2015-06-29T00:06:18 kiwichris : Okay. One question, To initialize the network for monkey, I am using the default-network-init.h from the rtems-libbsd/testsuite/include/rtems/bsd/test directories with some modifications of my own. It has a copyright notice at its top, and I don't know if directly using the file in monkey would lead to any conflicts of license etc. what do you think? 2015-06-29T00:20:06 What is the copyright ? 2015-06-29T00:22:05 kiwichris : Copyright (c) 2013-2014 embedded brains GmbH. All rights reserved. 2015-06-29T00:23:43 sujayraaj, from my understanding this is about the examples and user set up. 2015-06-29T00:24:06 sujayraaj, are you taking about the Monkey examples ? 2015-06-29T00:25:01 kiwichris : no. The one from libbsd. 2015-06-29T00:25:14 libbsd should be BSD licensed 2015-06-29T00:26:51 sujayraaj, where are putting this file from libbsd ? 2015-06-29T00:29:38 kiwichris : to set up the network on monkey, we shall need need to set up the network just like in the testsuites of libbsd, ping01 for an example. So we will have to provide a file to monkey that does initialization before the main for monkey is called. 2015-06-29T00:35:13 Let me take a loo and get back to you. 2015-06-29T00:41:18 kiwichris : okay, sure. Some detail of what I have done : From what I have understood, we need a Init(rtems_task_argument arg) for various initializations. So we should supply a file like rtems_init.c to monkey, which will contain the definition of Init, then init does various initializations, like setting up the nic and giving it an ip address etc (calling ifconfig) 2015-06-29T00:41:56 After everything has been set up, this Init function will call monkey's main function , to start running the server. 2015-06-29T00:43:28 sujayraaj, this seems correct, what we need to consider is how we add this support and make the changes to Monkey as small as possible. 2015-06-29T00:44:05 sujayraaj, the way we manage network set up is evolving in RTEMS and it is far from good. I think we need a better uniform way to manage this. 2015-06-29T00:47:12 kiwichris : one possible way out of this would be to create a header file in libbsd's include (instead of the testsuites) , this new header file will do all the required initializations, and will have an Init that calls a function pointer. So, in the programs that want to use the network, all it would need to do will be to point that function pointer to their main function, and run. 2015-06-29T00:49:56 This way, we will be able to make an abstract way to run any program that requires networking, not just monkey 2015-06-29T00:50:21 sujayraaj, this is an idea worth looking at. Often users have local specific ways. 2015-06-29T00:51:03 sujayraaj, we still need a way to import settings. Having them set in libbsd is only ok for those tests. In Monkey we need a way as well. 2015-06-29T00:51:23 sujayraaj, the same goes for examples-v2 if we add a Monkey example there. 2015-06-29T00:54:59 kiwichris : Yes, hard-coding settings in libbsd won't do. One way I am thinking is, to keep a plain init in that header, in libbsd, and various functions that do initialization. From the init, we call monkey, and monkey uses the functions in the header to set up the network with settings it wants. This would only require adding around 5-10 lines in monkey, while most of our functions will be in libbsd. Though some more thinking will 2015-06-29T00:56:43 sujayraaj, the support would need move out from the test suite tree in libbsd and become a full RTEMS API. 2015-06-29T00:56:52 still in libbsdb 2015-06-29T00:56:56 libbsd 2015-06-29T00:57:11 Yeah. exactly, that is what I am saying too. Because 2015-06-29T00:57:36 Today when I was configuring netdemo from the network-demo package, I had to add the same configurations from the testsuites 2015-06-29T00:58:02 So instead of using it as a makeshift way, a simple api would make things easier and portable. 2015-06-29T01:00:52 kiwichris : Moreover, one of my goals in gsoc proposal (in the second half) is to make an abstract way such that any program will be able to compile with either libbsd's network stack, or the default network stack , or the lwip port we have for rtems. That is why I once asked about lwip. Moreover since it was in the second phase in the proposal, and I was busy with monkey, I didn't think about it much untill today. 2015-06-29T01:02:26 sujayraaj, it would be nice to have a standard API for the simple cases. Some users will need more complex solutions. 2015-06-29T01:06:02 kiwichris : So the problem at our hand now is, how much code am I supposed to push into monkey? Will adding two or three header file and source file be okay? If that is the case then we will not have to worry about creating the above solution, for now. 2015-06-29T01:06:51 sujayraaj, if you can limit the code to a file or even a couple that would be good and then small changes in Monkey. 2015-06-29T01:08:03 kiwichris : presently I am adding a header file and a C file ( total two, that is ) . I can compress it and make it into one, it won't be a big problem. 2015-06-29T01:08:29 sujayraaj, that is fine. 2015-06-29T01:08:43 Just prefix with rtems-.... to make it clear. 2015-06-29T01:13:13 Okay. Then I will place all the extra functions that I created ( like timegm and sendfile ) and keep it in the header file as rtems-timegm and rtems-sendfile ( so we will only be dealing with changes in one source file, something like rtems-init.c ). 2015-06-29T01:13:38 Ok. 2015-06-29T03:37:37 *** monstr has joined #rtems 2015-06-29T04:48:26 *** monstr has quit IRC 2015-06-29T04:48:47 *** monstr has joined #rtems 2015-06-29T05:21:01 *** mkhoory-2 has quit IRC 2015-06-29T05:27:15 *** mkhoory has joined #rtems 2015-06-29T06:22:38 *** sebhub has joined #rtems 2015-06-29T06:59:17 good morning 2015-06-29T07:12:49 *** kiwichris has quit IRC 2015-06-29T07:41:49 *** mumptai has joined #rtems 2015-06-29T08:05:44 *** ketul has joined #rtems 2015-06-29T08:20:53 *** ketul has quit IRC 2015-06-29T08:44:11 *** mumptai has quit IRC 2015-06-29T08:44:26 *** sujayraaj has quit IRC 2015-06-29T08:50:38 *** ketul has joined #rtems 2015-06-29T10:47:21 hello beng-nl 2015-06-29T10:48:14 getting error while compiling latest rtems. 2015-06-29T10:48:25 http://pastebin.com/YiFW7Qzg 2015-06-29T10:48:40 sebhub: probably you can help :) 2015-06-29T11:24:03 yes, you need the latest tool chain 2015-06-29T11:28:47 sebhub: can you point me to the link for that tool chain (I mean steps ) :) 2015-06-29T11:29:04 how did you get your current tool chain? 2015-06-29T11:30:33 currently using previous tool chain. "export PATH=$HOME/development/rtems/4.11/bin:$PATH" 2015-06-29T11:31:28 ok, and how did you install it? 2015-06-29T11:31:58 umm. followed http://www.shrike-systems.com/beagleboard-xm-beaglebone-black-and-everything-else-rtems-on-the-beagles.html 2015-06-29T11:32:21 so now if I want to work with updated one. how can I ? 2015-06-29T11:33:00 do a git pull in the rsb and build the tools again 2015-06-29T11:33:30 no I didn't as now. so I should do it right ? 2015-06-29T11:33:50 ok will do 2015-06-29T11:46:24 *** kiwichris has joined #rtems 2015-06-29T12:30:04 *** ketul has quit IRC 2015-06-29T12:31:59 *** ketul has joined #rtems 2015-06-29T12:47:43 *** igutek has quit IRC 2015-06-29T12:56:13 *** gedare has joined #rtems 2015-06-29T12:56:13 *** ChanServ sets mode: +o gedare 2015-06-29T13:47:22 *** igutek has joined #rtems 2015-06-29T13:54:39 *** ketul has quit IRC 2015-06-29T14:08:42 *** ketul has joined #rtems 2015-06-29T14:34:55 *** gedare has quit IRC 2015-06-29T14:36:45 ketul: did it work? 2015-06-29T14:37:17 beng-nl: sorry are you talking about tool chain ? 2015-06-29T14:37:36 yeah 2015-06-29T14:37:55 ohh no. slow net speed here :( 2015-06-29T14:38:45 beng-nl: fixed some indentation issues for gpio as gedare suggested to gpio api . https://github.com/ketul93/rtems/tree/GPIO_API 2015-06-29T14:39:35 aha 2015-06-29T14:39:51 ketul: but the toolchain update is going forward? or do i need to update it on my side? 2015-06-29T14:40:37 beng-nl: don't know whether will it work or not. I am still at "building: arm-rtems4.11-gcc-4.9.2-newlib-2.2.0.20150423-x86_64-linux-gnu-1" 2015-06-29T14:40:53 so will let you know when it got completed :) 2015-06-29T14:42:32 and beng-nl have you checked gpio code. mailed you before some time 2015-06-29T15:02:39 the 1st time yes :) 2015-06-29T15:02:45 i'll take a look again 2015-06-29T15:04:08 yeah ok. 2015-06-29T15:04:43 tell me if more changes required :) 2015-06-29T16:25:37 beng-nl: i got same thing with updated tool-chain . http://pastebin.com/LrmCH1eb 2015-06-29T16:30:07 *** ragunath has joined #rtems 2015-06-29T16:45:25 damnit 2015-06-29T16:45:34 maybe i need to update the tools again 2015-06-29T16:45:39 *** Davidbrcz has joined #rtems 2015-06-29T16:45:53 ketul: i'll try to reproduce it, just a sec 2015-06-29T16:46:33 ah, no, that one is the same error as earlier 2015-06-29T16:49:27 you should actually be ok if you start building from scratch 2015-06-29T16:49:34 ketul: i'll try just to make sure 2015-06-29T16:49:58 doing ../source-builder/sb-set-builder --log=beagle.txt --prefix=$HOME/development/rtems/4.11 devel/beagle.bset 2015-06-29T16:50:01 now 2015-06-29T16:50:12 that takes 30 mins or so on my machine.. maybe a bit more now because there's more goin gon 2015-06-29T16:50:15 going on 2015-06-29T17:06:19 *** krohini has joined #rtems 2015-06-29T17:07:55 hi ben 2015-06-29T17:13:17 hi ragunath :) 2015-06-29T17:13:21 beng-nl: yes 2015-06-29T17:15:15 *** ragunath_ has joined #rtems 2015-06-29T17:15:23 beng-nl: sebastian has asked can we remove interrupt disable/enable in bsp_interrupt_dispatch function? 