| 00:00.00 | Tiersten | how about setting them to moderated? |
| 00:00.16 | rwhitby-bday | nope - wildly offtopic gets no mercy |
| 00:00.33 | Tiersten | heh |
| 00:01.16 | rwhitby-bday | bbiab |
| 00:01.18 | jacques | /dev/mtdblock4 5248 644 4604 12% /mnt/userdisk |
| 00:01.29 | jacques | looks like 644kB is the majoc number |
| 00:01.38 | jacques | magic that is |
| 00:02.13 | Tiersten | Oh |
| 00:02.17 | Tiersten | I know why it's doing that |
| 00:02.27 | jacques | do tell :-) |
| 00:02.30 | Tiersten | 640KB is the magic number BTW |
| 00:02.42 | Tiersten | JFFS2 reserves 5 x erase block for GC |
| 00:03.00 | jacques | ouch |
| 00:03.26 | *** join/#nslu2-linux CIA-7 (~CIA@to.je.spocco.com) |
| 00:03.55 | Tiersten | You can tweak that value if you really want to |
| 00:04.07 | Tiersten | It's just set at 5 because it should guarantee that it can do GC |
| 00:05.01 | jacques | tweak in the code I assume as it's not an option (that I can see) to mkfs.jffs2 |
| 00:05.53 | jacques | unless it's somewhere in /proc - now that would be kinda cool |
| 00:07.18 | Tiersten | You can't change it without recompiling I think... |
| 00:07.23 | Tiersten | Not looked too closely |
| 00:07.36 | jacques | ok - still it's good to know |
| 00:08.02 | jacques | I wonder what size would be appropriate for our partitions |
| 00:08.52 | Tiersten | It's just a fixed number. It doesn't get bigger as you increase the size of the partition |
| 00:10.11 | jacques | I know, let's just disable GC and recover all the blocks :-P |
| 00:10.51 | Tiersten | erm |
| 00:11.04 | Tiersten | JFFS2 won't work very well if you do that... |
| 00:11.54 | jacques | i figured :-) |
| 00:12.29 | Tiersten | Once you use up all blocks then you can't change anything apart from delete |
| 00:12.37 | Tiersten | That's why it needs the reserved blocks |
| 00:12.53 | Tiersten | It writes into an empty block and once that has been confirmed okay then it'll delete the old one |
| 00:13.23 | jacques | interesting |
| 00:13.30 | Tiersten | It reserves multiple ones because they're used for different things. That what I understand anyway |
| 00:14.01 | jacques | I remember on the tuxscreen, when the partition was nearly full, writes would take a lot of cpu |
| 00:14.21 | Tiersten | It was doing GC |
| 00:14.40 | Tiersten | This may prove a bit of a issue since we don't really have that much flash |
| 00:14.51 | kergoth | :P |
| 00:14.58 | Tiersten | heh |
| 00:15.10 | jacques | but then I wouldn't get to have this lovely conversation with Tiersten |
| 00:16.24 | jacques | kergoth, you referring to this site: http://www.linux-mtd.infradead.org/ ? |
| 00:16.47 | Christopher | YEAH!! |
| 00:16.55 | Tiersten | ? |
| 00:16.58 | jacques | Christopher, ? |
| 00:17.01 | Christopher | I got upslug to work, |
| 00:17.03 | Christopher | :-) |
| 00:17.06 | Tiersten | ah |
| 00:17.10 | jacques | Nice :-) |
| 00:17.19 | Christopher | I need to clean it up, but I just reflashed my box |
| 00:17.33 | kergoth | jacques: yes |
| 00:17.48 | Christopher | anyway, I was just excited |
| 00:18.17 | jacques | Christopher, sounds like you have good cause to be excited |
| 00:18.30 | Christopher | thanks |
| 00:18.38 | *** join/#nslu2-linux christopher_logg (~christoph@128-193-245-216.resnet.oregonstate.edu) |
| 00:21.23 | jacques | whoah - that was a great move captain archer just did |
| 00:22.04 | rwhitby-bday | Christopher: congrats - you are now the official owner of upslug :-) |
| 00:22.30 | rwhitby-bday | jacques: so you've flashed your other slug, or did you buy a third ? |
