| 15:44.24 | *** join/#edev purl (~purl@rikers.org) |
| 15:44.24 | *** topic/#edev is Discussion of embedded systems (Hardware/Software/Linux/Commercial Device Repurposing) : Busybox and Buildroot : http://www.elinux.org |
| 15:59.08 | *** mode/#edev [+o prpplague] by ChanServ |
| 16:06.49 | *** join/#edev T0mW (~Tom@24.115.148.35.res-cmts.t132.ptd.net) |
| 16:16.38 | prpplague | T0mW: hey bud |
| 16:26.43 | slashbeast | aargh. I cant build mdadm in aboriginal and I cant build lvm2 and busybox in buildroot, it does fail with segfault *or* undefinied __fini_array_start |
| 16:39.06 | *** join/#edev Mirell (~mirell@2607:fc88:1001:801e::2) |
| 16:39.07 | *** mode/#edev [+o Mirell] by ChanServ |
| 17:01.59 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 17:05.13 | *** join/#edev landley (~landley@140.242.26.2) |
| 17:05.51 | landley | All through the 20 minute presentation on branding where they refused to show us the new logo, I was going to myself "as long as it's not the brown ring of quality". |
| 17:05.55 | landley | It was the brown ring of quality. |
| 17:06.06 | landley | Nevermind. |
| 17:11.28 | av500 | polycom? |
| 17:11.33 | slashbeast | landley: yo, 'Finally got a couple hours to bang on open source again' <- did that moved up mdadm on your todo list? ;) |
| 17:13.33 | landley | slashbeast: aboriginal linux waits until I get the server set back up. |
| 17:13.48 | landley | av500: I plead the 5th. |
| 17:14.51 | slashbeast | landley: need any help with that? |
| 17:15.08 | av500 | landley: granted |
| 17:15.48 | landley | slashbeast: nah, I just need to get the right cables from the condo. Last night's run didn't have _any_ power cables. |
| 17:16.01 | landley | (Getting a desk to put it on would be nice too.) |
| 17:16.36 | landley | My #1 aboriginal todo item is actually trying to figure out why the fedora build doesn't work. I'm halfway through that one. |
| 17:16.46 | landley | Then mdadm, then making sure lfs builds all the way through. |
| 17:35.04 | slashbeast | landley: find: unrecognized: ( |
| 17:35.21 | slashbeast | landley: busybox 1.20 fail to build inside aboriginal, propably need CONFIG_FIND_XXX to be enabled first. |
| 17:35.43 | landley | slashbeast: release version or hg version? |
| 17:35.51 | landley | Because I was pretty sure I'd fixed that one over a month ago... |
| 17:36.00 | slashbeast | landley: hg version |
| 17:36.03 | slashbeast | I just bulit i586. |
| 17:36.17 | landley | Ah, 1.20. Haven't tried that one yet. |
| 17:36.51 | landley | I need to switch to the new busybox and the new uClibc++, but I'd like to fix the known bugs first. :) |
| 17:37.02 | *** join/#edev sndcrb2 (~sndcrb@dsl-hkibrasgw2-fef8de00-126.dhcp.inet.fi) |
| 17:51.47 | T0mW | prpplague: hey dude, what's new and exciting? |
| 17:58.29 | *** join/#edev florian (~fuchs@sign-4d094293.pool.mediaWays.net) |
| 17:58.29 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 18:02.07 | prpplague | T0mW: not a darn thing |
| 18:02.25 | prpplague | T0mW: i do like to complain that every time i need to ask you a question you aren't on the channel, hehe |
| 18:02.30 | T0mW | prpplague: getting married June 16th. |
| 18:02.51 | prpplague | T0mW: i will wear black that day for you and drink a good pint |
| 18:03.04 | T0mW | purl: t0mW |
| 18:03.04 | purl | from memory, t0mw is at tom@openhardware.net, or has a website at http://www.openhardware.net |
| 18:03.09 | T0mW | heh |
| 18:03.11 | prpplague | florian: what's cookin? |
| 18:03.42 | prpplague | T0mW: hehe, yea every time i plan on emailing you, but end up researching more instead |
| 18:03.45 | landley | purl: landley |
| 18:03.45 | purl | you are, like, like, just this guy, ya know |
| 18:03.56 | landley | purl: purl |
| 18:03.56 | purl | somebody said purl was the only good thing about sc2mapster |
| 18:04.04 | prpplague | purl: prpplague |
| 18:04.04 | purl | i guess prpplague is mailto:danders@amltd.com or Dave Anders |
| 18:04.08 | prpplague | yikes |
| 18:04.10 | prpplague | that is old |
