2013-11-04T00:32:23  *** 7CBAAB72H <7CBAAB72H!~monstr@nat-63.starnet.cz> has joined #rtems
2013-11-04T01:18:59  *** mwalle <mwalle!~mwalle@services.serverraum.org> has quit IRC (Ping timeout: 272 seconds)
2013-11-04T01:47:34  *** sebhub <sebhub!~sh@host-82-135-62-35.customer.m-online.net> has joined #rtems
2013-11-04T01:47:46  <sebhub> good morning
2013-11-04T02:16:41  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Ping timeout: 245 seconds)
2013-11-04T02:33:55  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-04T02:55:02  *** giusef <giusef!~Giuseppe@net-93-144-195-34.cust.dsl.teletu.it> has joined #rtems
2013-11-04T03:09:58  *** ysionneau <ysionneau!~fallen@ks3002254.kimsufi.com> has joined #rtems
2013-11-04T03:09:59  <ysionneau> hello
2013-11-04T03:10:47  <ysionneau> is it normal ubuntu packages in http://www.rtems.org/ftp/pub/rtems/linux/4.11/ubuntu/pool/rtems4.11/r/ seem to depend on binutils-common package, but this package is not available?
2013-11-04T03:11:00  <ysionneau> it seems to be available as RPM package though
2013-11-04T03:11:31  <ysionneau> *brb*
2013-11-04T03:25:15  *** guerby <guerby!~guerby@ip165-ipv6.tetaneutral.net> has quit IRC (Ping timeout: 252 seconds)
2013-11-04T03:26:31  *** guerby <guerby!~guerby@ip165-ipv6.tetaneutral.net> has joined #rtems
2013-11-04T03:38:51  <ysionneau> *back*
2013-11-04T04:48:15  *** giusef <giusef!~Giuseppe@net-93-144-195-34.cust.dsl.teletu.it> has quit IRC (Ping timeout: 272 seconds)
2013-11-04T05:03:35  <kiwichris> mkhoory: thanks; I will take a look tomorrow.
2013-11-04T05:58:33  *** asx15 <asx15!~asd@46.49.55.27> has joined #rtems
2013-11-04T05:58:43  <asx15> Hello
2013-11-04T05:58:59  <asx15> I have a very newbie question
2013-11-04T05:59:47  <asx15> Is it possible to run QtQuick on RTEMS OS?
2013-11-04T06:01:07  <asx15> Is there any GPU drivers i.MX QSB board? (AMD Z430 GPU)
2013-11-04T06:01:18  <sebhub> possible is a lot, but i don't think its pratical
2013-11-04T06:02:54  <asx15> sebhub: Actually I want to run any OpenGL accelerated UI engine (not exactly QtQuick), MyGUI (http://mygui.info/) also would be great. Just how to deal with GPU drivers?
2013-11-04T06:03:21  <sebhub> why not use linux for this?
2013-11-04T06:03:56  <asx15> sebhub: Linux has too poor kernel
2013-11-04T06:04:11  <asx15> sebhub: and also it is not a RTOS
2013-11-04T06:05:10  <sebhub> you can use graphics on rtems, but i think this requires a lot of work due to the lacking infrastructure support
2013-11-04T06:06:49  <asx15> sebhub: I can work a lot for that, just I have no idea where to get (or how to write myself) GPU driver.
2013-11-04T06:07:22  <sebhub> you can use the linux driver as a reference
2013-11-04T06:08:51  <asx15> sebhub: I tryied to use Solaris ARM port because of Solariss Japanese super quality kernel but it contains too much features unnecessery for me and it is too resource houngry and slow, in overall not the best option for Cortex A8
2013-11-04T06:11:32  <sebhub> rtems is a classic real-time os
2013-11-04T06:11:51  <sebhub> graphics support must be externally provided
2013-11-04T06:13:26  <asx15> sebhub: I am new with RTEMS, can you give me a link to any tutorial how to run RTEMS on any ARM board?
2013-11-04T06:15:38  <sebhub> i think we have not arm specific tutorials
2013-11-04T06:16:18  <sebhub> http://www.rtems.org/wiki/index.php/Board_Support_Packages
2013-11-04T06:16:55  <sebhub> we work currently on xilinx zync and altera cyclone v bsps
2013-11-04T06:18:59  <asx15> sebhub: :) Great, thank you. would report here if as soon as I would get any success with my ARM board
2013-11-04T06:19:13  <sebhub> which board do you have?
2013-11-04T06:19:30  <asx15> i.MX53 Quick Start Board
2013-11-04T06:23:21  <sebhub> i would use this bsp as a template:
2013-11-04T06:23:24  <sebhub> http://git.rtems.org/rtems/tree/c/src/lib/libbsp/arm/realview-pbx-a9/
2013-11-04T06:23:51  <sebhub> there is no out of the box support for this board
2013-11-04T06:29:51  <asx15> sebhub: Can you tell me in brief what I need to modify to get it running on board? heh, However I would add there UART debug after every code line to determine what I need to change.
2013-11-04T06:31:48  <sebhub> you need support for the low-level start, interrupt controller, console and clock tick
2013-11-04T06:33:00  <asx15> What did you mean in "low-level start", bootloader?
2013-11-04T06:33:27  <sebhub> something that intializes the memory, PLL, caches, etc
2013-11-04T06:37:07  <asx15> i.e. I must modify RTEMS initialization code for my board, however QEMU version would not contain this code, can you give me any other board BSP link, where from I can modify initialization code?
2013-11-04T06:39:36  *** sebhub <sebhub!~sh@host-82-135-62-35.customer.m-online.net> has quit IRC (Ping timeout: 246 seconds)
2013-11-04T06:39:48  *** sebhub2 <sebhub2!~sh@host-82-135-62-35.customer.m-online.net> has joined #rtems
2013-11-04T06:46:04  <asx15> seems I found everithyng needed :)
2013-11-04T06:49:41  *** asx151 <asx151!~asd@46.49.55.27> has joined #rtems
2013-11-04T06:51:45  *** asx15 <asx15!~asd@46.49.55.27> has quit IRC (Ping timeout: 272 seconds)
2013-11-04T07:10:18  *** asx151 <asx151!~asd@46.49.55.27> has quit IRC (Read error: Connection reset by peer)
2013-11-04T07:18:39  *** antgreen <antgreen!~green@dsl-173-206-41-123.tor.primus.ca> has quit IRC (Ping timeout: 260 seconds)
2013-11-04T07:26:39  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Ping timeout: 240 seconds)
2013-11-04T07:46:46  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-04T08:00:33  *** asx15 <asx15!~asd@46.49.55.27> has joined #rtems
2013-11-04T08:27:45  *** 7CBAAB72H <7CBAAB72H!~monstr@nat-63.starnet.cz> has quit IRC (Ping timeout: 246 seconds)
2013-11-04T08:47:35  *** antgreen <antgreen!~green@dsl-173-206-41-123.tor.primus.ca> has joined #rtems
2013-11-04T08:50:31  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-04T08:51:21  <gedare> good morning.
2013-11-04T08:51:28  <sebhub2> hi
2013-11-04T08:58:52  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Read error: Connection reset by peer)
2013-11-04T08:59:21  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-04T09:09:25  *** S_Somani <S_Somani!~Shubham@103.16.28.13> has joined #rtems
2013-11-04T09:21:34  <gedare> i dislike when i open my email to look something up, and get sucked into the vortex of neverending noise. ;)
2013-11-04T09:21:42  * gedare makes more noise
2013-11-04T09:21:56  <gedare> maybe i can quit email!