| 00:22.58 | jacques | rwhitby-bday, I'm flashing the second one :-) |
| 00:23.06 | Christopher | rwhitby-bday: thanks |
| 00:23.19 | jacques | no place in town has them in stock - radio shack caries them but are out |
| 00:23.30 | rwhitby-bday | kergoth: which particular PDF on that site? |
| 00:24.02 | jacques | there's a link to this page: http://sources.redhat.com/jffs2/ which has a pdf |
| 00:24.19 | kergoth | rwhitby-bday: you want the one thats linked from there |
| 00:24.20 | kergoth | .. right, that |
| 00:24.23 | jacques | more than one actually |
| 00:24.26 | kergoth | i always forget the jffs2 url |
| 00:24.51 | [g2]-away | kergoth any idea what's up with cvs.handhelds.org ? |
| 00:24.58 | jacques | I know I know |
| 00:25.19 | jacques | CRL building is without power due to schedules building maintenance |
| 00:25.49 | kergoth | yeah, kinda annoying |
| 00:25.51 | rwhitby-bday | that PDF says that 4 erase sectors should be enough |
| 00:25.58 | kergoth | particularly since i use kergoth@handhelds.org is my primary personal contact point |
| 00:26.05 | rwhitby-bday | but it's not been proven |
| 00:26.05 | kergoth | so i'm not gonna get jack for email till it comes back |
| 00:26.06 | kergoth | hehe |
| 00:26.43 | jacques | damn for some reason pasting into the input field of xchat is not working right |
| 00:26.45 | Tiersten | rwhitby: It's set at 5 because that is a safe number. In theory you can reduce it by 1 or 2 but you might run into problems when it's full |
| 00:27.06 | jacques | pasting from copies of xchat text taht is |
| 00:27.24 | jacques | (from other tabs) |
| 00:27.58 | jacques | hmm actually I just cant copy and paste part of the topic from hh.o to input |
| 00:28.46 | jacques | anyway I was trying to paste the notice - they expect if back by tomorrow morning |
| 00:31.00 | rwhitby-bday | well - that answers the question about 1 jffs2 partiton or two - we can't afford to have 2 cause that will cost more than the size of the kernel |
| 00:31.24 | rwhitby-bday | we were worrying about the 128Kb ramdisk! |
| 00:31.40 | rwhitby-bday | and nobody realised that jffs2 costs 640k anyway |
| 00:31.46 | jacques | good point - yeah I had the same thought about the 128k thing - even 4 blocks is 512kB - half a meg :-( |
| 00:32.15 | rwhitby-bday | the advantages of jffs2 are worth the 640k loss, but only one of them .... |
| 00:32.22 | jacques | I concur |
| 00:32.49 | jacques | too bad they can't share the GC space |
| 00:33.15 | jacques | i guess that makes no sense :-\ |
| 00:33.39 | rwhitby-bday | we'll have to do the ro/rw thing some other way to make it not-easy for people to screw up their unslung system files. |
| 00:34.14 | rwhitby-bday | back later |
| 00:34.21 | jacques | ok |
| 00:50.33 | rwhitby-bday | ok, so what do we need for the 128Kb ramdisk which can run /linuxrc and then switch between unslung disk root, jffs2 root or ramdisk root? |
| 00:51.28 | jacques | I would say static busybox with only minimal ash, mount, and ... ? |
| 00:51.48 | rwhitby-bday | yep - do we call it "switchbox" ? |
| 00:51.56 | jacques | heh |
| 00:54.34 | *** join/#nslu2-linux christopher_logg (~christoph@128-193-245-216.resnet.oregonstate.edu) |
| 00:55.56 | CIA-7 | 03edmondsc * 10upslug/upslug.c: Check in of upslug, it now has a --upload command that will allow the firmware to be flahsed, it still needs cleaned up and tested more. |
| 00:57.00 | dyoung-away | Thanks Christopher! |
| 00:57.08 | rwhitby-bday | Christopher: congrats! |
| 00:57.12 | Christopher | np |
| 00:57.17 | Christopher | It was fun |
