2011-02-28T00:26:30 *** projectgus has quit IRC 2011-02-28T01:12:27 *** sebhub has joined #rtems 2011-02-28T01:13:04 good morning 2011-02-28T01:43:59 *** core-ix has quit IRC 2011-02-28T02:18:46 *** bazinski has joined #rtems 2011-02-28T02:23:58 *** core-ix has joined #rtems 2011-02-28T02:44:01 *** arvind_khadri has joined #rtems 2011-02-28T04:03:27 *** arvind_k has joined #rtems 2011-02-28T04:03:59 *** arvind_khadri has quit IRC 2011-02-28T04:04:24 *** arvind_k has quit IRC 2011-02-28T04:06:17 *** arvind_khadri has joined #rtems 2011-02-28T06:15:58 *** core-ix has quit IRC 2011-02-28T06:19:03 *** core-ix has joined #rtems 2011-02-28T06:39:57 *** antgreen has joined #rtems 2011-02-28T06:53:05 *** core-ix has quit IRC 2011-02-28T07:09:36 *** core-ix has joined #rtems 2011-02-28T08:52:19 *** bazinski has quit IRC 2011-02-28T09:06:20 *** mmu_man has joined #rtems 2011-02-28T09:20:36 *** arvind_khadri has quit IRC 2011-02-28T09:49:18 *** arvind_khadri has joined #rtems 2011-02-28T10:47:54 * tuxmaniac notes that mentor organisations can start submitting applications from today 2011-02-28T10:48:01 * tuxmaniac needs to review the open ideas 2011-02-28T10:48:05 and see how much we can update 2011-02-28T10:54:54 *** dr__house has joined #rtems 2011-02-28T10:54:54 *** dr__house has joined #rtems 2011-02-28T11:07:29 *** sebhub has quit IRC 2011-02-28T11:14:01 *** dr__house has quit IRC 2011-02-28T11:14:05 *** dr__house` has joined #rtems 2011-02-28T11:14:05 *** dr__house` has joined #rtems 2011-02-28T11:30:20 *** dr__house` is now known as dr__house 2011-02-28T11:39:07 *** DrJoel has joined #rtems 2011-02-28T11:39:07 *** ChanServ sets mode: +o DrJoel 2011-02-28T11:39:58 * DrJoel waves hi 2011-02-28T11:52:09 DrJoel, hey :) 2011-02-28T11:57:44 hola 2011-02-28T12:01:36 hello.. back after a week helping my father-in-law recuperate from the flu 2011-02-28T12:02:03 today GSOC organization registration opens 2011-02-28T12:05:28 DrJoel, ohhh, you will bag it as always, btw you saw my mail? 2011-02-28T12:09:12 saw it .. haven't read it yet. Was very late (>2 hours) getting in this morning 2011-02-28T12:09:24 trying to catch up.. sigh 2011-02-28T12:11:44 *** madrazr1 has joined #rtems 2011-02-28T12:36:46 Oh :) 2011-02-28T12:39:03 *** core-ix has quit IRC 2011-02-28T12:39:45 *** madrazr1 is now known as madrazr 2011-02-28T12:39:52 *** madrazr has joined #rtems 2011-02-28T12:55:02 *** dr__house has quit IRC 2011-02-28T13:38:42 *** dr__house has joined #rtems 2011-02-28T13:38:47 *** dr__house has joined #rtems 2011-02-28T13:46:47 *** core-ix has joined #rtems 2011-02-28T14:10:26 *** dr__house` has joined #rtems 2011-02-28T14:10:26 *** dr__house has quit IRC 2011-02-28T14:11:03 *** dr__house` is now known as dr__house 2011-02-28T14:50:17 *** lcpfnyvc has quit IRC 2011-02-28T14:52:59 *** lcpfnyvc has joined #rtems 2011-02-28T15:07:37 *** arvind_khadri has quit IRC 2011-02-28T15:16:38 *** arvind_khadri has joined #rtems 2011-02-28T15:21:16 *** dr__house` has joined #rtems 2011-02-28T15:21:16 *** dr__house` has joined #rtems 2011-02-28T15:21:16 *** dr__house has quit IRC 2011-02-28T15:34:01 *** antgreen` has joined #rtems 2011-02-28T15:35:21 *** antgreen has quit IRC 2011-02-28T15:39:39 *** dr__house` is now known as dr__house 2011-02-28T15:54:21 *** arvind_khadri has quit IRC 2011-02-28T15:57:57 *** antgreen` has quit IRC 2011-02-28T15:59:47 *** projectgus has joined #rtems 2011-02-28T16:18:34 *** lekernel has quit IRC 2011-02-28T16:39:49 *** kiwichris has joined #rtems 2011-02-28T16:42:16 *** madrazr has left #rtems 2011-02-28T17:02:26 *** joel__ has joined #rtems 2011-02-28T17:08:00 *** core-ix has quit IRC 2011-02-28T17:08:01 *** DrJoel has quit IRC 2011-02-28T17:08:02 *** methril has quit IRC 2011-02-28T17:08:04 *** agrier has quit IRC 2011-02-28T17:08:19 *** core-ix has joined #rtems 2011-02-28T17:13:04 *** methril has joined #rtems 2011-02-28T17:33:38 *** joel__ has quit IRC 2011-02-28T18:09:34 * kiwichris wonders if Peter uses IRC 2011-02-28T18:26:24 *** dr__house has quit IRC 2011-02-28T19:25:56 *** dr__house has joined #rtems 2011-02-28T19:25:56 *** dr__house has joined #rtems 2011-02-28T19:52:36 *** dr__house has quit IRC 2011-02-28T19:52:49 *** dr__house has joined #rtems 2011-02-28T19:52:49 *** dr__house has joined #rtems 2011-02-28T20:29:12 *** dr__house has quit IRC 2011-02-28T21:06:10 *** dr__house has joined #rtems 2011-02-28T21:06:10 *** dr__house has joined #rtems 2011-02-28T21:20:21 *** rokka__ has quit IRC 2011-02-28T21:28:34 *** rokka_ has joined #rtems 2011-02-28T21:30:18 *** mmu_man has quit IRC 2011-02-28T21:58:11 *** kiwichris has quit IRC 2011-02-28T22:37:59 *** dr__house has quit IRC 2011-02-28T22:44:19 *** arvind_khadri has joined #rtems 2011-02-28T22:57:41 *** kiwichris has joined #rtems 2011-02-28T23:27:14 *** arvind_khadri has quit IRC 2011-03-01T00:17:09 *** projectgus has quit IRC 2011-03-01T00:25:19 *** arvind_khadri has joined #rtems 2011-03-01T00:29:10 *** arvind_khadri has quit IRC 2011-03-01T00:31:24 *** arvind_khadri has joined #rtems 2011-03-01T01:31:28 *** sebhub has joined #rtems 2011-03-01T01:31:33 good morning 2011-03-01T01:40:29 sebhub, morning, but its afternoon here :) 2011-03-01T02:15:56 *** arvind_khadri has quit IRC 2011-03-01T02:27:51 *** arvind_khadri has joined #rtems 2011-03-01T03:05:24 sebhub, hi 2011-03-01T03:12:36 hi chris 2011-03-01T03:12:48 Just catching up with the email. 