**** BEGIN LOGGING AT Fri Oct 07 02:59:56 2005
Oct 07 05:31:27 <CIA-8> 03florian 07org.openembedded.dev * r209a6256... 10/packages/gpe-ownerinfo/gpe-ownerinfo_0.28.bb: gpe-ownerinfo: Add new release.
Oct 07 05:41:37 <CIA-8> 03rw 07org.openembedded.dev * r634dd1df... 10/conf/bitbake.conf: bitbake.conf: add FREEBSD_MIRROR
Oct 07 05:41:40 <CIA-8> 03rw 07org.openembedded.dev * rcaf25ee1... 10/packages/rpm2cpio/rpm2cpio-native_1.2_2.bb: rpm2cpio: fix SRC_URI (uses newly introduced FREEBSD_MIRROR). fixes Bug #338
Oct 07 07:06:17 <CIA-8> 03koen 07org.openembedded.dev * r3b01989f... 10/packages/gpe-login/gpe-login_0.83.bb: packages/gpe-login/: add gpe-login 0.83
Oct 07 07:18:42 <dwery> NAiL: ping
Oct 07 07:20:31 <NAiL> pong
Oct 07 07:21:25 <CIA-8> 03koen 07org.openembedded.dev * r9eacd7bb... 10/packages/btscanner/btscanner_1.0.bb: packages/btscanner/btscanner_1.0.bb: fix compile with recent bluez headers, patch adapted from https://www.rocklinux.net/submaster/data//2005/06/0909132213173.patch
Oct 07 07:21:27 <CIA-8> 03koen 07org.openembedded.dev * rd158643d... 10/packages/btscanner/btscanner-1.0/0909132213173.patch: packages/btscanner/btscanner-1.0/0909132213173.patch: don't forget to commit the patch too.....
Oct 07 07:24:12 <dwery> NAiL: how's going on with the ds101?
Oct 07 07:25:51 <NAiL> just got up :P
Oct 07 07:26:52 <dwery> :)
Oct 07 07:26:58 <dwery> you want me to check the diffs?
Oct 07 07:27:33 <NAiL> yeah, I'm just reducing the size a bit ;)
Oct 07 07:28:25 <dwery> ok.. when you're done just put it in a place where I can wget it
Oct 07 07:28:34 <NAiL> sure
Oct 07 07:37:14 <NAiL> hmm
Oct 07 07:37:25 <NAiL> 2mb of the diff is MDocSys stuf
Oct 07 07:44:40 <dwery> ok
Oct 07 07:44:47 <dwery> that's good :)
Oct 07 07:47:15 <NAiL> http://osx.repvik.org/syno.diff.gz
Oct 07 07:47:36 <NAiL> There's probably a lot that can be removed, but I don't know what :P
Oct 07 07:48:07 <dwery> ok
Oct 07 07:48:36 <dwery> can you tell me which chip handles the ide ?
Oct 07 07:49:08 <NAiL> two sec
Oct 07 07:49:12 <dwery> found it
Oct 07 07:49:17 <NAiL> ah
Oct 07 07:49:26 <NAiL> all the hardware is listed on the wiki ;)
Oct 07 07:50:27 <dwery> ide-iops.c
Oct 07 07:50:31 <dwery> could be interesting
Oct 07 07:52:04 <NAiL> hmm.. why?
Oct 07 07:52:24 <NAiL> Isn't that for the ds101g+? 
Oct 07 07:52:34 <dwery> ops.
Oct 07 07:52:41 <dwery> the kernel; is the same for both?
Oct 07 07:52:46 <NAiL> yeah
Oct 07 07:52:57 <dwery> ok, let me check the wiki then :)
Oct 07 07:54:31 <dwery> is the driver the aec62xx or it821x?
