IRC log for #edev on 20070914

00:09.08*** join/#edev Off_Namuh104 (i=GPS@gateway/tor/x-a173a29feb4728f2)
00:18.33*** join/#edev chouimat (n=dieu@kde/developer/chouinard)
01:47.10*** join/#edev MothaWork_ (n=monmotha@host-69-95-169-16.cwon.choiceone.net)
01:47.41*** join/#edev MothaWork_ (n=monmotha@149.164.193.61)
02:02.23*** join/#edev redblue (i=star@ppp105.108-253-207.mtl.mt.videotron.ca)
02:34.39*** join/#edev FireEgl (i=FireEgl@Sebastian.Tcldrop.US)
03:06.34jbevrenhttp://www.surpluscomputers.com/store/main.aspx?p=ItemDetail&item=IMG10090
03:06.37jbevrenpalmpix $0.99
03:10.25agaffneywow
03:10.37agaffneyof course, the only palm device I have is a treo 650, which already has a camera
03:14.09ds2they still have them?!
03:14.28ds2they must be desperate to clear out their wear house
03:14.43ds2warehouse
03:28.21*** join/#edev joe_bleau (n=joe_blea@adsl-69-152-203-191.dsl.fyvlar.swbell.net)
03:31.56agaffneythey'd be better off just tossing them in the dumpster at that price
03:36.18ds2not in CA
03:36.32ds2they probally get charged a recycling disposal fee of $2 each ;)
03:36.39agaffneyhah
03:37.24ds2if they didn't close their store, I might go pick up a few but they did so too bad for them
03:38.46*** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1)
03:56.00*** join/#edev linac (n=lin@167.86.76.222.broad.fz.fj.dynamic.163data.com.cn)
04:23.29*** join/#edev redblue (i=star@ppp146.108-253-207.mtl.mt.videotron.ca)
06:27.05*** join/#edev ohsix (i=ohsix@bc111222.bendcable.com)
06:47.35*** join/#edev steliosk (n=Stelios@ipa226.211.tellas.gr)
06:55.19*** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de)
06:59.28*** join/#edev emil|k (n=emil|k@e179235211.adsl.alicedsl.de)
07:03.25*** join/#edev howtoo (n=howtoo@oslo-static64.bluecom.no)
10:29.24*** join/#edev redblue (i=star@ppp051.108-253-207.mtl.mt.videotron.ca)
10:51.07*** join/#edev emil|k_ (n=emil|k@e179239211.adsl.alicedsl.de)
10:57.27*** join/#edev lyakh_ (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de)
12:05.54*** join/#edev linac (n=lin@121.41.69.27)
12:33.04*** join/#edev Off_Namuh104 (i=GPS@gateway/tor/x-2c0bf158f8f05635)
12:58.58*** join/#edev linac (n=lin@121.41.79.209)
13:13.37*** join/#edev GPSFan (n=kenm@12.10.255.248)
13:13.37*** mode/#edev [+o GPSFan] by ChanServ
13:28.41*** join/#edev prpplague (n=DaveAnde@mail.americanmicrosystems.com)
13:29.08prpplagueho ho ho
13:29.10prpplaguemerry freakin friday
13:33.59*** mode/#edev [+o prpplague] by ChanServ
13:47.28*** join/#edev ico2 (n=ico2@modem-2104.cougar.dialup.pol.co.uk)
14:00.49*** join/#edev Crofton (n=balister@hc65217ba.dhcp.vt.edu)
14:12.15GPSFanprpplague: g'day. the hammer buildroot doesn't build g++ and thus programs that use it like iperf fail to compile. I tried enabling c++ cross compile support in buildroot, but for some reason it wouldn't build the g++ cross compiler.
14:14.56prpplagueGPSFan: yea i rarely use the c++ so i didn't have it enabled
14:15.02prpplagueGPSFan: i'll go through and test
14:16.30GPSFanprpplague: it maybe that I have to rebuild the whole toolchain from scratch. IIRC there was some issue with tryong to build a new part of the toolchain after the initial build was done.
14:17.18prpplagueGPSFan: that is possible\
14:18.02GPSFanI'll be around later, we've got about 50 tons of hay arriving today. bbl...
