2012-08-06T01:51:03 *** sebhub has joined #rtems 2012-08-06T01:52:18 good morning 2012-08-06T02:13:11 *** WikL has joined #rtems 2012-08-06T02:29:42 *** arvind_khadri has joined #rtems 2012-08-06T02:33:37 *** arvind_khadri has joined #rtems 2012-08-06T02:47:34 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2012-08-06T02:50:58 *** arvind_khadri has joined #rtems 2012-08-06T02:55:30 *** arvind_khadri has quit IRC (Quit: Leaving) 2012-08-06T03:00:29 *** arvind_khadri has joined #rtems 2012-08-06T04:12:07 *** WikL has quit IRC (Ping timeout: 240 seconds) 2012-08-06T04:14:17 *** WikL has joined #rtems 2012-08-06T04:34:48 *** krzysiekm13 has joined #rtems 2012-08-06T04:35:29 hello 2012-08-06T04:36:43 i am trying to do coverage analysis but i have some troubles. I followed instructions on wiki and bootstrap has beed done well but I don't know exactly how run coverage 2012-08-06T04:36:54 could you help me? 2012-08-06T04:40:22 *** krzysiekm13 has left #rtems 2012-08-06T04:40:43 *** krzysiekm13 has joined #rtems 2012-08-06T04:41:49 it's 5:41am you will have to wait around 5 hours to get some responses 2012-08-06T04:53:50 ok :) in Poland it's 11:50 so I hoped that here is anyone from Europe or Asia who can help me :) 2012-08-06T05:07:12 there are some.. your chances increase around 3PM GMT because that is when most people are around 2012-08-06T05:25:18 *** krzysiekm13 has left #rtems 2012-08-06T05:26:40 *** krzysiekm13 has joined #rtems 2012-08-06T05:30:45 *** krzysiekm13 has quit IRC (Ping timeout: 246 seconds) 2012-08-06T05:33:27 *** krzysiekm13 has joined #rtems 2012-08-06T05:33:35 *** krzysiekm13 has left #rtems 2012-08-06T07:31:25 *** WikL has quit IRC (Ping timeout: 272 seconds) 2012-08-06T09:01:12 *** QingPei has joined #rtems 2012-08-06T09:20:16 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-08-06T09:36:13 *** kristianpaul has quit IRC (Quit: leaving) 2012-08-06T09:41:57 *** kristianpaul has joined #rtems 2012-08-06T09:41:57 *** kristianpaul has joined #rtems 2012-08-06T09:53:08 *** QingPei has quit IRC (*.net *.split) 2012-08-06T09:53:08 *** sevikkk has quit IRC (*.net *.split) 2012-08-06T09:53:09 *** juli1 has quit IRC (*.net *.split) 2012-08-06T09:53:53 *** QingPei has joined #rtems 2012-08-06T09:55:24 *** QingPei has left #rtems 2012-08-06T09:55:32 *** arvind_khadri has quit IRC (Quit: Leaving) 2012-08-06T09:55:44 *** sevikkk has joined #rtems 2012-08-06T09:55:44 *** juli1 has joined #rtems 2012-08-06T09:55:52 *** arvind_khadri has joined #rtems 2012-08-06T09:56:32 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-06T09:57:58 *** arvind_khadri has joined #rtems 2012-08-06T09:59:58 *** arvind_khadri has joined #rtems 2012-08-06T10:01:23 *** arvind_khadri has joined #rtems 2012-08-06T10:01:59 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-06T10:02:23 *** arvind_khadri has joined #rtems 2012-08-06T10:04:34 *** WikL has joined #rtems 2012-08-06T10:31:56 *** gedare has joined #rtems 2012-08-06T11:05:23 *** WikL has quit IRC (Remote host closed the connection) 2012-08-06T11:12:30 *** WikL has joined #rtems 2012-08-06T11:15:04 *** DrJoel has joined #rtems 2012-08-06T11:15:04 *** ChanServ sets mode: +o DrJoel 2012-08-06T11:15:34 * DrJoel is here but still getting settled.. OAR is getting new carpet and paint.. so today was my day to empty my office. I am temporarily in the RTEMS lab. 2012-08-06T11:16:24 I haven't even done an email on my phone since I have been at the office. :( 2012-08-06T11:19:37 *** kristianpaul has quit IRC (Quit: Reconnecting) 2012-08-06T11:19:50 *** kristianpaul has joined #rtems 2012-08-06T11:19:50 *** kristianpaul has joined #rtems 2012-08-06T11:51:50 *** beng-nl has joined #rtems 2012-08-06T11:53:18 lo 2012-08-06T11:54:38 hi beng-nl 2012-08-06T12:00:21 hey gedare! 2012-08-06T12:00:25 nice to bump into you here 2012-08-06T12:00:35 i hope you remember me (minix-ben) from last gsoc mentor summit :) 2012-08-06T12:00:56 if not, saying no or pretending you do is both ok ;) 2012-08-06T12:09:55 ah, hi ben 2012-08-06T12:10:16 well i only recall one minix guy so it must be you. 2012-08-06T12:10:37 *** arvind_khadri has joined #rtems 2012-08-06T12:24:56 *** sebhub has quit IRC (Remote host closed the connection) 2012-08-06T12:27:01 :-) 2012-08-06T12:27:07 * beng-nl 'll take it 2012-08-06T13:14:01 *** deb has joined #rtems 2012-08-06T13:30:35 *** dr__hous1 has joined #rtems 2012-08-06T13:30:40 *** kristian1aul has joined #rtems 2012-08-06T13:30:41 *** kuzew_ has joined #rtems 2012-08-06T13:39:05 *** dr__house has quit IRC (Write error: Broken pipe) 2012-08-06T13:39:13 *** kristianpaul has quit IRC (Remote host closed the connection) 2012-08-06T13:39:13 *** kuzew has quit IRC (Remote host closed the connection) 2012-08-06T14:30:59 *** kristian1aul has quit IRC (Quit: Reconnecting) 2012-08-06T14:31:07 *** kristianpaul has joined #rtems 2012-08-06T14:31:07 *** kristianpaul has joined #rtems 2012-08-06T15:41:08 *** WikL has quit IRC (Remote host closed the connection) 2012-08-06T16:07:11 *** gedare has quit IRC (Quit: Leaving) 2012-08-06T17:54:15 *** DrJoel has quit IRC (Quit: Leaving) 2012-08-06T20:52:40 *** dr__hous1 is now known as dr__house 2012-08-06T20:52:50 *** dr__house has joined #rtems 2012-08-06T21:59:38 *** xiangfu has joined #rtems 2012-08-06T22:12:47 *** kuzew_ is now known as kuzew 2012-08-06T22:42:12 *** deb has left #rtems ("Leaving") 2012-08-06T22:49:54 *** arvind_khadri has quit IRC (Ping timeout: 264 seconds) 2012-08-06T23:04:27 *** xiangfu has quit IRC (Ping timeout: 240 seconds) 2012-08-06T23:37:42 *** kuzew has quit IRC (Ping timeout: 244 seconds) 2012-08-06T23:39:09 *** kuzew has joined #rtems 2012-08-06T23:49:31 *** kuzew has quit IRC (Ping timeout: 252 seconds) 2012-08-06T23:51:22 *** kuzew has joined #rtems 2012-08-07T01:35:54 *** WikL has joined #rtems 2012-08-07T01:38:50 *** xiangfu has joined #rtems 2012-08-07T01:42:18 *** arvind_khadri has joined #rtems 2012-08-07T01:42:57 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-07T01:43:27 *** arvind_khadri has joined #rtems 2012-08-07T01:44:53 *** arvind_khadri has joined #rtems 2012-08-07T01:46:03 *** arvind_khadri has joined #rtems 2012-08-07T01:46:40 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-07T01:49:49 *** sebhub has joined #rtems 2012-08-07T02:14:22 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-08-07T02:21:28 *** xiangfu has joined #rtems 2012-08-07T02:23:18 *** xiangfu has quit IRC (Read error: Connection reset by peer) 2012-08-07T03:49:00 *** arvind_khadri has joined #rtems 2012-08-07T04:17:56 *** xiangfu has joined #rtems 2012-08-07T07:06:13 *** WikL has quit IRC (Remote host closed the connection) 2012-08-07T08:21:23 *** deb has joined #rtems 2012-08-07T08:45:51 *** xiangfu has quit IRC (Ping timeout: 245 seconds) 2012-08-07T08:48:33 *** gedare has joined #rtems 2012-08-07T09:18:27 *** Hesham1 has joined #rtems 2012-08-07T09:23:06 *** Hesham1 has quit IRC (Ping timeout: 265 seconds) 2012-08-07T09:26:30 *** Hesham1 has joined #rtems 2012-08-07T10:05:23 hi gedare 2012-08-07T10:05:30 hi Hesham1 2012-08-07T10:05:36 how's it going? 2012-08-07T10:05:49 good 2012-08-07T10:06:09 i have modified comments 2012-08-07T10:06:17 and pushed them 2012-08-07T10:06:21 fantastic.. ok i'll have another look soon 2012-08-07T10:06:32 and pushed arm code too 2012-08-07T10:06:45 differnet branch? 2012-08-07T10:06:47 yes 2012-08-07T10:06:49 k 2012-08-07T10:06:50 arm_mmu 2012-08-07T10:07:01 want review on that yet, or still in progress? 2012-08-07T10:07:14 no still in progress 2012-08-07T10:07:22 and have some questions 2012-08-07T10:07:26 i agree with sebhub that we should not consider nested page table unless absolutely necessary 2012-08-07T10:07:47 er, multilevel 2012-08-07T10:07:47 sorry 2012-08-07T10:07:59 (nested too but that is something different.) 2012-08-07T10:08:30 because multilevel page table is more likely to cause page faults 2012-08-07T10:08:36 but i think powerpc have multi level 2012-08-07T10:08:36 is not it ? 2012-08-07T10:09:00 i haven't looked but ... i would still try to keep it simple 2012-08-07T10:09:03 yes i have handled that issue 2012-08-07T10:09:21 most architecture will support multilevel page table.. 2012-08-07T10:09:34 but if s/w manages the page table then you can keep it simpler 2012-08-07T10:09:41 if h/w does the page walk then you have to do more work 2012-08-07T10:09:50 for sparc64 we use only 4MB pages 2012-08-07T10:10:02 it is limiting but it makes life easier ;) 2012-08-07T10:10:16 i am managing page table from software 2012-08-07T10:10:20 every region has its own pagetable 2012-08-07T10:10:30 4MB ? 2012-08-07T10:10:45 what if we have to apply protection for 128KB only ? 2012-08-07T10:11:16 then there is fragmentation :) 2012-08-07T10:11:36 a big one 2012-08-07T10:12:16 and what if we have to apply different protection for a region that lies in the same 4MB Page ? 2012-08-07T10:12:18 but i think it is generally acceptable 2012-08-07T10:12:43 this is a problem for developer to resolve. if they need this they need to align their regions differently 2012-08-07T10:13:04 and regions overlaps ? 2012-08-07T10:13:49 depends on whether the hw can support it, but probably not 2012-08-07T10:14:17 ok i will let you know about how ARM MMU work 2012-08-07T10:14:27 ok 2012-08-07T10:14:34 if we used level 1 pages only 2012-08-07T10:14:51 every entry will define 1MB page/section 2012-08-07T10:16:12 that's fine. but you see this is why i said the test cases are non-portable 2012-08-07T10:16:26 if every arch has a different max page-size it supports then we have some small issues 2012-08-07T10:16:48 in general though, developers will be aiming toward a particular BSP / CPU, so they will know what the page size limits are and how to align their regions 2012-08-07T10:16:53 i know that makes life a lot easier 2012-08-07T10:18:16 and we may have some size checking / addresses aligning checking right ? 2012-08-07T10:19:16 according to a particular BSP/CPU 2012-08-07T10:19:47 yeah 2012-08-07T10:19:55 but again this is buried deep.. 2012-08-07T10:20:05 and can be done at init times, or put the burden on the developer a little 2012-08-07T10:20:16 right now we're just trying to bring some unity to how mmu is handled in rtems 2012-08-07T10:20:20 and give some higher-level interfaces 2012-08-07T10:21:38 ok 2012-08-07T10:24:41 *** Hesham2 has joined #rtems 2012-08-07T10:24:43 but if that is done at init time, libmm will not make use of that API 2012-08-07T10:24:51 and give some higher-level interfaces 2012-08-07T10:24:51 ok 2012-08-07T10:24:51 * Hesham2 (~hesham@197.132.136.119) has joined #rtems 2012-08-07T10:24:51 but if that is done at init time, libmm will not make use of that API 2012-08-07T10:24:54 did i miss anything? 2012-08-07T10:25:08 but if that is done at init time, libmm will not make use of that API 2012-08-07T10:25:13 ok. 2012-08-07T10:25:26 right ? 2012-08-07T10:25:52 i meant by init... when you first set up the memory mapping 2012-08-07T10:25:59 so for a static map 2012-08-07T10:26:03 yes 2012-08-07T10:26:08 *** Hesham1 has quit IRC (Ping timeout: 268 seconds) 2012-08-07T10:26:16 that still relies on libmm, or not? 2012-08-07T10:26:17 and when creating a new mpe ? 