irclog2html for #edev on 20060321

02:12.07*** join/#edev whickker (n=jamesdoe@carrell27.august.net)
02:58.13*** join/#edev MegaMahi (n=Greg@68.191.47.108)
04:18.53*** join/#edev joe_bleau (n=joe_blea@adsl-69-152-195-101.dsl.fyvlar.swbell.net)
04:39.17*** join/#edev redblue (i=star@ppp080.108-253-207.mtl.mt.videotron.ca)
04:46.46ruscalhave any of u guys seen behaviour with buildroot where even thought the dl/mtd_20050122.orig.tar.gz file exists if you do a make in the buildroot main it wont recognise the mtd_20050122.orig.tar.gz and download from ftp the same file again but call it mtd_20050122.orig.tar.gz.1?
04:47.47ruscalim using the latest snapshot of buildroot btw
04:56.42ruscalit works fine if i use the daily snapshot but if i dont then it downloads the same copy of mtd_20050122.orig.tar.gz with the above behaviour
05:11.01jbevrenyeah. happens a lot.  Sorry.  It's free software.  It's also open source, youre welcome to fix it :0
05:11.08jbevrenhehe
05:18.30ruscalyeah im trying to
05:18.39jbevrenhehe
05:18.45jbevrenthey'd love a patch if you figure it out
05:18.49jbevrenI just live with it cause i'm lazy
05:20.26ruscali think its got to do with the quotes that get stuck around "mtd_20050122.orig.tar.gz" when its put into BR2_PACKAGE_MTD_ORIG_STRING in mtd.mk  ,  i just hardcorded MTD_SOURCE to mtd_20050122.orig.tar.gz instead of $(strip $( BR2_PACKAGE_MTD_ORIG_STRING)) and its working fine
05:22.14ruscali think make is looking for a target dl/"mtd_20050122.orig.tar.gz" and not dl/mtd_20050122.orig.tar.gz and thats why its being retarded ?
05:22.59jbevrenunsure
05:23.12jbevrenmust sleep hehe
05:23.18jbevrentoo tired to figure anything out anymore
05:47.57ruscaldoes anyone know if theres a make target with buildroot just to build up the rootfs ?   im tweaking the skeleton dir and tar.gz for my needs and currently having to delete the target toolchain dir so make builds up the rootfs correctly with my changes
05:49.47MegaMahihmm, strange ive been just tweaking the build_<arch>/root/ folder and just running make again.. never had an issue
05:49.57MegaMahibut im using a older buildroot also
05:50.36ruscalyeah but you should lose those changes to /build_arch/root/ if you do a make clean
05:50.47MegaMahiah
05:51.02ruscalim making the changes to the skeleton dir and tar.gz so that if i do make clean and make again my tweaks are still there
05:51.07MegaMahiim just not doing cleans
05:52.24ruscalah
05:52.27MegaMahiya im only building to this single ppc so ive just been using the same directory havent bothered cleaning anything
05:54.28MegaMahistill having alsa issues though and its annoying the crap outta me, something about failing to set sw params, gonna try the 1.0.11rc3 packs now they seem to of patched the wordexp.h issue(never compiled uclibc with wordexp so ive been patching alsa)
06:04.31ruscalim pretty new to all this
06:04.50ruscalhaving trouble with openntp atm it compiles ok but segfaults when run on powerpc
06:05.04ruscaldunno if its something in the uClibc libraries or what
06:49.57*** join/#edev MegaMahi (n=Greg@68.191.47.108)
10:48.09*** join/#edev p-crat_ (n=p-crat@lns-bzn-45-82-65-184-186.adsl.proxad.net)
11:21.23*** join/#edev MegaMahi (n=Greg@68.191.47.108)
12:31.22*** join/#edev prpplague (n=billybob@72.22.146.238)
12:36.19*** mode/#edev [+o prpplague] by ChanServ
12:52.27*** join/#edev silverpower (n=shrike@adsl-212-228-96.cha.bellsouth.net)
13:42.23*** join/#edev GPSFan (n=Ken@12.10.255.248)
13:42.23*** mode/#edev [+o GPSFan] by ChanServ
14:45.12jbevren2gig CF at buy.com for $50 after rebate
14:45.19jbevren68.95 before
14:53.22jbevrenhttp://www.buy.com/prod/Kingston_Technology_2GB_Elite_Pro_Compact_Flash_Card_CF_2GB_S/q/loc/15579/10390647.html
14:53.25jbevrenoops wrong window heh
16:05.04MegaMahiyo GPS you around?