14:18.27prpplaguehehe
14:42.36*** join/#edev brad_mssw (n=brad@shop.monetra.com)
15:01.10*** join/#edev CosmicPenguin (n=nobody@crank.laptop.org)
15:09.57*** join/#edev prpplague (n=DaveAnde@mail.americanmicrosystems.com)
15:10.24*** mode/#edev [+o prpplague] by ChanServ
15:33.15chouimatmorning
15:36.10*** join/#edev ico2 (n=ico2@modem-2104.cougar.dialup.pol.co.uk)
15:45.51prpplaguechouimat: morning
15:45.54prpplagueCosmicPenguin: ping
15:46.39CosmicPenguinpong
15:47.13prpplagueCosmicPenguin: hey question
15:47.19CosmicPenguinshoot
15:47.31prpplagueCosmicPenguin: i've never seen this as an initial value in c/c++
15:47.46prpplagueCosmicPenguin: unsigned cs_change:1
15:47.59prpplagueCosmicPenguin: thats in a struct definition
15:48.15prpplagueCosmicPenguin: does that mean the cs_change always gets inited to 1 ?
15:48.25CosmicPenguinno, its a bitwise definition
15:49.56prpplagueCosmicPenguin: i guess i'm not familiar with that type of definition statement
15:50.35prpplagueCosmicPenguin: how does that work? url? doc?
15:50.46CosmicPenguinlet me find it in the spec
15:51.23prpplague(i mean i understand what bitwise is, but not the definition of the variable in this instance)
15:55.22CosmicPenguinI can't find it in the spec
15:55.27CosmicPenguinso I'll try to explain it myself
15:55.44CosmicPenguinbascially, its a a way to describe a bitmask in a struct
15:57.22CosmicPenguinso instead of reading a 32 bit register (for example), and doing something like this
15:57.34CosmicPenguinif ((reg & 0xF0) >> 8) == 3)
15:57.44CosmicPenguinyou could define a struct like this
15:57.46CosmicPenguinstruct {
15:57.59CosmicPenguinunsigned fielda:8;
15:58.08CosmicPenguinunsigned fieldb:8
15:58.08CosmicPenguin}
15:58.37CosmicPenguincast the register value into that, and then read struct->fieldb
15:59.05prpplagueahh ok
15:59.06prpplaguegotcha
15:59.11prpplagueCosmicPenguin: thanks
15:59.35CosmicPenguinMost people use it to cast really complex structur4es
15:59.46CosmicPenguinlike network packages
15:59.49CosmicPenguinpackets even
15:59.57CosmicPenguinto me, they're less interesting because bitmasks are so useful
16:00.09CosmicPenguinshifting and comparing uses less compiler magic
16:00.11prpplagueindeed, its the first time i've seen that before
16:00.13CosmicPenguinand compiler magic is bad
16:03.54*** join/#edev ico2 (n=ico2@modem-2104.cougar.dialup.pol.co.uk)
16:04.21prpplagueCosmicPenguin: indeed
16:14.49*** join/#edev ico2 (n=ico2@modem-2104.cougar.dialup.pol.co.uk)
17:20.44*** join/#edev chouimat (n=dieu@kde/developer/chouinard)
17:34.19*** join/#edev chouimat (n=dieu@kde/developer/chouinard)
17:55.19MothaWorkprpplague: I actually have the entire header for a network protocol we have here set up as a C struct.  Just typecast the buffer containing the packet to the struct and start picking up elements of the header.  VERY handy
17:57.18prpplagueMothaWork: interesting
18:03.34*** join/#edev Off_Namuh33 (i=GPS@gateway/tor/x-446aae6a61172418)
18:06.18GPSFanprpplague: I had to start from scratch and build the toolchain again with gcc cross compiler enabled. When I did I got an error that buildroot/project_build_arm/Hammer/root/usr/lib was missing. I mkdir'd that directory and the toolchain finished ok. IIRC there was an issue in buildroot regarding this a while ago. The fix is probably in the latest svn, but there maybe other issues.
18:06.48prpplagueGPSFan: ahh
18:08.17GPSFaniperf built as well, I'll be testing somewhat later this afternoon. Most of the hay is in the barn... ;>)
18:09.44prpplagueGPSFan: good deal, i was curious about the performance
18:09.54prpplagueGPSFan: there are several things that need to changed in the driver
18:10.11prpplagueGPSFan: the interrupt handler needs to be changed so that it uses a task instead
18:10.22prpplagueGPSFan: there is too much going on in that ISR
18:10.29GPSFanprpplague: any chance of a new apex?
18:10.41prpplagueGPSFan: i haven't had time to post it
18:10.54prpplagueGPSFan: but like i said yesterday, you can just change the option in your existing config
18:11.13GPSFancan apex be flashed from apex, or does one need jtag?
