| --- Log opened Wed Oct 23 00:00:52 2013 | ||
| poke53281 | LOOOOOLLLL This is unvvelievable ..... I have worked with qemu user mode for a while. And in one second I became the idea, if chroot might work. And really IT WORKS!!!1!!! I can chroot to my OpenRISC sysroot. Start any program I want. I freak out. | 03:57 |
|---|---|---|
| poke53281 | Ok, uname shows the wrong processor, the environment variables are totally wrong, and gcc does not link because of one unsupported syscall. But nothing which cannot be fixed. | 04:02 |
| hansfbaier | poke53281: I don't understand fully what you mean, but it seems to be cool. | 04:54 |
| poke53281 | you know the Linux command chroot. You know what it does? | 04:59 |
| stekern | bah, I wasted another hour on a non-existing problem this morning | 05:02 |
| stekern | I spent last evening setting up jackd/ardour so I can record into my workstation | 05:03 |
| stekern | and this morning I recorded an "ultimate pitchbend" (going through all notes available in the midi-spec, and stepping 1 cent at a time between them) | 05:04 |
| stekern | but when I recorded a sawtooth wave, it was "inverted" i.e., it rose from -max to max instead of falling from max to -max | 05:05 |
| hansfbaier | poke53281: yes | 05:06 |
| hansfbaier | poke53281: I don't understand the environment you are working in | 05:06 |
| hansfbaier | the context | 05:06 |
| stekern | turned out it's my mixer or my soundcard that did that, when scoping the signal on the lineout, it looks fine... | 05:06 |
| poke53281 | Lol, what mixer setting changes the sign of input signal? Is there a reason? | 05:09 |
| poke53281 | hansfbaier: You know what qemu user mode is doing? | 05:09 |
| poke53281 | I can compile binaries for OpenRISC. But they don't work on my computer. Of course, because it is an x86 compatible PC. | 05:10 |
| poke53281 | QEMU User mode translates the compiled binary into x86 machine code and executes it. | 05:10 |
| poke53281 | So with QEMU I can execute or1k binaries without an emulator. | 05:12 |
| poke53281 | Now combine this with chroot. | 05:12 |
| stekern | poke53281: well, it has an equalizer, so the signal is filtered | 05:12 |
| stekern | but I had it in "neutral" on all knobs, so I wouldn't have expected the signal to be altered that much | 05:13 |
| poke53281 | Me too. | 05:19 |
| hansfbaier | poke53281: so the openrisc chroot can execute x86 programs | 05:20 |
| hansfbaier | ? | 05:20 |
| poke53281 | No, the x86 chroot can execute OpenRISC binaries. | 05:21 |
| poke53281 | It works like chroot but you land in a OpenRISC environment | 05:21 |
| poke53281 | It runs on an x86-compiled-kernel. | 05:22 |
| poke53281 | hansfbaier: This can be used as an easy development environment without the cross-compile problems. | 05:25 |
| hansfbaier | poke53281: ah, ok. Very nice. | 05:26 |
| olofk_ | poke53281: That's awesome. Once that works smoothly it would make it much easier to create a distribution I guess | 06:12 |
| poke53281 | exactly. And also the gcc testsuite would be much easier to execute and much faster. | 06:18 |
| poke53281 | I am curious how fast I can make this emulator. At the moment it looks very good. | 06:19 |
| Powermaniac | poke53281: Your the creator of the Javascript emulation of OpenRISC right? | 06:24 |
| poke53281 | Yes | 06:24 |
| Powermaniac | I was wondering why did you choose Javascript? | 06:24 |
| poke53281 | you asked me the same question already a few days ago. | 06:24 |
| Powermaniac | Oh sorry | 06:24 |
| Powermaniac | It came up again in a conversation and I couldn't recall why | 06:25 |
| poke53281 | Why, I did it in Javascript? I think there are many reasons. | 06:26 |
| poke53281 | Because you can. Is this enough? | 06:26 |
| Powermaniac | Sure | 06:26 |
| poke53281 | In the beginning it was learning of Javascript. And learning about the OpenRISC CPU. | 06:30 |
| Powermaniac | Oh, interesting way to learn about both really | 06:30 |
| poke53281 | At the moment Javascript is the only true language that works everywhere. | 06:30 |
| poke53281 | Java was a failure in this sense. | 06:30 |
| Powermaniac | It is? Hmm news to me. For whatever reason the guys I was talking to about it, at UniSA were like Javascript?!? | 06:31 |
