[00:00] Sorry bout that. [00:01] Got a copy of those sources? [00:02] I searched for a half hour earlier while half asleep, and never found them... [00:02] Guess I gave up eventually. [00:03] Let me see if there is a link. [00:21] NCommander: Sure. [00:23] ScottK: thanks [00:33] hrw, y? === Lopi|dinner is now known as Lopi === prpplague^2 is now known as prpplague === lopi_ is now known as Lopi [03:11] GrueMaster: someone crashed my internet earlier, I assume it was the dogs, did you find that link bossmate? [03:11] Yes. http://images.barnesandnoble.com/PResources/download/Nook/source-code/nookcolor_1.2.tgz [03:12] Guess I gotta build it, what was the name of that tool? [03:15] Wow this internet blows. [03:31] =D [03:33] GrueMaster: what was that utility I needed to build it? [03:36] Oops, sorry. I think you can install gcc & gcc-arm-linux-gnueabi [03:36] Then cross compile. [03:36] (dinner called, had to eat) [03:37] Why did it call, was it still alive? [03:40] Daughter shut the internet [03:40] heh [03:40] StevenK: Hey! [03:41] StevenK: No, I don't eat Aussie style. :P [03:41] Hah [03:42] so gnu-gcc and gnu-arm-eabi? [03:43] I didn't quite read it before she shut the lid. [03:43] I'm going to open up irc on the desktop so I don't miss anything. [03:43] heh. gcc-arm-linux-gnueabi [03:44] I knew I was close. [03:44] And that has the utility to build the u-boot source? [03:44] Xase_: I have quasselcore installed on a server in my basement, and I connect to it with quasselclient. The server stays logged in, even when I travel. [03:45] I think it should, but you may need to tweek the makefile. I've never cross compiled before. [03:45] I'm working with unlimited space and resources unfortunately, but that sounds awesome. [03:45] rsalveti would know. [03:45] * rsalveti reading [03:45] READING?!?E!@ER@EWFSD [03:45] How dare you. [03:45] rsalveti: xcompile u-boot [03:45] :p [03:46] CROSS_COMPILE=arm-linux-gnueabi- make omap4_panda_config [03:46] CROSS_COMPILE=arm-linux-gnueabi- make u-boot.bin [03:46] for panda [03:47] Alright, I'll be sure to bother the room with numerous pastebins when I hit retardation walls. [03:48] Xase_: So just change omap4_panda_config for omap3621_evt1a [03:48] You know, I've never bothered trying to share files via windows & Linux... is it easy? [03:48] Xase_: samba should do it [03:49] easy to set [03:49] Good, because I have shite internet, and already downloaded the source to this computer... would rather just share things back and forth. [03:56] Nananana nananana hey hey, lets compile. [04:01] Hi Xase_ [04:01] Hi you sexy me. [04:31] I see internet weirdness has risen again. [07:29] lag: did you get your answer about the TI guy you were looking for? === ndec1 is now known as ndec [07:59] kgilmer: 01:33 is my sleeping time... [08:41] ndec: I did, thank you [08:41] lag: cool... thx.. and sorry about the delay ;-) [08:41] ndec: NP :) [09:16] cool [09:16] so, usb enumeration (and thus booting) on omap3 is broken [09:16] while it works on panda [09:17] at this point i wonder if i shall try with the linaro kernel too... [09:50] yeah hrw. i got it figured out, didn't have to downgrade. thx for the tip. [09:51] on upgrades, i wish there was a "leave my ui just like it is" option. [09:52] kgilmer: lot of people would love to [09:53] kgilmer: I saw that you mvoed to 35 kernel [09:53] yep [09:54] you -> izzy, others in nyc, not me :) [09:54] kgilmer: I meant BL [09:54] seems to be working alright [09:55] i'm also trying to support bug13 hw in my free time, maybe i told you [09:55] MACHINE="bug", basically w/ angstrom [09:56] kgilmer: I noticed on the list and blogs [09:56] ah cool [09:56] working on some refactoring now so at the app layer it is easier to combine different bits for different hw [09:57] so i can have something like "bug for beagleboard" ... our long lost dream [09:57] not that it's going to set the world on fire, but it's fun to work on. [10:04] ;D [10:04] kgilmer: I still have beaglebug board somewhere === Lopi is now known as Lopi|work [14:37] it seems my flyswatter didn't survive the trip to budapest :( === stenzel_ is now known as stenzel [14:47] ok, now i know we i've recompiled the x-loader :) [15:33] ppisati: http://focus.ti.com/pdfs/wtbu/OMAP4430_ES2.0_Public_TRM_vJ.pdf [15:33] ppisati: for the omap 4 trm [15:34] I believe you can find newer versions at the ti website [15:34] but that's the document you were