2013-11-04T09:23:07  <sebhub2> or filter them
2013-11-04T09:26:34  <gedare> yeah i already do that, a lot.
2013-11-04T09:27:05  <gedare> i just find it irresistable to go looking in those folders. guess i need better impulse control.
2013-11-04T09:27:07  <peerst> gedare: whats causing the noise, just random shit or mailinglists?
2013-11-04T09:27:41  <gedare> both really, and my Finnish pre-disposition to impulsivity. http://www.livescience.com/9197-impulsive-gene-identified-finnish-men.html
2013-11-04T09:27:51  <gedare> ;)
2013-11-04T09:28:41  <peerst> I subscribe my ML's from a junk gmail account and then read and post from a netnews reader via gmane
2013-11-04T09:29:25  <peerst> for the random commercial crap stuff I stopped extending my filers and start unsubscribing
2013-11-04T09:30:41  <gedare> i filter all ml's already. maybe i should try unsubscribing the commercial stuff, i don't get that much though.
2013-11-04T09:34:45  <gedare> how's Erlang+RTEMS doing? ready for public consumption? :)
2013-11-04T09:36:01  <gedare> i'm working through procedures to help users (like peerst) to transition their projects from personal use to RTEMS Project supported.
2013-11-04T09:36:42  <peerst> gedare: getting there slowly
2013-11-04T09:36:48  <gedare> that's the way of things.
2013-11-04T09:37:40  <gedare> Would it make sense to try getting a GSoC student to work on it next summer?
2013-11-04T09:37:51  <gedare> I admit I don't know all the ramifications of your project :)
2013-11-04T09:37:52  <peerst> gedare: have to move it to RTL to just have targets in Erlang for building each CPU arch and not arch+bsp+C-app
2013-11-04T09:38:00  <gedare> ahhh
2013-11-04T09:38:25  <gedare> that sounds like a concise student project to me!
2013-11-04T09:38:26  <peerst> also I need to disentangle it from some customer specific code
2013-11-04T09:38:32  <gedare> that is a deeper problem.
2013-11-04T09:38:57  <peerst> gedare: but the disentangle is going on at the same time as moving it to RTL
2013-11-04T09:39:04  <gedare> gotcha.
2013-11-04T09:39:38  <gedare> i have similar problems with some of my projects. except I'm the customer, so sometimes I just let my "proprietary" code leak a little.
2013-11-04T09:39:44  <gedare> (usually the customer)
2013-11-04T09:40:05  <gedare> but I definitely understand where you're coming from.
2013-11-04T09:40:35  <gedare> Anyway, think about whether there might be something worthy of an SOC project.
2013-11-04T09:40:52  <gedare> and, whether you want to commit to mentoring such an effort :)
2013-11-04T09:41:30  <peerst> can't leak because its also unclean
2013-11-04T09:41:42  * gedare nods
2013-11-04T09:41:53  <peerst> the Erlang/OTP project wouldn't accept my customer crap as patches
2013-11-04T09:43:34  <peerst> I plan to get it moving along until start of december where I'm speaking about it on the Codemesh conference in London
2013-11-04T09:43:59  <peerst> We'll see how well that will work out
2013-11-04T09:44:08  <gedare> you do know it is november? ;)
2013-11-04T09:44:19  <peerst> :-P yeah
2013-11-04T09:44:52  <peerst> well I already have my patches in a MQ (using mercurial here) that is rebased almost to master of Erlang/OTP
2013-11-04T09:45:31  <peerst> it mainly depends if building with RTL "just works" or doesn't
2013-11-04T09:46:03  <peerst> if it gets too messy I might invest into Erlang/OTP's build system to make it possible to build it as lib
2013-11-04T09:50:33  <gedare> ok. RTL got a lot of improvements this summer
2013-11-04T09:50:39  <gedare> there was a great student working on it.
2013-11-04T09:58:10  <peerst> great
2013-11-04T09:59:21  <peerst> gedare: I got a bit concerned when a few days ago mkhoory reported that its quite slow
2013-11-04T09:59:32  <peerst> not sure how that resolved?
2013-11-04T10:00:06  <gedare> I have not used RTL personally. you'd have to ask kiwichris he is the maintainer..
2013-11-04T10:00:17  <peerst> he said its unoptimized
2013-11-04T10:00:53  <peerst> currently I have aboz 2sec from poweron until Erlang shell
2013-11-04T10:01:34  <peerst> when I add 23 seconds to this it will get unacceptable for my customer at least (and many possible users from the Erlang world who were impressed with the startup time)
2013-11-04T10:01:45  <gedare> of course.
2013-11-04T10:02:18  <peerst> gedare: why I'm telling you this is it might be a GSoC project to optimize stuff
2013-11-04T10:02:24  <gedare> gotcha
2013-11-04T10:02:36  <gedare> RTL Performance Tuning
2013-11-04T10:02:52  <peerst> *if* it needs optimizing at all it might have been some artifact on the use case of mkhoory
2013-11-04T10:03:43  <peerst> afk for a bit
2013-11-04T10:03:44  <gedare> there are probably other opportunities.
2013-11-04T10:33:38  *** sebhub2 <sebhub2!~sh@host-82-135-62-35.customer.m-online.net> has quit IRC (Remote host closed the connection)
2013-11-04T11:21:49  *** asx15 <asx15!~asd@46.49.55.27> has quit IRC (Ping timeout: 272 seconds)
2013-11-04T11:43:35  <peerst> back
2013-11-04T11:54:46  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has joined #rtems
2013-11-04T13:26:07  <kiwichris> peerst: you concerns are noted and I will take a look when I have time
2013-11-04T13:26:47  <kiwichris> peerst: can you please post some details like the arch/bsp, the number of object files, symbols and relocation records
2013-11-04T13:27:13  <kiwichris> peerst: and I thank you for raising them
2013-11-04T13:28:22  <peerst> kiwichris: I was citing mkhoory's problem unfortunately there was no followup how it was resolved
2013-11-04T13:28:54  <peerst> when I switched Erlang building to RTL I'll have some numbers of my own
2013-11-04T13:29:27  <kiwichris> peerst: ah ok and thanks; he is here so maybe he will respond
2013-11-04T13:30:07  <kiwichris> peerst: maybe is cpu clock speed is slow, ie RAD hardened.
2013-11-04T13:30:22  <peerst> and then I'll use a Erlang RAP probably only to link it statically afterwards (loading some smaller dynamic libs later if thats then possible)
2013-11-04T13:30:49  <peerst> I hope to get some time working on this the next month
2013-11-04T13:34:37  <kiwichris> is Erlang built from clean sources for RTEMS, clean sources + patch or something special ?
2013-11-04T13:36:10  <peerst> Erlang built from clean sources with special settings
2013-11-04T13:36:30  <peerst> then linked with RTEMS and C-application
2013-11-04T13:36:34  <peerst> and BSP
2013-11-04T13:36:54  <kiwichris> Nice. Would you be interested in working with me to create an RSB config that builds it as a 3rd party library ?
2013-11-04T13:37:32  <kiwichris> This way a user can enter a single RSB command and have it built
2013-11-04T13:37:47  <kiwichris> need to go; back later
2013-11-04T13:40:34  *** S_Somani <S_Somani!~Shubham@103.16.28.13> has quit IRC (Quit: Leaving.)
