2015-03-09T00:07:02 *** idleguy has joined #rtems 2015-03-09T00:17:11 *** idleguy has quit IRC (Read error: Connection reset by peer) 2015-03-09T00:17:30 *** jarie has quit IRC (Quit: nah nah nah nah, nah nah nah nah, hey hey hey, goodbye) 2015-03-09T01:06:46 *** andrei__ has joined #rtems 2015-03-09T01:26:01 *** waleed has joined #rtems 2015-03-09T01:36:01 *** waleed has quit IRC (Ping timeout: 265 seconds) 2015-03-09T02:12:54 *** monstr has joined #rtems 2015-03-09T02:17:36 *** kiwichris has quit IRC () 2015-03-09T02:28:01 *** altairpearl has quit IRC (Remote host closed the connection) 2015-03-09T02:32:09 *** altairpearl has joined #rtems 2015-03-09T02:35:23 *** sebhub has joined #rtems 2015-03-09T02:46:02 *** zoso has joined #rtems 2015-03-09T02:52:25 [trac] #2286/defect (closed) updated by Sebastian Huber (In [changeset:"7b4e81d0c1c8ce23e1686db679630813ac02233e/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2286 2015-03-09T02:52:26 [git] 7b4e81d by Sebastian Huber: [score: Typo Close #2286.] http://devel.rtems.org/changeset/7b4e81d/rtems 2015-03-09T02:57:40 good morning 2015-03-09T03:04:22 sebhub: Hello! 2015-03-09T03:05:08 sebhub : did you get a chance to read the mail I sent on devel (Andrei Popescu ) ? 2015-03-09T03:14:08 *** altairpearl has quit IRC (Read error: Connection reset by peer) 2015-03-09T03:14:43 *** altairpearl has joined #rtems 2015-03-09T03:28:25 *** kiran4399 has joined #rtems 2015-03-09T03:44:32 *** idleguy has joined #rtems 2015-03-09T03:48:59 *** Hesham has joined #rtems 2015-03-09T03:51:38 *** idleguy has quit IRC (Ping timeout: 256 seconds) 2015-03-09T03:55:29 *** idleguy__ has joined #rtems 2015-03-09T04:02:14 *** idleguy__ has quit IRC () 2015-03-09T04:03:04 [trac] #2254/defect (new) updated by heshamelmatary (This is the set of patches [1] from 12-12-2014 until now. I tried to pull and apply ...) http://devel.rtems.org/ticket/2254 2015-03-09T04:22:58 [trac] #2205/defect (new) updated by sebastian.huber (Ok, nice yet another MSYS variant. http://sourceforge.net/projects/msys2/ I ...) http://devel.rtems.org/ticket/2205 2015-03-09T04:34:59 [trac] Developer/Projects/Open/Condition_Variables edited by sebastian.huber (Fix link) http://devel.rtems.org/wiki/Developer/Projects/Open/Condition_Variables 2015-03-09T04:36:12 hello andrei, I fixed the link in the wiki page 2015-03-09T04:41:14 *** guerby has quit IRC (Ping timeout: 252 seconds) 2015-03-09T04:41:37 sebhub: thanks ! I assume you're mentoring the project, right? 2015-03-09T04:42:15 *** guerby has joined #rtems 2015-03-09T04:42:43 *** andrei__ is now known as andrei_ 2015-03-09T04:42:51 *** andrei_ has joined #rtems 2015-03-09T04:42:51 *** andrei_ has joined #rtems 2015-03-09T04:44:27 *** namaste has joined #rtems 2015-03-09T04:45:52 hello got error "cannot open dl.tar for reading" while executing CONSOLE_POLLED=1 ../rtems-src/configure --target=arm-rtems4.11 --enable-rtemsbsp="beagleboneblack beagleboardxm" --enable-tests 2015-03-09T04:46:05 and make 2015-03-09T04:47:22 [trac] #2289/defect (v:4.11) created by sebastian.huber (rtems_ada_self is broken on SMP) http://devel.rtems.org/ticket/2289 2015-03-09T04:47:39 *** Davidbrcz_ has joined #rtems 2015-03-09T04:48:05 *** Davidbrcz has joined #rtems 2015-03-09T04:49:22 *** altairpearl has quit IRC (Ping timeout: 240 seconds) 2015-03-09T04:59:37 can I have solution for that? 2015-03-09T05:00:36 *** altairpearl has joined #rtems 2015-03-09T05:07:11 *** namaste has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-09T05:31:03 sorry andrei, i don't have time this year to be a mentor 2015-03-09T05:47:32 [git] 25ce352 by Sebastian Huber: [score: Fix for GCC version 5 and later] http://devel.rtems.org/changeset/25ce352/rtems 2015-03-09T06:01:11 sebhub: so I should pick a project with a different mentor, or is someone else going to mentor the project? 2015-03-09T06:19:23 *** jamieiles has joined #rtems 2015-03-09T06:45:43 *** Hesham has quit IRC (Ping timeout: 255 seconds) 2015-03-09T06:49:43 yes, you have to find a mentor, the project itself is useful 2015-03-09T06:50:48 sebhub: alright, thanks for letting me know ! 2015-03-09T07:05:48 *** Hesham has joined #rtems 2015-03-09T07:08:57 *** ragunath has joined #rtems 2015-03-09T07:16:22 [git] bf11690 by Sebastian Huber: [network: Avoid clash with FreeBSD ] http://devel.rtems.org/changeset/bf11690/rtems 2015-03-09T07:18:38 *** Hesham has quit IRC (Ping timeout: 246 seconds) 2015-03-09T07:26:41 *** Davidbrcz_ has quit IRC (Ping timeout: 246 seconds) 2015-03-09T07:26:52 *** Davidbrcz has quit IRC (Ping timeout: 240 seconds) 2015-03-09T07:28:36 [trac] Developer/OpenProjects edited by amar (Add a couple of python projects (to be filled in)) http://devel.rtems.org/wiki/Developer/OpenProjects 2015-03-09T07:29:47 [trac] Developer/OpenProjects edited by amar (Move a couple of python-related projects over.) http://devel.rtems.org/wiki/Developer/OpenProjects 2015-03-09T07:30:30 *** Hesham has joined #rtems 2015-03-09T07:31:50 [trac] PageTemplates/ProjectDescriptionTemplate edited by amar (Update replacables.) http://devel.rtems.org/wiki/PageTemplates/ProjectDescriptionTemplate 2015-03-09T07:31:57 what is the expected running time of spintrcritical tests? for ex tests 7,11,13,15? 2015-03-09T07:38:21 [trac] Developer/Projects/ConfigGUI created http://devel.rtems.org/wiki/Developer/Projects/ConfigGUI 2015-03-09T07:39:22 *** Hesham has quit IRC (Ping timeout: 240 seconds) 2015-03-09T07:39:33 [trac] waf/GUI edited by amar (Add link to options list.) http://devel.rtems.org/wiki/waf/GUI 2015-03-09T07:42:15 *** jamieiles has quit IRC (Ping timeout: 250 seconds) 2015-03-09T07:48:59 *** DrJoel_ has quit IRC (Quit: There's nothing dirtier then a giant ball of oil) 2015-03-09T07:53:37 [git] 19799fb by Sebastian Huber: [libtests/stringto01: Fix for GCC 5.0] http://devel.rtems.org/changeset/19799fb/rtems 2015-03-09T08:01:13 ragunath: on what platform? 2015-03-09T08:01:31 arm beagleboard? 2015-03-09T08:01:45 ragunath: most of the spintrcrit* tests can't actually tell if they hit the critical section or not and just run for a fixed period of time 2015-03-09T08:03:50 oh okay .... i am testing those tests and it takes more than 10 min to complete and it was taked as timeout by rtems-test script 2015-03-09T08:04:16 07 is just 06 compiled with a different priority order ifdef. How long does 06 take? 2015-03-09T08:04:27 i tried pc386 also.. it also took more than 10 min to complete 2015-03-09T08:05:07 one min let me check 2015-03-09T08:05:48 I am not sure but it may not be passing. 2015-03-09T08:06:52 6 is also identified as timeout Time: 0:03:04.226089.... but if run then individually the eventually pass and complte 2015-03-09T08:07:30 sebhub: I know the spintrcrit* tests can take a long time.. but any ideas? 2015-03-09T08:08:17 hm, the spintercrit* tests should not run long any more 2015-03-09T08:08:42 on a suitable target, qemu is a problem since in qemu the timing is imprecise 2015-03-09T08:09:24 on real hardware the should run fast and reliable 2015-03-09T08:09:45 in the hardware (BBB) it completed more quickly 2015-03-09T08:11:42 how long? 2015-03-09T08:12:30 [trac] waf/Config edited by amar (Info.) http://devel.rtems.org/wiki/waf/Config 2015-03-09T08:13:05 i havent checked exactly ... i will come up with a comparison for a particular test on qemu and hardware 2015-03-09T08:13:18 bye 2015-03-09T08:13:20 *** DrJoel has quit IRC (Quit: Leaving) 2015-03-09T08:13:25 qemu is a very bad simulator if it comes to timing 2015-03-09T08:14:50 [trac] Developer/Projects/BuildVariables created http://devel.rtems.org/wiki/Developer/Projects/BuildVariables 2015-03-09T08:15:32 [trac] Developer/Projects/BuildVariables edited by amar (Fill in template info.) http://devel.rtems.org/wiki/Developer/Projects/BuildVariables 2015-03-09T08:15:51 [trac] Developer/Projects/ConfigGUI edited by amar (Fix TOC.) http://devel.rtems.org/wiki/Developer/Projects/ConfigGUI 2015-03-09T08:20:56 [git] e5724f7 by Sebastian Huber: [atomic.h: Fix for GCC version 5 and later] http://devel.rtems.org/changeset/e5724f7/rtems-libbsd 2015-03-09T08:25:50 *** andrei__mobile has joined #rtems 2015-03-09T08:38:19 *** gedare has joined #rtems 2015-03-09T08:38:19 *** ChanServ sets mode: +o gedare 2015-03-09T08:46:22 *** Hesham has joined #rtems 2015-03-09T08:47:57 sebhub: spintrcritical6 takes on an average 50 secs on hardware , 7 takes 2 min 20 secs 2015-03-09T08:51:14 *** antgreen has joined #rtems 2015-03-09T08:52:40 *** andrei_ has quit IRC (Ping timeout: 246 seconds) 2015-03-09T09:03:22 *** andrei__mobile has quit IRC (Read error: Connection reset by peer) 2015-03-09T09:03:36 *** andrei__mobile has joined #rtems 2015-03-09T09:12:44 *** zoso has quit IRC (Ping timeout: 245 seconds) 2015-03-09T09:12:45 tmoverhd is outdated with commenting out #include "dumrtems.h" 2015-03-09T09:26:05 [git] 8552df6 by Sebastian Huber: [sysconf: Simplify] http://devel.rtems.org/changeset/8552df6/rtems 2015-03-09T09:26:05 [git] 40ecd11 by Sebastian Huber: [sysconf: Add _SC_NPROCESSORS_(CONF|ONLN)] http://devel.rtems.org/changeset/40ecd11/rtems 2015-03-09T09:28:20 *** Hesham has quit IRC (Ping timeout: 252 seconds) 2015-03-09T09:32:43 *** Hesham has joined #rtems 2015-03-09T09:36:52 *** monstr has quit IRC (Ping timeout: 240 seconds) 2015-03-09T09:44:53 *** sebhub has quit IRC (Quit: Konversation terminated!) 2015-03-09T09:46:17 *** altairpearl has quit IRC (Remote host closed the connection) 2015-03-09T09:47:23 *** andrei__mobile has quit IRC (Ping timeout: 246 seconds) 2015-03-09T09:47:29 *** Hesham has quit IRC (Ping timeout: 250 seconds) 2015-03-09T09:48:46 *** altairpearl has joined #rtems 2015-03-09T09:50:50 *** andreimobile19 has joined #rtems 2015-03-09T10:02:38 *** Hesham has joined #rtems 2015-03-09T10:15:03 *** Hesham has quit IRC (Quit: Leaving.) 2015-03-09T11:23:04 *** dr__house|away is now known as dr__house 2015-03-09T11:25:38 *** jamieiles has joined #rtems 2015-03-09T11:41:52 *** jamieiles has quit IRC (Ping timeout: 240 seconds) 2015-03-09T12:03:46 *** namaste has joined #rtems 2015-03-09T12:06:22 *** andreimobile19 has quit IRC (Read error: Connection reset by peer) 2015-03-09T12:13:14 *** namaste has quit IRC (Quit: Ex-Chat) 2015-03-09T12:24:06 *** namaste_ has joined #rtems 2015-03-09T12:25:40 *** andrei__mobile has joined #rtems 2015-03-09T12:43:20 *** DrJoel has joined #rtems 2015-03-09T12:43:21 *** DrJoel has joined #rtems 2015-03-09T12:43:21 *** ChanServ sets mode: +o DrJoel 2015-03-09T12:45:10 hi DrJoel 2015-03-09T12:46:35 Hello 2015-03-09T12:47:26 A's home sick today, so i'm working at about 1/10 capacity. ;) 2015-03-09T12:47:49 lol... I spent the morning with Michele at the doctor.. just a check up 2015-03-09T12:48:26 *** altairpearl has quit IRC (Read error: Connection reset by peer) 2015-03-09T12:49:51 DrJoel: about the gaisler pci and drvmgr, if we think of them as "Gaisler" products, we can merge them without much fuss. 2015-03-09T12:50:08 If we think of them as "RTEMS" products, we need to ensure the headers are "clean" and the code is consistent to RTEMS 2015-03-09T12:51:13 Yeah.. I realized that this mornign on the drive in. It would be expedient to merge them under libbsp/sparc and try to move them to a more general place or augment shared capabilities to replace them after we branch. 2015-03-09T12:51:21 It is not in anyone's best interests to not merge them though. 2015-03-09T12:51:26 This may be a time to be pragmatic. 2015-03-09T12:51:38 my instinct is to get them into a shape that makes them suitable for inclusion/promotion as RTEMS libraries.. my concern is (1) will danielh want to do it, and (2) will it break applications that rely on them using the RCC currently 2015-03-09T12:52:03 well we don't want to put them in 4.11 if we will change the public interface they export 2015-03-09T12:52:21 That is my instinct as well. (1) only danielh can answer. I think he is willing but there is the work load issues. 2015-03-09T12:52:27 There is also the fact that we want to branch. 2015-03-09T12:52:31 yes. 2015-03-09T12:52:45 I don't think (2) is a factor. This is buried in their BSPs. But should be asked anyway 2015-03-09T12:52:57 my feeling is to let it sit in RCC for 4.11 and work on a long-term, stable API for pci and drvmgr 2015-03-09T12:53:40 Unfortunately, that leaves a lot of leon3/4 and ngmp users unable to use our tree. Lots of missing drivers 2015-03-09T12:53:51 true.. 2015-03-09T12:54:11 i guess the big issue is going to be api compatibility 2015-03-09T12:54:17 That's why I am even willing to discuss the "expedient" step of merging as sparc specific with a plan to migrate up. 2015-03-09T12:54:19 Yep. 2015-03-09T12:54:36 And how much we can do before we branch. This doesn't have a ticket but could be the big blocker if we want it to. 2015-03-09T12:54:42 i.e. let it go into cpukit but ifdef it for sparc? 2015-03-09T12:54:52 That's an option. 2015-03-09T12:55:10 that is the easiest for having a release point without a huge hassle to relocate it later. 2015-03-09T12:55:17 plus it lets verm__ start to work on it with waf 2015-03-09T12:55:27 as opposed to merging it with/after waf changes things 2015-03-09T12:55:50 since it is going to add a couple subdirs in cpukit and a few more headers. not much i guess. 2015-03-09T12:56:48 Email the list and see if that's an acceptable option 2015-03-09T12:57:06 We have two bsps with warnings in bsp specific code left. None really in libcpu 2015-03-09T12:57:17 arm/nds and one of the mbx8xx variants. 2015-03-09T12:57:38 Most appear to be machine integer size of floating point 2015-03-09T12:57:55 crap.. integer size or printf() format 2015-03-09T13:03:43 *** Davidbrcz has joined #rtems 2015-03-09T13:07:53 Gedare.. if you get a chance, please look at the warning post I did and see if you can offer thuoghts on any of the printf() format warnings. So far, it just makes my head hurt with no progress 2015-03-09T13:08:06 Ignore the ones in hexdump and rtl mips specific code. 2015-03-09T13:19:30 *** Hesham has joined #rtems 2015-03-09T13:20:32 *** namaste___ has joined #rtems 2015-03-09T13:22:31 *** namaste_ has quit IRC (Ping timeout: 246 seconds) 2015-03-09T13:24:30 *sigh* my default email address for sending is not a subscriber so I keep getting bounces back when I forget to change it to the right address :p 2015-03-09T13:25:04 You can subscribe that address and not get email for it. Allows it through. 2015-03-09T13:25:14 hmm 2015-03-09T13:25:19 ok will check that option 2015-03-09T13:25:26 i lost the email where you attached the warnings log... 2015-03-09T13:25:30 know the subject? 2015-03-09T13:25:40 nvm i used has:attachment 2015-03-09T13:27:12 *** jamieiles has joined #rtems 2015-03-09T13:32:14 [trac] #2288/defect (new) updated by joel.sherrill (FWIW this looks like the set of tests using assert(). ./fstests/fstime/test.c ...) http://devel.rtems.org/ticket/2288 2015-03-09T13:32:45 do we really want to land a huge patch set like that just before a release? 2015-03-09T13:33:09 4.11 is already too huge, do it in 4.11.1 2015-03-09T13:33:30 release numbers aren't precious we can use as many as we like 2015-03-09T13:33:38 We have coding rules that fatal errors should never be triggered in cpukit/. The ones in cpukit/ should never have been allowed in. 2015-03-09T13:34:11 that makes sense 2015-03-09T13:34:20 i really want to move to a more rapid release cycle 2015-03-09T13:34:34 anytime more than a 'tiny patch' gets in we need to cut a release 2015-03-09T13:35:00 I looked at testsuites first since they are easy and safe to change. cpukit/ should be a bit harder to analyze and change. libbsp is likely too much to do at this point 2015-03-09T13:35:13 the rules for release in that scenerio are quite simple.. if you have to ask then make a release 2015-03-09T13:35:28 *** andrei__mobile has quit IRC (Ping timeout: 252 seconds) 2015-03-09T13:35:50 i got phabricator working last night with buildbot for the first time, so we'll have pre-review building now 2015-03-09T13:39:20 we'll see. it doesn't make sense to have a release cycle much tighter than user development cycles. 2015-03-09T13:40:01 *** kiran4399 has quit IRC (Ping timeout: 246 seconds) 2015-03-09T13:40:22 rtems users pick a version of rtems to use, and run with it for a decade or more. if we release 6 times a year, we'll have users on 30 different versions of rtems by 202. 2015-03-09T13:40:23 2020. 2015-03-09T13:40:40 that may be hyperbole, but it's fairly close to the truth. 2015-03-09T13:41:51 verm__ how often is the doxygen built? since I can't see the doxygen/ directory contents, I don't know when it was built. 2015-03-09T13:41:51 i'd like to see yearly releases, with support given for 2 previous releases = 2 years prior 2015-03-09T13:42:03 and paid support is always available ;) 2015-03-09T13:43:02 verm__: although the patch set is huge, the code has been distributed with RTEMS to users for about 5 years now 2015-03-09T13:43:40 and we don't permit new features into dot releases, only bug fixes 2015-03-09T13:43:53 getting it in the 4.11 release means the bugs can be applied to 4.11.1 2015-03-09T13:43:54 They had gotten more frequent before the move to git and SMP 2015-03-09T13:44:10 somewhat true. but there hasn't been one since i started :0 2015-03-09T13:44:21 We were doing a dot release every few months. 2015-03-09T13:44:40 Yeah.. no excuse for that except waiting for SMP to be stable. 2015-03-09T13:44:46 nd that's not a good one 2015-03-09T13:44:56 i'm happy with dot releases every quarter or biannually 2015-03-09T13:45:14 but major releases don't need to come out so often i don't think. 2015-03-09T13:45:26 DrJoel: it's not crontabbed yet 2015-03-09T13:45:33 still working on it do you want me to update it? 2015-03-09T13:45:57 gedare: that's beside the point, what users do with the code is fine the fact is we haven't done testing of it while it is in the repo 2015-03-09T13:46:12 DrJoel: change %x to %p to print pointers 2015-03-09T13:46:19 *** andrei_ has joined #rtems 2015-03-09T13:46:21 we've stacked so much into 4.11 if it's a big issue then we can release 4.11.1 and land that patch in 4.12 2015-03-09T13:46:50 the waf build will move to a more rapid release cycle it's not a big deal to start that with 4.x 2015-03-09T13:47:12 hmm 2015-03-09T13:47:57 i guess %x should be %lx for long unsigned int 2015-03-09T13:47:59 we can get 4.12 out in less than two weeks and just bump pwhatever we don't want in 4.11.1 2015-03-09T13:49:08 ok. i'm not a huge fan of merging the code anyway. but i want to solicit opinions. 2015-03-09T13:49:50 *** andrei_ has quit IRC (Changing host) 2015-03-09T13:49:50 *** andrei_ has joined #rtems 2015-03-09T13:49:55 Hello! 2015-03-09T13:51:07 hi andrei_ 2015-03-09T13:53:05 gedare: sorry to jump in the middle of a discussion, I'm working on getting "hello world" done in Rtems but I was wondering how should I proceed regarding the GSoC project 2015-03-09T13:53:30 should I wait to see if someone can mentor the one that I m most interested in, look for another project? 2015-03-09T13:54:39 DrJoel: should assert rules be documented in Coding Conventions? (What aRe they?) 2015-03-09T13:55:01 andrei_: what did sebhub say? 2015-03-09T13:55:20 gedare: he said that he doesn't have time to mentor this year 2015-03-09T13:56:26 ok. did he say if he thinks it is a good project? 2015-03-09T13:57:11 he said "the project in itself is useful" 2015-03-09T13:57:20 ok. 2015-03-09T13:57:31 you should proceed with it 2015-03-09T13:58:01 i think any of the core devs can mentor for it 2015-03-09T13:58:02 I don't mind switching to another, it's just that I m interested in putting more effort into the actual project research 2015-03-09T13:58:10 than switching 2015-03-09T13:58:18 you should propose a project that interests you the most first and foremost 2015-03-09T13:58:30 and do a good job convincing us that you can do it, and that all you need is a little guidance. :0 2015-03-09T13:59:06 sebhub will be helpful in reviewing your code and getting it merged, he just has no time to train you and work with you 2015-03-09T13:59:13 for that, we can find you a mentor. 2015-03-09T14:00:00 *** jarie has joined #rtems 2015-03-09T14:00:36 I think it is a pretty easy-to`-define project 2015-03-09T14:00:41 and that makes it a good0 project to work on 2015-03-09T14:00:56 my daughter is "helping" me work right now 2015-03-09T14:01:04 she wants to type 0s 2015-03-09T14:01:25 haha 2015-03-09T14:02:42 you should proceed with CV project. I have a good idea who could mentor 2015-03-09T14:03:14 for now you can ask questions here and on devel ml whoever answers answers 2015-03-09T14:03:44 gedare: I have some performance analysis for Context Switch for Epiphany code executing in DRAM and/or core's Local memory. What else may be useful to get timing analysis for? SMP locks? 2015-03-09T14:04:31 interrupt processing 2015-03-09T14:04:43 gedare: thanks! awesome that I get to "keep" it 2015-03-09T14:05:06 gedare: Including Clock_isr? 2015-03-09T14:05:34 Hesham: try to measure a "null' interrupt handler to see what the isr handling overhead is 2015-03-09T14:06:03 you might also see if any local resources are avialable for WCET analysis 2015-03-09T14:06:28 The trade-offs are a lot given that I can place the code and/or data in both local and external memory 2015-03-09T14:07:01 interrupt processing time is very important to WCET analyses 2015-03-09T14:07:12 I'll generate some analysis for interrupts yes 2015-03-09T14:07:14 because you have to account for their overhead in every task that might be interrupted 2015-03-09T14:07:23 I think there is a test case for it 2015-03-09T14:07:37 you may be right. 2015-03-09T14:07:52 those 2 are most important to wcet: ctx switch and isr handling 2015-03-09T14:08:42 lock performance matters too, but i don't think RTEMS has the state-of-the-art that would make measuring them interesting 2015-03-09T14:08:52 maybe after fine-grained locking is implemented, which sebhub is working on 2015-03-09T14:09:40 you could do some work on thread scalability though, and measure some performance-critical parts of the kernel like scheduling, that might be impacted by # of threads 2015-03-09T14:10:08 the deterministic scheduler shouldn't be impacted greatly, but if you increase # threads > 64 it might start to suffer especially if multiple wind up in the same priority level 2015-03-09T14:10:24 and if you go beyond 255 threads, you will definitely have collisions in priority leve3l 2015-03-09T14:10:34 for that you may need to create "random task sets" 2015-03-09T14:11:38 Which parts exactly should I get execution time for? 2015-03-09T14:11:57 I mean in this thread scalability test 2015-03-09T14:13:31 Scheduler-related functions? 2015-03-09T14:13:47 well, if you measure the clock isr that will be sufficient 2015-03-09T14:13:49 gedare Yes they should be in coding conventions. No they probably have never been captured. 2015-03-09T14:13:53 because that includes the call to scheduler 2015-03-09T14:14:23 I see, thanks gedare 2015-03-09T14:14:25 DrJoel: i think there is a line about using assert, so it should probably be clarified. you may like to draft a text to send to devel for appropval 2015-03-09T14:14:39 Hesham: sure. clock_tick() or whatever it is called will invoke the scheduler. 2015-03-09T14:15:00 Hesham: tm26 and tm27 should be of interest. They cover some of the times you are mentioning 2015-03-09T14:15:24 that captures any pre-emption overheads. it won't capture all scheduler overhead such as self-suspensions or initial thread creation scheduling 2015-03-09T14:15:25 DrJoel: Great this will save me a lot of time, thanks 2015-03-09T14:16:08 DrJoel: have you landed all your doxygen changes? 2015-03-09T14:16:29 Hesham: the tmtests are intended to be useful. If they don't capture the cases you are interested in or need adjusting, please make suggestions 2015-03-09T14:16:47 *** jamieiles has quit IRC (Ping timeout: 265 seconds) 2015-03-09T14:17:05 verm__: Yes. But I couldn't figure out how to fix the worst ones about group structure I just posted to the list about. Whoever worked on that file will have to look at them. 2015-03-09T14:18:06 ABout the first 10 lines of the warnings 2015-03-09T14:18:06 FWIW the deterministic priority scheduler should show no difference in performance across any mix of threads 2015-03-09T14:18:06 DrJoel: I tried to use tmoverhd, but it's outdated and most functions fail to compile 2015-03-09T14:18:06 *UNLESS* the bit scan is implemented in software 2015-03-09T14:18:07 DrJoel: should i rebuild doxygen then? 2015-03-09T14:18:07 I built it for every BSP over the weekend. 2015-03-09T14:18:11 can you give me a list of directories to ignore before that? 2015-03-09T14:18:12 verm__: Yes I suppose so. 2015-03-09T14:18:25 Not today .. about to disappear into meetings for the rest of the day. 2015-03-09T14:18:27 i'd like to make the doxygen 'warns clean' 2015-03-09T14:18:33 DrJoel: tmcontext uses cache function and Epiphany doesn't have caches, so I was forced to write it my own. 2015-03-09T14:19:00 DrJoel: any hints? 2015-03-09T14:19:02 third party code for sure. I can rattle off a few. md5, sha, libnetworking, mghttpd, libz 2015-03-09T14:19:08 ok 2015-03-09T14:19:15 i'll mark those at least then 2015-03-09T14:19:20 rpc/xdr, jffs2 2015-03-09T14:19:30 gedare: any other obivous third party code? 2015-03-09T14:19:47 DrJoel: tha'ts not true as the # threads grows espec ially if there are collisions inp riority 2015-03-09T14:20:10 gedare: Each priority is a FIFO. How does append take longer? 2015-03-09T14:20:20 hmm. 2015-03-09T14:20:38 rbtree might have issues based on priority mix but probably not collisions. 2015-03-09T14:21:08 might have an issue if the bitscan code is in software and early outs. But that shouldn't be more than a few instructions variance 2015-03-09T14:22:29 i'm going to take A to CostCo I think. 2015-03-09T14:22:34 DrJoel: we were discussing abt spintrcritical tests....spintrcritical06 on an average took 50 secs and 7 took 140 secs on beaglebone black... i guess thats ok 2015-03-09T14:22:36 can't think about this now :0 2015-03-09T14:23:15 so everything libnetworking/* should be excluded? 2015-03-09T14:23:23 verm__: Yes 2015-03-09T14:23:47 ragunath: that's OK I think. One is forward priorities, the other is backwards 2015-03-09T14:24:24 libcrypt if imported from third party 2015-03-09T14:24:36 pppd 2015-03-09T14:24:47 That should cover most of it. Let's see what that gets us down to 2015-03-09T14:24:59 going away from the keyboard 2015-03-09T14:25:06 ok 2015-03-09T14:25:07 thanks 2015-03-09T14:25:20 i'll edit the doxyfile and exclude them 2015-03-09T14:25:22 it won't fix all issues. :) 2015-03-09T14:25:26 i don't know why this wasn't done earlier 2015-03-09T14:25:31 well you can exclude code fragments, too 2015-03-09T14:25:36 we'll get them all 2015-03-09T14:26:08 It will be a huge effort to fix all the undocumented items. I think everything else should be fixed as soon as possible though. 2015-03-09T14:26:37 I fixed about 1/2 of those types of issues, bad command names, arguments not matching doxygen, undocumented single argument, etc. 2015-03-09T14:26:44 But an undocumented method is one I am not touching. 2015-03-09T14:27:03 The rtl code has something structurally wrong with the doxygen that generates a lot of warnings I **THINK** 2015-03-09T14:27:06 now leaving 2015-03-09T14:27:13 ah, hmm 2015-03-09T14:27:14 ok cya 2015-03-09T14:27:26 please comment on the migration plan when you have a chance 2015-03-09T14:27:33 i really want to start opening milestones/tickets/creating wiki pages 2015-03-09T14:27:35 *** gedare has quit IRC (Ping timeout: 256 seconds) 2015-03-09T14:29:22 *** jarie has quit IRC (Ping timeout: 252 seconds) 2015-03-09T14:37:20 *** Hesham has quit IRC (Ping timeout: 264 seconds) 2015-03-09T14:39:04 whoa 2015-03-09T14:39:13 there are a lot of files here that need to be excluded 2015-03-09T14:39:13 sheet 2015-03-09T14:39:17 er sheesh 2015-03-09T14:43:52 *** Hesham has joined #rtems 2015-03-09T14:48:25 *** Hesham has quit IRC (Ping timeout: 264 seconds) 2015-03-09T14:54:24 I'm building source builder and I followed this https://ftp.rtems.org/pub/rtems/people/chrisj/source-builder/source-builder.html 2015-03-09T14:54:31 and it worked on until the actual build cmd 2015-03-09T14:54:50 this is how I call it 2015-03-09T14:54:51 rtems-source-builder/rtems $ ../source-builder/sb-set-builder --log=l-sparc.txt 1 --prefix=/home/andrei/development/rtems/4.11 2 4.11/rtems-sparc 2015-03-09T14:55:18 and this is what I get 2015-03-09T14:55:19 RTEMS Source Builder - Set Builder, v0.5.0 2015-03-09T14:55:19 Build Set: 1 2015-03-09T14:55:19 error: no build set file found: 1.bset 2015-03-09T14:56:42 home/andrei/development/rtems exists as a path and I have rw over it 2015-03-09T14:57:44 walking by.. andrei_ you have a typo.. there is a 1 between the --log and the --prefix arguments 2015-03-09T14:59:30 DrJoel: got about 3,000 of the warnings gone 2015-03-09T15:00:31 I fugred that would do a lot of them. Is this run online? 