**** BEGIN LOGGING AT Sat May 15 02:59:57 2010
May 15 03:34:41 <Weedy> OutBackDingo: working ath5k
May 15 03:43:02 <OutBackDingo> Weedy: you on trunk ? whys it not working and whats the hardware?
May 15 03:44:02 <Weedy> not openwrt
May 15 03:44:15 <Weedy> OOOOOOOOOOOOOOOBAAAAAAAAAAAAAAAAANTUUUUUUUUU
May 15 11:04:59 <CIA-33> hauke * r21449 /trunk/ (18 files in 8 dirs):
May 15 11:04:59 <CIA-33> ssb: activate CONFIG_SSB_BLOCKIO in the kernel and not in compat-wireless.
May 15 11:04:59 <CIA-33> Before this patch CONFIG_SSB_BLOCKIO was set to y in compat-wireless
May 15 11:04:59 <CIA-33> but the in kernel ssb system had it set to n, this causes an invalid pointer.
May 15 11:04:59 <CIA-33> b43 now needs CONFIG_SSB_BLOCKIO=y
May 15 11:08:02 <CIA-33> swalker * r21450 /packages/net/tor/ (Makefile files/tor.init): [packages] tor: update to 0.2.1.26, fix log directory
May 15 11:33:59 <CIA-33> swalker * r21451 /packages/net/tor/files/tor.init: [packages] tor: better log directory permissions
May 15 11:43:34 <CIA-33> swalker * r21452 /packages/utils/tar/Makefile: [packages] tar: update to 1.23, use PKG_INSTALL
May 15 14:07:32 <CIA-33> swalker * r21453 /packages/utils/zile/Makefile: [packages] zile: update to 2.3.16
May 15 15:56:53 <CIA-33> nbd * r21454 /trunk/package/mac80211/ (18 files in 2 dirs): mac80211: update to wireless-testing 2010-05-12 plus some pending patches
May 15 15:57:01 <CIA-33> nbd * r21455 /trunk/include/package-bin.mk: only build bin packages for the selected build variants (same behaviour as with ipkg builds)
May 15 17:24:13 <[florian]> rtz2: ping?
May 15 17:24:25 <[florian]> rtz2: I have a slightly more intrusive solution to fix the mfd resource passing
May 15 17:24:37 <rtz2> [florian]: could you show it to me?
May 15 17:24:47 <[florian]> rtz2: yes hold a sec
May 15 17:25:52 <[florian]> this contains the refresh so it may not be really readable
May 15 17:25:54 <[florian]> http://openwrt.pastebin.com/AxU5MZKb
May 15 17:29:03 <rtz2> [florian]: it will work
May 15 17:29:26 <[florian]> rtz2: seems like it does here
May 15 17:29:37 <rtz2> [florian]: but I still think it's a better solution to use the regs directly
May 15 17:29:40 <[florian]> rtz2: I just forgot to change the platform_data into driver_daa
May 15 17:29:50 <rtz2> [florian]: right, that's also missing
May 15 17:29:53 <[florian]> rtz2: well yeah, might be easier in the end
May 15 17:29:59 <[florian]> it is actually ;~
May 15 17:30:08 <[florian]> do you have your last version of patch somewhere?
May 15 17:30:26 <rtz2> [florian]: you already have a base address in the form of the pci sb
May 15 17:30:29 <rtz2> one moment
May 15 17:32:30 <[florian]> rtz2: btw, sorry for not answering to your pings this week, I have been fairly busy
May 15 17:34:01 <rtz2> [florian]: no problem
May 15 17:34:26 <rtz2> [florian]: I wanted to ask about the broken image in the downloads dir
May 15 17:34:28 <rtz2> [florian]: http://openwrt.pastebin.com/D3QLbrra
May 15 17:34:43 <[florian]> rtz2: yes, well still have not figured out what is wrong
May 15 17:35:11 <rtz2> [florian]: this does remove the platform_data structure completly and instead uses rdc321x_gpio_dev->sb_pdev = to_pci_dev(pdev->dev.parent); to get the sb dev
May 15 17:35:19 <Weedy> nbd: SYN
May 15 17:35:28 <rtz2> [florian]: not sure, if this is a good idea
May 15 17:35:54 <[florian]> rtz2: does not seem too bad, the mfd maintainer was suggesting to pass the pdev pointer as platform/driver data, but I do not see the point
May 15 17:35:58 <rtz2> [florian]: did you check, wheather the kernel stripping option is enabled in the build config for the server?
