| 00:01.29 | lundman | when playing videos on mele, I get no sound, or, baaad stuttering |
| 00:10.26 | lundman | [ 349.130000] nand try to shutdown 0 time |
| 00:10.30 | lundman | I have a time machine! |
| 00:13.25 | *** join/#arm-netbook revident (~scott@fw.ss.org) |
| 00:16.08 | *** join/#arm-netbook DonkeyHotei (~danielg4@50-0-149-2.dsl.dynamic.sonic.net) |
| 00:21.20 | *** join/#arm-netbook tallnerd1985 (~tallnerd1@44.sub-166-250-3.myvzw.com) |
| 00:26.12 | Triffid_Hunter | hey all, the drive slot on the top of the mele A1000, is that for a 2.5" or a 3.5" drive? |
| 00:28.47 | lundman | damn, even with official ics4 kernel, it plays no sound |
| 00:28.53 | lundman | 2.5" |
| 00:29.11 | DonkeyHotei | lundman: really? |
| 00:29.18 | DonkeyHotei | i was quite hoping for 3.5" |
| 00:30.10 | hipboi | Mazon: yes, i think the uhost is Oval Elephant |
| 00:30.45 | hipboi | the reason it's expensive is because they made very few |
| 00:31.05 | hipboi | at this time, only 600 devices are made |
| 00:31.46 | Triffid_Hunter | lundman: thanks |
| 00:31.53 | revident | Triffid_Hunter, DonkeyHotei, the slot is a USM-SATA slot, which in plain terms is Seagates Goflex drive format spun out as an offical SATA form factor standard. |
| 00:32.00 | lundman | i sure have no problems fitting the 2.5ssd into it |
| 00:32.15 | Triffid_Hunter | revident: so potential problems fitting random 2.5" drives on it? |
| 00:32.18 | hipboi | Oval Elephant is pre-order now, when it hit 1k or 2k, the price goes down |
| 00:32.54 | DonkeyHotei | what's an oval elephant? |
| 00:32.57 | revident | The A1000 comes with an enclosure, just be mindful that your drive isn't one that wants more then 500mA |
| 00:32.59 | lundman | i dont think you'll have issues fitting any standard 2.5 drive, but, without a case, it just "sits there" and bumping it would not be great |
| 00:33.25 | Triffid_Hunter | revident: noted, thanks. I'm handy with an iron and a switching regulator so I can upgrade if necessary |
| 00:33.50 | DonkeyHotei | lundman: you mean for the A1000, or the A2000? |
| 00:36.10 | revident | DonkeyHotei, your going to need an enclourse of some kind that is USM-SATA spec. it's really just adds a bit of bulk to the drive and the mele's slot (A1000 and A2000) is expecting an encloused drive. So yeah, it will wobble on the connector with a bare drive. |
| 00:36.44 | DonkeyHotei | how does the A1000's enclosure apply to that? |
| 00:37.27 | lundman | I have A2000 and bare ssd sitting in it. but then, I'm just doing dev |
| 00:37.29 | revident | Triffid_Hunter, the main board has a standard sata connecter on the board, the top case has a back-place board. So yeah it will be easy to use a 3.5" drive if your creative. http://rhombus-tech.net/allwinner_a10/hacking_the_mele_a1000/A10_device_a1000_inside.jpg |
| 00:37.54 | lundman | so, can others get HDMI sound with Mele/ICS4 ? |
| 00:38.14 | DonkeyHotei | does sound work on GB? |
| 00:39.27 | revident | DonkeyHotei, the A1000's enclousure is a USM enclosure that you can put your own drive in. They simply made it easy. The USM spec really is just a casing formfactor, Seagate's Goflex dives simply exposed the drives SATA connector raw and use interchangeable SATA to (USB, USB3, FIREWIRE, etc) back plains. |
| 00:39.45 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 00:40.26 | lundman | great britain? |
| 00:40.36 | DonkeyHotei | gingerbread |
| 00:40.45 | lundman | downgrade? oh man |
| 00:41.41 | DonkeyHotei | revident: so, does the A1000's enclosure fit over the goflex enclosure, or over a bare drive? |
| 00:42.09 | revident | DonkeyHotei, bare drive |
| 00:42.38 | DonkeyHotei | and in contrast, the A2000 needs you to provide a goflex enclosure of your own? |
| 00:43.30 | DonkeyHotei | and no self-provided enclosure is expected for the A1000? |
| 00:44.14 | revident | The one thing to note is, the A1000 comes with the enclosure (which is a convenience), the A2000 does not. So with the A2000 you either have to get a Goflex or other USM form factor enclosure / pre-enclosed drive. |
| 00:44.19 | ssvb | DonkeyHotei: I guess it's this one - http://www.aliexpress.com/fm-store/511685/0-578825654/Mele-2-5-hard-drive-case.html |
