**** BEGIN LOGGING AT Mon Oct 27 03:00:01 2014 Oct 27 03:15:37 build #227 of bcm53xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/bcm53xx/builds/227 Oct 27 08:04:36 build #810 of brcm63xx is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/810 Oct 27 09:25:47 build #729 of ppc44x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/729 Oct 27 10:26:15 jow r43078 trunk/package/network/services/uhttpd/Makefile * uhttpd: fix HTTP incompatibilities in file handler Oct 27 10:26:16 karlp: ^ Oct 27 10:48:30 beautiful thanks. Oct 27 10:49:02 any chance of that going into BB as well? (at least the http parts, the mime types aren't essential, but shouldn't do any harm either) Oct 27 10:51:17 karlp: I'll eventually merge it, yes Oct 27 10:52:01 cool, no rush, thanks for looking into that for me, I was having a hard time working out what wsa going on Oct 27 11:03:30 nbd r43079 trunk/package/kernel/mac80211/patches/ 522-mac80211_configure_antenna_gain.patch 110-mac80211_keep_keys_on_stop_ap.patch 347-mac80211-flush-keys-for-AP-mode-on-ieee80211_do_stop.patch * mac80211: fix key flush handling in AP+STA Oct 27 11:08:47 jow r43080 trunk/package/ utils/px5g-standalone/Makefile utils/px5g-standalone/src/library/x509write.c * px5g-standalone: use /dev/urandom to initialize serial (#18232) Oct 27 11:08:53 nbd r43081 branches/barrier_breaker/package/ kernel/mac80211/patches/522-mac80211_configure_antenna_gain.patch kernel/mac80211/patches/520-mac80211_cur_txpower.patch kernel/mac80211/patches/300-pending_work.patch * mac80211: fix key flush handling in AP+STA Oct 27 11:10:17 jow r43082 branches/barrier_breaker/package/network/services/uhttpd/Makefile * BB: uhttpd: fix HTTP incompatibilities in file handler Oct 27 11:10:45 jow r43083 branches/barrier_breaker/package/ utils/px5g-standalone/src/library/x509write.c utils/px5g-standalone/Makefile * BB: px5g-standalone: use /dev/urandom to initialize serial (#18232) Oct 27 11:23:22 jow r43084 trunk/package/ utils/busybox/patches/270-libbb_make_unicode_printable.patch utils/busybox/Makefile * busybox: make high ASCII chars printable (#7993) Oct 27 11:23:50 jow r43085 branches/barrier_breaker/package/ utils/busybox/patches/270-libbb_make_unicode_printable.patch utils/busybox/Makefile * BB: busybox: make high ASCII chars printable (#7993) Oct 27 11:50:33 build #775 of orion is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/775 **** BEGIN LOGGING AT Mon Oct 27 12:03:18 2014 Oct 27 13:45:47 build #698 of sibyte is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/698 Oct 27 13:48:44 Kaloz: hi Oct 27 13:48:51 you have a WRT1900AC right? Oct 27 14:00:10 mripard: hi. yeah Oct 27 14:03:18 Kaloz: have you noticed any weird behaviour with the UART? Oct 27 14:03:57 I get something out of picocom/minicom only when I first boot the board and then start picocom Oct 27 14:04:06 and it hangs after a few seconds Oct 27 14:04:27 in order to have the UART back, I need to restart picocom Oct 27 14:04:35 and it's actually not really convenient :) Oct 27 14:04:58 I've only used minicom and cu, and had no isses with those: ) Oct 27 14:04:59 :) Oct 27 14:08:07 I have the same behaviour with minicom :) Oct 27 14:08:22 what UART<-> USB adapter chip are you using Oct 27 14:08:26 a pl2303? Oct 27 14:18:19 nope, ftdi Oct 27 14:19:58 hi Oct 27 14:20:56 Kaloz: what is the recommended way to add support for different targets for x86_64? Oct 27 14:21:07 I'm looking into adding a SMP-enabled target for the APU Oct 27 14:21:12 (PCEngines) Oct 27 14:22:01 hi. what's needed for it vs the normal target? Oct 27 14:23:37 Kaloz: smp Oct 27 14:23:59 the rest is more or less there, I run generic x86_64 (w/o smp) on an apu and it works well Oct 27 14:24:45 jow_laptop: well, then enabling smp is the way to go. only first-gen amd64 stuff were single core Oct 27 14:25:02 even those single cores should be able to run an smp-enabled kernel iirc Oct 27 14:25:08 jow_laptop: I'll do that when upgrading the kernel Oct 27 14:25:20 jow_laptop: