2012-03-26T00:42:27 *** arvind_khadri has quit IRC (Ping timeout: 260 seconds) 2012-03-26T01:17:10 *** arvind_khadri has joined #rtems 2012-03-26T01:17:11 *** arvind_khadri has joined #rtems 2012-03-26T01:23:14 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-26T01:37:11 *** xiangfu has joined #rtems 2012-03-26T01:42:38 *** Hesham has joined #rtems 2012-03-26T01:43:04 *** Hesham has left #rtems 2012-03-26T01:53:11 *** sebhub has joined #rtems 2012-03-26T01:53:19 good morning 2012-03-26T02:40:56 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-26T02:59:29 hi sebhub 2012-03-26T03:32:56 *** zw_yao has joined #rtems 2012-03-26T04:12:55 *** xiangfu has joined #rtems 2012-03-26T04:17:22 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-26T04:19:56 *** cdcs has joined #rtems 2012-03-26T04:29:38 *** xiangfu has joined #rtems 2012-03-26T06:07:03 *** zw_yao has quit IRC (Quit: Leaving) 2012-03-26T06:07:26 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-26T07:25:17 *** cdcs has quit IRC (Quit: Page closed) 2012-03-26T07:27:13 *** cdcs has joined #rtems 2012-03-26T07:58:48 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2012-03-26T08:17:08 *** zw_yao has joined #rtems 2012-03-26T10:02:11 *** zw_yao has quit IRC (Quit: Leaving) 2012-03-26T10:11:26 *** gedare has joined #rtems 2012-03-26T10:12:20 good morning 2012-03-26T10:13:23 hi 2012-03-26T10:13:28 hi 2012-03-26T10:14:31 sebhub: thanks for giving the early feedback on the lwip project. I didn't realize the driver sharing would be undesirable. 2012-03-26T10:14:38 that was my bad :) 2012-03-26T10:16:00 it is desirable, but a waste of time ;-) 2012-03-26T10:16:17 we have so many other good projects 2012-03-26T10:16:21 agreed 2012-03-26T10:16:28 with the lots of projects :) 2012-03-26T10:16:59 next time instead of doing high-priority we should have mentors "spotlight" projects they like... 2012-03-26T10:17:01 maybe. 2012-03-26T10:17:26 yes, high priority depends on whom you ask 2012-03-26T10:17:29 yeah 2012-03-26T10:41:36 re 2012-03-26T10:42:07 gedare: in addition, I think we should put efforts to the libbsd project 2012-03-26T10:42:23 having a smooth integration with freebsd would be really good in terms of potential new features 2012-03-26T10:44:04 sebhub: does the usb01 sample work on your side for the libbsd ? 2012-03-26T10:44:13 s/work/compile/g ? 2012-03-26T10:45:58 nevert tested this usb01 2012-03-26T10:46:21 i will have a look at this tomorrow 2012-03-26T10:48:39 ah, ok, I though you was working on it 2012-03-26T10:49:43 because I saw this test was designed by you in a commit on March 20 2012-03-26T10:51:18 *** sebhub has quit IRC (Ping timeout: 245 seconds) 2012-03-26T11:08:42 yes the freebsd/libbsd work is important. I just try to help students refine the ideas for the projects that interest them 2012-03-26T11:20:19 yep 2012-03-26T11:20:23 I understand 2012-03-26T11:20:25 :-) 2012-03-26T11:21:10 it might be nicer to organize the open projects page based on whether mentors are interested or not 2012-03-26T11:21:26 gedare: I also support this idea 2012-03-26T11:21:43 because it would say "ok, this is a project with a real need and somebody that can also drives you 2012-03-26T11:21:49 yes 2012-03-26T11:22:10 at least for "weaker" students the latter is quite important 2012-03-26T11:22:29 yes 2012-03-26T11:22:46 but also, for students, having a project that will be reused is nice 2012-03-26T11:22:59 on my side, we are looking for the new net stack from freebsd 2012-03-26T11:23:10 and it will be used in one in-house project 2012-03-26T11:23:58 :) 2012-03-26T11:24:11 where do you work? 2012-03-26T11:25:09 nvm google answered that question fo r me ;) 2012-03-26T11:26:59 gedare: rarf 2012-03-26T11:27:04 I am unmasked right now 2012-03-26T11:27:05 :-) 2012-03-26T12:10:34 Strange :P . Your employer usually likes older, less featured versions of RTEMS ;) 2012-03-26T12:15:41 *** arvind_khadri has joined #rtems 2012-03-26T12:28:05 cdcs: what do you mean ? 2012-03-26T12:28:55 * gedare dons his flame-retardant suit 2012-03-26T12:29:50 RTEMS 4.8? maybe i mistaken the place where you work :P 2012-03-26T12:31:04 cdcs: I do not go into flamewar 2012-03-26T12:31:16 ;-) 2012-03-26T12:31:27 neither do I. I actually work with 4.8 :P 2012-03-26T12:31:45 cdcs: you see, you start to feel the power of the force :-) 2012-03-26T12:31:53 lol 2012-03-26T12:32:11 * gedare makes popcorn 2012-03-26T12:32:36 I am joking 2012-03-26T12:32:44 in europe, this is the end of the day ou know 2012-03-26T12:33:04 on a monday no less! 2012-03-26T12:34:23 almost end of the day :P 2012-03-26T12:47:31 *** cdcs has quit IRC (Ping timeout: 245 seconds) 2012-03-26T14:20:53 is there a global way to specify rtems_resource_unlimited for all kinds resources at once? 2012-03-26T14:21:36 not that i'm aware 2012-03-26T14:21:51 so I have to put them in one by one 2012-03-26T14:22:38 i never tried :) 2012-03-26T14:23:03 but probably.. i don't see any place that will do it for you 2012-03-26T14:23:09 *** arvind_khadri has quit IRC (Ping timeout: 252 seconds) 2012-03-26T14:24:59 it would be nice to have maybe a #include that says ... I want to get this ported thing running and care about fixed resources later include 2012-03-26T14:25:18 or just a config option 2012-03-26T14:25:21 yeah a permissive configuration table 2012-03-26T14:25:39 send a request to rtems-devel :) 2012-03-26T14:25:58 kiwichris added the unlimited objects support i think 2012-03-26T14:26:26 but since each score Object is allocated/managed separately there is no notion of a global pool of resources 2012-03-26T14:26:34 although i think there could be... 2012-03-26T14:30:04 you wouldn't a global pool just a unlimited all config option 2012-03-26T14:30:16 ^need 2012-03-26T14:32:16 true. something to propagate the Unlimited bit to all of the objects 2012-03-26T14:32:25 That wouldn't be too hard to write actually. 2012-03-26T14:49:46 i took a stab at it. building and testing right now 2012-03-26T14:50:45 gedare: http://wiki.rtems.org/wiki/index.php/RTEMS_Testing 2012-03-26T14:50:46 comments? 2012-03-26T14:56:32 BSP-specific tests == tests specific to a Board 2012-03-26T14:56:46 rereading the email you sent to compare 2012-03-26T14:57:43 maybe..tests specific to a CPU is more likely-->things that target code in libcpu 2012-03-26T14:57:58 and then tests specific to a bSp --> tests that target code in libbsp 2012-03-26T14:58:36 verm__ looks pretty good other than ^ 2012-03-26T15:01:41 well when i say board i mean that literally a specific board from a manufacturer with a model # 2012-03-26T15:01:56 and a BSP is a bsp within rtems 2012-03-26T15:02:15 oh i see what you mean 2012-03-26T15:02:31 that does make sense. 2012-03-26T15:02:37 one bsp can run across dozens of boards we don't have a way to have tests specific to one single board from a manufacturer which is something we really need 2012-03-26T15:02:50 is there a better way for me to express that 2012-03-26T15:02:52 ? 2012-03-26T15:03:42 yeah. call it libbsp tests instead of BSP-specific tests. 2012-03-26T15:03:49 makes it more general 2012-03-26T15:04:37 maybe... i dunno :) 2012-03-26T15:04:47 or let whoever takes up the project figure it out with your help 2012-03-26T15:06:31 i changed "BSP" to "RTEMS BSP" 2012-03-26T15:06:41 ok that's good enough 2012-03-26T15:06:47 libbsp will only make sense to someone who has prior knowledge of rtems 2012-03-26T15:06:58 as long as someone is around to explain the intent i think it is fine 2012-03-26T15:07:08 *** antgreen has joined #rtems 2012-03-26T15:07:18 *** arvind_khadri has joined #rtems 2012-03-26T15:07:18 *** arvind_khadri has joined #rtems 2012-03-26T15:41:48 *** Hesham has joined #rtems 2012-03-26T15:42:02 hey gedare 2012-03-26T15:42:09 hello hesham 2012-03-26T15:43:03 should i send the main idea of memory protection only or also illustrate the concepts of Arena ? 2012-03-26T15:43:57 try to capture the essence of your project without being too detailed 2012-03-26T15:44:28 enough so that people who understand can go look for more details if they are curious. 2012-03-26T15:44:42 ok 2012-03-26T15:45:00 any advices about submitting my proposal to google ? 2012-03-26T15:49:09 gedare, yes 2012-03-26T15:49:15 I did add it 2012-03-26T15:49:31 i am sending a macro to allow users to do what peerst asked 2012-03-26T15:49:46 i tested it a bit..have the patch ready just need to email it 2012-03-26T15:50:26 hesham: just sign up, fill out your profile, upload your proposal (make sure to pick rtems project) and click 'save'... 2012-03-26T15:50:34 You still need to set the values in the config table. They define the allocation unit size 2012-03-26T15:50:47 yeah i allow user to specify a value or provide a default (8) 2012-03-26T15:50:49 So I am not sure what is being saved ? 2012-03-26T15:50:56 gedare: cool 2012-03-26T15:51:06 just give a single macro (or two) and user doesn't have to do much configuring 2012-03-26T15:51:18 It is such a big deal to set ? 2012-03-26T15:51:24 i dunno. peerst complained :p 2012-03-26T15:51:34 :-) 2012-03-26T15:51:48 complained not really 2012-03-26T15:51:50 doesn't seem like much...but maybe if we include posix also 2012-03-26T15:52:03 ah only for classic? 2012-03-26T15:52:14 i did it for classic. should be easy to extend 2012-03-26T15:52:16 Just interested. No opposition from me for the addition 2012-03-26T15:52:19 I'm running against posix limts 2012-03-26T15:52:37 let me look at the posix side. should be doable 2012-03-26T15:52:40 That may be due to some of them not being unlimited 2012-03-26T15:52:43 but maybe kiwichris is right ... not much to define 2012-03-26T15:52:46 yeah 2012-03-26T15:53:10 i'll send the patch to rtems-devel.. at least it will be out there ;) 2012-03-26T15:53:51 You can hide things. The allocation unit size is important. If you need deterministic task (or object) create the unlimited effects it. 2012-03-26T15:55:30 but if you are creating stuff you don't know how many units are left until it runs into notdeterministic timing 2012-03-26T15:56:06 how I understood it you give up determinism when using unlimted 2012-03-26T15:57:09 am I wrong? 2012-03-26T15:57:31 A little bit is given up. You get a block at a time. If you need fixed create times then my suggestion is to fix the number. Typically your system will be bounded. 2012-03-26T15:57:50 yeah 2012-03-26T15:58:03 you lose determinism because if you go past the allocation size it will do workspace allocation 2012-03-26T15:58:27 You get a block at init time then another block is added when that is used and that block stays until half the previous block is free 2012-03-26T15:58:35 gedare: so one chunk of units are created statically? 2012-03-26T15:58:45 Yes 2012-03-26T15:58:49 ah 2012-03-26T15:59:01 so you can have it partially both ways 2012-03-26T15:59:03 This sort of models the bounded set up 2012-03-26T15:59:06 Yes 2012-03-26T15:59:14 actually cool 2012-03-26T15:59:36 BTW: I was not intending using this in production 2012-03-26T16:00:05 I'm just porting the Erlang VM an have no clue yet how many of what they are using 2012-03-26T16:00:32 Sure. This is typical of why we need this feature 2012-03-26T16:00:42 for getting something like this up its nice to have a make everything unlimited switch 2012-03-26T16:00:55 Also dynamic loading will changes this area 2012-03-26T16:02:01 actually not having dynamic loading yet was a feature for me 2012-03-26T16:02:29 it got me get the foot in the door at the customer for selling them the Erlang VM 2012-03-26T16:02:54 because we can have dynamic loading in the VM 2012-03-26T16:03:28 i added posix but i don't have a test for it :p 2012-03-26T16:03:55 now Erlang ursurped all kinds of dynamic soft realtime stuff in the system 2012-03-26T16:03:55 Does all posix features handle unlimited ? 2012-03-26T16:04:01 good question 2012-03-26T16:04:17 I think keys are not. They did not use to be 2012-03-26T16:04:28 what happens if a feature doesn't and unlimited is used? 2012-03-26T16:06:43 No idea. 2012-03-26T16:06:52 heh. broke something 2012-03-26T16:07:13 one of the macros does not ignore the unlimited bit in confdefs.h, so probably if i find out what that is it will help 2012-03-26T16:07:14 I suspect not well. 2012-03-26T16:21:19 did not apply for keys and queued signals 2012-03-26T16:21:43 at least in confdefs. no idea if you tried to use it whether it will work with other posix objects :) 2012-03-26T16:23:35 I see 2012-03-26T16:25:14 *** Hesham has left #rtems 2012-03-26T16:34:22 ok peerst i sent a patch to rtems-devel 2012-03-26T16:34:42 * gedare back to real work ... :P 2012-03-26T16:36:07 nice thanks! 2012-03-26T16:36:17 I'll try it out 2012-03-26T16:36:40 great some positive feedback (assuming it works) will probably get it committed. 2012-03-26T16:36:58 the whole configuration / confdefs needs to be reworked some day. 2012-03-26T16:37:54 gedare: I don't know which resources will be tested by my port 2012-03-26T16:38:16 that's ok just make a comment about whether the unlimited patch makes your life easier 2012-03-26T16:38:26 can worry about testing it thoroughly later 2012-03-26T16:39:36 cool 2012-03-26T16:40:03 gedare: Mon Mar 26 16:28:15 CDT 2012 was the last patch? 2012-03-26T16:40:05 you're welcome...it's fun to solve problems :) 2012-03-26T16:40:47 yeah that should be right. if you're using git you can just copy/paste the email into a text file and use git-am on the file 2012-03-26T16:41:01 otherwise just pick out the diff part and use patch as usual 2012-03-26T16:41:22 oh joel...making me document things..baaaaaah 2012-03-26T16:51:29 gedare: thanks ... I've to try it out tomorrow ... bedtime here 2012-03-26T16:51:42 you're welcome good night 2012-03-26T16:51:45 hope it helps 2012-03-26T17:09:00 gedare: patch fails http://ideone.com/a4bdo 2012-03-26T17:09:37 I'm not on the normal git but on a internal tree from embedded brains 2012-03-26T17:09:57 oh must have modified the confdefs. 2012-03-26T17:10:02 that should be easy to apply manually 2012-03-26T17:10:13 just look for _Configure_POSIX_Named_Object_RAM 2012-03-26T17:11:45 eh ... could it be that I don't have this macro? _Configure_POSIX_Named_Object_RAM 2012-03-26T17:12:29 maybe... i'm not sure where / when it came from :) 2012-03-26T17:12:31 nevermind found it 2012-03-26T17:13:41 don't understand why patch couldn patch this ... looks exactly the same 2012-03-26T17:15:39 weird. 2012-03-26T17:20:39 *** cheryl has joined #rtems 2012-03-26T17:20:55 *** cheryl has left #rtems 2012-03-26T17:21:29 *** panx has joined #rtems 2012-03-26T17:22:05 peerst: compiled for you now? 2012-03-26T17:22:35 or just applied patch and don't have time to try it now 2012-03-26T17:23:24 compiling rtems now .... 