2014-07-14T00:19:45 *** monstr has joined #rtems 2014-07-14T01:14:38 *** mkhoory-2 has joined #rtems 2014-07-14T01:14:38 *** mkhoory has quit IRC (Read error: Connection reset by peer) 2014-07-14T01:16:00 *** mkhoory-2 has quit IRC (Read error: Connection reset by peer) 2014-07-14T01:16:26 *** mkhoory-2 has joined #rtems 2014-07-14T01:19:12 *** mkhoory-2 has quit IRC (Read error: Connection reset by peer) 2014-07-14T01:19:38 *** mkhoory has joined #rtems 2014-07-14T01:24:35 *** mkhoory has quit IRC (Ping timeout: 264 seconds) 2014-07-14T02:27:25 *** kiwichris has quit IRC () 2014-07-14T03:15:53 *** kiwichris has joined #rtems 2014-07-14T04:32:19 *** beng-nl_ has joined #rtems 2014-07-14T04:33:37 *** beng-nl has quit IRC (Read error: Connection reset by peer) 2014-07-14T05:40:59 *** atgreen has joined #rtems 2014-07-14T07:33:36 *** monstr has quit IRC (Remote host closed the connection) 2014-07-14T07:34:04 *** monstr has joined #rtems 2014-07-14T07:43:53 *** Hesham has joined #rtems 2014-07-14T09:43:08 *** Hesham has quit IRC (Read error: Connection reset by peer) 2014-07-14T09:56:58 *** monstr has quit IRC (Ping timeout: 240 seconds) 2014-07-14T10:11:10 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-14T10:31:25 *** edwardk has joined #rtems 2014-07-14T10:33:18 *** edwardk has quit IRC (Client Quit) 2014-07-14T10:38:13 *** gigetoo has quit IRC (Remote host closed the connection) 2014-07-14T10:38:45 *** gigetoo has joined #rtems 2014-07-14T10:44:23 *** edwardk has joined #rtems 2014-07-14T10:50:38 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-14T10:50:59 *** mattttttttt has joined #rtems 2014-07-14T10:51:06 *** mattttttttt has left #rtems 2014-07-14T10:56:29 *** edwardk has joined #rtems 2014-07-14T10:59:46 *** edwardk has quit IRC (Client Quit) 2014-07-14T11:19:26 *** edwardk has joined #rtems 2014-07-14T11:33:01 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-14T11:34:34 *** edwardk has joined #rtems 2014-07-14T11:36:02 *** edwardk_ has joined #rtems 2014-07-14T11:39:20 *** edwardk has quit IRC (Ping timeout: 260 seconds) 2014-07-14T11:40:33 *** edwardk_ has quit IRC (Ping timeout: 240 seconds) 2014-07-14T12:19:59 *** edwardk has joined #rtems 2014-07-14T13:02:31 *** Hesham has joined #rtems 2014-07-14T15:53:05 *** Hesham has quit IRC (Read error: No route to host) 2014-07-14T16:15:58 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-14T16:42:50 *** Hesham has joined #rtems 2014-07-14T17:13:55 *** edwardk has joined #rtems 2014-07-14T17:45:08 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-14T18:00:33 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-14T18:01:19 *** edwardk has joined #rtems 2014-07-14T18:05:38 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-14T18:06:59 *** edwardk has joined #rtems 2014-07-14T18:24:15 *** Hesham has joined #rtems 2014-07-14T18:30:53 *** kiwichris has quit IRC () 2014-07-14T19:08:03 *** kiwichris has joined #rtems 2014-07-14T19:13:33 *** Hesham has quit IRC (Ping timeout: 240 seconds) 2014-07-14T19:19:24 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-14T19:21:30 *** edwardk has joined #rtems 2014-07-14T19:26:56 *** Hesham has joined #rtems 2014-07-14T19:29:53 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-14T19:39:25 Hi kiwichris 2014-07-14T19:39:31 Hi 2014-07-14T19:39:44 I hope you're fine now 2014-07-14T19:40:01 I have some problems with clock driver 2014-07-14T19:40:31 Should I define the board frequency somewhere? 2014-07-14T19:40:52 and how rtems_configuration_get_microseconds_per_tick is implemented? 2014-07-14T19:41:25 I am running ticker.exe, the output is just 2014-07-14T19:41:26 "*** BEGIN OF TEST CLOCK TICK *** 2014-07-14T19:41:26 TA1 - rtems_clock_get_tod - 09:00:00 12/31/1988 2014-07-14T19:41:26 TA2 - rtems_clock_get_tod - 09:00:00 12/31/1988 2014-07-14T19:41:26 TA3 - rtems_clock_get_tod - 09:00:00 12/31/1988" 2014-07-14T19:41:42 and then it goes to the IDLE task 2014-07-14T19:42:20 and never context switched until I get Segmentation fault error 2014-07-14T19:42:56 *** kiwichris has quit IRC (Read error: Connection reset by peer) 2014-07-14T19:43:05 *** kiwichris has joined #rtems 2014-07-14T19:43:33 kiwichris: Did you miss anything? 2014-07-14T19:43:58 Sorry. I was distracted. Just reading .. 2014-07-14T19:44:37 tyt 2014-07-14T19:44:39 This looks ok for the start or Ticker. The crash must be related to interrupt. 2014-07-14T19:44:56 There are 3 tasks started which is the 3 messages 2014-07-14T19:45:16 the timer interrupt is entered the first time and returns fine 2014-07-14T19:45:43 The microsecs per tick is defined in the configuration table which uses confdefs.h and for this example is set up for you. 2014-07-14T19:45:59 The board's freq is specific to the BSP. 2014-07-14T19:46:48 I suggest you look at some ARM bsps eg zynq. Defaults are defined in configure.ac and cna be overridden by BSPOPS 2014-07-14T19:46:56 BSPOPTS that is. 2014-07-14T19:47:06 The timer scheme I use is restartable meaning that every time a tick interrupt occurs, it restart it self and start counting again 2014-07-14T19:47:16 I am not a fan of BSPOPTS but we have no better way to control things currently. 2014-07-14T19:47:39 May this (BSPOPTS) be the problem 2014-07-14T19:47:46 ? 2014-07-14T19:47:51 Nice and the clock driver sets the values based on the required tick length and the board's clock freq 2014-07-14T19:48:01 yeah 2014-07-14T19:48:12 And it's entered two times 2014-07-14T19:48:13 You need to code to manage BSPOPTS. 2014-07-14T19:48:23 What is entered 2 times ? 2014-07-14T19:48:32 _timer exception 2014-07-14T19:49:01 This means you are re-entering the interrupts and this should not happen if they have same priority. 2014-07-14T19:49:39 For example a tick enters and you break and while you wait the tick triggers, you should not see the second tick until the first exists. 2014-07-14T19:50:03 Here I mean a break in the tick interrupt handler 2014-07-14T19:50:24 the _timer exception is entered whenever a match between TTCR (timer counter which is added every clock cycle) and TTMR[PR] (Timer period set by me) 2014-07-14T19:51:01 Sure and if you set a break point on the interrupt handler and then wait it should not re-enter until you exit. 2014-07-14T19:51:24 If your driver or target interrupt code lets this happen it is a bug 2014-07-14T19:51:26 Yeah, I am just hitting continue to reach the second _timer break 2014-07-14T19:51:39 That is fine 2014-07-14T19:51:55 If you continue, then leave the interrupt and then return 2014-07-14T19:54:15 It does not return at the third _timer entry, just hang in IDLE task 2014-07-14T19:57:37 I suspect an interrupt but somewhere. 2014-07-14T19:57:58 You have new or1k interrupt support code and a new driver. :) 2014-07-14T19:58:19 Yeah, that's what I was working on this week 2014-07-14T19:58:31 Where is the crash you talked about ? 2014-07-14T19:59:29 How can I tell? I just wait after the second entry of _timer exception handler and the program hangs after that 2014-07-14T19:59:39 Segmentation fault (core dumped) 2014-07-14T20:00:46 By after that (I mean after hitting continue again) 2014-07-14T20:06:26 You need to step around in the debugger and check the interrupt entry and exit to make sure all the required regs etc are being saved. 2014-07-14T20:06:30 brb 2014-07-14T20:08:19 I did, I am stepping on each assembly instruction now to see where is the problem 2014-07-14T20:08:25 Thanks Chris 2014-07-14T20:14:38 *** edwardk has joined #rtems 2014-07-14T21:42:08 *** gigetoo has quit IRC (*.net *.split) 2014-07-14T21:43:18 *** gigetoo has joined #rtems 2014-07-14T21:44:07 kiwichris: Now the problem has been resolved, _timer is entered at intervals as programmed each time. The problem now is that context switch is happening with IDLE task only after printing the first message to console. 2014-07-14T21:45:10 I do not follow. If there is no console message at all there are no context switches ? 2014-07-14T21:49:42 There are console message and context switches 2014-07-14T21:49:59 consoles message is as before *** BEGIN OF TEST CLOCK TICK *** 2014-07-14T21:49:59 TA1 - rtems_clock_get_tod - 09:00:00 12/31/1988 2014-07-14T21:49:59 TA2 - rtems_clock_get_tod - 09:00:00 12/31/1988 2014-07-14T21:49:59 TA3 - rtems_clock_get_tod - 09:00:00 12/31/1988 2014-07-14T21:50:15 Context switch is just happens with IDLE task 2014-07-14T21:51:16 and when _timer exception occurs again, it's processed any switch to IDLE task again, and so on 2014-07-14T21:51:30 and* 2014-07-14T22:04:11 *** edwardk has quit IRC (Ping timeout: 264 seconds) 2014-07-14T22:04:47 *** edwardk has joined #rtems 2014-07-14T22:12:19 *** ChanServ has quit IRC (*.net *.split) 2014-07-14T22:12:20 *** beng-nl_ has quit IRC (*.net *.split) 2014-07-14T22:12:23 *** verm__ has quit IRC (*.net *.split) 2014-07-14T22:12:24 *** gigetoo has quit IRC (*.net *.split) 2014-07-14T22:12:27 *** bmatt has quit IRC (*.net *.split) 2014-07-14T22:20:49 *** bmatt has joined #rtems 2014-07-14T22:20:49 *** verm__ has joined #rtems 2014-07-14T22:20:49 *** beng-nl_ has joined #rtems 2014-07-14T22:20:49 *** gigetoo has joined #rtems 2014-07-14T22:23:58 *** ChanServ has joined #rtems 2014-07-14T22:23:58 *** sinisalo.freenode.net sets mode: +o ChanServ 2014-07-14T23:18:41 *** Hesham has quit IRC (Read error: No route to host) 2014-07-15T00:23:10 *** commander has quit IRC (Remote host closed the connection) 2014-07-15T00:25:06 *** monstr has joined #rtems 2014-07-15T01:02:56 *** edwardk has quit IRC (Ping timeout: 260 seconds) 2014-07-15T01:10:36 *** edwardk has joined #rtems 2014-07-15T02:13:15 *** kiwichris has quit IRC () 2014-07-15T03:08:50 *** kiwichris has joined #rtems 2014-07-15T03:18:55 *** kiwichris has quit IRC (Ping timeout: 252 seconds) 2014-07-15T04:56:04 *** beng-nl_ has quit IRC (*.net *.split) 2014-07-15T04:56:08 *** verm__ has quit IRC (*.net *.split) 2014-07-15T04:56:09 *** edwardk has quit IRC (*.net *.split) 2014-07-15T04:56:09 *** gigetoo has quit IRC (*.net *.split) 2014-07-15T04:56:11 *** bmatt has quit IRC (*.net *.split) 2014-07-15T05:00:46 *** beng-nl_ has joined #rtems 2014-07-15T05:02:16 *** gigetoo has joined #rtems 2014-07-15T05:03:15 *** verm__ has joined #rtems 2014-07-15T05:03:53 *** bmatt has joined #rtems 2014-07-15T05:08:32 *** edwardk has joined #rtems 2014-07-15T05:59:46 *** commander has joined #rtems 2014-07-15T06:54:16 *** Hesham has joined #rtems 2014-07-15T06:59:22 *** asuolen has joined #rtems 2014-07-15T07:34:42 *** AoLaD has joined #rtems 2014-07-15T07:36:05 *** jenniferA has joined #rtems 2014-07-15T07:41:13 *** krzysiekm13 has joined #rtems 2014-07-15T07:45:41 *** gedare has joined #rtems 2014-07-15T07:45:41 *** ChanServ sets mode: +o gedare 2014-07-15T07:50:58 *** JustJanek has joined #rtems 2014-07-15T07:56:42 *** Sched has joined #rtems 2014-07-15T08:00:30 Good morning everyone 2014-07-15T08:01:03 Hi, Gedare 2014-07-15T08:01:11 hey 2014-07-15T08:01:19 Good morning 2014-07-15T08:01:59 https://docs.google.com/document/d/1dp1ZTe-wwOEDRhBOSP16rtfTcTQt-1sGaAyo-PTRRsU/edit 2014-07-15T08:02:09 Sched: would you like to go first? 