**** BEGIN LOGGING AT Fri Jun 06 03:00:00 2014 Jun 06 03:14:39 <5EXAA8SU8> build #616 of ramips is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/616 Jun 06 05:23:26 <5EXAA8SU8> build #611 of ppc40x is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/ppc40x/builds/611 Jun 06 06:39:23 <5EXAA8SU8> build #480 of gemini is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/gemini/builds/480 Jun 06 09:21:53 nbd r41030 trunk/config/Config-kernel.in * kernel: fix duplicate KERNEL_PERF_EVENTS with wrong dependency Jun 06 09:33:22 nbd: my patch screwed up b43 for non-Broadcom :| Jun 06 09:33:28 sorry :( Jun 06 09:34:30 zajec: then please send a fix ;) Jun 06 09:34:38 nbd: no idea how :( Jun 06 09:34:44 huh? Jun 06 09:34:46 nbd: see my e-mail please Jun 06 09:34:58 hasn't arrived yet Jun 06 09:35:42 it'll in a second, or see https://lists.openwrt.org/pipermail/openwrt-devel/2014-June/025883.html Jun 06 09:36:07 blogic r41031 trunk/package/ system/procd/files/procd.sh system/procd/Makefile * procd: service instances can now report an error of why they failed to start. Jun 06 09:36:15 blogic r41032 trunk/package/network/services/igmpproxy/files/igmpproxy.init Jun 06 09:36:16 igmpproxy: report an error and dont start the instance if one of the interfaces is not up Jun 06 09:36:24 blogic r41033 trunk/package/ utils/ubi-utils/patches/130-lzma_jffs2.patch utils/ubi-utils/Makefile * ubi-utils: bump to version 1.5.1 for ubiblock support Jun 06 09:37:11 nbd: http://git.openwrt.org/?p=openwrt.git;a=blob;f=package/kernel/mac80211/Makefile;hb=HEAD#l1067 Jun 06 09:37:15 zajec: in DEPENDS: +PACKAGE_B43_BUSES_BCMA_AND_SSB:kmod-ssb +PACKAGE_B43_BUSES_BCMA_AND_SSB:kmod-bcma +PACKAGE_B43_BUSES_BCMA:kmod-bcma +PACKAGE_B43_BUSES_SSB Jun 06 09:37:21 :kmod-ssb Jun 06 09:37:31 try that Jun 06 09:37:33 ohh Jun 06 09:37:35 will see Jun 06 09:37:54 i didn't think about using sub-configs in DEPENDS Jun 06 09:39:36 working on it Jun 06 09:45:24 seems to be working for ar71xx Jun 06 10:53:17 nbd r41034 trunk/toolchain/binutils/ (8 files in 4 dirs) * toolchain/binutils: remove obsolete patches Jun 06 10:53:21 nbd r41035 trunk/package/ devel/oprofile/patches/100-use_insmod_instead_of_modprobe.patch devel/oprofile/Makefile devel/oprofile/patches/101-remove_24_support.patch devel/oprofile/patches/110-fix_compile_error.patch * oprofile: update to 0.9.9, fix compile error Jun 06 10:53:24 nbd r41036 trunk/package/devel/binutils/Makefile Jun 06 10:53:24 binutils: decouple from toolchain selection, use version 2.24, fix libiberty installation Jun 06 13:22:21 nbd: i think i'll need to resend V2 Jun 06 14:50:45 nbd: i think I'll have a working version soon Jun 06 15:15:42 damn Jun 06 15:15:46 I can't fix it Jun 06 16:04:45 KanjiMonster: nbd: any idea how to fix this circular dependency? :| Jun 06 16:06:33 KanjiMonster: nbd: is there any way to stop OpenWrt from watching this pseudo-broken-dependency? to silent this "Package kmod-b43 is missing dependencies for the following libraries:" Jun 06 16:08:52 KanjiMonster: your solution would still result in circular dependency I'm afraid Jun 06 16:19:18 zajec: almost done with a fix Jun 06 16:20:52 nbd: i'd love to see it Jun 06 16:20:57 this is killing me ;p Jun 06 16:23:14 i have to trick the