2015-06-29T17:16:22 *** ragunath has quit IRC 2015-06-29T17:17:14 i.e. we will not have recursive interrupts then? yes i suppose that's cleaner 2015-06-29T17:17:24 a cleaner workaround :) 2015-06-29T17:18:41 is it okay remove them? i can give it a try 2015-06-29T17:18:47 yep if it works 2015-06-29T17:20:26 ketul: gah. had to install python-dev for gdb. starting over. 2015-06-29T17:20:37 (this is a new install) 2015-06-29T17:20:52 beng-nl: ohh for the latest rtems 2015-06-29T17:21:07 beng-nl: can you explain why we need it in the first place? 2015-06-29T17:21:59 it's originally there because i wanted to be able to support recursive interrupts (interrupts happening before the current one's dispatch function is over) 2015-06-29T17:22:14 to minimize interrupt deafness time 2015-06-29T17:23:11 oh ok 2015-06-29T17:23:12 but re-enabling interrupts fired the current interrupt immediately (infinite recursion) iirc 2015-06-29T17:23:26 so i just masked this one, turned the thing on, then unmasked it once the function was done 2015-06-29T17:23:59 admittedly there may be a 'real' solution that sebhub alluded to.. 2015-06-29T17:24:20 but i haven't found it.. so i hacked it and now i got caught :-) 2015-06-29T17:25:35 the assumption there is that the irq is cleared in the handler function 2015-06-29T17:25:45 apprently that assumption isn't solid 2015-06-29T17:26:29 *** ragunath has joined #rtems 2015-06-29T17:26:34 *** krohini has joined #rtems 2015-06-29T17:26:36 perhaps the dispatch function should leave the interrupt masked and the handler function should unmask it explicitly 2015-06-29T17:26:40 that would work in both cases 2015-06-29T17:26:41 *** krohini has quit IRC 2015-06-29T17:26:43 beng-nl: sorry my web client is crashing 2015-06-29T17:27:19 the other bsp's dont support this 2015-06-29T17:27:34 *** ragunath_ has quit IRC 2015-06-29T17:28:11 no recursive interrupts?1 2015-06-29T17:28:12 no recursive interrupts?! 2015-06-29T17:28:21 the logic should work as it is if we directly call the interrupt handler 2015-06-29T17:29:00 since we have some other setup in libbsd we are facing the issue 2015-06-29T17:29:11 yes i understand the issue 2015-06-29T17:29:58 so let me remove the disable/enable and see if it works 2015-06-29T17:31:55 beng-nl: Angelo Fraietta asked on BBB app. I think he is building app on gpio but getting some prob. May be related to Makefile :) 2015-06-29T17:32:31 yep i emailed him 2015-06-29T17:32:46 ketul: i emailed him saying i'd help and asking if i can improve my tutorial based on his experience 2015-06-29T17:32:49 ragunath: great 2015-06-29T17:33:18 ohh but the app support would be great to community too :) 2015-06-29T17:35:18 beng-nl: and it would be good if you keep us in cc (I mean when query is related to our work) so we can also learn some hands on if possible :) 2015-06-29T17:37:44 good point 2015-06-29T17:37:56 :) hmm 2015-06-29T17:38:17 beng-nl: any progress for tool chain ? 2015-06-29T17:40:32 still building, i had to restart 2015-06-29T17:40:34 takes about 30 mins 2015-06-29T17:41:21 beng-nl: ok good. take your time. and can i know about gpio api ? 2015-06-29T17:43:17 yes one sec 2015-06-29T17:43:24 :) 2015-06-29T17:43:53 it actually looks good to me! well done 2015-06-29T17:44:10 ohh thanks. :D 2015-06-29T17:44:15 except for the RPi differences and drjoel's desire for dynamic registration 2015-06-29T17:44:35 ketul: can you post to rtems-devel for general review? if there's no objection i'm happy to merge it like this 2015-06-29T17:45:17 beng-nl: ok sure. let me. So should I divide it in 2 parts right ? 2015-06-29T17:46:12 which 2 parts? 2015-06-29T17:46:36 ketul: oh i just noticed; the gpio sample won't compile for other bsp's, so that probably shouldn't be in the list of samples unconditionally 2015-06-29T17:46:44 I mean dividing patch in 2. or should I just send link to my github 2015-06-29T17:47:25 git send-email is the preferred way to submit i ear 2015-06-29T17:47:26 hear 2015-06-29T17:47:35 ohh but gpio sample is bsp specific always 2015-06-29T17:48:11 but it will be built for every bsp with your change, no? 2015-06-29T17:49:16 "--enable-rtemsbsp="beagleboneblack beagleboardxm" Isn't for Beagleboards. 2015-06-29T17:50:49 so what happens if i'm building rtems with a different bsp.. 2015-06-29T17:50:57 and want to build the samples too? 2015-06-29T17:51:01 won't it explode? 2015-06-29T17:51:08 they are built by default after all 2015-06-29T17:51:13 *** igutek has quit IRC 2015-06-29T17:52:10 yup right so should I write condition like #ifdef ? 2015-06-29T17:52:22 for BSP specific sample 2015-06-29T17:53:32 and beng-nl what about the RTC. any sample added in that ? 2015-06-29T17:53:53 there wasn't i think 2015-06-29T17:54:04 it's nice to have though 2015-06-29T17:54:37 yes always because some demos make devel easy to make applications :) 2015-06-29T17:54:51 quite so 2015-06-29T17:54:59 ok, i've thought about it a bit :) 2015-06-29T17:55:14 :) 2015-06-29T17:55:17 i am inclined to include the whole sample, and #ifdef it to BBB-specific(!) 2015-06-29T17:55:41 and then to change the conditional to 'GPIO available' once the API is the same as the RPi and both can use the same sample 2015-06-29T17:56:05 this will need explanation in a comment in the code and commit message though 2015-06-29T17:56:23 yeah good idea 2015-06-29T17:57:00 *** igutek has joined #rtems 2015-06-29T17:57:05 so as of now I should add #ifdef it to BBB. right in sample ? 2015-06-29T17:57:13 so it does do something in all cases, but only for BBB does it actually do the GPIO. for oter bsp's there is a Init() etc but the code is empty or just prints something simple. 2015-06-29T17:57:22 ketul: yep. is there such a #ifdef? 2015-06-29T17:57:29 that selects BBB? 2015-06-29T17:57:53 don't know much about it. need to see 2015-06-29T17:58:11 I think you have idea if you have used it before 2015-06-29T17:58:54 i haven't; in the bsp there is only the SOC, not board 2015-06-29T17:59:27 you can add a macro to configure.ac though 2015-06-29T17:59:48 ketul: you can also submit the gpio patch without a sample as a first step if you prefer, so reviews can happen in parallel 2015-06-29T17:59:55 smaller patches are easier to review too 2015-06-29T18:00:09 yep good point. 2015-06-29T18:00:17 ok 2015-06-29T18:00:44 that I will do today and hope there are good reviews :) 2015-06-29T18:00:55 ketul: great :-) 2015-06-29T18:00:57 good going 2015-06-29T18:01:23 ty :) 2015-06-29T18:03:05 beng-nl: If you get some time have a look at adc too and make your comment(s) :) 2015-06-29T18:05:35 *** mumptai has joined #rtems 2015-06-29T18:05:51 beng-nl: now I am more used to work with TRM and developing drivers. this 6 7 weeks experience helps me a lot. happy to work with it . :) 2015-06-29T18:12:26 great :-) 2015-06-29T18:12:30 that's what GSOC is for :-) 2015-06-29T18:15:47 ketul: tools have built. trying to build rtems now. 2015-06-29T18:15:56 Build Set: Time 0:54:10.349422 urg 2015-06-29T18:17:18 *** monstr has quit IRC 2015-06-29T18:20:34 +1 :) 2015-06-29T18:31:31 ketul: i rebuilt everything from scratch and it works. 