2013-12-09T00:21:21 *** andyjin has left #rtems 2013-12-09T00:56:41 *** gigetoo has quit IRC (Remote host closed the connection) 2013-12-09T00:58:37 *** arvind_khadri has joined #rtems 2013-12-09T01:00:26 *** gigetoo has joined #rtems 2013-12-09T03:40:02 *** sebhub has joined #rtems 2013-12-09T05:37:12 *** YuGin has joined #rtems 2013-12-09T05:55:31 *** sebhub has quit IRC (Ping timeout: 260 seconds) 2013-12-09T07:18:09 *** antgreen has quit IRC (Ping timeout: 246 seconds) 2013-12-09T07:23:35 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2013-12-09T07:23:54 *** arvind_khadri has joined #rtems 2013-12-09T08:11:09 *** gcibot has quit IRC (Remote host closed the connection) 2013-12-09T08:11:41 *** marcinb has joined #rtems 2013-12-09T08:12:19 *** gcibot has joined #rtems 2013-12-09T08:14:40 *** gcibot has joined #rtems 2013-12-09T08:17:03 *** gcibot has joined #rtems 2013-12-09T08:17:39 *** gcibot has joined #rtems 2013-12-09T08:55:43 *** antgreen has joined #rtems 2013-12-09T08:59:14 *** gcibot has quit IRC (Remote host closed the connection) 2013-12-09T08:59:35 *** gcibot has joined #rtems 2013-12-09T09:01:03 *** mkhoory-2 has joined #rtems 2013-12-09T09:01:06 *** antgreen has quit IRC (Ping timeout: 252 seconds) 2013-12-09T09:01:28 *** javamonn has quit IRC (*.net *.split) 2013-12-09T09:01:29 *** mkhoory has quit IRC (*.net *.split) 2013-12-09T09:02:11 *** gedare has joined #rtems 2013-12-09T09:03:29 good morning 2013-12-09T09:03:58 *** javamonnn has joined #rtems 2013-12-09T09:07:34 hi gedare! 2013-12-09T09:07:43 peerst: how are you doing 2013-12-09T09:07:50 did you give your talk yet on erlang? 2013-12-09T09:08:14 which one? 2013-12-09T09:08:24 your project came up the other day in a meeting I was having. We were thinking about fault tolerance. sorry, erlang+rtems for space i think it was 2013-12-09T09:08:30 the workshop in germany or the one at codemesh 2013-12-09T09:08:46 yeah last week 2013-12-09T09:08:50 go well? 2013-12-09T09:08:57 very :-) 2013-12-09T09:09:05 awesome 2013-12-09T09:09:11 room full and people seemed to like it 2013-12-09T09:09:22 i think it is nice. different and possibly useful 2013-12-09T09:09:28 people tend to like those traits 2013-12-09T09:09:32 and very good networking on the conference 2013-12-09T09:09:38 sweet 2013-12-09T09:09:59 gedare: when the video is up I'll post a link 2013-12-09T09:10:09 conference was a bit on the strenous side 2013-12-09T09:10:20 from tu to fr 2013-12-09T09:10:41 great i look forward to watching it. 2013-12-09T09:10:57 20 horus of talking to people constantly while either drinking coffee or alcohol (depending on the time of day) 2013-12-09T09:11:02 hah 2013-12-09T09:11:06 and 4 hour sleep then repeat 2013-12-09T09:11:27 well you survived 2013-12-09T09:11:45 yeah … I mean nobody forced me to stay up that long 2013-12-09T09:12:17 btw interesting hardware you might like 2013-12-09T09:12:36 http://www.parallella.org 2013-12-09T09:13:13 one of the adapteva people was giving a talk and was hanging around 2013-12-09T09:13:21 also talked a lot to him 2013-12-09T09:13:42 nice 2013-12-09T09:14:19 gedare: remembers me of my transputer 2013-12-09T09:14:29 occam is already running on it 2013-12-09T09:14:43 but they are looking for ideas how to use something like it 2013-12-09T09:15:04 what is the epiphany processor? 