| 00:57.37 | Christopher | I logged the packets from the windows upgrad utility and then tried to recreate the same behavior |
| 00:57.40 | rwhitby-bday | it always is when it works :-) |
| 00:57.58 | Christopher | reverse engineering :-) |
| 00:58.25 | Christopher | k, I've got to go |
| 00:58.25 | dyoung-away | Cool, I'm glad someone did something with that Sercomm bit. I didnt get to adding it to the wiki yet. |
| 00:58.26 | rwhitby-bday | I couldn't get the windows upgrade utility that dyoung found to work for me - it got a hardware type mismatch, and also seemed to screw up other devices on my lan temporarily |
| 00:58.53 | Christopher | test it out and tell me about any issues |
| 00:58.57 | Christopher | bye all |
| 00:59.29 | dyoung-away | It gave me H/W mismatch until I replaced my redboot. |
| 00:59.40 | rwhitby-bday | this is with stock redboot |
| 00:59.45 | Christopher | hopefully mine isn't that intelligent :-) |
| 00:59.46 | dyoung-away | I'll test it later, I gotta get back to the smokehouse. |
| 00:59.57 | Christopher | k |
| 01:01.05 | dyoung-away | rwhitby, I dont know what exact bit its looking for when it says "H/W Mismatch". All I can describe is the stuff I had to do to make it go. It seemed to want to upgrade my ProdSlug. I wonder whats differnt about your DevSlug.... |
| 01:01.30 | rwhitby-bday | dunno |
| 01:01.31 | CIA-7 | 03rwhitby 07 * r1.2547 10packages/busybox/ (3 files in 2 dirs): Added switchbox - a minimal busybox for switching between root device options in /linuxrc on an NSLU2 device |
| 01:01.52 | rwhitby-bday | jacques: there's an initial switchbox, which is just a copy of slingbox - feel free to cut down the defconfig to what you think we need |
| 01:02.17 | jacques | rwhitby-bday, cool |
| 01:02.37 | jacques | i think I already have such a config |
| 01:03.02 | glc-away | Does this GC issue make cramfs look good for the readonly root? |
| 01:03.24 | jacques | the config I used to get this: 94208 Nov 11 21:59 tiny.squashfs* |
| 01:04.14 | rwhitby-bday | glc: dunno - we wanted a rootfs which was mounted ro, but could be remounted as rw easily to change it |
| 01:05.12 | rwhitby-bday | don't confuse the switchbox ramdisk with the rootfs - we need to have the 128Kb ramdisk anyway (to retain compatibility with stock redboot, and also cause it allows us to switch easily between root filesystem options) |
| 01:05.46 | rwhitby-bday | the switchbox ramdisk should be as compressed as possible. |
| 01:05.48 | rwhitby-bday | back later |
| 01:06.48 | *** join/#nslu2-linux soulwarrior (~konversat@83.222.38.101) |
| 01:08.15 | glc | You ruled out 2 jffs2 because of the GC blocks, and were worried about neophytes damaging a single jffs2 mounted rw. |
| 01:09.01 | glc | I was just thinking that essentials in a cramfs mounted ro with a jffs2 mounted rw in /usr might be workable for uprades, |
| 01:09.51 | glc | yet make for a safe reboot if the jffs was accidentally erased. Those pivoting to hdd won't care. |
| 01:11.44 | rwhitby-bday | glc: that was my original idea by Tiersten convinced me that we could do it with two jffs2 partitions instead. So maybe that need rethinking in light of the 640Kb cost of each jffs2 partitions |
| 01:12.32 | rwhitby-bday | jacques: we can probably put the image creation stuff in switchbox.oe, instead of having a whole new package, just for a /linuxrc and switchbox |
| 01:13.12 | jacques | rwhitby-bday, refresh my memory - what do I do after bk citool in order to checkin the changeset ? |
| 01:13.46 | rwhitby-bday | commit |
| 01:14.21 | jacques | that's what citool does |
| 01:14.26 | jacques | bk push ? |