2015-06-29T18:31:40 so the tools in my repo are uptodate 2015-06-29T18:33:04 *** krohini has joined #rtems 2015-06-29T18:34:12 so you'll have to rebuild more :) 2015-06-29T18:55:13 *** gedare has joined #rtems 2015-06-29T18:55:13 *** ChanServ sets mode: +o gedare 2015-06-29T19:04:09 *** dr__house|away is now known as dr__house 2015-06-29T19:04:22 beng-nl: okay I will do it 2015-06-29T19:53:42 *** Davidbrcz has quit IRC 2015-06-29T20:20:19 *** Davidbrcz has joined #rtems 2015-06-29T20:24:52 *** Davidbrcz has quit IRC 2015-06-29T20:25:06 *** Davidbrcz has joined #rtems 2015-06-29T20:40:40 *** ketul has quit IRC 2015-06-29T20:49:17 *** gedare has quit IRC 2015-06-29T20:57:15 *** krohini has quit IRC 2015-06-29T21:27:48 *** ragunath has quit IRC 2015-06-29T21:48:31 *** Davidbrcz has quit IRC 2015-06-29T21:48:40 *** Davidbrcz has joined #rtems 2015-06-29T22:29:32 [git] eb2caa2 by Ed Sutter: [cleanup monargs warning] http://devel.rtems.org/changeset/eb2caa2/umon 2015-06-29T22:29:32 [git] d6c7226 by Ed Sutter: [cleanup uninitialized variable warning] http://devel.rtems.org/changeset/d6c7226/umon 2015-06-29T22:29:32 [git] 2236db2 by Ed Sutter: [remove APPRAMBASE_OVERRIDE and include DM/PM commands] http://devel.rtems.org/changeset/2236db2/umon 2015-06-29T22:29:32 [git] 7921e2e by Ed Sutter: [remove ctype_ptr variables in prep for non-libc build] http://devel.rtems.org/changeset/7921e2e/umon 2015-06-29T22:29:32 [git] 08a34d8 by Ed Sutter: [add support for RTEMS toolset, make this the default] http://devel.rtems.org/changeset/08a34d8/umon 2015-06-29T22:29:33 [git] fc86a28 by Ed Sutter: [eliminate need for libc by eliminating need to pull in ctype array] http://devel.rtems.org/changeset/fc86a28/umon 2015-06-29T23:13:20 *** kiwichris has quit IRC 2015-06-29T23:16:26 *** mumptai has quit IRC 2015-06-29T23:31:07 *** kiwichris has joined #rtems 2015-06-30T00:02:23 *** Davidbrcz has quit IRC 2015-06-30T02:23:26 *** sujayraaj has joined #rtems 2015-06-30T02:24:10 kiwichis : hi 2015-06-30T02:27:09 kiwichris : hello, sorry, wrong spelling earlier. 2015-06-30T02:27:23 sujayraaj, hi 2015-06-30T02:32:11 kiwichris : I went through the dl01 test, I understood from it that it is supposed to load a tar file into memory and point to it with dl_tar pointer and is loaded by Untar_FromMemory . But I have been having trouble figuring out where exactly is the tar file, and how did it get loaded into memory in the first place. Plus, I don't think I have the "dl-tar.h" in my source. 2015-06-30T02:33:54 It is converted to an object (ELF) file and linked. 2015-06-30T02:34:16 The tests use bin2c and then compile the C file. 2015-06-30T02:36:53 Ah!! Now I finally get it! 2015-06-30T02:38:35 Thanks! 2015-06-30T04:04:49 *** monstr has joined #rtems 2015-06-30T04:13:19 *** sujayraaj has quit IRC 2015-06-30T04:15:58 *** sujayraaj has joined #rtems 2015-06-30T04:20:19 *** sujayraaj has quit IRC 2015-06-30T05:03:18 [trac] GSoC/2015 edited by gadia (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-06-30T05:23:31 *** ketul has joined #rtems 2015-06-30T06:03:31 *** mumptai has joined #rtems 2015-06-30T06:35:22 *** Davidbrcz has joined #rtems 2015-06-30T06:40:58 *** kiwichris has quit IRC 2015-06-30T07:10:39 *** Davidbrcz has quit IRC 2015-06-30T07:16:21 *** zoso has joined #rtems 2015-06-30T08:20:20 *** mumptai has quit IRC 2015-06-30T09:56:23 *** sebhub has quit IRC 2015-06-30T10:21:42 *** krohini has joined #rtems 2015-06-30T11:08:45 hi beng-nl :). sent all patches 2015-06-30T11:13:56 ketul: nice 2015-06-30T11:14:00 hi all 2015-06-30T11:18:23 *** krohini has joined #rtems 2015-06-30T11:21:45 *** sebhub has joined #rtems 2015-06-30T11:23:34 *** kulp has quit IRC 2015-06-30T11:28:35 *** kulp has joined #rtems 2015-06-30T11:33:41 *** asuolen has joined #rtems 2015-06-30T11:41:52 *** sujayraaj has joined #rtems 2015-06-30T12:23:36 *** guerby has quit IRC 2015-06-30T12:25:57 *** guerby has joined #rtems 2015-06-30T12:58:15 beng-nl: :) 2015-06-30T13:02:14 beng-nl: finally tool chain got built. Took too much time. 2015-06-30T13:33:23 *** salpha has joined #rtems 2015-06-30T13:34:50 *** salpha has quit IRC 2015-06-30T13:35:04 *** salpha has joined #rtems 2015-06-30T13:54:32 Hey everyone 2015-06-30T13:54:57 I'm having trouble building the STM32 BSP due to conflicting definitions of microtime 2015-06-30T13:55:46 This appeared recently after rebuilding the toolchiain using RSB 2015-06-30T13:55:54 Anyone have any ideas? 2015-06-30T13:56:26 More specifically, does anyone have experience with microtime issues in the nfsclient 2015-06-30T14:00:53 igutek: can you pastebin it ? 2015-06-30T14:01:47 http://pastebin.com/rEyDXcd7 2015-06-30T14:03:00 sys/time.h has a non-static declaration, while rtems/cpukit/libnetworking/rtems/rtems_bsdnet_internal.h 2015-06-30T14:03:00 has a static declaration. Removing the static results in a multiple declaration error during linking of nsfclient (I'll quickly re-build to get a paste) 2015-06-30T14:03:01 igutek: yesterday I had a issue like : http://pastebin.com/LrmCH1eb but it is for bb . 2015-06-30T14:04:57 Interesting. It looks slightly different 2015-06-30T14:05:43 igutek: i don't have much idea about it. But yesterday beng-nl updated his RSB repo and it got solved (something python needed for gdb). Probably another devel will help. :) 2015-06-30T14:06:01 Thanks! 2015-06-30T14:06:35 Here is the error after removing the static qualifier 2015-06-30T14:06:43 http://pastebin.com/T3Mer2fa 2015-06-30T14:16:58 *** yangqiao has joined #rtems 2015-06-30T14:26:09 [trac] GSoC/2015 edited by ketul (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-06-30T14:39:47 [trac] GSoC/2015 edited by krohini (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-06-30T14:43:29 *** ragunath has joined #rtems 2015-06-30T14:56:38 *** krohini has joined #rtems 2015-06-30T14:58:23 *** hermann_19829 has joined #rtems 2015-06-30T14:59:04 *** jrcatbagan has joined #rtems 2015-06-30T15:04:05 *** javamonn has joined #rtems 2015-06-30T15:05:50 *** beng-nl-droid has joined #rtems 2015-06-30T15:05:52 *** toxygen has quit IRC 2015-06-30T15:06:02 Hi all 2015-06-30T15:06:28 hello beng-nl-droid 2015-06-30T15:06:31 hi beng-nl 2015-06-30T15:06:51 hello beng-nl-droid :) 2015-06-30T15:11:37 *** sgworks has joined #rtems 2015-06-30T15:12:19 *** hermann_19829_ has joined #rtems 2015-06-30T15:14:36 *** hermann_19829 has quit IRC 2015-06-30T15:17:31 *** anandkp92 has joined #rtems 2015-06-30T15:17:39 when is the meeting? 2015-06-30T15:18:00 ragunath: 30 mins from the original time 2015-06-30T15:18:22 *** hermann_19829 has joined #rtems 2015-06-30T15:18:24 jrcatbagan:ok thanks 2015-06-30T15:18:33 ragunath: so in about 12 mins 2015-06-30T15:18:41 no problem 2015-06-30T15:22:11 *** gedare has joined #rtems 2015-06-30T15:22:11 *** ChanServ sets mode: +o gedare 2015-06-30T15:22:20 *** hermann_19829_ has quit IRC 2015-06-30T15:22:25 good morning 2015-06-30T15:22:34 good morning gedare 2015-06-30T15:22:42 hi gedare 2015-06-30T15:22:42 hello 2015-06-30T15:22:43 hi gedare GM 2015-06-30T15:23:00 morning 2015-06-30T15:23:34 hermann_19829: if your internet is spotty you can update us first 2015-06-30T15:24:36 I also have two announcements 2015-06-30T15:24:52 1) Fill out your evaluations, or you won't pass your midterm. 2015-06-30T15:25:04 2) We won't have a meeting next week. 2015-06-30T15:25:20 *** hermann_19829_ has joined #rtems 2015-06-30T15:25:33 hermann_19829_ you can start if you're here now 2015-06-30T15:25:34 I worked with DrJoel on debugging the byte size issue for the coverage report of Leon2 2015-06-30T15:25:36 *** salpha has joined #rtems 2015-06-30T15:25:54 yes i saw the emails, looks like you settled on a work-around for now 2015-06-30T15:26:11 and with you gedare :) 2015-06-30T15:26:28 if drjoel does not pick up your patches from the wiki page, you might want to submit them by email directly to him or to the devel mailing list, at least when you think they're "ready" 2015-06-30T15:26:33 yes DrJoel suggested the workaround 2015-06-30T15:26:35 hi all 2015-06-30T15:26:43 hello salpha we are just getting started. 2015-06-30T15:27:06 yes I can do that 2015-06-30T15:27:26 ok so you are working on implementing the workaround? 2015-06-30T15:27:31 plus doing your summer school thing 2015-06-30T15:27:47 as mentioned in the mail, I will also provide a blog post containing the instructions to obtain the coverage results 2015-06-30T15:28:19 that would be super. i think it is most important to document your process for this project 2015-06-30T15:28:21 yes, but I didn't start yet 2015-06-30T15:28:31 *** hermann_19829 has quit IRC 2015-06-30T15:28:45 ok thanks hermann_19829_ anything else? jrcatbagan you'll be next, and then yangqiao after that. 2015-06-30T15:28:56 no not at the moment 2015-06-30T15:29:04 hi 2015-06-30T15:29:07 thanks gedare! 2015-06-30T15:29:10 ok thank hermann_19829_ 2015-06-30T15:29:19 go ahead jrcatbagan 2015-06-30T15:29:31 *** hermann_19829_ has quit IRC 2015-06-30T15:29:39 ok so, patches I recently submitted are finally merged in 2015-06-30T15:29:48 uMon is now building with the RTEMS tools 2015-06-30T15:30:05 That's great work 2015-06-30T15:30:09 Ed, has already incorporated the necessary files that remove uMon's external dependence on the C library 2015-06-30T15:30:21 right now, him and I are working in tandem to get uMon to boot from uSD 2015-06-30T15:30:56 ok that is a nice goal to work toward. do you also have some ddr init work to do in the background / subsequently? 2015-06-30T15:31:04 yes 2015-06-30T15:31:11 I'll be pushing what I have this friday 2015-06-30T15:31:24 ok great. well so everyone can now, i'll be on vacation so won't be responsive 2015-06-30T15:31:31 but ed is doing a great job keeping up 2015-06-30T15:31:36 so i think you'll be fine 2015-06-30T15:31:40 ok anything else jrcatbagan? 2015-06-30T15:32:02 as of right now that is all I have 2015-06-30T15:32:24 ok thanks. yangqiao you're up next. yangqiao please remember to do your student evaluation. 