**** BEGIN LOGGING AT Sat Oct 28 02:59:57 2006 Oct 28 03:38:54 03bzhou * r4308 10optware/trunk/make/mod-python.mk: mod-python: upstream upgrade to 3.2.10 Oct 28 09:17:12 morning, anyone able to give some debianslug guidance? Oct 28 09:43:47 CieD, what do you mean by "guidance"? Oct 28 09:44:09 you know, like, spiritual guidane Oct 28 09:44:15 is Debian the one true way? Oct 28 09:44:22 is DebianSlug a mere imitation Oct 28 09:44:25 that sort of thing Oct 28 09:46:53 :) Oct 28 09:47:23 EvilDevil, I've just typed the problem out in #openslug Oct 28 09:47:37 I see you're in there too Oct 28 09:47:54 http://logs.nslu2-linux.org/livelogs/openslug.txt Oct 28 09:48:16 * CieD didn't know about logs Oct 28 09:48:26 cool Oct 28 09:49:01 don't worry, everything you say is recorded for posterity Oct 28 09:49:08 and then deleted the day after Oct 28 09:49:19 pretty useful though Oct 28 09:49:28 absolutely Oct 28 09:59:07 CieD, that's the way we make sure that people never become impolite ;) big brother's watching you Oct 28 10:00:03 good idea.. impolite people annoy me, so that'll not be me Oct 28 15:17:28 ~seen dwery Oct 28 15:17:35 dwery is currently on #nslu2-linux (4h 49m 33s) #openjtag (4h 49m 33s), last said: 'ok'. Oct 28 15:36:16 blaster8: pong Oct 28 15:36:25 hey there Oct 28 15:36:31 I got the avila patches in Oct 28 15:36:39 in svn? Oct 28 15:36:45 yes Oct 28 15:36:48 good. Oct 28 15:36:51 ok, i'll add Oct 28 15:36:54 thanks Oct 28 15:36:57 the cf specific section Oct 28 15:37:00 way ahead of me :) Oct 28 15:37:58 96-pata-loft.patch would be ok? Oct 28 15:48:46 I'd prefer it to be in08- Oct 28 15:49:13 they CF patch and the changes to Avila support should probably all go in at once Oct 28 15:51:39 dwery: just checking you saw that last bit Oct 28 15:51:55 blaster8: before having the CF platform patch upstream we should have the CF ixp4xx driver upstream Oct 28 15:52:08 and that one is going thru Jeff and Alan Oct 28 15:52:08 well, the two go separately, right Oct 28 15:52:40 I guess they shouul accept it for .20/.21 Oct 28 15:53:03 The CF driver to Linux-IDE and the platform stuff (separation out of Avila, adding in CF plaftorm support, Loft tweaks) to l-a-k Oct 28 15:53:12 no Oct 28 15:53:18 ? Oct 28 15:53:26 the cf driver must be sent to Jeff and Alan Oct 28 15:53:30 to lak Oct 28 15:53:38 oh ok Oct 28 15:53:42 we will send the separation of avila/loft Oct 28 15:53:56 when? Oct 28 15:54:23 even now. and then, once pata cf is upstream, we will send the CF platform patch to lak Oct 28 15:54:47 ok Oct 28 15:54:54 we need to check if the Avila patches work first Oct 28 15:54:56 we just need to have [g2] to confirm that he has tried the patch Oct 28 15:54:59 yep, right ;) Oct 28 15:55:03 indeed Oct 28 15:55:13 but last time he tested it, I don't think he updated his config Oct 28 15:55:50 I will test it in a while Oct 28 15:55:57 thanks Oct 28 16:14:23 <[g2]> dwery hey Oct 28 16:17:16 [g2]: pong Oct 28 16:17:33 <[g2]> dwery what's up ? Oct 28 16:17:56 we need to validate the latest svn in order to start pushing the patches Oct 28 16:20:13 [g2]: morinng. Oct 28 16:20:15 <[g2]> GPSFan hi Oct 28 16:20:36 <[g2]> dwery blaster8 and I were just talking about Avila testing with .19+ Oct 28 16:20:53 ok Oct 28 16:21:05 How close is Loft to Avila? Oct 28 16:21:06 I will compile and test against the loft in a few mins Oct 28 16:21:07 <[g2]> dwery blaster8 GPSFan just got an Avila board and has built .19-rc3 Oct 28 16:21:07 I got 2.6.19-rc3 running, there were a few issues with the patch set in svn 2.6.19 Oct 28 16:21:18 great! Oct 28 16:21:21 what are the problems? Oct 28 16:21:39 I'm not surprised - it was a completely blind change Oct 28 16:21:50 thanks for this, btw, GPSFan Oct 28 16:21:50 <[g2]> GPSFan did you get USB 2.0 or not ? Oct 28 16:21:56 there is an issue with redboot (maybe the one on my board) passing 0xf5 as the machine id Oct 28 16:22:09 GPSFan: oh yes, you need a shim. Oct 28 16:22:11 no usb2, I need all 4 mini-pci's Oct 28 16:22:22 <[g2]> GPSFan I've got a redboot if you want to flash it Oct 28 16:22:33 danger :) Oct 28 16:22:33 dwery: I just patched the mach-types file for the moment. Oct 28 16:22:40 GPSFan: That's not a kernel problem, but a bootloader issue Oct 