| poke53281 | I don't understand what you want to say. | 06:32 |
| poke53281 | It is true that Javascript is not a nice language. It has a lot of unfixable problems. A lot of developers don't like it and I understand why. | 06:32 |
| Powermaniac | Well they found it funny that Javascript was used to design an emulation of OpenRISC. | 06:33 |
| poke53281 | Ahh, yes. But one click and everyone can try it. Try this with or1ksim. | 06:33 |
| Powermaniac | Yeah it is pretty amazing what you have done | 06:34 |
| poke53281 | The last half year, after I realized how far I can go, when I realized the maximum speed that is possible, I decided to try to run X and wayland on the emulator. | 06:36 |
| poke53281 | And I have to admit, that one of my motivations was to get news on several news sites. And that happened the last two months. | 06:37 |
| Powermaniac | It sure did, ha | 06:37 |
| Powermaniac | I have to say all of you guys are pretty awesome, without this IRC channel I probably wouldn't be on my new path to getting an electrical and electronic engineering degree | 06:42 |
| poke53281 | You are welcome :). This is great! | 06:43 |
| poke53281 | By the way. I am physicist. | 06:44 |
| Powermaniac | Actually found out today I will be able to go into uni probably next year just have to sit a test first | 06:44 |
| Powermaniac | Oh you're a physicist that is pretty awesome, what do you specialise in? | 06:44 |
| poke53281 | Computer simulations. I can simulate everything. From a falling apple to the electronic structure of solids based on Quantum mechanics | 06:46 |
| poke53281 | And I can simulate CPUs. At least the instruction set :) | 06:47 |
| poke53281 | At the momen my specialization is X-ray spectroscopy of solids. | 06:48 |
| Powermaniac | Oh okay | 06:49 |
| Powermaniac | What about the brain... | 06:49 |
| Powermaniac | >_> heh heh | 06:49 |
| poke53281 | On the other hand. Three month ago I was in the synchrotron in Saskatoon doing measurements. | 06:49 |
| poke53281 | Neuronal Networks. Yes, I did it some time ago. | 06:49 |
| Powermaniac | Oh wow | 06:49 |
| poke53281 | Neuronal networks or synchrotron? | 06:50 |
| Powermaniac | Both | 06:50 |
| poke53281 | Well I am simply interested in all this kind of stuff. Hobby and Job in some way mixed. | 06:51 |
| poke53281 | Interestingly almost all physicist see there job more like a destiny. | 06:53 |
| Powermaniac | Oh okay interesting, they actually believe in having a destiny? | 06:53 |
| poke53281 | There are professors 80 year old still going every day to the institute | 06:53 |
| Powermaniac | Who wouldn't, what an amazing job to have | 06:53 |
| poke53281 | Oh sorry, Destiny is probably the wrong english word. | 06:54 |
| poke53281 | But I can't find the correct one right now. | 06:54 |
| Powermaniac | Basically on the bleeding edge of science. | 06:54 |
| Powermaniac | No destiny makes sense, but I just find it sort of odd that scientists would believe in having a destiny... | 06:55 |
| Powermaniac | See I'm agnostic and have had a very big focus on religion, to be more specific Christianity but I stopped beliving in it a couple years ago | 06:55 |
| poke53281 | 99% are atheists. | 06:55 |
| Powermaniac | In my life* | 06:56 |
| stekern | I can't see the connection between religion and have a feeling you're destined to do something | 06:59 |
| Powermaniac | I hope to one day emulate the brain with electronics basically, or make an artificial brain and eventually achieve the technological singularity | 06:59 |
| poke53281 | I stopped a long time ago in beliebing. In some way to be physicists excludes the believe in god. | 06:59 |
| Powermaniac | Yeah they tell you otherwise at Christian schools here, that being a scientist and Christian is totally possible but it doesn't really seem that way to me | 07:00 |
| poke53281 | This great. In my opinion the neuronal network science is one of the most fascinating topics you can choose. | 07:00 |
| poke53281 | And there is money now. Billions. | 07:00 |
| poke53281 | EU and USA will spend the next decase Billions in this topic. | 07:00 |
| poke53281 | ... decade... | 07:00 |
| Powermaniac | Well that is a bonus | 07:00 |