system a bit Jun 06 16:23:26 into processing the dependency only at build time, not at metadata dump time Jun 06 16:25:58 nbd r41037 trunk/package/kernel/mac80211/Makefile * b43: fix ssb/bcma dependencies Jun 06 16:26:04 zajec: there ;) Jun 06 16:34:21 nbd: warning: (PACKAGE_kmod-b44 && PACKAGE_B43_USE_SSB && PACKAGE_kmod-b43legacy) selects PACKAGE_kmod-ssb which has unmet direct dependencies (!TARGET_brcm63xx && PCI_SUPPORT && !TARGET_brcm47xx) Jun 06 16:34:45 odd, that warning didn't show up in my test Jun 06 16:34:54 i guess the PCI_SUPPORT stuff needs to be moved then Jun 06 16:34:57 this is with brcm63xx Jun 06 16:35:28 http://pastebin.com/LDfg1dMu <- this is what I did before you committed yours Jun 06 16:35:53 i tried that, but that was emitting the recursive dep warnings Jun 06 16:36:08 not for me Jun 06 16:36:12 interesting Jun 06 16:36:18 so we have different kinds of warnings ;) Jun 06 16:37:15 can you try adding depends on PCI_SUPPORT to PACKAGE_B43_USE_BCMA and PACKAGE_B43_USE_SSB and see if it makes the warning go away? Jun 06 16:37:20 the recursive warning was triggered as soon as it depended on more than one symbol (e.g. @(PCI_SUPPORT||TARGET_brcm47xx)), but not if it was only @PCI_SUPPORT Jun 06 16:39:59 later, now dinner Jun 06 16:47:22 nbd: i didn't know I can use if $(if ...), nice to know Jun 06 16:58:18 nbd: i can confirm that Jun 06 16:58:31 nbd: i did "rm .config", then "menuconfig" and selected brcm63xx Jun 06 16:58:43 kmod-ssb was on :( PACKAGE_kmod-ssb [=y] Jun 06 16:58:53 warning: (PACKAGE_kmod-b44 && PACKAGE_B43_USE_SSB && PACKAGE_kmod- Jun 06 16:58:58 this warning was in menuconfig Jun 06 16:59:01 I couldn't copy the rest Jun 06 17:16:32 nbd: warning: (PACKAGE_kmod-b44 && PACKAGE_B43_USE_SSB && PACKAGE_kmod-b43legacy) selects PACKAGE_kmod-ssb which has unmet direct dependencies (PCI_SUPPORT && !TARGET_brcm47xx && !TARGET_brcm63xx) Jun 06 17:16:44 it seems PACKAGE_B43_USE_SSB selected kmod-ssb Jun 06 17:17:18 nbd: maybe you missed brcm63xx target in Jun 06 17:17:22 depends on !TARGET_brcm47xx Jun 06 17:19:31 nbd: http://pastebin.com/y6rtJ5fh Jun 06 17:19:43 Hauke: hello :) thanks for pushing subtarget patches Jun 06 17:23:07 zajec: you can use spaces if you are in pure Kconfig (i.e. !TARGET_brcm47xx && !TARGET_brcm47xx) Jun 06 17:30:46 yeah, just noticed that Jun 06 17:30:55 i'm sure nbd knows that ;) Jun 06 17:38:34 <5EXAA8SU8> build #633 of orion is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/633 Jun 06 17:39:39 <5EXAA8SU8> build #522 of au1000 is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/au1000/builds/522 Jun 06 17:40:30 <5EXAA8SU8> build #617 of ramips is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/ramips/builds/617 Jun 06 17:40:43 <5EXAA8SU8> build #494 of iop32x is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/494 Jun 06 17:41:09 <5EXAA8SU8> build #587 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/587 Jun 06 17:41:14 <5EXAA8SU8> build #556 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/556 Jun 06 17:41:36 <5EXAA8SU8> build #493 of ep93xx is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/ep93xx/builds/493 Jun 06 17:41:47 <5EXAA8SU8> build #176 of mvebu is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/mvebu/builds/176 