**** BEGIN LOGGING AT Sun Aug 14 02:59:57 2011
Aug 14 03:23:09 <cshore> hey nbd, I've created a little utility called rereadpart to reread the partition tables after sysupgrade writes the image to the block devices (on x86).  Where should I stash it?
Aug 14 03:23:33 <cshore> it's needed e.g. if using ext4 rootfs on x86
Aug 14 03:24:48 <cshore> I'm also planning on making sure mtd refresh works on x86...right now it doesn't
Aug 14 07:31:27 <CIA-89> loswillios * r27981 /packages/utils/tmux/Makefile: [packages] tmux: fix librt dependency (#9938)
Aug 14 10:35:37 <EqUaTe> hm.. does freeswitch actually build at all these days? or not so much?
Aug 14 11:00:29 <swalker> updated openwrt/upstream, https://home.comcast.net/~sdwalker/uscan/index.html
Aug 14 11:41:02 <Hauke> RealOpty: do you have a ssb based bcm47xx device with serial flash?
Aug 14 14:37:22 <Chocky> moo
Aug 14 14:47:28 <EqUaTe> sorry about the connect spam guys.
Aug 14 14:49:13 <WillieNL> Chocky: using gcc 4.4.6 this problem do not happend to me https://dev.openwrt.org/ticket/7256
Aug 14 14:50:04 <WillieNL> it just works more or less :)
Aug 14 14:50:29 <Chocky> Unsure why you are telling me.  Tell a dev to close it; it looks old/irrelevant.
Aug 14 14:51:38 <WillieNL> dunno if you still need gcc inside the target image, else it posible can be helpfull
Aug 14 14:51:54 <Chocky> I don't know what that means.
Aug 14 14:52:13 <WillieNL> then i dunno
Aug 14 14:52:29 <Chocky> Versions of GCC have long since moved on from those mentioned in that bug.
Aug 14 14:52:41 <WillieNL> hmm 4.4.6 is still in the trunk
Aug 14 14:53:18 <Chocky> "stiil"?  4.4.6 is not mentioned in that bug report.
Aug 14 14:53:24 <Chocky> still
Aug 14 14:53:45 <WillieNL> but using gcc 4.4.6 i can finish the build without issues.
Aug 14 14:53:58 <Chocky> I know; you said.  hence my comment about having someone close it.
Aug 14 14:54:04 <WillieNL> just remove  depends (avr32 || ubicom32) from config.in and tada
Aug 14 14:54:06 <Chocky> but it's not my issue; nothing I can do about it
Aug 14 14:55:11 <WillieNL> all i wanted to say that it will work fine with eglibc 2.12/2.13 and glibc 2.12/2.13
Aug 14 14:55:30 * Chocky repeats himself again.  Have someone close it. 
Aug 14 14:56:06 <WillieNL> nope
Aug 14 14:56:13 <Chocky> whatever.
Aug 14 14:56:50 <WillieNL> hmm indeed never mind
Aug 14 14:58:25 <Chocky> sigh.
Aug 14 14:59:56 <Chocky> in the time you were repeating yourself, you could have easily have added a note to the bug to say you don't think it's an issue any longer.  So hard?
Aug 14 15:29:20 <Chocky> nbd: I see in the bootlog on the wiki, a message about this (presumably) code being disabled.    Hence your suggestion.  I guess someone thought it was fixed
Aug 14 15:29:39 <Chocky> running now with that code disabled. so we'll see
Aug 14 15:54:18 <Chocky> working so far.  Lasted much longer than before.
Aug 14 15:54:20 <builder_> build #67 of lantiq is complete: Failure [failed compile_6]  Build details are at http://buildbot.openwrt.org:8010/builders/lantiq/builds/67
Aug 14 16:26:51 <Chocky> hm.