2012-08-07T10:26:22 yeah 2012-08-07T10:26:32 probably then too, since you'd want to verify the region 2012-08-07T10:26:40 but not when installing/uninstalling 2012-08-07T10:26:44 pre-validate 2012-08-07T10:26:56 when creating an mpe that will no longer be static 2012-08-07T10:27:07 aha 2012-08-07T10:27:30 installing/uninstalling are dynamic operations 2012-08-07T10:28:07 ya 2012-08-07T10:28:23 in some sense you can think about setting up a "dynamic" map ahead of time though that never changes once it is installed 2012-08-07T10:28:28 this is neither static nor dynamic really 2012-08-07T10:29:03 and can not be installed ? 2012-08-07T10:29:09 uninstalled ** 2012-08-07T10:30:38 right that would be a way to set a permanent mapping 2012-08-07T10:30:49 you would then not induce any faults or overhead at runtime ideally, after installing 2012-08-07T10:31:06 this is more or less how i expect the mmu normally would be used 2012-08-07T10:31:19 until you need to do protection... all you want is a permanent translation / mapping 2012-08-07T10:31:32 once you want protection then you need to do mapping/unmapping dynamically 2012-08-07T10:31:35 but that is future work 2012-08-07T10:32:35 i see 2012-08-07T10:32:37 ok, so what are you working on? 2012-08-07T10:32:44 we have 1 week to the soft pencils-down 2012-08-07T10:32:55 but of course you can (and should) keep working past that 2012-08-07T10:33:06 just that for gsoc they want you to get your code ready and submitted 2012-08-07T10:33:23 i was working on level 2 page table to handle install_mpe 2012-08-07T10:33:25 the not-so-secret hope is that you will continue on and contribute as you are able :) 2012-08-07T10:33:34 ah, but now we're not sure abuot that 2012-08-07T10:33:42 i will keep working on the project of course 2012-08-07T10:33:59 dr__house: I see that you're Claas's mentor. Do you know if his work on the BeagleBoard is applicable to the BeagleBone? It's an AM3359 vs the BeagleBoard's OMAP3530 2012-08-07T10:34:33 and i hope we could still chat about libmm design and changes after GSoC 2012-08-07T10:35:07 for now 2012-08-07T10:35:24 mikeg: he is interested in beaglebone support, but i'm not sure how far he got with it. you might like to inquire (der_stuttgarter_25 at hotmail dot de 2012-08-07T10:35:34 i will implement install_mpe with 1 level page table only 2012-08-07T10:35:46 ok good 2012-08-07T10:35:50 it should be easier anyway 2012-08-07T10:36:02 yes a lot 2012-08-07T10:36:08 that will give you some idea about how to write tests for it too 2012-08-07T10:36:11 Great thanks. I've been trying to catch him on IRC one of these days (probably next meeting) 2012-08-07T10:36:59 mikeg: great 2012-08-07T10:37:15 at init time, i made 1:1 mapping 1MB pages with NO PROTECTION attributes 2012-08-07T10:37:42 does the BSP right now do anything with the mmu? 2012-08-07T10:37:48 I think what threw me was I saw something online saying "RTEMS running on LEON3 processor or ARM processor running on BeagleBoard" (http://rtjava.blogspot.com/2011/07/real-time-embedded-java-helloworld-with.html), which I don't think is actually working yet 2012-08-07T10:37:51 before your code changes? 2012-08-07T10:38:14 ahhh the IU folks.. 2012-08-07T10:38:15 yes 2012-08-07T10:38:39 mikeg: they tend to post early ,and you can't always find their code :) they do good work though 2012-08-07T10:38:56 it initialize level 1 page table with AP/CB attr embedded at a struct 2012-08-07T10:39:03 Hesham2: does your 1:1 mapping essentially do the same thing as what is done right now? 2012-08-07T10:39:06 i see 2012-08-07T10:39:14 Who do you mean they? The RTJava people? 2012-08-07T10:39:36 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2012-08-07T10:39:57 mikeg: yeah sorry. Jan Vitek's group out of Indiana U 2012-08-07T10:40:13 No 2012-08-07T10:40:36 the previous code does not create 1:1 mapping 2012-08-07T10:40:38 Hesham2: OK, I think getting something functionally equivalent to what is done right now might be important 2012-08-07T10:41:16 I think one benefit of this project should be to make BSP porting easier by showing how to deal with MMU.. 2012-08-07T10:41:23 right now there is no guidance 2012-08-07T10:41:41 i think there is 2012-08-07T10:42:28 Ah, I see. I got their software to work on RTLinux on the board, but not yet RTEMS :) Thanks again gedare 2012-08-07T10:42:46 yw mikeg good luck 2012-08-07T10:42:52 if anyone want to port a new MMU target, he has to do some steps at BSP code 2012-08-07T10:43:06 1- defining bsp_mm_mpe 2012-08-07T10:43:10 yes yes yes 2012-08-07T10:43:12 i meant before your project 2012-08-07T10:43:14 :) 2012-08-07T10:43:42 if your project defines a way for BSP porters to write MMU handling code then this is good 2012-08-07T10:43:50 one lingering issue is early-boot 2012-08-07T10:43:57 not sure what to do about that. 2012-08-07T10:44:00 You where right when you said that once the frame work is ready, the low-level work will not take a time to port new one 2012-08-07T10:44:10 ya 2012-08-07T10:44:30 but anyway making it easier to port new BSPs is just a side-benefit. so no reason to think about it too much 2012-08-07T10:44:44 i can write document for Porting MMU for new targets 2012-08-07T10:44:52 ok 2012-08-07T10:45:01 all right anything else? 2012-08-07T10:45:34 I will edit a little on libmm_working branch 2012-08-07T10:45:53 as well as porting libmm for ARM 2012-08-07T10:46:01 you tell me what else ? 2012-08-07T10:46:47 what else you doing on libmm_working? 2012-08-07T10:47:04 i thought that one is ready for review 2012-08-07T10:47:15 it's now 2012-08-07T10:49:00 *** sebhub has quit IRC (Remote host closed the connection) 2012-08-07T10:52:42 ok. continue with ARM, that should help you to smooth out the shared-layer code and tests 2012-08-07T10:53:37 and write up some documentation as well as you can 2012-08-07T10:54:43 ok, i will push code for arm regularly even it is not tested 2012-08-07T10:55:09 you can review it anytime 2012-08-07T10:56:19 ok 2012-08-07T10:56:45 i have limited time as you have noticed but will make an effort to review your code before end of this week 2012-08-07T10:59:23 *** Hesham2 has quit IRC (Ping timeout: 246 seconds) 2012-08-07T11:02:37 *** Hesham1 has joined #rtems 2012-08-07T11:13:01 *** QingPei has joined #rtems 2012-08-07T12:16:15 *** Hesham1 has quit IRC (Ping timeout: 276 seconds) 2012-08-07T12:18:30 *** Hesham1 has joined #rtems 2012-08-07T12:37:12 *** arvind_khadri has joined #rtems 2012-08-07T13:06:28 so.. rtems is related with latest mars rover? 2012-08-07T13:08:26 *** QingPei has left #rtems 2012-08-07T13:09:19 kristianpaul: yes it is being used on some components 2012-08-07T13:10:45 you have a reference for that? 2012-08-07T13:11:55 *** Hesham1 has quit IRC (Quit: Leaving.) 2012-08-07T13:12:37 not sure if it has been acknowledged / verified, hmm 2012-08-07T13:13:38 http://www.rtems.org/pipermail/rtems-users/2012-August/010316.html 2012-08-07T13:17:45 it's not clear from that if rtems is on the rover or the orbiter though. 2012-08-07T13:31:23 *** DrJoel has joined #rtems 2012-08-07T13:31:23 *** DrJoel has joined #rtems 2012-08-07T13:31:23 *** ChanServ sets mode: +o DrJoel 2012-08-07T13:35:08 gedare: there is a new orbiter also??? 2012-08-07T13:35:41 no the existing one i think 2012-08-07T13:35:54 already had electra radio system on it which i believe uses some rtems 2012-08-07T13:36:30 orbiter launched in 2005 and gedare is correct.. also uses Electra :) 2012-08-07T13:37:58 http://static.usenix.org/event/lisa06/tech/slides/byrne.pdf plus the presentation I posted on testing Electra is the most detailed info we have found 2012-08-07T13:38:31 the byrne presentation also gives a good hint that the Terminal Descent Sensor is running RTEMS (e.g. derived from Electra HW/SW) 2012-08-07T13:40:01 DrJoel: hi ! 2012-08-07T13:40:30 hey kristianpaul 2012-08-07T13:41:51 Can you point me to an example that ilustrate use of semaphors and event flags in rtems? 2012-08-07T13:42:03 * kristianpaul misses a rtems book 2012-08-07T13:42:16 * DrJoel has wanted to write one for a long time. 2012-08-07T13:42:50 Semaphores have many uses.. counting sems are often used to control allocation of a pool of multiple identical resources.. 2012-08-07T13:43:34 well seems i need to understand this use in eCos before portting to rtems.. 2012-08-07T13:43:54 simple binary semaphores are used to signal "events". In VxWorks, a task may block on a simple binary semaphore and an ISR post the semaphore to unblock it 2012-08-07T13:44:21 mutex binary semaphore.. mutual exclusion.. critical sections.. with optional priority inheritance and ceiling protocols 2012-08-07T13:44:51 events are "condition synchronization" with each bit (e.g. event flag) a distinct condition 2012-08-07T13:45:11 look in the examples-v2. I think there should be examples on how to use various types. If the explanation for any isn't good enough, ping me 2012-08-07T13:45:49 okay 2012-08-07T13:46:40 if you need a particular eCos object mapped to RTEMS, post a question to the users list. Trust me, plenty of people have done the same thing for a variety of RTOSes 2012-08-07T13:47:34 yeah just suscribed to user list some minutes ago :) 2012-08-07T13:50:58 having a guide on porting from RTOS X to RTEMS would be nice or porting kits with helpful code. But I have always thought it wouldn't be proper to host it at rtems.org. 2012-08-07T13:56:33 or pthreads support.. 2012-08-07T14:00:01 most of pthreads is supported 2012-08-07T14:00:10 lots of examples in testsuites/psxtests 2012-08-07T14:00:27 the main thing you can't do is fork. 2012-08-07T14:04:00 :-| 2012-08-07T14:22:04 eCos doesn't fork AFAIK .. so this shouldn't impact you :) 2012-08-07T14:25:05 lol no that no 2012-08-07T14:25:23 just was thinking in linux ;) 2012-08-07T14:28:12 still hard to me think in rtos :) 2012-08-07T14:28:24 rtems* :) 2012-08-07T14:31:38 As Gedare pointed out, RTEMS has a nice POSIX set. I have compared it recently against the POSIX profiles in http://www.opengroup.org/getinvolved/consortia/face/ and we look pretty good. 2012-08-07T14:32:06 For the largest of the four profiles they define for aviation systems, we are primarily missing complex math and some wide character routines 2012-08-07T14:32:26 for the three smaller profiles, we have between 5 and a dozen routines missing 2012-08-07T15:31:37 *** deb has quit IRC (Quit: Leaving) 2012-08-07T15:41:47 hi DrJoel, nice that you are here :) 2012-08-07T15:43:07 hey ben.. enjoying the pictures of the office stripped of carpet and paint? 2012-08-07T15:43:36 i am actualyl :-) 2012-08-07T15:43:38 actually* 2012-08-07T15:43:56 how good that i've met everyone i know of the rtems team on #rtems within just a few days 2012-08-07T15:44:20 DrJoel: although the MSL landing was at least as exciting 2012-08-07T15:44:22 :-) 2012-08-07T15:44:31 (which inspired my coming here) 2012-08-07T15:45:18 yay 2012-08-07T15:47:15 The MSL has been exciting.. 2012-08-07T15:47:38 Trying to find out what the Electra radio on MRO and MSL is used for. There are at least two radios on MSL 2012-08-07T15:47:54 hm-hm 2012-08-07T15:48:13 it's a tad surprising (to me) those guys don't tell the project that they're using it 2012-08-07T15:48:42 you have been doing FOSS a long time.. do you know everywhere Minix is used? 2012-08-07T15:48:57 well no, but i had assumed more basic reasons for that :) 2012-08-07T15:49:23 For space apps, People build these devices, finish them, the devices go into storage until a launch date.. years pass... they are on another project.. launches.. 