**** BEGIN LOGGING AT Mon Mar 01 02:59:57 2010 Mar 01 03:41:44 hmmm. playing with rootstock on karmic. seems to hang at "Extracting zlib1g..." after executing "sudo project-rootstock/rootstock --fqdn ubuntu --login ubuntu --password ubuntu --notarball --imagesize 3G" Mar 01 03:43:24 zlib1g and *-dev are installed. known issues...? Mar 01 03:46:05 There was some discussion about it in http://irclogs.ubuntu.com/2010/02/26/#ubuntu-arm.txt Mar 01 07:31:16 persia: thanks, yes i did see that. just wondering if the solution is still the same? ogra if you are around... any thoughts to share? Mar 01 07:32:32 * persia thinks it's very early in the morning there, and response may be delayed. Mar 01 07:34:12 lol. not a pressing issue. :D thanks persia... is that where you are geogrphically? Mar 01 07:35:42 No, it's late afternoon for me. Mar 01 07:36:24 bedtime for me. will check in manana. night! Mar 01 08:52:13 syadnom: Reason is performance Mar 01 08:52:38 syadnom: the distro targets netbooks, and potentially desktops and mids Mar 01 08:52:43 not older hardware Mar 01 08:53:00 the future is armv7 devices everywhere, and we want to optimize for that Mar 01 08:53:41 Well, the medium-term future. Mar 01 08:53:50 ARMv8 will come someday. Mar 01 08:54:03 But ARMv7 is current, and will last longer than prior generations. Mar 01 08:54:52 persia: True, let's start building for v8 Mar 01 08:56:31 lool: Can we wait until the next release of Ubuntu? I'd rather have at least one release available for each instruction set. Mar 01 11:58:58 lool, i assume your mono patch doesnt fix the assembly installation yet, right ? Mar 01 11:59:43 ogra: it hangs instead of crashing Mar 01 11:59:55 well, thats the karmic behavior Mar 01 12:00:24 Really? Mar 01 12:00:36 yes, in qemu-user it always hung Mar 01 12:00:54 thats why rootstock uses qemu-system for the second stage still Mar 01 12:01:20 its my only blocker to switching rootstock to use a chroot from start to end Mar 01 12:03:24 Apparently, it might be a non-trivial boehm-gc issue Mar 01 12:03:28 yes Mar 01 12:03:36 i was told so in karmic already Mar 01 12:03:53 (i think suihkulokki said it back then) Mar 01 12:05:41 ogra: Is there a bug open about it? Mar 01 12:06:04 i dont think so Mar 01 12:06:06 * ogra checks Mar 01 12:06:58 Hi Mar 01 12:07:27 Can somebody tell me whether I can install Ubuntu 9.04 on my Arm based Nokia N810 Mar 01 12:07:39 lool, not against mono at least Mar 01 12:08:05 Rishi1, only if you have your own kernel and bootloader Mar 01 12:08:39 Rishi1, for userspace creation you can just use rootstock (see RootfsFromScratch wikipage) Mar 01 12:10:02 Rishi1: We don't provide a kernel for it, but the userspace of Ubuntu releases 9.10 and earlier would work Mar 01 12:10:03 lool, should i file against qemu or mono ? Mar 01 12:10:18 ogra: it's a qemu bug Mar 01 12:10:26 It works unders system emulation Mar 01 12:10:45 well, but is boehm gc supposed to work with syscall translation ? :) Mar 01 12:13:34 qemu is supposed to emulate whatever is needed Mar 01 12:13:47 It might expose a boehm-gc bug, but I find that unlikely Mar 01 12:13:56 Anyway, speculation wont lead anywhere here Mar 01 12:15:54 hmm, looking at the mono documentation there seems to be an alternative to boaem Mar 01 12:15:59 *boehm Mar 01 12:16:14 "A new generational, precise and compacting GC is being developed and is currently available from SVN releases of Mono." Mar 01 12:16:45 http://www.mono-project.com/Compacting_GC Mar 01 12:17:21 I have installed Mer on it.. but now Im not able to connect my PenDrive to it Mar 01 12:17:39 "The code is still experimental and should not be used in any kind of production environment." Mar 01 12:17:40 :( Mar 01 12:21:56 Mer is I guess a ubuntu for ARM.. correct me if Im wrng Mar 01 12:22:30 Hmm it's based on Ubuntu sources and built