2013-03-11T00:44:23 *** arvind_khadri has quit IRC (Ping timeout: 260 seconds) 2013-03-11T01:45:22 If you are tracking the Source Builder, I have moved all patches into http://git.rtems.org/rtems-tools.git/ and the Source Builder now downloads them from this repo. 2013-03-11T02:18:35 *** monstr__ has joined #rtems 2013-03-11T02:29:36 *** shineworld has joined #rtems 2013-03-11T02:29:49 hi all 2013-03-11T02:30:23 any BSP (also beta) for low cost board like cubieboard (www.cubieboard.org) ? 2013-03-11T03:05:49 *** sebhub has joined #rtems 2013-03-11T03:06:01 good morning 2013-03-11T03:06:21 morning sebhub 2013-03-11T03:28:19 *** lcpfnvcy has joined #rtems 2013-03-11T04:19:17 *** the9a3eedi has quit IRC (Ping timeout: 256 seconds) 2013-03-11T04:21:59 *** lcpfnvcy has quit IRC (Quit: leaving) 2013-03-11T04:22:19 *** lcpfnvcy has joined #rtems 2013-03-11T04:24:33 *** shineworld has quit IRC (Quit: Leaving) 2013-03-11T05:31:40 *** MegaAlex|away has quit IRC (Read error: Operation timed out) 2013-03-11T05:33:42 *** MegaAlex|away has joined #rtems 2013-03-11T05:41:28 *** MegaAlex|away is now known as MegaAlex 2013-03-11T06:19:12 *** MegaAlex is now known as MegaAlex|away 2013-03-11T06:30:41 *** monstr__ has quit IRC (Ping timeout: 248 seconds) 2013-03-11T06:32:40 *** monstr__ has joined #rtems 2013-03-11T06:37:52 sebhub, around ? 2013-03-11T07:23:19 *** antgreen has quit IRC (Quit: Leaving) 2013-03-11T07:56:07 *** monstr__ has quit IRC (Remote host closed the connection) 2013-03-11T07:59:25 *** monstr__ has joined #rtems 2013-03-11T08:21:34 *** gedare has joined #rtems 2013-03-11T09:01:20 hi 2013-03-11T09:16:27 *** antgreen has joined #rtems 2013-03-11T09:36:02 *** Hesham has joined #rtems 2013-03-11T10:10:43 hey Hesham. 2013-03-11T10:11:04 hi gedare. How are you ? 2013-03-11T10:13:40 pretty good. busy 2013-03-11T10:13:46 most of the time :) 2013-03-11T10:18:05 Yeah I noticed that, specially you work to adjust rtems for the coming GSoC 2013-03-11T10:18:43 yeah. we're trying to get the repo and tools to stabilize before gsoc, and hopefully make an rtems release 2013-03-11T10:19:02 4.12 ? 2013-03-11T10:19:22 4.11 2013-03-11T10:20:12 Is not the current release 4.11 ? 2013-03-11T10:22:34 I always configure rtems with that command line "../rtems/configure --target=arm-rtems4.11" that's why I though the current release is 4.11 2013-03-11T10:25:20 4.11 is the current development version 2013-03-11T10:25:38 we usually name the development version what the next release will be 2013-03-11T10:25:57 ah, I see :) 2013-03-11T11:16:05 *** Hesham has left #rtems 2013-03-11T12:17:33 *** sebhub has quit IRC (Remote host closed the connection) 2013-03-11T12:21:05 *** antgreen has quit IRC (Ping timeout: 248 seconds) 2013-03-11T12:36:41 *** antgreen has joined #rtems 2013-03-11T12:52:19 *** MegaAlex|away is now known as MegaAlex 2013-03-11T13:08:15 *** monstr__ has quit IRC (Ping timeout: 260 seconds) 2013-03-11T13:24:29 *** monstr__ has joined #rtems 2013-03-11T14:01:26 *** gedare has quit IRC (Ping timeout: 252 seconds) 2013-03-11T14:16:33 *** monstr__ has quit IRC (Ping timeout: 240 seconds) 2013-03-11T15:18:16 *** antgreen has quit IRC (Quit: Leaving) 2013-03-11T15:26:19 *** DrJoel has joined #rtems 2013-03-11T15:26:20 *** DrJoel has joined #rtems 2013-03-11T15:26:20 *** ChanServ sets mode: +o DrJoel 2013-03-11T15:40:24 *** rtemsLogger` has joined #rtems 2013-03-11T15:47:14 *** rtemsLogger has quit IRC (Ping timeout: 246 seconds) 2013-03-11T16:13:47 Hi Joel 2013-03-11T16:20:35 hey kiwichris!!! 2013-03-11T16:21:00 I have a number of patches in rtems-tools.git now. Have you been able to check them ? 2013-03-11T16:21:10 We ordered the proof copy of Michele's book a few hours ago and I filed for a formal copyright 2013-03-11T16:21:26 Not yet. I was up to 2am this morning and every night this weekend getting her book finished. 2013-03-11T16:21:47 Also texinfo 5.0 had some pre-release snapshots I was testing. 2013-03-11T16:21:53 Ah ok. 2013-03-11T16:21:54 Do you want me to try a mips toolset? 2013-03-11T16:22:52 https://www.facebook.com/photo.php?fbid=10151470034800914&set=exp.10151470041050914.unitary&type=1&theater shows her cover 2013-03-11T16:22:54 No yet. Looks like the MIPS patches failed to apply in the build all test 2013-03-11T16:23:05 Just give me a list 2013-03-11T16:23:57 list of ? 2013-03-11T16:36:19 list of what to test in your desired order :) 2013-03-11T16:36:58 Microblaze please. I added the patch to fix the ldscript 2013-03-11T16:37:10 It builds but I have not tested the compiler. 2013-03-11T16:45:06 microblaze started 2013-03-11T16:45:12 Thanks. 2013-03-11T16:45:31 On mutex it takes about 10min on average, some are shorter. 2013-03-11T16:45:42 It is nice having the patches in git 2013-03-11T16:46:23 This is a big improvement in configuration management for the project. We just need to figure out what is needed and what needs to be updated. :) 2013-03-11T16:46:57 Yeah 2013-03-11T16:47:57 The long paths in gcc as a single tarball breaks mingw. It cannot delete them with the standard 'rm' command due to file name length 2013-03-11T16:50:10 Wow! That's an odd limitation. Do they know that? 2013-03-11T16:50:31 Not sure. I will raise a ticket on it. 2013-03-11T16:51:26 school drop off, back soon 2013-03-11T16:56:26 heading home myself 2013-03-11T16:56:28 later 2013-03-11T16:56:30 *** DrJoel has quit IRC (Quit: Light travels faster then sound, which is why some people appear bright, until you hear them speak) 2013-03-11T17:04:49 *** antgreen has joined #rtems 2013-03-11T19:21:38 *** MegaAlex is now known as MegaAlex|away 2013-03-11T23:22:58 *** krispin has joined #rtems 2013-03-11T23:40:58 *** krispin has quit IRC (Ping timeout: 245 seconds) 2013-03-11T23:53:16 *** arvind_khadri has joined #rtems 2013-03-11T23:56:47 *** arvind_khadri has quit IRC (Client Quit) 2013-03-11T23:56:57 *** arvind_khadri has joined #rtems 2013-03-12T01:00:11 *** krispin has joined #rtems 2013-03-12T01:04:06 *** krispin has quit IRC (Read error: No route to host) 2013-03-12T01:04:13 *** krispin1 has joined #rtems 2013-03-12T01:24:57 *** krispin has joined #rtems 2013-03-12T01:24:58 *** krispin1 has quit IRC (Read error: Connection reset by peer) 2013-03-12T01:43:33 *** the9a3eedi has joined #rtems 2013-03-12T01:47:16 *** krispin has quit IRC (Ping timeout: 246 seconds) 2013-03-12T01:53:05 *** monstr__ has joined #rtems 2013-03-12T02:28:10 *** arvind_khadri has quit IRC (Ping timeout: 260 seconds) 2013-03-12T02:51:38 *** arvind_khadri has joined #rtems 2013-03-12T02:54:00 *** lcpfnvcy has quit IRC (Ping timeout: 252 seconds) 2013-03-12T02:58:15 *** lcpfnvcy has joined #rtems 2013-03-12T03:02:39 *** sebhub has joined #rtems 2013-03-12T03:03:24 good morning 2013-03-12T03:04:22 sebhub, hi 2013-03-12T03:04:29 hi chris 2013-03-12T03:04:44 Have you managed to have a look at the rtems-tools.git repo ? 2013-03-12T03:04:51 Just for the layout. 2013-03-12T03:05:06 yes, i cloned it and committed the arm eabi patch 2013-03-12T03:05:11 Nice 2013-03-12T03:05:13 yesterday 2013-03-12T03:05:59 Becareful deleting patches. 2013-03-12T03:06:13 ok, this was my next question, why do we need them? 2013-03-12T03:06:18 Just because it is next does not mean there are configurations in the wild that may reference the old patch 2013-03-12T03:07:45 maybe we should add a README file which explains the patch life cycle 2013-03-12T03:08:47 It needs to be part of the RTEMS Tools Requirements 2013-03-12T03:08:58 which is a boarder document 2013-03-12T03:09:04 hm, yes 2013-03-12T03:09:07 It is the one in Google Docs I am working on 2013-03-12T03:09:30 I have started to add specifics about rtems-tools.git 2013-03-12T03:10:39 thats nice, i am sorry for not being more active in this area, but i am quite overloaded at the moment 2013-03-12T03:10:45 To update the RTEMS Source Builder to the new patch this file changes .... 2013-03-12T03:10:45 http://git.rtems.org/chrisj/rtems-source-builder.git/tree/rtems/config/4.11/rtems-arm.bset 2013-03-12T03:11:21 You are fine, all I need is checking of the RSB configs and any patches I need. 2013-03-12T03:11:48 I will switch to binutils 2.23.x once all targets are building. 2013-03-12T03:12:14 Hit this today http://sourceforge.net/p/mingw/bugs/1948/ 2013-03-12T03:12:25 I am not impressed with the answer. 2013-03-12T03:13:24 One thing that would help is building and testing the NIOS2 tools. Are you able to spare a few cycles to do this ? 2013-03-12T03:13:41 yes, but not this week 2013-03-12T03:14:05 Ok 2013-03-12T03:14:08 i posted some patches for the binutils to the list 2013-03-12T03:14:16 once they are integrated, we can do this 2013-03-12T03:14:30 I can add the patches to the RSB 2013-03-12T03:14:49 they are not approved yet 2013-03-12T03:14:54 I will look for the links. 2013-03-12T03:15:08 That is ok. Approval is one thing, being public is another. 2013-03-12T03:15:19 http://sourceware.org/ml/binutils/2013-03/ 2013-03-12T03:15:26 search for "NIOS2" 2013-03-12T03:15:56 Will do. 2013-03-12T03:16:13 ok, i should check the mails first 2013-03-12T03:16:22 some patches are already committed 2013-03-12T03:16:33 Nice. 2013-03-12T03:17:50 I have the LM32 building C and C++ using gcc-4.5.3 which the LM folk said is the one that works. 2013-03-12T03:18:14 Hit a horrible bug where it would not build if make had a -j option 2013-03-12T03:18:46 It is a good test because it is gcc-4.5.3 with newlib-2.0.0 2013-03-12T03:18:47 yes, the -j only works sometimes 2013-03-12T04:05:33 *** the9a3eedi has quit IRC (Ping timeout: 245 seconds) 2013-03-12T04:07:08 *** sevikkk has quit IRC (Ping timeout: 256 seconds) 2013-03-12T04:11:01 sebhub, switch to binutils-2.23.1 ? 2013-03-12T04:18:21 *** MegaAlex|away is now known as MegaAlex 2013-03-12T04:26:12 *** sevikkk has joined #rtems 2013-03-12T05:10:08 *** the9a3eedi has joined #rtems 2013-03-12T05:29:06 *** MegaAlex is now known as MegaAlex|away 2013-03-12T06:32:37 *** the9a3eedi has quit IRC (Ping timeout: 248 seconds) 2013-03-12T07:25:03 *** antgreen has quit IRC (Quit: Leaving) 2013-03-12T08:25:24 *** gedare has joined #rtems 2013-03-12T08:50:03 *** the9a3eedi has joined #rtems 2013-03-12T09:26:18 good morning. 2013-03-12T09:59:26 *** the9a3eedi has quit IRC (Ping timeout: 255 seconds) 2013-03-12T11:34:16 *** arvind_khadri has quit IRC (Ping timeout: 272 seconds) 2013-03-12T11:55:28 *** monstr__ has quit IRC (Read error: Operation timed out) 2013-03-12T11:59:01 *** DrJoel has joined #rtems 2013-03-12T11:59:01 *** DrJoel has joined #rtems 2013-03-12T11:59:01 *** ChanServ sets mode: +o DrJoel 2013-03-12T12:12:59 *** sebhub has quit IRC (Remote host closed the connection) 2013-03-12T14:33:46 gedare are you there? 2013-03-12T14:33:57 hi DrJoel 2013-03-12T14:34:03 for another..20 min or so 2013-03-12T14:34:25 After the ChangeLog deletion, I got messages about missing ChangeLog during bootstrap. 2013-03-12T14:34:36 I think I have corrected them all but would like a double-check 2013-03-12T14:35:15 ah 2013-03-12T14:35:47 bootstrapping. 