2012-02-06T00:34:52 *** arvind_khadri has joined #rtems 2012-02-06T01:32:25 *** lcpfnvcy has quit IRC (Quit: leaving) 2012-02-06T01:32:42 *** lcpfnvcy has joined #rtems 2012-02-06T02:20:51 *** sebhub has joined #rtems 2012-02-06T02:22:25 good morning 2012-02-06T03:28:08 *** arvind_khadri has quit IRC (Ping timeout: 240 seconds) 2012-02-06T03:42:39 *** arvind_khadri has joined #rtems 2012-02-06T03:55:10 *** arvind_khadri has quit IRC (Read error: Connection reset by peer) 2012-02-06T03:56:27 *** arvind_khadri has joined #rtems 2012-02-06T05:33:12 *** arvind_khadri has quit IRC (*.net *.split) 2012-02-06T05:34:59 *** arvind_khadri has joined #rtems 2012-02-06T06:07:28 *** lcpfnvcy has quit IRC (Ping timeout: 245 seconds) 2012-02-06T06:35:36 *** kuzew has quit IRC (Read error: Operation timed out) 2012-02-06T06:35:42 *** kuzew has joined #rtems 2012-02-06T07:05:49 *** lcpfnvcy has joined #rtems 2012-02-06T08:52:04 *** antgreen has joined #rtems 2012-02-06T09:17:38 *** arvind_khadri has quit IRC (Ping timeout: 240 seconds) 2012-02-06T09:48:46 *** antgreen has quit IRC (Read error: Connection reset by peer) 2012-02-06T09:49:24 *** antgreen has joined #rtems 2012-02-06T09:52:09 *** lcpfnvcy has quit IRC (Ping timeout: 240 seconds) 2012-02-06T10:00:23 *** antgreen has quit IRC (Remote host closed the connection) 2012-02-06T10:04:26 *** antgreen has joined #rtems 2012-02-06T10:29:42 *** sebhub has quit IRC (Ping timeout: 255 seconds) 2012-02-06T10:30:32 *** mwalle has quit IRC (Read error: Operation timed out) 2012-02-06T10:31:56 *** mwalle has joined #rtems 2012-02-06T11:01:46 *** gedare has joined #rtems 2012-02-06T11:09:44 *** dr__house has joined #rtems 2012-02-06T11:11:00 *** arvind_khadri has joined #rtems 2012-02-06T11:13:21 *** lcpfnvcy has joined #rtems 2012-02-06T11:20:24 *** lcpfnvcy has quit IRC (Ping timeout: 245 seconds) 2012-02-06T13:02:01 *** kristianpaul has quit IRC (*.net *.split) 2012-02-06T13:02:01 *** rokka has quit IRC (*.net *.split) 2012-02-06T13:02:08 *** rokka has joined #rtems 2012-02-06T13:02:12 *** kristianpaul has joined #rtems 2012-02-06T13:03:37 *** arvind_khadri has quit IRC (Quit: Leaving) 2012-02-06T13:18:19 *** th_d has joined #rtems 2012-02-06T13:19:04 *** methril has joined #rtems 2012-02-06T13:48:45 *** methril has quit IRC (Quit: Leaving) 2012-02-06T14:14:42 *** th_d has quit IRC (Quit: When two people dream the same dream, it ceases to be an illusion. KVIrc 3.4.2 Shiny http://www.kvirc.net) 2012-02-06T14:17:08 *** gedare has quit IRC (Quit: Ex-Chat) 2012-02-06T16:57:19 *** methril has joined #rtems 2012-02-06T17:50:58 *** DrJoel has joined #rtems 2012-02-06T17:50:58 *** DrJoel has joined #rtems 2012-02-06T17:50:58 *** ChanServ sets mode: +o DrJoel 2012-02-06T18:12:09 *** DrJoel has quit IRC (Remote host closed the connection) 2012-02-06T18:16:39 *** xiangfu has joined #rtems 2012-02-06T18:36:28 *** DrJoel has joined #rtems 2012-02-06T18:36:28 *** DrJoel has joined #rtems 2012-02-06T18:36:28 *** ChanServ sets mode: +o DrJoel 2012-02-06T18:39:12 *** DrJoel has quit IRC (Client Quit) 2012-02-06T18:43:57 *** lcpfnvcy has joined #rtems 2012-02-06T20:03:01 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-02-06T20:20:50 *** xiangfu has joined #rtems 2012-02-06T20:40:41 *** xiangfu has quit IRC (Ping timeout: 256 seconds) 2012-02-06T20:53:48 *** xiangfu has joined #rtems 2012-02-06T21:05:29 *** A2Sheds has quit IRC (Ping timeout: 240 seconds) 2012-02-06T21:09:08 *** mwalle has quit IRC (Ping timeout: 240 seconds) 