May 15 17:36:01 <[florian]> rtz2: since it's the parent device anyway
May 15 17:36:14 <[florian]> rtz2: got to check this with {Nico}
May 15 17:36:16 <rtz2> [florian]: seemed also more logical to me
May 15 17:36:47 <[florian]> rtz2: I will stick to your version and push this mainline later
May 15 17:36:54 <[florian]> rtz2: as well as your watchdog rewrite
May 15 17:37:17 <rtz2> ok, thanks
May 15 17:37:26 <[florian]> less patches to maintain
May 15 17:37:48 <[florian]> also I think I will drop 2.6.30 support soon
May 15 17:37:55 <[florian]> for rdc I mean
May 15 17:38:21 <nbd> Weedy: SYN|ACK
May 15 17:39:12 <rtz2> [florian]: it did produce a 830k kernel for me
May 15 17:39:35 <rtz2> [florian]: gcc 4.5 was smaller, but there is a pretty ugly bug in 4.5.0, it miscompiles busybox
May 15 17:40:05 <rtz2> [florian]: and backporting the patch was rather complicated
May 15 17:40:26 <Weedy> nbd: you hack on ath[59]k right?
May 15 17:40:50 <Weedy> like in wireless-dev not just openwrt
May 15 17:42:48 <nbd> ath9k only at the moment
May 15 17:42:50 <nbd> but yes
May 15 17:42:54 <Weedy> hmm
May 15 17:43:03 <Weedy> do you use the aircrack suite?
May 15 17:43:05 <[florian]> rtz2: ok, so we might want to wait a bit longer to drop support for it
May 15 17:43:07 <nbd> no
May 15 17:43:09 <[florian]> rtz2: or have a second stage loader
May 15 17:43:14 <Weedy> crap
May 15 17:43:57 <Weedy> i've noticed everything past 2009-12-11 breaks airodump
May 15 17:44:09 <nbd> weird
May 15 17:44:18 <Weedy> guess i'll go rage on the mailing list
May 15 17:44:23 <nbd> go ahead
May 15 17:44:57 <rtz2> [florian]: the second-stage bootloader would also have to have the rather odd file format of the x86 kernel, so this won't be straight forward
May 15 17:45:03 <Weedy> sorry i pinged you
May 15 17:45:18 <nbd> no problem
May 15 17:45:30 <[florian]> rtz2: yeah I know, you need to add it an x86 bootsector
May 15 17:50:09 <stintel> Weedy: breaks airodump how ?
May 15 17:50:29 <stintel> Weedy: I've been running airodump on recent openwrt/ath9k fine
May 15 17:54:37 <rtz2> [florian]: well, let's hope we can do without it
May 15 17:55:46 <[florian]> rtz2: hardly, the kernel will keep growing in size, and the bootloader will remain the same
May 15 17:55:54 <[florian]> rtz2: I do not feel confortable with asking users to change it
May 15 17:56:42 <_trine> [florian], when compiling for the TEW632BRP there used to be a .uni file compiled has this changed because it's not here anymore?
May 15 17:57:27 <[florian]> _trine: I do not know, I am not following closely ar71xx dev
May 15 17:58:02 <_trine> is it jushog that's looking after that?