2013-12-09T09:15:16 tile? 2013-12-09T09:15:28 yeah its one tile 2013-12-09T09:15:32 32k local ram 2013-12-09T09:15:44 but can address the ram of every other tile 2013-12-09T09:15:58 ahhh 2013-12-09T09:15:59 the farther away the slower 2013-12-09T09:16:12 interesting, and possibly useful ;-) 2013-12-09T09:16:18 er, different, and possibly useful 2013-12-09T09:16:19 :) 2013-12-09T09:16:22 yeah 2013-12-09T09:16:37 as they claim, a supercomputer-on-chip... 2013-12-09T09:16:38 NUMA 2013-12-09T09:16:40 and on each of the 4 sides high speed io 2013-12-09T09:16:50 with an ARM front-end 2013-12-09T09:17:02 yeah on their kickstarter board 2013-12-09T09:17:04 and usable FPGA fabric it looks like 2013-12-09T09:17:32 but it can run standalone and boot from a flash 2013-12-09T09:17:46 nice. 2013-12-09T09:18:17 yeah one of their io sides is hooked to the fpga on the board so people can implement their own protocols 2013-12-09T09:18:29 ahh 2013-12-09T09:18:38 currently they built a 16 and 64 core version 2013-12-09T09:18:46 1024 are planned 2013-12-09T09:18:52 sure 2013-12-09T09:19:03 and 2020 goal is 1024 cores with 1MB ram each 2013-12-09T09:19:32 thats the point where we erlangers will stop thinking about how to code this and just put a vm in 2013-12-09T09:19:39 i wonder how fundamentally different this is versus cache-coherent GPGPUs 2013-12-09T09:19:44 haha 2013-12-09T09:19:46 quite 2013-12-09T09:20:32 you can do OpenCL stuff 2013-12-09T09:20:32 but you also can do more branchy algorithms 2013-12-09T09:20:32 because: no simd 2013-12-09T09:20:40 ah, sure 2013-12-09T09:20:47 and its also very eneergy efficient 2013-12-09T09:20:48 the pipeline is gp 2013-12-09T09:21:26 aiming at embedded (e.g. replacing Signal Processors) 2013-12-09T09:21:29 it's an interesting platform. I'd be curious to try to implement a HW scheduler/dispatcher in the FPGA and have a no-OS supercomputer 2013-12-09T09:21:35 and supercomputing alike 2013-12-09T09:22:02 gedare: they are taking orders again in february 2013-12-09T09:22:04 with fpga implementations of event handling you can get high throughput, low latency message passing 2013-12-09T09:22:33 i'll have to think about it :) 2013-12-09T09:22:36 or , order one 2013-12-09T09:22:39 and then think about it :) 2013-12-09T09:22:51 most likely it will sit on a shelf with my other unused boards! 2013-12-09T09:23:21 btw: SMP hard realtime schduling is currently underresearched IMHO 2013-12-09T09:23:45 yeah I have this shelf too 2013-12-09T09:24:04 i have done some work in that area 2013-12-09T09:24:27 http://dl.acm.org/citation.cfm?id=2380382. 2013-12-09T09:24:29 http://dl.acm.org/citation.cfm?id=2380382 2013-12-09T09:24:31 no period :) 2013-12-09T09:25:28 i pushed my work into the application domain, but even special-purpose HW for scheduling matters for manycore systems 2013-12-09T09:26:33 i'll be back to it probably in February. I have another project with a deadline early Feb. 2013-12-09T09:26:49 interesting goes on my reading list (have acm lib :) 2013-12-09T09:27:10 ok good otherwise i can send the pdf 2013-12-09T09:27:21 I might be able to get funding to some EU uni for this topic in the future 2013-12-09T09:27:26 or you can google for it. i post all my papers on my website 2013-12-09T09:27:54 hmmm 2013-12-09T09:28:03 i know some folks in glasgow doing something along these lines 2013-12-09T09:28:14 thats waht i did (easier to put on reading list) 2013-12-09T09:28:21 :-O 2013-12-09T09:28:27 they apply this same HWDS concept to functional data structures for haskell 2013-12-09T09:28:31 glasgow would be very convenient 2013-12-09T09:28:39 and implement it with GPUs to simulate the hardware 2013-12-09T09:28:43 sec, i'll find the link 2013-12-09T09:28:56 hmm, I thinking more mundane 2013-12-09T09:29:00 http://www.dcs.gla.ac.uk/~jtod/ 2013-12-09T09:29:05 that's the guy 2013-12-09T09:29:38 but interesting that hes in glasgow because they might be in the project anyway 2013-12-09T09:30:24 heh 2013-12-09T09:30:31 not sure if its down to the application level we need 2013-12-09T09:30:47 http://www.dcs.gla.ac.uk/~jtod/research/esfa/ 