sure, just saying smp is the standard there now Oct 27 14:26:33 Kaloz: ok, thanks :) Oct 27 14:27:34 zorun: I'll probably get around that later this week/early next week. I want to bump up thing to .17 Oct 27 14:27:35 there are few other stuff, like setting the console rate to 115200 instead of 38400 (since the BIOS is at 115200) Oct 27 14:27:44 but this is minor Oct 27 14:28:05 ack Oct 27 14:28:40 zorun: yeah, the default will be 115200 there, 38400 is a remainder from the legacy stuff Oct 27 14:31:01 mripard: once kaloz or someone pushes in the generic 3.17, sunxi will be bumped to that one as well Oct 27 14:35:35 wigyori: cool Oct 27 14:35:52 any chance for the 3.16 patches to come in? Oct 27 14:36:47 jow r43086 trunk/config/Config-images.in * config: use PARTUUID by default on x86_64 Oct 27 14:36:52 jow r43087 trunk/package/network/utils/iwinfo/ Makefile src * iwinfo: switch to external git repo, support lookup by phyname & macaddr Oct 27 14:38:54 build #710 of avr32 is complete: Failure [failed compile_5] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/710 Oct 27 14:44:31 mripard: don't look at me for that, sorry :) Oct 27 14:44:52 mripard: my guess is that 3.16 will be left out with 3.17 already out Oct 27 15:01:53 wigyori: ok, thanks Oct 27 15:26:22 how does procd work at reboot? i had to wait for 2 minutes until the reboot happends while procd is busy(25~35% cpu load) Oct 27 15:27:10 I did change procd watchdog timer from 30 second to 2 minutes, is that related? 30 second is too short(sometimes one core got stuck for a while and the system reboots unnecessarily) Oct 27 15:28:21 not sure if those respawn process are related, procd might be killing them and they keep respawning thus make the cpu busy? Oct 27 15:28:39 procd in trunk is crashing a lot on me at least in uml, wonder if it is related Oct 27 15:28:52 i'm on BB though Oct 27 15:28:56 ah Oct 27 15:28:59 then not related Oct 27 15:29:02 haven't seen it crashing in BB ;) Oct 27 15:29:18 it's not crashing, just takes a long while for reboot Oct 27 15:29:32 and it takes around 30% cpu during the reboot process Oct 27 15:29:40 i had to remount /proc to top it Oct 27 15:30:10 that cpu usage sounds weird Oct 27 15:30:19 normally it should just fire the stop scripts and reboot Oct 27 15:30:24 (of course, some stop scripts are slow) Oct 27 15:30:37 or it's sensitive to some initscripts i wrote, but i wrote similar scripts for sysv/busybox and did not see the delay-reboot Oct 27 15:30:49 could be that Oct 27 15:30:59 because the init scripts are probably causing the slowdown for you Oct 27 15:31:13 idli: one thing I don't get is that after 'stop' how will those 'respawn' process work Oct 27 15:31:25 stop should stop their respawn Oct 27 15:31:30 if procd sets to respawn, how can a 'stop' remove that? Oct 27 15:32:24 i found some processes are still respawned after a stop, probably i should not 'kill -9' in the stop call Oct 27 15:34:03 checked again, /etc/init.d/test stop does not respawn, however procd/reboot let it respawn, i did have a K90test Oct 27 15:52:25 blogic r43088 trunk/package/base-files/files/bin/config_generate Oct 27 15:52:25 base-files: config_generate did not handle setups with more than 1 switch properly Oct 27 15:52:33 blogic r43089 trunk/package/libs/libiconv/src/iconv.c * libiconv: do not replace untranslatable characters with * or ? Oct 27 15:52:40 blogic r43090 trunk/package/system/fstools/Makefile * fstools: update to latest git head Oct 27 16:01:43 rmilecki r43091 trunk/target/linux/generic/patches-3.14/ 060-net-usb-hso-Add-support-for-Option-GTM671WFS.patch 060-hso_devices.patch 785-hso-support-0af0-9300.patch * kernel: 3.14: use patch that is really upstreamed for hso 0af0:9200 Oct 27 16:10:35 anyone know the backstory on r43089? Oct 27 16:36:49 jow r43092 trunk/package/network/utils/iwinfo/Makefile * iwinfo: add cli support for phy lookup from uci section Oct 27 16:38:38 karlp: well its the standards compliant behaviour Oct 27 16:39:11 