| 04:36 | cbxbiker61 | i have my carrier-one single core running geexbox quite good for right now, it's raw capabilities are way better than either the r-pi or the original cubox for xbmc. |
| 04:37 | cbxbiker61 | it's a huge advantage for the cubox-i to support EGL/framebuffer rather than EGL/XOrg on the original cubox |
| 04:38 | cbxbiker61 | i've also worked through all of the patches necessary to go from 3.0.35 to 3.0.101, after a bit more testing i'll release a 3.0.101. |
| 04:40 | cbxbiker61 | alsa back-ported f2fs into my 3.0 kernels, after having rolled out f2fs on all of my SSD/SD/USB-flash devices, it's really something i consider a must-have |
| 04:40 | cbxbiker61 | alsa/also |
| 04:51 | cbxbiker61 | hmmm, i can't decide whether i want to order 4 quad cubox-i's or 2 high-end dual cubox-i's and 2 quad cubox-i's |
| 04:53 | cbxbiker61 | it's pretty obvious from testing that even the single core cubox-i will handle most of what a person wants to do with xbmc |
| 04:54 | cbxbiker61 | the integrated wireless (if it works well) on the high-end models is desirable for my media playback boxes |
| 05:25 | jmontleo | _rmk_, thanks for the tips, I'll give it a try |
| 05:33 | Scotty2Shoes | So what caused the delays for the cuboxes? I know the demand has been high, but the email mentioned a material shortage |
| 08:34 | MikeSeth | hm |
| 09:01 | rabeeh | cbxbiker61: very good feedback |
| 09:02 | rabeeh | withregards f2fs; i totally see the potential. the main issue i don't have answer to is flashing raw images in Windows |
| 09:03 | cbxbiker61 | rabeeh, yeah prepping under windows is a problem |
| 09:03 | rabeeh | and then enlarging partition; can this be handled exactly as ext4 partition? (i.e. use parted or Windows equivalent to resize partition and then on first boot enlarge partition size on the fly?) |
| 09:03 | cbxbiker61 | no, as far as i know, we don't have a way to enlarge the f2fs partitions yet |
| 09:05 | cbxbiker61 | i'm running into an out-of-mem issue after a bit with GeeXBox/xbmc, small memory on the base model, i'll have to do a swap and see if that stabilizes the problem |
| 09:06 | hste | rabeeh: I have made a lxde version of jas-hacks_ jessie rootfs http://stende.no-ip.info/imx6/jessie-lxde.tar.xz |
| 09:12 | rabeeh | hste: lxde looks faster btw |
| 09:12 | rabeeh | hste: do you have an rootfs build script for that too? |
| 09:13 | rabeeh | cbxbiker61: sounds not ready then. |
| 09:13 | rabeeh | cbxbiker61: the main issue today is that we have lots of Windows / MAC users that because of other boards they are used to flashing a raw image and then resizing a partition |
| 09:14 | cbxbiker61 | yeah, it needs a bit of work, surprising to me that geexbox didn't support nfs4 as shipped either, which is the way I roll |
| 09:14 | jnettlet | cbxbiker61, running out of memory running xbmc on fbdev? |
| 09:14 | hste | rabeeh: no. I just used jas_hacks_ rootfs and installed lxde setting it to default |
| 09:14 | cbxbiker61 | yep, could be a mem leak in xbmc, at this point i'm not sure |
| 09:15 | MikeSeth | turns out I can write C |
| 09:15 | jnettlet | interesting. |
| 09:15 | rabeeh | MikeSeth: C that works? :) |
| 09:15 | jnettlet | I have only run xbmc under X |
| 09:16 | rabeeh | cbxbiker61: no memory leak for me |
| 09:16 | rabeeh | (ext4 partitions) |
| 09:16 | cbxbiker61 | yeah, i'll look at it more closely tomorrow, right at the moment i'm deep into season 3 of 24 |
| 09:16 | rabeeh | but there is an issue of allocating VPU buffers after a long uptime |
| 09:17 | cbxbiker61 | the fps in the gui is 60 or greater most of the time, which is way better than the original cubox |
| 09:17 | MikeSeth | rabeeh: that remains to be seen |
| 09:17 | jnettlet | ah, so that will be fixed with switching to the CMA backed bmm |