2014-07-15T08:02:16 phipse can't make it today 2014-07-15T08:02:21 Nice 2014-07-15T08:02:44 Last week I have finished some task in POK. But I encounter some problem. I give solution but I am not sure it's good. So I'd rather to discussed it before I push my code to github. 2014-07-15T08:03:11 Perhaps you should push your code and ask for code review along with discussing it 2014-07-15T08:03:11 Here is the job I do last week: 1. I add interrupt storage struct to vcpu, this struct will identify what interrupt is registered by this vcpu. And also will mask the pending state of this interrupt. 2014-07-15T08:04:54 ok 2014-07-15T08:04:58 2. I add the register function in Hypercall to register interrupt into vcpu when partitions startup. 2014-07-15T08:06:05 3. in the interrupt handler of POK, I add the function to mark corresponding vcpu which has register this interrupt. 2014-07-15T08:06:46 Sched: I think this might be better written as a blog post and dealt with as a code review, so all your mentors can participate and take some time to study the code. 2014-07-15T08:07:39 unless you have some specific question for me? 2014-07-15T08:08:04 I think the code is incomplete , I need to adjust it according to our discussion. If that's better, I will discuss it after. Here is the problem: how to manage the interrupt in vcpu. We don't know what interrupt will be register in vcpu. 2014-07-15T08:08:34 No, I did not predict the absent of phipse 2014-07-15T08:09:40 ok. maybe we can discuss it later? or by email 2014-07-15T08:09:53 OK, I will push the code and discuss with phipse later. but I think pushing incomplete code and make you review is waste your time. 2014-07-15T08:10:03 Right. 2014-07-15T08:10:05 not a waste if it helps clarify your design 2014-07-15T08:10:14 OK. 2014-07-15T08:10:16 Sorry. 2014-07-15T08:10:28 sometimes words are easier to understand, sometimes drawings are easier, sometimes code is easier 2014-07-15T08:10:36 usually, all 3 help ;) 2014-07-15T08:10:44 It is OK 2014-07-15T08:11:03 Thank you. That's all. 2014-07-15T08:11:09 thanks, I guess this week you will be continuing to fix this problem then? 2014-07-15T08:11:41 Yes and then write the handler function in partitions.. 2014-07-15T08:12:09 the handler of virtual interrupt. 2014-07-15T08:12:30 ok 2014-07-15T08:12:31 thanks Sched 2014-07-15T08:12:35 JustJanek: would you like to go now? 2014-07-15T08:12:45 yep, no problem 2014-07-15T08:13:17 The past week I have tried making a separate project for my example but I haven't been able to accomplish this so I think I will just upload the current files I am using for testing on rtems to github (currently just changed the default hello world in rtems to compile my project) 2014-07-15T08:13:44 I have also gotten arinc653 partition calls working on rtems and want to document it on either my blog or a wiki page so anybody can reproduce the steps I had to take. 2014-07-15T08:13:53 good 2014-07-15T08:14:21 It still takes a few unnecessary manual steps which I'd like to automate (regarding including header files to the build directory) so I might have to ask on the mailing list about that 2014-07-15T08:15:15 ok. so what is your next goal? 2014-07-15T08:16:52 longer term goal is getting more arinc653 calls to work but for this week id like to focus on documenting my steps in a blog/wiki, asking on the mailing list about a few problems/concern (e.g. automating a few things) and upload what I've gotten so far for rtems to my github 2014-07-15T08:17:24 ok 2014-07-15T08:17:59 after you have your current approach documented, perhaps we can help you to setup the separate project approach. 2014-07-15T08:18:30 yeah, it's rather that the current virtualpok build directory only compiles the examples in rtems 2014-07-15T08:19:08 well, it should be possible to build the virtualpok bsp, and then to build an application against that. 2014-07-15T08:19:17 we can work on that. 2014-07-15T08:19:49 I tried that using the seperate example github and compiling the examples in that 2014-07-15T08:19:55 ah, ok 2014-07-15T08:20:00 but haven't gotten that to work with virtualpok for i386 2014-07-15T08:20:19 maybe try it for another bsp first 2014-07-15T08:20:22 for normal i386 it was no problem but virtualpok just doesn't seem to like it 2014-07-15T08:20:24 ok 2014-07-15T08:20:35 probably need to ask philip 2014-07-15T08:20:51 it could be something about finding the library. 2014-07-15T08:20:57 Ok thanks JustJanek anything else? 2014-07-15T08:21:09 no, thats all for now 2014-07-15T08:21:20 ok thanks. Hesham can you go now? 2014-07-15T08:21:31 Yeah 2014-07-15T08:22:02 I have workedon generating binutils and newlib patch. Besides, I worked on implementing new interrupt (exception) handler; which is working fine now. Also, I have worked on a clock driver, it's working fine now. I have a problem with ticker sample, it's emitting the first three lines only to stdout, and hangs at the IDLE task, and the tick timer interrupt keep arising, processed, and return back to the IDLE task, and so on. 2014-07-15T08:22:58 That's where I am stuck at 2014-07-15T08:23:42 That is a broken context switch probably. 2014-07-15T08:24:05 Or, maybe broken watchdog chain 2014-07-15T08:24:25 I will debug more in context switch code 2014-07-15T08:24:27 Yeah, the tasks get put on the watchdog chain 2014-07-15T08:24:34 What about watchdog? 2014-07-15T08:24:42 when they go to sleep 2014-07-15T08:24:55 They should get dequeued from there after the right number of ticks 2014-07-15T08:25:04 You can check that number in their Watchdog_Interval 2014-07-15T08:25:51 It should be decreased at every tick ? 2014-07-15T08:25:52 and also find out what the Ticks_since_boot value is. I don't remember the exact name of that variable, but it is the global counter of clock ticks 2014-07-15T08:26:22 yeah, the head of the chain gets decremented each clock tick 2014-07-15T08:26:31 during Watchdog Tickle 2014-07-15T08:26:49 when it hits 0, that node should get dequeued / "Fired" 2014-07-15T08:26:57 along with any other nodes in the chain behind it with a value of 0. 2014-07-15T08:27:23 OK, where should I set break points for debugging this? 2014-07-15T08:27:31 Files or functions... ? 2014-07-15T08:27:40 So, take a look at (1) are the 3 tasks enqueued on the Watchdog Chain, (2) does the Watchdog Interval get adjusted, (3) what happens when it hits 0 2014-07-15T08:27:45 score/src/watchdog* 2014-07-15T08:27:58 There is one main function... 2014-07-15T08:28:04 I think it is called Watchdog_Tick 2014-07-15T08:28:18 OK 2014-07-15T08:28:23 There is a function I should implement 2014-07-15T08:28:24 or1ksim_clock_nanoseconds_since_last_tick 2014-07-15T08:28:33 But instead I return 0 2014-07-15T08:28:40 http://git.rtems.org/rtems/tree/cpukit/score/src/watchdogtickle.c 2014-07-15T08:28:41 May this be the problem? 2014-07-15T08:28:57 no, I don't think so 2014-07-15T08:28:58 I followed raspberry and arm implementation 2014-07-15T08:29:11 That value is for sub-tick interval timing 2014-07-15T08:29:19 so you can do fine-grained clocks 2014-07-15T08:29:25 Thanks, I will debug at the file you provided 2014-07-15T08:29:26 aka high-resolution timers 2014-07-15T08:29:39 ok 2014-07-15T08:29:57 you should be able to see that the watchdog chain holds the three tasks 2014-07-15T08:30:03 and check the interval values 2014-07-15T08:30:10 Anything else Hesham? 2014-07-15T08:30:11 And decremented every interrupt 2014-07-15T08:30:17 I see 2014-07-15T08:30:35 Next, I will work on timer benchmark, TLS 2014-07-15T08:31:08 ok. I think there is also the cpu counter? 2014-07-15T08:31:09 Ah, what should I do about newlib patch 2014-07-15T08:31:35 yes, um, about that, I'm not so sure. Maybe you can work to help improve the or1k-newlib, since Stefan thinks they will get its license fixed 2014-07-15T08:32:00 Or, you just push for your change with newlib upstream 2014-07-15T08:32:21 and the openrisc folks have to deal with the inconsistency 2014-07-15T08:32:27 I have sent out the patch and someone said it seems fine 2014-07-15T08:32:54 And asked for ChangLog and some explanation why or32, or16 have been removed 2014-07-15T08:33:11 Ok you should answer that 2014-07-15T08:33:15 Anyway, I will see what will they do, and delay this task. 2014-07-15T08:33:27 It is kind of a race. 2014-07-15T08:33:40 I did, but I do not know why my e-mails keep rejected there (to newlib mailing list) 2014-07-15T08:33:44 If you get your change into newlib, then the openrisc newlib maintainers have to handle your code when they merge with upstream 2014-07-15T08:33:47 Even it's a plain text 2014-07-15T08:34:02 you don't get an explanation about the rejection? 