**** BEGIN LOGGING AT Fri Sep 27 02:59:58 2013 Sep 27 03:07:00 tripolar r38227 packages/mail/dovecot/Makefile * [packages] dovecot: update to 2.2.6 Sep 27 03:20:02 build #342 of xburst is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/xburst/builds/342 Sep 27 04:03:46 build #334 of kirkwood is complete: Failure [failed compile_3] Build details are at http://buildbot.openwrt.org:8010/builders/kirkwood/builds/334 Sep 27 06:08:46 build #324 of au1000 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/324 Sep 27 06:11:13 build #298 of iop32x is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/298 Sep 27 07:05:00 build #297 of ep93xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/297 Sep 27 07:20:24 juhosg r38228 trunk/ (28 files in 27 dirs) * kernel: update 3.10 to 3.10.13 Sep 27 07:53:00 build #361 of ar7 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/ar7/builds/361 Sep 27 09:00:27 build #392 of atheros is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/atheros/builds/392 Sep 27 10:23:54 build #400 of brcm47xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx/builds/400 Sep 27 10:50:34 luka r38229 trunk/target/linux/ imx6/patches-3.10/0015-thermal-add-imx-thermal-driver-support.patch imx6/config-3.10 * imx6: backport thermal driver Sep 27 10:50:35 luka r38230 trunk/package/kernel/linux/modules/other.mk * kernel: add thermal modules Sep 27 10:50:37 luka r38231 trunk/target/linux/imx6/Makefile * imx6: enable thermal module by default Sep 27 10:50:38 luka r38232 trunk/target/linux/mvebu/profiles/100-Generic.mk * mvebu: enable thermal module by default Sep 27 10:50:39 luka r38233 trunk/target/linux/imx6/config-3.10 * imx6: enable cpufreq-imx6q by default Sep 27 11:16:36 juhosg r38234 trunk/target/linux/ar71xx/files/drivers/mtd/nand/ar934x_nfc.c * ar71xx: ar934x-nfc: disable subpage write when hardware ECC is used Sep 27 11:57:23 build #261 of octeon is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/261 Sep 27 12:05:28 hi, what will be the kernel used in the next (barrier breaker?) openwrt release? Sep 27 12:06:13 i saw some patches on the mailing list which move the kirkwood architecture to 3.10 Sep 27 12:06:32 will these patches be merged? Sep 27 12:07:16 3.10 is the target kernel, as it is a LTS kernel Sep 27 12:07:21 for BB Sep 27 12:07:22 i'm interested in adding support for the Linkstation LS-XHL to openwrt Sep 27 12:08:27 is it already supported upstream? Sep 27 12:08:37 (i.e. in the vanilla linux kernel) Sep 27 12:08:44 ok then it makes more sense to focus on the 3.10 because in that kernel the support is already integrated Sep 27 12:09:40 I can't say whether these exact patches will be merged, but kirkwood will be at 3.10 for BB Sep 27 12:11:57 ok, nice. do we know when BB will be released? just a rough estimation please. this year, next year? Sep 27 12:13:38 when it's ready ;p Sep 27 12:14:37 not this year Sep 27 12:25:05 jow_laptop: no ? Sep 27 12:25:06 :) Sep 27 12:27:30 ok then there is still plenty of time to get the support merged, right? Sep 27 12:28:18 we wont release with out all targets on 3.10 Sep 27 12:30:23 shouldn't -sysupgrade.bin images have the same header as factory? I tried to sysupgrade using -sysupgrade.bin but it said invalid image type, Image check platform_check_image failed Sep 27 12:31:40 Devastator: most often the (absence of the) vendor specific header is the difference between factory and sysupgrade Sep 27 12:32:39 KanjiMonster do I need to edit some file for it to generate sysupgrade images properly? Sep 27 12:33:20 luka r38235 trunk/tools/Makefile * [tools] quilt: add findutils as dependency Sep 27 12:33:44 Devastator: depends on the target, sometimes you need