2015-03-09T15:00:34 DrJoel: there was also a 2, thanks a lot, didn't see them 2015-03-09T15:00:37 works now 2015-03-09T15:00:41 2,803 so far 2015-03-09T15:00:54 DrJoel: i'll put the warnings file online 2015-03-09T15:01:00 unfortunatly it's crashing again so no full build 2015-03-09T15:01:05 andrei_ you cut and pasted from the web page and cut the little highlight numbers :) 2015-03-09T15:01:30 File a ticket to ask chrisj to add a warning about cutting and pasting below the example. :) 2015-03-09T15:01:30 https://docs.rtems.org/doxygen/warnings2.log 2015-03-09T15:01:36 https://docs.rtems.org/doxygen/exclude 2015-03-09T15:01:41 DrJoel: idd, but I did that in a pastebin, specifically to get rid of those 2015-03-09T15:03:20 DrJoel: https://docs.rtems.org/doxygen/count 2015-03-09T15:03:23 count of warnings per file 2015-03-09T15:03:49 that's the file to go through to add to the exclude the list.. i got most of the obvious ones 2015-03-09T15:04:50 the file list confirms my review of the warnings2.log. We own those. :( 2015-03-09T15:05:08 so i got all the 3rd party ones? 2015-03-09T15:05:36 Many of the files predate the existence of Doxygen so we haven't caught up on them yet. Some of those are just missing items in files we have touched. 2015-03-09T15:05:45 I don't see any third party files in that list 2015-03-09T15:07:29 i'm confused 2015-03-09T15:07:36 oh, no i meant in https://docs.rtems.org/doxygen/exclude 2015-03-09T15:07:39 these are all 3rd party? 2015-03-09T15:13:54 *** Hesham has joined #rtems 2015-03-09T15:14:42 I would bring back in rtems/* and ask Gedare and Chris for a double check 2015-03-09T15:19:44 *** jarie_ has joined #rtems 2015-03-09T15:26:16 *** jamieiles has joined #rtems 2015-03-09T15:27:19 *** jarie_ has quit IRC (Ping timeout: 245 seconds) 2015-03-09T15:28:15 DrJoel: you mean we own those files now? 2015-03-09T15:28:39 I mean we always owned them. and any issues in them. So to clear the list is going to take work. 2015-03-09T15:29:14 how much should the build take, it's on building: sparc-rtems4.11-gcc-4.9.2-newlib-b75c57950318ed657807f87a4c694e7b8d68fcae-x86_64-linux-gnu-1 2015-03-09T15:29:16 for quite a while 2015-03-09T15:29:24 Some of the files will be pretty easy to clean up. confdefs.h already has a lot of Doxygen and I noticed that the scheduler area just doesn't have the Doxygen markers on them. The commetns aren't great but they exist 2015-03-09T15:29:57 Depending on the machine, say an hour tops. I think I saw 70 minutes on one toolset this past week 2015-03-09T15:30:15 ah, cool, so it's not stuck 2015-03-09T15:30:25 thanks 2015-03-09T15:31:50 There is a file somewhere you could tail in another window. 2015-03-09T15:31:57 You could also do a top in another window 2015-03-09T15:33:20 top does the trick, thanks 2015-03-09T15:42:02 *** jarie_ has joined #rtems 2015-03-09T15:42:20 *** jamieiles has quit IRC (Ping timeout: 246 seconds) 2015-03-09T15:58:50 *** Hesham has quit IRC (Ping timeout: 252 seconds) 2015-03-09T16:07:45 *** jarie_ has quit IRC (Ping timeout: 244 seconds) 2015-03-09T16:12:47 https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=b0bc65729070b9cbdbb53ff042984a3c545a0e34 2015-03-09T16:14:39 [git] 3159267 by Joel Sherrill: [shared/include/grcan.h: Add prototypes to fix warnings] http://devel.rtems.org/changeset/3159267/rtems 2015-03-09T16:14:39 [git] a1148f3 by Joel Sherrill: [powerpc/shared/bootloader/pci.c: Remove warnings] http://devel.rtems.org/changeset/a1148f3/rtems 2015-03-09T16:14:39 [git] 52c56af by Joel Sherrill: [gensh4/include/bsp.h: Remove unused prototype and include] http://devel.rtems.org/changeset/52c56af/rtems 2015-03-09T16:14:39 [git] 867a54a by Joel Sherrill: [gensh1/include/bsp.h: Add needed prototype] http://devel.rtems.org/changeset/867a54a/rtems 2015-03-09T16:14:39 [git] 2a11a46 by Joel Sherrill: [or1k: Correct _CPU_Thread_Idle_body prototype] http://devel.rtems.org/changeset/2a11a46/rtems 2015-03-09T16:14:40 [git] 4f4e7bb by Joel Sherrill: [cpukit/score/cpu/sh/context.c: Correct name of _CPU_Context_switch] http://devel.rtems.org/changeset/4f4e7bb/rtems 2015-03-09T16:14:40 [git] 41a0074 by Joel Sherrill: [tools/cpu/sh/sci.c: Add include of bsp.h to generated file] http://devel.rtems.org/changeset/41a0074/rtems 2015-03-09T16:14:41 [git] ef598c3 by Joel Sherrill: [or1k/rtems/score/cpu.h: Add no_return attribute to _CPU_Context_restore] http://devel.rtems.org/changeset/ef598c3/rtems 2015-03-09T16:14:41 [git] 2468aa3 by Joel Sherrill: [vprintk.c: Reorder switch and add default to eliminate warning] http://devel.rtems.org/changeset/2468aa3/rtems 2015-03-09T16:14:42 [git] 500eb04 by Joel Sherrill: [mvme136.cfg: Remove -DNDEBUG flag] http://devel.rtems.org/changeset/500eb04/rtems 2015-03-09T16:14:42 [git] 058a92d by Joel Sherrill: [lpc1768_mbed.cfg: Remove -DNDEBUG flag] http://devel.rtems.org/changeset/058a92d/rtems 2015-03-09T16:14:43 [git] 120af80 by Joel Sherrill: [rtl22xx.cfg: Remove -DNDEBUG flag] http://devel.rtems.org/changeset/120af80/rtems 2015-03-09T16:14:43 [git] 9cc1892 by Joel Sherrill: [libcpu/powerpc/mpc8260/console-generic/console-generic.c: Include bsp.h to fix warning] http://devel.rtems.org/changeset/9cc1892/rtems 2015-03-09T16:14:44 [git] d9cda35 by Joel Sherrill: [cpukit/libdl/rtl-mdreloc-m32r.c: Add parentheses per Chris' suggestion] http://devel.rtems.org/changeset/d9cda35/rtems 2015-03-09T16:15:53 DrJoel: beaglebone is missig rtc support? i have made some changes to make it working ... do we have any coding guidelines sort of thing to follow before sending the patch 2015-03-09T16:16:51 There are some on the wiki.. devel.rtems.org and search for style 2015-03-09T16:17:53 https://devel.rtems.org/wiki/Developer/Coding/Conventions is the direct link. The top page also has some git guidelines as well. 2015-03-09T16:17:57 got it ... thanks 2015-03-09T16:18:24 For an RTC, there is some framework code. Check out any BSP with "tod" or "rtc" support 2015-03-09T16:18:27 *** jarie has joined #rtems 2015-03-09T16:18:55 *** namaste___ has quit IRC (Ping timeout: 246 seconds) 2015-03-09T16:19:59 *** jamieiles has joined #rtems 2015-03-09T16:24:09 DrJoel: I got the basic driver working... will add this to the framework 2015-03-09T16:25:05 DrJoel: not sure how linus is going to follow that 2015-03-09T16:27:05 Naming it code of conflict is a great indicator of how they expect people to behave. Code of Conduct at least starts on a positive note. 2015-03-09T16:28:14 ragunath: very nice! 2015-03-09T16:28:32 yeah.. well i don't know how much you peruse the kernel lists there it isn't very nice 2015-03-09T16:28:40 *** kiwichris has joined #rtems 2015-03-09T16:29:06 I didn't want to say anything will now.. the kernel list is a fine example of how people shouldn't be treated in a free/open source community. 2015-03-09T16:29:22 kiwichris: Good morning 2015-03-09T16:29:36 yep i agree 2015-03-09T16:29:42 DrJoel, hi 2015-03-09T16:32:24 DrJoel: so i should re-add rtems/jffs2.h rtems/libi2c.h rtems/motorola/mc68681.h rtems/pci.h? 2015-03-09T16:32:28 verm__, kiwichris: I don't see this Doxygen warning in the code: 2015-03-09T16:32:33 score/cpu/m68k/rtems/m68k/qsm.h 2015-03-09T16:32:34 score/cpu/m68k/rtems/m68k/sim.h 2015-03-09T16:32:34 score/cpu/m68k/rtems/m68k/m68302.h 2015-03-09T16:32:34 score/cpu/m68k/rtems/m68k/m68360.h 2015-03-09T16:32:45 Ignore those m68k lines 2015-03-09T16:32:50 /data/docs/build/b-doc/sparc-rtems4.11/leon3/lib/include/rtems/fsmount.h:233: warning: More #endif's than #if's found. 2015-03-09T16:33:07 verm__: yeah.. 2015-03-09T16:33:28 ok 2015-03-09T16:33:33 verm__, what am I looking at ? 2015-03-09T16:33:35 those files had at least a thousand between them i think 2015-03-09T16:33:55 I am dispersing the include/motorola and include/zilog directories though so they soon won't be an issue 2015-03-09T16:34:11 kiwichris: i'm trying to make an exclusion list for doxygen so we can be 'warns clean'. i want to link it to buildbot so doxygen errors make reviews / commits fail 2015-03-09T16:34:12 One file wasn't used at all. Most were used by a single BSP. Some by two 2015-03-09T16:34:24 new log: 2015-03-09T16:34:29 exclusion list: https://docs.rtems.org/doxygen/exclude 2015-03-09T16:34:34 verm__, sue 2015-03-09T16:34:36 sure 2015-03-09T16:34:42 count of errors: https://docs.rtems.org/doxygen/count 2015-03-09T16:34:52 joel said to re-add rtems/* so i've done that 2015-03-09T16:35:06 I don't see the issue in fsmount.h if someone can spot it, that would knock one off the list 2015-03-09T16:35:21 and /data/docs/build/b-doc/sparc-rtems4.11/leon3/lib/include/dev/i2c/i2c.h:431: warning: end of file while inside a group 2015-03-09T16:36:14 isn't a good one either. 2015-03-09T16:36:25 Fixing those MIGHT clean up some others 2015-03-09T16:36:41 rebuilding with the new exclude list one sec 2015-03-09T16:37:10 The grouping, ifdef, bad command issues sometimes take care of issues 2015-03-09T16:37:11 *** ragunath has left #rtems 2015-03-09T16:38:35 https://docs.rtems.org/doxygen/warnings3.log 2015-03-09T16:38:48 https://docs.rtems.org/doxygen/exclude (refresh) 2015-03-09T16:38:56 https://docs.rtems.org/doxygen/count (refresh) 2015-03-09T16:39:43 DrJoel: adding those 5 files back brought 2,133 warnings in :( 2015-03-09T16:41:40 [git] 5bc6237 by Joel Sherrill: [Move libcsupport/include/motorola/*.h to m68k/idp BSP These header files were only used by one BSP and they are hardware dependent. The hardware dependency always made them bad candidates for where they were in the tree. But this fixes that.] http://devel.rtems.org/changeset/5bc6237/rtems 2015-03-09T16:41:40 [git] 6796d9c by Joel Sherrill: [Remove unused cpukit/libcsupport/include/zilog/z8536.h] http://devel.rtems.org/changeset/6796d9c/rtems 2015-03-09T16:41:41 [git] 1aa7b8b by Joel Sherrill: [Move contents of libcsupport/include/zilog into libbsp/shared Only a few BSPs use this and it should not have been in libcsupport.] http://devel.rtems.org/changeset/1aa7b8b/rtems 2015-03-09T16:41:47 DrJoel: cat rtems/cpukit/dev/include/dev/i2c/i2c.h|egrep '@}|@{' 2015-03-09T16:41:49 unmatched 2015-03-09T16:42:10 run that on your local machine 2015-03-09T16:42:44 I just pushed a patch which moves some of those files closer to the few places they are used. They are HW dependent and I have hated their location for years. 2015-03-09T16:43:05 You just gave me the excuse earlier to move or remove them 2015-03-09T16:44:13 Clearly mismatched but I don't know what the right grouping is: :( 2015-03-09T16:44:14 * @{ 2015-03-09T16:44:15 * @{ 2015-03-09T16:44:15 * @{ 2015-03-09T16:44:15 * @{ 2015-03-09T16:44:31 well that sucked .. .. 4 opens, 3 closes 2015-03-09T16:45:52 yeah i was unsure what went where, too 2015-03-09T16:45:56 that's why i didn't fix it 2015-03-09T16:46:00 [git] 015211b by Joel Sherrill: [cpukit/dev/include/dev/i2c/i2c.h: Fix Doxygen grouping] http://devel.rtems.org/changeset/015211b/rtems 2015-03-09T16:46:38 Screw it. I took a guess and pushed it. The last open needed a close 2015-03-09T16:46:50 It seemed the most likely. :) 2015-03-09T16:47:43 One question is whether the BSP code should be considered grounds for failure. 2015-03-09T16:47:48 A lot of these includes are in the BSP 2015-03-09T16:48:12 I am not saying they should be excluded at the Doxygen level -- just not grounds for failure 2015-03-09T16:48:44 *** jarie has quit IRC (Ping timeout: 264 seconds) 2015-03-09T16:49:17 while compiling, it crashed with checking for python2.7... no 2015-03-09T16:49:17 configure: error: python is missing or unusable 2015-03-09T16:49:23 but if I do ll 2015-03-09T16:49:31 $ ll /usr/bin/python 2015-03-09T16:49:31 lrwxrwxrwx 1 root root 9 nov 9 00:57 /usr/bin/python -> python2.7 2015-03-09T16:49:33 all seems ok 2015-03-09T16:49:49 $ python --version Python 2.7.6 2015-03-09T16:49:59 DrJoel: we either exclude ir or include it 2015-03-09T16:50:05 *** jarie has joined #rtems 2015-03-09T16:50:15 i really like things to be clean that way there is no discussion about whether it stays or not 2015-03-09T16:50:32 The DanielH needs to know he has a lot of Doxygen issues :) 2015-03-09T16:51:50 At this point, we can't do much better than "not more than before" :) 2015-03-09T16:52:20 *** jamieiles has quit IRC (Ping timeout: 246 seconds) 2015-03-09T16:56:35 [git] 34e0a56 by Joel Sherrill: [cpukit/libmisc/fsmount/fsmount.h: Remove duplicate comments and copyright] http://devel.rtems.org/changeset/34e0a56/rtems 2015-03-09T16:57:26 still don't know why Doxygen thinks there are "More #endif's than #if's found." but spotted some issues and fixed them 2015-03-09T17:00:22 re-running 2015-03-09T17:00:29 *** jamieiles has joined #rtems 2015-03-09T17:00:42 I would love to know what's wrong with fsmount.h 2015-03-09T17:04:34 [trac] #2290/defect (v:4.10) created by chrisj (Move to gdb-7.9) http://devel.rtems.org/ticket/2290 2015-03-09T17:05:40 [trac] #2290/defect (new) updated by joel.sherrill (empty comment) http://devel.rtems.org/ticket/2290 2015-03-09T17:06:09 DrJoel: https://docs.rtems.org/doxygen/warnings4.log 2015-03-09T17:06:43 well the fsmount.h problem is gone. :) 2015-03-09T17:07:20 *** jamieiles has quit IRC (Ping timeout: 264 seconds) 2015-03-09T17:07:21 is the channel logged? 2015-03-09T17:07:22 Did you drop all the BSP include files? 2015-03-09T17:08:17 andrei_: yes but the logs are not currently public at the moment i'll be fixing it soon 2015-03-09T17:08:20 DrJoel: did i? 2015-03-09T17:08:26 https://docs.rtems.org/doxygen/exclude 2015-03-09T17:08:32 that's the exclude list i'm using now 2015-03-09T17:09:36 warnings3 starts with b1553brm which isn't in warnings4? But you didn't exclude that or any amba files 2015-03-09T17:09:43 OK.. time for me to leave and pick up kids. 2015-03-09T17:09:44 bye 2015-03-09T17:09:45 *** DrJoel has quit IRC (Quit: If you can't laugh at yourself, make fun of other people.) 2015-03-09T17:11:36 *** kiwichris has quit IRC (Read error: Connection reset by peer) 2015-03-09T17:11:49 *** kiwichris has joined #rtems 2015-03-09T17:23:56 *** andrei_ has quit IRC (Remote host closed the connection) 2015-03-09T17:51:45 *** jamieiles has joined #rtems 2015-03-09T17:53:14 *** jarie has quit IRC (Ping timeout: 246 seconds) 2015-03-09T18:01:29 *** jarie has joined #rtems 2015-03-09T18:12:59 *** ita has quit IRC (Remote host closed the connection) 2015-03-09T18:16:38 *** jamieiles has quit IRC (Quit: WeeChat 1.0.1) 2015-03-09T18:20:03 *** ita has joined #rtems 2015-03-09T18:30:23 *** gedare has joined #rtems 2015-03-09T18:30:23 *** ChanServ sets mode: +o gedare 2015-03-09T18:46:24 *** Davidbrcz has quit IRC (Ping timeout: 256 seconds) 2015-03-09T18:55:08 *** jarie has quit IRC (Quit: ":)") 2015-03-09T19:02:54 *** gedare has quit IRC (Ping timeout: 244 seconds) 2015-03-09T20:19:46 [trac] #2254/defect (new) updated by chrisj (Thanks for the update. It looks like a snapshot is our best path forward. Can you ...) http://devel.rtems.org/ticket/2254 2015-03-09T20:24:33 [trac] #2205/defect (new) updated by chrisj (I am moving to MSYS2 because the package management is more in line with what I ...) http://devel.rtems.org/ticket/2205 2015-03-09T20:44:48 *** namaste has joined #rtems 2015-03-09T20:56:31 beng-nl: while doing compilation for BBB BSP I got error " cannot open dl.tar for reading" 2015-03-09T20:56:49 Following http://www.shrike-systems.com/beagleboard-xm-beaglebone-black-and-everything-else-rtems-on-the-beagles.html 2015-03-09T21:12:46 [trac] #2291/defect (v:4.10) created by amar (Mention RSB in Getting Started Doc.) http://devel.rtems.org/ticket/2291 2015-03-09T21:14:28 [trac] Developer/Release edited by amar (Add some notes taken from 4.11_Release_Notes) http://devel.rtems.org/wiki/Developer/Release 2015-03-09T23:12:39 *** dr__house is now known as dr__house|away 2015-03-10T00:00:49 *** jarie has joined #rtems 2015-03-10T00:14:45 *** altairpearl_ has joined #rtems 2015-03-10T00:22:58 *** namaste has quit IRC (Ping timeout: 246 seconds) 2015-03-10T00:48:30 *** zoso has joined #rtems 2015-03-10T00:59:35 *** monstr has joined #rtems 2015-03-10T01:33:00 *** namaste has joined #rtems 2015-03-10T01:33:30 beng-nl: there? 2015-03-10T01:43:35 *** kiwichris has quit IRC () 2015-03-10T01:44:43 *** jarie has quit IRC (Quit: nah nah nah nah, nah nah nah nah, hey hey hey, goodbye) 2015-03-10T02:45:09 *** zoso has quit IRC (Ping timeout: 264 seconds) 2015-03-10T02:52:00 *** namaste has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-10T03:02:28 *** zoso has joined #rtems 2015-03-10T03:11:19 [git] f0ff18e by Sebastian Huber: [score: Add rtems_set_errno_and_return_value() Remove rtems_set_errno_and_return_minus_one_cast().] http://devel.rtems.org/changeset/f0ff18e/rtems 2015-03-10T03:11:19 [git] 60bded9 by Sebastian Huber: [posix: Use a value of 0 for SEM_FAILED This is the standard NULL pointer.] http://devel.rtems.org/changeset/60bded9/rtems 2015-03-10T03:11:19 [git] 77e682e by Sebastian Huber: [posix: Change sem_t to the 32-bit object type This change is also valid for 16-bit object type architectures since in this case POSIX_Semaphore_Control::Semaphore_id is used as a proxy.] http://devel.rtems.org/changeset/77e682e/rtems 2015-03-10T03:11:19 [git] 1f67156 by Sebastian Huber: [libcsupport: Delete superfluous _gettimeofday()] http://devel.rtems.org/changeset/1f67156/rtems 2015-03-10T03:11:19 [git] f244bfc by Sebastian Huber: [posix: Install only if not provided] http://devel.rtems.org/changeset/f244bfc/rtems 2015-03-10T03:25:47 *** zoso has quit IRC (Ping timeout: 246 seconds) 2015-03-10T03:32:48 *** zoso has joined #rtems 2015-03-10T03:40:30 *** kiwichris has joined #rtems 2015-03-10T03:49:37 *** Hesham has joined #rtems 2015-03-10T03:54:23 *** namaste has joined #rtems 2015-03-10T03:55:20 *** sebhub has joined #rtems 2015-03-10T03:56:37 good morning 2015-03-10T03:59:59 *** andrei__ has joined #rtems 2015-03-10T04:01:31 [trac] #2292/defect (v:4.10) created by dcb (msdos_misc.c:1059: array index used before sanity check ?) http://devel.rtems.org/ticket/2292 2015-03-10T04:29:33 [trac] #2205/defect (new) updated by sebastian.huber (The one from python.org, downloaded at the ticket creation date.) http://devel.rtems.org/ticket/2205 2015-03-10T04:52:21 *** Hesham has quit IRC (Quit: Leaving.) 2015-03-10T05:35:51 *** andrei__ has left #rtems 2015-03-10T05:35:58 *** andrei__ has joined #rtems 2015-03-10T05:39:17 hm, that's weird. $ git clone git://git.rtems.org/rtems-source-builder.git seems to hang 2015-03-10T05:39:46 *** andrei__ is now known as andrei_ 2015-03-10T05:40:04 *** andrei_ has joined #rtems 2015-03-10T05:40:04 *** andrei_ has joined #rtems 2015-03-10T05:40:32 on my home computer it succeded without issues 2015-03-10T05:45:55 *** namaste has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-10T05:46:59 *** namaste has joined #rtems 2015-03-10T06:59:12 *** namaste has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-10T07:37:28 *** antgreen has quit IRC (Ping timeout: 255 seconds) 2015-03-10T07:50:58 *** andrei_ has quit IRC (Ping timeout: 246 seconds) 2015-03-10T08:10:31 *** ragunath has joined #rtems 2015-03-10T08:16:10 *** andrei_ has joined #rtems 2015-03-10T08:25:41 *** gedare has joined #rtems 2015-03-10T08:25:41 *** ChanServ sets mode: +o gedare 2015-03-10T08:28:46 good morning 2015-03-10T08:29:42 hi, gedare 2015-03-10T08:37:02 *** kiran4399 has joined #rtems 2015-03-10T08:37:12 *** kiran4399 has left #rtems 2015-03-10T08:44:09 [trac] #2291/defect (new) updated by gedare (Which Getting Started doc? That link should also get propagated into wiki pages that ...) http://devel.rtems.org/ticket/2291 2015-03-10T08:53:37 [trac] #2291/defect (new) updated by amar (It's the 5th point listed here: ...) http://devel.rtems.org/ticket/2291 2015-03-10T08:55:28 [trac] 0001-dosfs-avoid-buffer-overread.-closes-2292.patch (Fix.) attached to #2292 by gedare http://devel.rtems.org/ticket/2292 2015-03-10T08:55:39 [trac] #2292/defect (new) updated by gedare (This is a possible out-of-bounds read access. I don't think this is a bug, but I've ...) http://devel.rtems.org/ticket/2292 2015-03-10T09:07:22 [trac] 0001-RSB-Use-gcc-5-20150301-snapshot.patch (Patch to use gcc snapshot) attached to #2254 by heshamelmatary http://devel.rtems.org/ticket/2254 2015-03-10T09:10:52 [git] b05549f by Joel Sherrill: [cpukit/libcrypt/misc.c: Reverse order of statements in ifdef] http://devel.rtems.org/changeset/b05549f/rtems 2015-03-10T09:18:39 [trac] TBR/UserManual/4.11_Release_Notes edited by joel.sherrill (Add some comments) http://devel.rtems.org/wiki/TBR/UserManual/4.11_Release_Notes 2015-03-10T09:19:39 *** DrJoel has joined #rtems 2015-03-10T09:19:39 *** DrJoel has joined #rtems 2015-03-10T09:19:39 *** ChanServ sets mode: +o DrJoel 2015-03-10T09:20:20 hi DrJoel 2015-03-10T09:20:33 Hey.. 2015-03-10T09:20:50 i'm pretty busy hacking these days, but i can spare a few minutes here and there to look at small things 2015-03-10T09:20:52 I didn't want to call you out on the last email but can you look at https://devel.rtems.org/wiki/TBR/UserManual/4.11_Release_Notes 2015-03-10T09:21:06 I think the Pok item listed in "nice to have" might need updating 2015-03-10T09:21:06 yes i glanced at it already when verm__ posted on the ticket 2015-03-10T09:21:10 ah 2015-03-10T09:21:28 heh 2015-03-10T09:21:30 sure. 2015-03-10T09:22:16 DrJoel: can we close off commits to master? 2015-03-10T09:22:24 No 2015-03-10T09:22:33 why?, they can still go into 4.11.1 2015-03-10T09:22:45 it's impossible to cut a release if we continously land unrelated patches to 4.11 2015-03-10T09:22:50 git checkout -b 4.11 2015-03-10T09:23:19 ? 2015-03-10T09:23:31 the rtems release procedure is to create a branch, and the master will continue to be the rolling development head 2015-03-10T09:23:36 Based on the tickets, there are still a few items and a couple are going to take time 2015-03-10T09:23:46 that too! 2015-03-10T09:23:49 i'm not saying to close off commits, period 2015-03-10T09:24:03 where is my list... 2015-03-10T09:24:04 i'm saying any new commits should be to 4.11.1 while any remaining issues for 4.11 are done in master 2015-03-10T09:24:13 [trac] TBR/UserManual/4.11_Release_Notes edited by gedare (Remove hypervisor patches from "like-to-have") http://devel.rtems.org/wiki/TBR/UserManual/4.11_Release_Notes 2015-03-10T09:24:14 Discuss it on devel 2015-03-10T09:24:18 once 4.11 is branched 4.11.1 can be merged in and continue it 2015-03-10T09:24:30 are the hypervisor patches in? 2015-03-10T09:24:33 this is my first time being part of an RTEMS release 2015-03-10T09:24:37 it seems very...haphazard 2015-03-10T09:24:41 We haven't settled the tools for one 2015-03-10T09:24:43 DrJoel: we merged what we wanted to merge, and dropped what we didn't 2015-03-10T09:25:08 Normally the tools would be settled before now. According to Chris, we don't have a gdb to branch with 2015-03-10T09:25:08 there were fundamental problems we couldn't resolve, having to do with externally-built binary blobs 2015-03-10T09:25:21 i'm not saying to close off fixes 2015-03-10T09:25:26 gedare: ok 2015-03-10T09:25:34 every morning sebastian lands a half a dozen patches 2015-03-10T09:25:37 Post it to devel for discussion 2015-03-10T09:25:47 we've had so many commits after joel did his run 2015-03-10T09:25:54 what should i post there? 2015-03-10T09:26:05 A question on when to branch? 2015-03-10T09:26:15 https://devel.rtems.org/query?status=accepted&status=assigned&status=new&status=reopened&milestone=4.11&col=id&col=summary&col=status&col=type&col=priority&col=milestone&col=component&order=priority 2015-03-10T09:26:20 I started the discussion in an email sometime in the past few days 2015-03-10T09:26:32 we can bump moxie tools they seem to be just broken. 2015-03-10T09:26:43 hesham and kiwichris were working on some fix but i don't think it will be good 2015-03-10T09:26:45 i don't know where anthony went he was here and said they worked 2015-03-10T09:26:53 sure, they work, if you build gcc head 2015-03-10T09:26:59 he said "use head" 2015-03-10T09:27:11 2290 is a blocker and 2291 will likely be fixed before Chris settles gdb 2015-03-10T09:27:27 gedare what about the deprecation ticket? 2015-03-10T09:27:40 i can work to close that today, at least for notepads. 2015-03-10T09:28:18 Chris wanted to deprecate task variables and I wanted to deprecate rtems_clock_get. I guess we need tickets for those. 2015-03-10T09:28:33 right. 2015-03-10T09:28:54 well, if you want me to use deprecate attribute, i'll have to add a macro for it too. 2015-03-10T09:29:00 [trac] #2293/defect (v:4.11) created by joel.sherrill (Deprecate and Obsolete Classic API task variables) http://devel.rtems.org/ticket/2293 2015-03-10T09:29:10 [trac] #2293/defect (new) updated by joel.sherrill (empty comment) http://devel.rtems.org/ticket/2293 2015-03-10T09:29:37 you can set me to own it i'll work my way through these deprecations today i guess 2015-03-10T09:30:11 [trac] #2294/defect (v:4.11) created by joel.sherrill (Deprecate and Obsolete Classic API rtems_clock_get()) http://devel.rtems.org/ticket/2294 2015-03-10T09:30:15 i don't really like Chris's justification, "because gdb-7.8.2 is broken on Windows." 2015-03-10T09:30:27 [trac] #2294/defect (assigned) updated by joel.sherrill (empty comment) http://devel.rtems.org/ticket/2294 2015-03-10T09:30:43 [trac] #2293/defect (assigned) updated by joel.sherrill (empty comment) http://devel.rtems.org/ticket/2293 2015-03-10T09:30:57 Just trying to make sure we have tickets to reflect the things we want. 2015-03-10T09:31:19 As you do this, take notes and create a wiki page. We don't do this often so should have notes. I barely remembered the function attribute 2015-03-10T09:32:59 https://devel.rtems.org/wiki/Developer/Release/Procedure 2015-03-10T09:33:01 I think there used to be something like --enable-deprecated but I don't see it on the head 2015-03-10T09:33:14 sent an email 2015-03-10T09:33:19 https://devel.rtems.org/wiki/Developer/Release 2015-03-10T09:33:24 some duplication on those two pages 2015-03-10T09:34:02 they should be consolidated and "fixed" as we do this release. 2015-03-10T09:34:10 i didn't even know that page existed 2015-03-10T09:34:19 i'm looking for the right place to put a page about deprecating apis. 2015-03-10T09:34:35 i will consolodate them later today, thanks 2015-03-10T09:34:47 okay. i'm skimming Developer/ for any suitable location and saw them. :) 2015-03-10T09:35:00 Ideally a deprecated API, should have to be enabled with a "--enable-deprecated", give a linking warning, maybe a warning when configured if applicable, and the Doxygen and documentation note that it is marked for removal. 2015-03-10T09:35:07 Anything else we might need to do? 2015-03-10T09:35:51 do you want a configure switch? I'd prefer to avoid one myself. 2015-03-10T09:37:50 6 years old: https://devel.rtems.org/wiki/TBR/UserManual/DriverManager 2015-03-10T09:39:31 verm__: why isn't waf/ nested under Developer? 2015-03-10T09:40:00 *** Davidbrcz has joined #rtems 2015-03-10T09:40:06 OK.. We used to have one for this purpose, but I didn't see it even in the 4.10 source. 2015-03-10T09:41:14 [trac] Developer/DeprecatingInterfaces created http://devel.rtems.org/wiki/Developer/DeprecatingInterfaces 2015-03-10T09:41:39 DrJoel: good smp projects off the top of your head? 2015-03-10T09:43:07 verm__: Developer/ is intended to be the "developer's manual" for RTEMS? 2015-03-10T09:43:18 evolving wiki-version of it anyway 2015-03-10T09:44:57 Nothing obvious. sebhub might want to comment. 2015-03-10T09:45:14 The issues I know are hit and miss unless something is on the SMP wiki page. 2015-03-10T09:45:43 gedare: yes 2015-03-10T09:45:57 [trac] Developer created http://devel.rtems.org/wiki/Developer 2015-03-10T09:46:05 gedare: because it's both user and developer content 2015-03-10T09:46:07 verm__: i didn't see an obvious entry point so i made one there 2015-03-10T09:46:09 There is the x86 context switch issue, Ada run-time switched to pthread_keys, probably a review of RTEMS reentrancy for newlib, maybe libc locks 2015-03-10T09:46:18 hopefully we can fill in from there. 2015-03-10T09:46:18 i didn't want to put it there, also waf/ will be emptied once it merges 2015-03-10T09:46:22 sebhub may have ideas for help on finer grained locks 2015-03-10T09:46:31 oh is there a good place to name something? 2015-03-10T09:46:45 he doesn't want to mentor 2015-03-10T09:46:52 so any projects should not overlap with his ongoing work 2015-03-10T09:47:14 i encouraged andrei_ to propose the CV project 2015-03-10T09:47:24 which is tangentially SMP related 2015-03-10T09:47:43 maybe to improve SMP bsp supports? 2015-03-10T09:47:44 The CV project is a good one. 2015-03-10T09:47:47 x86 needs work.. 2015-03-10T09:48:01 some other arch makes sense? 2015-03-10T09:48:15 definitely. x86 needs the context switch fixed, better support for modern PIC as minimum 2015-03-10T09:48:21 depends on the students comfort level with asm 2015-03-10T09:48:54 verm__: do you think having the entry page to the "Dev manual" at Developer/ is an ok thing? 