2012-08-07T15:49:54 hmmm 2012-08-07T15:50:07 it is only by googling a lot and asking questions when we meet people that we get any details. Scientific papers, budget reports, annual reports seem to be most telling 2012-08-07T15:50:19 my undergrad uni converted to google apps / gmail, so now i have another google account. 2012-08-07T15:50:23 for my alumni email 2012-08-07T15:50:37 * gedare talks at random 2012-08-07T15:50:55 DrJoel: it seems natural to me to want contact with the project itself for advice/support, that's all; apparently they want to do that all themselves 2012-08-07T15:50:57 just what you need... another email account 2012-08-07T15:51:25 it's ok i just get one email per month on it 2012-08-07T15:51:25 we do talk to the people but often we don't. I think Electra was on its way to Mars on the MRO before I knew about it. 2012-08-07T15:51:42 just a shame as the project doesn't even get credit (easily) for it in such cases 2012-08-07T15:51:42 I have gotten more agressive about asking questions, requesting visits to places, etc. 2012-08-07T15:52:10 That's the shame of it. RTEMS really has a great bunch of fielded applications and it is hard to pass them on. 2012-08-07T15:52:13 DrJoel: well i've done my part by re-tweeting the odd tweet to my 'massive' number of followers ;-) 2012-08-07T15:52:24 FYI we have a satellite launch later this month. Need to announce that. 2012-08-07T15:52:35 Are you into 3 digits? 2012-08-07T15:52:42 DrJoel: technically yes! 2012-08-07T15:52:57 lol.. Lady Gaga we are not 2012-08-07T15:53:10 i have so far successfully avoided twitter. 2012-08-07T15:53:20 DrJoel: i should resist telling you but yes i have exactly 100 followers for my personal twitter account 2012-08-07T15:53:52 i almost started a twitter feed last week... where i would post things that people should know but generally don't. and now and then things people "know" but shouldn't ;) 2012-08-07T15:53:54 to my surprise/satisfaction @minix3 keeps accruing followers steadily and is at 511 now 2012-08-07T15:54:05 My personal one only has 29 .. LOL 2012-08-07T15:54:36 That's way beyond RTEMS... 47.. but over 100 on Facebook 2012-08-07T15:54:55 strange world huh raelly 2012-08-07T15:54:56 raelly 2012-08-07T15:54:57 drat! 2012-08-07T15:55:02 r e a l l y 2012-08-07T15:55:36 i've been mulling over the idea of a minix3 blog.. or maybe i should stop talking about minix on #rtems :) 2012-08-07T15:56:19 anyway, i have a few good ideas for posts i could 'pre-load' it with to automatically publish on regular days.. that would be nice and professional 2012-08-07T15:56:24 the challenge would be to keep it going 2012-08-07T15:57:19 that's the challenge of blogs :) 2012-08-07T15:57:37 yeah 2012-08-07T15:57:42 i fill mine with random topics and multi-month gaps 2012-08-07T15:58:02 i actually finally posted today. about freezing food :) 2012-08-07T16:00:22 but maybe that is because the only thing to talk about in rtems right now is the hard reset of our repo ;) 2012-08-07T16:01:32 what's the repo reset? 2012-08-07T16:06:40 Keeping a blog active is hard work. I often use mine to post status reports. Or just write ups of interesting debug sessions. 2012-08-07T16:06:53 * DrJoel also has gaps 2012-08-07T16:07:44 a repo hard reset is when someone doesn't follow procedures, doesn't get reviews, and commits without discussion. Huge commit requires big hammer to revert 2012-08-07T16:09:15 i see, forcing the repo to have an older commit id? 2012-08-07T16:09:52 which is normally bad practice (hence hammer) because git repo's that have the bad commit id will get confused? 2012-08-07T16:10:24 yep.. but this series of commits was so egregious that no one committed afterwards and apparently no one updated after it was committed. :) 2012-08-07T16:10:33 okay 2012-08-07T16:11:57 you can commit the reverse commits on top of the bad ones to undo them and keep the data structure OK though actually (git revert). if that's what you want of course. 2012-08-07T16:15:22 These were LARGE commits.. 100s of generated files. 2012-08-07T16:15:53 i did revert a couple of them, because there were a few that happened months ago 2012-08-07T16:16:05 and i did not want to mess the history up quite so much 2012-08-07T16:16:20 but the decision was to make the massive commits go poof 2012-08-07T16:16:39 ok, just saying ;) 2012-08-07T16:16:43 all right... more or less done for the day 2012-08-07T16:16:44 yeah 2012-08-07T16:16:53 time for a jog! fresh air 2012-08-07T16:16:58 nice! enjoy 2012-08-07T16:16:59 ! 2012-08-07T16:17:22 *** gedare has quit IRC (Quit: Leaving) 2012-08-07T16:31:20 *** DrJoel has quit IRC (Quit: Leaving) 2012-08-07T17:02:14 seems posix is way go make all more portable if i run linux or rtems.. ;-), and seems fork no need :D 2012-08-07T17:07:11 *** Hesham has joined #rtems 2012-08-07T17:46:15 *** Hesham has quit IRC (Ping timeout: 265 seconds) 2012-08-07T17:48:04 *** Hesham has joined #rtems 2012-08-07T18:35:00 *** deb has joined #rtems 2012-08-07T20:03:52 *** xiangfu has joined #rtems 2012-08-07T20:32:58 *** Hesham1 has joined #rtems 2012-08-07T20:33:26 *** Hesham has quit IRC (Ping timeout: 256 seconds) 2012-08-07T20:49:53 *** Hesham has joined #rtems 2012-08-07T20:53:04 *** Hesham1 has quit IRC (Ping timeout: 255 seconds) 2012-08-07T21:03:29 *** Hesham has quit IRC (Quit: Leaving.) 2012-08-07T23:19:48 *** deb has left #rtems ("Leaving") 2012-08-07T23:44:59 *** arvind_khadri has quit IRC (Ping timeout: 256 seconds) 2012-08-07T23:45:56 *** kuzew has quit IRC (Read error: Operation timed out) 2012-08-07T23:48:58 *** kuzew has joined #rtems 2012-08-08T00:38:41 *** QingPei has joined #rtems 2012-08-08T01:11:00 *** arvind_khadri has joined #rtems 2012-08-08T01:12:51 *** arvind_khadri has joined #rtems 2012-08-08T01:14:34 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-08T01:40:01 *** arvind_khadri has joined #rtems 2012-08-08T01:52:24 *** arvind_khadri has quit IRC (Remote host closed the connection) 2012-08-08T01:58:13 *** arvind_khadri has joined #rtems 2012-08-08T01:59:30 *** arvind_khadri has joined #rtems 2012-08-08T02:04:59 *** arvind_khadri has quit IRC (Quit: Leaving) 2012-08-08T02:10:53 *** arvind_khadri has joined #rtems 2012-08-08T02:50:30 *** arvind_khadri has quit IRC (Ping timeout: 255 seconds) 2012-08-08T03:19:34 *** sebhub has joined #rtems 2012-08-08T03:21:26 good morning 2012-08-08T03:22:05 good morning sebhub 2012-08-08T03:42:55 *** QingPei has left #rtems 2012-08-08T04:38:15 *** arvind_khadri has joined #rtems 2012-08-08T04:39:06 *** arvind_khadri has joined #rtems 2012-08-08T04:39:37 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-08T04:40:03 *** arvind_khadri has joined #rtems 2012-08-08T04:41:51 *** arvind_khadri has joined #rtems 2012-08-08T04:48:39 *** arvind_khadri has quit IRC (Read error: Operation timed out) 2012-08-08T04:49:20 *** arvind_khadri has joined #rtems 2012-08-08T05:49:18 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2012-08-08T07:45:50 *** arvind_khadri has joined #rtems 2012-08-08T08:14:59 *** gedare has joined #rtems 2012-08-08T08:31:51 *** DrJoel has joined #rtems 2012-08-08T08:31:51 *** ChanServ sets mode: +o DrJoel 2012-08-08T09:29:52 *** deb has joined #rtems 2012-08-08T09:45:44 *** arvind_khadri has quit IRC (Ping timeout: 252 seconds) 2012-08-08T10:00:59 *** arvind_khadri has joined #rtems 2012-08-08T11:02:25 *** xiangfu has quit IRC (Ping timeout: 246 seconds) 2012-08-08T11:05:29 *** sebhub has quit IRC (Remote host closed the connection) 2012-08-08T11:56:11 *** Hesham has joined #rtems 2012-08-08T12:35:39 *** arvind_khadri has quit IRC (Ping timeout: 256 seconds) 2012-08-08T12:42:02 *** arvind_khadri has joined #rtems 2012-08-08T12:50:50 *** Hesham1 has joined #rtems 2012-08-08T12:54:21 *** Hesham has quit IRC (Ping timeout: 244 seconds) 2012-08-08T12:55:08 *** deb has quit IRC (*.net *.split) 2012-08-08T12:57:58 *** deb has joined #rtems 2012-08-08T13:07:44 *** Hesham1 has quit IRC (Quit: Leaving.) 2012-08-08T13:54:26 *** gedare has quit IRC (Quit: Leaving) 2012-08-08T15:23:38 *** antgreen has joined #rtems 2012-08-08T15:31:14 *** antgreen has quit IRC (Remote host closed the connection) 2012-08-08T15:43:30 *** deb has left #rtems ("Leaving") 2012-08-08T16:13:10 *** DrJoel has quit IRC (Quit: Leaving) 2012-08-08T16:38:24 *** deb has joined #rtems 2012-08-08T16:40:55 *** Hesham has joined #rtems 2012-08-08T18:44:25 *** xiangfu has joined #rtems 2012-08-08T18:51:12 *** xiangfu has quit IRC (Quit: Leaving) 2012-08-08T19:25:02 *** Hesham has quit IRC (Quit: Leaving.) 2012-08-08T20:20:01 *** xiangfu has joined #rtems 2012-08-08T23:31:16 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2012-08-09T00:32:23 *** arvind_khadri has joined #rtems 2012-08-09T00:39:36 *** arvind_khadri has quit IRC (Ping timeout: 244 seconds) 2012-08-09T00:41:22 *** deb has left #rtems ("Leaving") 2012-08-09T00:47:10 I see RTEMS's atomic project is featured on the GSoC home page. Nice to see. 2012-08-09T01:04:43 *** arvind_khadri has joined #rtems 2012-08-09T01:05:59 *** arvind_khadri has joined #rtems 2012-08-09T01:52:42 *** sebhub has joined #rtems 2012-08-09T02:41:17 :) 2012-08-09T02:51:38 good morning 2012-08-09T02:56:55 hi 2012-08-09T03:49:29 *** arvind_khadri has quit IRC (Ping timeout: 252 seconds) 2012-08-09T04:09:27 *** arvind_khadri has joined #rtems 2012-08-09T04:10:06 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-09T04:30:37 sebhub, did you see my post about atomic calls mapping to ISR enable and disable ? 2012-08-09T04:47:59 where is this post? 2012-08-09T04:48:10 On the devel list 2012-08-09T04:48:49 I am not sure about defaulting archs that cannot support atomic instructions to isr enable/disable 2012-08-09T04:48:49 didn't look at this thread yet 2012-08-09T04:49:25 I am concerned about polluting the atomic implementation with non-atomic functionality 2012-08-09T04:49:27 *** cdcs has joined #rtems 2012-08-09T04:49:29 for uniprocessor systems this enable/disable is ok 2012-08-09T04:49:50 Yes but is that the atomics layer or above 2012-08-09T04:50:35 For example critical section support is not atomic locks, it may contain them 2012-08-09T04:52:44 this interrup disable/enable make only sense on uniprocessor systems 2012-08-09T04:53:09 Why ? This means interrupts and that is not atomic locking 2012-08-09T04:53:14 on smp the atomic ops must not disable interrupts 2012-08-09T04:53:31 We need critical sections for interrupts and on SMP we need threading and contexts. 2012-08-09T04:53:36 on uniprocessors you cannot use spin locks 2012-08-09T04:53:39 We may need light weight contexts 2012-08-09T04:54:29 Should this be an score layering rather than embedded deep in the atomics layers 2012-08-09T04:55:23 Yes on UP no spin locks and things are different. What about a single SMP capable processor ? 2012-08-09T04:55:25 i don't understand your problem, the atomic operations provide simple operations that cannot be interrupted by other threads of execution 2012-08-09T04:55:43 Does include interrupts ? 2012-08-09T04:56:35 for example on powerpc you must clear the reservations during a context switch and an interrupt will clear the reservations implicitly 2012-08-09T04:57:07 so on powerpc atomic operations work without disabling the interrupts 2012-08-09T04:57:31 same is true for arm 2012-08-09T04:58:31 So interrupts spin on locks ? 2012-08-09T04:58:45 yes 2012-08-09T04:59:02 Where is this handled in relation to UP ? 2012-08-09T04:59:14 if the interrupt context holds the lock, then you have a dead lock 2012-08-09T04:59:22 interrupted context 2012-08-09T04:59:30 Yes 2012-08-09T04:59:51 in UP you have no spin locks 2012-08-09T04:59:51 Do the atomic primitives manage the spin ? 