| 05:23 | undo_ | yeah, running kernel with device tree |
| 05:24 | undo_ | the -a and -e flags are important, set to 0x80000 |
| 05:26 | undo_ | after trying different uboots, i didnt get a random ethaddr. until erased a few bits behind 0xd0000 |
| 05:45 | undo_ | now i can store the ethaddr in the environment 0xc0000-0xd0000 |
| 08:12 | undo_ | is the virtual terminal over hdmi only distorted on my dvi 1280x1024 display? |
| 08:13 | undo_ | with rabeehs kernel and forks |
| 08:26 | wumpus | seems to work fine here |
| 08:28 | wumpus | using rabeeh kernel commit 0d7ee41 |
| 08:29 | wumpus | my cubox is connected to an old 1280x1024 dvi monitor |
| 08:30 | wumpus | no obviously visible distortions, I use fbset 1280x1024-60 -vyres 1024 -depth 32 to set the video mode |
| 08:58 | undo_ | thanks, changed the active terminal (also from serial console). i believe busybox's fbset does not have the vyres option. maybe i can set this in early stage, before more vts get started. |
| 12:10 | RandomPixels | hello |
| 17:47 | _rmk | 17:47 * _rmk_ prods shesselba |
| 18:05 | _rmk_ | shesselba: I need you help to resolve the significant deltas between the si5351 driver in rabeeh's tree and the one you merged into mainline; otherwise I'm going to be forever stuck with v3.9 on the cubox |
| 18:06 | _rmk_ | it looks like the one in mainline is a regression over the one in rabeeh's kernel on a few things - missing .owner in the driver structure and the lack of remove function are two things I've spotted so far |
| 18:39 | dv_ | _rmk_: I am stuck with 3.6.9 on the cubox :) |
| 19:10 | _rmk | 19:10 * _rmk_ hates the apparant wilful rename of virtually everything in the si5351 driver |
| 19:13 | _rmk_ | like the rename of the interator 'i' to 'n' in si5351_i2c_probe - wtf was wrong with 'i' ? |
| 21:03 | dv_ | _rmk_: reminds me of some people who insisted on hungarian notation for everything, including local indexes |
| 21:03 | dv_ | the integer X became l_iX |
| 21:09 | _rmk_ | right, lets see if that merge worked... and whether I end up with a bootable 3.10 kernel |
| 21:11 | dv_ | is this kernel of yours available somewhere? do you consider it a working replacement for the 3.6.9 kernel ? |
| 21:11 | dv_ | because I could write openembedded recipes for it |
| 21:15 | _rmk_ | if it was, people wouldn't like it because I regularly redo all the merges in it (iow, it's not a stable tree) |
| 21:15 | _rmk | 21:15 * _rmk_ hmms, cubox uses a lot of old procfs stuff |
| 21:15 | _rmk_ | which is now gone |
| 21:38 | _rmk | 21:38 * _rmk_ now seems to have a v3.10 uImage to try |
| 21:42 | _rmk | 21:42 * _rmk_ checks the v3.10 stable tree; ISTR there's some important fixes in there |
| 21:48 | TrevorH1 | I'm running 3.10 ok though I don't use gfx nor the soundcard |
| 21:49 | _rmk_ | it looks to me like shesselba's DT configuration for the SI5351 is incorrect |
| 21:52 | _rmk_ | time to go bug hunting |
| 22:13 | _rmk_ | TrevorH1: do you boot with slab and lockdep enabled? |
| 22:15 | TrevorH1 | I just copied the .config from the debian 3.6.9 and did a make oldconfig then enabled a few other bits |
| 22:17 | _rmk_ | can you check on those two, because I'm seeing an oops with 3.10 |
| 22:20 | TrevorH1 | CONFIG_SLAB=y CONFIG_SLABINFO=y CONFIG_LOCKDEP_SUPPORT=y |
| 22:26 | _rmk_ | TrevorH1: its CONFIG_LOCKDEP=y |
| 22:27 | TrevorH1 | zgrep -i lockdep /proc/config.gz only gave me that one |
| 22:28 | TrevorH1 | that was the only one set in the 3.6.9 .config that I based it on too |
| 22:28 | _rmk_ | ok, so you don't have lockdep enabled |
| 22:28 | TrevorH1 | I'm running the debian wheezy image that was posted by one of the users |
| 22:31 | _rmk_ | found it, there's a patch in stable |
| 22:34 | _rmk_ | that worked |
| 23:26 | _rmk_ | bah. I need shesselba |