2013-11-04T13:51:33  *** antgreen <antgreen!~green@dsl-173-206-41-123.tor.primus.ca> has quit IRC (Ping timeout: 248 seconds)
2013-11-04T14:16:00  *** antgreen <antgreen!~green@dsl-173-206-41-123.tor.primus.ca> has joined #rtems
2013-11-04T14:35:54  <peerst> kiwichris: sure that'd be great, I'd rather write scripts than a lot of howto description
2013-11-04T14:43:09  *** antgreen <antgreen!~green@dsl-173-206-41-123.tor.primus.ca> has quit IRC (Read error: Operation timed out)
2013-11-04T15:36:55  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has quit IRC (Ping timeout: 265 seconds)
2013-11-04T16:01:52  *** arvind_k <arvind_k!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-04T16:05:39  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Ping timeout: 272 seconds)
2013-11-04T16:10:33  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 245 seconds)
2013-11-04T16:27:43  *** antgreen <antgreen!~green@dsl-173-206-84-94.tor.primus.ca> has joined #rtems
2013-11-04T17:37:43  <kiwichris> peerst: excellent, lets do that. How are you building it currently ?
2013-11-04T17:39:05  <peerst> with their normal cross building config + make (I patch some customer/application dep hacks into the tree at the moment still)
2013-11-04T17:40:46  <peerst> the configs for all supported RTEMS cross building archs will be part of the standard erlang distribution evenually (for this I need to isolate the BSP dependency of the current method, thats why I want to move to RTL)
2013-11-04T17:43:00  <peerst> when all is cleaned up there is one set of autoconf, configure, make install steps with one config value differing for each cpu arch
2013-11-04T18:23:08  <mkhoory> peerst, regarding RTL performance, the 23 seconds was with the RAP format. Switching to ELF cut down the load time by 8-9 seconds
2013-11-04T18:23:33  <mkhoory> I mean now it takes 8-9 seconds to load
2013-11-04T18:24:31  <mkhoory> my cpu is a RAD hardened LEON3 (LEON3-FT), 20MHz
2013-11-04T18:26:34  <mkhoory> The thing is, the module loading times on another competing RTOS with this board is still shorter (2-3 seconds), so I think the load times can be improved on RTL
2013-11-04T18:38:15  *** antgreen <antgreen!~green@dsl-173-206-84-94.tor.primus.ca> has quit IRC (Ping timeout: 252 seconds)
2013-11-04T20:03:32  *** antgreen <antgreen!~green@216.254.165.197> has joined #rtems
2013-11-04T20:46:41  *** arvind_k <arvind_k!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Ping timeout: 268 seconds)
2013-11-04T20:59:00  *** arvind_k <arvind_k!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-04T22:47:56  *** arvind_k <arvind_k!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-04T22:52:55  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
2013-11-04T23:03:26  *** S_Somani <S_Somani!~Shubham@103.248.173.61> has joined #rtems
2013-11-04T23:19:30  *** kiwichris <kiwichris!~textual@123.208.63.112> has joined #rtems
2013-11-05T00:23:38  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-05T00:26:11  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has joined #rtems
2013-11-05T00:31:17  *** kiwichris <kiwichris!~textual@123.208.63.112> has quit IRC (Read error: Connection reset by peer)
2013-11-05T00:50:17  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-05T02:37:30  <peerst> kiwichris: did you copy what mkhoory wrote? so it is a slow cpu partially
2013-11-05T02:40:03  <peerst> mkhoory: you produced the ELF also with RTL and load-linked it too?
2013-11-05T03:21:18  *** S_Somani <S_Somani!~Shubham@103.248.173.61> has quit IRC (Read error: Connection reset by peer)
2013-11-05T03:21:31  *** S_Somani <S_Somani!~Shubham@103.248.173.61> has joined #rtems
2013-11-05T05:11:15  *** S_Somani <S_Somani!~Shubham@103.248.173.61> has left #rtems
2013-11-05T05:13:44  <kiwichris> peerst: RTL can load ELF or RAP format files.
2013-11-05T05:40:54  <peerst> ah ok
2013-11-05T05:41:23  <peerst> so RAP was probably slower since it had to decompress also
2013-11-05T07:02:20  *** sebhub <sebhub!~sh@host-82-135-62-35.customer.m-online.net> has joined #rtems
2013-11-05T07:44:36  *** antgreen <antgreen!~green@216.254.165.197> has quit IRC (Ping timeout: 246 seconds)
2013-11-05T07:54:14  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-05T08:49:18  *** antgreen <antgreen!~green@216.254.165.197> has joined #rtems
2013-11-05T10:12:27  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-05T10:51:26  *** sebhub <sebhub!~sh@host-82-135-62-35.customer.m-online.net> has quit IRC (Remote host closed the connection)
2013-11-05T11:19:34  *** gedare_ <gedare_!~gedare@128.164.201.103> has joined #rtems
2013-11-05T13:02:56  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has quit IRC (Ping timeout: 245 seconds)
2013-11-05T14:30:01  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 245 seconds)
2013-11-05T16:43:16  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has joined #rtems
2013-11-05T16:59:52  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-05T17:09:28  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 272 seconds)
2013-11-05T17:24:01  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-05T17:24:33  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 246 seconds)
2013-11-05T18:25:26  <mkhoory> peerst, decompression might be it. I linked the ELF using regular ld, not with rtems-ld.
2013-11-05T18:26:17  <mkhoory> but since my board might be slower than your board, you should give it a try either way
2013-11-05T18:57:41  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-05T18:57:51  <gedare> hi kiwichri_ you around?
2013-11-05T19:00:40  <kiwichri_> gedare, yeah
2013-11-05T19:00:55  <gedare> cool, things going ok?
2013-11-05T19:01:01  <kiwichri_> I am working on doco for the RSB to show how to write config files :)
2013-11-05T19:01:06  <gedare> haha great!
2013-11-05T19:01:10  <gedare> i look forward to reading it.
2013-11-05T19:01:17  <kiwichri_> I hope soon.
2013-11-05T19:01:38  <kiwichri_> I am documenting the bare/config/dtc.bset
2013-11-05T19:30:43  <gedare> gedare_: go to sleep work computer!
2013-11-05T19:35:36  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 260 seconds)
2013-11-06T00:25:19  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has joined #rtems
2013-11-06T01:23:09  *** sebhub <sebhub!~sh@host-82-135-62-35.customer.m-online.net> has joined #rtems
2013-11-06T01:23:38  <sebhub> good morning
2013-11-06T01:54:21  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-06T03:36:53  *** rtemsLogger <rtemsLogger!~supybot@rtems.org> has joined #rtems
2013-11-06T03:36:56  *** rtemsLogger` <rtemsLogger`!~supybot@rtems.org> has quit IRC (Remote host closed the connection)
2013-11-06T05:26:01  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-06T06:35:47  *** microchip_sac <microchip_sac!~microchip@122.170.25.191> has joined #rtems
2013-11-06T06:37:24  <microchip_sac> what sort of tasks is RTEMS contributing for Google Code-In 2013?
2013-11-06T06:39:36  *** antgreen <antgreen!~green@216.254.165.197> has quit IRC (Ping timeout: 245 seconds)
2013-11-06T06:55:07  *** antgreen <antgreen!~green@216.254.165.197> has joined #rtems
2013-11-06T07:00:50  *** antgreen <antgreen!~green@216.254.165.197> has quit IRC (Ping timeout: 245 seconds)
2013-11-06T07:10:16  *** microchip_sac <microchip_sac!~microchip@122.170.25.191> has left #rtems ("Leaving channel!")
