#qi-hardware IRC log for Friday, 2011-06-24

kristianpaulhttp://cryptome.org/0004/whatsup-mesh.htm01:59
wpwrakkristianpaul: no problem. we're ready for ieee 802.15.4 ;-)02:18
qi-bot[commit] Werner Almesberger: at86rf230: assorted fixes http://qi-hw.com/p/qi-kernel/06f03a703:24
rjeffriesGNUtoo|bug20 what is the connectivity between BugBase and the add-on modules?03:46
whitequarkhm.07:43
kyakwow15:16
kyakit only took three days to deliver a parcel via Fedex15:16
kyakcan't believe it15:16
kyaki guess this is why delivery cost is bigger (50.98 EUR) than the parcel cost (50 EUR, atben+atusb)15:17
kyakunfortunately, they didn't catch me at home15:17
kyakso i'll get it on monday15:17
GNUtoovalhalla, hi18:04
valhallahi GNUtoo 19:02
GNUtooI'll PM you19:03
[g2]rjeffries ping.19:13
whitequarkjust wondering19:16
whitequark!ping19:16
[g2]hey whitequark 19:17
whitequark[g2]: hi20:23
[g2]whitequark, you were picking on me for "pinging" :)20:23
whitequark[g2]: nope, i was wondering if qi-bot will reply the command20:25
[g2]whitequark, Ah... didn't know a bot was running :)20:26
Action: whitequark doesn't actually know what "hey" can mean, depending on the context20:36
[g2]whitequark, It's like "hi"20:37
wpwrakGRRR. Maximum number of clients reachedError: Can't open display: :0.122:22
wpwraktraitor !!22:22
stefan_schmidtmuhahah22:28
stefan_schmidtRichardSharpe: hi22:36
stefan_schmidtwpwrak: hi22:36
stefan_schmidtwpwrak: lokking through your spi master driver now22:37
stefan_schmidtwpwrak: does it contain the irq handling we talked about that could also be used for atusb?22:37
wpwrakyes :)22:41
wpwrakand behold its beauty - it even allocates irq 0 on the ben :)22:42
wpwrakno idea how portable this is, though. maybe i was just incredibly lucky22:42
stefan_schmidt0 is always fun22:45
stefan_schmidton cc2420 one irq is also on 022:46
stefan_schmidtand trying to allocate the gpio freaks the gpio framework out22:46
stefan_schmidtnot sure if this is fixed by now22:46
stefan_schmidtAs the "alloc" is only for internal reference but not function I never came around to fix it :)22:47
wpwrakah ? didn't have any trouble with my irq 0. well, except in at86rf230.c22:47
stefan_schmidtwpwrak: hmm, let me check, moment22:47
wpwrakor perhaps we're talking about different frameworks :)22:47
stefan_schmidtgpio_request22:48
stefan_schmidtnot alloc22:48
stefan_schmidtnot sure its also used on mips and if it may differ there22:48
wpwraki think i found the source of my memory corruption. it's packets with a bad PHR (> 127)22:48
stefan_schmidtPHR?22:49
wpwraknow, if i throw them away and politely report back the problem, the whole thing hangs. hmm. more investigation needed. at least only the driver is dead, not the entire machine.22:49
wpwrakframe length22:49
stefan_schmidtwpwrak: (memory) It was funny to read this. I somehow implicit thought that with partial struct declaration all other would be 022:49
stefan_schmidt(frame length) ah22:50
stefan_schmidtover 127 is indeed bad in ieee802154 :D22:50
wpwrakand the 128 bytes skb hates it too22:51
stefan_schmidtwpwrak: (spi master driver) How would we integrate this with atusb? Mapping the spi calls to ATUSB_REG_WRITE, etc?22:52
stefan_schmidt(skb) And I can fully understand that :)22:52
wpwrakor ATUSB_SPI_...22:52
stefan_schmidthmm22:52
wpwraksee the dev_dbg statements in the classificator22:53
stefan_schmidtwpwrak: is there a list with available firmware commands? (I'm on latest here)22:54
wpwrakatusb/fw/include/atusb/ep0.h22:56
stefan_schmidtah, read the f*cking source, luke22:56
wpwrakaye ;-)22:56
wpwraknew error handling, new luck ... let's see how long until the first garbled packet this time22:58
stefan_schmidtwpwrak: greetings from Jan (shoragan)23:01
wpwrakthanks ! greetings, too ! :)23:02
wpwrakjapan has a funny extra wlan channel. channel number is 14, but it's 2.4 times the usual channel spacing from channel 1323:04
stefan_schmidtwpwrak: there is even more funny things about the wifi channels23:07
stefan_schmidtwpwrak: at one linux kongress, if have only been at one, they had the wifi on channel 13 IIRC23:07
stefan_schmidtand my driver did think he was in US and used the smallest amount of channels...23:07
stefan_schmidtwhich did not include 13...23:08
wpwrakcorrect. only 11 channels there :)23:09
stefan_schmidt_after_ the kongress I had a patch for it :)23:11
wpwraki wonder if i should color-code the critters for us/eu ...23:11
wpwrakcongresses are usually very fertile grounds for fixed for this sort of minor annoyances :)23:11
wpwrakmonday i should get my set of atben/atusb boards as well. finally, i'll have atusbs with a working reset line ;-)23:13
stefan_schmidtyay23:15
Action: stefan_schmidt puts in some single malt whisky and starts coding23:15