Aug 14 16:27:08 <Chocky> jffs2 eventually started complaining about errors.  on reboot, the FS is trashed
Aug 14 17:25:33 <Chocky> nbd: I will put in the line of code that was present in 2.6.32.10 et al.   if (mode == FL_WRITING) goto sleep;
Aug 14 18:56:17 <Chocky> nbd: looks better. will follow up to list
Aug 14 20:10:01 <builder_> build #64 of ps3 is complete: Failure [failed compile_6]  Build details are at http://buildbot.openwrt.org:8010/builders/ps3/builds/64
Aug 14 20:22:53 <philipp64|laptop> can we get 2.6.39.4 and the release in include/kernel-version.mk?
Aug 14 21:37:49 <philipp64|laptop> nbd: hmmm... added the following but it doesn't seem to be enough... http://fpaste.org/tm1V/
Aug 14 21:38:12 <philipp64|laptop> doing a "cat" of trigger doesn't show phy0tpt up...
Aug 14 23:38:58 <CIA-89> jogo * r27982 /trunk/package/kernel/Makefile:
Aug 14 23:38:58 <CIA-89> package/kernel: Fix AddDepends/SetDepends calls with make 3.82
Aug 14 23:38:58 <CIA-89> The module definitions depend on 001-depends.mk being the first included,
Aug 14 23:38:58 <CIA-89> but make 3.82 does not sort the results of wildcards anymore, so do an
Aug 14 23:38:58 <CIA-89> explicit sort of the includes.
Aug 14 23:38:59 <CIA-89> Found by Philip Prindeville, thanks!
Aug 14 23:39:00 <CIA-89> jogo * r27983 /trunk/package/kernel/modules/other.mk:
Aug 14 23:39:00 <CIA-89> package/kernel: kmod-hid: can't mix and match DEPENDS:= and $(call SetDepends/...)
Aug 14 23:39:01 <CIA-89> Since the SetDepends macro does a:
Aug 14 23:39:01 <CIA-89> DEPENDS:= ...
Aug 14 23:39:02 <CIA-89> it can't be used after setting DEPENDS:= explicitly, or it will clobber the previous value.
Aug 14 23:39:02 <CIA-89> Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
Aug 15 00:17:50 <CIA-89> nbd * r27984 /trunk/target/linux/generic/ (3 files in 3 dirs):
Aug 15 00:17:50 <CIA-89> kernel: disable cfi cmdset 0002 erase suspend
Aug 15 00:17:50 <CIA-89> on some platforms, erase suspend leads to data corruption and lockups when write
Aug 15 00:17:50 <CIA-89> ops collide with erase ops. this has been observed on the buffalo wzr-hp-g300nh.
Aug 15 00:17:50 <CIA-89> rather than play whack-a-mole with a hard to reproduce issue on a variety of devices,
Aug 15 00:17:50 <CIA-89> simply disable erase suspend, as it will usually not produce any useful gain on
Aug 15 00:17:51 <CIA-89> the small filesystems used on embedded hardware.
Aug 15 00:18:47 <nbd> Chocky: ^^
Aug 15 00:19:01 <Chocky> \o/
Aug 15 00:20:03 * Chocky hugs
Aug 15 00:20:55 <nbd> just arrived in vancouver, canada ;)
Aug 15 00:21:07 * Chocky waves over the border
Aug 15 00:36:56 <philipp64|laptop> nbd: I think I missed something in adding throughput support, but I'm not sure what.
Aug 15 00:37:53 <nbd> are you sure that ATH_STAT_LEDSOFT is set?
Aug 15 00:39:54 <philipp64|laptop> looking...
Aug 15 00:41:18 <nbd> btw. about the throughput table, the throughput that is checked against it is not the phy rate
Aug 15 00:41:26 <nbd> it's actual kbit/s throughput of rx/tx packets
Aug 15 00:43:55 <philipp64|laptop> ok. good to know. it seems that pci_match_id(&ath5k_led_devices, ... in order for the bit to get turned on. so if it's not a known device, it doesn't happen.
Aug 15 00:49:29 <philipp64|laptop> ok, I'm confused... ATH_LED() describes a GPIO associated with the card? but... I'm using a miniPCI card, and the GPIO # depends on which SBC I'm putting it in.
**** ENDING LOGGING AT Mon Aug 15 02:59:57 2011