Oct 07 07:56:11 <NAiL> AEC
Oct 07 07:57:26 <dwery> ok
Oct 07 07:59:37 <dwery> ixdp425-pci.c
Oct 07 07:59:41 <dwery> has the interrupt configuration
Oct 07 08:00:45 <dwery> i think you have to get the dmesg output of a working ds101
Oct 07 08:01:01 <dwery> and modify the 2.6 kernel to get the same irq mappings
Oct 07 08:01:07 <dwery> there are already debug stamentents in place
Oct 07 08:01:59 <NAiL> http://www.nslu2-linux.org/wiki/DS101/DS101Rev3BootLog <-- dmesg output ;)
Oct 07 08:02:48 <dwery> ixdp425_map_irq: PCI slot 0 , pin 1, assigned irq -1 ixdp425_map_irq: PCI slot 1 , pin 1, assigned irq 28 ixdp425_map_irq: PCI slot 2 , pin 1, assigned irq 27 ixdp425_map_irq: PCI slot 2 , pin 2, assigned irq 27 ixdp425_map_irq: PCI slot 2 , pin 3, assigned irq 26
Oct 07 08:03:01 <dwery> you have to check the 2.6 kernel assigns int in the same wat
Oct 07 08:03:02 <dwery> way
Oct 07 08:03:29 <dwery> you may want to run with dma off for initial debugging
Oct 07 08:04:14 <dwery> the kernel seems not so havily modified
Oct 07 08:16:03 <NAiL> mm
Oct 07 08:16:09 <NAiL> I gotta look at this later
Oct 07 08:17:45 <dwery> ok
Oct 07 08:17:47 <dwery> :)
Oct 07 08:19:27 <NAiL> yeah, it's friday :P
Oct 07 09:39:05 <dwery> anyone can tell me where to download devio ? google did not return anything useful...
Oct 07 09:40:03 <dwery> Jacmet: ping
Oct 07 09:47:38 <snotling> hello. I am trying to build an openslug kernel with BitBake but I've got an error with "quilt"
Oct 07 09:50:04 <eFfeM> snotling, i don't think you need quilt for building the kernel, just do bb openslug-kernel or bb openslug-image ; you might also be able to use make
Oct 07 09:54:03 <snotling> well, I actually do 'bb openslug-kernel'. But it seems to call quilt : + do_patchcleancmd /// ++ quilt applied ///+ quilt pop -a -R -f
Oct 07 09:54:23 <snotling> and then : "Usage: quilt [--trace[=verbose]] [--quiltrc=XX] command [-h] ..."
Oct 07 10:19:04 <eFfeM> snotling, was away for a while, haven't seen this myself
Oct 07 10:55:29 <dwery> nick dwery-away
**** BEGIN LOGGING AT Fri Oct 07 11:24:33 2005
Oct 07 13:10:30 <snotling> how can I edit the openslug-kernel's defconfig corr
Oct 07 13:10:44 <snotling> correctly and properly :-)
Oct 07 13:11:09 <CIA-8> 03rw 07org.openembedded.dev * r1b4d1903... 10/packages/bluez/ (4 files): bluez-utils: remove duplicate .inc. apply dbus.patch for -dbus variant.
Oct 07 13:20:54 <snotling> BTW I have a error when I try to build OpenSlug 2.7 and it reaches the kernel
Oct 07 13:21:08 <snotling> include/asm/glue.h:111:2: #error Unknown data abort handler type
Oct 07 13:21:11 <CIA-8> 03koen 07org.openembedded.dev * r2be87031... 10/packages/btscanner/btscanner_2.0.bb: packages/btscanner/btscanner_2.0.bb: add btscanner 2.0
Oct 07 13:21:19 <snotling> include/asm/page.h:92:2: #error Unknown user operations model
Oct 07 13:21:38 <snotling> what have I done wrong ?
Oct 07 13:28:10 <snotling> LOL, I reply to myself : it seems that I had f*cked my defconfig actualy...
Oct 07 13:32:08 <snotling> so now I have to find a way to produce a valid defconfig
Oct 07 13:36:25 <CIA-8> 03mreimer 07org.openembedded.dev * r4ba84c5c... 10/packages/tslib/ (tslib/h2200/tslib.sh tslib_cvs.bb): Add h2200 support
Oct 07 13:36:28 <CIA-8> 03rw 07org.openembedded.dev * r99a817d6... 10/packages/bluez/bluez-utils-dbus_2.21.bb: bluez-utils: re-enable dbus.patch for 2.21 as well.
Oct 07 13:53:45 <giel> hi there
Oct 07 15:29:05 <snotling> do I need udev to use an USB DVB-T device ?