2012-03-26T17:24:27 Hello, I am considering adding atomic operation to the proposal to application of SOC. 2012-03-26T17:24:33 hi panx 2012-03-26T17:25:02 that is a good project, and there already is a student who is planning to propose it 2012-03-26T17:25:15 not to worry there are lots of good projects to do 2012-03-26T17:25:33 what interests you? 2012-03-26T17:26:03 The idea of adding ck primitives. 2012-03-26T17:27:56 So, I try to compile it under the pc386 2012-03-26T17:28:07 panx: it is possible that integrating ck would be acceptable, although having two students work on quite similar projects is a minor concern 2012-03-26T17:28:26 OK. I got it. 2012-03-26T17:28:41 I am not so insisting on this. 2012-03-26T17:28:56 the other student is investigating how to provide atomic operations in the rtems kernel code though, whereas ck will live outside of rtems and so probably cannot be used in the kernel 2012-03-26T17:29:02 ok. there are a lot of good projects to go around 2012-03-26T17:29:11 what kind of projects / programming do you want to do? 2012-03-26T17:29:20 gedare: rtems build, application bulds now seeing if it works 2012-03-26T17:29:40 peerst: cool fingers crossed 2012-03-26T17:30:11 For the language, C, python and assembly. 2012-03-26T17:30:12 it seemed to have fixed some rsource limit problems ... now crashing with a different error ;-) 2012-03-26T17:30:33 panx: ok good. so anywhere from low-level to high-level project? 2012-03-26T17:30:45 peerst: haha then it has some worth! :) 2012-03-26T17:31:36 do you refer the drivers as low level? 2012-03-26T17:31:41 panx: i'd like to spotlight two projects: http://www.rtems.com/wiki/index.php/Bdbuf_improvements and http://www.rtems.com/wiki/index.php/RTEMS_Testing 2012-03-26T17:31:49 yeah. low-level will be closer and closer to hardware 2012-03-26T17:32:28 definitely ... the crash seems to be later ... now because of "pipe()" not working 2012-03-26T17:32:29 panx: if you have any questions on the testing project feel free to ask 2012-03-26T17:32:43 Thx. 2012-03-26T17:33:26 *** panzon_ has quit IRC (Ping timeout: 244 seconds) 2012-03-26T17:33:27 ok peerst, if you are on rtems-devel just send a short blurb about the utility of the patch... or we can wait to see if it gets ralfed first ;) 2012-03-26T17:33:56 have to get on rtems-devel first but will do 2012-03-26T17:34:15 no biggy.. joel wants me to revise the patch to include documentation in the user manual 2012-03-26T17:34:57 another question: we have select() but no poll() ? 2012-03-26T17:35:12 i think that may be correct...it's come up before on the ml 2012-03-26T17:37:07 i'm not aware if the libbsd work will change that or not 2012-03-26T17:41:58 For the testing project, what is the relationship of the testsuites of rtems and the creation of 4 seperate test groups? 2012-03-26T17:42:07 verm__ ^ 2012-03-26T17:42:53 Do you mean that maybe we can reuse some the tests in the testsuite and maybe we need to add something in need? 2012-03-26T17:43:16 panx; verm__ proposed that project so hopefully he'll come in but i'll try to explain a bit too 2012-03-26T17:43:31 here 2012-03-26T17:43:37 panx: yes exactly. most of the tests that are in testsuites right now are cpu-independent 2012-03-26T17:43:44 panx: the tests in the testsuite need to be refacorded the tests themselves are OK 2012-03-26T17:44:00 but the harness is poor, it needs to be modernised 2012-03-26T17:44:28 i should update the page that no actual tests need to be written 2012-03-26T17:44:36 if you get a group of tests that contains say, 15 tests 2012-03-26T17:44:52 that needs to either be put into a new harness or split into several tests, depending on the situation 2012-03-26T17:45:13 verm__: well.. i don't know about not writing new tests; if you want to have tests for the BSP or for board-specific those don't exist right now AFAIK 2012-03-26T17:45:46 unless you just mean to run the tests we have for a specific BSP or a specific board... 2012-03-26T17:45:59 what's important is to have the harness there 2012-03-26T17:46:05 so we *can* write tests 2012-03-26T17:46:07 true so that new tests can be fit in appropriately 2012-03-26T17:46:10 good point. 2012-03-26T17:46:12 as long as there is a few for an example that's fine :) 2012-03-26T17:46:37 the harness itself is a lot of work on it's own 2012-03-26T17:46:40 i'll leave you two to it for now... i'll keep an eye on this convo 2012-03-26T17:46:53 *** panzon_ has joined #rtems 2012-03-26T17:47:48 gedare: sent my praise to the ml :-) 2012-03-26T17:47:53 thanks peerst:) 2012-03-26T17:47:58 have a nice evening 2012-03-26T17:48:18 thanks, you 2 2012-03-26T17:48:59 For the modernization of tests, can you be more specific. I don't have much experience in testing. 2012-03-26T17:49:22 panx: the current harness or "framework" used was created by RTEMS developers it doesn't allow for any modern testing techniques 2012-03-26T17:49:59 I got it. So no modern testing techniques being used at all? 2012-03-26T17:50:05 OK 2012-03-26T17:50:07 softgware testing is generally broken into 4 major groups: Unit, Regresison, Operational and "Mock tests" which are actually used within the first three 2012-03-26T17:50:36 for instance, we want to run most tests in two different situations 2012-03-26T17:50:53 we want to test the lowlevel API on a host-based test (googletest) then we want to get that same test and run it within an instance of RTEMS 2012-03-26T17:51:07 in this way we test "is this API OK?" and "does the code using this API work correctly?" 2012-03-26T17:51:35 Is this regression testing? 2012-03-26T17:51:37 a lot of our tests can be duplicated between both situatiosn easily 2012-03-26T17:51:44 panx: no, this is unit testing 2012-03-26T17:51:53 a regression test is only put in place after you have fixed a bug 2012-03-26T17:52:01 once you fix that bug you add a test to make sure that bug stays fixed forever 2012-03-26T17:52:11 to stop a 'regression' of that bug. 2012-03-26T17:52:20 alright. 2012-03-26T17:52:33 they're only written on-demand.. a unit test involves taking a "unit" of something and testing it 2012-03-26T17:52:40 this can be an API call, but also the code behind that call 2012-03-26T17:52:55 a single API call could be a significant portion of code, you must test that API plus all the code inside of it 2012-03-26T17:53:13 most of this can be done on the host using googletest 2012-03-26T17:53:21 but if the API call is for instance, a file system call 2012-03-26T17:53:26 you want to do that test while running inside of RTEMS 2012-03-26T17:53:57 this project will get into some fairly advanced software testing techniques 2012-03-26T17:54:09 Cool. 2012-03-26T17:54:17 it's good knowledge, but be aware it has little to do with RTEMS itself 2012-03-26T17:54:25 it's all about software testing :) 2012-03-26T17:54:44 Yes. 2012-03-26T17:54:54 panx: just to clarify, some of our tests in testsuites are unit tests, regression tests, or integration tests. but they have all just sprung up and there is not much rhyme or reason. 2012-03-26T17:55:20 the harness is terrible 2012-03-26T17:55:27 it's basically ./run > output 2012-03-26T17:55:30 and all of the tests are run inside rtems (on a simulator usually) 2012-03-26T17:55:35 md5 output md5 2012-03-26T17:55:41 if a != b then fail 2012-03-26T17:55:42 yep 2012-03-26T17:55:53 that's what they used to do in the 80s 2012-03-26T17:56:05 that's our "current" approach to testing ;) 2012-03-26T17:56:19 can't go wrong with hash comparisons :) 2012-03-26T17:56:46 i have somewhere a classification of the testsuites where i copied their doc files into a google doc would you like it? might help with sorting them out I dont' know. 2012-03-26T17:57:12 that sounds useful 2012-03-26T17:57:23 ok i'll share with you. then you can share with your student when you get one 2012-03-26T17:57:39 sounds like panx might be interested :) 2012-03-26T17:58:20 Yes. I am interested. How long do you expect this testing project to last? 2012-03-26T17:59:00 that's a hard question to answer, basically we'll lay the project out in order 2012-03-26T17:59:04 starting from the lowest level 2012-03-26T17:59:23 however far you get in is as far as you complete you're more than welcome to stay around after GSoC and continue :) 2012-03-26T17:59:32 i have one student in NTP (http://www.ntp.org/) 2012-03-26T17:59:41 who will hopefully be coming for his 3rd year on the same project. 2012-03-26T18:00:13 NTP started with nothing, though.. RTEMS has a completely different set of requirements 2012-03-26T18:02:35 the important thing is to define enough small tasks that can be contributed back during the summer so that you aren't just working all summer and then trying to commit stuff at the end 2012-03-26T18:02:54 yeah, exactly 2012-03-26T18:03:22 i've always ensured my tasks are broken into little parts so in the end both parties have a useful result 2012-03-26T18:03:24 panx: while the whole project may take a long time to complete... you will lay the groundwork for it and if you enjoy the work you will be free to continue working on it 2012-03-26T18:03:35 yep 2012-03-26T18:04:56 Good. Even though I was attracted by the idea of embedded operating system to rtems, I am really happy to learn new stuff and do as many useful thing as I can 2012-03-26T18:05:55 But is it OK that I practically have little testing experience and only some knowledge of testing at higher level than this? 2012-03-26T18:06:14 having any knowledge of testing is not a prerequiset 2012-03-26T18:06:47 you do need to know C and demonstrate that you can setup RTEMS 2012-03-26T18:06:59 also cursory knowledge of C++ is required 2012-03-26T18:07:18 it's very basic C++ 2012-03-26T18:08:08 gedare: i added more info under "Additional Information": http://wiki.rtems.org/wiki/index.php/RTEMS_Testing 2012-03-26T18:08:53 Yes. A lot of my coursework is done in C. I have set up RTEMS. C++ is fine with me. 2012-03-26T18:09:50 looks good 2012-03-26T18:10:54 panx: just make sure you're truly interested in doing a testing project... there are plenty of projects available including more programming-oriented ones. Ultimately your success will hinge on how interested you are in your project. 2012-03-26T18:11:59 not trying to talk you out of it... just want you to be sure before committing yourself for a summer :) 2012-03-26T18:12:34 I understand. 2012-03-26T18:12:59 I am thinking about it. I am really interested about Embedded operating system. 2012-03-26T18:13:26 but not limited to it. 2012-03-26T18:13:56 ok. well, testing is important for any project, and the quirks of the embedded OS make testing it an interesting problem 2012-03-26T18:14:02 especially due to cross-compilation 2012-03-26T18:14:10 Yes. 2012-03-26T18:16:04 aside from refining the ideas in the wiki page for your proposal, I would suggest that you try to get RTEMS to run on as many simulators as possible 2012-03-26T18:17:38 common ones include: sparc/sis (sis bsp, sparc-gdb), psim (psim bsp, powerpc-gdb), qemu/x86 (pc386 bsp, qemu), qemu/arm 2012-03-26T18:17:45 qemu/arm has some special needs see http://www.rtems.org/pipermail/rtems-devel/2012-March/000446.html 2012-03-26T18:18:12 this way you'll be able to run the tests on a bunch of target hardware to get some belief that your testing harness is usable across platforms. 2012-03-26T18:18:48 panx: that is, if you do go ahead with proposing to do the testing project :) 2012-03-26T18:19:51 Thx. 2012-03-26T18:54:14 byte 2012-03-26T18:54:15 bye 2012-03-26T18:54:17 *** gedare has quit IRC (Quit: Ex-Chat) 2012-03-26T19:05:51 *** panx has quit IRC (Ping timeout: 245 seconds) 2012-03-26T19:11:10 *** panx has joined #rtems 2012-03-26T19:11:46 Where can I find the instruction regarding installing the various bsps? 2012-03-26T19:45:28 bye 2012-03-26T19:45:31 *** panx has left #rtems 2012-03-26T20:04:17 *** zw_yao has joined #rtems 2012-03-26T20:08:25 *** xiangfu has joined #rtems 2012-03-26T20:29:50 *** zw_yao has quit IRC (Ping timeout: 246 seconds) 2012-03-26T21:14:10 *** arvind_khadri has quit IRC (Ping timeout: 252 seconds) 2012-03-26T21:20:26 *** gedare has joined #rtems 2012-03-26T21:34:46 *** gedare has quit IRC (Quit: Ex-Chat) 2012-03-26T21:38:47 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-26T21:42:40 *** xiangfu has joined #rtems 2012-03-26T22:26:17 *** antgreen has quit IRC (Remote host closed the connection) 2012-03-27T00:49:05 *** arvind_khadri has joined #rtems 2012-03-27T01:13:11 *** sebhub has joined #rtems 2012-03-27T01:18:59 good morning 2012-03-27T02:11:12 *** silverio_di has joined #rtems 2012-03-27T03:18:07 *** panzon_ has quit IRC (Ping timeout: 246 seconds) 2012-03-27T03:31:27 *** panzon_ has joined #rtems 2012-03-27T03:57:00 morning sebhub :-) 2012-03-27T03:57:26 hi 2012-03-27T04:18:00 *** weiY has joined #rtems 2012-03-27T04:58:42 Hi all . Is there any potential mentor have time to discuss the atomic primitive project 2012-03-27T05:00:16 Now the peroid of submitting proposal is beginning. but it seems that this project has some change about its direction 2012-03-27T05:01:48 My proposal based on the orignal goal of this project is basically compeleted, so now it is time to determine the direction of this project and i can modify the proposal in time 2012-03-27T08:13:55 *** weiY has quit IRC (Ping timeout: 264 seconds) 2012-03-27T08:20:41 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-27T08:41:12 *** weiY has joined #rtems 2012-03-27T09:06:01 *** phipse has joined #rtems 2012-03-27T09:23:41 *** gedare has joined #rtems 2012-03-27T09:26:38 Hi gedare, good night 2012-03-27T09:26:47 hello weiY good morning :) 2012-03-27T09:27:06 hah, have you seen the mail about atomic project ? 2012-03-27T09:27:44 from yesterday yes.. did you upload your project to melange? 2012-03-27T09:27:49 it seems that the direction of this project is changing 2012-03-27T09:27:50 and apply to be a student 2012-03-27T09:27:55 yes but that is ok 2012-03-27T09:28:40 i have apply to be a student, but not yet upload my proposal to melange 2012-03-27T09:28:43 you have a bit of work to figure out how to use the C11 atomics but I think if the compiler support is strong enough it is good 2012-03-27T09:28:45 okay 2012-03-27T09:29:08 *** zw_yao has joined #rtems 2012-03-27T09:30:36 brb meeting 2012-03-27T09:30:43 yes, but now firstly we must determine the proposal of this project. 