28 16:22:41 ok Oct 28 16:22:54 <[g2]> GPSFan are you running BE or LE ? Oct 28 16:23:19 GPSFan: and did PCI work? Oct 28 16:23:33 there was also a MACH vs ARCH issue in the new avila patches that kept the pci bus from working. I changed to all MACH and it's working. Oct 28 16:23:42 BE Oct 28 16:23:43 ok, I'll fix that now Oct 28 16:23:57 GPSFan: do you have a compact flash? Oct 28 16:24:28 yes I had to mod the loft-cf patch (98) to go back to BE. Oct 28 16:24:42 is that the libata patch? Oct 28 16:24:52 there's no 98-xx ? Oct 28 16:24:58 <[g2]> no that's the CF IDE patch Oct 28 16:25:03 <[g2]> I think he means 96\ Oct 28 16:25:20 I haven't played 98-loft-ide-patch Oct 28 16:25:37 I made 98-avila-ide-be.patch Oct 28 16:25:55 is that the libata patch? or is it using the ide subsystem? Oct 28 16:25:56 GPSFan: 98-avila-ide-be.patch uses old IDE or new libata code? Oct 28 16:26:12 <[g2]> the 98- one uses the old IDE subsystem Oct 28 16:26:31 GPSFan: I might enable you to use the new libata CF driver Oct 28 16:26:42 [g2]: CF is wired the same way on Avila and Loft, right? Oct 28 16:26:51 <[g2]> dwery yeah Oct 28 16:26:54 dwery: the 98 patch came from the 2.6.18 patches dir on svn, but I used the rest of the patches from 2.6.19 Oct 28 16:26:55 great Oct 28 16:27:06 GPSFan: the 98- patch isn't going upstream Oct 28 16:27:07 <[g2]> it's really just an extra pci config Oct 28 16:27:11 I can read & write to the cf. Oct 28 16:27:36 <[g2]> GPSFan that's going to be replaced in .19+ with the new libata patches Oct 28 16:27:42 blaster8: that's ok I'm happy to leave that stuf to you guys. Oct 28 16:28:09 GPSFan: we'll sort it out so that you can test the new libata CF driver in about half an hour, if you're free... Oct 28 16:28:17 <[g2]> GPSFan dwery is running in LE we could use a BE tester Oct 28 16:28:20 I was just happy to be able to get it all running without toooo much effort. Oct 28 16:28:24 it would be great to be able to test it on BE :) Oct 28 16:28:28 as in, really great Oct 28 16:28:30 03azummo * r529 10kernel/trunk/patches/2.6.19/96-pata-ixp4xx.patch: driver updated to 0.1.1 Oct 28 16:29:32 I built a toolchain using buildroot with gcc 3.5.6, an binutils 2.16.1 Oct 28 16:29:49 cool, we use OpenEmbedded mostly Oct 28 16:30:14 where that problem of MACH vs ARCH was? Oct 28 16:30:30 I'm fixing that up Oct 28 16:30:32 ok Oct 28 16:30:52 built my own userland and it runs using busybox 1.2.1 and uClibc 0.9.29 Oct 28 16:31:11 that's good Oct 28 16:31:25 how hard was that? Oct 28 16:31:32 <[g2]> it's trivial Oct 28 16:31:52 cool Oct 28 16:31:54 dwery: in patch 8 and (my) 98. do a grep for ARCH-AVILA and MACH-AVILA Oct 28 16:32:09 <[g2]> blaster8 I've got a 2MB jffs2 rootfs that's just busybox and a simple program Oct 28 16:32:26 <[g2]> it's tiny little script Oct 28 16:32:33 [g2]: some mixes of binutils/uClibc/gcc don't work too well though. Oct 28 16:32:52 blaster8: 27-avila-setup-mac is not enabled in the series Oct 28 16:32:57 oops Oct 28 16:33:12 no real reason for that, just forgot Oct 28 16:33:15 yep Oct 28 16:33:17 it was 2 in the morning Oct 28 16:33:21 eheh Oct 28 16:33:44 maybe we should add cf platform infos before 27-avila-setup-mac Oct 28 16:33:56 otherwise the cf patch would depend on the net driver one Oct 28 16:34:39 that's a better way round Oct 28 16:35:05 ok, i'll fix this Oct 28 16:39:21 well, thanks guys, I'll be around (in & out) much of today. I'll be glad to help in testing & some debugging. I'll leave the sorting out of upstream submissions to you'all Oct 28 16:39:31 thanks Oct 28 16:39:51 we'll have patches supporting libata CF driver ready soon Oct 28 16:40:21 blaster8: ok, I've gotta do some outside work for a while, cul.. Oct 28 16:40:29 cya around Oct 28 16:41:09 i'm almost ready Oct 28 16:41:09 03blaster8 * r530 10kernel/trunk/patches/2.6.19/08-avila-loft-setup.patch: Fixds to Avila/Loft support Oct 28 16:41:53 ok, got 08. building. Oct 28 16:42:04 I'm updating the defconfig too Oct 28 16:42:07 (forgot that...) Oct 28 16:42:40 MACH_LOFT should depend on MACH_AVILA Oct 28 16:42:50 oops Oct 28 16:45:11 mmmm arch/arm/mach-ixp4xx/avila-setup.c:51: error: 'AVILA_SDA_PIN' undeclared here (not in a function) Oct 28 16:45:16 got it Oct 28 16:45:24 that