2014-07-15T08:34:10 double-check that your emailer isn't sending html code. 2014-07-15T08:34:19 gmail "plain-text" must be set 2014-07-15T08:34:26 Hi. This is the qmail-send program at sourceware.org. 2014-07-15T08:34:26 I'm afraid I wasn't able to deliver your message to the following addresses. 2014-07-15T08:34:26 This is a permanent error; I've given up. Sorry it didn't work out. 2014-07-15T08:34:26 : 2014-07-15T08:34:26 Invalid mime type "text/html" detected in message text or 2014-07-15T08:34:26 attachment.  Please send plain text messages only. 2014-07-15T08:34:27 See http://sourceware.org/lists.html#sourceware-list-info for more information. 2014-07-15T08:34:27 Contact newlib-owner@sourceware.org if you have questions about this. (#5.7.2) 2014-07-15T08:34:44 yeah you need to make sure your mailer isn't sending HTML 2014-07-15T08:34:56 OK I will check again 2014-07-15T08:35:13 if you use Gmail it is the "plain text" mode button you press 2014-07-15T08:35:27 ok thanks Hesham 2014-07-15T08:35:31 Thanks 2014-07-15T08:35:32 krzysiekm13: can you go now? 2014-07-15T08:36:23 yes, of course 2014-07-15T08:36:38 During last week I made quite good progress. I have working coverage within rtems-test framework, only by adding --coverage option to rtems-test command. I have also added checking if specified BSP supports coverage. This check is performed before test execution starts, so it will stop the process and notify user if requested coverage analysis can not be performed. After tests are run, I perform coverage analysis, which means t 2014-07-15T08:37:10 great 2014-07-15T08:37:16 After tests are run, I perform coverage analysis, which means that I load symbol sets configuration file, and for each symbol set I execute covoar. After that, final report is generated. This report contains index of symbol sets with proper links to its reports, and brief summary of total analysis. However, this report is only a stub now :) I will work on this in next days. 2014-07-15T08:37:26 Up to now, symbols are not generated within coverage module, because covoar will be responsible for generation of symbol list for each symbol sets. So, for tests I have prepared single symbol list file for each symbol sets and pass it as an command line argument to covoar. So in next weeks, the work will be also in this field. 2014-07-15T08:37:42 I think that in next week I will focus on final report generation and adding exception handling to coverage module. I will also refactor it a bit and I am going to publish it on my github to get some review from Chris. 2014-07-15T08:38:04 ok perfect. 2014-07-15T08:38:15 This all sounds good. 2014-07-15T08:38:31 any questions now? 2014-07-15T08:38:46 and I think that after that I will switch to extending covoar to allow for symbol list generation within it 2014-07-15T08:39:04 Now, was there some discussion about porting covoar to python? or not? 2014-07-15T08:39:35 from what i know, there was no such idea 2014-07-15T08:39:39 ah, ok. 2014-07-15T08:39:57 It is C++ right? 2014-07-15T08:40:27 but Chris suggested that generation of symbols should be performed within covoar with the use of some existing code from his private repo 2014-07-15T08:40:29 yes 2014-07-15T08:40:45 Ok 2014-07-15T08:40:54 Sounds good. Thanks for the update. 2014-07-15T08:41:10 asuolen: would you like to go now? 2014-07-15T08:41:21 :-) Thank you 2014-07-15T08:41:28 sure 2014-07-15T08:41:53 This past week I have moved all the GPIO API to the Raspberry Pi BSP directory, and I am currently working on the I2C/SPI buses. I am now in the begining of the SPI interface implementation because it seems a bit simpler than I2C. Right now I am working with the interface setup on the GPIO main header and the slave select. I still need to setup the transfer mode so data can be sent on the bus (set the clock and deal with the data FIFOs) and corre 2014-07-15T08:42:38 By next week I would like to have at least the SPI bus working in polled mode if nothing unexpected happens, so I will be working on that. 2014-07-15T08:43:35 ok. If the GPIO API is "stable", you can submit it? 2014-07-15T08:44:00 on a patch? 2014-07-15T08:44:07 yeah 2014-07-15T08:44:18 mail to rtems-devel 2014-07-15T08:44:22 *** krzysiekm13 has quit IRC (Quit: Page closed) 2014-07-15T08:45:07 I can do that. But maybe is best to hear a OK from alan or amar on the code before 2014-07-15T08:45:46 ok 2014-07-15T08:46:09 Will try to have that done then. 2014-07-15T08:46:14 thanks, anything else? 2014-07-15T08:46:39 I think not 2014-07-15T08:46:56 ok thanks. 2014-07-15T08:47:00 AoLaD: You can go now 2014-07-15T08:47:08 Hi Gedare 2014-07-15T08:47:30 I am afraid that I was not very productive last week. Besides promised no HW I was without connection and even without electricity (few hours during the storm). Most of the time, I was working offline with reference manuals and prepared routines for controlling timer, interrupts and serial. I am testing them with the board since today and I’ll try to put them in project right after they prove 2014-07-15T08:47:35 their correctness. 2014-07-15T08:48:52 ok. do you have any questions right now about it? 2014-07-15T08:49:13 do you still have the same bug as last week? 2014-07-15T08:49:22 not about the code. I may have few questions about the drivers when I'll be filling them in 2014-07-15T08:49:33 *** Hesham has quit IRC (Ping timeout: 240 seconds) 2014-07-15T08:49:51 ok. please send me emails, and CC joel and pavel 2014-07-15T08:50:04 *** Hesham has joined #rtems 2014-07-15T08:50:06 not really we have hello word running 2014-07-15T08:50:21 ok great. this was just from changing linkcmds? 2014-07-15T08:50:32 we managed to solve it day after the meeting. Since we are running from flash 2014-07-15T08:50:43 That's all that I saw you did on github, or is there more code you should push? 2014-07-15T08:50:54 we were rewriting it during the program 2014-07-15T08:51:14 OK, well hello-world is great progress :) 2014-07-15T08:52:00 Ok thanks AoLaD 2014-07-15T08:52:04 it is, but it is about 10 days behind the sheduce 2014-07-15T08:52:18 *** guerby has quit IRC (Quit: Leaving) 2014-07-15T08:52:25 thank you, I'll keep you noted as you wanted last week 2014-07-15T08:52:37 Yeah, some schedule slip is inevitable. Focus on the interrupt handling / clock timer now 2014-07-15T08:52:39 I'm not going home again anytime soon :) 2014-07-15T08:52:50 OK. The next milestone is to get ticker running. 2014-07-15T08:52:52 *** guerby has joined #rtems 2014-07-15T08:53:06 I'm on it, thank you 2014-07-15T08:53:22 *** gigetoo has quit IRC (Remote host closed the connection) 2014-07-15T08:53:27 OK. That's it for the meeting, please update the gdoc 2014-07-15T08:53:31 https://docs.google.com/document/d/1dp1ZTe-wwOEDRhBOSP16rtfTcTQt-1sGaAyo-PTRRsU/edit 2014-07-15T08:54:54 *** gigetoo has joined #rtems 2014-07-15T08:55:46 *** JustJanek has quit IRC (Quit: Leaving) 2014-07-15T09:15:02 *** asuolen has quit IRC (Quit: Lost terminal) 2014-07-15T09:17:03 *** AoLaD has quit IRC () 2014-07-15T09:28:49 *** beng-nl_ is now known as beng-nl 2014-07-15T09:33:12 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-15T09:48:34 *** Sched has quit IRC (Ping timeout: 246 seconds) 2014-07-15T09:57:13 *** gedare has quit IRC (Quit: Leaving) 2014-07-15T10:08:01 *** Hesham has joined #rtems 2014-07-15T10:09:10 *** sched has joined #rtems 2014-07-15T10:10:36 *** jenniferA has quit IRC (Quit: Leaving) 2014-07-15T10:15:52 *** sched has quit IRC (Ping timeout: 246 seconds) 2014-07-15T11:17:45 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-15T12:59:34 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-15T14:18:40 *** monstr has quit IRC (Quit: Leaving) 2014-07-15T14:38:01 *** mattttttttttttt has joined #rtems 2014-07-15T15:41:07 *** Hesham has joined #rtems 2014-07-15T15:57:01 *** edwardk has joined #rtems 2014-07-15T16:48:27 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-15T17:31:33 *** kiwichris has joined #rtems 2014-07-15T18:18:31 *** mattttttttttttt has quit IRC (Ping timeout: 246 seconds) 2014-07-15T18:42:42 beng-nl, just checking the header now 2014-07-15T18:43:03 okay 2014-07-15T18:43:20 you wrote it i believe :) 2014-07-15T18:43:27 beng-nl, I also think I managed to wipe out my OpenOCD MIO change for the bxm 2014-07-15T18:43:42 It was in a branch and I think I deleted it. 2014-07-15T18:44:02 if it was ever committed.. git reflog may be able to help you 2014-07-15T18:44:04 Did I happen to send it to you ? 2014-07-15T18:44:15 It is not in the repo 2014-07-15T18:44:19 oh the MLO 2014-07-15T18:44:32 Yeah to boot the core 2014-07-15T18:44:38 coulda sworn i saw it there 2014-07-15T18:45:47 where ? 2014-07-15T18:46:19 beagleboardxm.mc 2014-07-15T18:46:28 or.. oh.. openocd.. sorry 2014-07-15T18:48:27 Found a copy 2014-07-15T18:48:30 hmm what would .. okay 2014-07-15T18:48:36 (the filename be) 2014-07-15T18:49:17 It is the same as the one in git but it has 'proc beagleboard_xm_mlo' 2014-07-15T18:49:50 I have my black wired up and code built 2014-07-15T18:52:02 the file looks fine. Please send it in. 2014-07-15T18:52:18 Has Joel pushed your other changes ? 2014-07-15T18:52:36 the uart one yes 2014-07-15T18:52:44 Great 2014-07-15T18:53:29 the assert one jennifer did 2014-07-15T18:53:39 kiwichris: cool thanks 2014-07-15T18:53:51 i've done pretty good tests btw, i'm sure the functionality is ok 2014-07-15T18:53:55 i'll make you the author 2014-07-15T18:54:20 I do not mind who is on the code. 2014-07-15T18:56:52 mrm actually 2014-07-15T18:57:10 there is some mention of A9.. there, is that a leftover from something or another? 2014-07-15T19:11:05 Just looking. 2014-07-15T19:12:36 The A8 has the cp15 vbr ? 2014-07-15T19:13:33 It is the same vbr code from the a9mpcore 2014-07-15T19:14:06 i just mean the comments 2014-07-15T19:16:07 Please change the comment. 2014-07-15T19:16:11 ok 2014-07-15T19:16:12 did 2014-07-15T19:30:45 *** Hesham has joined #rtems 2014-07-15T19:45:45 kiwichris: sent 2014-07-15T19:46:23 great. 