for ARM Mar 01 12:26:44 oh, nice Mar 01 12:26:50 i filed bug 530000 Mar 01 12:26:51 Launchpad bug 530000 in qemu-kvm (Ubuntu) "mono assembly installation under qemu-arm-static hangs (affects: 1)" [Undecided,New] https://launchpad.net/bugs/530000 Mar 01 12:26:55 nice round number :) Mar 01 12:27:04 lool, ^^^ Mar 01 12:28:27 Rishi1: Mer is downstream from Ubuntu (as Ubuntu is downstream from Debian). It isn't Ubuntu for arm (that's just Ubuntu), but it is based on Ubuntu and compiled for arm. Mar 01 12:30:14 ok Mar 01 12:30:31 Thnx for info Mar 01 12:36:02 How could I enable usb host mode in Mer ?? Mar 01 12:40:09 Were the folk in #mer unable to help you? Mar 01 12:40:32 I have a feeling that it's a kernel-specific hack, and I'm not sure we have software that does that (although I could be wrong). Mar 01 12:40:46 * persia would like to help, but has no idea where to start Mar 01 12:44:02 lool, that gmane link doesnt open for me Mar 01 13:04:12 persia: #mer people are not responding... It must be a simple trick.. In Maemo it was so easy.. Mer is a more advanced than Maemo .. but Im not able to find any clue regarding it Mar 01 13:05:05 In Mer it has a provision that One can connect USB as ethernet .. Mar 01 13:18:49 * ogra takes a break Mar 01 13:52:15 ogra: works for me Mar 01 14:07:55 weird Mar 01 14:15:33 hi ogra Mar 01 14:15:40 hey Mar 01 14:16:07 ogra, is it possible to store two different magic number for the same platform in /usr/share/binfmts/qemu-arm? :-) Mar 01 14:16:22 i dont think so Mar 01 14:16:30 :-/ Mar 01 14:17:05 just create a second binfmt hook Mar 01 14:17:49 ogra, cool! i will figure out how to do that Mar 01 14:17:51 cp /usr/share/binfmts/qemu-arm /usr/share/binfmts/qemu-my-personal-binfmt Mar 01 14:17:53 thanks ogra Mar 01 14:18:07 edit that and do the binfmt-support magic Mar 01 14:18:16 ogra, and how it will load both? Mar 01 14:18:22 (see the qemu-static-user postinst for what to do) Mar 01 14:18:32 cool Mar 01 14:18:40 it will load one or the other Mar 01 14:18:47 based on your magic number Mar 01 14:19:11 ogra, and how do you get this magic number? Mar 01 14:19:57 i was looking for the magic binary but i did not find Mar 01 14:21:16 i just got a different output with the file cmd Mar 01 14:21:44 but i did not figure out how to get that strange string Mar 01 14:27:04 rbelem: Why do you need to do this? Mar 01 14:27:10 * persia expects to learn bundles Mar 01 14:27:14 ehhehe Mar 01 14:27:48 persia, i just got a different file cmd output Mar 01 14:28:05 persia, and the binary is not running Mar 01 14:28:36 ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV Mar 01 14:29:05 is it an i386? Mar 01 14:29:07 That's *supposed* to run under qemu-arm-static, or are you doing binfmt for 32-bit executables? Mar 01 14:29:17 i think it is Mar 01 14:29:29 it runs on my device Mar 01 14:29:34 Odd. Mar 01 14:30:04 :-/ Mar 01 14:30:19 Intel 80386 is definately an x86 binary Mar 01 14:31:00 qemu-arm-static cant run that but if your host is x86 it will just run nativelyx Mar 01 14:31:15 Doesn't mean that the device in question isn't using qemu-static-80386 to run it :) Mar 01 14:31:31 if someone builds qemu-static-80386 :) Mar 01 14:31:44 i dont think we do in ubuntu :) Mar 01 14:32:04 given that qemu only builds on x86 based arches atm Mar 01 14:34:05 hum... i will check it Mar 01 14:39:44 hmm, what is myigep.com ? Mar 01 14:40:29 ah, omap3530 boards Mar 01 14:46:30 wow, thats cute Mar 01 14:46:33 http://www.igep-platform.com/index.php?option=com_content&view=article&id=46&Itemid=55 Mar 01 14:47:33 that set of peripherials is awesome for $145 Mar 01 17:14:59 what do you use for uImage debian package creation? it seems make-kpkg does not support uImage. :-/ Mar 01 17:18:14 zumbi: uboot-mkimage Mar 01 17:18:35 zumbi: well, for kernel/initrd etc Mar 01 17:19:01 