**** BEGIN LOGGING AT Sat Jan 21 02:59:57 2012
Jan 21 06:04:38 <swalker> philipp64|laptop: build failures
Jan 21 06:05:00 <russell--> what does "NOHZ: local_softirq_pending 08" mean?
Jan 21 06:05:17 <russell--> it seems to be associated with ethernet not working
Jan 21 06:05:26 <russell--> (on a v2 wrt54g)
Jan 21 06:11:29 <russell--> weirdly, i'm seeing it on two versions of openwrt, a today-ish verion and an old one that has been running for a while (from last may)
Jan 21 06:11:40 <russell--> which makes me think hardware
Jan 21 06:18:55 <philipp64|laptop> walker: I don't understand. libcap built in the past. what's broken?
Jan 21 06:19:06 <philipp64|laptop> russell--: how's the weather there? :-P
Jan 21 06:19:22 <philipp64|laptop> it's rainy here in Boise, but nothing like Stumptown.
Jan 21 06:24:16 <swalker> philipp64|laptop: http://buildbot.openwrt.org:8010/broken_packages/x86/squid/compile.txt
Jan 21 06:26:51 <russell--> rainy, we got a little snow, but it was just slush by morning, never really got below 34F where i am.  pathetic.
Jan 21 06:33:19 <CIA-84> mirko * r29838 /feeds/xorg/lib/qt4/Makefile:
Jan 21 06:33:19 <CIA-84> Revert "qt4: Add basic font dependency to qt4-gui"
Jan 21 06:33:19 <CIA-84> This reverts commit 29192.
Jan 21 06:33:19 <CIA-84> Depending on font-basic implicitly makes qt4-gui depending on FEATURE_drawing-backend_libX11
Jan 21 06:33:19 <CIA-84> which is not set for makes qt4 on top of LinuxFB / DirectFB unusable (since unselectable).
Jan 21 09:51:05 <CIA-84> jow * r29839 /trunk/tools/libtool/patches/ (000-relocatable.patch 200-openwrt-branding.patch): [tools] libtool: fix accidental sed invocation in libtoolize
Jan 21 11:05:54 <glp> russell--: about the floss weekly
Jan 21 11:44:25 <builder_> build #129 of s3c24xx is complete: Failure [failed compile_10]  Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/129
Jan 21 12:57:31 <btsimonh> blogic: are you about?
Jan 21 12:57:56 <blogic> yes
Jan 21 12:58:01 <blogic> i just replied to your mail
Jan 21 12:58:14 <btsimonh> :) thanks.  i'll look there first
Jan 21 13:02:01 <btsimonh> i've got an hour or two before my wife recovers from here hangover :)  should I wait; or should i take a look at the nand & ebu stuff to see if I can get it to operate?, i feel I'm 'on a roll' after my recent small successes.
Jan 21 13:04:27 <builder_> build #126 of brcm47xx is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx/builds/126
Jan 21 13:08:22 <builder_> build #124 of atheros is complete: Failure [failed shell_10]  Build details are at http://buildbot.openwrt.org:8010/builders/atheros/builds/124
Jan 21 13:11:13 <btsimonh> blogic: do you have a danube manual/datasheet?
Jan 21 13:29:27 <blogic> why ?
Jan 21 13:29:34 <blogic> what info are you looking for ?
Jan 21 13:32:55 <btsimonh> I'm trying to understand the whole ebu thingy...  e.g. why both nand and GPIO are at 14000000.  gpio_ebu seems to use address region 1?
Jan 21 13:36:10 <blogic> ok
Jan 21 13:36:17 <blogic> ebu == externa bus unit
Jan 21 13:36:34 <blogic> tthere are 2 memory windows that will be mapped to te pins on this bus
Jan 21 13:36:38 <blogic> addrsel0/1
Jan 21 13:36:50 <blogic> normally a board has nand or nor
Jan 21 13:37:06 <blogic> so the kernel sets up addrsel0 and points either nand or nor at it
Jan 21 13:37:23 <blogic> i have so far not seen a board that has both
Jan 21 13:37:41 <blogic> so we need to simply make te nand driver always use addsel1 and then the 2 can co exist
Jan 21 13:37:56 <blogic> is that clearer ?
Jan 21 13:38:52 <btsimonh> :) I got that bit, but I'm not sure why the andn and gpio are at the same address...  GPIO seems to be working, so and can't be there?
Jan 21 13:39:08 <blogic> gpio ?
Jan 21 13:39:16 <btsimonh> gpio_ebu
Jan 21 13:39:23 <blogic> ah gpio_ebu will move to addsel2
Jan 21 13:39:25 <blogic> or
Jan 21 13:39:35 <blogic> does your board even have gpio_ebu ?
Jan 21 13:40:17 <btsimonh> good point...  I've got leds and buttones..  let me see..