2012-02-06T21:11:13 *** A2Sheds has joined #rtems 2012-02-06T21:21:00 *** mwalle has joined #rtems 2012-02-06T23:05:45 *** antgreen has quit IRC (Ping timeout: 256 seconds) 2012-02-07T01:15:47 *** arvind_khadri has joined #rtems 2012-02-07T01:15:47 *** arvind_khadri has joined #rtems 2012-02-07T02:10:05 *** sebhub has joined #rtems 2012-02-07T02:27:11 good morning 2012-02-07T03:49:21 *** xiangfu has quit IRC (Remote host closed the connection) 2012-02-07T03:57:22 morning 2012-02-07T03:59:17 *** arvind_khadri has quit IRC (Ping timeout: 248 seconds) 2012-02-07T04:11:41 *** arvind_khadri has joined #rtems 2012-02-07T08:12:42 *** panzon has joined #rtems 2012-02-07T10:04:46 *** arvind_khadri has quit IRC (Ping timeout: 272 seconds) 2012-02-07T10:28:38 *** sebhub has quit IRC (Ping timeout: 240 seconds) 2012-02-07T11:15:44 *** DrJoel has joined #rtems 2012-02-07T11:15:44 *** DrJoel has joined #rtems 2012-02-07T11:15:44 *** ChanServ sets mode: +o DrJoel 2012-02-07T11:53:59 hello DrJoel 2012-02-07T11:55:36 DrJoel : we have an open hardware project EOMA http://elinux.org/Embedded_Open_Modular_Architecture/EOMA-68 and are considering RTEMS for an ARM STM32 that does some odd and ends using the libopenc3 libs 2012-02-07T11:55:52 we are over at #arm-netbook 2012-02-07T12:09:16 cool! shame sebhub isn't in here. He is probably the most familiar developer with arm who shows up on irc. I can answer general questions. 2012-02-07T12:09:28 signing up for rtems-users is a great way to get answers 2012-02-07T12:09:31 but ask away 2012-02-07T12:12:29 since the goal of the project includes keeping as much as possible open... 2012-02-07T12:13:53 we are using http://www.libopencm3.org/wiki/Main_Page firmware library for http://www.st.com/internet/mcu/class/1734.jsp 2012-02-07T12:14:30 at this point the question is " What's the status of support for STM32 and ARM in general"? 2012-02-07T12:15:33 ARM is well supported and an old target.. but there are lots of ARMs 2012-02-07T12:15:44 so need to check on the STM32 in particular 2012-02-07T12:15:46 these are mostly cortex M3 2012-02-07T12:17:07 the STM32 is there to act like an EC found in laptop and tablets to handle things like power management, lid open/closed, gpio etc 2012-02-07T12:17:08 We have a BSP for http://www.ti.com/product/lm3s6965 2012-02-07T12:17:44 ok, so thats M3 core 2012-02-07T12:18:00 there are several devs involved 2012-02-07T12:18:40 whatever peripherals 2012-02-07T12:18:49 we might just need to add drivers to rtems for new peripherals in the SOC's 2012-02-07T12:19:05 but os and supporting tools issues should be dealt with already.. 2012-02-07T12:19:13 great 2012-02-07T12:19:24 plus that bsp should have startup code, linker scripts, etc that might be of use directly or as examples 2012-02-07T12:19:28 we have some high volume interest in China 2012-02-07T12:19:45 we are trading Linux support for hardware access 2012-02-07T12:20:31 so we need have the stm32's supported for factories there to pretty much drop in and then have us provide some minor tweaks 2012-02-07T12:21:30 so rtems could instantly have 1M users in a few months, well rtems embeded in products 2012-02-07T12:23:22 any idea if 512KB of flash and only 96KB of internal RAM is enough space? 