Jun 06 17:42:28 <5EXAA8SU8> build #587 of rb532 is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/587 Jun 06 17:42:40 <5EXAA8SU8> build #482 of mcs814x is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/mcs814x/builds/482 Jun 06 17:42:52 <5EXAA8SU8> build #486 of adm5120 is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/adm5120/builds/486 Jun 06 17:43:44 <5EXAA8SU8> build #99 of sunxi is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/sunxi/builds/99 Jun 06 17:43:56 <5EXAA8SU8> build #444 of octeon is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/octeon/builds/444 Jun 06 17:44:48 <5EXAA8SU8> build #82 of pxa is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/pxa/builds/82 Jun 06 17:45:01 <5EXAA8SU8> build #85 of cns21xx is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/cns21xx/builds/85 Jun 06 17:45:50 KanjiMonster: does this fix look OK for you? Jun 06 17:45:53 <5EXAA8SU8> build #83 of mpc83xx is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/mpc83xx/builds/83 Jun 06 17:45:54 KanjiMonster: did it fix your warning? Jun 06 17:46:05 <5EXAA8SU8> build #84 of bcm53xx is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/bcm53xx/builds/84 Jun 06 17:46:11 zajec: it does; I came to the same conclusion Jun 06 17:46:57 <5EXAA8SU8> build #82 of brcm2708 is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/brcm2708/builds/82 Jun 06 17:47:10 <5EXAA8SU8> build #84 of realview is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/realview/builds/84 Jun 06 17:48:01 <5EXAA8SU8> build #81 of omap is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/omap/builds/81 Jun 06 17:48:13 <5EXAA8SU8> build #82 of adm8668 is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/adm8668/builds/82 Jun 06 17:49:06 <5EXAA8SU8> build #82 of malta is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/malta/builds/82 Jun 06 17:49:18 <5EXAA8SU8> build #82 of cns3xxx is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/cns3xxx/builds/82 Jun 06 17:50:10 <5EXAA8SU8> build #83 of x86_64 is complete: Failure [failed svn compile_9] Build details are at http://buildbot.openwrt.org:8010/builders/x86_64/builds/83 Jun 06 18:46:41 <5EXAA8SU8> build #598 of uml is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/uml/builds/598 Jun 06 19:00:58 <5EXAA8SU8> build #627 of cobalt is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/627 Jun 06 19:19:58 <5EXAA8SU8> build #663 of brcm63xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/663 Jun 06 19:24:18 zajec: hm, not sure how much sense it makes to disable built-in ssb for 74k if b43 will pull it in again as a module (due to b43 defaulting to bcma+ssb); also won't this cause then the a missing ssb for bcm47xx_74k? Jun 06 19:24:23 -a Jun 06 19:25:57 <5EXAA8SU8> build #193 of imx6 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/imx6/builds/193 Jun 06 19:26:08 zajec: it should probably also default to bcma only for mips74k Jun 06 19:27:18 KanjiMonster: err, did you see my patch [PATCH] brcm47xx: mips74k: disable BCM47XX_SSB Jun 06 19:27:30 KanjiMonster: it simply dropps support for SSB bus in arch code Jun 06 19:27:53 KanjiMonster: but I also just sent [RFH][PATCH] brcm47xx: mips74k: drop CONFIG_SSB for kernel