| 01:14.30 | jacques | it's been too long |
| 01:14.47 | jacques | what do you do after commit ? |
| 01:15.07 | *** join/#nslu2-linux rwhitby-treo (~upirc@203.202.55.134) |
| 01:15.32 | rwhitby-bday | bk send to nslu-changesets@lists.sf.net |
| 01:15.37 | rwhitby-bday | then bk push |
| 01:15.50 | rwhitby-bday | VoodooZ_log: has a cheat sheet on the wiki |
| 01:16.03 | *** join/#nslu2-linux Tiersten (tman@ideal.trejan.com) |
| 01:16.12 | jacques | yeah bk send still does not work for me |
| 01:16.36 | dyoung-salmon | dont forget to reparent. |
| 01:16.41 | jacques | yeah I was looking for that but I cant find it |
| 01:16.53 | dyoung-salmon | http://www.nslu2-linux.org/wiki/OpenEmbedded/OECmdCheatSheet |
| 01:17.01 | jacques | ok but how does one find it? |
| 01:17.22 | jacques | it's not linked from info or howto |
| 01:17.32 | dyoung-salmon | OpenEmbedded Builds -> OECmdCheatSheet. |
| 01:19.10 | jacques | ok it's pushed - I dunno why CIA notification also does not work for me on bk |
| 01:20.53 | pb_ | the cia mail triggers for bk run locally (unlike with cvs, where they run on the server). check your email configuration on the machine where you do checkins. |
| 01:21.15 | jacques | my machine cannot send email |
| 01:21.23 | pb_ | that would be a problem |
| 01:22.38 | jacques | i spent hours on trying to get it to work a month or so ago and gave up |
| 01:22.50 | jacques | mail makes my heaad want to explode |
| 01:22.55 | pb_ | heh |
| 01:23.08 | pb_ | oh well. that'll be why cia isn't working for you, anyway |
| 01:23.19 | rwhitby-treo | jacques: you need to fix that - no email to changesets and no CIA notifications is no good |
| 01:23.36 | jacques | how do I do it? |
| 01:23.59 | Ksilebo | Messenger Pidgeons. |
| 01:24.12 | *** join/#nslu2-linux soulwarrior (~konversat@83.222.38.101) |
| 01:24.30 | Ksilebo | Telekenesis may be faster. |
| 01:25.24 | *** join/#nslu2-linux Christopher (~Christoph@plo-cr1-212-105.peak.org) |
| 01:28.12 | jacques | that's one reason I got the fastmail account |
| 01:28.23 | jacques | but I keep getting "Relay access deined" |
| 01:29.00 | Ksilebo | Do you just need a simple imap account? |
| 01:29.35 | jacques | probably because ssmtp has no way of specifying username/password :-( |
| 01:29.45 | rwhitby-treo | did you upgrade to member to get smtp? |
| 01:30.03 | rwhitby-treo | that should be all you need |
| 01:30.10 | jacques | I got the $15 lifetime account - is that member? they have like 5 account types with similar names |
| 01:30.43 | jacques | I also need a MTA that 1) I can configure and 2) supports their authentication mechanisms |
| 01:30.43 | Ksilebo | I think thats what I got a while back. |
| 01:31.47 | jacques | rwhitby-treo, ssmtp CAN use an RSA cert - do you know how I set that up with fastmail ? (or anyone else know) |
| 01:32.51 | rwhitby-treo | nope |
| 01:33.14 | rwhitby-treo | the single payment lifetime account is the one I $&91 |
| 01:33.19 | rwhitby-treo | have |
| 01:33.53 | jacques | rwhitby-treo, what do you use as an MTA ? |
| 01:34.27 | rwhitby-treo | so, my birthday is half over, and still no sight of my present: a CSR-enabled OpenSlug in OE ... |
| 01:35.02 | jacques | hmm kergoth can't find the changeset? |
| 01:35.17 | rwhitby-treo | exim I think - whatever is default in Debian. but I don't smtp through fastmail |
| 01:35.45 | jacques | oh |
| 01:35.59 | jacques | BTW I have read their FAQ like four times |
| 01:36.37 | jacques | they dont seem to support certs |
| 01:36.37 | rwhitby-treo | I smarthost through my isp |
| 01:36.49 | jacques | and my MTA *only* supports certs |