2012-03-27T09:30:45 ok 2012-03-27T09:31:45 should we use the C11 atomic support? i think there still are somethingto be discussed 2012-03-27T09:38:23 the benefit of using c11 atomic support is that it doesn't add a new dependency to rtems 2012-03-27T09:38:34 we can simply use the suff from gcc 2012-03-27T09:38:38 stuff 2012-03-27T09:38:59 and we don't have to maintain it 2012-03-27T09:40:01 we still have to look at this c11 atomics and figure out if they are in general suitable for us 2012-03-27T09:40:21 but at the moment i don't see why not 2012-03-27T09:40:48 yes, i agree with you about using c11 atomic support 2012-03-27T09:41:24 but first problem is whether it is stable enough for us to using it 2012-03-27T09:41:45 the standard is stable 2012-03-27T09:42:03 the implementation is currently experimental 2012-03-27T09:42:11 second we must investigate how to use it suitable for us as you said 2012-03-27T09:42:51 yeah, i belive the desigh of it include API is stable 2012-03-27T09:44:34 about how to use it i think we can refer the implementation of FreeBSD which provide a header wraping the atomic API from GCC 2012-03-27T09:45:13 instead of using it directly from compiler 2012-03-27T09:46:51 freebsd has already a wrapper to the c11 atomics? 2012-03-27T09:47:55 i think so. http://svnweb.freebsd.org/base/head/include/stdatomic.h?view=markup 2012-03-27T09:48:59 it just redefine the API on the C11 atomics 2012-03-27T09:51:16 the stdatomic.h is also part of GCC 2012-03-27T09:52:22 what i meant is http://www.freebsd.org/cgi/man.cgi?query=atomic&apropos=0&sektion=0&manpath=FreeBSD+9.0-RELEASE&arch=default&format=html 2012-03-27T09:52:35 but in the stdatomic.h it does not include any other header from gcc 2012-03-27T09:54:41 the linker above is atomic_ops of freebsd. what is your mean? 2012-03-27T09:55:11 this is the atomic API of the freebsd kernel 2012-03-27T09:57:43 yeah, i have look into its implementation. and it is different from the its application lib (header stdatomic.h http://svnweb.freebsd.org/base/head/include/stdatomic.h?view=markup) 2012-03-27T10:03:14 hi sebhub. in my understanding the C11 atomics is built-in GCC (libgcc) not supported by libc, right? 2012-03-27T10:03:33 yes 2012-03-27T10:05:08 and if i write a C application which want to use C11 atomic i just using its API directly and link with libgcc, it will ok? but where is the stdatomic.h stored? 2012-03-27T10:06:03 *** silverio_di has quit IRC () 2012-03-27T10:08:40 the stdatomic.h ships with the gcc 2012-03-27T10:09:32 ok, need we explicit include stdatomic.h or not? 2012-03-27T10:09:53 yes, of course 2012-03-27T10:12:06 i have a question about the above link http://svnweb.freebsd.org/base/head/include/stdatomic.h?view=markup 2012-03-27T10:13:28 this file stdatomic.h is provided by freebsd for application. in its contents it use __atomic_fetch_and like stuff which provided by compiler but it does not include any header, why? 2012-03-27T10:15:52 and i had find the path of my install gcc version 4.+, i found there is only a stdatomic.h in xxxx/include/c++/ dir 2012-03-27T10:17:18 try to use gcc 4.7 2012-03-27T10:18:00 the c++ standard had the atomics earlier 2012-03-27T10:23:57 *** zw_yao has quit IRC (Quit: Leaving) 2012-03-27T11:08:23 *** weiY has quit IRC (Ping timeout: 245 seconds) 2012-03-27T11:11:27 *** sebhub has quit IRC (Ping timeout: 244 seconds) 2012-03-27T11:27:33 *** arvind_khadri has quit IRC (Ping timeout: 248 seconds) 2012-03-27T12:13:31 *** gedare has quit IRC (Quit: Ex-Chat) 2012-03-27T12:45:44 *** arvind_khadri has joined #rtems 2012-03-27T12:45:44 *** arvind_khadri has joined #rtems 2012-03-27T14:13:25 *** arvind_khadri has quit IRC (Ping timeout: 248 seconds) 2012-03-27T14:44:19 *** xiaochen has joined #rtems 2012-03-27T14:51:09 Can you give the name of bsp for certain cpu as you want it to be? 2012-03-27T15:03:09 I'm not sure if I get you right, but see README.configure line 78 2012-03-27T15:05:30 *** Hesham has joined #rtems 2012-03-27T15:23:00 *** gedare has joined #rtems 2012-03-27T15:29:37 *** panzon_ has quit IRC (Ping timeout: 265 seconds) 2012-03-27T15:32:56 *** panzon_ has joined #rtems 2012-03-27T15:38:45 xiaochen: did you figure out what you wanted? 2012-03-27T15:39:40 You mean the project? Or the question I asked yesterday? 2012-03-27T15:39:59 [14:51:09] Can you give the name of bsp for certain cpu as you want it to be? 2012-03-27T15:40:57 I want to build bsp for arm. There are several bsps for it. I don't know which one to use. 2012-03-27T15:41:54 to run it on arm-gdb or or qemu? 2012-03-27T15:42:05 or do you have a real hardware 2012-03-27T15:42:16 qemu 2012-03-27T15:42:52 check http://www.rtems.org/pipermail/rtems-devel/2012-March/000446.html 2012-03-27T15:43:23 that would be the lm3s69xx bsp 2012-03-27T15:44:53 OK 2012-03-27T15:44:58 Thx. 2012-03-27T15:45:26 if you get it working maybe you can update the Qemu wiki page with instructions ( http://www.rtems.com/wiki/index.php/Qemu ) :) 2012-03-27T15:45:35 if you have time 2012-03-27T15:45:48 ok 2012-03-27T15:45:58 i did see that wiki page 2012-03-27T15:46:32 yeah it only applies for the pc386 / x86 target 2012-03-27T15:46:56 i tried skyeye but in vain 2012-03-27T15:47:01 and Armulator 2012-03-27T15:47:25 QEMU is only for lms3s69xx 2012-03-27T15:47:36 but i do not work for arm920 2012-03-27T15:47:43 it ** 2012-03-27T15:48:57 and when i tried to compile rtems for bsp = lms3s69xx , nothing happens !! 2012-03-27T15:49:19 arm-unknown-bsp 2012-03-27T15:51:45 try asking about it on rtems-users mailing list then 2012-03-27T15:54:39 i already did and someone gave me that link for lms3s69xx patches 2012-03-27T15:54:58 it seems that rtems have problems with simulators 2012-03-27T15:55:43 I think GDB is the most suitable one as it works for all BSPs i think 2012-03-27T15:56:39 qemu has some more features, just takes a little more to get it working 2012-03-27T15:57:09 it would be nice to get more thorough instructions on getting the different targets to run on whatever simulators 2012-03-27T15:57:14 hard to maintain all that info right now 2012-03-27T15:58:26 i am thinking of a solution 2012-03-27T15:58:41 when i run qemu-system-arm 2012-03-27T15:58:50 it asks for kernel image 2012-03-27T15:59:38 so i run qemu-system-arm -kernel hello.exe but it gets only black box 2012-03-27T15:59:48 did i miss any thing ? 2012-03-27T16:01:13 possibly it is how you configure rtems when you build i think you need to specify some thing about the console 2012-03-27T16:01:35 not sure exactly. ask on ml :) 2012-03-27T16:03:03 hesham: the README says to use hello.bin does that make a difference 2012-03-27T16:03:06 I did do that but when i tried to configure lms 2012-03-27T16:03:27 i do not have hello.bin 2012-03-27T16:03:35 only hello.exe & hello.ralf 2012-03-27T16:04:05 ok.. best bet is to ask on ml. sebhub can probably give some assist 2012-03-27T16:05:13 i will try some solution and if it did not work i will ask on ml 2012-03-27T16:05:17 another thing 2012-03-27T16:05:51 i am currently working on writing a simple OS for arm but i am not finished with it yet 2012-03-27T16:06:04 should i attach the link in the proposal ? 2012-03-27T16:11:05 Hesham: How you test it with hello.exe? 2012-03-27T16:11:27 qemu-system-arm -kernel hello.exe 2012-03-27T16:12:26 hesham: if you want to include a link to your other work that is ok.. in case someone wants to see. 2012-03-27T16:12:50 hesham: do you plan to put your simple OS implementation on hold if you are accepted for gsoc? 2012-03-27T16:13:03 of course 2012-03-27T16:13:40 just making sure:) 2012-03-27T16:13:50 actually i am working on my OS to prove that i can code in C and ARM if any body wants a prove :) 2012-03-27T16:14:07 well if you do this project then you will have more proof. 2012-03-27T16:14:12 Hesham: Have you run the command lms3s69xx directly? 2012-03-27T16:14:48 gedare:i am trying hard to finish it :) 2012-03-27T16:15:53 xiaochen: when i tried to compile lms3s69xx BSP nothing happens and no directories added in the build directory for that BSP 2012-03-27T16:16:33 How did you compile it? 2012-03-27T16:16:52 configured it first with that option included 2012-03-27T16:17:25 ./configure --enable-rtemsbsp=lms3s69xx 2012-03-27T16:17:25 hesham: what is your whole configure line? 2012-03-27T16:17:42 you should be configuring from outside of the source tree 2012-03-27T16:18:01 yeah i did that in the building directory 2012-03-27T16:18:08 yes. 2012-03-27T16:18:19 so it should be something like ../rtems/configure ... 2012-03-27T16:18:29 i did so :) 2012-03-27T16:18:39 hard to debug when you don't tell us exactly what you do :) 2012-03-27T16:18:45 also hard for someone else to follow your steps 2012-03-27T16:19:43 i am thinking now of re configuring arm920 with that option USE_COM1_AS_CONSOLE=1 BSP_PRESS_KEY_FOR_RESET=0 2012-03-27T16:20:01 and make another try with QEMU 2012-03-27T16:20:15 i dont' think it will work pretty sure USE_COM1_AS_CONSOLE is only for pc386 bsp 2012-03-27T16:20:26 i don't know why you only see blank screen. ask on ml ;) 2012-03-27T16:20:55 i will do now 2012-03-27T16:21:04 gedare: comments updated on the proposal 2012-03-27T16:21:13 and by the way, it is late, I have to sleep :D 2012-03-27T16:21:22 heh. thanks goodnight juli1 2012-03-27T16:23:32 gedare: I have decided to propose the testing project. But I am a little behind about what I should propose on it. I had little experience in testing. 2012-03-27T16:23:47 xiaochen: are you panx? 2012-03-27T16:23:54 yes 2012-03-27T16:24:01 ok sounds good just making sure 2012-03-27T16:24:18 it's ok to be a little behind right now 2012-03-27T16:24:44 make sure to save whatever conversations you had on IRC with verm__ and myself. 2012-03-27T16:25:10 OK 2012-03-27T16:25:22 a good place to start is to just start transforming the project description wiki page and the conversations we had into a basic proposal in a Google Doc 2012-03-27T16:25:35 invite us to review it so we can give some comments as you form your ideas 2012-03-27T16:25:49 and copy-paste from the Google Doc into the application form on Melange periodically just to keep that up-to-date 2012-03-27T16:26:21 OK 2012-03-27T16:26:28 make sure you get the simple hello world test run...and work on getting a couple other simulators to work if you can 2012-03-27T16:26:52 I am trying on other simulators today 2012-03-27T16:27:01 and other than that start to familiarize yourself with the projects that verm__ suggested to use for setting up the test harness 2012-03-27T16:27:01 should i make some some colors with the proposal ? or just keep it simple as it is ? 2012-03-27T16:27:11 simplify! 2012-03-27T16:27:57 xiaochen: and don't forget to breathe ;) 2012-03-27T16:28:13 :) 2012-03-27T16:28:51 it's better if you can get a first draft of your proposal done sooner so we can comment on it and you can revise it before the deadline next week 2012-03-27T16:29:12 OK. 2012-03-27T16:29:31 How to invite you to review it? 2012-03-27T16:29:49 invite gedare at gwmail dot gwu dot edu 2012-03-27T16:30:03 make sure to allow me to at least make comments. 2012-03-27T16:30:33 OK. Thx. And can I have verm's email? 2012-03-27T16:30:47 we can also make comments on your proposal when you upload it in Google Melange...but the interface is not as good 2012-03-27T16:31:07 i have it somewhere... 2012-03-27T16:31:45 i sent to you in pm 2012-03-27T16:41:19 there you go hesham, joel answered you 2012-03-27T16:41:34 i will check it 2012-03-27T16:45:47 well , i will try with connex now , wish me good luck :) 2012-03-27T16:47:13 Which bsp for mips should I build for qemu simulation? 2012-03-27T16:59:07 i don't know what the best simulator for mips is.. try asking on the rtems-users mailing list i guess 2012-03-27T17:00:28 xiaochen: qemu-system-mips -M ? 2012-03-27T17:00:29 Supported machines are: 2012-03-27T17:00:29 mipssim MIPS MIPSsim platform 2012-03-27T17:00:29 malta MIPS Malta Core LV (default) 2012-03-27T17:00:29 magnum MIPS Magnum 2012-03-27T17:00:29 pica61 Acer Pica 61 2012-03-27T17:00:31 mips mips r4k platform 2012-03-27T17:00:45 those are the machines that qemu supports. but i don't know what bsp will work with what machine, if any 2012-03-27T17:01:35 also my qemu is outdated so other ones might be available 2012-03-27T17:10:03 Gotta, thx. 2012-03-27T17:15:28 The bsp csb350 goes with skyeye 2012-03-27T17:15:41 Not qemu. 2012-03-27T17:16:30 And so does csb360 2012-03-27T17:19:12 They are bsp for mips. 2012-03-27T17:21:12 So, should I install skyeye? 2012-03-27T17:25:35 probably a good one to try using 2012-03-27T17:25:47 i personally haven't tried it 2012-03-27T17:26:01 but with a testing project the more targets you can run the better 2012-03-27T17:26:26 OK. 2012-03-27T17:27:01 xiaochen: please if you fine a suitable package for skyeye tell me 2012-03-27T17:27:29 Hesham: OK 2012-03-27T17:27:35 the last one i found was 2006 version , and when i tried to run its Make file it gives me an error 2012-03-27T17:31:29 gedare: should i delete build folder for a pre-compiled BSP to compile a new one ?? 2012-03-27T17:31:48 I think yes. 2012-03-27T17:31:56 for the same BSP i usually do 2012-03-27T17:31:59 http://sourceforge.net/projects/skyeye/files/skyeye/ 2012-03-27T17:32:03 I have one build folder for each BSP that I build 2012-03-27T17:32:22 although i think there might be other ways you can use the same build folder for multiple bsps 2012-03-27T17:32:24 i just don't bother 2012-03-27T17:32:54 i will try with a new build folder 2012-03-27T17:33:01 good looks like they just cut a skyeye release 2012-03-27T17:33:28 yeah the link for linux does not work any more 2012-03-27T17:33:49 we have a wiki page about how to use it, xiaochen did you find that page already? 2012-03-27T17:34:01 (http://www.rtems.com/wiki/index.php/ARM-EABI) Here it says that armcortexa9 is supported 2012-03-27T17:34:41 but when i searched for its library in armbsp library i did not fine it 2012-03-27T17:35:16 gedare: The page on skyeye? 2012-03-27T17:38:26 yeah 2012-03-27T17:39:16 Not yet. 2012-03-27T17:39:25 hesham: i don't know about that bsp. maybe it is an external/experimental one 2012-03-27T17:39:41 or maybe it got folded into some generic bsp. you'd have to ask on the list (i don't use arm much) 2012-03-27T17:40:05 http://www.rtems.com/wiki/index.php/SkyEye 2012-03-27T17:40:09 i will try another BSP 2012-03-27T17:40:14 it's mostly stubs :) 2012-03-27T17:40:22 Good. THx! 2012-03-27T17:40:41 well you can't just try random BSPs... they should match the hardware that is being simulated 2012-03-27T17:41:34 OK. 2012-03-27T17:42:16 I am doing with the hardware you recommended yesterday. 