18:11.15prpplagueGPSFan: make menuconfig -> environment -> kernel boot
18:11.22prpplagueGPSFan: you can do it from apex
18:11.29prpplagueGPSFan: xreceive 0x30008000
18:11.34prpplagueGPSFan: same location as the kernel
18:11.34GPSFanok, I'll give it a try.
18:11.45prpplagueGPSFan: then just do: erase nor:0x00+128k
18:11.57prpplagueGPSFan: then do: copy 0x30008000+128k nor:0x00
18:12.07prpplagueGPSFan: reboot and you should be good
18:12.18prpplagueGPSFan: if you screw up, you'll have to get jtag working of course
18:12.25GPSFan;>)
18:13.28GPSFanmake menuconfig in what? buildroot, the kernel?
18:14.16prpplagueapex
18:14.36GPSFanah...
18:16.53GPSFanso the current option is 256K+1536k  that should be changed to 256k+1024K ?
18:18.16prpplagueGPSFan: the current option should be 256k+768k for the kernel and 1024k+1024k for the ramdisk
18:18.32prpplagueGPSFan: you should have copied the hammer-config to .config
18:18.45prpplagueGPSFan: or jus set the machine type to hammer
18:18.45GPSFanprpplague:  that's probably it...
18:19.30prpplagueGPSFan: change the settings to 256k+1024k for the kernel and 1280k+1024k for the ramdisk
18:19.34GPSFanthere is no hammer-config in the apex-1.5.6 dir
18:19.43prpplagueodd should have been one
18:19.59prpplagueGPSFan: ok , just set the machine type to hammer, it will set all the defaults
18:20.08GPSFannot on the cd.
18:20.29prpplagueGPSFan: yea, must have missed it on the cvs export
18:20.38prpplagueGPSFan: so just do a make menuconfig in apex
18:20.48prpplagueGPSFan: and set the machine type to hammer
18:21.30GPSFanwhere is that in the menuconfig?
18:22.24prpplagueGPSFan: armplatform first
18:22.28prpplagueGPSFan: set it for s3c2410
18:22.49GPSFanah found it . you have to select the sharp s3c2410 first
18:22.57prpplagueGPSFan: righto
18:23.28prpplagueGPSFan: under general setup you can set the name of your cross tool, i.e. arm-linux-uclibc-
18:23.31GPSFanok now there are 2 regions in the environment
18:23.43GPSFanbueno..
18:24.03prpplagueyea, one for kernel and one for ramdisk
18:25.48GPSFanok kernel is changed to 256k+1024k and ramdisk to 1280k+1024k
18:26.02prpplagueokie dokie
18:26.09prpplagueshould be able to save it and do a make
18:26.25prpplaguethe bootloader will be in src/arch-arm/rom/apex.bin
18:26.47GPSFanafter changing the compiler path in the Makefile.
18:27.08prpplagueGPSFan: or just add the path to your $PATH
18:27.32prpplagueGPSFan: export PATH="$PATH:/home/gpsfan/hammer/buildroot/build_arm/staging_dir/bin"
18:27.58GPSFanI'll change the Makefile.
18:28.01prpplaguehehe
18:29.11GPSFanapex should be a buildroot package then it would be seemless
18:29.41prpplagueyep
18:29.49prpplaguejust need to add it all together
18:30.20prpplaguejust been too busy
18:35.25GPSFanthe binary is slightly different in size is that significant?
18:36.12GPSFan37480 vs the cd version of 36904
18:37.02prpplagueGPSFan: thats probably just text differences
18:37.13prpplagueshould be fine
18:37.53*** join/#edev chouimat|ibook (n=dieu@kde/developer/chouinard)
18:39.07GPSFanok, I'll flash it after lunch..
18:41.17prpplaguedamit do it NOW
18:41.21prpplague!!!!!!!!!!!!!!!!
19:07.24*** join/#edev Off_Namuh33 (i=GPS@gateway/tor/x-71c72092a05b9fcb)
19:24.42*** mode/#edev [+o CosmicPenguin] by ChanServ
19:56.18GPSFanprpplague: apex> xr 0x30008000
19:56.19GPSFanC
19:56.19GPSFanapex> erase nor:0x00+128k
19:56.19GPSFanapex> copy 0x30008000+128k nor:0x00
19:56.19GPSFan131072 bytes transferred
19:56.19GPSFanapex>
19:56.28GPSFanresult: brick. ;>(
19:58.40*** join/#edev redblue (i=star@ppp038.108-253-207.mtl.mt.videotron.ca)
20:02.05prpplagueGPSFan: doh
20:02.18prpplagueGPSFan: sounds like something didn't build right then
20:02.49prpplagueGPSFan: no console now?