didn't happen before, going to investigate Oct 28 16:45:48 it seems it's not defined in include/asm-/arch-ixp4xx/... Oct 28 16:46:00 I know why Oct 28 16:46:21 ok, I'll wait for new 08 Oct 28 16:47:15 fixed it Oct 28 16:47:19 building this time Oct 28 16:48:04 we should have an avila.h Oct 28 16:48:11 we do, but I forgot it Oct 28 16:48:12 and include it from hardware.h Oct 28 16:48:20 it's all done, I just forgot to diff that bit Oct 28 16:48:24 uh ok Oct 28 16:48:34 sorted Oct 28 16:48:42 great. me loves quilt :) Oct 28 16:49:44 this one builds Oct 28 16:49:55 03blaster8 * r531 10kernel/trunk/patches/2.6.19/ (08-avila-loft-setup.patch defconfig series): More fixes, and a complete patch this time Oct 28 16:53:31 building Oct 28 16:54:01 03blaster8 * r532 10kernel/trunk/patches/2.6.19/ (75-dsmg600-support.patch 96-fsg3-support.patch): Refresh board support patches Oct 28 16:59:50 ok, loft is running rc3 Oct 28 16:59:56 my setup loads in 30/40 sec via nfs. Oct 28 16:59:57 good ;) Oct 28 16:59:59 now trying pata Oct 28 17:00:33 yay Oct 28 17:01:29 ok, it works. Oct 28 17:01:38 nice Oct 28 17:01:40 updating svn Oct 28 17:02:50 I guess that, given that nfs is going that the ethernet connection is going fine? Oct 28 17:03:03 it's doing GREAT Oct 28 17:03:08 fair enough Oct 28 17:03:21 upstream ahoy Oct 28 17:03:30 maybe the code style of the driv isn't upstream ready, but it's awesome. Oct 28 17:03:43 apparently it's not the code style Oct 28 17:03:47 more the code location Oct 28 17:04:04 the NPE and QMGR drivers need to migrate into arch/arm/mach-ixp4xx Oct 28 17:04:38 or moved the arch/arm/common and then made more ixp2xxx friendly Oct 28 17:06:56 intendation and things like "if(" vs "if (" or "a=1" vs "a = 1" should be checked too. Oct 28 17:07:00 indentation. Oct 28 17:07:59 fair enough Oct 28 17:07:59 03azummo * r533 10kernel/trunk/patches/2.6.19/ (26-avila-setup-pata.patch series): Avila/Loft patcform setup for libata/CF driver Oct 28 17:08:43 patcform ;) Oct 28 17:08:47 [g2]: is up to you now ;) Oct 28 17:09:11 <[g2]> dwery up to me ? Oct 28 17:09:35 yep, testing svn on your loft ;) Oct 28 17:09:46 we've done all the work :p Oct 28 17:09:48 ;) Oct 28 17:10:41 blaster8: which board do you have? nslu2? Oct 28 17:10:48 * [g2] bows to the might blaster8 Oct 28 17:10:53 I'm joking don't worry Oct 28 17:10:53 * [g2] bows to the mighty blaster8 Oct 28 17:10:57 I can hardly even code C Oct 28 17:11:20 which is why me fixing up the setup code was a bit of an issue Oct 28 17:11:31 you did a great work Oct 28 17:11:33 fortunately, it was a copy/paste/grep job Oct 28 17:11:44 I have one NSLU2 Oct 28 17:11:55 is USB/USB2 working fine for you? Oct 28 17:11:57 I'll pick up a second in december Oct 28 17:12:08 Haven't got an USB ports on mine at the moment Oct 28 17:12:22 I removed them, and haven't yet soldered the new devices on Oct 28 17:12:29 ok. because I'm having probplems with .19 . USB seems unstable Oct 28 17:12:39 Also, it's about 100 miles from me Oct 28 17:12:44 and with no remote access Oct 28 17:12:44 :) Oct 28 17:12:54 poor lonely nslu2 Oct 28 17:12:55 that's one of my projects - remoteSlug Oct 28 17:13:04 <[g2]> blaster8 dwery r533 buils Oct 28 17:13:10 <[g2]> s/buils/builds/ Oct 28 17:13:10 [g2] meant: blaster8 dwery r533 builds Oct 28 17:13:14 yay! Oct 28 17:13:27 you've updated the config, now? Oct 28 17:13:41 <[g2]> the defconfg isn't appropriate yet but that's another issue Oct 28 17:13:47 gtg. see ya Oct 28 17:13:52 cya Oct 28 17:13:54 <[g2]> cheers THX Oct 28 17:14:09 [g2]: our defconfig is fine, but it has lots of modules where you want stuff built in Oct 28 17:14:23 <[g2]> blaster8 right that's the issue Oct 28 17:14:25 by fine, I mean, theoretically supports Loft Oct 28 17:14:35 that's a two minute fixup job, though Oct 28 17:14:52 <[g2]> blaster8 right but you know the difference between theory and practice right ? Oct 28 17:15:04 I can do it if you really want Oct 28 17:19:00 <[g2]> blaster8 you're doing a lot of good work Oct 28 17:19:12 well, it's mostly the easy stuff Oct 28 17:19:36 <[g2]> well it's stuff that needs to get done Oct 28 17:19:41 what do you have on the Loft? OHCI/UHCI? Oct 28 17:19:56 <[g2]> EHCI plus OHCI Oct 28 17:20:05 done Oct 28 