2011-03-01T03:13:18 yes, i receive a lot of mails during the night ;-) 2011-03-01T03:13:22 When you say my changes are not right, is it due to 1), 2) or both ? 2011-03-01T03:13:43 1 and 2 are independent i think 2011-03-01T03:13:59 Agreed. 2011-03-01T03:14:08 I have not look at the resize yet. 2011-03-01T03:14:29 most of the _Heap_* functions assume that the arguments are valid 2011-03-01T03:14:33 On the sparc extend called free with a NULL. 2011-03-01T03:14:39 Yes. 2011-03-01T03:15:08 I think the object extend change came about with a clean up by Joel and I related to coverage. 2011-03-01T03:15:14 this is ok, we should allow a _Heap_Free(_Heap_Allocate(any parameter)) 2011-03-01T03:15:30 On the sparc the 0 undertolled to the faulting address 2011-03-01T03:15:43 i was a bit too fast with my removal of the NULL check in _Workspace_Free() 2011-03-01T03:16:04 Well it has brought some attention to the area :) 2011-03-01T03:16:23 but only if debugging is enabled i hope 2011-03-01T03:16:30 it was not that clear from the mails 2011-03-01T03:16:42 The protection check needs to occur after the valid check. 2011-03-01T03:17:18 Yes and if I cannot enable it to use because API mutex init fails then it serves no useful purpose. 2011-03-01T03:17:29 I would like it to be present and be useful. 2011-03-01T03:18:05 I have no problem with the protection check after the valid pointer check. 2011-03-01T03:18:12 yes, it is a bug in _Heap_Protection_block_check_default() 2011-03-01T03:18:37 Just looking at the code. 2011-03-01T03:18:45 no, the valid pointer check has to be after the protection, because the protection is more effective 2011-03-01T03:19:56 hm, maybe you are right, we may encounter an exception due to the pointer dereference 2011-03-01T03:21:02 That code is ok. 2011-03-01T03:21:02 one problem was, that the return value of _Heap_Free will be ignored in most cases 2011-03-01T03:21:38 if you hit an exception you notice the problem 2011-03-01T03:21:47 Yes which makes me wonder why we have a return value. Returning something which is of little use is not good practice 2011-03-01T03:22:30 I see this as a system level problem where you monitor the heap or we raise a fatal error. There is no other solution 2011-03-01T03:22:31 free() uses the return value, but the check in the non-debug version is so vague, it is of little use 2011-03-01T03:23:02 Is this the free printk ? 2011-03-01T03:23:09 yes 2011-03-01T03:23:40 Agreed it should go. It is a hack. We need to raise a fatal error and let the BSP either stop the target or hook the handler and deal with it. 2011-03-01T03:23:42 it is better than nothing 2011-03-01T03:23:58 It can in the heap code and controlled by the RTEMS debug flag. 2011-03-01T03:24:15 *** core-ix has quit IRC 2011-03-01T03:24:21 Spreading this around is part of the problem we have. 2011-03-01T03:24:24 we have here several issues at once ;-) 2011-03-01T03:24:50 Yes. We have the protection - block header on NULL. 2011-03-01T03:25:04 kiwichris, hii, how are you, long time :) 2011-03-01T03:25:08 We also have the changed API. If I may just move to that one. I see this as simple. 2011-03-01T03:25:21 arvind_khadri, no too bad. 2011-03-01T03:25:49 The extend in the C users manual needs to be altered if the API is ok with Joel 2011-03-01T03:26:08 kiwichris, so how is it going ? All set for GSoC ? 2011-03-01T03:26:10 We need to consider the effect of scattered regions on the in heap check. 2011-03-01T03:26:27 am sorry if am interrupting. 2011-03-01T03:26:46 arvind_khadri, Joel asked me to review the application so it is getting underway. 2011-03-01T03:27:32 Once we have this resolved we need to update the code to document what we are doing. 2011-03-01T03:28:06 This all means it is ok to have this if we can dot the 'i's and cross the 't's. 2011-03-01T03:28:07 currently if you have disconnected heap area and pass a pointer inside the gap, it will be accepted, but this is ok from my point of view 2011-03-01T03:28:16 you get a feature and pay for it 2011-03-01T03:28:45 Fine and this is the case you need to present and we discuss it and decide what it means. 2011-03-01T03:29:12 This gets the design reviewed and on the record and users know what is happening. 