| 00:01.16 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 00:03.27 | *** join/#edev redblue (i=star@ppp130.108-253-207.mtl.mt.videotron.ca) |
| 00:05.26 | ds2 | gpio button fixed |
| 00:20.28 | *** join/#edev redb1ue (i=star@ppp035.108-253-207.mtl.mt.videotron.ca) |
| 00:25.18 | *** join/#edev davidc__ (n=davidc__@S01060002b360aacd.vc.shawcable.net) |
| 00:34.30 | *** join/#edev redblue (i=star@ppp035.108-253-207.mtl.mt.videotron.ca) |
| 01:29.53 | ds2 | solar: I have something that seems to work...a bit large but it is not segfaulting |
| 01:30.09 | ds2 | they are all static binaries to avoid the library issue |
| 01:33.57 | solar | sometimes that just the easisest route |
| 01:34.17 | ds2 | *nod* |
| 01:34.41 | solar | this will bug me later however so I'm sure I'll figure out how to make it submit to our wishes |
| 01:35.47 | ds2 | I know it cross compiles cuz the gumstix folks have it working and I did some hacking on it at MV |
| 01:36.11 | ds2 | FWIW, I am building the binaries with a cross tools generated compiler |
| 01:36.21 | solar | ya I think for the 3.x version I just need to tidy up some rpaths |
| 01:37.06 | ds2 | but that pulls in too much stuff |
| 01:39.18 | solar | I'll have to eval them all. |
| 01:39.27 | solar | anyway is that pppd bin working for you? |
| 01:39.40 | ds2 | yes, that seems to work fine |
| 01:39.53 | ds2 | ran it and it started spewing LCP requests as expected |
| 01:50.30 | solar | one down. bluez-utils-2.25 compiled for eabi |
| 01:51.53 | ds2 | Hmmmmmm I do not like what that suggests |
| 01:52.59 | solar | it's kinda good news. It means I just need to make a gcc-4.x cross compiler for uC |
| 01:53.10 | solar | thinks |
| 01:53.20 | ds2 | really? thought that suggets there is an OABI problem with userapps |
| 01:54.57 | ds2 | ack... these static binaries are doing weird things |
| 01:55.44 | solar | building gcc now |
| 01:56.11 | ds2 | scan is returning bogus BT addresses but seems to do the right things when giventhe right address |
| 02:04.37 | ds2 | before I go debug this weirdness, I think I'll wait and see if your new gcc works. got other stuff that needs to finish :/ |
| 02:05.24 | solar | my gcc failed one time already so I'm fixing that up and trying again. Might be 20mins or so. |
| 02:06.41 | ds2 | take your time, my list is long and I won't be blocked on this for a while; the original hurry was to try to verify the hardware. |
| 02:08.14 | solar | yeah it's going to take a bit. Might be a uclibc problem so I'll have to check what OE,BR has. http://rafb.net/p/luMsdI80.html |
| 02:08.58 | solar | guesses it needs to use some weak aliasing |
| 02:09.14 | ds2 | the mysterious libgcc :( |
| 04:16.50 | *** part/#edev pea (n=pea@75.58.4.163) |
| 04:34.22 | *** join/#edev mbuf (n=user@61.16.248.242) |
| 05:38.54 | *** join/#edev DrZimmerman (n=theo@77-56-242-58.dclient.hispeed.ch) |
| 05:39.25 | *** join/#edev roxfan2 (n=dunno@191.53-240-81.adsl-dyn.isp.belgacom.be) |
| 06:48.39 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 07:20.40 | *** join/#edev dijenerate (n=dijenera@65.48.164.209) |
| 07:47.53 | *** join/#edev chouimat (n=dieu@r2351064.cidc.net) |
| 08:01.40 | *** join/#edev redblue (i=star@ppp018.108-253-207.mtl.mt.videotron.ca) |
| 09:57.31 | *** join/#edev likewise (n=likewise@82-171-51-231.ip.telfort.nl) |
| 10:24.13 | *** join/#edev mbuf (n=user@61.16.248.242) |
| 10:34.26 | *** join/#edev g1powermac_PB (n=g1powerm@unaffiliated/g1powermac) |
| 10:54.29 | *** join/#edev g1powermac (n=g1powerm@unaffiliated/g1powermac) |
