**** BEGIN LOGGING AT Mon Nov 12 02:59:58 2012
Nov 12 06:43:02 <russell--> gah, wgt634u is broken?
Nov 12 06:43:40 <russell--> "Starting program at 0x80001000" is the last thing i see
Nov 12 06:43:55 <russell--> r34165
Nov 12 06:44:46 <russell--> was working as of r33891, i think ... confirming...
Nov 12 06:58:22 <russell--> yeah r33891 works
Nov 12 06:59:29 <russell--> trying rebuilding r34165 from a clean tree to confirm it's broken (that's bit my ass before)
Nov 12 07:20:40 <loswillios> russell--: noticed the same here
Nov 12 07:20:57 <loswillios> (wgt634u broken)
Nov 12 07:23:51 <russell--> r34026 is broken too
Nov 12 07:24:18 * russell-- commenses to bisect since such a brokenness is intolerable! ;-)
Nov 12 07:24:28 <russell--> commences?  something like that
Nov 12 07:28:09 <russell--> 6 steps, so i should be able to find it in about 3 hours ;-)
Nov 12 07:55:15 <russell--> r33955 is bad too
Nov 12 07:57:02 <OpenWrt-Commits> blogic r34166 trunk/target/linux/ (7 files in 3 dirs) * [atheros] move redboot boardconfig patch into target folder
Nov 12 08:14:44 <russell--> looks like 3 flash-related patches were made in the brcm47xx tree in the window of interest, r33935, r33920 and 33896.  i'm building r33920 right now.
Nov 12 08:20:59 <russell--> r33920 is bad too
Nov 12 08:22:03 <OpenWrt-Commits> kaloz r34167 trunk/target/linux/ (5 files in 4 dirs) * [cns3xxx]: split the i2c patch, merge the retry fixup
Nov 12 08:22:39 <OpenWrt-Commits> kaloz r34168 trunk/target/linux/ cns3xxx/patches-3.3/052-cns3xxx_spi.patch cns3xxx/files/drivers/spi/spi-cns3xxx.c cns3xxx/files/drivers/spi * [cns3xxx]: split spi patch, follow mainline file naming
Nov 12 08:45:53 <russell--> r33906 is okay
Nov 12 09:11:17 <russell--> pretty sure r33920 is the culprit, building r33919 now to confirm
Nov 12 09:11:43 <russell--> r33913 is okay
Nov 12 09:15:26 <blogic> russell--: you'll have to ping hauke then ;)
Nov 12 09:17:24 <russell--> jesus, r33920 is an 11k line patch!
Nov 12 09:18:19 <russell--> grrr
Nov 12 09:18:26 <OpenWrt-Commits> kaloz r34169 trunk/target/linux/cns3xxx/patches-3.3/ (43 files) * [cns3xxx]: reorder, rename and refresh patches
Nov 12 09:35:08 <russell--> confirmed, r33919 okay, r33920 bad
Nov 12 09:35:31 <russell--> can someone please hunt hauke down and break his kneecaps? ;-)
Nov 12 10:32:05 <russell--> sent a bug report to mailing list
Nov 12 10:41:05 <soma> i hope this gets fixed soon, one guy here uses wgt a lot
Nov 12 10:41:20 * russell-- has 50 or so deployed
Nov 12 10:42:41 <russell--> had it been reported already?  I didn't see it.
Nov 12 10:43:11 <soma> not to my knowledge
Nov 12 12:50:13 <OpenWrt-Commits> florian r34170 trunk/target/linux/ generic/patches-3.6/062-mips_decompressor_build_fix.patch ar7/patches-3.6/100-decompressor_memcpy_fix.patch * [kernel] MIPS: add a build fix for arch/mips/boot/compressed/decompress.c
Nov 12 13:17:46 <dape8708> dangole,  thanks! i have my dir 615 h1 on the table working fine since last friday !