| 18:04.13 | florian | T0mW: heh funny, I got married on a 16th of June as well :) |
| 18:04.26 | landley | purl: Mary had a little lamb, some white wine, and a salad. |
| 18:04.29 | prpplague | purl: prpplague is mailto:danders@tincantools.com or Dave Anders |
| 18:04.29 | purl | ...but prpplague is already something else... |
| 18:04.36 | prpplague | purl: forget prpplague |
| 18:04.36 | purl | prpplague: i forgot prpplague |
| 18:04.48 | florian | prpplague: hi plague :) |
| 18:04.52 | prpplague | purl: prpplague |
| 18:05.09 | T0mW | florian: was going to have been May 26th, but then I found out that I didn't have a cerified divorce decree so we couldn't get the license. |
| 18:05.18 | prpplague | purl: prpplague is mailto:danders@tincantools.com or Dave Anders |
| 18:05.18 | purl | prpplague: okay |
| 18:05.28 | prpplague | T0mW: doh |
| 18:05.56 | landley | Sigh. Why am I the only one to find things like http://www.slate.com/blogs/moneybox/2012/05/03/kindle_fire_the_fruitcake_of_tablets_.html flipping obvious? |
| 18:06.02 | T0mW | I thought that just a photocopy was fine... Hey, it has been almost 30 years since the divorce. |
| 18:06.05 | landley | A successful tablet is a big phone, not a small pc. |
| 18:06.30 | T0mW | landley: don't tell microsoft that, you'll burst their bubbles. |
| 18:06.34 | landley | mainframe -> minicomputer -> microcomputer -> smartphone. PC getting kicked up into the server space is called "the cloud" just like the mini getting kicked up was called "the server". |
| 18:06.47 | landley | Why is this not obvious to everybody else? |
| 18:08.02 | T0mW | landley: because most business admin degrees are given to people that are managers at MacDonalds first? That they have no clue as to what you just said. |
| 18:09.34 | prpplague | hehe |
| 18:10.13 | prpplague | i enjoy my kindle fire, even if the default OS is a little buggy |
| 18:10.16 | T0mW | florian: congrats on your marriage. I pretty much gave up finding a good woman, then Sandy suddenly came along... |
| 18:11.19 | florian | T0mW: Ok 30 years is quite some time... |
| 18:12.28 | prpplague | speaking from experience, if you are going to do it, might as well get it right on the first try |
| 18:13.51 | T0mW | florian: yeah, it is. But the time goes by if you're not paying attention. Marriage was not something I wanted to do again. |
| 18:15.09 | florian | T0mW: indeed... I am one who should pay attention a little bit more |
| 18:19.05 | landley | Fade and I were happy living in sin, we only really got married for the health insurance. |
| 18:19.28 | landley | (I admit there are some tax benefits too.) |
| 18:21.43 | T0mW | Yeah, I'm marrying Sandy for basically the same. I've got the house now, she is also paying into it and I want to make sure that if something happens to me that she can keep it. |
| 18:22.23 | T0mW | funny how life changes when you meet the right one. |
| 18:22.52 | T0mW | Look at what Donna did to Dave. heh |
| 18:36.27 | *** join/#edev eephillip (~eephillip@pdpc/supporter/student/eephillip) |
| 18:39.29 | *** join/#edev eephillip (~eephillip@pdpc/supporter/student/eephillip) |
| 18:55.29 | landley | Bleaugh. |
| 18:55.48 | landley | Having an hour long presentation on branding and then going to sit in a cubicle has sort of drained my will to live right now. |
| 18:55.56 | landley | Really hard to actually get any coding done. |
| 18:57.30 | *** join/#edev neale (neale-free@2600:3c00::f03c:91ff:fe96:544d) |
| 18:57.56 | landley | neale: the kernel not saying anything over the serial port could be a number of things. |
| 18:58.15 | landley | Could be toolchain, could be kernel config, could be kernel early init, could be serial. |
| 18:58.21 | landley | Do you have a bootloader running on the board? |
| 18:58.34 | landley | What board is it, by the way? |
| 19:20.52 | neale | hi, sorry |
| 19:21.10 | neale | it's a trimslice |
| 19:21.22 | neale | it has uboot in the firmware, from what I can tell. Boots off an SD card. |