| 11:18.08 | *** join/#edev roxfan2 (n=dunno@137.13-200-80.adsl-dyn.isp.belgacom.be) |
| 11:43.10 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 11:52.33 | *** join/#edev lyakh (n=lyakh@p57BD3A69.dip0.t-ipconnect.de) |
| 15:31.11 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 15:32.20 | *** join/#edev linac (n=lin@121.41.51.247) |
| 15:35.04 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 15:43.49 | *** join/#edev magnetron (n=magnetro@unaffiliated/magnetron) |
| 17:00.15 | *** join/#edev g1powermac (n=g1powerm@76.177.147.36) |
| 17:24.49 | *** join/#edev pea (n=pea@75.58.4.163) |
| 17:30.56 | ds2 | morning |
| 17:31.41 | *** join/#edev Ei5GTB_ (n=Paul@78.16.228.160) |
| 17:33.51 | solar | morn |
| 17:35.02 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 17:38.41 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 17:38.42 | ds2 | any luck |
| 17:38.44 | ds2 | ? |
| 17:40.03 | solar | ds2: nope. I had to start working on filtering a gallon of vodka for a friend last night. |
| 17:40.13 | solar | ~5 hr process |
| 17:40.39 | solar | but I did finish uploading all the eabi stuff. |
| 17:40.46 | solar | which contains all the stuff you needed. |
| 17:41.11 | solar | but glibc. So all those stupid locales will slip in there |
| 17:41.22 | ds2 | question now is can I fit it on the hammer |
| 17:41.27 | solar | yep |
| 17:41.48 | solar | 16M/32M onboard or so right? |
| 17:41.59 | ds2 | think so |
| 17:42.18 | ds2 | I am guessing root will have to be flash then |
| 17:42.33 | solar | so using jffs2 + X matchbox and all that jazz fits into about 25M for a full cellphone pda type of setup |
| 17:42.49 | solar | I've got a bunch of scripts handy for that if you want me to pass them on |
| 17:42.55 | ds2 | well, 16M I think is flash |
| 17:43.16 | ds2 | what's the script do? take a list of packages and spit out a jffs2 image suitable for flashing? |
| 17:44.33 | solar | take the .tbz2 dirs. install them into what we call a $ROOT then does all the staging cleanup. Overlays custom files. makes the jffs2 apex.bin and kernel. Then another script for easy uploading |
| 17:44.52 | solar | got it mostly tied to a single Makefile |
| 17:45.09 | ds2 | that sounds helpful... is it python? |
| 17:45.31 | solar | depends on gentoo's emerge(python util) for the installing staging of the ROOT dir. |
| 17:45.49 | ds2 | oh... that is not going to be useful for me |
| 17:45.55 | ds2 | no working python here |
| 17:46.26 | solar | alternative is pass me a binary copy of your kernel and desired pkglist |
| 17:46.37 | ds2 | I can do that |
| 17:46.44 | solar | apex.bin should be mostly the same |
| 17:46.49 | ds2 | not sure why you need akernel to generate a fs, but I can do that =) |
| 17:47.01 | solar | well I don't actually good point |
| 17:47.18 | ds2 | package list is the files I wnat in http://tinderbox.dev.gentoo.org/embedded/tct/gnueabi/All/? |
| 17:48.34 | solar | yep. So for the 8050 I'm slapping together a subset of those pkgs. http://rafb.net/p/IqTnV081.html |
| 17:48.55 | ds2 | I don't need to figure out dependancies, do I? |
| 17:49.27 | solar | it would help. But being I've got the tools over here. It's probably eaiser for me to figure them mostly out |
| 17:49.40 | solar | I wrote a handy little cross-elf-scannging tool |
| 17:49.40 | ds2 | okay, let me generate the list then |
| 17:50.03 | solar | so if some DT_NEEDED is missing. I'll say whats missing |
| 17:50.52 | ds2 | uh...bluez-utils-2.25 but bluez-libs-3.36? didn't know that is allowed |
| 17:52.43 | solar | sec let me build the older bluez-libs |