2013-12-09T09:30:49 I'm just thinking we need new scheduling methods for smp 2013-12-09T09:30:52 ahhh 2013-12-09T09:30:59 simple prio won't cut it in the future 2013-12-09T09:31:00 ok 2013-12-09T09:31:09 sure 2013-12-09T09:31:16 the guys at York do great work 2013-12-09T09:31:20 with scheduling and with architecture 2013-12-09T09:31:21 and SMP hard realtime scheduling with core affinity is unsolved 2013-12-09T09:31:48 I even think we need new scheduling paradigms 2013-12-09T09:31:53 also there is bjorn brandenburg at max planck, but he has his own path to follow i think. 2013-12-09T09:32:02 something like eariles deadline first for smp 2013-12-09T09:32:09 well, that exists 2013-12-09T09:32:17 oh it does? 2013-12-09T09:32:20 global EDF 2013-12-09T09:32:36 seems I have come to the right person with this topic :-) 2013-12-09T09:32:39 the problem is that contention on the global scheduling data structures kills 2013-12-09T09:32:46 sebastian has read a lot about this. 2013-12-09T09:33:33 yeah I know 2013-12-09T09:33:56 what do you think trigered my thinking .. we are regularly on the phone 2013-12-09T09:34:01 heh 2013-12-09T09:34:20 i also have read a lot about this :) 2013-12-09T09:34:29 yeah I mean something like EDF is a bit higher abstraction than prio based 2013-12-09T09:34:42 didn't mean this direct 2013-12-09T09:35:06 the core affinity is an issue 2013-12-09T09:35:16 that no longer is global scheduling 2013-12-09T09:35:21 a SMP variant of a higher abstraction for hard realtime shceduling which doesnt kill 2013-12-09T09:35:26 yeah 2013-12-09T09:35:41 most of the SMP scheduling work has been done in the soft real-time domain 2013-12-09T09:36:03 because the memory interference problems are hard to bound. 2013-12-09T09:36:04 but e.g. for erlang I want core affinity, becaus eit starts a thread per core I allow it to use 2013-12-09T09:36:24 well I have solved soft realtime inside erlang already 2013-12-09T09:38:17 well it sounds interesting ;) 2013-12-09T09:38:21 too bad you can't pay me to do it 2013-12-09T09:38:40 one easy way out is put erlang on N-1 cores and run anything oldschool HR on core N 2013-12-09T09:38:54 I can if you move to EU countries :-) 2013-12-09T09:39:06 Well lets say i might 2013-12-09T09:39:15 heh 2013-12-09T09:39:34 thats one of the opportinities that came out of my talks in the erlang/rtems area this conference 2013-12-09T09:39:46 gotcha 2013-12-09T09:40:08 so its not sure at the moment 2013-12-09T09:40:22 nothing ever is 2013-12-09T09:40:31 anyway i have enough on my plate :) 2013-12-09T09:40:36 but it might involve eu money, industry and unis and me (sin some not clear role) 2013-12-09T09:40:39 and, i'm not moving to EU 2013-12-09T09:40:51 heh 2013-12-09T09:45:28 gedare: btw slides are alerady up http://codemesh.io/#peer-stritzinger 2013-12-09T09:46:21 you miss my animations and my "I'm sorry dave I'm afraid I can't do this" soundclip 2013-12-09T09:46:24 though 2013-12-09T09:46:36 thanks added to my task list 2013-12-09T09:46:56 i'm sure i'll get the gist of grisp 2013-12-09T09:48:51 *** YuGin has quit IRC (Ping timeout: 260 seconds) 2013-12-09T10:19:11 hey cdesai 2013-12-09T10:19:25 hey gedare 2013-12-09T10:19:50 cdesai: go have a look at the draft document for BSP doxygen http://www.google-melange.com/gci/task/view/google/gci2013/5870434717270016 2013-12-09T10:19:51 Draft rules for BSP Doxygen 2013-12-09T10:19:52 Status: Claimed (3 days, 9 hours left) 2013-12-09T10:19:52 Mentor(s): Vipul Nayyar 2013-12-09T10:22:11 also, i uploaded some more bug fix tasks. 