2013-04-08T00:07:03 *** the9a3eedi-3 has quit IRC (Ping timeout: 256 seconds) 2013-04-08T00:10:36 *** arvind_khadri has joined #rtems 2013-04-08T02:03:43 *** sebhub has joined #rtems 2013-04-08T02:04:13 good morning 2013-04-08T02:17:33 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2013-04-08T02:37:27 *** the9a3eedi has joined #rtems 2013-04-08T03:00:36 *** MegaAlex|away is now known as MegaAlex 2013-04-08T03:17:49 *** arvind_khadri has joined #rtems 2013-04-08T03:21:13 sebhub, around ? 2013-04-08T03:37:34 sebhub, I will be back a little later. I have almost refactored the RSB to allow support for snapshots and pre-release testing. 2013-04-08T04:11:47 *** the9a3eedi has quit IRC (Ping timeout: 252 seconds) 2013-04-08T04:12:37 *** arvind_khadri has quit IRC (Remote host closed the connection) 2013-04-08T04:16:27 *** antgreen has joined #rtems 2013-04-08T04:22:13 *** antgreen has quit IRC (Quit: Leaving) 2013-04-08T04:30:07 *** arvind_khadri has joined #rtems 2013-04-08T05:01:05 *** verm__ has quit IRC (Ping timeout: 255 seconds) 2013-04-08T05:31:06 *** MegaAlex is now known as MegaAlex|away 2013-04-08T05:42:08 *** the9a3eedi has joined #rtems 2013-04-08T06:52:08 *** antgreen has joined #rtems 2013-04-08T07:01:15 *** verm__ has joined #rtems 2013-04-08T08:12:57 *** S_Somani has joined #rtems 2013-04-08T08:17:18 *** antgreen has quit IRC (Quit: Leaving) 2013-04-08T08:17:42 *** peerst is now known as l_p 2013-04-08T08:18:12 *** l_p is now known as p__l 2013-04-08T08:19:19 *** p__l is now known as peerst 2013-04-08T08:22:16 *** peerst is now known as hq2 2013-04-08T08:23:00 *** hq2 is now known as hq1 2013-04-08T08:28:06 *** hq1 is now known as Guest75225 2013-04-08T08:29:34 *** Guest75225 is now known as peerst1 2013-04-08T08:30:42 *** peerst1 is now known as peerst 2013-04-08T09:01:15 *** the9a3eedi has quit IRC (Read error: Connection reset by peer) 2013-04-08T09:02:18 *** the9a3eedi has joined #rtems 2013-04-08T09:07:46 *** arvind_khadri has quit IRC (Remote host closed the connection) 2013-04-08T09:07:51 *** the9a3eedi has quit IRC (Ping timeout: 276 seconds) 2013-04-08T09:17:27 *** arvind_khadri has joined #rtems 2013-04-08T09:17:56 *** weiY has joined #rtems 2013-04-08T09:20:17 *** arvind_khadri has quit IRC (Remote host closed the connection) 2013-04-08T09:22:38 Hi, everyone 2013-04-08T09:23:11 Is there a tool to debug the locks in the RTEMS? like witness for FreeBSD 2013-04-08T09:23:13 *** arvind_khadri has joined #rtems 2013-04-08T09:25:55 no 2013-04-08T09:26:11 but this would be nice 2013-04-08T09:27:48 *** arvind_khadri has quit IRC (Ping timeout: 252 seconds) 2013-04-08T09:36:16 sebhub, hi 2013-04-08T09:36:37 recently i am preparing the atomic project 2013-04-08T09:37:52 first i will complete other architectures atomic support. and other things is to make the original locks smp safe. so the lock debug tool is very useful 2013-04-08T09:39:03 So i wonder whether porting the witness to RTEMS is possible or useful? 2013-04-08T09:45:42 it is very hard to add proper smp locks in rtems 2013-04-08T09:45:58 this is more or less a complete reimplementation 2013-04-08T09:46:20 of the core operating system 2013-04-08T09:46:23 i think first can import the atomic ops 2013-04-08T09:46:53 and then make the exsit locks smp safe using atomic ops 2013-04-08T09:47:38 why it is very hard to add smp locks in rtems, sebhub? 2013-04-08T09:48:00 the score was never intended to run on a smp system 2013-04-08T09:48:28 blocking smp locks are hard to implement efficently 2013-04-08T09:49:18 if the smp schedule works ok, and smp locks will also be easy to import 2013-04-08T09:50:12 i don't think so, my impression is that proper smp locking will be very hard 2013-04-08T09:50:14 now in uniprocessor it just use disable/enable-irq. and in smp we can use atomic ops to do sync 2013-04-08T09:50:47 recently i see the road of freebsd smp project 2013-04-08T09:51:09 you have to keep in mind that we still want a real time system with fixed upper bounds for certain operations 2013-04-08T09:52:47 yeah, i know there is uncertain time about the locks such as spinlock which is used in smp. but if RTEMS plan to support SMP this is road to go 2013-04-08T09:53:39 so i wonder whether there is a plan of full SMP support in rtems 2013-04-08T09:54:56 yes, there are plans, the ESA sponsors a project in this area for example 2013-04-08T09:56:25 is this project open? and what is progress of this project? 2013-04-08T09:56:58 currently the ESA evaluates who gets the contract 2013-04-08T09:57:18 *** S_Somani has left #rtems 2013-04-08T09:57:27 I guess there will be an announcement on the rtems mailing list some time in the future 2013-04-08T09:58:10 project duration is several months 2013-04-08T09:59:39 *** wei has joined #rtems 2013-04-08T09:59:58 So could you give me some advice about my gsoc project about SMP support for rtems 2013-04-08T10:00:03 *** wei is now known as Guest83493 2013-04-08T10:00:53 which features are wanted by rtems developer 2013-04-08T10:00:58 at the moment I cannot give advice due to the ESA project evaluation 2013-04-08T10:01:24 *** Guest83493 is now known as ywei 2013-04-08T10:01:38 *** weiY has quit IRC (Ping timeout: 276 seconds) 2013-04-08T10:01:44 *** ywei is now known as weiy 2013-04-08T10:02:05 real smp support for rtems will be out of scope for a gsoc project from my point of view 2013-04-08T10:02:23 it might be possible to select certain subproblems 2013-04-08T10:03:52 yeah, i know the whole smp support is very large. so i want to select a certain task which is suitable for me to complete and also useful to rtems 2013-04-08T10:05:37 in my option a lock debug tool is useful for developer and also useful for future SMP project, what do you think? 2013-04-08T10:06:14 i thinks something like witness is mandatory 2013-04-08T10:07:04 i would use the freebsd system unless there are good reasons not to choose it 2013-04-08T10:07:33 *** weiwei has joined #rtems 2013-04-08T10:08:25 sebhub, sorry for my weak internet. i am lost again 2013-04-08T10:08:54 about your reply. mandatory? for what 2013-04-08T10:10:22 *** weiy has quit IRC (Ping timeout: 256 seconds) 2013-04-08T10:11:16 *** weiwei is now known as weiY 2013-04-08T10:23:01 mandatory for a useful smp support 2013-04-08T10:23:19 *** antgreen has joined #rtems 2013-04-08T10:25:17 sorry, i have to leave now 2013-04-08T10:25:25 ok, byb 2013-04-08T10:34:57 *** sebhub has quit IRC (Remote host closed the connection) 2013-04-08T10:35:00 *** antgreen has quit IRC (Quit: Leaving) 2013-04-08T11:04:00 *** weiY has quit IRC (Ping timeout: 256 seconds) 2013-04-08T11:39:41 *** DrJoel has joined #rtems 2013-04-08T11:39:41 *** DrJoel has joined #rtems 2013-04-08T11:39:41 *** ChanServ sets mode: +o DrJoel 2013-04-08T11:40:01 *** Hesham has joined #rtems 2013-04-08T11:45:36 *** MegaAlex|away is now known as MegaAlex 2013-04-08T11:47:39 *** MegaAlex is now known as MegaAlex|away 2013-04-08T12:13:10 *** Hesham has quit IRC (Ping timeout: 256 seconds) 2013-04-08T12:13:43 *** Hesham has joined #rtems 2013-04-08T12:18:44 1 hour 42 minutes until GSOC organization announcement 2013-04-08T12:21:35 waiting :) 2013-04-08T12:29:40 I was going to update the Student Application Template and make that available shortly. 2013-04-08T12:29:54 I want to add (1) What free and open source experience do you have? 2013-04-08T12:30:12 (2) How do you plan to stay active in RTEMS after GSOC? 2013-04-08T12:30:37 (3) Will you commit to ensuring that your code is merged even if that occurs AFTER the GSOC deadline? 2013-04-08T12:36:27 *** S_Somani has joined #rtems 2013-04-08T13:20:00 *** MegaAlex|away is now known as MegaAlex 2013-04-08T13:20:13 I think most GSoC templates include (1) and (2), not sure about (3) 2013-04-08T13:20:49 I thought it would be good to ask since the past two years we have had projects which took longer to settle and get merged than the program deadline 2013-04-08T13:35:10 Seems like there are many student interested in RTEMS than last year :) 2013-04-08T13:37:01 funny.. I was thinking the opposite .. but there are more people helping students so that lightens the load. I was on travel last week and barely answered any questions 2013-04-08T13:38:23 *** Hesham1 has joined #rtems 2013-04-08T13:39:33 *** Hesham has quit IRC (Ping timeout: 248 seconds) 2013-04-08T13:48:33 12 minutes 2013-04-08T13:55:23 *** antgreen has joined #rtems 2013-04-08T13:56:55 *** MegaAlex is now known as MegaAlex|away 2013-04-08T14:01:01 *** DrJoel changes topic to "Welcome to RTEMS (http://www.rtems.org). Please feel free to ask questions and, if you can, answer those from others. Google Summer of Code 2013 is ON and RTEMS is a participating organization! Students get your applications in soon." 2013-04-08T14:01:52 Congratulations :) 2013-04-08T14:15:05 Now to fill in the profile .. 2013-04-08T14:15:16 The mail hit my inbox and it was precisely 2:00 2013-04-08T14:37:13 *** Hesham1 has left #rtems 2013-04-08T15:04:04 *** Dhruv has joined #rtems 2013-04-08T15:04:20 *** Dhruv has left #rtems 2013-04-08T15:04:25 kiwichris awake yet? 2013-04-08T15:06:14 *** jedi__ has joined #rtems 2013-04-08T15:23:37 *** phipse has joined #rtems 2013-04-08T15:23:42 *** jedi__ has quit IRC (Quit: Page closed) 2013-04-08T15:38:47 * DrJoel looks around and notices it is quiet 2013-04-08T15:50:01 http://www.google-melange.com/gsoc/org/google/gsoc2013/rtems 2013-04-08T15:57:25 *** cdcs has joined #rtems 2013-04-08T15:58:35 *** cdcs has quit IRC (Client Quit) 2013-04-08T16:05:15 DrJoel: Yay! 2013-04-08T16:06:41 :) 2013-04-08T16:06:49 time to start putting in mentor applications 2013-04-08T16:07:24 is there a list of projects yet? 2013-04-08T16:07:27 http://www.google-melange.com/gsoc/homepage/google/gsoc2013 2013-04-08T16:07:47 RTEMS Student Projects? Or what Google calls organizations? 2013-04-08T16:08:00 RTEMS projects 2013-04-08T16:08:03 177 orgs were accepted.. about 1/2 have filled in their profile to show up 2013-04-08T16:08:44 http://www.rtems.org/wiki/index.php/Open_Projects but it can always stand work 2013-04-08T16:08:56 I am hawking the "Summer of Simulators" :) 2013-04-08T16:11:43 hmm 2013-04-08T16:12:17 didn't find anything I'd really like to mentor 2013-04-08T16:12:21 Basically add simulators to rtems-source-builder.. bump us to newest versions.. what has sim-scripts, works with Eclipse, has coverage, etc 2013-04-08T16:12:26 well.. be creative :) 2013-04-08T16:12:26 but I'm too busy anyway probably 2013-04-08T16:13:05 BeagleBone BSP ... 2013-04-08T16:16:44 Beagleboard was a project last year .. I know Bone is newer but this still needs merging https://google-melange.appspot.com/gsoc/project/google/gsoc2012/claasziemke/31001 2013-04-08T16:21:59 Does Beagleboard have the same SoC than BeagleBone? 2013-04-08T16:23:07 No AFAIK .. that's the issue. 2013-04-08T16:26:51 *** phipse has left #rtems 2013-04-08T16:28:25 So BeagleBone needs a completely new BSP 2013-04-08T16:28:42 that's the way it appears. :( 2013-04-08T16:29:01 How did the student who did Beagleboard get the board? 2013-04-08T16:30:18 There is a simulator and I think you can still buy them. Or could then 2013-04-08T16:30:26 qemu simulates it 2013-04-08T16:40:46 Well nothing beats real hardware when doing a BSP 2013-04-08T16:41:21 why wasn't the Beagleboard merged? Was it not completed? 2013-04-08T16:42:26 That's a good question.. I will email the author and see what the deal is 2013-04-08T16:46:24 he was asking questions in Dec 2012 so still at it.. 2013-04-08T16:47:08 ah ok 2013-04-08T16:51:35 Since I have BeagleBones on my table and plan to use them as RTEMS + Erlang image target for quick start I'm generally interested in BeagleBone BSP's 2013-04-08T16:51:44 I think kiwichris is also interested 2013-04-08T16:52:21 If he has ticker running, then it is time to merge 2013-04-08T16:52:27 maybe we can also get some input from GSoC to it 2013-04-08T16:52:30 cool 2013-04-08T16:53:39 I'm off to bed for now 2013-04-08T16:55:42 night 2013-04-08T16:59:50 *** verm__ has quit IRC (Changing host) 2013-04-08T16:59:50 *** verm__ has joined #rtems 2013-04-08T18:25:13 DrJoel, hi 2013-04-08T18:25:39 peerst, nice, I would like one. 2013-04-08T18:37:16 hey... kiwichris 2013-04-08T18:37:20 did you get your invites? 