2014-07-15T19:46:49 once that is merged, the rest of the beagle bsp is just new bsp-specific files.. 2014-07-15T19:53:50 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-15T20:56:40 *** Hesham has joined #rtems 2014-07-15T20:57:19 *** mattttttttt has joined #rtems 2014-07-15T21:01:10 *** Hesham has left #rtems 2014-07-15T23:15:06 beng-nl, I will push that file now. 2014-07-15T23:15:12 * bmatt does a happy dance 2014-07-15T23:15:29 bb?.+ for the win 2014-07-15T23:16:38 kiwichris: have you been able to do any extensive testing on your bbb? 2014-07-15T23:16:52 I don't have a jtag setup, so I've just been hammering qemu and doing spot checks on hardware 2014-07-15T23:16:54 Yes I have 2014-07-15T23:17:09 I have only been using real hardware and JTAG 2014-07-15T23:17:16 woot. 2014-07-15T23:17:45 change pushed 2014-07-15T23:18:02 I have a couple of Flyswatter2's 2014-07-15T23:18:32 One is connected to a client's zynq board and the other a beaglebone black 2014-07-15T23:19:40 Ben and I were getting only a few failures on real hardware with close to 500 tests 2014-07-15T23:20:59 I haven't actually found a failed test yet when doing things manually 2014-07-15T23:21:40 oh snap! tincantools has the adapter kit back in stock 2014-07-15T23:22:23 I got the kit from Rusty at Tincantools for the Beaglebone black. 2014-07-15T23:22:50 I got the debug lead from Dontronics 2014-07-15T23:24:30 now I just need Peer to respond to my emails 2014-07-15T23:27:53 Peer ? 2014-07-15T23:29:34 dude porting the erlang VM to RTEMS 2014-07-15T23:29:44 Ah ok. 2014-07-16T00:19:18 oh hah. looks like peerst is here! hahaha 2014-07-16T00:44:57 *** kiwichris has quit IRC () 2014-07-16T01:12:57 *** kiwichris has joined #rtems 2014-07-16T02:22:15 bmatt: ? 2014-07-16T02:22:20 bmatt: emails? 2014-07-16T02:32:54 bmatt: you sent to my gmail account: I don’t read this only subscribe to mailing lists with it 2014-07-16T02:33:19 answered your mail and gave you my real email 2014-07-16T02:33:41 happy to work with you getting Erlang on RTEMS on BeagleBone (I have some whites) running 2014-07-16T04:01:52 peerst: interesting 2014-07-16T04:02:04 peerst: 'we meet again' btw, hello :-) 2014-07-16T04:11:08 *** AoLaD has joined #rtems 2014-07-16T06:01:06 beng-nl: hi 2014-07-16T06:03:16 hows the state of the beaglebone bsp’s? Ready for prime time or something still missing? 2014-07-16T07:04:23 peerst: hmm, i'm not sure how things qualify for prime time, but i'm sure the basics work :) 2014-07-16T07:05:02 there are some failing tests that i'd still like to diagnose but because of the recent interest i prefer to merge it and then iron out the tests 2014-07-16T07:07:32 oh its not merged yet? 2014-07-16T07:08:04 no it's not.. i was being a bit too cautious for the initial merge i suppose 2014-07-16T07:08:17 probably 2014-07-16T07:08:29 since a bsp normally doesn’t break anything else 2014-07-16T07:08:33 a bit new to rtems & the community, big toe in water etc. 2014-07-16T07:08:52 don't want to make a bad first impression ;) 2014-07-16T07:09:00 hah 2014-07-16T07:09:25 well building a BSP for beaglebone is already a very good first impression 2014-07-16T07:09:32 nice to hear! 2014-07-16T07:09:51 witing for this for more than a year (not having any time at all doing it myself) 2014-07-16T07:09:52 all the dependencies are merged though, so just the bsp code remains 2014-07-16T07:10:21 yeah perfect, so everything that could break anyones elses stuff is already in 2014-07-16T07:10:29 so i'll reproduce everything from scratch to (a) document it and (b) verify it works from scratch and then submit it for merging 2014-07-16T07:10:37 peerst: indeed 2014-07-16T07:11:03 looks like bmatt and myself will be using it soon to get Erlang on it 2014-07-16T07:11:12 which is also a good test 2014-07-16T07:11:29 aha, what kind of dark corners of a bsp might it challenge? 2014-07-16T07:12:02 none specifically (unless you support SMP) 2014-07-16T07:12:09 gasp no 2014-07-16T07:12:20 they're single-core 2014-07-16T07:12:20 well Erlang on RTEMS neither 2014-07-16T07:12:45 but I plan to (have a P1020 eval box on my lab table for it) 2014-07-16T07:12:47 ah ok 2014-07-16T07:13:22 well when doing the first port I found POSIX issues triggered by Erlangs massive use of readv/writev 2014-07-16T07:13:33 and some NFS bugs 2014-07-16T07:13:43 always gratifying to find & fix bugs :-) 2014-07-16T07:14:09 when the rubber hits the road 2014-07-16T07:14:18 the intial port runs quite simple in one thread 2014-07-16T07:14:56 you can run erlang single threaded, multi threaded (for I/O) and in SMP mode 2014-07-16T07:15:04 so far I only did single thread 2014-07-16T07:15:14 aha, multithreaded is not the same as smp mode? 2014-07-16T07:15:21 no 2014-07-16T07:15:35 SMP requires multithreaded of course 2014-07-16T07:16:05 erlang does one scheduler thread per core usually 2014-07-16T07:16:19 and a bunch (default 10) async I/O threads 2014-07-16T07:16:19 i see 2014-07-16T07:16:49 and recently they have dirty schedulers on 10 further threads 2014-07-16T07:17:13 these can be used for long running NIFs (Native Implemented Functions) 2014-07-16T07:17:42 since normally these need to return within 1ms otherwise they disturb the erlang shcedulers and break soft realtime 2014-07-16T07:17:51 ahh 2014-07-16T07:18:11 so now you have the choice to either break up your C code in work chunks 2014-07-16T07:18:21 or move it to a dirty scheduler 2014-07-16T07:18:30 man. i gotta get into erlang 2014-07-16T07:18:38 its big fun 2014-07-16T07:19:05 read http://learnyousomeerlang.com its the best intro and advanced book at the same time 2014-07-16T07:19:17 sounds perfect 2014-07-16T07:19:33 and its author is very nice and helpful on #erlang (he’s MononcQc) 2014-07-16T07:19:49 I wish I had it when I learnt erlang 2014-07-16T07:20:41 beng-nl: ah btw one suspected issue with termios I maybe have discovered with Erlang 2014-07-16T07:21:02 when you use line mode on a UART console 2014-07-16T07:21:17 it seems to block the thread until it got a line 2014-07-16T07:22:04 at least I can freeze erlang by typing a character on the console and not terminating 2014-07-16T07:22:17 hmm 2014-07-16T07:22:36 which doesn’t hurt me at the moment since the production system doesn’t have a console 2014-07-16T07:23:02 it could also be a “select” vs “termios” interactin 2014-07-16T07:23:20 need to talk to sebhub about it sometime 2014-07-16T07:23:43 that does sound fishy 2014-07-16T07:23:44 he extended the current (crappy) select implementation to work on all devices 2014-07-16T07:24:01 before select only worked for network 2014-07-16T07:25:33 i can sympathize, i did select for minix and it's still a bit of a jungle, i think almost by nature 2014-07-16T07:25:49 has to be reasonable under unreasonable circumstances 2014-07-16T07:26:47 we have funding for porting FreeBSD kqueues to RTEMS 2014-07-16T07:27:00 very nice 2014-07-16T07:27:22 which also will probably somehow fix select a little bit more 2014-07-16T07:27:40 "somehow" :) 2014-07-16T07:27:55 or replace it with a kqueues based solution and also give us the possibility to do a poll call based on kqueues 2014-07-16T07:28:28 I think Erlang needs to be able to put the kqueue filehandles in a select 2014-07-16T07:28:42 iirc kqueue is a superset of select 2014-07-16T07:28:46 but I have not dived down that deep in the emulator loop 2014-07-16T07:28:51 yep 2014-07-16T07:28:57 selectanything 2014-07-16T07:29:07 you can implement slect and poll only using kqueue 2014-07-16T07:29:37 Erlang does use it only for its slect like subset as a speedup 2014-07-16T07:49:55 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-16T08:46:21 bmatt: did you get the backlog? 2014-07-16T09:47:01 *** bmatt has quit IRC (Ping timeout: 240 seconds) 2014-07-16T09:47:18 *** notbmatt has joined #rtems 2014-07-16T10:04:11 *** gigetoo has quit IRC (Ping timeout: 264 seconds) 2014-07-16T10:40:00 *** edwardk has joined #rtems 2014-07-16T10:53:13 hi peerst! just arriving at the office today :) 2014-07-16T10:53:58 *** notbmatt is now known as bmatt 2014-07-16T10:54:06 haha 2014-07-16T10:54:25 us californians like to sleep far into the CET day ;) 2014-07-16T10:54:37 some of us in CET too 2014-07-16T10:58:28 okay, I'm caught up 2014-07-16T10:59:02 it sounds like it's even further along than I'd hoped 2014-07-16T10:59:20 beng-nl: did chris push your bsp stuff into mainline or into his tree yesterday? 2014-07-16T10:59:52 the two things i submitted are in rtems mainline, i haven't submitted the actual bsp yet 2014-07-16T10:59:55 just the dependencies 2014-07-16T10:59:57 now just the bsp remains 2014-07-16T11:00:15 i just want to do some last-second rebase, testing, write the doc on how to fully reproduce it etc. 2014-07-16T11:00:29 great 2014-07-16T11:01:02 I know you're much closer to the code but if there's any help I can give, I'm eager to do :) 2014-07-16T11:01:54 thank you :) 2014-07-16T11:02:49 i wonder what is the common rtems way to describe how to get an rtems .exe booting on a target.. 