2013-12-09T10:23:05 *** arvind_khadri has quit IRC (Remote host closed the connection) 2013-12-09T10:25:39 I did read his thread on -devel, and that's why I looked at his commits before starting with the doxygen tasks 2013-12-09T10:27:40 yeah. just wanted to keep you informed 2013-12-09T10:27:56 he's trying to solve the issue about shared code 2013-12-09T10:28:00 thanks, great work javamonnn 2013-12-09T10:28:08 gedare: What exactly do you want me to submit for that task? Cindy sounds like she wants me to trim it down to just the structure rules, which I have no problems doings I can keep the rest of it in a .txt somewhere. 2013-12-09T10:28:33 i like what you wrote, as a guide for BSP Doxygen. 2013-12-09T10:28:49 cindy misinterprets this as a specification of an entire task for students i think. 2013-12-09T10:29:03 I got a bit confused regarding shared code (headers) and not having doxygen in C files, but then I read the 'doxygen.h' per BSP comments 2013-12-09T10:29:09 we can braek the tasks for students down into pieces if need be, but the whole documentation is needed. 2013-12-09T10:29:48 javamonnn: i prefer to put the whole thing into the wiki. we can trim and reorg it if needed 2013-12-09T10:30:15 okay cool. I just wasn't sure if I went way beyond what I was supposed to be doing or anything. I'll clean it up and finish it and submit in a few hours. 2013-12-09T10:30:49 the one issue she raised that i think is relevant is that you can link to the existing page on Doxygen_Recommendations 2013-12-09T10:30:54 rather than re-invent the wheel 2013-12-09T10:34:02 Yeah, the existing page of doxygen_recommendations very briefly covers some of the same things, I guess I could trim parts of the doyxgen basics section and link to it 2013-12-09T10:35:26 I forgot that page was around honestly 2013-12-09T10:35:39 I'm still trying to figure out how to do internal links on the wiki 2013-12-09T10:36:00 It's probably super simple but I might be misnaming things when I try it 2013-12-09T10:39:01 [[Page_name | this is a link]] i think 2013-12-09T10:39:02 gedare: thanks for uploading the bug fixing tasks, picked a quick simple one for a start (mainly because I have ss555 built) 2013-12-09T10:55:17 cdesai: pretty easy then, did you verify the warning no longer exists in the build? 2013-12-09T10:55:34 I did 2013-12-09T10:55:38 super 2013-12-09T10:56:05 now, pick a harder one ;) 2013-12-09T10:56:14 5 more days and it would be 7 years from the initial bug report 2013-12-09T10:56:18 sure :D 2013-12-09T10:56:19 haha 2013-12-09T10:56:21 yes 2013-12-09T10:56:24 there is one in there 2013-12-09T10:56:26 that is REALLY old 2013-12-09T10:56:35 number 300 something 2013-12-09T10:57:01 haha I got excited last night when I saw all the new bugfix tasks 2013-12-09T10:57:14 looks like somebody claimed it, not seeing any 3 digit bugs 2013-12-09T10:59:11 gedare: now sure what to do for https://www.rtems.org/bugzilla/show_bug.cgi?id=1323 2013-12-09T11:00:01 cdesai: this looks like a research task 2013-12-09T11:00:16 you can inquire on rtems-users to find someone who uses the MVME5500 2013-12-09T11:00:27 I believe there are a few real users of this board 2013-12-09T11:01:11 and determine if the -msoft-float is supposed to be in the libbsp/powerpc/mvme5500/make/custom/...cfg file 2013-12-09T11:01:35 http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MPC7455 suggests it does have a FPU 2013-12-09T11:03:06 there is unfortunately more to the story than whether or not a target has an FPU 2013-12-09T11:03:28 * cdesai gets it now 2013-12-09T11:03:59 cdesai: http://www.rtems.org/pipermail/rtems-users/2013-December/012230.html 2013-12-09T11:04:08 this is an illuminating post about FPU and the soft-float variant 2013-12-09T11:09:15 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2013-12-09T11:15:01 * cdesai subscribes to -user as well 2013-12-09T11:15:27 *-users it is 2013-12-09T11:16:33 "it's a trap!" 2013-12-09T11:19:43 too late 2013-12-09T11:19:43 off to bed, I'll read up more on this tomorrow 2013-12-09T11:19:47 night everyone 2013-12-09T11:20:04 goodnight 2013-12-09T12:26:22 *** marcinb has quit IRC (Read error: Operation timed out) 2013-12-09T13:29:04 *** marcinb has joined #rtems 2013-12-09T13:33:06 *** antgreen has joined #rtems 2013-12-09T13:43:14 *** bighead123 has joined #rtems 2013-12-09T13:59:14 *** edwardk has joined #rtems 2013-12-09T14:46:35 *** bighead123 has quit IRC (Quit: Leaving.) 