2013-04-08T18:37:32 I just registered as a mentor 2013-04-08T18:38:00 We can "connect" and invite by email. I haven't waded through everyone yet.. if you have gmail address for anyone to invite .. go ahead. I spent the afternoon announcing and filling out the org profile page 2013-04-08T18:39:00 Great. So I understand Melange has gone retro by allowing email sign up etc :) 2013-04-08T18:39:59 lol... and the student who implemented is off the grid right now so bug fixes are scarce. :) 2013-04-08T18:40:03 https://www.facebook.com/events/128850557300880/129191647266771/ check that out 2013-04-08T18:40:42 FYI Minix did NOT make it this year 2013-04-08T18:40:44 Parrot did 2013-04-08T18:41:15 Ooo I wonder what happened with them. I saw some posts on their group. 2013-04-08T18:41:46 I have no idea and neither did Ben. 2013-04-08T18:42:06 That is a real shame. 2013-04-08T18:44:14 I see MacPorts got accepted. Hmmm need to corner those guys and find out about the RTEMS ports being removed 2013-04-08T18:44:54 Plan 9 is in. 2013-04-08T18:45:07 I am pushing on getting the BeagleBoard merged if it is >= Pi in state 2013-04-08T18:45:21 The SoC is different to the BeagleBone. 2013-04-08T18:46:00 The BeagleBoard, BeagleBoard xM and BeagleBone have different OMAP versions 2013-04-08T18:46:20 I know that but it just needs to be done. It is a common enough board 2013-04-08T18:46:22 It is a real pain. Only Linux supports BeagleBoard and BeagleBoard xM 2013-04-08T18:46:57 FreeBSD has BeagleBone support 2013-04-08T18:47:19 The BSP was done last year as a GSOC project... why not merge it? 2013-04-08T18:47:28 Because we need to maintain it. 2013-04-08T18:47:43 I cannot run a BeagleBoard because I have a BeagleBoard xM. 2013-04-08T18:48:16 :( peerst has some variant.. wonder what it is 2013-04-08T18:48:40 Maybe Sebastian should be asked. If the RTEMS support is ok to allow the different OMAPs then we should. 2013-04-08T18:48:50 This is the key issue 2013-04-08T18:49:05 Yeah.. that's a good plan. 2013-04-08T18:49:17 It is builds, worst case, I can throw it on a branch on my repo so it isn't lost 2013-04-08T18:49:24 Good idea 2013-04-08T18:49:25 If it builds 2013-04-08T18:49:36 Let me check. 2013-04-08T18:51:26 It did not compile. 2013-04-08T18:51:37 I will email you a patch 2013-04-08T18:52:16 Hey on rtems.org touch is in /bin/touch and on rtbf64a it is in /usr/bin/touch. Has something changed in Fedora ? 2013-04-08T18:52:32 I don't have the BSP on my machine so it doesn't matter. If you want to file a PR with the pieces we have, that is cool 2013-04-08T18:53:00 It must have.. rtbf64a is likely a newer version. Jeff holds off on the public facing machines. The internal ones take every upgrade 2013-04-08T18:53:02 Check my personal repo 2013-04-08T18:53:17 oh.. you already have the beagleboard in it? 2013-04-08T18:53:27 That is a s*&t 2013-04-08T18:53:53 No a local branch 2013-04-08T18:54:57 why would something like that move. It shoudl be in the same place it was 30 years ago. :( 2013-04-08T18:55:07 POSIX ? 2013-04-08T18:55:22 I rather doubt that.. probably LSB 2013-04-08T18:55:35 FreeBSD is /usr/bin/touch 2013-04-08T18:55:57 This is a tough one to manage. 2013-04-08T18:55:57 rtems.org is an older distro if that means anything 2013-04-08T18:56:11 centos 6.4 has /bin/touch.. it moved! 2013-04-08T18:56:27 Yes so I need to make the RSB smarter managing the locations 2013-04-08T18:56:57 The older Fedora and CentOS are the same, the newer Fedora has changed. 2013-04-08T18:58:45 *** the9a3eedi has joined #rtems 2013-04-08T18:59:23 *** the9a3eedi-2 has joined #rtems 2013-04-08T18:59:42 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-08T19:03:13 *** the9a3eedi has quit IRC (Ping timeout: 256 seconds) 2013-04-08T19:05:29 Fedora = 'Fedora, eradicating documented old reliable architectures" 2013-04-08T19:05:50 :) 2013-04-08T19:06:10 If they keep going this way they will start using a BSD kernel 2013-04-08T19:06:33 now that's funny 2013-04-08T19:23:18 bye 2013-04-08T19:23:21 *** DrJoel has quit IRC (Quit: Say What?) 2013-04-08T23:26:33 *** S_Somani has joined #rtems 2013-04-08T23:59:57 *** arvind_khadri has joined #rtems 2013-04-09T00:28:08 *** S_Somani1 has joined #rtems 2013-04-09T00:28:44 Is there any "xmodem"-like functionality built into RTEMS? 2013-04-09T00:29:07 I would like to transfer a file via serial using a shell command, if it exists 2013-04-09T00:29:38 I do not think so in RTEMS. It may have floated around in the past 2013-04-09T00:30:17 I see. are any other methods of transferring a file to the debug host? 2013-04-09T00:30:22 *** S_Somani has quit IRC (Ping timeout: 256 seconds) 2013-04-09T00:30:25 only thing I could think of is FTP 2013-04-09T00:30:38 Networking enabled ? 2013-04-09T00:30:50 not yet 2013-04-09T00:30:54 Will it ? 2013-04-09T00:31:03 be supported ? 2013-04-09T00:31:15 I can I guess, I'm mostly experimenting 2013-04-09T00:31:31 Ah ok. If I have networking I tend to use NFS 2013-04-09T00:31:34 but I'd rather not use networking if there's a simpler method 2013-04-09T00:31:51 GDB running ? 2013-04-09T00:32:02 through eclipse, yes 2013-04-09T00:32:07 Ah ok 2013-04-09T00:32:17 JTAG, BDM stub ? 2013-04-09T00:32:26 I mean 2013-04-09T00:32:32 JTAG, BDM, or serial stub ? 2013-04-09T00:32:41 serial 2013-04-09T00:33:55 I think there are stub commands in the gdb remote protocol to transfer files. 2013-04-09T00:34:08 will take a look, thanks 2013-04-09T00:34:33 http://sourceware.org/gdb/onlinedocs/gdb/File_002dI_002fO-Remote-Protocol-Extension.html#File_002dI_002fO-Remote-Protocol-Extension 2013-04-09T00:34:41 http://sourceware.org/gdb/onlinedocs/gdb/File-Transfer.html found this. I'll try it out. thanks for the tip 2013-04-09T00:34:45 I have not used it. 2013-04-09T00:35:06 Looks interesting. If you get it working please make a wiki page. I am sure it will useful to othes 2013-04-09T00:35:51 Alright, though the last time I tried to edit a wiki page it told me I didn't have permission, so I assumed it's only for RTEMS developers 2013-04-09T00:36:22 No, just ask Joel for an account. We had to lock it down to stop it getting abused 2013-04-09T00:36:35 I see. Will send him an email then 2013-04-09T00:36:55 If you do this and figure it out you are an RTEMS developer :) 2013-04-09T00:36:59 or wait, there's a request form in the wiki. I'll just use that 2013-04-09T00:37:04 Ah ok 2013-04-09T00:37:30 I've become pretty good at working with RTEMS now :D I've ported it to the LEON3 hardware now, got it to work with CAN, and stuff 2013-04-09T00:37:44 Nice 2013-04-09T00:48:30 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2013-04-09T01:11:46 *** kiwichris has joined #rtems 2013-04-09T02:15:21 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2013-04-09T02:20:06 *** kiwichris has joined #rtems 2013-04-09T02:22:38 *** sebhub has joined #rtems 2013-04-09T02:23:26 good morning 2013-04-09T02:23:28 *** MegaAlex|away has quit IRC (Ping timeout: 260 seconds) 2013-04-09T02:25:25 *** MegaAlex|away has joined #rtems 2013-04-09T02:28:34 sebhub, hi 2013-04-09T02:30:13 DrJoel, kiwichris: I don't have a Beagleboard of any varian, only latest BeagleBone 2013-04-09T02:30:43 peerst, great. This is nice. It has FreeBSD support 2013-04-09T02:31:05 yeah, wouldn't have bought it otherwise :-) 2013-04-09T02:31:12 :) 2013-04-09T02:31:19 also has jtag built in 2013-04-09T02:31:41 because when I fail to put RTEMS on I allways can run stuff with FreeBSD 2013-04-09T02:31:48 jep 2013-04-09T02:32:11 that was the +1 over the Pi, that and more I/O and I like the shields 2013-04-09T02:32:28 Agreed for embedded yes 2013-04-09T02:32:32 need to run 2013-04-09T02:32:36 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2013-04-09T03:01:40 *** MegaAlex|away has quit IRC (Ping timeout: 256 seconds) 2013-04-09T03:02:52 *** MegaAlex|away has joined #rtems 2013-04-09T03:40:32 *** MegaAlex|away is now known as MegaAlex 2013-04-09T03:48:44 *** kiwichris has joined #rtems 2013-04-09T03:59:00 *** the9a3eedi-2 has quit IRC (Ping timeout: 264 seconds) 2013-04-09T04:28:23 *** S_Somani1 has quit IRC (Quit: Leaving.) 2013-04-09T04:29:15 *** freenix has joined #rtems 2013-04-09T04:31:15 *** freenix has left #rtems 2013-04-09T04:37:30 *** MegaAlex is now known as MegaAlex|away 2013-04-09T04:55:29 *** S_Somani has joined #rtems 2013-04-09T04:58:47 *** the9a3eedi has joined #rtems 2013-04-09T05:02:12 *** Dinesh_R_T has joined #rtems 2013-04-09T05:38:43 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-09T05:39:00 *** S_Somani has joined #rtems 2013-04-09T05:43:28 *** S_Somani has quit IRC (Ping timeout: 245 seconds) 2013-04-09T05:48:17 Hello, I am interested in working on RTEMS system events. I would like to know what is expected from this task . Also few pointers on what sheep/goat project i can attempt will be helpful . 2013-04-09T05:53:44 *** antgreen has quit IRC (Read error: Connection reset by peer) 2013-04-09T07:06:39 *** S_Somani has joined #rtems 2013-04-09T07:19:14 *** S_Somani has quit IRC (Ping timeout: 255 seconds) 2013-04-09T08:09:25 *** weiY has joined #rtems 2013-04-09T08:11:45 *** Hesham has joined #rtems 2013-04-09T08:18:59 *** Dinesh_R_T has left #rtems 2013-04-09T08:19:11 *** Dinesh_R_T has joined #rtems 2013-04-09T08:19:30 *** Dinesh_R_T has left #rtems 2013-04-09T08:24:42 *** S_Somani has joined #rtems 2013-04-09T08:35:15 *** weiwei has joined #rtems 2013-04-09T08:40:42 *** gedare has joined #rtems 2013-04-09T08:51:32 Hi, gedare 2013-04-09T08:55:20 yesterday i talked to sebhub and known that there will be a team to develope smp support for RTEMS. and the smp support is a too larget project for gsoc. But i am interested in SMP support for rtems project. So i want to discusse with you to select a suitable subproject of SMP support. 2013-04-09T08:56:00 now i have some idea. one is make exsiting lock primitive of rtems to smp safe using atomic ops 2013-04-09T08:57:08 another is to develop a lock debug tool like witness of freebsd and lockdep of linux. we can port the witness to rtems 2013-04-09T09:00:19 *** zhangwenjie has joined #rtems 2013-04-09T09:00:20 *** weiY has quit IRC (Read error: Connection reset by peer) 2013-04-09T09:00:42 *** zhangwenjie has left #rtems 2013-04-09T09:01:08 *** weiwei is now known as weiY 2013-04-09T09:01:17 hi weiY 2013-04-09T09:01:32 Hi gedare 2013-04-09T09:01:45 i'm not familiar with what aspects of SMP will be done that sebhub mentioned... 2013-04-09T09:01:54 so I cannot give much feedback on what is a suitable sub-project 2013-04-09T09:02:07 however, implementation of locks or SMP debugging aids do seem like good projects to me 2013-04-09T09:03:10 yeah, in my proposal first part is to complete the atomic support of other architecture. another is one of above idea 2013-04-09T09:03:15 good 2013-04-09T09:03:37 to me this is a good idea, just need to check with sebhub that your work is not going to be done already 2013-04-09T09:03:49 just send an email to rtems-devel with a short explanation of what you plan to do 2013-04-09T09:03:57 basically, what you just told me :) 2013-04-09T09:03:58 yeah, this is what i am wonder 2013-04-09T09:04:13 ok 2013-04-09T09:08:03 good I think. 2013-04-09T09:08:06 Hi Hesham. 2013-04-09T09:08:18 hi gedare 2013-04-09T09:08:29 Hesham, I'm reading your doc right now. it seems ok at first glance though. 2013-04-09T09:09:06 i wish we can get the arm platform tested somehow... :( 2013-04-09T09:09:07 I wanted to discuss some other issues related to PPC too, do not want to embed it at the proposal right now 2013-04-09T09:09:14 like what? 2013-04-09T09:09:38 Like BAT ( Block Address Translation ) 2013-04-09T09:09:47 As we use one level page tables on ARM with 1MB sections, why not using BAT at powerpc that can range from 128Kb to 256Mb in size. BAT translation is also faster, simpler and has higher precedence than page translations. 2013-04-09T09:10:22 I don't really care at the start.. as long as the static map can support what is already defined by the BSP it does not matter.... 2013-04-09T09:10:39 but for when users can override it, if we support some dynamic maps, then it may be necessary to have smaller pages 2013-04-09T09:10:56 I agree 2013-04-09T09:10:58 I'd suggest asking on rtems-devel about the need for small page sizes, though.. 