16:14.01sorphinhehe
16:31.20jbevrenafter dealing with me, I wonder if he answers strangers :)
16:31.23jbevrenlol j/k
16:31.53MegaMahilol
18:21.58*** join/#edev T0mW (n=Tom@24.115.193.215.res-cmts.sth.ptd.net)
18:22.22T0mWprpplague: quiet over on elinux today, what did you do / say?
18:24.45prpplagueT0mW: wasn't me
18:24.52T0mWnod
18:25.27prpplagueelinux imho has been in decline for awhile
18:26.38T0mWtypecasting + pointers will get you into so much trouble with an ARM processor!
18:27.16T0mWunaligned acesses silently load the wrong values
18:27.43T0mWI'd rather deal with a data abort than a silent failure
18:30.27jbevrenHeya tom
18:30.41T0mWjbevren: afternoon
18:30.56T0mWlife sux then you program on ARM
18:31.13jbevrenso ARM doesnt have exceptions for bus errors?
18:33.45T0mWapparently the ARM7TDMI-S cores will allow you to access a uint16_t on a odd boundry.
18:34.07T0mWno abort, but it gives an erroneous value from the fetch
18:34.16T0mWit silently does it
18:35.57T0mWjbevren: interesting problem to track down, just where you are getting an unexpected value, especially when you look at the asm code and it is "correct"
18:48.01jbevrentom: kinda dumb imho
18:48.34jbevren10+ year old cpu (m68000) for the win
18:48.40jbevrenit has alignment exceptions :)
18:55.15T0mWyeah, I had thought the ARM did also
18:55.29jbevrennod, you would -think- a modern cpu would be able to detect alignment errors
18:56.19T0mWhmm, I wonder if there is a warning switch that I have to enable in gcc? maybe gcc can warn me of this.
18:56.26T0mWor, gas
18:57.39jbevrenmight be impossible to detect..
18:57.44jbevrenyou never know the value of your reference
18:57.49jbevren(if youre casting)
18:58.03T0mWmight be in the linker then
18:58.54jbevrenmaybe
18:58.59jbevrenI dont know compiler tech that well. :)
18:59.08jbevrenI just know assemblers on a basic level
19:00.45T0mWah
19:00.52T0mW-Wcast-align
19:01.12jbevren<PROTECTED>
19:01.15T0mWdo
19:01.15jbevren:)
19:01.17T0mWdoh
19:01.27T0mWit is already set. hmmm
19:01.38jbevren:(
19:02.19T0mWunless, that option is not being propogated down into that subdir
19:03.29T0mWno, it is there, it just doesn't detect the access.
19:04.19T0mWthe var is an array of a packed struct.  But the linker puts it on an uneven boundry, then when the code runs, it loads the wrong value.
19:04.58T0mWThank God for gdb + insight + abatron bdi2000.  I could have been searching for that one for a while with printf + an LED
19:05.04T0mWheheheheh
19:05.25T0mWI glad I paid the money for that abatron
19:05.27T0mWunit
19:09.47jbevrenlol
19:09.55jbevrenwhy in the heck would the linker put a struct on an uneven boundary?
19:10.06jbevrenthat sounds like a linker bug imho
19:10.13T0mWno
19:10.29jbevrenor is the struct misaligned internally
19:10.34T0mWthe linker has to give the user some control, it cannot make all the decisions.
19:11.10T0mWif it insisted on 32bit alignment, we'd have an Intel x86 processor situation: holes all over the place
19:11.48T0mWan __attrubute__((packed),(align(4))) took care of the problem.
19:11.56jbevrencool
19:12.17*** join/#edev tux1800 (n=eric@c66.110.147-33.clta.globetrotter.net)
19:12.20T0mWyeah, it then placed it onto a 32bit (4 byte) boundry
19:18.27sorphinT0mW: hey old timer
19:48.22GPSFansorphin: hey, how's it...
19:53.57sorphinGPSFan: hey ken.. it's goin.. picked up a 3ware ide raid card last night.. just waiting for it to ship and ordered 2 of those 2G cf cards from buy.com
19:54.00sorphinGPSFan: yourself?
19:56.59GPSFansorphin: fine now that it's stopped raining/snowing for a while. trying to crosscompile some stuff with buildroot, and having libc issues. gotta learn this software crap someday ;>)
19:58.24sorphinGPSFan: indeed.. it snowed here too.. blech.. looking forward to this card so i can get my 4x300G drives doing something.... my temp controlled iron is supposed to be here today, but UPS's tracking is either wrong or they misrouted
19:59.05GPSFansorphin: what iron did (will) you get?
