| 00:39.41 | *** join/#edev kergoth (n=kergoth@ip24-251-170-95.ph.ph.cox.net) |
| 00:44.03 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 01:40.02 | *** join/#edev landley (n=landley@cpe-66-68-109-179.austin.res.rr.com) |
| 01:40.03 | *** mode/#edev [+o landley] by ChanServ |
| 02:02.55 | Mirell | Mmm, this mStand is a beaut. |
| 02:20.06 | landley | Mirell: what's an mStand? |
| 02:22.59 | Mirell | landley: http://www.raindesigninc.com/mstand.html |
| 02:35.19 | *** join/#edev ohsix (i=ohsix@66.220.111.99) |
| 02:37.31 | *** join/#edev SandGorgon (n=OmNomNom@122.173.249.152) |
| 02:47.27 | *** join/#edev olivdt_ (n=x@modemcable206.214-57-74.mc.videotron.ca) |
| 02:59.10 | landley | Mirell: Ooh, that's really cool. |
| 02:59.18 | landley | How does it do on keyboard wrist-strain? |
| 02:59.28 | landley | (Or do you use an external keyboard with it?) |
| 03:06.02 | Mirell | landley: You use an external keyboard and mouse with it |
| 03:06.32 | Mirell | landley: It's mainly designed to bring it up to the right height for viewing, like an external monitor. |
| 03:33.56 | rektide | there are so many fucking douche bags in #linux |
| 03:34.10 | rektide | yet there are essentially no other "shooting the shit" #linux channels on the net |
| 03:34.20 | landley | rektide: Sturgeon's Law is universal. |
| 03:34.23 | Mirell | Never tried that channel. |
| 03:34.31 | Mirell | #linux is way, way, way to general a channel topc. |
| 03:34.33 | landley | Mirell: I have. Traffic's too high. |
| 03:35.07 | rektide | erp, irssi failure. i thought i was talking to a different channel, on irc.arstechnica.com |
| 03:35.07 | landley | It's a bit like the door64 Happy Hours. If you can hear yourself think, you're early. |
| 03:35.29 | rektide | theres no pretense of problem solving / helping people there, its just linux bullshit. |
| 03:35.39 | rektide | ruled over by insane power hungry insulting mods |
| 03:36.21 | rektide | how small a system footprint can linux fit on / |
| 03:36.42 | Mirell | Eh, probably a size 2 |
| 03:36.48 | Mirell | I really don't know women's shoe sizes. |
| 03:37.18 | landley | rektide: I've seen it in 4 megs dram and 2 megs flash. |
| 03:37.22 | rektide | could you get a usable system out of a meg of storage and half a meg of ram |
| 03:37.31 | landley | Generally that small you go with a nommu system to eliminate the overhead of page tables. |
| 03:37.32 | Mirell | rektide: Not a 2.6 kernel |
| 03:37.33 | rektide | landley: i could do that on openwrt |
| 03:37.37 | landley | rektide: no. |
| 03:37.41 | landley | Not even with older kernels. |
| 03:37.45 | rektide | landley: the 4m ram / 2m flash |
| 03:37.54 | landley | The smallest anybody's _ever_ managed to get working was 2 megs of ram, and that was like a 1.x kernel. |
| 03:38.18 | landley | Although if your kernel is running _out_ of the flash, it might be theoretically possible. |
| 03:38.31 | rektide | woo xip |
| 03:38.43 | landley | But for a gzipped kernel image to extract itself requires 2 megs of ram due to the memory mappings. The kernel loads at the 1 meg mark and then gunzips itself to address 0. |
| 03:39.13 | rektide | oh man |
| 03:39.19 | landley | An "allnoconfig" kernel is over a megabyte uncompressed these days. (Which is sad, but true.) |
| 03:39.47 | landley | For years I've wanted to do a "hello world" kernel that has just the boot wrapper stuff and the early_printk() driver to write "hello world" to the console. |
| 03:39.53 | rektide | are there any quasi mainstream open source posix-y embedded kernejls? |
| 03:40.03 | landley | And then add back in the task switching and vmalloc and vfs and so on, incrementally. |
| 03:40.16 | landley | rektide: yes. |
| 03:40.16 | rektide | :effort: |
| 03:40.49 | landley | Some people use bsd's for this, but they're not that much of an improvement. (The main advantage they have is that they're old and only weakly maintained, so some of 'em haven't changed much since systems were smaller.) |
| 03:41.00 | landley | netbsd, for example. Preserved in amber. |
| 03:41.51 | rektide | i wonder if l4 + whatever that linux userland is would/could be smaller |