2013-03-12T14:36:17 I have microblaze building again using Chris' rtems-source-builder for the tools. 2013-03-12T14:36:29 If I can get it to run hello world on the gdb simulator, I am pushing it upstream 2013-03-12T14:36:51 that puts it on par with a few other ports which do not have interrupt support. And it is a base to add interrupt support and real HW drivers to. 2013-03-12T14:37:04 k 2013-03-12T14:38:27 we'll need an arch/bsp status matrix for status of differnet ports and bsps 2013-03-12T14:39:24 yep.. as painful as it may be, it also needs to be in the wiki since it is tied to the release and release notes have always gone in the wiki 2013-03-12T14:48:30 ./contrib/rtemsrpms 2013-03-12T14:48:30 Makefile.am: error: required file './ChangeLog' not found 2013-03-12T14:48:30 autoreconf: automake failed with exit status: 1 2013-03-12T14:48:30 grep: Makefile.in: No such file or directory 2013-03-12T14:49:33 Can you tell which Makefile.am? 2013-03-12T14:49:45 not yet..peering around 2013-03-12T14:49:59 I grep'ed for ChangeLog at one point and didn't see it. I wonder if it is required by the autotools. 2013-03-12T14:50:10 i think maybe 2013-03-12T14:50:21 contrib/rtemsrpms/config.sub 2013-03-12T14:50:21 33:# diff and a properly formatted GNU ChangeLog entry. 2013-03-12T14:50:44 you can ask kiwichris about this directory anyway 2013-03-12T14:50:47 we might be able to nuke it 2013-03-12T14:50:56 The `--gnu' option (or `gnu' in the `AUTOMAKE_OPTIONS' variable) causes 2013-03-12T14:50:56 `automake' to check the following: 2013-03-12T14:50:56 * The files `INSTALL', `NEWS', `README', `AUTHORS', and `ChangeLog', 2013-03-12T14:50:56 plus one of `COPYING.LIB', `COPYING.LESSER' or `COPYING', are 2013-03-12T14:50:56 required at the topmost directory of the package. 2013-03-12T14:53:08 Ralf felt free to remove the other parts and move to his own world.. 2013-03-12T14:53:23 this is old stuff 2013-03-12T14:53:25 Perhaps I should ask on devel... my day to be the petrol delivery man. :) 2013-03-12T14:53:43 commit 4b88b73b1bb039b3d088ea3dfe9ae19d511197b5 2013-03-12T14:53:44 Author: Chris Johns 2013-03-12T14:53:44 Date: Sun Sep 3 10:50:36 2006 +0000 2013-03-12T14:53:44 Default to not building bsps. Too many problems 2013-03-12T14:53:50 that's the most recent relevant commit to this directory 2013-03-12T14:54:07 yeah. 2013-03-12T14:54:12 We have two votes.. should I nuke it and ask forgiveness later :) 2013-03-12T14:54:17 or send an email first 2013-03-12T14:54:26 send an email, let chris or ralf weigh in 2013-03-12T14:54:48 kiwichris will probably be around in a little while 2013-03-12T14:55:15 Should I wait to send the email until then? 2013-03-12T14:55:17 there are probably other garbage directories bit rotting too. 2013-03-12T14:55:21 sure if you're hanging out 2013-03-12T14:55:28 I am here for a while 2013-03-12T14:55:32 i have to go..may check in if i have time 2013-03-12T14:55:41 ok .. 2013-03-12T14:56:28 I emailed you with a set of options ... I expect the best is to try for the July date unless the April date magically happens and is Ok for you 2013-03-12T14:57:47 ok. it might be better for me to attend at jhuapl rather than travel right now 2013-03-12T14:58:14 i'll try to figure it out. 2013-03-12T14:58:18 bbl 2013-03-12T14:58:24 bye 2013-03-12T15:07:06 *** MegaAlex|away is now known as MegaAlex 2013-03-12T15:30:50 back for a bit 2013-03-12T15:36:09 DrJoel: Do you know if it is possible to "lift" the rtems Chain implementation without having to pull all of RTEMS? I reimplemented a subset of it for something, but it might be nice to have it as a module. 2013-03-12T15:36:45 i imagine if i look around some more i can find something suitable instead 2013-03-12T15:37:32 ah.. the ISR synchronization would be a problem 2013-03-12T15:37:42 hmmmm 2013-03-12T15:37:59 Long ago, I pulled it out for use on linux.. If you handle the isr stuff, the rest is painless. 2013-03-12T15:38:10 yeah that's kind of what it looks like. 2013-03-12T15:38:39 i subsetted prepend and append.. 2013-03-12T15:38:46 just need a deque 2013-03-12T15:40:13 That would be get or some variant. 2013-03-12T15:40:57 yeah. i only needed about 50 LOC or so from the chain code 2013-03-12T15:41:27 That's about the core of it 2013-03-12T15:41:32 heh. yep 2013-03-12T15:41:55 i think someone mentioned something on one of the mls related to this and pointed to one of the *bsd implementations of lists 2013-03-12T15:42:37 OAR has a high school student here next week shadowing/interning... Jennifer is going to try to teach them enough to get the Livermore posix examples ported. 2013-03-12T15:42:47 Then some ported to Classic API as alternative examples. 2013-03-12T15:43:06 If he is quick, we have other posix tutorials on net to let him work through. 2013-03-12T15:43:37 nice 2013-03-12T15:43:54 i'm having to rewrite some code from Classic into pthreads 2013-03-12T15:44:06 so i can run it in linux too 2013-03-12T15:46:00 ahh.. sys/queue.h 2013-03-12T15:46:02 man queue 2013-03-12T16:03:27 I have a branch waiting to kill contrib 2013-03-12T16:04:19 good 2013-03-12T16:05:20 we should consider axing the tools/ subdirectory too. 2013-03-12T16:05:42 gtg. bye! 2013-03-12T16:06:38 some of those are used.. need to audit 2013-03-12T16:09:29 in fact all of the tools/cpu are used. Off the top of my head I know at least 1/2 of tools/build are used. 2013-03-12T16:10:02 *** gedare has quit IRC (Ping timeout: 246 seconds) 2013-03-12T16:12:41 DrJoel, hi 2013-03-12T16:13:19 hey. look back.. gedare and I were talking about you.. see about contrib/ 2013-03-12T16:14:22 what is in it ? 2013-03-12T16:15:18 mingw and rtemsrpm infrastructure that apparently hasn't been updated in 7 years 2013-03-12T16:15:38 The last change was yours, ironically enough 2013-03-12T16:16:34 Oh ha rtemsrpms ? hmmm ? I created it to attempt to build BSPs and CPU kit and Ralf sent me a "special" private email so it died and in CVS you could not remove a dir 2013-03-12T16:16:54 Thought I was helping ? haha 2013-03-12T16:17:11 It can goes 2013-03-12T16:17:15 It cah go 2013-03-12T16:17:38 That's three votes.. I am killing it 2013-03-12T16:17:57 The MinGW can go 2013-03-12T16:18:17 Do you see any reason to discuss this on the list? I really don't because that opens the door to comments I don't want. 2013-03-12T16:18:23 No need 2013-03-12T16:18:32 Better to kill it and say "unmaintained" and clearly out of date. 2013-03-12T16:19:23 done.. have your flame retardent suit ready .. just in case. 2013-03-12T16:19:35 I'm trying to kill cruft 2013-03-12T16:20:13 It is fine. 2013-03-12T16:21:36 Nice to see the microblaze is progressing. The issue was building with a -j option. Removed that and it worked 2013-03-12T16:23:26 great! 2013-03-12T16:24:12 I just want to get it to run hello world and then push it upstream. That makes it have the same level of functionality as the m32c and m32r. If someone wants to add interrupts, they can do it as a volunteer or pay someone. But it gets a lot of the basics in place. 2013-03-12T16:24:16 brb 2013-03-12T16:31:04 *** antgreen has joined #rtems 2013-03-12T16:37:03 DrJoel, take a look at https://sourceforge.net/p/mingw/bugs/1948/ 2013-03-12T16:39:37 Could rm check that the path is above the PATH_MAX and split it into a basename and dirname.. then chdir, rm and change back in the worst case 2013-03-12T16:39:56 but the resolution sucks.. we plan to reimplement everything so f**k the current bugs.. instant close 2013-03-12T16:40:16 I am sure it could. The rm is a Unix tool and I suppose never had this limit 2013-03-12T16:40:54 Yeah it is not endearing 2013-03-12T16:41:26 I think it is just an edge case and could be addressed at some level. but the way they closed it was not. No thought into what could be done.. whether other systems might have a limit 2013-03-12T16:41:34 even a 32k limit could be exceeded. 2013-03-12T16:42:06 Agreed, and I will keep posting day after day 2013-03-12T16:43:15 The key point is the purpose of MinGW is to build open source packages and a main one gcc is much more difficult because you cannot automatically delete it 2013-03-12T16:44:56 I wonder if there is any polite way to get gcc involved. Are the mingw gcc maintainers also the mingw maintainers? 2013-03-12T16:46:55 Maybe. Lets just let the current path run a while. I know Earnie will grow tied of my postings but I first want to see if there the rm is POSIX compatible 2013-03-12T16:47:49 I think it is not because I suspect the standard when it is talking about "exceed system limitation" it is things like memory 2013-03-12T16:51:39 *** antgreen has quit IRC (Quit: Leaving) 2013-03-12T17:14:31 *** gedare has joined #rtems 2013-03-12T17:17:03 did not know the tools/ was being used. 2013-03-12T17:18:59 tools/nios2 is likely used given the new nature of the port. 2013-03-12T17:19:54 tool/build/ has a number of things line rtems-bin2c and packhex which I know are used. Somethings in there may not be used. And may be able to be moved to the merge-helpers, or maintenance helpers in rtems-testing but more is used than not I think 2013-03-12T17:19:57 It would have to be audited 2013-03-12T17:20:13 Not sure. It is a tool to convert the NIOS2 project thingy data into a BSP and while it is a nice idea I think it is not upto date 2013-03-12T17:20:23 Do you think a GSOC project could be to convert the RTEMS Scheduler Simulator to waf? It is a native app that just pulls in files from rtems 2013-03-12T17:20:39 The NIOS2 comes with a chip builder thing and this is suppose to be the output. 2013-03-12T17:20:43 Each item would have to be discussed. 2013-03-12T17:21:04 i think there are probably more fruitful projects available 2013-03-12T17:21:18 Yes it could. I would like to another component added to the project, like add some reporting etc 2013-03-12T17:22:05 is there a way to generalize the scheduler simulator to build a generic test harness for rtems components/modules? 2013-03-12T17:22:45 hm yeah i think i have used rtems-bin2c myself for prepping imfs disk images or something 2013-03-12T17:22:59 yep on rtems-bin2c gedare.. 2013-03-12T17:23:10 but it seems host tools belong somewhere else :) 2013-03-12T17:23:24 I do not use bin2c rather objcopy 2013-03-12T17:24:23 gedare.. I didn't plan it that way but it is just a scripting language (derived from rtems shell) that lets you drive "events" like task creation, clock ticks, priority changes, etc. 2013-03-12T17:24:48 So it should be possible to add other commands that do other things. If it makes sense to test them that way. 2013-03-12T17:27:08 DrJoel: i would like something along those lines, and polishing up what we have in the schedsim. maybe waf will make it easier to maintain and add new commands and tests though 2013-03-12T17:27:56 i think it could be a nice way to do some unit and integration tests, together with the UNIX bsp it would be a nice testing package 2013-03-12T17:28:50 I am happy with a move to waf and if these others bits could be added it would be great. 