Oct 07 15:37:37 <cheef_daniel> where are the sources for php-thttpd? I can't find them with google
Oct 07 15:37:50 <cheef_daniel> need it for openslug
Oct 07 15:56:52 <rwhitby> dwery-away: devio is on sf.net, project "devio".
Oct 07 16:26:42 <CIA-8> 03justinp 07org.openembedded.dev * r417b6b1a... 10/packages/lineak/lineak_0.8.3.bb: lineak: fix S
Oct 07 17:36:23 <CIA-8> 03jbowler 07org.openembedded.dev * rbc8ebf4b... 10/packages/ (6 files in 5 dirs): 
Oct 07 17:36:23 <CIA-8> ixp-osal_2.0.bb ixp4xx-csr_2.0.bb ixp400-eth_1.4.bb: fixed LE implementation
Oct 07 17:36:23 <CIA-8> This checkin includes a patch to add working le support in these modules
Oct 07 18:36:52 <dwery-away> rwhitby-away: thanks
Oct 07 18:50:07 <rwhitby-away> dwery-away: np - hopefully someday we will make it a debian package
Oct 07 18:50:20 <dwery-away> seems a pretty nice utility
Oct 07 18:50:49 <dwery-away> jacmet will need a way to install the debian le kernel on the nslu2
Oct 07 18:51:12 <dwery-away> with both le swap code
Oct 07 18:51:14 <dwery-away> and machine id
Oct 07 18:52:50 <rwhitby-away> yeah, devio is very powerful.
Oct 07 18:53:15 <dwery-away> having the same features in upslug2 could also be useful
Oct 07 18:53:18 <rwhitby-away> so how's the CVS repo working out?  I see that both you and jbowler are now commiting to it.
Oct 07 18:54:48 <rwhitby-away> is it increasing the speed and convergence of the development ?
Oct 07 18:56:08 <dwery-away> i can avoid doing wrong tar.gz's this way, yes. i think it's useful.
Oct 07 18:56:54 <dwery-away> once we are set with the le/mtd patches i'll send them upstream
Oct 07 18:59:22 <rwhitby-away> yep, that's been great collaboration.
Oct 07 18:59:31 <rwhitby-away> back later ...
Oct 07 18:59:57 <dwery-away> going to bed...
Oct 07 19:00:00 <rwhitby-away> BTW, you might be interested in this: http://lwn.net/Articles/154571/ 
Oct 07 19:00:07 <rwhitby-away> we're going to try and be there.
Oct 07 19:00:07 <dwery-zzz> ok, thansk
Oct 07 19:00:34 * rwhitby-away is gone
Oct 07 20:51:20 <CIA-8> 03jbowler 07org.openembedded.dev * r140ea215... 10/packages/linux/ (4 files in 3 dirs): (log message trimmed)
Oct 07 20:51:20 <CIA-8> nslu2-kernel_2.6.14-rc3.bb: better fix for LE flash access
Oct 07 20:51:20 <CIA-8> patch "packages/linux/nslu2-kernel/2.6.14/10-ixp4xx-copy-from.patch"
Oct 07 20:51:20 <CIA-8>  Change to use the cfi16 APIs to byte swap the 16 bit values read from
Oct 07 20:51:20 <CIA-8>  the flash as required.
Oct 07 20:51:22 <CIA-8> patch "packages/linux/nslu2-kernel/2.6.14/defconfig"
Oct 07 20:51:24 <CIA-8> patch "packages/linux/openslug-kernel-2.6.14-rc3/defconfig"
Oct 07 20:52:14 <CIA-8> 03jbowler * 10upslug2/ (configure.ac nslu2_image.cc nslu2_image.h upslug2.cc): 
Oct 07 20:52:14 <CIA-8> upslug2_8: add -L and -B options for various wacky ways of writing the
Oct 07 20:52:14 <CIA-8> flash on an LE system. The standard recommendation is -B - this writes an
Oct 07 20:52:14 <CIA-8> LE kernel but writes the flash data 'BE', the standard kernel config makes
Oct 07 20:52:14 <CIA-8> things such at the bytes read match those written.