2012-03-27T17:42:59 i know , i am trying to fine a match and i think i found one , i am trying with it now 2012-03-27T17:45:08 *** phipse has quit IRC (Quit: Leaving) 2012-03-27T17:54:00 Hesham, what compilation error did you encounter? 2012-03-27T17:54:15 in skyeye ? 2012-03-27T17:54:20 yes 2012-03-27T17:54:59 i do not actually remember , wait for a moment i will recompile it and get the error 2012-03-27T17:55:27 OK. You can just use the package from this link. It is 2012 version 2012-03-27T17:55:46 http://www.rtems.com/wiki/index.php/SkyEye 2012-03-27T18:03:09 Sorry the compilation error was on Armulator 2012-03-27T18:03:34 OK 2012-03-27T18:03:59 if it worked with you successfully tell me please 2012-03-27T18:29:35 *** xiaochen has left #rtems 2012-03-27T18:59:32 undefined reference to bsp_interrupt_hanlder_default 2012-03-27T19:00:45 that is what i get when compiling gumstix bsp , i think there is an error with it 2012-03-27T19:38:06 is it spelt bsp_interrupt_hanlder_default or did you make that typo 2012-03-27T19:38:11 (should be handler) 2012-03-27T19:38:28 we dont' hear much about gumstix so it may be outdated a bit i dunno 2012-03-27T19:39:05 that what happens when i tried to compile it 2012-03-27T19:39:13 i am trying another bsp noew 2012-03-27T19:41:18 **now 2012-03-27T19:49:15 *** zw_yao has joined #rtems 2012-03-27T19:51:52 hi,all. Since I've put my GSoC proposal on google-melange.com, where should I discuss: on the comments below my proposal or the mail list? 2012-03-27T19:56:44 address the comments that mentors give you on the proposal there directly, but otherwise put questions and general discussions about your project on the mailing list 2012-03-27T19:56:53 i'm giving some feedback on your google doc right now 2012-03-27T19:56:59 zw_yao: ^ 2012-03-27T20:00:31 gedare:oh, I see it just now. Thanks! 2012-03-27T20:01:32 I also have a comment below my proposal from Julen Delange, I don't know where I should continue the discuss with Julen. 2012-03-27T20:01:52 respond to that comment right there 2012-03-27T20:02:19 I see, thanks:) 2012-03-27T20:03:35 gedare: has anyone else expressed interest in the testing project? 2012-03-27T20:04:23 verm__ the same person 2012-03-27T20:04:31 just changed their irc nick 2012-03-27T20:04:34 oh 2012-03-27T20:04:44 confused me for a minute too :) 2012-03-27T20:04:48 xiaochen = panx 2012-03-27T20:05:04 ahh, was a proposal submitted? 2012-03-27T20:05:22 not yet 2012-03-27T20:05:34 ok good i was wondering if i went blind :) 2012-03-27T20:05:42 i said to work on that while also trying out some more simulators and getting to know some of the tools you suggest using 2012-03-27T20:06:07 ah 2012-03-27T20:56:53 *** Hesham has left #rtems 2012-03-27T21:02:30 *** xiangfu has joined #rtems 2012-03-27T22:36:41 *** gedare has quit IRC (Quit: Ex-Chat) 2012-03-28T00:43:55 *** panzon_ has quit IRC (Ping timeout: 252 seconds) 2012-03-28T01:09:35 *** silverio_di has joined #rtems 2012-03-28T01:22:48 *** arvind_khadri has joined #rtems 2012-03-28T01:22:53 *** arvind_khadri has joined #rtems 2012-03-28T01:46:39 *** shine_ has joined #rtems 2012-03-28T01:53:20 *** shine_ has quit IRC (Remote host closed the connection) 2012-03-28T01:56:41 *** sebhub has joined #rtems 2012-03-28T02:08:27 good morning 2012-03-28T02:11:28 g'morning you ;) 2012-03-28T02:33:34 sebhub, hi 2012-03-28T03:28:00 *** arvind_khadri has quit IRC (Ping timeout: 272 seconds) 2012-03-28T03:31:29 *** sevikkk has quit IRC (Ping timeout: 246 seconds) 2012-03-28T03:41:34 *** Hesham has joined #rtems 2012-03-28T03:42:04 hey sebhub 2012-03-28T03:46:54 *** sevikkk has joined #rtems 2012-03-28T03:51:19 *** zw_yao has quit IRC (Ping timeout: 244 seconds) 2012-03-28T04:11:01 morning everybody :-) 2012-03-28T04:19:24 hi juli1 2012-03-28T05:06:58 good morning :) 2012-03-28T05:50:26 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-28T05:55:47 morning 2012-03-28T06:04:16 *** arvind_khadri has joined #rtems 2012-03-28T06:04:16 *** arvind_khadri has joined #rtems 2012-03-28T06:25:01 do we have some support for pipes in rtems? 2012-03-28T06:25:22 like e.g. a pipe() api function 2012-03-28T07:02:06 *** Hesham has left #rtems 2012-03-28T07:05:28 *** weiY has joined #rtems 2012-03-28T07:21:02 peerst: look at libcsupport/src/pipe.c 2012-03-28T07:21:31 this is implemented in the libfs 2012-03-28T07:21:45 see also pipe_create() in libfs/src/pipe/pipe.c 2012-03-28T07:24:06 *** panzon has joined #rtems 2012-03-28T08:07:44 *** xiangfu has joined #rtems 2012-03-28T08:25:50 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-28T08:26:10 *** cdcs has joined #rtems 2012-03-28T08:45:40 *** silverio_di has quit IRC () 2012-03-28T09:14:10 *** zw_yao has joined #rtems 2012-03-28T10:12:58 *** zw_yao has quit IRC (Quit: Leaving) 2012-03-28T10:31:56 *** sebhub has quit IRC (Ping timeout: 252 seconds) 2012-03-28T10:53:21 *** cdcs has quit IRC (Ping timeout: 245 seconds) 2012-03-28T13:21:13 *** arvind_khadri has quit IRC (Ping timeout: 260 seconds) 2012-03-28T13:23:12 *** arvind_khadri has joined #rtems 2012-03-28T13:23:13 *** arvind_khadri has joined #rtems 2012-03-28T14:35:11 juli1: oh great there are pipes ... now only need to find out why my pipe() call fails ... thanks! 2012-03-28T14:56:29 *** Hesham has joined #rtems 2012-03-28T15:47:54 peerst: you have to include the appropriate functionalities 2012-03-28T16:05:33 juli1: yeah I've found the CONF.*PIPES stuff 2012-03-28T16:17:55 *** Hesham has quit IRC (Read error: Connection reset by peer) 2012-03-28T16:18:36 peerst: :-) 2012-03-28T16:18:50 so, it is late 2012-03-28T16:18:56 have a good night all 2012-03-28T16:35:01 *** Hesham has joined #rtems 2012-03-28T17:34:41 *** gedare has joined #rtems 2012-03-28T17:45:31 gedare: seems like nobody is interested in my project :) 2012-03-28T17:53:28 hesham: i haven't had a chance to see.. do you have some comments or no comments? 2012-03-28T17:56:48 no comments 2012-03-28T17:56:55 patience then 2012-03-28T17:57:44 Ok :) 2012-03-28T18:24:10 *** Hesham has left #rtems 2012-03-28T19:24:36 and gedare chases another person away 2012-03-28T19:38:15 *** xiangfu has joined #rtems 2012-03-28T19:45:10 i do what i can 2012-03-28T20:00:02 *** zw_yao has joined #rtems 2012-03-28T20:23:14 *** weiY has quit IRC (Ping timeout: 272 seconds) 2012-03-28T20:58:05 *** xiangfu has quit IRC (Quit: Leaving) 2012-03-28T20:58:13 *** xiangfu has joined #rtems 2012-03-28T21:37:21 *** zw_yao has quit IRC (Read error: Connection reset by peer) 2012-03-29T00:08:03 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2012-03-29T01:50:25 *** sebhub has joined #rtems 2012-03-29T01:54:54 good morning 2012-03-29T01:55:15 good morning 2012-03-29T02:34:38 *** arvind_khadri has joined #rtems 2012-03-29T02:34:38 *** arvind_khadri has joined #rtems 2012-03-29T02:48:31 *** Hesham has joined #rtems 2012-03-29T03:59:03 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2012-03-29T04:12:09 morning everybody :D 2012-03-29T04:18:26 morning 2012-03-29T04:55:25 *** arvind_khadri has joined #rtems 2012-03-29T04:55:25 *** arvind_khadri has joined #rtems 2012-03-29T04:58:36 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2012-03-29T04:59:28 *** arvind_khadri has joined #rtems 2012-03-29T04:59:29 *** arvind_khadri has joined #rtems 2012-03-29T05:07:22 *** antgreen has joined #rtems 2012-03-29T05:08:38 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2012-03-29T05:15:46 *** Hesham has left #rtems 2012-03-29T05:26:15 *** arvind_khadri has joined #rtems 2012-03-29T05:26:15 *** arvind_khadri has joined #rtems 2012-03-29T05:41:42 *** camcelme3 has quit IRC (Read error: Connection reset by peer) 2012-03-29T05:42:06 *** camcelme3 has joined #rtems 2012-03-29T05:44:06 *** arvind_khadri has joined #rtems 2012-03-29T05:44:06 *** arvind_khadri has joined #rtems 2012-03-29T05:44:16 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2012-03-29T06:00:53 *** arvind_khadri has joined #rtems 2012-03-29T06:18:38 *** arvind_khadri has joined #rtems 2012-03-29T06:36:43 *** arvind_khadri has joined #rtems 2012-03-29T06:36:43 *** arvind_khadri has joined #rtems 2012-03-29T07:37:57 *** zw_yao has joined #rtems 2012-03-29T07:47:22 *** weiY has joined #rtems 2012-03-29T07:47:27 *** zw_yao has joined #rtems 2012-03-29T07:50:48 *** zw_yao has quit IRC (Client Quit) 2012-03-29T07:51:34 gedare:Hi, good morning 2012-03-29T08:37:50 *** xiangfu has quit IRC (Remote host closed the connection) 2012-03-29T08:42:06 *** zw_yao has joined #rtems 2012-03-29T08:44:05 *** zw__yao has joined #rtems 2012-03-29T08:46:08 *** zw_yao has quit IRC (Ping timeout: 240 seconds) 2012-03-29T09:13:31 *** arvind_khadri has joined #rtems 2012-03-29T09:14:50 *** zw__yao has quit IRC (Ping timeout: 252 seconds) 2012-03-29T09:17:10 *** zw_yao has joined #rtems 2012-03-29T09:30:24 good morning 2012-03-29T09:31:52 hello weiY 2012-03-29T09:35:54 hi gedare 2012-03-29T09:36:23 now i am reading metarial about gcc built-in atomic operations 2012-03-29T09:37:31 and i am also considering add the ISO C11 atomic operations support related informations to my proposal 2012-03-29T09:38:16 But as a starting point i think we should design a atomic API for RTEMS 2012-03-29T09:39:22 which is compatible with ISO C11 atomic for evolution 2012-03-29T09:39:41 what about your opinion? 2012-03-29T09:41:40 weiY: i think that sounds good. in your proposal make an argument for why just using C11 would not work 2012-03-29T09:41:46 GCC only 4.7.0 or later can support C11 atomic 2012-03-29T09:42:13 ok, i will list the reasom why not using C11 directly 2012-03-29T09:42:53 and now its implementation is not stable enough. 2012-03-29T09:43:11 the implementation on powerpc is not as good as i hoped: 2012-03-29T09:43:14 http://gcc.gnu.org/ml/gcc-help/2012-03/msg00377.html 2012-03-29T09:46:04 sebhub: looks like someone finally bit on bdbuf project 2012-03-29T09:46:42 yes, i am pleased 2012-03-29T09:47:04 i hope i have the time to mentor it 2012-03-29T09:51:43 sebhub, good work for atomic. i see in the standard quite"No other type requires lock-free operations, so the atomic_flag type is the minimum hardware-implemented type needed to conform to this International standard" 2012-03-29T09:52:17 yes, but on powerpc this is fullfulled by a 32-bit integer also 2012-03-29T09:52:31 in powerpc for atomic_flag operation it seems it not use lwarw and stwcx 2012-03-29T09:53:05 yes, but GCC uses 8-bit for the flag, thus we bit operations 2012-03-29T09:54:11 yeah, have you test this on the Clang/LLVM? 2012-03-29T09:54:30 it seem clang has already a basic support for atomic 2012-03-29T09:56:28 weiY: I think (as I emailed) the best way forward is to design an internal RTEMS atomic API (at the Score-layer). Eventually we might be able to use C11 to implement it, but for now I think the implementation will have to make calls into some libcpu/ implementation 2012-03-29T09:56:48 I'm actually not sure where the atomic api should live yet 2012-03-29T10:00:09 these internal RTEMS atomic API is only servered for RTEMS score-layer or other compenent also will use? 2012-03-29T10:00:36 like application 2012-03-29T10:01:02 not application 2012-03-29T10:01:25 probably libcpu/libbsp code 2012-03-29T10:01:36 maybe... 2012-03-29T10:01:44 really it should just be in score 2012-03-29T10:01:54 and score should use them to provide synchronization primitives like spinlock 2012-03-29T10:02:07 that will be used by the rest 2012-03-29T10:02:25 yes i think most usage is to build synchronization primitives 2012-03-29T10:02:50 and then synchronization primitives will be used by other components 2012-03-29T10:04:17 i think the implementation of architecture-dependent should put into the score/libcpu 2012-03-29T10:05:29 correct which makes the project somewhat harder 2012-03-29T10:05:43 you can see how the cache manager works...that is split in a couple layers 2012-03-29T10:06:28 or you can see how I made an attempt at memory protection manager http://code.google.com/p/gsoc2011-rtems-mmu-support-project/ 2012-03-29T10:06:54 yeah i will see what you do 2012-03-29T10:08:27 you can also see how some architectures implement their context switch and interrupt handling... 2012-03-29T10:08:50 some are implemented at cpukit/score/cpu/@CPU@/, and some implement in libcpu/@CPU@ 2012-03-29T10:09:05 though eventually the implementation will be replaced by C11 atomic, so they should be place in somewhat concerntrated 2012-03-29T10:09:13 *** cdcs has joined #rtems 2012-03-29T10:09:15 maybe. 2012-03-29T10:09:55 if you make an attempt to be general (like the cache manager does) then it will be easier to remove implemntations that become unnecessary due to C11 2012-03-29T10:10:25 the advantage of C11 is that it (should) be implementable in score layer since it relies only on C langauge features it should be generic and multilib-safe 2012-03-29T10:11:40 yeah, the implementation of architecture-dependent which is also belong to OS closely related is the same as atomic 2012-03-29T10:13:52 we can just using a thin wraper of C11 atomc which is what the freebsd do now 2012-03-29T10:16:58 Hi gedare http://code.google.com/p/gsoc2011-rtems-mmu-support-project/ there is no code to be download? 2012-03-29T10:17:18 i cannot find anything to be downloadable 2012-03-29T10:25:24 *** arvind_khadri has quit IRC (Read error: No route to host) 2012-03-29T10:26:03 *** DrJoel has joined #rtems 2012-03-29T10:26:04 *** DrJoel has joined #rtems 2012-03-29T10:26:04 *** ChanServ sets mode: +o DrJoel 2012-03-29T10:26:36 sebhub, I know it is close to closing there .. but can you hang on a minute 2012-03-29T10:26:54 Rather than restore a backup, can we do a git reset to a specific version 2012-03-29T10:27:43 This command is useful if you notice some small error in a recent 2012-03-29T10:27:43 commit (or set of commits) and want to redo that part without showing 2012-03-29T10:27:44 the undo in the history. 2012-03-29T10:28:25 *** weiY has quit IRC (Ping timeout: 260 seconds) 2012-03-29T10:33:34 *** weiY has joined #rtems 2012-03-29T10:33:46 normally you can to these reset operations only on non-remote repositories 2012-03-29T10:34:17 How about a single patch patch to reapply the changes between those two dates or simply to revert everything I did except the last sparc changes 2012-03-29T10:34:20 once you publish a change to a remote repository it is immutable 2012-03-29T10:34:43 then "un-f*cking" with a single large patch is probably the least evil 2012-03-29T10:34:45 what did you actually do to reach the current state? 