plars: yes, kernel/initrd is what i look for, does it replaces make-kpkg? or you use that afterwards? Mar 01 17:19:35 nice, uboot-mkimage - generate kernel image for U-Boot :-) Mar 01 19:21:09 zumbi: fail! Mar 01 20:21:12 https://wiki.ubuntu.com/ARM states: Ubuntu jaunty targets the ARM EABI, with an expectation of minimum compliance with the ARMv5t instruction set. Mar 01 20:21:39 my question is: can i compile to armv5t target from karmic? Mar 01 20:22:00 Ox83: Yes Mar 01 20:22:10 Ox83: But all libs are built for armv6 + vfp in karmic Mar 01 20:22:40 So if you build a static binary which links to any lib, it might have some v6 + vfp instructions in it Mar 01 20:24:37 hmmm. any workarounds to avoid that? Mar 01 20:24:48 :D Mar 01 20:25:48 build your own libs Mar 01 20:26:19 lol Mar 01 20:27:13 heh. well... any way to emulate jaunty envt to avoid 1) building libs from scratch and 2) ending up with armv6 libs? Mar 01 20:31:40 short answer: install jaunty. any others? Mar 01 20:31:55 build under chroot? Mar 01 22:14:38 Ox83: Sticking with Jaunty is really the only option if you need to stay with ARMv5 Mar 01 22:16:00 ok. so a dual boot install of jaunty is on its way. Mar 01 22:16:33 btw: anyone try a jaunty build for arm5 system? any pros cons to that vs. deb? Mar 01 22:30:59 Ox83: I run a jaunty system with some extra backports. It's a bit faster for some stuff than Debian, if you have the right hardware. There are more bugs than in Debian. Unless you anticipate an upgrade path, or need vendor support on a preinstall, I'd recommend Debian for ARMv5 systems. Mar 01 22:31:37 (vendor support being available for stuff like SheevaPlug, SmartQ5, etc.) Mar 01 22:34:39 sorry to make you repeat yourself persia. : / Mar 01 22:35:45 Ox83: No worries. Probably 50-60% of my IRC traffic is duplicate to my previous statements :) Mar 01 22:35:49 but i am grateful for the input. i have been playing with OE to get images for this machine working, but with varied luck. it is nice to have (some) options :D Mar 01 22:36:01 Which machine again? Mar 01 22:36:07 zaurus sl-5600 Mar 01 22:36:34 Oh, right. I remember now. Yeah, even if you jam jaunty on that, you'll likely not be happy with the results. Mar 01 22:37:02 The lxde stack didn't get happy until karmic, and what GUI stacks are in jaunty weren't tested at that resolution. Mar 01 22:37:49 heh. wisdom. thank you. will play with deb then (another repeat sry). Mar 01 22:40:13 With luck, someone will release a replacement device that can run future code. The form-factor seems to have become unpopular (well, except for the RAON Digital products). Mar 02 01:22:24 persia: mainly I wanted an arm netbook for the battery life, and so I had an excuse to play with arm Mar 02 01:23:02 Would you accept a 5" Netbook with a 7-8 hour battery life? Mar 02 01:23:16 But I've basically been sold on the new AMD64 Atom that came out in December, which has 8 to 10 hours. Mar 02 01:24:11 What's the mass of a shipping product? At least for me, grams matter more than architecture. Mar 02 01:26:52 persia: my netbook's primary job is to connect to an external monitor and keyboard, and run Emacs and SSH to connect to "real" computers (i.e. stuff in the racks) for sysadmin work and for writing code. Mar 02 01:27:53 It's secondary role is to let me do the same stuff while I'm not at my desk, e.g. when I'm at a friend's place or on a bus. Mar 02 01:28:50 Well, if you're willing to run Lucid, you can use a USB dock with an arbitrary device to take care of the primary job. Mar 02 01:29:23 The secondary role becomes more interesting: and then it's a matter of size/mass/time Mar 02 01:30:35 It'll run Debian, not Ubuntu. Sorry. Mar 02 01:31:40 Well, I hope to find time to get displaylink working for squeeze, but I'm not sure I'll finish testing prior to freeze. It will