2012-02-07T12:24:18 the TI device has 256 kB flash and 64 kB SRAM. So it must be! 2012-02-07T12:26:57 *** noughth has joined #rtems 2012-02-07T12:30:07 That should be enough RAM/ROM depending on the functionality required. You could obviously write an app that wouldn't fit. :) 2012-02-07T12:30:39 I am interested in learning more about how RTEMS schedules threads, particularly how it decides which threads should be mapped to which cores on multicore hardware 2012-02-07T12:30:44 openEC would be very similar to what we would run 2012-02-07T12:30:54 what's the best way to learn more? 2012-02-07T12:31:20 http://wiki.laptop.org/go/OpenEC git clone http://dev.laptop.org/git/users/rsmith/ec-1.75 2012-02-07T12:39:34 *** methril has quit IRC (Read error: Operation timed out) 2012-02-07T12:41:06 noughth, Unfortunately for SMP stuff asking on the -devel list or reading the source. Highest priority N threads are spread across N cores in accordance with basic priority scheduling with FIFOs per priority 2012-02-07T12:41:18 scheduling algorithm is pluggable so you can replace it if you have a better one 2012-02-07T12:41:47 A2Sheds: I will look at that in a bit.. tracking down qemu/driver interaction and shoulders deep :( 2012-02-07T12:42:43 rjoel: no rush, once the cards are ready there will be several devs taking the plunge 2012-02-07T12:45:21 the main reason for rtems is for when the STM32 is used for audio playback and say the battery reaches charge, you don't want the audio playback interrupted while the battery charging routine sends a message over I2C to the main cpu 2012-02-07T12:47:27 DrJoel: Okay, I'll try to check out the source. 2012-02-07T12:47:38 This may be a stupid question, but where is the scheduler source? 2012-02-07T12:48:06 The README at http://git.rtems.org/rtems/tree/c/src/README seems to tell me to look in the nonexistant exec directory 2012-02-07T12:50:03 noughth, If you don't mind, please file a PR for that. :) c/src/exec has been gone a long time.. cpukit/score/ .. look for "scheduler*" .. should be simplesmp for the smp scheduler .. gedare is good on scheduling also 2012-02-07T12:50:25 A2Sheds, sebhub is the man for M3.. in Munich .. so watch his time for irc.. email -devel for questions 2012-02-07T12:50:48 ok 2012-02-07T13:25:49 *** methril has joined #rtems 2012-02-08T08:29:11 *** rtemsLogger has joined #rtems 2012-02-08T08:57:42 *** antgreen has joined #rtems 2012-02-08T09:50:27 *** harlan has quit IRC (Ping timeout: 252 seconds) 2012-02-08T09:57:46 *** harlan has joined #rtems 2012-02-08T10:26:22 *** sebhub has quit IRC (Read error: Operation timed out) 2012-02-08T12:27:11 *** lcpfnvcy has quit IRC (Read error: Operation timed out) 2012-02-08T12:27:23 *** lcpfnvcy has joined #rtems 2012-02-08T14:13:59 *** mumptai has joined #rtems 2012-02-08T16:04:16 *** arvind_khadri has joined #rtems 2012-02-08T16:15:57 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2012-02-08T16:24:35 *** mumptai has quit IRC (Ping timeout: 265 seconds) 2012-02-08T16:35:46 *** arvind_khadri has joined #rtems 2012-02-08T17:38:38 *** arvind_khadri has quit IRC (Ping timeout: 240 seconds) 2012-02-08T17:55:34 *** rokka has quit IRC (Ping timeout: 245 seconds) 2012-02-08T17:56:34 *** rokka has joined #rtems 2012-02-08T18:55:49 *** xiangfu has joined #rtems 2012-02-08T20:52:59 *** mumptai has joined #rtems 2012-02-08T21:18:48 *** xiangfu has quit IRC (Ping timeout: 252 seconds) 2012-02-08T21:25:15 *** xiangfu has joined #rtems 2012-02-08T21:35:21 *** kristianpaul has quit IRC (Quit: leaving) 2012-02-08T21:37:58 *** kristianpaul has joined #rtems 2012-02-08T22:46:52 *** antgreen has quit IRC (Read error: Connection reset by peer) 2012-02-09T01:18:29 *** xiangfu has quit IRC (Ping timeout: 245 seconds) 2012-02-09T01:37:13 *** xiangfu has