| 09:18 | jnettlet | cbxbiker61, are you running at 1080p? |
| 09:19 | jnettlet | with dirty regions enabled? |
| 09:19 | cbxbiker61 | pretty sure i am, i didn't tweek the default for geexbox |
| 09:20 | cbxbiker61 | uses about 20% cpu when things settle down |
| 09:21 | jnettlet | I think the default geexbox setup is for 720p |
| 09:21 | hste | rabeeh: whats the naming of the devices? hb-1, I1 , I2, I2ultra and I4pro ? |
| 09:22 | rabeeh | cbxbiker61: fps on i4pro is 100 to 130fps |
| 09:22 | rabeeh | geexbox default is 1080p |
| 09:23 | rabeeh | geexbox uses the framebuffer directly (i.e. on X) |
| 09:23 | rabeeh | so whatever you boot with remains |
| 09:23 | rabeeh | hste: hb-i1, hb-i2 etc... |
| 09:24 | cbxbiker61 | yeah, i think geexbox runs lower res on some other platforms, it's configured in the build |
| 09:25 | cbxbiker61 | rabeeh, if i order tomorrow, am i looking at end of january for the i4pro? |
| 09:25 | hste | rabeeh: is there some default uEnv.txt on the wiki? |
| 09:26 | rabeeh | cbxbiker61: i don't know what are the current numbers. |
| 09:29 | jnettlet | rabeeh, you can use fbset to change resolutions. I don't know if xbmc settings can control that. |
| 09:31 | cbxbiker61 | another thing that i think you can do with fbset, is to set the bits at a particular resolution, i haven't done it yet, but i think i need to on one of my older 1080p tv's, as the darks are too dark |
| 09:32 | cbxbiker61 | it did look good on my newer TV though |
| 09:37 | rabeeh | cbxbiker61: which geexbox did you try? |
| 09:37 | rabeeh | i mean did you build? or downloaded from somewhere? |
| 09:39 | cbxbiker61 | for now i'm running 20131231 snapshot, i expect to do my own builds in the next few days or so, so i can get into the code |
| 09:40 | cbxbiker61 | one of the issues with the current imx xbmc is that it's apparently ignoring the pts/dts, i had previously patched the xbmc for the original cubox, so i just need to work those patches in |
| 10:45 | jnettlet | wanted 16368 bytes, got 16368 |
| 10:45 | jnettlet | random bytes@: 8c310000: 684ccad6 3769ea6c 2c137103 a6e2f93a ..Lhl.i7.q.,:... |
| 10:45 | jnettlet | random bytes@: 8c310010: cc446beb 2e390a01 e0cb63cb 633d6199 .kD...9..c...a=c |
| 10:45 | jnettlet | random bytes@: 8c310020: c0663fcc d4456da6 c7c0f1d5 ed2384b9 .?f..mE.......#. |
| 10:45 | jnettlet | crypto is working in the 3.10.17 kernel now |
| 10:50 | hste | jnettlet: pushed to some git? |
| 10:54 | hste | jnettlet: still 26 days till u should have the kernelwork finished :) 15:37 * jnettlet is making is new years resolution to get all his 3.10 lts kernel work out by the end of Jan. |
| 11:54 | jnettlet | hste, yep working hard at it. |
| 17:06 | MikeSeth | rabeeh: my tracking number still doesn't show in post office tracker after 3 days, is that normal? |
| 17:23 | _rmk_ | well, v3.13-rc7 is broken |
| 17:36 | hste | _rmk_: for hummingboard? |
| 17:37 | _rmk_ | yep |
| 17:38 | _rmk_ | fec 2188000.ethernet eth0: no PHY, assuming direct connection to switch |
| 17:38 | _rmk_ | libphy: PHY fixed-0:00 not found |
| 17:38 | _rmk_ | fec 2188000.ethernet eth0: could not attach to PHY |
| 17:38 | _rmk_ | Kernel BUG at c0437b70 [verbose debug info unavailable] |
| 17:38 | _rmk_ | Internal error: Oops - BUG: 0 [#1] SMP ARM |
| 17:38 | _rmk_ | PC is at fec_enet_open+0x4d0/0x500 |
| 18:00 | _rmk_ | rmk@rmk-PC:[linux-rmk]:<1595> git log -p v3.13-rc6..v3.13-rc7 drivers/net/ethernet/freescale/fec_main.c |
| 18:00 | _rmk_ | bah |
| 18:01 | _rmk_ | 7a399e3a2e05bc580a78ea72371b3896827f72e1 is the culpret |
| 18:09 | _rmk_ | yep, reverting that fixes it |
| 18:31 | jnettlet | _rmk_, what commit was that? |
| 18:34 | _rmk_ | a very broken one |
| 18:36 | jnettlet | have you publicly shamed whomever committed it yet? |
