| 01:27.42 | *** join/#edev prpplague (~prpplague@107-206-64-184.lightspeed.rcsntx.sbcglobal.net) |
| 01:32.51 | *** join/#edev JoeLlama (~snork@unaffiliated/joellama) |
| 02:01.28 | *** join/#edev Belliash (~belliash@asiotec/constitutive/belliash) |
| 02:28.36 | *** join/#edev Lars_G (~Lars@unaffiliated/lars-g/x-000001) |
| 03:39.32 | *** join/#edev orhan89 (~orhan89@103.10.171.14) |
| 03:40.22 | *** join/#edev NightMonkey (~NightrMon@173-164-139-193-SFBA.hfc.comcastbusiness.net) |
| 03:40.22 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 04:31.43 | *** join/#edev [7] (~quassel@rockbox/developer/TheSeven) |
| 05:22.11 | *** join/#edev kuldeepdhaka (~kuldeepdh@117.254.216.202) |
| 05:27.10 | *** join/#edev MonMotha (~monmotha@2001:470:c5f2:1:224:7eff:fedc:893e) |
| 05:59.04 | *** join/#edev ffio (~fire@unaffiliated/security) |
| 06:23.03 | *** join/#edev clio (~andrej@85.159.109.222) |
| 06:36.58 | *** join/#edev djinni (~djinni@li125-242.members.linode.com) |
| 07:05.06 | *** join/#edev mozmck (~moses@client-67.210.159.209.dfwtx.partnershipbroadband.com) |
| 07:42.42 | *** join/#edev Belliash (~belliash@gateway/shell/bellsrv/x-qydbqavfevvtzivj) |
| 07:42.42 | *** join/#edev Belliash (~belliash@asiotec/constitutive/belliash) |
| 08:17.19 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 08:36.02 | *** join/#edev sjhill (~sjhill@50-81-92-201.client.mchsi.com) |
| 08:36.03 | *** mode/#edev [+o sjhill] by ChanServ |
| 08:36.25 | *** join/#edev spectrum_ (~regrax@host178-58-dynamic.25-79-r.retail.telecomitalia.it) |
| 08:36.42 | spectrum_ | hi all |
| 09:55.25 | *** join/#edev n13z (~iosick@unaffiliated/n13z) |
| 10:09.35 | *** join/#edev mbuf (~user@49.248.86.150) |
| 10:29.54 | *** join/#edev devilsadvocate (~quassel@109.200.19.188) |
| 10:38.22 | *** join/#edev devgiant (~devgiant@unaffiliated/devgiant) |
| 12:31.10 | *** join/#edev g1powermac (~g1powerma@unaffiliated/g1powermac) |
| 13:38.44 | *** join/#edev n13z (~iosick@unaffiliated/n13z) |
| 14:28.30 | *** join/#edev mozmck (~moses@client-67.210.159.209.dfwtx.partnershipbroadband.com) |
| 14:54.16 | *** join/#edev prpplague (~prpplague@107-206-64-184.lightspeed.rcsntx.sbcglobal.net) |
| 14:55.38 | *** join/#edev ffio (~fire@unaffiliated/security) |
| 15:01.02 | *** join/#edev ccole (~cole@97.107.140.186) |
| 15:18.45 | *** join/#edev prpplague (~prpplague@107.206.64.184) |
| 15:37.46 | *** join/#edev sndcrb (~sndcrb@dsl-hkibrasgw1-58c385-172.dhcp.inet.fi) |
| 15:43.36 | *** join/#edev luffe48 (~luffe48@0407ds1-esp.0.fullrate.dk) |
| 15:51.36 | luffe48 | so i'm trying to get a new linux kernel (and later, a new userland) on to this 7 year old nas |
| 15:51.45 | luffe48 | but i can't get the stupid thing to boot it |
| 15:53.30 | luffe48 | how am i supposed to set up the kernel, apart from selecting a somewhat matching board type and churning out a .config? |
| 15:54.41 | luffe48 | it's a mpc8343 based board with an old (pre dt, i think) u-boot |
| 15:55.51 | SpeedEvil | 'somewhat' may not cut it. |
| 15:56.25 | SpeedEvil | It's not like x87 really - out of tree patches that are required to boot at all are not uncommon. |
| 15:56.33 | SpeedEvil | x86 |
| 15:56.42 | SpeedEvil | Or - to be more accurate - PCs. |
| 15:57.07 | SpeedEvil | What do you mean by 'can't boot' - how are you observing it. |
| 15:57.18 | luffe48 | i have an ancient kernel source from from the manufacturer that's patched all to hell |
| 15:57.28 | luffe48 | it just resets immediately |
| 15:58.26 | luffe48 | i was hoping to just be able to grab some offsets and whatnot out of some headers, plop that into a device tree and be on my way |
| 15:58.34 | luffe48 | but i have absolutely no idea where to start |
| 15:59.42 | SpeedEvil | resets at what point - have you got a serial console? |
| 15:59.52 | SpeedEvil | uboot? |
| 15:59.57 | luffe48 | yeah, i have a serial console |