2013-04-09T09:11:22 Yes, Peng argues with me about having tow-level page tables 2013-04-09T09:11:25 It also depends on how much RAM that the board has 2013-04-09T09:11:30 heh 2013-04-09T09:12:14 Ok I think we will keep page translation in that way 2013-04-09T09:12:44 Yes. I think right now pages can be large... 2013-04-09T09:12:55 covering areas like all of Heap, all of .text, etc... 2013-04-09T09:13:11 *** S_Somani1 has joined #rtems 2013-04-09T09:13:32 Pages are only 4 KB at PPC unlike ARM which page sizes can vary 2013-04-09T09:13:34 But this is still arch-dependent too... 2013-04-09T09:13:39 right, exactly 2013-04-09T09:14:06 Another issue is that, for powerpc, the UISA is mostly unified for PowerPC CPUs, I could try to run the current code for different boards. Is that ok ? 2013-04-09T09:14:09 i see, so you can combine BAT with pages to make larger regions? 2013-04-09T09:14:15 Yes 2013-04-09T09:14:18 there is no PPC MMU, there are many variants 2013-04-09T09:14:44 yes, and you can ask for some assistance in testing from users/developers too who might be willing to try 2013-04-09T09:14:53 that was what I feared :P 2013-04-09T09:14:59 I know some ppc also use MPU 2013-04-09T09:15:19 yes, they check also dma access 2013-04-09T09:15:36 so, the low-level implementation needs to be more or less BSP-specific for the PPC family 2013-04-09T09:16:02 Some PPC implementations treats instructions (like MMU,MPU related) as no-ops if the there is no HW/Unit support 2013-04-09T09:16:13 *** S_Somani has quit IRC (Ping timeout: 240 seconds) 2013-04-09T09:17:04 it would be better to actually not execute the instructions in those cases though 2013-04-09T09:17:27 Should I add the part of sharing PPC MMU code between families ? 2013-04-09T09:17:34 at any rate, the low-level infrastructure has to take into account the variability in MMU and MPU, even within one architecture family. 2013-04-09T09:18:03 sure, but also that you need to identify which PPC BSPs use which MMU/MPU capabilities 2013-04-09T09:18:34 Ok 2013-04-09T09:18:37 btw I think there are also some different ARM BSPs that do not use the CP15 approach, even some that use MPU also in ARM 2013-04-09T09:18:53 so, finding out what is the MMU/MPU for each BSP is necessary 2013-04-09T09:18:58 then you can determine what can be shared 2013-04-09T09:20:33 and what are the requirements of the HW, e.g. does it support paging, does it have variable-size pages, or does it provide just some fixed segmentation 2013-04-09T09:20:54 When I say MPU I mean something that defines only a fixed number of regions... 2013-04-09T09:21:07 usually, MMU means page-based in modern hw 2013-04-09T09:21:24 (Although, we could provide segmentation-based mmu support for x86) 2013-04-09T09:21:26 Yes it's like BAT except to address translation requires and it's fixed sizes 2013-04-09T09:21:44 right. 2013-04-09T09:21:48 Did you have the chance to ask about ARM simulators problems ? I think that will raise some problems unlike PPC. 2013-04-09T09:22:13 in fact, it would be good if you can find a BSP that has an MPU to give you another type of system 2013-04-09T09:22:55 I have not.. it seems like there are not many for the type of ARM we use 2013-04-09T09:23:17 The "heavier" ARM processors have simulators.. because of Android developers I would guess. 2013-04-09T09:23:28 I intended to work on PPC for most parts of the project. PSIM is reliable and gives me whatever I need for testing purpose 2013-04-09T09:23:41 just remember psim only simulates one kind of ppc 2013-04-09T09:23:58 that is one problem in this project.. the wide variety among boards is hard to test and implement 2013-04-09T09:24:17 but, if you can get a few kinds of BSPs with different MMU/MPU requirements, hopefully the infrastructure will be easily extended to other BSPs 2013-04-09T09:24:39 iirc you said skyeye does not simulate the MMU for ARM? 2013-04-09T09:25:06 Nope, I said I am not sure whether it handles MMU/Caches the right way 2013-04-09T09:25:19 But they said it supports MMU operations 2013-04-09T09:26:13 I have changes some attributes when I was testing gumstix code last year, but seems like it did not recognize these changes 2013-04-09T09:26:26 ah, i see. 2013-04-09T09:26:28 *** jedi__ has joined #rtems 2013-04-09T09:27:18 another possibility is to buy a raspberry pi, it is cheap enough and should have enough support in RTEMS now to do some basic MMU work with it... 2013-04-09T09:27:39 what about beaglebone ? 2013-04-09T09:28:36 you have to check with Claas since we did not get his code merged yet. and I think he was working on beagleboard first before beaglebone, but i'm not sure 2013-04-09T09:29:31 Does raseberry BSP is out there for rtems ? 2013-04-09T09:29:39 http://alanstechnotes.blogspot.com/ 2013-04-09T09:29:55 yes, it was merged last month, those are some notes by the author of the BSP 2013-04-09T09:30:34 i don't know what MMU it uses. 2013-04-09T09:32:38 Seems like I would buy Raspberry Pi if it supports MMU ( and I think It does ) 2013-04-09T09:32:47 That's for ARM, what about PPC ? 2013-04-09T09:33:19 i'm not sure... probably most boards cost $200+ 2013-04-09T09:34:09 but you might be able to find someone who has MMU code for a ppc board.. i know peter dufualt was doing some work with a ppc board with an mpu 2013-04-09T09:34:40 I wanted to test the current MMU code for PPC 2013-04-09T09:35:12 besides on psim? 2013-04-09T09:35:31 Yes if it's possible. 2013-04-09T09:39:30 i don't know. 2013-04-09T09:39:37 raspberry uses arm11 core (armv6 family) 2013-04-09T09:40:20 ARM1176JZF-S 2013-04-09T09:40:37 *** jedi__ has quit IRC (Quit: Page closed) 2013-04-09T09:40:38 i have a ppc p2020rdb board which can test your code 2013-04-09T09:40:54 but this board is not a full time test board. haha 2013-04-09T09:41:46 Can I test code for mpc6xx on it ? 2013-04-09T09:42:41 i think mpc6xx code will not run successfully 2013-04-09T09:42:51 p2020 uses E500 core 2013-04-09T09:43:30 :( 2013-04-09T09:43:33 you can try qemu. 2013-04-09T09:43:45 now qemu also support e500 core 2013-04-09T09:44:14 so it should also support mpc6xx which is earler than e500 core 2013-04-09T09:44:17 there is also psim (gdb simulator) 2013-04-09T09:44:27 heh, I spent a lot of time trying to find a working simulator for ARM (including qemu), but I was not so lucky 2013-04-09T09:44:44 mpc6xx is running on psim 2013-04-09T09:44:58 I just wanted a real boards to test the code 2013-04-09T09:45:00 i think that is ok for now... better to try some other kind of boards 2013-04-09T09:45:49 if qemu/e500 supports the MMU, then that might be a good one to check out 2013-04-09T09:45:54 Ok lets give PPC higher priority than ARM 2013-04-09T09:46:18 Also raspberry PI ARM11 core supports MMU 2013-04-09T09:46:25 and also i can help you test e500 code 2013-04-09T09:46:41 on my real board 2013-04-09T09:46:50 weiY, have you used qemu/e500? 2013-04-09T09:46:54 Great 2013-04-09T09:47:01 yeah 2013-04-09T09:47:20 ok. that might be a good place for Hesham to go with the ppc side... 2013-04-09T09:47:21 i have run successfully p1020/p2020 code on qemu 2013-04-09T09:47:33 have you run RTEMS on it? 2013-04-09T09:47:36 Yes that's good news 2013-04-09T09:47:57 but i run this code last year in the qemu which i apply a patch 2013-04-09T09:48:14 i think this patch should be included in latest version 2013-04-09T09:48:42 no, i just try linux on the qemu 2013-04-09T09:48:59 gedare: Could we give PPC higher priority than ARM (Since variants simulators and boards work fine) ? 2013-04-09T09:49:55 sure. 2013-04-09T09:50:08 Hesham, first you'll want to try to get RTEMS to run on Qemu/e500 2013-04-09T09:50:13 if i run RTEMS there will be some modifies about boot rom, i think 2013-04-09T09:50:16 the right BSP will be one of the qoriq ones 2013-04-09T09:50:56 what is the qemuppc bsp.. 2013-04-09T09:51:11 ah, also mpc6xx 2013-04-09T09:51:46 and Hesham you can order a raspberrypi if you like, and that will give you an ARM to play with too. 2013-04-09T09:51:47 gedare, how about e500 support in RTEMS 2013-04-09T09:52:02 weiY, yes.. the qoriq BSPs should support it I suppose 2013-04-09T09:52:11 e500 is well supported 2013-04-09T09:52:13 Ok, I will order raspberry anyway 2013-04-09T09:52:45 sebhub: did you run it on Qemu before ? 2013-04-09T09:52:50 good, and sounds like you can use e500 and psim for powerpc 2013-04-09T09:52:52 ok, time later i can try the rtems in my p2020rdb board 2013-04-09T09:53:19 i didn't use qemu for ppc 2013-04-09T09:53:20 it would still be nice to get a BSP that uses an MPU. 2013-04-09T09:55:43 I will search for the current rtems PPC BSPs and hopefully find one that use MPU 2013-04-09T09:56:04 i think MPC5554 does. 2013-04-09T09:56:29 nevermind.. it migh be regular mmu hmm 2013-04-09T09:57:00 mpc5643l has a mpu for example 2013-04-09T09:57:07 or mpc5674f 2013-04-09T09:57:12 *** S_Somani1 has quit IRC (Quit: Leaving.) 2013-04-09T09:58:34 gedare: yes I searched for it and it's mmu based 2013-04-09T09:59:09 sebhub pointed out some that might.. you'll have to see if any have affordable development boards or simulator support 2013-04-09T09:59:12 sebhub: is there rtems BSP port for it ? 2013-04-09T09:59:55 I think the mpc55xxevb BSPs 2013-04-09T10:00:02 yes 2013-04-09T10:01:03 good, I will search about these boards too 2013-04-09T10:02:22 gedare: Is the part of redesigning libmm to include Caching, Allocators, Protection OK to be embedded into the proposal ? 2013-04-09T10:03:09 i found some development kits but they don't seem to have the chip prepackaged. 2013-04-09T10:03:33 yes,but you will need to have a good reason to refactor allocation... 2013-04-09T10:04:36 I don't really understand why you would want to do anything with the allocators 2013-04-09T10:05:19 in fact, probably it is better to leave them alone for now, and to focus on the low-level infrastructure to make the interface between BSP and the upper-levels of RTEMS more uniform. 2013-04-09T10:05:54 btw you can also find some mmu related code in some of the different powerpc BSPs already to see what they provide/support 2013-04-09T10:07:46 i think the best thing to focus on is to make it easier for BSP developers to understand how to fit MMU/MPU handling code into RTEMS... 2013-04-09T10:08:00 So right now we just need to focus on the initialization and setting up the static mapping 2013-04-09T10:08:29 then, some very few architectures might also support dynamic mapping, which can be used by higher layers to provide protection between allocators, or between tasks, or similar. 2013-04-09T10:09:12 *** arvind_khadri has quit IRC (Ping timeout: 264 seconds) 2013-04-09T10:09:12 I'll guide the other interested student toward those high-level services, which we can maybe test on mpc6xx/psim... or convince him to look at another project :) 2013-04-09T10:09:43 but your focus should be the low-level services, and fixing the interface "glue" to the upper layers. 2013-04-09T10:11:15 Sorry I was not here 2013-04-09T10:11:38 it's ok. Maybe the "libmm" should just exist in parallel to the existing cache manager code.. i don't know what is the best thing :) 2013-04-09T10:12:24 Yes I was about to ask about that 2013-04-09T10:12:54 Caching operations are few, but also almost always interacts with MMU 2013-04-09T10:13:15 the question to answer is whether there are different instructions for cache operations when the MMU differs. 2013-04-09T10:13:17 It will not take too much time 2013-04-09T10:13:33 if cache operations are the same despite differences in MMU, then it does not make sense to combine the two 2013-04-09T10:14:05 I think It may be related to cache attributes for pages itself 2013-04-09T10:14:19 ok 2013-04-09T10:14:38 Most archs/CPUs defines write-through, write-back attributes 2013-04-09T10:14:51 yeah looking at c/src/lib/libcpu/powerpc/shared/src/cache_.h for example shows it actually divides the implementation among different ppc variants 2013-04-09T10:15:02 These are common for any CPU that has cache unit 2013-04-09T10:15:21 Yes I noticed that 2013-04-09T10:15:35 And some of them are not implemented yet for other CPUs 2013-04-09T10:16:32 ok. i think just avoid biting off more than you can chew. 2013-04-09T10:16:46 but cleaning up the cache manager code is a good thing 2013-04-09T10:16:59 We can define common attributes for the sake of portability, and give other ones default values 2013-04-09T10:17:18 Ok. 2013-04-09T10:17:40 ah one other thing.. 