2012-08-09T05:00:22 they can do this 2012-08-09T05:01:02 This is what concerns me. We are layering into the primitives things that we may layer else where. 2012-08-09T05:01:58 It is not clear to me where this next layering is. It the spin is an score layer then isr enable/disable could also be outside. 2012-08-09T05:02:18 the atomic primitives 2012-08-09T05:02:34 we have to separate UP from SMP here 2012-08-09T05:02:39 Yes. 2012-08-09T05:03:07 no the isr disable/enable must be part of the spin lock implementation 2012-08-09T05:03:15 In the user land I have no problem with the idea of ISR disable/enable on UP but that maybe a layer made from the atomic primitives 2012-08-09T05:03:53 http://git.rtems.org/rtems/tree/c/src/lib/libbsp/powerpc/qoriq/shmsupp/lock.S 2012-08-09T05:04:17 this is a primitive example for a powerpc spin lock 2012-08-09T05:04:43 if you move the interrupt disable out of the implementation then the time of disable interrrupts may be huge 2012-08-09T05:04:44 Is this part of the atomic primitives ? 2012-08-09T05:04:55 no, is special for one bsp 2012-08-09T05:05:33 it is based on Freescale Book E recommendations 2012-08-09T05:06:18 GCC can generate this in 4.7 in a very similar way using atomic operations 2012-08-09T05:06:25 Ok. Just concerned about things being added in the wrong layer in the atomic arch layer 2012-08-09T05:06:49 I am only talking about the score 2012-08-09T05:07:01 http://gcc.gnu.org/ml/gcc-help/2012-03/msg00377.html 2012-08-09T05:08:30 this spin lock is totally useless for the score, here we need a spin lock with fifo properties 2012-08-09T05:32:44 ok, and this will evolve as the work is done so adding is adding isr enable/disable worth it 2012-08-09T07:35:14 *** monstr_ has joined #rtems 2012-08-09T07:37:31 Hi, someone there? 2012-08-09T07:38:14 I would like to compile toolchain by rtems-crossrpms 2012-08-09T07:38:31 I haven't found any manual on the wiki. Can someone tell me how to use it? 2012-08-09T07:38:57 I would expect to run several commands and this package will download all necessary files and will compile them 2012-08-09T07:39:05 On what host ? 2012-08-09T07:39:30 ubuntu 10.04 64bit 2012-08-09T07:39:51 They are spec files for RPM based distros. 2012-08-09T07:40:11 You could use the SpecBuilder on ubuntu. 2012-08-09T07:40:12 I know that but would like to compile it myself 2012-08-09T07:40:30 http://git.rtems.org/rtems-tools.git/ 2012-08-09T07:40:53 Check that package out and search the wiki for SpecBuilder 2012-08-09T07:40:55 ok. let me clone it 2012-08-09T07:41:37 It is a python based package that can read the crossrpm spec files and build the tools. I use it on MacOS and FreeBSD and I have some Windows patches 2012-08-09T07:47:33 ok. thanks it does something. I will read it 2012-08-09T08:02:01 *** weiY has joined #rtems 2012-08-09T08:05:58 Ok. 2012-08-09T08:06:08 I am about to head off 2012-08-09T08:08:08 hi chris, good night 2012-08-09T08:08:16 hi 2012-08-09T08:08:49 Please check the IRC logs. Seb and I had a chat about the atomics layering. It is still not clear to me 2012-08-09T08:12:43 ok, i will see the irc logs 2012-08-09T08:13:14 *** DrJoel has joined #rtems 2012-08-09T08:13:15 *** DrJoel has joined #rtems 2012-08-09T08:13:15 *** ChanServ sets mode: +o DrJoel 2012-08-09T08:27:49 Hi sebhub, good night. are you around? 2012-08-09T08:28:05 What chris said about the devel post is http://www.rtems.org/pipermail/rtems-devel/2012-August/001559.html 2012-08-09T08:34:02 *** gedare has joined #rtems 2012-08-09T08:34:40 gedare, hi 2012-08-09T08:34:49 hey kiwichris been awhile 2012-08-09T08:34:57 you're up early/late 2012-08-09T08:35:12 yeah, rather late here. Watch grown men leaping over rather high bars at the moment 2012-08-09T08:35:18 ahhh 2012-08-09T08:35:38 Did the 4.10 repo get cleaned up ? I thought I saw some changes go in. 2012-08-09T08:36:21 oh good point i should take a look... there were a couple of files i think similar to the ones that i manually reverted 2012-08-09T08:36:36 More than a few I think 2012-08-09T08:36:49 oh! your'e right 2012-08-09T08:37:06 Only sebhub committed something I think since your work 2012-08-09T08:37:13 that's going to get a harder axe then... 2012-08-09T08:37:44 some of the commits there are wrong too... 2012-08-09T08:38:00 gedare, I would like to do a coverage run, then do whatever (git pull) and do coverage on the pthread key work. We should be able to review that and merge it quickly. Not nearly as much code as it was hard algorithmic and analysis work. 2012-08-09T08:38:16 agreed 2012-08-09T08:39:41 weiY, my issue is the purpose of adding ISR enable/disable in the atomics. Yes in UP with interrupts it is how you lock but would we really use that in the score. 2012-08-09T08:39:55 i'm going to reset 4.10 to "bsp/genmcf548x: Enable FPU in BSP startup code" and add back "libblock: Fix purge device tree traversal" and "Use http://www.rtems.org/bugzilla as bug-URL" 2012-08-09T08:40:07 I am wondering if on UP and arch with no locking support we do not compile. 2012-08-09T08:43:04 I am off for the night, cya 2012-08-09T08:43:12 night chrijs. 2012-08-09T08:44:03 hi chris, i think add ISR enable/disable is only suitable for UP which is also lack of atomic operations 2012-08-09T08:44:40 his point is maybe code that uses atomic operations should fail to compile if processor support does not exist 2012-08-09T08:44:44 and this method is just a stage method 2012-08-09T08:44:52 (I think that is his point) 2012-08-09T08:45:37 ok, i understand 2012-08-09T08:46:28 but in that way the atomic operations will not be support for UP architecture which is lack of atomic instructions 2012-08-09T08:52:50 DrJoel: did you / can you send meeting notice? 2012-08-09T08:57:52 https://docs.google.com/document/d/1nsUvrGyRJfUVKf4tLL2PkR-jgD2cne4DempIecCtwb8/edit 2012-08-09T08:57:58 that should be the weekly report 2012-08-09T08:58:11 * DrJoel grimaces.. construction guys drilling into metal .... 2012-08-09T08:58:15 DrJoel: permissions 2012-08-09T08:58:20 hi 2012-08-09T08:58:25 hello sebhub 2012-08-09T08:58:43 done.. thanks gedare.. why that doesn't stick on a copy I don't know 2012-08-09T08:58:48 4.10 just got a forced push reset... you may wish to update your repos 2012-08-09T08:59:23 *** tuxmaniac has joined #rtems 2012-08-09T08:59:23 did my student show up last week in the irc? 2012-08-09T08:59:39 hmmm 2012-08-09T08:59:45 let me check 2012-08-09T08:59:58 http://www.facebook.com/MorpheusLander?ref=ts is doing a test today. Hopefully one day it will get to the moon 2012-08-09T09:00:07 *** zw_yao has joined #rtems 2012-08-09T09:00:12 sebhub: it looks like he did 2012-08-09T09:00:30 oh yes.. he said basically he was going to be implementing clock-pro 2012-08-09T09:00:55 and he had a pointer to some linux code for it.. i was concerned with the complexity of the linux code to be studying 2012-08-09T09:01:20 *** zw_yao has quit IRC (Remote host closed the connection) 2012-08-09T09:01:23 it wasn't clear to me why he wanted to implement "clock-pro 2" rather than "clock-pro" 2012-08-09T09:01:49 its not clear to me on what he works at all 2012-08-09T09:01:59 there are only two weeks left? 2012-08-09T09:02:21 yes.. and really monday is supposed to be soft pencils down 2012-08-09T09:02:34 that is mainly to encourage students to get code submitted 2012-08-09T09:03:02 that's why I want to review, do coverage, and merge the pthread key work. It should be very close. 2012-08-09T09:04:04 *** zw_yao has joined #rtems 2012-08-09T09:04:30 i am not sure if the bdbuf project completes successfully 2012-08-09T09:04:41 *** JenniferA has joined #rtems 2012-08-09T09:05:04 understood.. we'll evaluate it in 2 weeks.. anything that might be done to save it yet 2012-08-09T09:05:15 other than personal heroics 2012-08-09T09:05:20 hey JenniferA .. carpet is outside on a truck.. good day for you to be at home.. electrician is back and drilling into the metal studs.. 2012-08-09T09:05:26 (which are certainly a valid option, but unlikely I suppose) 2012-08-09T09:05:49 the libbsd project is in the same state 2012-08-09T09:06:14 *** zw_yao has quit IRC (Remote host closed the connection) 2012-08-09T09:06:17 *** monstr_ has quit IRC (Ping timeout: 245 seconds) 2012-08-09T09:06:21 Good Morning, glad I'm not there in the chaos 2012-08-09T09:07:33 Ron is actually here today .. lol 2012-08-09T09:09:16 *** zw_yao has joined #rtems 2012-08-09T09:15:16 *** xiangfu has quit IRC (Ping timeout: 244 seconds) 2012-08-09T09:15:17 *** zw_yao has quit IRC (Remote host closed the connection) 2012-08-09T09:27:31 *** cdcs has quit IRC (Ping timeout: 245 seconds) 2012-08-09T09:27:33 *** xiangfu has joined #rtems 2012-08-09T09:27:55 *** alseh has joined #rtems 2012-08-09T09:37:55 *** zw_yao has joined #rtems 2012-08-09T09:39:20 *** WikL has joined #rtems 2012-08-09T09:46:02 I was wondering if we should like to consider providing release snapshots to facilitate RTEMS beginners and users. Snapshots can include fully generated sources, and perhaps pre-built manuals. 2012-08-09T09:46:34 yes this is a good idea 2012-08-09T09:46:47 we should also release rtems 4.11 soon 2012-08-09T09:46:59 at least before more smp work starts 2012-08-09T09:47:14 good idea 2012-08-09T09:47:36 i am quite satisfied with the current master 2012-08-09T09:49:45 *** claas_ziemke has joined #rtems 2012-08-09T09:52:41 maybe DrJoel can make another call for feature requests or freeze 2012-08-09T09:53:00 *** weiY has quit IRC (Ping timeout: 260 seconds) 2012-08-09T09:53:03 but i agree... the master has been static for awhile now 2012-08-09T09:53:05 good time to cut 2012-08-09T09:53:22 it is time to do a release BUT.. 2012-08-09T09:53:25 * DrJoel begins to rant 2012-08-09T09:53:32 lol 2012-08-09T09:53:44 The conversion to git broke all the release scripts and procedures 2012-08-09T09:54:12 oh 2012-08-09T09:54:40 I have asked for help from the author to update Makefile.maint in the rtems git module but they appear to have more pressing things to do. You can check the author history on that 2012-08-09T09:55:56 For the other git modules, I just need some help in knowing what the right procedure to tag, export, etc. like we did for cvs. The other git modules use a release script I wrote years ago and used to use for the rtems module. Until Makefile.maint was pushed. 2012-08-09T09:56:11 * DrJoel picks up a clue bat. :) 2012-08-09T09:56:15 I need some help :) 2012-08-09T09:56:21 * DrJoel puts down the clue bat 2012-08-09T09:56:53 I think the next 4.9 and 4.10 minor releases are overdue. We were releasing minor releases pretty frequently before git conversion. 2012-08-09T09:57:41 4.11 needs to head to feature freeze and out the door. It is probably mostly up to a few people to let me know if their app is happy with 4.11's state. And their pet RTEMS project ready for prime-time 2012-08-09T09:57:47 * DrJoel steps down.. 2012-08-09T09:58:23 sebhub is going on holiday.... 2012-08-09T09:58:42 *** weiY has joined #rtems 2012-08-09T09:59:00 gedare, can you help me with the release procedure update to git? We can work through the issues as we put out 4.9.x and 4.10.x, then work on pushing for 4.11 to branch 2012-08-09T09:59:21 maybe. i'm T minus 3 weeks and sans 1 thesis. 