2015-06-30T15:32:29 I've moved some of the graphic libraries into rsb so that we can use rsb to build graphic libraries instead of rtems graphic tool kits. I've got some problem when trying to apply my custom patches using sb-build and I've sent an mail to mailing list. 2015-06-30T15:32:29 For the rpi graphics, I would like to invite other rpi developpers to try it so that the memory table for framebuffer can cover more situations. It seems that I've got always the same address with the same display. I'm not sure if it will change with others and how. So it would be great if other rpi developpers can have a try with my works and help me update the memory setup for framebuffer. That would have no conflict with others I think, except th 2015-06-30T15:32:29 at I'm not sure if it works on rpi2. Thats'all. I've cleaned up the "soucre license" problem . Should I send an patch with all my works or I just send an link to my github ? That's my works for the first half . 2015-06-30T15:32:30 thank you gedare 2015-06-30T15:32:53 ok yangqiao hopefully your problem can be resolved with some help there in RSB. 2015-06-30T15:32:56 yes, and the blogs I'll complete in these days 2015-06-30T15:33:09 Hello gedare, add me to the queue too , unlike last time :D 2015-06-30T15:33:19 yangqiao: to get other rpi folks to try it out, you should create a blog post with a clear tutorial for how to use it 2015-06-30T15:33:30 including any hw hookups that are needed, some photos, screen shots, etc 2015-06-30T15:33:51 The problem is that: I need to patch the source code library, 2015-06-30T15:34:05 and if there are any patches that haven't been committed to RTEMS, either push to get them committed (ping us if they haven't received feedback) by emailing to devel ml, or post the patches on github or as a branch for checking out. 2015-06-30T15:34:26 yangqiao: you may search for beng-nl blog post about using bbb with rtems, that is a good example of what i mean you can do 2015-06-30T15:34:41 [trac] SOCIS/2015 edited by salpha (empty comment) http://devel.rtems.org/wiki/SOCIS/2015 2015-06-30T15:34:45 hi sujayraaj i'll try to remember. ;) 2015-06-30T15:34:54 after yangqiao will be salpha 2015-06-30T15:35:08 when I give the path to "patch add", it asked me for some "protocol", what I've found is that the existing patches are from rtems-tools 2015-06-30T15:35:36 [trac] SOCIS/2015 edited by salpha (empty comment) http://devel.rtems.org/wiki/SOCIS/2015 2015-06-30T15:35:46 yangqiao: you should send some separate patches that are sensible commits to review. i'm the only one that really looks at the github code, so you should probably just email to the devel ml next 2015-06-30T15:35:51 what I want is to add a patch, for example,, placed in the directory sourcebuilder/patches 2015-06-30T15:36:07 ok, I'll take care 2015-06-30T15:36:11 yangqiao: you'll have to either look more in the RSB manual, or ask on email. 2015-06-30T15:36:37 rsb looks in some patches directories by default, if you don't provide it with a URL 2015-06-30T15:36:43 ok thanks yangqiao. salpha you're up 2015-06-30T15:37:01 after salpha will be ragunath 2015-06-30T15:37:13 *** yangqiao has quit IRC 2015-06-30T15:37:13 yangqiao: remember to do your student evaluation. 2015-06-30T15:37:17 oops he left. 2015-06-30T15:37:27 in this week i've updated the blog with some technical background and terminology from highly cited paper: basic concepts and taxonomy of dependable and secure computing. 2015-06-30T15:37:37 i've re-structured github repo and also updated it with latest modifications. 2015-06-30T15:37:47 grinder-rtems maven project has been created and implements a empty body for essential functions of QemuAbstraction. 2015-06-30T15:37:57 also a patch has been created to modify grinder to adapt it to rtems. 2015-06-30T15:38:12 build instructions are included in the readme file so grinder-rtems should be compiled on new machines however db should be populated (population instructions not included now) which causes run-time exceptions. 2015-06-30T15:38:24 is grinder an active project that the patch might get submitted to it? 2015-06-30T15:38:34 okay 2015-06-30T15:38:36 it is active 2015-06-30T15:38:51 but a patch is required to extend/adapt it to new targets 2015-06-30T15:38:54 alright see about how that might be done 2015-06-30T15:39:02 grinder itself should be used as submodule 2015-06-30T15:39:07 ah 2015-06-30T15:39:21 now i have to mainly focus on implementing "runExperiment" function to run the tests. it might take longer than a week to implement it because storing back results to the db should also be considered. 2015-06-30T15:39:21 and then see about using RSB to build it w/ qemu 2015-06-30T15:39:36 ok that makes sense too 2015-06-30T15:39:47 as you said, the test currently is rtems hello without any injected faults just to keep me focused on grinder first. however hello output should be stored back to db. 2015-06-30T15:40:14 in the end, i'll consider merging to RSB. 2015-06-30T15:40:35 i mean making it automatic with rsb 2015-06-30T15:40:55 ok great thanks salpha anything else? 2015-06-30T15:41:04 no, thanks! 2015-06-30T15:41:12 ok, note there won't be a meeting next week. 2015-06-30T15:41:24 hi ragunath you're up 2015-06-30T15:41:32 hello everyone 2015-06-30T15:41:37 This week patch has been submitted to RTEMS. I also addressed various comments made on the patch. 2015-06-30T15:41:42 I have also been working on why cache needs to be disabled for networking to work. 2015-06-30T15:41:51 Cache has to be disabled to make it work. So instead of disabling cache, we have to look if there is any alternative options available. 2015-06-30T15:41:59 Also we need to find why it is not working with cache enabled. 2015-06-30T15:42:11 Sebastian also made a suggestion for intr hang issue. 2015-06-30T15:42:23 In the next week, I am going to work on these two issues. 2015-06-30T15:42:25 after ragunath will be krohini. 2015-06-30T15:42:40 yeah i saw your email that was well-done, too bad marcos did not have any insight. 2015-06-30T15:42:42 Write blog post covering the progress made till now 2015-06-30T15:43:00 ok good, putting your patches somewhere with a tutorial would be good at this point also for you ragunath 2015-06-30T15:43:11 since we can't get the official versions working yet 2015-06-30T15:43:19 yes i thought they might have debugged 2015-06-30T15:44:08 is this okay for the next week? 2015-06-30T15:44:21 yes i think it will be plenty of work for even the next 2 weeks probably 2015-06-30T15:44:35 debugging the cache issue is open-ended who knows how long it may take? :) 2015-06-30T15:44:48 ok thanks ragunath. krohini you're up 2015-06-30T15:44:58 thank you 2015-06-30T15:45:04 hi 2015-06-30T15:45:07 I have finally been able to get some lead into where to start from for SMP implementation. 2015-06-30T15:45:27 with help from Sebastian Huber 2015-06-30T15:46:10 I have finally been able to get some lead into where to start from for SMP implementation. Earlier in the I was working on organizing the code for a patch for cache improvement. But there is another issue we need to work out 2015-06-30T15:46:44 as discussed.. 2015-06-30T15:46:57 ok. yes you are working to re-do the default implementations of the mminit code, right/ 2015-06-30T15:47:36 *** sgworks has quit IRC 2015-06-30T15:47:48 *** sgworks has joined #rtems 2015-06-30T15:47:52 umm,the mminit code. ya that is to make it more shareable with other bsp variants 2015-06-30T15:48:25 but even the raspberry pi mmu config table for the .text section configurations 2015-06-30T15:48:38 oh, yes 2015-06-30T15:48:42 you are digging into that, good 2015-06-30T15:48:56 if you get stumped, send a report of what you've done and found out. 2015-06-30T15:49:33 yeah.. I spent some time on the smp front so that it keeps moving ahead.. 2015-06-30T15:49:54 ok that's good too. thanks krohini anything else? sujayraaj you'll be next 2015-06-30T15:50:08 okay, sure 2015-06-30T15:50:08 thats it for now 2015-06-30T15:50:16 ok thanks, hi sujayraaj 2015-06-30T15:50:18 thanks.. 2015-06-30T15:50:34 hi. This week was a bit unusual, lot of work, little results -_- . After spending around 3 days studying for creating syscalls on bsd and mbufs, we decided to implement the sendfile later and use the present hack , because it will just be an optimization. My patches are ready to be pushed upstream, but I am waiting a bit to test on qemu. 2015-06-30T15:50:44 An hour ago I got monkey to read configuration files, and am now debugging minor errors. I think I can get it running on qemu within 5 - 6 hours. Once that gets done, I will send a pull request. 2015-06-30T15:52:09 And there is going to be a tonne of documentation generated by this whole project. 2015-06-30T15:52:18 heh good.. 2015-06-30T15:52:20 pull request to where? 