| 00:45.20 | revident | ssvb, correct. That is a USM enclousure, sans drive, plus USB3 backplane. |
| 00:46.57 | DonkeyHotei | i have a question about this: http://www.aliexpress.com/fm-store/511685/0-566678042/Mele-A1000-TV-box-debug-board-pack.html |
| 00:47.52 | DonkeyHotei | it says "It's the same product as Mele website, I get the boxes from their factory, the package includes |
| 00:47.53 | DonkeyHotei | the tv box, remote, cvbs cable." but lower it says "Maybe you also need the all-in-one air mouse + wireless mouse and keyboard + remote." |
| 00:48.07 | ssvb | revident: so it is $70 for A1000 vs. $75 + $13 for A2000 |
| 00:48.23 | DonkeyHotei | so does it come with a remote different from the F10 remote? |
| 00:48.24 | ssvb | revident: A1000 was too good, so it had to be downgraded :) |
| 00:48.49 | revident | DonkeyHotei, yes, it comes with a simple IR remote. |
| 00:49.08 | DonkeyHotei | i wonder how different they are |
| 00:49.17 | revident | ssvb, yes, if you plan to use an HDD the A1000 is the better value. |
| 00:49.48 | DonkeyHotei | i'm told that android with ANY remote is much less convenient than with a pointer and keyboard |
| 00:49.55 | revident | DonkeyHotei, hugely, the F10 remote has a keyboard and use RF signaling and I think also has a click wheel, but my memory might be faulty there. |
| 00:49.56 | Turl | [offtopic] does any of you know about continuous ink supply systems? |
| 00:50.03 | ssvb | revident: too late, got A2000 (instead of the already out of stock A1000) |
| 00:50.42 | DonkeyHotei | Turl: they're only useful if you print a ton of pages every day without fail |
| 00:51.09 | revident | ssvb, yeah you make do with what you got, in my case with the GoflexNet (arm based NAS) I made do with a laser cutter. http://openrebel.ss.org/images/2011-07-29-GoFlexNet-HDD-stablizers.jpg |
| 00:53.11 | Turl | DonkeyHotei: say I print 50 pages a month in bursts, any use? |
| 00:53.25 | hno | Turl, watchdog seems to work fine from uboot. |
| 00:53.31 | DonkeyHotei | Turl: probably not |
| 00:53.47 | Turl | I might start printing more stuff @ home if it becomes cheaper though |
| 00:53.52 | revident | ssvb, If I was going to buy a second Mele, it would be the A2000, I like the look of the smaller casing and I could design something to keep a drive from wobbling like I did in that photo. |
| 00:53.55 | CIA-121 | rhombus-tech: hno master * ra34dc9128f11 /allwinner_a10/A10_register_guide/A10_TIMER.mdwn: |
| 00:54.04 | Turl | DonkeyHotei: the thing is the damn cartidges cost like 30US$ each |
| 00:54.05 | DonkeyHotei | "bursts" is just not amenable to continuous ink |
| 00:54.07 | Turl | and you need 2 |
| 00:54.20 | Turl | hno: interesting |
| 00:54.31 | DonkeyHotei | do you have canon, epson, or other? |
| 00:54.36 | Turl | DonkeyHotei: HP |
| 00:54.54 | Turl | this one, to be more precise DonkeyHotei http://www.viper550.be/DSCF1344.JPG |
| 00:54.55 | DonkeyHotei | HP you can refill for a while |
| 00:55.17 | Turl | DonkeyHotei: I was considering it |
| 00:55.26 | Turl | but all I hear about refills are horror stories |
| 00:55.32 | DonkeyHotei | the only time you really need to replace HP cartridges is when they clog |
| 00:55.37 | Turl | about how they kill the printer or print crappily |
| 00:56.11 | DonkeyHotei | you can get pre-refilled ones, because refilling yourself makes a godawful mess |
| 00:56.12 | hno | Turl, it's not a 100% perfect watchdog as there is no way to reset the counter without temporarily disabling the watchdog, but better than nothing. |
| 00:56.27 | hipboi | DonkeyHotei: you want to buy mele a1000/2000 from my shop? |
| 00:56.38 | hno | hi hipboi |
| 00:56.45 | Turl | hey hipboi :) |
| 00:56.46 | hipboi | hno: hi |
| 00:57.01 | Turl | hno: but does it work to some extent? |
| 00:57.02 | DonkeyHotei | killing the printer is not something a cartridge can do on HP. canon maybe, but HP is designed differently |
| 00:57.18 | Turl | DonkeyHotei: great |
| 00:57.34 | DonkeyHotei | hipboi: strongly considering it, but also weighing against the MK802 |
| 00:57.37 | Turl | DonkeyHotei: how hard is it to refill them? |
| 00:57.46 | hno | Turl, yes, it works fine. Only that there is a slight race window when resetting the watchdog where it's temporarily disabled. If the system crashes in that instant then you are smoked. |