2015-03-10T09:48:57 Given that it is probably <25 lines of asm and uses the same algorithm as arm, ppc, and sparc, it shouldn't prove too challenging. :) 2015-03-10T09:48:57 https://devel.rtems.org/wiki/Developer 2015-03-10T09:49:03 gedare, DrJoel, I started building RSB yesterday, but I ran into some issue, I think I'm missing python-dev, will check when I get home 2015-03-10T09:49:14 DrJoel helped me in the process 2015-03-10T09:49:20 ok the python versions can be an issue too 2015-03-10T09:49:30 it said that it can't find usr/bin/python 2015-03-10T09:49:35 andrei_: kiwichris is in Sydney and is often here when gedare and I call it a day 2015-03-10T09:50:00 Sounds like Python is missing or not where it is expected. Is python on your machine at all? 2015-03-10T09:50:08 yes 2015-03-10T09:50:13 ll /usr/bin/python 2015-03-10T09:50:19 was a simlink to 2015-03-10T09:50:29 python2.7 2015-03-10T09:50:46 it either is missing python2.7-dev, or I don't know 2015-03-10T09:50:59 I'm at work now, so I can't check, I'm GMT+2 .. 2015-03-10T09:51:06 Hmm... no idea. weird though 2015-03-10T09:51:14 I have to go away from the keyboard. bbl 2015-03-10T09:51:37 andrei_: this is a trick that sometimes works: 2015-03-10T09:51:37 This is a shell programming trick to find out where the desired program is and invoke it. I have used this to use a non-default Python version. 2015-03-10T09:51:37 #! /usr/bin/env bash 2015-03-10T09:52:02 DrJoel : thanks, I ll try 2015-03-10T09:52:04 That finds bash whereever it is located. Is the RSB doing something similar or does it have a hard-coded path 2015-03-10T09:52:12 well replace bash with python :) 2015-03-10T09:52:17 DrJoel: release with or without a commit freeze? your $0.02 2015-03-10T09:52:29 gedare: regarding ASM, where you talking about the CV project? 2015-03-10T09:52:37 ok later DrJoel 2015-03-10T09:52:40 no andrei_ 2015-03-10T09:52:43 ah 2015-03-10T09:52:59 i'd expect that will be purely in C. 2015-03-10T09:53:14 although the SMP-safe version will require using some atomics, we provide wrappers for those 2015-03-10T09:53:21 I asked because afaik mutex have hardware implementation 2015-03-10T09:53:26 (test-and-set) 2015-03-10T09:53:31 but I don t know about monitors, however 2015-03-10T09:53:35 correct 2015-03-10T09:53:49 which I assume are involved in CV project 2015-03-10T09:53:52 every synchronization primitive must use a hardware atomic operation 2015-03-10T09:54:23 but CV is more comparable to POSIX CV's but in Classic API. 2015-03-10T09:54:24 if you want to read/write the same location in parallel you need atomic support 2015-03-10T09:54:33 we should only need HW atomics at lower levels 2015-03-10T09:54:48 right. i think we want to implement CV in supercore though 2015-03-10T09:54:52 POSIX spin lock API may be worth a revisit not that SMP is in place 2015-03-10T09:55:09 Yes .. but shuoldn't need HW atomics necessarily except maybe to test condition 2015-03-10T09:56:29 I'll try to get rtems going asap, it'll probably take me some time to 'chew' details on how it works, even at a high level view 2015-03-10T09:56:41 yup. 2015-03-10T09:57:14 once you get tools working and getting started done, you can tinker some 2015-03-10T09:57:45 check out the posix condvar test cases, and maybe try to write a test case for a classic condvar 2015-03-10T09:57:58 if one were to exist. ;) 2015-03-10T09:58:38 this will help you to think about the API for it, and give a solid place to start from using a TDD approach to begin your effort. 2015-03-10T09:58:53 Hi 2015-03-10T09:59:02 I'll look into it once I get it running, thanks 2015-03-10T09:59:21 i'm using one the makefile a found at https://github.com/RTEMS/examples-v2 2015-03-10T09:59:50 I want to add a custom Include directory with the -I option and a custom link path with the -L one 2015-03-10T10:00:00 thanks for heloping out, need to go now, will be back later ! 2015-03-10T10:00:24 I've tried to change LINK_LIBS, LD_PATH, LD_LIBS, CXXFLAGS, CPPFLAGS 2015-03-10T10:00:37 nothing works and I'm out of idea =/ 2015-03-10T10:00:37 hi Davidbrcz 2015-03-10T10:00:41 hi gedare =) 2015-03-10T10:01:07 If you need, I can paste online my current makefile 2015-03-10T10:01:13 and the layout of my project 2015-03-10T10:01:29 you might like to consider using the waf version, i think it will pick up your environments CFLAGS 2015-03-10T10:01:39 btw did you try CFLAGS? or LDFLAGS? 2015-03-10T10:02:03 you can see README.waf for instructions to build examples-v2 with waf 2015-03-10T10:02:17 not LDFLGAS 2015-03-10T10:03:33 but I cant use CFLAGS since I'm using C++ 2015-03-10T10:04:19 *** andrei_ has quit IRC (Ping timeout: 246 seconds) 2015-03-10T10:04:33 oh i see 2015-03-10T10:05:22 and I dont understand how waf helps me =/ 2015-03-10T10:05:30 (I'm kind of a newbie with RTEMS) 2015-03-10T10:05:44 examples-v2 has two ways to be compiled 2015-03-10T10:05:53 one way is with 'make' and the "RTEMS_MAKEFILE" 2015-03-10T10:06:14 https://devel.rtems.org/wiki/TBR/UserManual/Example_Application_Compiling 2015-03-10T10:06:28 somewhat documented there 2015-03-10T10:06:39 another way you can compile it is with waf, which is documented in the README.waf 2015-03-10T10:06:46 the waf approach ignores the Makefiles 2015-03-10T10:06:51 i'm using the first way 2015-03-10T10:06:54 and instead uses the wscripts to compile 2015-03-10T10:07:20 gedare, I is a bit more complex than that 2015-03-10T10:07:42 I'm porting an HLA simulation on a embedded device with RTEMS 2015-03-10T10:07:46 i see! 2015-03-10T10:07:54 I can afford to introduce a new building system 2015-03-10T10:08:08 I'm already fighting to use CMake and rtems 2015-03-10T10:08:17 *can't afford? 2015-03-10T10:08:19 (it seems tp be working) 2015-03-10T10:08:26 sure 2015-03-10T10:08:27 not enough time 2015-03-10T10:08:36 you wrote can, just checking 2015-03-10T10:08:44 The deadline is in 2 weeks (I'm a student) 2015-03-10T10:08:57 ok 2015-03-10T10:09:56 where did you add CXXFLAGGS? 2015-03-10T10:10:31 before including RTEMS's makefile 2015-03-10T10:10:42 I'll upload my makefile, 1s 2015-03-10T10:11:09 http://paste.fedoraproject.org/196267/14260002 2015-03-10T10:11:36 and the executed line is sparc-rtems4.10-g++ -B/opt/rtems-4.10/sparc-rtems4.10/leon3/lib/ -specs bsp_specs -qrtems -g -Wall -g -mcpu=cypress -msoft-float -c -o o-optimize/init.o init.cc 2015-03-10T10:12:00 That looks OK but if you added args, they didn't show up. 2015-03-10T10:12:26 indeed 2015-03-10T10:12:27 put the flags after the include 2015-03-10T10:12:39 include Makefile.inc will probably wipe out some of the flags settings 2015-03-10T10:12:54 damn It works ! 2015-03-10T10:13:08 yeah 2015-03-10T10:13:12 thank you gedare =) 2015-03-10T10:13:14 *** zoso has quit IRC (Ping timeout: 272 seconds) 2015-03-10T10:13:23 also, LDFLAGS not LDFLGAS 2015-03-10T10:13:37 I've lost a bit of hair on that x) 2015-03-10T10:13:53 i've fought with just about every build system in common use. 2015-03-10T10:14:00 i know the pain 2015-03-10T10:14:48 and honnestly, I dont find a lot of ressources on the internet 2015-03-10T10:15:03 feel free to improve our wiki. :) 2015-03-10T10:15:14 When I'll have time x) 2015-03-10T10:15:22 i don't think a lot of folks use include in makefiles that much 2015-03-10T10:15:42 or when they do it is very simple 2015-03-10T10:15:47 ok good luck! 2015-03-10T10:16:10 if you're not funded through the summer, you should consider applying to GSOC 2015-03-10T10:16:35 * gedare gets back to work 2015-03-10T10:18:55 gedare, I'll be at UCB as short term scholar 2015-03-10T10:19:10 can't apply for ESA Summer school 2015-03-10T10:19:21 =/ 2015-03-10T10:19:34 Can apply to Google SoC but work load may not permit 2015-03-10T10:20:11 DrJoel, while being in the USA ? 2015-03-10T10:21:08 Yes. GSOC allows participants from many more countries than ESA SOCIS 2015-03-10T10:21:23 DrJoel, I'm french at the begining 2015-03-10T10:21:50 How long will you be at UCB? 2015-03-10T10:21:55 6 monts 2015-03-10T10:21:58 months* 2015-03-10T10:22:04 Starting? 2015-03-10T10:22:04 from April to September 2015-03-10T10:22:07 ok 2015-03-10T10:22:18 that precludes effective participation in GSOC really. 2015-03-10T10:22:20 so summer is already busy ^^ 2015-03-10T10:22:26 you should be focused on doing well in that effort. 2015-03-10T10:22:27 yeah. 2015-03-10T10:23:20 *** Hesham has joined #rtems 2015-03-10T10:36:55 hum 2015-03-10T10:39:54 *** monstr has quit IRC (Remote host closed the connection) 2015-03-10T10:45:36 *** sebhub has quit IRC (Quit: Konversation terminated!) 2015-03-10T10:49:01 FYI NASA MMS launches Thursday.. details at: https://www.facebook.com/joel.sherrill/posts/10152789988536589 2015-03-10T10:50:05 It is hard to believe that it has been over 14 years since I went to Goddard to teach an RTEMS class to them. They were working on ST-5 at the time. That was the proof of technology/risk reduction mission for this. 2015-03-10T10:55:47 [trac] Developer/DeprecatingInterfaces edited by gedare (Write new advice for deprecating features.) http://devel.rtems.org/wiki/Developer/DeprecatingInterfaces 2015-03-10T10:57:11 heh 2015-03-10T10:58:29 i spent too much time on the notepad deprecation. i dno't think i'll have time to do the others today but hopefully they won't take long now i have the basics down. 2015-03-10T10:58:54 DrJoel: Can I get a written permission for using some figure part of RTEMS docs from OAR corporation, or do you suggest I imitate them and draw my own? 2015-03-10T10:59:48 Hesham: you should probably send him an email with a link to the figure you want to use, and where you want to use it. 2015-03-10T11:00:41 gedare: Ok, I'll, thanks 2015-03-10T11:03:01 also, it is odd to read a contraction without a following verb 2015-03-10T11:04:46 Ok, I'll add a verb next time ;) 2015-03-10T11:04:57 exactly. 2015-03-10T11:05:25 or spell it out, "Ok, I will, thanks" 2015-03-10T11:05:45 i don't know that its grammatically incorrect, it just "sounds weird". 2015-03-10T11:06:18 Thanks, please keep telling me about these minor errors, I appreciate it 2015-03-10T11:08:42 Hesham.. Yeah verily you can use the figure. Which one? Do you need an alternate format? 2015-03-10T11:09:26 DrJoel: I'll send you an e-mail with the paper containing the figures. Thanks! 2015-03-10T11:10:21 Also just note what format you want them in. 2015-03-10T11:32:00 DrJoel: verm__: ppisa makes a good point about not centralizaing headers, to improve code modularity 2015-03-10T11:33:50 *** ragunath_ has joined #rtems 2015-03-10T11:36:22 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-10T11:48:37 *** ragunath_ has quit IRC (Ping timeout: 246 seconds) 2015-03-10T11:53:04 gedare: he's confused about what we're trying to do i've been thinking about how to respond 2015-03-10T11:53:15 the headers we're moving are ones that we consume internally as well as public 2015-03-10T11:53:36 his idea of using multiple -I to include headers is the hack we're specifically trying to avoid 2015-03-10T11:53:46 as that lets you 'cheat' and have non-absolute includes 2015-03-10T11:53:47 Epiphany folks merged the binutils and GCC pull requests, still there are some problems with limits.h definitions when building RTEMS which I didn't have time to investigate more. 2015-03-10T11:53:51 i already tripped up twice in the current repo 2015-03-10T11:53:56 due to #include 2015-03-10T11:54:19 it tried to include the compiler pthreads. in waf it's #include 2015-03-10T11:56:27 ok. take your time to answer his comments and i'll have a careful read through. you may like to mention that BSP headers are specifically kept separate for modularity. 2015-03-10T11:56:54 that's a good idea 2015-03-10T11:56:56 i'm responding now 2015-03-10T12:03:04 hum 2015-03-10T12:03:22 I though the hack with LDFLAG was working but it is 2015-03-10T12:03:24 not* 2015-03-10T12:04:09 I have undefined symbols. A teacher of mine told me it could be because the libraries are not at the end 2015-03-10T12:04:23 yes 2015-03-10T12:05:04 never mind 2015-03-10T12:05:07 found the hack 2015-03-10T12:05:25 ok good. the linker command line can be fragile. 2015-03-10T12:05:31 using LDFLAGS to set up the libraries was placing them in the middle of the command 2015-03-10T12:05:48 using LINK_LIBS put them at the end 2015-03-10T12:06:47 ah. 2015-03-10T12:08:30 *** namaste has joined #rtems 2015-03-10T12:10:06 I was wondering 2015-03-10T12:10:19 does any of you know ISAE and/or Supaero ? 2015-03-10T12:11:12 DrJoel: is it just rtems_clock_get() or also rtems_clock_get_tod()? what is the suggested alternative? just gettimeofday? 2015-03-10T12:11:20 never heard of 'em. 2015-03-10T12:11:28 gedare, ok 2015-03-10T12:11:34 good to know ^^ 2015-03-10T12:11:54 but i'm not really up to par with that community 2015-03-10T12:12:02 I know of EPFL... ;) 2015-03-10T12:12:23 yeah =) 2015-03-10T12:13:58 is there any list of all the #define I can switch on/off for configuring RTMES ? 2015-03-10T12:14:31 there is a manual.. um 2015-03-10T12:14:55 i just accidentally closed my browser taking me a sec. 2015-03-10T12:15:21 https://docs.rtems.org/doc-current/share/rtems/html/c_user/Configuring-a-System.html#Configuring-a-System 2015-03-10T12:17:11 do you think thah "Unexpected trap ( 4) at address 0x40071CFC p disabled" 2015-03-10T12:17:26 has anythng to do with a bad RTEMS's set up 2015-03-10T12:17:39 or its somethng completly different 2015-03-10T12:17:47 and tanks for the link gedare 2015-03-10T12:19:31 what target is this? 2015-03-10T12:20:27 it is a GR-XC6S from gaisler (http://www.gaisler.com/index.php/products/boards/gr-xc6s-lx75) 2015-03-10T12:20:54 gedare: just rtems_clock_get(). The first parameter is an enum indicating which format to get the time in. The method is now implemented by calling the replacement methods for each enum 2015-03-10T12:21:04 leon3? 2015-03-10T12:21:10 ok 2015-03-10T12:21:12 Davidbrcz: How far did it run? 2015-03-10T12:21:31 * gedare consults his handy-dandy sparc trap table translator 2015-03-10T12:21:44 DrJoel, not far 2015-03-10T12:21:57 I've not been able yet to run a remote gdb on it 2015-03-10T12:22:12 BAD_TRAP; ! 04 fp disabled 2015-03-10T12:22:26 Sounds like your tasks need to have floating point turned on for them. 2015-03-10T12:22:35 but I have not been able to see any of my printf/std::cout 2015-03-10T12:22:52 even if it is the first command in the Init_task 2015-03-10T12:23:22 and I find that weird 2015-03-10T12:23:41 what is the instruction at that address and what method is it in? 2015-03-10T12:24:10 you can use sparc-rtemsXXX-objdump -da -s to dump the .exe with source interleaved 2015-03-10T12:24:12 I don'k know =/ 2015-03-10T12:24:15 could be a constructor 2015-03-10T12:24:59 global constructors run before the first user thread gets control. Right after the first context switch and before your Init task 2015-03-10T12:25:01 the binary is 1.7 Mo 2015-03-10T12:25:13 I'll need a minute or two ^^ 2015-03-10T12:26:06 #define CONFIGURE_INIT_TASK_ATTRIBUTES RTEMS_FLOATING_POINT 2015-03-10T12:26:21 Add that to the configure directives before including 2015-03-10T12:26:27 But let's check what really failed first 2015-03-10T12:26:46 after running objdump I do not have the interleaved source with the dump 2015-03-10T12:27:05 I'll switch to debug first 2015-03-10T12:27:21 It's ok for now.. what was the instruction at that address 2015-03-10T12:28:28 *** gedare has quit IRC (Remote host closed the connection) 2015-03-10T12:28:44 Capital -S not lower case -s :) 2015-03-10T12:29:51 DrJoel: i re-ran the doxygen and fixed the crashes i was having 2015-03-10T12:30:06 https://docs.rtems.org/doxygen/warnings5.log 2015-03-10T12:30:08 How bad? 2015-03-10T12:30:38 we've got many thousands of undocumented bits of code 2015-03-10T12:30:55 *** siddharth7 has joined #rtems 2015-03-10T12:31:28 *** gedare has joined #rtems 2015-03-10T12:31:29 *** ChanServ sets mode: +o gedare 2015-03-10T12:32:43 Hi mentors, anyone around? 2015-03-10T12:32:51 That's no surprise really. RTEMS is older than Doxygen and we probably only started 2015-03-10T12:33:03 DrJoel, the instruction is 40071cfc: d1 18 40 00 ldd [ %g1 ], %f8 2015-03-10T12:33:07 I am not sure when we started converting to Doxygen :( 2015-03-10T12:33:12 Well that's a floating point instruction. 2015-03-10T12:33:25 Take my suggestion to add to the configuration lines and try again 2015-03-10T12:33:26 hi diginet 2015-03-10T12:33:29 hi siddharth7 2015-03-10T12:33:40 sorry diginet my fingers slipped. :0 2015-03-10T12:33:45 it is from std::numeric_limits::epsilon() 2015-03-10T12:33:45 gedare: no problem 2015-03-10T12:34:17 sounds like an fp issue for sure! 2015-03-10T12:34:29 hi gedare, I am willing to contribute to Configuration Gui project, can you help me get started? 2015-03-10T12:35:04 siddharth7: that project is being actively developed by verm__ but you may like to discuss with him about possible projects related 2015-03-10T12:35:06 DrJoel: maybe we can add a sudden death feature to buildbot, if you touch a file your commit won't go in until it's doxygen clean? 2015-03-10T12:35:20 heh, i'm sure everyone will avoid touching the largest files 2015-03-10T12:35:56 gedare, ok thanks 2015-03-10T12:36:11 sed -i -e 's/\/\*.*\*\//g' 2015-03-10T12:36:14 no more doxygen problems. 2015-03-10T12:36:45 hi verm__, can you help me get started on Configuration Gui project? 2015-03-10T12:37:54 verm__: I am actually prone to attach the smallest numbers and largest numbers. One shortens the list quickly. The other drops the number. That's how I worked on coverage and pushed that. 2015-03-10T12:38:08 I will have some more Doxygen commits after my meeting I think. 2015-03-10T12:38:22 Mostly knocking out more of the old format comments. :( 2015-03-10T12:38:28 Not sure if it will help the results or not 2015-03-10T12:38:29 same result 2015-03-10T12:38:41 with the define =/ 2015-03-10T12:38:41 siddharth7: do you have experience working with apertium? 2015-03-10T12:38:42 Where did you add the line? 2015-03-10T12:38:51 *** ragunath has joined #rtems 2015-03-10T12:39:00 gedare, yes 2015-03-10T12:39:23 We really need to know what method owns that address. you can look at the .num file. it is sorted by address 2015-03-10T12:39:33 have you worked with the apertium community? know people in it? 2015-03-10T12:39:39 DrJoel, here is my main http://paste.fedoraproject.org/196327/26009132 2015-03-10T12:40:34 Not much, I just started working on a light weight desktop Gui some weeks back 2015-03-10T12:40:35 it is a bit messy because I'm trying to port on a embedded target a HLA simulation 2015-03-10T12:41:05 oh.. I missed that it was a POSIX init thread. It is already FP. That line isn't needed 2015-03-10T12:41:08 ok siddharth7 thanks just curious. 2015-03-10T12:41:48 gedare, ok 2015-03-10T12:41:50 siddharth7: are you interested only in python-gui projects, or does other coding interest you too? 2015-03-10T12:42:24 I know python,c, c++ 2015-03-10T12:42:49 David -mcpu=cypress -msoft-float 2015-03-10T12:42:55 i just learnt to make gui in my holidays, so wanted to do some project on it 2015-03-10T12:43:12 were on your gcc line. How did this code get compiled? 2015-03-10T12:43:34 Are you using rtems from rtems.org or gaisler.com? 2015-03-10T12:43:36 but yeah, any other project that may use c or c++ would also be great 2015-03-10T12:43:50 I'm using the one from rtems.org 2015-03-10T12:44:01 OK. 2015-03-10T12:44:03 I did a git clone of the 4.10.2 branch 2015-03-10T12:44:14 OK.. just wanting to know 2015-03-10T12:44:30 siddharth7: we have a number of projects that use python, c, and c++. the rtems-source-builder and rtems-tester projects are heavy on python/scripting side 2015-03-10T12:44:38 there is also the python-gdb project 2015-03-10T12:44:42 I configure rtems with ../rtems-4.10.2/configure --target=sparc-rtems4.10 --enable-rtemsbsp=leon3 --enable-tests=samples --enable-posix --enable-networking --enable-languages=c,c++ 2015-03-10T12:44:55 We need to know what code that address is associated with 2015-03-10T12:45:08 and verm__ is working on rolling out a waf-based build system for rtems, which is python-backed and uses build rules that are similar to python scripting. 2015-03-10T12:45:12 That's a fine configure line. Just trying to know what version you are using. 2015-03-10T12:45:44 I'm linking with my binary a homemade version of CERTI 2015-03-10T12:45:59 Davidbrcz: you can use sparc-rtems4.10-objdump -d blah.exe > blah.txt and then you can search for the address where the trap happens 2015-03-10T12:45:59 (an open Source RTI) wich I compiled myself 2015-03-10T12:46:43 gedare, the instruction is the instruction is 40071cfc: d1 18 40 00 ldd [ %g1 ], %f8 2015-03-10T12:46:56 *** altairpearl_ has quit IRC (Remote host closed the connection) 2015-03-10T12:46:56 it is one from std::numeric_limits::epsilon() 2015-03-10T12:47:07 a standard c++ function 2015-03-10T12:47:23 and what function is it in? 2015-03-10T12:47:43 %f8 is a floating point register. most likely the floating point unit is not enabled 2015-03-10T12:47:44 ok 2015-03-10T12:48:01 if you enable the floating point unit probably this will work then 2015-03-10T12:49:02 there is #define CONFIGURE_INIT_TASK_ATTRIBUTES RTEMS_FLOATING_POINT 2015-03-10T12:49:35 But you have a POSIX init thread. That applies to Classic API Init tasks. 2015-03-10T12:50:26 My big issue is that I don't know where that code was generated from. The BSP is set for soft-float. Any code compiled using the same flags would avoid the floating point register 2015-03-10T12:50:32 I want a posix application because I porting an unix application 2015-03-10T12:50:45 That's fine. We just need to know what method that instruction belongs to 2015-03-10T12:50:52 I'm seeing your point 2015-03-10T12:51:07 gedare, well the project "Improve RTEMS Source Builder" is a potential project that i would like to work upon, any inputs? 2015-03-10T12:51:12 -msoft-float means : "emulate flot" with integer instruction ? 2015-03-10T12:51:52 yes.. 2015-03-10T12:52:02 be back in a bit.. have meeting in 9 minutes 2015-03-10T12:52:23 ok 2015-03-10T12:55:26 siddharth7: kiwichris is the maintainer of that, and he lives in AU, so not awake at the moment. i think the big need in RSB nowadays is to build other third-party packages for including to RTEMS 2015-03-10T12:56:04 siddharth7: https://devel.rtems.org/wiki/Developer/Projects/Open/Toolkits 2015-03-10T12:56:42 kiwichris recently added Google Protocol Buffers which may be useful as an example 'howto'. 2015-03-10T12:57:48 I added libjpeg which wasn't too bad. The packages in rtems-addon-packages and graphics-toolkit git repos are the top of the list. Others will likely come from HOWTOs and notes. 2015-03-10T12:57:50 bbl 2015-03-10T12:58:45 beng-nl: hi ben 2015-03-10T13:01:49 gedare, thanks for the help, i will go through the link in mean time and come back if I have any issues 2015-03-10T13:02:16 siddharth7: i am replying to your email on the list 2015-03-10T13:02:52 ok verm__ 2015-03-10T13:03:55 ok siddharth7 you should also go through our "getting started for GSOC students" guide to familiarize yourself with building and running RTEMS. 2015-03-10T13:04:09 sure! 2015-03-10T13:05:13 siddharth7: Did you apply to RTEMS for GSOC in the past or talk to us? Nickname sounds familiar 2015-03-10T13:06:01 DrJoel, nope..This is my first time, I am a freshman student 2015-03-10T13:06:04 Davidbrcz: Once we get this characterized, it probably needs to be a question on devel@rtems.org. CC' danielh@gaisler.com on it. I think this may turn out to be a common SPARC issue he can answer quickly 2015-03-10T13:06:25 siddharth7: OK .. welcome aboard. Must have been someone with a similar name .. 2015-03-10T13:06:42 DrJoel, Thanks 2015-03-10T13:06:43 ! 2015-03-10T13:08:03 [trac] Developer/Projects/ConfigGUI edited by amar (Fix 2nd reference.) http://devel.rtems.org/wiki/Developer/Projects/ConfigGUI 2015-03-10T13:08:56 DrJoel, thanks for the lead 2015-03-10T13:09:06 I'll ask my teachder first 2015-03-10T13:09:20 FWIW I think I may be near the end of fixing warnings based on the current build parameters (no networking, only sample tests). The remaining ones have me stumped or I am sick of looking at them. 2015-03-10T13:09:44 I am going to build with all tests overnight and see if something interesting shows up in the tests that I can tackle. 2015-03-10T13:09:45 ok.. back to waiting for folks to show for meeting 2015-03-10T13:09:47 and if he is clueless I'll go to devel@rtems.org (on which he may be on x)) 2015-03-10T13:10:27 The Gaisler folks know the answer. The FPU handling on the SPARC's is a bit odd. But we need to know the method that did that. 2015-03-10T13:10:52 It may also be worth the cycles to switch to the 4.11 head. Build new tools, build RTEMS. Shouldn't take but a couple of hours tops 2015-03-10T13:16:33 [trac] waf/GUI edited by amar (Add some details.) http://devel.rtems.org/wiki/waf/GUI 2015-03-10T13:16:39 siddharth7: https://devel.rtems.org/wiki/waf/GUI 2015-03-10T13:19:36 verm__, going through it 2015-03-10T13:31:25 *** andrei_ has joined #rtems 2015-03-10T13:37:52 *** Hesham has quit IRC (Ping timeout: 240 seconds) 2015-03-10T13:47:03 verm__: in the new layout, to use pthread, we must #include ? 2015-03-10T13:52:09 ? 2015-03-10T13:52:15 oh, sorry yes 2015-03-10T13:52:21 it needs to be explicit 2015-03-10T13:52:29 i actually didn't even notice the error at first 2015-03-10T13:52:39 because trhere was no -Iinclude/rtems/posix it was picking up pthread.h from the compiler 2015-03-10T13:54:18 hmmm... 2015-03-10T13:54:45 this trivially breaks the casual observers expectation of how to use pthreads. 2015-03-10T13:54:55 and makes it harder to port code to rtems 2015-03-10T13:56:53 then move the header 2015-03-10T13:57:00 move *and* rename 2015-03-10T13:57:18 naming an internal header the same as a well known header that's installed, essentially everywhere is a very bad idea 2015-03-10T13:57:44 if it's public though it will start with rtems/.. no matter what 2015-03-10T13:57:58 it's the only header in posix/ i think 2015-03-10T13:58:44 oh, no it's not i was thinking of another header 2015-03-10T14:08:45 *** Hesham has joined #rtems 2015-03-10T14:09:33 ohhh, that pthread.h 2015-03-10T14:09:44 :q 2015-03-10T14:10:34 that is included in #include currently, so that part doesn't chagne, i see. 2015-03-10T14:11:28 not sure what code you had that was picking it up... 2015-03-10T14:11:55 anywho. 2015-03-10T14:14:12 https://git.rtems.org/amar/waf.git/commit/?h=fix 2015-03-10T14:14:26 those 2015-03-10T14:14:31 that's probably not all of them, either 2015-03-10T14:20:21 that's weird, I installed python2.7-dev and I still got the same error 2015-03-10T14:20:44 I ll paste the error right away 2015-03-10T14:21:01 checking for python2.7... no 2015-03-10T14:21:01 configure: error: python is missing or unusable 2015-03-10T14:21:43 shell cmd failed: /bin/sh -ex /home/andrei/rtems-source-builder/rtems/build/sparc-rtems4.11-gdb-7.8.2-x86_64-linux-gnu-1/doit 2015-03-10T14:24:40 *** siddharth7 has quit IRC (Read error: Connection reset by peer) 2015-03-10T14:25:08 also, is there a way to not restart the whole build process 2015-03-10T14:29:29 andrei_: it always starts fresh because there is no sane way to ensure the environment is exactly the same 2015-03-10T14:30:10 umm i think kiwichris posted recently about python issues...hmm 2015-03-10T14:30:16 gedare: makes sense. That error is a common gdb error, but people on other debian based distro's solved it by installing python2.7-dev package 2015-03-10T14:30:42 which I did and it failed, now I'm trying with an all version compatible package to see if it works 2015-03-10T14:30:45 oh that was in the win build. 2015-03-10T14:30:51 oh 2015-03-10T14:30:59 i think rsb looks for a very specific python 2015-03-10T14:31:08 not necessarily for any good reason... 2015-03-10T14:31:45 the log says checking for python... /usr/bin/python 2015-03-10T14:32:36 https://devel.rtems.org/ticket/2275 might have relevant background info 2015-03-10T14:32:43 and what version is that for you? 2015-03-10T14:33:17 $ python -V 2015-03-10T14:33:17 python: /usr/local/lib/libz.so.1: no version information available (required by python) 2015-03-10T14:33:18 Python 2.7.6 2015-03-10T14:33:45 I keep getting error that my libz is newer than the one used to linked some stuff 2015-03-10T14:33:47 *** siddharth7 has joined #rtems 2015-03-10T14:33:59 but I believe it wouldn't cause an issue as my python behavior is normal 2015-03-10T14:35:31 the issue is similar to what I found on stack overflow 2015-03-10T14:35:39 and Chris Jones pointed to the same solution (install dev package) 2015-03-10T14:35:44 in my case python2.7-dev 2015-03-10T14:36:32 *** Hesham has quit IRC (Ping timeout: 245 seconds) 2015-03-10T14:40:27 [trac] Projects/Open/SimulatorUpdates edited by chrisj (empty comment) http://devel.rtems.org/wiki/Projects/Open/SimulatorUpdates 2015-03-10T14:43:40 *** acenos has joined #rtems 2015-03-10T14:44:41 Hello 2015-03-10T14:47:42 hi acenos 2015-03-10T14:47:44 I have some doubts about the GSoC, specifically about the Runtime Tracing and/or the CPU Statistic part: someone only I can consult? : ) 2015-03-10T14:47:45 good morning kiwichris 2015-03-10T14:47:51 hey gedare 2015-03-10T14:47:55 kiwichris: any advice for andrei_ ? 2015-03-10T14:48:15 acenos: what are your doubts? 2015-03-10T14:49:01 hmmm, i see andrei_. 2015-03-10T14:50:03 I has been toying with the cpu_usage to see if I can make a proposal, but all the test I make give me "0 seconds" of usage. So I think I am making some stupid mistake 2015-03-10T14:50:05 DrJoel: if the rtems_clock_get and task variables patches look good i'll push 'em all and deprecate the lot today. 