2015-06-30T15:52:43 monkey , i meant getting that patch upstream to monkey 2015-06-30T15:53:08 I am trying to ensure that there should be least modification to monkey source 2015-06-30T15:53:09 ah, ok 2015-06-30T15:53:29 Though me and chris discussed the possibility of a networking api 2015-06-30T15:53:40 so what is your plan after the pull req? 2015-06-30T15:54:04 Monkey running on qemu presently is just barebones 2015-06-30T15:54:16 it has tonnes of features in form of its plugins 2015-06-30T15:54:34 they are in migration phase according to edsiper, so he has adviced me not to touch them now 2015-06-30T15:54:47 but slowly would get them running on rtems too 2015-06-30T15:54:51 secondly, 2015-06-30T15:55:19 Next phase of my gsoc project ( on my proposal ) was something like a networking api for rtems 2015-06-30T15:55:54 (though I didn't exactly use the word API ) but I am supposed to make things modular, such that any program can run either from the old stack, libbsd, or lwip 2015-06-30T15:56:13 yeah, talk with Chris about this. Certainly being able to interchange libbsd and lwip is nice 2015-06-30T15:56:28 i don't know how much attention we should be paying to the old stack, although "it works" on a lot of useful targets 2015-06-30T15:56:39 I have figured out the networking of the old stack, and libbsd, I have even made modifications in the network demos to use libbsd , so two of them are down 2015-06-30T15:57:16 Even I think that. we don't even have a functioning select. so sticking to old stack wouldn't be good. 2015-06-30T15:57:34 [trac] GSoC/2015 edited by jrcatbagan (Added status update for June 30, 2015) http://devel.rtems.org/wiki/GSoC/2015 2015-06-30T15:57:51 yeah, exactly. hard to say i guess 2015-06-30T15:57:54 ok thanks sujayraaj 2015-06-30T15:58:06 You're welcome gedare 2015-06-30T15:58:11 ketul: you can go next 2015-06-30T15:58:17 gedare: hi 2015-06-30T15:58:18 hi all this week I worked on GPIO and ADC drivers. for gpio fixed some issues suggested by you and beng-nl and submitted the finall patches using git send :) with pushing to my rtems repo in gpio_api branch. Probably GPIO drivers can be merged after community review as beng-nl suggested. Also build a new toolchain for the updated rtems. For ADC ba 2015-06-30T15:58:19 sic test done and going to have some more testing and making a video. Still some blog posts remaining on ADC and troubleshooting guide. 2015-06-30T16:00:39 ok thanks i see your patches waiting for review... 2015-06-30T16:00:51 sure me too :) 2015-06-30T16:00:51 so what is your next immediate steps while you wait for review or merge? 2015-06-30T16:01:15 working on ADC. and next target is on I2C 2015-06-30T16:01:27 so mostly I am with my timeline. 2015-06-30T16:03:04 ok, focus on polishing what you have got done then. are you thinking of supporting the touchscreen? 2015-06-30T16:03:31 gedare: ya same thing gonna ask. 2015-06-30T16:03:53 I mean is it good support for rtems community ? 2015-06-30T16:04:17 then I can reschedule my work and can have more focus on touchscreen. 2015-06-30T16:04:20 probably. i think you could consider pushing off i2c in favor of it. 2015-06-30T16:04:29 chat with beng-nl about his thoughts on it 2015-06-30T16:04:57 ya ok. I had a thought about some sensors interfacing using ADC 2015-06-30T16:06:27 ok well, it is best to get some concrete examples to demo if you think of something 2015-06-30T16:06:29 *** yangqiao has joined #rtems 2015-06-30T16:06:33 then I will ask beng-nl about his views and reschedule the work flow if required 2015-06-30T16:06:46 thanks ketul and i'll try to review your code before I leave hopefully. 2015-06-30T16:06:53 yaa i will post some videos 2015-06-30T16:06:54 anandkp92: you can go next 2015-06-30T16:06:59 ok 2015-06-30T16:07:09 hi :) 2015-06-30T16:07:16 so last week, I've created the gui and the display and event hanlders for Boolean and Integer options are done. Next step would be to include String and String List options. Currently, the options are put into tabs based on type, but that will be changed to groups like we had discussed while preparing the drafts. 2015-06-30T16:07:25 Thanks gedare. waiting for review 2015-06-30T16:08:11 though the groups are already present, we are currently waiting to finish all the different types of options 2015-06-30T16:08:14 and then include it 2015-06-30T16:08:56 i had mailed to devel about this and also updated my blog with the report too. 2015-06-30T16:09:09 were you able to check out the gui? 2015-06-30T16:09:15 ok good. i haven't been able to run it yet, no 2015-06-30T16:09:24 i probably won't get to it for awhile 2015-06-30T16:09:24 oh ok. 2015-06-30T16:09:49 we've run it on mac and linux and it looks decent. 2015-06-30T16:09:56 so you'll be working on strings next? 2015-06-30T16:10:01 we would like to get a review from someone who uses windows 2015-06-30T16:10:02 yes. 2015-06-30T16:10:15 it wouldn't take much time - including strings and string list. 2015-06-30T16:10:33 ok great. are there any more types after that? 2015-06-30T16:10:33 after that we'll work towards creating the config.cfg file - completing the purpose of gui 2015-06-30T16:10:37 okay 2015-06-30T16:10:56 so far no more types. but we could be including. 2015-06-30T16:11:04 havent discussed that with my mentors yet. 2015-06-30T16:12:42 thats it for now. 2015-06-30T16:12:49 ok, i guess i would focus on getting all the functionality in place and making it easy to incrementally add the content. 2015-06-30T16:12:58 ok thanks anandkp92. asuolen you'll be up next, anything else anandkp92 ? 2015-06-30T16:13:01 ok 2015-06-30T16:13:13 no, thanks gedare 2015-06-30T16:13:19 hello 2015-06-30T16:13:21 I have been addressing the issues/suggestions noted on the GPIO API, and am at this point writing my midterm report, which details the current state of the GPIO API. 2015-06-30T16:13:27 One of the reasons for this delay is a problem with the generic_isr that arised when I changed the pin state data structure, which for some reason causes a crash for no aparent reason when the application uses interrupts (e.g.: just adding a new field to the struct causes a crash when the rtems interrupt handler is called). 2015-06-30T16:13:34 I am not sure if this is something with the RPI BSP, but if I have the clock driver enabled the application blocks immediatly, while if I have the clock driver disabled the application runs until I cause an interrupt. For now I have put a dummy rtems_chain_control which is not needed, but by having it there I can continue the work, so until now I have ignored it but in the next days I will have to address it. 2015-06-30T16:15:09 *** yangqiao has quit IRC 2015-06-30T16:15:57 ok some kind of corruption i guess 2015-06-30T16:16:12 it's good to hunt that down in case it is a general bug in your code 2015-06-30T16:17:14 yes, although i have had problems with the pi's clock driver in the past also while loading a filesystem i think 2015-06-30T16:17:42 ok well if you get a reproducible test case that can be something worth submitting / discussing 2015-06-30T16:17:50 anything else asuolen? sgworks you'll be next 2015-06-30T16:18:25 yes thats what i do 2015-06-30T16:18:51 last time was with rtems_fsmout which also blocked 2015-06-30T16:19:25 ok thank asuolen it's always tricky to debug interrupts, especially on a target without a good debugger :) 2015-06-30T16:19:29 so, good luck! 2015-06-30T16:19:34 *** anandkp92 has quit IRC 2015-06-30T16:19:34 hi sgworks 2015-06-30T16:19:42 hi, 1. work on creation, solution and testing of spsem03 test case suggested by Sebastian. 2. try to find alternative for big lock 3. With big locks try to test under different combination of threads and locks using Verify() 4. Annotating JPF-Code 2015-06-30T16:19:52 *** zoso has quit IRC 2015-06-30T16:19:54 submitted the midterm evaluation 2015-06-30T16:20:21 Do I need to have seperate blog post related to midterm evaluation? 2015-06-30T16:20:42 i think a midterm report on your blog post would be nice, to summarize your progress and your goals for the second half 2015-06-30T16:20:48 if so I will aggregate my stuff and put a blog about it. 2015-06-30T16:20:52 okay 2015-06-30T16:21:43 *** toxygen has joined #rtems 2015-06-30T16:21:54 mailed regarding the all the stuff last week 2015-06-30T16:22:13 ok i guess you have your goals for the next 2 weeks at least, and your mentors have been in good contact with you and expressed that you should continue your work at the current pace 2015-06-30T16:22:35 yes 2015-06-30T16:22:37 no problem 2015-06-30T16:22:45 ok 2015-06-30T16:22:46 yeah sounds good 2015-06-30T16:22:50 thanks sgworks 2015-06-30T16:22:52 was problem with raghunath 2015-06-30T16:22:57 in rx buffer 2015-06-30T16:22:59 ? 2015-06-30T16:23:05 What about it? 2015-06-30T16:23:19 ?? 