config Jun 06 19:28:32 zajec: true, but then wouldn't then a "normal" all-m build cause b43 to be built with BUSES_BCMA_SSB, thus selecting kmod-ssb as =m (which btw isn't available for brcm47xx)? Jun 06 19:28:36 KanjiMonster: I wanted to allow hacker to compile firmware without SSB and b43 without SSB support Jun 06 19:29:32 so unless you go the extra effort with also changing b43 to bcma-only, you will actually increase in space Jun 06 19:29:35 KanjiMonster: we can't be bcma-only for mips74k, because there are some BCMA SoCs with SSB PCIe cards attached Jun 06 19:29:50 will that increase size? Jun 06 19:29:53 huh, that's .. interesting Jun 06 19:30:08 yes, ssb as a module will be bigger than if included in the kernel Jun 06 19:30:37 I thought all mips74k devices use bcma based wifis for their secondary wifi Jun 06 19:30:39 oh :( Jun 06 19:30:53 KanjiMonster: BCM43222 Jun 06 19:31:20 at least i think so Jun 06 19:31:21 zajec: huh, never saw a pcie bcm43222, only pci Jun 06 19:31:21 let me check Jun 06 19:31:36 maybe i'm wrong Jun 06 19:31:47 also did you test that this doesn't break usb-brcm47xx? Jun 06 19:32:00 (becuase that kmod includes both ssb and bcma drivers) Jun 06 19:35:38 KanjiMonster: https://wikidevi.com/wiki/Linksys_E3000 https://wikidevi.com/wiki/Linksys_WRT610N_v2 (BCMA SoC with SSB cards) Jun 06 19:38:36 zajec: heh. must have been before the introduction of bcm43224. Jun 06 19:38:43 so I stand corrected ;) Jun 06 19:39:46 :) i still have to look at this USB, etc. Jun 06 19:40:21 KanjiMonster: can you try to help me anyway with this [RFH][PATCH] brcm47xx: mips74k: drop CONFIG_SSB for kernel config Jun 06 19:40:39 why this doesn't work? KCONFIG:=SSB_DRIVER_MIPS=n Jun 06 19:42:18 register_pci_controller isn't EXPORT_SYMBOL'd, so you can't use it from modules Jun 06 20:02:51 zajec: no idea why it doesn't work; I guess m/y from kernel config trumps n from module config. Jun 06 20:04:39 KanjiMonster: but kernel config has "n" Jun 06 20:05:07 I've added in config-default following line: # CONFIG_SSB is not set Jun 06 20:06:06 zajec: but doesn't target/bcm47xx/config-3.1x still have SSB_DRIVER_MIPS=y set? Jun 06 20:06:12 oh Jun 06 20:06:13 wait Jun 06 20:23:53 too bad we can not get rid of ssb when the SoC used bcma Jun 06 20:24:02 ther are some devices with 16MB ram Jun 06 20:31:53 <5EXAA8SU8> build #567 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/567 Jun 06 20:35:38 Hauke: yeah Jun 06 20:36:35 i forgot ssb.ko will be loaded by b43.ko, even if there is no SSB device Jun 06 20:36:41 i wanted to save some RAM Jun 06 20:37:26 Hauke: this helped: # CONFIG_SSB_PCICORE_HOSTMODE not set Jun 06 20:38:20 you can still deactivate some of the parts for ssb which are not needed by pcie devices Jun 06 20:38:40 the mips and flash part Jun 06 20:39:01 i'm working ont hat Jun 06 20:39:04 comparing sizes right now Jun 06 20:40:21 you can also remove these config options: CONFIG_FIXED_PHY CONFIG_GPIO_WDT CONFIG_ADM6996_PHY Jun 06 20:40:48 i didn't think about FIXED, thx Jun 06 20:41:43 it is only used for the adm switch Jun 06 21:23:56 Hauke: it seems I can save only ~4 KB by droppping some SSB options Jun 06 21:24:02 I hoped for a better savings :( Jun 06 21:29:33 anybidy understands how static dhcpv6 