| 17:55.56 | ds2 | do I need hotplug or will this setup default to a sane static /dev? |
| 18:07.19 | ds2 | I think this should be bootable and have what I need - |
| 18:07.26 | ds2 | http://rafb.net/p/Kka9Nc76.html |
| 18:10.24 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 18:13.50 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 18:14.06 | *** join/#edev JoeLlama (n=snork@unaffiliated/joellama) |
| 18:14.59 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 18:16.33 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 18:17.41 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 18:21.30 | solar | ds2: hrmm my config files sets up mdev. Assuming you are a busybox user. |
| 18:22.36 | solar | ds2: that list might need some love. Mind if I rework it a bit? Cuz out of some fluff? |
| 18:22.45 | solar | s/Cuz/Cut |
| 18:23.20 | solar | well no. First lets do it your way |
| 18:24.58 | ds2 | feel free to strip it down |
| 18:25.37 | ds2 | the most obvious being e2fs stuff... that's only there for the plan Z emergency where I have dig out a hub + flash drive |
| 18:32.15 | *** join/#edev DrZimmerman (n=theo@77-56-242-58.dclient.hispeed.ch) |
| 18:32.17 | *** join/#edev kergoth (n=kergoth@ip72-201-11-30.ph.ph.cox.net) |
| 18:42.41 | solar | note. One of the gsm pkgs is for Lossy speech compression library and tool. |
| 18:42.49 | solar | while the other is openmoko's gsm. |
| 18:42.58 | solar | did you really intend for both? |
| 18:49.43 | solar | ds2: ok with all the pkgs you requested and not following any deps. It's -rw-r--r-- 1 root root 13M Sep 28 11:49 rootfs.jffs2 |
| 18:50.46 | solar | basic cleanup of *.a *.o *.la *.pl |
| 18:51.08 | ds2 | whoa |
| 18:51.09 | solar | personal experience says if you are willing to use the busybox tools. that could be close to 4-6M |
| 18:51.27 | ds2 | bah... I don't need GSM then.... thought it was for AT modem support |
| 18:51.52 | ds2 | busybox is fine... wasn't sure if shtools, etc really meant the actual gnu packages or if they were the set of symlinks |
| 18:53.19 | solar | so. for iproute2 are you after the 'ip' command or 'tc' for QOS stuff? If not QOS then bb tends to be fine |
| 18:53.39 | ds2 | ip command |
| 18:53.48 | ds2 | and that can go if there is no space either |
| 18:54.02 | ds2 | BB is indeed fine |
| 18:54.47 | *** join/#edev likewise (n=likewise@82-171-51-231.ip.telfort.nl) |
| 18:55.27 | solar | ok test #2 |
| 18:56.17 | solar | - 5.4M Sep 28 11:55 rootfs.jffs2 |
| 18:56.28 | ds2 | much more tolerable |
| 18:56.49 | solar | http://rafb.net/p/F4Kbrm64.html |
| 18:57.30 | solar | =evieext-1.0.2 says it's an xproto. Wonder what that is needed for. |
| 18:57.33 | ds2 | evieext should be a better vi then the busybox vi right? |
| 18:57.55 | solar | x11-proto/evieext X.Org EvIE protocol headers |
| 18:57.55 | ds2 | never heard of it but neither vim, elvis, nor nvi is on the list |
| 18:58.02 | ds2 | oh that can go then |
| 18:58.25 | solar | oh woops we did not need procps either |
| 18:58.38 | ds2 | ah right |
| 18:59.03 | ds2 | sash is optional but seems like there is some room for it |
| 18:59.17 | solar | yeah. But the default shell with be busybox (ash) |
| 18:59.25 | solar | so seems a lot of duped functionality |
| 18:59.40 | solar | sash is kind nice however. thats the static one right? |
| 18:59.44 | ds2 | true, chuck it... if I loose /lib/ld.so I am hosed for other reasons |
| 18:59.54 | ds2 | yeah... sash is good for recovering from list /lib |
| 19:00.50 | solar | almost. Real pita when the libc is broke. |