2013-12-09T14:50:14 *** javamonnn has quit IRC (Ping timeout: 240 seconds) 2013-12-09T14:50:36 *** javamonnn has joined #rtems 2013-12-09T14:55:23 *** edwardk has quit IRC (Ping timeout: 272 seconds) 2013-12-09T15:23:27 *** antgreen has quit IRC (Ping timeout: 240 seconds) 2013-12-09T15:32:46 *** gedare has quit IRC (Quit: Leaving) 2013-12-09T16:52:21 *** Acespace has joined #rtems 2013-12-09T17:03:17 *** marcinb has quit IRC (Quit: Leaving) 2013-12-09T17:03:25 *** Acespace has quit IRC (Ping timeout: 250 seconds) 2013-12-09T17:05:49 *** edwardk has joined #rtems 2013-12-09T19:14:45 *** peerst has quit IRC (Read error: Connection reset by peer) 2013-12-09T19:15:06 *** peerst has joined #rtems 2013-12-09T19:16:37 *** mkhoory-2 is now known as mkhoory 2013-12-09T19:35:01 *** javamonnn has quit IRC (Ping timeout: 248 seconds) 2013-12-09T19:50:01 *** YuGin has joined #rtems 2013-12-09T19:54:50 *** javamonnn has joined #rtems 2013-12-09T22:37:34 *** javamonnn is now known as javamonn 2013-12-09T23:23:10 *** arvind_khadri has joined #rtems 2013-12-10T01:04:03 *** sebhub has joined #rtems 2013-12-10T02:18:38 *** javamonn has quit IRC (Ping timeout: 240 seconds) 2013-12-10T02:18:59 *** javamonnn has joined #rtems 2013-12-10T02:49:31 *** YuGin has quit IRC (Read error: Connection reset by peer) 2013-12-10T03:03:30 good morning 2013-12-10T03:53:32 *** antgreen has joined #rtems 2013-12-10T06:41:40 the day of truth has come... I have a working bootloader and gdb stub on real moxie hardware. Today I'll see if I can run my first RTEMS app on hardware. 2013-12-10T06:56:25 *** javamonnn has quit IRC (Ping timeout: 272 seconds) 2013-12-10T07:48:44 *** gedare has joined #rtems 2013-12-10T08:00:25 good morning 2013-12-10T08:01:46 Morning 2013-12-10T08:06:30 *** phipse has joined #rtems 2013-12-10T08:06:50 hi 2013-12-10T08:07:03 kiwichris: when you commit a patch written by someone else you can use --author="First Last " to give the proper metadata to git 2013-12-10T08:07:06 hello phipse 2013-12-10T08:09:34 hi gedare are there any news on the status of my patches? 2013-12-10T08:12:20 well it only is 2 business days 2013-12-10T08:12:32 but no one complained so maybe i will commit them soon 2013-12-10T08:13:10 ah business days, didn't think about that ... students ;P 2013-12-10T08:13:30 heh yeah 2013-12-10T08:13:57 do you have more code too? or will this be it 2013-12-10T08:15:14 this should be it for now. I will work on doxygen when this is upstream. 2013-12-10T08:15:20 ok no worries.. 2013-12-10T08:15:38 i can also try to entice a GCI student to do the basic parts for you ;) 2013-12-10T08:20:40 well, ckinit.c console.c irq.c, should be close to the common ones. So if you are in need of tasks, if the package structure is done for these, I would appreciate it :) 2013-12-10T08:32:04 :) 2013-12-10T09:14:53 *** gigetoo has quit IRC (Remote host closed the connection) 2013-12-10T09:16:26 *** gigetoo has joined #rtems 2013-12-10T09:30:59 *** sebhub has quit IRC (Remote host closed the connection) 2013-12-10T09:50:42 *** arvind_khadri has quit IRC (Remote host closed the connection) 2013-12-10T10:02:51 *** phipse has quit IRC (Quit: leaving) 2013-12-10T10:52:03 *** antgreen has quit IRC (Ping timeout: 260 seconds) 2013-12-10T10:59:52 *** edwardk has quit IRC (Quit: Computer has gone to sleep.) 