Nov 12 13:18:56 <dangole> dape8708: i'm working on a new set of patches which can also go upstream
Nov 12 13:19:25 <dangole> the current version contained a misunderstanding about when the differentiated rf_vals have to be set
Nov 12 13:19:25 <dape8708> awesome, im here if you need me testing
Nov 12 13:19:39 <dangole> i'm working on fixing it now, i'd be glad if you can test stuff in ~0.5h
Nov 12 13:23:48 <dape8708> SURE
Nov 12 16:24:02 <DonkeyHotei> [Sun 2012-11-11 11:21:48 AM PST] <blogic> ok, gotta run, might have time to hold your hand tomorrow <------ Y/N?
Nov 12 18:28:45 <blogic> DonkeyHotei: you are complicated
Nov 12 18:28:53 <blogic> kindly asking if i am here came to your mind ?
Nov 12 18:29:32 <DonkeyHotei> i checked your idle time instead :P
Nov 12 18:32:00 <DonkeyHotei> anyway, i don't think it's really an lzma error, because the same thing happens with the flash image in the bootloop
Nov 12 18:38:46 <russell--> Hauke: how do i build with early_printk enabled?
Nov 12 18:44:46 <DonkeyHotei> blogic: i hope i haven't offended you
Nov 12 18:47:50 <blogic> russell--: its a kernel option
Nov 12 18:47:55 <blogic> which target ?
Nov 12 18:48:04 <blogic> DonkeyHotei: you'd have to try harder
Nov 12 18:48:23 <DonkeyHotei> heh
Nov 12 18:48:52 <russell--> blogic: thanks
Nov 12 18:48:56 <DonkeyHotei> anyway, got any more ideas how to make the ram img boot?
Nov 12 18:49:35 <blogic> menuconfig Target Images  --->  [*] ramdisk  --->
Nov 12 18:49:42 <blogic> then generat the image
Nov 12 18:49:56 <blogic> tftp 0x81000000 my_image; bootm 0x81000000
Nov 12 18:50:11 <KanjiMonster> russell--: you need to remove bcm47xx patch 116 first, which disables early printk
Nov 12 18:50:49 <blogic> DonkeyHotei: you will ahve to add spi-gpio to the image i think
Nov 12 18:52:38 <DonkeyHotei> but why is it failing lzma decode?
Nov 12 18:52:56 <blogic> because you loaded the image wrong most likely
Nov 12 18:53:20 <DonkeyHotei> i used "tftpboot"
Nov 12 18:54:01 <blogic> tftp 0x81000000 my_image; bootm 0x81000000
Nov 12 18:54:31 <DonkeyHotei> i will try
Nov 12 18:56:00 <DonkeyHotei> ok, that worked
Nov 12 18:58:01 <russell--> KanjiMonster: thanks also, going to try with just Hauke's mailing list patch first, as suggested
Nov 12 18:58:01 <blogic> i know
Nov 12 18:58:30 <blogic> DonkeyHotei: did it load the flash ?
Nov 12 19:00:52 <DonkeyHotei> it loaded the ram img
Nov 12 19:05:28 <DonkeyHotei> i will now rebuild with spi-gpio
Nov 12 19:05:54 <blogic> include it in the kernel
Nov 12 19:05:59 <blogic> not just a module
Nov 12 19:06:12 <DonkeyHotei> i included in the image; not enough?
Nov 12 19:06:15 <DonkeyHotei> ok
Nov 12 19:11:00 <DonkeyHotei> so, i add it to target/linux/lantiq/ar9/config-default?
Nov 12 19:11:33 <blogic> make kernel_menuconfig CONFIG_TARGET=subtarget
Nov 12 19:11:46 <blogic> then go to drivers -> spi -> add spi_gpio
Nov 12 19:12:00 <DonkeyHotei> with subtarget being ar9?
Nov 12 19:12:39 <blogic> no
Nov 12 19:12:41 <blogic> that line as is
Nov 12 19:12:46 <blogic> "make kernel_menuconfig CONFIG_TARGET=subtarge"
Nov 12 19:12:49 <blogic> "make kernel_menuconfig CONFIG_TARGET=subtarget"
Nov 12 19:13:10 <DonkeyHotei> literally?