| 18:36 | jnettle | 18:36 * jnettlet is wondering if he should make popcorn |
| 18:36 | _rmk_ | yes, as has ojn now too :) |
| 18:37 | jnettlet | and I missed it :-\ |
| 19:14 | jmontleo | even with APPENDED_DTB all I get is: Starting kernel ... |
| 19:14 | jmontleo | Uncompressing Linux... done, booting the kernel. |
| 19:15 | jmontleo | something about console settings maybe? |
| 19:15 | _rmk_ | you need APPENDED_DTB _and_ you need to combine the zImage with the appropriate dtb image before turning it into a uImage |
| 19:16 | _rmk_ | cat $builddir/arch/arm/boot/zImage $builddir/arch/arm/boot/dts/imx6dl-hummingboard.dtb > zImage.appended |
| 19:16 | _rmk_ | mkimage -A arm -O linux -C none -a 0x10008000 -n "Linux $v" -d zImage.appended my-uImage |
| 19:17 | jmontleo | I did that, save it was imx6dl-cubox-i-carrier-1.dtb it looks like it is called imx6dl-cubox-i-carrier-1.dtb in the 3.12-4c4 kernel |
| 19:18 | _rmk_ | yep, that's the old name for the dtb blob |
| 19:18 | _rmk_ | if it helps, I boot with this: "console=ttymxc0,115200n8 console=tty1 video=mxcfb0:dev=hdmi,1920x1080M@60,if=RGB24,bpp=32 root=/dev/nfs rw cma=64M" |
| 19:19 | jnettlet | _rmk_, you shouldn't need that video entry for KMS |
| 19:23 | jnettlet | oh great my uhs1 card is really causing the 3.10 kernel to throw fits |
| 19:24 | jnettle | 19:24 * jnettlet keeps lighting up a new green light on the hb1 when it crashes, still wondering if that is intentional or accidental |
| 19:25 | jmontleo | aha! _rmk_ that got me output :) Now, I'm seeing mmcblk0: error -110 transferring data, sector 0, nr 8, cmd response 0x900, card status 0xb00, which I guess might be my SD Card... |
| 19:30 | _rmk_ | hmm, not sure what's causing that. it says it timed out, but 0x900 would be right for the command response, but a status of 0xb00 shows that the card was in data mode. |
| 19:36 | jmontleo | looks like it spit out 0xb00 once and then card status 0x0 repeatedly |
| 19:36 | jmontle | 19:36 * jmontleo has a second card he can try |
| 19:43 | jnettlet | anyone know how to shut systemd the hell up on geexbox? |
| 19:43 | _rmk_ | avoid systemd? :) |
| 19:44 | tomlohave | why ? |
| 19:44 | tomlohave | what's the problem with systemd ? |
| 19:47 | jnettle | 19:47 * jnettlet is not going to get started on the systemd debacle |
| 19:47 | _rmk_ | me neither - the issues about why some people hate it are well known |
| 19:48 | jnettlet | jmontleo, is your card a uhs1 card. |
| 19:48 | jmontleo | yes |
| 19:48 | jnettlet | that seems to be the problem. |
| 19:49 | jnettlet | I am having similary problems, I just switched over from my class 8 to uhs1 |
| 19:49 | jnettlet | jmontleo, which kernel are you running? |
| 19:50 | jmontleo | 3.12-rc4 ; from git://ftp.arm.linux.org.uk/~rmk/linux-cubox.git - the latest I was able to find. |
| 19:51 | jmontleo | the card I am using is "SanDisk Ultra 32 GB microSDHC Class 10 UHS-1"; unfortunately I don't have a non uhs1 card to try on, but I'm sure I can get one |
| 19:51 | jmontleo | same on the second card (same model) |
| 19:52 | jnettlet | well i just used device-tree to disable the 1.8v probing and am still seeing problems so I will need to dig a bit deeper |
| 19:54 | jnettlet | worst case we can lean on cjb a bit. |
| 19:55 | jmontleo | no worries. I can dig up a class 8 card or something for now; i feel considerably less nuts now that I can see console output :) |
| 19:57 | jnettle | 19:57 * jnettlet has learned sometimes seeing console output just proves you are nuts :-) |
| 20:00 | jmontleo | fair enough :) |
| 21:05 | jmontleo | jnettlet, found a class 4 card and get the same, interestingly |
| 21:05 | jmontleo | can rebuild with the -rc7 stuff and see if it makes a difference |