2015-06-30T16:23:19 I faced the same problem while developing network driver from scratch 2015-06-30T16:23:32 ok, no gtament today, did i miss anyone else? 2015-06-30T16:23:43 *** dougk_ has quit IRC 2015-06-30T16:23:55 Hi raghunath we can have private chat 2015-06-30T16:24:02 ok sure 2015-06-30T16:24:09 Thats all from me 2015-06-30T16:24:21 doesn't look like it to me, ok thanks everyone have a nice 2 weeks and everyone should be passing the midterm as far as I know 2015-06-30T16:24:40 yay 2015-06-30T16:24:48 :-) 2015-06-30T16:24:52 *** salpha has quit IRC 2015-06-30T16:26:00 thats nice :) 2015-06-30T16:26:54 *** beng-nl-droid has quit IRC 2015-06-30T16:28:10 *** sgworks has quit IRC 2015-06-30T16:30:08 *** mumptai has joined #rtems 2015-06-30T16:30:13 *** mumptai_ has joined #rtems 2015-06-30T16:31:08 :) gedare have a great vacation :D 2015-06-30T16:35:04 thanks 2015-06-30T16:35:48 *** dougk_ has joined #rtems 2015-06-30T16:36:11 *** mumptai_ has quit IRC 2015-06-30T16:48:59 *** asuolen has quit IRC 2015-06-30T16:49:24 *** jrcatbagan has quit IRC 2015-06-30T17:05:28 *** ketul has quit IRC 2015-06-30T17:06:44 *** ketul has joined #rtems 2015-06-30T17:06:50 hi beng-nl 2015-06-30T17:12:17 *** ketul has quit IRC 2015-06-30T17:16:45 *** Davidbrcz has joined #rtems 2015-06-30T17:22:13 *** Davidbrcz has quit IRC 2015-06-30T17:22:20 *** Davidbrcz has joined #rtems 2015-06-30T17:28:18 gedare : are you online? I had a small ( and a bit irrelevant ) question to ask. 2015-06-30T17:37:22 *** krohini has quit IRC 2015-06-30T17:42:02 *** sujayraaj has quit IRC 2015-06-30T17:46:28 *** sujayraaj has joined #rtems 2015-06-30T17:52:03 hi sujayraaj 2015-06-30T17:53:39 hi gedare, just wanted to ask , do you know if people use rtems for game development? We do have a nds and gba bsp with us. 2015-06-30T17:55:25 I was a bit involved into game development for handhelds (read GBA and nds) an year ago, and most of the tutorials started right from creating their own cross compiled gcc 2015-06-30T17:56:31 *** krohini has joined #rtems 2015-06-30T17:56:37 given that now I know how powerful rtems is, game development really seems something that can be done on it. Its surprising if that is not the case 2015-06-30T17:57:24 sujayraaj: Not that I'm aware of but it's possible. Probably some hobbyists have done small projects. Most likely those ports came about just because someone wanted to give it a try though. 2015-06-30T18:01:44 Oh. They don't seem active in development. I have a few half-boiled gba games I created. I was thinking of getting them on rtems. Would be a nice tutorial of sorts to get others involved :) 2015-06-30T18:03:37 yeah those targets have probably bit-rotted by now 2015-06-30T18:03:57 *** ita_ is now known as ita 2015-06-30T18:04:13 *** ita has joined #rtems 2015-06-30T18:06:17 Hi all 2015-06-30T18:07:41 I haven't quite kept track of who's logged in since 4 hours ago, but I've just got back into trying to solve the microtime issue 2015-06-30T18:08:20 Has anyone seen this issue before? 2015-06-30T18:08:32 http://pastebin.com/rEyDXcd7 2015-06-30T18:08:54 *** ita has quit IRC 2015-06-30T18:08:58 sys/time.h has a non-static declaration, while rtems/cpukit/libnetworking/rtems/rtems_bsdnet_internal.h has a static declaration. Removing the static results in a multiple declaration error during linking of nsfclient (I'll quickly re-build to get a paste) 2015-06-30T18:09:10 ttp://pastebin.com/T3Mer2fa 2015-06-30T18:12:22 that seems familiar 2015-06-30T18:13:45 igutek: ok so the easy workaround is to use --disable-networking 2015-06-30T18:13:57 when you ./configure 2015-06-30T18:14:24 this problem stems from changes made to support the newer net stack in libbsd, i guess there is a conflict now 2015-06-30T18:15:25 however, this may also be a problem of an out-of-sync compiler tools and rtems version, since many changes happened to the time code about 1-1.5 months ago 2015-06-30T18:18:56 igutek: I see where this problem was introduced. sebhub imported sys/time.h from FreeBSD and it had non-static declaration 2015-06-30T18:20:11 but it looks like it should have been fixed already by https://git.rtems.org/rtems/commit/?id=bf11690ca24743008dc659a74d3ea7b13c3b29fe 2015-06-30T18:20:46 igutek: is your version of RTEMS missing that^? 2015-06-30T18:21:12 If I use an older compiler (about 1-1.5 months ago), it works 2015-06-30T18:23:01 I think so. Is that in master? 2015-06-30T18:24:53 yeah should be 2015-06-30T18:25:10 *** ragunath has quit IRC 2015-06-30T18:25:25 *** ragunath has joined #rtems 2015-06-30T18:26:39 igutek: actually master has another update to it 2015-06-30T18:26:41 7cd2484c (Alexander Krutwig 2015-05-12 14:32:47 +0200 65) #define microtime(tv) 2015-06-30T18:27:57 I think I setup a faulty local mirror, so I'm going to pull rtems.org master again 2015-06-30T18:30:12 *** ita has joined #rtems 2015-06-30T18:34:01 *** monstr has quit IRC 2015-06-30T18:34:21 Looks like I am having git troubles, and this is already fixed. Thanks for your help! 2015-06-30T18:38:56 ok good luck 2015-06-30T18:43:48 *** gedare has quit IRC 2015-06-30T18:45:41 *** gedare has joined #rtems 2015-06-30T18:45:41 *** ChanServ sets mode: +o gedare 2015-06-30T19:03:57 *** ragunath has quit IRC 2015-06-30T19:34:21 *** sujayraaj has quit IRC 2015-06-30T19:36:51 *** krohini has quit IRC 2015-06-30T19:58:28 *** Davidbrcz has quit IRC 2015-06-30T20:09:58 gedare: how nice you think my post is a good example :) 2015-06-30T20:10:09 (seriously) 2015-06-30T21:20:00 *** cdesai has quit IRC 2015-06-30T21:23:52 *** cdesai has joined #rtems 2015-06-30T21:31:19 *** Davidbrcz has joined #rtems 2015-06-30T21:32:11 *** Davidbrcz has joined #rtems 2015-06-30T21:49:34 *** kiwichris has joined #rtems 2015-06-30T22:22:37 *** stryx` has quit IRC 2015-06-30T22:23:01 *** mumptai has quit IRC 2015-06-30T22:26:00 *** gedare has quit IRC 2015-06-30T22:30:54 *** stryx` has joined #rtems 2015-06-30T23:01:34 *** toxygen has quit IRC 2015-06-30T23:05:01 *** jkridner has quit IRC 2015-07-01T00:02:51 *** jkridner has joined #rtems 2015-07-01T02:22:17 *** Davidbrcz has quit IRC 2015-07-01T03:11:44 *** ketul has joined #rtems 2015-07-01T03:31:53 *** dr__house is now known as dr__house|away 2015-07-01T03:40:30 *** javamonn has quit IRC 2015-07-01T03:57:00 *** sujayraaj has joined #rtems 2015-07-01T04:02:15 kiwichris : hi! Finally I was succesful in creating and untaring the files. It had to do with the options I was passing to tar , which should be '-cvf' . Nonetheless, that is done. So monkey starts fine but when it gets to its event loop, I get a runtime error pipe: Function not implemented, errno=88. is 'pipe' functional in libbsd? Because it compiles and links very well, but gives a runtime error. 2015-07-01T04:03:01 sujayraaj, nice work on the tar debugging. It can be hard. 2015-07-01T04:03:11 sujayraaj, what is Monkey using pipe for ? 2015-07-01T04:06:51 _mk_event_channel_create ? 2015-07-01T04:08:04 Comment says, it Registers the scheduler channel to signal active workers. 2015-07-01T04:08:31 In your github repo which branch is the code you are using on ? 2015-07-01T04:09:33 I am using monkey head , though it may be about 5-7 days old. by the way, the comment on its use was in : mk_server/mk_scheduler.c 2015-07-01T04:09:53 where it mk_event_channel_create is called in mk_sched_launch_worker_loop 2015-07-01T04:11:42 Branch in _your_ repo is ? 2015-07-01T04:12:46 line number in mk_server/mk_scheduler.c ? 2015-07-01T04:13:21 branch is * master , the line number is 346 2015-07-01T04:15:32 basically, mk_event_channel_create is a frontend to the _mk_event_channel_create. 2015-07-01T04:15:32 https://github.com/sujayraaj/monkey/blob/master/mk_core/mk_event_kqueue.c#L196 is static so how is the other file seeing it ? 2015-07-01T04:16:12 Is there some sort of event wrapper happening ? 2015-07-01T04:16:57 yeah. in mk_core/mk_event.c , there is a function defined mk_event_channel_create, which wraps _mk_event_channel_create 2015-07-01T04:17:00 pipes just create overhead 2015-07-01T04:17:46 So what options do we have? 2015-07-01T04:18:24 We should use it. Pipes exist in RTEMS proper. 2015-07-01T04:18:43 Is the pipe call working on line 203 of the create function ? 2015-07-01T04:18:52 We can address performance issues later 2015-07-01T04:19:40 Yeah. line 203 of mk_event_kqueue.c 2015-07-01T04:20:12 Ok, I will dig out something for you, hang on ... 2015-07-01T04:21:21 Try adding to your init.c where condefs.h is included .... 2015-07-01T04:21:22 #define CONFIGURE_FIFOS_ENABLED 2015-07-01T04:21:35 #define CONFIGURE_MAXIMUM_FIFOS 8 2015-07-01T04:21:48 #define CONFIGURE_MAXIMUM_PIPES 8 2015-07-01T04:22:10 okat, let me check 2015-07-01T04:22:12 and see what happens. I am not sure how many pipes Monkey wants 2015-07-01T04:22:15 thanks 2015-07-01T04:29:26 *** monstr has joined #rtems 2015-07-01T04:29:46 *** sujayraaj__ has joined #rtems 2015-07-01T04:30:42 *** sujayraaj has quit IRC 2015-07-01T04:31:20 Thanks kiwichris, that was exactly what was needed. Though, I am getting an error 22 (kevent :Invalid argument) at mk_event_kqueue.c:95. 