2012-08-09T10:00:11 ok.. how about I put together a google doc with the cvs commands and steps used to make releases. Then we can slowly work through the git version of the procedure 2012-08-09T10:00:13 no idea what my time is going to be like after the next 3 weeks go by. 2012-08-09T10:00:17 and do the releases when you are ready 2012-08-09T10:00:18 * tuxmaniac slowly rises his tired hands and waves. Grad school is tiring :-) 2012-08-09T10:00:20 yeah i should be able to chip away 2012-08-09T10:00:33 heh. 2012-08-09T10:00:35 But I won't do anything "real" until you are ready. 2012-08-09T10:00:40 let's start this meeting 2012-08-09T10:00:46 * DrJoel remembers finishing dissertation with 4 young kids.. 2012-08-09T10:00:51 who's up? 2012-08-09T10:00:52 where did i put that docu 2012-08-09T10:00:56 i usually go by the order in the doc 2012-08-09T10:01:05 * tuxmaniac bows to DrJoel (for dissertaition with 4 kids) 2012-08-09T10:01:06 as listed by attending 2012-08-09T10:01:15 so it's me, is it? 2012-08-09T10:01:30 ok WikL 2012-08-09T10:01:36 saw a lot of updates but little progress 2012-08-09T10:01:52 for any other students: https://docs.google.com/document/d/1nsUvrGyRJfUVKf4tLL2PkR-jgD2cne4DempIecCtwb8/edit 2012-08-09T10:01:53 yeah, got stuck for a while with the interrupt thing 2012-08-09T10:01:58 still stuck on it? 2012-08-09T10:02:03 first I was looking for the bug, but later the concept was reviewed 2012-08-09T10:02:15 a new approach has appeared 2012-08-09T10:02:18 ok 2012-08-09T10:02:37 well, not exactly appeared - it was arrived at during a conversation on the ml 2012-08-09T10:02:49 hopefully you can get it resolved... once you have a clock tick "working" you at least have a functional bsp to submit 2012-08-09T10:03:14 anyway, I'm clear about the concept and I like it since it can be used with other interrupts as well 2012-08-09T10:03:19 ok good 2012-08-09T10:03:27 i may have missed the details but that is ok 2012-08-09T10:03:34 if you and your mentors feel good about it then go to it 2012-08-09T10:03:53 I'm going to submit a specification for it 2012-08-09T10:03:56 great 2012-08-09T10:04:07 for your project we need some good documentation to go with the 'bsp' 2012-08-09T10:04:11 and later figure out some implementation details specific to the x86 2012-08-09T10:04:14 hmm 2012-08-09T10:04:24 what will that include? 2012-08-09T10:04:28 that can come later 2012-08-09T10:04:35 ok 2012-08-09T10:04:37 *** cdcs has joined #rtems 2012-08-09T10:04:39 get your code in working order first 2012-08-09T10:04:48 but keep track of what you do and why 2012-08-09T10:04:51 sure 2012-08-09T10:04:51 so that you can report it later 2012-08-09T10:04:57 for any who follow :) 2012-08-09T10:05:06 and keep an eye out for any "pok"isms 2012-08-09T10:05:14 all right.. anything else? 2012-08-09T10:05:31 not now 2012-08-09T10:05:33 make sure clock tick is propagted to all partitions.. also please write up the design of this when it works. :) It was hard to get there :) 2012-08-09T10:06:06 I have an idea for the interface, but have to make sure if it can be used for multiple partitions of RTEMS at once 2012-08-09T10:06:06 *** soh_cah_toa has joined #rtems 2012-08-09T10:06:23 I know you mentioned how important it is before, but somehow slipped my mind recently 2012-08-09T10:06:24 https://docs.google.com/document/d/1nsUvrGyRJfUVKf4tLL2PkR-jgD2cne4DempIecCtwb8/edit 2012-08-09T10:06:51 so yeah, that's it from me I guesss 2012-08-09T10:06:51 ok. well that is pok's problem though 2012-08-09T10:06:58 or, it should be 2012-08-09T10:07:10 great. 2012-08-09T10:07:15 yeah, *should be* 2012-08-09T10:07:25 weiY: you're up 2012-08-09T10:07:30 ok 2012-08-09T10:07:51 i have add the atomic support for PowerPC architecture 2012-08-09T10:07:53 I have seen your discussion, and kiwichris concern about the UP implementation of "atomic" 2012-08-09T10:08:31 you shall need to determine whether it is appropriate to use ISR disable/enable to implement atomic. it does seem to be a little complex. 2012-08-09T10:08:36 we discussed this on irc today 2012-08-09T10:08:38 And misleading for developers 2012-08-09T10:08:46 it should be ok now 2012-08-09T10:08:47 ok good. send the resolution to rtems-devel ml 2012-08-09T10:08:48 yeah, about the kiwichris concern the add ISR able/disable is just for compatible atomic for UP and SMP 2012-08-09T10:09:05 did you test atomic on ppc? 2012-08-09T10:09:26 test = run ;) 2012-08-09T10:09:28 yeah, i have added seven test case for UP also 2012-08-09T10:09:31 ok great 2012-08-09T10:09:37 push a branch and make a pull request 2012-08-09T10:09:48 what's next? 2012-08-09T10:10:04 and compiler PSIM BSP and run all the seven UP test case and all pass successully 2012-08-09T10:10:29 ok , i will push a clean new branch and make pull request 2012-08-09T10:11:15 next is to make sych primitives on RTEMS SMP safely 2012-08-09T10:11:26 but this work need more discussion 2012-08-09T10:11:39 ok. start that ml thread too 2012-08-09T10:11:55 OK 2012-08-09T10:12:23 it seems no other updates 2012-08-09T10:12:33 again.. try to document the design decisions .. these tradeoffs and decisions get lost over time 2012-08-09T10:12:33 ok 2012-08-09T10:12:41 next up: zw_yao 2012-08-09T10:12:51 Yeah, I'm ready. 2012-08-09T10:13:06 super: go 2012-08-09T10:13:09 ok, next week i will try to add all the design document to wiki 2012-08-09T10:13:15 thanks weiY 2012-08-09T10:13:20 as soon as possible 2012-08-09T10:14:13 Ok, My status: 2012-08-09T10:15:00 post the summary of the pull request and discussed the alternatives on ml 2012-08-09T10:15:44 It seems that other approach is not worth to implement... 2012-08-09T10:15:53 ok. DrJoel is hoping to do a base coverage run and then pull your key's work and test that. 2012-08-09T10:16:02 do you have something else to code? 2012-08-09T10:16:10 * soh_cah_toa sneaks in quietly 2012-08-09T10:16:23 soh_cah_toa: https://docs.google.com/document/d/1nsUvrGyRJfUVKf4tLL2PkR-jgD2cne4DempIecCtwb8/edit 2012-08-09T10:16:30 I think no... 2012-08-09T10:16:53 ok you can start on documentation 2012-08-09T10:16:54 And I found a serious bug today. 2012-08-09T10:16:57 uh ohs. 2012-08-09T10:17:01 and fixing that bug :) 2012-08-09T10:17:20 Haven't fixed yet, but should be fixed soon I think. 2012-08-09T10:18:26 So, I have only documentation to do next? 2012-08-09T10:18:33 please describe the bug in reply to your pull request 2012-08-09T10:18:38 so that we do not pull your code until you fix the bug 2012-08-09T10:18:54 sure, I will. 2012-08-09T10:19:33 it seems nothing else to report. 2012-08-09T10:19:41 That's all:) 2012-08-09T10:20:00 the other thing to look at with your project is whether there is some way to implement the per-task memory in supercore in order to replace classic task variables 2012-08-09T10:20:44 and then sharing the score implementation between classic task vars and posix key manager. that is something you can investigate a little bit 2012-08-09T10:20:55 perhaps sketch a design for it. 2012-08-09T10:20:58 thanks zw_yao 2012-08-09T10:21:09 next up: claas_ziemke 2012-08-09T10:21:29 OK, I've read classic task variables's code, that seems not as complex as POSIX key:) 2012-08-09T10:21:41 I'll take a deep look at it. 2012-08-09T10:21:51 alright... 2012-08-09T10:22:10 first what i did: 2012-08-09T10:22:47 someone asked me this week if your project will be / is supporting beaglebone. 2012-08-09T10:22:53 as stated in the last irc i attended (2 weeks ago) i was on a short vacation, where i did not have interneet, so sorry for not attending last week... 2012-08-09T10:23:11 anyway i worked on the i2c driver and the rtc driver 2012-08-09T10:23:41 the i2c driver is basically working now, i still have to integrate it whith the rtems/libi2c 2012-08-09T10:24:27 the rtc is working and i cann access the registers where the time is stored, but i still have to implement the rtc driver 2012-08-09T10:24:43 thats what i did last weeek... 2012-08-09T10:24:52 now for the plans for next week: 2012-08-09T10:25:34 first i will integrate the i2c driver in the rtems/libi2c 2012-08-09T10:25:41 next i will code the rtc driver 2012-08-09T10:26:23 is i2c prerequisite to rtc? 2012-08-09T10:26:28 the question is shall i write it in the bsp older first or do a driver for the TPS65950 chip in the shared folder? 2012-08-09T10:26:50 yes, the TPS65950 companion chip of the OMAP35xx is accessible only through i2c 2012-08-09T10:26:59 i would suggest you get the bsp working first 2012-08-09T10:27:07 it hosts the rtc, the power managment the audio the usb otg etc... 2012-08-09T10:27:27 ok, i will keep my code in the bsp folder then 2012-08-09T10:27:28 less struggling with being generic or dealing with build system infrastructure quite so much 2012-08-09T10:27:34 you can even keep your i2c in the bsp for now 2012-08-09T10:27:40 i would urge to get the rtc in the bsp 2012-08-09T10:27:46 just to get something working 2012-08-09T10:27:58 because you are running out of time to get something that is submittable 2012-08-09T10:28:14 then after gsoc you can make it more general if you would like to 2012-08-09T10:28:20 ok 2012-08-09T10:28:24 does that make sense? 2012-08-09T10:28:29 it sure does 2012-08-09T10:28:37 the libi2c is quite hard to understand 2012-08-09T10:29:20 ok. anything else for you? 2012-08-09T10:29:21 and as i stated before i am VERY willing to do more coding/clean up after gsoc, so that seems a reasonable plan: implement more features now in the bsp and clean up/make generic after gsoc... 2012-08-09T10:29:27 did i get that right?\ 2012-08-09T10:29:27 great 2012-08-09T10:29:34 yes 2012-08-09T10:29:40 sebhub: i know, i read into it... 2012-08-09T10:29:43 try to get clock working on your bsp 2012-08-09T10:29:53 that should give you a bsp that can demo ticker at minimum 2012-08-09T10:29:54 so i will get the rtc to work next... then libi2c later 2012-08-09T10:30:07 (you did get hello world already?)_ 2012-08-09T10:30:08 gedar: thats the plan 2012-08-09T10:30:24 YES, UART IS WORKING FLAWLESSLY 2012-08-09T10:30:28 ok good :) 2012-08-09T10:30:31 ups sorry for the caps 2012-08-09T10:30:42 next up: alseh 2012-08-09T10:30:52 hi all 2012-08-09T10:30:55 so then thats all, i will update the googledoc, but now its not working.. 2012-08-09T10:31:03 thanks claas_ziemke 2012-08-09T10:31:16 my status is almost finished on changes for RTEMS Graphics Toolkit 2012-08-09T10:31:33 the bugs that I detected, are fixed 2012-08-09T10:31:39 yay! 2012-08-09T10:31:42 keyboard works? 2012-08-09T10:31:52 yes :) 2012-08-09T10:31:54 yay! 2012-08-09T10:32:08 it was a little problem that nxlib was expecting scancode 2012-08-09T10:32:23 and rtems pc_kbd driver provides keysym 2012-08-09T10:32:25 prepare some pull requests? 2012-08-09T10:32:54 right now I am cleaning NXlib, and it's the last one I clean for now 2012-08-09T10:32:57 ok 2012-08-09T10:33:52 one question I had, is why the zlib implementation in rtems have z_functionname instead of functionname ? 2012-08-09T10:34:25 it's so that external zlib can be added ? 2012-08-09T10:34:45 no, to avoid namespace pollution 2012-08-09T10:35:38 every global symbol is visible in an rtems application, so we need to be careful about what global symbols are defined / used to avoid conflicts with user code 2012-08-09T10:35:39 because fltk had a problem on detecting it and provided it's own .... now it's ok ... changed detection to z_function 2012-08-09T10:35:57 that's what I thought :) 2012-08-09T10:36:00 ok, libpng had the same problem 2012-08-09T10:36:19 it would be good to document this in whatever notes you have about udpating these libraries 2012-08-09T10:36:25 its really nice that fltk works now 2012-08-09T10:36:30 there is where the problem appears :) 2012-08-09T10:36:39 yes, and almost all the tests are building 2012-08-09T10:36:56 is there any advantage on porting the fluid toolkit also from fltk ? 