2013-11-06T07:14:19  *** antgreen <antgreen!~green@216.254.165.197> has joined #rtems
2013-11-06T07:48:48  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-06T07:51:15  *** guerby <guerby!~guerby@ip165-ipv6.tetaneutral.net> has quit IRC (Ping timeout: 252 seconds)
2013-11-06T07:56:10  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Read error: Operation timed out)
2013-11-06T09:14:53  *** guerby <guerby!~guerby@ip165-ipv6.tetaneutral.net> has joined #rtems
2013-11-06T09:19:48  *** guerby <guerby!~guerby@ip165-ipv6.tetaneutral.net> has quit IRC (Ping timeout: 252 seconds)
2013-11-06T09:21:42  *** guerby <guerby!~guerby@ip165-ipv6.tetaneutral.net> has joined #rtems
2013-11-06T09:33:47  *** sebhub <sebhub!~sh@host-82-135-62-35.customer.m-online.net> has quit IRC (Ping timeout: 260 seconds)
2013-11-06T09:56:07  *** gedare_ <gedare_!~gedare@128.164.201.103> has quit IRC (Quit: Leaving)
2013-11-06T09:56:20  *** gedare_ <gedare_!~gedare@128.164.201.103> has joined #rtems
2013-11-06T09:57:02  *** gedare_ <gedare_!~gedare@128.164.201.103> has quit IRC (Client Quit)
2013-11-06T09:58:31  *** gedare <gedare!~gedare@128.164.201.103> has joined #rtems
2013-11-06T10:05:24  <gedare> howdy
2013-11-06T11:47:54  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has quit IRC (Remote host closed the connection)
2013-11-06T13:27:46  <kiwichri_> gedare, hi
2013-11-06T14:06:34  <gedare> kiwichri_, hi
2013-11-06T14:57:18  *** gedare <gedare!~gedare@128.164.201.103> has quit IRC (Ping timeout: 240 seconds)
2013-11-06T15:41:16  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-06T16:03:49  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 248 seconds)
2013-11-06T17:22:14  <peerst> mkhoory: do you know why these LEON's run so slow? Is that the result of space hardening or because of power constraints?
2013-11-06T18:09:25  <mkhoory> peerst, I'm still new, but I think it's because of the space hardening. EDAC memory is slow due to its complexity for example. It's also expensive to make
2013-11-06T18:09:40  <mkhoory> at least this is how I understood it. I might be wrong
2013-11-06T18:10:25  <mkhoory> expensive to make the chips I mean, not just the RAM. Power constraints is probably another big thing
2013-11-06T19:23:18  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-06T19:30:55  <kiwichri_> peerst, the MMU NASA project uses a RAD Coldfire v2 running at 32MHz. Heat is a problem in the RAD casing
2013-11-06T19:31:22  <kiwichri_> Sorry NASA MMS project ... http://mms.gsfc.nasa.gov/
2013-11-06T19:40:26  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 245 seconds)
2013-11-06T19:52:55  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 245 seconds)
2013-11-06T19:53:13  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-06T20:54:02  *** antgreen <antgreen!~green@216.254.165.197> has quit IRC (Read error: Operation timed out)
2013-11-06T21:49:27  *** antgreen <antgreen!~green@216.254.165.197> has joined #rtems
2013-11-06T22:02:37  *** antgreen <antgreen!~green@216.254.165.197> has quit IRC (Read error: Operation timed out)
2013-11-06T22:43:03  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 272 seconds)
2013-11-06T22:44:06  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-06T22:56:08  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
2013-11-06T23:10:08  *** kiwichris <kiwichris!~textual@110.150.180.113> has joined #rtems
2013-11-06T23:49:43  *** kiwichris <kiwichris!~textual@110.150.180.113> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
2013-11-07T00:09:12  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-07T00:27:28  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-07T01:20:39  *** monstr <monstr!~monstr@nat-63.starnet.cz> has joined #rtems
2013-11-07T03:00:32  <peerst> kiwichris: ah ok I see, do you know if there is thought/possibility to use COTS hardware and using failure tolerance techniques and multiple "networked" mothrboards to get failure safety?
2013-11-07T03:25:23  <kiwichris> Then you get power and weight issues.
2013-11-07T04:16:23  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has joined #rtems
2013-11-07T04:38:55  <peerst> hmm
2013-11-07T04:40:21  <peerst> is there any literature about the design issues of hardware in space?
2013-11-07T06:17:03  *** ysionnea1 <ysionnea1!~fallen@ks3002254.kimsufi.com> has joined #rtems
2013-11-07T06:21:45  *** ysionneau <ysionneau!~fallen@ks3002254.kimsufi.com> has quit IRC (*.net *.split)
2013-11-07T06:54:30  *** guerby <guerby!~guerby@ip165-ipv6.tetaneutral.net> has quit IRC (Ping timeout: 253 seconds)
2013-11-07T07:09:58  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-07T07:18:48  *** guerby <guerby!~guerby@ip165.tetaneutral.net> has joined #rtems
2013-11-07T07:34:47  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has quit IRC (Ping timeout: 240 seconds)
2013-11-07T07:35:46  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has joined #rtems
2013-11-07T08:31:39  *** monstr <monstr!~monstr@nat-63.starnet.cz> has quit IRC (Remote host closed the connection)
2013-11-07T08:32:37  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has joined #rtems
2013-11-07T08:45:01  *** monstr_ <monstr_!~monstr@nat-63.starnet.cz> has quit IRC (Remote host closed the connection)
2013-11-07T09:21:26  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has quit IRC (Ping timeout: 240 seconds)
2013-11-07T09:35:27  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-07T10:01:50  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has quit IRC (Ping timeout: 240 seconds)
2013-11-07T10:02:46  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has joined #rtems
2013-11-07T12:23:21  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-07T13:37:01  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-07T13:47:46  *** antgreen <antgreen!~green@out-on-138.wireless.telus.com> has joined #rtems
2013-11-07T13:58:59  *** antgreen <antgreen!~green@out-on-138.wireless.telus.com> has quit IRC (Ping timeout: 260 seconds)
2013-11-07T14:49:09  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has joined #rtems
2013-11-07T15:38:58  *** gedare_ <gedare_!~gedare@128.164.201.103> has joined #rtems
2013-11-07T15:51:35  *** guerby <guerby!~guerby@ip165.tetaneutral.net> has quit IRC (Ping timeout: 252 seconds)
2013-11-07T15:56:38  *** guerby <guerby!~guerby@ip165.tetaneutral.net> has joined #rtems
2013-11-07T17:44:36  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 252 seconds)
2013-11-08T00:27:38  *** qingpei <qingpei!~qingpei@58.212.38.38> has joined #rtems
2013-11-08T00:40:59  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-08T01:44:55  *** qingpei <qingpei!~qingpei@58.212.38.38> has quit IRC (Quit: qingpei)
2013-11-08T01:56:50  *** qingpei <qingpei!~qingpei@58.212.38.38> has joined #rtems
2013-11-08T02:51:54  *** mkhoory <mkhoory!~mkhoory@211.117.165.123> has quit IRC (Read error: Connection reset by peer)
2013-11-08T03:28:22  *** antgreen_ <antgreen_!~green@dsl-207-112-125-176.tor.primus.ca> has joined #rtems
2013-11-08T03:29:43  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has quit IRC (Ping timeout: 252 seconds)
2013-11-08T06:23:41  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has quit IRC (Ping timeout: 272 seconds)
2013-11-08T06:24:07  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has joined #rtems
2013-11-08T06:38:08  *** antgreen_ <antgreen_!~green@dsl-207-112-125-176.tor.primus.ca> has quit IRC (Ping timeout: 265 seconds)
2013-11-08T06:58:53  *** qingpei <qingpei!~qingpei@58.212.38.38> has quit IRC (Quit: qingpei)
2013-11-08T07:27:53  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-08T07:32:12  <gedare> hello
2013-11-08T07:37:34  <gedare> peerst: planning to put erlang in space? ;)
2013-11-08T07:38:52  <peerst> gedare: well they talked me into this talk title: http://codemesh.io/#peer-stritzinger
2013-11-08T07:39:26  <peerst> which somehow mutates the talk I wanted to give
2013-11-08T07:39:54  <peerst> now I'm actually thinking what Erlang would bring to the table in space applications
2013-11-08T07:40:08  <peerst> how it could disrupt the old way of doing things
2013-11-08T07:40:14  <gedare> haha,so yes.:)
2013-11-08T07:40:30  <gedare> hmm.