2013-04-09T10:18:06 linkcmds provides a lot of the definitions of useful regions for static mapping 2013-04-09T10:18:21 so as you are looking through this, think about how you might make use of the linkcmds regions to support useful static maps 2013-04-09T10:18:51 I could implement that at testing phase 2013-04-09T10:19:20 And when the code is mature enough, we can embed it at BSPs init time or something 2013-04-09T10:19:26 BRB... 2013-04-09T10:19:42 ok. 2013-04-09T10:20:04 i think there may be some opportunity to add a new BSP call-out during initialization to deal with MMU/MPU setup 2013-04-09T10:26:58 Yes that will be nice. 2013-04-09T10:27:30 Does rtems work entirely at super-user mode ? 2013-04-09T10:27:57 *** S_Somani has joined #rtems 2013-04-09T10:33:47 *** sebhub has quit IRC (Remote host closed the connection) 2013-04-09T10:36:40 Hesham, yes. RTEMS is a single mode, single address space operating system. 2013-04-09T10:37:17 and the mm needs differ from general purpose OSs like linux and bsd 2013-04-09T10:38:01 byb, i have to bed 2013-04-09T10:38:45 I see, but I think Chris will need some level of dynamic mm & page allocations to deal with dynamic apps loading 2013-04-09T10:39:16 weiY: have nice dreams 2013-04-09T10:42:59 *** weiY has quit IRC (Ping timeout: 276 seconds) 2013-04-09T10:52:23 I have to go... Bye 2013-04-09T10:52:27 *** Hesham has left #rtems 2013-04-09T11:05:20 *** S_Somani has quit IRC (Ping timeout: 256 seconds) 2013-04-09T12:14:52 *** S_Somani has joined #rtems 2013-04-09T12:50:58 *** gedare has quit IRC (Ping timeout: 245 seconds) 2013-04-09T13:07:11 *** MegaAlex|away is now known as MegaAlex 2013-04-09T13:20:41 *** Dinesh_R_T has joined #rtems 2013-04-09T13:21:52 *** jedi__ has joined #rtems 2013-04-09T13:22:47 *** Dinesh_R_T has left #rtems 2013-04-09T13:46:34 *** DrJoel has joined #rtems 2013-04-09T13:46:34 *** DrJoel has joined #rtems 2013-04-09T13:46:34 *** ChanServ sets mode: +o DrJoel 2013-04-09T13:54:33 *** S_Somani has left #rtems 2013-04-09T14:14:12 *** jedi__ has quit IRC (Ping timeout: 245 seconds) 2013-04-09T14:19:07 *** gedare has joined #rtems 2013-04-09T14:20:09 hey DrJoel regarding the OA issue it's fine I just wanted to let you know 2013-04-09T14:20:15 you're "it" for the time being ;) 2013-04-09T14:20:43 the connection crap is broken badly badly badly in Melange right now 2013-04-09T14:21:32 yeah 2013-04-09T14:22:04 but i think that is well known, so we just should sit back and wait a bit. 2013-04-09T14:22:51 yep.. just tend to student request 2013-04-09T14:22:59 nice to see Matthew from GCI is interested in GSOC 2013-04-09T14:24:39 Do you have a few minutes for a voice chat gedare? 2013-04-09T14:25:50 i don't think i have it configured on this computer 2013-04-09T14:25:57 I was going to call your cell 2013-04-09T14:26:06 ah.. that's fine. i have to go in about 15 min 2013-04-09T14:26:14 so let's shoot for 5 minutes :) 2013-04-09T14:38:11 sorry if it was too long.. thanks 2013-04-09T14:40:13 it's ok.. i still have a few minutes 2013-04-09T14:44:11 Did I brag that Michele has a book signing at an art gallery in Chattanooga next Friday night? 2013-04-09T14:44:25 nice 2013-04-09T14:44:33 i saw the fb request 2013-04-09T14:44:46 Not expecting you to come.. invited you to brag on her :) 2013-04-09T14:45:05 A freind who owns a print shop is making posters and post cards for it. :) 2013-04-09T14:45:18 well the post cards are for her book in general 2013-04-09T14:49:25 cool 2013-04-09T14:49:52 ok time for me to run. later. 2013-04-09T14:50:53 bye 2013-04-09T14:53:33 *** gedare has quit IRC (Read error: Operation timed out) 2013-04-09T15:12:29 *** phipse has joined #rtems 2013-04-09T15:14:51 hi 2013-04-09T15:15:25 hi phipse 2013-04-09T15:17:32 I have a question about the proposal. I haven't worked with google docs before. 2013-04-09T15:18:06 I just have to give you read access via file/clearance and send the link to the mailinglist, right? 2013-04-09T15:18:35 Basically.. there is a share button. And you list the gmail addresses of the mentors who want to see it. 2013-04-09T15:18:59 Announce who wants to review it and then add them. 2013-04-09T15:19:07 Okay, thanks :) 2013-04-09T15:19:47 Some projects ask you to put your proposal in a publicly visible place. I personally don't like that since it could let someone copy it or simply leverage it to write a better proposal. That would be unfair. 2013-04-09T15:20:24 FWIW mentors are prone to fix grammatical errors but insert comments for bigger issues. When all is Ok, just cut and paste into the application form on the website 2013-04-09T15:20:54 Yeah, we still got 3 weeks for that, don't we ;) 2013-04-09T15:21:35 Yep. And working now let's mentors spend more time with you. I promise that a student who first pops up Thursday before the deadline will not get near the amount of feedback you will get by starting early 2013-04-09T15:25:26 :) I've seen you have been co-mentor on last years ARINC 653 API project. I am building on the results of that with the paravirtualization project. 2013-04-09T15:25:46 Did you find the code or do we need to help dig? 2013-04-09T15:26:03 Gedare already sent me a link 2013-04-09T15:27:03 But the goal of this project was slightly different to what I have in mind 2013-04-09T15:28:18 but I haven't read enough about POK and ARINC, so that might change 2013-04-09T15:30:32 So if you are interested in the mentoring the paravirt project, too, I would be happy to send you a link to the proposal 2013-04-09T15:31:09 how so.. The first order goal was to get RTEMS runnign in a Pok ARINC 653 partition. With a paravirtualized interface to the Pok hypervisor. Although the end of the project would given an RTEMS ARINC-653 FOSS solution, all the work was on paravirtualization 2013-04-09T15:31:17 I am always intersted in things like this. 2013-04-09T15:34:31 I just take your oarcorp.com address or does it need to be a google one? 2013-04-09T15:53:02 joel.sherrill gmail.com 2013-04-09T15:53:13 the google drive stuff needs a gmail address 2013-04-09T15:55:32 thanks, you should've got a mail. 2013-04-09T15:56:47 And I added folks to the review list. :) 2013-04-09T16:00:14 Unfortunately ARINC 653 is an expensive standard. Do you have access to it through your university library? I can give you links and/or full titles 2013-04-09T16:01:11 https://www.arinc.com/cf/store/catalog.cfm?prod_group_id=1&category_group_id=3 list the 600 series standards. 2013-04-09T16:01:18 I haven't checked, yet. 2013-04-09T16:02:02 you are looking at 653P1-3 primarily But P2 (extended services) and P3 (testing 653) are probably useful if they have them or can get them 2013-04-09T16:03:22 * DrJoel is thrilled to know the Bartendro Kickstarter project met its goal.. Now I need to naem a drink and provide the recipe 2013-04-09T16:04:03 wohow, 400$ ... that's a lot 2013-04-09T16:04:07 *** MegaAlex is now known as MegaAlex|away 2013-04-09T16:04:23 yeah. hence the immediate recommendation for a university library :) 2013-04-09T16:10:29 well, at least not the library directly, maybe I can find one in a chair library. 2013-04-09T16:14:28 The -3 is fairly recent but the -2 isn't that different for understanding purposes 2013-04-09T16:20:37 Thanks I try to find one tomorrow. 2013-04-09T16:24:09 it's getting late. thanks for the help :) 2013-04-09T16:25:02 night.. just ask questions 2013-04-09T16:26:18 *** phipse has quit IRC (Quit: leaving) 2013-04-09T16:42:44 *** rokka has quit IRC (Ping timeout: 240 seconds) 2013-04-09T16:43:50 *** rokka has joined #rtems 2013-04-09T16:52:57 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2013-04-09T16:53:38 *** kiwichris has joined #rtems 2013-04-09T17:04:27 *** antgreen has joined #rtems 2013-04-09T17:27:44 * DrJoel should have locked off but is back 2013-04-09T17:53:26 Now to pick up Justin.. for real 2013-04-09T17:53:35 And we still need that drink named RTEMS :) 2013-04-09T17:53:48 Eric N. Suggested E for Ethanol.. but I don't think that's something most bars have 2013-04-09T17:53:52 *** DrJoel has quit IRC (Quit: Depression is merely anger without enthusiasm) 2013-04-09T18:43:07 *** SYCrane has joined #rtems 2013-04-09T19:00:28 *** the9a3eedi has quit IRC (Ping timeout: 256 seconds) 2013-04-09T19:17:14 *** the9a3eedi has joined #rtems 2013-04-09T19:19:30 *** the9a3eedi-2 has joined #rtems 2013-04-09T19:23:09 *** the9a3eedi has quit IRC (Ping timeout: 276 seconds) 2013-04-09T20:19:50 *** SYCrane has quit IRC (Ping timeout: 256 seconds) 2013-04-09T21:25:24 *** the9a3eedi has joined #rtems 2013-04-09T22:01:15 *** SYCrane has joined #rtems 2013-04-09T22:02:33 *** the9a3eedi has quit IRC (Ping timeout: 240 seconds) 2013-04-09T23:29:58 *** arvind_khadri has joined #rtems 2013-04-09T23:42:57 *** the9a3eedi has joined #rtems 2013-04-10T00:28:01 *** dhananjay has joined #rtems 2013-04-10T00:28:19 Hi all 2013-04-10T00:46:41 *** Hesham has joined #rtems 2013-04-10T00:53:47 *** dhananjay has quit IRC (Ping timeout: 245 seconds) 2013-04-10T01:12:52 *** jassi has joined #rtems 2013-04-10T01:16:43 *** jassi has quit IRC (Client Quit) 2013-04-10T01:52:30 *** sebhub has joined #rtems 2013-04-10T01:52:48 good morning 2013-04-10T01:59:38 good morning 2013-04-10T02:06:46 morning 2013-04-10T02:22:41 *** Hesham has left #rtems 2013-04-10T03:05:24 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2013-04-10T03:27:43 *** MegaAlex|away is now known as MegaAlex 2013-04-10T03:28:39 *** arvind_khadri has joined #rtems 2013-04-10T03:37:58 *** freenix has joined #rtems 2013-04-10T04:09:04 *** the9a3eedi has quit IRC (Ping timeout: 252 seconds) 2013-04-10T04:47:11 *** Dinesh_R_T has joined #rtems 2013-04-10T05:09:58 *** the9a3eedi has joined #rtems 2013-04-10T05:29:42 *** MegaAlex is now known as MegaAlex|away 2013-04-10T06:11:54 hello, 2013-04-10T06:11:56 I am interested in working on the RTEMS System events task . It will be helpful to know what is the goal of the project. Also pointers on sheep/goat project that will help me go through with this will be helpful. 2013-04-10T06:12:10 Hi and welome 2013-04-10T06:12:31 Events are fast and have a low overhead so are nice to use. 2013-04-10T06:12:43 The classic API only defines 32 events. 2013-04-10T06:13:05 I have read the documentation for event manager 2013-04-10T06:13:14 what are system events? 2013-04-10T06:13:16 This means system level services such as the file system cache or the network stack have to borrow some events from the user to work 2013-04-10T06:13:54 A user could be in a socket read call and have a task that sends an event to the task in the network stack and things go wrong 2013-04-10T06:14:03 we can reserve some events for the system level services 2013-04-10T06:14:26 http://git.rtems.org/rtems/commit/?id=0edf263139088e8ac0ff1f0d52513f6fc85677d2 2013-04-10T06:14:57 We have defined in the user manual which you read that these are available for users. 2013-04-10T06:15:08 sebhub yeah that is the system level events 2013-04-10T06:15:09 Thanks seb 2013-04-10T06:18:03 i would like to know whats the change needed from the present code 2013-04-10T06:18:09 condition variables for the classic api would be a nice project 2013-04-10T06:18:19 Yeah great idea. 2013-04-10T06:18:59 the existing system events are fine from my point of view 2013-04-10T06:19:06 no need to change something 2013-04-10T06:19:47 Condition variables would be most welcome and we could remove the horrible disable pre-emption hacks 2013-04-10T06:20:17 sebhub, there must be a project for system events still listed somewhere 2013-04-10T06:20:35 i will have a look at the wiki 2013-04-10T06:20:44 *** antgreen has quit IRC (Ping timeout: 255 seconds) 2013-04-10T06:20:45 thanks 2013-04-10T06:20:45 there was also some event object project 2013-04-10T06:20:56 huh ? 