looking for [15:36] when using the linux gadget framework what exactly does "platform code" mean.... [15:36] does that mean creating your own module and associated kernel options that get set in the .config? [15:39] http://www.linuxquestions.org/questions/linux-newbie-8/linux-usb-gadget-hid-driver-problem-840733/ [15:39] basically on this link... the guy responds to his own question, but I don't know where the platform_register_device should go? [16:04] PXE patches are ready [16:04] where should I post the u-boot PXE patches? [16:04] server mailing list? [16:14] martyn: did you post at the u-boot ml? [16:14] jcrigby: can probably be interested on that too [16:14] This patch is meant for us, first [16:15] so we can get working now .. it will need to be upstreamed as it gets to be higher quality [16:15] martyn: jcrigby is maintaining the u-boot we're using at our images [16:15] jcrigby: You online? [16:16] martyn, yes I'm here [16:16] patch coming your way :) [16:16] martyn, thanks! [16:16] I'll do a git patch, then email? [16:17] martyn, sounds good john.rigby@linaro.org [16:19] okay .. know what's going on with the patches? [16:19] this is for the arm server 11.10 .. we need PXE, so calxeda is putting in our PXE patches [16:20] * hrw thought that tftp is enough [16:20] Nope [16:20] we need to be able to do everything that x86 server can do [16:21] that includes pxe install :) [16:21] heck, I'm even going to take a stab at implementing TCP so we can do HTTP install [16:21] (gpxe) [16:21] but the PXE patch is ready now [16:21] I had arm boards which booted uboot from flash and then tftp-ed kernel to boot [16:22] hrw : Refer to notes from UDS : https://blueprints.launchpad.net/ubuntu/+spec/server-o-arm-server [16:22] ok [16:31] ... GrueMaster what was that target rsalveti mentioned, stupid webchat crashed, and my irc client hadn't joined the room till after he mentioned it. [16:32] [19:47:31] CROSS_COMPILE=arm-linux-gnueabi- make omap4_panda_config [16:32] [19:47:42] CROSS_COMPILE=arm-linux-gnueabi- make u-boot.bin [16:32] And I noted that you will want omap3621_evt1a [16:35] Yeah thanks :D GrueMaster gonna compile that soon as I get samba up [16:36] jcrigby: Do you want the patch we used to change the configuration on ca9x4_ct_vxp as well? It's a good reference [16:37] (I'm just using git diff to generate the patch) [16:37] martyn, thats fine [16:47] Man... [16:47] Samba sucks to set up [16:48] I got it working on windows to access ubuntu... but ubuntu can't open the network for some reason [16:54] THERE [16:54] Finally, here comes the source... [16:54] I'll figure out why I can't see the computer from Ubuntu's end later... [17:01] Windows doesn't play well with others. [17:05] Great and now some how I've misplaced my natty img [17:09] oops [17:24] wtf... now that mount command asks for filetype [17:26] GrueMaster: what was that mount command again -_-? I have this: sudo mount /home/jase/wrkspc/natty-preinstalled-netbook-armel+omap.img /home/jase/wrkspc/NATTYHAXYO/ -o loop,offset$((512*63)) [17:27] should be offset=$((512*63)). Missing = [17:27] You are mighty sir@ GrueMaster [17:27] heh [17:28] now, I'll only have to replace u-boot.bin? @ GrueMaster , or will I have to compile some other files? [17:28] x-loader as well. [17:29] so build x-loader and u-boot, and just copy over my resultant files? [17:31] make: *** No rule to make target `omap3621_evt1a'. Stop. [17:33] Xase: I am not sure what the name of the config is for the nook. I haven't had a chance to look through it. [17:33] Would it be in the nook source? [17:34] I'll ask in NookDevs and Cyanogen mod [17:34] It should be [17:38] I assume I'm looking for a .mk file@ GrueMaster? [17:38] Let me look. Give me a sec. [17:39] Well I see several. [17:39] Take your time. [17:41] there's config.mk i386_*.mk m68k_*.mk microblaze_*.mk mips_*.mk nios2_*.mk nios_*.mk and ppc_*.mk here [17:41] oh and arm_*.mk [17:41] and a blackfin one [17:43] make -j2 omap3621_evt1a_config @ GrueMaster according to Nookdevs.com [17:44] jase@xike:~/wrkspc/distro/u-boot$ CROSS_COMPILE=arm-linux-gnueabi- make omap3621_evt1a_config [17:44] Configuring for omap3621_evt1a board... [17:45] Compiling now @ GrueMaster [17:45] yep. [17:46] Can't believe all I was missing was _config :D [17:47] Do the same for x-loader once you have u-boot. [17:47] Don't I have to also build the kernel ? [17:47] Yea, probably. [17:48] Like I said