| 01:38.35 | rwhitby-treo | which MTA? |
| 01:40.17 | jacques | ssmtp |
| 01:44.57 | jacques | coool |
| 01:45.08 | jacques | why didnt I think of this before - rwhitby-treo reminded me of it |
| 01:45.14 | jacques | I can use my isp |
| 01:45.21 | jacques | (normally I use them for a little as possible) |
| 01:46.42 | jacques | now to set it up on all three of my build machines |
| 01:56.21 | *** join/#nslu2-linux Ksilebo (~russ@dialup-4.225.219.249.Dial1.Denver1.Level3.net) |
| 02:04.01 | jacques | OK! I think I have all three machines sending through my isp :-) |
| 02:04.19 | jacques | now I feel like making a gratuitous change to test (already tested manually but want to test bk) |
| 02:05.07 | rwhitby-treo | add linuxrc to switchbox |
| 02:06.24 | jacques | ok |
| 02:07.42 | rwhitby-treo | the script that is ... |
| 02:07.53 | *** part/#nslu2-linux rwhitby-treo (~upirc@203.202.55.134) |
| 02:08.18 | jacques | hmm |
| 02:09.50 | jacques | add it where? not sure what you mean - write the linuxrc that does what we want ? |
| 02:11.37 | rwhitby-bday | take the one from unslung ramdisk for a start |
| 02:12.53 | rwhitby-bday | then add it just like the OE busybox adds files |
| 02:13.58 | jacques | OK I'm trying not to be dense, I found packages/nslu2-binary-only/unslung-ramdisk-2.3r25/linuxrc |
| 02:14.27 | jacques | but I don't understand "then add it just like the OE busybox adds files" |
| 02:14.56 | peteru-home | Just add a README file - that's pretty low risk :) |
| 02:17.56 | peteru-home | anyone recall the proper #ifdef expression to check for KERNEL 2.6 or higher? |
| 02:18.57 | jacques | not I :-\ |
| 02:20.42 | Tiersten | #if LINUX)VERSION_CODE >= KERNEL_VERSION(2,6,0) ? |
| 02:20.45 | Tiersten | uh |
| 02:20.48 | Tiersten | ) should be _ |
| 02:22.04 | peteru-home | Thanks Tiersten |
| 02:22.13 | jacques | rwhitby-bday, you mean something like: "install -m 0755 ${WORKDIR}/linuxrc ${D}/" ? |
| 02:23.59 | jacques | put that into switchbox_1.00.oe ? (and add a linuxrc into packages/busybox/switchbox-1.00) ? |
| 02:27.24 | rwhitby-bday | yep |
| 02:28.54 | jacques | ok |
| 02:29.08 | peteru-home | is it just my environment, or is it expected that the headers in cross/armeb-linux/include/linux are for kernel 2.6, even though we build 2.4 ? |
| 02:29.21 | kergoth | yes. |
| 02:29.28 | kergoth | those headers are only used to build the C library |
| 02:30.03 | peteru-home | so which headers are used when doing armeb-linux-gcc ? |
| 02:30.05 | jacques | why won't emacs let me edit a SCCS controlled file? sometimes it does and other times it doesn't |
| 02:30.08 | kergoth | peteru-home: huh? |
| 02:30.32 | kergoth | if a userspace application relies on <linux/*> or <asm/*> as matching the runtime kernel, they're *broken* |
| 02:30.37 | kergoth | period. |
| 02:30.38 | kergoth | with very few exceptions |
| 02:30.44 | kergoth | (namely, glibc, wireless-tools, pcmcia-cs, ..) |
| 02:30.57 | rwhitby-bday | jacques: C-x C-q is what I use to do bk edit |
| 02:31.34 | peteru-home | I'm writing for the usbdevfs and the name of a member has changed from MaxPower to bMaxPower |
| 02:31.49 | peteru-home | I don't see anything broken with catching that case |
| 02:32.04 | kergoth | you *cannot* rely on those headers matching the kernel you're going to run on. |
| 02:32.19 | kergoth | if you need *real* kernel headers, you have to add that path explicitly to your include search path |
| 02:32.30 | kergoth | (i.e. -I${KERNELDIR}/include in the Makefile) |
| 02:32.31 | peteru-home | The headers describe data sent over the USB bus |