2011-03-01T03:29:40 The previous heap allocator was review by ESA and Joel updated the code to deal with the results of the review. 2011-03-01T03:30:45 Sure and I understand the use case or a fragmented heap. 2011-03-01T03:30:55 *** arvind_khadri has quit IRC 2011-03-01T03:33:23 ok, how do we continue now? for the heap extend we have a PR, this is ok 2011-03-01T03:33:37 one issue is the _Heap_Free(*, NULL) 2011-03-01T03:33:42 For example a rename of the _Heap_Is_block_in_heap to _Heap_Is_block_ptr_valid then some checks of the address for correct alignment catches a few more cases 2011-03-01T03:34:09 The issue is ? 2011-03-01T03:34:18 return true or false 2011-03-01T03:34:40 I say remove returning the bool and the problem goes away :) 2011-03-01T03:34:59 We need a better error management 2011-03-01T03:35:32 my intention for the heap protection was to keep the non-debug code simple and to provide a debug-version with a good check 2011-03-01T03:36:22 changing _Heap_Is_block_in_heap() into something more sophisticated will not yield much 2011-03-01T03:36:28 I think you have done this. I think following the malloc/free/realloc model has benefits 2011-03-01T03:36:58 Why not ? 2011-03-01T03:37:52 the real evil situation is a usage after free, and you will not catch this 2011-03-01T03:38:39 We cannot be expected to catch bugs in a production system. We should attempt to verify the parameters are correct 2011-03-01T03:38:59 This of course is a perform/overhead vs outcome balance. 2011-03-01T03:39:49 If heap is [start, end) and we know all blocks are aligned to uint32_t then only 1 in 4 addresses can be valid when passed to free 2011-03-01T03:41:31 *** arvind_khadri has joined #rtems 2011-03-01T03:41:42 yes, will be the parameters uniformly distributed? 2011-03-01T03:41:57 Sorry ? 2011-03-01T03:42:22 passing a word-aligned value seems to be more likely 2011-03-01T03:42:47 we can improve the check, but i think the current one is sufficient 2011-03-01T03:43:27 I think any check like this helps and costs little. 2011-03-01T03:44:29 ok, should we open a PR for the _Heap_Free()? 2011-03-01T03:45:29 I think the problem has been side stepped for now with 1746. The 1747 handles the API change and discontinuous regions. 2011-03-01T03:45:40 I say just score heap. 2011-03-01T03:45:57 I am just looking the score fatal error code 2011-03-01T03:46:19 These are not internal errors but fatal errors. 2011-03-01T03:47:18 what is the difference? 2011-03-01T03:47:57 internal would be fault in the heap allocator while a bad address to free is a fatal error 2011-03-01T03:48:27 The last means the system state is invalid 2011-03-01T03:48:50 yes, this is definitely the case 2011-03-01T03:51:12 We should call _User_extensions_Fatal 2011-03-01T03:52:15 We can call _Internal_error_Occurred with is_internal true or false depending on the nature of the error 2011-03-01T03:53:05 _Internal_error_Occurred() is the right function 2011-03-01T03:53:24 I will look at this tomorrow. 2011-03-01T03:53:37 ok 2011-03-01T03:54:10 We should fix the heap to use fatal error then we should deal with 1747 2011-03-01T03:55:37 good 2011-03-01T03:55:52 I will raise a PR when I have patch. 2011-03-01T03:56:30 i am out of office by next Monday 2011-03-01T03:56:40 Ok. 2011-03-01T03:57:02 The code is working for me which means I can work on run time link editor 2011-03-01T03:57:19 from 2. up 6. march 2011-03-01T03:57:30 Holiday ? 2011-03-01T03:57:33 yes 2011-03-01T03:57:44 Nice. Going away ? 2011-03-01T03:58:25 not far, but we have carnival 2011-03-01T03:58:42 or Fasching like its called in Bavaria 2011-03-01T03:59:04 Lovely. Have a relaxing break 2011-03-01T04:01:36 thanks 2011-03-01T04:03:23 *** core-ix has joined #rtems 2011-03-01T04:04:08 *** arvind_khadri has quit IRC 2011-03-01T04:17:06 *** arvind_khadri has joined #rtems 2011-03-01T05:56:50 *** mmu_man has joined #rtems 2011-03-01T06:44:34 *** antgreen has joined #rtems 2011-03-01T07:07:01 *** core-ix has quit IRC 2011-03-01T07:37:55 *** kornbluth has joined #rtems 2011-03-01T07:38:02 *** kornbluth has joined #rtems 2011-03-01T07:40:02 *** verm__ has quit IRC 2011-03-01T07:48:43 *** kornbluth is now known as verm__ 2011-03-01T07:48:49 *** verm__ has joined #rtems 2011-03-01T08:44:24 *** dr__house has joined #rtems 2011-03-01T08:44:24 *** dr__house has joined #rtems 2011-03-01T08:50:47 *** arvind_khadri has quit IRC 2011-03-01T09:54:53 *** DrJoel has joined #rtems 2011-03-01T09:54:53 *** ChanServ sets mode: +o DrJoel 2011-03-01T10:25:32 *** core-ix has joined #rtems 2011-03-01T10:45:38 *** sebhub has quit IRC 2011-03-01T11:03:56 *** dr__house has quit IRC 2011-03-01T11:04:02 *** dr__house has joined #rtems 2011-03-01T11:04:03 *** dr__house has joined #rtems 2011-03-01T11:31:02 *** core-ix has quit IRC 2011-03-01T13:31:59 *** agrier has joined #rtems 2011-03-01T13:41:12 *** methril has quit IRC 2011-03-01T13:53:17 *** methril has joined #rtems 2011-03-01T14:50:08 * kiwichris waves to Joel 2011-03-01T14:52:14 * DrJoel waves back.. 2011-03-01T14:52:19 feeling better about heap_free? 2011-03-01T14:57:44 Yes and no. Seb and I had a chat about it. We are thinking the return bool should go. If we get a failure then internal error handler should be called. 