2013-11-08T07:41:12  <peerst> the whole fault tolerance and hot code upgrade thing looks quite usefull for space
2013-11-08T07:41:29  <gedare> yes, they already do that in application logicxs
2013-11-08T07:41:38  <gedare> FT is usually done at hw level
2013-11-08T07:41:48  <peerst> also Erlang hosts a bunch of the most powerful testing methods (Erlangs Version of Quickcheck)
2013-11-08T07:42:08  <gedare> either with circuit-level redundancy (TMR) or with replicated single-board computers and output validation
2013-11-08T07:42:22  <peerst> which btw can be used to test C code
2013-11-08T07:42:43  <gedare> fault detection would be useful.
2013-11-08T07:42:47  <peerst> the area of replicated single board computers is right in erlangs ballpark
2013-11-08T07:42:50  <gedare> is useful, really.
2013-11-08T07:43:10  <gedare> well, those are from my understanding the two main methods used in space hw designs.
2013-11-08T07:43:13  <peerst> Erlang has builtin distribution because you can't do fault tolerance without it
2013-11-08T07:43:14  <gedare> i'm no expert..
2013-11-08T07:43:18  <gedare> but i've sat in rooms with some
2013-11-08T07:43:39  <gedare> sure. good. does erlang adopt a component-based model internally?
2013-11-08T07:43:56  <gedare> or even in application design?
2013-11-08T07:44:09  <peerst> Erlang is very good at fault detection (through pattern matching) -- think you have implicit multiple C assert() statements per code line
2013-11-08T07:44:20  <peerst> and even better at fault isolation
2013-11-08T07:44:52  <gedare> i think fault detection is one of the trickier things.
2013-11-08T07:44:56  <peerst> well whatever component based means yeah
2013-11-08T07:45:04  <gedare> a lot of modern research still assumes fail-crash/fail-safe
2013-11-08T07:45:21  <peerst> what does that mean?
2013-11-08T07:45:34  <gedare> it means a fault causes the system to abort in an obvious way
2013-11-08T07:45:53  <peerst> yeah thats one of the Erlang paradigms
2013-11-08T07:45:58  <peerst> "let it crash"
2013-11-08T07:46:09  <gedare> it is a way for researchers to circumvent solving the fault detection problem
2013-11-08T07:46:13  <gedare> and to focus on fault recovery methods
2013-11-08T07:46:55  <peerst> for fault recovery Erlang has lowlevel things (like linking and monitoring of processes, having exit signals out of band)
2013-11-08T07:47:05  <peerst> and building upon that supervision trees
2013-11-08T07:47:13  <gedare> what is a supervision tree
2013-11-08T07:47:34  <peerst> a supervisor is a very simple and well tested process (everything in Erlang is a process)
2013-11-08T07:47:43  <peerst> which gets a config table
2013-11-08T07:47:58  <peerst> either statically at startup or dynamically
2013-11-08T07:48:22  <gedare> ah
2013-11-08T07:48:56  <gedare> :q
2013-11-08T07:49:00  <gedare> wrong window...
2013-11-08T07:49:13  <gedare> did i expose my editor preference??
2013-11-08T07:49:17  <peerst> which contains child specifications, that tell it how to start, type of child temporary/permanent/transient
2013-11-08T07:49:29  <peerst> hehe yeah
2013-11-08T07:49:42  <gedare> ok interesting.
2013-11-08T07:49:53  <peerst> and you can tell it how to shutdown, and some general strategy things
2013-11-08T07:49:59  <gedare> does erlang map its processes onto OS processes?
2013-11-08T07:50:30  <peerst> like one_for_one (restart each child in isolation) and rest_for_one (terminate all childs behind the failed and restart all in order)
2013-11-08T07:50:45  <peerst> or one_for_all
2013-11-08T07:50:58  <peerst> no its like green threads
2013-11-08T07:51:09  <peerst> Erlang has very lightweight processes internally
2013-11-08T07:51:28  <peerst> which behave like real processes (shared nothing = good for fault isolation)
2013-11-08T07:51:32  <gedare> ok
2013-11-08T07:51:39  <peerst> a process costs about 500bytes
2013-11-08T07:51:51  <gedare> but a global address space?
2013-11-08T07:51:55  <peerst> it has builtin schedulers
2013-11-08T07:51:58  <peerst> no
2013-11-08T07:52:02  <peerst> nothing global
2013-11-08T07:52:11  <gedare> language-based isolation?
2013-11-08T07:52:19  <peerst> yep
2013-11-08T07:52:24  <gedare> type safety?
2013-11-08T07:52:24  <peerst> only message passing
2013-11-08T07:52:31  <peerst> dynamic typed
2013-11-08T07:52:39  <gedare> maybe i should just read about it :)
2013-11-08T07:52:43  <peerst> strict but dynamic
2013-11-08T07:52:57  <peerst> it has a special type system built for it
2013-11-08T07:53:12  <peerst> you can optionaly specify types and have them checked by a tool
2013-11-08T07:53:33  <gedare> ok. i'm assuming the type system prevents forging references to symbols in another process..?
2013-11-08T07:53:36  <peerst> gedare: http://learnyousomeerlang.com
2013-11-08T07:54:02  <peerst> thats the best intro and in addition the best advanced text (because it gose all the way)
2013-11-08T07:54:13  <gedare> i tried to learn about haskell last week and gave up after 20 minutes, so hard to say. :)
2013-11-08T07:54:28  <peerst> no the type system has nothing to do with it
2013-11-08T07:54:39  <gedare> how is isolation between processes enforced/
2013-11-08T07:54:43  <peerst> its a strict functional language with message passing added
2013-11-08T07:54:52  <gedare> Ahhhhhh
2013-11-08T07:55:02  <peerst> (and a few workarounds to its strictness we don't like to talk about in public)
2013-11-08T07:55:08  <gedare> lol.
2013-11-08T07:55:29  <peerst> there is a process global process dict
2013-11-08T07:55:51  <gedare> are there scalability problems with the message passing?