to define two image macros, one for factory, one for sysupgrade, sometimes the macro already creates both Sep 27 12:34:11 hum.. this is ramips by the way.. Sep 27 12:34:23 don't know if makes any difference] Sep 27 12:40:41 KanjiMonster I showed my bootlog to Gabor, which said these jffs2 messages were ugly and if it happened after sysupgrade also.. here is the messages: http://codepad.org/DAJsuZ0R Sep 27 12:41:32 Devastator: is this a jffs2-only image or a squashfs+jffs2 one? Sep 27 12:42:07 KanjiMonster squashfs+jffs2 I guess, I only selected squashfs as target image Sep 27 12:45:07 hm Sep 27 12:45:25 does the image contain the deadc0de marker? Sep 27 12:45:47 KanjiMonster how do I check? hex editor? Sep 27 12:46:05 yeah Sep 27 12:46:15 KanjiMonster which offset? Sep 27 12:46:16 (it's 0xdeadc0de of course ;) Sep 27 12:46:24 somwhere at the end Sep 27 12:46:27 oh :) Sep 27 12:46:34 let's see.. Sep 27 12:48:18 factory doesn't have it.. checking sysupgrade Sep 27 12:49:37 sysupgrade have it, but it's not accepted by sysupgrade Sep 27 12:50:35 Does anyone here know why ARM v5TE has been replaced with ARM XSCALE on the latest SVN Trunk? Sep 27 12:50:37 luka r38236 trunk/tools/Makefile * [tools] patch: other tools might depend on this one Sep 27 12:52:08 mazilo: where exactly? Sep 27 12:53:03 nbd: I just updated to SVN r38234 and now the compilation on toolchains changes to ARM XSCALE for Marvel Kirkwood. It used to be ARM v5TE Sep 27 12:53:32 "the compilations on toolchains"? Sep 27 12:53:54 please describe what you see, not what you read into it Sep 27 12:54:02 Well, when I launched 'make', the process rebuilt toolchains. Sep 27 12:54:10 right Sep 27 12:54:23 that currently happens with all targets, as i did some rework there Sep 27 12:54:28 that renames the toolchain directory Sep 27 12:55:00 i introduced a CPU_TYPE variable to eliminate false sharing of toolchains between incompatible arm targets (or ones with different optimizations) Sep 27 12:55:46 IC. But, the toolchains fails to compile. Sep 27 12:56:09 you could have started with that ... ;) Sep 27 12:56:20 and provided some details on the build error Sep 27 12:56:26 I did a 'make dirclean' first. Sep 27 12:56:44 http://pastebin.com/7UvUAndC Sep 27 12:57:10 did you do make oldconfig as well after updating? Sep 27 12:57:10 now I have to figure out why sysupgrade doesn't accept my sysupgrade image, any pointers? :) Sep 27 12:57:40 either way, that log already indicates that the real error message is inside config.log Sep 27 12:57:48 so please put that in a pastebin as well Sep 27 12:57:54 so i don't have to do random guesses ;) Sep 27 12:58:39 nbd: I did a 'make dirclean' first. Sep 27 12:58:59 ok, let me know when you have the info Sep 27 13:00:20 http://pastebin.com/NnYf8wdi Sep 27 13:00:59 What I don't know is which 'configure' file. Sep 27 13:01:38 I'm guessing ramips has only one macro for both factory and sysupgrade images Sep 27 13:03:11 mazilo: that pastebin contains the path to where the config.log is Sep 27 13:03:17 at least it looks like it Sep 27 13:03:20 nbd: i see the same errors Sep 27 13:03:53 nbd: there seems to be a typo in the xscale optimization flags, "march=" vs "-march=" Sep 27 13:04:16 ah, right Sep 27 13:04:17 will fix that Sep 27 13:04:20 thanks Sep 27 13:04:25 nbd: and i guess CPU_TYPE=arm926ej-s should be set in the kirkwood makefile Sep 27 13:04:31 nbd: Yes. But, in that path '/opt/openwrt-svn-trunk/build_dir/toolchain-arm_xscale_gcc-4.8-linaro_uClibc-0.9.33.2_eabi/gcc-linaro-4.8-2013.08-minimal/arm-openwrt-linux-uclibcgnueabi/libgcc', I don't see a 'configure' file. Sep 27 13:04:47 nbd r38237 trunk/include/target.mk * build: fix typo in arm cflags