joined #rtems 2012-02-09T01:54:02 *** sebhub has joined #rtems 2012-02-09T02:01:52 good morning 2012-02-09T02:04:12 morning 2012-02-09T02:06:31 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-02-09T02:19:06 *** xiangfu has joined #rtems 2012-02-09T02:31:48 *** mumptai has quit IRC (Ping timeout: 245 seconds) 2012-02-09T02:52:45 *** arvind_khadri has joined #rtems 2012-02-09T03:31:43 *** xiangfu has quit IRC (Ping timeout: 252 seconds) 2012-02-09T03:36:22 *** xiangfu_ has joined #rtems 2012-02-09T03:41:59 *** xiangfu has joined #rtems 2012-02-09T03:47:52 *** xiangfu has quit IRC (Ping timeout: 252 seconds) 2012-02-09T03:49:54 *** xiangfu has joined #rtems 2012-02-09T03:59:33 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2012-02-09T04:12:13 *** arvind_khadri has joined #rtems 2012-02-09T04:27:33 *** methril has quit IRC (Ping timeout: 248 seconds) 2012-02-09T04:40:07 *** methril has joined #rtems 2012-02-09T05:02:40 *** methril has quit IRC (Ping timeout: 252 seconds) 2012-02-09T05:14:56 *** methril has joined #rtems 2012-02-09T05:27:09 *** methril has quit IRC (Ping timeout: 252 seconds) 2012-02-09T05:39:34 *** methril has joined #rtems 2012-02-09T05:48:10 *** methril has quit IRC (Ping timeout: 276 seconds) 2012-02-09T06:01:37 *** methril has joined #rtems 2012-02-09T06:33:40 *** xiangfu has quit IRC (Ping timeout: 276 seconds) 2012-02-09T06:39:44 *** xiangfu has joined #rtems 2012-02-09T06:39:45 *** arvind_khadri has quit IRC (Ping timeout: 252 seconds) 2012-02-09T07:05:57 *** sevikkk has quit IRC (Ping timeout: 248 seconds) 2012-02-09T07:07:36 *** sevikkk has joined #rtems 2012-02-09T07:39:07 *** arvind_khadri has joined #rtems 2012-02-09T07:39:07 *** arvind_khadri has joined #rtems 2012-02-09T07:44:57 *** arvind_khadri has quit IRC (Ping timeout: 246 seconds) 2012-02-09T07:57:19 *** arvind_khadri has joined #rtems 2012-02-09T07:58:17 *** DrJoel has joined #rtems 2012-02-09T07:58:17 *** ChanServ sets mode: +o DrJoel 2012-02-09T08:27:18 *** xiangfu has quit IRC (Quit: Leaving) 2012-02-09T08:38:28 *** arvind_khadri has quit IRC (Ping timeout: 276 seconds) 2012-02-09T08:50:44 *** arvind_khadri has joined #rtems 2012-02-09T09:00:09 sebhub, Has A2Sheds asked you any questions? 2012-02-09T09:00:31 yes, we have chatted! 2012-02-09T09:41:21 cool! 2012-02-09T09:46:02 *** arvind_khadri has quit IRC (Ping timeout: 244 seconds) 2012-02-09T09:48:49 *** mumptai has joined #rtems 2012-02-09T09:52:53 *** methril has quit IRC (Ping timeout: 248 seconds) 2012-02-09T09:58:51 *** arvind_khadri has joined #rtems 2012-02-09T10:18:29 *** arvind_khadri has quit IRC (Ping timeout: 245 seconds) 2012-02-09T10:21:20 sebhub? 2012-02-09T10:21:47 ja 2012-02-09T10:22:06 are there netdb functions somewhere in the tcp stack ? 2012-02-09T10:22:20 like getnetbyaddr etc 2012-02-09T10:22:42 I'm trying to port stuff that does #include 2012-02-09T10:22:56 they should be available 2012-02-09T10:23:09 can't find them in the headers ... 2012-02-09T10:24:25 in the netdb.h? they are there 2012-02-09T10:25:32 do you have a netdb.h in your rtems install? 2012-02-09T10:25:48 yes 2012-02-09T10:25:54 there is one in the newlib sources subdir sys/linux/include ... 2012-02-09T10:26:22 no, it is part of the BSP 2012-02-09T10:26:40 yeah I see it in another install 2012-02-09T10:26:45 seems this one is broken 2012-02-09T10:27:10 every network specific stuff is part of the BSP (unfortunately) 2012-02-09T10:27:25 was a long time when I built this ... forgot if it worked ... 