| 03:42.21 | landley | Some people use newlib+libgloss. |
| 03:42.42 | landley | You basically link your program against a libc and some device drivers and run on the bare metal. |
| 03:42.45 | rektide | hadnt heard of either |
| 03:42.50 | landley | Not task switching, no nothing. |
| 03:42.54 | rektide | heh |
| 03:43.01 | rektide | i would like an actual OS on my embedded gear |
| 03:43.02 | landley | Somebody once got busybox to work linked that way. Well, a subset of the applets anyway. |
| 03:43.44 | rektide | newlib is... redhat? weird |
| 03:43.53 | landley | This was the guy who made busybox work: http://sourceware.org/ml/newlib/2005/msg00138.html |
| 03:44.07 | landley | Red Hat acquired an embedded company a few years ago. |
| 03:44.21 | landley | They didn't develop it, they got it in an acquisition. |
| 03:45.00 | landley | Keep in mind that Red Hat also acquired cygnus, which did a lot of this stuff. |
| 03:45.01 | rektide | thats slightly more understandable. i dont really think of rh as an embedded player. |
| 03:45.42 | landley | Although most of the cygnus developers quit, and about half of 'em eventually re-coalesced under Code Sourcery. |
| 03:46.01 | landley | They wanted cygnus because it did toolchain development. The gcc and glibc maintainers worked there. |
| 03:47.25 | landley | There's a linux-tiny project. Matt Mackall used to maintain it, then whatsisface took over. |
| 03:47.33 | landley | Name... name... Guy who runs CELF... |
| 03:47.37 | landley | Tim Bird. |
| 03:48.03 | landley | http://elinux.org/Linux_Tiny |
| 03:48.28 | landley | Go read up on that. |
| 03:48.33 | landley | Also, there's an embedded mailing list on vger. |
| 03:49.10 | landley | Waiting for kmail to open the folder. |
| 03:49.14 | landley | I REALLY need a new email client... |
| 03:49.51 | landley | linux-embedded@vger.kernel.org |
| 03:50.11 | rektide | cygnus, commercial arm of cygwin, yeah, right |
| 03:53.20 | landley | Actually cygnus came first, cygwin was just one of their projects (a port of the gnu development environment to windows). |
| 03:53.43 | landley | The maintainer of cygwin left cygnus after the red had acquisition, and the project sort of forked. I didn't really keep track of it... |
| 03:54.04 | landley | cygnus stands for "Cygnus, Your Gnu Support". |
| 03:54.08 | landley | A recursive acronym, like gnu. |
| 03:54.14 | rektide | i want to migrate from this Scenix/Ubicom platform for my home automation network to something with 802.15.3 onboard |
| 03:54.41 | rektide | and that doesnt have <512 bytes of ram and 4k of flash |
| 03:54.54 | landley | http://oreilly.com/catalog/opensources/book/tiemans.html |
| 03:55.01 | landley | History of cygnus, from its founder. :) |
| 03:55.11 | rektide | dont really know what to look at... contiki is one of the few i've heard of |
| 03:55.36 | landley | Generally, the low-end Linux system most people bother with is 4 megs dram, 2 megs flash. |
| 03:55.37 | rektide | oh sweet, nice article, thanks for the link |
| 03:55.46 | landley | It's possible to go below that, but you really have to know what you're doing. |
| 03:55.53 | landley | rektide: computer history is one of my hobbies. |
| 03:55.54 | rektide | yeah, thats still a fat requirement for embeeded |
| 03:56.09 | landley | http://landley.net/history/mirror/ |
| 03:56.18 | landley | rektide: depends on your definition of embedded. |
| 03:56.58 | landley | You can buy a low-end router that has way more than that for under $50, but buying systems that _aren't_ routers with those specs that cheaply turns out to be difficult. |
| 03:57.06 | landley | It's the old "cost is entirely a matter of unit volume" problem. |
| 03:57.21 | rektide | this is for building, not buying |
| 03:57.55 | landley | The systems that knock out a million units a quarter have a cost basis of maybe $15 for this hardware. The ones that knock out 50,000 units it costs $50 for the same hardware. The ones that do 10,000 units it costs $100. :( |
| 03:57.57 | rektide | i have a stack of wgt634u's, a sheevaplug and openrd |
| 03:58.03 | rektide | they're all good linux systems |