Sep 27 13:04:53 mazilo: that's normal Sep 27 13:05:06 mazilo: the configure file for that is in the gcc source directory Sep 27 13:05:13 mazilo: i just stumpled about the same error Sep 27 13:05:14 not in the build one Sep 27 13:05:22 mwalle: it's fixed now Sep 27 13:05:26 (see above) Sep 27 13:05:32 need to run make oldconfig after updating Sep 27 13:05:34 nbd: thanks ;) Sep 27 13:05:38 nbd: OK. Thanks. Sep 27 13:05:43 but still kirkwood defaults to xscale, right? Sep 27 13:05:51 right, just like before my changes Sep 27 13:06:01 just the directory name is different now Sep 27 13:06:17 nbd: I am now doing the 'make oldconfig'. Sep 27 13:06:39 nbd: ok i see, but shouldn't be the correct one the arm926ej-s ? Sep 27 13:07:09 there is probably a better match than xscale, but i'm not sure if it's arm926ej-s Sep 27 13:10:26 nbd: Same problem after the 'make oldconfig' (http://pastebin.com/mLcqb9AQ). Sep 27 13:12:26 KanjiMonster do you think deadc0de marker is all I need to stop seeing these ugly jffs2 messages? Sep 27 13:12:47 Devastator: that or a complete format of the flash Sep 27 13:12:58 s/format/erasure Sep 27 13:13:02 KanjiMonster I see.. Sep 27 13:13:12 nbd: i guess you are right, the linux kernel does the follwoing: tune-$(CONFIG_CPU_FEROCEON) =$(call cc-option,-mtune=marvell-f,-mtune=xscale) Sep 27 13:13:14 mazilo: right, make dirclean is necessary as well. Sep 27 13:13:29 nbd: I just SVN updated to 38237 and am now doing 'make dirclean; make'. Sep 27 13:13:44 nbd: will be back in 1hr. Sep 27 13:14:55 KanjiMonster thanks for your helped, I've reported the sysupgrade problem to Gabor as well, let's see what he says heheh Sep 27 13:15:14 I'm not sure if sysupgrade needs it Sep 27 13:15:30 but maybe he has some input Sep 27 13:15:40 ooops, help Sep 27 13:16:24 KanjiMonster factory image doesn't have it, only sysupgrade has it Sep 27 13:43:29 how many targets are not in 3.10? Sep 27 14:05:50 mstenber@mstenber-ubuntu-1204 ~/hnet/openwrt/dist/target/linux>egrep '^LINUX_VERSION:=' */Makefile | cut -d '=' -f 2 | sort | uniq -c | sort -n Sep 27 14:05:50 2 3.6.11 Sep 27 14:05:50 2 3.9.11 Sep 27 14:05:50 8 3.8.13 Sep 27 14:05:50 12 3.3.8 Sep 27 14:05:50 21 3.10.13 Sep 27 14:06:33 at least 3.10 is the most common one.. ;) Sep 27 14:08:07 24 targets to migrate hehe Sep 27 14:08:10 a lot of work.. Sep 27 14:09:27 porting a device is easy, I need to learn C for real contribution Sep 27 18:14:29 Hauke: hey, i saw some ppl asking about ADM on ML Sep 27 18:14:40 Hauke: someone said he sent you one router with ADM switch Sep 27 18:14:49 yes I have it Sep 27 18:14:53 Hauke: did you start writing switch driver for ADM? or anyone else? Sep 27 18:15:10 no I haven't started :-( Sep 27 18:15:30 ok, just wanted to now :) to don't duplicate your work in case i decide to start ;) Sep 27 18:15:49 Hauke: recently AMD released specs for their audio engine, so I'll be busy with that Sep 27 18:15:58 but after testing that I may try to work on ADM Sep 27 18:16:11 I also found specs for the adm switch Sep 27 18:16:26 oh, nice Sep 27 18:16:33 just the gpio interface is a little strange Sep 27 18:16:52 but we should be able to use "swtich" as code reference, right? Sep 27 18:16:59 yes Sep 27 18:17:03 ok :) Sep 27 18:17:44 there is already a driver for such a switch or a similar one for swconfig, but it uses a different interface to talk to the switch Sep 27 18:18:28 oh, thanks for info Sep 27 18:18:31 i didn't know tha Sep 27 18:32:14 zajec hi, I've found a router with BCM43224 and thought about buying, but after talking to Hauke and KanjiMonster, it won't be much useful.. Sep 27 18:33:07 the problem with bcm43224 is mostly political, not technical Sep 27 18:37:02 