2013-11-08T07:56:03  <peerst> and there is a highly efficient table structure that behaves a bit like a process
2013-11-08T07:56:23  <peerst> well scaling is what erlang is used for nowadays
2013-11-08T07:56:25  <peerst> so no
2013-11-08T07:56:31  <gedare> heh, ok
2013-11-08T07:56:39  <peerst> not if you don't misdesign your system
2013-11-08T07:56:41  <gedare> any particular bottlenecks in the runtime?
2013-11-08T07:56:48  <peerst> whatsapp uses erlang
2013-11-08T07:57:04  <peerst> yeah the bottleneck is sequential performance
2013-11-08T07:57:28  <gedare> seq. perf. of what?
2013-11-08T07:57:56  <peerst> interestingly because of its soft realtime reactiveness many real world applications perform much better than other (sequentially faster languages)
2013-11-08T07:58:03  <peerst> seq perf of everything
2013-11-08T07:58:47  <peerst> if you write a number crunching loop you'll find that erlang was not built for this
2013-11-08T07:58:53  <gedare> ok.
2013-11-08T07:59:01  <gedare> i understand
2013-11-08T07:59:20  <peerst> bu it has good interface to OpenCL I worked with already
2013-11-08T07:59:36  <gedare> it does really well for message-passing thread/data parallel workloads, and degrades as parallelism decreases?
2013-11-08T08:00:01  <peerst> Ercison built a ATM swicth with erlang that got nine nines availability
2013-11-08T08:00:09  <gedare> interesting.
2013-11-08T08:00:20  <peerst> thats milliseconds per year
2013-11-08T08:00:29  <peerst> and it included maintainence
2013-11-08T08:00:43  <peerst> five nines is normal for a telecom equipment
2013-11-08T08:01:09  <peerst> its not only parallelism
2013-11-08T08:01:39  <peerst> since you have a process waiting for everything that needs to be done and fast switch times its very reactive
2013-11-08T08:01:48  <gedare> ah. ok
2013-11-08T08:02:01  <peerst> and resilent as hell when overloaded
2013-11-08T08:02:07  <gedare> interesting and worth coming back to.
2013-11-08T08:02:15  <peerst> its actually hard to compare apache to erlang webservers
2013-11-08T08:02:23  <gedare> know if it is used on any robotics platforms yet?
2013-11-08T08:02:31  <peerst> because apache disappears at so weak load
2013-11-08T08:02:43  <peerst> there are some projects
2013-11-08T08:03:02  <peerst> and of course my customer project I did the port for
2013-11-08T08:03:18  <peerst> which is transport system for industrial
2013-11-08T08:04:02  <gedare> one aspect of my research is heading toward robotics systems, with an eye toward fault tolerance and performance in distributed settings.
2013-11-08T08:04:14  <peerst> but especially if combined with RTEMS for the hard realtime things of motion control it would be very useful for robots
2013-11-08T08:04:27  <peerst> haha
2013-11-08T08:04:31  <peerst> perfect fit
2013-11-08T08:04:39  <peerst> you need it
2013-11-08T08:04:55  <gedare> one of my main thrusts though is improving the performance (throughput/latency) of data structures, hence my question about runtime bottlenecks :)
2013-11-08T08:05:03  <gedare> yeah it sounds like it.
2013-11-08T08:05:04  <peerst> transparent distribution!
2013-11-08T08:05:31  <peerst> the throughput for communication is good
2013-11-08T08:05:41  <peerst> it has a binary blob datatype
2013-11-08T08:05:48  <gedare> most of the prior art seems to use either domain specific languages or complex procedural/OO languages.
2013-11-08T08:06:04  <peerst> with binary pattern matching (whch is soooo great for protocol stuff)
2013-11-08T08:06:13  <gedare> hmm.
2013-11-08T08:06:30  <peerst> and these binaries are only stored on the process heap if =< 64 bytes
2013-11-08T08:06:46  <peerst> large binaries are on their own global heap
2013-11-08T08:07:03  <peerst> so when you pass them (or parts of them) around you only pass on a pointer
2013-11-08T08:07:22  <gedare> sure, otherwise they get marshalled/copied?
2013-11-08T08:07:33  <peerst> message passing = copying
2013-11-08T08:07:37  <gedare> yep.
2013-11-08T08:07:43  <peerst> which sounds slow at first
2013-11-08T08:07:55  <peerst> but isn't when you have multicore
2013-11-08T08:08:10  <peerst> because it keeps everything very lock sparse
2013-11-08T08:08:21  <gedare> software designers tend to figure out how to reduce the size/number of parameters they send too.
2013-11-08T08:08:38  <peerst> and in current NUMA architectures it gets message passed around anayway
2013-11-08T08:09:07  <peerst> well and you can just send a process Id
2013-11-08T08:09:17  <peerst> and talk to the process for the data later
2013-11-08T08:10:02  <peerst> the VM starts one OS thread per CPU core normally
2013-11-08T08:10:19  <peerst> and maps the processes on these schedulers
2013-11-08T08:10:39  <peerst> one OS thread = one scheduler
2013-11-08T08:10:48  <peerst> = one virtual core
2013-11-08T08:10:50  <gedare> got it
2013-11-08T08:11:12  <peerst> so the VM does share memory but your processes don't
2013-11-08T08:11:14  <gedare> yeah the language/runtime is definitely of interest to me.
2013-11-08T08:11:41  <peerst> it is a leap though because Erlang systems are built differently
2013-11-08T08:12:04  <peerst> the language is easy to learn if you don't mind its unusual syntax
2013-11-08T08:12:23  <peerst> but the implications and how to design systems is harder
2013-11-08T08:15:05  <gedare> my intuition is that Erlang systems look a lot like other systems based on a message-passing runtime
2013-11-08T08:15:38  <peerst> not sure (don't know other message based systems … except Occam)
2013-11-08T08:16:18  <gedare> well, there's MPI from the parallel supercomputing world
2013-11-08T08:16:25  <peerst> its less centered around message based but more on the actor based
2013-11-08T08:16:33  <gedare> yeah
2013-11-08T08:16:39  <gedare> you break your system up into logical processes, and model your application with something along the lines of communicating sequential processes.
2013-11-08T08:16:46  <peerst> creating and destroying processes is cheap and fast
2013-11-08T08:17:09  <gedare> what you end up with is a graph of software components that communicate with each other
2013-11-08T08:17:23  <peerst> you can model your system also with processes that are created dynamically on demand
2013-11-08T08:18:03  <gedare> right
2013-11-08T08:18:11  <peerst> thats the actor thing
2013-11-08T08:18:20  <gedare> most microkernels exhibit this message passing approach
2013-11-08T08:18:41  <peerst> true
2013-11-08T08:18:56  <peerst> Erlang looks a bit like a OS
2013-11-08T08:19:28  <peerst> Erlang on RTEMS looks like a OS on a nanokernel
2013-11-08T08:20:11  <gedare> makes sense to me.