Oct 07 20:57:02 <CIA-8> 03jbowler * 10kernel/2.6.14/10-ixp4xx-copy-from.patch: Better patch for flash endianness problems.
Oct 07 21:03:41 <cheef_daniel> I downloaded the Master Makefile for openslug, build openslug-image, but how do I customnize the kernel-modules, which should created?
Oct 07 21:04:17 <cheef_daniel> e.g I want to enable the stv680.ko module
Oct 07 21:09:41 <jbowler> cd tmp/work/openslug-kernel*; cp temp/run.do_conf* dconf; sed -i -e 's/oldconfig/menuconfig' -e 's/yes |//' dconf; ./dconf 
Oct 07 21:10:29 <jbowler> Then copy the result back into packages/linux/openslug-kernel-2.6.14-rc3/defconfig (or wherever it is, can't remember  what openslug builds by default)
Oct 07 21:14:17 <CheefDaniel> thx I will try it
Oct 07 21:22:02 <CheefDaniel> how do I copy it back?
Oct 07 21:22:35 <CheefDaniel> paths are correct
Oct 07 21:24:09 <CheefDaniel> you mean I should copy the defconfig folder to packages/linux/openslug-kernel-2.6.14-rc3/defconfig
Oct 07 21:24:16 <CheefDaniel> right?
Oct 07 21:28:06 <jbowler> I mean the *13/.config file which is what menuconfig generates - overwrite the defconfig with it
Oct 07 21:32:10 <cheef_daniel> ok, I have understand, first I edit the *13/.config-file and after that I do cp *13/.config-file to packages/linux...
Oct 07 21:34:53 <cheef_daniel> is it right now?
Oct 07 21:41:40 <jbowler> Yes, assuming the cp does overwrite the correct defconfig for your kernel.
Oct 07 21:44:43 <cheef_daniel> assuming I want to enable stv680, then I must only modify a parameter in the .config file?
Oct 07 21:45:25 <jbowler> Run menuconfig, use the interface.
Oct 07 21:45:46 <jbowler> It manages any dependencies correctly.
Oct 07 22:00:53 <cheef_daniel> the problem is that I can't use menueconfig without ncurses-dev, but on my system there is libnurses5 installed, and I must remove them to install ncurses-dev
Oct 07 22:01:58 <jbowler-away> Then edit defconfig in work/openslug-kernel* and run an unmodified temp/run.do_config* - this will copy the defconfig, sanitise it, and write *13/.config
Oct 07 22:02:30 <jbowler-away> Search defconfig for the thing you want, if in doubt read *13/.../Kconfig which contains help for each option
Oct 07 22:05:30 <cheef_daniel> ok thanks, now I have a point to search for
Oct 07 22:08:49 <cheef_daniel> gn8
Oct 07 23:23:28 <rwhitby-away> jbowler-away: nice work on the flash patches!
Oct 07 23:25:47 <jbowler-away> It's neat - I can build ucslugc arm, thumb uclibc, glibc le, be...
Oct 07 23:26:26 <jbowler-away> I'm not sure that thumb will work with either le or glibc, but all the others should.
Oct 07 23:27:14 <rwhitby-away> is the cvs repo allowing better collaboration between you and Alessandro?
Oct 07 23:28:14 <rwhitby-away> hopefully Jacmet will be able to build his LE debian kernel from the same patches
Oct 07 23:28:16 <jbowler-away> I think so, there's no ambiguity any longer about what the current state is.
Oct 07 23:28:51 <rwhitby-away> and when we start building official nslu2 debian kernels, we will get the patches from there too.
Oct 07 23:29:48 <jbowler-away> There's no way of publishing the Intel ixp stuff though - I put the patches into monotone, but even that is on the borderline.
Oct 07 23:30:08 <jbowler-away> ('cause they are 'licensed according to the original Intel license')
Oct 07 23:30:49 <rwhitby-away> jbowler-away: worst case we could always distribute 'ed' patches with no context :-)
Oct 07 23:32:27 <rwhitby-away> then there is no intel content at all in the ed script.
Oct 07 23:33:05 <jbowler-away> Well, there's actually almost as much as in a diff -c, since the patches often just change part of a line.  I don't think that's likely to be an issue.