Nov 12 19:14:10 <OpenWrt-Commits> jow r34171 trunk/package/network/ (6 files in 2 dirs)
Nov 12 19:14:10 <OpenWrt-Commits> ppp:
Nov 12 19:14:11 <OpenWrt-Commits>   - fix multilink ppp with custom ifnames, patch by George Kashperko <george@znau.edu.ua>
Nov 12 19:14:11 <OpenWrt-Commits>   - refresh patches
Nov 12 19:17:35 <OpenWrt-Commits> jow r34172 branches/attitude_adjustment/ (6 files in 2 dirs) * AA: ppp: backport r34171
Nov 12 19:18:22 <dangole> ping dape8708
Nov 12 19:18:39 <dangole> i posted some updated patches for dir-615h1 wifi on the mailing list
Nov 12 19:19:07 <dangole> remove the old file from package/mac80211/622*
Nov 12 19:19:15 <dangole> and apply those two to the openwrt instead
Nov 12 19:19:17 <dape> hey, ah, i arrived home, the router is at work on the desk, is tomorrow morning too late to test?
Nov 12 19:19:29 <dangole> not at all
Nov 12 19:19:34 <dangole> no worries
Nov 12 19:19:53 <dape> cool, thanks, so remove the manual patch and put your two from the mailing list, got it, writing memory
Nov 12 19:20:12 <dangole> :p
Nov 12 19:20:33 <dape> (was wondering whats with this ding in this textual new irc client,,back to sleep, thanks dangole)
Nov 12 19:21:17 <dangole> the second one changes some bits in the kernel to detect that 20MHz crystal already in the board setup phase
Nov 12 19:21:29 <dangole> so you'll have to rebuild everything, not just mac80211
Nov 12 19:21:35 <dangole> i'm here tomorrow
Nov 12 19:21:37 <dangole> good night
Nov 12 19:21:38 <dape> make clean; make
Nov 12 19:21:50 <dangole> make target/clean
Nov 12 19:21:51 <dangole> would do
Nov 12 19:21:54 <dape> oki
Nov 12 19:27:55 <OpenWrt-Commits> luka r34173 trunk/target/linux/cns3xxx/patches-3.3/075-spi_support.patch * cns3xxx: fix build failure after r34168
Nov 12 19:27:56 <OpenWrt-Commits> luka r34174 trunk/target/linux/cns3xxx/patches-3.3/ 100-gpio_irq.patch 095-gpio_support.patch * cns3xxx: merge gpio patches
Nov 12 19:39:25 <russell--> Hauke: your patch fixed it ... i'm booting again
Nov 12 19:45:51 <russell--> here is the bootlog: http://sprunge.us/KHZU
Nov 12 19:46:42 <luka> russell--: you should reply to the mailing list too ;)
Nov 12 19:46:57 <russell--> lots of "can not parse nvram name"
Nov 12 19:47:04 <russell--> luka, will do
Nov 12 19:52:31 <DonkeyHotei> blogic: ok, now what difference was spi-gpio supposed to make?
Nov 12 19:52:46 <blogic> mae the flash work
Nov 12 19:53:04 <DonkeyHotei> /proc/mtd is still empty
Nov 12 19:53:38 <DonkeyHotei> /proc/mtd is still empty
Nov 12 19:57:25 <DonkeyHotei> blogic: what could still be missing?
Nov 12 20:00:23 <DonkeyHotei> perhaps something in the platform init?
Nov 12 20:02:02 <blogic> DonkeyHotei: i am fixing dgn300 in the next 48 hours
Nov 12 20:02:05 <blogic> its already on my desk
Nov 12 20:02:12 <DonkeyHotei> 300?
Nov 12 20:03:11 <blogic> 3500
Nov 12 20:03:23 <DonkeyHotei> would you like any help?
Nov 12 20:03:29 <blogic> not really
Nov 12 20:05:43 <DonkeyHotei> i'd like to help anyway, if possible
Nov 12 20:06:03 <blogic> i will tell you when you can start testing
Nov 12 20:06:26 <DonkeyHotei> maybe some discussion of what could be wrong might yield quicker ideas for fixes
Nov 12 20:06:43 <DonkeyHotei> not just testing
Nov 12 20:07:01 <blogic> ?