2012-03-29T10:34:46 weiY: look at "Source" link 2012-03-29T10:35:15 http://git.rtems.org/rtems/commit/?id=c90350485cb9a2aee3fc0a91aec32f09eba160c9 is the first patch to keep 2012-03-29T10:35:16 hi gedare. i have see the source code in the RTEMS latest repository 2012-03-29T10:35:25 another option is to delete the master branch and i check in my local version of it 2012-03-29T10:35:36 DrJoel: you should do git revert instead to generate a new commit that undoes things 2012-03-29T10:35:43 if i am understanding what you two are talking about 2012-03-29T10:36:03 oh yes you probably are. I accidentally undid every commit since about 15-March 2012-03-29T10:36:15 lol ooo 2012-03-29T10:36:26 i think you cannot repair this without causing severe problems for everyone with a git checkout 2012-03-29T10:36:31 i have a tree that is valid from last night 2012-03-29T10:36:41 you did you do this? 2012-03-29T10:36:43 revert is for a single patch - Given one existing commit, revert the change the patch introduces, 2012-03-29T10:37:07 best option is to restore from a backup. 2012-03-29T10:37:20 I have not attempted to undo it. But it seems that a single large patch will be best. 2012-03-29T10:37:29 OK.. let me see when backup is dated... brb 2012-03-29T10:37:43 well, every developer has a backup 2012-03-29T10:38:15 the tree i have on my desktop right now is probably the most recent valid one since i checked in the last change before this set was pushed by joel 2012-03-29T10:38:15 anyone who has the commits that disappeared will have its tree in a inconsistent state if you do a single patch or a revert. 2012-03-29T10:38:29 the big question is what did you do to loose the commits, this should be impossible for remote repositories 2012-03-29T10:38:50 hi gedare: i think we can put the API definition of atomic to score. and them implement a layer like cache manager to libcpu/share, and the architecture-dependent implementation are place libcpu/@CPU@, eventually we turn to C11 implementation we can just use a header which wrap the C11 atomic in the score. remove all the implementation under libcpu 2012-03-29T10:38:56 i agree with sebhub. that is very awkward. 2012-03-29T10:39:28 the script to do the backup ran at 7pm last night.. is that OK? 2012-03-29T10:39:41 DrJoel, can we see the history of commands that lead to this event? 2012-03-29T10:39:53 no.. I am not sure what I did. :( 2012-03-29T10:40:07 on which machine did you do this? 2012-03-29T10:40:19 weiY: sounds reasonable. 2012-03-29T10:40:25 was it on the same machine which contains the git master? 2012-03-29T10:40:38 No. It was from my local machine 2012-03-29T10:40:51 hm, then the git master configuration is broken 2012-03-29T10:41:27 I can restore a backup from last night at 7pm.. how would that be? 2012-03-29T10:41:50 yes, this is ok 2012-03-29T10:41:56 *** arvind_khadri has joined #rtems 2012-03-29T10:41:56 *** arvind_khadri has joined #rtems 2012-03-29T10:42:08 I will tar up the one in /data 2012-03-29T10:42:57 if you want to push something to a remote repo and get a non-fast-forward warning, you should be very careful 2012-03-29T10:43:33 apparently so. :( 2012-03-29T10:43:49 it is restored but the web view will take a minute to restore .. I don't know how to clear cache 2012-03-29T10:44:01 we should also make sure that such an operation is impossible in the future, this is a configuration option 2012-03-29T10:44:09 Agreed.. 2012-03-29T10:44:12 good night . i have to go to bed. see you 2012-03-29T10:44:16 gedare's personal git was updated 13 hours ago. 2012-03-29T10:44:23 now what to do to our local clones? 2012-03-29T10:44:28 This only impacted the master 2012-03-29T10:44:28 i'll re-push my changes 2012-03-29T10:44:53 good night weiY. 2012-03-29T10:45:34 hold on gedare.. let's see if I can pull it 2012-03-29T10:45:38 okay 2012-03-29T10:45:57 i'm not going near it until you claim it is once again stable ;) 2012-03-29T10:46:32 sebhub, when I pull, it isn't changing locally 2012-03-29T10:46:56 git pull --rebase upstream master 2012-03-29T10:47:43 still not looking different in gitk 2012-03-29T10:48:01 the git clone still returns the broken state 2012-03-29T10:48:28 I am cloning it again to see if it is better that way ... grrr. 2012-03-29T10:49:47 I didn't do something right .. I think I copied the bad onto the bad... hold on 2012-03-29T10:50:09 better than bad on good ;-) 2012-03-29T10:51:03 I am so stressed out.. my fingers are tingling.. :( 2012-03-29T10:51:21 relax 2012-03-29T10:51:35 i've always found that telling someone to relax just adds more stress 2012-03-29T10:52:19 I do have some valium here and that may just have been earned.. one can only expect so much from blood pressure meds 2012-03-29T10:55:48 it is restored.. I am test cloning it.. web view shows permission error :( 2012-03-29T10:56:56 i guess you lost your command line for what you did? no 'history'? 2012-03-29T10:57:26 no.. it is in a random window on one of 4 desktops.. damn it 2012-03-29T10:57:36 was just trying to get Daniel's change in before a meeting 2012-03-29T10:58:02 i sucessfully cloned it 2012-03-29T10:58:14 * gedare afraid to try ;) 2012-03-29T10:58:18 I was cloning my personal repo and it has the problem 2012-03-29T10:58:24 switching to master 2012-03-29T10:59:50 master looks ok.. now what to do about personal one 2012-03-29T11:00:16 git clone ssh://sebh@git.rtems.org/data/git/rtems.git git-rtems-tmp 2012-03-29T11:00:17 Initialized empty Git repository in /home/sh/tmp/git-rtems-tmp/.git/ 2012-03-29T11:00:18 fatal: '/data/git/rtems.git' does not appear to be a git repository 2012-03-29T11:00:20 fatal: The remote end hung up unexpectedly 2012-03-29T11:00:30 git fetch upstream 2012-03-29T11:00:30 fatal: '/data/git/rtems.git' does not appear to be a git repository 2012-03-29T11:00:40 guess i should have to reclone the copy? 2012-03-29T11:00:47 rather than try to resume my work 2012-03-29T11:00:49 DrJoel: telco done :-) 2012-03-29T11:01:08 hmm 2012-03-29T11:01:14 same problem as you sebhub. 2012-03-29T11:01:32 juli1, yes.. but in the process of hurrying to commit some patches, I did something horrible to the repository 2012-03-29T11:01:47 DrJoel: like what ? 2012-03-29T11:02:01 erverted two weeks of patches and now trtying to restore thigns 2012-03-29T11:06:05 permissions on rtems.git are overly permissive 2012-03-29T11:06:57 in fact permissions seem screwed 2012-03-29T11:08:48 I can clone the master and it looks to be restored 2012-03-29T11:08:52 I can clone my personal repo 2012-03-29T11:09:35 [joel@iceland rtems]$ git remote add upstream ssh://joel@git.rtems.org/data/git/rtems.git 2012-03-29T11:09:35 [joel@iceland rtems]$ git fetch upstream 2012-03-29T11:09:35 fatal: '/data/git/rtems.git' does not appear to be a git repository 2012-03-29T11:09:35 fatal: The remote end hung up unexpectedly 2012-03-29T11:10:30 go look ls -lt /data/git/rtems.git 2012-03-29T11:10:45 ok.. what's broken 2012-03-29T11:11:12 compare ls -l /data/git/rtems.git and ls -l ~/git/rtems.git 2012-03-29T11:11:28 the backup had some kind of wonky permission bits being set 2012-03-29T11:12:00 agreed.. they are not surviving the backup/restore :( 2012-03-29T11:13:09 from rtems.git ... find . -type d | xargs -e chmod 755 2012-03-29T11:13:13 then readd the -s 2012-03-29T11:13:56 find . -type d | xargs -e chmod g+s 2012-03-29T11:14:02 find . -type f | xargs -e chmod 644 2012-03-29T11:14:13 then go re-add the x bit to hooks 2012-03-29T11:14:36 chmod +x hooks/* 2012-03-29T11:14:44 visible again.. 2012-03-29T11:14:47 on web I mean 2012-03-29T11:15:23 cloning my repo again 2012-03-29T11:16:03 chmod 444 objects/* -r 2012-03-29T11:16:10 remove the write permission for the objects files 2012-03-29T11:16:20 well, for the files not directorys 2012-03-29T11:16:23 so you should use a find ;) 2012-03-29T11:16:31 find objects -type f | xargs -e chmod 444 2012-03-29T11:16:44 not sure whether that matters, but those were the permissions set before 2012-03-29T11:16:56 I think the backup script needs to tar up the contents of each git repo. :( 2012-03-29T11:17:41 that also would be good 2012-03-29T11:17:59 sebhub: clone worked now for me 2012-03-29T11:18:20 safer with these -S/-s set everywhere ... sigh.. it is going across a network with rsync and maybe something is getting lost 2012-03-29T11:19:24 gitk looks OK on personal.. I am now 3 commits ahead of master 2012-03-29T11:19:30 *** weiY has quit IRC (Ping timeout: 252 seconds) 2012-03-29T11:19:41 error: insufficient permission for adding an object to repository database ./objects 2012-03-29T11:19:41 fatal: failed to write object 2012-03-29T11:19:41 error: unpack failed: unpack-objects abnormal exit 2012-03-29T11:19:41 To ssh://joel@git.rtems.org/data/git/rtems.git 2012-03-29T11:19:41 ! [remote rejected] master -> master (n/a (unpacker error)) 2012-03-29T11:19:41 error: failed to push some refs to 'ssh://joel@git.rtems.org/data/git/rtems.git' 2012-03-29T11:19:51 so objects directory is still not correct permissions 2012-03-29T11:21:05 ok, the clone looks good 2012-03-29T11:21:31 got it.. missed a couple of permissions 2012-03-29T11:21:42 chmod -R g+ws * 2012-03-29T11:21:45 chgrp -R gitrw 2012-03-29T11:22:10 ok, have to leave now, cu 2012-03-29T11:22:16 now the damn backup script needs to be rewritten :( 2012-03-29T11:22:36 gedare.. are you able to pull my commits of the three patches from Daniel 2012-03-29T11:23:03 the files themselves in objects/ should only need read-permissions i think 2012-03-29T11:23:13 $ git status 2012-03-29T11:23:13 # On branch master 2012-03-29T11:23:14 # Your branch and 'origin/master' have diverged, 2012-03-29T11:23:14 # and have 68 and 48 different commit(s) each, respectively. 2012-03-29T11:23:14 # 2012-03-29T11:23:14 nothing to commit (working directory clean) 2012-03-29T11:24:31 How do I sync so that is gone? 2012-03-29T11:25:10 ok.. only readable 2012-03-29T11:25:14 hmmm 2012-03-29T11:25:30 so..you pushed whatever badness to your personal repo 2012-03-29T11:25:44 offtopic: i think the problem began with these patches: Add missing BSD sections; bsp/gen5200:* 2012-03-29T11:26:24 cdcs, It was a bad git command that effectively reverted two weeks of patches.. :( 2012-03-29T11:26:56 the easiest thing to do will be to wipe out your personal repo and replace it with a new mirror 2012-03-29T11:27:16 not only. you also reabased everything since 09th of march 2012-03-29T11:27:29 ssh to the server and rm your ~/git/rtems.git 2012-03-29T11:27:31 grrr.. I have a couple of branches that have code on them.. should I just generate patches for them and recreate 2012-03-29T11:27:39 generate patches 2012-03-29T11:27:54 every commit since "Merge branch 'master' of ssh://git.rtems.org/data/git/rtems" from the 06 of march was wrong. 2012-03-29T11:27:56 use git format-patch and give the commit sha1 should work 2012-03-29T11:28:10 ok.. I can always move that to rtems-bad.git and do that :) 2012-03-29T11:28:19 true. 2012-03-29T11:28:24 probably a good idea to do that. 2012-03-29T11:29:00 good chance to merge those branches anyway .. if I don't screw it up again 2012-03-29T11:29:04 just... move to a private directory or change its perms so that it does not appear on the web interface (beyond this post-mortem) 2012-03-29T11:29:37 ok.. that just impacts me.. does it look like you all have working clones? 2012-03-29T11:29:57 i'm going to test drive now 2012-03-29T11:30:31 hmmmm 2012-03-29T11:30:48 *** sebhub has quit IRC (Ping timeout: 272 seconds) 2012-03-29T11:31:02 i just issued git fetch upstream and I think it pulled all the badness into my local git 2012-03-29T11:31:17 oh nvm 2012-03-29T11:31:26 its ok I was just confused about what the log should look like 2012-03-29T11:31:41 i do see the 3 commits from daniela 2012-03-29T11:31:42 daniel 2012-03-29T11:32:08 I have three branches to deal with: multi_port_serial_boards ucontext and virtex 2012-03-29T11:32:29 virtex should really be closed to being on master. but not quite there 2012-03-29T11:32:37 multi_port_serial_boards definitely should be ready 2012-03-29T11:32:47 any pointer on what's the procedure to merge a branch? 2012-03-29T11:32:57 everytime I rebase one of them, it gives 100s of conflicts 2012-03-29T11:33:52 ok i re-sent the commit that I made between the time the backup was pulled and right now. 2012-03-29T11:34:32 Jennifer may still have issues .. her personal repo was not updating to master. What's her magic command? 2012-03-29T11:36:20 git pull --rebase upstream master 2012-03-29T11:36:40 that updates your local repository with the current upstream 2012-03-29T11:36:42 git push 2012-03-29T11:36:49 will push your local repository to your personal repo 2012-03-29T11:37:00 assuming you have set things up like the wiki suggests for using personal repos 2012-03-29T11:37:19 We are supposed to have. I cut and paste the commands and change user name and repo 2012-03-29T11:37:28 you have a personal repo that contains publicly viewable code, a local repo, and the rtems upstream 2012-03-29T11:37:42 the local repo kind of acts like a go-between 2012-03-29T11:38:10 even though it is logically (in my head) at the lowest end.. it is really the middle 2012-03-29T11:38:16 right 2012-03-29T11:38:20 so... 2012-03-29T11:38:27 when you git push/pull you are working on your personal repo 2012-03-29T11:38:34 this is safest by default... 2012-03-29T11:38:46 and you have to specify upstream to manipulate the rtems repo 2012-03-29T11:39:01 I should be OK. Hopefully Jennifer was not impacted 2012-03-29T11:39:06 my guess... 2012-03-29T11:39:17 is that you did something like reset on your master branch and then pushed it upstream 2012-03-29T11:39:37 I broke a rule and was trying to do someting in a hurry.. never do something in a hurrry 2012-03-29T11:39:42 yeah 2012-03-29T11:39:47 and never start something on a Friday afternoon or before a holiday 2012-03-29T11:40:06 Server upgrades on Fridays are BAD BAD idea 2012-03-29T11:40:10 and don't talk about religion, politics, or money :) 2012-03-29T11:40:45 those are ok.. we all know that no one has enough money, politicians are lying when their lips move, and you can't change someone's mind on religion :) 2012-03-29T11:41:29 I hope the rest of the world understands that us lowly Americans really have a "best of the worst" decision for President .. lol 2012-03-29T11:41:36 brb.. going to help Jennifer 2012-03-29T11:41:39 kj 2012-03-29T11:42:47 drjoel: want to make a brief announcement on the mls about this repo issue? 