2013-11-08T08:20:32  <peerst> need to get this usable for everyone
2013-11-08T08:20:44  <peerst> cleaned up my patches today
2013-11-08T08:20:53  <peerst> now trying to build it with RTL
2013-11-08T08:21:49  <gedare> cool
2013-11-08T08:21:54  <peerst> maybe play a bit with RTEMS SMP and Erlang later this konth
2013-11-08T08:22:05  <gedare> i'll give it a whirl when it comes out, and i have time. :)
2013-11-08T08:22:15  <peerst> gut the same P1020 board SMP is developed on for RTEMS
2013-11-08T08:23:07  <peerst> get the book I linked (or read it online if you prefer) its light and fun reading
2013-11-08T08:42:37  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-08T09:18:13  *** qingpei <qingpei!~qingpei@121.237.191.6> has joined #rtems
2013-11-08T10:34:23  *** qingpei <qingpei!~qingpei@121.237.191.6> has quit IRC (Quit: qingpei)
2013-11-08T10:42:21  *** qingpei <qingpei!~qingpei@121.237.191.6> has joined #rtems
2013-11-08T10:52:53  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has quit IRC (Ping timeout: 248 seconds)
2013-11-08T10:54:41  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has joined #rtems
2013-11-08T10:59:17  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has quit IRC (Ping timeout: 248 seconds)
2013-11-08T11:00:41  *** MegaAlex|away <MegaAlex|away!~MegaAlex@85.130.71.44> has joined #rtems
2013-11-08T11:13:25  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-08T11:52:45  <peerst> kiwichris: how would I build the base image RTL links against?
2013-11-08T11:56:11  *** qingpei <qingpei!~qingpei@121.237.191.6> has quit IRC (Quit: qingpei)
2013-11-08T11:59:12  *** qingpei <qingpei!~qingpei@121.237.191.6> has joined #rtems
2013-11-08T12:21:32  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-08T12:26:16  <peerst> kiwichris: Do I always have to specify --base with RTL
2013-11-08T12:27:46  <peerst> I would like to build Erlang just for a architecture e.g. powerpc and get something usable by every RTEMS bsp + app that runs on powerpc
2013-11-08T12:27:53  <peerst> Is that possible at all?
2013-11-08T12:28:56  <peerst> Currently there is a cross compilation target config file that tells Erlang how to be built
2013-11-08T12:29:13  <peerst> (it can't be built as lib currently)
2013-11-08T12:29:30  <peerst> so I wanted to replace the actual link command with rtems-ld
2013-11-08T12:29:47  <peerst> and make sure it doesn't build a main function
2013-11-08T12:30:28  <peerst> but if I need to specify a base image that would be a RTEMS+bsp+otherstuff image
2013-11-08T12:31:07  <peerst> so the Erlang RAP again gets the bsp dependency or not?
2013-11-08T12:32:32  <peerst> I hoped RTL would take the burden off me to modify Erlangs build process so it can be built as a lib
2013-11-08T12:33:19  <peerst> BTW I don't need to load the Erlang VM during runtime, just looking for a way to build something liblike
2013-11-08T12:34:10  <peerst> I need the Erlang VM then to be able to do dlopen on some libs (drivers and C-modules mainly)
2013-11-08T12:34:19  <peerst> but thats the second stage
2013-11-08T12:34:58  <peerst> First I need a way to just build a Erlang binary "package" for e.g. powerpc and RTEMS
2013-11-08T12:58:04  <gedare> makes sense to me. you want Erlang as a RAP module, without building the application
2013-11-08T12:59:16  <gedare> do you really need Erlang to be a module though? or is it the erlang application that needs to be "modularized"?
2013-11-08T13:01:27  <gedare> at any rate, your question relates to something i have been curious about, which would be generating dynamic modules for parts of the RTEMS OS.
2013-11-08T13:02:20  *** qingpei <qingpei!~qingpei@121.237.191.6> has quit IRC (Quit: qingpei)
2013-11-08T13:04:34  <kiwichris> peerst: --base is needed so we know what symbols are in the kernel; it allows a partial link against the current symbol base to collect the symbols missing. If you have access to all the libraries from the target you can manage this at runtime with an extra runtime overhead.
2013-11-08T14:36:30  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has joined #rtems
2013-11-08T15:50:44  <peerst> kiwichris: yeah I feared this
2013-11-08T15:51:21  <peerst> hmm
2013-11-08T15:52:17  <peerst> and the kernel being rtems + bsp and thats it
2013-11-08T15:52:51  <peerst> or do you have a way to build the "kernel" of just a base rtems without the bsp and the user app
2013-11-08T15:53:16  <gedare> pieces build as libraries right now
2013-11-08T15:54:59  <gedare> e.g. librtemscpu.a
2013-11-08T16:05:36  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 245 seconds)
2013-11-08T16:29:42  *** ysionnea1 is now known as ysionneau
2013-11-08T16:39:51  <peerst> I still don't understand what a kernel actually consists of … a bunch of libraries?
2013-11-08T16:40:13  <peerst> a fully built and linked RTEMS image
2013-11-08T16:40:20  <peerst> or what?
2013-11-08T17:05:16  <ysionneau> FYI, I think gcc 4.5.4 does not build on modern systems
2013-11-08T17:05:19  <ysionneau> like ubuntu 13.10
2013-11-08T17:06:08  <ysionneau> because of texi issues and the fact that modern texinfo (version >= 5) will report them as errors and not anymore as warnings
2013-11-08T17:06:31  <ysionneau> With this patch I got it to compile on my Ubuntu 13.10 and on another one: https://gist.github.com/fallen/7378887
2013-11-08T17:07:06  <ysionneau> I'm using gcc-core-4.5.4 from http://www.rtems.org/ftp/pub/rtems/SOURCES/4.11/ with rtems patch
2013-11-08T17:07:41  <ysionneau> you might want to consider adding my patch to RTEMS patch if you experience the same issue
2013-11-08T17:08:23  <ysionneau> gn8!
2013-11-08T17:10:25  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has quit IRC (Ping timeout: 272 seconds)
2013-11-08T17:12:48  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has joined #rtems
2013-11-08T19:48:45  *** qingpei <qingpei!~qingpei@121.237.191.6> has joined #rtems
2013-11-08T20:02:45  *** qingpei <qingpei!~qingpei@121.237.191.6> has quit IRC (Quit: qingpei)
2013-11-08T23:42:24  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-08T23:42:24  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Read error: Connection reset by peer)
2013-11-09T00:43:20  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 245 seconds)
2013-11-09T00:47:23  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-09T00:53:56  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 245 seconds)
2013-11-09T01:50:43  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-09T02:12:07  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-09T03:32:07  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-09T03:51:46  <peerst> kiwichris: how would you want to build Erlang for RTEMS with source builder?  I think maybe thats the best way to design how to build Erlang
2013-11-09T03:53:22  <peerst> For source builder there will not be a kernel or does source builder build something that can be used by RTL --base ?
2013-11-09T03:55:00  <peerst> For the Erlang runtime I have three choices: 1. build it as RAP with RTL or 2. build it like I do now linking it with RTEMS to a runnable image or 3. build it as lib
2013-11-09T03:55:51  <peerst> additional problem: there are certain modules of Erlang that need a dynamically loaded piece of C code
2013-11-09T03:56:33  <peerst> so when the whole tree is built I need to use RTL at least for these, so I probably need a kernel
2013-11-09T03:56:48  <peerst> before that
2013-11-09T03:57:38  <peerst> That doesn't select between 1, 2 or 3 if I understand the RTL approach correctly (which I might as well not)
2013-11-09T03:59:34  <peerst> So how would we build Erlang in source builder?  Probably easiest to integrate would be as lib (this requires work on Erlangs build sytstem) but then I don't have any kernel to specify for the C-modules that need to be loaded with dlopen
2013-11-09T03:59:49  <peerst> since its all only libs
2013-11-09T04:00:08  <peerst> except if I could specify a bunch of libs as "kernel"
2013-11-09T06:55:38  *** qingpei <qingpei!~qingpei@111.193.224.124> has joined #rtems
2013-11-09T08:24:26  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-09T08:52:13  *** qingpei <qingpei!~qingpei@111.193.224.124> has quit IRC (Quit: qingpei)
2013-11-09T09:09:57  *** qingpei <qingpei!~qingpei@111.193.224.124> has joined #rtems
2013-11-09T09:36:31  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 272 seconds)
2013-11-09T10:05:05  *** qingpei <qingpei!~qingpei@111.193.224.124> has quit IRC (Quit: qingpei)
2013-11-09T14:24:23  <kiwichris> peerst: I have almost finished doco for the RSB on adding new configurations.