| poke53281 | So, you want to build Skynet ;) | 07:01 |
| Powermaniac | Good I'm not going to be caught out without a job in the next few decades atleast | 07:01 |
| Powermaniac | HAHA basically...But a non-evil one haha | 07:01 |
| Powermaniac | And a smarter one. | 07:01 |
| poke53281 | Well, Skynet was not really evil. From our view maybe. | 07:01 |
| Powermaniac | Hoping it will do what Ray Kurzweil believes it will do and become a god like entity that knows all, and humans being able to combine with it...And yeah all futuristic stuff. | 07:02 |
| Powermaniac | I personally think IBM is the closest currently to making it happen it seems | 07:02 |
| poke53281 | Yes, our smartness is limited by other constrains like energy ressources. | 07:02 |
| Powermaniac | With there SyNAPSE project working on making brain like hardware, and there 'electric blood' project etc. So they are solving the hardware problems | 07:03 |
| poke53281 | Yes, IBM simulated the brain of cat a few years ago. | 07:03 |
| poke53281 | But with a lot of critics. because they excluded a lot. | 07:03 |
| Powermaniac | I want to work for IBM the only thing I would like to change is for them to create more open source content. | 07:04 |
| poke53281 | There is also the Blue Brain project. | 07:04 |
| Powermaniac | Yeah and Watson | 07:04 |
| poke53281 | Haha, yes Watson. But this was only programmed by intelligent people. | 07:04 |
| Powermaniac | True | 07:04 |
| poke53281 | But very fascinating. Watched all episodes on youtube. | 07:04 |
| Powermaniac | Although some people think we won't be able to achieve the technological singularity with average computers, we will instead have to use quantum computers. What do you think about that? | 07:05 |
| poke53281 | I have also calcuated the amount of space we need to save our brain. Only a few PB | 07:05 |
| Powermaniac | Yes only a few O_O... | 07:06 |
| poke53281 | This is definitely wrong. But there are two ways to answer why. | 07:07 |
| poke53281 | This question could be related to the assumption that the brain works by quantum mechanics. | 07:07 |
| Powermaniac | Ahh okay | 07:07 |
| poke53281 | This is true on the level chemical reactions. But there is no sign that these are important for functioning. You could approximate what happens on a cell level. And a cell is macroscopic. | 07:08 |
| Powermaniac | But could the chemical reactions be somewhat random if affected by quantum mechanics? | 07:09 |
| poke53281 | You can also say, that you need quantum mechanics to simulate a falling apple. | 07:09 |
| Powermaniac | MAking the rest of the system random as well? | 07:09 |
| poke53281 | No one would do it, but you could do it | 07:09 |
| poke53281 | Yes, but this randomness would be the noise you are observing. | 07:10 |
| Powermaniac | Oh okay, hmm that is very interesting. Never been able to ask an actual physicist questions like this before | 07:10 |
| poke53281 | Believe me, that you can average over all quantum mechanic effects. | 07:10 |
| poke53281 | The reason for this is that no ones understands quantum mechanics. And it is the hope of some people that it is based on quantum mechanics that we will maybe never really understand it. | 07:11 |
| poke53281 | To keep the "magic" | 07:11 |
| poke53281 | Classical mechanics can produce such complexity, believe me. | 07:12 |
| Powermaniac | Oh you might be able to answer one of my questions I've had for a long time. There is said to be a top speed anything can travel at, the speed of light. Now is there a limit on temperature though, a top temperature? Because if there is not couldn't you just heat something to pass the speed of light as temperature is just the movement of partciles? | 07:12 |
| Powermaniac | No teacher at high school here has ever been able to answer this question for me so =\ | 07:13 |
| poke53281 | Before I answer this. The second point. At the moment you need tremendous computing power to simulate a human brain and the power of a nuclear power plant. So people could say, that we need a quantum computer to achieve such speed necessary. But our classical brain proves exactly the opposite. It is just that our computer are not very good in simulating a brain. | 07:14 |