| 00:57.48 | Turl | hipboi: did you get the A100s? |
| 00:58.01 | hipboi | Turl: not yet |
| 00:58.12 | Turl | hno: you're talking about uboot still right? |
| 00:58.24 | hipboi | Turl: they are at my friend's place |
| 00:58.33 | DonkeyHotei | Turl: i've tried only with epson, not hp, and it's not too hard, but makes a big mess, and i heard HP inks are even messier |
| 00:58.35 | Turl | hipboi: do you know if they have sata? |
| 00:58.36 | hno | Turl, kernel is no different. But can not reach it's register from userspace. |
| 00:58.45 | hipboi | Turl: not yet |
| 00:58.58 | revident | hipboi, is there any marketing material or photos of this A100 yet? I'm curious about it since you mentioned it's existence. |
| 00:59.04 | lundman | well, this is a bit of a problem.. any suggestions what i can look at for the audio problem? unless its a hardware problem, should try spdif |
| 00:59.30 | Turl | hipboi: let me know when you find out |
| 00:59.49 | Turl | hipboi: I want to buy a mele but with sata |
| 00:59.51 | hipboi | revident: here -> http://detail.tmall.com/item.htm?id=16964088540&wwdialog=bbxxbbmc&ad_id=&cm_id=&pm_id=&am_id=&asker=wangwang |
| 01:00.05 | hipboi | Trul: yes, from the picture, it has sata |
| 01:00.17 | revident | hipboi, thanks |
| 01:00.18 | hno | Turl, A1000/A2000 have sata. |
| 01:00.48 | Turl | hipboi: the picture say "A2000" though |
| 01:00.51 | Turl | not A100 |
| 01:01.13 | lundman | I should have go the real remote too, the standard one is a pain to try to use |
| 01:01.35 | Turl | hno: if the A100 have no sata I'll get a 1000/2000 |
| 01:01.58 | hno | what's the benefit of A100? |
| 01:02.03 | Turl | hno: cheaper |
| 01:02.06 | DonkeyHotei | other than the sata, what are the major functional differences between the mele and the mk802? |
| 01:02.22 | Turl | DonkeyHotei: mk802 has hdmi only |
| 01:02.35 | Turl | mele has vga and even composite I think |
| 01:02.38 | hno | yes |
| 01:02.39 | hipboi | DonkeyHotei: the main difference is the quality |
| 01:02.47 | DonkeyHotei | ok. anything else? |
| 01:02.52 | DonkeyHotei | hipboi: oh? |
| 01:02.57 | Turl | DonkeyHotei: ethernet |
| 01:02.57 | hipboi | personally i think the mk802 is not well made |
| 01:03.03 | DonkeyHotei | ah |
| 01:03.11 | hipboi | maybe the same with uHost |
| 01:03.13 | Turl | and SPDIF too right? |
| 01:03.19 | hno | mele have everything you need, except maybe memory. |
| 01:03.35 | DonkeyHotei | are there major differences in the factory android firmware too? |
| 01:03.35 | hno | even including ESD protection on everything that matters. |
| 01:03.48 | hipboi | the mele device is good quality comparing to the kirf devices |
| 01:04.18 | DonkeyHotei | hno: i thought the mk802 had the same amount of ram as the mele |
| 01:04.23 | hipboi | because it has a brand |
| 01:04.35 | hno | MK802 exists in 1GB version. Mele only 512MB. |
| 01:04.45 | DonkeyHotei | ok |
| 01:04.45 | Turl | DonkeyHotei: mk802 comes on 2 versions, 512 and 1G |
| 01:05.24 | Turl | hipboi: the Mele brand is not too known outside of china I suppose |
| 01:05.48 | hipboi | mele is not too known in China also |
| 01:05.57 | *** join/#arm-netbook gsilvis (~almostsix@50.12.163.241) |
| 01:06.09 | hipboi | but they keep mading a good brand |
| 01:06.29 | Turl | their products look like good quality on the pictures indeed :) |
| 01:06.44 | lundman | mele is a great name, well, in english :) |
| 01:06.56 | hipboi | well, when i go to Sangda electronic market, i saw the mk802 was be packaged |
| 01:07.07 | DonkeyHotei | is it pronounced may-lay? |
| 01:07.37 | Turl | Mele reminds me of https://en.wikipedia.org/wiki/Super_Smash_Bros._Melee |
| 01:07.49 | lundman | 乱戦! |
| 01:08.00 | lundman | well, thats japanese though |
| 01:08.46 | DonkeyHotei | mele is pronounced ransen? |
| 01:08.52 | lundman | guess there is nothing for me to do now, I'll switch back to cubox dev |
| 01:09.03 | lundman | yep |
| 01:10.17 | Turl | lundman: ooh, cubox has CEC |
| 01:10.20 | Turl | interesting :) |
| 01:10.32 | lundman | not really, I htink its broken :) |
| 01:10.35 | DonkeyHotei | i'll never know how they keep track of all those on/kun readings |
| 01:10.58 | lundman | how is xbmc looking for mele? anything I can test yet? |