2012-03-29T11:53:07 *** Jennifer has joined #rtems 2012-03-29T11:53:32 hey gedare.. joel and jennifer here .. at her desk and crisis .. to many damned crises today 2012-03-29T11:53:45 ]$ git status # On branch master # Your branch and 'origin/master' have diverged, # and have 82 and 1 different commit(s) each, respectively. # 2012-03-29T11:54:33 That was multiple lines.,, irc crunched them 2012-03-29T11:54:44 git pull --rebase upstream master was done immediatey before that 2012-03-29T12:03:12 hmmm 2012-03-29T12:03:38 you could try git push to update origin/master 2012-03-29T12:04:51 Jennifer: but i'm unsure -- if you pulled in the badness from upstream earlier today then you might still have that in your local repo 2012-03-29T12:05:31 she committed a patch onto her local master about 2 weeks ago and it is getting in the way of rebasing.. I am reverting it now to simplfify things 2012-03-29T12:06:36 use git reset HEAD^ 2012-03-29T12:06:45 that should undo the local commit 2012-03-29T12:07:10 assuming that commit is the first thing you see from git log 2012-03-29T12:09:26 I don't think the web view is tracking her repo right :( 2012-03-29T12:09:49 have you done git push recently? 2012-03-29T12:10:32 looks like the malta branch was updated 2 days ago that was the most recent thing to get pushed to personal repo 2012-03-29T12:11:18 Jennifer: It doesn't look like the personal repo has been updated from upstream.. side-effect of not viewing the local repo as a go-between 2012-03-29T12:11:19 Joel here.. Jennifer committed two bug fixes we misse merging the cvs code 2012-03-29T12:11:30 somehow this should be more clear in the wiki i guess 2012-03-29T12:11:55 but her clone is missing a change by sebastian on 3/24 to add a file to c/src/lib/libbsd/shared/src/irq 2012-03-29T12:11:57 http://git.rtems.org/jennifer/rtems.git/commit/?id=7e6640236981a1263fdb61e76ebfb15e41178936 2012-03-29T12:11:59 that? 2012-03-29T12:12:06 oh 2012-03-29T12:13:24 shuld we assume her personal repo is screwed up somehow and treat hers like mine? Malta has been announced. 2012-03-29T12:13:40 probably. her personal repo is badly outdated is really the problem 2012-03-29T12:13:46 you have to 'git push' in order to propagate local repo changes to personal repo 2012-03-29T12:14:34 i think the mental model was just a bit off 2012-03-29T12:15:08 I don't know what just changed in git's mind. We switched branch to malta, then back to master 2012-03-29T12:15:10 i'm about to have lunch. 2012-03-29T12:15:32 did a push .. then a tetch upstream master .. and now it looks ok 2012-03-29T12:15:53 thanks.. I don't know what went wront ..sigh 2012-03-29T12:15:56 ok 2012-03-29T12:16:51 partially I think it is a workflow issue, and just familiarity with tools still is lacking 2012-03-29T12:17:07 afk 2012-03-29T12:17:15 yep.. won't argue that.. 2012-03-29T12:21:47 * DrJoel is back at his keyboard and wants lunch 2012-03-29T13:10:31 heh. did you get some 2012-03-29T13:15:44 *** Hesham has joined #rtems 2012-03-29T13:22:51 finally.. then had to update slides for my 3rd telecon of day in about 90 minutes 2012-03-29T13:23:18 I am giving GSOC talk at UA Huntsville tonight .. 2012-03-29T13:23:32 did you see email for what's the http address to clone for rtems git? 2012-03-29T13:28:34 yeah but i never tried it so i am no help 2012-03-29T13:32:03 i got it. 2012-03-29T13:32:13 git clone http://git.rtems.org/rtems.git rtems works. 2012-03-29T13:32:41 sry. git clone http://git.rtems.org/rtems rtems works! :P 2012-03-29T13:33:02 he should have caught the moment that you where updating the repo. 2012-03-29T13:33:23 i'll respond to his email 2012-03-29T13:33:32 suggest he might try again the git clone due to earlier issues today 2012-03-29T13:33:46 i was doing that. 2012-03-29T13:33:52 :P 2012-03-29T13:34:01 ok then you do it :) 2012-03-29T13:34:09 ok :) 2012-03-29T13:38:30 gedatre: should i try to propose another project? , testing for example , or just keep on my current one 2012-03-29T13:40:18 hesham: keep on current one 2012-03-29T13:40:42 hesham: no news is good news for right now. 2012-03-29T13:40:51 that's good 2012-03-29T13:41:04 when i was exploring library of rtems 2012-03-29T13:41:15 i found mmu file for arm920 2012-03-29T13:41:44 i thought there were no mmu implementations for arm yet 2012-03-29T13:43:08 hesham: I'm unaware. maybe some low-level code exists. 2012-03-29T13:43:17 who knows if it works or not... 2012-03-29T13:43:37 did you solve your gumstix problem? 2012-03-29T13:44:03 no , i sent the problem to rtems.users as you told me 2012-03-29T13:44:30 aslo i had tried other BSPs to work on Qemu 2012-03-29T13:44:31 did you? I did not see that email. 2012-03-29T13:44:42 armcortex 2012-03-29T13:44:45 check to make sure it did not bounce 2012-03-29T13:44:51 i sent it an hour ago i think 2012-03-29T13:45:39 make sure you are signed up for that list, only list members can send mail 2012-03-29T13:46:07 oh i think i miss it 2012-03-29T13:47:03 looks like the arm920 mmu code is a little old but that should help you get going even faster 2012-03-29T13:47:11 make a reference to it in your proposal. 2012-03-29T13:48:09 looks like it is mainly for virt->phy translation and cache attributes. 2012-03-29T13:48:27 you will have to see how much of it will be useful and also how to add protection bits. 2012-03-29T13:48:39 i did only see that file but actually i did not explore it 2012-03-29T13:49:02 so i will have a look on it now 2012-03-29T13:49:07 see how can it helps 2012-03-29T13:49:46 and summarize it 2012-03-29T13:50:20 good. 2012-03-29T13:51:00 about the issue of simulator ( QEMU ) i tried to build many BSPs and the ones Joel told me about 2012-03-29T13:51:24 but it was the same problem 2012-03-29T13:52:07 except that some BSPs just gave me a blank screen with no errors 2012-03-29T13:52:15 report the steps you took (bootstrap, configure, make) and whether there were warnings/errors at any steps 2012-03-29T13:52:51 (warnings during make are ok to ignore though) 2012-03-29T13:53:12 we had make warnings down to 0 but that was awhile ago 2012-03-29T13:53:26 please give me the link to subscribe to rtems-users list 2012-03-29T13:54:54 http://www.rtems.org/mailman/listinfo/rtems-users 2012-03-29T13:55:58 thanks 2012-03-29T13:58:52 *** cdcs has quit IRC (Quit: Page closed) 2012-03-29T14:25:57 hm. i had to go to my personal repo and manually go back in time so that I dont' have a weird history due to resetting the repo to its backup 2012-03-29T14:36:44 *** Hesham has quit IRC (Ping timeout: 246 seconds) 2012-03-29T14:43:59 FWIW git clone of http isn't in the wiki .. or I didn't see it 2012-03-29T14:45:32 gedare, sebastian has patch for qemu in the lm3s69xx so he must be running that on there 2012-03-29T14:46:20 yeah i know 2012-03-29T14:46:32 i pointed students to it before ;) 2012-03-29T14:49:36 poor drjoel..git'in sad 2012-03-29T14:52:20 lol.. well.. it is an excuse to learn and git rid of some work branches that should have been merged already 2012-03-29T14:52:35 A0Sheds, did you send me anything for news for the website? 2012-03-29T14:53:46 DrJoel: we didn't get picked up for GSOC but the project is still moving ahead 2012-03-29T14:54:47 there may not be time to do anything about it. But we would have been happy to have had a "co project" which would have benefitted both 2012-03-29T14:54:48 http://rhombus-tech.net./gsoc2012/ideas/openec32/ 2012-03-29T14:54:53 We have done this in the past 2012-03-29T14:55:44 that is the jist of it 2012-03-29T14:55:48 Add that to our wiki open projects list with a page for it. 2012-03-29T14:55:55 did you have any students interested? 2012-03-29T14:56:26 well about 20 companies in China are :) ..... 2012-03-29T14:56:37 we have been so busy 2012-03-29T14:57:17 we did have some interested students, but we actually have large volume OEM's lining up for the solution 2012-03-29T14:58:19 this will be a long term project that will grow to expand into several applications and ARM soc's 2012-03-29T14:58:30 if you had a good student you can send them our way and sign up to co-mentor if you're interested. 2012-03-29T14:58:42 let me check 2012-03-29T15:01:23 we will let you know by Monday or Tuesday 2012-03-29T15:04:06 A0Sheds: OK. Just direct the student to submit the proposal and to do our RTEMS - GSOC getting started tasks if you do since the timeline will be short 2012-03-29T15:04:42 ok 2012-03-29T15:04:42 carving out an appropriately sized project will be the important part i think. 2012-03-29T15:05:09 well RTEMS already works on STM32 cortex M3 2012-03-29T15:05:32 it's just adding peripherlas/register support 2012-03-29T15:05:36 *** xiaochen has joined #rtems 2012-03-29T15:05:38 ok. i'm not familiar with the projects just making general statements :) 2012-03-29T15:05:41 so this could be rather quick 2012-03-29T15:05:45 ok 2012-03-29T15:06:10 yeah for someone with enough knowledge on the hw side, plus with a comentor with rtems bsp knowhow i could see this as a good project 2012-03-29T15:06:14 it's basically just getting RTEMS working with ARM cortex M3 by either ST or NXP 2012-03-29T15:06:35 the rest is porting libraries and routines form other projects 2012-03-29T15:06:44 ok cool 2012-03-29T15:07:04 *** xiaochen has quit IRC (Client Quit) 2012-03-29T15:07:08 there are $10 ST stm32 boards 2012-03-29T15:07:28 Mouser, digikey, farnell etc etc 2012-03-29T15:07:44 I put out the call 2012-03-29T15:08:03 thanks for the gracious offer 2012-03-29T15:09:08 alright. its a benefit to both our communities 2012-03-29T15:09:19 the arm soc's running RTEMS will replace the proprietary firmware used in netbook, laptops and tablets 2012-03-29T15:09:50 they are used for ACPI, battery charging, lid open closed, audio playback, GPIO etc etc 2012-03-29T15:10:47 I know and like to cross-pollinate projects when it makes sense 2012-03-29T15:10:50 BIOS cos' like Award. Phoenix and AMI are trying to convince customers that their ARM soc needs a BIOS and the EC's need to be using their closed source fimware as well 2012-03-29T15:12:09 either way this will get done by the summer 2012-03-29T15:12:22 notebook vendors can't wait 2012-03-29T15:12:57 we are trying to show these off at Computex Taipei in June 2012-03-29T15:13:48 there might suddenly be 20+ more RTEMS devs in China by fall 2012-03-29T15:15:03 most of the devs are currently working on Linux kernel support for the allwinner a10 2012-03-29T15:15:35 once that smooths out, many will move to support the ARM stm32 EC's 2012-03-29T15:15:55 it should be a busy summer 2012-03-29T15:16:42 exciting stuff 2012-03-29T15:17:44 :) Which distro is looking popular on these 2012-03-29T15:19:12 a few 2012-03-29T15:19:55 somebody has ported Ret Hat enterprise to it, Red Sleeves Linux ;p 2012-03-29T15:20:04 also debian is popular 2012-03-29T15:21:26 Gentoo even has hard float 2012-03-29T15:23:25 I'm trying to stick with hardware engineering and leave the rest to everyone interested :) 2012-03-29T15:24:03 my other job is being a polymer Chemist http://code.google.com/p/lemoncurry/wiki/main 2012-03-29T15:35:11 nice 2012-03-29T15:35:43 So these are just regular netbooks with arm support chips? now arm netbooks 2012-03-29T15:54:35 DrJoel: yes ARM powered with smaller ARM devices to handle the power management 2012-03-29T15:54:58 also x86 powered with smaller ARM devices to handle the power management 2012-03-29T15:55:13 it's the small ARM soc that runs RTEMS 2012-03-29T15:56:04 cool! so both arm and x86 as the main cpu. :) 2012-03-29T15:58:03 yes, with ARM always as the EC (embedded controller) 2012-03-29T16:04:35 DrJoel, hi 2012-03-29T16:05:37 I have some concerns about PR 2038. I do not understand how this change can just be made with out specific testing. 2012-03-29T16:06:19 The PR contains the comment "Newlib is suppose to be OK". I am not sure this is good enough 2012-03-29T16:07:07 If newlib and gcc are the same then why this sudden change ? 2012-03-29T16:08:05 my understanding is that we were using either our own or newlib's stdint.h which was in newlib. Now we switch to gcc's 2012-03-29T16:08:19 but that is just an impression 2012-03-29T16:08:24 If they are the same why change ? 2012-03-29T16:13:36 I think we had our "not gcc" version to be ahead (again not feeding upstream right). they were not in sync and gcc's apparently was right, ours was wrong 2012-03-29T16:17:48 that still makes no sense, because newlibs stdint was supplaned by gccs 2012-03-29T16:18:05 without any testing or regard the report simply states "it should be ok" 2012-03-29T16:18:28 newlib has gone to great lengths to avoid using gccs stdint i'm highly skeptical it can be simply dropped in 2012-03-29T16:19:04 if anything a massive amount of testing should be done -before- this change happens, a simple statement that testing will be done at some point in the future and "it should work" is not good enough 2012-03-29T16:19:19 <-no fan of retroactive testing 2012-03-29T16:20:15 Also there is no staging area for tools. It is just bang here you go and tell me if it works. 2012-03-29T16:21:01 Another issue is why suddenly with some gcc tests running we move ? Seems like we need full gcc testing before a repo push 2012-03-29T16:21:09 s/supplaned/supplanted/ 2012-03-29T16:50:33 I have updated the pr .... http://www.rtems.org/pr2038 2012-03-29T16:55:56 *** DrJoel has quit IRC (Quit: Leaving) 2012-03-29T17:20:55 heh.. took out the stick this time, no carrot 2012-03-29T17:21:26 this is just another issue of our decrepit state of tools 2012-03-29T17:27:29 Yeah stick or maybe just frustrated at the adhoc manner the tools get treated. Comparing this change to ARM EABI change leave me wondering what that was all about. 2012-03-29T17:33:55 did you catch what happened earlier? Joel borked rtems.git; it was a fun time 2012-03-29T17:34:23 I did see but I am still in the dark on what happened. 2012-03-29T17:34:48 truly we all are... somehow Joel pushed something into rtems.git that caused it to drop a few weeks worth of commits 2012-03-29T17:35:01 my guess is that he somehow rebased upstream from his personal repo 2012-03-29T17:35:01 Impressive. 2012-03-29T17:35:09 Ah ok. 2012-03-29T17:35:24 but he doesn't know and didn't have his command history available 2012-03-29T17:35:44 so he restored a backup from yesterday evening, which then had to be manually fixed for permissions but otherwise seems "ok" 2012-03-29T17:35:46 It is distributed so it is ok. 2012-03-29T17:35:58 We all have a copy of the changes 2012-03-29T17:36:09 yeah the big issue is that my personal repo became out of sync as a result of restoring the backup 2012-03-29T17:36:38 because i had committed some things upstream after the backup 2012-03-29T17:36:43 That is what I feared. Restore of a backup is not a great solution 2012-03-29T17:37:32 It is if the machine fails. 