| 19:21.41 | neale | I have an SD card with some Ubuntu image that someone here obtained from someplace, and it boots up just fine. |
| 19:22.10 | neale | I made a uimage with buildroot, copied over the ubuntu's .scr file, and that's where I am now. |
| 19:22.58 | *** join/#edev bizhanMona (~bizhan@nat/cisco/x-fblmguoblkrgekkf) |
| 19:26.43 | slashbeast | landley: do you happen to know what does leeding '-' in mdev? like -net/.* root:root 600 @nameif |
| 19:26.48 | landley | So u-boot loads the kernel, but the kernel doesn't produce output? |
| 19:27.10 | slashbeast | landley: googled, 'dont stop on this line' |
| 19:27.10 | landley | slashbeast: it sounds vaguely familiar, but no. |
| 19:27.12 | slashbeast | ;-) |
| 19:27.14 | neale | that's correct. The last thing I see on the serial console is "Starting kernel ..." |
| 19:27.25 | slashbeast | * Leading minus in 1st field means "don't stop on this line", otherwise |
| 19:27.25 | slashbeast | * search is stopped after the matching line is encountered. |
| 19:27.38 | landley | neale: what kernel .config are you using? |
| 19:28.00 | landley | slashbeast: sounds about right. That would be one of solar's suggestions. |
| 19:28.23 | landley | He came up with most of the config file stuff. |
| 19:28.51 | landley | I just went "need somewhere to specify permissions and ownership"... |
| 19:29.06 | neale | it's a fully custom .config from an x86 build of the same embedded distro. I of course changed the architecture, but it very well could be I did that wrong. |
| 19:29.11 | landley | The actual conversation (on #uclibc here) is archived. I think I linked to it here a week or two back. :) |
| 19:29.40 | landley | neale: you need to enable EARLY_PRINTK=y in your kernel so you can see output as soon as possible. |
| 19:29.53 | landley | And you need to enable support for the serial chip you're using, and having the console on it. |
| 19:30.16 | slashbeast | landley: I noticed that tun[0-9]* also match 'tun' |
| 19:30.17 | slashbeast | odd. |
| 19:30.24 | landley | (Most boards use a 16550 variant because it's just TRIVIALLY easy to program that. But sometimes they write a driver that pretends it isn't a 16550 even though it is. TI did that, for example.) |
| 19:30.37 | landley | slashbeast: that's regexes for you. * means "0 or more of". |
| 19:30.56 | landley | QUick and dirty fix: tun[0-9][0-9]* |
| 19:31.12 | landley | THere's almost certainly a more graceful way to do that, but I'd have to look it up. |
| 19:31.22 | ohsix | {1,} |
| 19:31.42 | neale | great, thanks landley, the serial driver sounds like it might be the problem. |
| 19:31.47 | landley | ohsix: last I tried that it didn't work, but I think that was in uClibc without the full gnu crap enabled. |
| 19:31.58 | slashbeast | landley: right. can I use there + to match 1 or more? |
| 19:32.10 | landley | You also have to say "console=ttyS0" or whatever your serial devise is. |
| 19:32.17 | neale | here's another newbie question, what needs to be enabled for EARLY_PRINTK to be an option? I'm not seeing it in .config. |
| 19:32.22 | neale | well, yes :) |
| 19:32.25 | landley | ON the Netra boards I've been banging on recently it's ttyO2 for some reason. |
| 19:32.38 | landley | neale: CONFIG_EMBEDDED |
| 19:32.39 | neale | the working ubuntu kernel uses ttyS0 |
| 19:32.48 | neale | ah |
| 19:33.45 | ohsix | well, they're "extended", but basic doesn't even have + or ? so, yea |
| 19:34.13 | ohsix | but both are posix.2 |
| 19:34.16 | landley | neale: in menuconfig, hit forward slash and type EARLY_PRINTK |
| 19:34.24 | landley | IT should show you where it is and what dependencies it needs. |
| 19:36.21 | *** join/#edev strnx (~strnx@199.127.225.243) |
| 19:36.21 | landley | slashbeast: I debugged the busybox 1.20.0 build, but can't check it in until evening. |
| 19:37.08 | slashbeast | landley: it just failed because the system's find did not like ( or { |
| 19:37.17 | slashbeast | landley: propably thats what happen after you stripped the aboriginal |