Oct 07 23:33:23 <jbowler-away> I can't imagine Intel worrying about this, since the patch is useless on its own.
Oct 07 23:33:29 <rwhitby-away> yeah, I agree
Oct 07 23:34:00 <rwhitby-away> aren't there rumours of later versions of the IAL (not the microcode) being GPL ?
Oct 07 23:34:30 <jbowler-away> I was hoping they might do that as they had split the microcode out, but there is no sign of it.
Oct 07 23:34:51 <jbowler-away> The split may have just been to avoid the unnecessary overhead of duplicating the microcode in the kernel.
Oct 07 23:35:28 <rwhitby-away> jbowler-away: have you heard about this: http://lwn.net/Articles/154571/ ?
Oct 07 23:35:32 <jbowler-away> I wondered if it might be possible to persuade them to release a de-configured version under an MIT license.
Oct 07 23:35:51 <rwhitby-away> dunno whether you are interested or not, but a few of us are going to turn up to support the arm architecture
Oct 07 23:37:25 <jbowler-away> Would be difficult for them not to support it I would have thought...
Oct 07 23:38:36 <rwhitby-away> there's been talk of dropping arm and mipsel and s390 and even m68k
Oct 07 23:38:52 <jbowler-away> m68k?
Oct 07 23:39:13 <jbowler-away> I can understand them only supporting arm and x86
Oct 07 23:39:29 <jbowler-away> That would cover almost all the processors in the world.
Oct 07 23:39:29 <rwhitby-away> going by these figures (http://popcon.debian.org/) we (armeb) have more machines than arm and mipsel and m68k and s390.
Oct 07 23:41:57 <jbowler-away> Not very statiscally significant I think...
Oct 07 23:42:25 <rwhitby-away> yeah, but good for a jibe :-)
Oct 07 23:43:20 <jbowler-away> One of the bottom lines is that pretty much all the handhelds seem to be arm
Oct 07 23:46:37 <jbowler-away> ARM Ltd made announcements recently which suggest an attempt to push into markets requiring more power: http://www.arm.com/news/10548.html
Oct 07 23:46:57 <rwhitby-away> so can we take a virgin 2.6.14-rcX kernel source, just apply the patches from repo, and have a fully working system?
Oct 07 23:47:23 <jbowler-away> Apart from the NPE issue yes.
Oct 07 23:48:04 <jbowler-away> You need to futz the command line to boot of something appropriate (e.g. NFS), and you might have to fine tune the kernel to make it fit.
Oct 07 23:48:05 <rwhitby-away> BTW, we need to talk about the nslu2_defconfig that we need to push upstream (to go in arch/arm/configs)
Oct 07 23:48:54 <rwhitby-away> obviously openslug, ucslugc, debian, etc. will all have their own defconfigs, but I think that every arm arch needs to have a <foo>_defconfig in the kernel source (as a starting point for new distros)
Oct 07 23:49:05 <jbowler-away> The one which is checked in to monotone works fine with both LE and BE (the .bb file just hacks the ENDIAN setting).
Oct 07 23:49:37 <rwhitby-away> I was figuring we'd just take the existing ixp4xx_defconfig, make the required changes for the slug hardware, and just push that upstream. 
Oct 07 23:50:09 <rwhitby-away> (assuming that would be the least contentious to deepak)
Oct 07 23:50:10 <jbowler-away> That works.
Oct 07 23:50:28 <jbowler-away> It resolves what to put in for USB and networking.
Oct 07 23:51:07 <jbowler-away> It should have MTD on for NSLU2, with the BE config.
Oct 07 23:51:34 <rwhitby-away> yeah
Oct 07 23:52:46 <rwhitby-away> for the debian kernels we'll be working from the other direction - take the debian kernel defconfig and mod that for the slug.  Each distro will have differing requirements.
Oct 07 23:54:52 <rwhitby-away> anyway, catch ya later.
Oct 07 23:55:10 <jbowler-away> bed for me...
Oct 08 01:57:24 <CIA-8> 03rwhitby * 10kernel/2.6.12/ (9 files): Back-ported from 2.6.14 patchset
**** ENDING LOGGING AT Sat Oct 08 02:59:56 2005