Jan 21 13:40:28 <blogic> buttons are not gpio_ebu
Jan 21 13:40:38 <blogic> those are output only gpios
Jan 21 13:41:24 <btsimonh> ok, leds are gpio_stp
Jan 21 13:42:12 <btsimonh> so for nand, before each access, it will need to change ebu to addsel1, like the gpio_ebu does?
Jan 21 13:45:12 <btsimonh> but the nand.c in xway, I don't see how all accesses are done... there does not seem enough callbacks to operate the whole chip from here..
Jan 21 13:47:03 <russell--> glp: yes?
Jan 21 13:48:25 <blogic> ?
Jan 21 13:48:37 <blogic> it works on addrsel0 so it will work on the other aswell
Jan 21 13:50:44 <builder_> build #122 of cobalt is complete: Failure [failed compile_10]  Build details are at http://buildbot.openwrt.org:8010/builders/cobalt/builds/122
Jan 21 13:52:24 <russell--> glp: what about floss weekly?
Jan 21 13:53:52 <btsimonh> it's strange; it's as though it's already on adsel1, from xway_nand_probe?
Jan 21 13:55:02 <btsimonh> is all we need to do to un-protect the region by writing to BUSCON1 in sysctl.c?
Jan 21 13:56:00 <blogic> 1 sec
Jan 21 13:57:57 <blogic> yes, that could work
Jan 21 13:58:19 <blogic> show me your mach file please
Jan 21 13:58:26 <glp> russell--: you mentioned floss weekly the other day, and I just wanted to mention that I've been working on setting up a date for an openwrt edition
Jan 21 13:59:20 <btsimonh> https://sourceforge.net/p/hh2b4ever/code/ci/0ef5fe3abce50ecd85cbd6db66e90105aa5ec944/tree/trunk/build_dir/linux-lantiq_danube/linux-3.1.4/arch/mips/lantiq/xway/mach-bthomehubv2b.c
Jan 21 14:01:41 <blogic> ok
Jan 21 14:02:22 <blogic> you will need this patch for a start http://dev.phrozen.org/gitweb/?p=lantiq-queue.git;a=blobdiff;f=arch/mips/lantiq/xway/nand.c;h=604ac6d028cfa2248c9bc8166746ee9dc2e5d637;hp=4768d9b153bc560a9317f971b72238f6dab55905;hb=7c132829e63fd38c243609291474945cd84b6dab;hpb=3bf8730930890aba5f3258e753d08dea9c940f17
Jan 21 14:02:48 <blogic> that way you can pass the nand init code a partition map
Jan 21 14:02:56 <blogic> that bit is missing from trunk at the moment
Jan 21 14:04:17 <btsimonh> Thankyou!  I have a partition table defines and ready,... :)
Jan 21 14:07:36 <btsimonh> for consistancy, before comitting this one, can we have those partitons passed as physmap_flash_data, same a NOR (don't want to be picky, will work as is :) ).
Jan 21 14:08:38 <blogic> ?
Jan 21 14:08:45 <blogic> dont get the question
Jan 21 14:08:55 <russell--> glp: cool, have you mail'd merlyn@stonehenge.com?
Jan 21 14:09:11 <blogic> btsimonh: basically... that driver runs on 1 board
Jan 21 14:09:22 <blogic> and makes assumption about the ebu being in some state
Jan 21 14:09:46 <blogic> ... set by the bootloader. so it must be a diffeent default state and some setting is missing
Jan 21 14:10:05 <blogic> and show me a bootlog so i see the error message please
Jan 21 14:10:09 <btsimonh> (compare the new xway_register_nand with ltq_register_nor - the arguments could the the same type; like i said, not problem :) )
Jan 21 14:10:17 <btsimonh> ok, just re-building now
Jan 21 14:18:10 <btsimonh> a few more moments...
Jan 21 14:24:38 <btsimonh> I see where I call register, but don't see a failure or anything saying it's looking for nand.  Maybe I have nand disabled in config, I've added a printk in probe, and will check config and let you know.
Jan 21 14:25:39 <blogic> ah
Jan 21 14:25:50 <blogic> enable plat_nand
Jan 21 14:25:57 <blogic> and add the partition to it
Jan 21 14:26:02 <blogic> maybe that is all that is missing
Jan 21 14:26:57 <btsimonh> nor is wokring, so have mtd.  can't find nand in make menuconfig yet - is it only in kernel_menuconfig?
Jan 21 14:27:41 <blogic> kenrel
Jan 21 14:27:44 <blogic> wait
Jan 21 14:29:35 <blogic> the option is called "Support for generic platform NAND drive
Jan 21 14:29:50 <blogic> drivers->mtd->nand->
Jan 21 14:30:42 <btsimonh> so with our local xway registration (which we know we have), the generic is enough extra....  got it..
Jan 21 14:31:33 <btsimonh> ok, building. hope a make target/install is enough.
Jan 21 14:31:51 <blogic> yes
Jan 21 14:34:57 <btsimonh> ok, hangs after NOR adds it's partitions.  my probe printk was after the probe; I'll move it to before
Jan 21 14:36:32 <btsimonh> what are the ltq_gpio_request in nand.c?  how are these referenced?