2011-03-01T14:59:53 that sounds reasonable. You can't do anythign about it anyway.. spwkspace already needs fixing 2011-03-01T15:00:47 What do you mean by spwkspace needs fixing ? 2011-03-01T15:04:43 Joel, there is ieee stuff in the FreeBSD powerpc directory. 2011-03-01T15:04:49 and e500 stuff 2011-03-01T15:05:14 spwkspace is broken by the recent mods. NULL passed in now returns a different value 2011-03-01T15:05:32 Oh I see the test. 2011-03-01T15:05:32 Oh I see the test is broken 2011-03-01T15:06:04 Post the link to the PR. I am only doing "good maintainership" but if that code is the answer, then maybe one of the interested parties can merge it. 2011-03-01T15:06:19 FWIW what happened to THomas and Sebastian's Thumb-2 multilibs? 2011-03-01T15:06:31 Sure. I have not looked at the code and do not know it is ok but they can. 2011-03-01T15:06:44 No idea 2011-03-01T15:07:03 that is more important (to me) to get dealt with since it is a barrier to the BeagleBoard. :) 2011-03-01T15:07:19 Do you have one ? 2011-03-01T15:08:06 Not yet but it is supported by Qemu and could be a GSoC project. It has 3 issues right now... tool support, cpukit and then the bsp itself 2011-03-01T15:08:24 Nice idea 2011-03-01T15:09:09 I tried to write up the FIle system test suite as a project idea last night 2011-03-01T15:14:42 The unlimited failure is on the clean up and not the create so this is when the memory is being released 2011-03-01T15:16:01 I still think it is a memory layout problem in the BSP.. the corruption of running is hit when freeing 2011-03-01T15:16:13 Yes I agree 2011-03-01T15:16:35 any idea how to (from CLI) get PR title give PR number? 2011-03-01T15:17:33 Does this help http://www.bugzilla.org/docs/2.16/html/cmdline.html 2011-03-01T15:17:53 reading it now.. i hope it helps 2011-03-01T15:18:42 tools listed are not in RPM. :( 2011-03-01T15:23:40 hmmm.. Python help request .. ping 2011-03-01T15:23:48 http://code.google.com/p/pybugz/downloads/detail?name=pybugz-0.9.0_rc1.tar.gz&can=2&q= 2011-03-01T15:24:04 looks promising but I am missing some python fu 2011-03-01T15:29:42 I will take a look today and send you an email. I am about to head out for a while 2011-03-01T15:31:09 Thanks.. now really at https://github.com/williamh/pybugz 2011-03-01T15:32:28 *** dr__house has quit IRC 2011-03-01T17:31:15 *** kiwichris has quit IRC 2011-03-01T17:33:35 *** core-ix has joined #rtems 2011-03-01T17:39:22 *** projectgus has joined #rtems 2011-03-01T18:06:00 *** kiwichris has joined #rtems 2011-03-01T18:08:08 kiwichris: just about to leave.. got pybugz to run but had to install stuff by hand to /usr/lib/python2.7/site-packages/ 2011-03-01T18:08:25 Oh that is not good. Is this on rtems.org ? 2011-03-01T18:08:49 no.. was doing it on my laptop. I was looking for something to help me write the PR lists at each release 2011-03-01T18:09:15 grep Changelog diff for PR numbers, then look them up with script from Bugzilla and write reports. :-D 2011-03-01T18:09:21 Ah ok. I suspect you should be able to fiddle an environment variable or option to add a path 2011-03-01T18:09:30 Nice 2011-03-01T18:09:33 I saw PYTHONPATH but never got that set right 2011-03-01T18:09:42 I used to do this with GNATS but have been doing it by hand. 2011-03-01T18:09:44 With removing BSPs from RTEMS do we have a page ? 2011-03-01T18:10:01 It should go in the 4.11 release notes and on each BSPs page 2011-03-01T18:10:25 It would be nice to see somewhere the date so you can get it back from CVS 2011-03-01T18:10:31 i386ex was for Erik Ivenanko? 2011-03-01T18:10:41 Ahh.. 2011-03-01T18:15:50 Now at least on the 4.11 release notes with 1 March 2011 2011-03-01T18:16:11 night 2011-03-01T18:16:14 *** DrJoel has quit IRC 2011-03-01T18:16:17 bye 2011-03-01T19:41:10 *** madrazr has joined #rtems 2011-03-01T20:44:05 *** arvind_khadri has joined #rtems 2011-03-01T20:44:23 *** madrazr has quit IRC 2011-03-01T21:35:48 *** arvind_khadri has quit IRC 2011-03-01T21:48:46 *** arvind_khadri has joined #rtems 2011-03-01T21:48:48 *** arvind_khadri has joined #rtems 2011-03-01T22:02:43 *** arvind_khadri has quit IRC 2011-03-01T23:24:18 *** kiwichris has quit IRC 2011-03-01T23:58:57 *** mmu_man has quit IRC 2011-03-02T00:11:11 *** projectgus has quit IRC 2011-03-02T00:52:05 *** kiwichris has joined #rtems 2011-03-02T01:10:15 *** core-ix has quit IRC 2011-03-02T02:04:23 *** verm__ has quit IRC 2011-03-02T02:25:55 *** arvind_khadri has joined #rtems 2011-03-02T02:49:39 *** lcpfnyvc has quit IRC 2011-03-02T02:50:05 *** lcpfnyvc has joined #rtems 2011-03-02T03:05:12 *** arvind_khadri has quit IRC 2011-03-02T04:08:57 *** arvind_khadri has joined #rtems 2011-03-02T04:12:40 *** peerst has joined #rtems 2011-03-02T04:12:51 hi 2011-03-02T05:47:32 *** core-ix has joined #rtems 2011-03-02T05:54:14 *** methril_work has joined #rtems 2011-03-02T07:01:22 *** verm__ has joined #rtems 2011-03-02T08:03:06 *** antgreen has quit IRC 2011-03-02T08:36:23 *** mmu_man_ has joined #rtems 2011-03-02T09:41:57 *** arvind_khadri has quit IRC 2011-03-02T11:36:18 *** core-ix has quit IRC 2011-03-02T12:35:31 *** mmu_man_ has quit IRC 2011-03-02T12:49:30 *** c0smikdebris has joined #rtems 2011-03-02T12:54:48 *** dr__house has joined #rtems 2011-03-02T13:32:59 *** kristianpaul has quit IRC 2011-03-02T13:36:53 *** madrazr has joined #rtems 2011-03-02T13:37:25 *** DrJoel has joined #rtems 2011-03-02T13:37:25 *** ChanServ sets mode: +o DrJoel 2011-03-02T14:20:45 *** lcpfnyvc has quit IRC 2011-03-02T14:58:28 *** mmu_man has joined #rtems 2011-03-02T15:06:59 *** madrazr has quit IRC 2011-03-02T15:40:26 *** core-ix has joined #rtems 2011-03-02T15:55:27 *** projectgus has joined #rtems 2011-03-02T16:04:54 *** bswimley has joined #rtems 2011-03-02T16:05:39 Dr Joel: Hello. 2011-03-02T16:07:37 hey bswimley .. 2011-03-02T16:08:23 * DrJoel has to run.. son needs to be picked up.. catch me tomorrow 2011-03-02T16:08:26 *** DrJoel has quit IRC 2011-03-02T16:40:18 *** c0smikdebris has quit IRC 2011-03-02T16:52:56 *** bswimley has quit IRC 2011-03-02T17:41:26 *** dr__house` has joined #rtems 2011-03-02T17:41:48 *** dr__house has quit IRC 2011-03-02T17:52:11 *** kiwichris has quit IRC 2011-03-02T18:19:20 *** kiwichris has joined #rtems 2011-03-02T19:27:09 *** dr__house` has quit IRC 2011-03-02T19:43:53 *** kristianpaul has joined #rtems 2011-03-02T21:40:58 *** arvind_khadri has joined #rtems 2011-03-02T21:54:34 *** kiwichris has quit IRC 2011-03-02T22:34:43 *** arvind_khadri has quit IRC 2011-03-02T22:35:52 *** kiwichris has joined #rtems 2011-03-02T22:40:35 *** mmu_man has quit IRC 2011-03-02T22:47:33 *** arvind_khadri has joined #rtems 2011-03-02T23:23:12 *** arvind_khadri has quit IRC 2011-03-02T23:35:01 *** arvind_khadri has joined #rtems 2011-03-02T23:43:36 *** lcpfnyvc has joined #rtems 2011-03-02T23:47:10 *** arvind_khadri has quit IRC 2011-03-03T00:07:55 *** projectgus has quit IRC 2011-03-03T01:15:56 *** arvind_khadri has joined #rtems 2011-03-03T04:04:35 *** arvind_khadri has quit IRC 2011-03-03T04:07:07 *** arvind_khadri has joined #rtems 2011-03-03T05:31:14 *** methril_work has quit IRC 2011-03-03T05:41:28 *** mmu_man has joined #rtems 2011-03-03T08:04:32 *** antgreen has joined #rtems 2011-03-03T08:23:33 *** arvind_khadri has quit IRC 2011-03-03T08:27:13 *** dr__house has joined #rtems 2011-03-03T08:27:13 *** dr__house has joined #rtems 2011-03-03T09:19:14 *** bswimley has joined #rtems 2011-03-03T10:19:35 *** arvind_khadri has joined #rtems 2011-03-03T11:39:36 *** dr__house` has joined #rtems 2011-03-03T11:39:36 *** dr__house` has joined #rtems 2011-03-03T11:39:45 *** dr__house has quit IRC 2011-03-03T11:39:55 *** dr__house` is now known as dr__house 2011-03-03T11:40:45 *** bswimley has quit IRC 2011-03-03T11:40:45 *** agrier has quit IRC 2011-03-03T11:49:19 *** bswimley has joined #rtems 2011-03-03T12:23:14 *** DrJoel has joined #rtems 2011-03-03T12:23:14 *** ChanServ sets mode: +o DrJoel 2011-03-03T12:34:57 *** arvind_khadri has quit IRC 2011-03-03T13:41:12 *** methril has quit IRC 2011-03-03T13:54:00 *** methril has joined #rtems 2011-03-03T14:12:28 Dr Joel: hello again! 2011-03-03T14:26:21 you have great timing... I was just heading out the door again 2011-03-03T14:26:25 how are you? 2011-03-03T14:26:33 Doing well. Thanks. 2011-03-03T14:26:44 I don't want to keep you. Do you have 3 minutes? 2011-03-03T14:26:49 Sure .. ask away 2011-03-03T14:27:36 Our company is looking at an ARM processor for a new product. Looking at Possibly the Cortex variants from NXP or Atmel. 2011-03-03T14:27:51 The ones with Thumb2? 2011-03-03T14:28:13 I *think* so - I'm just getting my feet wet in the ARM world. 2011-03-03T14:28:43 Just wondering about RTEMS support for the ARM at this point. Looks to be pretty good. 2011-03-03T14:29:40 Yes. The support is very good. The only issue I know is that we still need to add THumb2 support. Thomas/Sebastian have a tool patch to add the multilib and there is likely some tinkering but that's the biggest issue. Works well. 2011-03-03T14:30:03 http://www.rtems.org/ftp/pub/rtems/people/joel/coverage/ includes a couple of ARM BSPs :-D 2011-03-03T14:30:34 Good to know. My ideal scenario would be similar to how I'm developing for coldfire. MingGW with Eclipse and JTag Wiggler debugging. 2011-03-03T14:30:42 Under Windows. 2011-03-03T14:30:55 Thanks for the links - I will look at the BSPs! 2011-03-03T14:30:59 that's pretty common.. 