| Powermaniac | Yeah, that's a good point | 07:15 |
| poke53281 | If you could simulate a neuron just with a few transistors you could build a brain like ours and use only a few watts. | 07:15 |
| poke53281 | The question with the temperature is indeed an interesting problem. Well, the problem is first, how to define temperature. | 07:16 |
| Powermaniac | Which is what the IBM SyNAPSE project is trying to do | 07:16 |
| Powermaniac | I've tried contacting IBM as I want to know if they will make it in anyway open source but they never contact me back | 07:17 |
| Powermaniac | =\ | 07:17 |
| poke53281 | And the definitions I know are based on classical mechanics. | 07:17 |
| Powermaniac | Oh okay | 07:17 |
| poke53281 | But in relativistic mechanics you define the temperature over the average energy. | 07:17 |
| poke53281 | And there is no limit. | 07:17 |
| Powermaniac | Oh? | 07:18 |
| poke53281 | So, it depends how you exactly define the temperature on the relativistic scale. | 07:18 |
| poke53281 | As average speed or average energy. | 07:18 |
| poke53281 | If you define average speed than there is indeed a limit. | 07:19 |
| poke53281 | Thats the reason they switch to energy. | 07:19 |
| Powermaniac | Hmm okay, so it sounds like to me there is a top speed and that is still the speed of light | 07:19 |
| poke53281 | Yes | 07:19 |
| Powermaniac | Thank you, always wanted to know the answer to that one. Now I know it! | 07:20 |
| poke53281 | Sorry, but I can't answer the question exactly, because I don't know the relativistic definition of temperature | 07:20 |
| Powermaniac | What do you think about perpetual motion, do you think it is impossible? | 07:20 |
| Powermaniac | Oh okay well, I now have a rough idea that the top speed is still the speed of light though I guess? | 07:21 |
| poke53281 | http://en.wikipedia.org/wiki/Relativistic_plasma | 07:21 |
| poke53281 | the temperatures are then defined as energies as I said. | 07:22 |
| Powermaniac | Ahh okay | 07:22 |
| poke53281 | At the moment the is no experimental evidence for perpetual motion. I don't believe it is possible. But that doesn't mean it isn't possible. | 07:24 |
| Powermaniac | Oh okay | 07:25 |
| poke53281 | Last year there was this experimental proof of the neutrinos faster than light. | 07:26 |
| poke53281 | Of course everyone was sceptical. And it the end it was wrong. But if the experiments would be true we would accept it and try to change our theories. | 07:27 |
| poke53281 | By the way this is the difference between scientists and people believing in god. | 07:28 |
| Powermaniac | Hmm | 07:28 |
| Powermaniac | Yeah someone has supposedly proved the Bible wrong recently and is going to reveal his findings as well | 07:28 |
| Powermaniac | It turns out that Jesus was basically a character representing things that were meant to be fake history of what a Caeser did that traveled through there at that time | 07:29 |
| Powermaniac | A Roman trick to control the populace | 07:29 |
| poke53281 | You can't really proof it wrong. But there is something important in the scientific method. You know the sentence "The theory is so bad. It can't even be wrong" | 07:29 |
| Powermaniac | Well I know similar sentences yes. | 07:30 |
| poke53281 | Ok, this is another problem. But I am talking about proofs about god. | 07:30 |
| poke53281 | A theory must give the possibility to prove if it is right or wrong. | 07:31 |
| Powermaniac | Yes, I know that atleast xD | 07:31 |
| poke53281 | All theories in physics do it. If you find one experiment which fails the theory, than the theory is wrong. | 07:31 |
| poke53281 | With god you can't do it. | 07:31 |
| poke53281 | And that makes it a complete useless theory. | 07:32 |
| Powermaniac | Yeah =\ | 07:32 |
| Powermaniac | I find it sad that my parents still believe in it all, sort of why I want the technological singularity to happen before they die, as then it might be possible to make us all immortal essentially... | 07:32 |
| Powermaniac | Then of course there is the problem of entropy (The Last Question, Isaac Asimov) | 07:33 |
| poke53281 | There is a tenth dimension with small little green men and then love about us. But there is know interaction between there dimension and ours. ... This theory is bad, it can't even be proven wrong. | 07:33 |