| 01:10.59 | Turl | lundman: heh |
| 01:11.25 | DonkeyHotei | lundman: probably not until the x11 drivers get freed |
| 01:11.52 | DonkeyHotei | at least that's my guess |
| 01:12.30 | lundman | I thought they were doing xbmc with FB, not x11 |
| 01:12.35 | lundman | I'd rather not have to run X11 |
| 01:13.04 | Turl | yeah, no X required I heard |
| 01:13.37 | DonkeyHotei | i highly doubt xbmc could ever run without x11 |
| 01:13.41 | Turl | I heard the UI was working and video decoding too, but sluggish still or something |
| 01:14.17 | lundman | someone dcc me xbmc alpha to test :) |
| 01:14.30 | DonkeyHotei | but if xbmc can run on the mele, it can run on the mk802 also |
| 01:15.09 | Turl | DonkeyHotei: https://github.com/xbmc/xbmc/pull/454 |
| 01:15.22 | Turl | lundman: nothing released I think still |
| 01:16.09 | lundman | yes, but it's *me*, I can help :) |
| 01:23.28 | DonkeyHotei | Turl: "Letting this die. While it is interesting as a concept, this would bring in no additional hardware support at the moment. Much more interesting would be skipping the dfb layer and using mesa+kms directly. See here:" |
| 01:23.54 | DonkeyHotei | at the bottom |
| 01:24.16 | Turl | yeah, but I heard from the people here they weren't using X, so I suppose they revived that or something |
| 01:26.20 | DonkeyHotei | hipboi: a higher-priority purchase than a mele, though, would be some A10 tablet similar to the ZaTab but with built-in gps and bluetooth, but looking all over aliexpress finds such animals to be quite rare |
| 01:27.17 | hipboi | yes, bluetooth on a10 is very rare |
| 01:28.28 | DonkeyHotei | gps seems far more rare, looking at aliexpress |
| 01:28.41 | DonkeyHotei | quite a few have bluetooth but no gps |
| 01:29.55 | hipboi | well, a10 have built-in gps engin |
| 01:30.15 | hipboi | i don't know why they do not make a usage of it |
| 01:30.24 | DonkeyHotei | right |
| 01:30.37 | Turl | hipboi: what about the antenna? |
| 01:30.37 | DonkeyHotei | still needs a receiver, but few include it |
| 01:31.12 | hipboi | yes, addtional rf module is needed |
| 01:33.03 | DonkeyHotei | if you run across a 1GB-ram A10 tablet that's at least 8" with a 8000mAh-ish battery and both gps and bluetooth, PLEASE let me know |
| 01:35.53 | lundman | Status |
| 01:35.53 | lundman | Currently there is no X11 acceleration of mali on A10, this page describe attempting get the mali x11 driver working. At the time of writing the mali x driver is not working. This is a reference for who will work on it later. |
| 01:36.04 | lundman | Is that true? since it goes into detail on how to get it to work |
| 01:40.36 | hipboi | lundman: how to try to get it work |
| 01:40.56 | lundman | ah so it almost works, but not quite? |
| 01:41.19 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 01:41.25 | hipboi | but the glmark-es shows that the 3d acceleration does not work |
| 01:43.10 | lundman | Not really a field I am into, I want media player :) |
| 01:43.14 | *** join/#arm-netbook tallnerd1985 (~tallnerd1@c-67-168-239-239.hsd1.wa.comcast.net) |
| 01:53.49 | Turl | lundman: mpd? |
| 01:54.06 | Turl | + groovebasin :) https://github.com/superjoe30/groovebasin |
| 01:54.53 | lundman | mpd is terrible |
| 01:56.12 | Turl | lundman: why? |
| 01:56.17 | Turl | it's awesome imo |
| 01:56.25 | Turl | as long as you have a vfp :) |
| 01:57.50 | lundman | its mostly for music anyway, and terribly implemented protocol, when there are already funny designed and impletemented ones |
| 01:58.09 | lundman | but most people dont care about that, as long as it "looks pretty" :) |
| 02:00.25 | lundman | hmm others are complaining about no HDMI sound |
| 02:01.44 | Turl | lundman: name another mpd like protocol? |
| 02:01.54 | Turl | and don't tell me DLNA :) |
| 02:02.43 | lundman | I'm not a fan of DLNA no, over engineered :) |
| 02:03.01 | lundman | but UPNP does work, and can do more than mpd |
| 02:03.44 | Turl | isn't UPNP and DLNA the same thing? |
| 02:04.08 | lundman | no, DLNA is UPNP with someone puking over it |
| 02:04.16 | lundman | wait, there is an audio patch for ICS4? |
| 02:04.18 | Turl | UPNP uses http stuff, it's a mess on itself |
| 02:04.40 | Turl | I tried it once |
| 02:04.45 | Turl | no player played nice with it :P |