| 19:37.25 | slashbeast | 's busybox config |
| 19:38.57 | landley | busybox 1.19 didn't need FIND_PAREN and FIND_PRUNE enabled. |
| 19:39.13 | landley | Don't ask me why every single find feature has its own config symbol. |
| 19:39.33 | ohsix | because find is super huge and you must conquer it |
| 19:57.06 | slashbeast | I lol'd at the modprobe where I had to add extra switch for support blacklists |
| 20:02.01 | landley | I admit I'm guilty of this stuff with the toysh config. But I got better. |
| 20:02.19 | landley | This is something I totally changed my mind on since I did the first pass on toysh: most of this config options are GOING TO DIE when I get back around to banging on that. |
| 20:02.50 | landley | Even something like the command shell should have maybe a half-dozen config options, total. |
| 20:03.08 | landley | And that's the most complex command in posix. (SO much so it has its own section of the spec.) |
| 20:04.41 | *** join/#edev Posterdati (~tapioca@host75-208-dynamic.11-87-r.retail.telecomitalia.it) |
| 20:04.45 | Posterdati | hi |
| 20:04.58 | landley | Posterdati: hello. |
| 20:05.13 | Posterdati | please help: it is possible to use pointer to functions on sdcc for pic14? |
| 20:08.45 | slashbeast | landley: I think the toybox will be major success because BSD license. |
| 20:09.08 | landley | slashbeast: that would be nice. :) |
| 20:09.17 | landley | Posterdati: what are sdcc and pic14? |
| 20:10.01 | Posterdati | sdcc is a free compiler for microchip pic16f and pic18f micros |
| 20:12.30 | slashbeast | landley: also, Do you have any notes on gcc >=4.3 cross compile? I am getting always an error with x86_64-walrus-ar: command not found. For some reason you hardcoded it as CROSS_HOST variable. |
| 20:13.03 | slashbeast | would be nice if I could build aboriginal with 4.4 so I can test if the gcc is to blame or not |
| 20:26.36 | *** join/#edev prpplague^2 (~danders@192.94.92.14) |
| 20:29.50 | slashbeast | /usr/lib/gcc/x86_64-pc-linux-gnu/4.5.3/../../../../x86_64-pc-linux-gnu/bin/as: error while loading shared libraries: libc.so.0: cannot open shared object file: No such file or directory |
| 20:29.59 | slashbeast | what the... frack? buildroot x86_64. |
| 20:40.58 | landley | slashbeast: I audited the build output and every time it was using the wrong tool, I hardwired in the right tool through the relevant environment variable. |
| 20:41.10 | landley | I had to greatly change that moving from 4.1.2 to 4.2.1, so I |
| 20:41.24 | landley | I'm not surprised 4.3 has it again. (I vaguely recall making one or two changes there when I did a script for that...) |
| 20:43.08 | landley | http://landley.net/hg/aboriginal/rev/749 |
| 20:43.56 | landley | slashbeast: it's using the host as. |
| 20:44.02 | landley | It should be using the target as. |
| 20:44.04 | landley | Search path's wrong. |
| 20:44.33 | landley | gcc really sucks at cross compiling from x86-64 to x86-64. It keeps thinking it's doing a native compile. |
| 20:44.40 | landley | Which glibc->musl ain't. |
| 20:44.44 | slashbeast | I was doing i586 build |
| 20:44.53 | slashbeast | you use uname -m in functions.sh |
| 20:45.19 | landley | slashbeast: to determine the host tuple, yes. |
| 20:45.52 | landley | I note that the toybox uname will say "i686" for a 32 bit build running on a 64 bit system. :) |
| 20:45.53 | slashbeast | gcc is nightmare |
| 20:46.15 | landley | gcc 4.2 is a nightmare. gcc 4.3 and later are increasing levels of lovecraftian horror. |
| 20:46.20 | slashbeast | do you plan to switch to llvm in aboriginal someday? |
| 20:46.27 | landley | I'm tracking it. |
| 20:46.51 | slashbeast | i remember swtich from 3.3.6 to 3.4 |
| 20:46.59 | slashbeast | that was nightmare. New ABI. |
| 20:47.03 | landley | What I'd _really_ like to do is ask Fabrice Bellard how much he'd charge to glue qemu's TCG onto TCC as a code generator and get tccboot working with current kernels, and then do a kickstarter for that amount. |