2012-02-09T10:27:59 oh and there also is a memory.h 2012-02-09T10:28:24 needs to redo the build 2012-02-09T10:28:42 which bsp do you use? 2012-02-09T10:31:16 was trying to build a lib without mentioning the bsp 2012-02-09T10:31:20 *** arvind_khadri has joined #rtems 2012-02-09T10:31:37 which can obviously not work 2012-02-09T10:32:44 is there a solution to build such generic libraries? 2012-02-09T10:34:04 you can use the multilib cpu options and copy the header files into the top level include directory 2012-02-09T10:34:41 what are the multilib cpu options?? 2012-02-09T10:35:07 the libc is generated with specific options 2012-02-09T10:35:16 ok 2012-02-09T10:35:33 see find /opt/rtems-4.11/powerpc-rtems4.11/ -name libc.a 2012-02-09T10:36:23 the actual options are defined in the gcc sources 2012-02-09T10:36:50 I see 2012-02-09T10:37:31 but aren't the header files bsp specific? 2012-02-09T10:38:00 all header files of the cpukit are not bsp specific in the powerpc arch 2012-02-09T10:38:14 so its arch dependent? 2012-02-09T10:38:38 yes, some arch ports of rtems are broken in this case 2012-02-09T10:39:13 it is probably easier to just start building for one or two bsps 2012-02-09T10:39:22 yes ;-) 2012-02-09T10:39:24 and build a multilib later 2012-02-09T10:39:37 or just forget about the multilib 2012-02-09T10:40:01 it is on my todo list, i need this for boost and the Google C++ testing framework 2012-02-09T10:40:44 to make porting non hardware dependent stuff easier? 2012-02-09T10:40:54 yes 2012-02-09T10:40:57 cool 2012-02-09T10:42:31 so if I build it for one powerpc bsp I can reuse it anyway? 2012-02-09T10:45:08 no 2012-02-09T10:45:20 the instruction set must match 2012-02-09T10:45:31 thus all the multi-libs 2012-02-09T10:45:59 have to go now, bye 2012-02-09T10:47:49 bye, thanks! 2012-02-09T10:50:06 *** sebhub has quit IRC (Ping timeout: 246 seconds) 2012-02-09T11:05:05 *** arvind_khadri has quit IRC (Ping timeout: 244 seconds) 2012-02-09T11:43:48 *** gedare has joined #rtems 2012-02-09T12:14:03 *** arvind_khadri has joined #rtems 2012-02-09T13:40:22 *** gedare has quit IRC (Quit: Ex-Chat) 2012-02-09T14:06:54 *** gedare has joined #rtems 2012-02-09T14:13:11 * gedare pokes DrJoel 2012-02-09T14:17:58 * DrJoel is up to his eyeballs in qemu and pain 2012-02-09T14:20:24 drjoel: read this blurb real quick give me your opinion 2012-02-09T14:20:31 RTEMS is an open source real-time operating system for embedded applications across a range of processors. RTEMS is used in deployed products in homes, factories, hospitals, particle accelerators, and space satellites. 2012-02-09T14:20:50 for the mentors page i guess they asked for a short description 2012-02-09T14:22:04 I have a file I keep the answers to the GSOC, GCI, and ESA SOCIS questions in .. hold on and I will share it with you. :) 2012-02-09T14:23:20 Sure. based on the others we just need two sentences that give an idea for students who might be scanning this wiki page 2012-02-09T14:24:01 Hmmm... you have access to it already :) 2012-02-09T14:24:21 ok i'll look through it for an appropriate sentence or two 2012-02-09T14:24:27 Feel free to add to it. It makes it easier to answer the same questions. 2012-02-09T14:24:31 what's it called 2012-02-09T14:24:40 RTEMS SOC Proposal Information 2012-02-09T14:24:49 last updated July 5 2011 if that helps 2012-02-09T14:24:57 got it thanks 2012-02-09T14:25:22 fyi I am going to Embedded World in Nuremberg and working a boot with Thomas, Sebastian, etc 2012-02-09T14:25:35 a boot huh 2012-02-09T14:25:38 sounds tasty 2012-02-09T14:25:55 lol.. booth would be more appropriate 2012-02-09T14:26:05 cool 2012-02-09T14:26:41 peerst, are you going? 