20:01.35sorphinGPSFan: Xytronics 137ESD
20:01.48sorphinand got a Mini-Wave tip
20:02.14sorphinsolder gets pulled up into a little cup, then you just run it over the pins, and the flux pulls it out to the pins
20:02.26sorphinfor SMD stuff
20:02.36sorphinalso picked up some ChipQuik
20:02.51sorphinso i don't have to fight w/ any more flash chips
20:07.32GPSFansorphin: nice, just googled it, not too expensive either..
20:30.57*** join/#edev JoseJX (n=JoseJX@c-67-171-66-14.hsd1.pa.comcast.net)
20:37.31jbevrenso this chipquik stuff
20:39.40jbevrenPleasant odor
20:39.46jbevrendoes it smell like the old model glue? :D
21:11.42*** join/#edev redblue (i=star@ppp216.108-253-207.mtl.mt.videotron.ca)
21:21.49*** join/#edev MegaMahi (n=Greg@68.191.47.108)
22:33.29*** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net)
22:42.03sorphinjbevren: no idea, but it melts cooler than regular solder, which is why it works
22:42.25jbevrenheh
22:42.36jbevrenold model glue has a nice citrus odor
22:45.26sorphinthis has bismuth in it
22:45.46jbevrentill beer thirty!
22:45.48jbevren20 mins!
22:45.50sorphinjbevren: uh
22:45.54sorphinno /proc/sensors
22:46.00jbevren<PROTECTED>
22:46.05jbevrenfind /proc -type d -name sensors :)
22:46.37sorphinroot@fweep:/proc# find /proc -type d -name sensors
22:46.37sorphinroot@fweep:/proc#
22:46.42jbevrenyou suck :0
22:46.43jbevren<PROTECTED>
22:46.44jbevrenj/k
22:46.50jbevrenyou have lm_sensors up?
22:46.55sorphinno, but your m.....  nevermind
22:47.03sorphini just ran sensors -f :P
22:47.05jbevrenmy mom's dead. say anything you like, I know its not true :)
22:47.20sorphinok, your sister then
22:47.20sorphinlol
22:47.29jbevrendepending on which sister, I -might- believe it
22:47.32sorphinLOL
22:47.39sorphinanyways
22:47.49jbevrencontinuing :)
22:47.51sorphinsensors -f reports back my sensor data :P
22:47.53sorphinso...
22:47.54jbevrenfind /proc |grep -i sensor :)
22:48.06sorphinnothing
22:48.13jbevrenwtf
22:48.17jbevrenusing udev or sysfs?
22:48.42sorphin<PROTECTED>
22:48.52jbevrenfind /sys |grep pwm
22:48.53jbevrenhehe
22:49.21sorphin<PROTECTED>
22:49.25jbevrenyay
22:49.25sorphin<PROTECTED>
22:49.29jbevrenthat be it
22:49.38jbevrenwhen you get home, echo numbers into em :)
22:49.39sorphinonly 2 of them
22:49.43sorphin3 fans
22:49.50sorphinguess they don't let you touch the cpu fan
22:49.52sorphin;)
22:49.59jbevrenits more likely cpu and chassis fan
22:50.33sorphinheh
22:52.29sorphinalmost wish i had more fan hookups
22:52.36sorphinsince there's one fan that's not hooked up :P
22:52.46jbevrenheh
22:52.47sorphinthis is a 2U case
22:53.17sorphinhooking 2 fans up to the same connector would confuse the chip
22:55.36sorphinanything else? :P
22:56.33jbevrenhook two fans up
22:56.36jbevrenbut only one fan's tach line
22:56.51sorphini meant anything else to check?
22:56.57jbevrenhm
22:57.03jbevrennot really, I know the rest works :)
22:57.08sorphink
22:57.18jbevrenbtw I won the board
22:57.18jbevren$25
22:57.22sorphinyeah, both ethers work
22:57.24sorphincool
22:57.52jbevrenmy niece will be thrilled to get off her 128M k6/450 lol
22:58.38sorphinya
22:59.02jbevrenI'll stick one of my celeron 1300s on the SiS board with some ram and hand it over
22:59.03jbevrenlol
22:59.13jbevrenshe'll be in high heaven
23:02.04jbevren37.80 after shipping :X
23:02.14jbevrenstill not bad.. I get to put another of my handful of S370 cpu's to work
23:14.53*** join/#edev catastrop (n=catastr0@d14-69-178-65.try.wideopenwest.com)
23:15.37sorphinyeah
23:16.07sorphinat home now btw
23:18.18sorphinok
23:18.22sorphinpwm2 is cpufan
23:18.28sorphinpwm1 is the other 2 fans
23:58.19*** join/#edev p-crat (n=p-crat@lns-bzn-45-82-65-184-186.adsl.proxad.net)

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.