2013-11-09T14:25:08  <kiwichris> The difference with this lib is building it for RTEMS; this brings a new range is problems to solve
2013-11-09T14:26:05  <peerst> Ill probably just pass the BSP name to Erlangs cross configuration
2013-11-09T14:26:28  <peerst> and make it build RTEMS + Erlang as one image
2013-11-09T14:26:43  <peerst> like I do now with hardcode BSP name
2013-11-09T14:27:46  <peerst> Then I can try to setup the building of the dynamic loaded C-modules with this generated image as --base
2013-11-09T14:28:10  <kiwichris> If you look here http://www.rtems.org/ftp/pub/rtems/people/chrisj/rtems-waf/ I use what will be come a standard or arch/bsp
2013-11-09T14:30:15  <peerst> yeah from this I can either autogenerate a cross config or just pass it to a generic rtems config
2013-11-09T14:30:59  <kiwichris> The python code I have here finds all the installed BSPs and a user can then supply a filter which can be 1:1 for a single BSP.
2013-11-09T14:32:33  <peerst> If someone wants to add their own objects libs they either need to attach them as additional object/lib names or just use the dynamic loading controlled from Erlang
2013-11-09T20:00:16  <kiwichris> peerst: I have updated the RSB doco with a simple example of a new configuration file.
2013-11-09T20:37:26  *** kiwichris <kiwichris!~textual@CPE-58-172-130-193.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 264 seconds)
2013-11-09T21:12:33  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-09T22:20:25  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 245 seconds)
2013-11-09T23:08:22  *** kiwichris <kiwichris!~textual@1.149.23.124> has joined #rtems
2013-11-10T00:13:36  *** kiwichris <kiwichris!~textual@1.149.23.124> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
2013-11-10T00:48:32  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-10T01:18:02  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has joined #rtems
2013-11-10T01:19:59  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-10T01:22:13  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 248 seconds)
2013-11-10T01:30:32  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has quit IRC (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
2013-11-10T01:36:02  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-10T02:40:50  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-10T02:41:02  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 264 seconds)
2013-11-10T05:29:22  *** qingpei <qingpei!~qingpei@123.116.120.57> has joined #rtems
2013-11-10T05:35:44  *** qingpei <qingpei!~qingpei@123.116.120.57> has quit IRC (Quit: qingpei)
2013-11-10T05:55:38  *** skm <skm!~suleman@39.41.164.189> has joined #rtems
2013-11-10T05:56:26  *** skm <skm!~suleman@39.41.164.189> has left #rtems
2013-11-10T06:07:15  *** qingpei <qingpei!~qingpei@221.219.112.117> has joined #rtems
2013-11-10T07:03:02  *** qingpei <qingpei!~qingpei@221.219.112.117> has quit IRC (Quit: qingpei)
2013-11-10T07:09:22  *** qingpei <qingpei!~qingpei@221.219.112.117> has joined #rtems
2013-11-10T07:28:43  <peerst> kiwichri_: cool
2013-11-10T08:35:14  *** qingpei <qingpei!~qingpei@221.219.112.117> has quit IRC (Quit: qingpei)
2013-11-10T08:37:42  *** qingpei <qingpei!~qingpei@221.219.112.117> has joined #rtems
2013-11-10T09:07:45  *** qingpei <qingpei!~qingpei@221.219.112.117> has quit IRC (Quit: qingpei)
2013-11-10T10:45:01  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-10T10:46:11  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has quit IRC (Read error: Connection reset by peer)
2013-11-10T11:45:44  *** qingpei <qingpei!~qingpei@123.116.42.108> has joined #rtems
2013-11-10T11:57:03  *** arvind_khadri <arvind_khadri!~arvind@unaffiliated/arvind-khadri/x-2237230> has quit IRC (Remote host closed the connection)
2013-11-10T12:03:58  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has joined #rtems
2013-11-10T12:09:37  *** gedare <gedare!~gedare@ip98-169-179-252.dc.dc.cox.net> has quit IRC (Ping timeout: 272 seconds)
2013-11-10T12:33:28  *** qingpei <qingpei!~qingpei@123.116.42.108> has quit IRC (Quit: qingpei)
2013-11-10T15:35:11  *** antgreen <antgreen!~green@dsl-207-112-125-176.tor.primus.ca> has quit IRC (Ping timeout: 272 seconds)
2013-11-10T15:36:58  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-10T15:39:53  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 272 seconds)
2013-11-10T16:25:41  *** kiwichris <kiwichris!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has joined #rtems
2013-11-10T16:26:07  *** kiwichri_ <kiwichri_!~textual@CPE-58-172-129-85.nfcz1.ken.bigpond.net.au> has quit IRC (Ping timeout: 252 seconds)
2013-11-10T17:16:18  *** Guest64805 <Guest64805!~daniel@2602:306:ce88:6290:2cb7:7f88:6516:f41d> has joined #rtems
2013-11-10T18:17:04  *** mkhoory <mkhoory!~mkhoory@211.117.165.123> has joined #rtems
2013-11-10T18:17:24  <mkhoory> morning
2013-11-10T18:28:15  *** mkhoory <mkhoory!~mkhoory@211.117.165.123> has left #rtems ("Leaving")
2013-11-10T18:28:20  *** mkhoory <mkhoory!~mkhoory@211.117.165.123> has joined #rtems
2013-11-10T20:51:26  *** Guest64805 <Guest64805!~daniel@2602:306:ce88:6290:2cb7:7f88:6516:f41d> has quit IRC (Ping timeout: 240 seconds)
2013-11-10T20:53:44  *** Guest64805 <Guest64805!~daniel@2602:306:ce88:6290:2cb7:7f88:6516:f41d> has joined #rtems
2013-11-10T22:12:57  *** qingpei <qingpei!~qingpei@125.33.126.9> has joined #rtems
2013-11-10T22:49:01  *** qingpei <qingpei!~qingpei@125.33.126.9> has quit IRC (Quit: qingpei)
2013-11-10T23:03:05  *** qingpei <qingpei!~qingpei@125.33.126.9> has joined #rtems
2013-11-10T23:08:58  *** mkhoory <mkhoory!~mkhoory@211.117.165.123> has quit IRC (Ping timeout: 240 seconds)
2013-11-10T23:14:22  *** mkhoory <mkhoory!~mkhoory@211.117.165.123> has joined #rtems
2013-11-10T23:14:29  *** qingpei <qingpei!~qingpei@125.33.126.9> has quit IRC (Quit: qingpei)
2013-11-10T23:33:29  *** Guest64805 <Guest64805!~daniel@2602:306:ce88:6290:2cb7:7f88:6516:f41d> has quit IRC (Remote host closed the connection)
2013-11-10T23:41:44  *** qingpei <qingpei!~qingpei@125.33.126.9> has joined #rtems