2012-02-09T14:35:08 *** antgreen has joined #rtems 2012-02-09T14:35:20 RTEMS (Real-Time Executive for Multiprocessor Systems) is a free real-time operating system designed for deeply embedded systems including automotive, medical devices, science, industrial control, and space flight. RTEMS has been to Venus, circles Mars, is going to the asteroid belt, and can be found in high energy physics research labs around the world. 2012-02-09T14:35:24 Sending that. seems nice. 2012-02-09T14:35:33 hi 2012-02-09T14:35:47 hello kiwichris 2012-02-09T14:37:02 Just want to borrow a comment I came across yesterday .... 'OAR, You don't wanna be up a creek without them!' 2012-02-09T14:37:08 http://stackoverflow.com/questions/7888894/ada-multitasking-rtos-supported-with-opensource-cross-compilers 2012-02-09T14:37:10 lol 2012-02-09T14:37:13 made me laugh 2012-02-09T14:38:02 i like this too: RTEMS is very mature (it was open source before Linux was invented) 2012-02-09T14:38:22 Makes me feel old cause I was around then 2012-02-09T14:38:45 i was in ... grade school 2012-02-09T14:39:05 you are not making me feel any better ;) 2012-02-09T14:39:07 * gedare does some math 2012-02-09T14:39:34 probably rtems is older than most of our gsoc students. 2012-02-09T14:40:47 * kiwichris old man chuckle 2012-02-09T14:49:37 *** antgreen has quit IRC (Ping timeout: 252 seconds) 2012-02-09T15:11:46 rtems likely is 2012-02-09T15:15:21 how old is it? 2012-02-09T15:18:22 ahh, 1988 2012-02-09T15:22:08 The cvs history that is now in git goes back to May 1995. Before that it was a research project which offered the source freely. 2012-02-09T15:22:23 mumptai, you disappeared earlier this week before I had a chance to say hello 2012-02-09T15:25:17 mumptai, I think April 2004 if I remember 2012-02-09T15:26:42 DrJoel: do you still have the archives going back that far? 2012-02-09T15:27:01 releases images? 2012-02-09T15:27:02 it's not too late to add all the history going back to 1988 :) 2012-02-09T15:27:38 verm__, could you import tarballs ? 2012-02-09T15:27:40 no.. we don't really have that.. I have a few qic-150 tapes and... a printout of the entire RTEMS source as a technical report :) 2012-02-09T15:28:03 gsoc project, type it back in :) :) 2012-02-09T15:28:12 :) 2012-02-09T15:28:52 kiwichris: yep 2012-02-09T15:29:13 are there release images going back that far? 2012-02-09T15:29:16 kiwichris, drjoel: i added your IRC nicks to the mentor's page 2012-02-09T15:29:54 nah 2012-02-09T15:30:07 too bad 2012-02-09T15:32:42 I don't know if you can get it electronically but the entire source code was printed and included as a final report for one of the Army contracts. That made the code a matter of public record. I wish I could find it online. :) 2012-02-09T15:33:16 that'd be great. i could spawn a fork 2012-02-09T15:33:17 ;) 2012-02-09T15:33:58 gedare and esa... lol 2012-02-09T15:34:33 Amar... 2012-02-09T15:34:41 Subject: [Contact] RTEMS at github.com 2012-02-09T15:34:41 Hi Joel, 2012-02-09T15:34:41 I've made the username available, so you can use it. 2012-02-09T15:34:41 Cheers! 2012-02-09T15:34:41 Sonya 2012-02-09T15:34:50 is that what we needed? 2012-02-09T15:35:12 gedare, thanks 2012-02-09T15:37:23 oh 2012-02-09T15:37:25 who did that? 2012-02-09T15:37:44 DrJoel: you got that taken care of? 2012-02-09T15:38:21 * DrJoel takes a bow.. yes.. a polite email to support today .. and it is done 2012-02-09T15:38:36 what do we need to do to get rtems as an account? 