| 02:32.40 | kergoth | you're missing the point. |
| 02:32.43 | peteru-home | as defined by USB2.0 spec |
| 02:32.55 | kergoth | the C library kernel headers are not guaranteed to match runtime. |
| 02:33.00 | kergoth | in oe or on any modern linux distribution. |
| 02:33.01 | kergoth | period. |
| 02:33.13 | peteru-home | it is only the name of the member that changes, not it's bit definition which is independent of the Linux kernel. |
| 02:33.17 | kergoth | generally the Makefile for someone that really needs the real kernel headers would d something like the following: |
| 02:33.30 | kergoth | KERNELDIR=/lib/modules/$(shell uname -r)/build |
| 02:33.32 | peteru-home | the kernel only passes the data received from usb, it does not even correct for endian issues |
| 02:33.39 | kergoth | CFLAGS += "-I$(KERNELDIR)/include" |
| 02:33.59 | kergoth | and oe defines a variable that points to the real kernel headers, that you're welcome to use to override such things. |
| 02:34.15 | kergoth | let's take an example |
| 02:34.22 | kergoth | an application that uses the framebuffer. |
| 02:34.27 | kergoth | it needs defines for the ioctls on /dev/fb* |
| 02:34.40 | kergoth | yet for it to include <linux/fb.h> would be bad. |
| 02:34.50 | kergoth | for itd be relying on that matching the running kernel, which isnt guaranteed. |
| 02:35.05 | kergoth | generally the best solution in that particular case would be to bring the ioctl definitions local, into its own headers. |
| 02:35.11 | rwhitby-bday | jacques: any reason why switchbox needs DEVPTS ? |
| 02:35.20 | rwhitby-bday | and SUID? |
| 02:36.14 | kergoth | peteru-home: so you have two options. 1) bring the definitions local, or 2) explicitly have a -I to the *real* kernel headers, which you'd override in the .oe. |
| 02:36.20 | peteru-home | kergoth: I am NOT using the kernel headers to talk to the kerel, I am using it to talk to a remote USB device. The fact that the header is defined in the kernel space is just a covenience for me, so that I don't have to ship my own version of the same thing. The data structure is set in stone by the USB2.0 spec. I take your point in general case where you have a data structure defined by the kernel, but that is not the case here. |
| 02:36.26 | jacques | rwhitby-bday not that I can think of |
| 02:36.45 | kergoth | it doesnt matter. if you're relying on the fact that the linux/* headers must be 2.4, your application is broken. |
| 02:36.48 | kergoth | period. |
| 02:37.34 | CIA-7 | 03jacques 07 * r1.2549 10packages/busybox/ (switchbox-1.00/linuxrc switchbox_1.00.oe): added linuxrc and linuxrc install to switchbox (linuxrc is just a copy of the unslung-ramdisk one ATM) |
| 02:37.34 | jacques | yay |
| 02:37.58 | peteru-home | nope, I am saying that if it's 2.4 the member is "MaxPower" and in 2.6 and later it is called "bMaxPower" |
| 02:38.22 | kergoth | then use the kernel version define to conditionally alter your behavior |
| 02:38.31 | rwhitby-bday | jacques: do we need INIT support? Do we want this to be able to do anything other than /linuxrc? Same for HALT, REBOOT and POWEROFF .. ? |
| 02:38.33 | peteru-home | that's EXACTLY what I did. |
| 02:38.37 | kergoth | good :) |
| 02:38.53 | kergoth | gee, ya cant tell i've been over this with a billion developers in the past, can you? |
| 02:38.53 | peteru-home | I think we both want the same thing :) |
| 02:38.56 | kergoth | :) |
| 02:39.06 | jacques | rwhitby-bday, I thought the halt reboot might be useful but I don;'t have a strong reason |
| 02:39.16 | jacques | and I think we do *not* need init |