17:20:24 building in USB mass storage support Oct 28 17:20:26 <[g2]> blaster8 please stop Oct 28 17:20:27 you never know Oct 28 17:20:30 ok Oct 28 17:20:43 back, does the current patch set in svn 2.6.19 include the cf driver? I'm about to try building the kernel? Oct 28 17:20:45 if you want to fix that up, that's cool Oct 28 17:20:57 GPSFan: yes it does Oct 28 17:21:14 GPSFan: have you tried the ethernet? Oct 28 17:21:51 blaster8: ok, I'll let you know what happens, I haven't tried building any modules yet I wanted to get the built-in stuff going first. Oct 28 17:21:59 right Oct 28 17:22:23 it's probably best to build the ethernet as a module for now, when you want to get it going Oct 28 17:22:47 <[g2]> GPSFan where's rootfs ? Oct 28 17:22:51 <[g2]> jffs2 ? Oct 28 17:22:57 you will also need a patched mdev or udev running to handle firmware upload Oct 28 17:23:07 cya all later Oct 28 17:24:53 [g2]: I've had several rootfs's going cf, nfs, flash. for current testing I'm using the one that came in the flash, and an image of it on cf. Oct 28 17:25:47 <[g2]> GPSFan ok so your booting directly to the CF Oct 28 17:26:43 the current defconfig has the cf driver as a module. Oct 28 17:27:08 so you either boot to flash or CONFIG_PATA_IXP4XX_CF=y . Oct 28 17:27:14 I have been, but the last test I did I booted from the onboard flash for rootfs, loading the kernel from the cf. then I did a mount of the cf to do read/write testing. Oct 28 17:27:46 I'll be there for another 5/10 min if help is needed Oct 28 17:29:13 the patches for non-avila, is is safe/neccessart to apply them. In my pervious tests I only applied what was neccessary for avila. Oct 28 17:30:14 some of them may require other patches Oct 28 17:30:44 checking Oct 28 17:30:54 your should be safe with *avila* Oct 28 17:31:00 and *pata* Oct 28 17:31:12 dwery: thanks .. Oct 28 17:31:21 and, if you want, the ixp4xx net driver Oct 28 17:36:10 of the patches I applied against a clean kernel tree 27 applied with offset "Hunk #1 succeeded at 132 with fuzz 1 (offset 28 lines)" Oct 28 17:37:18 <[g2]> dwery the defconfig file in r533 doesn't enable the CONFIG_PATA_IXP4XX_CF at all Oct 28 17:38:48 533 has Oct 28 17:38:53 CONFIG_ATA=m Oct 28 17:38:58 CONFIG_PATA_IXP4XX_CF=m Oct 28 17:40:50 GPSFan: should be ok Oct 28 17:41:07 it needs a refresh Oct 28 17:41:24 27 is required only if you want the ixp4xx net driver Oct 28 17:41:33 otherwise i might even not build Oct 28 17:43:43 dwery: ok, the previous 27 parch applied without offset. haven't started compiling quite yet. Oct 28 17:44:11 ok gtg Oct 28 17:44:26 see ya and thanks for your support. I'd appreciate having feedback on the following issues: Oct 28 17:44:48 a) libata compact flash driver in BE Oct 28 17:45:13 b) ixp4xx network driver with 33-ixp4xx-net-driver-mtd-load-fw.patch that should load the firmware from redboot Oct 28 17:45:26 i'd also appreciate having dmesg output for both :) Oct 28 17:45:27 thanks! Oct 28 17:47:48 compiling now. Oct 28 17:48:33 <[g2]> http://pastebin.ca/226629 Oct 28 17:48:58 <[g2]> dwery http://pastebin.ca/226629 I'll double check the series file, but that should work right ? Oct 28 17:49:32 root=/dev/sda1 Oct 28 17:50:17 <[g2]> heh... old habits die hard Oct 28 17:50:34 <[g2]> sda1 is usually reserved for the usb boots Oct 28 17:50:43 that's one big issue, actually Oct 28 17:52:20 GPSFan: I would suggest checking out our Kernel repository and building your kernel from that Oct 28 17:52:34 <[g2]> VFS: Cannot open root device "sda1" or unknown-block(0,0) Oct 28 17:52:34 <[g2]> Time: OSTS clocksource has been installed. Oct 28 17:52:58 time to boot from flash and do some diagnostics Oct 28 17:53:34 <[g2]> dwery where's your .confing ? Oct 28 17:53:53 <[g2]> oh... wait you're NFS booting Oct 28 17:57:37 ok, it compiled and runs, the cf is ata1 though. no dev for that. http://pastebin.ca/226641 Oct 28 17:58:47 ata1? Oct 28 17:58:49 how strange Oct 28 17:59:44 with the stuff I did yesterday using the gateworks driver it came out hda1, I haven't looked too carefully at this new code though. Oct 28 18:00:42 the new code runs under the libata umbrella Oct 28 18:00:56 so, I was expecting it to be /dev/sda1... Oct 28 18:01:30 # mount /dev/sda1 /mnt Oct 28 