| poke53281 | s/love/laugh/ | 07:33 |
| poke53281 | Yes, immortality is nice goal. | 07:34 |
| poke53281 | And I hope too that it will happen during my lifetime. | 07:34 |
| Powermaniac | Isn't it thought that there has to be aliens as it is a statistical improbable that there isn't. As there is just so many solar systems and planets that could be inhabited. | 07:34 |
| poke53281 | No, I am not talking about aliens. | 07:34 |
| poke53281 | The story important point the non interacting dimensions. | 07:34 |
| poke53281 | When there is no interaction I can never proof there existence. | 07:35 |
| Powermaniac | Oh right | 07:35 |
| Powermaniac | Yeah I see what you mean | 07:35 |
| Powermaniac | Oaky well I have to go for a bit, got dinner. So yeah... | 07:36 |
| Powermaniac | Anyway thanks for the great conversation | 07:36 |
| poke53281 | And I have to go to bed. Enjoy your meal. | 07:36 |
| hansfbaier | stekern: http://i.imgur.com/rtXUmGV.png | 08:27 |
| hansfbaier | That explains something. | 08:27 |
| stekern | I'm not sure what, I'm not familiar with looking at those schematics. Are there wires missing? | 08:31 |
| _franck_ | no i2c ports ? | 08:31 |
| _franck_ | hansfbaier: you must be aware that stekern does not understand GUI things :) | 08:32 |
| stekern | no "a lot of things" | 08:32 |
| stekern | that's why I suspect that it's in how that GUIy thing represents stuff | 08:33 |
| stekern | ;) | 08:33 |
| stekern | now, this explains it: https://github.com/hansfbaier/orpsoc-cores/blob/master/systems/de0_nano/rtl/verilog/wb_intercon.vh | 08:38 |
| stekern | there are no wb wires for i2c | 08:39 |
| stekern | why has he edited that by hand? | 08:39 |
| _franck_ | may be because we didn't tell him how to use the generator... | 08:40 |
| _franck_ | that no in the orpsoc user's guide | 08:41 |
| stekern | it was told in the interactive orpsoc userguide | 08:43 |
| stekern | aka irc | 08:43 |
| rokka | lol | 08:46 |
| hansfbaier | stekern: Hmm. The viewer is weird: http://i.imgur.com/RQAOPas.png | 08:48 |
| hansfbaier | I wonder though why in signaltap no i2c related stuff shows up. | 08:48 |
| stekern | hansfbaier: https://github.com/hansfbaier/orpsoc-cores/blob/master/systems/de0_nano/rtl/verilog/wb_intercon.vh | 08:49 |
| stekern | why did you hand-edit that? | 08:49 |
| hansfbaier | stekern: Is that generated? | 08:52 |
| _franck_ | yes it is, from this file: https://github.com/openrisc/orpsoc-cores/blob/master/systems/de0_nano/data/wb_intercon.conf | 08:54 |
| hansfbaier | _franck_: with wb_intercon_gen? How? | 08:55 |
| _franck_ | using this tool: https://github.com/openrisc/orpsoc-cores/tree/master/cores/wb_intercon/sw | 08:55 |
| hansfbaier | maybe I should look at the source again... | 08:55 |
| _franck_ | wb_intercon_gen wb_intercon.conf wb_intercon.v | 08:55 |
| stekern | hansfbaier: yes, and it's missing the i2c wires, which might explain the problems | 08:56 |
| hansfbaier | _franck_: But isnt the .v different from the .vh? | 08:57 |
| _franck_ | both are generated at the same time | 08:57 |
| hansfbaier | stekern: What about https://github.com/hansfbaier/orpsoc-cores/blob/master/systems/de0_nano/rtl/verilog/wb_intercon.vh#L240 | 08:57 |
| hansfbaier | ah yes, the wires.... | 08:58 |
| stekern | it's your turn to curse at verilog allowing implicit wires | 09:00 |
| stekern | I did it last | 09:00 |
| olofk_ | I did it the time before that | 09:00 |
| stekern | http://juliusbaxter.net/openrisc-irc/%23openrisc.2013-10-21.log.html#t06:30 | 09:00 |
| stekern | olofk_: irc log paste or it didn't happen! | 09:01 |
| olofk_ | http://juliusbaxter.net/openrisc-irc/%23openrisc.2013-07-31.log.html | 09:02 |
| hansfbaier | stekern: I read your comment, but didn't really know what you meant. Now I do. | 09:02 |
| stekern | olofk_: it happened! | 09:03 |
| stekern | it's a bit worrying that we depend on juliusb for things to happen... | 09:06 |
| hansfbaier | stekern: You are my hero: http://pastie.org/8423500 | 09:15 |
| hansfbaier | olofk: pushed it | 09:18 |
| _franck_ | question: is the i2c core you are using compatible with this driver ? http://lxr.free-electrons.com/source/drivers/i2c/busses/i2c-ocores.c | 09:21 |