| 20:47.33 | landley | I don't care if the result is efficient, I just want a simple compiler that WORKS. |
| 20:47.38 | landley | Efficient can be natively built on target. |
| 20:47.45 | slashbeast | +1 |
| 20:50.20 | ohsix | add a compiler to toybox, add build flags that change the language \m/ |
| 20:50.33 | landley | ohsix: no. |
| 20:53.23 | landley | I am aware that SUSv4 specifies c99 and ld and such as commands. They are not my problem. |
| 20:53.33 | ohsix | that was more a joke about find from earlier than actually including a compiler |
| 20:54.34 | slashbeast | I cant stand the mozilla stupindess |
| 20:54.39 | slashbeast | they bumping firefox and thunderbird together. |
| 20:54.40 | slashbeast | also, when I have started tb, I cant start ff |
| 20:54.45 | slashbeast | because it says that ff is already running, when ff is running, tb refuse to sart |
| 20:55.16 | ohsix | that sounds more like a packaging problem |
| 20:56.18 | slashbeast | it used to work |
| 20:56.18 | slashbeast | sometimesI can start ff when tb is running |
| 20:56.18 | ohsix | tb and ff are just hosted on mozilla, so as far as producing versions you go with the underlying thing |
| 20:57.33 | slashbeast | also, chromium require libudev to compile. |
| 20:57.35 | ohsix | did thunderbird start heeding .mozilla/firefox/<currentprofile>/profile.lock? :D |
| 20:57.44 | slashbeast | so I can't drop udev from my system even that I dont use it. |
| 20:57.44 | landley | slashbeast: do a libudevstub. |
| 20:58.07 | slashbeast | I may eventualy |
| 20:58.08 | ohsix | or build chromium without webcam (and what else?) support |
| 21:02.46 | slashbeast | libudev dep is *hardcoded* |
| 21:02.46 | slashbeast | there is if statement, if os=bsd, then it does need libudev. |
| 21:28.56 | *** join/#edev purl (~purl@rikers.org) |
| 21:28.56 | *** topic/#edev is Discussion of embedded systems (Hardware/Software/Linux/Commercial Device Repurposing) : Busybox and Buildroot : http://www.elinux.org |
| 21:31.00 | *** join/#edev Spida (~timo@ns1.spinnennetz.org) |
| 21:52.47 | slashbeast | /usr/lib/gcc/x86_64-pc-linux-gnu/4.5.3/../../../../x86_64-pc-linux-gnu/bin/as: error while loading shared libraries: libc.so.0: cannot open shared object file: No such file or directory |
| 21:52.54 | slashbeast | win. |
| 22:24.27 | neale | I figured it out. |
| 22:24.44 | neale | Apparently you have to tell the kernel what *type* of ARM you have, and not just accept the defaults. |
| 22:24.53 | neale | So, I did that, and it boots now :) |
| 22:25.02 | neale | go figure, right? |
| 22:33.52 | Posterdati | please help: it is possible to use pointer to functions on sdcc for pic16f886? |
| 22:36.38 | roxfan | try it and see? |
| 22:39.30 | Posterdati | not working |
| 22:40.16 | roxfan | guess not then |
| 22:40.58 | Posterdati | well no errors, but wrong address when jumping at the function |
| 22:41.13 | roxfan | oh that's weird |
| 22:42.10 | Posterdati | using static declaration for struct containing the pointer to function works for jumping, but it is impossible to pass a parameter to the function itself |
| 22:43.00 | Posterdati | the value is trashed |
| 22:43.29 | roxfan | hmm i wonder if they use the same register for indirect jump |
| 22:43.35 | roxfan | check the asm? |
| 22:45.02 | Posterdati | not checked |
| 22:46.35 | Posterdati | that's why I hate microchip stuff |
| 22:47.24 | *** join/#edev av500 (~av500@b2b-46-252-131-98.unitymedia.biz) |
| 22:57.25 | Posterdati | microchips stuffs are completely unusable using freesoftware solutions |
| 22:59.51 | roxfan | you can always go back to asm :) |
| 23:05.08 | Posterdati | no |
| 23:16.11 | neale | is there a document anywhere on the tubes that explains how uboot works? |
| 23:16.30 | neale | the uboot wiki spends a long time on recipes for various things but I want to know what it is doing. |
| 23:17.45 | neale | actually maybe I should just explain what I'm trying to do. |
| 23:18.00 | neale | I'm trying to boot a squashfs initrd, which lives as a file in / |