2013-04-10T06:21:32 http://www.rtems.org/wiki/index.php/RTEMSSystemEvents 2013-04-10T06:21:45 this page is already up to date 2013-04-10T06:22:06 hm, should i remove it from the open projects list? 2013-04-10T06:22:17 Yes 2013-04-10T06:22:46 *** mgahlot has joined #rtems 2013-04-10T06:22:52 kiwichris sebhub Thanks for this dicussion 2013-04-10T06:23:18 Dinesh_R_T, interested in condition variables ? 2013-04-10T06:23:27 Adding them to the classic API ? 2013-04-10T06:23:30 yes 2013-04-10T06:23:56 We need to generate a project page how-ever until then here is a brief outline... 2013-04-10T06:24:17 yeah that would be great ! 2013-04-10T06:25:54 Look at the POSIX implementation and see how it maps to the score. 2013-04-10T06:26:08 Just looking I see things like _POSIX_Condition_variables_Wait_support which is not very score'ish 2013-04-10T06:27:39 sebhub, would a classic CV mean something implemented specific to the classic API. The POSIX CV looks specific to it. 2013-04-10T06:28:31 Dinesh_R_T, the first stage would be documentation for the API, implementation of the API and finally tests 2013-04-10T06:29:04 kiwichris okay. so i would have to mirror the POSIX CV Implementation 2013-04-10T06:30:00 That I do not know. You would need to mirror the specification of a CV how-ever it would relate to classic API resources such as a classic API semaphore 2013-04-10T06:30:24 It would be similar at a guess 2013-04-10T06:30:29 Joel would know more 2013-04-10T06:30:36 okay 2013-04-10T06:31:45 i will dig at the classic CV API 2013-04-10T06:32:15 *** prasannatsm has joined #rtems 2013-04-10T06:54:50 *** prasannatsm has left #rtems 2013-04-10T07:14:23 you should investigate if a priority inversion protection algorithm exists for CVs 2013-04-10T07:15:44 the hardest part of such a project will be the api definition 2013-04-10T07:17:41 the CONDVAR(9) api from freebsd is also interesting in this area 2013-04-10T07:17:44 http://www.freebsd.org/cgi/man.cgi?query=condvar 2013-04-10T07:18:02 it allows CVs with various lock objects (not only mutexes) 2013-04-10T07:24:15 okay 2013-04-10T07:51:18 *** gedare has joined #rtems 2013-04-10T07:51:58 *** mgahlot has quit IRC (Remote host closed the connection) 2013-04-10T07:52:16 *** mgahlot has joined #rtems 2013-04-10T07:59:52 *** Dinesh_R_T has left #rtems 2013-04-10T08:24:55 *** weiY has joined #rtems 2013-04-10T08:45:15 *** gedare has quit IRC (Ping timeout: 276 seconds) 2013-04-10T08:47:02 *** monstr_ has joined #rtems 2013-04-10T09:02:30 *** mgahlot has quit IRC (Remote host closed the connection) 2013-04-10T09:04:08 *** freenix has left #rtems ("Leaving") 2013-04-10T09:15:43 *** SYCrane has quit IRC (Quit: ChatZilla 0.9.90 [Firefox 20.0/20130326150557]) 2013-04-10T09:37:07 *** sevikkk has quit IRC (Ping timeout: 264 seconds) 2013-04-10T09:37:52 *** Hesham has joined #rtems 2013-04-10T09:45:52 *** Hesham has left #rtems 2013-04-10T10:29:15 *** gedare has joined #rtems 2013-04-10T10:46:35 the student left, but it might be better to provide score CV that supports posix and classic. this is more "RTEMS"-like. 2013-04-10T10:46:50 although solving the classic CV first and then refactoring is good too. if possible 2013-04-10T10:47:12 i'm just glad to see a student interested in a project that I'm not the likely mentor for! 2013-04-10T10:48:03 *** weiY has quit IRC (Ping timeout: 260 seconds) 2013-04-10T10:55:28 *** sebhub has quit IRC (Remote host closed the connection) 2013-04-10T11:06:01 *** DrJoel has joined #rtems 2013-04-10T11:06:01 *** DrJoel has joined #rtems 2013-04-10T11:06:01 *** ChanServ sets mode: +o DrJoel 2013-04-10T11:08:14 *** S_Somani has joined #rtems 2013-04-10T11:16:55 hey DrJoel there was an interesting conversation and new open project listed to add condition variables to the classic API implementation. 2013-04-10T11:17:10 and a student expressing interest in said project. 2013-04-10T11:17:28 That's a pet project of sebastian's :) 2013-04-10T11:17:49 makes sense, it was he and chris that were talking about it with the student. 2013-04-10T11:18:05 i was just glad to see a student interested in a project that is not under my purview! 2013-04-10T11:18:07 My view is that CVs need to migrate down to the supercore and both APIs derive their implementation. Avoid duplicating API code 2013-04-10T11:18:24 yeah i said this, but i was late to the party 2013-04-10T11:18:31 (by at least 6 hours I think) 2013-04-10T11:18:36 lol... 2013-04-10T11:18:47 it depends on how closely the posix and classic CVs are 2013-04-10T11:21:11 POSIX CVs are pretty simple. Hopefully we avoid code duplication 2013-04-10T11:21:25 What would a drink named RTEMS contain? Bartendro drink naming time. :) 2013-04-10T11:22:21 i saw your fb post. i don't have any grand ideas 2013-04-10T11:23:50 Google #OSPO Google Hangout information session tomorrow on #GSOC and archived on YouTube afterwards. 2013-04-10T11:23:50 https://plus.google.com/events/cpooa4srhdkp7o6ttsu88tcuhl4 2013-04-10T11:26:30 rum, tonic, espresso, maple syrup 2013-04-10T11:26:31 ;-) 2013-04-10T11:26:54 busy tomorrow touristy stuff 2013-04-10T11:27:39 raspberries, triple sec, eggnog, maraschino cherries, sake 2013-04-10T11:27:53 and, i'm out! 2013-04-10T11:28:05 *** gedare has quit IRC (Quit: Leaving) 2013-04-10T11:32:41 *** zjedwin has joined #rtems 2013-04-10T11:34:07 *** sevikkk has joined #rtems 2013-04-10T11:51:18 *** zjedwin has quit IRC (Ping timeout: 245 seconds) 2013-04-10T12:02:44 * DrJoel looks around 2013-04-10T12:07:39 *** monstr_ has quit IRC (Ping timeout: 252 seconds) 2013-04-10T12:12:45 DrJoel: I did some work on application GUI. Created a sample script which loads configuration parameters dynamically. Solves all the 3 cases you specified in the mail 2013-04-10T12:16:27 Cool! What format was the input for the parameter definition in? 2013-04-10T12:17:12 I took it in text. Is it acceptable? Does it need to be XML? 2013-04-10T12:17:43 We haven't defined anything. It is an open topic. The need to do the UI dynamically with no hard coding is the key requirement. 2013-04-10T12:17:50 okay :) 2013-04-10T12:18:10 This is the link to the sample text parser http://pastebin.com/FWfcCe6m 2013-04-10T12:18:25 Personally I would like to see the conf.t file (doc/user/conf.t) documented in regular enough fashion that all the information could be derived from that 2013-04-10T12:18:41 Ok :) 2013-04-10T12:18:58 could you please review this flow of control https://docs.google.com/file/d/0B41ApxXt-m4jYkgwOEdQRnhfcEU/edit 2013-04-10T12:19:17 this IRC client sucks at clicking on URLs.. hard to do. :( 2013-04-10T12:19:34 should I mail the links? 2013-04-10T12:20:00 no .. I am getting there. Did you share the Google doc with my gmail address? That puts it in my doc list 2013-04-10T12:20:27 I shared this publically 2013-04-10T12:20:34 so i guess anyone can see it 2013-04-10T12:20:42 I am also having trouble. The end of a finger on my right hand has a tender spot. I think I have bruised it or something. Wearing a bandaid for padding so it can heal and my typing is slow 2013-04-10T12:21:21 Up to you.. but we only ask that it be editable by possible mentors/reviewers. Just in case there is stealing of ideas. 2013-04-10T12:22:19 what would anyone gain by stealing my idea...in any case it is for the benifit of the community :P 2013-04-10T12:22:42 OK.. I just am sensitive to plagarism 2013-04-10T12:22:48 what is the format of config-v1.txt 2013-04-10T12:23:05 its text only 2013-04-10T12:23:44 Could the conf.t chapter be processed by a program to produce it? Even if we had to add comments 2013-04-10T12:24:28 FYI there has to be some grouping of related parameters.. the @section should be enough to derive that. 2013-04-10T12:24:55 Ok..will modify accordingly 2013-04-10T12:25:13 this was just a naive script to get a hang of the basic idea 2013-04-10T12:25:25 And FWIW there is a tool "bmenu" -- build menu which reads .t files to insert @node links automatically for next, prev, up, etc. It may or may not be of use. It is in C and old 2013-04-10T12:26:07 Prototyping and evolving is what we need to do. But the conf.t rewrite was trying to make the UI input easier to derive. If we need to do more and make it more rigid, that's ok. 2013-04-10T12:26:45 Probably easier to be rigid in conf.t so it can be processed, than to convert chunks to XML and write an XML processor to produce the conf.t and UI input. 2013-04-10T12:27:24 We could add "Question" to the parameter desciription if needed. "How many Classic API tasks do you want to configure?" 2013-04-10T12:27:36 Everything is on the table. 2013-04-10T12:28:37 I'll do some research on bmenu and see if it can be put to use 2013-04-10T12:29:47 I honestly don't know if it is useful. I think this should all be in Python. :) 2013-04-10T12:30:00 Ok . Would WxPython be fine on the GUI front? 2013-04-10T12:30:00 And I wrote bmenu would wouldn't consider it great code either. 2013-04-10T12:31:08 That needs to be discussed on the list. kiwichris wants to make sure we have a text based UI and I think we all would really like to have a TUI and GUI version. Identifying implementation paths for both in Python would be good. That way, the implementation could be sharing a lot of code. 2013-04-10T12:32:09 *** arvind_khadri has quit IRC (Remote host closed the connection) 2013-04-10T12:35:04 S_Somani: One of our top Google Code In students is not 18 yet so can't do GSOC but is interested in this project and offered to help. I have no idea how to take advantage of that offer yet but wanted to pass it along 2013-04-10T12:37:25 DrJoel: Umm... I have no clue on how to respond to that :P 2013-04-10T12:38:25 Me either. Make a plan and then we will go from there. He has to graduate and do AP exams so is too busy to do anything for a while. 2013-04-10T12:39:44 DrJoel: If you want an additional eclipse implementation then probably he could work on that as I have no clue of how things would go in eclipse 2013-04-10T12:40:49 Me either.. Sebastian did the current plugin. But that's the way to think.. Given a core set of code and definitions.. only the final UI should vary 2013-04-10T12:54:00 DrJoel: I have a small doubt, how is conf.t different from the config-v1.pdf u sent me via mail? 2013-04-10T12:54:24 It is the input file for that 2013-04-10T12:54:28 cd rtems/doc 2013-04-10T12:54:31 ../bootstrap 2013-04-10T12:54:34 ./configure 2013-04-10T12:54:35 make 2013-04-10T12:54:41 ls user/c_user.pdf 2013-04-10T12:55:06 assuming you have tools installed to build docs 2013-04-10T12:56:13 umm...so i have to read and modify conf.t and make it more rigid...correct? 2013-04-10T12:58:58 maybe... :) 2013-04-10T12:59:43 texinfo is pretty rigid and we only use a small set of the markup 2013-04-10T13:00:16 Ok. :) 2013-04-10T13:01:25 It may not take much more than adding a "@c BEGIN PARAMETER" and "@c END PARAMETER" The content between those two points is based on a template 2013-04-10T13:01:52 Then noting what @section you are in while reading forward in the file 2013-04-10T13:03:31 Ok. Thanks for helping out :) 2013-04-10T13:15:59 *** MegaAlex|away is now known as MegaAlex 2013-04-10T13:34:11 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-10T14:51:10 *** arvind_khadri has joined #rtems 2013-04-10T15:34:54 *** antgreen has joined #rtems 2013-04-10T15:44:55 antgreen were you the one who tried -ffunction-sections? 2013-04-10T15:45:13 yes, it works fine 2013-04-10T15:45:23 I think I've got it as the standard build option for the moxie port 2013-04-10T15:45:33 Did it make much difference? 2013-04-10T15:45:44 I thought so, but I don't recall of the top of my head now. 2013-04-10T15:45:47 I have tried a powerpc and the sparc/sis BSPs and didn't get more than 32 bytes on any test 2013-04-10T15:45:57 I thought you had a big difference 2013-04-10T15:46:04 That's what I remember. 2013-04-10T15:46:20 Still, it may be useful for real user applications 2013-04-10T15:46:27 where the code is not so carefully partitioned. 2013-04-10T15:46:51 But I guess that doesn't impact how RTEMS is built. 