2015-03-10T14:50:32 acenos: some possibilities are that the tests are not running long enough, or that your target is not accounting for time accurately. 2015-03-10T14:50:37 *** namaste has quit IRC (Ping timeout: 246 seconds) 2015-03-10T14:50:43 some simulators may "cheat" time. 2015-03-10T14:51:10 and if you use a sleep()-like call, you might not get any results either if idle is not being accounted for. 2015-03-10T14:51:24 I thought about that, and made a long stupid program writting randomly in a vector. The simulation takes 1 minute, but still reporting 0 seconds 2015-03-10T14:51:29 now, i think there is a shell cmd that will print a bunch of info including cpu usage stats including for idle 2015-03-10T14:51:38 ok. what is your simulator? 2015-03-10T14:51:44 sparc/sis? 2015-03-10T14:51:56 aye, I used the Hello World tutorial 2015-03-10T14:52:09 and added the rtems_cpu_usage_report(); call 2015-03-10T14:53:16 gedare: http://stackoverflow.com/questions/10792844/python-missing-or-unusable-error-while-cross-compiling-gdb this is what I'm looking at 2015-03-10T14:55:41 acenos: sis should give timing results... so i'm not sure 2015-03-10T14:56:36 acenos: did you build with --enable-tests? 2015-03-10T14:56:53 let me check 2015-03-10T14:57:12 acenos: do you have sparc-rtems4.11/c/sis/testsuites/libtests 2015-03-10T14:57:13 in your build directory 2015-03-10T14:57:53 this was the configuration line (copy paste from the tutorial) ../rtems/configure --target=sparc-rtems4.11 --enable-rtemsbsp=sis \ --enable-tests=samples --disable-posix 2015-03-10T14:57:57 hmm let me check 2015-03-10T14:58:34 ok acenos remove the "=samples" (and remove --disable-posix) 2015-03-10T14:58:44 ok 2015-03-10T14:58:47 you'll need to re-build it all 2015-03-10T14:58:53 so you can rm -rf your build tree 2015-03-10T14:58:58 and then rerun configure 2015-03-10T14:59:01 and then make 2015-03-10T14:59:26 so "nuke and rebuild from the scratchs!" ok 2015-03-10T14:59:28 when it finishes, go into sparc-rtems4.11/c/sis/testsuites/libtests/cpuuse and run cpuuse.exe 2015-03-10T15:00:07 which comes from the code in rtems/testsuites/libtests/cpuuse 2015-03-10T15:01:25 andrei_: ok looks like a pretty flaky problem 2015-03-10T15:02:38 I looked in config.log 2015-03-10T15:02:43 and it seems like a zlib issue 2015-03-10T15:02:44 libpython2.7.so: undefined reference to `inflatecopy@zlib_1.2.0' 2015-03-10T15:03:18 andrei_: is it possible to use python 2.6.6 instead? 2015-03-10T15:03:25 hmm. 2015-03-10T15:03:33 you have zlib-devel installs? 2015-03-10T15:05:38 let me check 2015-03-10T15:06:36 it's a compression library, right? 2015-03-10T15:07:18 apparently, I didn't. 2015-03-10T15:07:30 yes it is 2015-03-10T15:07:49 let's see if it works now 2015-03-10T15:10:55 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-10T15:19:10 [trac] TBR/UserManual/4.11_Release_Notes edited by gedare (Add deprecations.) http://devel.rtems.org/wiki/TBR/UserManual/4.11_Release_Notes 2015-03-10T15:20:41 [trac] #2293/defect (closed) updated by Gedare Bloom (In [changeset:"cf4045630e1dd59221c67005ca1652b4350ad8eb/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2293 2015-03-10T15:20:41 [git] 3ac6811 by Gedare Bloom: [cpukit: deprecate notepads Deprecate Classic API Notepads. Mark task_set/get_note() with the deprecated attribute, and also mark the notepads field. Replace disable with enable option for notepads in confdefs.h, and make notepads disabled by default. The previous option CONFIGURE_DISABLE_CLASSIC_API_NOTEPADS is now unused and 2015-03-10T15:20:41 will emit a compile-time wa...] http://devel.rtems.org/changeset/3ac6811/rtems 2015-03-10T15:20:41 [git] 23e43f6 by Gedare Bloom: [cpukit: deprecate rtems_clock_get(). closes #2294.] http://devel.rtems.org/changeset/23e43f6/rtems 2015-03-10T15:20:41 [git] cf40456 by Gedare Bloom: [cpukit: deprecate task variables. closes #2293.] http://devel.rtems.org/changeset/cf40456/rtems 2015-03-10T15:22:08 hey verm__ thoes last two commits' "closes" commands didn't get picked up by trac. probably because they were in the first line of the commit msg? 2015-03-10T15:22:45 gedare: do you happen to know which fail I can tail to see where the build is? 2015-03-10T15:22:53 er, nvm.... 2015-03-10T15:22:54 damn 2015-03-10T15:22:55 which file 2015-03-10T15:23:42 well verm__ trac didn't close 2294 nor 2265. I didn't put the period after 2265 so I guess that was the problem there... 2015-03-10T15:24:09 For 2294, i don't know why, since it closed 2293 and the commit messages are almost identical... 2015-03-10T15:24:24 strange i'll setup a debug environment and figure out why 2015-03-10T15:24:30 maybe the () in the commit message threw it off. 2015-03-10T15:24:33 this has been an ongoing thing with trac i've had issues elsewhere 2015-03-10T15:24:39 it shouldn't 2015-03-10T15:24:51 it's either a bug or an issue with how i've set it up 2015-03-10T15:24:54 i think it's the later 2015-03-10T15:24:59 i'll see about making it more robust 2015-03-10T15:25:00 that, and underscores are like the only difference in the commit messages. 2015-03-10T15:25:02 ok 2015-03-10T15:26:13 on the plus side, copying the commit hash from the vc is easier now ;) 2015-03-10T15:26:58 [trac] #2294/defect (closed) updated by gedare ([23e43f6af1ba5d8a1947422ff76fbc211a8c8cb0]) http://devel.rtems.org/ticket/2294 2015-03-10T15:27:03 [trac] #2294/defect (closed) updated by Gedare Bloom (In [changeset:"23e43f6af1ba5d8a1947422ff76fbc211a8c8cb0/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2294 2015-03-10T15:27:15 it's my script 2015-03-10T15:27:26 [trac] #2293/defect (closed) updated by Gedare Bloom (In [changeset:"cf4045630e1dd59221c67005ca1652b4350ad8eb/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2293 2015-03-10T15:27:32 [trac] #2265/enhancement (closed) updated by gedare ([3ac681191ec76f7c0a876f2f2cc33adad9a99cc0]) http://devel.rtems.org/ticket/2265 2015-03-10T15:27:33 because if i do it manually it works 2015-03-10T15:28:22 ok do you want me to reopen the ticket? 2015-03-10T15:31:22 oh nvm it is already reopend. 2015-03-10T15:31:43 "I'd like to buy a vowel" 2015-03-10T15:32:01 chose the a 2015-03-10T15:32:01 i know why it's happening now 2015-03-10T15:32:08 no idea how to fix it 2015-03-10T15:32:12 acenos: did your build work? 2015-03-10T15:32:19 andrei_: did your build work? 2015-03-10T15:32:21 i'll probably have to write a daemon to handle it 2015-03-10T15:32:26 * ita waves 2015-03-10T15:32:32 or a trac plugin 2015-03-10T15:32:34 ita: hey 2015-03-10T15:32:40 gedare: it's still building, so I hope so 2015-03-10T15:32:48 gedare: I am on it, I patched some problem. For example testsuites/libtests/dl01 was not compiling so I took it away 2015-03-10T15:33:11 acenos: yes that is a known problem. you can just go into libtests directly and 'make' there. 2015-03-10T15:33:16 err, cpuuse rather 2015-03-10T15:35:23 okey, it looks like the cpuuse.exe work... hmm for some reason the timestamp thinks it is back in 1988 2015-03-10T15:35:51 hmm I took it from the makefiles.am and .in 2015-03-10T15:35:53 that's an RTEMS "Epoch" 2015-03-10T15:36:08 oh! 2015-03-10T15:36:26 of sorts. :0 2015-03-10T15:36:26 and... still giving 0 2015-03-10T15:36:43 https://www.rtems.org/node/6 2015-03-10T15:37:16 well i guess you have something to work with now though, since you can look at the cpuuse code 2015-03-10T15:38:13 oh! RTEMS has been around since the 80's interesting 2015-03-10T15:45:16 same error after installing zlib 2015-03-10T15:45:22 and zlib-dev 2015-03-10T15:45:56 :( ok i dno't have any other ideas for you right now andrei_ I guess you should write up a complete report and send to the users mailing list 2015-03-10T15:46:17 gedare: do you know any distro on which you re more or less certain it works? 2015-03-10T15:46:27 I use a VM right now, so I could switch 2015-03-10T15:47:31 *** siddharth7 has quit IRC (Quit: Leaving) 2015-03-10T15:48:52 ahh.. 2015-03-10T15:49:28 if not, I ll try on a fresh ubuntu 2015-03-10T15:49:33 that's worth trying 2015-03-10T15:49:41 you'll probably get decent support from it. 2015-03-10T15:49:47 I use ubuntu, some lts version 2015-03-10T15:50:40 I ll also keep the old VM, if it doesn't work, I ll submit a report. 2015-03-10T15:50:40 acenos: when did you do a pull last? I worked real hard over the weekend and thought every BSP built every test (or disabled the feature) except moxie which has a tool issue 2015-03-10T15:50:57 If dl01 failed, why? I think you need pax version of tar installed. Look at the failure 2015-03-10T15:51:15 on friday 2015-03-10T15:51:38 I pushed after that I am pretty sure. Like Sunday or Monday 2015-03-10T15:51:58 But sparc/sis isn't a failure that I fixed. It should almost always build everything. 2015-03-10T15:52:01 the compilation error was: cannot open dl.tar for reading 2015-03-10T15:52:06 Just cut and past the failure if you don't mind 2015-03-10T15:52:20 DrJoel: it's the same problem i have 2015-03-10T15:52:55 well.. cut and paste it for you. :) 2015-03-10T15:53:51 i pushed the deprecations 2015-03-10T15:54:30 and now it is time to clock-out. 2015-03-10T15:54:50 I commented the lines where it was mentioned in the Makefile.in and am, and it continue compiling 2015-03-10T15:56:22 *** andrei_ has quit IRC (Remote host closed the connection) 2015-03-10T15:57:18 Thanks gedare. 2015-03-10T15:57:27 If you have the error handy, please paste it 2015-03-10T15:58:40 let me a second to reproduce it 2015-03-10T15:59:28 *** gedare has quit IRC (Ping timeout: 252 seconds) 2015-03-10T16:01:01 Making all in dl01 2015-03-10T16:01:01 gmake[6]: se ingresa al directorio `/home/acenos/development/b-sis/sparc-rtems4.11/c/sis/testsuites/libtests/dl01' 2015-03-10T16:01:01 w -f dl.tar dl-o1.o 2015-03-10T16:01:01 21:59:15 up 13:02, 10 users, load average: 0,00, 0,02, 0,18 2015-03-10T16:01:01 USER TTY FROM LOGIN@ IDLE JCPU PCPU WHAT 2015-03-10T16:01:02 ../../../../../../tools/build/rtems-bin2c -C dl.tar dl-tar.c 2015-03-10T16:01:04 cannot open dl.tar for reading 2015-03-10T16:01:06 gmake[6]: *** [dl-tar.c] Error 1 2015-03-10T16:01:19 sorry for the spanish 2015-03-10T16:02:36 Using the configuration in https://devel.rtems.org/wiki/TBR/UserManual/Quick_Start, it does not appear 2015-03-10T16:02:52 Using this one: ../rtems/configure --target=sparc-rtems4.11 --enable-rtemsbsp=sis --enable-tests , it does 2015-03-10T16:03:02 ../../../../../../tools/build/rtems-bin2c -C dl.tar dl-tar.c 2015-03-10T16:03:02 ../../../../../../tools/build/rtems-bin2c -H dl.tar dl-tar.h 2015-03-10T16:03:18 /usr/bin/pax -w -f dl.tar dl-o1.o dl-o2.o 2015-03-10T16:03:24 that was the list that failed for you 2015-03-10T16:05:51 if you have a build log, search for "checking for pax" and let me know what it found 2015-03-10T16:07:33 DrJoel, do you mean in the output of the make? 2015-03-10T16:07:53 may be in that. Since the configure process continues during make 2015-03-10T16:08:16 My build log has this: 2015-03-10T16:08:17 checking whether ln -s works... yes 2015-03-10T16:08:17 checking for pax... /usr/bin/pax 2015-03-10T16:09:07 ok, let me do a clean and recompile, I will save the log and the err 2015-03-10T16:09:35 Obviously installing pax will fix this but I am filing a ticket so we can decide what to really do 2015-03-10T16:10:53 [trac] #2295/defect (v:4.11) created by joel.sherrill (dl01/dl02 fail build when pax not installed) http://devel.rtems.org/ticket/2295 2015-03-10T16:15:45 I dont find anything about the "pax" nor during the configuration nor the make 2015-03-10T16:16:36 Hmm.. do you have pax installed? That's what is invoked .. notice the pax command I cut and pasted 2015-03-10T16:17:30 /usr/bin/pax11publish ? 2015-03-10T16:18:26 That's a new one on me. : 2015-03-10T16:18:27 :) 2015-03-10T16:19:08 to be sincere I dont know what is the pax... aside of the roman pax : ) 2015-03-10T16:19:28 No reason any human should know one beyond that LOL 2015-03-10T16:19:40 It is an open source standards compliant tar implementation 2015-03-10T16:20:02 oh! I will go to sleep a bit more wiser 2015-03-10T16:20:10 wise* 2015-03-10T16:36:15 well, thanks DrJoel and good night : ) 2015-03-10T16:36:51 good night.. you must be in Europe .. at least east of US. 2015-03-10T16:37:25 aye, Spain : ) 2015-03-10T16:37:27 *** acenos has quit IRC (Quit: Saliendo) 2015-03-10T16:58:07 *** ita is now known as itazzzz 2015-03-10T17:22:36 *** andrei_ has joined #rtems 2015-03-10T17:23:53 hmm, tried to install on a freshly installed ubuntu 14.12 2015-03-10T17:23:54 same issue 2015-03-10T17:24:21 and there wasn't this issue around there /usr/local/lib/libz.so.1: no version information available (required by python) 2015-03-10T17:26:02 http://stackoverflow.com/questions/11961400/getting-rid-of-git-usr-local-lib-libz-so-1-no-version-information-available 2015-03-10T17:26:05 might help 2015-03-10T17:27:54 hm, I tried that, zlib1g and zlib1g-dev are installed on my Mint distro 2015-03-10T17:27:54 I 2015-03-10T17:28:45 on the ubuntu 14.12 which I freshly installed on a vm, I don t get that error 2015-03-10T17:28:47 well my google-fu fails 2015-03-10T17:28:55 but I still do get checking for python2.7... no 2015-03-10T17:28:55 configure: error: python is missing or unusable 2015-03-10T17:29:28 I m currently running a grep for "checking for python2.7" 2015-03-10T17:29:31 to see what it looks for 2015-03-10T17:29:38 Can you see in the code what is being looked for and check by hand? OK you are doing that. :) 2015-03-10T17:32:40 after I get done with the hello world task 2015-03-10T17:32:54 ah, nvm 2015-03-10T17:33:04 I think you already have. I was thinking of setting up a VM with rtems already installed 2015-03-10T17:33:17 but I remembered seeing it on the wiki 2015-03-10T17:36:06 The VM referenced on the wiki is way out of date and if you saw that, I need to remove it. :( 2015-03-10T17:36:18 ah, so it would be useful to make one? 2015-03-10T17:37:07 They are huge to share and seemed to be regualrly over 3GB the last few times I tried to make one to upload. Easier to just let people build their own and then build the tools. Downloading virtualbox and an ISO is much smaller 2015-03-10T17:38:55 fair enough 2015-03-10T17:45:57 *** kiwichris has quit IRC () 2015-03-10T17:48:53 DrJoel: we have plenty of bandwidth 2015-03-10T17:49:00 we're perfectly fine for hosting files of any size even if it's 150gb 2015-03-10T17:49:14 we can do files 1-2TB if we need to 2015-03-10T17:50:00 OK. I am still leaning to making people build their own VMs if they want. The RSB makes the setup so easy. :) 2015-03-10T17:50:25 hmm, DrJoel, this is weird 2015-03-10T17:50:36 apparently, after I installed python2.7-dev on Ubuntu as well 2015-03-10T17:50:37 it worked 2015-03-10T17:50:38 FWIW I have added Doxygen comments to a couple of files and am working on confdefs.h. 2015-03-10T17:55:12 ok, so I got RSB installed 2015-03-10T17:55:43 the next step would be to git clone examples-v2 2015-03-10T17:55:44 ? 2015-03-10T17:59:03 You haven't built rtems. clone, bootstrap, and build rtems 2015-03-10T17:59:14 fyi put the RTEMS tools/bin at the head of your PATH 2015-03-10T18:03:02 if you mean /development/rtems/4.11/bin I already did, cloning atm 2015-03-10T18:03:27 that path is relative to my clone directory ,it's not under / 2015-03-10T18:04:19 thanks 2015-03-10T18:08:47 OK. 2015-03-10T18:08:51 tiome for me to head home 2015-03-10T18:08:52 night 2015-03-10T18:08:55 *** DrJoel has quit IRC (Quit: Few women admit their age. Few men act theirs.) 2015-03-10T18:13:09 *** Davidbrcz has quit IRC (Ping timeout: 245 seconds) 2015-03-10T18:17:43 *** andrei_ has quit IRC (Remote host closed the connection) 2015-03-10T18:18:51 *** kiwichris has joined #rtems 2015-03-10T18:19:54 [trac] #2295/defect (assigned) updated by chrisj (empty comment) http://devel.rtems.org/ticket/2295 2015-03-10T18:40:01 [trac] Developer/Simulators/QEMU edited by chrisj (Moved and updated.) http://devel.rtems.org/wiki/Developer/Simulators/QEMU 2015-03-10T18:46:33 [trac] Developer/Simulators/Qemu/Preparation edited by chrisj (Moved and updated.) http://devel.rtems.org/wiki/Developer/Simulators/Qemu/Preparation 2015-03-10T18:49:30 [trac] Developer/Simulators/QEMU/CAN Emulation edited by chrisj (Update TOC path.) http://devel.rtems.org/wiki/Developer/Simulators/QEMU/CAN Emulation 2015-03-10T19:07:58 [trac] Developer/Simulators/QEMU/CAN Emulation edited by chrisj (Fix formatting.) http://devel.rtems.org/wiki/Developer/Simulators/QEMU/CAN Emulation 2015-03-10T19:08:30 [trac] Developer/Simulators/QEMU/CAN Emulation edited by chrisj (empty comment) http://devel.rtems.org/wiki/Developer/Simulators/QEMU/CAN Emulation 2015-03-10T19:09:25 [trac] Developer/Simulators/QEMU/CANEmulation edited by chrisj (Fix TOC.) http://devel.rtems.org/wiki/Developer/Simulators/QEMU/CANEmulation 2015-03-10T19:10:50 [trac] Developer/Simulators/QEMU edited by chrisj (Fix links.) http://devel.rtems.org/wiki/Developer/Simulators/QEMU 2015-03-10T21:12:32 *** fds has joined #rtems 2015-03-10T21:50:45 *** dr__house|away is now known as dr__house 2015-03-10T22:11:04 *** antgreen has joined #rtems 2015-03-11T00:05:40 *** siddharth7 has joined #rtems 2015-03-11T00:08:57 *** dr__house is now known as dr__house|away 2015-03-11T00:32:50 *** siddharth7 has quit IRC (Quit: Leaving) 2015-03-11T00:44:13 *** richashi has joined #rtems 2015-03-11T00:47:17 Hi, I sent out a mail on the mailing list regarding my interest in the project Run Time Tracing project in GSOC 2015. I wanted to ask whom I can contact for discussing more about this project. Thanks 2015-03-11T01:04:13 *** zoso has joined #rtems 2015-03-11T01:11:46 *** richashi has quit IRC (Quit: richashi) 2015-03-11T01:25:21 *** monstr has joined #rtems 2015-03-11T01:36:05 *** richashi has joined #rtems 2015-03-11T01:40:22 *** namaste has joined #rtems 2015-03-11T01:41:25 Yay, I successfully altered the Hello World program. :-) 2015-03-11T01:45:52 *** namaste has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-11T02:15:41 *** kiwichris has quit IRC () 2015-03-11T02:59:58 *** zoso has quit IRC (Read error: Connection reset by peer) 2015-03-11T03:01:08 *** sebhub has joined #rtems 2015-03-11T03:27:30 good morning 2015-03-11T03:37:47 *** siddharth7 has joined #rtems 2015-03-11T03:56:32 *** sebhub has quit IRC (Ping timeout: 245 seconds) 2015-03-11T04:15:22 *** richashi has quit IRC (Read error: Connection reset by peer) 2015-03-11T04:53:51 *** andrei_ has joined #rtems 2015-03-11T05:08:07 Hello ! 2015-03-11T05:57:05 *** siddharth7 has quit IRC (Quit: Leaving) 2015-03-11T06:05:23 *** avenger_ has joined #rtems 2015-03-11T06:13:59 beng-nl : there? 2015-03-11T06:20:40 *** avenger_ has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-11T06:21:03 *** ragunath has joined #rtems 2015-03-11T06:21:17 *** avenger_ has joined #rtems 2015-03-11T06:33:38 I got error while doing cross compilation 2015-03-11T06:33:46 can anyone help me out? 2015-03-11T07:06:25 beng-nl: hi ben 2015-03-11T07:11:50 *** avenger_ has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-11T07:25:07 *** Hesham has joined #rtems 2015-03-11T08:02:46 *** gedare has joined #rtems 2015-03-11T08:02:46 *** ChanServ sets mode: +o gedare 2015-03-11T08:38:12 andrei_: thanks for reporting to the ml 2015-03-11T08:57:20 *** Davidbrcz has joined #rtems 2015-03-11T09:07:52 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-11T09:14:04 gedare: no problem, any suggestions on how to continue? If not, I'll read about condition variables in general, how they work, how are they implemented and such 2015-03-11T09:22:23 that's a good start. get all the background info / theory. 2015-03-11T09:22:31 look at the posix spec and implementation too 2015-03-11T09:22:53 and search for references in the mailing list archives.. 2015-03-11T09:27:43 *** DrJoel has joined #rtems 2015-03-11T09:27:43 *** DrJoel has joined #rtems 2015-03-11T09:27:43 *** ChanServ sets mode: +o DrJoel 2015-03-11T09:49:54 *** ragunath has joined #rtems 2015-03-11T09:50:44 ragunath: hi 2015-03-11T09:53:05 ran the tests for beagleboard on qemu 2015-03-11T09:53:32 beng-nl: only one failed ... minimum.exe .. reason being not having end of test line 2015-03-11T09:54:11 beng-nl: there were some timeouts spintrcritical tests... they ran fast on the hardware 2015-03-11T09:55:35 hmm 2015-03-11T09:57:08 to get started... tried to get rtc driver working for BBB 2015-03-11T09:57:54 They spin waiting to synchronize with a clock tick edge. Consume CPU time watching ticks since boot. I can see them being very slow on qemu 2015-03-11T09:58:28 beng-nl: i have sent a patch to the mailing lists... if possible pls take a look and give your comments 2015-03-11T10:01:32 ragunath: in general we prefer to get patches submitted by 'git send-email' command. but attaching a patch like you did is fine for now. 2015-03-11T10:02:15 gedare: okay .... will use it from now on 2015-03-11T10:02:35 there are instructions on the wiki git for users page 2015-03-11T10:40:45 *** idleguy has joined #rtems 2015-03-11T10:44:49 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-11T10:50:00 *** siddharth7 has joined #rtems 2015-03-11T11:00:32 *** siddharth7 has quit IRC (Read error: Connection reset by peer) 2015-03-11T11:09:33 verm__: I fixed all the Doxygen warnings in three files. 2015-03-11T11:21:51 *** idleguy has joined #rtems 2015-03-11T11:22:03 *** idleguy has left #rtems 2015-03-11T11:33:24 *** siddharth7 has joined #rtems 2015-03-11T11:33:48 *** Davidbrcz has quit IRC (Quit: Leaving) 2015-03-11T11:38:18 *** andrei_ has quit IRC (Ping timeout: 252 seconds) 2015-03-11T12:00:59 *** ragunath has joined #rtems 2015-03-11T12:01:36 lost connectivity.. campus proxy server has gone down.... 2015-03-11T12:01:59 you didn't miss anything. :) 2015-03-11T12:02:51 gedare: cool :) 2015-03-11T12:03:13 gedare I spot checked powerpc/psim after doing the Doxygen comments and psxualarm fails at line 95 of init.c. Not really sure what is going on there. 2015-03-11T12:03:48 sleep(10) is returning a value of 1000 and stepping through the underlying nanosleep(), it looks like the end time may be less than the start time for the timer. 2015-03-11T12:06:50 hm. 2015-03-11T12:07:24 maybe because of the modification to __times()? 2015-03-11T12:08:03 no that shouldn't matter... 2015-03-11T12:08:11 I don't think it would be related. The ticks math and conversion is nanosleep.c 2015-03-11T12:09:12 I suspect the test is firing the timer before it is completely initialized. So start_time is > end_time. Or it isn't being initialized at all. Like the insert is aborted. 2015-03-11T12:10:29 the return of 1000 seems wierd 2015-03-11T12:11:09 Yep 2015-03-11T12:11:25 Tracing it looks like _Thread_Delay_ended may be firing while we are blocking 2015-03-11T12:13:37 Value it somewhat consistent 2015-03-11T12:13:38 Breakpoint 5, nanosleep (rqtp=rqtp@entry=0x3e108, rmtp=rmtp@entry=0x3e108) 2015-03-11T12:13:38 at ../../../../../../rtems/c/src/../../cpukit/posix/src/nanosleep.c:99 2015-03-11T12:13:38 99 _Timespec_From_ticks( ticks, rmtp ); 2015-03-11T12:13:38 (gdb) p executing->Timer 2015-03-11T12:13:38 $4 = {Node = {next = 0x2aa24 <_POSIX_signals_Ualarm_timer>, 2015-03-11T12:13:39 previous = 0x2c3f4 <_Watchdog_Ticks_chain>}, state = WATCHDOG_INACTIVE, 2015-03-11T12:13:39 initial = 1000, delta_interval = 0, start_time = 300, stop_time = 1304, 2015-03-11T12:13:40 routine = 0x14808 <_Thread_Delay_ended>, id = 0, user_data = 0x32a30} 2015-03-11T12:13:40 (gdb) p _Watchdog_Ticks_since_boot 2015-03-11T12:13:41 $5 = 1304 2015-03-11T12:13:56 [trac] GSoC/2015 edited by ragunath (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-03-11T12:14:00 1004 ticks from start to end 2015-03-11T12:14:30 Converts to 10 seconds. 2015-03-11T12:16:51 Very weird. 2015-03-11T12:17:09 unsigned sleep(unsigned seconds) 2015-03-11T12:17:09 { 2015-03-11T12:17:09 struct timespec ts; 2015-03-11T12:17:09 ts.tv_sec = seconds; 2015-03-11T12:17:09 ts.tv_nsec = 0; 2015-03-11T12:17:10 if (!nanosleep(&ts,&ts)) return 0; 2015-03-11T12:17:10 if (errno == EINTR) return ts.tv_sec; 2015-03-11T12:17:11 return -1; 2015-03-11T12:17:11 } 2015-03-11T12:17:14 yeah i know 2015-03-11T12:17:22 nanosleep() returns -1 and EINTR 2015-03-11T12:17:34 but ts isn't filled in correctly when I look at it gdb 2015-03-11T12:17:34 hmm 2015-03-11T12:17:51 ts.tv_sec should be 10... 2015-03-11T12:17:55 why is it 1000 2015-03-11T12:18:02 (gdb) p ts 2015-03-11T12:18:02 $8 = {tv_sec = 4294, tv_nsec = 927296000} 2015-03-11T12:18:09 .... 2015-03-11T12:18:16 The timespec in nanosleep is correct 2015-03-11T12:18:30 The timespec passed in via sleep isn't 2015-03-11T12:19:27 I think I will file this and see if Sebastian can try it on real PowerPC HW. Doesn't look right but needs someone to confirm 2015-03-11T12:19:45 Could be simulator issue, compiler issue, ghost, etc 2015-03-11T12:19:46 the ts in sleep() was what you printed? 2015-03-11T12:20:00 that was $8 above.. yes 2015-03-11T12:20:23 (gdb) p *time 2015-03-11T12:20:23 $6 = {tv_sec = 10, tv_nsec = 0} 2015-03-11T12:20:38 and after returning you get the above? 2015-03-11T12:20:39 was what I printed at the end of the _Timespec_From_ticks() conversion 2015-03-11T12:21:03 After returning from nanosleep() into sleep(), that's what I see. 2015-03-11T12:21:27 oh um 2015-03-11T12:21:36 ok i gotcha... 2015-03-11T12:21:47 and 4294 gets printed at the test level. 2015-03-11T12:21:49 i'm finally grokking it 2015-03-11T12:22:10 So that's consistent with the sleep() layer but not the underlying nanosleep() 2015-03-11T12:23:11 what is ticks at nanosleep.c:99? 2015-03-11T12:24:04 and what is it at line 64 2015-03-11T12:25:01 *** DrJoel_ has joined #rtems 2015-03-11T12:25:09 what is ticks at nanosleep.c:99? 2015-03-11T12:25:09 and what is it at line 64 2015-03-11T12:26:15 [trac] #2296/defect (v:4.11) created by joel.sherrill (powerpc psxualarm01 Failure) http://devel.rtems.org/ticket/2296 2015-03-11T12:28:40 Hmmm.. it is underflowing 2015-03-11T12:28:41 (gdb) p ticks 2015-03-11T12:28:41 $5 = 1000 2015-03-11T12:28:41 (gdb) n 2015-03-11T12:28:41 97 ticks -= executing->Timer.stop_time - executing->Timer.start_time; 2015-03-11T12:28:41 (gdb) 2015-03-11T12:28:41 99 _Timespec_From_ticks( ticks, rmtp ); 2015-03-11T12:28:42 (gdb) 2015-03-11T12:28:42 108 if ( ticks ) 2015-03-11T12:28:43 (gdb) p ticks 2015-03-11T12:28:43 $6 = 4294967292 2015-03-11T12:28:44 (gdb) p *tmtp 2015-03-11T12:28:45 No symbol "tmtp" in current context. 2015-03-11T12:29:36 yup. 2015-03-11T12:29:40 [trac] #2296/defect (new) updated by joel.sherrill (Hmmm... maybe an underflow: (gdb) p ticks $5 = 1000 (gdb) n 97 ticks -= ...) http://devel.rtems.org/ticket/2296 2015-03-11T12:30:12 Perhaps computer an intermediate value and if >= ticks, set ticks to 0? 2015-03-11T12:31:02 4294967292 / TOD_MICROSECONDS_PER_SECOND = 4294 2015-03-11T12:31:31 Yep.. subtracted below zero and it is explainable beyond that 2015-03-11T12:32:24 you can do 2015-03-11T12:32:34 yes 2015-03-11T12:32:46 uint32_t elapsed = executing->Timer.stop_time - executing->Timer.start_time; 2015-03-11T12:33:45 if (elapsed < 0 ) { ticks -= elapsed; ... } 2015-03-11T12:33:53 er 2015-03-11T12:33:56 elapsed < ticks 2015-03-11T12:34:52 Also cleaning up the EINTR path. Looks like a bug where it could only happen it the user wanted the time remaining. 2015-03-11T12:37:22 I appear to have fixed it. I am going to run all the sis, psim, and jmr3904 tests and post a patch 2015-03-11T12:37:48 *** dr__house|away is now known as dr__house 2015-03-11T12:37:51 *** DrJoel_ has quit IRC (Read error: Connection reset by peer) 2015-03-11T12:37:55 *** DrJoel has quit IRC (Read error: Connection reset by peer) 2015-03-11T12:38:07 *** DrJoel has joined #rtems 2015-03-11T12:38:07 *** DrJoel has joined #rtems 2015-03-11T12:38:07 *** ChanServ sets mode: +o DrJoel 2015-03-11T12:38:22 This client weirdly gets lost from the task bar and I get a hidden instance. :( 2015-03-11T12:38:41 [trac] #2296/defect (assigned) updated by joel.sherrill (empty comment) http://devel.rtems.org/ticket/2296 2015-03-11T12:38:54 [trac] #2296/defect (assigned) updated by joel.sherrill (I have a pending patch and am testing this.) http://devel.rtems.org/ticket/2296 2015-03-11T12:44:16 DrJoel: nice! 2015-03-11T12:45:11 The bug I just located or that I am about to commit Doxygen fixes for three files? 2015-03-11T12:45:45 * DrJoel says someone else is going to have to help on Doxygen. (1) I don't have enough time to do them all. and (2) I don't know all the files well enough to document them. 2015-03-11T12:46:30 From https://docs.rtems.org/doxygen/count 2015-03-11T12:46:35 (1) is Chris' and old 2015-03-11T12:46:45 (2) is Gaisler and relatively new 2015-03-11T12:46:55 (3) Not mine, not sure who to blame :) 2015-03-11T12:47:00 i don't have time for doxygen presently. i already over-committed to RTEMS this week 2015-03-11T12:47:07 (4) I eliminated this by movign it to BSPs 2015-03-11T12:47:43 Not asking Gedare. Most I would ask you to do is look at very specific files which are fairly low in count 2015-03-11T12:47:50 i'm just telling. :) 2015-03-11T12:48:03 88 rtems/score/atomic.h is the highest one I would even begin to ask you about 2015-03-11T12:48:54 verm__: can we drop libchip from the set we worry about for now? I am WAY more concerned about things in cpukit/ which the RTEMS project is the originator of 2015-03-11T12:49:33 * DrJoel says lunch time (finally) 2015-03-11T12:50:24 [trac] GSoC/2015 edited by andrei (Added myself in Students' proposal table) http://devel.rtems.org/wiki/GSoC/2015 2015-03-11T12:50:29 hear hear for lunch... 