2012-08-09T10:37:18 (it's the tool used to build GUI graphically ) 2012-08-09T10:37:45 hm, to run it on rtems? 2012-08-09T10:37:57 i don't want to develop on rtems 2012-08-09T10:38:09 lol 2012-08-09T10:38:29 just asking because is included in fltk :) ... 2012-08-09T10:38:37 but did not compile it 2012-08-09T10:38:59 ok.. what's next? 2012-08-09T10:39:14 so, I think next is cirrus driver 2012-08-09T10:39:20 and VBE after that 2012-08-09T10:39:24 ok 2012-08-09T10:39:44 I had no success on running on skyeye .... someone on maillist provided a patch that did not help :( 2012-08-09T10:39:50 hmm 2012-08-09T10:40:00 the gumstix pathc? 2012-08-09T10:40:05 yes 2012-08-09T10:40:19 * zw_yao says good night and is going to bed... 2012-08-09T10:40:20 you might try older versions of skyeye with older patches ;) 2012-08-09T10:40:37 i personally don't have experience with it so i cannot give input 2012-08-09T10:40:50 *** cdcs has quit IRC (Ping timeout: 245 seconds) 2012-08-09T10:40:58 anyway as long as you have some target that runs that is good 2012-08-09T10:41:14 ok.. good for now? 2012-08-09T10:41:29 yes ... cirrus driver in the next days :) 2012-08-09T10:41:29 alseh, should we start working through merging your library updates in the graphics toolkit? One at a time. :) 2012-08-09T10:41:32 thank you 2012-08-09T10:41:35 next up: soh_cah_toa JenniferA DrJoel 2012-08-09T10:42:15 he's finishing cleaning them up.. i don't mind taking these as a batch 2012-08-09T10:42:15 k 2012-08-09T10:42:18 since they are separate 2012-08-09T10:42:41 i'd rather see more progress on the fb before "pencils down" 2012-08-09T10:42:51 thanks alseh 2012-08-09T10:42:55 hi soh_cah_toa 2012-08-09T10:42:59 on the fb I want to go no 2012-08-09T10:43:02 *now 2012-08-09T10:43:12 the toolkit is updated on my github 2012-08-09T10:43:38 before i start, i literally have to walk to the end of the bay to get a moderate wifi connection. i'm lagging pretty bad here (2.1 seconds) so my responses may be a bit delayed. everything's kinda coming in waves 2012-08-09T10:43:56 but aside from my terrible connection, i've had some good progress and a few patches to show for it 2012-08-09T10:44:35 so far, i've got patches for av5282, csb360, gen68302, gen68360, and i'll have genmcf548x in 2012-08-09T10:44:38 just a few minutes 2012-08-09T10:44:56 any other architectures done? Or just plowing through m68k? 2012-08-09T10:45:28 i have mips done but i have to wait until next week when i'm back from vacation to get to my desktop machine 2012-08-09T10:45:36 so yeah, i'm just plowing through m68k right now 2012-08-09T10:45:48 gedare: thanks for asking about the bone, i talked to him :) 2012-08-09T10:46:03 i would've had more but that stupid libmpc and autoconf/automake 2.69 issue gave me a serious headache for a few days. which btw, after gsoc when i'm not crunched for time, i'll definitely help get that centos image up to 2012-08-09T10:46:04 mikeg: great 2012-08-09T10:46:05 date 2012-08-09T10:47:07 now that i kinda understand what's going on, i can bang them out pretty quickly. i've been adding the .robsdsets section and also cleaning them up a bit since a lot of them were near cryptic 2012-08-09T10:47:20 I am in the process of updating it this week. Email me privately and I can send you the instructions. Tools added a dependency on another repository to pick up libmpc.so.2 2012-08-09T10:47:21 i'll definitely have m68k done when i can submit my mips patches 2012-08-09T10:47:49 awesome :) 2012-08-09T10:48:28 ok.. i386 is only one bsp and i did that. 2012-08-09T10:48:40 how many BSPs does that leave? 2012-08-09T10:49:27 k 2012-08-09T10:49:54 18 2012-08-09T10:50:17 but if they're anything like the previous ones, that won't take long at all 2012-08-09T10:51:35 while i have a connection now and can get to my email, i'll submit the patches to the ml after the meeting 2012-08-09T10:52:10 ok... anything else? 2012-08-09T10:52:44 18 m68k or total? I was askign across all architectures. 2012-08-09T10:53:07 about the ro/rw rtems linkcmds files. i'm starting to see which parts are redundant in a lot of bsps but i wanna get all these bsps passing the testsuite first. just to get done what i said i was gonna get done 2012-08-09T10:53:21 process of updating it this week. Email me privately and I can send you the instructions. Tools added a dependency on another repository to pick up libmpc.so.2 2012-08-09T10:53:24 oops 2012-08-09T10:53:27 *** xiangfu has quit IRC (Ping timeout: 240 seconds) 2012-08-09T10:53:31 sorry 2012-08-09T10:53:39 oh, i meant 18 m68k left to do 2012-08-09T10:54:03 i'll update the spreadsheet i made as well 2012-08-09T10:54:05 I would like to see all BSPs updated, then a patch to make that piece shared. Plow through them quickly 2012-08-09T10:54:59 anything else? Please hurry.. you got off to a slow start but can still finish all bsps 2012-08-09T10:55:48 yeah, i know. :( it's getting easier now 2012-08-09T10:56:25 you can see how these got cut and pasted over the years to their current state 2012-08-09T10:56:33 and why we want to get them to a shared state :) 2012-08-09T10:56:57 definitely 2012-08-09T10:57:27 where should the common linkcmds file go? 2012-08-09T10:57:40 like, within the libbsp directory 2012-08-09T10:58:34 "shared" subdirectory 2012-08-09T10:58:40 if shared only with a single cpu architecture, the shared for that cpu 2012-08-09T10:58:46 the farthest common ancestor that makes sense 2012-08-09T10:58:52 if across 2 or more.. then higher shared 2012-08-09T10:58:55 what gedare said 2012-08-09T10:59:49 ok, that was my first guess but wasn't sure since it looks a bunch of c files 2012-08-09T11:00:04 arm and sparc have linkcmds in them 2012-08-09T11:00:10 not many converted to this yet 2012-08-09T11:00:26 * DrJoel notes that the carpet glue smell is starting to get to him :( 2012-08-09T11:01:01 ewww... :P 2012-08-09T11:01:13 anything else? 2012-08-09T11:01:55 meeting adjourned! 2012-08-09T11:02:15 i'll be hanging around here a bit 2012-08-09T11:03:10 anyone know..is there an environment variable that GCC reads to get extra include paths? 2012-08-09T11:03:20 please start getting us code to review 2012-08-09T11:03:35 other than jamming them into $CFLAGS 2012-08-09T11:03:57 GCC_EXEC_PREFIX has an impact on that but could break things 2012-08-09T11:04:37 not too many calls to getenv in the source 2012-08-09T11:05:04 re 2012-08-09T11:05:10 so then i'll just keeping doing what i've been doing and crank up the speed. i'll email the patches i have now in a minute. update my spreadsheet and the gsoc 2012-08-09T11:05:10 status report 2012-08-09T11:05:10 and... 2012-08-09T11:05:10 i'll start seeing how i can refactor the redundant parts for .robsdsets 2012-08-09T11:05:10 and the other rtems parts 2012-08-09T11:05:13 and i guess that's it\nthen 2012-08-09T11:05:14 sorry, I was not able to come before 2012-08-09T11:05:22 WikL: are you still here ? 2012-08-09T11:08:50 agh, sorry. lag just shot up to 20 seconds :O 2012-08-09T11:12:15 *** JenniferA has quit IRC (Quit: Page closed) 2012-08-09T11:15:35 soh_cah_toa, work first on getting all BSPs done and a patch for that. Refactoring can continue 2012-08-09T11:16:51 *** weiY has quit IRC (Ping timeout: 272 seconds) 2012-08-09T11:17:54 *** soh_cah_toa has quit IRC (Ping timeout: 276 seconds) 2012-08-09T11:21:33 *** soh_cah_toa has joined #rtems 2012-08-09T11:27:40 *** sebhub has quit IRC (Remote host closed the connection) 2012-08-09T11:27:40 *** zw_yao has quit IRC (Remote host closed the connection) 2012-08-09T11:34:29 *** soh_cah_toa has quit IRC (Quit: [(--)]ZZzzz...) 2012-08-09T11:53:44 *** claas_ziemke has left #rtems 2012-08-09T11:54:15 *** DrJoel has quit IRC (Quit: Leaving) 2012-08-09T12:19:02 *** monstr_ has joined #rtems 2012-08-09T12:49:19 *** alseh has quit IRC (Quit: Leaving) 2012-08-09T13:12:54 *** beng-nl has quit IRC (Ping timeout: 244 seconds) 2012-08-09T13:23:14 *** gedare has quit IRC (Quit: Leaving) 2012-08-09T14:18:27 *** monstr_ has quit IRC (Ping timeout: 268 seconds) 2012-08-09T14:46:17 *** WikL has quit IRC (Remote host closed the connection) 2012-08-09T15:40:41 *** beng-nl has joined #rtems 2012-08-09T17:10:52 *** peerst has quit IRC (Ping timeout: 246 seconds) 2012-08-09T17:38:27 *** deb has joined #rtems 2012-08-09T17:46:01 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2012-08-09T18:20:38 *** kiwichris has joined #rtems 2012-08-09T18:54:29 *** deb has quit IRC (Quit: Leaving) 2012-08-09T19:32:58 *** peerst has joined #rtems 2012-08-09T19:34:45 *** xiangfu has joined #rtems 2012-08-09T20:49:18 *** peerst has quit IRC (Ping timeout: 245 seconds) 2012-08-09T21:01:01 *** peerst has joined #rtems 2012-08-09T22:15:19 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2012-08-09T22:34:44 *** kiwichris has joined #rtems 2012-08-09T23:26:23 *** kiwichris has quit IRC (Read error: Connection reset by peer) 2012-08-09T23:27:20 *** kiwichris has joined #rtems 2012-08-09T23:49:11 *** monstr_ has joined #rtems 2012-08-10T00:09:17 *** alseh has joined #rtems 2012-08-10T00:35:54 *** krzysiekm13 has joined #rtems 2012-08-10T02:37:59 *** WikL has joined #rtems 2012-08-10T03:10:08 *** sebhub has joined #rtems 2012-08-10T03:11:14 good morning 2012-08-10T04:13:51 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-10T04:26:00 *** xiangfu has joined #rtems 2012-08-10T04:31:40 *** QingPei has joined #rtems 2012-08-10T05:15:41 hi, I'm trying to do coverage analysis for erc32 bsp. I wonder if it is possible using gdb-sis or only with tsim? 2012-08-10T05:17:25 i think this only works with tsim 2012-08-10T05:17:36 lo 2012-08-10T05:47:58 *** bacon has joined #rtems 2012-08-10T05:48:21 *** bacon is now known as Guest34512 2012-08-10T05:49:18 *** Guest34512 has quit IRC (Client Quit) 2012-08-10T05:49:56 *** BaconXu has joined #rtems 2012-08-10T05:51:45 Hi, Masters, Does RTEMS support MMU? 2012-08-10T05:52:46 what do you mean with "support MMU"? 2012-08-10T05:54:24 To protect the kernel, and user application, separate them. 2012-08-10T05:54:52 no, rtems is a single address space system 2012-08-10T05:55:13 Ok, I see, thank you. 2012-08-10T05:59:53 *** BaconXu has left #rtems 2012-08-10T06:03:50 *** BaconXu has joined #rtems 2012-08-10T06:05:37 *** BaconXu has left #rtems 2012-08-10T06:30:44 do you know if tsim is available for free? according to the gaisler's website it isn't unfortunately.. 2012-08-10T06:36:57 krzysiekm13, that is something you need ask them 2012-08-10T08:05:31 *** alseh has quit IRC (Remote host closed the connection) 2012-08-10T08:27:17 *** monstr_ has quit IRC (Quit: Leaving) 2012-08-10T08:32:54 *** WikL has quit IRC (Remote host closed the connection) 2012-08-10T08:49:27 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-10T09:14:43 *** xiangfu has joined #rtems 2012-08-10T09:18:47 *** xiangfu has quit IRC (Ping timeout: 240 seconds) 2012-08-10T10:11:08 *** gedare has joined #rtems 2012-08-10T10:28:35 *** sebhub has quit IRC (Remote host closed the connection) 2012-08-10T11:02:54 *** kristianpaul has quit IRC (Ping timeout: 248 seconds) 2012-08-10T11:14:14 *** krzysiekm13 has left #rtems 2012-08-10T11:17:29 *** kristianpaul has joined #rtems 2012-08-10T11:17:29 *** kristianpaul has joined #rtems 2012-08-10T11:31:06 *** BaconXu has joined #rtems 2012-08-10T11:32:16 *** BaconXu has quit IRC (Client Quit) 2012-08-10T12:30:43 *** gedare has quit IRC (Quit: Ex-Chat) 2012-08-10T12:56:31 *** tuxmaniac has quit IRC (Quit: Lost terminal) 2012-08-10T14:53:26 *** krzysiekm13 has joined #rtems 2012-08-10T15:04:11 *** gedare has joined #rtems 2012-08-10T15:06:47 *** Hesham has joined #rtems 2012-08-10T15:19:39 hi Hesham 2012-08-10T15:19:46 hi gedare 2012-08-10T15:20:20 how are you? 2012-08-10T15:20:25 good. yourself? 