2013-12-10T11:00:42 gedare: Hi 2013-12-10T11:00:47 hey cdesai 2013-12-10T11:00:49 i see your email 2013-12-10T11:00:53 re: bug 1323 2013-12-10T11:01:01 anything else that I could do? 2013-12-10T11:01:34 you should follow-up to clarify that you want to know if this BSP requires -msoft-float or not 2013-12-10T11:01:45 ok 2013-12-10T11:01:55 really.. 2013-12-10T11:02:02 sorry :( 2013-12-10T11:02:03 you want to know if there are users using the git HEAD 2013-12-10T11:02:10 since it does not have msoft-float 2013-12-10T11:02:14 if users are using it without a problem... 2013-12-10T11:02:19 then it seems unnecessary to have it 2013-12-10T11:02:43 however, if the powerpc context switch code does lazy FPU switching, then there could be a bug if the compiler utilizes floating point registers for optimizations 2013-12-10T11:02:48 i do not know if it is the case however. 2013-12-10T11:07:23 cdesai: I wonder if there might be a useful test case that could be made to check for this floating-point issue.. 2013-12-10T11:07:36 i'm thinking out loud for a minute :) 2013-12-10T11:07:58 actually, be right back.. gotta clear some snow, and think through this 2013-12-10T11:08:31 * cdesai nods and re-reads the thread you linked yesterday 2013-12-10T11:23:44 Gedare: I gotta go, cya tomorrow 2013-12-10T12:04:04 *** antgreen has joined #rtems 2013-12-10T12:13:42 *** bighead123 has joined #rtems 2013-12-10T12:40:03 *** antgreen has quit IRC (Ping timeout: 272 seconds) 2013-12-10T13:27:24 *** edwardk has joined #rtems 2013-12-10T13:56:51 *** gedare has quit IRC (Ping timeout: 260 seconds) 2013-12-10T14:11:28 gedare, the PR contains all the related info. I also asked the user and he did not want anything else. 2013-12-10T14:14:23 *** antgreen has joined #rtems 2013-12-10T15:21:40 *** gedare has joined #rtems 2013-12-10T15:37:57 hi kiwichris 2013-12-10T15:38:07 Hi 2013-12-10T15:38:41 do you have any time available to relocate rtems_waf into a separate git repository somewhere either on git.rtems.org/chrisj or your github account would do? 2013-12-10T15:39:09 *** bighead123 has quit IRC (Read error: Connection reset by peer) 2013-12-10T15:39:19 i would like to try out using git submodule for it in the examples-v2. i'm not opposed to committing rtems_waf with examples-v2 as-is... 2013-12-10T15:39:30 but i think having a git submodule example would be nice. 2013-12-10T15:40:04 if you have no time i will just push the version of rtems_waf that I have into examples-v2.git 2013-12-10T15:41:09 Yes I can do this. 2013-12-10T15:41:18 ok thanks just drop me a line when you get around to it 2013-12-10T15:41:30 I will set it up today. It is really great to see this all happenng. 2013-12-10T15:41:53 yes i agree it was a nice little project. i'm trying to think of more little projects like it. 2013-12-10T15:42:04 A big thanks from me for this. I am so impressed with the students in GCI this year. 2013-12-10T15:42:19 yes we have a good group 2013-12-10T15:43:00 my current set of tasks are trying to clean out bugzilla 2013-12-10T15:43:07 Great. 2013-12-10T15:43:18 there's also a student working on a video tutorial how to do the getting started with a VM and with RSB 2013-12-10T15:43:29 Really ? Cool 2013-12-10T15:43:43 yeah we'll see how it turns out. might be a good contribution to RTEMS-TV 2013-12-10T15:43:47 When ready let me know and I will upload to the RTEMS Youtube channel 2013-12-10T15:43:49 or whatever the youtube channel is :) 2013-12-10T15:43:52 will do 2013-12-10T15:44:23 An idea which currently does not have any where to go is the test screen traces. 