| 03:58.10 | rektide | but this home automation system |
| 03:58.12 | rektide | again |
| 03:58.18 | rektide | i'm coming from 512 bytes of ram and 2k flash |
| 03:58.27 | landley | prpplague put together a system called the "hammer board", which cost about $150 for the dev kit I think. |
| 03:58.48 | rektide | oh man thats nice hardware |
| 03:59.00 | rektide | really nice hardware |
| 03:59.02 | landley | http://www.tincantools.com/product.php?productid=16143 |
| 03:59.05 | rektide | yeah |
| 03:59.11 | landley | http://www.tincantools.com/product.php?productid=16133 |
| 03:59.23 | landley | At the moment, prpplague_afk :) |
| 03:59.36 | landley | But you can ask him in the morning, if he shows up. He's the guy who designed most of that. |
| 03:59.57 | landley | (The nail board is based on my complaints about their original dev kit for the hammer module. :) |
| 04:00.10 | landley | I told 'em what I really wanted, and they produced it. :) |
| 04:00.19 | landley | Alas, nobody's ever bought hammer modules in high enough quantity to get the price down. |
| 04:00.53 | landley | I believe if they could produce a 50,000 unit run they could sell 'em for under $50 each. (Doing 'em for $20 each would require at least doubling that, probably a whole lot more.) |
| 04:01.15 | landley | Unfortunately, I believe they sell in the hundreds, not the thousands. :( |
| 04:01.42 | landley | Still, if you wanted a physically small, low power system: there you go. |
| 04:02.00 | rektide | and capable |
| 04:02.06 | landley | (You could also go with the "beagle board", which needs something like 3x the wattage but is a lot more powerful. That's even got video. Not really any cheaper, though.) |
| 04:02.15 | rektide | even at 200mhz, i think it may well kick the living pants off my wgt634u's wifi routers |
| 04:02.16 | landley | Mirell has a beagle board. It's nice. |
| 04:02.28 | rektide | i have an Pandora on order |
| 04:02.41 | rektide | hand held unit with the same SoC as the beagle |
| 04:02.43 | landley | It's also physically much larger than the hammer module, and weighs maybe 10x as much. |
| 04:03.09 | landley | But I haven't tried to take the dev kit apart, so I dunno what the SoC is like on its own. |
| 04:03.22 | landley | (I'm a software person, not a hardware person.) |
| 04:03.37 | landley | My pet project (Firmware Linux) creates native development environments you run under QEMU. |
| 04:03.47 | landley | Native compilation under emulation: because cross compiling sucks rocks. |
| 04:04.21 | landley | And if you have your native compiler call out to the cross compiler via distcc, it's not _too_ slow... |
| 04:04.53 | rektide | the emdebian cross compilers have been really nice |
| 04:05.26 | landley | Cool. |
| 04:05.29 | rektide | havent used em yet via distcc ; that would be convenient |
| 04:05.45 | landley | emdebian, buildroot, code sourcery, crosstool-ng... |
| 04:05.49 | landley | I should put together a list. |
| 04:05.58 | rektide | need to find a lighter weight OS now |
| 04:06.31 | landley | All that really comes to mind is netbsd. |
| 04:06.39 | landley | And all I know about it is that it's A) portable, B) stagnant. |
| 04:06.59 | rektide | smaller |
| 04:07.04 | rektide | contiki, for example |
| 04:09.15 | rektide | http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MC13224V&fsrch=1 |
| 04:09.44 | rektide | maximally integrated SoC, no space for a "real" OS |
| 04:10.07 | landley | Poke at contiki then? I haven't got a better suggestion. |
| 04:10.26 | rektide | my knowledge is rusty is the issue |
| 04:10.35 | rektide | i dont know what other options are on the field |
| 04:11.09 | rektide | that freescale is a great wireless sensor board |
| 04:11.11 | rektide | http://www.redwirellc.com/store/ |
| 04:11.37 | landley | I'm not the person to ask. :) |
| 04:12.17 | rektide | no worries. i enjoyed our chat. |
| 04:12.37 | rektide | plenty of valuable tid-bits, even if not directly applicable to this particular demented quest |
| 04:13.30 | rektide | honestly it'd be much more convenient to go 802.11bgn instead of 802.15.3, but that leaves the SoC niche, i think |