2014-07-16T11:03:09 i want to make it un-fragile 2014-07-16T11:03:15 and repeatable 2014-07-16T11:03:32 I think at that point, it's more or less implementation-specific, no? 2014-07-16T11:03:42 true 2014-07-16T11:03:45 well, I guess you usually use objcopy to convert it into whatever format your writer expects 2014-07-16T11:04:47 I can't think of a platform that expects relocatables, but iirc AVRisp wants hex, and MLO wants bare binaries, for example 2014-07-16T11:12:38 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-16T12:30:47 *** edwardk has joined #rtems 2014-07-16T13:35:58 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-16T13:45:48 *** edwardk has joined #rtems 2014-07-16T13:50:45 *** edwardk has quit IRC (Ping timeout: 272 seconds) 2014-07-16T14:13:16 *** edwardk has joined #rtems 2014-07-16T14:17:33 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-16T14:27:32 *** edwardk has joined #rtems 2014-07-16T15:37:54 lo 2014-07-16T16:07:11 bmatt, beng-nl: its deffo application specific 2014-07-16T16:07:27 since there are all kinds of wasy a rtems image gets booted 2014-07-16T16:07:54 it boots from reset, … does so from nor flash, nand flash whatever 2014-07-16T16:08:19 there is a bootloader that expects format X in filename Y with config in Z 2014-07-16T16:08:53 e.g. the curent project where I did the erlang port for runs on a MPC8309 2014-07-16T16:09:31 it boots a rtems based bootloader from nand flash (the way to get it there the first time is arcane with boot block and whatnot) 2014-07-16T16:10:11 this rtems already has the flash filesystem included and load another RTEMS from a naked binary to a predefined adress 2014-07-16T16:10:49 on the last project for the same customer I wrote a elf loader and loded a zipped elf directly 2014-07-16T16:10:56 and boote from nor flash 2014-07-16T16:11:05 (also bootloader in RTEMS) 2014-07-16T16:11:11 then there is uboot 2014-07-16T16:11:21 and any kind of proprietary bootloader 2014-07-16T16:11:38 and we can’t just all use uboot because its gpl3 2014-07-16T16:12:03 which means its just not in question for 90% of my embedded customers to use it 2014-07-16T16:13:05 becasue the secure way to do remote upgrades is to have signed software packets that you load and prevent anything else from going on the machine 2014-07-16T16:13:17 which looks just like DRM and is forbidden by GPL3 2014-07-16T16:14:21 and tehre is also issue with having patents and doing GPL3, which are unclear enough that company lawyers are scared, which is enough for us to just not being able to use it 2014-07-16T16:15:04 for my prducts I tend to avoid GPLed code altogether no matter whether 2 or 3 2014-07-16T16:15:17 but I digress 2014-07-16T16:15:32 booting and loading is very diverse 2014-07-16T16:15:47 so no standard way 2014-07-16T16:16:26 beng-nl: sounds like peerst concurs - feel free to put implementation-specific details in your writeup ;) 2014-07-16T16:18:28 bmatt: with the erlang stuff, I think the best way would be for me to get a basic RTEMS with network running on my beglebones 2014-07-16T16:19:05 then I can see which differences need to be addressed between my MPC8309 and this one 2014-07-16T16:19:26 a easy first shot would be to have a cross target for beaglebone for Erlang 2014-07-16T16:20:14 later I want to make this parametrizable from configure, so there need only be one cross config file per arch and not per BSP 2014-07-16T16:20:37 but with the only 2 targets currently I’d rather keep it simple 2014-07-16T16:21:38 there is still some customer specific stuff that I patch into erlang/otp tree to get a main and confdefs 2014-07-16T16:22:23 might just remove the customer specific stuff for the start and still keep it as patch in 2014-07-16T16:22:44 eventually will need to get rid of this setup though 2014-07-16T16:24:50 bmatt, beng-nl: if you have a rtems version I can get running on my beaglebone white either with your help or writeups we can get this started 2014-07-16T16:25:14 probably best would be writeups, then I can play testuser to fix the missing parts 2014-07-16T16:26:05 peerst: the quickest link I can get to you right now is https://github.com/bengras/rtems-source-builder/pull/1 2014-07-16T16:26:16 check out beng-nl's branches 2014-07-16T16:26:40 if you use vagrant, you can also use https://github.com/thenewwazoo/rtems-vagrant 2014-07-16T16:26:55 edit the branches and repo URLs to point to bengras :) 2014-07-16T16:27:49 not using vagrant at the moment … planning to 2014-07-16T16:28:04 but I’d rather get this running on my FreeBSD boxes 2014-07-16T16:28:23 for one because I try to avoid Linux 2014-07-16T16:28:39 and the next is that this gets the stuff better tested 2014-07-16T16:29:34 but for distribution a vagrant quick setup will be preferable 2014-07-16T16:30:58 I’ll look at it tomorrow 2014-07-16T16:31:17 *** kiwichris has quit IRC () 2014-07-16T16:31:30 whats the qemu stuff? 2014-07-16T16:31:41 don’t you run it on real boards? 2014-07-16T16:32:07 ah just for testing, nice 2014-07-16T16:44:03 bmatt: oh and if you want to write blogposts on www.grisp.org, just mail me either a markdown or asciidoc file 2014-07-16T16:44:31 its currently only a static site builder, so there is no remote CMS interface 2014-07-16T16:45:23 beng-nl: that blog is open to you too if you start Erlang on RTEMS or have anything related, e.g. write about the beaglebone BSP 2014-07-16T16:45:45 or anyone else doing Erlang+RTEMS related things 2014-07-16T16:46:29 peerst: yeah, I was doing a bunch of testing with qemu since I hadn't been able to get my hands on jtag stuff for my target 2014-07-16T16:51:38 peerst: I'm not sure what the state of ethernet on the beagleb* is, I haven't had time to test it 2014-07-16T16:51:40 perhaps I'll do that tonight 2014-07-16T16:51:52 beng-nl: do you know? 2014-07-16T16:52:32 *** atgreen has quit IRC (Remote host closed the connection) 2014-07-16T16:57:44 bmatt, peerst: agreed (target specific) 2014-07-16T16:58:21 peerst: thank you (blog) 2014-07-16T16:58:58 bmatt: the ethernet works on the beagleb*, not on rtems though :) 2014-07-16T16:59:25 bmatt: on the beagleboard xm it's only accessible as a usb device; on the beaglebones it is directly accessible as a memory mapped ethernet controller 2014-07-16T16:59:36 minix has support for it 2014-07-16T17:00:08 so with any luck that can be used to implement it for rtems, but i was planning to look into that after the bsp is merged and some of the wrinkles (if any) are ironed out 2014-07-16T17:11:38 *** edwardk has quit IRC (Ping timeout: 240 seconds) 2014-07-16T17:18:19 *** kiwichris has joined #rtems 2014-07-16T18:14:26 *** edwardk has joined #rtems 2014-07-16T18:20:30 *** edwardk has quit IRC (Quit: Leaving...) 2014-07-16T18:51:00 I haven't looked yet, but is there a canonical RTEMS interface for ethernet devices? 2014-07-16T18:51:40 ah, libbsd? 2014-07-16T21:29:43 *** kiwichris has quit IRC (Read error: Connection reset by peer) 2014-07-16T22:40:30 *** kiwichris has joined #rtems 2014-07-17T00:19:07 *** monstr has joined #rtems 2014-07-17T01:29:24 *** beng-nl_ has joined #rtems 2014-07-17T01:32:05 *** notbmatt has joined #rtems 2014-07-17T01:34:02 *** bmatt has quit IRC (Ping timeout: 240 seconds) 2014-07-17T01:34:08 *** beng-nl has quit IRC (Ping timeout: 240 seconds) 2014-07-17T02:57:08 *** Hesham has joined #rtems 2014-07-17T02:57:24 Hi 2014-07-17T04:45:40 *** monstr has quit IRC (Remote host closed the connection) 2014-07-17T05:01:39 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-17T05:18:59 *** edwardk has joined #rtems 2014-07-17T06:11:47 *** monstr has joined #rtems 2014-07-17T06:34:13 *** edwardk has quit IRC (Ping timeout: 250 seconds) 2014-07-17T06:34:49 *** edwardk has joined #rtems 2014-07-17T06:38:32 *** jenniferA has joined #rtems 2014-07-17T06:38:35 hi 2014-07-17T07:28:53 *** kiwichris has quit IRC (Ping timeout: 272 seconds) 2014-07-17T07:41:41 *** beng-nl_ is now known as beng-nl 2014-07-17T08:29:52 lo 2014-07-17T09:08:14 *** atgreen has joined #rtems 2014-07-17T10:11:59 *** atgreen has quit IRC (Ping timeout: 264 seconds) 2014-07-17T10:23:27 *** atgreen has joined #rtems 2014-07-17T11:19:03 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-17T11:20:12 *** notbmatt is now known as bmatt 2014-07-17T11:26:54 *** edwardk has joined #rtems 2014-07-17T11:29:11 *** Hesham has joined #rtems 2014-07-17T11:35:59 *** atgreen has quit IRC (Ping timeout: 264 seconds) 2014-07-17T11:37:13 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-17T11:55:43 *** atgreen has joined #rtems 2014-07-17T13:01:25 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-17T13:14:30 *** atgreen has quit IRC (Remote host closed the connection) 2014-07-17T13:15:20 *** atgreen has joined #rtems 2014-07-17T14:09:38 *** monstr has quit IRC (Ping timeout: 240 seconds) 2014-07-17T14:16:50 bmatt: ping? 2014-07-17T14:16:59 hi! 2014-07-17T14:18:03 *** jenniferA has quit IRC (Quit: Leaving) 2014-07-17T14:19:49 ah you are there 2014-07-17T14:20:03 re doing beaglebone and x86 in parallel 2014-07-17T14:21:04 I don’t want to suppres your enthusiasm 2014-07-17T14:21:38 but I think it makes more sense to port it to one non proprietary target (currently it only runs on customer hardware) 2014-07-17T14:22:17 after this port my goal is to have a target independent RTEMS port 2014-07-17T14:22:29 but this needs work in the makesystem 2014-07-17T14:22:31 of erlang 2014-07-17T14:22:52 because currently each target needs a cross compile config file 2014-07-17T14:23:02 i.e. each RTEMS bsp 2014-07-17T14:23:19 and we need to reduce this, to either one config for all 2014-07-17T14:23:37 or to at least only one config per arch and not per bsp 2014-07-17T14:24:04 besides x86 is a peculiar target for rtems 2014-07-17T14:24:34 that's fair 2014-07-17T14:24:35 who wants a x86 for small embedded systems that don’t run Unix/Linux/Windows 2014-07-17T14:24:50 ? 2014-07-17T14:24:53 I sort of assumed x86 was (as is common for some embedded OSes) a default sort of sandbox target 2014-07-17T14:25:16 you build for x86, run on qemu, do your testing, and then work with the hardware where the iteration time may be higher 2014-07-17T14:25:37 well since Intel has a very bad rep in embedded circles, its rather not 2014-07-17T14:25:50 why not run on arm qemu 2014-07-17T14:25:54 ? 2014-07-17T14:26:01 tests aren't entirely reliable 2014-07-17T14:26:08 ? 2014-07-17T14:26:27 why not reliable on arm but reliable on x86? 