2013-03-12T17:29:36 I just don't want the scheduler simulator to die. and it needs converting out of auto* 2013-03-12T17:29:58 I expect it might not take one of US all summer but a student it easily could. If we have a couple of bonus activities 2013-03-12T17:30:18 Waf is not that difficult 2013-03-12T17:30:21 :) 2013-03-12T17:31:15 well.. we could also add converting the examples, rtems-addon, rtems-testing, etc to it 2013-03-12T17:31:34 And maybe even the idea of how to rework the simulator scripts.. and rewrite them in Python. 2013-03-12T17:31:57 Hmm interesting. That is a more complex task. It would expose some issues in the BSPs which would need fixing 2013-03-12T17:32:28 Just tinkering on the simulator scripts in rtems-testing? 2013-03-12T17:32:35 Sure. The Source Builder has an execute module which handles running things. 2013-03-12T17:32:47 http://www.gpsworld.com/first-galileo-only-position-fix-performed/ 2013-03-12T17:32:56 Also would like to add simulators to the source builder but this could be complex. 2013-03-12T17:33:19 qemu uses pkgconfig and glib and they are not present on all hosts, ie macos 2013-03-12T17:34:23 i think converting the schedsim to waf is a promising idea to giving it new life. it would also be nice to move a lot of the Makefile magic that schedsim uses (VPATHs) into waf 2013-03-12T17:35:12 i think that run scripts for simulators is a fantastic project. 2013-03-12T17:35:49 Waf wants all the code under the single tree so it might require having rtems source under the top of the schedsim tree 2013-03-12T17:36:04 ln -s work? 2013-03-12T17:36:17 We also need the simulators, some need patches and not all hosts have them 2013-03-12T17:36:30 That is not portable. 2013-03-12T17:36:31 yeah 2013-03-12T17:36:45 I was thinking about my git.py module and just checking out an rtems source tree 2013-03-12T17:36:57 ah. 2013-03-12T17:37:01 No sure 2013-03-12T17:37:15 want to also point to a development tree 2013-03-12T17:37:58 It might need some waf magic to make it happen. 2013-03-12T17:38:17 Which means extending the framework to handle it 2013-03-12T17:40:07 I had another project in mind for that. Along with the audit we discussed.. update simulators.. find breakages.. git bisect.. report.. basically do to the simulators what we are doing with the development tools. 2013-03-12T17:40:44 Integrating them into rtems-source-builder -- even with limitations -- would be a BIG improvement over the utter lack of configuration management system we have now 2013-03-12T17:43:09 jot down notes in open projects, if i find time i'll try to flesh out 2013-03-12T17:43:18 if not...hopefully interested students will inquire 2013-03-12T17:43:23 time for me to head out 2013-03-12T17:44:32 hug the little one for me. 2013-03-12T17:47:57 *** gedare has quit IRC (Ping timeout: 245 seconds) 2013-03-12T18:04:43 lm32/milkymist builds ok with new tools from source builder 2013-03-12T18:08:04 Nice. 2013-03-12T18:08:22 Did you see if down the patches directly from git ? :) 2013-03-12T18:10:47 No.. but i guess it had to. I just ran the VM and minimized it. :) 2013-03-12T18:24:52 *** DrJoel has quit IRC (Quit: It's a dud! It's a dud! It's a du...) 2013-03-12T19:14:50 *** the9a3eedi has joined #rtems 2013-03-12T19:19:07 *** the9a3eedi-2 has joined #rtems 2013-03-12T19:22:32 *** the9a3eedi has quit IRC (Ping timeout: 245 seconds) 2013-03-12T19:23:13 *** gedare has joined #rtems 2013-03-12T19:25:44 *** the9a3eedi-3 has joined #rtems 2013-03-12T19:26:41 *** the9a3eedi-2 has quit IRC (Read error: Connection reset by peer) 2013-03-12T19:35:04 *** the9a3eedi-3 has quit IRC (Quit: Leaving) 2013-03-12T19:35:28 *** the9a3eedi has joined #rtems 2013-03-12T19:45:12 *** gedare has quit IRC (Ping timeout: 264 seconds) 2013-03-12T20:36:21 *** MegaAlex is now known as MegaAlex|away 2013-03-12T22:33:22 *** antgreen has joined #rtems 2013-03-12T23:11:11 *** arvind_khadri has joined #rtems 2013-03-13T01:05:00 *** sevikkk has quit IRC (Read error: Connection reset by peer) 2013-03-13T01:05:25 *** sevikkk has joined #rtems 2013-03-13T01:59:42 *** monstr__ has joined #rtems 2013-03-13T02:06:00 *** sebhub has joined #rtems 2013-03-13T02:06:09 good morning 2013-03-13T03:35:49 morning 2013-03-13T03:37:18 is it possible in any way that writing out of bounds on a buffer obtained from an rtems partition can mess up the stack? or perhaps can it cause a segfault? 2013-03-13T03:38:06 because I am 90% sure that's what's causing my segfault, because increasing the buffer size from my partition to twice its old size fixes the problem I'm having 2013-03-13T03:38:51 out of bounds writes to a buffer are always bad in a single address space system 2013-03-13T03:39:09 yes, not doing it intentionally :P 2013-03-13T03:39:25 just trying to figure out where the segfault is coming from 2013-03-13T04:08:06 *** the9a3eedi has quit IRC (Ping timeout: 250 seconds) 2013-03-13T04:29:54 *** MegaAlex|away is now known as MegaAlex 2013-03-13T04:36:20 *** arvind_khadri has quit IRC (*.net *.split) 2013-03-13T04:42:49 *** arvind_khadri has joined #rtems 2013-03-13T04:44:27 *** the9a3eedi has joined #rtems 2013-03-13T05:36:55 *** MegaAlex is now known as MegaAlex|away 2013-03-13T06:06:41 *** MegaAlex|away has quit IRC (Ping timeout: 248 seconds) 2013-03-13T07:10:24 *** antgreen has quit IRC (Ping timeout: 245 seconds) 2013-03-13T07:25:14 *** sebhub has quit IRC (Ping timeout: 258 seconds) 2013-03-13T07:31:23 *** gedare has joined #rtems 2013-03-13T07:48:29 *** sebhub has joined #rtems 2013-03-13T07:50:28 *** gedare has quit IRC (Ping timeout: 248 seconds) 2013-03-13T07:58:20 *** SYCrane has joined #rtems 2013-03-13T07:59:00 *** sebhub has quit IRC (Ping timeout: 248 seconds) 2013-03-13T08:21:21 *** sebhub has joined #rtems 2013-03-13T08:31:48 *** antgreen has joined #rtems 2013-03-13T08:58:33 *** arvind_khadri has quit IRC (Read error: Operation timed out) 2013-03-13T08:58:50 *** MegaAlex|away has joined #rtems 2013-03-13T09:54:19 *** SYCrane has quit IRC (Quit: ChatZilla 0.9.90 [Firefox 19.0.2/20130307023931]) 2013-03-13T10:36:30 *** the9a3eedi has quit IRC (Ping timeout: 240 seconds) 2013-03-13T12:28:43 *** monstr__ has quit IRC (Ping timeout: 246 seconds) 2013-03-13T12:36:16 *** sebhub has quit IRC (Remote host closed the connection) 2013-03-13T12:45:36 *** DrJoel has joined #rtems 2013-03-13T12:45:36 *** DrJoel has joined #rtems 2013-03-13T12:45:36 *** ChanServ sets mode: +o DrJoel 2013-03-13T13:49:49 *** MegaAlex|away is now known as MegaAlex 2013-03-13T14:53:23 *** gedare has joined #rtems 2013-03-13T15:07:31 *** antgreen has quit IRC (Quit: Leaving) 2013-03-13T18:38:40 *** MegaAlex is now known as MegaAlex|away 2013-03-13T19:32:36 *** gedare has quit IRC (Ping timeout: 264 seconds) 2013-03-13T19:42:23 *** antgreen has joined #rtems 2013-03-13T22:16:58 *** sevikkk has quit IRC (Ping timeout: 252 seconds) 2013-03-13T22:20:49 *** sevikkk has joined #rtems 2013-03-14T00:46:23 *** arvind_khadri has joined #rtems 2013-03-14T01:23:07 *** the9a3eedi has joined #rtems 2013-03-14T02:13:22 *** monstr__ has joined #rtems 2013-03-14T02:19:52 *** arvind_khadri has quit IRC (Ping timeout: 272 seconds) 2013-03-14T02:21:51 *** monstr__ has quit IRC (Ping timeout: 260 seconds) 2013-03-14T02:24:06 Is there any documentation on using the capture engine from within code? (and not from the shell) 2013-03-14T02:24:26 can't seem to find anything 2013-03-14T02:44:27 *** arvind_khadri has joined #rtems 2013-03-14T03:02:45 *** 31NACZISB <31NACZISB!~monstr@nat-63.starnet.cz> has joined #rtems 2013-03-14T03:04:40 *** MegaAlex|away is now known as MegaAlex 2013-03-14T03:18:11 *** DrJoel has left #rtems 2013-03-14T03:22:34 *** sebhub has joined #rtems 2013-03-14T03:22:58 good morning 2013-03-14T04:15:54 nevermind, I found the README. I should look for READMEs more often 2013-03-14T04:23:17 *** the9a3eedi has quit IRC (Ping timeout: 255 seconds) 2013-03-14T04:54:48 *** 31NACZISB <31NACZISB!~monstr@nat-63.starnet.cz> has quit IRC (Ping timeout: 252 seconds) 2013-03-14T05:11:31 *** the9a3eedi has joined #rtems 2013-03-14T05:16:28 *** monstr has joined #rtems 2013-03-14T05:20:26 *** MegaAlex is now known as MegaAlex|away 2013-03-14T07:26:44 *** gedare has joined #rtems 2013-03-14T07:45:58 morning 2013-03-14T08:52:43 *** the9a3eedi has quit IRC (Ping timeout: 264 seconds) 2013-03-14T09:10:16 *** antgreen has quit IRC (Ping timeout: 272 seconds) 2013-03-14T09:48:53 *** lcpfnvcy has quit IRC (Ping timeout: 245 seconds) 2013-03-14T09:55:20 *** lcpfnvcy has joined #rtems 2013-03-14T10:03:30 *** antgreen has joined #rtems 2013-03-14T10:34:35 *** arvind_khadri has quit IRC (Read error: Operation timed out) 2013-03-14T11:08:26 *** antgreen has quit IRC (Ping timeout: 252 seconds) 2013-03-14T11:29:53 *** sebhub has quit IRC (Remote host closed the connection) 2013-03-14T12:00:32 *** sevikkk has quit IRC (Remote host closed the connection) 2013-03-14T12:00:55 *** sevikkk has joined #rtems 2013-03-14T12:01:26 *** sevikkk has joined #rtems 2013-03-14T12:17:36 *** antgreen has joined #rtems 2013-03-14T12:38:36 *** antgreen has quit IRC (Ping timeout: 264 seconds) 2013-03-14T12:42:12 *** monstr has quit IRC (Ping timeout: 264 seconds) 2013-03-14T13:03:38 *** antgreen has joined #rtems 2013-03-14T13:30:31 *** antgreen has quit IRC (Ping timeout: 264 seconds) 2013-03-14T13:59:46 *** Hesham has joined #rtems 2013-03-14T14:34:54 *** DrJoel has joined #rtems 2013-03-14T14:34:54 *** DrJoel has joined #rtems 2013-03-14T14:34:54 *** ChanServ sets mode: +o DrJoel 2013-03-14T14:35:11 hey DrJoel 2013-03-14T14:35:12 Hey Gedare http://youtu.be/X_mzgdQGwY4 2013-03-14T14:35:33 silly. 2013-03-14T14:35:49 plus, i think sheldon is actually a gentoo user. 2013-03-14T14:36:49 I am putting an SSD in my old laptop (dual-core, 4GB RAM). Centos 6 made the / partition read-only during a system update. I tried multiple times but it was too unstable to update after the OS install 2013-03-14T14:37:05 Fedora 18 locked up during the installation when I tried to click to enter the root password. 2013-03-14T14:37:33 I looked on the net and you need a REALLY recent kernel with this disk and Ubuntu has it. So there you go. 2013-03-14T14:38:02 My new laptop has Centos VM, a Win7/Centos dual boot and my old one has a Win7/Ubuntu dual boo. 2013-03-14T14:38:02 My claim remains. 2013-03-14T14:38:26 But does he have time for Gentoo with Amy wanting attention. :) 2013-03-14T14:39:36 4 days to org application phase 2013-03-14T14:40:07 hmm... carol hasn't posted the questions yet has she? I don't recall seeing them. 2013-03-14T14:41:21 not that i saw. 2013-03-14T14:41:33 i just happened to hit the gsoc page today and noticed. 2013-03-14T14:42:41 but the period is 2 weeks long so it shouldn't be too bad 2013-03-14T14:42:55 I am using a new irc client. Did you get a private channel request? 