| 04:16.52 | landley | I dunno what 802.15.3 is. |
| 04:42.00 | *** join/#edev mbuf (n=mbuf@118.102.130.6) |
| 04:48.43 | rektide | 802.15.3 is the basis of zigbee, which is a protocol stack implemented on 802.15.3, which itself is 250kbit/s wireless, no specified topology. |
| 05:25.05 | *** join/#edev SandGorgon (n=OmNomNom@122.160.41.129) |
| 06:29.32 | *** join/#edev MonMotha (n=monmotha@cpe-24-208-213-87.indy.res.rr.com) |
| 07:18.42 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 07:32.37 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 07:36.11 | *** join/#edev linac (n=lin@59.61.122.146) |
| 07:37.54 | *** join/#edev olivdt (n=x@modemcable206.214-57-74.mc.videotron.ca) |
| 07:57.07 | *** join/#edev wmat (n=btraynor@gromit.mixdown.ca) |
| 08:00.02 | *** join/#edev roxfan (i=dunno@143.131-246-81.adsl-dyn.isp.belgacom.be) [NETSPLIT VICTIM] |
| 08:01.58 | *** join/#edev cjameshuff (n=cjameshu@pool-71-97-195-26.aubnin.dsl-w.verizon.net) [NETSPLIT VICTIM] |
| 08:02.15 | *** join/#edev Crofton|work (n=balister@pool-71-171-31-89.nwrknj.east.verizon.net) [NETSPLIT VICTIM] |
| 08:03.46 | *** join/#edev tsjsieb (n=tsjsieb@dejongbeheer.nl) |
| 08:04.01 | *** join/#edev solar (n=solar@140.211.166.183) |
| 08:04.06 | *** join/#edev Crofton (n=balister@pool-71-171-31-89.nwrknj.east.verizon.net) |
| 08:17.23 | *** join/#edev MonMotha (n=monmotha@cpe-24-208-213-87.indy.res.rr.com) |
| 09:22.07 | *** join/#edev florian_kc (n=fuchs@Maemo/community/contributor/florian) |
| 09:26.05 | *** join/#edev insomniac (n=insomnia@slackware.it/staff/insomniac) |
| 09:26.10 | insomniac | hi to everyone :) |
| 09:27.17 | insomniac | has anyone ever debugged a BFLT binary on-target (ARM7/uClinux) via gdbserver? |
| 09:28.35 | insomniac | I'm having issues while trying to set breakpoints |
| 09:29.51 | insomniac | I start gdbserver on the target, run gdb on the host, load the symbols from the generated .gdb executable, then.. |
| 09:29.54 | insomniac | (gdb) b main |
| 09:29.56 | insomniac | Cannot access memory at address 0x28 |
| 09:39.44 | florian | good morning |
| 09:42.06 | insomniac | good morning florian |
| 10:05.10 | roxfan | sounds like you need to tell gdb the actual load base of your binary |
| 10:08.38 | roxfan | maybe "add-symbol-file filename address" ? |
| 10:11.11 | insomniac | I guessed that, too, but I don't know how to get the base of the binary |
| 10:11.25 | insomniac | can you help on that, or point me to some documentation? |
| 10:12.04 | roxfan | sorry, not familiar with uclinux |
| 10:12.21 | insomniac | thank you anyway :) I'll try more |
| 10:27.24 | *** join/#edev eox425 (n=root@82.207.127.231) |
| 10:54.10 | *** join/#edev pirho (i=pirho@gateway/gpg-tor/key-0x2CEEC9CB) |
| 11:27.02 | *** join/#edev bsdfox_ (n=h36sa@c-67-161-174-21.hsd1.ca.comcast.net) |
| 11:36.37 | *** join/#edev bsdfox_ (n=h36sa@unaffiliated/bsdfox) |
| 11:54.19 | *** join/#edev bsdfox_ (n=h36sa@unaffiliated/bsdfox) |
| 12:11.45 | *** join/#edev gustavoz (n=gustavoz@host108.201-253-243.telecom.net.ar) |
| 12:14.31 | *** join/#edev bsdfox_ (n=h36sa@c-67-161-174-21.hsd1.ca.comcast.net) |
| 12:41.14 | *** join/#edev dijenerate (n=dijenera@64.210.44.37) |
| 12:47.37 | *** join/#edev gustavoz (n=gustavoz@host108.201-253-243.telecom.net.ar) |
| 12:56.55 | *** join/#edev dijenerate (n=dijenera@64.210.44.37) |
| 13:01.38 | *** join/#edev dijenerate_ (n=dijenera@64.210.44.37) |
| 13:02.25 | *** join/#edev dijenerate (n=dijenera@64.210.44.37) |
| 13:56.57 | *** join/#edev linac (n=lin@122.90.86.204) |
| 14:14.08 | *** join/#edev GPSFan (n=kenm@64.92.145.112) |
| 14:14.08 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:15.49 | *** join/#edev mbuf (n=mbuf@118.102.130.6) |
| 14:37.10 | *** join/#edev kergoth (n=kergoth@nat/montavista/x-tdlnsyhviifqusnd) |
| 14:40.07 | kergoth | morning |
| 14:43.02 | prpplague | kergoth: greetings |
| 14:44.01 | *** join/#edev kazaa_lite (n=msaleem@94-193-98-124.zone7.bethere.co.uk) |
| 14:45.57 | chouimat | morning |
| 14:46.23 | Mirell | So this morning is awesome.. |