2015-03-11T12:51:31 :) 2015-03-11T12:52:01 any feedback from CRC press article? 2015-03-11T12:53:03 nah 2015-03-11T12:53:29 these things tend to drag out a bit. 2015-03-11T12:53:45 i expeced that 2015-03-11T12:53:56 still owe Hesham figures after I eat 2015-03-11T12:53:57 DrJoel: i cut all the gcc build instructions now in getting started 2015-03-11T12:54:01 and put a link to RSB docs 2015-03-11T12:54:03 *** andrei_ has joined #rtems 2015-03-11T12:54:11 i know you want full printed manuals 2015-03-11T12:54:16 my recommendation is to print the RSB doc too 2015-03-11T12:54:28 DrJoel: Thanks :) 2015-03-11T12:54:39 i don't care to duplicate its instructions 2015-03-11T12:54:41 Hello 2015-03-11T12:54:44 hi andrei_ 2015-03-11T12:55:03 I've updated my entry in https://devel.rtems.org/wiki/GSoC/2015 2015-03-11T12:55:06 good 2015-03-11T12:55:23 make your proposal private 2015-03-11T12:55:24 but I was curious about one fact. Should I make my proposal public, or is there a mentor list to "allow" on my gdoc 2015-03-11T12:55:32 there is not a list right now... 2015-03-11T12:55:45 i should perhaps make one 2015-03-11T12:55:55 right now, I just copied your format, so it's not much of an issue 2015-03-11T12:56:00 We won't be duplicating the RSB docs in texinfo 2015-03-11T12:56:14 4.10 needs updating and likely the RPM part stays 2015-03-11T12:56:18 When I submit a proposal, should I add an entry for both ESA and GSoC? 2015-03-11T12:56:20 but 4.11 needs the rpm stuff removed 2015-03-11T12:56:40 Hesham: yes. That gives us flexibility. Apply to both. 2015-03-11T12:56:44 andrei_: there should be a list of people to share it with already 2015-03-11T12:56:48 But know you will not be accepted to 1 for sure. :) 2015-03-11T12:56:59 andrei_: by copying from the template, if you kept the "sharing" settings 2015-03-11T12:57:25 DrJoel: Yeah of course, it doesn't make sense to be accepted at both ;) 2015-03-11T12:57:45 gedare did you update the list? andrei_ it may still need someone added to it based on the topic. Core mentors should already be on it 2015-03-11T12:58:12 the list is in ok shape for now 2015-03-11T12:58:19 i just removed someone from it *ahem* 2015-03-11T13:01:40 *** monstr has quit IRC (Remote host closed the connection) 2015-03-11T13:01:49 uhm, sorry, I'm a bit confused, I'm not sure about what list are you guys talking. I found https://devel.rtems.org/wiki/GSoC/Mentors , is this the list you mentioned? 2015-03-11T13:06:16 andrei_: You've already added an entry here https://devel.rtems.org/wiki/GSoC/2015 2015-03-11T13:06:42 I know, I was asking about what list of mentors' email to add them to my proposal 2015-03-11T13:09:29 andrei_: the mentors should already be shared if you copied the sharing settings from the template. 2015-03-11T13:09:29 Ah I see, just add gedare and DrJoel and they will add potential mentors for your project if you don't already know there e-mails. 2015-03-11T13:15:46 got it, I'll get it sorted soon 2015-03-11T13:22:40 do we have a framework for ethernet driver? any reference ethernet driver to check? 2015-03-11T13:30:37 Hesham figures on their way 2015-03-11T13:30:48 ragunath: no, but there are two useable networking stacks, lwIP and a recent port from freeBSD 2015-03-11T13:30:59 some users reported they have run lwIP on a beagleboard. 2015-03-11T13:31:18 DrJoel: Thanks, I am catching the bus now, will thank you again when I'm home ;) 2015-03-11T13:31:53 Gotta go, bye. 2015-03-11T13:34:27 gedare: which one is expected to be used for new bsp's? 2015-03-11T13:35:34 it depends on the platform size. lwip is appropriate for smaller boards that can't handle the freebsd stack, which needs quite a bit of memory. 2015-03-11T13:36:08 For Beagle, both should be an option. LWIP is supposed to already work on Beagle. 2015-03-11T13:36:37 it would be nice to get the new freebsd stack to work on it, to give a nice testing option 2015-03-11T13:36:44 *** Hesham has quit IRC (Ping timeout: 272 seconds) 2015-03-11T13:36:58 IMO ideally we would split the build into RTEMS + (TCP/IP stack == NONE, old IPV4, new IPV4/V6, or LWIP) as addon. 2015-03-11T13:37:49 Yep. Agreed. LWIP is supposed to work out of the box and FreeBSD is supposed to support the BB so may be tinkering, setup, polishing and ensuring all upsrteam and documented. 2015-03-11T13:39:25 hmmm let me check.... currently which bsp's use lwIP? 2015-03-11T13:39:48 None officially. It is a separate project. 2015-03-11T13:40:29 *** ketul has joined #rtems 2015-03-11T13:40:45 http://savannah.nongnu.org/projects/lwip/ is the main site I think 2015-03-11T13:41:57 If there isn't a driver there, apparently TI supplies one 2015-03-11T13:43:42 DrJoel: i saw this page https://devel.rtems.org/wiki/Packages/LWIP and thought some bsp might use it 2015-03-11T13:45:05 There was a port a few years ago. It is in my ftp directory on rtems.org. The newer one is supposed to work with RTEMS as POSIX threads with no real changes. 2015-03-11T13:45:36 Our idea of a package is something you build using the RSB. Just fetch upstream, apply any needed packages and build for a BSP. That's the idea for it. 2015-03-11T13:45:49 hello I am interested for project on BBB BSPs. 2015-03-11T13:46:27 can I know how to proceed? 2015-03-11T13:48:59 There is a hello world project everyone needs to do to ensure you have a functional RTEMS development environment. 2015-03-11T13:49:26 For the Beagle specifically, you should have one. I am not the right person to speak about the variants and what features are missing in the BSP 2015-03-11T13:49:51 For the purposes of the hello world exercise, if you don't have hardware, just run the RTEMS tests on qemu. 2015-03-11T13:50:20 Look at the description of the project, see what is not supported on the BB and work from there. The goal is to fill in the missing pieces 2015-03-11T13:50:26 beng-nl is the author of the BSP 2015-03-11T13:51:30 ya I read that blog and started cross compilation 2015-03-11T13:51:48 and I do have BBB board with me right now 2015-03-11T13:53:07 but I got an error "cannot open dl.tar for reading" 2015-03-11T13:53:24 DrJoel can you help me out? 2015-03-11T13:54:01 You do not have pax installed on your host OS 2015-03-11T13:54:14 gedare.. did you see the ticket I filed yesterday on this? 2015-03-11T13:54:24 https://devel.rtems.org/ticket/2295 2015-03-11T13:54:52 ok thanks DrJoel 2015-03-11T13:55:00 https://devel.rtems.org/ticket/2295 2015-03-11T13:55:22 The decision in that ticket is to fail the build or skip the tests. 2015-03-11T13:55:52 the workaround is to install pax :) 2015-03-11T13:57:49 Yep. Eventually some other tests would fail as well. 2015-03-11T13:57:50 ./libtests/dl01/Makefile.am 2015-03-11T13:57:50 ./libtests/tar01/Makefile.am 2015-03-11T13:57:50 ./libtests/dl02/Makefile.am 2015-03-11T13:57:50 ./libtests/mghttpd01/Makefile.am 2015-03-11T13:57:50 ./libtests/tar02/Makefile.am 2015-03-11T13:59:08 ya but after pax installation it will be done right? 2015-03-11T13:59:25 build should finish yes 2015-03-11T13:59:51 ok let me check and will revert you back 2015-03-11T14:00:03 :) 2015-03-11T14:06:10 DrJoel: as per this page TI did not provide the driver for BBB http://lwip.wikia.com/wiki/Available_device_drivers 2015-03-11T14:08:53 oh... someone mentioned in a thread that it was in the TI Starterkit 2015-03-11T14:09:37 ragunath: https://lists.rtems.org/pipermail/devel/2014-December/009591.html 2015-03-11T14:11:08 DrJoel: sure 2015-03-11T14:11:25 Email devel@ and cc' Daniel Guston. He may just need a poke. It has been 3 months. I am sure he is busy 2015-03-11T14:12:21 *** ankur has joined #rtems 2015-03-11T14:13:19 DrJoel: i found this thread https://lists.rtems.org/pipermail/devel/2014-November/008965.html .... ben was mentioned in it .. 2015-03-11T14:13:32 ben may have some inputs i guess 2015-03-11T14:13:47 Yep.. beng-nl is always going to be mentioned when BB comes up :) 2015-03-11T14:14:08 Just post to devel@ and cc both of them. An answer will pop out. Daniel has the code though. 2015-03-11T14:14:30 Will "closes #2296." on the last line of a commit message close a ticket? 2015-03-11T14:14:58 DrJoel: ok sure 2015-03-11T14:24:00 DrJoel: i think it should. 2015-03-11T14:24:17 depends on whether verm__'s script decides it wants to or not. flip a coin 2015-03-11T14:24:39 My side of the coin never shows up. :) 2015-03-11T14:25:06 i'm going to push my doc change soon 2015-03-11T14:25:24 it builds and clicks through 2015-03-11T14:28:07 *** Davidbrcz has joined #rtems 2015-03-11T14:29:28 https://docs.rtems.org/doxygen/count <-refresh new count list 2015-03-11T14:29:37 https://docs.rtems.org/doxygen/warnings6.log 2015-03-11T14:29:38 latest warnings 2015-03-11T14:30:02 i removed libchip 2015-03-11T14:30:30 *** Hesham has joined #rtems 2015-03-11T14:30:55 jeez pci.h .. every damn pci constant needs a message. :( 2015-03-11T14:31:16 Makes the 74 in rtems/confdefs.h I got rid of look like nothing 2015-03-11T14:31:50 to be fair each one really should have a description... 2015-03-11T14:32:06 Not arguing on that.. just wow! 2015-03-11T14:32:15 At least that one file has a huge impact :) 2015-03-11T14:32:31 yeah we're down to 3,400! woo 2015-03-11T14:32:47 I am more likely to nibble at the bottom of the count after I commit these. 2015-03-11T14:33:10 yeah reducing the number of files is a good thing 2015-03-11T14:33:47 is there no way for us to parse some of these ids? 2015-03-11T14:33:59 most of these should be pretty obvious what they are 2015-03-11T14:34:11 if we can just quickly get the description from the ID for certain blocks 2015-03-11T14:35:28 *** ankur has quit IRC (Ping timeout: 246 seconds) 2015-03-11T14:40:01 DrJoel : got another error "error: conflicting types for 'sysconf' long _EXFUN(sysconf, (int __name )); ^" 2015-03-11T14:40:53 gedare or Hesham: Is that a recent tool change? Like late last week? 2015-03-11T14:41:07 I think you may need to pull rtems-source-builder to update and unfortunately rebuild the tools. 2015-03-11T14:41:14 But let's wait for Hesham and gedare to comment 2015-03-11T14:41:31 I did it just yesterday 2015-03-11T14:41:39 ok 2015-03-11T14:41:48 Is your RTEMS up to date? 2015-03-11T14:42:29 DrJoel: Which tool? RSB? 2015-03-11T14:42:39 My arm gcc was built on 6 March 2015-03-11T14:42:51 ya RTEMS is up to date 2015-03-11T14:42:58 Does his error look familiar? I know there was some sysconf() change 2015-03-11T14:43:07 just cloned yesterday 2015-03-11T14:43:35 was that compiling sysconf.c? 2015-03-11T14:43:36 or1k is built also last week 2015-03-11T14:43:53 shouldn't have been a type error. sebhub added some new options is all 2015-03-11T14:44:02 for probing sysconf for. 2015-03-11T14:44:40 just wait I am giving you pastebin link 2015-03-11T14:46:42 http://pastebin.com/yvBewgVz 2015-03-11T14:49:12 related to '_SC_CLK_TCK' problem would be there. 2015-03-11T14:50:31 the toolchain is using an outdated newlib 2015-03-11T14:50:49 * gedare takes a peek 2015-03-11T14:51:31 ketul: from where did you clone rtems-source-builder? 2015-03-11T14:52:23 https://github.com/bengras/rtems-source-builder.git 2015-03-11T14:52:59 :( wonder when that was last updated. Everything should really come from git.rtems.org 2015-03-11T14:53:00 please clone the upstream version at https://git.rtems.org/rtems-source-builder/ 2015-03-11T14:53:18 *** ankur has joined #rtems 2015-03-11T14:53:28 even the github.com/rtems/rtems-source-builder won't work it is 8 days out of sync 2015-03-11T14:53:45 verm__: ping on github... :( 2015-03-11T14:53:58 i know it's running from buildbot 2015-03-11T14:54:02 ohh 2015-03-11T14:54:41 https://git.rtems.org/rtems-source-builder/ is updated right? 2015-03-11T14:54:48 verm__, I have build rtems on my system, can you guide to the next step? 2015-03-11T14:59:07 gedare : whether is it a problem of variable not declared . If so I can add if you will guide me. 2015-03-11T15:01:13 ketul: use the updated rsb and it should work 2015-03-11T15:02:22 ok I will but that toolchain build, qemu and uboot really takes too much time 2015-03-11T15:02:42 always it happens from starting 2015-03-11T15:10:53 yes um 2015-03-11T15:10:53 that's because there's no reliable way to be sure nothing changed in the user environment, so the safest thing is to start over 2015-03-11T15:10:53 otherwise it is hard to get repeatable results 2015-03-11T15:11:30 And unfortunately, starting from a mirror that is out of date is unfortunate :( 2015-03-11T15:13:24 gotta go .... bye 2015-03-11T15:13:24 *** ragunath has left #rtems 2015-03-11T15:13:24 ok 2015-03-11T15:16:11 in git://git.rtems.org/rtems-source-builder.git I think no build set file for beagle.bset 2015-03-11T15:17:33 gedare: got error "error: no build set file found: devel/beagle.bset" 2015-03-11T15:19:07 ahh.. i think you must build the arm 2015-03-11T15:19:27 it will build the beagle (along with all the other arm) tools. i guess it should work, beng-nl you still around? 2015-03-11T15:23:29 gedare: how can I build the arm. Can you give some input? 2015-03-11T15:26:34 *** ankur has quit IRC (Ping timeout: 246 seconds) 2015-03-11T15:30:37 *** siddharth7 has quit IRC (Quit: Leaving) 2015-03-11T15:31:42 yeah 2015-03-11T15:31:49 replace beagle with arm 2015-03-11T15:32:04 hmm 2015-03-11T15:36:22 gadere: same error "error: no build set file found: devel/arm.bset" 2015-03-11T15:36:51 *gedare 2015-03-11T15:39:36 ketul: use 4.11/rtems-arm 2015-03-11T15:39:52 see also https://docs.rtems.org/rsb/ 2015-03-11T15:40:00 with RTEMS you have to think in terms of architecture and board. 2015-03-11T15:40:06 Arm could be beagle, pi, etc. 2015-03-11T15:41:04 ya I know 2015-03-11T15:41:04 but I just wanted to know the commmand 2015-03-11T15:41:04 [trac] #2291/defect (closed) updated by Gedare Bloom (In [changeset:"d84995d28c141fd6eb0f04606ea35caa316b9eb7/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2291 2015-03-11T15:41:04 [git] d84995d by Gedare Bloom: [doc/started: simplify and fix * fix and remove some macros in rtems.texi.in. * refer to devel mailing list. * remove reference to Debian packaging in requirements section. * remove section on prebuilt tools. * replace toolset build instructions with link to RSB docs. * Add a note in building RTEMS section about using RSB. * Fix 2015-03-11T15:41:04 URLs Closes #2291.] http://devel.rtems.org/changeset/d84995d/rtems 2015-03-11T15:48:30 gedare as I started to say earlier, does any of this need to go onto 4.10? 2015-03-11T15:48:37 Did you remove the RPM docs from 4.11? 2015-03-11T15:49:13 DrJoel: that is in the master. 2015-03-11T15:49:27 i'm not interested in updating 4.10 docs. 2015-03-11T15:49:47 RPMs work there. RSB does too 2015-03-11T15:50:59 [trac] #2297/defect (v:4.11) created by joel.sherrill (sparc dl02 execution failure) http://devel.rtems.org/ticket/2297 2015-03-11T15:51:17 DrJoel: someone interested in monkey on devel list. that's all you. 2015-03-11T15:52:17 [trac] #2298/defect (v:4.11) created by joel.sherrill (sptls01 fails on sis) http://devel.rtems.org/ticket/2298 2015-03-11T15:54:34 bye 2015-03-11T15:56:12 *** Hesham has quit IRC (Ping timeout: 256 seconds) 2015-03-11T15:57:43 *** andrei_ has quit IRC (Ping timeout: 246 seconds) 2015-03-11T15:59:13 *** gedare has quit IRC (Ping timeout: 255 seconds) 2015-03-11T16:15:47 verm__ someone mentioned that the user's manuals don't have figures. They need epstopdf to generate them. Perhaps that wasn't installed on the machine when you ran the build 2015-03-11T16:18:49 i haven't rebuilt the manuals at all 2015-03-11T16:18:51 what is the url? 2015-03-11T16:18:52 [git] bdf23b6 by Joel Sherrill: [cpukit: Remove old DESCRIPTION: in comments These were remnants of pre-Doxygen comment style.] http://devel.rtems.org/changeset/bdf23b6/rtems 2015-03-11T16:18:52 [git] a575af8 by Joel Sherrill: [cpukit/include/rtems/bspIo.h: Add Doxygen comments] http://devel.rtems.org/changeset/a575af8/rtems 2015-03-11T16:18:52 [git] a2fe43b by Joel Sherrill: [cpukit/include/rtems/btimer.h: Add DOxygen comments] http://devel.rtems.org/changeset/a2fe43b/rtems 2015-03-11T16:18:52 [git] 78d5448 by Joel Sherrill: [mvme147s: Install z8530.h This was missed when the file was moved.] http://devel.rtems.org/changeset/78d5448/rtems 2015-03-11T16:18:52 [git] bfe6391 by Joel Sherrill: [confdefs.h: Improve the Doxygen] http://devel.rtems.org/changeset/bfe6391/rtems 2015-03-11T16:21:03 DrJoel: i crontabed github to run every hour i think one of the paths was wrong i fixed it 2015-03-11T16:22:51 great! One less student issue 2015-03-11T16:47:12 [trac] #2299/defect (v:4.11) created by joel.sherrill (DOSFS test failures on v850sim) http://devel.rtems.org/ticket/2299 2015-03-11T16:48:10 * DrJoel is leaving.. night folks 2015-03-11T16:48:13 *** DrJoel has quit IRC (Quit: Clap on! , Clap off! Clap@#&$NO CARRIER) 2015-03-11T16:54:05 [trac] #2299/defect (new) updated by joel.sherrill (empty comment) http://devel.rtems.org/ticket/2299 2015-03-11T17:53:34 *** ketul has quit IRC (Ping timeout: 246 seconds) 2015-03-11T17:54:01 *** kiwichris has joined #rtems 2015-03-11T18:18:36 *** Davidbrcz has quit IRC (Ping timeout: 246 seconds) 2015-03-11T18:57:20 [trac] #2205/defect (new) updated by chrisj (I have taken a look and there must be something specific to your Windows setup we ...) http://devel.rtems.org/ticket/2205 2015-03-11T20:47:29 *** jpc has joined #rtems 2015-03-11T20:47:58 hi everyone 2015-03-11T20:53:02 i guess everybody goes to sleep early here... 2015-03-11T20:57:24 nope 2015-03-11T21:01:01 oh hey! 2015-03-11T21:01:41 i'm guessing most people here are on the east coast, though. 2015-03-11T21:01:55 (of the US) 2015-03-11T21:11:16 [git] 020389f by Chris Johns: [Add MSYS2 documentation.] http://devel.rtems.org/changeset/020389f/rtems-source-builder 2015-03-11T21:11:16 [git] 0f6afd9 by Chris Johns: [4.11/moxie: Fix the broken dtc-1.4.1 build.] http://devel.rtems.org/changeset/0f6afd9/rtems-source-builder 2015-03-11T21:11:58 well, if anybody wants to chat about free (libre) embedded tools, i'm all ears! i'm currently working on setting up embedded toolchains using entirely free tools, and i'm looking for some collaborators to aid in choosing the best tools, platforms, and practices. right now, my toolchain is not purely free/libre, but that's the end-goal. 2015-03-11T21:14:58 for the record, my setup so far: GNU/Linux, OpenOCD, CMSIS-DAP, GCC, GDB, Emacs, RTEMS, Freescale FRDM 2015-03-11T21:17:40 we're all over the world 2015-03-11T21:17:48 europe, us, canada 2015-03-11T21:18:37 of course, sorry if that sounded closed-minded 2015-03-11T21:18:38 jpc: nice 2015-03-11T21:18:45 :-D 2015-03-11T21:19:48 verm__: what's your platform of choice? 2015-03-11T21:19:58 freebsd 2015-03-11T21:20:28 rpm is shaking his head :P 2015-03-11T21:20:34 sorry, rps* 2015-03-11T21:20:38 gah 2015-03-11T21:20:38 chris uses osx, others use various linux flavours and windows 2015-03-11T21:20:40 rms* 2015-03-11T21:20:54 i started using unix about 6 years before linux existed 2015-03-11T21:20:55 jpc, hi 2015-03-11T21:21:04 hello, chris 2015-03-11T21:21:07 i've never used it, never caught on with me 2015-03-11T21:21:12 lol, back in '86? 2015-03-11T21:21:15 yep 2015-03-11T21:21:20 nice 2015-03-11T21:21:26 irix and sunos 2015-03-11T21:21:29 and some hpux 2015-03-11T21:21:41 jpc, I am on MacOS, Windows, Ubuntu, and FreeBSD and all free 2015-03-11T21:21:52 i never really started using it until 1988 though 2015-03-11T21:22:05 well, my experience in historical computing is somewhat limited... although, i did work on building the very first home computer 2015-03-11T21:22:11 heh i have osx and windows here too no linux though 2015-03-11T21:22:20 i really need to get a linux machine 2015-03-11T21:22:29 i should jsut buy one and throw debian on it 2015-03-11T21:22:31 I have clients on Linux, none close to me 2015-03-11T21:22:38 same 2015-03-11T21:22:45 Linux machines that is. 2015-03-11T21:23:05 i have access to quite a few linux machines which is why i've never bothered there are so many different versions 2015-03-11T21:23:12 jpc, we have moved to build from source for tools 2015-03-11T21:23:14 i can ssh to at least 30 of them (different distributions/versions) 2015-03-11T21:23:27 jpc: yeah? vic20? 2015-03-11T21:23:31 yes, i'm very please with rtems-source-builder 2015-03-11T21:23:39 verm__: lol, no the Mark-8 2015-03-11T21:23:40 Great 2015-03-11T21:23:53 look it up. it's the first computer based on the i8008 2015-03-11T21:24:01 well, one of the first 2015-03-11T21:24:05 yep i know the mark-8 2015-03-11T21:24:08 i've seen several of them 2015-03-11T21:24:10 nice! 2015-03-11T21:24:33 i've seen several in shambles... i've almost built one, but i'm having trouble getting it running 2015-03-11T21:24:39 i was thinking commercially available ones when you said home computer 2015-03-11T21:24:55 well, i'm impressed you've heard of the mark-8 2015-03-11T21:25:06 most people just say "huh?" 2015-03-11T21:25:13 verm__, showing your age ? 2015-03-11T21:25:19 hey i'm not that old 2015-03-11T21:25:24 haha or historical savvy 2015-03-11T21:25:29 but i feel better knowing i'm half your age 2015-03-11T21:26:01 are ya'll with OAR? 2015-03-11T21:26:05 well the mark-8 was fairly popular among enthusists at the time 2015-03-11T21:26:07 nope 2015-03-11T21:26:08 verm__, haha what ever makes you happy 2015-03-11T21:26:18 verm__: hardly anyone is 2015-03-11T21:26:18 I work with them as the need arises 2015-03-11T21:26:22 a couple of people 2015-03-11T21:26:49 or actually isn't it just joel? 2015-03-11T21:27:02 No there are others including Jennifier 2015-03-11T21:27:13 i haven't seen her here in a while 2015-03-11T21:27:18 i was talking about irc 2015-03-11T21:27:32 Oh ok 2015-03-11T21:27:45 company/affiliations? 2015-03-11T21:27:47 Maybe she has a life ... 2015-03-11T21:28:21 Me, self employed consulting to people for tools, OS support for realtime embedded 2015-03-11T21:28:41 self employed too, consulting to people for testing systems both hardware and software 2015-03-11T21:28:49 so, kind of an FAE? 2015-03-11T21:29:00 but independent 2015-03-11T21:29:29 FAE? 2015-03-11T21:29:34 We tend to view it as an embedded team member who is on demand 2015-03-11T21:29:37 field applications engineer 2015-03-11T21:30:00 i guess that's industry jargon 2015-03-11T21:30:09 sorry, i'm kind of new to the scene 2015-03-11T21:30:24 oh, no not really 2015-03-11T21:30:58 I find most clients need specific things to complement a team without a full time person plus they want a direct support channel. 2015-03-11T21:31:16 yep that's exactly what i do, too 2015-03-11T21:31:23 esp the 'direct support channel' 2015-03-11T21:31:46 Some clients such as NASA what specific features so they have the "embedded team manage" who works on a specific feature 2015-03-11T21:31:50 also project managment.. i don't do much of that anymore takes too much time 2015-03-11T21:31:58 manage -> memeber 2015-03-11T21:32:23 nice. so "guru" ? 2015-03-11T21:32:33 in a sense. 2015-03-11T21:32:43 same goes for testing.. when you have new software, project or device you need a system setup to test it properly 2015-03-11T21:32:51 it's fairly adhoc 2015-03-11T21:32:56 Hmm I am not a fan of that word, just across the parts the team does not need to be across 2015-03-11T21:33:07 yeah 2015-03-11T21:33:10 agreed 2015-03-11T21:33:53 cool. well, i'm in support, too. i'm with a small company called micrium. 2015-03-11T21:34:28 Hows the POSIX layer going ? 2015-03-11T21:34:44 for tcpip or fs? 2015-03-11T21:34:53 the kernel ? 2015-03-11T21:35:24 ...i'm not aware of any plans for POSIX compliance in the kernel 2015-03-11T21:36:00 oh, i know micrium 2015-03-11T21:36:47 we have a bsd api for uc/tcpip, and there is a POSIX wrapper for our FS module, but i don't know of any plans to support pthreads or sem/mutex functions 2015-03-11T21:37:40 Have you looked at RTEMS's SMP support ? 2015-03-11T21:38:37 i'm aware of it. i'm still getting familiar with the whole build environment. understanding the code is a whole other beast! right now, i'm trying to get my feet wet with a FRDM board. 2015-03-11T21:39:14 i'm actually not quite up-and-running yet 2015-03-11T21:40:04 like i said, i'm trying to use 100% free (libre) tools. i'm not quite there, but that's the end-goal. 2015-03-11T21:40:33 jpc, nice 2015-03-11T21:40:52 :) 2015-03-11T21:41:01 *** jpc has left #rtems ("Ex-Chat") 2015-03-11T21:41:17 portability across hosts is well supported so you get the same on any host 2015-03-11T21:41:34 *** jpc has joined #rtems 2015-03-11T21:41:48 portability across hosts is well supported so you get the same on ny host 2015-03-11T21:41:52 you missed that 2015-03-11T21:42:07 thanks 2015-03-11T21:42:39 i may have missed more before that 2015-03-11T21:42:46 i got kicked right after the :) 2015-03-11T21:44:50 are you guys familiar with OpenJTAG? 2015-03-11T21:46:28 Hmmm not sure, I use OpenOCD 2015-03-11T21:46:53 oh, OpenJTAG is an open-source hardware debugging interface 2015-03-11T21:47:13 i believe it's supposed to compete with J-Link 2015-03-11T21:47:42 No I have not used it 2015-03-11T21:47:54 So it is the pod side of the USB connection ? 2015-03-11T21:48:07 pod? 2015-03-11T21:48:55 oh, you mean the dongle? 2015-03-11T21:49:08 The board etc with the USB connector and the pins for hooking to JTAG 2015-03-11T21:49:08 Yes 2015-03-11T21:49:25 yes, it's the usb->debug interface of the target MCU 2015-03-11T21:50:00 i've never used it but someone at a clients were trying it out 2015-03-11T21:50:18 yeah, they sold a few thousand prototypes 2015-03-11T21:50:27 verm__, been regression testing 4.11/rtems-all in the RSB ... some things have broken 2015-03-11T21:53:48 well guys, i'm signing off. nice talking to you. 2015-03-11T21:57:43 *** jpc has quit IRC (Quit: Ex-Chat) 2015-03-12T00:08:24 *** rtemsLogger has joined #rtems 2015-03-12T00:10:22 *** rtemsLogger` has quit IRC (Ping timeout: 240 seconds) 2015-03-12T00:35:18 *** dr__house is now known as dr__house|away 2015-03-12T01:25:49 *** andrei_ has joined #rtems 2015-03-12T01:33:49 *** monstr has joined #rtems 2015-03-12T01:44:49 [git] 7f7d388 by Chris Johns: [Add gdb-7.8.2 hash.] http://devel.rtems.org/changeset/7f7d388/rtems-source-builder 2015-03-12T01:45:15 *** kiwichris has quit IRC () 2015-03-12T02:22:28 *** ketul has joined #rtems 2015-03-12T02:25:14 *** andrei_ has left #rtems 2015-03-12T02:25:20 *** andrei_ has joined #rtems 2015-03-12T02:45:37 *** andrei_ has quit IRC (Quit: Page closed) 2015-03-12T02:46:10 *** andrei_ has joined #rtems 2015-03-12T03:35:42 hello 2015-03-12T03:35:55 I am trying to implement hello word on BBB' 2015-03-12T03:36:48 but getting stuck 2015-03-12T03:36:55 can anyone available to help? 2015-03-12T03:39:19 hi, ketul, I'm not really an Rtems expert but just say what's wrong, maybe I can help(or anyone else reading this) 2015-03-12T03:42:59 hi anderi_ I have executed dd if=bone_hello.exe-sdcard.img of=/dev/mmcblk0 bs=4096 2015-03-12T03:43:15 and tried to boot on BBB 2015-03-12T03:43:39 but it couldn't 2015-03-12T03:43:57 the 4 usr leds don't blink 2015-03-12T03:46:00 and thats why I am not able to connect it through serial device like USB. 2015-03-12T03:46:54 I have no idea how a black beagleboard works but I m generally familiar with booting from a board 2015-03-12T03:47:05 do you use a uimage? 2015-03-12T03:47:30 I mean, are you booting from an sdcard using an someimage.bin ? 2015-03-12T03:47:36 and some other file (MLO, I think) ? 2015-03-12T03:47:46 ya first I created bone_hello.exe-sdcard.img. 2015-03-12T03:47:56 and then it was written on sd 2015-03-12T03:48:03 then trying to boot it 2015-03-12T03:48:12 *** Hesham has joined #rtems 2015-03-12T03:48:35 http://stackoverflow.com/questions/9823699/beagleboard-erased-nand-flash-and-now-not-booting 2015-03-12T03:48:52 can you check TheCodeArtist's answer, maybe that'll get it working 2015-03-12T03:49:02 ok let me check 2015-03-12T03:54:00 *** siddharth7 has joined #rtems 2015-03-12T03:54:54 my sd-card has MLO,rtems-app.img,u-boot.img,uEnv.txt 2015-03-12T03:55:20 but it doesn't have u-boot.bin 2015-03-12T03:57:07 I'm not sure it needs u-boot.bin 2015-03-12T03:57:18 that depends on how rtems boots, which I m not familiar with 2015-03-12T03:58:08 ok let me check after copying it 2015-03-12T03:58:31 thanks andrei_ 2015-03-12T03:58:32 also, I ran into http://www.shrike-systems.com/beagleboard-xm-beaglebone-black-and-everything-else-rtems-on-the-beagles.html 2015-03-12T03:58:51 sorry, I really don't know much more than that, I have a general idea on how to boot things on some boards 2015-03-12T04:05:51 verm__, are you there? 2015-03-12T04:08:50 andrei_ are you mentor in rtems org? 2015-03-12T04:09:42 ketul: nah, just a prospective student 2015-03-12T04:09:58 ok (y) 2015-03-12T04:14:33 then andrei_ you are working in which flow? 2015-03-12T04:14:59 ketul: I'm sorry, I don't really understand your question 2015-03-12T04:15:08 what do you mean by in which flow? 2015-03-12T04:16:09 I mean you are working on which project? 