| 23:18.34 | neale | I know how Linux early userspace works but I can't figure out where in RAM I'm supposed to ext2load my uRamdisk |
| 23:23.32 | roxfan | i think it rather depends on kernel support (boot options etc) than uboot itself |
| 23:25.12 | neale | unless I'm mistaken, the boot loader is what brings the initrd into RAM |
| 23:26.09 | roxfan | yes, but it's the kernel which then reads it |
| 23:26.15 | neale | so I need to figure out where in RAM to put this thing. When I bring up uboot, it's got a bunch of environment variables set already, and one of them is "loadaddr" which the boot.scr I got uses to load the kernel. But I have no idea where to put the ramfs. |
| 23:26.47 | roxfan | so it's probably either at some fixed address determined at compile time or some special boot option |
| 23:26.57 | roxfan | (kernel compile time, that is) |
| 23:27.00 | neale | no, it gets passed in on the kernel command line. |
| 23:27.10 | roxfan | the address? |
| 23:27.21 | neale | yeah, with initrd_start and initrd_end |
| 23:27.37 | neale | I mean, if I'm reading the wiki correctly. |
| 23:27.41 | roxfan | then it probably doesn't matter much as long as it doesn't overwrite the kernel image |
| 23:28.01 | neale | so there's no strategy about where to load that in? I just pick a number I like? |
| 23:28.30 | roxfan | well you need to check the ram layout of your board |
| 23:28.39 | roxfan | maybe there's slower and faster ram |
| 23:28.58 | roxfan | and you need to check where the kernel gets loaded and decompresses itself |
| 23:29.11 | neale | golly. |
| 23:29.56 | roxfan | you don't have any example BSPs or something? |
| 23:31.01 | neale | I don't know what a BSP is. I have only the vaguest idea what I'm doing here :) |
| 23:31.21 | roxfan | board support package |
| 23:31.25 | roxfan | what's your target? |
| 23:31.32 | neale | Something called a "TrimSlice" |
| 23:31.44 | neale | It's a Tegra 2 board, apparently. |
| 23:32.19 | neale | I'm trying to see if we can use these things as thin clients, and the ARM boot process is not something I have any experience with. |
| 23:33.20 | roxfan | you should probably inspect the factory configuration, it should have some reasonable defaults |
| 23:33.43 | neale | would the "factory configuration" be these uboot environment variables that are set? |
| 23:33.59 | roxfan | yeah |
| 23:34.12 | neale | okay, so there's an address set for the kernel, but nothing for an initrd. |
| 23:34.31 | roxfan | hmm probably it has an attached initrd |
| 23:34.32 | neale | there's also this: "mem=mem=384M@0M nvmem=128M@384M mem=512M@512M" |
| 23:35.33 | neale | the like example OS it comes with is Ubuntu |
| 23:35.50 | roxfan | what's the kernel load addr? |
| 23:36.21 | neale | loadaddr=0x40800 |
| 23:36.33 | neale | er, 408000 |
| 23:37.01 | roxfan | okay, so they load the kernel at around 4M |
| 23:37.04 | neale | where are these environment variables coming from? |
| 23:37.18 | neale | are they squirreled away in CMOS somewhere? |
| 23:37.32 | roxfan | there's a non-volatile storage used by uboot (a flash or something) |
| 23:38.29 | roxfan | maybe you could try putting your ramdisk somewhere in the second half of ram |
| 23:38.29 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 23:38.50 | roxfan | a lot of things depend on how the kernel manages the ram after booting... |
| 23:39.39 | *** join/#edev mpk (mp@aggr.com) |
| 23:40.46 | neale | I suspect the Ubuntu kernel has an initramfs built into it. |
| 23:41.03 | roxfan | <roxfan> hmm probably it has an attached initrd |
| 23:41.11 | neale | the cpio thing, yeah. |
| 23:42.02 | neale | I didn't realize the RAM on these things might be different across different address spaces. |
| 23:42.20 | neale | maybe it would be best if I just dd'ed the initrd out to a partition |
| 23:46.39 | neale | sorry if that sounded like I was ignoring your advice. I really appreciate the help, I'm just learning a ton very quickly and haven't assembled it all in my head yet. |