18:01:30 mount: Mounting /dev/sda1 on /mnt failed: No such device or address Oct 28 18:01:54 ls /dev ? Oct 28 18:02:03 do you have a static dev? Oct 28 18:02:11 # ls /dev/sd* Oct 28 18:02:11 /dev/sda /dev/sda2 /dev/sda4 /dev/sdb1 /dev/sdb3 Oct 28 18:02:11 /dev/sda1 /dev/sda3 /dev/sdb /dev/sdb2 /dev/sdb4 Oct 28 18:03:01 hmm Oct 28 18:04:04 you could poke around in /sys/drivers/ata if you want Oct 28 18:04:12 indeed, I ahall... Oct 28 18:04:18 shall even ;>) Oct 28 18:04:20 I need to go, so dwery will be your man to fix this up Oct 28 18:04:29 k thanks.. Oct 28 18:04:50 <[g2]> GPSFan it mounts as /dev/ata1 ? Oct 28 18:05:11 I don't think so Oct 28 18:05:12 [g2]: no, nor as sda1 Oct 28 18:05:23 I think the ata1 bit is internal to the libata system Oct 28 18:05:37 <[g2]> ok so it's not mounting at all Oct 28 18:06:55 <[g2]> GPSFan with the old GW driver no root delay was needed for /dev/hda1 Oct 28 18:07:12 # pwd Oct 28 18:07:12 # cat model Oct 28 18:07:12 SAMSUNG CF/ATA Oct 28 18:07:26 <[g2]> for /dev/sda1 there was a rootdelay needed for all the drivers to come up Oct 28 18:07:44 oops missed a few lines.. Oct 28 18:08:00 the driver is seeing the card Oct 28 18:08:00 I'm not sure why mount isn't going Oct 28 18:08:24 [g2]: so what does that mean? whouls I add something to the mount command> Oct 28 18:08:48 * GPSFan can't type... time for lunch maybe... Oct 28 18:08:57 <[g2]> ehe Oct 28 18:09:09 <[g2]> :) Oct 28 18:10:33 <[g2]> ok root delay doesn't make a difference Oct 28 18:10:41 <[g2]> I'm not seeing the drive probed Oct 28 18:11:15 <[g2]> so IMHO the driver isn't fully up an hence the rootfs isn't available Oct 28 18:12:10 [g2]: double check you are actually compiling the driver into the kernel Oct 28 18:12:16 have you run make menuconfig Oct 28 18:12:35 [g2]: so we are seeing basically the same thing, I have my root in ram, so I can look at /proc & /sys. Oct 28 18:12:36 <[g2]> blaster8 you saw my svn diff right ? Oct 28 18:12:36 you are not only missing the driver probing, but the driver even registering *its* existance Oct 28 18:12:41 yup Oct 28 18:13:12 compare and contrast your bootlog with GPSFan Oct 28 18:13:13 <[g2]> GPSFan nod. I could boot ram or jffs2 also Oct 28 18:13:28 on yours, the driver does not print any messages at all Oct 28 18:13:39 and it should, even if it sees no CF card, I think Oct 28 18:13:44 <[g2]> blaster8 well he's also running Avila and I'm running Loft Oct 28 18:14:09 shouldn't make a difference, given the patch, but I suppose there may be some error Oct 28 18:14:33 triple check the config, with make menuconfig, then talk to dwery Oct 28 18:14:35 I didn't use the defconfig from svn, I used my .config from yesterday and did make oldconfig. Oct 28 18:14:42 I need to get to a concert Oct 28 18:14:52 blaster8: have fun ;>) Oct 28 18:14:58 <[g2]> enjoy Oct 28 18:15:00 will do, I'm singing in it Oct 28 18:15:05 much better than watching Oct 28 18:15:07 cool.. Oct 28 18:15:19 good luck with this Oct 28 18:17:43 [g2]: I'll poke at this after lunch, I've never used libata before, so it might be educational. Oct 28 18:18:48 <[g2]> GPSFan I think libata is heading towards mainstream Oct 28 18:19:25 I really don't know anything about it, will have to do some googling. Oct 28 18:19:44 <[g2]> really I think we just have some minor config issues Oct 28 18:20:00 do you want my .config? Oct 28 18:20:56 it isn't too different from the defconfig except I've trimmed out a lot of the modules. and it's BE. Oct 28 18:20:58 <[g2]> GPSFan see the PM ? Oct 28 18:21:16 will do.. Oct 28 18:21:50 <[g2]> nod, I've got a loft specific config that patches/2.6.18 dir Oct 28 18:22:09 <[g2]> it's trimmed down a lot of the modules and has a bunch of builtins Oct 28 18:23:57 well, more info can't hurt, and mine did build a functioning driver (not fully though) Oct 28 18:24:27 I started with a clean un-taring of the 2.6.19-rc3 tarball Oct 28 18:24:52 * GPSFan lunch Oct 28 18:24:57 <[g2]> GPSFan cheers Oct 28 18:26:01 <[g2]> GPSFan dwery has a working CF driver in LE on the Loft. He's not booting to it and has his own config. I really think it a minor config issue or a minor bug somewhere. Oct 28 18:42:01 <[g2]> bbl Oct 28 18:43:27 [g2]: you are probably right, if /sys is