| 02:04.48 | lundman | UPNP uses SOAP to get directory listings. After that, it doesn't have to use URL for media, but generally it is easier to |
| 02:05.16 | lundman | I'm not here to try to change your mind, if you are happy about mpd, stick with it |
| 02:05.40 | Turl | lundman: UPNP fails if you can't autodiscover the other end on most apps |
| 02:05.54 | Turl | they just say there's nothing here and won't let you set up ip:port |
| 02:06.39 | Turl | I like mpd because it has a low footprint and plays my music |
| 02:07.39 | lundman | The auto discover is nice, when it works. But many upnp software lets you add manual address to add devices too |
| 02:07.50 | lundman | which is really nice if you are on non-broadcast networks |
| 02:08.09 | lundman | for example, PlugPlayer has "manual device url" |
| 02:08.27 | lundman | llink has smaller food print, and plays videos from RAR archives |
| 02:08.28 | Turl | lundman: another UPNP problem is the naming |
| 02:08.34 | lundman | now, that's important to me, probably not you :) |
| 02:08.37 | Turl | provider, player, and all that mess |
| 02:09.05 | Turl | if you play a remote file on a remote speaker, how is it called? :P |
| 02:09.26 | Turl | besides most players just stream the remote media to you and don't let you control the remote device or viceversa |
| 02:10.06 | Turl | lundman: btw, any control application as nice as https://dl.dropbox.com/u/553461/screenshot-1316717729227.png for UPNP? :) |
| 02:10.38 | lundman | now i do agree about the name |
| 02:10.48 | lundman | seriously, trigger a remote client to play from a remote server |
| 02:10.51 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 02:10.53 | lundman | UPNPMediaRenderer... serious |
| 02:11.43 | lundman | turl: eyecon maybe, plugplayer does too. but I write a cmdline mediarenderer anyway, and control from shell/cron/cgi |
| 02:13.03 | Turl | I've never been much of a cli guy for the 'traditional' stuff |
| 02:13.28 | Turl | I use a graphical music player, im client |
| 02:13.40 | Turl | a graphical code editor at tmes |
| 02:13.42 | Turl | times* |
| 02:13.52 | Turl | the exception is for code editing, I use nano most of the time |
| 02:15.04 | lundman | i never listen to music |
| 02:15.26 | lundman | but yeah, similarly mpd/upnp, that you can watch youtube, and just send it to mele to play on tv is nice |
| 02:16.25 | *** join/#arm-netbook tallnerd1985 (~tallnerd1@c-67-168-239-239.hsd1.wa.comcast.net) |
| 02:16.44 | lundman | ok found audio patch |
| 02:19.19 | Turl | I liked elisa media center back then |
| 02:20.25 | Turl | too bad they killed it :( |
| 02:21.08 | Turl | https://www.youtube.com/watch?v=PrAslOCLYwA |
| 04:22.57 | *** join/#arm-netbook gimli_ (~gimli@xbmc/staff/gimli) |
| 04:52.56 | *** join/#arm-netbook tallnerd1985 (~tallnerd1@c-67-168-239-239.hsd1.wa.comcast.net) |
| 05:29.21 | *** join/#arm-netbook eFfeM (~frans@g235028.upc-g.chello.nl) |
| 05:47.21 | gimli_ | what libMali.so, libGLESv2.so, libEGL.so. libUMP.so do you use for the mali driver in https://github.com/amery/linux-allwinner/commits/ ? |
| 05:52.00 | specing | steev: :( |
| 06:03.25 | hipboi | gimli_: http://rhombus-tech.net/allwinner_a10/Compile_X11_driver_for_A10/ |
| 06:04.13 | specing | hipboi: Have you investigated opencart? :) |
| 06:04.14 | gimli_ | to bad it's hardfp. my toolchain is softfp |
| 06:04.41 | hipboi | specing: yes, zencart seems nice |
| 06:04.58 | hipboi | specing: i will set up tonight |
| 06:05.37 | hipboi | gimli_: i think you can stay the r2p4 for now |
| 06:05.40 | specing | Ah I see there is a .cn zencart |
| 06:05.45 | hipboi | and not update the kernel |
| 06:06.02 | hipboi | because arm released the r3p0 mali driver |
| 06:06.11 | hipboi | the kernel has not updated |
| 06:06.23 | specing | Although I did not like it on the first sight, the .com site seems too comercialized |
| 06:07.04 | lundman | http://www.engadget.com/2012/06/25/google-tv-goes-international-sony-ns7-gs7-set-top-box-up-for-pr/ |
| 06:07.35 | lundman | zencart is nice, I've used it.. good multi-language support too |
| 06:13.08 | TomNL | gimli_:how is xbmc going? anything we can do to help? |
| 06:13.39 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 06:13.40 | lundman | yeah |