2012-08-10T15:20:32 fine 2012-08-10T15:20:37 as you saw.. i finally got a chance to actually run your code ;) 2012-08-10T15:20:57 yes i noticed that great action ;) 2012-08-10T15:21:07 well the amalgam of aanjhan's, quanming's, yours and my own 2012-08-10T15:21:32 do you understand my request for different tests? 2012-08-10T15:21:52 i do not think so 2012-08-10T15:22:07 what i got is making one test for "passed tests" 2012-08-10T15:22:18 and another one for "failures" 2012-08-10T15:22:21 i want tests to terminate 2012-08-10T15:22:30 at least ones that we know should terminate 2012-08-10T15:22:48 the test right now hangs, so that makes it harder to "test" 2012-08-10T15:23:07 i see 2012-08-10T15:23:17 so make the conditions that cause hanging be in a different test 2012-08-10T15:23:24 which will be more easily separated 2012-08-10T15:23:34 but quanming write it to raise exceptions and never terminates 2012-08-10T15:23:36 then i can run all the tests that i expect will terminate successfully 2012-08-10T15:23:42 yeah that is not right to do :) 2012-08-10T15:24:21 i do not know why it hangs, i tried to debug that issue, i found it's not entering exception handler 2012-08-10T15:24:48 i asked aanjhan about that, he told me maybe it's stuck to infinite loop 2012-08-10T15:24:49 that is odd, but can be dealt with later.. the cases that cause exceptions should be made separate 2012-08-10T15:25:11 you mean separate test case ? 2012-08-10T15:25:20 that will also reduce the possible sources of the bug . yes separate test cases 2012-08-10T15:25:58 ok, i will make one for high-level/Mid level changes that does not deal with exceptions 2012-08-10T15:26:05 and another one that does 2012-08-10T15:26:12 *** antgreen has joined #rtems 2012-08-10T15:26:21 mainly the high-level one will be portable 2012-08-10T15:26:47 and the second one will vary for each target 2012-08-10T15:26:58 because of address mappings and such issues 2012-08-10T15:27:03 is that make sense ? 2012-08-10T15:27:53 yes 2012-08-10T15:28:11 that will be helpful 2012-08-10T15:28:11 why do you think the programs hangs ? 2012-08-10T15:28:23 don't know 2012-08-10T15:30:16 at first i was hoping that the program hangs and do not reach "this line should never be printed". After i refactored the work, i am hoping to know why the program hangs and do not print this line :) 2012-08-10T15:30:30 anyway, i have some questions 2012-08-10T15:31:02 oki 2012-08-10T15:31:22 if the linker script define a variable, i.e pagetable base address..... 2012-08-10T15:31:57 _ttbl_base = ORIGIN (SDRAM_MMU); 2012-08-10T15:32:02 how should c code reference it ? 2012-08-10T15:32:29 i have write extern uint32_t _ttbl_base; 2012-08-10T15:32:47 "extern uint32_t _ttbl_base;" 2012-08-10T15:33:20 _ttbl_base is not a pointer right? 2012-08-10T15:33:53 usually i've seen it done like... 2012-08-10T15:34:05 extern char _ttbl_base[]; 2012-08-10T15:34:40 uintptr_t _ttbl = (uintptr_t) _ttbl_base; 2012-08-10T15:35:06 int base = *(_ttbl); 2012-08-10T15:35:09 or something like that 2012-08-10T15:35:20 i don't know whwat is the best proper way 2012-08-10T15:35:43 also the linkcmds have to export that symbol... like 2012-08-10T15:35:46 *** krzysiekm13 has left #rtems 2012-08-10T15:35:57 PROVIDE ( _ttbl_base = ORIGIN(SDRAM_MMU)); 2012-08-10T15:36:09 yes i agree, but what is odd is the following code ( which is the existing mmu code for arm ) 2012-08-10T15:36:11 PROVIDE tells linker to put the symbol in global sym table 2012-08-10T15:36:22 lvl1_base = (mmu_lvl1_t *)&_ttbl_base; 2012-08-10T15:36:56 does that make sense ? 2012-08-10T15:37:14 i mean specially this part "&_ttbl_base;" 2012-08-10T15:37:18 _ttbl_base is a symbol 2012-08-10T15:37:26 so it's saying to figure out what the address for that symbol is 2012-08-10T15:37:36 and to make a pointer from that 2012-08-10T15:37:51 then you can dereference that address 2012-08-10T15:37:55 to get the value 2012-08-10T15:38:05 i think! 2012-08-10T15:38:15 that's my best guess anyways 2012-08-10T15:38:20 the code does not do so 2012-08-10T15:38:29 what is lvl1_base? 2012-08-10T15:38:50 it must be an mmu_lvl1_t* 2012-08-10T15:39:09 it's saying.... 2012-08-10T15:39:27 a pointer to level1 page table base address 2012-08-10T15:39:30 an mmu_lvl1_t structure starts at &_ttbl_base 2012-08-10T15:40:07 err 2012-08-10T15:40:08 sorry 2012-08-10T15:40:17 an mmu_lvl1_t structure starts at _ttbl_base 2012-08-10T15:40:39 so you want a pointer to the mmu_lvl1_t, so you find what is the address of _ttbl_base 2012-08-10T15:41:11 then you have a pointer to a struct mmu_lvl1_t 2012-08-10T15:41:21 is that any more clear? 2012-08-10T15:42:00 i think that line should be like that 2012-08-10T15:42:03 lvl1_base = (mmu_lvl1_t *) _ttbl_base; 2012-08-10T15:42:59 not lvl1_base should contains _ttbl_base ( which contains memory address for level 1 page table base) 2012-08-10T15:43:22 lvl1_base should contains _ttbl_base ( which contains memory address for level 1 page table base) 2012-08-10T15:43:48 right ? 2012-08-10T15:45:50 does _ttbl_base have memory address for lvl1 page table base, or is _ttbl_base the actual start of the structure? 2012-08-10T15:46:34 _ttbl_base is defined at linker script to have the memory address for lvl1 page table base 2012-08-10T15:46:47 MEMORY { 2012-08-10T15:46:48 SDRAM_MMU : ORIGIN = 0x20100000, LENGTH = 16k 2012-08-10T15:46:48 SDRAM : ORIGIN = 0x20104000, LENGTH = 63M - 16k 2012-08-10T15:46:48 SRAM : ORIGIN = 0x00200000, LENGTH = 16k 2012-08-10T15:46:48 NIRVANA : ORIGIN = 0, LENGTH = 0 2012-08-10T15:46:48 } 2012-08-10T15:46:53 _ttbl_base = ORIGIN (SDRAM_MMU); 2012-08-10T15:46:59 i think _ttbl_base is the starting address of the lvl1 page table 2012-08-10T15:47:06 this is quoted from linker script 2012-08-10T15:47:10 yes it is 2012-08-10T15:47:25 So, &_ttbl_base is a pointer to the lvl1 page table 2012-08-10T15:47:57 and _ttbl_base interprets the first word at that location as an address 2012-08-10T15:48:07 you want a pointer to the lvl1 page table 2012-08-10T15:48:26 not to interpret the word at the _ttbl_base address as a pointer 2012-08-10T15:49:13 say you have... void *_ttbl_base = 0x20100000; 2012-08-10T15:49:52 if you do... mmu_lvl1_t *table = (mmu_lvl1_t*)_ttbl_base; 2012-08-10T15:50:18 then *table = *0x20100000; 2012-08-10T15:50:34 which does not point to the lvl1 page table... 2012-08-10T15:50:42 it probably points to some random location in ram 2012-08-10T15:50:51 you mean that when typing "_ttbl_base = ORIGIN (SDRAM_MMU);" at linker script, that corresponds to typing that in .C file " void *_ttbl_base = 0x20100000;" ?! 2012-08-10T15:50:57 if you do mmu_lvl1_t *table = (mmu_lvl1_t*) &_ttbl_base; 2012-08-10T15:51:04 then *table = 0x20100000; 2012-08-10T15:51:17 basically 2012-08-10T15:51:42 it's setting the symbol named _ttbl_base to be equal to the address of SDRAM_MMU 2012-08-10T15:53:15 that means _ttbl_base is not a pointer a void pointer as you stated here "say you have... void *_ttbl_base = 0x20100000;" 2012-08-10T15:55:30 ? 2012-08-10T15:56:48 the linker creates a symbol labelled _ttbl_base, places the symbol in the executable, and puts the value 0x20100000 at the location of that symbol 2012-08-10T15:56:50 ok what i got, is a symbol defined at linker script "_ttbl_base = ORIGIN (SDRAM_MMU);" which contains the base address of a pagetable 2012-08-10T15:57:40 extern uint32_t _ttbl_base; at .c file 2012-08-10T15:58:22 if i referenced _ttbl_base value it should be 0x20100000, right ? 2012-08-10T15:58:35 if you do... printf("%x\n", (uint32_t)_ttbl_base); you should get 0x20100000 2012-08-10T15:58:43 yes 2012-08-10T15:58:59 hmm 2012-08-10T15:59:14 if i want to make a pointer from that address it should be "(mmu_lvl1_t *)_ttbl_base;" 2012-08-10T15:59:24 mmu_lvl1_t is just uint32_t 2012-08-10T16:00:35 :q 2012-08-10T16:00:37 yeah 2012-08-10T16:01:21 well that means 2012-08-10T16:01:22 "lvl1_base = (mmu_lvl1_t *)&_ttbl_base;" is wrong. 2012-08-10T16:03:44 maybe. i feel like i'm missing something though 2012-08-10T16:04:12 if you compile an application, and do objdump -d on the binary, see if there is a value given for _ttbl_base 2012-08-10T16:05:07 ok 2012-08-10T16:05:27 i think the existing code for mmu is not used/tested till now 2012-08-10T16:05:58 maybe... if so you should lay this out clearly to the ml 2012-08-10T16:06:08 i think something is missing though 2012-08-10T16:06:57 *** antgreen has quit IRC (Read error: Connection reset by peer) 2012-08-10T16:08:00 if you are not busy you may tell what's missing 2012-08-10T16:08:28 can you see what is _ttbl_base from objdump? 2012-08-10T16:08:36 or add some printk code to print the value 2012-08-10T16:08:48 to verify what you think 2012-08-10T16:10:15 i think no functions at mmu.c code is called to check for that value, so i will try to compile it and run objdump 2012-08-10T16:10:49 hmm if the code is never called then it won't be in objdump... 2012-08-10T16:11:00 but that would make it dead code which would surprise me. 2012-08-10T16:11:24 i think it's 2012-08-10T16:11:36 csb337/start/start.S 2012-08-10T16:11:36 86: bl mmu_init 2012-08-10T16:11:36 gp32/start/start.S 2012-08-10T16:11:36 139: bl mmu_init 2012-08-10T16:11:36 csb336/start/start.S 2012-08-10T16:11:37 93: bl mmu_init 2012-08-10T16:11:39 gumstix/start/start.S 2012-08-10T16:11:41 86: bl mmu_init 2012-08-10T16:12:01 looks like it gets called during boot 2012-08-10T16:12:07 :q 2012-08-10T16:12:24 i gotta quit doin that. 2012-08-10T16:13:14 i will check it 2012-08-10T16:14:43 it's either the case that _ttbl_base = 0x20100000, or *_ttbl_base = 0x20100000 2012-08-10T16:15:03 if it is the latter, then it makes sense to use &_ttbl_base 2012-08-10T16:21:16 _ttbl_base should exist when i run "arm-rtems4.11-objdump -d hello.exe" right ? 2012-08-10T16:21:33 yeah i think so. 2012-08-10T16:21:40 redirect the output to a file 2012-08-10T16:21:52 i did 2012-08-10T16:22:05 and i searched for it and i got no result ! 2012-08-10T16:22:13 what about mmu_init 2012-08-10T16:22:22 it exist 2012-08-10T16:22:29 ok. might not be dumping all symbols 2012-08-10T16:22:50 i will do printk 2012-08-10T16:23:00 ok 2012-08-10T16:23:15 otherwise try objdump -D 2012-08-10T16:23:18 it prints more 2012-08-10T16:23:42 i think you need -D to see data sections 2012-08-10T16:25:06 i did but i have no result for _ttbl_base also 2012-08-10T16:25:50 ok 2012-08-10T16:32:19 i think i understand... the mmu lvl1 table is passed around by address as a uint32_t. the trick is the code uses an array 2012-08-10T16:32:33 so it needs to get a base address for the array 2012-08-10T16:33:47 lvl1_base = (mmu_lvl1_t *) &_tttbl_base; 2012-08-10T16:33:59 lvl1_base[0] = 0x20100000 2012-08-10T16:34:24 lvl1_base[1] = 0x20100004 2012-08-10T16:34:31 ... 2012-08-10T16:34:48 i think this might be what is happening. 