| 16:00.11 | luffe48 | after u-boot uncompresses the kernel image, it just resets |
| 16:00.36 | luffe48 | so i guess the kernel does /something/, but not enough to get the serial going |
| 16:05.18 | luffe48 | the image im trying to boot is one with a dtb baked in - cuImage.mpc834x_mds on a mpc8343 based board |
| 16:06.15 | luffe48 | the soc and uart seems to have the right addresses, so i'm not sure where i'm screwing up |
| 16:07.30 | luffe48 | it's my understanding that all the board-specific stuff goes in the device tree now, not sure if that's right either |
| 16:09.18 | luffe48 | oh, and just to be clear, it's a brand new kernel source i'm building |
| 16:30.05 | *** join/#edev orbarron (~orbarron@nat/ti/x-qfztiglblndpnjig) |
| 16:50.03 | *** join/#edev addams__ (~addams@i577B6E41.versanet.de) |
| 17:19.12 | *** join/#edev prpplague (~danders@rrcs-97-77-26-26.sw.biz.rr.com) |
| 17:21.38 | SpeedEvil | http://interviews.slashdot.org/story/13/12/16/1457256/interview-ask-forrest-mims-about-rockets-electronics-and-engineering |
| 17:21.54 | *** join/#edev ffio (~fire@unaffiliated/security) |
| 17:27.45 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 17:40.08 | *** join/#edev florian (~fuchs@Maemo/community/contributor/florian) |
| 18:35.36 | prpplague | SpeedEvil: nice! |
| 18:45.24 | clever | luffe48: have you tried turning on early printk? |
| 18:46.38 | luffe48 | clever: it's on in the kernel .config |
| 18:47.14 | clever | cant think of anything else right now |
| 18:47.27 | clever | i was trying to get device tree to boot on a raspberrypi, i had to turn on early printk to get anything out of it |
| 18:47.53 | clever | the problem seems to be related to mailboxes in my case, its locking up before it can init the framebuffer or the usual uart channels |
| 18:55.04 | luffe48 | this is sort of frustrating. i've been at it for 3 days now, with not as much as a single character from the kernel yet |
| 19:26.08 | luffe48 | SUCCESS! i needed to add a chosen/linux,stdout-path node to my dts - now i just got some output! |
| 19:26.25 | luffe48 | it still dies horribly, but at least i get something |
| 19:27.44 | SpeedEvil | Woo! |
| 20:20.07 | *** join/#edev devilsadvocate (~quassel@109.200.19.188) |
| 20:21.10 | *** join/#edev devilsadvocate (~quassel@109.200.19.188) |
| 20:22.51 | clever | luffe48: i had hard-coded that into my .config file |
| 20:23.02 | clever | but it may differ by the base platform you pick |
| 20:27.57 | *** join/#edev NightMonkey (~NightrMon@173-164-139-193-SFBA.hfc.comcastbusiness.net) |
| 20:27.57 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 20:29.19 | luffe48 | i wonder if i might have more fun patching the old kernel up from 2.6.11 to current, because this really blows |
| 20:37.35 | clever | in my case, i had to disable device tree, and boot a working kernel |
| 20:37.44 | clever | then modify that to printk stuff at key points, and learn how it works |
| 20:37.55 | clever | then switch to a device tree one, and see what does work |
| 20:37.59 | clever | and where it failed |
| 20:39.47 | luffe48 | that sounds unpleasant |
| 20:40.14 | clever | i traced it down to either interupts not firing |
| 20:40.22 | clever | the memory mapped io for the mailbox is wrong (causing the event for the irq to never start) |
| 20:40.30 | clever | or mutex problems |
| 20:43.01 | luffe48 | ugh, yeah, i've sort of glossed over interrupts |
| 20:43.34 | luffe48 | it's going to be a pain digging that out of the old kernel |
| 20:43.41 | clever | in my case, the mailbox api writes some data to a special register |
| 20:43.48 | clever | then starts a blocking mutex, to wait for the irq |
| 20:43.53 | clever | and then just hangs |
| 21:03.37 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |
| 21:33.30 | *** join/#edev phoenixxx (~phoenixxx@91.210.101.152) |
| 22:28.19 | *** join/#edev NightMonkey (~NightrMon@pdpc/supporter/professional/nightmonkey) |