20:03.32prpplagueGPSFan: you'll have to get your flyswatter working then
20:04.10GPSFannope nothing. I'm working on the jtag dongle. openocd wants a lib that looks it should be a symlink from what you provided.
20:04.37prpplagueGPSFan: yea, you'll need to moves the libs into the appropriate dir and do symlinks
20:05.03GPSFanthat's what I figured, just doing it now..
20:07.15GPSFanprpplague: which is the config file, .ocd or .cfg ?
20:08.45prpplagueGPSFan: cfg
20:09.30prpplagueGPSFan: if execute in the same dir with the .cfg and .ocd it should init and copy the apex.bin to flash
20:10.22GPSFanwhich way does the jtag conn go on?
20:11.08prpplagueif you look closely on the cpu board you can see one corner is marked as pin 1 with it notched
20:11.37prpplagueGPSFan: pin1 of the jtag os on the side next to VD0
20:12.10GPSFanah, I see it
20:12.45GPSFanpower on to the hammer before powering up the jtag?
20:13.39prpplagueGPSFan: usually you'll power on the jtag first, power the board, then start up openocd
20:13.59GPSFanok.
20:16.05*** join/#edev chouimat (n=dieu@kde/developer/chouinard)
20:16.10GPSFanprpplague:  after doing ./openocd -f hammer.cfg I get:   Error:   ft2232.c:967 ft2232_init(): No matching layout found for flyswatter
20:22.32GPSFanit looks like the ftdi_sio driver grabbed the interface.
20:23.16prpplagueGPSFan: yea, you have to make sure that ftdi_sio doesn't load when you connect up the
20:23.28GPSFanprpplague: this is Ubuntu 6.06 LTS
20:24.22GPSFanhow do I do that?
20:29.35cbrakeSCO files CH11: http://www.prnewswire.com/cgi-bin/stories.pl?ACCT=104&STORY=/www/story/09-14-2007/0004663029&EDATE=
20:31.37Croftonhttp://www.reuters.com/article/marketsNews/idUSBOM16661920070914
20:31.39Croftonarg
20:32.41GPSFannot totally unexpected, trial starts (would have started) monday.
20:33.29Croftonshould still start
20:33.33GPSFanone would hopw so, but I wouldn't bet on it.
20:37.56MothaWorkhey, anyone with some actual experience...
20:38.29MothaWorkwhich is likely to last longer?  Aluminum electrolytic running at a full 2x the voltage rating, or a tantalum running at about 1.5x the rating?
20:38.40MothaWorkthis is the output cap on a SMPS
20:42.20MothaWorkerm
20:42.26MothaWorkrunning at 1/2 and 75% of
20:42.39MothaWorki.e. overspec'd 2x and 1.5x, respectively
20:44.42prpplaguewahooo
20:44.59prpplagueGPSFan: you should have a blacklist file somewhere
20:45.04prpplagueGPSFan: for the hotplug
20:45.16GPSFanin userspace?
20:45.18prpplagueGPSFan: or you can plug in the device, then rmmod the ftdi_sio
20:45.30prpplagueGPSFan: yea, probably something like /etc/hotplug
20:46.38GPSFannothing there
20:46.59GPSFanI'll try rmmoding as root.
20:48.45GPSFansame thing.
20:48.54prpplagueGPSFan: ?
20:49.03prpplagueGPSFan: when you rmmod ftdi_sio what do you get?
20:49.55GPSFanrmmoding succeededs, and it's no longer in the lsmod list, running openocd restlts in the same error message as before.
20:50.24prpplagueGPSFan: ok, what options did you use to configure openocd?
20:50.34prpplagueGPSFan: or are you using the binary included?
20:50.45GPSFanusing the included binary
20:51.02prpplagueGPSFan: hmm exiles report problems with the included binary as well
20:51.16prpplagueGPSFan: you'll probably need to compile it from the source included
20:51.30GPSFan<PROTECTED>
20:51.39prpplagueone sec
20:51.43prpplaguegrabing notes
20:52.37GPSFanwell, there is no makefile, and no configure.
20:53.19GPSFanmust use automake or some such garbage.
20:53.33prpplagueGPSFan: yea, you have to run the automake stuff
20:53.49prpplagueGPSFan: ./bootstrap
20:53.55prpplagueGPSFan: should generate the configure
20:54.18GPSFanclear as day..