| 14:46.48 | Mirell | I did laundry last night, except I forget to put my clothes in the dryer before falling asleep |
| 14:46.58 | Mirell | So now, drying my clothes, since I have no non-wet pants, and going to be late for work. |
| 14:47.19 | chouimat | Mirell: I hope you work for "clothing optional" company ;) |
| 14:48.45 | Mirell | chouimat: Eh, no. Hence why I'm sitting at home currently. |
| 14:48.55 | chouimat | k |
| 14:49.08 | prpplague | Mirell: doh |
| 14:49.13 | prpplague | Mirell: i hate it when that happens |
| 14:49.14 | kergoth | apparently Mirell is unfamiliar with the concept of humor |
| 14:49.29 | Mirell | kergoth: No, just kind of sorta annoyed at myself. |
| 14:49.48 | Mirell | Also, the sheer lameness of the excuse for being late |
| 14:49.52 | Mirell | "I had no dry pants" |
| 14:49.55 | prpplague | Mirell: thats why i got some closet elves, i leave dirty clothes in the hamper and the elves clean them and put them on hangers in the closet |
| 14:50.00 | kergoth | chuckles |
| 14:50.07 | kergoth | that sucks |
| 14:50.25 | Mirell | prpplague: That's called a "wife" in most countries |
| 14:50.33 | prpplague | Mirell: are you sure? |
| 14:50.48 | kergoth | no emergency fallback clothes? i usually keep at least a few things that i never wear under normal circumstances for laundry emergencies |
| 14:50.49 | chouimat | prpplague: step 1) dirty underwear step2)?????? step 3)world domination |
| 14:51.17 | Mirell | kergoth: I should do that... |
| 14:52.02 | chouimat | kergoth:I do that too, so the day I show up at work wearing a suit is a) got a job interview or b) forgot the laundry |
| 14:52.09 | kergoth | hehe |
| 14:52.49 | prpplague | Mirell: i was off to met koen in richardson a few weeks ago and someone pulled out in front of me causing me to spill coffee all over my cloths, since then i've started caring an extra set of clothes in my trunk |
| 14:53.15 | Mirell | Smart idea. |
| 14:53.46 | prpplague | has to tell his customer that the project is 40 to 80 hours behind schedule :( |
| 14:53.56 | kergoth | :( |
| 14:54.34 | prpplague | i don't think they will be too upset since it involved alot of reverse engineering |
| 14:54.45 | prpplague | still i just don't like breaking that kind of news |
| 14:55.21 | prpplague | Mirell: oh and btw, silver is at 18.34 and gold at 1150 |
| 14:55.26 | Mirell | prpplague: ...thanks |
| 14:55.33 | prpplague | Mirell: hehe |
| 14:55.53 | *** topic/#edev by Mirell -> A channel for the discussion of embedded systems (Hardware/Software/Linux/WinCE/Commercial Device Repurposing) :: Silver 18.34 :: Gold 11150 |
| 14:56.08 | *** topic/#edev by Mirell -> A channel for the discussion of embedded systems (Hardware/Software/Linux/WinCE/Commercial Device Repurposing) :: Silver 18.34 :: Gold 1150 |
| 14:56.26 | prpplague | hehe |
| 14:57.00 | prpplague | thinks gold will hit 1800 by the end of the year |
| 14:57.28 | Mirell | That seems a bit much... |
| 14:58.03 | prpplague | i think by this time next year it willbe 2500 |
| 14:58.35 | SpeedEvil | I still think that aluminium purchased in ~march this year will beat it |
| 14:59.53 | prpplague | tries to get motivated today |
| 15:00.08 | prpplague | gotta stop at 10:30 an take my step daughter to the dentist |
| 15:34.06 | *** join/#edev SandGorgon (n=OmNomNom@122.173.249.152) |
| 15:36.17 | prpplague | wmat: ping |
| 15:36.37 | wmat | prpplague: pong |
| 15:37.52 | prpplague | wmat: hey, enlighten me..... on the wiki when you create a new page using the: == item == and === subitem === , how do you get the contents section at the top of the page? |
| 15:39.10 | wmat | it's automatic |
| 15:39.41 | prpplague | wmat: hmm doesn't seem to be working |
| 15:39.53 | prpplague | ahh wait looks like you have to have atleast 2 subitems |
| 15:39.56 | prpplague | wmat: nm, thanks |
| 15:40.07 | wmat | np |
| 15:46.29 | *** join/#edev mbuf (n=mbuf@122.169.140.76) |
| 15:47.30 | Mirell | Mmm. |
| 16:00.37 | *** join/#edev brolin (n=brolin@iner.udea.edu.co) |
| 16:20.02 | *** join/#edev likewise (n=chatzill@82-171-51-231.ip.telfort.nl) |
| 17:02.40 | *** join/#edev CosmicPenguin (n=nobody@66-162-99-237.static.twtelecom.net) |
