#qi-hardware IRC log for Saturday, 2013-03-30

xiangfuwpwrak: they symlinked to each other.02:25
kyakwpwrak: hm, why would you need the SDK?06:21
kyaktoolchain should be sufficient, plus the STAGING_DIR export (you can point it to toolchain, it doesn't matter)06:23
kyakwpwrak: btw, are you testing atusb on 3.8?06:24
kyaklarsc: nice hint!11:48
wpwrakkyak: the SDK has some libraries I use (e.g., SDL)11:50
wpwrakand i'm testing atusb on net-next, which is ... checking ... 3.9.0-rc4-01196-ga0b8667-dirty11:51
wpwraknow let's see if i can make a 3.9 kernel for the ben ...11:54
wpwrakis gcc getting senile ? "mm/page_alloc.c:5022:34: warning: array subscript is below array bounds [-Warray-bounds]"12:10
viricwhat gcc?12:11
viricmh this damn fuloong kernel somehow doesn't boot. I've a registers dump in pmon.12:12
wpwrakmipsel-openwrt-linux-uclibc-gcc.bin (Linaro GCC 4.6-2012.02) 4.6.3 20120201 (prerelease)12:15
wpwrakusing the openwrt .config, i get a rather peculiar mix of devices. ps/2 mouse, (USB) HID, etc. didn't check the origin of these settings, though. some could have been forced by other items.12:17
wpwraksomething to clean up later ...12:17
viricmaybe some defconfig12:17
viricI use 4.6.3 here too, and I don't remember that message (building 3.8 now)12:18
viricdo mips64 kernels boot with early mips32 instrsuctions?12:20
wpwraknice typo ;-)12:22
wpwrak arch/mips/boot/compressed/vmlinux.bin.z ... gzip-compressed. very peculiar extension. wasn't .z "pack", the predecessor of "compress" ?12:26
viric.Z, no?12:27
wpwrakchecking wikipedia ... yeah. darn, that stuff was old when i started on unix12:27
wpwrakcompress was .Z12:27
viricmh I don't understand how to make this pmon output to the serial port12:28
wpwrakhmm, seems that  flash_eraseall /dev/mtd1 && nandwrite -p /dev/mtd1 uImage  no longer works :-(12:34
Action: wpwrak wonders just how much of the system he just killed12:34
wpwrakor maybe it's just the kernel that's bad12:37
viriccheck the kernel size12:45
viricand the part size12:45
larschm, isn't mtd1 the bootloader?12:46
viriclarsc: do you work with pmon?12:48
larscluckily we don't have that on jz474012:53
wpwraku-boot is still there. it got me into usb boot mode. so it must be the kernel.13:07
wpwrakare the essential drivers in mainline ?13:08
wpwrakthe owrt kernel has a collection of some 27 patches13:09
Action: wpwrak undusts his idbg-equipped ben. let's see what goes on in that kernel ...13:34
viricI'm investigating an early crash in the fuloong kernel13:37
wpwrakhmm. NAND: "2048 MiB", followed by total silence. 13:48
virica kerne line?13:56
virichere the pmon memory dump instructions only dump 0xfffffffff13:56
viricinstead of memory contents. no idea why.13:57
wpwrakno kernel. that's still u-boot13:59
wpwrakhmm, the patch set looks intimidating. i thought the ben stuff had been pushed upstream ?14:34
larscit has, what's the size of your kernel?14:39
wpwrakarch/mips/boot/vmlinux.bin.gz is 2396147 bytes14:50
wpwraka bit larger than the one on downloads.qi-hardware.com14:51
larsctry to get it to less than 2MB14:51
larsclots of kernel bload these days14:51
wpwrakunbloating ...14:51
larscI think we started out with 1MB back in the old days14:52
wpwrak"8250/16550 and compatible serial support" hmm ...14:54
larscyes, the jz4740 serial is a 1655014:55
larscwith minor changes14:55
larscdisable ext3/4 support or vfat14:55
wpwraki cleaned out a few other things. let's see how far i get ...14:58
wpwrakstill 2353883 :-(14:59
wpwrakwithout "disk" filesystems, 2141620. closer :)15:00
viricwpwrak: my troubles are gone; I had to disable FTRACE in the kernel. no more crashes.15:03
larscwpwrak: jffs2 can be disabled as well15:04
viricFTRACE may be rotten in mips15:04
wpwrakand creepy codepages15:06
viricWhy sysrq doesn't work on kernel panic?15:07
viricI always found this very annoying15:07
viricno, it works :)15:08
viricI was used to 'cu' (~# break), and in minicom the break is ctrl-a f.15:08
wpwrak1804699 bytes ;-)15:09
viricgzipped, right?15:09
viricthere should be a contest on kernel configuration, seeing who makes the smallest working kernel15:09
viricwpwrak: http://viric.name/cgi-bin/nanonixos/doc/trunk/doc/kerneldbg/index.wiki15:10
wpwrakstill no luck, though15:11
wpwraklet's try the "official" kernel ...15:12
wpwrak"OpenWrt powered". nice. then it trips over something (never gets past the logo). maybe the rootfs took a hit15:13
wpwrakblargh. no readline in usbboot. how lame is that ? :)15:20
viricrlwrap :)15:29
wpwrakoh. you learn a new thing every day ;-)15:29
viricFLOSS is a wide field15:30
viricwould the nanonote work with this? https://code.google.com/p/google-authenticator/15:35
kyakwpwrak: (atusb) very nice, hope you'd get it ro work16:00