| * _franck_ has never used any opencores i2c core | 09:21 | |
| hansfbaier | _franck_: Yes, that's what I used. From latest julius | 09:26 |
| hansfbaier | it works. | 09:26 |
| stekern | latest julius what? | 09:31 |
| _franck_ | I mean can I use the upstream driver ? | 09:33 |
| stekern | _franck_: there's no "downstream" changes to that afaik, so yes | 09:37 |
| olofk_ | hansfbaier: Great! | 09:42 |
| hansfbaier | Bitcoin price goes crazy ATM | 09:45 |
| olofk_ | Crazy good or crazy bad? | 09:45 |
| hansfbaier | $207 | 09:46 |
| hansfbaier | another parabolic rise | 09:46 |
| hansfbaier | very strong uptrend | 09:46 |
| hansfbaier | as steep as the last spike | 09:47 |
| hansfbaier | just gain $50 today | 09:47 |
| hansfbaier | gained | 09:55 |
| nvmind | good morning | 09:59 |
| olofk_ | morning | 10:05 |
| stekern | afternoon | 10:08 |
| nvmind | I am experiencing some strange behavior with mor1kx and test program I wrote | 10:15 |
| nvmind | the same program is working perfectly on or1200 | 10:15 |
| * nvmind looking for yes/no answer (once again) | 10:21 | |
| nvmind | are you able to stop at break point? | 10:21 |
| nvmind | I am using openocd + or1k-elf-gdb | 10:22 |
| hansfbaier | nvmind: no | 10:24 |
| hansfbaier | using mor1kx not | 10:24 |
| hansfbaier | didn't try with or1200 | 10:24 |
| hansfbaier | same toolchain | 10:24 |
| hansfbaier | if there's a break point the PC goes to nirvana | 10:24 |
| nvmind | hansfbaier: good for the PC... because we are stuck in hell :) | 10:26 |
| nvmind | hansfbaier: so at the moment mor1kx can't be debugged | 10:27 |
| hansfbaier | _franck_: another one with the same problem ^ | 10:31 |
| juliusb | hansfbaier: yeah, what "latest" are you talking about getting some i2c stuff from? | 10:38 |
| hansfbaier | linux kernel -> master | 10:38 |
| juliusb | oh.... you mean jonas perhaps? | 10:39 |
| stekern | juliusb: does the mor1kx-devenv tests set breakpoints? | 10:39 |
| hansfbaier | sorry jonas, yes ... :] | 10:39 |
| juliusb | stekern: good question, i have one back; sw or hw breakpoiunts? | 10:40 |
| hansfbaier | stekern: # hwclock | 10:40 |
| hansfbaier | Wed Oct 23 10:39:47 2013 0.000000 seconds | 10:40 |
| stekern | juliusb: sw should be enough | 10:40 |
| stekern | since people say they don't work | 10:41 |
| juliusb | yes i have a test for something in pronto i believe | 10:41 |
| stekern | I never use them... so I haven't noticed | 10:41 |
| stekern | I'll take a look if I can get cappuccino to run that too then | 10:42 |
| stekern | hansfbaier: ??? | 10:43 |
| stekern | I didn't ask what time it is ;) | 10:43 |
| hansfbaier | stekern: I2C RTC | 10:44 |
| hansfbaier | stekern: input: ADXL34x accelerometer as /devices/a0000000.ocores/i2c-0/0-001d/input/input0 | 10:44 |
| stekern | ah, great! | 10:45 |
| stekern | juliusb: ok, you have one for espresso as well, is that *really* espresso specific? | 10:53 |
| stekern | https://github.com/juliusbaxter/mor1kx-dev-env/blob/master/boards/generic/mor1kx-espresso/sw/tests/mor1kx/sim/mor1kx-debugbkpnttest.S | 10:56 |
| stekern | I'll test just copying it over and see what happens | 11:00 |
| juliusb | not really, there looks to be some explicit checks for delay slot traps which wouldnt work on pronto but for cappuccino it should he appropriatr | 11:01 |
| stekern | but that's for espresso sans pronto | 11:01 |
| juliusb | yep | 11:02 |
| stekern | ah, I now see what you actually wrote | 11:05 |
| stekern | hmm, it's not even going past the tb write thing | 11:08 |
| stekern | ah, you have to hardcode some things related to that in the cappuccino specific related testbench ;) | 11:12 |
| stekern | ok, now it runs... and fails miserably | 11:15 |
| stekern | nvmind: no | 11:15 |
| stekern | I'll take a closer look later tonight | 11:15 |
| stekern | but it looks like it just executes the trap instruction instead of giving control to the debug unit | 11:18 |
| stekern | err, executes the trap exception, I mean | 11:18 |
| stekern | or nirvana exception as hans calls it ;) | 11:19 |
| nvmind | well I decided to rollback to or1200 for the moment | 12:08 |
| nvmind | but we will keep one eye on mor1kx | 12:09 |