seeing it and can read it's model then it should work, esp, considering that dwery has it working on le loft. Oct 28 18:44:13 arg, it's too nice a day outside, gotta run out again and get some stuff done before winter really arives. cul.. Oct 28 20:55:52 03rwhitby * r4309 10optware/trunk/ (make/openssh.mk sources/openssh/rc.openssh): openssh: On startup, kill process in /var/run/sshd.pid if that file exists Oct 28 21:38:01 03bzhou * r4310 10optware/trunk/Makefile: updated PATCH_LIBTOOL sed script Oct 28 21:42:43 03bzhou * r4311 10optware/trunk/make/dspam.mk: dspam: bumped IPK_VERSION to build with updated PATCH_LIBTOOL sed script Oct 28 21:43:02 03bzhou * r4312 10optware/trunk/make/svn.mk: svn: bumped IPK_VERSION to build with updated PATCH_LIBTOOL sed script Oct 28 21:45:23 03bzhou * r4313 10optware/trunk/make/lighttpd.mk: lighttpd: bumped IPK_VERSION to build with updated PATCH_LIBTOOL sed script Oct 28 21:47:40 03rwhitby * r4314 10optware/trunk/ (make/denyhosts.mk sources/denyhosts/rc.denyhosts): denyhosts: re-arranged directory locations to better match optware standards Oct 28 21:49:42 03bzhou * r4315 10optware/trunk/make/imagemagick.mk: imagemagick: bumped IPK_VERSION to build with updated PATCH_LIBTOOL sed script Oct 28 21:53:05 rwhitby: around? Oct 28 21:56:27 i'd be really interested in a chat with you about the topfield some time Oct 28 21:56:40 what about topfield? Oct 28 21:57:19 well, just that i met an enthusiast last week, and im interested myself in a digital terrestrial pvr Oct 28 21:57:52 I'm pretty satisfied with mine 5100C Oct 28 21:58:47 does yours run taps? Oct 28 21:58:54 of course Oct 28 21:59:00 or do you mean do I use'em? Oct 28 21:59:11 i meant did that version 5100c Oct 28 21:59:28 is yours for sat or dtb ? Oct 28 21:59:32 C Oct 28 21:59:34 cable Oct 28 21:59:37 aah Oct 28 21:59:52 and does it rock? Oct 28 21:59:57 pretty much Oct 28 22:00:11 03bzhou * r4316 10optware/trunk/make/ (template-cvs.mk template.mk): template*: added -check target to save some typing Oct 28 22:00:20 some nation specific probs - but doesn't really apply anywhere else than in Finland Oct 28 22:00:29 dvb texts mostly Oct 28 22:00:53 i think the 5800 160gb was the one used by this guy i met Oct 28 22:00:54 which two channels here still persistently use Oct 28 22:01:46 and do you find you move much stuff off the topfield via slug? Oct 28 22:02:09 damn good idea - could record a movie and watch it later on without any texts blocking the view...but not many manufacturers know how to follow the standard Oct 28 22:02:29 yes, daily Oct 28 22:02:42 it gets full pretty easily ;) Oct 28 22:02:45 heh Oct 28 22:02:47 and I have 300gig drive installed Oct 28 22:02:52 320 actually Oct 28 22:02:52 i see Oct 28 22:03:02 and puppy is for getting data off? Oct 28 22:03:08 nah, not puppy Oct 28 22:03:09 or is that the ftp? Oct 28 22:03:12 ftpd-topfield Oct 28 22:03:19 what exactly is puppy for? Oct 28 22:03:48 well, it's for the same thing - but own tool for it - not an ftp-interface like ftpd-topfield Oct 28 22:03:57 oh i see Oct 28 22:04:02 it started with puppy Oct 28 22:04:07 and I guess people still use it Oct 28 22:04:08 and requires some tap running on topfield? Oct 28 22:04:11 nope Oct 28 22:04:16 out-of-the-box Oct 28 22:04:19 kk Oct 28 22:04:29 just usb cable Oct 28 22:04:37 so, it requires usb out from topfield and into slug? Oct 28 22:04:42 yup Oct 28 22:04:52 not sure if topfield has already rj45 models released Oct 28 22:04:59 which slug os are you using? Oct 28 22:05:12 unslung5.5b still Oct 28 22:05:18 k Oct 28 22:05:26 so, im using debianslug Oct 28 22:05:39 i compiled the apps and they run ok Oct 28 22:05:44 puppy and ftpd Oct 28 22:06:00 but, you think there may be issues plugging in usb to debianslug? Oct 28 22:06:09 why would there? Oct 28 22:06:12 no idea ;-) Oct 28 22:06:29 do you have issues using usb-hdd under debianslug? ;) Oct 28 22:06:38 no Oct 28 22:07:12 but i mounted that usb hd once (right back on day one slug ownership) and havent done anything similar since Oct 28 22:07:41 I don't think you'll run into any complications Oct 28 22:08:04 cool, sounds like the topfield is going on my xmas list Oct 28 22:08:24 i probly need a good 10 mtr usb cable though Oct 