2013-03-14T14:43:00 nope 2013-03-14T14:43:05 i got a DCC request ;p 2013-03-14T14:43:43 I don't understand this client completely. but it seems OK enough. I don't have a "chat" button. :( 2013-03-14T14:44:34 your gwu account was on the document. Try again 2013-03-14T14:44:39 ah 2013-03-14T14:44:58 yes i had this in my drive 2013-03-14T14:45:01 *** antgreen has joined #rtems 2013-03-14T14:45:02 could have just told me to find it ;) 2013-03-14T14:45:45 Sorry.. not everyone I work with is as brilliant as you :) 2013-03-14T14:46:20 FYI http://www.opengroup.org/getinvolved/consortia/face released Edition 2.0. I am now listed as a principal author. :) 2013-03-14T14:46:20 i just know how to use ctrl+f 2013-03-14T14:46:28 nice 2013-03-14T14:54:29 gtg 2013-03-14T14:54:34 happy pi day, btw 2013-03-14T14:54:39 i baked apple. 2013-03-14T14:54:45 Veronica clicked like :) 2013-03-14T14:54:50 it is a fav in our house 2013-03-14T14:54:52 take care 2013-03-14T14:58:41 *** MegaAlex|away is now known as MegaAlex 2013-03-14T15:10:57 *** antgreen has quit IRC (Ping timeout: 276 seconds) 2013-03-14T15:17:58 *** Hesham has quit IRC (Ping timeout: 258 seconds) 2013-03-14T15:54:53 DrJoel, hi 2013-03-14T16:02:45 Hey.. scroll up to see what I have been doing the past few days on my old laptop 2013-03-14T16:03:10 on a positive note... Ubuntu 12.10 and source builder go together nicely 2013-03-14T16:03:45 After a standard install, the check found this: 2013-03-14T16:04:00 $ ./source-builder/sb-check error: exe: not found: (__autoconf) autoconf 2013-03-14T16:04:00 error: exe: not found: (__autoheader) autoheader 2013-03-14T16:04:00 error: exe: not found: (__automake) automake 2013-03-14T16:04:00 error: exe: not found: (__bison) /usr/bin/bison 2013-03-14T16:04:00 error: exe: not found: (__cxx) /usr/bin/g++ 2013-03-14T16:04:01 error: exe: not found: (__flex) /usr/bin/flex 2013-03-14T16:04:01 error: exe: not found: (__makeinfo) /usr/bin/makeinfo 2013-03-14T16:04:02 RTEMS Source Builder environment is not correctly set up 2013-03-14T16:04:02 apt-get install automake autoconf bison texinfo g++ flex 2013-03-14T16:04:15 I think that's close to what you had in your notes. 2013-03-14T16:04:31 Plus it didn't install git which is kind of useful in this case. 2013-03-14T16:04:38 We can debate gvim. :) 2013-03-14T16:05:43 Did you check the doco for the host bits to install ? :) 2013-03-14T16:05:49 http://www.rtems.org/ftp/pub/rtems/people/chrisj/source-builder/source-builder.html 2013-03-14T16:06:39 Please let me know if the list is not complete 2013-03-14T16:07:15 What does `*-rtem4.11-gcc --version` say ? 2013-03-14T16:07:29 I checked it and they were different. Yours looked longer.. will check again.. 2013-03-14T16:08:12 Did you get the build reports installed ? 2013-03-14T16:08:22 Look at this: 2013-03-14T16:08:28 joel@JoelD830:~/rtems-4.11-work$ ./tools/bin/sparc-rtems4.11- 2013-03-14T16:08:28 sparc-rtems4.11-addr2line sparc-rtems4.11-nm 2013-03-14T16:08:28 sparc-rtems4.11-ar sparc-rtems4.11-objcopy 2013-03-14T16:08:28 sparc-rtems4.11-as sparc-rtems4.11-objdump 2013-03-14T16:08:28 sparc-rtems4.11-c++filt sparc-rtems4.11-ranlib 2013-03-14T16:08:28 sparc-rtems4.11-elfedit sparc-rtems4.11-readelf 2013-03-14T16:08:28 sparc-rtems4.11-gprof sparc-rtems4.11-size 2013-03-14T16:08:29 sparc-rtems4.11-ld sparc-rtems4.11-strings 2013-03-14T16:08:29 sparc-rtems4.11-ld.bfd sparc-rtems4.11-strip 2013-03-14T16:08:30 joel@JoelD830:~/rtems-4.11-work$ ./tools/bin/sparc-rtems4.11-^C 2013-03-14T16:08:30 joel@JoelD830:~/rtems-4.11-work$ ls tools/bin/*gcc 2013-03-14T16:08:31 ls: cannot access tools/bin/*gcc: No such file or directory 2013-03-14T16:08:31 joel@JoelD830:~/rtems-4.11-work$ ls tools/bin/*gdb 2013-03-14T16:08:32 ls: cannot access tools/bin/*gdb: No such file or directory 2013-03-14T16:08:32 joel@JoelD830:~/rtems-4.11-work$ 2013-03-14T16:08:37 I have binutils but not gcc or gdb. 2013-03-14T16:09:14 Did the build fail ? 2013-03-14T16:10:01 I didn't have build-dep which looks to be huge -- 300MB.. that may have given you texinfo, bison, etc 2013-03-14T16:10:21 I think so. Mick from Perth gave me that line 2013-03-14T16:10:29 Failed with missing zlib.h 2013-03-14T16:10:37 The Raspbian list might work. It is debian 2013-03-14T16:11:10 Hmm CentOS lists zlib-devel and python-devel 2013-03-14T16:11:27 Pi looks similar to what I did 2013-03-14T16:12:02 On a personal note, check this out: http://www.amazon.com/-/e/B00BU3IQXE 2013-03-14T16:12:47 Congradulations and well done 2013-03-14T16:13:47 Also on a personal note ? http://my.leukaemiafoundation.org.au/personalPage.aspx?Referrer=http%3a%2f%2fworldsgreatestshave.com%2f®istrationID=473233 2013-03-14T16:14:14 Come Sunday your time I will very little hair left 2013-03-14T16:14:28 Paperback proof arrived about noon yesterday.. about 3pm we approved it, by 8pm it was on Amazon and Kindle to be purchased. 2013-03-14T16:14:47 This is great. 2013-03-14T16:25:03 I don't know how to scroll back but I don't recall anything in the output that made me think it did not complete successfully. The log is clear about the failure 2013-03-14T16:27:01 An error typically looks like this ... 2013-03-14T16:27:01 error: shell cmd failed: /bin/sh -ex /usr/home/chris/development/rtems/src/sb/rtems-source-builder.chrisj/rtems/build/moxie-rtems4.11-binutils-2.23.1-1-2.23.1-1/doit 2013-03-14T16:27:23 Did you use the --log option ? 2013-03-14T16:28:59 yes and there was information there -- but no clear indication on the console. 2013-03-14T16:29:34 I would need to see the console output 2013-03-14T16:30:09 Alway happy to improve the information given to users. 2013-03-14T16:30:44 I will have to remove something in a minute and see if it fails 2013-03-14T16:31:02 Ok 2013-03-14T17:01:26 DrJoel: apt-get install build-essential 2013-03-14T17:01:43 What do these do ? 2013-03-14T17:02:03 Thanks gedare.. just trying to remove zlib-dev to break the gcc build and see how the source-builder reacts. 2013-03-14T17:02:13 But apt is too handy and it is hard to get a "broken" system 2013-03-14T17:02:35 Just rename zlib.h 2013-03-14T17:03:23 ouch.. too late for that. let's see if this fails 2013-03-14T17:03:25 I have added "Build FAILED" if a general exception is caught which is how all errors are handled 2013-03-14T17:03:48 Not committed or pushed. 2013-03-14T17:04:19 Have you seen http://www.etalabs.net/compare_libcs.html ? 2013-03-14T17:05:21 DrJoel, is this the 4.9 spec file package http://git.rtems.org/rtems-crossrpms.git/snapshot/rtems-4-9-6.tar.bz2 ? 2013-03-14T17:05:23 i shaved my head once... in winter. it was cold 2013-03-14T17:05:52 It is cold where you are in winter. Not as bad here 2013-03-14T17:06:39 wonder how newlib compares 2013-03-14T17:06:48 Yeah 2013-03-14T17:06:48 i know someone who uses dietlibc in his rtos 2013-03-14T17:07:46 it would be nice if his scripts for how he got these numbers were available. 2013-03-14T17:08:01 Ask ? 2013-03-14T17:08:09 mm, dietlibc is gpl2 2013-03-14T17:08:39 i don't really have the time/motivation right now ;) 2013-03-14T17:08:41 plus... 2013-03-14T17:08:48 Nor me. 2013-03-14T17:08:53 i'm not interested in that aspect of rtems 2013-03-14T17:09:00 if it works and doesn't get in my way.... 2013-03-14T17:09:01 Hmmm gpl2 2013-03-14T17:09:41 time for me to head home.. it currently just fails with an exit code. If I hadn't wrapped it in a shell script, it wouldn't have been obvious 2013-03-14T17:10:05 config: tools/rtems-gcc-4.7.2-newlib-2.0.0-1.cfg 2013-03-14T17:10:05 package: sparc-rtems4.11-gcc-4.7.2-newlib-2.0.0-1 2013-03-14T17:10:05 building: sparc-rtems4.11-gcc-4.7.2-newlib-2.0.0-1 2013-03-14T17:10:05 error: shell cmd failed: /bin/sh -ex /home/joel/rtems-4.11-work/rtems-source-builder/rtems/build/sparc-rtems4.11-gcc-4.7.2-newlib-2.0.0-1-4.7.2-1/doit 2013-03-14T17:10:05 Command exited with non-zero status 1 2013-03-14T17:10:07 That does not make sense. 2013-03-14T17:10:13 misread it.. that's your doit script 2013-03-14T17:10:23 So it prints a little .. just nothing that would scare you. :) 2013-03-14T17:10:24 The 'error: shell cmd ?' is the error 2013-03-14T17:10:39 I have add 'Build FAILED' now 2013-03-14T17:10:56 as the next line 2013-03-14T17:11:18 DrJoel: do you know if you ever saw the posix keys code with the "Preallocation" feature? 2013-03-14T17:12:13 DrJoel, before you head off is the crossrpms 4.9 version the correct one ? 2013-03-14T17:12:47 gedare.. I thought I looked at it long ago but will again 2013-03-14T17:13:03 kiwichris I have no idea whether it is in there or the new repo he set up 2013-03-14T17:13:08 *** DrJoel has quit IRC (Quit: Depression is merely anger without enthusiasm) 2013-03-14T17:15:39 I should tell him that his signoff message is a wee bit..odd 2013-03-14T17:16:05 It is a bit odd, maybe it is automatic and changes 2013-03-14T17:16:29 maybe. he's using a new irc client presumably on his ubuntu laptop i guess 2013-03-14T17:16:38 sometimes i hate to review code in email 2013-03-14T17:16:42 sometimes it is ok 2013-03-14T17:17:00 * gedare random gripe 2013-03-14T17:33:43 *** MegaAlex is now known as MegaAlex|away 2013-03-14T17:51:36 Does anyone even use posix keys :-/ 2013-03-14T18:11:04 I have but I am not working that app any more 2013-03-14T18:18:56 *** antgreen has joined #rtems 2013-03-14T20:04:36 *** the9a3eedi has joined #rtems 2013-03-14T20:07:07 *** gedare has quit IRC (Ping timeout: 264 seconds) 2013-03-14T21:57:13 *** dhananjay has joined #rtems 2013-03-14T22:17:14 I am interested to apply for gsoc thisyear, and I run Archlinux. Which way should be better for me? Switch to a RPM based distro or build tool chain in my distro? 2013-03-14T22:24:35 Try http://www.rtems.org/ftp/pub/rtems/people/chrisj/source-builder/source-builder.html 2013-03-14T22:24:53 I have not tried Archlinux so any info would be welcome 2013-03-14T22:31:09 This is definitly much more welcoming than the original build from source wiki article. Shouldnt this link be on the additional referance there? 2013-03-14T22:31:44 I am trying this. I will keep you posted about the outcome. 2013-03-14T22:31:52 It will be in time. I am just working to finish it off and then start to push into the main stream. 2013-03-14T22:32:25 Please do. If you need packages please send me a patch to the doco for Archlinux. The more distros supported the better. 2013-03-14T22:55:36 kiwichris: Why do we have to explicitly define the binary path? In arch we have gzip at /usr/bin/gzip but linux.py defines them as /bin/gzip 2013-03-14T22:56:19 doesnt a simple which solve the problem, esp since we are not enforcing the host versions of those programs? 2013-03-14T22:56:21 So they can be tracked. If you look in windows.py the path is used 2013-03-14T22:57:03 I dont understand what you mean by tracked. 2013-03-14T22:57:06 Using the path is IMO a sign of a bad design. This tool is about making a repeatable process for building from source. 2013-03-14T22:57:26 Anything in the environment is dangerous because you need have it set to make things works 2013-03-14T22:57:53 This a lesson learnt from managing verification and validation processes in large complex systems. 2013-03-14T22:58:36 okay, so shall I alter the variation dictionary and build? 2013-03-14T22:58:39 Yes you are correct about the version of the tool how-ever there is a certain system level of confidence you need 2013-03-14T22:58:56 Please. Does the linux.py make sense for you to change ? 