20:54.30prpplaguehehe
20:54.48prpplagueGPSFan: i can see that i'll need to add openocd to buildroot as well
20:55.00prpplague~quote scox
20:57.16prpplagueExiles: ping
20:58.04GPSFanprpplague: http://www.pastebin.ca/697948 using the newly compiled openocd
20:59.27prpplagueGPSFan: ok, you did the bootstrap, what options did you pass to the configure?
20:59.58GPSFannone, that's what I was asking for hints about.
21:00.12prpplague./configure  --enable-ft2232-ftd2xx
21:00.20prpplagueGPSFan: sorry doing to many things at once
21:00.23prpplague~lart prpplague
21:00.23purlmakes a balloon animal out of prpplague
21:00.25GPSFannp
21:01.19GPSFanft2232.c:45:20: error: ftd2xx.h: No such file or directory
21:03.07prpplagueyou have got to be kidding
21:03.23prpplagueyou need to extract the libftdxx tarball
21:03.23*** join/#edev Off_Namuh33 (i=GPS@gateway/tor/x-ae5f02e9fd599321)
21:03.52prpplagueGPSFan: i can see i need to make some changes to the software layout to make it easier
21:04.12prpplagueGPSFan: the libftd2xx tarball will have the .h files you need
21:04.18prpplagueGPSFan: just copy them over into that
21:09.59GPSFanprpplague: not sure where to put the .h file, the readme in the tarball just talks about wher th put the lib (usr/local/lib) and how to make the appropriate symlinks.
21:11.23prpplagueGPSFan: its probably looking for it in the local dir, one set i'll tell ya
21:12.21prpplagueGPSFan: you should be able to put it in the src/jtag dir
21:12.41GPSFank, what I'm trying to do is do all this stuff blindly, as if a novice user was trying it without any knowledge.
21:13.16prpplagueGPSFan: indeed, i'm going to do a fresh build this weekend
21:13.34prpplagueGPSFan: you will need to copy the WinTypes.h file too
21:14.26GPSFanyep.
21:15.07prpplagueGPSFan: exactly why i wanted to do a beta run with the eval kit
21:15.13prpplagueGPSFan: find out the good, bad and  ugly
21:15.55GPSFanError:   ft2232.c:1341 ft2232_init_ftd2xx(): unable to open ftdi device: 18
21:15.55GPSFaninitialise_usb failed
21:16.26GPSFanprpplague: well, I hope I'm helping with it all. ;>)
21:16.39prpplagueGPSFan: indeed
21:16.59prpplagueGPSFan: do this: cat /proc/bus/usb/devices | grep -e Product
21:17.36GPSFanah, no usbdevfs
21:17.46prpplagueahh
21:17.53prpplagueshould be able to mount that
21:20.25GPSFanyeah, it's this version of ubuntu doesn't have it mounted by default, I have to do it every time I play with usb stuff.
21:20.36prpplagueGPSFan: fun fun
21:23.08GPSFanwell it's listed in the proc entry, but I still get Error:   ft2232.c:1341 ft2232_init_ftd2xx(): unable to open ftdi device: 2
21:23.35GPSFansilverpower:  Product=Flyswatter
21:23.41prpplagueGPSFan: ok one sec
21:24.57prpplaguehmm
21:25.14prpplagueGPSFan: do a ldd on your openocd binary
21:25.35*** join/#edev leoncamel (n=leoncame@p6024-ipbfp305fukuokachu.fukuoka.ocn.ne.jp)
21:26.22GPSFanhttp://www.pastebin.ca/697988
21:26.56prpplaguehmm ok so all the libs are there
21:27.15GPSFanyeah, the libs came from the CD
21:27.39prpplaguehmm that should be correct
21:28.51GPSFanprpplague: did Exiles get the flyswatter working?
21:28.54prpplagueGPSFan: yea
21:29.07prpplagueGPSFan: sounds like similiar issues
21:29.24prpplagueGPSFan: sounds like the product string isn't an exact match
21:29.46GPSFanI gotta run out for a while. bbl
21:29.55prpplagueGPSFan: okie dokie
22:46.52*** join/#edev chouimat (n=dieu@kde/developer/chouinard)
23:54.38*** join/#edev joe_bleau (n=joe_blea@adsl-76-213-136-138.dsl.fyvlar.sbcglobal.net)
23:56.57*** join/#edev Off_Namuh33 (i=GPS@gateway/tor/x-092f9d0fe744db0d)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.