May 15 17:58:20 <_trine> I may not have his nick correct
May 15 17:58:35 <[florian]> your nick is correct, he maintains ar71xx
May 15 17:58:59 <_trine> ok I will ask him
May 15 17:59:02 <_trine> thanks
May 15 17:59:47 <rtz2> [florian]: so the course is clear: we need a magic compression algorithm :)
May 15 18:02:03 <[florian]> rtz2: yeah, magic is definitively the word, as it has no appeared to us yet :p
May 15 18:02:52 <rtz2> [florian]: nanozip is pretty good, from what I know, but not open source
May 15 18:03:24 <rtz2> [florian]: it got the kernel down to 550k or so
May 15 18:04:05 <[florian]> rtz2: I did not know about it
May 15 18:05:11 <rtz2> [florian]: there isn't much to know: http://www.nanozip.net/
May 15 18:06:04 <[florian]> rtz2: unless we have an open source implementation there is not much we can do with booting stuff with that algo
May 15 18:07:33 <rtz2> [florian]: right
May 15 18:07:42 <rtz2> [florian]: and i kinda doubt he will hand it over
May 15 18:07:55 <rtz2> [florian]: hmmm, maybe if I'm really bored, I will ask
May 15 18:08:02 <[florian]> rtz2: yeah, ask
May 15 18:11:47 <[florian]> rtz2: have I somehow broken gpio for rdc321x recently?
May 15 18:12:30 <rtz2> [florian]: let me check
May 15 18:15:13 <rtz2> [florian]: yup
May 15 18:15:40 <rtz2> [florian]: not sure if you are using my patch or not, but maybe i did mix up the regs (crtl1 and crtl2)
May 15 18:15:59 <[florian]> rtz2: I am using your patch
May 15 18:22:07 <stintel> argh this console issue causing openwrt to hang on kvm is really making me crazy :(
May 15 18:22:25 <rtz2> [florian]: hmm, looks correct to me
May 15 18:23:34 <rtz2> [florian]: on first glance
May 15 18:24:02 <rtz2> brb
May 15 18:33:54 <Weedy> stintel: it stops seeing data packets
May 15 19:14:52 <CIA-33> florian * r21456 /trunk/target/linux/generic-2.6/patches-2.6.30/971-ocf_20100325.patch: [kernel] do not export find_task_by_vpid twice on 2.6.30
May 15 19:15:05 <CIA-33> florian * r21457 /trunk/target/linux/ar7/patches-2.6.32/140-uart_port_ar7.patch: [ar7] use correct port type, which sets the correct receive fifo trigger options (#7095)
May 15 19:15:11 <CIA-33> florian * r21458 /trunk/target/linux/rdc/patches-2.6.30/011-tune_lzma_options.patch: [rdc] fix 2.6.30 lzma tuning patch to output file size at end of kernel, otherwise decompressor remains stuck
May 15 19:18:41 <[florian]> rtz2: as to the mfd driver, the correct flags to have been set were IORESOURCE_IO, that would have prevented all the issues from the beginning ;)
May 15 19:19:19 <rtz2> [florian]: ok
May 15 19:19:34 <rtz2> [florian]: good to know in case this comes up again
May 15 19:21:19 <[florian]> rtz2: pci_write_config_dword does not fail
May 15 19:21:36 <[florian]> rtz2: and the values in gpch->data_reg[reg] are correctly updated by the gpio functions
May 15 19:22:08 <[florian]> rtz2: so I am starting to wonder if this is not due to adding the input gpio button
May 15 19:25:20 <rtz2> [florian]: this did work originally
May 15 19:26:24 <[florian]> rtz2: yes I know, caching the value is not an issue I think
May 15 19:31:24 <[florian]> rtz2: that's weird, setting the led works fine, but clearing its status does not work