2013-03-14T23:00:57 When you get a binary the '*-rtems-gcc --version' will contain the git hash for the source builder. This lets you track the extract process used to build the tools. 2013-03-14T23:03:09 Developers are great at using modified environments, by extact paths these tend to be removed from the equation. 2013-03-14T23:03:46 Can I make a suggestion? What if we make it more conditional, that the paths are adjusted according to the distro? 2013-03-14T23:03:46 Having said this some of the scripts use 'rm' rather than '%{rm}'. In time these will be tightned. 2013-03-14T23:04:05 yes please do. Example ? 2013-03-14T23:04:51 make a switch on content of /etc/issue? 2013-03-14T23:05:42 provide conditions for major linux distros on where they put the binary? A simple starting point for users? 2013-03-14T23:06:07 so the variations becomes a dict keyed on a distro name ? 2013-03-14T23:07:14 yes. 2013-03-14T23:07:39 Great. If you cannot find or parse the file you set a default distro ? 2013-03-14T23:08:21 and the default paths are the linux standard thingy that exists somewhere ? 2013-03-14T23:08:50 This sounds nice. Send me a git patch and you can be the first other contributor to this tool :) 2013-03-14T23:09:33 :-) 2013-03-14T23:29:42 Adding support for older RTEMS version ?. 2013-03-14T23:29:43 $ ~/development/rtems/4.9/bin/sparc-rtems4.9-gcc --version 2013-03-14T23:29:45 sparc-rtems4.9-gcc (GCC) 4.3.2 20080827 (RTEMS 4.9-RSB(1efc3f79b3d452701549ae6c80a21548dcac7f21-modified)-1,gcc-4.3.2/newlib-1.16.0) 2013-03-14T23:30:08 Built on FreeBSD-9.1 2013-03-14T23:43:18 kiwichris: Where should I sent thw patches? 2013-03-14T23:43:42 chrisj at rtems.org 2013-03-14T23:48:00 As of now, if it cant find a distro, it does do any variations. (Or should we ask for a user input?) 2013-03-14T23:48:26 *** arvind_khadri has joined #rtems 2013-03-14T23:49:41 I saw that. There should be a LSB set of paths and this should be a "distro". If a distro matches LSB great use that 2013-03-14T23:49:53 If something is not support we need to get it supported 2013-03-14T23:50:32 It is not difficult to add so after a period of time we will cover most distros 2013-03-14T23:50:45 :-) 2013-03-14T23:52:17 Also it would be nice to have a progress report while downloading the sources, not neccesary though. 2013-03-14T23:53:38 Yeah it would be nice, but I have not looked into it. It is using the standard urlib2 so there may exist something on the next. 2013-03-14T23:53:47 next -> net 2013-03-15T00:03:18 Found. 2013-03-15T00:03:19 *** the9a3eedi-2 has joined #rtems 2013-03-15T00:03:48 *** the9a3eedi has quit IRC (Ping timeout: 264 seconds) 2013-03-15T00:12:04 dhananjay, no embedded tabs please. Just spaces 2013-03-15T00:15:56 dhananjay, pushed. Many thanks. 2013-03-15T00:16:13 Can you please test for me ? 2013-03-15T00:17:23 Hmm CentOS is giving errors. 2013-03-15T00:19:05 that means it's working 2013-03-15T00:19:06 _haha_ 2013-03-15T00:19:50 kiwichris: Sure. 2013-03-15T00:19:55 hang on 2013-03-15T00:20:45 verm__, haha .. for that you can sponsor me :) 2013-03-15T00:21:51 verm__, getting trimmed this Sunday my time 2013-03-15T00:21:55 kiwichris: its working fine for me, and I think I know the problem. The old script was importing all the variations blindly 2013-03-15T00:22:35 I moved them only to import specific to distro. 2013-03-15T00:22:54 kiwichris: Which editor do you use? 2013-03-15T00:23:09 I will remove the centos warnings 2013-03-15T00:23:11 emacs 2013-03-15T00:31:16 kiwichris: oh! i forgot about that 2013-03-15T00:45:25 dhananjay, is archlinux really 'Arch' in python's 'platform.dist()' ? 2013-03-15T00:46:02 cause Fedora is 'fedora' and CentOS is 'centos' 2013-03-15T00:47:14 Arch is not LSB compatible, platform.dist does not work in archlinux. 2013-03-15T00:49:38 Ah ok. 2013-03-15T00:49:44 So /etc/issue does ? 2013-03-15T00:50:54 so it comes from /etc/issue - its Arch there. But now thinking about it, it might be better to keep everything lowercase - in the future possibility of arch adopting LSB. 2013-03-15T00:51:25 Good idea. I am playing with the code and will make the change 2013-03-15T00:51:26 /etc/issue contains the distro name. 2013-03-15T00:52:23 Raspbian is 'debian' 2013-03-15T00:54:27 Hmmm Ubuntu returns 'Ubuntu' in Python 2013-03-15T00:54:38 Hm.. but since all the binaries follow debian conventions, shouldnt be a probelm 2013-03-15T00:55:18 I will add aliases, so centos, fedora == redhat, and 'ubuntu == debian 2013-03-15T00:56:32 Mm. 2013-03-15T00:57:00 Maybe we should warn the users about the misspelled distro names, if nothing matches. 2013-03-15T00:57:43 Too hard 2013-03-15T01:01:40 committed a change; please pull and test 2013-03-15T01:14:54 Testing, no problems so far. 2013-03-15T01:27:15 Nice 2013-03-15T01:48:12 I have got classes now. I should run. I am running the build in parallel, will let you know the result 2013-03-15T01:51:34 Excellent. I may be gone for the night so just email me if I do not respond. Please post here as well for the record. 2013-03-15T02:04:24 *** the9a3eedi-2 has quit IRC (Ping timeout: 264 seconds) 2013-03-15T02:09:23 *** monstr has joined #rtems 2013-03-15T02:38:01 *** the9a3eedi-2 has joined #rtems 2013-03-15T03:03:50 *** sebhub has joined #rtems 2013-03-15T03:03:55 good morning 2013-03-15T04:13:54 *** the9a3eedi-2 has quit IRC (Ping timeout: 264 seconds) 2013-03-15T04:25:43 *** MegaAlex|away is now known as MegaAlex 2013-03-15T04:34:53 kiwichris: The buld fails while compiling gcc. Heres my log file https://gist.github.com/anonymous/5168618 2013-03-15T06:25:49 *** YatharthROCK has joined #rtems 2013-03-15T06:26:49 *** MegaAlex is now known as MegaAlex|away 2013-03-15T06:28:40 I have a n00b question. I want to be able to control whether power is supplied to a motor or not using software. I don't care how it is controlled, via bluetooth on a phone or wifi on c omputer. Thing is, I can't get any special hardware, so I can't use embedded programing by just flashing something onto some chip. Are there an ways I could do this? I just want to be able to control whether power flows or not. 2013-03-15T07:04:50 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-15T07:16:55 *** YatharthROCK has joined #rtems 2013-03-15T07:27:05 *** monstr has quit IRC (Ping timeout: 245 seconds) 2013-03-15T07:29:21 *** monstr has joined #rtems 2013-03-15T07:40:23 *** antgreen has quit IRC (Ping timeout: 255 seconds) 2013-03-15T08:10:27 *** dhananjay has quit IRC (Read error: Connection reset by peer) 2013-03-15T08:14:19 *** sebhub has quit IRC (Ping timeout: 264 seconds) 2013-03-15T08:18:18 *** gedare has joined #rtems 2013-03-15T08:27:02 *** dhananjay has joined #rtems 2013-03-15T08:35:49 good morning 2013-03-15T08:35:59 I have a n00b question. I want to be able to control whether power is supplied to a motor or not using software. I don't care how it is controlled, via bluetooth on a phone or wifi on c omputer. Thing is, I can't get any special hardware, so I can't use embedded programing by just flashing something onto some chip. Are there an ways I could do this? I just want to be able to control whether power flows or not. 2013-03-15T08:36:13 hi YatharthROCK 2013-03-15T08:36:45 I guess I'll have to get some hardware. What sort of stuff should I get? I don't mind doing embedded programming in any language but something like Python would be nice 2013-03-15T08:37:15 Python works if you have language/runtime support for your platform 2013-03-15T08:37:18 gedare, Hey. Sorry for just asking it here instead of researching it properly first. Can you give me any helpful links? 2013-03-15T08:37:46 i can't really think of anything off the top of my head 2013-03-15T08:37:49 gedare, Which platform are you talking about? The embedded chip or w/e'? 2013-03-15T08:37:57 gedare, How did you start? 2013-03-15T08:37:58 yeah the processor you're usuing 2013-03-15T08:38:04 Just tell me something to Google 2013-03-15T08:38:11 well... 2013-03-15T08:38:28 So, what platforms are there? Do you mean like AMD or PowerPC? 2013-03-15T08:38:30 look up the specs for your motor, but i'm guessing it is just a regular "power" line to turn on / off 2013-03-15T08:39:28 gedare, Specs for my motors? They're just simple ones I rummaged from my Mechinics Set, so I dunno 2013-03-15T08:39:30 if you power it from an embedded board you should be able to "drive" the pin on or off programatically 2013-03-15T08:39:36 I guess with a relay it doesn't matter 2013-03-15T08:39:51 hmmmm 2013-03-15T08:39:58 gedare, drive the pin? Sorry, not understanding... 2013-03-15T08:40:19 ok the basic idea is that you have to hook up a power line to the motor, right? 2013-03-15T08:40:34 so make the source of that power line be something attached to a cpu 2013-03-15T08:40:53 in embedded systems that could be a GPIO or power pin on a development board... 2013-03-15T08:41:31 or maybe you can get power from something else like a USB port, but that seems complicated to me too... 2013-03-15T08:42:10 gedare, Hmm... the USB port 2013-03-15T08:42:32 That's a good idea as I need to build a prototype *now* 2013-03-15T08:42:34 if you look at a development board for an embedded system, for example http://arduino.cc/en/Main/arduinoBoardUno 2013-03-15T08:42:42 well in that case... 2013-03-15T08:42:43 But I'll need proper stuff later 2013-03-15T08:42:48 you want the simplest way to get power 2013-03-15T08:43:08 you should determine how much power you need to supply to the motor 2013-03-15T08:43:47 i think the simplest thing to do would be to build a very small circuit in a breadboard 2013-03-15T08:44:17 and run a power line from a battery into the breadboard 2013-03-15T08:44:28 then have a switch in the breadboard 2013-03-15T08:44:43 hmm, but you want to control it programmatically 2013-03-15T08:44:54 so you will also need to run a line out from a computer 2013-03-15T08:45:02 a serial port might work for this... 2013-03-15T08:45:18 i don't know, i haven't done too much hardware hacking like this :p 2013-03-15T08:48:12 Uhh.... I'll just read up on my terminolgy and get back to you 2013-03-15T08:48:17 ok 2013-03-15T08:48:31 if you want to do it on linux... the linux device drivers 3 chapter 9 has a nice example using a parallel port 2013-03-15T08:49:00 What do you mean by a serial port here? 2013-03-15T08:49:04 i don't think it could power the motor directly, but it could be used to control the power switch 2013-03-15T08:49:23 So a breadboard is just a amounting board? 2013-03-15T08:49:23 http://en.wikipedia.org/wiki/RS-232 2013-03-15T08:49:35 http://en.wikipedia.org/wiki/Breadboard 2013-03-15T08:49:50 a breadboard... lets you build circuits easily 2013-03-15T08:50:19 it has horizontal and vertical connecting wires, so you plug something into a line, and plug something else in the same line and they will be in a circuit together 2013-03-15T08:51:06 so... plug a power wire in one side of a line, add a switch and plug a wire from your computer to the switch's control, and route the power to the switch output, and send the output to your motor 2013-03-15T08:51:20 that's probably what i would do 2013-03-15T08:51:33 if you're in the US, most radio shack carry the components you need 2013-03-15T08:53:16 gedare, 10X for all the info. I gtg now but I'll be back with morewe quesions later :) 2013-03-15T08:56:05 *** YatharthROCK is now known as Yatharth_gtg 2013-03-15T08:58:20 I am compiling the RTEMS toolchain using the rtems-source-builder, I get an error while running make install - log are here http://pastebin.ca/2333058. Any thoughts? 