2012-03-29T17:37:39 i wound up having to fix up my personal repo but otherwise it's business as usual 2012-03-29T17:38:10 his and jennifer's personal repos also got broken, just not sufficient git training yet 2012-03-29T17:38:20 it happens 2012-03-29T17:38:22 yeah 2012-03-29T17:38:27 especially when moving from a centralised RCS to DCVS 2012-03-29T17:38:31 yeah and as a community we help out 2012-03-29T17:38:36 we got it resolved 2012-03-29T17:38:46 we're small enough that it's actually easier to just jump in 2012-03-29T17:38:51 yeah and all before I got up hehe 2012-03-29T17:38:53 rather than doing long and bloated planning 2012-03-29T17:38:57 but restoring from backup is not ideal, it would be nice to have a smoother solution, but hopefully this is very rare 2012-03-29T17:39:07 yep that is true 2012-03-29T17:39:35 You would need a snapshot type solution. 2012-03-29T17:40:08 gedare: joel could have used git reset locally 2012-03-29T17:40:11 then pushed that 2012-03-29T17:40:14 to undo whatever he did 2012-03-29T17:40:44 http://stackoverflow.com/questions/1270514/undoing-a-git-push <-found this in a cursory search 2012-03-29T17:41:07 git reflog is your friend, you can undo anything sometimes going back months as long as you don't do a 'git gc' very often 2012-03-29T17:41:22 it'll also give an indication as to what he did 2012-03-29T17:41:37 Addition to the committer's wiki page ? 2012-03-29T17:41:54 we should definitely have something in there..not sure how to phrase it though 2012-03-29T17:42:21 If you suff up, "stop" "raise the issue", then look at this. 2012-03-29T17:42:40 yeah 2012-03-29T17:43:05 this force push is a nicer solution than what we did 2012-03-29T17:43:07 It is human to "err" 2012-03-29T17:43:28 Yeah we need to not have a restore from back up again 2012-03-29T17:43:31 or even what i ended up doing on my personal repo, which was to ssh to the git box and use git update-ref or something similar 2012-03-29T17:49:19 well I added a few words at least, and a link to the stack overflow article 2012-03-29T17:49:46 http://wiki.rtems.org/wiki/index.php/Git_Committers#Ooops.21 2012-03-29T18:04:45 Just for me with git. To avoid the 'merge branch' messages do you rebase and not merge a branch ? 2012-03-29T18:11:01 ya 2012-03-29T18:11:22 i shall just change these instructions to simplify them 2012-03-29T18:11:33 i always do 'git pull --rebase upstream master' 2012-03-29T18:15:05 kiwichris: as long as you are working with code you haven't shared with someone else it's just fine to rebase 2012-03-29T18:15:47 and to commit code back to rtems probably you'll end up rebasing anyway to make the history look nicer... but your collaborators would need to be aware of that since they might have branched from your work already 2012-03-29T18:16:28 for long-running publicly visible work it may be necessary to do a merge...but you can also amend the merge commit message to make it a little nicer to read 2012-03-29T18:16:39 or use --no-commit 2012-03-29T18:16:48 If I do "git co -b foo" then 'git co master' do I 'git merge foo' or 'git rebase foo' (or something like that) ? 2012-03-29T18:16:52 then commit it manually which is what i do if i want to change it 2012-03-29T18:17:24 you want to merge foo onto your local master? 2012-03-29T18:17:37 Yes 2012-03-29T18:17:46 Ah ok I could push from the branch. 2012-03-29T18:17:53 I see 2012-03-29T18:17:59 yeah i always push the branch directly to upstream master 2012-03-29T18:18:05 and then pull from upstream to verify ;) 2012-03-29T18:18:11 verm__, what does --no-commit do ? 2012-03-29T18:18:22 gedare, pull on master ? 2012-03-29T18:18:31 yeah 2012-03-29T18:18:35 here's my usual workflow 2012-03-29T18:18:39 git co -b am 2012-03-29T18:18:44 git am ../patches/am/* 2012-03-29T18:18:53 git pull --rebase upstream master 2012-03-29T18:18:58 git push upstream am:master 2012-03-29T18:19:01 git co master 2012-03-29T18:19:06 git pull upstream master 2012-03-29T18:19:07 git log 2012-03-29T18:19:13 git branch -d am 2012-03-29T18:19:20 for applying a bunch of patches for example 2012-03-29T18:19:24 Nice. 2012-03-29T18:20:20 ooo, git help has been updated with a pic about bootstrap 2012-03-29T18:20:28 i think seb did that 2012-03-29T18:20:34 i just added one in git committers too 2012-03-29T18:20:47 Needs a big huge warning about branches and configure output 2012-03-29T18:20:53 lol yes 2012-03-29T18:21:15 Like "Warning: git and autotools generate output do not work well together" 2012-03-29T18:21:44 that's easy enough... just say the red circles do not track properly in git ;) 2012-03-29T18:22:25 Yeah and need to be regenerated via a bootstrap because autotools cannot cleanly track them 2012-03-29T18:23:27 oooo pics in committers page as well, very up market 2012-03-29T18:24:55 gedare, your little "apply a patch" sequence should be added to the committers page as a simple recipe on committing a patch 2012-03-29T18:25:17 Maybe with a little bit about the patches in the patch dir 2012-03-29T18:38:41 will do 2012-03-29T18:38:41 kiwichris: those "merge XXX" messages is an automessage when you merge from a branch 2012-03-29T18:38:50 really bad idea on the part of the git designers 2012-03-29T18:38:59 i also 'git push' at the end to make my personal repo match up 2012-03-29T18:39:02 you have two ways to change it, merge then use ammend to change the message 2012-03-29T18:39:10 --no-commit does the merge but skips the autocommit 2012-03-29T18:39:12 verm__, I agree on that 2012-03-29T18:39:23 so you can then use 'git commit' with your own message 2012-03-29T18:39:41 verm__, nice. I like that as well. 2012-03-29T18:39:47 kiwichris: they're stuck with it now but i belive it's ranked as one of the top (if not _the_ top) worst choice in git 2012-03-29T18:40:00 most dcvs default to fast-forward 2012-03-29T18:40:23 there's talk on the git ml to default push to --rebase 2012-03-29T18:40:30 i have merge aliased with --no-comment 2012-03-29T18:40:42 gedare: yeah that's auto-ff but nobody is sure that it can ever happen 2012-03-29T18:40:44 That is a good idea. 2012-03-29T18:40:50 too many are used to the default to merge 2012-03-29T18:40:54 i agree. 2012-03-29T18:41:16 linus has fended off talks about this for years heh 2012-03-29T18:46:14 the main wiki page is so crowded 2012-03-29T18:46:40 it'd be nice to put some useful links on the left-hand side too. 2012-03-29T18:46:47 most of that space is wasted 2012-03-29T18:46:47 hm 2012-03-29T18:46:51 Agreed 2012-03-29T18:49:20 kiwichris did you request to be a mentor for rtems 2012-03-29T18:49:35 i'm not sure if the OAs automatically are or not 2012-03-29T18:49:37 No, I was just made a mentor 2012-03-29T18:49:47 oh isee 2012-03-29T18:49:50 Joel asked and I checked and I was 2012-03-29T18:49:54 oh ok 2012-03-29T18:50:16 ahh i see it on a different table 2012-03-29T19:46:41 https://www.rtems.org/bugzilla/show_bug.cgi?id=2043 2012-03-29T19:46:43 nice 2012-03-29T20:04:06 *** xiangfu has joined #rtems 2012-03-29T20:38:00 *** panzon has quit IRC (Ping timeout: 260 seconds) 2012-03-29T22:35:18 *** panzon has joined #rtems 2012-03-29T22:39:15 *** gedare has quit IRC (Quit: Leaving) 2012-03-29T22:49:50 *** panzon has quit IRC (Ping timeout: 260 seconds) 2012-03-29T23:21:20 *** arvind_khadri has quit IRC (Ping timeout: 260 seconds) 2012-03-30T00:05:50 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-30T00:47:37 *** arvind_khadri has joined #rtems 2012-03-30T01:23:37 *** antgreen has quit IRC (*.net *.split) 2012-03-30T02:36:19 *** xiangfu has joined #rtems 2012-03-30T02:57:35 *** sebhub has joined #rtems 2012-03-30T02:57:35 *** antgreen has joined #rtems 2012-03-30T06:06:21 *** panzon has joined #rtems 2012-03-30T06:07:39 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2012-03-30T06:20:15 *** arvind_khadri has joined #rtems 2012-03-30T06:37:14 *** Hesham has joined #rtems 2012-03-30T06:40:17 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-30T07:47:08 *** sevikkk has quit IRC (Ping timeout: 245 seconds) 2012-03-30T07:51:40 *** sevikkk has joined #rtems 2012-03-30T08:29:29 *** antgreen has quit IRC (Read error: Connection reset by peer) 2012-03-30T08:41:58 *** weiY has joined #rtems 2012-03-30T08:43:51 *** xiangfu has joined #rtems 2012-03-30T08:56:00 *** xiangfu has quit IRC (Quit: Leaving) 2012-03-30T09:16:52 *** Hesham has quit IRC (Ping timeout: 276 seconds) 2012-03-30T09:52:54 *** gedare has joined #rtems 2012-03-30T10:39:08 *** sebhub has quit IRC (Remote host closed the connection) 2012-03-30T10:55:30 *** arvind_khadri has quit IRC (Ping timeout: 260 seconds) 2012-03-30T11:00:28 *** weiY has quit IRC (Ping timeout: 245 seconds) 2012-03-30T11:20:42 *** gedare has quit IRC (Quit: Ex-Chat) 2012-03-30T12:37:15 *** Hesham has joined #rtems 2012-03-30T12:52:45 *** xiaochen has joined #rtems 2012-03-30T12:53:32 *** arvind_khadri has joined #rtems 2012-03-30T13:15:51 *** xiaochen has quit IRC (Ping timeout: 245 seconds) 2012-03-30T13:59:18 *** gedare has joined #rtems 2012-03-30T14:09:11 *** Hesham has quit IRC (Quit: Leaving.) 2012-03-30T14:25:47 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2012-03-30T15:55:10 gedare, did Ralf really respond to you doc configure but with that ? 2012-03-30T15:55:20 kiwichris: the proof is in the pr. 2012-03-30T15:55:22 grr but -> bug 2012-03-30T15:55:39 That sort of response is not on 2012-03-30T15:56:20 Feed up 2012-03-30T15:56:31 I have responded 2012-03-30T15:56:33 i almost flipped the bug to wontfix. i should have gone and tried to build the doc using a different build system. 2012-03-30T15:57:46 No response on the tools PR questions I raised. 2012-03-30T15:58:13 yep...but i'm pretty sure the change has been propagated by now since i got an rpm update for tools today. 2012-03-30T15:58:47 It has. RTEMS's stdint.h is decided by an 'rm'. Joy 2012-03-30T15:59:07 Hmmm I did not get that update ..... oh that right .... I build from source. 2012-03-30T16:03:10 heh. yeah.. and even if you wanted that update it would be a pain to get 2012-03-30T16:27:49 *** gedare has quit IRC (Quit: Leaving) 2012-03-30T17:21:00 *** xiaochen has joined #rtems 2012-03-30T18:39:34 *** xiaochen_ has joined #rtems 2012-03-30T19:11:02 *** xiaochen_ has quit IRC (Quit: Page closed) 2012-03-30T20:32:00 *** xiangfu has joined #rtems 2012-03-30T21:00:36 *** A0Sheds has quit IRC (Quit: puff of smoke) 2012-03-30T21:02:21 *** A1Sheds has joined #rtems 2012-03-30T21:02:22 *** A1Sheds has joined #rtems 2012-03-30T21:40:56 *** somebody_ has joined #rtems 2012-03-30T21:43:23 *** xiaochen has quit IRC (Quit: Page closed) 2012-03-30T22:10:39 *** xiaochen has joined #rtems 2012-03-30T22:21:26 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-30T22:22:03 *** xiangfu has joined #rtems 2012-03-30T22:36:07 *** xiaochen has quit IRC (Quit: Page closed) 2012-03-30T22:48:32 *** somebody_ has quit IRC (Quit: Leaving) 2012-03-30T22:48:47 *** deb has joined #rtems 2012-03-30T22:49:05 *** debby has joined #rtems 2012-03-30T23:24:11 *** debby has quit IRC (Quit: Leaving) 2012-03-30T23:24:14 *** deb has quit IRC (Quit: Leaving) 2012-03-30T23:29:25 *** panzon_ has joined #rtems 2012-03-30T23:31:45 *** panzon has quit IRC (Ping timeout: 252 seconds) 2012-03-30T23:37:55 *** Hesham has joined #rtems 2012-03-30T23:40:38 *** Hesham has left #rtems 2012-03-31T01:18:53 *** arvind_khadri has joined #rtems 2012-03-31T03:56:45 *** xiangfu has quit IRC (Quit: Leaving) 2012-03-31T06:03:30 *** Dhruva92 has joined #rtems 2012-03-31T06:03:50 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2012-03-31T06:04:10 *** Dhruva92 has left #rtems 2012-03-31T06:20:39 *** arvind_khadri has joined #rtems 2012-03-31T07:46:39 *** weiY has joined #rtems 2012-03-31T07:51:19 *** gedare has joined #rtems 2012-03-31T07:52:12 good morning 2012-03-31T08:22:19 *** debby has joined #rtems 2012-03-31T08:22:46 *** Deb has joined #rtems 2012-03-31T08:40:31 good morning gedare. me night 2012-03-31T08:40:45 heh. yep 2012-03-31T08:41:53 the git repository is ok now. it seems that yestoday there is a error on git 2012-03-31T08:41:58 ? 2012-03-31T08:49:47 yes there was a problem 2012-03-31T08:50:05 in fact anyone who updated from it may have a slightly inconsistent repo 2012-03-31T09:02:49 yeah, now every developer git clone the branch from server and git commit , then git push? or every one setup own branch based on server and commit ,then merge? 2012-03-31T09:05:24 developers with commit access push to the primary rtems.git 2012-03-31T09:11:09 every developers with commit access push to the same branch? 2012-03-31T09:11:19 on the rtems.git 2012-03-31T09:11:49 yes 2012-03-31T09:12:07 the rtems.git master branch is the primary rtems 2012-03-31T09:12:19 and there are some refs for past releases on rtems.git 2012-03-31T09:12:40 all of the developer repos are just experimental work / staging areas 2012-03-31T09:14:18 ok, but there are some ugly log if one developer git pull the master branch and develop to commit on the local branch, meantime other developers also commit and git push to the rtems.git, then this developer git push to rtems.git. this will lead some ugly merge log 2012-03-31T09:15:42 we suggest/impose using rebase before pushing 2012-03-31T09:16:12 this way the master rtems.git has a linear history 2012-03-31T09:16:45 at the cost of making developer repo's sometimes inconsistent (but it should only be inconsistent on branches) 2012-03-31T09:16:53 you mean git fetch and git rebase and git commit ? 2012-03-31T09:18:34 yeah, that mean the developer local branches is inconsistent 2012-03-31T09:20:11 correct. 2012-03-31T09:21:54 are there any formal materials of the processes of using git on the rtems FOR developers 2012-03-31T09:23:23 http://wiki.rtems.org/wiki/index.php/Git_Committers 2012-03-31T09:23:47 this is how committers are supposed to push their work upstream... suggested 2012-03-31T09:24:05 we're still working on this process but it seems to work decently well 2012-03-31T09:28:45 ok, thank you gedare, i think as more developers work for rtems this process will be more better. i will also get familiar with the git develop process for the coming GSOC 2012-03-31T09:36:26 yes we will need to develop procedures for developers w/o commit access to contribute patches and make pull requests 2012-03-31T09:36:51 right now the way to go is to submit patch by email (git send-email) or to file a bug report on bugzilla 2012-03-31T09:37:25 ok, i think the git send-email is very suitable 2012-03-31T09:39:11 and if i generate a patch and send it to email, then it is ok after review. and who is resposble for push to branches on the rtems.git? joel? 2012-03-31T09:39:40 or push by myself 2012-03-31T09:47:03 someone with commit access. See MAINTAINERS file in rtems/ 2012-03-31T09:48:10 by the way..feel free to write up some of these notes on the wiki somewhere if you think the answers will help others understand... 