| 06:19.31 | TomNL | hipboi_: did you manage to ship the mele? |
| 06:32.40 | gimli_ | TomNL: we can talk about xbmc in the evening. have leave for work :( |
| 06:34.09 | lundman | dutch or german, so cet, i guess +8 hours from now |
| 06:34.44 | TomNL | gimli_: ok i will be there, have a nice day |
| 06:36.20 | hipboi | specing: i have a qurestion about the zencart, does it support paypal? |
| 06:36.44 | hipboi | TomNL: i will wait the post man picking up the package today |
| 06:36.51 | lundman | it has a module |
| 06:37.50 | TomNL | thank you hipboi_ |
| 06:38.16 | specing | hipboi: ^ lundman |
| 06:39.04 | *** join/#arm-netbook antony (~Adium@2001:67c:2e8:13:e948:252f:5cb1:32df) |
| 06:48.10 | *** join/#arm-netbook Quarx (~Quarx@46.233.252.170) |
| 06:55.52 | *** join/#arm-netbook eebrah (~chatzilla@41-139-199-130.safaricombusiness.co.ke) |
| 06:57.31 | *** join/#arm-netbook asdffdsa (~ubuntu@koln-4d0b5199.pool.mediaWays.net) |
| 06:58.39 | *** join/#arm-netbook rell (~chatzilla@p5B0785C5.dip0.t-ipconnect.de) |
| 07:14.20 | *** join/#arm-netbook eFfeM (~frans@g235028.upc-g.chello.nl) |
| 07:14.44 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 07:17.14 | *** join/#arm-netbook eFfeM (~frans@g235028.upc-g.chello.nl) |
| 07:22.02 | *** join/#arm-netbook hno (~hno@squid/developer/hno) |
| 07:23.38 | *** join/#arm-netbook antony (~Adium@guest196.guestnet.ripe.net) |
| 07:32.14 | *** join/#arm-netbook CiHO (~CiHO@koln-4d0b5199.pool.mediaWays.net) |
| 07:32.39 | *** join/#arm-netbook technologov (~user@israel.vpn.websense.com) |
| 07:32.44 | technologov | hi all ! |
| 07:33.01 | lundman | hello isreal |
| 07:33.44 | lundman | wonder if djmount is worth porting to mele |
| 07:34.42 | technologov | One of the biggest blockers for ARM hardware acceleration is the lack of video stack on Linux: there are too many libs / media frameworks on it: xine, ffmpeg, vlc, libva, gstreamer... and codecs (libvpx for VP8, ...) any ideas how developing A10 VPU drivers is possible ? |
| 07:35.02 | technologov | Is patching dozens of libs will be necessary ? |
| 07:38.43 | rm | how about getting the Xv support for a start |
| 07:39.38 | rm | talking about h/w decoding without even having Xv is like we can't even drive, and already think about participating in nascar |
| 08:09.12 | *** join/#arm-netbook dfletcher (~fletch@108-196-222-251.lightspeed.sntcca.sbcglobal.net) |
| 08:09.20 | *** join/#arm-netbook dfletcher (~fletch@drupal.org/user/72475/view) |
| 08:31.48 | *** join/#arm-netbook chiques_ (~tony@pool-173-60-8-237.lsanca.fios.verizon.net) |
| 08:36.39 | *** join/#arm-netbook antony (~Adium@2001:67c:2e8:13:95bf:5275:5fcb:cd57) |
| 08:55.22 | hno | crap, the watchdog is more broken than I thought. Needs a RTC clock cycle to reset the counter. If not it happily carries on where it was. |
| 08:57.38 | CIA-121 | rhombus-tech: Simon master * r6689ebb7d7bd /allwinner_a10/hacking_the_mele_a1000/android.mdwn: |
| 09:03.45 | CIA-121 | rhombus-tech: hno master * reef6ef5189ad /allwinner_a10/A10_register_guide/A10_TIMER.mdwn: |
| 09:44.09 | *** join/#arm-netbook tallnerd1985_ (~tallnerd1@c-67-168-239-239.hsd1.wa.comcast.net) |
| 09:48.15 | specing | Did anyone get usbnet on their tablet/mele? |
| 09:48.20 | specing | (to work) |
| 09:48.52 | lundman | usb works but maybe usbnet is something else |
| 09:49.01 | specing | networking over usb |
| 09:49.45 | *** join/#arm-netbook Quarx (~Quarx@46.233.252.170) |
| 09:49.45 | specing | Its is in lsmod on the device and I've modprobe just about everything under the "USB networking" category on the host |
| 09:50.22 | lundman | ah i see. not tried, have wire and wifi, why do you need usbnet? |
| 09:50.32 | lundman | also, usbethernet i tried, also works :) |
| 09:50.33 | lundman | 9from |
| 09:50.35 | lundman | err |
| 09:51.03 | specing | tablet |
| 09:51.10 | specing | ? |
| 09:51.31 | specing | There is no ethernet plug on my tablet, ofcourse; and wifi is terribly slow |
| 09:52.15 | specing | Also I've made a hacked-up heatsink for the A10 :P |
| 09:53.03 | CIA-121 | rhombus-tech: Eric master * rf285406dd141 /allwinner_a10/orders/831.mdwn: |
| 09:59.48 | hno | Where can I find lyf1-2.3.4-20120209-0.7.1-a721.img? |