definitely be in squeeze+1. Mar 02 01:31:50 But then this isn't the right channel anymore :) Mar 02 01:33:05 Yeah, I figured I'd just stick my head in to hear what you had to say re. the netwalker Mar 02 01:34:18 I'm happy with mine, so I tell people who want ARM netbooks that there is one currently in retail and they should go buy it. Mar 02 01:34:31 I'm not convinced that lots of others will appear in retail if the first one doesn't sell so well. Mar 02 01:35:39 What was it retailing going? Mar 02 01:35:43 What was it retailing *for? Mar 02 01:35:47 EPARSE Mar 02 01:35:59 (Edited out the wrong word.) Mar 02 01:36:25 Depends on where you get it. Somthing like 500 AUD Mar 02 01:36:42 +shipping / etc. Mar 02 01:37:14 Blergh. If it was AUD300 I'd be more interested. Mar 02 01:37:45 wait, arm netboook where? the touchbook? Mar 02 01:37:47 Yeah well. This is why the rest of the world has to wait a couple years for Japanese electronics :) Mar 02 01:37:54 DanaG: The NetWalker Mar 02 01:37:57 DanaG: we're talking about the Sharp Netwalker Mar 02 01:38:01 DanaG: http://en.wikipedia.org/wiki/Netwalker Mar 02 01:38:04 ah. Mar 02 01:38:30 Nice little box. I find the positioning of the "A" key a bit frustrating, but otherwise very usable. Mar 02 01:39:06 Sturdy too: I've watched it bounce down a flight of concrete&steel stairs, and dropped it from ~3.5 meters a few times onto hardwood. I don't recommend doing this, but mine still works. Mar 02 01:40:11 hmm, does it do mobile broadband? Mar 02 01:40:16 (not that I have a data plan...) Mar 02 01:40:50 DanaG: It's advertised as doing so with a USB dongle. Having split my case open, I *think* there's enough space to wedge one inside, but it's a bit of a hack. Mar 02 01:41:07 (plus you'd be down to one USB port) Mar 02 01:41:33 bah, I'd want mini-pcie.... and you could use it for gps, while you were at it anyway. Mar 02 01:42:30 Finding PCI or PCIe on ARM boards is *hard*. I have a collection of ARM devices, but only one of them has PCI, and I'm not convinced it isn't a hack like PCI-over-USB on the boards. Mar 02 01:42:59 The bus usually isn't there at all. Mar 02 01:43:35 You should be able to work it out by checking if the chipsets on the board implement pcie at all Mar 02 01:44:10 I'm pretty sure the i.MX51 series doesn't. Mar 02 01:44:37 Yeah Mar 02 01:45:22 Erm, most "Mini-PCIe" WWAN devices are really USB-based. Mar 02 01:45:49 And most "Mini-PCIe" SSDs abuse mini-pcie pins for SATA or PATA. Mar 02 01:46:28 Sure, but if the bus isn't present, you can't expect the pinout to be available to be abused. Mar 02 01:46:57 At least for the Eee hacks, stripping the case from a USB dongle and sticking it internal seemed a popular solution. Mar 02 01:47:01 Is mini-pcie that crufty mishmash of pcie and cardbus? Mar 02 01:47:08 twb: Yes. Mar 02 01:47:13 Righto Mar 02 01:47:28 Not quite as confusing as XpressCard, but close :) Mar 02 01:49:12 Wrongo. Mar 02 01:49:23 Mini-PCIe is the internal slot -- equivalent of mini-PCI. Mar 02 01:49:24 Please share? Mar 02 01:49:35 ExpressCard is the PCIe equivalent of CardBus. Mar 02 01:49:59 Both mini-PCIe and ExpressCard happen to offer both USB and PCIe interfaces in the standard, though. Mar 02 01:50:05 Sorry, I got confused. Mar 02 01:50:29 OK. So mini-PCIe is USB+PCIe and ExpressCard is CardBus+PCIe+USB ? Mar 02 01:50:49 nope, ExpressCard is just hotpluggable PCIe+USB in a cardbus-ish shape. Mar 02 01:50:58 yep Mar 02 01:51:03 convenient too Mar 02 01:51:05 So the old CardBus ATA hacks don't work? Mar 02 01:51:33 although the cardbus-like connector puts some serious constraints on the types of peripherals you can create Mar 02 01:51:49 And ExpressCard 54mm actually offers LESS space than cardbus. Mar 02 01:51:54 Martyn: Bandwidth excepted, how? Mar 02 01:52:12 persia : Mini-PCIe can handle larger