2014-07-17T14:26:30 the results of running rtems tests in linaro's qemu-arm aren't... deterministic :) 2014-07-17T14:27:00 because rtems bsp is buggy or what? 2014-07-17T14:27:18 I know that sebhub does a lot of arm qemu testing for smp 2014-07-17T14:27:19 I suspect it's because qemu doesn't emulate the entire SOC quite right 2014-07-17T14:27:33 yeah well, run it on a real chip 2014-07-17T14:27:36 I haven't diagnosed it 2014-07-17T14:27:54 because the same problem remains when you have a x86 qemu doesn’t it? 2014-07-17T14:28:15 qemu x86 is a lot more mature; in any case, I'm not married to the idea, I just thought it might be worthc onsidering 2014-07-17T14:29:06 I'll happily work directly on the BB* port if you think it's a better use of time 2014-07-17T14:30:43 well there is the tests you can do on a emulation 2014-07-17T14:31:00 and there is the stuff that needs real unfaked hardware 2014-07-17T14:31:33 right 2014-07-17T14:31:37 problem with BB is that without ethernet working in RTEMS its kind of no fun having Erlang on 2014-07-17T14:31:57 depends on your use case - I have no need for ethernet :) 2014-07-17T14:32:32 well one of the neat things having Erlang on a embedded node is that you can remsh into the node during development 2014-07-17T14:32:48 thats one of the reasons the development is speeded up 2014-07-17T14:32:52 fair enough 2014-07-17T14:33:04 you play with your hardware drivers from the Erlang shell 2014-07-17T14:33:34 getting out of the edit compile link upload via jtag run crash repeat loop 2014-07-17T14:34:04 so while its no hold up for starting with the port 2014-07-17T14:34:19 it takes the fun out of having the port quite a bit 2014-07-17T14:34:33 beng-nl: ^;-) 2014-07-17T14:36:05 bmatt: evaluating the emulator based things for getting people to try out stuff quickly without hardware is another thing 2014-07-17T14:36:44 but even then emulating already on one of the relevant embedded platforms might be more interesting for people 2014-07-17T14:37:09 I very much believe in a "more eyes" approach 2014-07-17T14:37:27 having a x86 machine with qemu-x86 and on it rtems + erlang looks to peaople like a crappy Erlang on Xen version which already exists 2014-07-17T14:37:50 bmatt: well I lost my faith in more eyes and shallow bugs 2014-07-17T14:37:55 heh I vehemently disagree; Erlang on Xen is a totally different beast 2014-07-17T14:38:10 since OpenSSL I think we are through with this eric raymond saying 2014-07-17T14:38:23 it's very tightly coupled to Xen, which is not designed to be portable 2014-07-17T14:38:37 bmatt: yeah it is different but erlang on rtems on x86 quemu looks similar 2014-07-17T14:38:39 that's why I was so excited by GRISP, because RTEMS is explicitly intended to be portable 2014-07-17T14:38:46 yeah 2014-07-17T14:38:52 similar to you, different to me :) 2014-07-17T14:39:18 different also to me, I’m just talking about impressions people may get 2014-07-17T14:40:05 * bmatt shrugs 2014-07-17T14:40:09 Erlang itself is portable to the main embedded archs so its a no brainer to have whatever arch you want 2014-07-17T14:40:23 sparc, powerpc and arm covers a big part 2014-07-17T14:40:32 right, except I have to carry Linux around too 2014-07-17T14:40:33 not sure if erlang’s ported to mips 2014-07-17T14:40:43 which I really want to stop doing :) 2014-07-17T14:40:53 no I mean for Grips 2014-07-17T14:40:55 Grisp 2014-07-17T14:40:59 ah, I see 2014-07-17T14:41:23 It is running very stable already 2014-07-17T14:41:28 well, my single goal right now is to convince you to share your source ;) 2014-07-17T14:41:51 customer has two embedded systems running a distributed erlang app on MPC 8309 on his table since a year 2014-07-17T14:41:58 constantly running never crashed 2014-07-17T14:42:18 I don’t need to be convinced 2014-07-17T14:42:19 my table is empty... won't you help? ;) 2014-07-17T14:42:22 I wnat to 2014-07-17T14:42:22 hahaha 2014-07-17T14:43:18 it needs a little bit of disentangling from the proprietary environment and run on a interesting embedded non proprieatry and cheap platform 2014-07-17T14:43:34 thats why I aim for BB 2014-07-17T14:44:20 my next hardware target is Freescale P1020 (already have it here) 2014-07-17T14:44:29 running SMP Erlang on SMP RTEMS 2014-07-17T14:44:47 sebhub is using P1020 to work on SMP 2014-07-17T14:44:57 thats why I bought one too 2014-07-17T14:45:50 Then I want to get my hands on a 12 core (24 virtual) http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=T4240 2014-07-17T14:46:23 imagine having erlang on a machine like this running right on the bare metal (with a thin RTEMS layer) 2014-07-17T14:47:11 bmatt: and re source: my goal is that you can build Erlang for RTEMS without any patches 2014-07-17T14:48:02 so the port to BB and disentangling from customer code is also meant to make a set of patches to OTP which will be submitted back 2014-07-17T14:48:24 right. so I'd like to help! ;) 2014-07-17T14:48:38 BTW what BB do you have? 2014-07-17T14:48:44 white or black? 2014-07-17T14:48:46 I think beng-nl used the phrase "high-bar paralysis" 2014-07-17T14:48:53 black at the moment, but I've got a white on order 2014-07-17T14:49:00 ok 2014-07-17T14:49:09 I have whites at the moment 2014-07-17T14:49:17 and will probably order a black 2014-07-17T14:49:45 scine it will be the more interesting platform for new users (cheaper + more bang) 2014-07-17T14:51:26 yep; that plus the PRU modules is what got my attention 2014-07-17T14:51:49 PRU? 2014-07-17T14:52:05 the AM335x has two small 200MHz coprocessors 2014-07-17T14:52:36 ok 2014-07-17T14:52:49 and you plan to do what with them? 2014-07-17T14:53:00 flip bits, obviously ;) 2014-07-17T14:53:47 they're really good for doing things like protocol decoding in software and other such lightweight DSP-type tasks 2014-07-17T14:54:11 do they already work from RTEMS? 2014-07-17T14:54:25 oh, no, I'd have to write a driver for them - they're very unique to the AM335x 2014-07-17T14:54:31 but that's entirely secondary 2014-07-17T14:54:34 ah ok 2014-07-17T14:54:51 the "1GHz at $50" got my attention first! ;) 2014-07-17T14:55:02 yeah 2014-07-17T14:55:17 BB is kick ass 2014-07-17T14:55:27 much better than rPi for embedded 2014-07-17T14:55:42 mine is so lonely... all it wants is some Erlang without Linux.... ;) 2014-07-17T14:55:50 hehe 2014-07-17T14:56:43 ok I promise to slip in a timeslice for this tomorrow 2014-07-17T14:56:52 probably setting up the build tools for one 2014-07-17T14:57:42 I'm happy to help with that - most of my work right now is building repeatable build environments for people 2014-07-17T14:58:01 in fact, I started on a FBSD vagrant setup last night but ran out of time 2014-07-17T14:58:07 ah 2014-07-17T14:58:19 I *really* hate doing things more than once :) 2014-07-17T14:58:25 I’m *very* interested in that 2014-07-17T14:58:39 I'll push my branch tonight and you can play with it 2014-07-17T14:58:46 cool 2014-07-17T14:58:57 post me a link here or by mail 2014-07-17T14:59:01 will do 2014-07-17T14:59:03 have to run now 2014-07-17T14:59:38 bmatt: and keep bugging me until the stuff is out 2014-07-17T14:59:57 peerst: I intend to 2014-07-17T15:00:08 I’m multitasking 4 projects with deadlines and and no reduction in workload in sight 2014-07-17T15:00:27 so without any external driving … nothing will happen 2014-07-17T15:01:12 peerst: git init;git add *; git commit -m "initial commit"; git add remote [etc]; git push 2014-07-17T15:01:17 ;) 2014-07-17T15:04:38 haha 2014-07-17T15:05:01 * bmatt whistles innocently 2014-07-17T15:05:12 there is no separate code 2014-07-17T15:05:20 its all patches to OTP 2014-07-17T15:05:30 and some patches to RTEMS to fix stuff 2014-07-17T15:05:41 but those should be already in RTEMS by now 2014-07-17T15:06:49 well, I'm handy with diff! ;) 2014-07-17T15:06:57 okay, I'll leave you alone for now 2014-07-17T15:55:56 *** atgreen has quit IRC (Ping timeout: 250 seconds) 2014-07-17T16:16:36 *** edwardk has joined #rtems 2014-07-17T16:21:18 *** Hesham has joined #rtems 2014-07-17T16:52:01 peerst: got it 2014-07-17T16:52:03 bmatt: got it 2014-07-17T16:53:06 beng-nl: it looks to me like RTEMS already has a TCP and ethernet stack, it's just a "simple" (ha-ha) matter of getting the hardware talking to the existing libraries, right? 2014-07-17T16:54:08 afaik, "yes" :) 2014-07-17T16:54:23 bmatt: i.e. "simply" (ha-ha) the ethernet driver 2014-07-17T16:56:25 i was planning to 'simply' (ha-ha) do that at some point as it have the hardware support from minix, with the appropriate lincense 2014-07-17T16:56:50 but i'm sure it'd be pretty hard actually, it's a first for me 2014-07-17T16:57:01 same here :) 2014-07-17T16:57:07 :) 2014-07-17T17:01:46 bmatt: "Can you fly her?" asked Ford pleasantly. 2014-07-17T17:01:46 "No, can you?" 2014-07-17T17:01:46 "No." 2014-07-17T17:01:46 "Trillian, can you?" 2014-07-17T17:01:46 "No." 2014-07-17T17:01:48 "Fine," said Zaphod, relaxing. "We'll do it together." 2014-07-17T17:01:51 "I can't either," said Arthur, who felt it 2014-07-17T17:01:53 was time he began to assert himself. 