| 17:17.43 | *** join/#edev pirho (i=pirho@gateway/gpg-tor/key-0x2CEEC9CB) |
| 17:28.15 | *** join/#edev sjhill (n=sjhill@home.bethel-hill.org) |
| 17:28.15 | *** mode/#edev [+o sjhill] by ChanServ |
| 18:09.27 | calculus | rektide: you mentioned the beagle board, there is gumstix which is the same SoC, but smaller form factor and some models have wireless/bluetooth on board |
| 18:13.33 | *** join/#edev gustavoz (n=gustavoz@host108.201-253-243.telecom.net.ar) |
| 18:44.36 | Mirell | Oh I do not like Tcl... |
| 18:44.55 | *** join/#edev prpplague (n=Dave@adsl-99-57-149-34.dsl.rcsntx.sbcglobal.net) |
| 19:06.25 | *** join/#edev roxfan2 (i=dunno@38.223-240-81.adsl-dyn.isp.belgacom.be) |
| 19:13.10 | *** join/#edev florian (n=fuchs@Maemo/community/contributor/florian) |
| 19:39.23 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 19:39.23 | *** join/#edev cjameshuff (n=cjameshu@pool-71-97-195-26.aubnin.dsl-w.verizon.net) |
| 20:00.49 | Mirell | Buildbot is not a pretty thing |
| 20:07.46 | CosmicPenguin | Oh? I thought it was pretty sexy myself |
| 20:14.30 | Mirell | ...really |
| 20:15.53 | calculus | uses hudson |
| 20:16.23 | calculus | it is quite simple to use for our use, slaves are quite easy to add |
| 20:16.48 | calculus | mainly, out of the box support for java/ant/maven |
| 20:17.49 | Mirell | We have a mixture of Java, C, C++, Flash, buildroot... |
| 20:19.06 | prpplague | misses being able to post his finds on elinux.org |
| 20:19.21 | calculus | oh, and I have seen plugins for msbuild and nant things |
| 20:19.38 | prpplague | had to set up a wiki on my local box just so i could keep notes like i'm used to on the wiki |
| 20:20.16 | CosmicPenguin | Mirell: tell more - what concerns you about buildbot? |
| 20:20.38 | SpeedEvil | prpplague: Seen tiddlywiki? |
| 20:20.48 | prpplague | SpeedEvil: sorry no |
| 20:20.57 | SpeedEvil | prpplague: js wiki - totally in-browser - though it can integrate with other stuff. tiddlywiki.com |
| 20:21.07 | CosmicPenguin | xim is also a local wiki thingy |
| 20:21.10 | CosmicPenguin | though I don't like its markup |
| 20:22.15 | prpplague | yea i setup mediawiki like whats on elinux.org so that if i get the approval i can just add the pages to elinux.org |
| 20:22.16 | Mirell | CosmicPenguin: How can I kick off a build remotely. |
| 20:22.19 | Mirell | CosmicPenguin: Through a script |
| 20:22.30 | CosmicPenguin | I'm not sure on the particulars |
| 20:23.00 | Mirell | Nor am I, hence my current annoyance. |
| 20:23.36 | Mirell | If you have a way for a human to do it on the website, there should be an API, even just simple JSON, I don't care. |
| 20:46.11 | *** join/#edev ja2 (n=a@81-233-153-190-no83.tbcn.telia.com) |
| 20:46.48 | ja2 | Hi, can anyone give me a hint about what cacheline sharing is all about? |
| 20:47.46 | prpplague | ja2: hard to say without some context |
| 20:48.01 | ja2 | The thing is I lose the USB-stick, it disconnects. |
| 20:48.08 | ja2 | And then reconnects. |
| 20:48.22 | ja2 | But by then the rootfs is gone. :/ |
| 20:48.36 | ja2 | So Im fumbling in the dark to find a cause. |
| 20:49.05 | prpplague | ja2: need to provide some background information about what platform you are working with and what you are trying to do |
| 20:49.12 | ja2 | I think dont really see this if the rootfs is else where, not on USB. |
| 20:49.38 | ja2 | This is a dev.board in the making. EP9307 ARM. |
| 20:49.56 | ja2 | Runs linux 2.6.24. Im just a random tester in the project. |
| 20:51.08 | ja2 | What more can I supply you with? Error dumps? |
| 20:52.21 | ja2 | One of the strange thing with this is that whenever a USB fault has occured. The next boot will be some garbage on the serial console. |
| 20:52.53 | ja2 | Sometimes only garbage, sometimes a little. But there seem to be a connection with the USB disconnect. |
| 20:53.58 | prpplague | ja2: most likely from what you describing, from what little there is, you probably have a USB issue where the rootfs fails in some manner generating incorrect cached data |
| 20:54.03 | ja2 | There's also a SD-card on SPI bus. But that is slow. So rootfs on USB-stick would be nice. |