leases should work under odhcpd? Jun 06 21:29:51 i can't figure out how to set it up (to work with dnsmasq for v4) Jun 06 21:33:15 Hauke: CONFIG_FIXED_PHY saves 500 B ;) Jun 06 21:43:29 Hauke: B43_PHY_LP eats 15 KB, finally something Jun 06 21:59:14 I hope that mips16 compiler support will save some more space Jun 06 22:06:55 Hauke: what exactly is that? Jun 06 22:07:52 the 74k cpu supports a new extension Jun 06 22:07:53 http://www.linux-mips.org/wiki/MIPS16 Jun 06 22:08:23 gcc can generate smaller instructions and make the binary smaller this way Jun 06 22:11:04 how we can enable that? Jun 06 22:11:14 is MIPS16 supported on MIPS74K only? Jun 06 22:14:26 it is also supported by some older CPUs I think, but not by all cpus used in bcm47xx Jun 06 22:14:50 how to enable it? Jun 06 22:15:40 set FEATURES:=mips16 in the makefile Jun 06 22:15:51 see: target/linux/ar71xx/Makefile Jun 06 22:33:28 Hauke: i'll try to test it over weekend on some devices Jun 06 22:34:07 thanks Jun 06 22:35:10 Hauke: is this possible to set some FEATURES for one subtarget only? Jun 06 22:35:19 I do not know Jun 06 22:35:26 find ./target/ -name Makefile -exec grep -H --color "FEATURES" {} \; Jun 06 22:35:37 it seems we set FEATURES for the whole target :| Jun 06 22:37:41 zajec: there are several targets that set features in the subtargets Jun 06 22:39:21 KanjiMonster: how/where? Jun 06 22:39:24 KanjiMonster: can't find Jun 06 22:39:36 ./target/linux/malta/be/target.mk Jun 06 22:39:44 zajec: grep -r FEATURES --include=target.mk Jun 06 22:40:37 ahh, i was looking for Makefile-s only, my bad Jun 06 22:43:58 do i need some make clean? Jun 06 22:44:19 it seems adding mips16 feature didn't change size Jun 06 22:46:03 ok, I'll try the hard way... rm -fR build_dir/ Jun 06 22:46:18 you also have to activate mipsr2 optimized code Jun 06 22:48:16 do you mean CONFIG_CPU_MIPS32_R2 ? Jun 06 22:50:37 zajec: try this: http://pastebin.com/jH3nESqR Jun 06 22:54:52 * zajec compiling Jun 06 22:55:45 nothing Jun 06 22:55:47 same size Jun 06 22:55:52 rm -fR build_dir Jun 06 22:58:52 did you ran make oldconfig? Jun 06 22:58:59 it should build a new gcc version Jun 06 22:59:16 zajec: do you have build_dir/toolchain-mipsel_74kc_gcc-4.8-linaro_uClibc-0.9.33.2/ Jun 06 22:59:29 i did "rm .config" Jun 06 23:00:26 hm, i don't Jun 06 23:00:32 target-mipsel_mips32_uClibc-0.9.33.2 Jun 06 23:04:59 wait, maybe this time i forgot oldconfig or sth? Jun 06 23:05:40 ok, now I got target-mipsel_74kc_uClibc-0.9.33.2 Jun 06 23:05:42 sorry for that Jun 06 23:18:21 i've a feeling it takes much more time to compile the toolchain Jun 06 23:20:29 <5EXAA8SU8> build #632 of atheros is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/atheros/builds/632 Jun 06 23:25:03 will continue tomorrow Jun 06 23:25:05 night Jun 06 23:28:57 i'm back ;) Jun 06 23:29:14 right after my "goodnight" I've noticed toolchain is complete Jun 06 23:29:41 so my squashfs+vmlinux size decreased from 2 993 360 to 2 948 144 Jun 06 23:30:08 well, it's always something Jun 06 23:30:19 but I hoped for even more Jun 06 23:30:49 night #2 Jun 07 02:48:31 <5EXAA8SU8> build #297 of mpc85xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/mpc85xx/builds/297 **** ENDING LOGGING AT Sat Jun 07 02:59:59 2014