| 10:00.05 | *** join/#arm-netbook piezo (~piezo@pdpc/supporter/active/piezo) |
| 10:00.37 | hno | google do not want to talk to me today... |
| 10:03.50 | rm | it's not "usb net", it's not "networking over USB" |
| 10:04.07 | rm | you just mean usb to ethernet devices |
| 10:04.25 | rm | while the former two may also mean something completely bizzare |
| 10:05.00 | specing | I do not mean usb to ethernet devices |
| 10:05.19 | specing | I mean networking over the USB cable going into my computer |
| 10:05.38 | lundman | oh |
| 10:05.44 | hno | have not tried. |
| 10:05.48 | specing | A similar term is "tethering", but it is not the same |
| 10:06.04 | rm | oh <- seconded |
| 10:06.11 | rm | so you do mean something completely bizzare |
| 10:06.16 | lundman | but I got partial sound working today, so there is that |
| 10:06.26 | lundman | slip over usb! |
| 10:06.45 | rm | on the subject of usb ethernet, # modprobe asix |
| 10:06.47 | rm | ERROR: could not insert 'asix': Exec format error |
| 10:06.47 | specing | No, usbnet is a different protocol |
| 10:06.56 | rm | [49632.720000] usbnet: disagrees about version of symbol module_layout |
| 10:06.59 | rm | what's up with that? |
| 10:07.33 | rm | am I using a kernel that is different from the root fs /lib/modules/ that I use |
| 10:07.37 | hno | rm, mismatch between kernel & modules. |
| 10:07.38 | rm | sounds likely |
| 10:07.59 | lundman | yeah it is, and its a pain |
| 10:08.24 | hno | it's a pain to not have matching kernel & modules indeed. Don't do that. |
| 10:08.34 | lundman | are you using the right config for your kernel/module build |
| 10:08.56 | lundman | i found a way to binary patch it, i put that on stackoverflow, but I doubt its worth doing |
| 10:09.01 | specing | What is an .axf doing in the middle of .ko's in /drv? |
| 10:09.12 | hno | specing, which one where? |
| 10:09.23 | specing | x_mac.axf |
| 10:09.27 | specing | /drv/x_mac.axf |
| 10:09.31 | hno | <PROTECTED> |
| 10:09.34 | specing | yeah |
| 10:09.44 | specing | Im stuck with android atm |
| 10:10.09 | lundman | androidn , you use the _crane_ conf to build |
| 10:11.38 | hno | specing, where is that Android image from? |
| 10:12.12 | rm | so I am looking for a newer prebuilt kernel (for now) + modules, then |
| 10:12.20 | rm | found debs posted at http://rhombus-tech.net/allwinner_a10/hacking_the_mele_a1000/debian_kernel/ |
| 10:12.20 | specing | hno: stock |
| 10:12.29 | lundman | mmm salmiakki |
| 10:12.34 | hno | specing, stock what? |
| 10:12.38 | rm | anyone knows if these *really* require flashing kernel into the internal flash and changing the uboot as described? |
| 10:12.47 | specing | the "ROM" that shipped with the device |
| 10:12.56 | lundman | none of them suggest flashing nand |
| 10:13.01 | lundman | they all are to boot SD directly |
| 10:14.01 | rm | hm, okay. I just used "flash-kernel" that it tells to install on a DNS-323, and it flashed NAND there |
| 10:16.19 | Turl | specing: it's usually called RNDIS |
| 10:16.57 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 10:17.35 | specing | Turl: RNDIS is M$'s crap |
| 10:18.34 | Turl | yeah, the true standard is named CDC-Ethernet iirc |
| 10:18.54 | Turl | but everyone just implements RNDIS because otherwise win craps on itself |
| 10:19.37 | specing | I've loaded all the modeules I could and no usbX interface is showing up :/ |
| 10:20.30 | Turl | on your PC or ARM device? |
| 10:20.42 | specing | PC |
| 10:20.57 | specing | the device already has all the modules inserted |
| 10:21.10 | Turl | if you have a decent distro it should be plug&play |
| 10:21.22 | Turl | did you switch the usb mode on the device? |
| 10:21.30 | specing | How? |
| 10:21.51 | Turl | echo 1 > /sys/......../rndis/enable |
| 10:21.56 | Turl | usually |
| 10:22.01 | specing | I've looked around setting but there is only config for hardware ethernet |
| 10:22.03 | Turl | (on the device) |
| 10:22.10 | specing | ok, I'll try that |
| 10:22.34 | Turl | on android, go to the tethering section |
| 10:22.41 | specing | There is none |
| 10:22.42 | Turl | and enable usb tethering |
| 10:23.00 | Turl | specing: are you using the android gadget? |
| 10:23.27 | specing | idk |
| 10:24.19 | Turl | /sys/class/android_usb/f_rndis |