| stekern | nvmind: I will at least ask you to test the breakpoint support I just fixed | 12:26 |
| stekern | not pushed yet though | 12:27 |
| stekern | http://pastie.org/8423875 | 12:28 |
| nvmind | stekern: to be onest I was wondering if there is a guide to put together the development environment of mor1kx | 12:28 |
| stekern | basically just merged in what espresso was doing (turned out the support for breakpoints was completely missing in cappuccino) | 12:29 |
| nvmind | not that I have a lot of free time... but maybe I can contribute something :) | 12:29 |
| stekern | nvmind: this is what we have been using so far: https://github.com/juliusbaxter/mor1kx-dev-env | 12:29 |
| stekern | we're trying to unify the test environments though | 12:30 |
| stekern | hansfbaier: could you give the patch in the pastie a go too? | 12:30 |
| hansfbaier | stekern: mom | 12:32 |
| hansfbaier | stekern: BINGO. It works! Amazing! Thanks! | 12:47 |
| stekern | mom? | 12:47 |
| hansfbaier | stekern: http://pastie.org/8423931 | 12:48 |
| hansfbaier | mom = moment please | 12:48 |
| hansfbaier | Weeeeeeeeeeeeeellll, the breakpoints work now, but the LEDs don't blink | 12:49 |
| hansfbaier | But when I delete all breakpoints and cont, then the LEDs blink | 12:49 |
| hansfbaier | stekern: http://pastie.org/8423941 | 12:51 |
| hansfbaier | Hey wait, now the leds shift around, but the output of the variable leds is still constant 1. Very weird. Anyway great progress so far. Thanks! | 12:53 |
| hansfbaier | nvmind: ^ | 12:53 |
| nvmind | hansfbaier: this evening I will give it a try | 12:56 |
| nvmind | sorry :) | 12:56 |
| nvmind | but I am going to integrate a generated accelerator in my soc | 12:56 |
| nvmind | to test if everything is working :) | 12:57 |
| stekern | hansfbaier: hmm, ok. thanks for testing! | 13:13 |
| stekern | and credit go to juliusb, I just copied his espresso code over to cappuccino | 13:14 |
| stekern | maybe I can try running the ledblinker in verilator to see what happens | 13:15 |
| mor1kx | [mor1kx] skristiansson pushed 1 new commit to master: https://github.com/openrisc/mor1kx/commit/89da9cbe393749c1b1337a117889144e474b9df1 | 13:55 |
| mor1kx | mor1kx/master 89da9cb Stefan Kristiansson: cappuccino/ctrl: add support for sw breakpoints... | 13:55 |
| stekern | time to read up on http://www.linux-usb.org/gadget/ | 13:58 |
| stekern | so I can write a usb-midi device driver for sockit | 13:59 |
| stekern | ah, looks like there's not much I need to do there, just enable the existing driver for that | 14:42 |
| stekern | and then write a midi driver for sublime that I can connect it to | 14:43 |
| olofk | I'm going to take care of a few pull requests now. How can I combine two commits into one? | 18:17 |
| stekern | olofk: git rebase -i | 18:27 |
| stekern | and then mark the lower of the two commits you want to combine with an s | 18:27 |
| olofk | Yeah. I just tried that. Tree is now messed up | 18:28 |
| stekern | as in squash | 18:28 |
| olofk | You are not currently on a branch | 18:28 |
| olofk | I thought I should add 's' to the beginning of each commit I wanted to merge with the previous in the emacs window that cae up | 18:29 |
| olofk | So I did, saved and exited | 18:29 |
| stekern | umm, yes | 18:29 |
| olofk | Then I got sent back to the terminal with a message that said I was in the middle of a rebase and offered me to do rebase --continue or commit --amend | 18:30 |
| olofk | I suspected something was messed up and assumed that commit --amend was "undo" | 18:30 |
| stekern | and then when it comes to the commit that is about to be squashed you edit the commit message | 18:30 |
| olofk | emacs came up again with a commit message from the commit before the start of my rebase. | 18:31 |
| stekern | yes, amend is not "undo" | 18:31 |
| stekern | ;) | 18:31 |
| olofk | So now it seems that I have lost the HEAD of my tree | 18:31 |
| stekern | don't worry, we'll find your head | 18:32 |
| olofk | Last commit is identical to the last commit from the github tree, but it doesn't seem to be neither head, origin or master | 18:32 |
| olofk | Decapitated by git | 18:32 |
| stekern | if you want to abort an rebase, it's git rebase --abort | 18:33 |