Jan 21 14:43:15 <btsimonh> blogic: it hangs AFTER the probe. I added ltq_ebu_w32(ltq_ebu_r32(LTQ_EBU_BUSCON1) & ~EBU_WRDIS, LTQ_EBU_BUSCON1); to the top of the probe fn
Jan 21 15:00:50 <glp> russell--: what was the stonehenge.com thing about?
Jan 21 15:02:04 <glp> I used another mail adress for randal ...
Jan 21 15:02:46 <btsimonh> blogic: it hangs in xway_chip_select(0)
Jan 21 15:03:08 <btsimonh> now building with the enable at the start of every function as a test
Jan 21 15:50:33 <CIA-84> jow * r29840 /packages/admin/htop/Makefile: [packages] htop: disable unicode to remove libncursesw requirement
Jan 21 15:57:24 <Olipro> mazilo ^
Jan 21 16:23:33 <btsimonh> blogic: http://pastebin.com/mr8X9rve is kernel log.  I've got nowhere, but added lots of diags.  if you get a moment to look, email me.  Wife's up and need a break :)
Jan 21 16:23:44 <btsimonh> nand.c is here http://pastebin.com/kwXCcxRz
Jan 21 16:24:07 <blogic> ok
Jan 21 16:27:29 <Llamafarmers> btsimonh: so, that device has both a NOR and NAND chip?
Jan 21 16:27:40 <Llamafarmers> I thought the AMD AM29LV400BB was NOR
Jan 21 18:50:22 <CIA-84> jow * r29841 /packages/libs/curl/patches/100-cross-compile.patch: [packages] curl: restore 100-cross-compile.patch, the current version also does LD_LIBRARY_PATH crap
Jan 21 19:39:02 <nbd> jr-: ping
Jan 21 19:40:03 <kristianpaul> pong
Jan 21 19:40:52 <CIA-84> mirko * r29842 /trunk/toolchain/eglibc/patches/2.14/ (4 files): [toolchain/eglibc] eglibc 2.14 needs our eglibc patchset as well
Jan 21 19:41:07 <nbd> kristianpaul: ?
Jan 21 19:41:34 <kristianpaul> oops
Jan 21 19:41:42 <kristianpaul> i misread ;)
Jan 21 19:41:44 <kristianpaul> sorry
Jan 21 19:41:49 <nbd> ;)
Jan 21 19:41:50 <kristianpaul> also jr- :)
Jan 21 19:42:38 <philipp64|laptop> swalker: are you building with CAP enabled?
Jan 21 19:46:53 <philipp64|laptop> if libcap has been enabled, then you should probably compile and link against it conditionally...
Jan 21 20:01:39 <russell--> glp: oh, when i was suggesting nbd, that's the one randal gave me to use.
Jan 21 20:20:52 <jr-> me? what did i do this time? :)
Jan 21 20:31:50 <nbd> jr-: we talked about some netifd issues a while back, right?
Jan 21 20:37:39 <jr-> yes
Jan 21 20:49:05 <nbd> i've reworked the code to hopefully fix the interface restart issues that you had
Jan 21 20:49:59 <nbd> http://nbd.name/netifd-update.patch
Jan 21 20:50:05 <nbd> could you please test it?
Jan 21 20:51:45 <jr-> sounds great. will need to wait until tomorrow, but i'll get back to you with results/questions/whatever then.
Jan 21 20:53:32 <jr-> i also have new item to netifd wishlist. :)
Jan 21 20:53:42 <jr-> http://www.spinics.net/lists/netdev/msg186287.html
Jan 21 22:40:37 <CIA-84> hauke * r29843 /trunk/ (12 files in 2 dirs):
Jan 21 22:40:37 <CIA-84> brcm47xx: add support for pcie host controller on bcma based SoCs.
Jan 21 22:40:37 <CIA-84> This also adds minimal support for the BCMA43224 pcie wireless card.
Jan 21 22:45:24 <CIA-84> juhosg * r29844 /trunk/target/linux/generic/files/drivers/net/phy/ (rtl8366_smi.c rtl8366_smi.h rtl8366rb.c rtl8366s.c):
Jan 21 22:45:24 <CIA-84> generic: rtl8366: add rtl8366_smi_write_reg_noack helper
Jan 21 22:45:24 <CIA-84> After issuing a soft reset on the RT8366{S,RB}
Jan 21 22:45:24 <CIA-84> switch, waiting for the last acknowlegement fails
Jan 21 22:45:24 <CIA-84> in rtl8366_smi_write_reg. Add a _noack version of
Jan 21 22:45:24 <CIA-84> the function and use that for issuing a soft reset.
Jan 22 00:51:00 <builder_> build #89 of mpc52xx is complete: Failure [failed compile_10]  Build details are at http://buildbot.openwrt.org:8010/builders/mpc52xx/builds/89
**** ENDING LOGGING AT Sun Jan 22 02:59:57 2012