| 20:55.02 | ja2 | About the cacheline sharing it was just something I saw in linux-2.6.24.7/drivers/input/serio.c |
| 20:55.03 | CosmicPenguin | hmm - is there a way to turn off the "did you mean this?" spam from git? |
| 20:55.26 | ja2 | So that is not anything that the acctual error dumps have showed. |
| 20:56.22 | ja2 | I figured maybe the serial and USB got collisions. |
| 20:56.29 | ja2 | Would that be possible? |
| 20:56.36 | ja2 | buffer overwrites? |
| 20:56.42 | ja2 | *overruns |
| 20:57.57 | ja2 | hub 1-0:1.0: port 1 disabled by hub (EMI?), re-enabling... |
| 20:57.58 | ja2 | usb 1-1: USB disconnect, address 3 |
| 20:58.08 | ja2 | That's the actual error mesg. |
| 20:58.23 | ja2 | I dont believe EMI, though. But hey, anything is possible. |
| 20:59.42 | ja2 | That err mesg is from 2.6.24/drivers/usb/host/ehci-xxx.c (something) |
| 20:59.52 | ja2 | ehci-q.c |
| 21:00.46 | ja2 | I think newer kernels (2.6.31) have this code reworked pretty much. But this project is stuck at .24 |
| 21:00.50 | prpplague | ja2: yea that definetly would cause a problem if your rootfs is on that usb device |
| 21:00.59 | ja2 | :) |
| 21:01.55 | ja2 | Finding a way for a immediate fix would be ideal, to continue work to reach .31 :) |
| 21:02.23 | prpplague | ja2: need to look at why it thinks it need to disable the root hib |
| 21:02.24 | prpplague | hub |
| 21:03.10 | ja2 | If I go dd if=/dev/zero of=file count=131070 I randomly get this disconnect, or not. |
| 21:03.25 | ja2 | if count=191000, so far it has alwas disconnected. |
| 21:03.56 | ja2 | Sometimes at boot, disconnect is immediatly. Sometimes you can work for a couple of minutes, 10-20 min perhaps. |
| 21:04.38 | prpplague | ja2: like i said you need to look at the ehci driver and turn on some debug to find the reason it thinks that the root hub has a problem |
| 21:04.40 | ja2 | Secondary stackdumps from fs/buffer.c:1169 mark_buffer_dirty() is available. |
| 21:05.08 | ja2 | yes. Ok, I try to find more. |
| 21:05.21 | prpplague | ja2: i only have time to give you pointers, cant do your work for you |
| 21:05.39 | ja2 | Sure. Your help was appreciated. |
| 21:05.45 | ja2 | It always help to discuss. |
| 21:06.09 | ja2 | newbie, no shit. |
| 21:07.29 | prpplague | ja2: just a note, this channel is frequented by many professionals who are on their company networks, with that in mind, please mind the use of your language |
| 21:07.42 | ja2 | Sometimes the error is not where the error mesg is. |
| 21:07.46 | ja2 | Sorry for being rude. |
| 21:18.27 | CosmicPenguin | Now I can't get Ferris Bueller's Day Off out of my head |
| 21:18.28 | CosmicPenguin | thanks prpplague |
| 21:19.41 | prpplague | CosmicPenguin: ?? |
| 21:24.27 | CosmicPenguin | Relax sir - I'm a professional... |
| 21:24.30 | CosmicPenguin | A professional what? |
| 21:26.32 | chouimat | hehe |
| 21:26.57 | *** join/#edev brolin (n=brolin@201.232.107.129) |
| 21:27.45 | prpplague | ahh some movie reference |
| 21:28.07 | CosmicPenguin | not just some move |
| 21:28.09 | CosmicPenguin | movie |
| 21:28.50 | kergoth | prpplague hasn't seen ferris bueller's day off? this must be dealt with |
| 21:29.17 | CosmicPenguin | Seriously |
| 21:29.18 | prpplague | kergoth: i've seen it, just years and years ago and i didn't find it that memorable |
| 21:29.26 | kergoth | gasps |
| 21:29.29 | CosmicPenguin | heh |
| 21:29.34 | CosmicPenguin | I was about to type the same thing |
| 21:30.14 | chouimat | can't stand ferris bueller's day off anymore ... thanks to his sister who watched it way too much back then |
| 21:30.15 | kergoth | put "Oh Yeah" from that movie on his itunes wishlist the other day :D |
| 21:30.24 | Mirell | OOOOOOOOOOOHHH YEEEEEEEEEEEAH |
| 21:30.29 | Mirell | *boom boom* |
| 21:30.33 | Mirell | You want some of my gum? |
| 21:31.04 | kergoth | has the kool-aid guy scenes from family guy popping into his head now |
| 21:31.12 | CosmicPenguin | Now I do too |
| 21:31.14 | CosmicPenguin | damn |
| 21:31.21 | prpplague | as do i |
| 21:32.36 | chouimat | the court one? |