| 19:01.11 | solar | at gentoo we moved away from sash and opted to install a busybox.dynamic and a fallback busybox.static |
| 19:01.33 | solar | which is named /bin/bb. So init=/bin/bb ; and we get all those wonderful rescue utils |
| 19:01.35 | *** join/#edev redblue (i=star@ppp099.109-253-207.mtl.mt.videotron.ca) |
| 19:01.54 | solar | tar/bzip2 that one often needs to preform real repair work |
| 19:02.07 | ds2 | so /bin/bb knows to start a shell? the thing that worries me is not being able to control arg0 |
| 19:02.41 | ds2 | in reality if I am sitting at a corner 10 mins before the talk fixing things that need sash functionality, I am AFU |
| 19:03.13 | solar | 4.8M |
| 19:03.23 | ds2 | usable |
| 19:03.45 | solar | ok gimme a few mins to now cross check all the link libs to make sure we did not forget anything |
| 19:03.51 | ds2 | how configureable is the busybox syslogd? |
| 19:03.59 | ds2 | never used that one |
| 19:04.11 | ds2 | I need to be able to have it dump pppd.debug to a file |
| 19:07.07 | solar | haha doh. This for sure would of not worked. Something about missing a libc :) |
| 19:10.07 | *** join/#edev cortez (n=egoe@demolitionman.xs4all.nl) |
| 19:11.43 | solar | few things to track down. But here is what it's looking like. http://rafb.net/p/K0a3YK95.html |
| 19:11.46 | ds2 | hahahha |
| 19:13.51 | solar | ok here is what's missing. I'll track down the providers libbluetooth.so.2 libgcc_s.so.1 liblzo2.so.2 libusb-0.1.so.4 libz.so.1 |
| 19:13.52 | *** join/#edev mpk (n=mpk@aggr.com) |
| 19:14.13 | ds2 | bluetooth should be bluez-libs |
| 19:14.21 | ds2 | who needs libusb? |
| 19:15.52 | solar | you had usb-utils in there. So that |
| 19:16.47 | ds2 | all I wanted from that is lsusb... chuck usb-utils then since libusb i think is for writing userland drivers |
| 19:18.08 | solar | ET_EXEC libusb-0.1.so.4,libc.so.6 ./usr/sbin/lsusb |
| 19:18.31 | ds2 | that's too "expensive", I can work around that with some hacks |
| 19:18.39 | solar | I have an alias for it :) |
| 19:18.51 | solar | alias lsusp='car /proc/bus/usb' |
| 19:19.01 | solar | or so. I spell it right in the alias |
| 19:19.12 | ds2 | so that's less typing then cat /proc/.... :) |
| 19:19.30 | solar | so pull those two for now? |
| 19:19.33 | ds2 | yep |
| 19:20.01 | ds2 | bluez has to say... donno who would provide the runtime libgcc though |
| 19:20.07 | solar | gcc |
| 19:20.08 | solar | I |
| 19:20.09 | ds2 | has to stay |
| 19:20.18 | solar | I'll have to carefully select that bin only. |
| 19:20.23 | ds2 | installing a compiler seems overkill |
| 19:20.27 | ds2 | OH you can do that? Nice |
| 19:20.30 | solar | cuz otherwise all of gcc gets masked just for the one lib |
| 19:20.47 | solar | yep. I'm overlaying a bunch of config files that I use. |
| 19:20.57 | solar | to setup mdev and act like a hotplug |
| 19:21.00 | ds2 | is this stock gentoo embedded? |
| 19:21.25 | solar | not all of it. Some stuff I still need to push to our tree officiall |
| 19:21.36 | *** join/#edev Ian_Daniher (n=it@69.61.230.246) |
| 19:21.40 | *** join/#edev Ian_Daniher|srv1 (n=it@69.61.230.246) |
| 19:21.53 | ds2 | so far what I am seeing, I am liking it... unlike the other stuff, there seems to be real control |
| 19:23.27 | solar | it's a well kept secret. OE and Gentoo are the only two that offer this kind of flexiablity as far I know. |
| 19:24.03 | solar | ds2: were you at the last CELF in mtv? If so I was the gentoo guy on the panel that did not talk much |
| 19:26.07 | ds2 | solar: no, I was at the conflicting event, ESC |