2013-03-15T08:58:38 (make install of gc 2013-03-15T08:59:16 *(make install of gcc for spark) 2013-03-15T09:00:56 *** Yatharth_gtg has quit IRC (Ping timeout: 255 seconds) 2013-03-15T09:08:12 hi dhananjay 2013-03-15T09:10:14 did you have ok output from sb-check? 2013-03-15T09:10:17 what build set are you using? 2013-03-15T09:10:33 also, what platform is your host? 2013-03-15T09:10:45 and can you show the command line for your call to sb-set-builder? 2013-03-15T09:20:55 *** arvind_khadri has quit IRC (Ping timeout: 264 seconds) 2013-03-15T09:21:50 the install feature is relatively new so it might not be well tested yet 2013-03-15T09:22:12 source builder used to just create a tar of the compiled tools and you would have to untar them by hand in your install point 2013-03-15T09:22:56 Yes. sb-check is fine. I ran builds on archlinux 64bit. - The actual error from sb-builder was sh -ex failed on gcc path /doit script 2013-03-15T09:22:59 ../source-builder/sb-set-builder --log=l-sparc.txt \ git:master 2013-03-15T09:23:00 > --prefix=$HOME/build/rtems/4.11 4.11/rtems-sparc 2013-03-15T09:23:09 thats my actual command 2013-03-15T09:23:58 I am guesssing th texinfo 5.0-1 is causing the problem 2013-03-15T09:24:01 yea 2013-03-15T09:24:18 you might be able to modify the buildset to disable building doc... 2013-03-15T09:24:22 there might be an option for it but i'm not sure 2013-03-15T09:25:00 do we need the docs? its just standarad gcc manuals right? 2013-03-15T09:25:20 right 2013-03-15T09:25:24 don't need doc... 2013-03-15T09:25:43 you can suggest the default build set configurations have doc disabled 2013-03-15T09:26:05 source builder bugs and feedback can be routed to rtems-devel, or sent to kiwichris directly i imagine 2013-03-15T09:28:39 configure doesnt help much http://gcc.gnu.org/install/configure.html :-( I dont see any directives 2013-03-15T09:30:55 may be I should try eith a legacy version of texinfo 2013-03-15T09:40:28 gedare:I am just starting off with RTEMS, thought I sohluld build the toolchain mayself, also 3.1 GB is kind of hard to download.:-( 2013-03-15T09:41:21 3.1 GB of what? 2013-03-15T09:41:41 the virtualbox image 2013-03-15T09:41:46 ah. 2013-03-15T09:41:57 i'll inquire if that can be shrunk at all 2013-03-15T09:42:21 but i think it is a good idea to at least have a clue about how the tools are built... 2013-03-15T09:42:29 source builder is still in beta but it is the best way to go i think 2013-03-15T09:42:35 I assume it contains a full blown DM? 2013-03-15T09:42:36 you're just an early adopter :) 2013-03-15T09:42:53 i don't know. but i imagine it can be cut down quite a bit 2013-03-15T09:43:33 kiwichris suggeested it earlier this day. I had a chance to beautify it little too. :-) 2013-03-15T09:43:56 good.. kiwichris is the creator/maintainer of it. 2013-03-15T09:44:05 and.. yes i would say use an older texinfo 2013-03-15T09:44:11 texinfo 5 is pretty much cutting edge 2013-03-15T09:47:42 Yes. running the builds with 4.13 now 2013-03-15T09:47:49 (texinfo) 2013-03-15T09:51:50 great 2013-03-15T10:01:04 *** jenniferA has joined #rtems 2013-03-15T10:04:34 *** DrJoel has joined #rtems 2013-03-15T10:04:34 *** DrJoel has joined #rtems 2013-03-15T10:04:34 *** ChanServ sets mode: +o DrJoel 2013-03-15T10:08:03 hey DrJoel 2013-03-15T10:08:21 i have a couple minutes to look at the soc proposal page 2013-03-15T10:08:29 i think that is the best way to spend these minutes, hmmm 2013-03-15T10:09:16 Great! I will be at Georgia Tech next Tues-Thurs. I will try to fill in enough of the application Monday so it can be saved. Lots of TBDs. :) 2013-03-15T10:09:53 DrJoel: what do you think is the difficulty of having RTEMS code (at least, new contributions) be under a standard OSI license like the 2-clause BSD or MIT? 2013-03-15T10:10:15 or are you a strong proponent of the GPLv2 with linking exception 2013-03-15T10:10:45 just a random curiousity. I see "preferred license" in this doc 2013-03-15T10:10:51 GA Tech huh 2013-03-15T10:11:25 I am not a license bigot. I like the GPLv2 with linking exception because it is a nice blend of free software ideals and practicals of embedded world. 2013-03-15T10:11:43 ok 2013-03-15T10:12:00 But in reality I don't care which license something is submitted under AS LONG AS it has no impact on the statically linked applications -- viral or advertising 2013-03-15T10:12:26 yeah 2013-03-15T10:12:43 that's the main problem with gpl2... 2013-03-15T10:13:00 and it could be a perceptual problem to a casual observer that might see rtems as a "gpl2" product 2013-03-15T10:13:25 because in reality... 2013-03-15T10:13:30 we can't accept any gpl code to rtems 2013-03-15T10:13:36 so it is kind of a pointless license :) 2013-03-15T10:13:43 well, not pointless 2013-03-15T10:13:43 It is preferred only in the sense that it is a pull down in Melange.. We could change it in the GSOC application and update whatever we have on licensing to reflect our open attitude to licensing given protection of embedded free software ideals 2013-03-15T10:14:09 that's the thing, does melange have a gpl2 w/ linking exception option? 2013-03-15T10:14:23 I think we have a very well defined goal for our licensing 2013-03-15T10:14:28 ok 2013-03-15T10:14:33 No.. so maybe the 2 clause BSD is better 2013-03-15T10:14:35 but not a well-defined OSI license that fits 2013-03-15T10:14:37 for GSOC 2013-03-15T10:15:04 because.. LGPL is wrong too 2013-03-15T10:15:16 for some reason 2013-03-15T10:15:19 i don't remember what :) 2013-03-15T10:15:23 Yep.. Horrible license for us. 2013-03-15T10:15:30 but... 2013-03-15T10:15:51 does it really matter if rtems core is copyleft or not? 2013-03-15T10:16:02 End user must be able to relink app with new LGPL code that they can modify. So apps would have to ship RTEMS source, binary blobs for apps and a relinking kit 2013-03-15T10:16:05 i guess it could matter if we got a huge influx of new corporate users... 2013-03-15T10:16:22 Not particularly I think but there would have to be wide buy-in. 2013-03-15T10:16:47 well.. it would require OAR and EB buy-in, and probably me and chris, maybe a handful of other people 2013-03-15T10:16:58 hard to say..just ruminating 2013-03-15T10:17:04 Potential topic for the next Hangout.. plus OAR and Embedded Brains would have to discuss offline with employee contributors 2013-03-15T10:17:12 this "preferred license" being a non-standard license just throws me 2013-03-15T10:17:18 i remember it threw me the first time i started with rtems too 2013-03-15T10:17:19 That would cover a lot of code between the "core". 2013-03-15T10:17:28 because i was using google code, and had to pick a license, but there was not a good match. 2013-03-15T10:17:36 yeah 2013-03-15T10:17:41 There is a certain tool packager who will likely blow their stack 2013-03-15T10:17:57 somehow i think the amount of code that will affect is small 2013-03-15T10:18:22 although it could be troublesome 2013-03-15T10:18:29 i also wonder ... 2013-03-15T10:18:29 This may be something to also discuss in terms of when a non-profit might be ready. 2013-03-15T10:18:35 yeah 2013-03-15T10:18:48 what the legality of just changing the terms of LICENSE are ;) 2013-03-15T10:18:55 i mean, for every file that just refers to the LICENSE file.. 2013-03-15T10:19:17 but this is a topic to push down the road a bit 2013-03-15T10:19:25 for now we should consider a standard license for gsoc though 2013-03-15T10:19:28 You would have to change individual files and double check that it can be changed based upon the set of authors listed 2013-03-15T10:19:36 i'm making a comment in the doco 2013-03-15T10:19:36 2 clause BSD is fine by me 2013-03-15T10:19:38 ok 2013-03-15T10:20:04 we might also have to delve the git log to determine if the authors listed is complete 2013-03-15T10:20:05 yeah.. we are a multi-license project and picking only one acceptable one sucks.. that's the crux 2013-03-15T10:20:10 to be traceable 2013-03-15T10:20:29 yep. could be a rathole and we have a big enough one with the tools right now 2013-03-15T10:20:39 imo if the core can become a permissive license, it makes publicity and contributing easier 2013-03-15T10:20:39 on a different topic.. haven't discussed this with Chris.. 2013-03-15T10:20:47 k 2013-03-15T10:20:58 converting rtems-graphics-toolkit and rtems-addon-packages to something like the rtems-source-buidler. 2013-03-15T10:21:04 Begin to have our own ports collection. 2013-03-15T10:21:18 i had this thought myself last night. 2013-03-15T10:21:35 really.. a framework for the various kit ideas i have listed on thewiki 2013-03-15T10:21:46 Email Chris and I .. we might be able to flesh this out as a GSOC project. 2013-03-15T10:21:50 i think so. 2013-03-15T10:22:02 Every third party library that supports RTEMS under one "ports" kit 2013-03-15T10:22:09 let it be fetched, patched, etc. 2013-03-15T10:22:11 yes 2013-03-15T10:22:12 rtems-ports 2013-03-15T10:22:18 or something 2013-03-15T10:22:34 There is a middle point between cross-tools, simulators and on-target code. Building RTEMS. :) 2013-03-15T10:22:50 But figuring out the other pieces makes a nice view 2013-03-15T10:23:38 ya 2013-03-15T10:23:49 hold on.. phone 2013-03-15T10:31:12 *** the9a3eedi has joined #rtems 2013-03-15T10:44:02 gedare: Texinfo did the trick, gcc build is over.:-) 2013-03-15T10:46:42 dhananjay: with the source builder 2013-03-15T10:46:56 Yes with the source builder - 2013-03-15T10:46:57 dhananjay: great! DrJoel yes, he is building on archlinux 2013-03-15T10:48:02 Awesome! I was wondering if it was a different distro. Are there notes to add to the documentation? 2013-03-15T10:48:13 Chris has been keen on listing as many distros/OSes as possible 2013-03-15T10:48:51 Maybe we should include it in the builder - to enforce for the texinfo version. I have mailed Chris about this 2013-03-15T10:49:10 super 2013-03-15T10:50:19 thanks! 2013-03-15T10:51:09 DrJoel: I have some suggestions for the usability of the tool - 2013-03-15T10:51:35 like, It doesnt pickup builds from where it left, 2013-03-15T10:51:44 Feel free to pass them along to Chris. It is working but new and needs feedback 2013-03-15T10:51:52 also not much verbose. 