2011-03-03T14:31:23 Excellent. I didn't see any red flags, but it never hurts to ask the experts :) 2011-03-03T14:31:26 If you have specific models, I can give more advice. The BeagleBoard is a good target. 2011-03-03T14:31:59 In the open source world, all you ever hear are "it's broken".. but in reality most projects aren't near as buggy as the email would indicate .. LOL 2011-03-03T14:31:59 OK, I will get back to you. Just wanted to make sure there weren't any roadblocks immediately thrown up. 2011-03-03T14:32:22 Yeah, people never seem to post "it works!" 2011-03-03T14:32:53 FYI I am soon to make an announcement of a new support web site. We have legacy releases now for 4.6, 4.7, and 4.8. After the open project drops them, we still track them internally and backport patches 2011-03-03T14:33:37 For example, the legacy support 4.6 branch has > 40 patches not in 4.6.6. So you don't have to fear not getting bug fixes on an older version 2011-03-03T14:34:02 OK. I'm actually moving to 4.11 at this point. Really like some of the new features. RFS is way cool. Chris was a lot of help in finding one last bug that was biting me in the RFS. 2011-03-03T14:34:54 The new shell is slick as well - and SD card support! 2011-03-03T14:35:26 :) A lot of the newer features really improve the experience.. :) 2011-03-03T14:35:29 * DrJoel waves bye 2011-03-03T14:35:31 *** DrJoel has quit IRC 2011-03-03T14:37:17 *** madrazr has joined #rtems 2011-03-03T15:02:00 *** dr__house has quit IRC 2011-03-03T15:02:15 *** bswimley has quit IRC 2011-03-03T15:21:15 *** madrazr has quit IRC 2011-03-03T15:39:47 *** kiwichris has quit IRC 2011-03-03T16:49:37 *** projectgus has joined #rtems 2011-03-03T16:54:16 *** kiwichris has joined #rtems 2011-03-03T20:44:31 *** dr__house has joined #rtems 2011-03-03T20:44:31 *** dr__house has joined #rtems 2011-03-03T20:53:07 *** methril has quit IRC 2011-03-03T21:28:41 *** kiwichris has quit IRC 2011-03-03T21:45:36 *** arvind_khadri has joined #rtems 2011-03-03T22:05:24 *** dr__house has quit IRC 2011-03-03T22:40:15 *** arvind_khadri has quit IRC 2011-03-03T23:01:09 *** kiwichris has joined #rtems 2011-03-03T23:31:27 *** mmu_man has quit IRC 2011-03-04T00:22:36 *** projectgus has quit IRC 2011-03-04T03:36:01 *** mmu_man has joined #rtems 2011-03-04T03:57:12 *** mmu_man has quit IRC 2011-03-04T04:20:40 *** mmu_man has joined #rtems 2011-03-04T04:24:00 *** core-ix has quit IRC 2011-03-04T04:38:54 *** methril has joined #rtems 2011-03-04T07:18:47 *** DrJoel has joined #rtems 2011-03-04T07:18:49 *** DrJoel has joined #rtems 2011-03-04T07:18:49 *** ChanServ sets mode: +o DrJoel 2011-03-04T07:19:37 * DrJoel waves hi 2011-03-04T07:38:34 hi! 2011-03-04T07:39:25 hello Fallenou 2011-03-04T07:41:22 helou 2011-03-04T07:43:14 if task with highest priority has while(1){sleep(1)} and it has an ASR should that task ASR handler be invoked soon after signal has been sent or only after sleep has completed 2011-03-04T07:44:26 and how should task modes set to this task and other tasks so that highest priority asr can be served as fast as possible 2011-03-04T07:47:43 im using arm7tdmi 2011-03-04T07:47:54 and rtems 4.10 2011-03-04T07:51:24 rokka_: version and cpu doesn't matter. If the signal is sent via Classic API (rtems_signal_send), then the task is NOT unblocked by a signal. If you want it to wake up via a signal, then you need to use POSIX signals. 2011-03-04T07:54:38 oh 2011-03-04T07:55:32 does it wake up from sleep? 2011-03-04T07:57:37 seems strange that it is not possible with classical api :l 2011-03-04T07:58:09 how does the posix api do it ? 2011-03-04T07:59:08 *** lcpfnyvc has quit IRC 2011-03-04T08:00:42 can task_restart called from a posix signal handler? 2011-03-04T08:03:44 That's just the way the Classic API was defined.. signals are different from POSIX 2011-03-04T08:04:11 yes on task restart.. don't know that there is an explicit test for it. If you want to submit an spXX test for that case, it would be appreciated. 2011-03-04T08:10:00 DrJoel: hi 2011-03-04T08:10:31 DrJoel: Can you point me to an example wich uses the region manager, i mean a common use, i guess before driver devolopment for debugging? or can be used un production? 