2015-07-01T04:32:46 Is 22 the errno ? 2015-07-01T04:33:52 yeah. kevent: Invalid argument, errno=22 2015-07-01T04:36:28 Note: In my init file, I set #define CONFIGURE_LIBIO_MAXIMUM_FILE_DESCRIPTORS 512 , to match with what monkey was using ( I handcoded that to 512 too). Is it too large? 2015-07-01T04:36:57 man page for kevent says "[EINVAL] The specified time limit or filter is invalid." 2015-07-01T04:40:46 Can you step into the kevent and see where the error is happening ? 2015-07-01T04:41:01 sebhub will about soon and he might need to help here 2015-07-01T04:43:39 kiwichris :I think I may have figured it out! What version of FreeBSD rtems uses for its libbsd? Because I saw the man page of kqueue, and the historic man page of kqueue https://people.freebsd.org/~jmg/kqueue.historic.man.html 2015-07-01T04:43:59 and definition of both are different. They exchanged second and third parameters 2015-07-01T04:44:42 ooh, we are a 9 series kernel 2015-07-01T04:45:06 Try the change and see then we will need to find the magic to allow Monkey work on both versions 2015-07-01T04:46:07 Yeah. I am doing that right now. 2015-07-01T04:59:09 kiwichris : I verified, that is not the case. Our libbsd uses the current definition of kqueue. So there must be something else that is wrong. 2015-07-01T04:59:40 *current definition of kevent 2015-07-01T05:01:49 Ok 2015-07-01T05:01:55 Can you step in ? 2015-07-01T05:02:06 sebhub, are you about yet ? 2015-07-01T05:03:24 I am doing that right now. 2015-07-01T05:31:37 *** sujayraaj__ has left #rtems 2015-07-01T05:38:28 *** sujayraaj has joined #rtems 2015-07-01T05:44:07 *** mumptai has joined #rtems 2015-07-01T06:20:57 good morning 2015-07-01T06:22:24 pipes are not supported with the libbsd kqueue 2015-07-01T06:23:00 it is possible to add this, but the pipe implementation needs to get rewritten from scratch if you ask me 2015-07-01T06:25:07 [git] fdb45d6 by Sebastian Huber: [score: Freechain handler API changes Replace the extend function with an allocator since this fits better to the current use case.] http://devel.rtems.org/changeset/fdb45d6/rtems 2015-07-01T06:25:07 [git] f59f2fe by Sebastian Huber: [score: Accept NULL pointer in _Freechain_Put() With this a _Freechain_Put( _Freechain_Get() ) works always.] http://devel.rtems.org/changeset/f59f2fe/rtems 2015-07-01T06:32:43 [git] 372bee8 by Sebastian Huber: [score: Simplify _SMP_ticket_lock_Release() Add a SMP lock statistics pointer to SMP_lock_Stats_context and drop the SMP lock statistics parameter from _SMP_ticket_lock_Release().] http://devel.rtems.org/changeset/372bee8/rtems 2015-07-01T06:34:59 *** ketul has quit IRC 2015-07-01T06:35:58 good morning sebhub, So, what other alternatives can we use instead of pipe? Because I don't think we have other alternatives for kqueue. 2015-07-01T06:36:36 you can add kqueue support to the pipes and fix the pipe implementation 2015-07-01T06:37:55 *** ketul has joined #rtems 2015-07-01T08:00:52 sujayraaj, can you please chat to Eduardo about this and see if we can sort it out in Monkey ? 2015-07-01T08:01:17 sujayraaj, can this please be cc'ed to devel list; thanks 2015-07-01T08:03:43 *** mumptai has quit IRC 2015-07-01T08:52:28 kiwichris : I was out having lunch. :) . By the way, what about poll? 1. does rtems has working poll support? 2. Will poll work with pipe? Because as it seems communication between threads is necessary for monkey. I will chat with eduardo too about this. 2015-07-01T08:53:07 kiwichris : further, if poll+pipe combination would work, I would just need to back poll in the mk_event_* functions , which isn't a big deal 2015-07-01T08:53:19 and can be done within a day or two 2015-07-01T08:56:01 poll kqueue select, this is all the same under the hood 2015-07-01T09:14:05 [git] 635ed82 by Sebastian Huber: [bsp/qoriq: Update due to API changes] http://devel.rtems.org/changeset/635ed82/rtems 2015-07-01T09:29:53 sujayraaj, what data is Monkey passing between threads ? 2015-07-01T09:32:37 kiwichris : I would have to look deeper into that. 2015-07-01T09:33:45 sujayraaj, please do. Using pipes for comms between threads is easy but heavy even on Unix. If it is for exec'ed processes for CGI type functions then RTEMS does not need this. 2015-07-01T09:50:07 *** krohini has joined #rtems 2015-07-01T10:04:22 kiwichris : Well, I disabled the mk_channel create completely, ran monkey , and it run to the point " [ Info ] HTTP Server started" , which got my hopes really high, but to process the connection, monkey has to call up its workers ( which are basically threads). and then those threads handle our connections. In normal run on my linux box, monkey calls up 8 workers initially, and as a connection arrives it assigns that connection 2015-07-01T10:06:02 sujayraaj, good start; I assume a worker and so a thread is given a socket when a connection comes in. Is this correct ? 2015-07-01T10:06:28 Yes. Exactly 2015-07-01T10:07:51 So what role is the pipe doing ? 2015-07-01T10:07:57 And without the mk_channel_create, those worker threads never start. 2015-07-01T10:08:29 Start or run ? 2015-07-01T10:08:43 They map block forever 2015-07-01T10:08:51 may that is 2015-07-01T10:11:36 kiwichris : From the look of it at the moment, they don't start. Do try once running monkey on your machine with WITH_TRACE option in CMakelist set to Yes. Provides good insight. On my linux box, when a worker starts there is an output of " Worker started (SIGNAL_START)" for i from 0 to 7. Without mk_channel_create, we never reach this stage 2015-07-01T10:12:26 I suggest you keep digging in and see if you can find the reason for not starting. 2015-07-01T10:12:47 I am travelling for a couple of days and my access will be limited. 2015-07-01T10:14:01 *** ketul has quit IRC 2015-07-01T10:14:09 Oh! Okay. Sure. Btw, when monkey starts in qemu, I can ping as usual to 10.1.2.5 ( which is its address ) but I can also telnet it on the port 2001 ( which is the port monkey is listening to ) 2015-07-01T10:15:57 But no replies to browser requests. So I think everything that we have done uptil now is good enough, and perhaps this is the last hurdle to get monkey running. ( this makes it both exciting and frustrating :D ) 2015-07-01T10:26:21 *** sujayraaj has quit IRC 2015-07-01T10:34:02 *** sujayraaj has joined #rtems 2015-07-01T11:52:07 *** jkridner has quit IRC 2015-07-01T11:58:45 *** kiwichris has quit IRC 2015-07-01T12:14:10 *** igutek has quit IRC 2015-07-01T12:55:42 *** gedare has joined #rtems 2015-07-01T12:55:43 *** ChanServ sets mode: +o gedare 2015-07-01T14:07:02 *** igutek has joined #rtems 2015-07-01T14:39:05 *** rtems-irc` has joined #rtems 2015-07-01T14:39:50 *** rtems-trac_ has joined #rtems 2015-07-01T14:41:50 *** commande1 has joined #rtems 2015-07-01T14:42:35 *** gigetoo_ has joined #rtems 2015-07-01T14:43:27 *** gedare has quit IRC 2015-07-01T14:45:32 *** monstr has quit IRC 2015-07-01T14:46:23 *** gedare has joined #rtems 2015-07-01T14:46:24 *** ChanServ sets mode: +o gedare 2015-07-01T14:46:35 *** guerby has quit IRC 2015-07-01T14:46:36 *** gigetoo has quit IRC 2015-07-01T14:46:37 *** commander has quit IRC 2015-07-01T14:46:38 *** rtems-trac has quit IRC 2015-07-01T14:46:38 *** rtems-irc has quit IRC 2015-07-01T14:46:39 *** peerst has quit IRC 2015-07-01T14:46:42 *** gigetoo_ is now known as gigetoo 2015-07-01T14:55:12 *** gedare has quit IRC 2015-07-01T14:55:15 *** guerby has joined #rtems 2015-07-01T15:03:52 *** op_ has joined #rtems 2015-07-01T15:04:52 Hi, i need help on getting my bsp driver running. I have read the doc but something is missing i think. Anybody keen enough? 