2013-03-15T10:52:15 That was one of my complaints. I left a build going overnight. It failed and I didn't even realize it. :) 2013-03-15T10:52:58 Yes, I havemade some alternations and he accepted it earliar today. I am writing up a patch for a verbose downloader now. :-) 2013-03-15T10:53:09 It may be possible now but I often need all stable but one unstable component like gdb. 2013-03-15T10:53:48 And we discussed tarballs. Which would help me a lot. I switch from stable to unstable fairly randomly. Need easy access to both without rebuilding or keeping it all on disk. 2013-03-15T10:54:01 I already have to have a 1.5 TB external drive for VMs. :) 2013-03-15T10:55:07 Why do we need to have VMs? At the most a chrrot envrinment is fine right? 2013-03-15T10:56:17 They should but this way you can do FreeBSD, OpenSolaris (or whatever it is called), Minix, various distributions, etc. Easy to manage and be sure about the configuration control. It is REALLY that OS on HW as far as it can tell 2013-03-15T10:58:12 About the gdb versions:cant we already edit the versions of tools in the config before building right? 2013-03-15T10:58:53 Oh yeah, I was thinking about diffrent versions of toolchains, doesnt think about platform variations. 2013-03-15T10:59:16 You should be able to but I just want to know how Chris wants to manage it. There are formally stable and unstable toolsets now. I want to mix and match 2013-03-15T11:00:07 The host/target matrix is large. :) The source-builder is a great step toward better support for more hosts. 2013-03-15T11:00:12 And better configuration control IMO 2013-03-15T11:01:48 *** MegaAlex|away is now known as MegaAlex 2013-03-15T11:01:57 Hmm.. okay 2013-03-15T11:02:26 sb used to build tar balls by default 2013-03-15T11:02:31 so i'm sure the option is in there 2013-03-15T11:02:45 i requested it change to installing directly instead by default because that seems the more likely use case 2013-03-15T11:04:02 The build is complete. :) 2013-03-15T11:05:31 Yeah! 2013-03-15T11:07:35 Cant fidn the sample programs, I suppoes sb doesnt download the sample programs. 2013-03-15T11:08:02 Is there any repo? I am looking for spark ones. 2013-03-15T11:13:41 git.rtems.org/rtems.git 2013-03-15T11:13:46 The source-builder only builds the cross development toolset -- binutils, gcc, gdb (w/sim) 2013-03-15T11:14:04 see the rtems wiki getting started it guides through sparc setup and running... i'll link 2013-03-15T11:14:18 building rtems is separate from building tools 2013-03-15T11:14:39 This is kind of what Gedare and I were chatting about before you came in. There are 3-4 build phases depending on how you look at things: 2013-03-15T11:14:55 + Cross toolset (possibly with simulator) 2013-03-15T11:14:57 DrJoel: i don't really think the gci table on the main page of the wiki is appropriate 2013-03-15T11:15:03 anything we can do to simplify the main page is good 2013-03-15T11:15:08 + RTEMS itself 2013-03-15T11:15:12 and complifying it is bad :p 2013-03-15T11:15:13 + add-on libraries 2013-03-15T11:15:25 and your application 2013-03-15T11:15:35 dhananjay: http://wiki.rtems.org/wiki/index.php/GSoC_Getting_Started 2013-03-15T11:15:45 that shows you how to configure and build RTEMS for SPARC/SIS 2013-03-15T11:15:58 We need to take a good cross view of that and see how it all can be presented better, configuration managed better, and easier for the end user 2013-03-15T11:16:28 yep. 2013-03-15T11:16:48 OK.. I am bailing early to have lunch with Michele.. any idea on the name of a new page to move it to 2013-03-15T11:17:59 Hmm. Thanks, I will look into it, and will patch the builder for texinfo. 2013-03-15T11:19:26 dhananjay just keep asking questions and making suggestions. Your first impressions matter. We have really been working for 5+ years to improve that. Any hints on rough edges are apperciated 2013-03-15T11:20:38 Sure. 2013-03-15T11:20:58 I will help the way I can. 2013-03-15T11:38:10 *** the9a3eedi has quit IRC (Ping timeout: 256 seconds) 2013-03-15T12:20:12 I am trying to build rtems (w/ sparc arch). Make returns an error - https://gist.github.com/anonymous/5171512 2013-03-15T12:21:00 What was your configure command. 2013-03-15T12:21:20 Is sparc-rtems4.11-gcc in your PATH? 2013-03-15T12:21:37 I followed the quick start guide - 2013-03-15T12:21:42 blindly - 2013-03-15T12:21:44 ../rtems/configure --target=sparc-rtems4.11 --enable-rtemsbsp=sis --enable-tests=samples --disable-posix 2013-03-15T12:22:31 Hmmm.. that looks correct. What was the output of the configure command? 2013-03-15T12:22:50 yes. 2013-03-15T12:22:51 /home/dhananjay/build/rtems/4.11/bin/sparc-rtems4.11-gcc 2013-03-15T12:23:03 ok.. that is a common issue 2013-03-15T12:23:11 but should have broken before where yuo did 2013-03-15T12:26:07 The former was the output of which sparc-rtems4.11-gcc 2013-03-15T12:26:42 that looks ok 2013-03-15T12:27:47 I Here is the configure o/p http://pastebin.ca/2333173 2013-03-15T12:28:57 did you install the rtems autotools? 2013-03-15T12:30:13 I guess not. I just built the tools and then moved on to building rtems 2013-03-15T12:30:13 And put the RTEMS prefix/bin at the head of your PATH? 2013-03-15T12:30:32 look around line 21 and following and somewhere later.. looks like an autotools issue 2013-03-15T12:30:42 Not at the head. is it nec? 2013-03-15T12:31:11 yes.. otherwise you get the OSes autoconf/automake which don't match the source tree expectations. 2013-03-15T12:31:35 Amd believe when I say, I am sorry you will have to bootstrap -c and bootstrap again. :( 2013-03-15T12:32:31 Okay, But I havent built them - rtems-autotools 2013-03-15T12:35:48 then we can be certain we know what the problem is. 2013-03-15T12:36:23 And we look forward to a Python based build system... 2013-03-15T12:36:33 I assume these tools are not patched, autotools and automake and I have newer version installed already. Automake 1.13.1 and autoconf 2.69-1 2013-03-15T12:37:20 from my distro. so is it really neccessary? Are we certain the problem is autoconf and automake? 2013-03-15T12:46:27 I built autotools and conf from the source tree, checking now. And yes, we should include this in the build system. 2013-03-15T12:47:48 Have you looked at the AUR? its an arch repository fwhere the package builidng scripts are stored, kinda similiar to pkgsrc 2013-03-15T12:49:01 There was a thread about automake 1.13.1. I asked why we were not using it and I honestly didn't read the answer. But there was one. :) 2013-03-15T12:49:31 No.. Chris has been excited about this and he has run. I am trying to help on the requirement side. 2013-03-15T12:51:42 Hmm. wouldnt help other distro users, but an AUR script can cover rtems builds (IMHO with as far as I have seen rtems) 2013-03-15T12:51:51 in archlinux 2013-03-15T12:52:35 *** monstr has quit IRC (Ping timeout: 252 seconds) 2013-03-15T12:57:50 DrJoel: About what you told me earlier - So an actual RTEMS deployment contains Core RTEMS + Extra Libs + Application 2013-03-15T12:58:05 and sim if we are testing? 2013-03-15T12:58:33 All compiled into a single thing and deployed in the target. 2013-03-15T12:59:13 Is this complete? Or is there anything else? 2013-03-15T12:59:49 Also, Extra libs means the h/w drivers and other applications extentions right? 2013-03-15T12:59:59 That all resides on the host. Ultimately you get a single linked application that you put on the target. (a single .exe) 2013-03-15T13:00:24 We do have dynamic loading but the purpose is more for optional features and upgrades than on Linux. 2013-03-15T13:00:31 hm 2013-03-15T13:00:38 Target code can be very small 2013-03-15T13:00:51 ralf didn't explain why 1.13.1 was broke, just said there were subtle well-known problems 2013-03-15T13:01:29 dhananjay: we do not want to provide upstream distro scripts/packages because that becomes a maintenance and compatibility burden especially if the packager decides to stop updating 2013-03-15T13:01:44 then the distro has outdated tool packages that break with newer rtems = bad 2013-03-15T13:01:53 this happened with mac ports that i know of 2013-03-15T13:02:18 I used the rtems autotools and rebuilt the code. Now the thing compiles but fails at linker 2013-03-15T13:02:28 however, if someone wants to provide such distro-specific packaging we can serve them from RTEMS servers 2013-03-15T13:02:30 gedare: I understand. 2013-03-15T13:02:38 that way if the packager moves on we can just take them down when they are out of date 2013-03-15T13:02:56 what is the link failure 2013-03-15T13:03:10 regarding deployment... 2013-03-15T13:03:25 core rtems +optional libs + application gets statically linked to a single binary that is executed on the target platform 2013-03-15T13:03:29 Output http://pastebin.ca/2333189 2013-03-15T13:03:30 and that platform is either real hw or a simulator 2013-03-15T13:04:25 this is a newlib "bug" that we do not have resolution on yet 2013-03-15T13:04:43 newlib now requires pthread support in 2.0.x for thread cancellation 2013-03-15T13:04:50 Hmm.. Then there should be a loader right? In the real hw 2013-03-15T13:04:51 so currently if we use newlib 2.0 we need to enable posix 2013-03-15T13:04:58 it is a problem and hopefully will get fixed soon 2013-03-15T13:05:04 try to configure with --enable-posix for rtems 2013-03-15T13:05:49 depends on the hw platform.. some use a loader, some let you load your application code at a pre-set address that is executed whenever the board is powered on 2013-03-15T13:05:53 why are we not using posix? 2013-03-15T13:05:53 I will let gedare handle things. My wife and kids are here and it is time for lunch out.. 2013-03-15T13:06:02 *** DrJoel has quit IRC (Quit: For Sale: Parachute. Only used once, never opened, small stain.) 2013-03-15T13:06:15 posix pulls in extra stuff (=overhead) that is not necessary for non-posix applications 2013-03-15T13:06:15 DrJoel: have a good time. 2013-03-15T13:07:03 posix is preferred if you need an application that will run across multiple POSIX-compliant OSs... 2013-03-15T13:07:13 otherwise the "classic" API in RTEMS is better for RTEMS-specific applications 2013-03-15T13:07:17 because the overhead is less 2013-03-15T13:07:42 unfortunately there is no other OS that provides the classic API afaik 2013-03-15T13:08:00 it would be a fun (?) project to provide a compatibility layer for it say in posix... 2013-03-15T13:08:15 but a lot of work :) 2013-03-15T13:09:40 Can you link me to the classic API documentation? 2013-03-15T13:10:11 so you dont have to use posix at all if you are deploying only on rtems? 2013-03-15T13:11:11 is posix app == portability across other rt oses? 2013-03-15T13:12:55 Hmm.. do you mean like, if the app hass some posix calls that can be emulated well enough eith rtems api calls, convert them dynamically? 2013-03-15T13:13:15 http://rtems.org/onlinedocs/doc-current/share/rtems/html/c_user/index.html 2013-03-15T13:13:37 depends if the other rtos supports posix, but i think many do 2013-03-15T13:14:06 really either way... convert posix into rtems, or rtems into posix 2013-03-15T13:14:21 what i was thinking was to make it easy to move an rtems application that uses classic API to another OS using posix 2013-03-15T13:14:24 okay 2013-03-15T13:14:27 i'm not sure if anyone cares though :) 2013-03-15T13:14:30 the build still fails 2013-03-15T13:14:41 hmm same error? 2013-03-15T13:14:41 with same error 2013-03-15T13:14:45 yes 2013-03-15T13:14:52 did you rm -rf your build, or make distclean? 