| 19:26.39 | ds2 | OE's recipes didn't not seem appealing |
| 19:26.54 | ds2 | btw, strace'ing the X server from an xterm is a bad thing ;) |
| 19:27.13 | *** join/#edev keesj (n=keesj@ip49-193-210-87.adsl2.static.versatel.nl) |
| 19:36.54 | solar | I may of learned that one the hardware myself in the past |
| 19:39.26 | solar | Ok. 7.3M all libs should be happy |
| 19:39.44 | ds2 | not bad |
| 19:39.49 | ds2 | a bit large but not bad |
| 19:41.10 | solar | you handy with loop back mounting jffs2 images? |
| 19:41.58 | ds2 | sort of but I don't have a kernel handy with the right options |
| 19:42.07 | solar | if not I'll just post the ROOT from which the jffs2 image is made incase you need to do some editing. I'll be in sunnyvale most of the day at a BBQ drinking this vodka I've been working on all night |
| 19:42.28 | solar | you are using the Hammer carrier and not the Nail right? |
| 19:43.19 | solar | or this is your custom carrier? |
| 19:43.54 | ds2 | custom |
| 19:43.57 | ds2 | but I have the hammer carrier |
| 19:44.14 | ds2 | give me both the jffs2 and the tarball and I should be able to go from there assuming I can at least get a shell |
| 19:44.47 | solar | for now here is the jffs2 image. http://dev.gentoo.org/~solar/embedded/tct/hammer-beagle.jffs2 |
| 19:44.59 | solar | cuz I think your doing something with the beagle+hammer iirc. |
| 19:45.08 | ds2 | yep |
| 19:45.34 | ds2 | I should be able to do all this myself.. just lacking time |
| 19:45.50 | ds2 | thanks for all the work |
| 19:47.03 | solar | all good. I needed to solve this bluez thing anyway for the 8050 |
| 19:47.21 | ds2 | heh... |
| 19:47.32 | solar | ds2: one thing to note. I've not actually tested this yet. Still waiting for them to ship me that unit |
| 19:48.00 | ds2 | solar: I can deal with those "surprises", usually |
| 19:48.13 | ds2 | since this is a rootfs, I'll keep the initrd intact so I can have a shell |
| 19:48.49 | solar | http://dev.gentoo.org/~solar/embedded/tct/ROOT-ds2-20080928.tar.bz2 |
| 19:49.18 | solar | and the cmdline to make that into the jffs2 image is. |
| 19:49.19 | solar | mkfs.jffs2 -x lzo -e 0x20000 -l -s 0x1000 -d . -q -o ../rootfs.jffs2 |
| 19:50.13 | ds2 | 'k |
| 19:50.14 | ds2 | wget'ing |
| 19:50.35 | *** join/#edev prpplague^2 (n=prpplagu@ppp-70-244-162-142.dsl.rcsntx.swbell.net) |
| 19:51.23 | solar | If i stripped out to many locales (which I might of) you will need to snag the glibc pkg and put a few of them back |
| 19:51.43 | solar | which I have the feeling might be the case with some of the stuff in the gconv dir. |
| 19:51.46 | ds2 | don't think I care about locale |
| 19:52.08 | ds2 | it'll bring up 2 PPP connections and the logs will be analyzed |
| 19:52.49 | solar | cool good luck. When all is said and done. If you come up with any handy config files. Plz share. |
| 19:53.14 | ds2 | will do |
| 19:54.55 | prpplague | solar: i'll have the orcad schematic for you late tomorrow, the layout guy had a couple issues that had to be corrected |
| 20:23.08 | chouimat | hi ds2 solar prpplague |
| 20:47.38 | solar | prpplague: cool thanks |
| 21:27.36 | *** join/#edev likewise (n=likewise@82-171-51-231.ip.telfort.nl) |
| 21:41.59 | *** join/#edev Ian_Daniher|srv1 (n=it@69.61.230.246) |
| 21:42.47 | *** join/#edev Ian_Daniher (n=it@69.61.230.246) |
| 22:09.24 | *** join/#edev flavioribeiro (n=flavior@189.71.54.183) |
| 22:47.10 | *** join/#edev likewise (n=likewise@82-171-51-231.ip.telfort.nl) |
| 22:51.47 | *** join/#edev [g2] (n=tom@nslu2-linux/g2) |