2012-03-31T09:49:16 you're only new once, so your perspective is important since you understand how new users/developers might think :) 2012-03-31T09:50:55 yeah, along the processes of study i will add my confues and some notes to wiki. 2012-03-31T09:52:58 i have to go to bed, good nigth. see you olo 2012-03-31T09:54:42 later weiY. 2012-03-31T10:09:04 *** leyyin has joined #rtems 2012-03-31T10:09:43 *** arvind_khadri has quit IRC (Ping timeout: 264 seconds) 2012-03-31T10:11:01 *** panzon__ has joined #rtems 2012-03-31T10:13:55 *** panzon_ has quit IRC (Ping timeout: 264 seconds) 2012-03-31T10:16:37 *** weiY has quit IRC (Ping timeout: 248 seconds) 2012-03-31T10:26:00 *** Deb has left #rtems ("Leaving") 2012-03-31T10:26:00 *** debby has quit IRC (Quit: Leaving) 2012-03-31T10:55:35 *** arvind_khadri has joined #rtems 2012-03-31T11:12:12 *** arvind_khadri has quit IRC (Ping timeout: 272 seconds) 2012-03-31T11:42:40 *** deb_ has joined #rtems 2012-03-31T11:42:41 *** Deb has joined #rtems 2012-03-31T11:43:04 *** Deb has quit IRC (Client Quit) 2012-03-31T12:21:17 *** leyyin has quit IRC (Remote host closed the connection) 2012-03-31T12:24:50 *** deb_ has left #rtems ("Leaving") 2012-03-31T13:07:10 *** arvind_khadri has joined #rtems 2012-03-31T13:46:49 *** arvind_k has joined #rtems 2012-03-31T13:48:57 *** arvind_khadri has quit IRC (Ping timeout: 250 seconds) 2012-03-31T13:53:15 *** Deb has joined #rtems 2012-03-31T13:53:30 *** panx has joined #rtems 2012-03-31T16:09:02 *** gedare has quit IRC (Quit: Leaving) 2012-03-31T17:09:08 *** Deb has quit IRC (Quit: This computer has gone to sleep) 2012-03-31T17:09:08 *** panx has quit IRC (Quit: This computer has gone to sleep) 2012-03-31T18:05:09 *** panx has joined #rtems 2012-03-31T18:05:09 *** Deb has joined #rtems 2012-03-31T18:22:56 *** deb_ has joined #rtems 2012-03-31T18:23:03 *** panx has quit IRC (Read error: Connection reset by peer) 2012-03-31T18:23:04 *** Deb has quit IRC (Write error: Connection reset by peer) 2012-03-31T18:52:11 *** deb_ has quit IRC (Quit: This computer has gone to sleep) 2012-03-31T18:55:26 *** Deb has joined #rtems 2012-03-31T19:45:37 *** xiangfu has joined #rtems 2012-03-31T19:50:50 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-03-31T20:02:36 *** xiangfu has joined #rtems 2012-03-31T20:08:57 *** xiangfu has quit IRC (Ping timeout: 250 seconds) 2012-03-31T20:09:20 *** xiangfu has joined #rtems 2012-03-31T22:23:12 *** Deb has left #rtems ("Leaving") 2012-03-31T22:59:48 *** Deb has joined #rtems 2012-04-01T00:47:46 *** Deb has left #rtems ("Leaving") 2012-04-01T00:48:26 *** Deb has joined #rtems 2012-04-01T00:48:59 *** Deb has quit IRC (Client Quit) 2012-04-01T00:57:08 *** arvind_k has quit IRC (Ping timeout: 245 seconds) 2012-04-01T01:13:23 *** arvind_k has joined #rtems 2012-04-01T01:51:07 *** arvind_k has quit IRC (Ping timeout: 264 seconds) 2012-04-01T02:03:39 *** arvind_k has joined #rtems 2012-04-01T04:29:14 *** panzon__ has quit IRC (Ping timeout: 244 seconds) 2012-04-01T04:41:35 *** panzon__ has joined #rtems 2012-04-01T05:10:08 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-04-01T05:24:56 *** xiangfu has joined #rtems 2012-04-01T05:27:29 *** Hesham has joined #rtems 2012-04-01T05:27:41 *** Hesham has left #rtems 2012-04-01T06:30:00 *** arvind_k has quit IRC (Ping timeout: 260 seconds) 2012-04-01T06:30:46 *** panzon__ has quit IRC (Ping timeout: 244 seconds) 2012-04-01T06:44:35 *** panzon__ has joined #rtems 2012-04-01T07:19:43 *** panzon__ has quit IRC (Ping timeout: 265 seconds) 2012-04-01T07:29:09 *** arvind_k has joined #rtems 2012-04-01T08:03:02 *** kristianpaul has quit IRC (Ping timeout: 272 seconds) 2012-04-01T08:03:38 *** kristianpaul has joined #rtems 2012-04-01T08:18:51 *** panzon has joined #rtems 2012-04-01T08:23:16 *** Deb has joined #rtems 2012-04-01T08:31:43 *** panzon has quit IRC (Ping timeout: 245 seconds) 2012-04-01T08:33:28 Hi Amar 2012-04-01T09:00:53 *** panzon has joined #rtems 2012-04-01T09:03:35 *** xiangfu has quit IRC (Quit: Leaving) 2012-04-01T09:17:02 *** Deb has left #rtems ("Leaving") 2012-04-01T10:11:45 *** Hesham has joined #rtems 2012-04-01T10:16:21 *** joe9 has joined #rtems 2012-04-01T10:38:15 *** Deb has joined #rtems 2012-04-01T10:54:34 *** gedare has joined #rtems 2012-04-01T11:01:52 Deb: if you want verm__'s attention you have to use his name or open a private dialog 2012-04-01T11:02:07 Deb: are you panx/xiaochen? 2012-04-01T11:04:48 *** Hesham has quit IRC (Remote host closed the connection) 2012-04-01T11:08:10 *** arvind_k has quit IRC (Ping timeout: 276 seconds) 2012-04-01T11:42:27 Yeah, I mistakenly typed in here 2012-04-01T11:42:53 gedare: I emailed you yesterday about the testing part 2012-04-01T11:42:54 ah ok did you get answers you needed? 2012-04-01T11:43:02 yes i just got the email 2012-04-01T11:43:37 dont' get thrown into too much panic by all the tests there 2012-04-01T11:43:54 this is one of the reasons we need better testing infrastructure...current approach is a mess 2012-04-01T11:44:19 I have trouble writing the proposal 2012-04-01T11:44:49 And verm told me that I should talk to you for suggestions on which testing part to start with 2012-04-01T11:45:27 the sptests are the main testing we use for "integration" testing of rtems 2012-04-01T11:46:12 Here is what I am confused about 2012-04-01T11:46:15 sptests stands for single-processor tests, but really those are the main tests used to try to cover all of the code in cpukit/score. 2012-04-01T11:46:47 Testing seems to be a job that tests the integration of different parts and see if they can work together correctly 2012-04-01T11:47:16 But for the testing projects, it wants to create four seprate tests groups 2012-04-01T11:47:40 correct.. 2012-04-01T11:48:45 right now all of the tests we use would be classified as "Operational" using the terminology that verm__ adopts 2012-04-01T11:49:18 and they tend to be a mix of integration and unit tests 2012-04-01T11:49:34 OK 2012-04-01T11:49:47 verm__ envisions that we should have a better understanding of our own tests, and make it easier to write/maintain tests 2012-04-01T11:49:56 and to run them automatically and get the results 2012-04-01T11:49:56 OK 2012-04-01T11:50:06 in order to do that we need to have a more standardized approach to testing 2012-04-01T11:50:09 that is the motivation of the project 2012-04-01T11:50:13 OK 2012-04-01T11:50:21 That's the big goal of this project 2012-04-01T11:50:25 yes exactly. 2012-04-01T11:50:26 I understand that 2012-04-01T11:50:32 ok good 2012-04-01T11:50:42 so now the hard part is figuring out how to get there ;) 2012-04-01T11:50:49 But what I don't understand is the framework we are trying to build 2012-04-01T11:50:58 For the host-based tests 2012-04-01T11:51:41 host-based tests would run rtems code in isolation from the rest of rtems on the developer's desktop 2012-04-01T11:52:02 Yes, I understand this too 2012-04-01T11:52:12 this has been done before with limited success by DrJoel but it was a mess to get working right 2012-04-01T11:52:32 so you will need mentor guidance to develop the 'how' of that 2012-04-01T11:52:49 Yes. This is what I am confused about 2012-04-01T11:53:06 Good; it is ok to be confused at this point 2012-04-01T11:53:19 I don't have an idea what the framework would look like 2012-04-01T11:53:40 He suggests me to take a tour on the online webguide for googletests and googlemock 2012-04-01T11:53:47 I looked through it. 2012-04-01T11:54:04 So I tried to understand what I can do with these two tools 2012-04-01T11:54:21 to build the framework 2012-04-01T11:54:41 But I don't have any idea what is a framework 2012-04-01T11:54:47 What is a good framework 2012-04-01T11:54:54 ahh hmmmm 2012-04-01T11:55:28 one that requires minimal work to maintain once implemented, allows for easy writing of new tests, and reports test results efficiently ;) 2012-04-01T11:55:37 tbh I am not entirely certain myself 2012-04-01T11:55:55 but the framework we have right now is not a very good one 2012-04-01T11:56:32 Without a clear thought on the work I am trying to do. I doubt I can write a good proposal 2012-04-01T11:56:33 it's ok when you tie it in to other tools, like coverage ( http://wiki.rtems.org/wiki/index.php/RTEMS_Test_Coverage ) 2012-04-01T11:56:42 yes I understand your concerns 2012-04-01T11:57:21 first I think the googletest/googlemock stuff might be too complicated to worry about right now 2012-04-01T11:57:33 OK 2012-04-01T11:57:43 they are just possible tools you might end up using to achieve the goals 2012-04-01T11:58:11 OK 2012-04-01T11:58:22 the real questions to answer are: how will tests be run/results reported and how will new tests be written 2012-04-01T11:58:40 part of these answers may include tools such as googletest/mock 2012-04-01T11:59:07 For google tests, it can groups common test cases as test fixtures 2012-04-01T11:59:27 And by organizing the tests, it can have a good test reports 2012-04-01T11:59:42 Automation can be achived in this way 2012-04-01T12:00:32 Yes. What I would focus on first is how to refactor the existing test cases we have implemented to support the 4 test groups verm__ identified 2012-04-01T12:01:09 Yes. The documentation for the existing tests suites are limited 2012-04-01T12:01:15 i agree completely 2012-04-01T12:01:28 I kind of lost looking through them 2012-04-01T12:01:51 i have a table of them in a google doc I can share with you 2012-04-01T12:02:06 It's still a lot of data but a little easier to read than just the files/directories 2012-04-01T12:02:08 Thank you. 2012-04-01T12:03:50 ok I shared it with you. This is something I did about 2 years ago so it is a little outdated but not too bad 2012-04-01T12:05:07 Thanks! 2012-04-01T12:05:09 Got it 2012-04-01T12:05:42 So what do you suggest I need to do in order to right a good proposal 2012-04-01T12:06:53 Write the high-level goals in the abstract, make sure you fill in all the bio stuff that matters, and a link to the proof that you did the hello world test... 2012-04-01T12:07:16 for the body of the proposal itself, try to decompose the project into stages 2012-04-01T12:07:36 Can I just paste the picture onto the proposal? 2012-04-01T12:07:45 the screenshot 2012-04-01T12:07:49 of the proof 2012-04-01T12:07:54 yeah at least on the google doc 2012-04-01T12:07:58 OK 2012-04-01T12:07:59 it might not come through when you copy-paste into melange 2012-04-01T12:08:07 OK 2012-04-01T12:08:12 Thanks 2012-04-01T12:08:17 also: make sure you go to melange and sign up to be a student and submit your initial proposal there 2012-04-01T12:08:26 and just update it by copy-pasting from your google doc periodically 2012-04-01T12:08:31 OK 2012-04-01T12:08:33 that way you will have at least something in place 2012-04-01T12:08:37 OK 2012-04-01T12:08:45 Thank you. 2012-04-01T12:09:08 I have to grab a bite. See you. 2012-04-01T12:09:17 *** Deb has left #rtems ("Leaving") 2012-04-01T12:12:53 *** arvind_k has joined #rtems 2012-04-01T12:24:03 *** kristianpaul has quit IRC (Quit: leaving) 2012-04-01T12:29:46 *** Deb has joined #rtems 2012-04-01T12:35:25 *** kristianpaul has joined #rtems 2012-04-01T12:35:25 *** kristianpaul has joined #rtems 2012-04-01T12:57:29 *** Deb has quit IRC (Quit: This computer has gone to sleep) 2012-04-01T13:10:29 *** Deb has joined #rtems 2012-04-01T13:44:59 *** Hesham has joined #rtems 2012-04-01T13:45:24 gedare: Hey 2012-04-01T13:45:31 hi hesham. 2012-04-01T13:46:25 i have compiled lm3s69xx BSP but there were no samples to test although configuration included --enable-tests=samples 2012-04-01T13:47:05 so i could not simulate it with Qemu as there were no samples 2012-04-01T13:48:18 sounds like maybe compiling failed 2012-04-01T13:48:54 go into your build tree... under arm-rtems4.11/c/lm3s69xx/testsuites/samples/hello 2012-04-01T13:48:55 and type make 2012-04-01T13:49:03 but i got no compilation error with that BSP 2012-04-01T13:51:12 no samples dir there 2012-04-01T13:53:59 something went wrong. 2012-04-01T13:54:21 what was your configure command? 2012-04-01T13:55:07 --enable-tests=samples 2012-04-01T13:55:19 entire line 2012-04-01T13:57:39 ../configure --target=arm-rtems4.11 --enable-rtemsbsp=lm3s69xx --disable-posix --disable-itron --enable-tests=samples 2012-04-01T13:58:23 and you got no error from that? 2012-04-01T13:58:51 no i got no errors 2012-04-01T13:59:07 but i did not find samples directory 2012-04-01T14:00:24 broken configure... the bsp is actually lm3s6965 2012-04-01T14:00:50 it failed silently 2012-04-01T14:01:14 ohh 2012-04-01T14:01:24 i will compile it again now 2012-04-01T14:01:52 did you apply the patches in the bsp directory also? 2012-04-01T14:02:05 yeah 2012-04-01T14:02:26 to your qemu right 2012-04-01T14:02:46 and qemu built Ok? 2012-04-01T14:03:24 yeah 2012-04-01T15:10:37 *** Hesham has quit IRC (Ping timeout: 276 seconds) 2012-04-01T15:12:45 *** Deb has quit IRC (Quit: This computer has gone to sleep) 2012-04-01T15:20:24 *** Deb has joined #rtems 2012-04-01T15:40:38 *** Deb has left #rtems ("Leaving") 2012-04-01T15:51:55 *** kristianpaul has quit IRC (Quit: leaving) 2012-04-01T15:53:38 *** kristianpaul has joined #rtems 2012-04-01T15:53:38 *** kristianpaul has joined #rtems 2012-04-01T16:13:09 *** Hesham has joined #rtems 2012-04-01T16:21:12 gedare: i got that message when i run make for lm3s6965 2012-04-01T16:21:17 gmake[2]: Entering directory `/home/hesham/lm3s6965/arm-rtems4.11/c/lm3s6965' 2012-04-01T16:21:18 gmake[2]: *** No rule to make target `all'. Stop. 2012-04-01T16:25:31 *** Deb has joined #rtems 2012-04-01T16:39:41 *** Hesham has quit IRC (Quit: Leaving.) 2012-04-01T16:50:15 *** soh_cah_toa has joined #rtems 2012-04-01T16:57:23 *** soh_cah_toa has quit IRC (Read error: Connection reset by peer) 2012-04-01T16:57:43 *** soh_cah_toa has joined #rtems 2012-04-01T16:58:22 *** soh_cah_toa has joined #rtems 2012-04-01T17:40:35 *** Deb has left #rtems ("Leaving") 2012-04-01T18:04:11 *** Jennifer has quit IRC (Ping timeout: 245 seconds) 2012-04-01T18:24:49 *** arvind_k has quit IRC (Ping timeout: 246 seconds) 2012-04-01T20:52:36 *** Deb has joined #rtems 2012-04-01T21:33:20 *** joe9 has quit IRC (Quit: leaving) 2012-04-01T22:45:49 *** Deb has quit IRC (Quit: Leaving) 2012-04-01T23:55:53 *** soh_cah_toa has quit IRC (Quit: hasta la pasta!)