2015-03-12T04:17:03 ah 2015-03-12T04:17:09 Condition Variables 2015-03-12T04:19:01 ok 2015-03-12T05:05:22 *** kiwichris has joined #rtems 2015-03-12T05:24:12 ketul: I don't know if beng-nl used this way before or not, but in case you can boot linux from your SD card on the board, you may get the RTEMS ELF file (for example hello.exe) in a binary format using objcopy, rename it to kernel.img, and overwrite the corresponding file on the SD card containing linux. This is how RTEMS run on Raspberry (back in 2013 since I last used it). 2015-03-12T05:26:38 [trac] #2121/defect (assigned) updated by chrisj (Should this bug stay open given newlib has been converted to git from CVS ?) http://devel.rtems.org/ticket/2121 2015-03-12T05:28:39 [trac] #2119/enhancement (closed) updated by chrisj (To download just the source and not build anything use '--dry-run --with-download'.) http://devel.rtems.org/ticket/2119 2015-03-12T05:40:11 *** mkhoory has quit IRC (Ping timeout: 252 seconds) 2015-03-12T05:40:12 [trac] #2300/defect (v:4.11) created by chrisj (MIPS Tools fail to build.) http://devel.rtems.org/ticket/2300 2015-03-12T05:41:30 [trac] rsb-report-mips-rtems4.11-gcc-4.9.2-newlib-b75c57950318ed657807f87a4c694e7b8d68fcae-x86_64-freebsd10.1-1.txt.gz () attached to #2300 by chrisj http://devel.rtems.org/ticket/2300 2015-03-12T05:52:27 Hesham: do you have any proper documentation for that.? 2015-03-12T05:53:42 [trac] #2279/defect (assigned) updated by chrisj (I cannot build a MIP set of tools. See #2300.) http://devel.rtems.org/ticket/2279 2015-03-12T05:54:14 so that I can implement in proper way. 2015-03-12T05:54:35 ketul, just dropping by, what are you wanting to do ? 2015-03-12T05:55:42 kiwichris: I have executed dd if=bone_hello.exe-sdcard.img of=/dev/mmcblk0 bs=4096 2015-03-12T05:55:42 and tried to boot on BBB 2015-03-12T05:55:42 but it couldn't 2015-03-12T05:55:42 the 4 usr leds don't blink 2015-03-12T05:55:57 and thats why I am not able to connect it through serial device like USB. 2015-03-12T05:56:58 Been a while since I worked in the BBB, but I seem to remember you need to load the MIO loader that uboot builds. 2015-03-12T06:03:44 ketul, have you looked at Ben Gras's page on using RTEMS on the beagle board ? 2015-03-12T06:04:53 kiwichris: I had a look, it contains a reference to beagle.bset which is not at RSB. 2015-03-12T06:05:12 It is in his repo on github 2015-03-12T06:05:41 Didn't check his repo. 2015-03-12T06:06:02 I thought it was at RSB master 2015-03-12T06:08:26 And, there isn't any reference or hints that beagle.bset is at beng-nl's repo. 2015-03-12T06:08:54 I think Gedare was telling about this issue yesterday 2015-03-12T06:12:16 ketul: You can check this for RPi, and try to imitate for your BBB http://alanstechnotes.blogspot.co.uk/2013/03/running-your-first-rtems-program-on.html But not sure it's applicable as I didn't work with BBB before. 2015-03-12T06:22:19 Hesham: Let me check and will revert back 2015-03-12T06:30:18 *** ragunath has joined #rtems 2015-03-12T06:44:19 *** kiran4399 has joined #rtems 2015-03-12T07:01:54 *** ketul has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-12T07:02:33 *** ketul has joined #rtems 2015-03-12T07:40:14 *** ketul has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-12T07:49:40 *** monstr has quit IRC (Ping timeout: 265 seconds) 2015-03-12T07:50:04 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-12T07:55:18 *** gedare has joined #rtems 2015-03-12T07:55:18 *** ChanServ sets mode: +o gedare 2015-03-12T07:58:20 *** monstr has joined #rtems 2015-03-12T08:07:59 good morning 2015-03-12T08:08:42 hello 2015-03-12T08:09:58 I looked over how condition variables work and how they're implemented 2015-03-12T08:10:06 and I found this mail chain I tracked this mail chain https://lists.rtems.org/pipermail/devel/2014-March/005643.html regarding the Condition Variables 2015-03-12T08:10:13 oops, paste fail. 2015-03-12T08:11:17 I looked over this proposal from last year, https://github.com/cloud-hot/proposal/blob/master/proposal.md 2015-03-12T08:12:05 If it's ok to ask, why wasn't it selected? 2015-03-12T08:12:52 Lack of spots/ priority, or was something wrong with it ? I've never mentored in GSoC so far, but to me it seemed alright, a bit too generic but alright. 2015-03-12T08:21:24 *** kiran4399 has left #rtems 2015-03-12T08:24:03 *** monstr has quit IRC (Remote host closed the connection) 2015-03-12T08:26:40 *** gedare has quit IRC (Remote host closed the connection) 2015-03-12T08:27:25 *** monstr has joined #rtems 2015-03-12T08:27:28 *** gedare has joined #rtems 2015-03-12T08:27:28 *** ChanServ sets mode: +o gedare 2015-03-12T08:28:06 *** ragunath has joined #rtems 2015-03-12T08:31:56 *** antgreen has quit IRC (Ping timeout: 252 seconds) 2015-03-12T08:32:25 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-12T08:35:36 [trac] TBR/UserManual/Quick_Start edited by gedare (Remove link to prebuilt tools.) http://devel.rtems.org/wiki/TBR/UserManual/Quick_Start 2015-03-12T08:47:56 [trac] #2121/defect (closed) updated by gedare (empty comment) http://devel.rtems.org/ticket/2121 2015-03-12T08:50:39 *** andrei_ has quit IRC (Quit: Page closed) 2015-03-12T08:51:22 [trac] #2300/defect (new) updated by gedare (xgcc: error: addsf3: No such file or directory xgcc: error: subsf3: No such file or ...) http://devel.rtems.org/ticket/2300 2015-03-12T09:02:47 [git] d88b0b7 by Daniel Cederman: [posix: Move function to file with matching name mmap was previously in munmap.c and munmap was in mmap.c.] http://devel.rtems.org/changeset/d88b0b7/rtems 2015-03-12T09:02:47 [git] b0f8bb4 by Daniel Cederman: [posix: Return error code if mmap is used] http://devel.rtems.org/changeset/b0f8bb4/rtems 2015-03-12T09:37:34 *** ragunath has joined #rtems 2015-03-12T09:49:25 [trac] #2300/defect (new) updated by gedare (Tools build fine on Ubuntu 14.04. I am unable to reproduce your bug.) http://devel.rtems.org/ticket/2300 2015-03-12T09:50:47 [trac] #2300/defect (new) updated by gedare (Check your gcc version. My host gcc uses 4.8.2) http://devel.rtems.org/ticket/2300 2015-03-12T09:59:32 [trac] Developer/Coding/Conventions edited by gedare (Fix internal link.) http://devel.rtems.org/wiki/Developer/Coding/Conventions 2015-03-12T10:04:35 [trac] Developer/Projects/Obsolete/SimpleScalar edited by gedare (fix toc.) http://devel.rtems.org/wiki/Developer/Projects/Obsolete/SimpleScalar 2015-03-12T10:05:02 [trac] Developer/Projects/Obsolete/Libbsd-port edited by gedare (delete toc.) http://devel.rtems.org/wiki/Developer/Projects/Obsolete/Libbsd-port 2015-03-12T10:05:27 [trac] Developer/Projects/Obsolete/CEXP edited by gedare (delete toc.) http://devel.rtems.org/wiki/Developer/Projects/Obsolete/CEXP 2015-03-12T10:07:08 [trac] Developer/Projects/Obsolete/Canadian_Cross_Compiler edited by gedare (Fix TOC) http://devel.rtems.org/wiki/Developer/Projects/Obsolete/Canadian_Cross_Compiler 2015-03-12T10:07:25 [trac] Developer/Projects/Obsolete/DebianHostedTools edited by gedare (Fix TOC) http://devel.rtems.org/wiki/Developer/Projects/Obsolete/DebianHostedTools 2015-03-12T10:09:59 [trac] Developer/OpenProjects edited by gedare (Fix links to obsolete projects.) http://devel.rtems.org/wiki/Developer/OpenProjects 2015-03-12T10:10:39 verm__: please move https://devel.rtems.org/wiki/Developer/Projects/ConfigGUI into Developer/Projects/Open/ConfigGUI 2015-03-12T10:10:43 and fixup 2015-03-12T10:10:54 same with https://devel.rtems.org/wiki/Developer/Projects/BuildVariables 2015-03-12T10:11:35 there is a 'rename page' button at the bottom 2015-03-12T10:11:42 it's not working? 2015-03-12T10:14:24 [trac] Building/RTEMS edited by gedare (Replace build advice with pointer to RSB.) http://devel.rtems.org/wiki/Building/RTEMS 2015-03-12T10:15:25 [trac] TBR/UserManual/RTEMS_Source_Builder edited by gedare (Fix link to RSB docs) http://devel.rtems.org/wiki/TBR/UserManual/RTEMS_Source_Builder 2015-03-12T10:16:46 [trac] Building/RTEMS edited by gedare (tweak.) http://devel.rtems.org/wiki/Building/RTEMS 2015-03-12T10:21:21 [trac] Developer/Projects/Open/BuildVariables edited by gedare (fix verms page) http://devel.rtems.org/wiki/Developer/Projects/Open/BuildVariables 2015-03-12T10:21:47 [trac] Developer/Projects/Open/ConfigGUI edited by gedare (Fix verm__s page) http://devel.rtems.org/wiki/Developer/Projects/Open/ConfigGUI 2015-03-12T10:22:20 [trac] Developer/OpenProjects edited by gedare (fix verm__s links) http://devel.rtems.org/wiki/Developer/OpenProjects 2015-03-12T10:23:27 i fixed them for you as passive-aggressively as I knew how 2015-03-12T10:32:20 *** monstr has quit IRC (Remote host closed the connection) 2015-03-12T10:44:34 thanks! i didn't know where they went since not all open projects are in /Open/ 2015-03-12T10:44:45 they should be. 2015-03-12T10:45:01 i tried to fix all of them before gsoc 2015-03-12T10:58:49 *** ragunath has left #rtems 2015-03-12T10:59:57 *** ketul has joined #rtems 2015-03-12T11:04:02 beng-nl: are you there? I have some queries regarding to beagle BSP project. 2015-03-12T11:06:26 i'm here 2015-03-12T11:06:40 *** soniadalal has joined #rtems 2015-03-12T11:07:17 *** Davidbrcz has joined #rtems 2015-03-12T11:08:03 beng-nl: I have executed dd if=bone_hello.exe-sdcard.img of=/dev/mmcblk0 bs=4096 and tried to boot on BBB but it couldn't the 4 usr leds don't blink 2015-03-12T11:10:48 Instead of devel/beagle.bset I took 4.11/rtems-arm 2015-03-12T11:10:52 *** ragunath has joined #rtems 2015-03-12T11:12:31 because yesterday devel/beagle.bset didn't work and error. So when I had done chat on IRC some one suggested to change bset and not to clone from github 2015-03-12T11:13:09 hm, the leds might not do much 2015-03-12T11:13:19 no i/o on console at 115k2? 2015-03-12T11:13:53 yep I am not able to login in BBB 2015-03-12T11:14:12 oh it does rely on an internal uboot i think 2015-03-12T11:14:26 ya 2015-03-12T11:14:29 have you tried to force a boot from the sdcard? by pressing the boot button on poweron? 2015-03-12T11:14:57 ya I tried but leds are not even blinking 2015-03-12T11:15:25 eh are they supposed to? 2015-03-12T11:15:31 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-12T11:16:16 *** ragunath has joined #rtems 2015-03-12T11:16:29 ya I think because that means it is booting 2015-03-12T11:19:53 let me test an image on my bbb 2015-03-12T11:19:57 and i'll send it to you 2015-03-12T11:20:02 ok sure 2015-03-12T11:23:48 beng-nl : After writing the image on sd card I do have 4 files named : MLO, rtems-app.img, u-boot.img, uEnv.txt in 33MB volume 2015-03-12T11:23:59 ok good 2015-03-12T11:24:00 so till there all is ok. right? 2015-03-12T11:24:03 yep 2015-03-12T11:24:09 ketul: i have this image i just tested: http://www.shrike-systems.com/bone_hello.exe-sdcard-ok.img.gz 2015-03-12T11:24:16 sha1 is 61cec0cd910ee45f4197d530af94d8ef4e796fa0 2015-03-12T11:25:39 so if that works, it's a build problem; if it doesn't.. a boot/serial/hw etc problem 2015-03-12T11:26:20 ok let me check 2015-03-12T11:26:51 so I should simply copy your image file to my sd right? 2015-03-12T11:27:06 with dd yes 2015-03-12T11:27:17 yep ok 2015-03-12T11:27:19 i.e. just like you did earlier, with of=/dev/mmcblk0 2015-03-12T11:27:26 ya 2015-03-12T11:41:12 *** soniadalal has quit IRC (Ping timeout: 246 seconds) 2015-03-12T11:41:48 *** soniadalal has joined #rtems 2015-03-12T11:44:01 beng-nl: tried on both hw BBB and BeaglexM 2015-03-12T11:44:08 but didn't work 2015-03-12T11:44:33 even the serial port is also not coming like ttyACM0 2015-03-12T11:46:18 and on the same BBB debian was running just a few hours ago. So I think no hw or serial prob is there. 2015-03-12T11:47:54 verify the sha1, and list the commands you used to load it to your BB 2015-03-12T11:52:17 shal is md5 checksum right? 2015-03-12T11:52:23 sha1sum 2015-03-12T11:52:38 oh k 2015-03-12T11:52:40 use sha1sum command on the img 2015-03-12T11:52:46 and make sure it is what he wrote above 2015-03-12T11:53:15 ya 2015-03-12T11:54:34 4de20ab890297f5c51f55d5f040640a0223f8b2e bone_hello.exe-sdcard-ok.img.gz I got that is different 2015-03-12T11:54:41 decomprseed 2015-03-12T11:55:30 61cec0cd910ee45f4197d530af94d8ef4e796fa0 bone_hello.exe-sdcard-ok.img 2015-03-12T11:55:33 ya thats fine 2015-03-12T11:55:41 good.. and how do you load it? 2015-03-12T11:57:46 sudo dd if=bone_hello.exe-sdcard-ok.img of=/dev/sdb bs=4096 2015-03-12T12:02:57 and after that I also see that "8192+0 records in 8192+0 records out 33554432 bytes (34 MB) copied, 1.40011 s, 24.0 MB/s" 2015-03-12T12:03:36 geez 2015-03-12T12:05:03 are you sure the sdcard boot is forced? by holding down the boot button on power(!)up? and are you sure the serial cable for console is i/o? 2015-03-12T12:05:13 (is ok) 2015-03-12T12:05:52 does the same hw (sdcard, cable) work with e.g. linux? 2015-03-12T12:06:34 ya I only changed sd card 2015-03-12T12:06:47 all other things are same 2015-03-12T12:07:44 let me check one more time by running debian on that BBB 2015-03-12T12:07:54 ok, same sdcard 2015-03-12T12:08:00 would be better 2015-03-12T12:08:03 ya 2015-03-12T12:08:08 let me check 2015-03-12T12:08:10 btw any chance the i/o didn't complete to the card? 2015-03-12T12:08:17 i.e. do a sync and wait for a few secs before ejecting 2015-03-12T12:08:27 perhaps try and read it back after removing & inserting 2015-03-12T12:09:12 ok I will 2015-03-12T12:09:14 ok 2015-03-12T12:23:56 * gedare <3s hw 2015-03-12T12:25:34 *** siddharth7 has quit IRC (Read error: Connection reset by peer) 2015-03-12T12:28:40 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-12T12:44:23 beng-nl : tried debian on the same setup (along with sd card and cable) . It worked with debian ( I do get ttyACM0). But didn't work for bone_hello 2015-03-12T12:45:00 even while inserting sd it shows all the 4 files 2015-03-12T12:45:12 but didn't work for BBB. 2015-03-12T12:52:55 *** siddharth7 has joined #rtems 2015-03-12T12:53:36 *** ragunath has joined #rtems 2015-03-12T12:54:02 a quick question, what is waf? 2015-03-12T12:56:53 beng-nl : Are you using Beagle xM? 2015-03-12T13:06:54 siddharth7: https://code.google.com/p/waf/ 2015-03-12T13:07:28 we are in the process of transitioning RTEMS to be built using waf 2015-03-12T13:07:52 gedare, thanks 2015-03-12T13:13:42 gedare & beng-bl: any inputs from your side? 2015-03-12T13:16:43 ketul: i don't have the hw to give it a go myself 2015-03-12T13:17:21 *** siddharth7 has quit IRC (Quit: Leaving) 2015-03-12T13:18:03 i'm pretty sure beng-nl is using BBB. 2015-03-12T13:18:12 *** soniadalal has quit IRC (Read error: Connection reset by peer) 2015-03-12T13:18:25 ok gedare 2015-03-12T13:18:42 *** hashm has joined #rtems 2015-03-12T13:19:20 ketul: can you try netbooting? 2015-03-12T13:19:35 @gerade, i was building rtems on my system 2015-03-12T13:19:35 hashm: Error: "gerade," is not a valid command. 2015-03-12T13:20:08 i am using this(https://docs.rtems.org/rsb/) 2015-03-12T13:20:12 to build it 2015-03-12T13:20:12 hi hashm are you siddharth? 2015-03-12T13:20:35 nevermind. hmm 2015-03-12T13:21:10 ragunath: I think for that one OS should be installed . Right like debian .? 2015-03-12T13:21:16 yeah, actually my xchat stopped working, turned on webchat.freenode 2015-03-12T13:21:38 ok hashm. so what's up? 2015-03-12T13:22:02 i have gone till 2.4 index 2015-03-12T13:22:28 okay 2015-03-12T13:22:34 that step can take a long time. 2015-03-12T13:22:51 yeah 2015-03-12T13:22:53 done with that 2015-03-12T13:23:05 tried this command: $HOME/development/rtems/4.11/bin/sparc-rtems4.11-gcc -v 2015-03-12T13:23:11 its working 2015-03-12T13:23:30 whats next? 2015-03-12T13:23:59 ketul: not necessary ... you need to stop at uboot .... set ipaddr serverip and you can tftp the image 2015-03-12T13:24:45 hashm: continue with https://devel.rtems.org/wiki/GSoC/GettingStarted from "Configure and Build RTEMS for SPARC/SIS" 2015-03-12T13:25:00 Can I have any document related to that ragunath please? 2015-03-12T13:27:11 ketul: i dont have any document .... to access uboot u need console access .. do you have one? 2015-03-12T13:27:23 if you have its pretty straight forward 2015-03-12T13:27:59 hi Davidbrcz how is your porting effort going? 2015-03-12T13:28:12 Hi gedare 2015-03-12T13:28:23 I've have a step forward 2015-03-12T13:28:47 I didn't compile all the libraries I was linking with the -msoft-float 2015-03-12T13:28:57 thus trigering the trap 2015-03-12T13:29:15 great. I meant to tell you, that you should use CONFIGURE_UNLIMITED_OBJECTS in your configuration 2015-03-12T13:29:18 silly mistake but I took me 2 days to figure it ^^ 2015-03-12T13:29:41 I need now to focus on using the network and the network stack 2015-03-12T13:29:47 ahh. 2015-03-12T13:29:53 you are using rtems 4.10? 2015-03-12T13:30:04 maybe that configure option did not exist.. I think I added it during 4.11 2015-03-12T13:30:46 but you might consider using the git master since 4.11 should be released soon. then you can look at the updated freebsd networking stack too as an option. 2015-03-12T13:34:02 [trac] #2296/defect (closed) updated by Joel Sherrill (In [changeset:"d937d364af95fec2e9a7a4af4dca6ac7d23397b2/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2296 2015-03-12T13:34:02 [git] d937d36 by Joel Sherrill: [posix/src/nanosleep.c: Address issue when delay is longer than desired This resulted in the elapsed time going below 0 and an arbitrarily large number returned as the time remaining. closes #2296.] http://devel.rtems.org/changeset/d937d36/rtems 2015-03-12T13:35:49 :gedare, there are lots of warning coming while running ./bootstrap, is that ok? 2015-03-12T13:37:48 [trac] #2299/defect (new) updated by joel.sherrill (I am wondering if this is a broader issue of more tests not working on simulators ...) http://devel.rtems.org/ticket/2299 2015-03-12T13:38:24 hashm: it depends, but probably you should not get warnings.. make sure autoconf/automate are getting picked up from the tools you built with RSB, and not from your host 2015-03-12T13:38:26 which autoconf 2015-03-12T13:38:52 on the command line will tell you where it is being found in your PATH 2015-03-12T13:39:33 does i386 have any ethernet driver working under qemu? 2015-03-12T13:40:14 *** Hesham has quit IRC (Ping timeout: 246 seconds) 2015-03-12T13:40:34 ragunath: yes it should 2015-03-12T13:40:36 *** soniadalal has joined #rtems 2015-03-12T13:40:59 yes it does :) 2015-03-12T13:41:14 i have such a build working on current rtems mainline 2015-03-12T13:41:53 beng-nl: you using the in-tree net stack? 2015-03-12T13:42:11 yes.. i didn't know of any other actually 2015-03-12T13:42:16 i've never(!) run an rtems networking stack. :) 2015-03-12T13:42:26 gedare: i can see drivers 3c509, ne2000, wd8003? which one ? 2015-03-12T13:42:27 :) 2015-03-12T13:42:29 ragunath: -device ne2k_isa 2015-03-12T13:43:00 ragunath: BSP_NE2000_NETWORK_DRIVER_NAME 2015-03-12T13:43:10 beng-nl: oh ok ben.... i was 3c509 .. will try it 2015-03-12T13:43:31 *trying 3c509 2015-03-12T13:43:56 :gedare, done with ./bootstrap but this(../rtems/configure --target=sparc-rtems4.11 --enable-rtemsbsp=sis \ --enable-tests=samples --disable-posix) command gives error 2015-03-12T13:44:00 * gedare thinks up a new game---driver roulette 2015-03-12T13:44:01 ragunath: i then also configured the base addr & irq just in case, setting the same values on the qemu cmdline -device ne2k_isa,irq=$(IRQ),iobase=$(IOBASE) (from my makefile) 2015-03-12T13:44:11 gedare: ha 2015-03-12T13:44:24 gedare: driver roulette drinking game ;-) 2015-03-12T13:44:25 hashm: what is the error? it's possible that the bootstrap with the wrong autotools could be to blame 2015-03-12T13:44:34 i don't drink...much 2015-03-12T13:44:41 bash: ../rtems/configure: No such file or directory 2015-03-12T13:44:43 *** soniadalal has quit IRC (Quit: Leaving) 2015-03-12T13:44:48 bootstrap failed in that case 2015-03-12T13:44:52 gedare: smart 2015-03-12T13:44:57 hashm: what does the command `which autoconf` give you 2015-03-12T13:45:45 bash: ../rtems/configure: No such file or directory 2015-03-12T13:45:59 sorry 2015-03-12T13:46:08 ^this 2015-03-12T13:46:16 i am trying to get a feel of RTEMS networking .... so trying i386 ... 2015-03-12T13:46:36 *** soniadalal has joined #rtems 2015-03-12T13:46:48 beng-nl : can you try this one on your BBB? http://goo.gl/6jJMGG 2015-03-12T13:46:51 it gives this /usr/bin/autoconf 2015-03-12T13:47:16 ketul: ok 2015-03-12T13:47:17 whether it works or not? 2015-03-12T13:47:26 :) 2015-03-12T13:48:04 hashm: you will need to add the install point from RSB to your PATH, like this: 2015-03-12T13:48:08 ketul: i'm doing it now 2015-03-12T13:48:23 ya 2015-03-12T13:48:31 export PATH=$HOME/development/rtems/4.11/bin:$PATH 2015-03-12T13:48:35 ketul: yes, it works 2015-03-12T13:48:59 hashm: if you like you may add that command to your .bashrc file in your home directory so it always gets picked up 2015-03-12T13:49:14 hi soniadalal 2015-03-12T13:49:34 ok 2015-03-12T13:49:47 so i need to the bootstrap thing again? 2015-03-12T13:49:59 hashm: yes. this time you should see no warnings 2015-03-12T13:50:05 beng-nl : ok. thanks Now can you guide me how can I go further? 2015-03-12T13:51:06 ok, let me see 2015-03-12T13:52:31 ketul: w a i t a s e c o n d 2015-03-12T13:52:55 ok 2015-03-12T13:53:05 "I do get ttyACM0" 2015-03-12T13:53:14 do you mean over the bbb's usb port? 2015-03-12T13:54:10 ya 2015-03-12T13:54:15 that won't work 2015-03-12T13:54:21 rtems doesn't talk on that usb port 2015-03-12T13:54:39 ohhh 2015-03-12T13:55:03 then through which port it supposed to? 2015-03-12T13:55:10 A problem identified is a problem half solved. 2015-03-12T13:55:13 *** DrJoel has joined #rtems 2015-03-12T13:55:13 *** DrJoel has joined #rtems 2015-03-12T13:55:13 *** ChanServ sets mode: +o DrJoel 2015-03-12T13:55:31 ketul: serial headers on the device 2015-03-12T13:55:41 *** hashm_ has joined #rtems 2015-03-12T13:56:02 ketul: http://www.armhf.com/wp-content/uploads/2013/06/ftdi_serial_debug.jpg 2015-03-12T13:56:18 that coloured cable is connecting the serial headers to a usb cable 2015-03-12T13:56:22 to a pc 2015-03-12T13:56:35 ketul: so you need a (e.g.) https://www.tigal.com/product/3466 2015-03-12T13:57:00 ok beng-nl 2015-03-12T13:57:07 my friend has it 2015-03-12T13:57:19 * beng-nl just felt like http://img2.wikia.nocookie.net/__cb20081021023219/muppet/images/a/ad/SimonSoundman.jpg 2015-03-12T13:57:19 So I will check it tomorrow 2015-03-12T13:57:22 ketul: excellent 2015-03-12T13:57:28 ketul: then you should see the i/o on the serial 2015-03-12T13:57:34 *** hashm has quit IRC (Ping timeout: 246 seconds) 2015-03-12T13:57:38 hopefully with your own build 2015-03-12T13:58:08 ya sure 2015-03-12T13:59:23 beng-nl : but want to ask why rtems doesn't talk on usb? 2015-03-12T13:59:59 it's just not implemented to act like a usb device.. 2015-03-12T14:00:15 ok 2015-03-12T14:00:30 the console i/o is ready to go, just a uart, so that's what it does 2015-03-12T14:01:20 beng-nl: after that what will be my next task? 2015-03-12T14:01:46 writing! 2015-03-12T14:02:19 to apply for gsoc, you have to finish a 'hello world' assignment, it's linked to from https://devel.rtems.org/wiki/Developer/Projects/Open/ImproveBeagleBSP 2015-03-12T14:07:01 *** soniadalal has quit IRC (Ping timeout: 264 seconds) 2015-03-12T14:07:46 hmm, ./bootstrap worked without error 2015-03-12T14:08:33 [trac] Developer/Projects/Open/ImproveBeagleBSP edited by beng (added more explicit note about the beaglebone black needing a seperate console cable. corrected hello world project link syntax.) http://devel.rtems.org/wiki/Developer/Projects/Open/ImproveBeagleBSP 2015-03-12T14:10:06 should i do this(../rtems/configure --target=sparc-rtems4.11 --enable-rtemsbsp=sis \ --enable-tests=samples --disable-posix) now? 2015-03-12T14:10:31 Sounds like time for that 2015-03-12T14:10:42 You will need to run that outside the RTEMS tree, 2015-03-12T14:11:05 in a build directory usually at the same level as the rtems source 2015-03-12T14:11:24 My lazy name is usually b-@BSP@ so b-sis in this case 2015-03-12T14:14:01 [trac] Developer/Projects/Open/ImproveRaspberryPiBSP edited by gedare (Add the list of peripherals.) http://devel.rtems.org/wiki/Developer/Projects/Open/ImproveRaspberryPiBSP 2015-03-12T14:14:32 hmm, wish we had a nice description of using the rpi like we do for bb 2015-03-12T14:14:40 just not enough beng-nls in the world. 2015-03-12T14:16:24 the new directory(b-sis) should be made under rtems only? 2015-03-12T14:16:31 50% of CS professors in the US come from 18 universities. 2015-03-12T14:17:00 no, make it in parallel to rtems 2015-03-12T14:17:03 so you did ./bootstrap 2015-03-12T14:17:05 now do 2015-03-12T14:17:09 cd .. ; mkdir b-sis 2015-03-12T14:17:11 cd b-sis 2015-03-12T14:17:16 ../rtems/configure ................. 2015-03-12T14:17:30 gedare that's interesting. 2015-03-12T14:17:37 ok 2015-03-12T14:17:42 50% of history profs come from 8 unis. 2015-03-12T14:18:00 50% of people are below average. :) 2015-03-12T14:18:07 it's been well known academia is not a meritocracy, but these numbers point to deep-seeded hierarchy and elitism 2015-03-12T14:18:19 Yep. 2015-03-12T14:18:47 it would be interesting to see how a "Blinded" faculty search would compare, where the faculty don't know the candidates' institutions. 2015-03-12T14:19:15 anyway, i'm a computer scientist, not a social scientist, so i'll get back to work 2015-03-12T14:19:40 LOL... says a lot about connections though. 2015-03-12T14:20:01 I have seen some reports that Silicon Valley hiring is biased to a few institutions 2015-03-12T14:20:09 "old boys" 2015-03-12T14:20:35 the numbers are worse when looking at female profs. 2015-03-12T14:20:44 Yep. But is the production of PhDs dominated by a few universities 2015-03-12T14:22:20 probably, but the indicators are that a phd only gets a job as a prof in an institution that "places" as many or fewer profs as the phd's alma mater 2015-03-12T14:22:37 here i'll find you the article you can have a read if you like :0 2015-03-12T14:23:09 sure.. hesham doesn't mind his paper getting farther down the queue ;) 2015-03-12T14:23:21 https://chroniclevitae.com/news/929-academia-s-1-percent?cid=VTEVPMSED1 2015-03-12T14:27:04 thank you gedare, it now up and running 2015-03-12T14:27:49 great! 2015-03-12T14:30:46 Interesting. The comments are even more interesting :) 2015-03-12T14:30:50 Need to do real work now 2015-03-12T14:31:30 i hear ya. 2015-03-12T14:32:03 oh, DrJoel, if rtems lab doesn't have a bb and rpi, they should perhaps be provisioned and setup 2015-03-12T14:33:15 There is currently no hardware setup for automated test but we have at least two BBBs and a couple or Pi Model Bs. 2015-03-12T14:33:25 I probably should order a Pi V2 2015-03-12T14:33:39 getting them set-up would be a nice job for potential co-mentors 2015-03-12T14:34:29 Yep. Jeff, Josh and Lou have all done the Pi. Jeff and Josh have also done the BBB and Edison 2015-03-12T14:36:17 *** Hesham has joined #rtems 2015-03-12T14:40:14 ok great to know. 2015-03-12T14:40:33 Hesham: are you planning to propose to do rump kernels? 2015-03-12T14:41:24 [trac] Developer/OpenProjects edited by gedare (delete raspberry pi2 as separate open project. it can fall under the rapsberry pi umbrella.) http://devel.rtems.org/wiki/Developer/OpenProjects 2015-03-12T14:41:25 gedare: Yes. 2015-03-12T14:42:13 ok. it may be helpful if you can find a mentor with rump kernel experience 2015-03-12T14:43:41 They are active on the their IRC list, and Antti said he would be happy to help with any questions but he doesn't want to apply as a mentor. 2015-03-12T14:44:16 ok 2015-03-12T14:44:29 good enough for me 2015-03-12T14:44:40 DrJoel: see msg 2015-03-12T14:47:26 *** acenos has joined #rtems 2015-03-12T14:47:45 hello 2015-03-12T14:48:23 hi acenos 2015-03-12T14:51:24 gedare do you remember the current magic for a BSP to exit back to the invoking environment? 2015-03-12T14:56:01 no 2015-03-12T14:56:17 the hack that schedsim uses? 2015-03-12T14:57:21 no.. the normal way. Sebastian reworked it and m32csim bsp doesn't exit now 2015-03-12T15:01:29 oh 2015-03-12T15:01:48 Oh.. combination of configure.ac and adding bsp_reset() 2015-03-12T15:05:45 does RTEMS support intel EDISON? 2015-03-12T15:06:07 DrJoel made some steps in that direction, but there are some problems with it 2015-03-12T15:08:52 The basic problem is that the documentation is under NDA. If you can answer enough questions about it, the BSP has identifiable steps 2015-03-12T15:09:09 The code -- as it is -- is all in git. 2015-03-12T15:10:23 http://rtemsramblings.blogspot.com/2014/12/intel-edison-and-rtems-road-forward.html 2015-03-12T15:10:37 I kept a blog of my week hacking on it. 2015-03-12T15:11:01 The APIC work should also improve pc386 SMP and help Edison. 2015-03-12T15:11:23 DrJoel: oh okay .... Will RTEMS try anything with Internet of Things (IoT)? 2015-03-12T15:11:37 There is AFAIK no way to debug on this board for bare metal applications without buying the Wind River JTAG 2015-03-12T15:11:52 Depends on which version of the IoT vision you are buying into. 2015-03-12T15:12:27 Some folks say the IoT is just old school embedded systems being re-marketed. Many already had network connections 2015-03-12T15:13:13 From a practical perspective, we probably need Wifi support. Beyond that, it is just normal embedded hardware with apps that talk to things that talk to the net or apps 2015-03-12T15:15:26 DrJoel: yeah but everyone is providing their own SDK for IoT.... the demo solutions the companies provide also is nothing new 2015-03-12T15:15:46 RTEMS provides support for standard interfaces. 