernlooking at pinouts, that would mean middle mouse press06:23
stekernI found the 'schematic' for the converter:
stekernlooks like I only did 1 button06:24
-!- Netsplit *.net <-> *.split quits: poke53282, slp, simoncoo1, arokux, ssvb, jeremybennett_, olofk08:10
_franck_that should be jsp ready10:57
_franck_and UART over JTAG is now part of openocd10:57
maxpalnok, so about to attack this fusesoc implementation in anger - from the comments so far it seems my first hurdle will be the fact that our Diamond/Synplify tool flow isn't part of the existing commands.14:43
maxpalnSo what's the easiest way to get the tools to give me the RTL I want - that way I can manually run it through the tools, check it works, then backin the commands etc?14:43
maxpalnjust starting to familiarise myself with the other systems14:54
maxpalnSome of it makes sense - for example, I can see the purpose of the <>.system and <>.cores files, similarly the data/wb_intercon.conf appears to define what peripherals will appear on the bus14:55
maxpalnbut the <>.core includes an RTL section - I thought this would be just the manually created files (since it includes the orpsoc_top.v, clkgen.v ddr2_*.v for the atlys board)14:57
maxpalnbut also in this list is the rtl/verilog/wb_intercon.v - which I understand is auto generated by the tools.14:57
maxpalnSo perhaps this is the list of files that aren't covered by a repository somewhere - which would make sense, since the repository stored files can be downloaded by the tools from the latest version.14:58
maxpalnBut it brings me back to the question of how to get the tools to generate the wb_intercon.v file....14:58
maxpalnis this done by a different utility? The options under 'fusesoc --help' don't seem to cover this - the closest would be 'build', perhaps this generates the wb_intercon.v if it doesn't already exist....15:01
maxpalnor is out of date with respect to the data/wb_intercon.conf file15:02
maxpalnmakes sense - doesn't mean it's right!!15:02
-!- Netsplit *.net <-> *.split quits: rah, xlro, maxpaln, zama, wallento, pecastro, fotis2, blueCmd, ysionneau, LoneTech, (+23 more, use /NETSPLIT to show all of them)15:19
-!- Netsplit over, joins: masoud__, O01eg, maxpaln, stekern, pecastro, ssvb, olofk, poke53281, aburgess, kiwichris (+23 more)15:23
stekernmaxpaln: the wb_intercon.v file is generated by a utility in orpsoc-cores15:27
stekerncalled wb15:27
maxpalnaha, great - I'll check it out15:28
maxpalnhmmm, my orpsoc-cores/cores folder doesn't contain that utility -15:31
maxpalnI have wb_intercon but noth the _gen version!15:31
maxpalnshould it be in the orpsoc-cores folder itself? I only have 'cores', 'systems' and in that folder15:32
maxpalnaha - its in cores/wb_intercon/sw/15:41
maxpaln :-)15:41
maxpalnwell, that seems to work - this shouldn't be that difficult, at least getting the RTL generated and compiling. Debugging any issues that arise is another thing...15:45
olofkIt's really annoying that I must have a clean tree to rebase a git branch18:32
stekernolofk: git stash; git rebase; git stash pop18:36
stekernprboom runs pretty smooth when running mor1kx at 100MHz18:37
stekernand my netcat/uinput hack worked pretty well too18:37
olofkGrr... my rebase failed. I got two diverging branches now18:39
olofkI thought it would apply the patches from my branch on top of the other branch18:40
stekernit will... if you do it right18:41
olofkOr maybe not... it might just be telling me that the remote branch (on github) diverges with my local18:41
olofkSo I just need to force-push my local branch to github, right?18:42
stekernerr... I don't know what you're doing ;)18:42
stekernforce-pushing when there are question marks around sounds like a bad idea18:42
olofkI have a branch of orpsoc-cores called de0_fixes, that I pushed to github some time ago18:44
olofkNow I want to resume work on that branch in my local tree, and I started by rebasing it to the master branch18:44
olofkI guess that made the remote and local branches diverge. I don't really need the remote branch however as that was mainly used to get comments for the pull request18:46
stekernah you are force-pushing to the de0_fixes branch?18:48
stekernthat should be ok18:48
stekernbut if you just want to work on it locally, as you say, you don't really need to push it (yet)18:49
stekernand the diverge thing is just a status info, telling you about the difference between your local branch and the remote branch18:50
olofkYeah, that's true18:50
olofkGit messages just tend to scare me in general18:51
olofkIt usually means I have fucked something up18:51
stekernthe nice thing is, that even if it's pretty easy to fuck things up, it's usually not irrepealable18:53
olofkVery true18:54
olofkhmm... icarus doesn't like the i2c controller18:57
-!- Netsplit *.net <-> *.split quits: kiwichris18:59
stekernno openrisc regression in neither 3.16 nor 3.17-rc219:01
-!- Netsplit over, joins: kiwichris19:01
stekernor actually, there was one ethoc regression in 3.16, but that commit got reverted later19:01
olofkstekern: I created a mor1kx-2.2 core for orpsoc-cores that checks out the v2.2 tag. I'm thinking of letting the existing systems depend on that. Whaddayasay?19:27
olofkExcept for mor1kx-generic19:27
stekernpoke53281: did you get scummvm working with musl?19:50
poke53281type "help"21:49
poke53281The options are all from uclibc. Not sure if you need them anymore21:51
poke53281especially the static-libgcc. Can't even remember why.21:51
poke53281Do you would like to run dosbox too?21:52
-!- FreezingCold is now known as NegativeForty22:14
--- Log closed Thu Aug 28 00:00:26 2014

Generated by 2.15.2 by Marius Gedminas - find it at!