2013-04-10T15:47:18 I remember when eCos was built with -ffunction-sections it made a huge differerence for sure 2013-04-10T15:47:25 fileio.exe is ~500K exe and there was no change in size (543840 both) 2013-04-10T15:48:06 oh.. linker argument.. hold on 2013-04-10T15:48:38 btw - I'm struggling w/ the memory controller now on the FPGA because RTEMS programs won't fit in the on-board FPGA RAM. Once I sort that out (I'm close!) then I'll run my first RTEMS app on h/w. 2013-04-10T15:49:05 very cool!!! 2013-04-10T16:02:40 *** gedare has joined #rtems 2013-04-10T16:05:08 gedare.. got a second? 2013-04-10T16:09:09 DrJoel: hi 2013-04-10T16:09:25 DrJoel: playing put the baby down for a nap... right now i'm winning, but who knows 2013-04-10T16:09:33 hey.. I thought I had the connection on melange worked out. 2013-04-10T16:09:42 But all I did was manage to delete you :( 2013-04-10T16:09:55 Can you try to initiate the connection to be org-admin again? 2013-04-10T16:10:09 hah 2013-04-10T16:10:19 i kinda figured that would be the "solution" anyway btw 2013-04-10T16:10:39 FYI if you are admin, then by assumption you are mentor. So no dupe role 2013-04-10T16:10:57 hmmmmm 2013-04-10T16:11:04 i think you b0rked my account! 2013-04-10T16:11:14 on the RTEMS org page it says i'm already a mentor 2013-04-10T16:11:15 that's what I was afraid of :( 2013-04-10T16:11:20 emailing melange 2013-04-10T16:11:20 but i don't have a "connection" available for RTEMS 2013-04-10T16:11:22 k 2013-04-10T16:11:29 be sure to include my link_id 2013-04-10T16:11:57 oh wait 2013-04-10T16:12:02 DrJoel: it just showed up now 2013-04-10T16:12:09 email didn't work .. link id did 2013-04-10T16:12:19 hmmmm 2013-04-10T16:12:22 I have four options 2013-04-10T16:12:27 accept/reject and mentor/OA 2013-04-10T16:12:31 guess i should accept OA? 2013-04-10T16:12:37 accept org admin and reject the others :) 2013-04-10T16:12:52 ok it looks like i'm an OA 2013-04-10T16:13:02 and implicitly a mentor. 2013-04-10T16:13:14 i see it all fine on the "members" page 2013-04-10T16:15:10 ok.. we need more mentors! ;) 2013-04-10T16:15:52 half the listed projects are oriented toward me... and i'm not convinced i want to primary ;-) 2013-04-10T16:16:02 yep we do need more mentors 2013-04-10T16:19:59 I sent an email but don't know who else I missed 2013-04-10T16:20:00 brb 2013-04-10T16:20:14 we need to fix kiwichris and cynthia to be org admins as well 2013-04-10T16:40:29 antgreen looks like that combination of options worked. smallest exe went from 82K to 4K. Largest from 543K to 511K. Some had 50% reductions. ticker dropped 30% 2013-04-10T16:42:02 10% smaller looks pretty consistent on test exe's larger than 200K on powerpc/icecube 2013-04-10T16:51:37 *** DrJoel has quit IRC (Quit: We be chillin - IceChat style) 2013-04-10T18:08:41 *** MegaAlex is now known as MegaAlex|away 2013-04-10T18:40:59 *** pkjha has joined #rtems 2013-04-10T18:41:03 Hi, 2013-04-10T18:41:14 I am gsoc aspirant and want to work with this organization. 2013-04-10T18:41:32 I have good testing skills, I am looking for some projects that requires improving / writing automated tests - unit testing in python or c. 2013-04-10T18:41:43 Is there any scope of such projects here? 2013-04-10T18:43:05 *** gedare has quit IRC (Read error: Operation timed out) 2013-04-10T18:44:11 I got it here http://www.rtems.org/wiki/index.php/Open_Projects who is the mentor for testing related projects? 2013-04-10T18:46:44 *** SYCrane has joined #rtems 2013-04-10T18:48:43 *** the9a3eedi has quit IRC (Ping timeout: 260 seconds) 2013-04-10T18:55:23 Hi, all. I want to participate in the GSoC, however I don't know whether should I propose a new one or just choose one? 2013-04-10T19:31:58 *** the9a3eedi has joined #rtems 2013-04-10T20:02:32 *** pkjha has quit IRC (Ping timeout: 245 seconds) 2013-04-10T20:18:18 SYCrane, we are always interested in new projects. What did you have in mind ? 2013-04-10T20:42:24 I study CAN recently, it used widely in industry. So I want to add the can driver to RTEMS, just like network driver, or other drivers 2013-04-10T20:44:00 I do not know if this is ok 2013-04-10T20:44:25 Are you talking about a CAN bus stack ? 2013-04-10T20:44:34 Yes 2013-04-10T20:45:09 A driver would be a little light weight, but a whole CAN API, stack and low level driver interface would be nice. 2013-04-10T20:45:28 Silly question, but I can't seem to find an answer to this: If a task allocates some memory using malloc, then the task gets deleted, what happens to the allocated memory? Is it reclaimed? 2013-04-10T20:45:50 No, it is leaked. RTEMS is a single address space. 2013-04-10T20:46:03 The pattern is normally to ask a task to delete itself. 2013-04-10T20:46:16 yeah, but I'm currently experimenting and I'm too lazy to do that lol 2013-04-10T20:46:31 It is the same model as threads in an application. 2013-04-10T20:47:01 Oh well if memory is cheap then do not bother, then again it will not run for long. 2013-04-10T20:47:14 kiwichris, that is what i want to do 2013-04-10T20:47:34 well, I'm spawning and deleting tasks to see how well it can handle that sort of thing 2013-04-10T20:47:47 SYCrane, excellent. I suggest you write this up as a proposal and send it to the list. Do not forget to include documentation 2013-04-10T20:48:20 the9a3eedi, RTEMS is deterministic. The times should be fixed no matter how many tasks exist 2013-04-10T20:48:51 I realize that, but this is more meant for demonstration purposes 2013-04-10T20:49:29 Years ago when testing a large C++ application I created about 10,000 tasks and 3000 sockets on a PC with 1G of RAM. 2013-04-10T20:50:00 that's a lot of tasks 2013-04-10T20:51:03 kiwichris, i am beginner, the list you said is rtems-devel mailing list? 2013-04-10T20:51:27 and also, how to include documentation? 2013-04-10T20:51:30 The rtems-user list is fine 2013-04-10T20:52:06 "The CAN API project shall include documentation for the user interface and the CAN bus low level driver." 2013-04-10T20:52:26 Just include a statement the project needs to provide it. 2013-04-10T20:54:03 the9a3eedi, you could randomally connect to any of the sockets and measured latency did not change from100 sockets and 300 threads. 2013-04-10T20:56:34 kiwichris, you mean after the CAN API have coded, the documentation should be provided, or before? 2013-04-10T20:58:02 Documenting the API first is a good idea. Implementing it will expose issues and require changes and finally the doco needs to match the finished code. 2013-04-10T20:58:18 You should also consider suitable tests for the code. 2013-04-10T20:58:58 you will need code to test the API you write so making them suitable for the testsuite means the API can be tested by anyone into the future 2013-04-10T21:00:15 ok, i know 2013-04-10T21:00:59 but i am not faimilar with the RTEMS, maybe it's a little hard for me to write a API ducumentation. 2013-04-10T21:05:32 i want to write a proposal firstly, just describe what i'm going to do, and define the CAN API latter. is it ok? 2013-04-10T21:07:05 oh, i mean a proposal for GSoC2013. 2013-04-10T21:07:14 Part of the project is to document the API, not part of the proposal 2013-04-10T21:07:30 Sorry if I was not clear about this 2013-04-10T21:07:50 I get it 2013-04-10T21:07:52 I do not expect the API in the proposal 2013-04-10T21:07:55 thanks 2013-04-10T21:09:08 Have you reviews the GSoC 2013 block on the www.rtems.org main page ? 2013-04-10T21:09:16 reviews -> reviewed 2013-04-10T21:10:10 back later 2013-04-10T21:11:10 kiwichris, nice. I guess I should be getting similar results with my experiment 2013-04-10T21:12:17 maybe I should just implement the whole "ask the task to delete itself" pattern 2013-04-10T21:13:35 kiwichris, i read it before, i will check it later. 2013-04-10T21:42:33 *** pram has joined #rtems 2013-04-10T21:45:56 *** pram_ has joined #rtems 2013-04-10T21:46:59 *** pram_ has quit IRC (Client Quit) 2013-04-10T21:47:03 *** pram has quit IRC (Client Quit) 2013-04-10T21:50:03 Is there any way to have the RTEMS shell call any C function that was written in the user code? kinda like how vxworks does it 2013-04-10T21:50:13 or does it have to be configured manually? 2013-04-10T21:50:31 The shell has an API that allows you to add user commands. 2013-04-10T21:50:39 Will find an example ... 2013-04-10T21:51:05 I've seen that, but it requires that I fill out a structure for every command 2013-04-10T21:51:20 whereas I was wondering if there's a way to call any C function I've written via the shell directly 2013-04-10T21:51:24 http://git.rtems.org/chrisj/rtl.git/tree/main.c#n429 2013-04-10T21:51:37 No it does not 2013-04-10T21:52:08 I see 2013-04-11T00:01:55 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2013-04-11T00:14:55 *** antgreen has quit IRC (Ping timeout: 264 seconds) 2013-04-11T00:16:46 *** arvind_khadri has joined #rtems 2013-04-11T00:30:34 *** monstr_ has joined #rtems 2013-04-11T01:01:21 *** mgahlot has joined #rtems 2013-04-11T01:25:50 *** Hesham has joined #rtems 2013-04-11T01:26:36 *** mgahlot has quit IRC (Remote host closed the connection) 2013-04-11T01:33:39 *** kiwichris has quit IRC (Quit: This computer has gone to sleep) 2013-04-11T01:34:33 *** kiwichris has joined #rtems 2013-04-11T01:55:43 *** SYCrane has quit IRC (Ping timeout: 256 seconds) 2013-04-11T01:56:37 http://www.rtems.org/wiki/index.php/MMU_Support was looking at this and was wondering what the status of it was.. does RTEMS still not have MMU support? 2013-04-11T01:57:28 *** SYCrane has joined #rtems 2013-04-11T01:57:31 It was a GSoC project last year. It is on going. 2013-04-11T01:58:16 MMU management is just on area of the whole design. If you step back from the specifics of the MMU you need to think about what you want to do with it. 2013-04-11T01:58:55 it's not really essential for certain embedded stuff though, I think. Probably just makes things more complicated 2013-04-11T01:59:05 You can do application specific management but is usually just some custom set up to protect some memory or to manage cache mappings 2013-04-11T01:59:22 I was just curious though 2013-04-11T01:59:23 There is a domain of applications where it is important. 2013-04-11T02:00:17 true 2013-04-11T02:00:21 In an embedded system you may have a sub-system of code and that code has a few threads and you may want to provide restricted access to a peice of memory 2013-04-11T02:00:41 We have called the peice of memory an arena. 2013-04-11T02:01:03 You alloc is somehow and then set the attributes and then allow specific tasks access 2013-04-11T02:01:39 This is an API. Each processor has specific limitations. For example the number of TLBs in the device 2013-04-11T02:01:58 you do not want an exception to load a TLB in a real-time system 2013-04-11T02:02:51 http://www.rtems.org/pipermail/rtems-devel/2012-May/001182.html 2013-04-11T02:04:12 *** sebhub has joined #rtems 2013-04-11T02:04:39 good morning 2013-04-11T02:04:42 I see. I'm not exactly very familiar with MMUs so the stuff you write is insightful. 2013-04-11T02:04:46 morning 2013-04-11T02:29:58 *** the9a3eedi has quit IRC (Read error: Connection reset by peer) 2013-04-11T02:30:23 *** the9a3eedi has joined #rtems 2013-04-11T02:36:03 the9a3eedi: I was the student who worked on the project last year, and hopefully would continue working on it this year 2013-04-11T02:39:36 Hesham, I've seen you on IRC earlier. how complex is the work so far? 2013-04-11T02:39:37 kiwichris: According to Ric feedback we discarded task/arena relationship last year, it was only about memory protection, So the design at the rtems-devel discussion is out-dated. 2013-04-11T02:41:05 Currently we want to give a unifrom API for handling memory protection / Caching whatever the memory unit is (MMU,MPU, segmenets, etc) 2013-04-11T02:42:39 There are few BSPs that support MMU currently. PPC (mpc6xx) and gumstix/arm920 through our libmm API 2013-04-11T02:45:40 The main challenge is to handle architectures variants and how they apply memory protection. For example, different page sizes, Blocks, Regions, Segments or whatever the memory unit protection is. 2013-04-11T02:48:09 Hesham, was that because the concept was too high level and too big for a GSoC project ? 2013-04-11T02:49:00 Hesham, I will be interested to see how you manage the variant and differences without a solid top level abstraction as an API. 2013-04-11T02:49:11 *** S_Somani has joined #rtems 2013-04-11T02:49:52 Hesham, where is it document that the design was given away ? 