| 21:33.21 | prpplague | my favorite of all family guy episodes is the one where he goes the brewery and its a take off of "willie wonka and the chocolate factory" |
| 21:33.35 | kergoth | thats a classic |
| 21:33.46 | chouimat | prpplague: why I'm not surprised |
| 21:33.48 | CosmicPenguin | I wish they had done the whole episode |
| 21:33.50 | kergoth | hehe |
| 21:33.54 | CosmicPenguin | There was so much more they could have done |
| 21:34.03 | prpplague | CosmicPenguin: that was my thought exactly |
| 21:34.44 | CosmicPenguin | See also - the slurm episode from Futurama |
| 21:34.59 | prpplague | hehe indeed, i love that one as well |
| 21:35.26 | prpplague | although i like the one when fry at the worm infested egg salad sandwich |
| 21:35.36 | prpplague | s/at/ate |
| 21:36.08 | CosmicPenguin | That one is a classic |
| 21:36.25 | CosmicPenguin | I think the top two fan picks are that one and the luck of the fryish |
| 21:36.28 | prpplague | must go collect his yoots |
| 21:45.27 | *** join/#edev MonMotha (n=monmotha@cpe-24-208-213-87.indy.res.rr.com) |
| 21:46.33 | CosmicPenguin | didn't think FBDO was memorable - that worries me |
| 21:53.58 | *** part/#edev ja2 (n=a@81-233-153-190-no83.tbcn.telia.com) |
| 21:55.27 | *** join/#edev Madsy (n=Madsy@fu/coder/madsy) |
| 22:08.28 | *** join/#edev pirho (i=pirho@gateway/gpg-tor/key-0x2CEEC9CB) |
| 22:25.35 | prpplague | CosmicPenguin: FBDO ? |
| 22:38.07 | SpeedEvil | (09:28:50 PM) kergoth: prpplague hasn't seen ferris bueller's day off? this must be dealt with |
| 22:38.25 | prpplague | SpeedEvil: ahh |
| 22:40.43 | prpplague | CosmicPenguin: just didn't find it that interesting |
| 22:40.56 | Mirell | It's proof that you are a failure, then! |
| 22:41.01 | Mirell | FBDO is mandatory for the human race. |
| 22:41.05 | Mirell | :D |
| 22:41.16 | prpplague | Mirell: must be |
| 22:43.14 | kergoth | damnit |
| 22:43.23 | kergoth | i keep losing track of where i was working on various things |
| 22:43.35 | prpplague | has always had different opinions about such things |
| 22:43.51 | Mirell | I keep not being motivated to work on Tcl anymore |
| 22:43.53 | kergoth | i know i was doing this in a git branch.. but which repo is the question.. |
| 22:43.58 | prpplague | kergoth: that happens when you try to do dev work on your laptop at the nudie bar |
| 22:44.05 | kergoth | heh |
| 22:44.12 | kergoth | just has way too many trees |
| 22:44.18 | Mirell | Nudie bars here have free wireless |
| 22:45.56 | Crofton|work | where is here? |
| 22:46.21 | Mirell | Austin, TX |
| 22:49.28 | *** join/#edev archae0pteryx (n=snewman@207.47.42.130.static.nextweb.net) |
| 22:58.00 | rektide | anyone familiar with embedded os's? |
| 22:58.27 | prpplague | rektide: i'm sure someone here is |
| 22:58.28 | rektide | i'm interested in the freescale mc1322, which is an arm chip with 802.15.3 |
| 22:58.32 | rektide | 80k ram |
| 22:58.34 | rektide | iirc |
| 22:58.40 | rektide | little more flash |
| 22:59.08 | rektide | i'd like to be able to have some starting platform for it, rather than diy everything, as i've done with all my other micro projects |
| 22:59.33 | prpplague | rektide: not going to get much of an OS on that |
| 22:59.46 | calculus | prpplague: don't feel so bad, I haven't even seen this movie that has been spoken of |
| 23:00.17 | rektide | prpplague: obviously nothing as full fledged as a posix os, |
| 23:00.30 | rektide | more just the form of having some infrastructure for doing development with |
| 23:01.06 | kergoth | hmm |
| 23:11.35 | MonMotha | 802.15.3 or 802.15.4? |
| 23:11.50 | MonMotha | and FreeRTOS will fit on there |
| 23:12.05 | MonMotha | not that FreeRTOS is much more than a scheduler |
| 23:12.32 | florian | true |
| 23:12.44 | florian | NutOS or Contiki might be interesting |
| 23:13.25 | MonMotha | You can also roll your own. A basic OS (scheduler, some locking primitives, and some memory management of sorts) isn't terribly complicated. |
| 23:14.00 | MonMotha | remember that you don't have an MMU, though |
| 23:14.46 | calculus | maybe RTEMS (I only know of it since someone mentioned it to me) |
| 23:45.18 | *** join/#edev MonMotha (n=monmotha@cpe-24-208-213-87.indy.res.rr.com) |