Nov 12 20:07:05 <blogic> i know what is wrong
Nov 12 20:07:08 <blogic> the spi flash needs a fix
Nov 12 20:07:14 <blogic> and the gpios for the siwtch need to be set
Nov 12 20:07:26 <blogic> and i need to test wifi
Nov 12 20:07:31 <DonkeyHotei> will the switch support vlans?
Nov 12 20:07:42 <blogic> its a switch
Nov 12 20:07:45 <blogic> of course it will
Nov 12 20:08:06 <DonkeyHotei> i meant with existing drivers
Nov 12 20:08:15 <DonkeyHotei> and testing wifi should not require spi to work
Nov 12 20:08:16 <blogic> yes
Nov 12 20:08:22 <blogic> it does
Nov 12 20:08:28 <blogic> because the calibration data is in the spi flash
Nov 12 20:08:34 <DonkeyHotei> oh.
Nov 12 20:22:49 <DonkeyHotei> blogic: i found that if i name the image "uImage" i can simply type "run net_ram" in uboot
Nov 12 20:27:19 <Hauke> russell--: I am back, thanks for testing the patch
Nov 12 20:39:16 <mirko> written_tabletpc: pong
Nov 12 20:58:58 <russell--> Hauke: no problem, glad to help
Nov 12 21:32:58 <OpenWrt-Commits> hauke r34175 trunk/target/linux/brcm47xx/patches-3.3/820-wgt634u-nvram-fix.patch * brcm47xx: fix cfe config detection on WGT634U broken in r33920
Nov 12 21:34:25 <OpenWrt-Commits> hauke r34176 trunk/target/linux/ brcm47xx/profiles/110-Bcm4705-b43.mk brcm47xx/profiles/115-Bcm4705-none.mk * brcm47xx: remove double definition of kmod-tg3 added in r34150
Nov 12 21:35:27 <OpenWrt-Commits> blogic r34177 trunk/target/linux/ ramips/image/Makefile ramips/files/drivers/net/ethernet/ramips/ramips_main.c * [ramips] move ethernet hw init to init/uninit functions
Nov 12 21:36:06 <OpenWrt-Commits> hauke r34178 branches/attitude_adjustment/target/ linux/brcm47xx/profiles/110-Bcm4705-b43.mk linux/brcm47xx/profiles/115-Bcm4705-none.mk * brcm47xx: backport of r34176
Nov 12 21:36:48 <OpenWrt-Commits> blogic r34179 trunk/target/linux/ramips/image/Makefile * [ramips] previous commit pushed 1 file too many ;)
Nov 12 22:16:26 <russell--> hmm
Nov 12 22:16:29 <russell--> new problem
Nov 12 22:17:04 <russell--> if i flash from the now booting image, the next boot comes up broken, drops to CFE
Nov 12 22:18:28 <russell--> most recently, i built r34179, flashed from a way earlier version, boots, reboots, then sysupgrade to *the same image* and it's busted
Nov 12 22:20:24 <russell--> Loading: Failed.
Nov 12 22:20:24 <russell--> Could not load : Not an ELF-format executable
Nov 12 22:20:24 <russell--> CFE>
Nov 12 22:21:11 * russell-- going to try the completely vanilla image to see if that does the same thing
Nov 12 22:21:30 <russell--> (this was a custom image)
Nov 12 23:20:45 <russell--> yeah, confirmed with vanilla image
Nov 12 23:20:49 <russell--> wtf?
Nov 12 23:22:41 * russell-- guesses it has something to do with where it thinks linux partition starts
Nov 12 23:36:00 <russell--> ah yeah:
Nov 12 23:36:02 <russell--> [    0.884000] 0x000000000000-0x000000080000 : "cfe"
Nov 12 23:36:03 <russell--> [    0.900000] 0x000000080000-0x0000007e0000 : "linux"
Nov 12 23:36:06 <russell--> that's wrong
Nov 12 23:36:42 <russell--> should be:
Nov 12 23:36:44 <russell--> 0x000000000000-0x0000000a0000 : "cfe"
Nov 12 23:36:45 <russell--> 0x0000000a0000-0x0000007e0000 : "linux"
Nov 13 02:45:57 <written_tabletpc> what about the fltk2 compile error?
**** ENDING LOGGING AT Tue Nov 13 02:59:57 2012