28 22:08:37 do you have usb hub in your slug? Oct 28 22:08:44 nope Oct 28 22:08:57 10 meters...asking for probs really Oct 28 22:08:57 1hd, 1topfield? Oct 28 22:09:00 why not move slug closer to it? Oct 28 22:09:13 mm Oct 28 22:09:16 I have it set right behind my topfield - 30cm cable or such ;P Oct 28 22:09:21 maybe possible Oct 28 22:09:35 slug is close to router Oct 28 22:09:40 and router to phone line Oct 28 22:09:45 i guess its easier to just extend that Oct 28 22:09:51 well, rj45 can always be longer Oct 28 22:09:54 usb-cable.. not Oct 28 22:10:04 yeah lol, i have a 10 mtr ethernet from xbox to router Oct 28 22:10:10 from those - definitely rj45 Oct 28 22:10:38 it can be closer to 100m to have real probs :P Oct 28 22:11:29 an ethernet upnp type topfield would be ideal i guess Oct 28 22:12:25 upnp? Oct 28 22:12:47 well, i mean just something to share the pvr disk on the network Oct 28 22:12:47 explain me why it would be something ideal? Oct 28 22:13:33 yes, it would also read external storages via ethernet.. Oct 28 22:13:45 for me, just to shove everything in the router is my ideal Oct 28 22:14:48 mm yes, i guess im using the xbox for that purpose at the moment. im just interested in being able to record some tv (although i havent done that since i binned my vcr 5 years ago) Oct 28 22:15:17 believe me when you get your toppy - you will ;) Oct 28 22:15:48 heh, yes this guy was telling me how its nice and easy to record any programmes with certain 'strings' in the metadata Oct 28 22:16:06 he recommended an epg tap Oct 28 22:16:19 i momentarily forget its name Oct 28 22:16:42 do you use your slug to send programme information to the topfield? Oct 28 22:17:00 or the stream data is reliable enough on your cable? Oct 28 22:18:54 I've tried toying with those epg taps a bit.. but yes, stream data is realiable enough Oct 28 22:20:42 well thanks for the chat! im pretty sure the topfield will be now on my christmas list Oct 28 22:21:49 laters :) Oct 28 22:22:10 oh, the ftpd Oct 28 22:22:16 yes? Oct 28 22:22:20 your run that on the slug? Oct 28 22:22:25 yes Oct 28 22:22:29 package is ftpd-topfield Oct 28 22:22:35 and ftp to yourself ? Oct 28 22:22:42 and it communicates down the usb? Oct 28 22:22:42 for most reliable results yes Oct 28 22:22:54 yes Oct 28 22:22:56 gotcha Oct 28 22:24:49 when you start the ftpd Oct 28 22:24:59 do you specify the location of the usb? Oct 28 22:25:05 err the mount point? Oct 28 22:25:45 nope Oct 28 22:25:58 just connect localhost port 2121 (as default) Oct 28 22:26:30 03nail 07slugos-3.10-beta * r380 10slugos/openembedded/packages/sipsak/sipsak_0.9.6.bb: sipsak: Fix wrong linewrap Oct 28 22:26:41 /opt/etc/init.d startup script executes the daemon Oct 28 22:26:47 /opt/sbin/ftpd-topfield -A -b -D -P 2121 Oct 28 22:26:51 oh you always have it running? Oct 28 22:26:55 yes Oct 28 22:27:07 whats turbo mode? Oct 28 22:27:09 I need to restart slug and/or daemon from time to time though Oct 28 22:27:32 not the most stabile machine around that topfield in that sense.. Oct 28 22:27:42 gotcha Oct 28 22:27:44 sorry..no slug Oct 28 22:27:46 slug=toppy Oct 28 22:27:47 :P Oct 28 22:28:04 nicknames getting mixed up there ;) Oct 28 22:28:23 03nail 07slugos-3.10-beta * r381 10slugos/openembedded/packages/ (3 files in 2 dirs): Add package dircproxy Oct 28 22:28:44 when i compiled mine, the executable was just named 'ftpd' Oct 28 22:29:10 ~/ftpd-topfield-0.6.7 Oct 28 22:29:13 that is Oct 28 22:33:56 [g2]: ping Oct 28 22:33:59 GPSFan: ping Oct 28 22:34:27 [g2]: you might be missing 26-avila-setup-pata.patch Oct 28 22:36:17 GPSFan: check that you haveCONFIG_SCSI=y Oct 28 22:36:17 CONFIG_SCSI_PROC_FS=y Oct 28 22:36:17 CONFIG_BLK_DEV_SD=y Oct 28 22:36:41 the partition name will be /dev/sdaX Oct 28 22:36:50 no root delay required Oct 28 22:55:10 03nail 07slugos-3.10-beta * r382 10slugos/openembedded/packages/asterisk/asterisk_1.2.13.bb: asterisk: Upstream upgrade to 1.2.13 Oct 28 22:59:34 dwery: had CONFIG_SCSI=y but not the other 2, rebuilding now.. Oct 28 23:00:00 GPSFan: SD is the scsi disk layer Oct 28 23:00:10 great! Oct 28 23:00:27 gotcha, still building ... Oct 28 23:03:03 dwery: seems to be working /dev/sda1 mounts and there is stuff there!! Oct 28 23:03:34 