2015-07-01T15:05:27 *** peerst has joined #rtems 2015-07-01T15:16:22 [trac] GSoC/2015 edited by Asuol (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-07-01T15:21:23 *** op_ has quit IRC 2015-07-01T15:49:20 *** dr__house|away is now known as dr__house 2015-07-01T15:56:41 *** sujayraaj has quit IRC 2015-07-01T16:04:05 *** sujayraaj has joined #rtems 2015-07-01T16:08:54 *** jkridner has joined #rtems 2015-07-01T16:09:56 *** jkridner has joined #rtems 2015-07-01T17:05:42 *** jkridner|work has joined #rtems 2015-07-01T17:06:33 *** Davidbrcz has joined #rtems 2015-07-01T17:08:46 *** jkridner|work has quit IRC 2015-07-01T17:09:45 *** jkridner has quit IRC 2015-07-01T18:16:31 *** ketul has joined #rtems 2015-07-01T19:26:34 *** mumptai has joined #rtems 2015-07-01T19:40:57 *** Davidbrcz has quit IRC 2015-07-01T20:05:56 *** ketul has quit IRC 2015-07-01T20:41:30 *** Davidbrcz has joined #rtems 2015-07-01T21:02:56 *** mumptai_ has joined #rtems 2015-07-01T21:03:04 *** mumptai has quit IRC 2015-07-01T21:22:59 *** guerby has quit IRC 2015-07-01T21:26:41 *** guerby has joined #rtems 2015-07-01T22:11:41 *** timothyh has quit IRC 2015-07-01T22:13:10 *** shelf has joined #rtems 2015-07-01T22:13:10 *** shelf has joined #rtems 2015-07-01T22:50:00 *** sujayraaj has quit IRC 2015-07-01T23:20:17 *** kiwichris has joined #rtems 2015-07-01T23:39:56 *** kiwichris has quit IRC 2015-07-01T23:40:23 *** kiwichris has joined #rtems 2015-07-01T23:49:21 *** kiwichris has quit IRC 2015-07-02T00:23:40 *** gigetoo has quit IRC 2015-07-02T00:40:38 *** gigetoo has joined #rtems 2015-07-02T01:27:54 *** sujayraaj has joined #rtems 2015-07-02T01:59:04 *** Davidbrcz has quit IRC 2015-07-02T02:07:48 *** sujayraaj has quit IRC 2015-07-02T02:58:18 *** kiwichris has joined #rtems 2015-07-02T03:55:45 *** dr__house is now known as dr__house|away 2015-07-02T05:20:44 *** monstr has joined #rtems 2015-07-02T05:41:38 *** sujayraaj has joined #rtems 2015-07-02T07:01:45 good morning 2015-07-02T07:01:45 kiwichris : hi 2015-07-02T07:02:13 good morning sebhub 2015-07-02T07:19:57 sebhub : About the issue we talked about yesterday , pipe with kqueue , I think presently fixing it would be the only way out, the other way would have been using something like a socketpair, but presumably we don't have socketpair. So, I think I should get down to fix it. Can you help me get a start, like maybe citing a reference from where I can read or maybe tell me what I am supposed to look at, because I am a bit inexperienced 2015-07-02T07:23:19 i would look at the existing pipe implementation and move it to libbsd, then re-implement it with condition variables, then add the kqueue and poll handlers 2015-07-02T07:23:46 before this i would write a test case that covers all your requirements 2015-07-02T07:25:31 *** shelf is now known as timothyh 2015-07-02T07:26:58 Okay. I will try to do it. Thanks 2015-07-02T08:47:23 *** hesham has joined #rtems 2015-07-02T09:33:23 *** kiwichris has quit IRC 2015-07-02T09:55:31 *** mkhoory-2 has joined #rtems 2015-07-02T09:57:46 *** mkhoory has quit IRC 2015-07-02T09:57:46 *** sebhub has quit IRC 2015-07-02T09:57:55 *** sebhub has joined #rtems 2015-07-02T09:58:36 *** igutek has quit IRC 2015-07-02T10:00:06 *** igutek has joined #rtems 2015-07-02T11:10:45 *** hesham has quit IRC 2015-07-02T12:41:42 *** sujayraaj has quit IRC 2015-07-02T12:48:18 *** jkridner has joined #rtems 2015-07-02T13:18:51 *** sujayraaj has joined #rtems 2015-07-02T13:42:04 *** kiwichris has joined #rtems 2015-07-02T13:51:54 *** monstr has quit IRC 2015-07-02T14:00:26 *** kiwichris has quit IRC 2015-07-02T14:12:21 *** jkridner has quit IRC 2015-07-02T14:40:39 *** sujayraaj has quit IRC 2015-07-02T14:56:01 *** sujayraaj has joined #rtems 2015-07-02T15:25:27 *** sujayraaj has quit IRC 2015-07-02T15:31:59 *** dr__house|away is now known as dr__house 2015-07-02T16:25:44 *** mumptai has joined #rtems 2015-07-02T17:27:03 *** ketul has joined #rtems 2015-07-02T17:27:59 hi beng-nl :) 2015-07-02T17:41:12 *** Davidbrcz has joined #rtems 2015-07-02T18:31:34 *** giusef has joined #rtems 2015-07-02T18:35:45 *** monstr has joined #rtems 2015-07-02T18:38:02 Hi everyone 2015-07-02T18:38:16 Anyone have experience getting the IMFS working on the STM32F4? 2015-07-02T18:41:06 Or, in general, what issues might arise when trying to get the example to work? I'm suspecting linker script issues 2015-07-02T19:02:22 *** ketul has quit IRC 2015-07-02T19:07:00 *** monstr has quit IRC 2015-07-02T20:08:26 *** Davidbrcz has quit IRC 2015-07-02T20:43:39 *** Davidbrcz has joined #rtems 2015-07-02T21:44:03 *** kiwichris has joined #rtems 2015-07-02T22:04:08 *** kiwichris has quit IRC 2015-07-02T22:35:48 *** giusef has quit IRC 2015-07-02T23:17:40 *** mumptai has quit IRC 2015-07-02T23:37:43 *** kiwichris has joined #rtems 2015-07-03T00:26:57 *** sujayraaj has joined #rtems 2015-07-03T01:55:05 *** Davidbrcz has quit IRC 2015-07-03T04:59:48 *** dr__house is now known as dr__house|away 2015-07-03T05:21:18 *** monstr has joined #rtems 2015-07-03T06:10:40 good morning 2015-07-03T06:47:14 *** giusef has joined #rtems 2015-07-03T06:55:49 *** kiwichris has quit IRC 2015-07-03T07:09:14 *** ita_ has joined #rtems 2015-07-03T07:11:06 *** ita has quit IRC 2015-07-03T07:11:13 *** edsiper has quit IRC 2015-07-03T07:11:23 *** edsiper has joined #rtems 2015-07-03T07:15:06 *** mumptai has joined #rtems 2015-07-03T07:20:11 *** giusef has quit IRC 2015-07-03T07:25:38 *** zoso has joined #rtems 2015-07-03T08:04:26 *** mumptai has quit IRC 2015-07-03T08:18:51 *** kiwichris has joined #rtems 2015-07-03T08:47:14 *** kiwichris has quit IRC 2015-07-03T10:46:27 *** zoso has quit IRC 2015-07-03T13:14:16 *** sebhub has quit IRC 2015-07-03T13:41:53 *** monstr has quit IRC 2015-07-03T16:39:29 *** sujayraaj has quit IRC 2015-07-03T17:00:01 *** commande1 is now known as commander 2015-07-03T17:15:22 *** mumptai has joined #rtems 2015-07-03T19:12:10 *** Davidbrcz has joined #rtems 2015-07-03T21:10:26 *** Davidbrcz has quit IRC 2015-07-03T22:21:26 *** kiwichris has joined #rtems 2015-07-04T01:07:44 *** sujayraaj has joined #rtems 2015-07-04T01:10:23 hi kiwichris 2015-07-04T01:32:19 *** sujayraaj has left #rtems 2015-07-04T04:51:49 *** Davidbrcz has joined #rtems 2015-07-04T07:38:47 *** Davidbrcz has quit IRC 2015-07-04T12:44:03 *** sujayraaj has joined #rtems 2015-07-04T13:47:41 [trac] GSoC/2015 edited by anandkp92 (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-07-04T13:49:46 [trac] GSoC/2015/GUI edited by anandkp92 (empty comment) http://devel.rtems.org/wiki/GSoC/2015/GUI 2015-07-04T14:30:14 *** sujayraaj has quit IRC 2015-07-04T17:47:20 hi all 2015-07-04T19:09:31 *** ankur has joined #rtems 2015-07-04T19:40:29 *** Davidbrcz has joined #rtems 2015-07-04T19:47:01 *** ankur has quit IRC 2015-07-04T21:13:36 *** Davidbrcz has quit IRC 2015-07-04T21:22:50 *** Davidbrcz has joined #rtems 2015-07-04T21:55:36 *** Davidbrcz has quit IRC 2015-07-04T23:13:36 *** mumptai has quit IRC 2015-07-05T01:06:36 *** Davidbrcz has joined #rtems 2015-07-05T01:19:24 [git] 37e453e by Jarielle Catbagan: [BBB: am335x.h: Fix/update * Fixed file header * Formatted code to be atleast below the 80 character limit * Incorporated defines for CM_WKUP registers * Included defines for UART0 RXD/TXD Control Module registers * Added defines for pad control bits] http://devel.rtems.org/changeset/37e453e/umon 2015-07-05T01:19:24 [git] a5f94c8 by Jarielle Catbagan: [BBB: Migrate custom exception handler mapping from rom_reset.S to ram_vector_install() in cpuio.c] http://devel.rtems.org/changeset/a5f94c8/umon 2015-07-05T01:19:24 [git] d77ed25 by Jarielle Catbagan: [BBB: cpuio.*: Add UART0 initialization and set DEFAULT_BAUD_RATE to 115200] http://devel.rtems.org/changeset/d77ed25/umon 2015-07-05T01:19:24 [git] 27db08e by Jarielle Catbagan: [BBB: cpuio.h: Remove MONARGV0] http://devel.rtems.org/changeset/27db08e/umon 2015-07-05T01:19:24 [git] 40b5966 by Jarielle Catbagan: [BBB: cpuio.c: Set devInit() to return 0 by default devInit() is not currently being used. With the current state of uMon, all initializations unique to the AM335x are performed in initCPUio() for now] http://devel.rtems.org/changeset/40b5966/umon 2015-07-05T01:19:24 [git] dd9f290 by Jarielle Catbagan: [BBB: Makefile: Remove irrelevant source files from CPUCSRC, IODEVSRC, and FLASHSRC] http://devel.rtems.org/changeset/dd9f290/umon 2015-07-05T01:19:24 [git] f94045f by Jarielle Catbagan: [BBB: xcmddcl.h/xcmdtbl.h: Remove irrelevant commands] http://devel.rtems.org/changeset/f94045f/umon 2015-07-05T01:19:25 [git] f344230 by Jarielle Catbagan: [BBB: Build an MLO file when booting from non-XIP devices (e.g. SD) An MLO file will be created from the concatenation of a corresponding GP header and the uMon image. A GP header must be prepended to the uMon image when booting from non-XIP devices. Refer to the AM335x TRM, Section 26.1.9 for more information. These type 2015-07-05T01:19:26 of images have been tested...] http://devel.rtems.org/changeset/f344230/umon 2015-07-05T01:31:20 *** Davidbrcz has quit IRC 2015-07-05T04:04:00 *** kiwichris has quit IRC 2015-07-05T04:35:50 *** crazy3k has joined #rtems 2015-07-05T06:56:46 *** kiwichris has joined #rtems 2015-07-05T07:11:01 *** Davidbrcz has joined #rtems 2015-07-05T07:43:36 *** Davidbrcz has quit IRC 2015-07-05T09:26:43 *** mumptai has joined #rtems 2015-07-05T10:06:27 *** crazy3k has quit IRC 2015-07-05T10:26:24 *** mumptai has quit IRC 2015-07-05T12:00:36 *** sujayraaj has joined #rtems 2015-07-05T12:16:29 *** ketul has joined #rtems 2015-07-05T13:37:21 *** sujayraaj has quit IRC 2015-07-05T13:49:23 *** guerby has quit IRC 2015-07-05T13:52:41 *** guerby has joined #rtems 2015-07-05T14:31:13 *** sujayraaj has joined #rtems 2015-07-05T14:35:27 *** sujayraaj has quit IRC 2015-07-05T14:39:33 *** ketul has quit IRC 2015-07-05T14:40:40 *** ketul has joined #rtems 2015-07-05T14:42:18 *** ketul has quit IRC 2015-07-05T14:43:12 *** ketul has joined #rtems 2015-07-05T15:26:29 *** ketul has quit IRC 2015-07-05T18:18:17 *** Davidbrcz has joined #rtems 2015-07-05T21:31:14 *** Davidbrcz has quit IRC 2015-07-05T22:07:43 *** kiwichris has quit IRC 2015-07-05T22:31:29 *** kiwichris has joined #rtems