before, I haven't really had time to explore this. [17:48] Well I'll just wing it, and reference nookdevs.com. Get back to what you're doing boss man. [17:52] heh [17:55] GrueMaster: Building that kernel now, then I'll dd the image to my SD Card and reply with the results [17:56] cool [18:06] GrueMaster: one more teensy thing... that command to mount didn't give read write access did it? [18:06] or only to root? [18:06] It should have, but it may be root only write access. You could also just flash and SD and mount it. [18:07] brb [18:07] Meaning like... if I already have a prepared sd card? [18:39] Flash the image to the SD, then mountthe first partition and copy your files to it. [18:40] has anyone had some success at making a platform hid device with the linux gadget framework? [18:40] I've got my device up but no transfer between /dev/hidg0 (device side) and /dev/hidraw0 (host side) [18:50] When I plug in the device, and boot it with the card in, it shows up as an omap3630 GrueMaster [18:51] No display comes up though [18:51] Hmm. [18:51] Plug it in where? [18:51] I wish there was a way to communicate with the device over usb... but I've been googling for hours and haven't found anything. [18:52] My windows PC [18:52] Oh. [18:52] It's just what my nook cord happened to be plugged into. [18:52] Have you flashed the sd card with the natty image? [18:52] Yup [18:52] ANd rebooted? [18:52] Yep [18:52] But nothing? Interesting. [18:52] Devmgr didn't report an omap 3630 last time though I think. [18:53] I built mlo uboot and uimage [18:56] I hear windows detecting it, and then failing to register it. [18:57] Is there perchance a way to see if anything is happening over usb? [18:57] Via linux of course. [18:58] not really. But when I looked at u-boot, it looked like I would need to make some changes so that it loads initrd properly along with our boot.scr [18:58] Hmm [18:58] So more work to tweak. [18:58] Yay [18:58] As I said, I haven't had a chance to really dive in for the last 2 weeks. [18:59] Stuff that's over my head :), Well let me know what I can do to help and whatnot, and I can be your ears and eyes [18:59] And I really won't be able to focus on it until next week. [18:59] Yeah, I'll see if I can equip my self with some developers of +1 intelligence :p [18:59] Sounds good. BTW: there is a thread on xdadevelopers that has a rudamentary ubuntu running on it through vnc. [19:00] Yeah, that is an old trick. [19:00] I used to do that on my G1 [19:00] look for nookbuntu [19:00] Think that ubuntu image would help some how? [19:00] Ah, so you are familiar with it. [19:00] Might for the boot portion. [19:03] Yeah, I saw that, and said PFFT [19:03] It's rather useless [19:03] But I'll take a look at it. [19:09] hey rsalveti, do you know if the patch for panda to take ethernet MAC addr from bootargs is gone? [19:10] guy next to me just switched to same kernel, and now we both have the same MAC addr on our pandas, which is a bit inconvenient.. [19:15] ahh, yeah, looks like the patch for macaddr in cmdline is reverted [19:21] GrueMaster / rsalveti question for you guys [19:21] Shoot. [19:23] GrueMaster / rsalveti if you had an LCD adapter available for the pandaboard, would you prefer to have a large +5v wall wart that is used for the +5 and generates +12v for the lcd backlight, or would you prefer to have support for +9V or larger wall wart that drives the backlight and regulates down to +5v for the panda? [19:24] Personally I would prefer a brick. No room for wall warts left (I already have 5 powerstips in my office). [19:25] But one power source for both would be easier to manage. [19:25] GrueMaster: right, but something like +12v and drop to +5v , or +5v push up to +12v [19:25] GrueMaster: i'm stuck on tht [19:26] I don't know enough about electronics to know the pros/cons of either. [19:27] I do know most laptops use 19v DC and down convert for devices. [19:28] GrueMaster: yea it is easier to down convert [19:28] Plus most cars are 12v, so someone doing in-car dev work would prefer that I think. [19:28] GrueMaster: indeed [19:43] What's NCommander's canonical email add? [19:43] * martyn is sending out pxe patches now [19:54] patches sent [19:54] enjoy [20:43] So the pre-installed image goes to OEM setup right? [20:43] Anyway to set ssh or vnc to start and connect over to it that way? [20:44] To see if anything is actually functioning? [21:02] martyn: thanks [21:03] It's a start :) [21:08] Beh, now this is no fun :( [21:11] robclark: I think it was reverted, but it's now using the omap die id to create the mac address [21:11] Huh .. I think Java is broken on my natty installation [21:11] I know the first version of this patch was getting the wrong values from the omap die id, so all macs were all the same [21:11] *grumpy* [21:11] robclark: which kernel are you using? [21:12] rsalveti, yeah, I saw that in the code.. [21:12] I'm using your drm branch ;-) [21:12] not sure if we have same die-id (at least in the bits it is reading)? [21:12] robclark: it should be unique [21:13] at least with my current kernel I'm getting different macs for all my 3 pandas [21:13] where does the OMAP normally get it's MAC address? [21:13] it doesn't store it in nvram? [21:13] robclark: could be that my branch is missing this fix [21:13] OMAP doesn't have a MAC.. [21:13] OMAP/pandaboard [21:13] sorry :) [21:13] martyn: missing eeprom at the smsc hardware [21:13] urk [21:13] * robclark just being over-precise [21:13] rsalveti, do you have a pointer to the right patch? I can double check that.. [21:13] No, it's good. There certainly are lots of OMAP boards out there [21:14] robclark: sure, hold on [21:14] thx [21:14] rsalveti: It uses that damned SMSC usb->ethernet hardware? [21:14] martyn: yup [21:14] same as beagle xm [21:14] yuk [21:14] martyn, fwiw, there is no eeprom/nvram, which makes life a bit more difficult.. [21:14] I like the ones that use the ram interface better [21:14] 1911 [21:15] anyone know what snowball or the other low cost arm boards are doing? I assume it is a common problem.. [21:15] BRB -- I'm taking a look at the pandaboard that Trevor has here [21:15] because that's just a silly decision :) [21:15] snowball has an SRAM interface ethernet [21:15] I'll bet you it stores a mac address in flash :) [21:16] ok, so has eMMC or something like that on the board? [21:16] looking at the CALAO website now [21:16] Ethernet is a SMSC LAN9221 on the Snowball [21:16] Wireless is the Azurewave NH580 [21:18] 8GByte eMMC [21:19] martyn: picking one up? [21:19] ten [21:19] robclark: snowball doesn't have eeprom either [21:19] prpplague: But it does have eMMC [21:20] martyn: right [21:20] which means you can load a MAC address in via u-boot [21:20] robclark: just apply this patch: http://rsalveti.net/tmp/0001-OMAP2-Common-CPU-DIE-ID-reading-code-reads-wrong-reg.patch [21:20] Most dev boards don't have eeprom/nvram for mac addresses. It requres registration with a standards body to get a mac range. [21:22] martyn: have they updated the ship dates? [21:22] no. [21:22] * prpplague wishes the #igloo channel was on freenode instead of oftc [21:22] but we pre-ordered [21:22] right as did i [21:23] Hmm .. according to the SMSC spec sheet for the LAN9221 .. we -can- burn a MAC into it [21:23] seems trivial to do so [21:24] Ah, I see [21:24] "optional EEPROM interface" [21:24] boooo [21:25] And they save a -whooooole- eight bytes of storage [21:26] GrueMaster: any tips for where I should head from here? [21:33] Not yet. I was going to work from a post-installed image. Do you get video? [21:40] I get nothing on the screen. === Xase1 is now known as Xase [23:04] ogra_: Damnit man .. your serial cable is STILL IN MY BAG [23:04] AaaaaaaaaAAAaa!@ [23:05] I found it this morning [23:05] LOL [23:05] It's like the famous cat from the song 'the cat came back' [23:05] PXE patches submitted [23:05] ncommander and jcrigby have 'em [23:05] just keep it :) [23:06] I'll keep it in the 'to lend' hardware pile [23:10] hi, i did a 'do-release-upgrade' on my NSLU2 running jaunty, and it died in the ass in a big way: [23:10] ERROR got an error from dpkg for pkg: '/var/cache/apt/archives/libc6_2.10.1-0ubuntu19_armel.deb': 'subprocess new post-removal script killed by signal (Illegal instruction)' [23:11] i've read that support for the NSLU2 was dropped after jaunty, but do you know how i can clean this thing up so i can use it again? as in roll back to jaunty [23:13] reinstall is your only option i fear === Lopi|work is now known as Lopi [23:16] apt and dpkg still seem to function OK, but it is missing things like ldconfig and locale, lol [23:17] well, if the new libc is already installed you will definitely run into probs, karmic was all compiled for armv6 [23:24] tim, that sounds pretty grave [23:36] ah well, gotta keep it interesting somehow :)