kyakcurrently the most i get from atusb are segfaults on 3.8 :)16:01
wpwrakatusb is happy on 3.9. the user space tools too. but they appear to be broken in owrt. at least all i got from them there were complaints. actually .. lemme check ...16:09
kyakwpwrak: are you using the driver from ben-wpan branch of qi-kernel git?16:17
kyakwith the latest updates for the firmware, and it works on 3.9 - do i understand correctly?16:17
wpwrakthe driver is only loosely based on the old one. my tree is here: https://github.com/wpwrak/ben-wpan-linux16:18
kyakdo you have plans to push it to the upstream?16:19
wpwrakhmm. fragile system. reinstalled owrt. saw it boot. then installed the new kernel. complained about spi and the LCD chip. back to the owrt kernel. booted to the gui but got no usbnet. booted again, and this time it doesn't even make it to the gui :-(16:20
wpwrakwell, let's reinstall again ...16:21
kyakby reinstall do you mean reflashing?16:22
wpwrakyes. reflash_ben.sh16:22
wpwrakregarding the atusb driver, it's not really usable yet. produces a lot of debug output, some details are missing, and i also need to check for some races.16:23
wpwrakalso, i haven't tested any proper IEEE 802.15.4 communication, just IEEE 802.15.4 to atrf-txrx. so i know it can send and receive but i don't know if the data is right.16:24
wpwrakfor testing wpan to wpan, i need another station. that's why i'm now trying to get atben to run16:24
wpwrakviric: does look as if one should be able to port it16:25
kyakbut it's still something, which is good.. atben/atusb hasn't got any development for some time, so now probably soemthing with change16:25
kyaki have a router with USB port, and have big plans for that :)16:26
kyakopenwrt router, of course16:26
viricwpwrak: I guess the ios and android 'clients' are just *extras*16:28
viricnot the only ones16:28
wpwrakwolfspraul: btw, have you noticed that sharism.cc has bitten the dust once more ? (someone mentioned it a while ago but i don't know if you were around at that time)16:35
rohwpwrak: seen this? http://www.heise.de/newsticker/meldung/ZigBee-IP-beherrscht-IPv6-1832899.html16:45
wpwrakroh: yeah. and i posted a comment warning about it :)16:50
wpwraki also added a link about the non-commercial license to our ben-wpan wiki page. after all, our rants are a source referenced by wikipedia ;-)16:51
wpwraki eat my words - the wpan tools are fine in owrt. just a bit dated. (don't have "iz set")16:54
viricso you fixed the kernel things?16:54
wpwrakno. just installed the "official" distribution16:55
wpwrakat86rf230 in owrt is rather chatty. that can't be good for performance ...16:56
viricshame :)16:56
wpwrakpicking one fight at a time :)16:58
wpwraknice. the wpan stack in 3.9 just oopsed without me even helping :)16:59
wpwrakah well. time to do a bit of shopping. and then do as any good starship trooper would do ...16:59
virickill the bugs!17:03
wpwrakyeah :)17:04
viricone of the best films ever17:05
rohwell.. i got loads of openwrt routers distributed in this building.. if there would be other devices to talk to...17:05
rohanybody ever tried to send/recieve other stuff in the same band? like weather-stations and similar17:06
viricmh mpg123 doesn't build easily well on mips. It makes a very wrong sound17:09
viricfuloong, not the nanonote.18:03
viricin the nanonote, it works, cross-building18:03
larscwpwrak: 'Und heute abend im Programm bei Werner Almesberger deckt auf!, Vorsicht Falle: Zigbee'18:05
viricnice, someone updated kicad in nixpkgs and it wasn't me!19:24
wpwraklarsc: yeah. subtitle: patent trolls, proprietary standards, and phishers.21:13
wpwrakkewl. izchat works.23:33
wpwrakaddress assignment is a bit of a nightmare, though23:33
wpwrakseems that the net-next stack dropped support for non-data frames. thus, we can't exchange administrative frames and therefore association can't be implemented (though, strangely, the API still seems to be there)23:35
wpwrakfortunately, there is now a manual override with "iz set". so the atusb side on net-next is covered23:35
wpwrakbut ... on the ben, with the owrt kernel that has the old wpan stack WITH support of administrative frames, the wpan user space tools are a bit older and don't have "iz set". so the manual override isn't available there.23:36
wpwrakhowever, if firing up a izcoordinator on the ben, which is of course dysfunctional because the other side can't handle administrative frames, that izcoordinator will also set the local address, thus in the end having the same effect as "iz set"23:37
wpwraki'll now wash my hands after typing this :)23:38
--- Sun Mar 31 201300:00

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