2011-03-04T08:11:58 Second, about an Analog Driver, what should i consider if rtems will be reading about 2MSPS from a ADC, for later processing 2011-03-04T08:13:06 I mean what should i care about in the define in my rtems program so i dont get with a slow or bad perfomance system 2011-03-04T08:13:46 Of course the ADC will use DMA for sure. 2011-03-04T08:14:05 kristianpaul: can be used in production. Best use case I know of was of a satellite ground station. It got messages of varying sizes as the satellite was in view. Allocated memory for messages from Region, placed them on message queue for later processing and freeing. Coudl get up to 100K messages queued 2011-03-04T08:14:55 If you can rotate the DMA buffers, then fill one with N and send a message to a task to consume it. This lets the next buffer get full. 2011-03-04T08:15:02 I would lead to a set of N fixed size buffers 2011-03-04T08:15:55 At 2MSPS, you don't have much processing time per sample... what do you plan to do with them? 2011-03-04T08:16:27 Inially just save it to a flash memory 2011-03-04T08:16:36 Later the idea is implement some correlation system 2011-03-04T08:16:48 I'm working on a GPS/GNSS receiver 2011-03-04T08:17:13 so the 2MSPS are the I/Q data from the RF fronted 2011-03-04T08:21:36 the N buffering should work good for that. 2011-03-04T08:23:01 I have a sample app that does something similar if you are curious.. fakes reading a bunch of analogs into a buffer, then sends a message to another task which sends them via a socket.. similar but not exact 2011-03-04T08:27:04 can i send posix signal to a spesific task? 2011-03-04T08:27:07 how can i get pid for that task? 2011-03-04T08:27:51 getpid? 2011-03-04T08:28:59 pthread_self() in the task and save it in a variable is the official POSIX way 2011-03-04T08:29:11 but you can send a posix signal to a classic API task IF it enables signals 2011-03-04T08:31:24 sigaction = enabling? 2011-03-04T08:33:13 rokka_: yes.. plus maybe some sigmask .. 2011-03-04T08:34:08 there should be some psxsignal* tests which show how to do it 2011-03-04T08:35:41 *** dr__house has joined #rtems 2011-03-04T08:35:41 *** dr__house has joined #rtems 2011-03-04T08:35:55 hey dr__house .. how's school? 2011-03-04T08:36:16 DrJoel: hi :) School's hectic :P 2011-03-04T08:37:11 DrJoel: how are you doing? 2011-03-04T08:38:50 well.. kidney stuff is in remission with "normal" test results now. yeah! going to Spring Awakening broadway show tonight.. and hoping we get accepted in GSOC.. fingers crossed 2011-03-04T08:39:45 DrJoel: cool :). Good to see that you're getting well :) 2011-03-04T08:41:10 undefined reference to `siggetmask' 2011-03-04T08:41:14 undefined reference to `sigsetmask' 2011-03-04T08:41:15 :o 2011-03-04T08:41:39 did you --enable-posix? 2011-03-04T08:42:03 dr__house: I just hope I feel like being more active this spring/summer... finally went hiking a few weeks ago with Boy Scouts 2011-03-04T08:42:44 DrJoel: Cool! 2011-03-04T08:45:21 We went on a geocaching hike/contest on the Land Trust "Monte Sano Preserve" http://landtrustnal.org/trail-maps/ .. a cool thing here is that a state park and a lot of preserves are within a couple of miles of downtown 2011-03-04T08:47:13 DrJoel: (sample app) oh yes, is it in rtems repo? 2011-03-04T08:47:16 DrJoel: Seems like a great place :) 2011-03-04T08:49:00 kristianpaul: hmmm... probably not. I did it for a customer and have been meaning to turn it into a white paper 2011-03-04T08:49:19 email me and I will cut you a tarball 2011-03-04T08:50:03 ok 2011-03-04T09:06:59 *** mwalle has quit IRC 2011-03-04T09:18:49 DrJoel: i continue programmin next week .. thank you for help :) 2011-03-04T09:19:39 *** mmu_man_ has joined #rtems 2011-03-04T09:25:08 *** mmu_man has quit IRC 2011-03-04T09:26:56 *** mw1 has joined #rtems 2011-03-04T09:27:24 rokka_: no problem 2011-03-04T09:27:32 * DrJoel goes awk 2011-03-04T10:55:24 *** rtemsLogger` has joined #rtems 2011-03-04T10:56:30 *** rtemsLogger has quit IRC 2011-03-04T12:36:39 *** madrazr has joined #rtems 2011-03-04T12:37:04 *** mmu_man has quit IRC 2011-03-04T12:37:50 *** madrazr has quit IRC 2011-03-04T12:37:50 *** madrazr has joined #rtems 2011-03-04T12:38:12 *** mmu_man has joined #rtems 2011-03-04T12:57:40 *** dr__house has quit IRC 2011-03-04T13:23:23 *** arvind_khadri has joined #rtems 2011-03-04T14:34:08 *** rtemsLogger has joined #rtems 2011-03-04T14:34:21 *** rtemsLogger` has quit IRC 2011-03-04T14:55:12 *** kristianpaul has quit IRC 2011-03-04T15:02:03 *** kristianpaul has joined #rtems 2011-03-04T15:02:04 *** kristianpaul has joined #rtems 2011-03-04T16:00:01 *** DrJoel has quit IRC 2011-03-04T16:07:32 *** madrazr has left #rtems 2011-03-04T16:17:58 *** dr__house has joined #rtems 2011-03-04T16:17:58 *** dr__house has joined #rtems 2011-03-04T17:11:26 *** core-ix has joined #rtems 2011-03-04T17:44:05 *** kristian1aul has joined #rtems 2011-03-04T17:46:54 *** kristianpaul has quit IRC 2011-03-04T17:51:28 is there a memory allocation function in rtems kernel ? 