2013-04-11T02:49:52 Ric said tasks data structures do not have to embed anything about Arena, He also suggested to just create a static mapping 2013-04-11T02:50:52 What is the use case for the static mappings ? 2013-04-11T02:51:10 We have three APIs that helps abstracting the low-level API 2013-04-11T02:51:33 What is the top level API ? 2013-04-11T02:52:40 kiwichris: https://github.com/heshamelmatary/rtems/tree/master/testsuites/libtests/mmtest1 2013-04-11T02:52:55 that's a use case 2013-04-11T02:53:11 from my GSoC 2012 repo 2013-04-11T02:53:22 What concerns me is the wording "So the design at the rtems-devel discussion is out-dated". It not relate to the GSoC project but it is a valid concept. 2013-04-11T02:53:59 That is just setups to configure memory management. I am referring to the how an application uses this in a dynamic manner 2013-04-11T02:54:31 The API here is important and there is a real need. 2013-04-11T02:54:57 What I am talking about is the linking of memory management to task and context switching. 2013-04-11T02:55:03 I did not mean it's not valid, only we did not implement it, sorry for misunderstanding 2013-04-11T02:55:20 I am fine with not doing it, and breaking the task down into smaller parts 2013-04-11T02:56:21 Some RTOS systems provide protected sub-systems where tasks operate in a pseudo-fork type setup. 2013-04-11T02:56:43 This is a runtime model I do not like and we can do better. 2013-04-11T02:56:44 Yes I did not work on the linking between task/memory after ric feedback and some discussions with gedare 2013-04-11T02:57:12 Great 2013-04-11T02:57:15 *** arvind_khadri has quit IRC (Ping timeout: 276 seconds) 2013-04-11T02:57:30 That part is hard because of being in a kernel hot spot. 2013-04-11T02:58:08 Gedare preferred that we work on providing an abstract well-designed API for memory management 2013-04-11T02:58:32 Yes and I agreed. Without that being stable nothing else is worth considering. 2013-04-11T02:59:00 It was a goal for the last year and I think I was able to achieve it. 2013-04-11T02:59:09 Great. 2013-04-11T02:59:31 Was your email about the 1M vs smaller ARM allocations ? 2013-04-11T02:59:49 That's why I intend to propose a project this year that implement that API for different targets 2013-04-11T03:00:01 Yes that was me 2013-04-11T03:03:31 I am happy with 1M or 1K for the loader. The critical issue is not [ab]using the MMU resources. I am not sure how this will fall out. We need more info about the what can be done to know. 2013-04-11T03:04:37 It might be the user sets up a loader allocator that ha 1, 2 or 3 megs of memory and then the allocator dishes it out as read/execute memory 2013-04-11T03:04:46 brb 2013-04-11T03:09:08 But what if the application that need to be loaded is too small (few Kbs) and the RAM memory is relatively small 2013-04-11T03:09:44 *** monstr_ has quit IRC (Remote host closed the connection) 2013-04-11T03:10:10 I asked about that, because current code one-level page table can only support 1MB sections. 2013-04-11T03:27:07 *** monstr_ has joined #rtems 2013-04-11T03:37:42 *** lcpfnvcy has quit IRC (Ping timeout: 252 seconds) 2013-04-11T03:44:48 *** lcpfnvcy has joined #rtems 2013-04-11T03:55:45 *** S_Somani has quit IRC (Ping timeout: 276 seconds) 2013-04-11T04:02:02 If I wanted my tasks to do timeslicing, am I also supposed to manually configure a timer ISR to call rtems_clock_tick()? or is this done automatically? 2013-04-11T04:14:48 I g2g, will continue tomorrow 2013-04-11T04:19:54 *** the9a3eedi has quit IRC (Ping timeout: 264 seconds) 2013-04-11T04:20:21 *** mgahlot has joined #rtems 2013-04-11T04:42:34 *** antgreen has joined #rtems 2013-04-11T04:47:33 kiwichris: I'm interested in mentoring BeagleBone BSP works, however I'm a bit scared to do it alone 2013-04-11T04:48:15 would you be interested in co-metoring? 2013-04-11T05:04:52 *** S_Somani has joined #rtems 2013-04-11T05:18:26 *** the9a3eedi has joined #rtems 2013-04-11T06:01:36 *** arvind_khadri has joined #rtems 2013-04-11T06:21:57 *** mgahlot has quit IRC (Remote host closed the connection) 2013-04-11T06:52:59 peerst, it is difficult for me to say yes because it comes down to all the projects offered and accepted. It might Sebastian who helps. He knows the ARM code base well. 2013-04-11T06:53:31 That is not a no, more a case of us all being there to help 2013-04-11T06:58:41 my main concern is that I might get overloaded due to customer projects and then not having enough time. problem is that it is not easyly predictable 2013-04-11T06:59:54 peerst, I understand, I am in the same situation. 2013-04-11T07:00:08 yeah we all are proably 2013-04-11T07:00:45 mentoring is on extra time 2013-04-11T07:00:54 but sometimes there is no extra time 2013-04-11T07:00:56 yeap 2013-04-11T07:02:05 what would be important for me is that a BeagleBone GsOC project doesn't depend on me 2013-04-11T07:02:16 otherwise I'd rather not mentor 2013-04-11T07:02:25 and work on a BSP myelf 2013-04-11T07:02:38 because this I can postpone indefinitely if there is no time 2013-04-11T07:03:38 I understand. The mentoring effort is about help the student plan and complete the work they agree to do. The amount of time need can vary over the whole GSoC process. 2013-04-11T07:05:15 do you know how work on real targets is handled? how do the students get the targets? 2013-04-11T07:06:10 For this BSP I suppose they would have to. Maybe the beagleboard project would donate one to the student. I had not considered it. 2013-04-11T07:07:29 so Google is not sponsoring stuff like this? 2013-04-11T07:08:17 No. Either RTEMS would sponsor it or beagleboard. It would be worth sending Jason Kridner an email to find out. 2013-04-11T07:08:38 Beagleboard has been in GSoC in the past. 2013-04-11T07:11:15 ah ok 2013-04-11T07:12:13 well they are not that expensive, sonsoring one would be possible for me also 2013-04-11T07:14:09 And I would help also. 2013-04-11T07:15:05 do I need to do anything at this time if I want to mentor this? 2013-04-11T07:15:29 Just register to be a mentor. The details will be sorted out later. 2013-04-11T07:15:52 ok 2013-04-11T07:19:36 *** arvind_khadri has quit IRC (Read error: Operation timed out) 2013-04-11T07:36:48 *** antgreen has quit IRC (Remote host closed the connection) 2013-04-11T07:47:15 *** Hesham has left #rtems 2013-04-11T07:50:13 *** antgreen has joined #rtems 2013-04-11T08:38:06 *** DrJoel has joined #rtems 2013-04-11T08:38:06 *** DrJoel has joined #rtems 2013-04-11T08:38:07 *** ChanServ sets mode: +o DrJoel 2013-04-11T09:17:01 *** SYCrane has quit IRC (*.net *.split) 2013-04-11T09:40:20 * DrJoel looks around 2013-04-11T10:03:52 *** weiY has joined #rtems 2013-04-11T10:10:26 *** weiwei has joined #rtems 2013-04-11T10:12:35 *** weiY has quit IRC (Ping timeout: 276 seconds) 2013-04-11T10:26:45 Hi drjoel. 2013-04-11T10:26:54 *** weiwei is now known as weiy 2013-04-11T10:34:27 *** sebhub has quit IRC (Remote host closed the connection) 2013-04-11T10:45:21 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-11T11:01:24 *** weiy has quit IRC (Ping timeout: 264 seconds) 2013-04-11T11:38:27 *** S_Somani has joined #rtems 2013-04-11T11:40:32 *** MegaAlex|away has quit IRC (Ping timeout: 256 seconds) 2013-04-11T11:49:57 *** MegaAlex|away has joined #rtems 2013-04-11T13:02:20 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-11T13:48:21 *** mgahlot has joined #rtems 2013-04-11T14:17:19 * DrJoel looks around again 2013-04-11T14:31:42 *** arvind_khadri has joined #rtems 2013-04-11T14:42:27 https://plus.google.com/events/cpooa4srhdkp7o6ttsu88tcuhl4 2013-04-11T14:42:42 GSOC On Air Hangout 2013-04-11T15:31:02 *** monstr_ has quit IRC (Ping timeout: 252 seconds) 2013-04-11T16:55:26 *** antgreen has joined #rtems 2013-04-11T17:04:11 *** mgahlot has quit IRC (Remote host closed the connection) 2013-04-11T17:04:47 *** mgahlot|Away has joined #rtems 2013-04-11T17:05:51 *** mgahlot|Away is now known as mgahlot 2013-04-11T17:33:47 *** mgahlot is now known as mgahlot|away 2013-04-11T18:07:49 *** mgahlot|away is now known as mgahlot 2013-04-11T18:13:52 *** antgreen has quit IRC (Quit: Leaving) 2013-04-11T18:13:54 *** mgahlot|away has joined #rtems 2013-04-11T18:17:31 *** mgahlot has quit IRC (Ping timeout: 245 seconds) 2013-04-11T18:17:56 *** mgahlot|away is now known as mgahlot 2013-04-11T18:33:35 *** SYCrane has joined #rtems 2013-04-11T18:38:19 *** the9a3eedi has quit IRC (Ping timeout: 245 seconds) 2013-04-11T19:02:02 *** the9a3eedi has joined #rtems 2013-04-11T19:18:36 I'm having an issue with timeslicing.. It seems that if I set the ticks per timeslice configuration to anything greater than one, then timeslicing doesn't seem to work. I've also set microseconds per tick to 1000, but no dice. Is there anything else I should know? 2013-04-11T20:38:36 *** mgahlot is now known as mgahlot|away 2013-04-11T21:04:27 *** S_Somani has joined #rtems 2013-04-11T21:06:16 *** Robbie has joined #rtems 2013-04-11T21:10:27 *** Robbie has quit IRC (Ping timeout: 245 seconds) 2013-04-11T21:10:42 *** Robbie1 has joined #rtems 2013-04-11T21:17:00 *** Robbie has joined #rtems 2013-04-11T21:18:21 *** mgahlot|away has quit IRC (Ping timeout: 276 seconds) 2013-04-11T21:21:06 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-11T21:22:43 *** Robbie1 has quit IRC (Quit: Page closed) 2013-04-11T21:22:44 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2013-04-11T21:36:04 *** arvind_khadri has joined #rtems 2013-04-11T23:03:35 *** antgreen has joined #rtems 2013-04-12T00:00:11 *** arvind_khadri has quit IRC (Ping timeout: 258 seconds) 2013-04-12T00:28:46 *** monstr_ has joined #rtems 2013-04-12T01:16:43 *** S_Somani has joined #rtems 2013-04-12T01:21:03 *** S_Somani has quit IRC (Ping timeout: 258 seconds) 2013-04-12T01:25:06 *** S_Somani has joined #rtems 2013-04-12T01:25:58 *** arvind_khadri has joined #rtems 2013-04-12T01:42:08 *** S_Somani has quit IRC (Ping timeout: 258 seconds) 2013-04-12T01:43:17 *** S_Somani has joined #rtems 2013-04-12T01:44:29 *** sebhub has joined #rtems 2013-04-12T01:44:51 good morning 2013-04-12T01:56:08 *** MegaAlex|away has quit IRC (Ping timeout: 255 seconds) 2013-04-12T01:58:33 *** MegaAlex|away has joined #rtems 2013-04-12T02:03:14 *** S_Somani has quit IRC (Read error: Connection reset by peer) 2013-04-12T02:30:00 *** S_Somani has joined #rtems 2013-04-12T02:40:57 *** S_Somani has quit IRC (Ping timeout: 252 seconds) 2013-04-12T03:32:22 *** freenix has joined #rtems 2013-04-12T03:34:21 *** freenix has quit IRC (Client Quit) 2013-04-12T03:58:49 *** S_Somani has joined #rtems 2013-04-12T04:02:56 *** S_Somani has quit IRC (Ping timeout: 245 seconds) 2013-04-12T04:09:01 *** the9a3eedi has quit IRC (Quit: gotta run) 2013-04-12T06:18:21 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2013-04-12T07:01:57 *** mgahlot has joined #rtems 2013-04-12T07:02:17 *** mgahlot is now known as mgahlot|away 2013-04-12T07:14:15 *** Hesham has joined #rtems 2013-04-12T07:15:30 Hi 2013-04-12T07:16:01 Anyone knows what boards qemuppc BSP represents ? 2013-04-12T07:34:36 it is for qemu 2013-04-12T07:43:03 *** weiY has joined #rtems 2013-04-12T07:55:29 *** weiY has quit IRC (Ping timeout: 258 seconds) 2013-04-12T07:56:15 *** weiY has joined #rtems 2013-04-12T08:10:38 *** Hesham has left #rtems 2013-04-12T08:28:09 *** mgahlot has joined #rtems 2013-04-12T08:30:42 *** mgahlot|away has quit IRC (Ping timeout: 252 seconds) 2013-04-12T08:50:34 *** mgahlot is now known as mgahlot|away 2013-04-12T09:02:23 *** mgahlot|away has quit IRC (Ping timeout: 260 seconds) 2013-04-12T09:06:22 *** SYCrane has quit IRC (Remote host closed the connection) 2013-04-12T09:11:49 *** Robbie has quit IRC (Ping timeout: 246 seconds) 2013-04-12T09:16:47 *** mgahlot has joined #rtems 2013-04-12T09:22:27 *** mgahlot|away has joined #rtems 2013-04-12T09:24:58 *** mgahlot has quit IRC (Ping timeout: 252 seconds) 2013-04-12T10:16:43 *** mgahlot has joined #rtems 2013-04-12T10:48:23 *** Hesham has joined #rtems 2013-04-12T10:52:00 *** sebhub has quit IRC (Remote host closed the connection) 2013-04-12T11:13:07 *** weiY has quit IRC (Ping timeout: 252 seconds) 2013-04-12T11:24:21 *** dhananjay has joined #rtems 2013-04-12T11:55:40 *** mgahlot|away_ has joined #rtems 2013-04-12T11:55:42 *** mgahlot|away__ has joined #rtems 2013-04-12T11:57:46 *** mgahlot|away__ has joined #rtems 2013-04-12T11:58:21 *** mgahlot has quit IRC (Ping timeout: 245 seconds) 2013-04-12T11:58:48 *** mgahlot|away has quit IRC (Ping timeout: 260 seconds) 2013-04-12T12:11:36 *** Hesham has quit IRC (Quit: Leaving.) 