2012-08-10T16:35:27 i do not think so 2012-08-10T16:35:30 oh ok 2012-08-10T16:35:34 because later in the code 2012-08-10T16:35:49 it assign PTEs pattern to lvl1_base[vbase] 2012-08-10T16:36:32 which means lvl1_base is not array of addresses, it's array for level-1 PTEs 2012-08-10T16:36:39 ah. 2012-08-10T16:36:59 er, &lvl1_base[0] = 0x20100000 ? 2012-08-10T16:37:09 that should be 2012-08-10T16:37:15 that makes more sense anyway 2012-08-10T16:37:33 yes 2012-08-10T16:37:51 and that is what it means... to do lvl1_base = (mmu_lvl1_t *) &_ttbl_base 2012-08-10T16:38:06 maybe :P 2012-08-10T16:38:09 but that line "lvl1_base = (mmu_lvl1_t *)&_ttbl_base;" violates that logic 2012-08-10T16:38:10 you can verify by printk 2012-08-10T16:38:16 i did 2012-08-10T16:38:21 ok what did you print 2012-08-10T16:38:40 printk(("%x\n", (uint32_t)_ttbl_base); 2012-08-10T16:39:04 but the problem is of simulators to show that value now :P 2012-08-10T16:39:16 lol 2012-08-10T16:39:39 ok get your simulator working. 2012-08-10T16:41:27 anyway i intended to write mmu cp-15 header file handling for pxa250 processor ( which is embedded at gumstix) and apply xiang patched for skyeye. I consider that as a future work or something. 2012-08-10T16:41:49 well... we need to test your arm work somehow or it cannot be accepted 2012-08-10T16:42:49 does gdbarmsim not work? 2012-08-10T16:42:57 well, that way i think i will begin to apply patches for gumstix skyeye hope it works 2012-08-10T16:43:01 ok 2012-08-10T16:43:09 long weekend ahead of you i guess? 2012-08-10T16:44:48 I do not have weekends, it was because of traveling 2012-08-10T16:45:36 ok 2012-08-10T16:45:50 gdbarmsim is not working 2012-08-10T16:46:00 is that a known problem? 2012-08-10T16:46:22 (gdb) run 2012-08-10T16:46:23 Starting program: /home/hesham/csb337/arm-rtems4.11/c/csb337/testsuites/samples/hello/hello.exe 2012-08-10T16:46:45 that's what i always get when trying any script at sim-script 2012-08-10T16:46:58 that's the case for all arm BSPs 2012-08-10T16:47:46 and the program stuck at this phase 2012-08-10T16:49:46 Are there anything has an importance over writing cp-15 for pxa250 now ? 2012-08-10T16:50:16 nothing immediate comes to mind 2012-08-10T16:50:25 oh, splitting up the test cases 2012-08-10T16:51:51 yes, nothing else? 2012-08-10T16:52:53 i can't think of anything, can you? 2012-08-10T16:54:54 just one function implementation for arm920 code 2012-08-10T17:01:30 ok... good luck 2012-08-10T17:03:24 thanks gedare 2012-08-10T17:44:37 *** lcpfnvcy has quit IRC (Ping timeout: 244 seconds) 2012-08-10T17:55:03 gedare: gumstix calls mmu_init, but i do not see where exactly this function is implemented for pxa255 or for gumstix in general. 2012-08-10T18:12:35 *** Hesham has quit IRC (Ping timeout: 260 seconds) 2012-08-10T18:16:37 *** Hesham has joined #rtems 2012-08-10T18:31:24 *** Hesham1 has joined #rtems 2012-08-10T18:33:09 *** Hesham has quit IRC (Ping timeout: 248 seconds) 2012-08-10T18:33:31 *** Hesham has joined #rtems 2012-08-10T18:35:51 *** Hesham1 has quit IRC (Ping timeout: 245 seconds) 2012-08-10T18:47:44 *** Hesham has quit IRC (Quit: Leaving.) 2012-08-10T20:54:54 *** kristianpaul has quit IRC (Ping timeout: 248 seconds) 2012-08-10T21:52:37 *** gedare has quit IRC (Quit: Leaving) 2012-08-10T23:31:02 *** kristianpaul has joined #rtems 2012-08-10T23:31:02 *** kristianpaul has joined #rtems 2012-08-11T00:11:05 *** kuzew has quit IRC (Read error: Operation timed out) 2012-08-11T00:13:18 *** kuzew has joined #rtems 2012-08-11T00:28:20 *** arvind_khadri has joined #rtems 2012-08-11T00:32:37 *** arvind_khadri has quit IRC (Ping timeout: 240 seconds) 2012-08-11T00:33:55 *** arvind_khadri has joined #rtems 2012-08-11T00:36:58 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-11T00:37:25 *** arvind_khadri has joined #rtems 2012-08-11T00:41:48 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2012-08-11T01:02:07 *** arvind_khadri has joined #rtems 2012-08-11T01:23:11 *** arvind_khadri has quit IRC (Ping timeout: 272 seconds) 2012-08-11T01:23:16 *** krzysiekm13 has joined #rtems 2012-08-11T01:56:04 *** arvind_khadri has joined #rtems 2012-08-11T02:01:10 *** arvind_khadri has quit IRC (Ping timeout: 244 seconds) 2012-08-11T02:05:26 *** arvind_khadri has joined #rtems 2012-08-11T02:06:23 *** arvind_khadri has joined #rtems 2012-08-11T02:07:13 *** arvind_khadri has quit IRC (Max SendQ exceeded) 2012-08-11T02:07:32 *** arvind_khadri has joined #rtems 2012-08-11T02:51:08 *** lcpfnvcy has joined #rtems 2012-08-11T03:24:37 *** arvind_khadri has quit IRC (Ping timeout: 268 seconds) 2012-08-11T03:36:25 *** arvind_khadri has joined #rtems 2012-08-11T03:51:14 *** arvind_khadri has quit IRC (Ping timeout: 260 seconds) 2012-08-11T04:25:47 *** arvind_khadri has joined #rtems 2012-08-11T04:40:28 *** arvind_khadri has quit IRC (Ping timeout: 268 seconds) 2012-08-11T05:02:37 *** xiangfu has joined #rtems 2012-08-11T05:09:44 *** arvind_khadri has joined #rtems 2012-08-11T05:10:27 *** WikL has joined #rtems 2012-08-11T05:27:17 *** arvind_khadri has quit IRC (Ping timeout: 244 seconds) 2012-08-11T06:26:11 *** alseh has joined #rtems 2012-08-11T07:06:18 hi, i'm trying to prepare environment to use rtems-testing with qemu for i386/pc386 2012-08-11T07:06:30 and i have some problems 2012-08-11T07:07:54 i have already installed qemu and it seems to working good 2012-08-11T07:17:11 but i don't understand what is doing pc386 script in rtems-testing/sim-scripts 2012-08-11T07:20:09 hi, I am using lately the pc386 script 2012-08-11T07:20:49 it copies the executable file to ~/qemu/hd/test.exe 2012-08-11T07:21:32 and runs qemu-laucher with a grub image that sends the command multiboot (hd0,0)/test.exe 2012-08-11T07:21:51 if you run the script verbose, you can see exactly what the comand is 2012-08-11T07:49:14 *** WikL has quit IRC (Ping timeout: 260 seconds) 2012-08-11T08:00:14 ok, thank you 2012-08-11T08:00:16 :) 2012-08-11T08:18:57 *** krzysiekm13 has left #rtems 2012-08-11T08:25:51 *** xiangfu has quit IRC (Ping timeout: 245 seconds) 2012-08-11T08:26:56 *** xiangfu has joined #rtems 2012-08-11T08:57:46 *** deb has joined #rtems 2012-08-11T09:52:54 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-11T09:54:06 *** QingPei has quit IRC (Remote host closed the connection) 2012-08-11T10:01:01 *** xiangfu has joined #rtems 2012-08-11T10:05:46 *** xiangfu has quit IRC (Ping timeout: 244 seconds) 2012-08-11T10:24:58 *** WikL has joined #rtems 2012-08-11T10:29:29 *** xiangfu has joined #rtems 2012-08-11T10:35:48 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-11T10:43:53 *** kuzew has quit IRC (Ping timeout: 245 seconds) 2012-08-11T10:47:09 *** xiangfu has joined #rtems 2012-08-11T10:54:00 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-11T11:13:55 *** Hesham has joined #rtems 2012-08-11T11:21:55 *** WikL has joined #rtems 2012-08-11T11:31:24 *** Hesham has quit IRC (Quit: Leaving.) 2012-08-11T11:38:03 *** Hesham has joined #rtems 2012-08-11T13:30:46 *** Amadiro has joined #rtems 2012-08-11T13:32:22 Evening. I'm interested in the RTEMS AVR port. The wiki indicates that some features are still missing. Which chips does it run on? How much memory and flash space does it minimally require? 2012-08-11T13:36:19 The wiki mentions the atmega128, which has more flash & sram than the 32u4 (which I'd like to target) but I could go with an AT90USB as well, which roughly matches the atmega128s capabilities 2012-08-11T14:05:02 *** krzysiekm13 has joined #rtems 2012-08-11T14:10:08 *** krzysiekm13 has left #rtems 2012-08-11T14:29:47 *** WikL has quit IRC (Ping timeout: 240 seconds) 2012-08-11T14:31:59 *** WikL has joined #rtems 2012-08-11T14:37:19 *** alseh has quit IRC (Remote host closed the connection) 2012-08-11T15:25:56 *** WikL has quit IRC (Remote host closed the connection) 2012-08-11T15:52:12 *** krzysiekm131 has joined #rtems 2012-08-11T15:53:28 *** krzysiekm13 has joined #rtems 2012-08-11T15:53:51 *** krzysiekm13 has left #rtems 2012-08-11T15:54:34 *** krzysiekm131 has left #rtems 2012-08-11T15:57:21 *** krzysiekm13 has joined #rtems 2012-08-11T15:57:42 *** krzysiekm13 has left #rtems 2012-08-11T16:15:27 *** krzysiekm13 has joined #rtems 2012-08-11T16:15:39 *** krzysiekm13 has left #rtems 2012-08-11T16:20:33 *** deb has quit IRC (Quit: This computer has gone to sleep) 2012-08-11T16:38:22 *** deb has joined #rtems 2012-08-11T18:07:38 *** dr__house has quit IRC (Read error: Operation timed out) 2012-08-11T18:10:40 *** dr__house has joined #rtems 2012-08-11T18:10:41 *** dr__house has joined #rtems 2012-08-11T19:32:53 *** kuzew_ has joined #rtems 2012-08-11T19:33:14 *** kuzew has joined #rtems 2012-08-11T19:35:19 *** jahf has quit IRC (Ping timeout: 240 seconds) 2012-08-11T19:36:19 *** jahf has joined #rtems 2012-08-11T19:53:28 *** Amadiro has quit IRC (Ping timeout: 245 seconds) 2012-08-11T20:47:06 *** Amadiro has joined #rtems 2012-08-11T20:49:04 *** xiangfu has joined #rtems 2012-08-11T21:13:51 *** xiangfu has quit IRC (Ping timeout: 244 seconds) 2012-08-11T21:40:14 *** xiangfu has joined #rtems 2012-08-11T21:44:43 *** xiangfu has quit IRC (Ping timeout: 245 seconds) 2012-08-11T21:47:40 *** xiangfu has joined #rtems 2012-08-11T22:39:27 *** kristianpaul has quit IRC (Quit: leaving) 2012-08-11T22:45:07 *** kristianpaul has joined #rtems 2012-08-11T22:45:07 *** kristianpaul has joined #rtems 2012-08-11T23:18:15 *** deb has left #rtems ("Leaving") 2012-08-11T23:49:18 *** xiangfu has quit IRC (Ping timeout: 248 seconds) 2012-08-12T00:24:40 *** xiangfu has joined #rtems 2012-08-12T00:29:45 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-12T00:41:25 *** xiangfu has joined #rtems 2012-08-12T01:05:28 *** WikL has joined #rtems 2012-08-12T03:16:48 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-12T03:57:50 *** xiangfu has joined #rtems 2012-08-12T04:02:57 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-12T04:06:01 *** xiangfu has joined #rtems 2012-08-12T04:07:47 *** WikL has quit IRC (Remote host closed the connection) 2012-08-12T08:18:10 *** kristianpaul has quit IRC (Quit: leaving) 2012-08-12T08:18:22 *** kristianpaul has joined #rtems 2012-08-12T08:18:28 *** kristianpaul has joined #rtems 2012-08-12T09:18:29 *** deb has joined #rtems 2012-08-12T10:29:11 *** xiangfu has quit IRC (Ping timeout: 245 seconds) 2012-08-12T10:41:37 *** QingPei has joined #rtems 2012-08-12T11:26:20 *** QingPei has left #rtems 2012-08-12T16:55:54 *** kristianpaul has quit IRC (Quit: leaving) 2012-08-12T16:56:06 *** kristianpaul has joined #rtems 2012-08-12T16:56:06 *** kristianpaul has joined #rtems 2012-08-12T17:40:55 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2012-08-12T18:29:47 *** kristian1aul has joined #rtems 2012-08-12T18:30:22 *** kristianpaul has quit IRC (Ping timeout: 248 seconds) 2012-08-12T18:36:14 *** kiwichris has joined #rtems 2012-08-12T18:51:42 *** Amadiro has quit IRC (Ping timeout: 248 seconds) 2012-08-12T19:13:52 *** Amadiro has joined #rtems 2012-08-12T19:17:18 *** xiangfu has joined #rtems 2012-08-12T19:56:57 *** kristian1aul has quit IRC (Quit: Reconnecting) 2012-08-12T19:57:06 *** kristianpaul has joined #rtems 2012-08-12T19:57:06 *** kristianpaul has joined #rtems 2012-08-12T20:32:54 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-08-12T20:42:51 *** xiangfu has joined #rtems 2012-08-12T21:13:41 *** kuzew has quit IRC (Quit: leaving) 2012-08-12T21:14:03 *** kuzew has joined #rtems 2012-08-12T23:22:46 *** deb has left #rtems ("Leaving") 2012-08-12T23:32:21 *** krzysiekm13 has joined #rtems