2014-07-17T17:01:58 hahah yes! 2014-07-17T17:02:06 bmatt: let's go ;-) 2014-07-17T17:06:37 bmatt: it's a delight to hear you are both of a similiar attitude and experience level :-) 2014-07-17T17:10:34 beng-nl: everything I've ever done, I've done for the first time at least once ;) 2014-07-17T17:51:16 *** Schoumi has quit IRC (*.net *.split) 2014-07-17T17:52:34 *** ysionneau has quit IRC (*.net *.split) 2014-07-17T17:52:35 *** peerst has quit IRC (*.net *.split) 2014-07-17T17:52:36 *** edwardk has quit IRC (*.net *.split) 2014-07-17T17:52:38 *** verm__ has quit IRC (*.net *.split) 2014-07-17T17:52:39 *** Hesham has quit IRC (*.net *.split) 2014-07-17T17:52:40 *** beng-nl has quit IRC (*.net *.split) 2014-07-17T17:52:46 *** ChanServ has quit IRC (*.net *.split) 2014-07-17T17:52:48 *** bmatt has quit IRC (*.net *.split) 2014-07-17T17:52:49 *** commander has quit IRC (*.net *.split) 2014-07-17T17:52:50 *** rokka has quit IRC (*.net *.split) 2014-07-17T17:53:30 *** Hesham has joined #rtems 2014-07-17T17:53:30 *** edwardk has joined #rtems 2014-07-17T17:53:30 *** bmatt has joined #rtems 2014-07-17T17:53:30 *** beng-nl has joined #rtems 2014-07-17T17:53:30 *** commander has joined #rtems 2014-07-17T17:53:30 *** verm__ has joined #rtems 2014-07-17T17:53:30 *** ChanServ has joined #rtems 2014-07-17T17:53:30 *** peerst has joined #rtems 2014-07-17T17:53:30 *** ysionneau has joined #rtems 2014-07-17T17:53:30 *** rokka has joined #rtems 2014-07-17T17:53:30 *** sinisalo.freenode.net sets mode: +o ChanServ 2014-07-17T17:53:36 *** Schoumi has joined #rtems 2014-07-17T17:54:38 *** bmatt has quit IRC (*.net *.split) 2014-07-17T17:54:39 *** commander has quit IRC (*.net *.split) 2014-07-17T17:54:40 *** rokka has quit IRC (*.net *.split) 2014-07-17T17:55:13 *** bmatt has joined #rtems 2014-07-17T17:55:13 *** commander has joined #rtems 2014-07-17T17:55:13 *** rokka has joined #rtems 2014-07-17T17:57:58 *** beng-nl_ has joined #rtems 2014-07-17T18:00:20 *** beng-nl has quit IRC (Write error: Broken pipe) 2014-07-17T18:31:46 *** beng-nl_ is now known as beng-nl 2014-07-17T18:32:04 lo 2014-07-17T18:47:58 lo 2014-07-17T21:29:23 *** Hesham has quit IRC (Ping timeout: 264 seconds) 2014-07-17T21:48:00 *** Hesham has joined #rtems 2014-07-17T23:12:49 alright, I give up on freebsd 2014-07-17T23:13:17 peerst: sorry, freebsd doesn't play nicely enough with vagrant :) 2014-07-18T01:20:45 *** gigetoo has joined #rtems 2014-07-18T01:36:52 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-18T02:04:12 *** Hesham has joined #rtems 2014-07-18T02:37:53 bmatt: it doesn’t? People tell me it works with FreeBSD 2014-07-18T02:38:26 my opinion of the usefulnes of vagrant is in free fall 2014-07-18T02:39:12 so vagrant is not architecture neutral, its just a tool to build linux in virtualbox environments 2014-07-18T02:39:35 probably using vmware won’t work either? 2014-07-18T02:40:41 https://vagrantcloud.com/chef/freebsd-9.2 2014-07-18T02:40:43 ? 2014-07-18T02:41:37 There is even https://github.com/wunki/vagrant-freebsd 2014-07-18T02:42:09 but I think that one is overdoing it 2014-07-18T02:42:21 for a dev box 2014-07-18T02:42:43 its more a play with ZFS and jails and all the neat stuff you can’t have on Linux box 2014-07-18T02:49:10 bmatt: did you have trouble with synced folders? 2014-07-18T03:05:39 well there are a few boxes with freebsd https://vagrantcloud.com/search?utf8=✓&sort=&provider=&q=freebsd 2014-07-18T03:06:45 https://github.com/opscode/bento 2014-07-18T03:41:11 *** Hesham has quit IRC (Read error: No route to host) 2014-07-18T04:08:25 *** Hesham has joined #rtems 2014-07-18T04:31:20 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-18T06:04:23 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-18T10:04:50 *** edwardk has joined #rtems 2014-07-18T10:38:54 *** utzig has joined #rtems 2014-07-18T11:13:47 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-18T11:32:45 *** phipse has joined #rtems 2014-07-18T12:03:36 peerst: I've had trouble at every step of the way :) 2014-07-18T12:54:11 *** edwardk has joined #rtems 2014-07-18T13:33:34 *** edwardk has quit IRC (Ping timeout: 264 seconds) 2014-07-18T13:57:52 *** edwardk has joined #rtems 2014-07-18T14:56:31 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-18T15:04:24 *** Hesham has joined #rtems 2014-07-18T15:24:39 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-18T15:31:13 *** edwardk has joined #rtems 2014-07-18T15:36:06 *** utzig_ has joined #rtems 2014-07-18T15:37:09 *** utzig has quit IRC (Ping timeout: 250 seconds) 2014-07-18T16:07:59 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-18T16:25:54 *** Hesham has joined #rtems 2014-07-18T16:33:55 *** utzig_ has left #rtems 2014-07-18T17:25:42 *** edwardk has joined #rtems 2014-07-18T17:30:17 *** edwardk has quit IRC (Ping timeout: 245 seconds) 2014-07-18T18:18:21 *** Hesham has quit IRC (Quit: Leaving.) 2014-07-18T18:26:20 *** edwardk has joined #rtems 2014-07-18T18:31:02 *** edwardk has quit IRC (Ping timeout: 250 seconds) 2014-07-18T19:28:57 *** Hesham has joined #rtems 2014-07-18T19:59:39 *** Hesham has left #rtems 2014-07-18T20:23:28 *** edwardk has joined #rtems 2014-07-19T03:52:27 bmatt: peerst: I've had trouble at every step of the way <- quite unspecific 2014-07-19T03:53:24 have a vagrant freebsd 9.2 up and running by now and didn’t have troubles besides, steps I needed to do and didn’t know yet but were explained in the documentation 2014-07-19T06:32:53 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-19T10:17:33 *** edwardk has joined #rtems 2014-07-19T10:47:43 *** edwardk has quit IRC (Read error: Connection reset by peer) 2014-07-19T12:07:11 peerst: oh, that's good to hear - do me a favor and shoot me a pull request for your vagrantfile changes? 2014-07-19T15:20:20 beng-nl: it just occurred to me that Das U-Boot initializes the phy on the BBB 2014-07-19T15:20:47 which, if we consider u-boot to be a requirement of RTEMS on the BBB, means the driver can probably be much slimmer 2014-07-19T15:49:36 ooh, it looks like RTEMS' mii driver was ported from FBSD, which may make it easier to start with FBSD's LAN8710A driver 2014-07-19T17:26:36 *** edwardk has joined #rtems 2014-07-19T17:57:02 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-19T20:11:18 *** edwardk has joined #rtems 2014-07-20T05:35:07 *** gigetoo has quit IRC (Remote host closed the connection) 2014-07-20T05:36:44 *** gigetoo has joined #rtems 2014-07-20T05:43:58 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T05:45:55 *** gigetoo has joined #rtems 2014-07-20T05:53:04 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T05:55:14 *** gigetoo has joined #rtems 2014-07-20T06:04:18 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T06:04:52 *** gigetoo has joined #rtems 2014-07-20T06:12:01 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T06:14:18 *** gigetoo has joined #rtems 2014-07-20T06:19:22 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T06:25:47 *** gigetoo has joined #rtems 2014-07-20T06:34:11 *** gigetoo has quit IRC (Ping timeout: 264 seconds) 2014-07-20T06:36:39 *** gigetoo has joined #rtems 2014-07-20T06:44:42 *** gigetoo has quit IRC (Ping timeout: 255 seconds) 2014-07-20T06:46:55 *** gigetoo has joined #rtems 2014-07-20T06:53:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T06:55:35 *** gigetoo has joined #rtems 2014-07-20T07:03:35 *** gigetoo has quit IRC (Ping timeout: 264 seconds) 2014-07-20T07:05:48 *** gigetoo has joined #rtems 2014-07-20T07:13:22 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T07:16:07 *** gigetoo has joined #rtems 2014-07-20T07:23:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T07:26:34 *** gigetoo has joined #rtems 2014-07-20T07:34:47 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T07:37:14 *** gigetoo has joined #rtems 2014-07-20T07:44:52 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T07:47:49 *** gigetoo has joined #rtems 2014-07-20T07:56:01 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T07:58:44 *** gigetoo has joined #rtems 2014-07-20T08:06:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T08:09:17 *** gigetoo has joined #rtems 2014-07-20T08:19:27 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T08:20:02 *** gigetoo has joined #rtems 2014-07-20T08:26:47 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T08:28:37 *** gigetoo has joined #rtems 2014-07-20T08:36:35 *** gigetoo has quit IRC (Ping timeout: 264 seconds) 2014-07-20T08:39:24 *** gigetoo has joined #rtems 2014-07-20T08:49:21 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T08:49:55 *** gigetoo has joined #rtems 2014-07-20T08:56:52 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T08:58:23 *** gigetoo has joined #rtems 2014-07-20T09:08:56 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T09:10:05 *** gigetoo has joined #rtems 2014-07-20T09:18:08 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T09:18:43 *** gigetoo has joined #rtems 2014-07-20T09:26:35 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T09:29:22 *** gigetoo has joined #rtems 2014-07-20T09:38:07 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T09:40:59 *** gigetoo has joined #rtems 2014-07-20T09:48:45 *** gigetoo has quit IRC (Ping timeout: 255 seconds) 2014-07-20T09:51:51 *** gigetoo has joined #rtems 2014-07-20T09:59:47 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T10:02:30 *** gigetoo has joined #rtems 2014-07-20T10:10:22 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T10:12:53 *** gigetoo has joined #rtems 2014-07-20T10:19:52 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T10:21:20 *** gigetoo has joined #rtems 2014-07-20T10:29:17 *** gigetoo has quit IRC (Ping timeout: 256 seconds) 2014-07-20T10:31:33 *** gigetoo has joined #rtems 2014-07-20T10:39:27 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T10:42:02 *** gigetoo has joined #rtems 2014-07-20T10:49:52 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T10:52:49 *** gigetoo has joined #rtems 2014-07-20T11:00:45 *** gigetoo has quit IRC (Ping timeout: 255 seconds) 2014-07-20T11:03:15 *** gigetoo has joined #rtems 2014-07-20T11:11:53 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T11:14:43 *** gigetoo has joined #rtems 2014-07-20T11:20:32 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T11:28:25 *** gigetoo has joined #rtems 2014-07-20T11:38:57 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T11:42:32 *** gigetoo has joined #rtems 2014-07-20T11:52:22 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T11:53:53 *** gigetoo has joined #rtems 2014-07-20T12:01:52 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T12:04:43 *** gigetoo has joined #rtems 2014-07-20T12:12:33 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T12:15:26 *** gigetoo has joined #rtems 2014-07-20T12:23:03 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T12:26:08 *** gigetoo has joined #rtems 2014-07-20T12:35:23 *** gigetoo has quit IRC (Ping timeout: 264 seconds) 2014-07-20T12:37:08 *** gigetoo has joined #rtems 2014-07-20T12:45:09 *** gigetoo has quit IRC (Ping timeout: 255 seconds) 2014-07-20T12:48:08 *** gigetoo