wpwrakinteresting programming fuel :)23:16
stefan_schmidtwpwrak: normally its club mate (the icetea mate)23:17
stefan_schmidtJust remembered I have some left23:17
wpwrakme, diet coke for a nice boost. else, tea. well, last night is was about a liter of coffee ... today was jittery23:18
stefan_schmidtI don't like the taste of coffee.23:19
wpwraki think that's part of the effect :)23:19
stefan_schmidtheh23:20
wpwrakwell, it was instant cappuccino. so the first half liter or so went down nicely.23:20
stefan_schmidtRichardSharpe: I would start on some work work with atusb now. based on wpwrak spi master driver.23:23
stefan_schmidtRichardSharpe: Any conflicts with this on your side?23:23
wpwrakah, i haven't mentioned the dark side yet23:24
wpwrakfor interrupt synchronization, i think we need to enqueue an URB to read, then wait for the maximum time within which it would get serviced23:25
stefan_schmidtwpwrak: you alway wait with this until I start...23:25
wpwrak(this will get more predictable when switching to an interrupt EP. for now, 10 ms seems to be a good value.)23:25
stefan_schmidtnot sure I really will hit irqs tonight23:26
wpwraknow, the interrupt synchronization would have to be triggered by detecting writes to TRX_STATE23:26
stefan_schmidtwould be happy to have some basic skeleton running for spi read and write23:26
wpwrakso the driver would have to decode some of the requests23:26
wpwrakyeah, with spi read/write, you can already make it through the chip identification23:27
wpwrakand you'll also need reset23:27
stefan_schmidtand tx should not involve any sending as well, right?23:27
wpwrakhaven't hacked this into spi_atben yet23:27
wpwraktx ?23:27
stefan_schmidttransmit23:27
wpwrakwhy would transmit not involve any sending ?23:28
wpwrakhow much single malt did you have left ? ;-)23:28
stefan_schmidthuh, you lost me23:29
stefan_schmidttransmit would not involve any irq23:29
wpwrakah ! well, there's an interrupt at the end of it23:29
wpwrakTRX_END fires when a packet has been received and when a packet has been sent23:30
stefan_schmidtyeah, sure, but I can send without caring for it now :)23:30
wpwrakyes, but at86rf230.c will not be very happy when it doesn't see the interrupt :)23:30
stefan_schmidtI'm not here to make at86rf230.c happy, hmm, atcualy I am ;)23:31
wpwrakdo not ask what at86rf230 can do for you - ask what you can do for at86rf230 ! ;-)23:31
qi-bot[commit] Stefan Schmidt: ieee802154/atusb: Update command enum and comment. http://qi-hw.com/p/qi-kernel/047c9c823:40
wpwrak2770 pings and still no bad PHR. boring. the last time, i got one after 1080 ...23:44
Action: stefan_schmidt would be happy to have 2770 pings23:45
wpwraki get a few "security support is not implemented". so there's some mild unhappiness. but nothing in the PHR yet23:46
wpwrakah, wrong. there's one. hehe. at86rf230 spi2.0: PHR 0xb2 >= buffer 128 bytes23:47
wpwrakand nothing died. good boy ;-)23:47
stefan_schmidtheh23:47
stefan_schmidthmm23:48
wpwrakah yes, of course. i turned off the serial console as much as possible. that's why such minor things didn't pop up there.23:48
stefan_schmidtwpwrak: is platform_driver an embedded only thing or can I use it on x86 as well?23:49
stefan_schmidtThere is no place for platform_data23:49
Action: stefan_schmidt is confused.23:49
stefan_schmidtTo much embedded development and to less x8623:50
stefan_schmidtHmm, on a second thought that does not need to get passed to the driver as it is wired in hw and can get set in the driver directly23:51
wpwrakshould work everywhere. the platform data should be inside spi->dev23:51
wpwrakso you can tweak it there :)23:51
wpwrakthat's what i'll do for atben_reset. so if you want to wait a little ...23:51
stefan_schmidtwait, wait, I only hear wait around me ;)23:52
stefan_schmidtsure, can wait for it. :)23:52
wpwrak(wait) first i want to see no PHR_related trouble for a bit. then i'll hack on the reset. so ~2 hours.23:53
stefan_schmidtwpwrak: heh, take your time23:55
wpwrakhmm. do arcs look good of cheesy ? cheesy, i think23:55
wpwrakhttp://downloads.qi-hardware.com/people/werner/wpan/tmp/rssi-wlan-arcs.png23:56
stefan_schmidtlooks overloaded but gets the message over23:57
stefan_schmidtFISL slides?23:57
wpwraka bit of overload is good. makes you feel it does something important :)23:58
wpwrakno no, that's atrf-rssi -g23:58
wpwrakslides are tomorrow or so23:59
qi-botThe build has FAILED, see log here: http://fidelio.qi-hardware.com/~xiangfu/compile-log/openwrt-xburst.full_system-06232011-2104/23:59
stefan_schmidtah23:59
wpwraknaw, i don't like the arcs. i'll try lines instead23:59
--- Sat Jun 25 201100:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!