2015-03-12T15:15:59 When a standard interface for IoT emerges, RTEMS will probably try to support it. 2015-03-12T15:18:22 I don't see actually how IoT is a new thing. 2015-03-12T15:18:43 me neither 2015-03-12T15:19:10 not that that says much - i've thought the same of now-billion-dollar companies 2015-03-12T15:19:29 Someone gave a presentation here about it, and still I think it's some embedded system applications with WiFi access 2015-03-12T15:21:38 I have heard IoT is a great way to get papers published : D 2015-03-12T15:21:45 :) 2015-03-12T15:21:49 but every big company i know is doing somethingwith it ..... 2015-03-12T15:23:43 I think it's about marketing, so when people see that title "This is an IoT" product, they would say waw, and buy it. 2015-03-12T15:26:18 it's really about _every_ embedded system having internet access and being internet accessible. 2015-03-12T15:27:16 instead of the traditional view of a client as a human, a server as a machine in a rack... a client could be an ES, and so could a server. this isn't new, but increasing the scale could be. 2015-03-12T15:27:52 *** antgreen has joined #rtems 2015-03-12T15:30:41 is the libnetworking code under cpukit is the latest ip stack ? i could not find any additional package for ip stack 2015-03-12T15:32:28 no that is the "old stack", which should work but is...old 2015-03-12T15:32:55 the new one is at https://git.rtems.org/rtems-libbsd/ 2015-03-12T15:35:21 gedare: okay ... with the old one i have networking enabled and access the interface on qemu 2015-03-12T15:35:32 probably try the new stack tomo 2015-03-12T15:36:19 gotta go .... bye 2015-03-12T15:36:46 *** ragunath has left #rtems 2015-03-12T15:38:49 *** Hesham has quit IRC (Ping timeout: 264 seconds) 2015-03-12T15:46:11 [trac] #2300/defect (new) updated by chrisj (It fails on FreeBSD so that is a 4.11 issue.) http://devel.rtems.org/ticket/2300 2015-03-12T15:47:10 *** kiwichris has quit IRC () 2015-03-12T15:59:52 *** ketul has quit IRC (Quit: Page closed) 2015-03-12T16:00:58 *** kiwichris has joined #rtems 2015-03-12T16:03:12 *** gedare has quit IRC (Ping timeout: 245 seconds) 2015-03-12T16:05:22 *** kiwichris has quit IRC (Ping timeout: 240 seconds) 2015-03-12T16:05:56 *** siddharth7 has joined #rtems 2015-03-12T16:06:13 *** hashm_ has left #rtems 2015-03-12T16:11:25 *** kiwichri_ has joined #rtems 2015-03-12T16:26:38 [trac] #2299/defect (new) updated by chrisj (The output path waits a period of time before flushing the data. It stops the output ...) http://devel.rtems.org/ticket/2299 2015-03-12T16:40:48 [trac] #2299/defect (new) updated by joel.sherrill (This is a case where the simulator does not have a clock tick source. So we just may ...) http://devel.rtems.org/ticket/2299 2015-03-12T16:47:57 *** acenos has quit IRC (Remote host closed the connection) 2015-03-12T16:54:25 *** siddharth7 has quit IRC (Quit: Leaving) 2015-03-12T16:54:28 *** jpc has joined #rtems 2015-03-12T17:17:09 [trac] #2299/defect (new) updated by joel.sherrill (OK. This is not related to lack of a clock tick ISR source. The gdb arm simulator ...) http://devel.rtems.org/ticket/2299 2015-03-12T17:38:40 *** Davidbrcz has quit IRC (Ping timeout: 255 seconds) 2015-03-12T17:43:09 [git] 71fb99b by Joel Sherrill: [Add configurations for all v850 GDB simulator BSP variants] http://devel.rtems.org/changeset/71fb99b/rtems-tools 2015-03-12T17:43:09 [git] b938414 by Joel Sherrill: [Add configuration for m32rsim BSP for simulator in GDB] http://devel.rtems.org/changeset/b938414/rtems-tools 2015-03-12T17:43:09 [git] ab5e0c5 by Joel Sherrill: [Add configuration for m32csim BSP for simulator in GDB] http://devel.rtems.org/changeset/ab5e0c5/rtems-tools 2015-03-12T17:43:09 [git] 8abbe80 by Joel Sherrill: [Add configurations for all arm GDB simulator BSP variants] http://devel.rtems.org/changeset/8abbe80/rtems-tools 2015-03-12T17:43:09 [git] a30ec56 by Joel Sherrill: [Add configurations for all h8 GDB simulator BSP variants] http://devel.rtems.org/changeset/a30ec56/rtems-tools 2015-03-12T17:44:36 [git] f1b4680 by Joel Sherrill: [m32cbsp: Add bsp_reset() so the BSP exits] http://devel.rtems.org/changeset/f1b4680/rtems 2015-03-12T18:00:20 *** gedare has joined #rtems 2015-03-12T18:00:26 *** ChanServ sets mode: +o gedare 2015-03-12T18:05:52 [trac] #2299/defect (new) updated by chrisj (Are those BSP's which are passing simulating a clock tick using a BSP option and so ...) http://devel.rtems.org/ticket/2299 2015-03-12T18:07:22 *** gedare has quit IRC (Ping timeout: 240 seconds) 2015-03-12T18:08:26 DrJoel: is there any issue with how i've done it in the waf build? see my reply to gedare 2015-03-12T18:08:56 [trac] #2299/defect (new) updated by joel.sherrill (arm920-testsuite.tcfg only has this: include: testdata/require-tick-isr.tcfg ...) http://devel.rtems.org/ticket/2299 2015-03-12T18:09:16 I answered. The "not link" is a first pass filter. 2015-03-12T18:10:12 i see, i can incorporate this as well 2015-03-12T18:10:27 You can have a tiny test which creates 10000 tasks. It will link and fit in RAM but not run because it needs 10000 task stacks and TCBs. Probably 40000k (~45MB) to allocate all them. 2015-03-12T18:11:11 My script catches those cases if we know the amount of memory available for the work space. That can be determined though. Just requires at least architecture specific information for linker symbol names 2015-03-12T18:12:38 Agreed. You can incorporate it. But at this point, we would have to rewrite this in Python (no biggie) and dig the symbols up. I am hoping it is just architecture specific information. I am afraid sometimes it might be BSP specific. :( 2015-03-12T18:12:59 For architectures with shared linkcmds, this is pretty easy since the symbols are common :) 2015-03-12T18:19:03 [trac] #2299/defect (new) updated by chrisj (I would check the BSPOPTS for the passing vs the failing. Some BSPOPTS default to ...) http://devel.rtems.org/ticket/2299 2015-03-12T18:19:33 DrJoel, looking at the script to catch failing tests. 2015-03-12T18:19:52 DrJoel, why is the failure at test run time no good ? 2015-03-12T18:20:40 The ones I saw are silent so no clue what went wrong. They waste download time and test time. 2015-03-12T18:21:00 Oh there is no error showing the reason ? 2015-03-12T18:21:02 Just would be quicker to avoid putting them on the target at all since we can guarantee they will fail 2015-03-12T18:21:18 No. This fails before rtems_initialize_data_structures is called. Very early in boot_card(). 2015-03-12T18:21:39 I suppose _Terminate has a useful error given to it. 2015-03-12T18:21:52 I just thought it was easy to script and avoid ever downloading them. 2015-03-12T18:21:53 Should the BSP printk the failure ? 2015-03-12T18:22:13 Understood. But this means we have another class of test failure to manage. 2015-03-12T18:22:16 It could. 2015-03-12T18:22:33 Yes. I understand that. But we have that anyway. 2015-03-12T18:22:39 The BSP should be nice to all users. 2015-03-12T18:23:11 But we have no way to undo the adding of tests which is a flaw in the current approach 2015-03-12T18:23:41 We do not see a test could pass now due to improved memory usage 2015-03-12T18:24:03 On the technical level the Configuration.workspace is a const table ? 2015-03-12T18:25:11 ie the workspace space must be a const somewhere in the executable ? 2015-03-12T18:25:23 workspace space -> workspace size 2015-03-12T18:26:28 Yes. Configuration is a const table. 2015-03-12T18:26:43 I was not proposing to add these to a static list. 2015-03-12T18:27:03 I was just detecting tests which won't pass the initial work space size tests. 2015-03-12T18:27:47 I have no idea mechnically how to avoid pushing them to the target except for rtems-tester to do this inspection before it runs them and say "workspace too large for target". 2015-03-12T18:28:03 I do **NOT** want them on a permanent black list for the BSP. More dynamic 2015-03-12T18:28:13 You could fail it or NA it 2015-03-12T18:28:17 I don' 2015-03-12T18:28:37 I don't know how to account for it. Just thought rtems-tester could avoid putting it on target 2015-03-12T18:30:56 leaving to get Justin. If you want to chat, call my cell in a few minutes. Got to lock up 2015-03-12T18:31:00 *** DrJoel has quit IRC (Quit: I used to think I was indecisive, but now I'm not too sure.) 2015-03-12T19:03:19 ok 2015-03-12T20:06:50 *** dr__house|away is now known as dr__house 2015-03-12T20:11:10 *** shelf is now known as timothyh 2015-03-12T20:11:34 *** gedare has joined #rtems 2015-03-12T20:11:34 *** ChanServ sets mode: +o gedare 2015-03-12T20:23:20 [trac] #2301/enhancement (v:4.10) created by johill (cpukit is built in BSP, and BSP custom, specific part of build tree but isnt built using BSP's spec file) http://devel.rtems.org/ticket/2301 2015-03-12T20:27:18 NASA TV is showing MMS on the launch pad. 2015-03-12T20:27:25 yep very cool 2015-03-12T20:29:06 hmm 2015-03-12T20:30:35 nice 2015-03-12T20:30:44 that's an interesting ticket johill filed 2015-03-12T20:32:06 btw i never modified an installed header, because i almost never install rtems. ;) 2015-03-12T20:32:42 Google Code is shutting down. when I first saw the email I didn't see Code, and was a bit surprised. 2015-03-12T20:34:24 kiwichri_: I don't see the need to hold-up 4.11 release for one toolset build failure that is host-specific 2015-03-12T20:36:29 anyway good-night 2015-03-12T20:36:38 hope its a great launch! 2015-03-12T20:38:00 [trac] #2301/enhancement (new) updated by johill (empty comment) http://devel.rtems.org/ticket/2301 2015-03-12T20:38:31 gedare, so what do MIP people on FreeBSD do. 2015-03-12T20:39:00 gedare, also gdb needs to move to a later release 2015-03-12T20:40:52 *** gedare has quit IRC (Ping timeout: 252 seconds) 2015-03-12T21:38:30 mms is launching!! 2015-03-12T23:20:09 *** jpc has quit IRC (Quit: Ex-Chat) 2015-03-13T00:48:28 *** dr__house is now known as dr__house|away 2015-03-13T01:38:10 *** monstr has joined #rtems 2015-03-13T02:16:18 *** kiwichri_ has quit IRC () 2015-03-13T02:19:39 *** andrei_ has joined #rtems 2015-03-13T02:32:41 *** acenos has joined #rtems 2015-03-13T03:12:44 *** ketul has joined #rtems 2015-03-13T03:40:36 *** soniadalal has joined #rtems 2015-03-13T03:53:42 *** Hesham has joined #rtems 2015-03-13T04:01:12 hello I want to ask that rtems works on which uart from 1-4? 2015-03-13T04:09:22 *** kiwichris has joined #rtems 2015-03-13T04:11:57 *** ketul has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-13T04:12:36 *** ketul has joined #rtems 2015-03-13T04:57:00 [trac] bintime_requests.png (Number of bintime requests with hardware periphery) attached to #2271 by AlexKrEB http://devel.rtems.org/ticket/2271 2015-03-13T04:57:21 [trac] bintime_requests_wo.png (Number of bintime requests without hardware periphery) attached to #2271 by AlexKrEB http://devel.rtems.org/ticket/2271 2015-03-13T04:57:46 [trac] uptime_requests.png (Number of uptime requests with hardware periphery) attached to #2271 by AlexKrEB http://devel.rtems.org/ticket/2271 2015-03-13T04:58:03 [trac] uptime_requests_wo.png (Number of uptime requests without hardware periphery) attached to #2271 by AlexKrEB http://devel.rtems.org/ticket/2271 2015-03-13T05:13:42 *** ketul has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-13T05:29:02 *** soniadalal has quit IRC (Ping timeout: 246 seconds) 2015-03-13T07:24:28 [trac] #2271/enhancement (accepted) updated by AlexKrEB (== Current status == Several tests were executed to compare the performance ...) http://devel.rtems.org/ticket/2271 2015-03-13T07:31:06 [trac] #2271/enhancement (accepted) updated by sebastian.huber (Newlib patch for is available upstream: ...) http://devel.rtems.org/ticket/2271 2015-03-13T08:11:57 *** Hesham has quit IRC (Quit: Leaving.) 2015-03-13T08:12:43 *** gedare has joined #rtems 2015-03-13T08:12:44 *** ChanServ sets mode: +o gedare 2015-03-13T08:35:49 *** andrei_ has quit IRC (Ping timeout: 246 seconds) 2015-03-13T08:48:36 *** ragunath has joined #rtems 2015-03-13T09:20:25 *** itazzzz has quit IRC (Quit: Lost terminal) 2015-03-13T09:42:52 *** soniadalal has joined #rtems 2015-03-13T09:55:02 i am trying to compile rtems-libbsd ... i got the following error "error: expected specifier-qualifier-list before 'ISR_LOCK_MEMBER'" 2015-03-13T09:55:15 i am following the steps mentioned in libbsd.txt 2015-03-13T10:09:12 got it my repo is not sync to the latest .. the change was added 8 days ago 2015-03-13T10:53:16 *** Hesham has joined #rtems 2015-03-13T11:01:57 *** Hesham has quit IRC (Ping timeout: 245 seconds) 2015-03-13T11:10:29 *** ita has joined #rtems 2015-03-13T11:11:21 kiwichris: no news from the mingw folks on your side? 2015-03-13T11:16:19 *** ketul has joined #rtems 2015-03-13T11:16:51 *** Hesham has joined #rtems 2015-03-13T11:16:57 ita, he is probably not awake yet, he is on aussie time. 2015-03-13T11:19:46 beng-nl: I got ftdi. Now I want to ask you that whether rtems talks on which uart ? 2015-03-13T11:21:54 gedare: ? 2015-03-13T11:22:02 hi ketul I don't know the answer to your question. 2015-03-13T11:22:09 ok 2015-03-13T11:22:20 probably ben will answer sometime :) 2015-03-13T11:22:37 ok no prob. 2015-03-13T11:47:09 ketul: i'm here 2015-03-13T11:47:30 ketul: eh, i don't totally get it.. if you build the bsp in beaglebone mode, it'll select the righ uart 2015-03-13T11:47:34 right* 2015-03-13T11:48:10 while referring BBB gpio diagram I found 4 uarts are there 2015-03-13T11:48:10 *** ankur has joined #rtems 2015-03-13T11:48:25 http://beagleboard.org/support/bone101 2015-03-13T11:48:38 ok 2015-03-13T11:48:38 but 2015-03-13T11:48:41 *** Hesham has quit IRC (Ping timeout: 250 seconds) 2015-03-13T11:48:48 the rtems code selects the right one 2015-03-13T11:50:06 but I want to ask you that on the BBB on which GPIOs I should connect ftdi 2015-03-13T11:50:24 aha 2015-03-13T11:50:32 there's just one place right, those serial pins/headers 2015-03-13T11:50:49 http://www.armhf.com/wp-content/uploads/2013/06/ftdi_serial_debug.jpg 2015-03-13T11:50:56 there, the colourful wires 2015-03-13T11:51:51 *** Hesham has joined #rtems 2015-03-13T11:52:00 ya that I referred but didn't get which wire should connect to which no. of gpio of BBB. 2015-03-13T11:52:10 just want gpio no. 2015-03-13T11:52:25 like 24rx-with ftdi's tx 2015-03-13T11:55:13 if you're not wiring it up using those pins, i don't know; but you can tell the uart number from the rtems bsp code, perhaps that helps? 2015-03-13T12:02:10 ok beng-nl: let me try with that 6 pins 2015-03-13T12:14:31 *** ragunath has quit IRC (Quit: Page closed) 2015-03-13T12:14:58 *** acenos has quit IRC (Quit: Leaving) 2015-03-13T12:15:17 *** ragunath has joined #rtems 2015-03-13T12:17:52 beng-nl: Finally got the output 2015-03-13T12:18:00 you can check here. 2015-03-13T12:18:00 http://pastebin.com/YXiUkHAT 2015-03-13T12:19:06 great! 2015-03-13T12:20:10 beng-nl: Currently I am using just uart 2015-03-13T12:20:17 rx,tx and gnd 2015-03-13T12:21:17 okay 2015-03-13T12:22:34 really it took 2-3 days to have output after successful img built. Anyways at last happy to have that 2015-03-13T12:26:15 that's kinda normal isn't it :) 2015-03-13T12:26:47 2-3 days is good progress to get an ES running. 2015-03-13T12:27:06 and, it is why we require students to do it before they finish the application. 2015-03-13T12:33:31 thanks gedare and beng-nl. 2015-03-13T12:46:42 Before wasting more time on debugging, I was wondering if this line may have some issues, because it's returning with invalid Thread address > https://github.com/RTEMS/rtems/blob/master/testsuites/tmtests/tm27/task1.c#L195 2015-03-13T12:48:00 On Epiphany... 2015-03-13T12:53:45 *** ankur_ has joined #rtems 2015-03-13T12:53:46 *** ankur has quit IRC (Ping timeout: 246 seconds) 2015-03-13T12:54:53 beng-nl: now trying to run with Beagle xM over JTAG 2015-03-13T13:04:46 hi Hesham 2015-03-13T13:07:48 it's possible, it is cheating by accessing the Scheduler_priority_Context structure. 2015-03-13T13:07:59 *** Hesham has quit IRC (Ping timeout: 250 seconds) 2015-03-13T13:08:01 and generally calling scheduler operations directly 2015-03-13T13:17:54 *** soniadalal has left #rtems ("Leaving") 2015-03-13T13:20:55 beng-nl : trying to execute "arm-rtems4.11-gdb $HOME/development/rtems/b-beagle/arm-rtems4.11/c/beagleboardxm/testsuites/samples/ticker/ticker.exe" 2015-03-13T13:21:18 but arm-rtems4.11-gdb: command not found 2015-03-13T13:21:42 I do have gdb-arm-none-eabi 2015-03-13T13:26:55 sorry found it. 2015-03-13T13:36:13 *** ankur_ has quit IRC (Quit: Page closed) 2015-03-13T13:36:58 *** ankur has joined #rtems 2015-03-13T13:38:42 *** monstr has quit IRC (Remote host closed the connection) 2015-03-13T13:39:14 *** gedare has quit IRC (Remote host closed the connection) 2015-03-13T13:42:42 *** Hesham has joined #rtems 2015-03-13T13:54:40 *** ketul has quit IRC (Ping timeout: 246 seconds) 2015-03-13T13:58:45 [trac] #2302/defect (v:4.11) created by joel.sherrill (beagle BSPs - warning because of bad symbol name choice) http://devel.rtems.org/ticket/2302 2015-03-13T13:59:39 *** ankur has left #rtems 2015-03-13T14:01:38 [trac] #2303/defect (v:4.11) created by joel.sherrill (Arm/rtl22xx* BSPs use common symbol FLASH_SIZE conflicts) http://devel.rtems.org/ticket/2303 2015-03-13T14:03:00 [trac] #2304/defect (v:4.11) created by joel.sherrill (or1ksim warnings in spcontext01) http://devel.rtems.org/ticket/2304 2015-03-13T14:14:47 [trac] #2304/defect (accepted) updated by heshamelmatary (empty comment) http://devel.rtems.org/ticket/2304 2015-03-13T14:38:59 *** gedare has joined #rtems 2015-03-13T14:38:59 *** ChanServ sets mode: +o gedare 2015-03-13T14:49:48 [trac] #2304/defect (closed) updated by Hesham ALMatary (In [changeset:"a764d673876f8e7b61b1a13fced85e4126a4b00d/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2304 2015-03-13T14:49:49 [git] a764d67 by Hesham ALMatary: [score/or1k: Fix warnings. Add missing prototypes Closes #2304] http://devel.rtems.org/changeset/a764d67/rtems 2015-03-13T15:04:38 *** mazzoo has joined #rtems 2015-03-13T15:24:07 beng-nl: hi ben .. do you have info running lwIP on BBB? came to know that someone already did some work on that 2015-03-13T15:30:33 *** Hesham has quit IRC (Ping timeout: 250 seconds) 2015-03-13T15:35:36 you might contact daniel gutson or inquire on the mailing list about it. 2015-03-13T15:36:52 gedare: ok i will send a mail 2015-03-13T16:01:34 gedare: sent a mail asking for details 2015-03-13T16:05:13 *** gedare has quit IRC (Ping timeout: 250 seconds) 2015-03-13T16:10:13 *** ragunath has left #rtems 2015-03-13T17:21:03 *** jpc has joined #rtems 2015-03-13T18:16:00 *** dr__house|away is now known as dr__house 2015-03-13T18:19:43 ita, no I have no heard anything. 2015-03-13T18:43:58 [trac] #2271/enhancement (accepted) updated by chrisj (Is the watchdog change and freebsd timer counter change linked in any way or are ...) http://devel.rtems.org/ticket/2271 2015-03-13T19:47:28 [trac] Developer/Projects/Open/TraceTool edited by chrisj (Fix image.) http://devel.rtems.org/wiki/Developer/Projects/Open/TraceTool 2015-03-13T19:55:57 *** gedare has joined #rtems 2015-03-13T19:55:57 *** ChanServ sets mode: +o gedare 2015-03-13T20:04:25 hey kiwichris i know you're busy just wanted to let you know, spot on with the drvmgr/pci stuff. 2015-03-13T20:04:37 gedare, hi 2015-03-13T20:04:43 cheers 2015-03-13T20:04:50 it's a political decision to merge it, as much as anything. "real" sparc users are already using the code 2015-03-13T20:05:09 and sebastian stated his support of drvmgr as an alternate to something like busspace. 2015-03-13T20:05:51 but i completely agree about the testing situation, and it is extremely an odd case, to have sparc-specific code in cpukit, and also to merge a large piece of code before release. 2015-03-13T20:05:53 I think the code in a good thing. All I am asking is some public commitment to support it. Yes there are a few about but it exists and there are users and drivers. 2015-03-13T20:06:20 Bus space may not everything to all people. It comes with bagage 2015-03-13T20:06:26 yep 2015-03-13T20:06:33 Be back in a bit; Need to drop by daughter at dance 2015-03-13T20:09:57 ok i'm heading out anyway. hopefully danielh will get back to us with his plan, last i heard he needed a few weeks time to address the issues. maintaining this is a low-priority issue for them, but they do seem to put some effort into it. 2015-03-13T20:15:07 *** gedare has quit IRC (Ping timeout: 256 seconds) 2015-03-13T22:16:13 *** jpc has quit IRC (Quit: Ex-Chat) 2015-03-14T03:17:59 *** zoso has joined #rtems 2015-03-14T03:52:10 *** dr__house is now known as dr__house|away 2015-03-14T04:06:33 *** waleed has joined #rtems 2015-03-14T05:25:29 *** Hesham has joined #rtems 2015-03-14T05:28:19 *** altairpearl has joined #rtems 2015-03-14T05:31:09 *** altairpearl has quit IRC (Remote host closed the connection) 2015-03-14T05:32:45 *** altairpearl has joined #rtems 2015-03-14T05:39:27 *** zoso has quit IRC (Ping timeout: 250 seconds) 2015-03-14T05:42:56 [trac] TBR/BSP/Orp edited by heshamelmatary (empty comment) http://devel.rtems.org/wiki/TBR/BSP/Orp 2015-03-14T05:57:24 *** zoso has joined #rtems 2015-03-14T06:42:54 *** gedare has joined #rtems 2015-03-14T06:42:55 *** ChanServ sets mode: +o gedare 2015-03-14T06:54:41 *** gedare has quit IRC (Ping timeout: 246 seconds) 2015-03-14T06:56:22 *** zoso has quit IRC (Ping timeout: 240 seconds) 2015-03-14T06:57:41 *** waleed has quit IRC (Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/) 2015-03-14T07:01:53 *** gedare has joined #rtems 2015-03-14T07:01:53 *** ChanServ sets mode: +o gedare 2015-03-14T07:20:52 [trac] GSoC/2015 edited by heshamelmatary (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-03-14T07:25:55 *** antgreen` has joined #rtems 2015-03-14T07:54:32 *** Hesham has quit IRC (Ping timeout: 246 seconds) 2015-03-14T08:32:25 *** Hesham has joined #rtems 2015-03-14T09:14:08 *** gedare has quit IRC (Ping timeout: 256 seconds) 2015-03-14T09:25:16 *** Hesham has quit IRC (Ping timeout: 252 seconds) 2015-03-14T09:51:37 *** ragunath has joined #rtems 2015-03-14T10:07:05 [trac] #2302/defect (new) updated by beng (testing a change to fix this now.) http://devel.rtems.org/ticket/2302 2015-03-14T10:17:30 hey guys. i'm seeing a build problem for psxclock01 - missing _SC_CLK_TCK and sysconf prototype. anyone seen this / working on this? i think it's just a missing #include in the test 2015-03-14T10:25:12 [trac] #2305/defect (v:4.11) created by joel.sherrill (sp07 needs to be split into an user extensions and a notepad test) http://devel.rtems.org/ticket/2305 2015-03-14T10:26:15 *** DrJoel has joined #rtems 2015-03-14T10:26:15 *** DrJoel has joined #rtems 2015-03-14T10:26:15 *** ChanServ sets mode: +o DrJoel 2015-03-14T10:44:47 beng-nl: i am also seeing the error 2015-03-14T10:44:57 which error? 2015-03-14T10:48:18 DrJoel: error: '_SC_CLK_TCK' undeclared (first use in this function) in psxclock01 2015-03-14T10:48:33 DrJoel: beagleboneblack compilation 2015-03-14T10:48:40 Updating the RSB and new tools should fix that 2015-03-14T10:48:46 oops. ok. 2015-03-14T10:48:56 we should rebuild all tools? 2015-03-14T10:49:10 aha, it's a newlib thing i suppose 2015-03-14T10:49:11 okay 2015-03-14T10:49:27 Just for the target(s) you care about 2015-03-14T10:49:31 right ok 2015-03-14T10:49:36 *** ketul has joined #rtems 2015-03-14T10:49:54 DrJoel: oh okay 2015-03-14T10:50:14 ragunath: i will rebase & push new to the beagle forks of RSB & tools 2015-03-14T10:50:22 (branches) 2015-03-14T10:50:55 beng-nl: okay ben thanks 2015-03-14T10:51:27 will take an hour or two before it's done, tested and pushed 2015-03-14T10:51:36 yep. :( 2015-03-14T10:52:00 I am working through turning off all the deprecated warnigns Gedare introduced and updating code as needed to not use deprecated methods. :( 2015-03-14T10:52:25 so be it 2015-03-14T10:53:08 beng-nl: hi 2015-03-14T10:53:36 ideally i shouldn't be maintaining a seperate RSB & rtems-tools branch.. 2015-03-14T10:54:07 but kiwichris didn't like all of the changes which i totally accept but making the changes takes time 2015-03-14T10:55:18 want to tell that https://www.rtems.org/irclogs/index.html is not working. 2015-03-14T10:55:55 verm__: has a ticket for them. We moved servers and the bot hasn't moved yet. It is still logging on an old machine but not public facing 2015-03-14T11:05:38 *** ketul has quit IRC (Quit: Page closed) 2015-03-14T11:06:21 *** ketul has joined #rtems 2015-03-14T11:12:28 Well.. it will take at least 10 hours to finish the build. I am building all BSPs with all tests to see how many warnings are left. Hopefully I didn't break any code eliminating use of deprecated methods 2015-03-14T11:15:56 Looks like the Pi 2 just showed up at my door. :) 2015-03-14T11:17:04 *** starprince has joined #rtems 2015-03-14T11:17:13 *** starprince has left #rtems 2015-03-14T11:19:08 cool! 2015-03-14T11:24:26 I get Saturday (and Sunday) Amazon shipping. Wouldn't get to the office until Monday. :) 2015-03-14T11:25:23 but I need to go away from the keyboard and let Michele read her book to me. her third book is getting very to finished and she reads it out loud to me as I read it on the screen. If the words she says don't match, the pacing is off, or either or us spot something, we stop and address it. 2015-03-14T11:25:27 be back later 2015-03-14T11:27:26 good process! enjoy 2015-03-14T11:29:54 The draft is in Google Docs so we get live edits. :) 2015-03-14T11:30:41 The first book we did this at the office and used the big screens in the OAR conference room. Google docs is pretty safe for not losing anything and nice for collaboration 2015-03-14T11:31:13 exactly 2015-03-14T11:31:21 very beneficial to have a single master copy 2015-03-14T11:31:32 or.. instance.. 2015-03-14T11:32:34 yeah.. Michele went to a lot of trouble moving a Word doc from notebook to desktop to me.. 2015-03-14T11:49:22 *** altairpearl has quit IRC (Ping timeout: 252 seconds) 2015-03-14T12:20:42 Watching RTEMS Talk by Joel Sherrill @ VU University in Amsterdam in January 2014 . Really, very much informative and realized practical implementation what I studied in my course. 2015-03-14T12:33:41 *** Hesham has joined #rtems 2015-03-14T12:36:02 :-) 2015-03-14T12:36:08 ketul: nice you watched it! 2015-03-14T12:36:17 ketul: i subtitled it with my own fingers! 2015-03-14T12:36:42 *** ankur has joined #rtems 2015-03-14T12:36:45 ya seen it. Thanks for uploading 2015-03-14T12:48:13 :-) 2015-03-14T12:48:15 cool 2015-03-14T12:48:53 ragunath: i merged RSB & rtems-tools mainline, tested a new from scratch build, and pushed the result, so we're ok again 2015-03-14T12:51:03 [trac] #2302/defect (new) updated by beng (tested; submitted the change for merging.) http://devel.rtems.org/ticket/2302 2015-03-14T12:51:26 Thanks for the ticket Ben. Was just going to ask. :) 2015-03-14T12:51:53 As things come together, we need tickets for the remaining issues. Like settling the gdb patches issue 2015-03-14T12:52:37 beng-nl: thanks ben ... it is available in the beagle repo 2015-03-14T12:52:41 ? 2015-03-14T12:53:22 ragunath: yes, that's what i meant by having pushed the branches 2015-03-14T12:53:50 beng-nl: oh ok... let me sync to the latest 2015-03-14T12:53:54 ok 2015-03-14T12:54:09 i merged (instead of rebased) for pull-friendliness 2015-03-14T12:54:24 beng-nl: cool 2015-03-14T13:07:07 *** ragunath_ has joined #rtems 2015-03-14T13:07:21 *** richashi has joined #rtems 2015-03-14T13:08:48 32 BSPs into the build.. no errors.. only 160 left to go. 2015-03-14T13:09:04 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-14T13:09:19 *** ragunath_ is now known as ragunath 2015-03-14T13:10:36 * DrJoel hopes the bulbs he just ordered for the light under the microwave work. $10 for 25 bulbs is cheap. But $7 for two Philips bulbs is also very expensive. :) 2015-03-14T13:17:38 Hi, I wanted to ask what is the IRC nick for Chris Johns 2015-03-14T13:20:18 kiwichris is his nick. He is in Sydney Australia so there are timezone issues. But he is here regularly 2015-03-14T13:20:28 richashi: I can take a stab at a question 2015-03-14T13:23:04 *** acenos has joined #rtems 2015-03-14T13:23:39 * DrJoel finally had to order Michele a new food processor. Enough parts needed to be replaced on her 8 year old one that it was time to get a new one. I had replaced the bowl 4 years ago on that one. She cooks from scratch a lot and ends up wearing things out. 2015-03-14T13:55:03 *** dr__house|away is now known as dr__house 2015-03-14T14:00:00 * DrJoel waves at dr__house 2015-03-14T14:00:16 * dr__house waves at DrJoel back 2015-03-14T14:00:22 DrJoel: how are you? 2015-03-14T14:02:57 great. Michele is making an Apple Pie for Pi Day and she is wrapping up the third book. It looks like it will be over quite a bit longer than the previous two 2015-03-14T14:03:04 are you still in the Atlanta area? 2015-03-14T14:03:12 DrJoel: yep :) 2015-03-14T14:03:20 michele rocks, making an applie pie for pi day 2015-03-14T14:03:25 apple* 2015-03-14T14:03:42 Hopefully we will have a book signing before my trip to Atlanta for work in June. 2015-03-14T14:04:12 Or we can find our way to the Georgia Renaissance Festival. That's always a blast 2015-03-14T14:04:28 DrJoel: Cool! Looking forward to meeting you 2015-03-14T14:04:30 beng-nl: she has a tradition of making one every year for Pi Day. Not always apple 2015-03-14T14:04:41 It has been a while. 