| 10:27 | RandomPixels | hello |
| 11:36 | jnettlet | RandomPixels, hey |
| 11:37 | RandomPixels | hello |
| 11:37 | jnettlet | dv_, is there a reason you don't have a rootwait or rootdelay= option in your kernel options? |
| 11:38 | jnettlet | I had to add it to get my openembedded image to boot with a 3.10 device-tree kernel |
| 14:32 | dv_ | jnettlet: just didnt need it |
| 14:32 | dv_ | I only ever used rabeeh's kernel |
| 14:33 | jnettlet | interesting. I will have to figure out what is different at some point. |
| 18:31 | neofob | hi all, i haven't been up-to-date with cubox lately |
| 18:31 | neofob | so what is latest kernel that can run cubox? 3.10? |
| 18:33 | cbxbiker61 | http://www.xilka.com/kernel/3/3.11/3.11.0/ |
| 18:33 | cbxbiker61 | http://www.xilka.com/kernel/UPDATE-KERNEL.sh |
| 18:40 | jnettle | 18:40 * jnettlet is working on a linaro 3.10 lts kernel with newest galcore/drm/vmeta goodness. |
| 18:41 | dv_ | yay |
| 18:41 | jnettlet | Oh and iwmmxt support on dv_'s openembedded layer |
| 18:49 | jnettlet | rabeeh, dv_, otavio, Any of you guys know what the deal is with Freescale's binaries for the Vivante GPU is? I was reading their EULA and it states that they can only be used on Freescale's devices. I don't know about SolidRun but the OLPC license is very different. They are licensed free to distribute and re-use unmodified. |
| 18:49 | jnettlet | Wondering if this is a Freescale thing, or a Vivante thing. |
| 18:50 | dv_ | nope. although it makes sense if they are customized for the imx6 |
| 18:51 | jnettlet | ^ /s/GPU is/GPU/g shouldn't talk on the phone and type at the same time |
| 18:51 | jnettlet | I can't imagine they are. The drivers are very abstracted from the platform |
| 18:52 | dv_ | hmm |
| 18:52 | dv_ | wumpus ? any idea? |
| 18:53 | bencoh | "Freescale's devices" ? |
| 18:53 | bencoh | you mean, SoCs ? |
| 18:54 | jnettlet | It should be devices that have Freescale SoCs. The wording is "The device for which they are intended" |
| 18:54 | jnettlet | So they don't want the iMX6 registered wayland drivers running on the XO-4 or Cubox |
| 18:55 | dv_ | what? no wayland drivers for cubox |
| 18:55 | dv_ | ? |
| 18:56 | jnettlet | well for the Cubox-i possibly, but as of right now not the original Cubox |
| 18:57 | jnettlet | I guess I need to push on the contacts I have made in Vivante to try to come up with something better. |
| 18:57 | dv_ | I think the drivers apply to the SoC |
| 18:57 | dv_ | not the device |
| 18:57 | dv_ | that wouldnt make sense |
| 18:57 | cbxbiker61 | amd arm64/radeon coming in second half of 2014 |
| 18:58 | dv_ | but I anyway believe the drivers cannot be used that easily with another SoC |
| 18:58 | jnettlet | yeah according to Apple everything with be 40x faster with arm 64-bit :-) |
| 18:59 | cbxbiker61 | i wonder if apple adhered to arm's 64bit spec, or used custom instructions |
| 18:59 | jnettlet | that has got to be the most abusive marketing towards non-technical people I have seen in a while. Just raping their lack of understanding to try and sell phones. |
| 19:00 | dv_ | I can then have a 64bit address space on an ARM box. finally, I can build firefox and chrome on an embedded device! :P |
| 19:00 | cbxbiker61 | well actually they're claiming 2x performance, but most of that is probably due to doubling register count |
| 19:01 | jnettlet | but all their 32-bit pointers are taking up twice the space in cache |
| 19:01 | dv_ | btw. we are tweaking chrome for the imx6 here. for linking chrome 31, ld uses up 7 GB RAM, and the resulting non-stripped binary is 2 GB big. after stripping, 60 MB. |
| 19:01 | jnettlet | dv_, yeah I build it here for my chromium OS on XO-4 testing. It is a beast to build |
| 19:02 | dv_ | it really is impossible to build chrome with a 32bit machine |
| 19:02 | cbxbiker61 | for my tastes the arm64/radeon combo sound good, since the open source radeon drivers are getting quite good |
| 19:03 | jnettlet | cbxbiker61, do they have ARM support yet? I know that some tegra support just landed in the nouveau stack |
| 19:04 | cbxbiker61 | jnettlet, i'd say no, but keeping in mind the hardware should be pretty near the same arm-vs-x86, it should be mostly a case of recompile on the target platform |
| 19:06 | cbxbiker61 | i'd think they'll just put the radeon on an internal pci-e bus |
| 19:32 | jnettlet | cbxbiker61, which SoC manufacturer is planning on supporting this/ |
| 19:32 | jnettlet | ? |
| 20:57 | cbxbiker61 | jnettlet, amd will produce the soc's |