sorry, got disconnected Sep 27 18:37:46 by political I guess it is broadcom internal politics?! Sep 27 18:38:28 kernel rules don't allow more than one driver for the same hardware Sep 27 18:38:42 and there already is a driver for bcm43224; brcmsmac Sep 27 18:39:45 that is what I heard Sep 27 18:40:45 are there any comparisons between brcmsmac and b43? Sep 27 18:41:20 comparisons? Sep 27 18:41:36 performance comparisons, I mean Sep 27 18:43:30 Devastator: http://sprunge.us/PgEM (as of r38226) Sep 27 18:44:46 b43 is broadcom driver reverse engineered by zajec and brcmsmac is basically broadcom source, kernel independent, work of Hauke, is this correct? Sep 27 18:44:48 brcmsmac support ieee80211n rates Sep 27 18:45:51 brcmsmac was done by broadcom I just added beconing Sep 27 18:47:27 beconing? Sep 27 18:47:51 sending Beacon frames Sep 27 18:47:57 http://en.wikipedia.org/wiki/Beacon_frame Sep 27 18:48:08 this is needed for ap mode Sep 27 18:49:24 right! that's what nbd said when I asked if brcmsmac was sta only Sep 27 18:49:32 that you were adding ap mode Sep 27 18:49:35 now I remember Sep 27 18:51:25 Devastator: policy was already violated, b43 and brcmsmac, both support BCM43224 Sep 27 18:51:54 b43 support is poorer / more poor Sep 27 18:56:05 zajec I see.. anyway, when I saw you and KanjiMonster talking, I thought it would be helpful and got excited in the pursuit of a device with that chip or BCM43222 Sep 27 18:56:45 bcm43222 is fine politically Sep 27 18:57:16 hummm.. back to the search :) Sep 27 18:57:18 Devastator: by working on BCM43224 you would probably imrpove b43 for BCM43222 and BCM4331 and others too Sep 27 18:57:35 BCM43222 is SSB based (it may be hard to google) Sep 27 18:57:40 brcmsmac doesn't support SSB Sep 27 18:57:45 only BCMA Sep 27 18:58:00 I've found it but only on usb dongles.. Sep 27 18:58:42 http://www.ebay.de/itm/BroadCom-BCM94322-BCM43222-Mini-PCI-WLAN-WIFI-KaRTE-for-Acer-Asus-Toshiba-/180946693310?pt=DE_Computing_Interne_Netzwerkkarten&hash=item2a21436cbe Sep 27 18:59:17 zajec: right, but it will be hard to get that accepted if state that its tested/based on bcm43224 because bcm43224 is already quite well supported with brcmsmac (and will improving 43224 support really help 4331 that much? these are different PHYs) Sep 27 18:59:47 anyway, if the seller answer my questions and it is in fact v1.0 (the one with BCM43224), I'm gonna buy it anyway, it's dirty cheap for my country Sep 27 19:00:50 KanjiMonster: b43 need 802.11n support in general code, probably not much N-PHY specific Sep 27 19:00:53 BCM43224 has pcie and BCM4222 has PCI Sep 27 19:04:59 zajec: might be true. also I think 4331 still has the rx/tx headers issue Sep 27 19:14:15 well, I think my WRT310N has a 4331... Sep 27 19:17:50 Devastator: no the wrt310n has a BCM4321 (in ver 1.0) or uses the integrated wifi in the BCM4716 (in ver 2.0) Sep 27 19:19:06 humm, correct Sep 27 19:19:15 mine is v1 Sep 27 19:19:42 useless then hehe Sep 27 19:27:27 blogic: ping? Sep 27 19:39:31 I really want to do this mod: http://wl500g.info/showthread.php?30205-Asus-RT-N16-256MB-RAM-amp-Wi-Fi-2-4-5-GHz-DualBand Sep 27 19:46:10 Devastator: lol Sep 27 19:47:17 that's great, I love this router Sep 27 19:48:33 the only hardware modifications I do is adding serial ;-) Sep 27 19:50:52 build #294 of adm5120 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/adm5120/builds/294 Sep 27 19:52:34 hi Sep 27 19:52:43 jow_laptop: any new about the socket for ubus? Sep 27 19:52:59 jow_laptop: any news about the socket for ubus? Sep 27 20:30:18 serial++ luci is for lusers! ;-) Sep 27 20:31:46 build #393 of atheros is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/atheros/builds/393 