2013-12-10T15:44:46 these have object id's in them yet these can vary from BSP to BSP. 2013-12-10T15:45:02 the whole test output validation is broken. 2013-12-10T15:45:09 i've known this since i started working with rtems 2013-12-10T15:45:37 We need a way to manage any id but with some checking 2013-12-10T15:46:03 ie it must be a task id 2013-12-10T15:46:05 yes. a test harness for checking the correctness 2013-12-10T15:46:51 this is hard to do without a reference implementation however 2013-12-10T15:46:55 We need all the screens updated to include some metadata 2013-12-10T15:47:10 yes and why I said it has no where to go 2013-12-10T15:47:14 ;) 2013-12-10T16:01:25 later. 2013-12-10T16:01:30 *** gedare has quit IRC (Quit: Leaving) 2013-12-10T16:02:57 *** antgreen has quit IRC (Ping timeout: 252 seconds) 2013-12-10T17:12:19 *** edwardk has quit IRC (Ping timeout: 260 seconds) 2013-12-10T17:41:44 *** antgreen has joined #rtems 2013-12-10T19:39:45 *** javamonn has joined #rtems 2013-12-10T20:18:14 *** javamonn has quit IRC (Ping timeout: 240 seconds) 2013-12-10T20:23:20 *** javamonnn has joined #rtems 2013-12-10T20:36:16 *** bighead123 has joined #rtems 2013-12-10T21:23:02 *** javamonnn has quit IRC (Ping timeout: 240 seconds) 2013-12-10T21:32:32 *** javamonnn has joined #rtems 2013-12-10T21:33:19 *** bighead123 has quit IRC (Quit: Leaving.) 2013-12-10T22:12:40 woo! I'm running my first moxie RTEMS app on hardware. 2013-12-10T22:49:26 *** javamonnn has quit IRC (Ping timeout: 240 seconds) 2013-12-10T23:17:04 *** javamonnn has joined #rtems 2013-12-10T23:21:16 cdesai: edx is offering an embedded systems course that starts right after GCI ends, if you're interested 2013-12-10T23:21:23 cdesai: https://www.edx.org/course/utaustinx/utaustinx-ut-6-01x-embedded-systems-1172 2013-12-10T23:22:38 It looks pretty cool. 2013-12-10T23:39:02 *** javamonnn has quit IRC (Ping timeout: 240 seconds) 2013-12-11T00:02:20 *** javamonnn has joined #rtems 2013-12-11T00:02:47 antgreen, nice 2013-12-11T00:16:14 *** javamonnn has quit IRC (Ping timeout: 240 seconds) 2013-12-11T00:20:31 *** javamonnn has joined #rtems 2013-12-11T00:29:57 *** arvind_khadri has joined #rtems 2013-12-11T01:43:37 *** sebhub has joined #rtems 2013-12-11T03:14:13 sebhub, hi 2013-12-11T06:30:31 *** antgreen has quit IRC (Ping timeout: 246 seconds) 2013-12-11T06:34:49 *** peerst has quit IRC (Ping timeout: 272 seconds) 2013-12-11T06:39:14 hello chris 2013-12-11T07:33:45 *** sebhub has quit IRC (Ping timeout: 245 seconds) 2013-12-11T07:33:57 *** sebhub has joined #rtems 2013-12-11T07:57:54 *** gedare has joined #rtems 2013-12-11T08:00:57 *** edwardk has joined #rtems 2013-12-11T08:56:02 *** sebhub has quit IRC (Ping timeout: 264 seconds) 2013-12-11T09:26:28 *** antgreen has joined #rtems 2013-12-11T09:38:12 *** peerst has joined #rtems 2013-12-11T09:44:35 *** antgreen has quit IRC (Ping timeout: 245 seconds) 2013-12-11T09:55:08 hi, gedare 2013-12-11T09:55:13 hello cdesai 2013-12-11T09:55:24 how's it going? 2013-12-11T09:58:40 good. you? 2013-12-11T09:58:54 decent 2013-12-11T09:59:15 cdesai: I wonder if there might be a useful test case that could be made to check for this floating-point issue.. - can you elaborate more on that? 2013-12-11T09:59:20 ah, yes 2013-12-11T09:59:57 i wonder if it makes sense to write some floating point tasks 2013-12-11T10:00:02 or rather, floating point tests 2013-12-11T10:00:30 that might be useful to detect problems like mismatch between soft and hard FPU, the definition of CPU_HAS_FPU, and such 2013-12-11T10:00:34 this is beyond your task though 2013-12-11T10:01:35 i think you have done enough for your task now, there is not much else to investigate or fix 2013-12-11T10:01:44