2011-03-04T17:51:51 like kmalloc 2011-03-04T18:55:34 *** dr__house has quit IRC 2011-03-04T19:03:15 *** antgreen has quit IRC 2011-03-04T19:03:31 *** antgreen has joined #rtems 2011-03-04T20:27:11 *** agrier has joined #rtems 2011-03-04T21:00:00 *** methril has joined #rtems 2011-03-04T21:02:56 *** mmu_man has quit IRC 2011-03-04T21:30:46 *** dr__house has joined #rtems 2011-03-04T22:00:38 *** kiwichris has quit IRC 2011-03-04T22:05:45 *** kiwichris has joined #rtems 2011-03-04T23:34:29 *** dr__house has quit IRC 2011-03-05T00:46:26 *** peerst has quit IRC 2011-03-05T01:05:52 *** madrazr has joined #rtems 2011-03-05T01:38:37 *** mmu_man has joined #rtems 2011-03-05T01:40:15 *** madrazr has quit IRC 2011-03-05T06:31:19 *** arvind_khadri has quit IRC 2011-03-05T07:13:52 *** methril has quit IRC 2011-03-05T07:18:53 *** madrazr has joined #rtems 2011-03-05T07:19:23 *** methril has joined #rtems 2011-03-05T07:24:24 *** arvind_khadri has joined #rtems 2011-03-05T07:24:24 *** arvind_khadri has joined #rtems 2011-03-05T08:44:44 *** madrazr has quit IRC 2011-03-05T09:40:01 *** kristian1aul has quit IRC 2011-03-05T09:40:06 *** kristianpaul has joined #rtems 2011-03-05T09:40:06 *** kristianpaul has joined #rtems 2011-03-05T09:43:37 *** methril has quit IRC 2011-03-05T09:47:04 *** methril has joined #rtems 2011-03-05T09:48:42 *** methril has quit IRC 2011-03-05T09:51:38 *** methril has joined #rtems 2011-03-05T10:07:16 *** dr__house has joined #rtems 2011-03-05T11:00:41 *** arvind_khadri has quit IRC 2011-03-05T11:07:37 *** arvind_khadri has joined #rtems 2011-03-05T11:44:39 *** arvind_khadri has quit IRC 2011-03-05T11:46:54 *** arvind_khadri has joined #rtems 2011-03-05T11:53:35 *** madrazr has joined #rtems 2011-03-05T12:14:56 *** arvind_khadri has quit IRC 2011-03-05T12:59:53 *** dr__house` has joined #rtems 2011-03-05T12:59:53 *** dr__house` has joined #rtems 2011-03-05T12:59:56 *** dr__house has quit IRC 2011-03-05T13:17:47 *** madrazr has quit IRC 2011-03-05T13:41:47 *** methril has quit IRC 2011-03-05T13:53:43 *** methril has joined #rtems 2011-03-05T16:42:57 *** kiwichris has quit IRC 2011-03-05T17:38:09 *** dr__house` has quit IRC 2011-03-05T21:03:25 *** mmu_man has quit IRC 2011-03-05T21:35:00 *** arvind_khadri has joined #rtems 2011-03-05T21:41:52 *** dr__house has joined #rtems 2011-03-05T21:41:52 *** dr__house has joined #rtems 2011-03-05T23:29:56 *** dr__house has quit IRC 2011-03-06T00:00:21 *** lcpfnyvc has joined #rtems 2011-03-06T00:04:13 *** arvind_khadri has quit IRC 2011-03-06T00:17:18 *** arvind_khadri has joined #rtems 2011-03-06T00:17:18 *** arvind_khadri has joined #rtems 2011-03-06T00:48:12 *** madrazr has joined #rtems 2011-03-06T01:04:00 *** lcpfnyvc has quit IRC 2011-03-06T01:17:09 *** arvind_khadri has quit IRC 2011-03-06T01:28:47 *** arvind_khadri has joined #rtems 2011-03-06T01:28:47 *** arvind_khadri has joined #rtems 2011-03-06T01:34:34 *** antgreen has quit IRC 2011-03-06T03:19:09 *** madrazr has quit IRC 2011-03-06T03:34:22 *** madrazr has joined #rtems 2011-03-06T03:39:05 *** verm__ has quit IRC 2011-03-06T03:43:26 *** madrazr has quit IRC 2011-03-06T03:47:03 *** arvind_khadri has quit IRC 2011-03-06T04:00:13 *** arvind_khadri has joined #rtems 2011-03-06T04:22:34 *** verm__ has joined #rtems 2011-03-06T04:22:41 *** verm__ has joined #rtems 2011-03-06T04:58:35 *** mmu_man has joined #rtems 2011-03-06T06:42:09 *** core-ix has quit IRC 2011-03-06T07:41:47 how about virtualized rtems i386 ? 2011-03-06T07:41:50 with RT-XEN 2011-03-06T07:41:57 https://sites.google.com/site/realtimexen/home 2011-03-06T08:57:00 *** arvind_khadri has quit IRC 2011-03-06T09:35:37 *** arvind_khadri has joined #rtems 2011-03-06T09:59:37 *** arvind_k has joined #rtems 2011-03-06T09:59:58 *** arvind_khadri has quit IRC 2011-03-06T10:04:52 *** arvind_k is now known as arvind_khadri 2011-03-06T10:05:07 *** arvind_khadri has joined #rtems 2011-03-06T11:01:58 *** dr__house has joined #rtems 2011-03-06T11:01:59 *** dr__house has joined #rtems 2011-03-06T11:02:23 *** madrazr has joined #rtems 2011-03-06T11:28:27 *** mw1 is now known as mwalle 2011-03-06T13:10:01 *** dr__house has quit IRC 2011-03-06T13:13:03 *** arvind_khadri has quit IRC 2011-03-06T13:16:48 *** core-ix has joined #rtems 2011-03-06T13:21:09 *** madrazr has quit IRC 2011-03-06T14:12:14 *** mmu_man has quit IRC 2011-03-06T15:49:18 *** calvino has joined #rtems 2011-03-06T15:51:59 *** calvino has joined #rtems 2011-03-06T15:52:51 *** verm__ has quit IRC 2011-03-06T16:17:29 *** dr__house has joined #rtems 2011-03-06T16:17:29 *** dr__house has joined #rtems 2011-03-06T17:03:32 *** calvino is now known as verm__ 2011-03-06T17:03:39 *** verm__ has joined #rtems 2011-03-06T18:52:23 *** methril has quit IRC 2011-03-06T19:50:44 *** dr__house has quit IRC 2011-03-06T20:57:57 *** dr__house has joined #rtems 2011-03-06T20:57:57 *** dr__house has joined #rtems 2011-03-06T21:43:22 *** dr__house` has joined #rtems 2011-03-06T21:43:22 *** dr__house has quit IRC 2011-03-06T21:52:43 *** arvind_khadri has joined #rtems 2011-03-06T22:01:25 *** dr__house` is now known as dr__house 2011-03-06T22:22:41 *** dr__house has quit IRC 2011-03-06T22:53:26 *** arvind_khadri has quit IRC 2011-03-06T23:13:20 *** arvind_khadri has joined #rtems