karlp: I suppose the previous "silently replace with ?" broke some programs that rely on EILSEQ to detect invalid or not representable encodings Oct 27 16:41:34 now you'll just get the programs that break because they don't expect to ever see EILSEQ :) Oct 27 16:41:43 I'm sure the EILSEQ path is the right way forward, no complaints there Oct 27 16:41:56 in this case they would break with gnu gettext as well Oct 27 16:42:02 just would have been nice to see a ticket link or anything to something that was breaking. Oct 27 16:42:05 note that the change above patched our own stub library Oct 27 17:04:51 do we need this patch? https://dev.openwrt.org/browser/trunk/target/linux/generic/patches-3.14/061-arm_xz_decompressor_build_fix.patch Oct 27 17:04:58 take a look at this file: https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/arch/arm/boot/compressed/decompress.c?h=linux-3.14.y Oct 27 17:05:08 it already includes Oct 27 17:05:09 extern char * strstr(const char * s1, const char *s2); Oct 27 17:43:14 zajec: it seems you don't need it Oct 27 17:43:33 great :) Oct 27 17:43:47 it's not used anyway Oct 27 17:44:24 zajec: are you at patch 061 already? Oct 27 17:44:44 i was just looking at random stuff Oct 27 17:44:46 give me a sec Oct 27 17:44:54 glad you're interested in it :) Oct 27 17:45:14 0xx patches are easy, they are upstremed mostly Oct 27 18:15:13 build #729 of ar7 is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/ar7/builds/729 Oct 27 18:15:47 rmilecki r43093 trunk/target/linux/ (170 files in 2 dirs) * kernel: start working on 3.18 support Oct 27 18:17:54 rmilecki r43094 trunk/include/kernel-version.mk * kernel: make it possible to select experimental 3.18 (3.18-rc2) Oct 27 18:19:11 build #627 of iop32x is complete: Failure [failed shell_10] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/627 Oct 27 21:06:01 Just did upgrade my local OpenWRT git trunk and noticed package/network/utils/iwinfo has undergone through an overhaul. I don't have this package selected, yet buildroot tried to compile it to no avail as shown here (http://pastebin.com/ihJqnRVa). I hope jow_laptop will be able to fix it. Oct 27 21:06:50 The error is on line #5 (iwinfo_cli.c:(.text.startup+0xd4): undefined reference to `iwinfo_backend_by_name'). Oct 27 21:20:07 zajec: 3.18's 220-gc_sections.patch looks busted Oct 27 22:09:31 swalker: of course Oct 27 22:09:38 swalker: that's what I call early/experimental Oct 27 22:10:15 swalker: it's a bit complex for 1 developer to update all patches, so I did what I could Oct 27 22:10:22 and I'll ask other devs for help Oct 27 22:22:15 maybe they should get upstream ;) I'm still merging in my local rc1 changes Oct 27 22:22:37 what do you mean by get upstream? Oct 27 22:25:13 in the kernel proper so openwrt devs don't have the maintenance burden Oct 27 22:25:59 swalker: no, no way, we don't do that, we don't send any patches upstream, we want them PRIVATE ;P Oct 27 22:56:47 hauke r43095 trunk/target/linux/generic/patches-3.18/ (38 files) * kernel: make the kernel 3.18 patches apply and boot on arm. Oct 27 22:59:20 hauke r43096 trunk/target/linux/generic/config-3.18 * kernel: add generic config for kernel 3.18 Oct 27 23:03:16 hauke r43097 trunk/target/linux/ (35 files in 2 dirs) * bcm53xx: initial support for kernel 3.18 Oct 27 23:05:15 hauke r43098 trunk/target/linux/bcm53xx/Makefile * bcm53xx: remove support for fpu Oct 27 23:07:46 Hauke: 259-regmap_dynamic.patch is missing a trailing " for REGMAP_I2C's tristate (and has for awhile iirc) Oct 27 23:08:44 swalker: thanks Oct 27 23:15:29 hauke r43099 trunk/target/linux/ generic/patches-3.13/259-regmap_dynamic.patch generic/patches-3.18/259-regmap_dynamic.patch generic/patches-3.14/259-regmap_dynamic.patch * kernel: add closing " to REGMAP_I2C's tristate Oct 28 00:02:13 build #708 of xburst is complete: Failure [failed compile_8] Build details are at http://buildbot.openwrt.org:8010/builders/xburst/builds/708 **** ENDING LOGGING AT Tue Oct 28 03:00:00 2014