IRC log of #cubox of Thu 09 Jan 2014. All times are in CET < Back to index

00:06 jmontleo jnettlet, is USB MXS USB PHY required for USB to work?
05:59 _dab_ I have a new i4pro and a new build machine, what to test?
06:03 jmontleo hmm, did not make a difference; see these in dmesg: [ 0.132071] usb_h1_vbus: 5000 mV and [ 0.132286] usb_otg_vbus: 5000 mV
06:03 jmontleo other than that no sign of usb
06:46 jnettlet jmontleo, as soon as I finish my 3.10 lts kernel I will test HEAD
07:20 jnettlet Hey guys, bunny from the Chumby project is using the IMX6 as well on his OSS laptop project. http://makezine.com/magazine/building-an-open-source-laptop/
07:25 _dab_ booted up my i4pro into an old (Dec 12) xbmc, I love it
07:26 _dab_ jnettlet: Do you have kernel or uboot you want me to test?
07:27 jnettlet _dab_, not yet. Still finalizing bits and pieces
07:30 _dab_ i4 is quicker than I expected and is barely warm running xbmc
07:40 _dab_ rabeeh: I have the i4pro, anything you want me to try?
07:40 jnettlet _dab_, just wait. we are working on making it quicker
07:41 _dab_ dab quivers in anticipation
09:21 MikeSeth _dab_: do yo have jas' jessie debian image?
09:23 _dab_ no
09:23 _dab_ I will try it
10:45 MikeSeth _dab_: if you do, take note of the bogomips counter, then set gpumem=64M in /uEnv.txt, reboot and take note again
10:45 MikeSeth just something weird
10:45 _dab_ thanks
10:46 MikeSeth oh have you tried the android 4.3 image?
10:46 MikeSeth i wonder if it works fine on i4pro, on my i1 it majorly misbehaved
13:05 xraxor cant wait to use xbmc with all the work you guys are doing.
13:05 xraxor atm i find the UI not as fast using CEC remote
13:07 xraxor CEC does not speed up scrolling when you have a long list of items, can that be changed? the Pi had the same problem really slow to scroll, Yatse remote for android does does this faster and why cant CEC do so as well
13:35 MarcusVinter Jon, how may I go aobut fixing bluetooth?
13:43 MarcusVinter lsmod lists nothing
14:34 jnettlet MarcusVinter, post your output to dmesg and I will take a look when I get some free time.
14:49 MarcusVinter Thank you. Ill try grab some errors through dmesg now.
17:04 jmontleo jnettlet, thanks. I tried a few more tweaks to the usb configuration and rebuilt the kernel, but still nothing. It's not killing me though; just thought it would be a nice to have.
17:04 jnettlet jmontleo, unfortunately I am still hacking at MMC. It is on my list
17:31 hste jnettlett: how is it going with mmc?
17:46 jnettlet hste, think I found a bug in the 3.0.35 kernel that is forcing the SDHC slot down to 1-bit transfers
17:47 hste jnettlett: is the same bug in 3.10?
17:48 jnettlet hste, no that is a different problem.
17:49 hste jnettlett: nice if u can fix it :)
18:02 jnettlet hey that is better. 16.4MB/s read speed
18:03 hste what did u have before?
18:04 jnettlet 5.x I believe
18:05 jnettle 18:05 * jnettlet has to retest. I have been testing too much today
18:06 hste is it a big patch?
18:06 jnettlet nah a couple of lines.
18:07 jnettlet let me double check it is even necessary
18:39 jnettlet could somebody with nothing better to do run this command 3-4 times in their shell on a Cubox-i, preferably a HB1/C1
18:40 hste what command?
18:41 MarcusVinter I could run it on an i2U
18:47 tomlohave-away jnettlet: on i4pro if you want
18:47 tomlohave-away or c1
18:47 Coolgeek c1 ? is it the name of the first cubox ?
18:48 tomlohave-away HB1
18:48 Coolgee 18:48 * Coolgeek is lost with all those name
18:48 tomlohave-away hb1=c1
18:49 jnettlet tomlohave-away, either is fine. Just as long as they are running the 3.0.35 kernel
18:49 tomlohave-away command line ?
18:50 tomlohave-away please
18:52 jnettlet ha it didn't paste okay
18:52 jnettlet dd if=/dev/mmcblk0 of=/dev/null bs=1M count=1000
18:52 jnettlet the speed doesn't matter, more curious if the kernel panis
18:52 jnettlet panics
18:53 hste c1: 104857600 bytes (105 MB) copied, 6.7185 s, 15.6 MB/s
18:53 jnettlet it is usually the 3-4th run that blows up right now
18:54 hste but freeze io for a time
18:54 tomlohave-away 000+0 records in
18:54 tomlohave-away 1000+0 records out
18:54 tomlohave-away 1048576000 bytes (1000.0MB) copied, 62.923957 seconds, 15.9MB/s
18:54 hste running it many times : 104857600 bytes (105 MB) copied, 0.294293 s, 356 MB/s
18:55 tomlohave-away 1000+0 records in
18:55 tomlohave-away 1000+0 records out
18:55 tomlohave-away 1048576000 bytes (1000.0MB) copied, 2.841233 seconds, 352.0MB/s
18:56 jnettlet 356MB/s on your sdhc card?
18:56 tomlohave-away the same for me
18:56 hste no panic on my c1
18:56 tomlohave-away no panic on i4 pro
18:56 jnettlet okay
18:56 jnettlet hmmm interesting
18:56 hste yes looks like its buffered after first run
18:57 hste ls
19:00 hste same results on my gk802 with 3.0.35_4.1.0
19:03 Coolgeek it's not buffered on my cubox (the first one) with kernel 3.6.9 (geexbox)
19:03 Coolgeek but same time and no panic
19:03 jnettlet okay sounds good thanks everyone
19:04 jnettlet wow, I just ran graphics under the default kernel for the firs time in a while. the 2d redraws are really laggy
19:05 jnettlet Who this helps depends on the card you are using, but regardless. http://fpaste.org/67122/90700138/
19:06 jnettlet basically if you don't specify some sort of CAP mmc in the older kernel just chooses 1_BIT capabilities
19:23 hste c1: without patch 104857600 bytes (105 MB) copied, 6.7185 s, 15.6 MB/s. with patch: 1048576000 bytes (1.0 GB) copied, 67.6862 s, 15.5 MB/s
19:25 jnettlet hste, well it will be card dependent as well.
19:25 hste jnettlet: mmc0: new high speed SDHC card at address aaaa
19:29 jnettlet that is good it means 4-bit more is enabled
19:31 hste On old kernel on my gk802 mmc0: new high speed SDHC card at address 59b4
19:33 jnettlet hste, yeah that aaaa pops up alot. something it reads off the card I think...not sure
23:10 MikeSeth the mini-usb on i1 is serial console, correct?