May 15 19:33:03 <rtz2> [florian]: got it
May 15 19:33:09 <rtz2> I think
May 15 19:33:19 <rtz2> check rdc_gpio_set_value_impl
May 15 19:33:33 <rtz2> this was in your code: reg ? gpch->reg1_data_base : gpch->reg2_data_base,
May 15 19:34:08 <rtz2> [florian]: this is, how it should be: reg ? RDC321X_GPIO_DATA_REG2 : RDC321X_GPIO_DATA_REG1
May 15 19:34:32 <[florian]> right, this is reverted
May 15 19:34:58 <rtz2> [florian]: I don't know, if you switched the values for reg1 and reg2, I did map reg1 to the define for reg1
May 15 19:36:02 <rtz2> [florian]: it's probably wrong for all regs
May 15 19:36:46 <[florian]> rtz2: I guess so
May 15 19:38:53 <[florian]> rtz2: works ok now
May 15 19:40:36 <rtz2> ok
May 15 20:50:52 <builder> build #56 of octeon is complete: Failure [failed compile_4]  Build details are at http://tksite.gotdns.org:8010/builders/octeon/builds/56
May 15 21:08:21 <[florian]> rtz3: I get such errors with 2.6.32 on rdc: http://openwrt.pastebin.com/J7ZwEr52
May 15 21:08:37 <[florian]> rtz3: on the second reboot it is ok though
May 15 21:09:48 <[florian]> rtz3: I spoke too quickly, happens on the second reboot too
May 15 21:10:01 <rtz3> rtz3: could you poste complete dmesg?
May 15 21:11:16 <[florian]> rtz3: here it is: http://openwrt.pastebin.com/xqpmhGyu
May 15 21:11:26 <[florian]> rtz3: let me check one thing
May 15 21:15:48 <rtz3> [florian]: you remember the small change which was supposed to fix a problem with the sitecom partition size?
May 15 21:16:08 <[florian]> rtz3: yes, is that applicable to ar525w as well?
May 15 21:17:50 <rtz3> [florian]: nope
May 15 21:18:39 <rtz3> [florian]: but you did add the fix at the wrong place
May 15 21:19:01 <rtz3> [florian]: only at the ar525w instead of sitecom
May 15 21:19:14 <[florian]> rtz3: ah ah ok let me fix that
May 15 21:23:31 <[florian]> rtz3: the change is not required for 2.6.32 right?
May 15 21:24:18 <rtz3> [florian]: it is, but mtd doesn't work anyway, so in the end, it doesn't really matter
May 15 21:24:44 <[florian]> rtz3: ok
May 15 21:26:57 <[florian]> ok, rdc should be in a much better shape now
May 15 21:26:58 <CIA-33> florian * r21459 /trunk/target/linux/rdc/ (6 files in 2 dirs): [rdc] fix panic on boot due to invalid IORESOURCE for MFD cells, fix gpio value setting
May 15 21:27:03 <CIA-33> florian * r21460 /trunk/target/linux/rdc/patches-2.6.32/100-rdc_boards.patch: [rdc] fix wrong hunk introduced by r21309, thanks rtz2
May 15 21:28:27 <[florian]> rtz3: I am off to bed now
May 15 21:28:52 <rtz3> [florian]: ok, bye
May 15 22:37:00 <CIA-33> jow * r21461 /trunk/target/linux/brcm-2.4/patches/200-fix_ipv6_receiving_with_ipv4_socket.patch:
May 15 22:37:00 <CIA-33> [brcm-2.4] fix a kernel bug where IPv6 packets are received on IPv4 sockets, this fixes dual stack software like OLSRd when running on Kernel 2.4
May 15 22:37:00 <CIA-33> Signed-off-by: Alina Friedrichsen <x-alina@gmx.net>
May 16 00:20:20 <CIA-33> nbd * r21462 /trunk/ (10 files in 6 dirs): get rid of CRLF encoding in various files
May 16 00:23:08 <CIA-33> nbd * r21463 /branches/backfire/ (9 files in 7 dirs): [backfire] merge CRLF changes from r21462
**** ENDING LOGGING AT Sun May 16 02:59:57 2010