2012-02-09T15:38:50 do we setup an rtems.org email alias for it? 2012-02-09T15:38:54 DrJoel: hi, sorry was not around when you pinged me earlier 2012-02-09T15:38:58 you emailed support?! damn i need to do that 2012-02-09T15:39:01 where did you email? 2012-02-09T15:39:13 web form on the bottom of the page 2012-02-09T15:39:21 thanks i will try that i'm going to take the name now 2012-02-09T15:39:44 DrJoel: how long did it take for them to respond? 2012-02-09T15:39:48 I assume the access information will be shared with at least Chris and I 2012-02-09T15:39:56 yep 2012-02-09T15:39:58 I did it today.. 2012-02-09T15:40:16 i'll do it when i wake up later thanks for the tip 2012-02-09T15:40:47 I have an email from someone who has large chunks of stuff to submit. I thought they would be a good candidate for the github hosted personal repo and we review/pull model you described 2012-02-09T15:41:18 https://github.com/rtems/ done, it's ours 2012-02-09T15:41:42 yep using github in this way will make things very easy 2012-02-09T15:41:48 i also wanted to get patches working for rtems 2012-02-09T15:42:00 http://patches.libav.org/project/libav-devel/list/ 2012-02-09T15:42:08 it's a system that listens to mailing lists and makes patches available 2012-02-09T15:42:15 http://patches.libav.org/patch/16687/ <-example patch 2012-02-09T15:42:19 you can approve/deny/commit them 2012-02-09T15:42:22 there's also a commandline tool 2012-02-09T15:42:29 that will let you pull and automatically apply patches (it does a git am) 2012-02-09T15:43:02 this way we can track submissions to the lists trivially and make sure they get taken care of (applied/denied) 2012-02-09T15:43:40 awesome. 2012-02-09T15:44:22 the commandline tool makes things much easier (works via XMLRPC) 2012-02-09T15:47:03 bye 2012-02-09T15:47:08 *** gedare has quit IRC (Quit: Ex-Chat) 2012-02-09T15:49:09 Cool! 2012-02-09T15:49:51 i'll get the github thing going this weekend you and chris should sign up 2012-02-09T15:50:17 I have an account on github 2012-02-09T15:50:22 what is it? 2012-02-09T15:50:31 https://github.com/joelsherrill 2012-02-09T15:50:49 Don't remember why I got it.. but have had it awhile.. better to get your name than not to. :) 2012-02-09T15:53:28 added 2012-02-09T15:53:46 oh chris is on there too 2012-02-09T15:54:07 yeah chrisjohns 2012-02-09T15:54:09 added him as well 2012-02-09T15:54:13 you 2012-02-09T15:54:21 No idea why I created. 2012-02-09T15:54:26 Forgot I had it :) 2012-02-09T15:54:31 I recall we had to have them to keep our Parrot commits with real names 2012-02-09T15:54:45 Ok 2012-02-09T15:58:33 dammit the save button is missing from the profile settings page 2012-02-09T16:01:58 i need to sleep i'll do this later 2012-02-09T16:45:19 *** mumptai has quit IRC (Ping timeout: 276 seconds) 2012-02-09T16:50:29 * DrJoel needs to head home 2012-02-09T16:54:43 DrJoel: catch you later 2012-02-09T16:54:46 bye 2012-02-09T16:54:58 bye 2012-02-09T17:09:06 *** sevikkk has quit IRC (*.net *.split) 2012-02-09T17:10:06 *** sevikkk has joined #rtems 2012-02-09T18:32:31 *** gedare has joined #rtems 2012-02-09T18:33:31 hi 2012-02-09T18:33:40 hi 2012-02-09T18:34:03 anything exciting happening? :) 2012-02-09T18:34:36 not really 2012-02-09T18:34:52 well no news is good news 2012-02-09T18:39:53 i was thinking about "arenas"; we should try to write/find some good definitions of what we might want 2012-02-09T18:42:33 An arena is a section of the address with specific memory attributes that can be shared by a tasks that are residents 2012-02-09T18:43:15 Access to an arena by tasks not residents is determined by the specific memory attributes 2012-02-09T18:43:30 What are the memory attributes? 