current draw Mar 02 01:52:17 Aha! The EeePC has a *special* *custom* miniPCIe that is PCIe+USB+SATA Mar 02 01:52:19 it's designed for it Mar 02 01:52:33 persia: which eee is that? I'll avoid it. Mar 02 01:52:36 Which pins do they overload to get SATA? Mar 02 01:52:36 Martyn: That makes sense. Mar 02 01:53:12 http://beta.ivancover.com/wiki/index.php/Eee_PC_Research is cited in wikipedia as the reference for the odd custom port. Mar 02 01:53:25 And a bummer with HP: they whitelist stuff, so you can't use the wwan usb-only slot for anything but HP wwan module. Mar 02 01:54:14 must be pins 51,49,47,45,43,41,39,37 Mar 02 01:54:16 those are "reserved" Mar 02 01:55:00 DanaG: But seriously, you shouldn't worry about PCIe if looking for an ARM netbook. It's unlikely to be implemented in most solutions (I would expect PCIe to only be exposed for "server" type devices (which might include plug-servers and NAS boxes). Mar 02 01:55:43 persia : *cough* *COUGH* Mar 02 01:55:50 * Martyn can't say .. but *COOOOOOOUUUGH* Mar 02 01:56:18 (in fact, most people implementing cortex-A9's are implementing Synopsys' PCIe) Mar 02 01:56:22 Martyn: Really? That's very encouraging. Mar 02 01:56:34 I saw somebody on #radeon talking about sticking a Radeon on Marvell ARM thingy. Mar 02 01:57:14 DanaG : most manufacturers of multicore ARM solutions are looking at PowerVR Mar 02 01:57:23 although nVidia is getting hot 'n heavy with Tegra2 Mar 02 01:57:37 Does PowerVR have open-source drivers yet? Mar 02 01:57:37 Good luck getting open drivers for that, though. =þ Mar 02 01:58:13 DanaG : Open drivers now exist Mar 02 01:58:25 Oh excellent! Mar 02 01:58:36 DanaG : They went through the trouble of creating an ABI + shim just like nVidia Mar 02 01:58:53 so they can publish a binary driver, and open source (linux) solutions only have to implement the shim Mar 02 01:59:05 ah, though, that's not entirely open. Mar 02 01:59:20 Sure, but open drivers DO exist, just not optimized, nor with great 3D accel Mar 02 01:59:37 random link: http://www.techeye.net/software/amd-and-nvidia-bitchfight-over-open-source-support Mar 02 01:59:41 A shim solution is sufficient to start. Mar 02 01:59:58 ah. Mar 02 02:00:00 A parallel to the various more-open driver can wait. Mar 02 02:00:18 My criterion for "good" 3D is "can run compiz". Mar 02 02:00:36 =þ Mar 02 02:00:41 hell, there are some OPEN SOURCE drivers that can't run compiz Mar 02 02:00:50 for well supported cards (points at nouveau) Mar 02 02:00:51 Lots of them. Mar 02 02:01:02 nouveau can run compiz for some cards. Mar 02 02:01:13 in fact, I can't think of any that can support the full compiz requirements Mar 02 02:01:26 closed source shims do so pretty well Mar 02 02:01:41 They have their own problems Mar 02 02:01:42 check the last comment on that thread... that's my post. Mar 02 02:01:45 <-- is _not_ biased towards open source. Mar 02 02:01:47 * persia has a shim-melted card Mar 02 02:02:16 What I call bad: when your binary drivers fail to do anything but segfault the X server -- and have been broken in that way for 2 years. Mar 02 02:02:18 But shims are sufficiently free to ship, which is the key bit. Mar 02 02:02:40 DanaG: That's usually a case of full binary drivers, rather than shims. Mar 02 02:03:14 And they've "updated it to support new x servers" -- but I say, s/support/segfault/ Mar 02 02:03:18 =þ Mar 02 02:03:41 anyway, how well does the SGX in beagleboard work? Mar 02 02:04:01 Oh, and OpenGL ES (OpenGL on bare console?) seems like 'black voodoo magic' to me. Mar 02 02:04:06 Give me a Matrox GPU any day :-) Mar 02 02:04:23 DanaG : That's a pretty good description :) Mar 02 02:04:39 OpenGL ES uses the framebuffer as the base canvas Mar 02 02:04:43 it's not THAT much voodoo Mar 02 02:04:50 but it's hard to grok **** ENDING LOGGING AT Tue Mar 02 02:59:58 2010