has joined #rtems 2014-07-20T12:56:03 *** gigetoo has quit IRC (Ping timeout: 256 seconds) 2014-07-20T12:58:46 *** gigetoo has joined #rtems 2014-07-20T13:08:27 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T13:10:13 *** gigetoo has joined #rtems 2014-07-20T13:18:12 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T13:18:47 *** gigetoo has joined #rtems 2014-07-20T13:27:38 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T13:29:46 *** gigetoo has joined #rtems 2014-07-20T13:39:22 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T13:41:29 *** gigetoo has joined #rtems 2014-07-20T13:50:10 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T13:52:11 *** gigetoo has joined #rtems 2014-07-20T14:02:06 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T14:02:43 *** gigetoo has joined #rtems 2014-07-20T14:04:00 *** xMYTHICx has joined #rtems 2014-07-20T14:05:21 *** xMYTHICx has quit IRC (Client Quit) 2014-07-20T14:10:51 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T14:13:21 *** gigetoo has joined #rtems 2014-07-20T14:21:27 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T14:24:03 *** gigetoo has joined #rtems 2014-07-20T14:34:16 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T14:34:51 *** gigetoo has joined #rtems 2014-07-20T14:42:42 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T14:45:49 *** gigetoo has joined #rtems 2014-07-20T14:54:35 *** gigetoo has quit IRC (Ping timeout: 264 seconds) 2014-07-20T14:56:46 *** gigetoo has joined #rtems 2014-07-20T15:05:00 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T15:07:35 *** gigetoo has joined #rtems 2014-07-20T15:16:51 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T15:17:30 *** gigetoo has joined #rtems 2014-07-20T15:25:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T15:28:21 *** gigetoo has joined #rtems 2014-07-20T15:38:41 *** gigetoo has quit IRC (Ping timeout: 256 seconds) 2014-07-20T15:40:20 *** gigetoo has joined #rtems 2014-07-20T15:48:31 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T15:49:08 *** gigetoo has joined #rtems 2014-07-20T15:52:39 *** gigetoo has joined #rtems 2014-07-20T15:59:22 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T16:01:49 *** gigetoo has joined #rtems 2014-07-20T16:08:52 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T16:11:01 *** gigetoo has joined #rtems 2014-07-20T16:18:22 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T16:20:25 *** gigetoo has joined #rtems 2014-07-20T16:28:32 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T16:30:23 *** gigetoo has joined #rtems 2014-07-20T16:41:20 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T16:42:52 *** gigetoo has joined #rtems 2014-07-20T16:53:47 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T16:55:25 *** gigetoo has joined #rtems 2014-07-20T17:04:36 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T17:05:11 *** gigetoo has joined #rtems 2014-07-20T17:12:24 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T17:14:27 *** gigetoo has joined #rtems 2014-07-20T17:24:11 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T17:26:03 *** gigetoo has joined #rtems 2014-07-20T17:33:56 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T17:34:32 *** gigetoo has joined #rtems 2014-07-20T17:41:59 *** gigetoo has quit IRC (Ping timeout: 264 seconds) 2014-07-20T17:43:03 *** gigetoo has joined #rtems 2014-07-20T17:52:41 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T17:53:16 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2014-07-20T17:54:34 *** gigetoo has joined #rtems 2014-07-20T18:01:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T18:04:21 *** gigetoo has joined #rtems 2014-07-20T18:09:02 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T18:13:00 *** gigetoo has joined #rtems 2014-07-20T18:20:17 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T18:28:50 *** gigetoo has joined #rtems 2014-07-20T18:34:47 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T18:34:52 bmatt: hmmm 2014-07-20T18:34:59 oh hey 2014-07-20T18:35:06 hey yourself! :) 2014-07-20T18:35:12 so it looks like FBSD has a mature MII stack 2014-07-20T18:35:17 Minix doesn't use MII at all 2014-07-20T18:35:17 i don't think we should count on uboot tbh 2014-07-20T18:35:23 I agree :) 2014-07-20T18:35:26 ok 2014-07-20T18:35:44 I'm inclined to avoid trying to take on the task of implementing a general-purpose MII layer on RTEMS 2014-07-20T18:35:58 and instead just copy the structure of the CS8900 driver and stick Minix code in there, more or less 2014-07-20T18:36:20 oh i didn't realize the minix driver depended on the mii being initialized 2014-07-20T18:36:29 u-boot does that and then it's usable? 2014-07-20T18:37:22 I don't think minix cares about MII at all 2014-07-20T18:38:09 or, at least, it doesn't have any sort of MII abstraction 2014-07-20T18:38:19 no indeed 2014-07-20T18:38:31 the registers are all named MDIO*, but it's just reg read/write :) 2014-07-20T18:39:06 minix also doesn't appear to do DMA, but I'm not sure about that 2014-07-20T18:39:14 I think it just gets an interrupt and copies the buffer itself 2014-07-20T18:40:10 I am amused by Minix' driver model, though 2014-07-20T18:40:18 everything is a process! :) 2014-07-20T18:43:23 *** gigetoo has joined #rtems 2014-07-20T18:45:30 just like life ;-) 2014-07-20T18:46:13 bmatt: the driver does dma into its own address space, then the data is copied around to where it's needed 2014-07-20T18:46:39 bmatt: ehmm you don't have a bbxm right? 2014-07-20T18:46:50 correct 2014-07-20T18:46:54 hm-hm 2014-07-20T18:47:56 i'm working down my beagle bsp checklist, reproducing all the work from scratch and testing it works again.. and my bbxm sd card is giving me some trouble.. but on the emulator it's ok and the device isn't booting ANY sd card 2014-07-20T18:48:10 so i maaaayyy have fucked it up 2014-07-20T18:48:17 doh! 2014-07-20T18:49:04 i'll try jtag, i.e. without the sd card, and call it working if that works 2014-07-20T18:49:10 fair enough don't you think? 2014-07-20T18:49:20 definitely 2014-07-20T18:49:25 okay 2014-07-20T18:50:54 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T18:51:43 ahhh, it sets up DMA on line 524 and passes it the address of the head of the RX buffer 2014-07-20T18:51:45 right. 2014-07-20T18:52:17 *** gigetoo has joined #rtems 2014-07-20T19:01:11 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T19:02:44 *** gigetoo has joined #rtems 2014-07-20T19:10:48 *** gigetoo has quit IRC (Ping timeout: 255 seconds) 2014-07-20T19:12:29 *** gigetoo has joined #rtems 2014-07-20T19:23:51 *** gigetoo has quit IRC (Ping timeout: 255 seconds) 2014-07-20T19:27:02 *** gigetoo has joined #rtems 2014-07-20T19:34:52 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T19:36:24 *** gigetoo has joined #rtems 2014-07-20T19:44:05 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T19:45:35 *** gigetoo has joined #rtems 2014-07-20T19:53:24 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T19:54:53 *** gigetoo has joined #rtems 2014-07-20T20:02:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T20:04:46 *** gigetoo has joined #rtems 2014-07-20T20:12:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T20:14:06 *** gigetoo has joined #rtems 2014-07-20T20:21:52 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T20:23:56 *** gigetoo has joined #rtems 2014-07-20T20:32:08 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T20:33:48 *** gigetoo has joined #rtems 2014-07-20T20:41:28 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T20:43:07 *** gigetoo has joined #rtems 2014-07-20T20:50:56 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T20:52:35 *** gigetoo has joined #rtems 2014-07-20T21:00:36 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T21:02:00 *** gigetoo has joined #rtems 2014-07-20T21:09:22 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T21:11:34 *** gigetoo has joined #rtems 2014-07-20T21:19:27 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T21:21:13 *** gigetoo has joined #rtems 2014-07-20T21:28:33 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T21:30:22 *** gigetoo has joined #rtems 2014-07-20T21:40:11 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T21:42:57 *** gigetoo has joined #rtems 2014-07-20T21:51:52 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T21:53:22 *** gigetoo has joined #rtems 2014-07-20T21:58:37 *** gigetoo has quit IRC (Ping timeout: 245 seconds) 2014-07-20T22:05:03 *** gigetoo has joined #rtems 2014-07-20T22:12:43 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T22:14:51 *** gigetoo has joined #rtems 2014-07-20T22:17:39 *** edwardk has joined #rtems 2014-07-20T22:21:01 *** gigetoo has quit IRC (Ping timeout: 256 seconds) 2014-07-20T22:29:14 *** gigetoo has joined #rtems 2014-07-20T22:36:42 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T22:39:08 *** gigetoo has joined #rtems 2014-07-20T22:48:29 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T22:49:04 *** gigetoo has joined #rtems 2014-07-20T22:56:43 *** gigetoo has quit IRC (Ping timeout: 256 seconds) 2014-07-20T22:58:25 *** gigetoo has joined #rtems 2014-07-20T23:08:56 *** gigetoo has quit IRC (Ping timeout: 260 seconds) 2014-07-20T23:11:13 *** gigetoo has joined #rtems 2014-07-20T23:18:44 *** gigetoo has quit IRC (Ping timeout: 255 seconds) 2014-07-20T23:21:05 *** gigetoo has joined #rtems 2014-07-20T23:28:52 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T23:30:44 *** gigetoo has joined #rtems 2014-07-20T23:38:22 *** gigetoo has quit IRC (Ping timeout: 240 seconds) 2014-07-20T23:40:23 *** gigetoo has joined #rtems 2014-07-20T23:49:34 *** gigetoo has quit IRC (Read error: Connection reset by peer) 2014-07-20T23:50:19 *** gigetoo has joined #rtems 2014-07-20T23:53:06 *** kiwichris has joined #rtems 2014-07-20T23:58:08 *** gigetoo has quit IRC (Ping timeout: 250 seconds) 2014-07-20T23:59:31 *** gigetoo has joined #rtems