| olofk | ahh ok | 18:33 |
| olofk | But how do I get back to a branch? | 18:33 |
| stekern | git checkout master | 18:33 |
| olofk | Is it just git branch master? | 18:33 |
| olofk | Great! Back on the branch again | 18:34 |
| olofk | It seems that there is already a rebase-merge directory, and | 18:34 |
| olofk | I wonder if you are in the middle of another rebase | 18:35 |
| olofk | How the fuck should _I_ know? | 18:35 |
| stekern | haha | 18:35 |
| olofk | Please git, don't ask me these things | 18:35 |
| stekern | run git rebase --abort now then | 18:35 |
| olofk | Just did | 18:35 |
| olofk | Round 2. Fight! | 18:35 |
| olofk | ok.. so I got emacs up with a helpful message that tells me that the command for squashing is 's' | 18:37 |
| olofk | How do I run 's' on a commit? | 18:37 |
| stekern | I wonder how I should do the midi interface between the arm core and or1k... | 18:37 |
| stekern | you change the 'pick' to 's' | 18:37 |
| olofk | ahhh | 18:37 |
| olofk | And then save and exit? | 18:38 |
| stekern | yup | 18:38 |
| olofk | Amazing! It worked | 18:40 |
| olofk | Thanks | 18:40 |
| stekern | my options are 1) a 'midi-bridge' core, with wb interface that just generates an interrupt when a register is written and read 2) some shared memory thing 3) just hook-up uarts from both sides, like a "normal" midi interface | 18:40 |
| olofk | Is MIDI usually sent over UART? | 18:41 |
| stekern | of course it worked, git always works ;) | 18:41 |
| stekern | traditional MIDI at least, nowdays everything seems to be USB | 18:41 |
| olofk | aha | 18:42 |
| olofk | UART sounds like a nice solution. You should be able to cat around MIDI streams as well then, right? | 18:42 |
| stekern | well, you kind of can do that with the midistreams under alsa anyway | 18:43 |
| olofk | Didn't know. I've only had problems with my MIDI stuff | 18:43 |
| stekern | the only pro with the uart would be that you easily could connect it to a physical midiport | 18:44 |
| stekern | i.e. perhaps more portable to other boards | 18:44 |
| olofk | Got my fancy new 6 channel sound card with MIDI now that I haven't connected since I bought it in april :( | 18:44 |
| stekern | I've got a fancy sockit with MIDI now ;) | 18:44 |
| olofk | I think that sounds like a big advantage. | 18:44 |
| olofk | I want one too!!!! | 18:45 |
| stekern | but otoh, it's just down to a small part of software in the baremetal or1k program that will differ | 18:45 |
| stekern | so I'm not sure it's a good reason enough to instantiate two extra uarts just to pass some MIDI messages between two CPUs | 18:46 |
| stekern | I never had much trouble with MIDI, but I always seem to have trouble with all other sound under Linux | 18:48 |
| olofk | I had a midi dongle and a USB sound card. Last time I tried they didn't mix well at all | 18:48 |
| stekern | I can't get my internal soundcard to work at all, so I'm using a usb soundcard too now | 18:49 |
| stekern | and I have a midi dongle as well | 18:52 |
| stekern | olofk: hmm, does a non-local core that have an orpsoc driven compatible testbench work? | 19:11 |
| stekern | apparently | 19:13 |
| olofk | stekern: wb_sdram_ctrl is a good example | 19:16 |
| stekern | true | 19:19 |
| stekern | and now my little i2s core: https://github.com/skristiansson/orpsoc-cores/commit/4021669fb18ae568592cef517a37f11b0d6845e8 | 19:19 |
| stekern | heh, now I see an "error" in the core file | 19:20 |
| stekern | wb_bfm isn't really used in that | 19:20 |
| stekern | it doesn't even have a wb interface | 19:21 |
| stekern | it might or might not in the future though | 19:21 |
| stekern | I've been thinking a lot how to handle cores producing sound streams | 19:24 |
| stekern | I'm thinking about perhaps creating a "mixer" core, and let that have a wb master port, so cpu's can use that as a DMA engine | 19:25 |
| olofk | _franck_: Is SIGINT when you press ctrl-c ? | 19:27 |
| _franck_ | yes | 19:33 |
| _franck_ | if you start a build and you ctrl+c because you want to stop it then python gives you a backtrace | 19:34 |
| olofk | Great. I've always wondered how to do that | 20:01 |
| --- Log closed Thu Oct 24 00:00:53 2013 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!