Sep 27 20:44:52 build #395 of orion is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/395 Sep 27 20:45:18 nbd r38238 trunk/package/kernel/linux/modules/other.mk Sep 27 20:45:19 kernel: mark kmod-thermal menuconfig symbol as hidden to avoid building it on targets with no thermal drivers Sep 27 20:48:11 nbd r38239 trunk/package/kernel/linux/modules/pcmcia.mk * kernel: remove old kernel compat code for pcmcia Sep 27 20:48:29 build #401 of brcm47xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx/builds/401 Sep 27 20:54:33 build #382 of cobalt is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/382 Sep 27 20:55:03 nbd r38240 trunk/target/linux/ generic/config-3.8 generic/config-3.6 * kernel: add missing symbols for iio on older kernel versions Sep 27 20:57:40 build #363 of rb532 is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/363 Sep 27 21:00:35 build #383 of cobalt is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/383 Sep 27 21:00:44 build #363 of ppc44x is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/363 Sep 27 21:07:36 Hauke I usually install serial and jtag header, but I like hardware mod hehe Sep 27 21:07:49 specially because these devices price are quite high over here.. Sep 27 21:11:53 build #385 of lantiq is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/385 Sep 27 21:13:03 nbd r38241 trunk/target/linux/ generic/config-3.8 generic/config-3.9 generic/config-3.10 * kernel: add another missing symbol for iio Sep 27 21:13:07 nbd r38242 trunk/package/kernel/linux/modules/sound.mk * kernel: make sound-soc-core depend on kmod-ac97 Sep 27 21:13:10 nbd r38243 trunk/target/linux/lantiq/config-default * lantiq: add a default for CONFIG_MTD_NAND_XWAY to fix build errors Sep 27 21:13:14 nbd r38244 trunk/target/linux/octeon/config-default * octeon: remove CONFIG_FW_LOADER override to fix build errors Sep 27 21:13:18 nbd r38245 trunk/package/kernel/linux/modules/other.mk * kernel: exclude regmap-spi.ko if SPI support cannot be selected Sep 27 21:13:21 nbd r38246 trunk/target/linux/ generic/config-3.8 generic/config-3.9 generic/config-3.6 generic/config-3.10 * kernel: add missing config symbol Sep 27 21:13:40 nbd: thanks Sep 27 21:13:59 no problem Sep 27 21:14:10 just went through all buildbot failure logs and fixed whatever showed up there ;) Sep 27 21:14:20 let's see how many still fail in the next round ;) Sep 27 21:24:40 build #415 of brcm63xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/415 Sep 27 21:26:55 build #343 of sibyte is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/343 Sep 27 21:28:33 a seller that doesn't answer questions for 3 days can't be trusted.. :( Sep 27 21:30:30 bcm4322 is not the same as bcm43222 right? Sep 27 21:43:14 right Sep 27 21:43:15 http://wireless.kernel.org/en/users/Drivers/b43 Sep 27 21:43:37 BCM4322 is N rev 4, BCM43222 is N rev 6 Sep 27 21:43:59 both are SSB based Sep 27 21:59:29 build #276 of mpc52xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/mpc52xx/builds/276 Sep 27 22:00:02 build #376 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/376 Sep 27 22:05:24 build #411 of at91 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/411 Sep 27 23:07:05 build #366 of x86 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/x86/builds/366 Sep 28 00:36:57 build #353 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/353 Sep 28 01:33:07 build #325 of au1000 is complete: Failure [failed compile_4] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/325 Sep 28 01:35:02 build #299 of iop32x is complete: Failure [failed shell_12] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/299 **** ENDING LOGGING AT Sat Sep 28 02:59:58 2013