2013-04-12T12:13:40 *** mgahlot|away__ has quit IRC (Remote host closed the connection) 2013-04-12T12:14:19 *** mgahlot has joined #rtems 2013-04-12T12:15:07 *** monstr_ has quit IRC (Ping timeout: 260 seconds) 2013-04-12T12:23:35 *** S_Somani has joined #rtems 2013-04-12T12:32:09 *** mgahlot has quit IRC (Read error: Connection reset by peer) 2013-04-12T12:32:14 *** mgahlot|away_ has quit IRC (Read error: Connection reset by peer) 2013-04-12T12:46:15 my talk at the Erlang user conference about porting Erlang on RTEMS is accepted :-) 2013-04-12T12:46:26 http://www.erlang-factory.com/conference/ErlangUserConference2013/speakers/PeerStritzinger 2013-04-12T12:51:54 *** mgahlot has joined #rtems 2013-04-12T12:58:02 *** mgahlot is now known as mgahlot|away 2013-04-12T13:02:44 *** mgahlot has joined #rtems 2013-04-12T13:05:51 *** mgahlot|away has quit IRC (Ping timeout: 245 seconds) 2013-04-12T13:10:09 peerst CONGRATULATIONS! If you want to write up something to post, we can make it news on rtems.org and put it out via social media 2013-04-12T13:17:48 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-12T14:21:38 *** DrJoel has quit IRC (Quit: Give a man a fish and he will eat for a day. Teach him how to fish, and he will sit in a boat and drink beer all day) 2013-04-12T14:39:03 *** mgahlot has quit IRC (Read error: Connection reset by peer) 2013-04-12T14:40:59 *** mgahlot has joined #rtems 2013-04-12T15:22:38 *** mgahlot|away has joined #rtems 2013-04-12T15:26:27 *** mgahlot has quit IRC (Ping timeout: 276 seconds) 2013-04-12T15:32:55 *** MegaAlex|away is now known as MegaAlex 2013-04-12T16:06:06 *** mgahlot|away has quit IRC (Ping timeout: 276 seconds) 2013-04-12T16:28:12 *** mgahlot has joined #rtems 2013-04-12T17:09:16 *** Hesham has joined #rtems 2013-04-12T17:31:05 *** mgahlot has quit IRC (Read error: Connection reset by peer) 2013-04-12T17:43:13 *** mgahlot has joined #rtems 2013-04-12T18:00:56 *** mgahlot has quit IRC (Read error: Connection reset by peer) 2013-04-12T18:03:46 *** mgahlot has joined #rtems 2013-04-12T18:34:40 *** mgahlot has quit IRC (Quit: No Ping reply in 180 seconds.) 2013-04-12T18:35:06 *** mgahlot has joined #rtems 2013-04-12T18:45:26 *** mgahlot has quit IRC (Ping timeout: 245 seconds) 2013-04-12T18:59:52 *** mgahlot has joined #rtems 2013-04-12T19:07:16 *** mgahlot has quit IRC (Quit: No Ping reply in 180 seconds.) 2013-04-12T19:10:04 *** mgahlot has joined #rtems 2013-04-12T19:16:42 *** MegaAlex is now known as MegaAlex|away 2013-04-12T19:18:49 *** SYCrane has joined #rtems 2013-04-12T19:28:09 *** Hesham has left #rtems 2013-04-12T19:55:40 *** SYCrane has quit IRC (Read error: Connection reset by peer) 2013-04-12T19:55:52 *** SYCrane has joined #rtems 2013-04-12T20:04:11 *** mgahlot has quit IRC (Ping timeout: 245 seconds) 2013-04-12T20:08:32 *** mgahlot has joined #rtems 2013-04-12T22:31:24 *** mgahlot is now known as mgahlot|away 2013-04-12T23:59:23 *** S_Somani has joined #rtems 2013-04-13T01:13:13 *** mgahlot|away has quit IRC (Ping timeout: 240 seconds) 2013-04-13T02:27:58 *** mgahlot has joined #rtems 2013-04-13T02:30:24 *** freenix has joined #rtems 2013-04-13T03:06:33 *** MegaAlex|away is now known as MegaAlex 2013-04-13T03:14:20 *** freenix has quit IRC (Quit: Leaving) 2013-04-13T04:00:53 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-13T04:01:27 *** Robbie has joined #rtems 2013-04-13T04:11:10 *** S_Somani has joined #rtems 2013-04-13T05:20:01 *** SYCrane has quit IRC (Read error: Connection reset by peer) 2013-04-13T06:50:33 *** S_Somani has quit IRC (Ping timeout: 252 seconds) 2013-04-13T07:09:56 *** mgahlot|away has joined #rtems 2013-04-13T07:13:47 *** mgahlot has quit IRC (Ping timeout: 260 seconds) 2013-04-13T08:29:51 *** Robbie_ has joined #rtems 2013-04-13T08:51:40 *** weiY has joined #rtems 2013-04-13T08:58:55 *** MegaAlex is now known as MegaAlex|away 2013-04-13T09:05:27 *** Robbie_ has quit IRC (Ping timeout: 245 seconds) 2013-04-13T09:05:51 *** Robbie has quit IRC (Ping timeout: 245 seconds) 2013-04-13T09:35:37 *** Hesham has joined #rtems 2013-04-13T10:04:18 *** sevikkk has quit IRC (Ping timeout: 245 seconds) 2013-04-13T10:17:03 *** S_Somani has joined #rtems 2013-04-13T10:21:46 *** S_Somani1 has joined #rtems 2013-04-13T10:24:37 *** sevikkk has joined #rtems 2013-04-13T10:24:42 *** S_Somani has quit IRC (Ping timeout: 256 seconds) 2013-04-13T10:50:51 *** phipse has joined #rtems 2013-04-13T11:21:13 *** S_Somani1 has quit IRC (Quit: Leaving.) 2013-04-13T11:21:25 *** S_Somani has joined #rtems 2013-04-13T11:22:59 *** S_Somani has quit IRC (Client Quit) 2013-04-13T11:24:01 *** S_Somani has joined #rtems 2013-04-13T11:24:14 *** weiY has quit IRC (Ping timeout: 276 seconds) 2013-04-13T11:59:52 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-13T12:05:34 *** sevikkk has quit IRC (Ping timeout: 256 seconds) 2013-04-13T12:12:03 *** sevikkk has joined #rtems 2013-04-13T12:16:16 *** sevikkk has quit IRC (Ping timeout: 245 seconds) 2013-04-13T12:19:35 *** sevikkk has joined #rtems 2013-04-13T12:20:58 *** Hesham has quit IRC (Ping timeout: 245 seconds) 2013-04-13T12:32:30 *** Hesham has joined #rtems 2013-04-13T12:36:23 *** S_Somani has joined #rtems 2013-04-13T12:56:34 *** phipse has left #rtems 2013-04-13T13:11:16 *** mgahlot has joined #rtems 2013-04-13T13:13:13 *** mgahlot|away has quit IRC (Ping timeout: 240 seconds) 2013-04-13T13:16:45 *** mgahlot has joined #rtems 2013-04-13T13:19:30 *** Hesham-- has joined #rtems 2013-04-13T13:24:24 *** MegaAlex|away has quit IRC (Quit: ZNC - http://znc.in) 2013-04-13T13:58:58 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-13T14:51:57 *** Hesham1 has joined #rtems 2013-04-13T14:55:10 *** Hesham has quit IRC (Ping timeout: 256 seconds) 2013-04-13T15:00:08 *** Hesham has joined #rtems 2013-04-13T15:03:48 *** Hesham1 has quit IRC (Ping timeout: 264 seconds) 2013-04-13T15:47:53 *** Hesham has quit IRC (Quit: Leaving.) 2013-04-13T16:57:48 *** mgahlot has quit IRC (Read error: Connection reset by peer) 2013-04-13T17:56:29 *** mgahlot has joined #rtems 2013-04-13T19:10:58 *** mgahlot|away has joined #rtems 2013-04-13T19:14:13 *** mgahlot has quit IRC (Ping timeout: 240 seconds) 2013-04-13T19:23:43 *** Robbie has joined #rtems 2013-04-13T19:47:29 *** Robbie has quit IRC () 2013-04-13T19:57:52 *** noselasd has joined #rtems 2013-04-13T20:17:40 *** mgahlot|away has quit IRC (Read error: Network is unreachable) 2013-04-13T20:17:56 *** mgahlot has joined #rtems 2013-04-13T20:21:37 *** mgahlot has quit IRC (Read error: Connection reset by peer) 2013-04-13T20:21:53 *** mgahlot has joined #rtems 2013-04-13T20:22:40 *** mgahlot is now known as mgahlot|away 2013-04-13T20:24:28 *** SYCrane has joined #rtems 2013-04-13T20:28:27 *** mgahlot|away has quit IRC (Ping timeout: 276 seconds) 2013-04-13T21:08:10 *** mgahlot has joined #rtems 2013-04-13T22:03:29 *** S_Somani has joined #rtems 2013-04-13T22:37:58 *** mgahlot|away has joined #rtems 2013-04-13T23:32:19 *** weiY has joined #rtems 2013-04-13T23:45:17 *** S_Somani1 has joined #rtems 2013-04-13T23:47:37 *** S_Somani2 has joined #rtems 2013-04-13T23:48:39 *** S_Somani has quit IRC (Ping timeout: 276 seconds) 2013-04-13T23:49:42 *** S_Somani1 has quit IRC (Ping timeout: 252 seconds) 2013-04-14T00:37:30 *** S_Somani2 has quit IRC (Quit: Leaving.) 2013-04-14T01:11:31 *** mgahlot|away_ has joined #rtems 2013-04-14T01:12:06 *** mgahlot|away__ has joined #rtems 2013-04-14T01:14:23 *** mgahlot|away has quit IRC (Ping timeout: 258 seconds) 2013-04-14T01:15:06 *** mgahlot has quit IRC (Ping timeout: 276 seconds) 2013-04-14T01:42:32 *** mgahlot has joined #rtems 2013-04-14T01:42:42 *** mgahlot|away has joined #rtems 2013-04-14T01:45:03 *** mgahlot|away__ has quit IRC (Ping timeout: 258 seconds) 2013-04-14T01:45:03 *** mgahlot|away_ has quit IRC (Ping timeout: 258 seconds) 2013-04-14T01:46:56 *** mgahlot|away has quit IRC (Read error: Connection reset by peer) 2013-04-14T01:49:06 *** mgahlot has quit IRC (Quit: No Ping reply in 180 seconds.) 2013-04-14T01:51:23 *** sevikkk has quit IRC (Quit: Leaving.) 2013-04-14T03:28:21 *** weiY has quit IRC (Ping timeout: 245 seconds) 2013-04-14T03:34:52 *** weiY has joined #rtems 2013-04-14T03:40:46 *** weiY has quit IRC () 2013-04-14T04:30:54 *** sevikkk has joined #rtems 2013-04-14T04:33:56 *** sevikkk has quit IRC (Read error: Connection reset by peer) 2013-04-14T04:34:10 *** sevikkk has joined #rtems 2013-04-14T06:13:20 *** mgahlot has joined #rtems 2013-04-14T06:13:25 *** mgahlot|away has joined #rtems 2013-04-14T06:37:42 *** S_Somani has joined #rtems 2013-04-14T06:53:23 *** MegaAlex|away has joined #rtems 2013-04-14T07:17:53 *** Robbie has joined #rtems 2013-04-14T07:18:38 *** Robbie has quit IRC (Client Quit) 2013-04-14T08:34:36 *** mgahlot|away has quit IRC (Ping timeout: 245 seconds) 2013-04-14T08:34:36 *** mgahlot has quit IRC (Ping timeout: 245 seconds) 2013-04-14T09:18:22 *** mgahlot has joined #rtems 2013-04-14T09:18:23 *** mgahlot|away has joined #rtems 2013-04-14T09:53:12 *** SYCrane has quit IRC (Read error: Connection reset by peer) 2013-04-14T10:22:54 *** sevikkk has quit IRC (Ping timeout: 264 seconds) 2013-04-14T10:25:26 *** sevikkk has joined #rtems 2013-04-14T10:46:10 *** weiY has joined #rtems 2013-04-14T11:06:47 *** weiY has quit IRC (Ping timeout: 260 seconds) 2013-04-14T13:04:08 *** mgahlot|away has quit IRC (Read error: Connection reset by peer) 2013-04-14T13:04:08 *** mgahlot has quit IRC (Read error: Connection reset by peer) 2013-04-14T13:45:55 *** S_Somani has quit IRC (Quit: Leaving.) 2013-04-14T15:19:32 *** vishrut009 has joined #rtems 2013-04-14T17:07:32 *** vishrut009 has quit IRC (Ping timeout: 252 seconds) 2013-04-14T17:09:36 *** vishrut009 has joined #rtems 2013-04-14T17:26:15 *** vishrut009 has quit IRC (Ping timeout: 252 seconds) 2013-04-14T18:19:12 *** mgahlot has joined #rtems 2013-04-14T18:25:53 *** MegaAlex|away has quit IRC (Ping timeout: 258 seconds) 2013-04-14T19:01:13 *** the9a3eedi has joined #rtems 2013-04-14T19:50:59 *** mgahlot has quit IRC (Ping timeout: 258 seconds) 2013-04-14T20:37:12 What's the difference between the RTEMS Monitor and the Shell? 2013-04-14T20:37:25 history 2013-04-14T20:37:35 the monitor existed before the shell 2013-04-14T20:37:51 I figured that was the case lol 2013-04-14T20:37:51 The shell is more generic while the monitor is just the monitor commands 2013-04-14T20:38:18 so I guess the monitor is a "simple shell"? 2013-04-14T20:38:42 Yes. 2013-04-14T20:40:07 so if the monitor is older, does that mean that I shouldn't be using it? 2013-04-14T20:40:16 or it's deprecated in any way? 2013-04-14T20:40:26 The shell has commands that map to the monitor 2013-04-14T20:40:41 It is just the CLI type interface that is the issue 2013-04-14T20:47:14 I see.. I'm just trying to figure out how I can resolve my issue with the shell interfering with the timeslicing of other tasks.. was hoping that the monitor could be an alternative maybe 2013-04-14T20:47:47 Lower its priority to below the other tasks. 2013-04-14T20:55:48 I tried that. The thing is the other tasks never block, so the shell doesn't get a chance to run. 2013-04-14T20:57:03 according to sebastian the shell goes into a busy loop waiting for input, so I'm looking to see if I could slightly modify the code a bit 2013-04-14T20:57:15 shouldn't this sort of thing be event based anyway? lol 2013-04-14T20:57:51 You need an interrupt driven console driver. 2013-04-14T20:58:22 so it's a BSP thing I guess? 2013-04-14T20:58:26 This is the only way to avoid dropping characters. 2013-04-14T20:58:28 Yes, 2013-04-14T20:58:37 I'll look there then, thanks 2013-04-14T22:32:14 *** mgahlot has joined #rtems 2013-04-14T23:25:32 *** mgahlot has quit IRC (Remote host closed the connection)