2013-03-15T13:15:14 if you don't have --enable-maintainer-mode, it might not pick up the new configure option. 2013-03-15T13:15:22 i'm not certain :) 2013-03-15T13:15:54 and yes, you should not need posix to run an rtems application 2013-03-15T13:16:02 just right now the libc requires it 2013-03-15T13:16:12 I did a make clean 2013-03-15T13:18:23 libc in the rtems distribution? 2013-03-15T13:19:43 Built. :) 2013-03-15T13:19:48 complete. 2013-03-15T13:20:04 *** jenniferA has quit IRC (Quit: Leaving) 2013-03-15T13:20:20 libc=newlib 2013-03-15T13:20:35 okay. 2013-03-15T13:20:35 it gets compiled statically into the toolchain 2013-03-15T13:21:11 So we cant compile without it. or we will have to provide another standard lib implementation. 2013-03-15T13:22:25 yeah. but since libc gets linked into the application we need to avoid strong copyleft (viral) licensed software 2013-03-15T13:22:39 we're pretty much tied to newlib right now 2013-03-15T13:22:54 but they do pretty well about accepting patches we push upstream 2013-03-15T13:24:06 mm. I dont understand the clause about licenses. 2013-03-15T13:24:43 I executed the ticker.exe sucessfully. :) 2013-03-15T13:26:08 Whats with the data ? 12/31/1988 or is it a date? 2013-03-15T13:28:42 I see that ticker is written in ada. Cant decipher much from the code 2013-03-15T13:29:00 Which all languages are supported in rtems? 2013-03-15T13:29:07 c/ada - anything else? 2013-03-15T13:31:16 that's a date 2013-03-15T13:31:26 ticker is written in C, but there is an ada version I tihnk 2013-03-15T13:32:41 C++. Some people have worked with python, lua, go, perl, java and some interpreters like Parrot 2013-03-15T13:32:52 out of the box, C and C++ are easiest to use 2013-03-15T13:32:57 Ada takes a little bit more work 2013-03-15T13:33:06 so do these others 2013-03-15T13:35:07 the ticker date has to do with when RTEMS was publicly released I think 2013-03-15T13:35:09 hmm.. but why would someone want to use these others in an RTOS? 2013-03-15T13:35:15 i forget the historical significance exactly 2013-03-15T13:35:17 :) 2013-03-15T13:35:28 http://www.rtems.com/pipermail/rtems-users/2009-December/006372.html 2013-03-15T13:35:52 for programmability perhaps 2013-03-15T13:36:02 or because programmers like languages they know. i don't know 2013-03-15T13:36:12 probably only for background / best-effort tasks 2013-03-15T13:36:22 http://rtems.org/wiki/index.php/Timeline 2013-03-15T13:36:25 found it 2013-03-15T13:36:30 ada and c are the best for real-time probably 2013-03-15T13:37:21 the c code is at testsuites/samples 2013-03-15T13:40:42 the licensing issue has to do with how different open-source licenses handle linking with other code 2013-03-15T13:41:04 saw that- why are the braces put in seperate line, coding guideline? 2013-03-15T13:41:16 GPL code seems to require that if you link statically with something else, that something else must be under GPL or a similarly permissive license 2013-03-15T13:41:29 so you cannot statically link proprietary software with GPL code, legally 2013-03-15T13:41:42 this is a problem for embedded systems, which tend to be statically linked binaries 2013-03-15T13:42:00 http://wiki.rtems.org/wiki/index.php/Coding_Conventions 2013-03-15T13:42:16 braces for functions are, for other structures are not 2013-03-15T13:42:39 I guess we don't really specify the braces in the conventions 2013-03-15T13:44:29 LGPL is not as bad but close... it imposes a relinking requirement that LGPL code can be relinked with other LGPL code, so if you link to LGPL code you need to provide everything necessary to relink... meaning source code, binary blobs, and linker support. 2013-03-15T13:44:42 so, we right now say we use GPL 2 that permits static linking... 2013-03-15T13:45:05 in practice, we can only accept code from other projects that use something permissive like the BSD 2/3 clause, MIT license, or public domain 2013-03-15T13:45:50 in the future we might have a limited form of dynamic linking available which would make it possible for users to add pure GPL code, but RTEMS itself will not be able to do so 2013-03-15T13:49:16 Hmm.. understood 2013-03-15T13:49:35 *** antgreen has joined #rtems 2013-03-15T13:49:43 So I have a working rtems tree+tools and simulator now 2013-03-15T13:50:04 Also I have a patch for the texinfo version. 2013-03-15T13:50:12 (Need to test it though_ 2013-03-15T13:51:19 good progress 2013-03-15T13:51:29 now i'm going to head out though myself and enjoy some fresh air 2013-03-15T13:51:32 can you point me to things which I should explore? 2013-03-15T13:51:49 And I should get sleep too 2013-03-15T13:51:52 if you run into more trouble you can try the rtems-users mailing list, or for development ideas and patches the rtems-devel mailing list is the place to go 2013-03-15T13:51:56 where are you located? 2013-03-15T13:52:01 India 2013-03-15T13:52:11 Ah. Quite... early 2013-03-15T13:52:13 ;) 2013-03-15T13:52:18 :D 2013-03-15T13:52:25 check out open projects... 2013-03-15T13:52:34 right now we're working our way toward a release 2013-03-15T13:52:51 so compiling rtems and finding/fixing warnings is a good way to get involved 2013-03-15T13:53:11 we like to eliminate compiler warnings for releases 2013-03-15T13:53:38 okay 2013-03-15T13:53:52 take care, good luck 2013-03-15T13:54:01 thanks 2013-03-15T13:54:21 see you. 2013-03-15T13:56:18 *** gedare has quit IRC (Read error: Operation timed out) 2013-03-15T15:57:32 *** ppisa has joined #rtems 2013-03-15T17:24:03 *** ppisa has left #rtems ("Kopete 0.12.7 : http://kopete.kde.org") 2013-03-15T20:10:51 *** MegaAlex is now known as MegaAlex|away 2013-03-16T01:12:15 *** Yatharth_gtg has joined #rtems 2013-03-16T01:29:35 *** Yatharth_gtg has quit IRC (Ping timeout: 255 seconds) 2013-03-16T02:02:14 *** Yatharth_gtg has joined #rtems 2013-03-16T02:03:34 *** Yatharth_gtg is now known as YatharthROCK 2013-03-16T03:33:58 *** the9a3eedi has joined #rtems 2013-03-16T03:52:28 *** the9a3eedi has quit IRC (Ping timeout: 256 seconds) 2013-03-16T04:18:20 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-16T04:39:48 *** YatharthROCK has joined #rtems 2013-03-16T04:49:42 *** MegaAlex|away is now known as MegaAlex 2013-03-16T05:07:29 *** YatharthROCK has quit IRC (Read error: Connection reset by peer) 2013-03-16T05:07:36 *** YatharthROCK has joined #rtems 2013-03-16T05:49:54 Can anyone help me? I want to do embedded programming. Thing is, I don 2013-03-16T05:50:06 't have any idea of what hardware I may need. 2013-03-16T05:51:45 Will a Raspberry Pi or Arduino help? I just want to control whether power flows through a curcuit or not. Using a relay, that should require very little current. I want to be able to use GPS and infrared sensors, so should I communicaye with a mobile using bluettoh? 2013-03-16T06:19:37 *** YatharthROCK has quit IRC (Read error: Connection reset by peer) 2013-03-16T06:19:52 *** YatharthROCK has joined #rtems 2013-03-16T06:28:05 Can anyone help me? I want to do embedded programming. Thing is, I don 2013-03-16T06:28:05 't have any idea of what hardware I may need. 2013-03-16T06:28:05 Will a Raspberry Pi or Arduino help? I just want to control whether power flows through a curcuit or not. Using a relay, that should require very little current. 2013-03-16T06:32:06 *** YatharthROCK has quit IRC (Read error: No route to host) 2013-03-16T06:32:23 *** YatharthROCK has joined #rtems 2013-03-16T06:32:59 I have a n00b question. I want to be able to control whether power is supplied to a motor or not using software. I don't care how it is controlled, via bluetooth on a phone or wifi on c omputer. Thing is, I can't get any special hardware, so I can't use embedded programing by just flashing something onto some chip. Are there an ways I could do this? I just want to be able to control whether power flows or not. 2013-03-16T10:52:50 *** the9a3eedi has joined #rtems 2013-03-16T10:53:26 *** dhananjay has quit IRC (Quit: leaving) 2013-03-16T11:01:40 *** dhananjay has joined #rtems 2013-03-16T11:06:34 *** dhananjay has quit IRC (Quit: Vrooomm..) 2013-03-16T11:08:24 *** dhananjay has joined #rtems 2013-03-16T12:15:08 *** dhananjay has quit IRC (Ping timeout: 245 seconds) 2013-03-16T13:22:24 *** the9a3eedi has quit IRC (Read error: Connection reset by peer) 2013-03-16T13:32:44 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-16T14:10:13 *** Hesham has joined #rtems 2013-03-16T14:26:22 *** Hesham has left #rtems 2013-03-16T17:07:38 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2013-03-16T17:46:34 *** dhananjay has joined #rtems 2013-03-16T17:51:48 *** dhananjay has quit IRC (Ping timeout: 245 seconds) 2013-03-16T19:44:52 *** kiwichris has joined #rtems 2013-03-16T19:47:26 *** kiwichris has quit IRC (Client Quit) 2013-03-16T20:06:31 *** MegaAlex is now known as MegaAlex|away 2013-03-16T20:27:20 *** kiwichris has joined #rtems 2013-03-16T21:36:54 *** dhananjay has joined #rtems 2013-03-16T22:27:21 *** dhananjay has quit IRC (Ping timeout: 256 seconds) 2013-03-16T23:16:55 *** dhananjay has joined #rtems 2013-03-16T23:33:29 *** dhananjay has quit IRC (Ping timeout: 255 seconds) 2013-03-17T00:13:23 *** YatharthROCK has joined #rtems 2013-03-17T00:27:29 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-17T00:28:19 *** dhananjay has joined #rtems 2013-03-17T01:48:50 Hi. I was trying to determin the depended packages needed to build in my distro (Archlinux),Since I build pretty much everything, I didnt have to install anything except texinfo 2013-03-17T01:49:14 Is there any way I can do this without testing on a fresh install? 2013-03-17T03:09:48 *** YatharthROCK has joined #rtems 2013-03-17T03:47:17 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-17T03:51:35 *** YatharthROCK has joined #rtems 2013-03-17T04:14:47 *** MegaAlex|away is now known as MegaAlex 2013-03-17T05:01:52 *** MegaAlex is now known as MegaAlex|away 2013-03-17T05:37:04 *** monstr has joined #rtems 2013-03-17T05:44:11 *** monstr has quit IRC (Remote host closed the connection) 2013-03-17T06:25:41 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-17T06:35:43 *** YatharthROCK has joined #rtems 2013-03-17T07:40:26 *** YatharthROCK is now known as JordigGH 2013-03-17T07:40:53 *** JordigGH is now known as JordiGH 2013-03-17T07:41:01 *** JordiGH is now known as YatharthROCK 2013-03-17T09:55:23 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-17T10:34:07 *** MegaAlex|away is now known as MegaAlex 2013-03-17T11:45:35 *** antgreen_ has joined #rtems 2013-03-17T11:48:22 *** antgreen has quit IRC (Ping timeout: 245 seconds) 2013-03-17T13:35:57 *** YatharthROCK has joined #rtems 2013-03-17T14:14:08 *** YatharthROCK has quit IRC (Ping timeout: 255 seconds) 2013-03-17T16:00:54 *** antgreen_ is now known as antgreen 2013-03-17T18:56:48 *** MegaAlex is now known as MegaAlex|away 2013-03-17T19:06:55 dhananjay, I do not know without testing. Just add something which should work and others can update when we learn more. 2013-03-17T19:07:09 *** the9a3eedi has joined #rtems