2015-03-14T14:09:06 :q 2015-03-14T14:11:49 the saddest smiley - the quit vi smiley 2015-03-14T14:12:53 You are right! I try to not post that. But I think even Michele recognizes it in chat from me 2015-03-14T14:19:25 The Pi symbol is on the pie and she is working on 3.14 :) 2015-03-14T14:28:30 *** ketul__ has joined #rtems 2015-03-14T14:29:13 *** ketul has quit IRC (Ping timeout: 246 seconds) 2015-03-14T14:29:34 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-14T14:29:34 *** ankur has quit IRC (Ping timeout: 246 seconds) 2015-03-14T14:30:04 *** ragunath has joined #rtems 2015-03-14T14:32:43 *** ketul__ has quit IRC (Ping timeout: 246 seconds) 2015-03-14T14:34:28 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-14T14:35:59 *** ragunath has joined #rtems 2015-03-14T14:59:06 DrJoel: for bbb ethernet driver, we need it work on both lwIP and new bsd ip stack right? 2015-03-14T15:04:33 ragunath: yes.. lwip *SHOULD* be more of a packaging/polishing issue 2015-03-14T15:05:11 ragunath: For the new stack, look in FreeBSD 9.x source and see if you can find BBB NIC driver. beng-nl may know exactly where it is. :) 2015-03-14T15:05:28 DrJoel: there is a driver in freebsd fir bbb 2015-03-14T15:05:33 *for 2015-03-14T15:05:36 I am assuming that it exists and all that is needed is to provide a bus driver for BBB 2015-03-14T15:06:28 https://git.rtems.org/rtems-libbsd/ is the new TCP/IP stack from FreeBSD 9.x. It includes USB support so that should be easy to get working as well. 2015-03-14T15:07:05 DrJoel: I got the new ip stack and tested it with arm realview on qemu 2015-03-14T15:07:22 I spoke with Chris yesterday and he likes the idea of RTEMS without network stack, RSB support building LWIP, Old one as separate package, or new one.. then a collection of network services 2015-03-14T15:07:40 very cool! Then you are well on your way to success if you have that much working 2015-03-14T15:07:44 walking away 2015-03-14T15:15:12 DrJoel: bbb has musb otg controller .... i guess it is not ported to rtems libbsd port ... on top that controller we need to add bbb specific driver code 2015-03-14T15:16:37 Hmm.. I thought only the Pi did that but I have no real idea :( 2015-03-14T15:16:40 DrJoel: not sure how much effort it needs 2015-03-14T15:17:14 beng-nl: should be able to answer this one. He is also a Minix core developer and knows the BB very well from both perspectives 2015-03-14T15:19:00 DrJoel:speaking of minix ... minix also has their own ethernet driver for bbb .. on the outline it looked simple compared to freebsd one 2015-03-14T15:19:17 beng-nl: probably had a hand in that. :) 2015-03-14T15:19:30 :) 2015-03-14T15:19:48 DrJoel: http://wiki.minix3.org/doku.php?id=developersguide:minixonarm minix bbb status 2015-03-14T15:26:35 :) then it should be a matter of porting/providing the right underlying infrastructure. All should be at the BSP level 2015-03-14T15:26:43 beng-nl: really does know more though. 2015-03-14T15:35:20 DrJoel: how should we be testing lwIP ? since networking tests are linked with the ip stack we use 2015-03-14T15:38:59 DrJoel, ragunath: i do know about that driver yes 2015-03-14T15:39:46 the only thing there is that the phy & mii stuff is a single thing and it relies on some uboot initialization, so it might not be totally suitable for rtems integration 2015-03-14T15:39:51 haven't looked into it 2015-03-14T15:39:56 the license is good though :) 2015-03-14T15:40:34 beng-nl: the minix driver? 2015-03-14T15:40:42 yep 2015-03-14T15:42:14 beng-nl: freebsd driver uses fdt to get phy address ? i think fdt is also related to uboot i guess? i am not sure though 2015-03-14T15:42:58 [trac] #2302/defect (closed) updated by Ben Gras (In [changeset:"86204bdeb554289c3eaf66efbc5057d0614ad076/rtems"]: {{{ ...) http://devel.rtems.org/ticket/2302 2015-03-14T15:42:59 [git] dbe8595 by Ben Gras: [autoconf changes] http://devel.rtems.org/changeset/dbe8595/rtems 2015-03-14T15:42:59 [git] 86204bd by Ben Gras: [beagle bsp: delete TIMER_FREQ, TIMER_COUNT unused and poorly named (no prefix) and colliding with sp68. Closes #2302.] http://devel.rtems.org/changeset/86204bd/rtems 2015-03-14T15:43:27 *** acenos has joined #rtems 2015-03-14T15:43:32 drat. i accidentally also merged the autoconf commit. 2015-03-14T15:53:22 *** andrei_il has joined #rtems 2015-03-14T15:55:38 *** richashi has quit IRC (Quit: richashi) 2015-03-14T16:19:13 *** richashi has joined #rtems 2015-03-14T16:23:56 *** acenos has quit IRC (Quit: Saliendo) 2015-03-14T16:30:03 [git] 275a9bf by Ben Gras: [Revert "autoconf changes" This reverts commit dbe85952f9adac28a2dc84473eac24ae2c85e260. Accidentally merged.] http://devel.rtems.org/changeset/275a9bf/rtems 2015-03-14T16:44:53 *** Hesham has quit IRC (Ping timeout: 244 seconds) 2015-03-14T16:48:52 *** andrei_il has quit IRC (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client) 2015-03-14T17:11:19 *** ragunath has left #rtems 2015-03-14T17:18:41 *** DrJoel_ has joined #rtems 2015-03-14T17:19:11 * DrJoel_ is enjoying apple pie on Pi Day https://www.facebook.com/michele.pendleton.sherrill/posts/10152869607590914 2015-03-14T17:42:03 *** richashi has quit IRC (Quit: richashi) 2015-03-14T17:55:20 *** gedare has joined #rtems 2015-03-14T17:55:20 *** ChanServ sets mode: +o gedare 2015-03-14T18:12:25 *** gedare has quit IRC (Ping timeout: 264 seconds) 2015-03-14T19:03:12 DrJoel_, an infinite Pi is a nice idea, slice after slice 2015-03-14T19:03:38 :) 2015-03-14T19:34:47 DrJoel_, adding Trace wiki pages. 2015-03-14T19:39:02 OK. I chatted with rish.. this afternoon and sent him the FSW presentaiton 2015-03-14T19:39:35 Is that availiable online ? 2015-03-14T19:40:18 not yet. I need to email and ask when. it has been almost 3 months 2015-03-14T19:43:00 [trac] Developer/Tracing created http://devel.rtems.org/wiki/Developer/Tracing 2015-03-14T21:14:17 [trac] #2306/defect (v:4.11) created by joel.sherrill (powerpc/mvme5500/vectors/exceptionhandler.c uses task variables) http://devel.rtems.org/ticket/2306 2015-03-14T22:53:21 *** zoso has joined #rtems 2015-03-14T23:11:32 *** zoso has quit IRC (Ping timeout: 245 seconds) 2015-03-14T23:14:15 *** zoso has joined #rtems 2015-03-14T23:34:31 *** dr__house is now known as dr__house|away 2015-03-14T23:57:11 [trac] rtems-trace.png (RTEMS Trace using CTF) attached to #Developer/Tracing by chrisj http://devel.rtems.org/ticket/Developer/Tracing 2015-03-15T00:39:23 *** zoso has quit IRC (Ping timeout: 246 seconds) 2015-03-15T00:42:17 [trac] Developer/Tracing edited by chrisj (Add CTF workflow.) http://devel.rtems.org/wiki/Developer/Tracing 2015-03-15T00:49:54 [trac] rtems-trace-printk.png (RTEMS Trace using Printk) attached to #Developer/Tracing by chrisj http://devel.rtems.org/ticket/Developer/Tracing 2015-03-15T01:02:48 [trac] Developer/Tracing edited by chrisj (Add Printk workflow.) http://devel.rtems.org/wiki/Developer/Tracing 2015-03-15T01:09:29 [trac] Developer/Tracing/Trace Linker created http://devel.rtems.org/wiki/Developer/Tracing/Trace Linker 2015-03-15T01:17:16 *** ketul has joined #rtems 2015-03-15T01:35:11 [trac] Developer edited by chrisj (Start adding developer pages.) http://devel.rtems.org/wiki/Developer 2015-03-15T01:37:46 [trac] Developer/Git/Committers edited by chrisj (Fix renaming and image.) http://devel.rtems.org/wiki/Developer/Git/Committers 2015-03-15T01:38:15 [trac] Developer/Git/Committers edited by chrisj (Scale down the image.) http://devel.rtems.org/wiki/Developer/Git/Committers 2015-03-15T01:39:09 [trac] Developer/Git edited by chrisj (Update renamed pages.) http://devel.rtems.org/wiki/Developer/Git 2015-03-15T01:41:01 [trac] Developer/Git/Users edited by chrisj (Update after rename.) http://devel.rtems.org/wiki/Developer/Git/Users 2015-03-15T01:42:27 [trac] Developer/Git edited by chrisj (Rename updates.) http://devel.rtems.org/wiki/Developer/Git 2015-03-15T01:50:01 [trac] Coding created http://devel.rtems.org/wiki/Coding 2015-03-15T01:51:51 [trac] Developer/Coding/GenerateAPatch edited by chrisj (empty comment) http://devel.rtems.org/wiki/Developer/Coding/GenerateAPatch 2015-03-15T02:11:55 [trac] Developer/Tools created http://devel.rtems.org/wiki/Developer/Tools 2015-03-15T02:15:15 [trac] Developer/Tracing edited by chrisj (Update trace linker line.) http://devel.rtems.org/wiki/Developer/Tracing 2015-03-15T02:16:17 [trac] Developer/Tracing edited by chrisj (empty comment) http://devel.rtems.org/wiki/Developer/Tracing 2015-03-15T02:17:51 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Update Tools link.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T02:18:38 [trac] Developer/Tools edited by chrisj (Update trace linker link.) http://devel.rtems.org/wiki/Developer/Tools 2015-03-15T02:23:46 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Add more detail.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T02:26:09 *** siddharth7 has joined #rtems 2015-03-15T03:44:10 *** Hesham has joined #rtems 2015-03-15T04:01:55 *** ketul has quit IRC (Ping timeout: 246 seconds) 2015-03-15T04:29:26 [git] 91c250f by Sebastian Huber: [bsps/sparc: Fix linker section] http://devel.rtems.org/changeset/91c250f/rtems 2015-03-15T04:29:26 [git] 6db5e65 by Sebastian Huber: [score: Add _Objects_ISR_disable_and_acquire()] http://devel.rtems.org/changeset/6db5e65/rtems 2015-03-15T05:03:31 *** ketul has joined #rtems 2015-03-15T05:09:39 [trac] GSoC/2015 edited by ketul (empty comment) http://devel.rtems.org/wiki/GSoC/2015 2015-03-15T05:26:13 *** Hesham has quit IRC (Ping timeout: 264 seconds) 2015-03-15T05:52:53 *** gedare has joined #rtems 2015-03-15T05:52:53 *** ChanServ sets mode: +o gedare 2015-03-15T06:00:49 *** Hesham has joined #rtems 2015-03-15T07:19:24 *** gedare has quit IRC (Ping timeout: 252 seconds) 2015-03-15T08:04:33 *** siddharth7 has quit IRC (Read error: Connection reset by peer) 2015-03-15T08:04:58 *** siddharth7 has joined #rtems 2015-03-15T09:20:10 *** ragunath has joined #rtems 2015-03-15T09:47:43 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-15T10:14:43 *** ragunath_ has joined #rtems 2015-03-15T10:16:03 *** siddharth7 has quit IRC (Remote host closed the connection) 2015-03-15T10:35:40 *** ragunath_ has quit IRC (Ping timeout: 246 seconds) 2015-03-15T10:43:33 *** ragunath has joined #rtems 2015-03-15T10:53:31 *** ragunath has quit IRC (Ping timeout: 246 seconds) 2015-03-15T11:00:46 *** ragunath has joined #rtems 2015-03-15T11:02:01 *** gedare has joined #rtems 2015-03-15T11:02:02 *** ChanServ sets mode: +o gedare 2015-03-15T11:31:40 *** ketul has quit IRC (Ping timeout: 246 seconds) 2015-03-15T11:32:52 *** ketul has joined #rtems 2015-03-15T11:41:25 hello everyone 2015-03-15T11:41:38 *** gedare has quit IRC (Ping timeout: 264 seconds) 2015-03-15T11:41:46 hey beng-nl 2015-03-15T11:41:53 ahh MrDrJoel :-) 2015-03-15T11:41:56 et tu ? 2015-03-15T11:42:14 all well thank you 2015-03-15T11:42:27 i trust you enjoyed apple pie on pi day :-) 2015-03-15T11:42:39 Ides of March :) 2015-03-15T12:24:12 lo 2015-03-15T12:44:40 We have tickets to see Hozier Tuesday in Nashville. Have you heard of him? 2015-03-15T12:45:31 It is his birthday, he is Irish, and it is St Patrick's Day. Plus at the Ryman which is an incredibly historic and small place to see a show 2015-03-15T13:40:39 *** Davidbrcz has joined #rtems 2015-03-15T14:05:02 *** Hesham has quit IRC (Ping timeout: 264 seconds) 2015-03-15T14:23:31 beng-nl: are you there? 2015-03-15T14:35:45 i am 2015-03-15T14:37:41 ketul: ^ 2015-03-15T14:38:41 beng-nl: hi 2015-03-15T14:38:44 hi 2015-03-15T14:38:46 done something 2015-03-15T14:38:46 https://github.com/ketul93/RTEMS-on-BBB 2015-03-15T14:38:50 :) 2015-03-15T14:45:30 :) 2015-03-15T14:45:33 i love the YT link :) 2015-03-15T14:47:02 yeh . thanks. 2015-03-15T14:54:21 Looks like you added yourself to the list on the rtems wiki too. :) 2015-03-15T14:55:45 yep DtJoel_ 2015-03-15T14:55:59 *** gtament has joined #rtems 2015-03-15T14:56:01 preparing proposal 2015-03-15T14:56:53 DrJoel_ * 2015-03-15T14:58:53 I have a problem sb-set-builder fails and I don't why. Here is the log http://paste.kolibrios.org/show/373/ Seems like the problem is in python but I have python 2.6 and it is default version (running cmd 'python' invokes exactly 2.6.6 version). Also is it possible to disable all these checks, because sb-set-builder runs more than one hour till it fails?? Linux distro: CentOS 6. Thanks in advance! 2015-03-15T14:59:41 gtament: please open a ticket on devel.rtems.org 2015-03-15T15:00:25 I did yesterday, but I still got no answer 2015-03-15T15:02:04 gtament: just want to tell you can run 'apt-get update' 2015-03-15T15:03:04 ketul: how it is supposed to help in my case? 2015-03-15T15:05:31 I use Centos 6.x every day. 2015-03-15T15:06:05 *** Hesham has joined #rtems 2015-03-15T15:06:24 https://docs.rtems.org/rsb/ has a yum install command 2015-03-15T15:06:32 yum install autoconf automake binutils gcc gcc-c++ gdb make patch \ 2015-03-15T15:06:32 bison flex xz unzip ncurses-devel texinfo zlib-devel python-devel git 2015-03-15T15:10:09 DrJoel_: sb-check says environment is ok 2015-03-15T15:10:44 DrJoel_: And I definitely have all these installed 2015-03-15T15:12:07 which ticket 2015-03-15T15:13:32 DrJoel_: My message "I have a problem sb-set-builder fails and I don't why. Here is the log http://paste.kolibrios.org/show/373/ Seems like the problem is in python but I have python 2.6 and it is default version (running cmd 'python' invokes exactly 2.6.6 version). Also is it possible to disable all these checks, because sb-set-builder runs more than one hour till it fails?? Linux distro: CentOS 6. Thanks in advance!" 2015-03-15T15:15:57 search for rsb* to see details on what happened 2015-03-15T15:16:25 There is an option to dry run and download all the files. They sometimes fails 2015-03-15T15:16:40 DrJoel_: Here is the log http://paste.kolibrios.org/show/373/ 2015-03-15T15:16:50 I mean you can run the download first. Then run the buld 2015-03-15T15:17:23 DrJoel_: Ok, I'll try. Also is it possible to disable checks in set-builder? 2015-03-15T15:17:45 That is a buid failure in gdb. You are missing a python library 2015-03-15T15:19:31 DrJoel_: Thanks. I'll check it. Hopefully it will help 2015-03-15T15:21:09 there should be a config.log in the directory of the failure (in the build tree) with details 2015-03-15T15:23:22 DrJoel_: What about checks disabling? 2015-03-15T15:23:58 nott an rsb issue 2015-03-15T15:45:50 *** Davidbrcz has quit IRC (Ping timeout: 264 seconds) 2015-03-15T15:51:23 *** ragunath has left #rtems 2015-03-15T15:54:52 *** Hesham has quit IRC (Ping timeout: 245 seconds) 2015-03-15T15:57:21 *** gtament has quit IRC () 2015-03-15T16:01:00 *** kiwichris has quit IRC (Ping timeout: 246 seconds) 2015-03-15T16:02:08 beng-nl: You wrote about Booting without uboot. right? 2015-03-15T16:14:24 *** kiwichris has joined #rtems 2015-03-15T16:23:13 *** ketul has quit IRC (Ping timeout: 246 seconds) 2015-03-15T16:35:28 kiwichris: Do you remember the cause of this gdb configure error? http://paste.kolibrios.org/show/373/ 2015-03-15T16:35:41 DrJoel, evening 2015-03-15T16:35:45 A student asked on the list about it and was chatting earlier asking about it. I didn't remember 2015-03-15T16:35:56 Answered on the list. It is no devel lobs 2015-03-15T16:35:56 And good morning to you 2015-03-15T16:36:01 libs that is 2015-03-15T16:36:24 Michele is burned out from reading to me. She is reading her latest book out loud as a near final proofreading exercise 2015-03-15T16:36:25 this is such a common failure I am wondering if I need a way to test for libs in the RSB 2015-03-15T16:36:44 He claimed to have installed python-dev RPM. This is on Centos and I use that daily 2015-03-15T16:37:05 we have discussed this before and zlib is the other big pain 2015-03-15T16:37:57 where was this discussed ? 2015-03-15T16:38:23 You and I have before. not sure when or how on rsb and lib checking 2015-03-15T16:38:28 The devel libs that is 2015-03-15T16:38:31 he installed 2015-03-15T16:38:49 it looks like he might have it working based on that last email. He was on irc earlier 2015-03-15T16:38:58 Ah ok 2015-03-15T16:39:44 GDB detected the version of python but not the libraries. The gdb error message could be more helpful because people think all they need is python which they have 2015-03-15T16:39:53 FWIW I am going to put a keyword "warning" on all tickets about warnings. I am filtering all warnings which are known with tickets out of my analysis and focusing on ones not addressed by solving or tickets. 2015-03-15T16:40:28 The one for not having 32-bit libs on a 64-bit host are confusing (and have spelling errors). :) 2015-03-15T16:40:29 Should the depreciate wiki page go under Coding ? 2015-03-15T16:40:33 I submitted a patch for that 2015-03-15T16:40:51 Yes. And I probably need to add a section on what I have been doing this weekend 2015-03-15T16:41:19 I will move it. I have moved and updated a number of wiki pages. I added a top level Tracing page. 2015-03-15T16:41:56 Awesome. I probably need to add a news item on the FSW papers or add the PDFs to a news item on it if we have one :) 2015-03-15T16:42:39 [trac] Developer/Coding/Deprecating edited by chrisj (Update TOC.) http://devel.rtems.org/wiki/Developer/Coding/Deprecating 2015-03-15T16:43:02 I think we are down to primarily warnings with tickets or on 16/64 bit targets. 2015-03-15T16:44:11 [trac] Developer edited by chrisj (Remove depreciating ) http://devel.rtems.org/wiki/Developer 2015-03-15T16:46:19 [trac] Developer/Coding edited by chrisj (Add deprecating interfaces link.) http://devel.rtems.org/wiki/Developer/Coding 2015-03-15T16:47:13 verm__, in Trac's index page some heading are black and others are not. What determines this ? 2015-03-15T16:48:25 [trac] Developer/VirtualMachines/VirtualBox edited by chrisj (Update TOC after rename.) http://devel.rtems.org/wiki/Developer/VirtualMachines/VirtualBox 2015-03-15T16:55:51 [trac] Developer edited by chrisj (Add tracing and tools.) http://devel.rtems.org/wiki/Developer 2015-03-15T16:57:01 DrJoel, I have been adding pages are each point in a wiki path so you can click back to that level and see what is below. 2015-03-15T16:58:15 [trac] Developer/Tools edited by chrisj (Add RSB Link.) http://devel.rtems.org/wiki/Developer/Tools 2015-03-15T17:02:45 great! 2015-03-15T17:04:05 DrJoel, can you please review the Tracing page ? 2015-03-15T17:04:14 I just sent an email to newlib@. Interesting that adding CPU flags can change the sizeof(ptr) and thus break the inttypes.h logic on h8 and m32c. I think this may be the root cause of most of our printf warnings on 16 bit targets 2015-03-15T17:05:18 Oh really, I thought intypes.h was a very carefully thought out piece of work ? 2015-03-15T17:05:31 lol.. 2015-03-15T17:05:52 hehe 2015-03-15T17:06:12 It works great with default arguments. My script to test doesn't cycle over all multilib variants. I write a 6 line C program with a printf() and push it through all the compilers. Need to cycle over all the flags 2015-03-15T17:06:31 Nice. 2015-03-15T17:06:46 I would raise a bug on gcc 2015-03-15T17:06:53 The file is in there. 2015-03-15T17:07:17 It might improve the signal to noise ratio 2015-03-15T17:13:31 *** gedare has joined #rtems 2015-03-15T17:13:31 *** ChanServ sets mode: +o gedare 2015-03-15T17:13:37 [trac] Developer/Coding/Deprecating edited by joel.sherrill (Add information on disabling deprecated warnings and updating support/test code.) http://devel.rtems.org/wiki/Developer/Coding/Deprecating 2015-03-15T17:13:57 What's the link? 2015-03-15T17:14:18 Goto the Developer page and then follow the links :) 2015-03-15T17:16:52 It is 22C here today and I am supposed to follow links? 2015-03-15T17:16:58 Et Tu kiwichris ? 2015-03-15T17:17:13 About the same temp here 2015-03-15T17:17:16 brrrr 2015-03-15T17:17:33 http://devel.rtems.org/wiki/Developer then the tracing link 2015-03-15T17:17:39 Ides of March you know :) 2015-03-15T17:17:54 niced 2015-03-15T17:17:58 [trac] Developer/Tools/RSB edited by chrisj (Clean up the page.) http://devel.rtems.org/wiki/Developer/Tools/RSB 2015-03-15T17:18:18 [trac] Developer/Tracing edited by joel.sherrill (Add Oxford comma :)) http://devel.rtems.org/wiki/Developer/Tracing 2015-03-15T17:18:36 rpi2 patch can go in.. joel you able to test it? 2015-03-15T17:18:42 oh the comma police 2015-03-15T17:18:48 oh boy. 2015-03-15T17:19:18 DrJoel, https://devel.rtems.org/wiki/Developer/Tools/RSB has a CentOS 7 section. Is that still valid ? 2015-03-15T17:19:18 Tomorrow I can get Jeff to. I just had the boards shipped here because I get Saturday and Sunday delivery :) 2015-03-15T17:19:31 ok 2015-03-15T17:19:59 Is there harm in just pushing it? Alan said it was functional. If there aren't style issues, just push it 2015-03-15T17:20:29 i don't see any problems its all bsp-specific anyway so i don't care much 2015-03-15T17:20:50 welp, it's almost all build-ssytem changes tbh 2015-03-15T17:21:06 Let's Eat Grandma Or Let's Eat, Grandma Commas Save Lives 2015-03-15T17:23:28 Can we generate CTF now? 2015-03-15T17:26:29 [trac] Developer/Tracing edited by joel.sherrill (Minor corrections) http://devel.rtems.org/wiki/Developer/Tracing 2015-03-15T17:27:08 I am not sure the page emphasizes that the CTF support in the trace-linker, gdb python code to fetch, etc is still a plan 2015-03-15T17:33:34 *** gedare has quit IRC (Ping timeout: 245 seconds) 2015-03-15T17:37:18 It says this is still be done. 2015-03-15T17:37:29 It if changes we change the page 2015-03-15T17:38:04 *** jpc has joined #rtems 2015-03-15T17:38:34 hey, all! 2015-03-15T17:38:39 Just wanting to make sure it is clear that it is "to be done" to everyone 2015-03-15T17:40:56 It does say it under the heading. Is that not clear enough ? 2015-03-15T17:41:34 I need to provide enough details so GSoC and others can clearly see the work to be done. 2015-03-15T17:43:04 Extracting data from the target with Python should be possible. We will need to make sure we have a simple memory based protocol with the capture engine. 2015-03-15T17:43:28 hey guys. are we talking about a FOSS GSoC project? 2015-03-15T17:43:58 jpc, Hi. Introduce yourself to Joel. 2015-03-15T17:45:13 [trac] Developer/Tracing edited by chrisj (Update the CTF status line.) http://devel.rtems.org/wiki/Developer/Tracing 2015-03-15T17:46:23 [trac] Developer/Tools edited by chrisj (Add link to tracing page.) http://devel.rtems.org/wiki/Developer/Tools 2015-03-15T17:47:00 Hi Joel. My name's Jim, and I just joined the IRC channel recently. I'm really interested in the RTEMS platform, though my current employer may restrict my ability to contribute (unfortunately). I'm a software engineer at Micrium, and a big advocate of FOSS! 2015-03-15T17:48:03 jpc, we are part of GSoC and have been for a number of years. Currently students are looking at RTEMS and the projects and putting forward proposals. 2015-03-15T17:49:48 yes, i'm aware of that. I'm a recent graduate of FGCU, and actually heard about RTEMS through my professor there. 2015-03-15T17:49:55 jpc, experiences users or anyone suitable can mentor students which is a rewarding thing to do. 2015-03-15T17:50:30 well, i'm not an experienced user, but I would like to be. and, I would like to mentor students back at my uni. 2015-03-15T17:50:46 *in embedded systems in general* 2015-03-15T17:51:22 jpc, even giving a talk at the Uni about GSoC and promoting is a really good first step. 2015-03-15T17:52:51 well, my professor actually has a series of guest lecturers every year. if there is anyone at OAR or in the RTEMS community that would like to talk to students and faculty about GSoC and RTEMS, I might be able to arrange something. 2015-03-15T17:54:14 personally, I do not feel experienced enough to give a full lecture. But I feel like I'm moving in that direction 2015-03-15T17:57:19 is there a recommended platform/eval board for beginning RTEMS development? 2015-03-15T18:02:58 alright, i may be copping out by saying i'm not experienced enough. If I were to be a mentor for some students back at FGCU, where should I start? Are there any good resources or guidlines out there for mentoring? 2015-03-15T18:05:09 hi jpc. feel free to join in however you can 2015-03-15T18:05:36 where if FGCU? 2015-03-15T18:05:53 Fort Myers, FL 2015-03-15T18:06:25 jpc .. starting with simulators is cheap and easy. Then moving on to something like the ARM Zynq, Beagle, or Pi (or old PC) is a cheap real HW alternative 2015-03-15T18:06:36 that's not driving distance for me unfortunately. :( 2015-03-15T18:06:41 I am in Huntsville 2015-03-15T18:06:54 And kiwichris would get tired swimming from Sydney 2015-03-15T18:06:56 yep, i saw on the website 2015-03-15T18:07:07 hah! yeah, that's a bit of a distance 2015-03-15T18:08:25 Contrary to the normal view some kiwi's can fly ! 2015-03-15T18:09:08 from NZ, yeah? 2015-03-15T18:09:25 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Update.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T18:09:34 bb 2015-03-15T18:09:46 i've always wanted to visit 2015-03-15T18:09:49 brb 2015-03-15T18:10:25 mostly for snowboarding, not just to see the shire 2015-03-15T18:12:52 jpc, long ago 2015-03-15T18:16:16 so, have you been to the states much? 2015-03-15T18:34:02 jpc, over the years yes; worked on a large project in NY and Otawa in the '90s; NATO AWACS voice switch that runs on RTEMS. 2015-03-15T18:44:46 oh, nice. what was working for the gov't like? were you in management or devlopment/engineering? 2015-03-15T18:53:20 jpc, working with contractors providing the equipment used. 2015-03-15T19:20:31 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Update.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T19:35:22 *** jpc has quit IRC (Ping timeout: 246 seconds) 2015-03-15T20:29:26 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Update.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T20:33:01 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Remove excess white space) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T21:05:41 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Update.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T21:06:16 [trac] Developer/Tracing/Trace_Linker edited by chrisj (INI formatting.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T21:07:59 [trac] Developer/Tracing/Trace_Linker edited by chrisj (Update formatting.) http://devel.rtems.org/wiki/Developer/Tracing/Trace_Linker 2015-03-15T21:16:52 [trac] Developer/Tools edited by chrisj (Fix links.) http://devel.rtems.org/wiki/Developer/Tools 2015-03-15T21:17:51 [trac] Developer/Tools edited by chrisj (Minor fixes.) http://devel.rtems.org/wiki/Developer/Tools 2015-03-15T21:30:07 kiwichris: what do you mean? 2015-03-15T21:30:28 i'm not sure what area(s) you're talking about 2015-03-15T21:32:56 https://devel.rtems.org/wiki/TitleIndex then Debugging, Coding etc 2015-03-15T21:33:01 They are black 2015-03-15T21:33:18 that's just the subheading 2015-03-15T21:33:20 you can't edit that page 2015-03-15T21:33:23 it's automatically generated 2015-03-15T21:34:21 Sure. What is the bit that causes this ? I cannot see it. 2015-03-15T21:34:40 causes what? the title index? 2015-03-15T21:34:46 just edit the page and you'll see 2015-03-15T21:35:44 Ah thanks. I see now 2015-03-15T21:37:03 Is 2 too small ? 2015-03-15T21:37:48 It looks ok until TBR but so is that ok ? 2015-03-15T21:42:43 why do you want to lower it? 2015-03-15T21:42:54 the titleindex isn't meant to be used as a way to navigate the site 2015-03-15T21:43:07 it's meant to see the hierarchy when laying it out or finding orphaned pages 2015-03-15T21:43:13 so if you lower that it will make it less useful 2015-03-15T21:50:57 But 4 seems a little more. 2015-03-15T21:51:19 oh wait it's only 2 right now? 2015-03-15T21:51:28 ahhh 2015-03-15T21:51:35 sorry i thought the reverse 2015-03-15T21:51:56 No it is not 2 it is 4. 2015-03-15T21:52:09 And we have a few areas, say git which is 3 pages 2015-03-15T21:53:32 have you tried setting it to 2 and doing a preview? 2015-03-15T21:55:19 Yes and I thought it looked ok until TBR 2015-03-15T21:55:35 if you think it looks better then change it 2015-03-15T21:55:43 Will do, thanks 2015-03-15T21:55:57 [trac] TitleIndex edited by chrisj (Developer lists show up better.) http://devel.rtems.org/wiki/TitleIndex 2015-03-15T21:57:10 [trac] Developer/Coverage/Theory edited by chrisj (empty comment) http://devel.rtems.org/wiki/Developer/Coverage/Theory 2015-03-15T22:01:16 [trac] Coverage created http://devel.rtems.org/wiki/Coverage 2015-03-15T22:02:29 verm__, we have Projects and Developer/Projects in the wiki and this is a bit confusing 2015-03-15T22:12:05 there are two different types 2015-03-15T22:12:15 projects that are developer only and ojnes suitable for users 2015-03-15T22:21:17 *** mkhoory has joined #rtems 2015-03-15T22:44:33 *** mkhoory has quit IRC (Read error: Connection reset by peer) 2015-03-15T23:26:28 *** mkhoory has joined #rtems