2012-02-09T18:43:35 Note, a memory attribute can include address translation 2012-02-09T18:43:38 * gedare playing role of interviewer 2012-02-09T18:43:58 cache, read, write and virtual address and physical address 2012-02-09T18:44:13 execute also 2012-02-09T18:44:19 Ah yes 2012-02-09T18:45:04 Aside from being able to "join" an arena, what else might a task do with an arena, or what might an arena do with its tasks? 2012-02-09T18:45:55 At the lowest level not a lot. We leave the management of all memory to high layers. 2012-02-09T18:46:23 For example an allocator that works in an arena 2012-02-09T18:47:00 The full address space can be considered a single global arena with read,write, execute and the default cache set up 2012-02-09T18:47:27 I tend to use the wording 'set up the C arena' in boot code 2012-02-09T18:47:41 You may initialise it. 2012-02-09T18:47:46 So an arena can be viewed somewhat analagous to the Heap or Workspace? 2012-02-09T18:48:13 An area of memory that (some) tasks may share, but with the addition of attributes 2012-02-09T18:48:16 The whole idea is remove restrictions and let the memory management work with the application to solve system issues 2012-02-09T18:48:23 Yes. 2012-02-09T18:48:39 And without hard-coding an allocator that manages said memory region 2012-02-09T18:48:46 Well those structures operate in the global arena by default 2012-02-09T18:49:25 Do you view arenas as being subsets of other arenas, or as being distinct from each other? 2012-02-09T18:49:26 For example you could create an arena where the cache is disabled to allow access to DMA'ed data 2012-02-09T18:49:53 Not sure. It did cross my mind but I see TLB as flat and open. 2012-02-09T18:49:59 ((From a technical perspective it may be difficult to enforce subsets)) 2012-02-09T18:50:16 Yes and for what purpose ? To help the application fix errors. 2012-02-09T18:50:49 Or to provide for carving up a global arena into smaller arenas 2012-02-09T18:50:49 Maybe a option to enforce non-overlapping arenas may help. This is an allocation things and not runtime 2012-02-09T18:51:09 Yes that seems like it may be a requirement of the hardware 2012-02-09T18:51:14 a restriction and that is what I would like to avoid. This is low level stuff 2012-02-09T18:51:51 You get is wrong, boom. Once working in a given target it will always work. 2012-02-09T18:52:31 There is no point creating rules to help development only to have them eat cycles in production hardware that should never see overlap or subset issues 2012-02-09T18:52:33 I see. But does that imply that every hardware can support the same lowest-level mechanisms? 2012-02-09T18:52:47 Not sure. 2012-02-09T18:53:25 May the API is the superset and errors are returned when certain features are not supported 2012-02-09T18:53:55 If you select a device and then get an error cause a feature is not present, you have bigger problems 2012-02-09T18:54:14 like looking for a new job 2012-02-09T18:55:25 OK. The issue I see is that different MMU hardware will support different memory protection, but I agree that the API layer should handle the mismatch. 2012-02-09T18:55:43 Since these arenas are not for security a permissive approach will work 2012-02-09T18:55:49 Agrede. 2012-02-09T18:55:51