ok, so I did the right byte swapping in the driver :) Oct 28 23:05:36 seems so I just tested writing and it woeks. all the files I wrote before and by the PC were there and had the right stuff inside. Oct 28 23:06:23 ok. I will submit the driver for upstream inclusion in the next few days. How's the speed? Oct 28 23:06:45 haven't tested the speed, what have youall been seeing? Oct 28 23:07:38 with my card I've god ~1MB/sec for raw read speed Oct 28 23:07:56 and ~512K/sec in writing (IIRC) Oct 28 23:07:56 tested with dd Oct 28 23:09:41 # ls -la /mnt/z3 Oct 28 23:09:41 -rwxr-xr-x 1 0 0 1366620 /mnt/z3 Oct 28 23:09:41 # time cp /mnt/z3 /dev/null Oct 28 23:09:41 real 0m1.272s Oct 28 23:09:41 user 0m0.010s Oct 28 23:09:41 sys 0m0.640s Oct 28 23:10:05 03bzhou * r4317 10optware/trunk/make/denyhosts.mk: denyhosts: unbreak the build Oct 28 23:16:55 dwery: dd in busybox doesn't have a timer function, but it looks like I get about 1Mbytes/sec with my old trusty 64Mbyte card. Oct 28 23:17:06 ok Oct 28 23:33:28 GPSFan: thanks for your tests.. Oct 28 23:33:33 bed's time here... Oct 28 23:33:35 see ya Oct 28 23:34:03 dwery: night, I'm starting to play with the ethernet, somewhat confusing so far.. Oct 28 23:34:36 should be easy Oct 28 23:34:54 CONFIG_IXP4XX_QMGR=y Oct 28 23:34:54 CONFIG_IXP4XX_NPE=y Oct 28 23:34:54 CONFIG_IXP4XX_NPE_FW_LOAD=y Oct 28 23:34:54 CONFIG_IXP4XX_NPE_FW_MTD=y Oct 28 23:34:54 CONFIG_IXP4XX_NPE_FW_MTD_PARTITION="RedBoot" Oct 28 23:34:54 CONFIG_IXP4XX_MAC=y Oct 28 23:35:11 30-ixp4xx-0.2.1-driver.patch Oct 28 23:35:27 31-ixp4xx-2.6.19-cleanup.patch Oct 28 23:35:31 32-ixp4xx-net-driver-fix-qmgr.patch Oct 28 23:35:35 33-ixp4xx-net-driver-improve-mac-handling.patch Oct 28 23:35:35 built the modules, need the modprobe order, and firmware. Oct 28 23:35:42 33-ixp4xx-net-driver-mtd-load-fw.patch Oct 28 23:36:04 33-ixp4xx-net-driver-mtd-load-fw.patch Oct 28 23:36:04 will load the firmware from redboot Oct 28 23:36:32 need to check the .config thanks Oct 28 23:37:46 remember to save the dmsg output for me :) Oct 28 23:37:53 will do Oct 28 23:47:07 ah, there are 2 33-*** patches, I only had one.. Oct 28 23:58:06 dwery: it works.. http://pastebin.ca/227042 Oct 29 00:00:29 dwery: well it seemed to work the MAC address is wrong, and I can't ping anything. cu tomorrow. Oct 29 00:47:30 03rwhitby * r4318 10optware/trunk/ (3 files in 2 dirs): denyhosts: Added config file, renamed log file, added data dir. Oct 29 00:54:09 eno: thx for unbreaking denyhosts - I forget to svn add the new file Oct 29 00:54:39 mwester: good email topic (unslung future) Oct 29 01:28:35 rwhitby: no problem Oct 29 01:29:05 eno: do you run dspam yourself? Oct 29 01:29:17 no Oct 29 01:29:35 rwhitby: i'm thinking about adding optware/host/ directory Oct 29 01:29:47 what's that do? Oct 29 01:30:12 for python-host, perl-host, erlang-host, ... Oct 29 01:30:32 so that various platform does not need to do host build multiple times Oct 29 01:31:09 under host can have builds/ staging/ Oct 29 01:31:10 I see, and symlink it from the subdirs, like downloads. Then have the autobuilder build that first. Oct 29 01:31:15 yes Oct 29 01:31:53 ok, as long as it works in both cases (optware/fsg/... autobuilds and optware/... single platform builds) Oct 29 01:32:09 it will Oct 29 01:32:30 any how does it work for native autobuilds? Oct 29 01:33:31 i can have host/.configured actions do nothing Oct 29 01:33:43 ok. Oct 29 01:34:08 also, at the same time, can you think about how foo-stage targets should work for native builds? should they run ipkg install ? Oct 29 01:34:55 i probably will have separate foo-stage and foo-host-stage Oct 29 01:35:19 ok Oct 29 01:35:23 foo-host-stage is meaningful only for cross Oct 29 01:35:57 does this allow us to move anything out of NATIVE_PACKAGES into CROSS_PACKAGES ? Oct 29 01:36:11 no Oct 29 01:36:34 it's only a bit cleaner than the current way Oct 29 01:36:58 since perl-host-build happens for every platform, where it only needs to be built once Oct 29 01:37:42 ok, so it's mainly time-saving for multi-platform builds. Oct 29 01:38:46 yes Oct 29 01:47:40 03bzhou * r4319 10optware/trunk/Makefile: added HOST_* variables and host/.configured target **** ENDING LOGGING AT Sun Oct 29 02:59:56 2006