| 02:39.32 | rwhitby-bday | jacques: do they work properly for the slug anyway? |
| 02:39.41 | peteru-home | my case really is benign: |
| 02:39.43 | peteru-home | #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,0) |
| 02:39.43 | peteru-home | <PROTECTED> |
| 02:39.44 | peteru-home | #else |
| 02:39.44 | peteru-home | <PROTECTED> |
| 02:39.44 | peteru-home | #endif |
| 02:40.01 | jacques | one thing that's going to be "challenging" is teaching oe to not install the shared uclibc libs (BTW - I think this only have a chance of working with uclibc) |
| 02:40.18 | jacques | rwhitby-bday, reboot does work |
| 02:40.19 | kergoth | ahh, cool |
| 02:40.20 | rwhitby-bday | I'll leave reboot - that and /bin/sh should allow us to at least get back to RedBoot without having to powercycle |
| 02:40.41 | rwhitby-bday | what does EXTRA_QUIET do ? |
| 02:40.58 | rwhitby-bday | and does LFS support increase the size of busybox at all ? |
| 02:41.13 | jacques | I wondered about the LFS thing myself - I dunno |
| 02:41.30 | rwhitby-bday | I'll turn it off |
| 02:41.39 | rwhitby-bday | sane for FDISK_SUPPORT_LARGE_DISKS |
| 02:42.22 | rwhitby-bday | building now, then will push |
| 02:42.27 | jacques | cool |
| 02:42.50 | rwhitby-bday | this is not tested, right jacques :-) |
| 02:43.15 | rwhitby-bday | don't go flashing it ..... ;-) |
| 02:43.45 | jacques | rwhitby-bday, d`oh! :-P |
| 02:44.07 | rwhitby-bday | 75708 for switchbox |
| 02:44.41 | jacques | nice |
| 02:44.47 | jacques | nice nice |
| 02:44.53 | jacques | how stripped is it? |
| 02:45.43 | rwhitby-bday | only what I said in here as far as feature stripping |
| 02:45.59 | jacques | i remember on the tuxscreen we had a long line of options to strip to get things as small as possible - but that might have been mailny for the shared libs |
| 02:46.07 | jacques | sorry I meant binary strip |
| 02:46.43 | rwhitby-bday | file tmp/work/switchbox-1.00-r3/install/switchbox/bin/switchbox |
| 02:46.44 | rwhitby-bday | tmp/work/switchbox-1.00-r3/install/switchbox/bin/switchbox: ELF 32-bit MSB executable, ARM, version 1 (ARM), for GNU/Linux 2.4.0, d\ |
| 02:46.44 | rwhitby-bday | ynamically linked (uses shared libs), stripped |
| 02:46.53 | rwhitby-bday | ls -l tmp/work/switchbox-1.00-r3/install/switchbox/bin/switchbox |
| 02:46.54 | rwhitby-bday | -rwxr-xr-x 1 rwhitby rwhitby 75708 Nov 14 13:17 tmp/work/switchbox-1.00-r3/install/switchbox/bin/switchbox |
| 02:46.59 | jacques | hmg |
| 02:47.03 | jacques | why is it no tstatic ? |
| 02:47.11 | rwhitby-bday | hmm - we told busybox STATIC ... |
| 02:47.21 | jacques | I dont think we can reach our goal otherwise |
| 02:47.26 | *** join/#nslu2-linux Ksilebo|WARK (~russ@dialup-4.225.219.249.Dial1.Denver1.Level3.net) |
| 02:47.29 | rwhitby-bday | I gotta go out soon - so I will push this and you can continue |
| 02:47.37 | jacques | ok |
| 02:47.47 | jacques | I'll try to find out why it's not building it statis |
| 02:47.52 | jacques | static |
| 02:50.04 | rwhitby-bday | think you need to add linuxrc to SRC_URI ... |
| 02:50.14 | CIA-7 | 03rwhitby 07 * r1.2550 10packages/busybox/switchbox_1.00.oe: |
| 02:50.16 | CIA-7 | Merge bk://nslu2-linux@nslu2-linux.bkbits.net/packages |
| 02:50.18 | CIA-7 | into home.(none):/home/openembedded/packages |
| 02:50.25 | CIA-7 | 03rwhitby 07 * r1.2548.1.1 10packages/busybox/ (switchbox-1.00/defconfig switchbox_1.00.oe): Stripped more features from switchbox |
| 02:54.11 | rwhitby-bday | ok, back much later - let's see if my birthday present arrives in the meantime ,,, :-) |
| 02:54.24 | jacques | have fun rwhitby-bday! |