| 10:24.23 | Turl | check around there |
| 10:24.35 | Turl | my intel gpu hung fu :( |
| 10:24.39 | specing | Also none of the rndis /sys entries have an enable file :( |
| 10:24.47 | Turl | let me reboot this |
| 10:24.48 | Turl | brb |
| 10:24.58 | specing | ls /sys/class/android_usb/f_rndis |
| 10:25.00 | specing | /sys/class/android_usb/f_rndis: No such file or directory |
| 10:25.43 | Turl | do you even have the android directory there? |
| 10:27.17 | specing | nope |
| 10:27.26 | specing | Not on the device and not on the host |
| 10:28.05 | Turl | you aren't using the android gadget I suppose then |
| 10:28.20 | Turl | specing: check that you have CONFIG_G_ANDROID set |
| 10:28.27 | Turl | (off the top of my head, name may vary) |
| 10:30.33 | specing | Where? on the host? |
| 10:31.03 | Turl | no |
| 10:31.05 | Turl | on the device |
| 10:31.32 | specing | How do I check that? |
| 10:32.05 | Turl | open your kernel config and check? :P |
| 10:32.19 | specing | kernel is stock |
| 10:32.24 | specing | I don't have the config |
| 10:32.38 | Turl | ah |
| 10:32.42 | hno | my googlefu is failing me. no lyf1-2.3.4-20120209-0.7.1-a721.img anywhere? |
| 10:32.51 | Turl | maybe they just didn't enable it? |
| 10:33.06 | Turl | try with a new kernel |
| 10:33.14 | Turl | time to reboot this hanged gpu |
| 10:33.18 | specing | Turl: the device does have an usbnet module in lsmod |
| 10:34.03 | rm | http://rhombus-tech.net/allwinner_a10/hacking_the_mele_a1000/debian_kernel/ this mentions reconfiguring uboot, I suppose that's because it's not configured to use uInitrd by default, right? |
| 10:34.17 | rm | the kernel in the prebuilt images I seen is monolithic |
| 10:34.22 | rm | no initrd |
| 10:34.51 | Turl | specing: usbnet != rndis afaik |
| 10:36.15 | specing | I dont care if it is usbnet or rndis or whatever, I just want it to work... |
| 10:37.17 | lundman | so.. magic? |
| 10:37.47 | specing | if it works... |
| 10:38.06 | lundman | tilts his hand back and forth |
| 10:38.48 | Turl | specing: usbnet is 10Mbit tops iirc |
| 10:39.18 | specing | It should be 480 MBit raw |
| 10:39.28 | Turl | tethering from a tablet is.. weird btw |
| 10:39.41 | Turl | that's probably why they didn't enable/implement it |
| 10:40.52 | specing | I want it to access my computer where I have lots of NFS stores, not the other way around |
| 10:41.02 | specing | NFS works through wifi though |
| 10:41.06 | specing | its slow as hell |
| 10:41.15 | Turl | so you want reverse tethering? |
| 10:41.19 | lundman | nfs yeah.. use tcp at least |
| 10:41.49 | specing | Turl: I guess? |
| 10:41.56 | Turl | specing: rndis as configured by android won't give you what you need, it's set up on the inverse way |
| 10:42.16 | Turl | your tablet is the gateway and your pc the client |
| 10:43.17 | hno | technically it can be used either way, but usually everyone expects the gadget (i.e. tabled) to be the gateway. |
| 10:43.45 | Turl | specing: have you tried cifs+smb? I played a video the other day over wifi with it |
| 10:43.51 | specing | Also is there a way to have all the UARTs show up in android? |
| 10:43.56 | Turl | worked good, the video was like 480p though |
| 10:44.18 | specing | Turl: Its pointless if the wifi is slow |
| 10:44.43 | lundman | nfs is faster than smb, but make sure to mount with tcp |
| 10:44.59 | Turl | well my wifi can pump ~3MB/s |
| 10:45.25 | lundman | not enough for 1080p :) |
| 10:45.26 | specing | WOW, thats A LOT!!!!111oneone |
| 10:45.43 | Turl | lundman: 1080p content is usually ~1.5G |
| 10:45.58 | lundman | heh. no |
| 10:46.17 | Turl | (1.5 GB) / (3 (MB / s)) = 8.53333333 minutes |
| 10:46.45 | lundman | LOTR for example, peaks at 57MB/s, which is about 6MB/s |
| 10:46.55 | Turl | lowut? :P |
| 10:46.57 | Turl | (4.7 GB) / (3 (MB / s)) = 26.7377778 minutes |
| 10:47.03 | Turl | even for DVD-sized mkv releases |
| 10:47.20 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 10:47.31 | lundman | MKV is heavily compressed, compared to a 50GB .ts :) |
| 10:47.49 | Turl | who downloads 50GB .ts? :P |
| 10:48.02 | lundman | i have a few |
| 10:48.09 | lundman | 100mbit fiber, no limits |
| 10:48.15 | Turl | besides they're expensive to keep |
| 10:48.23 | Turl | oh well, I have 6Mbit down |
| 10:48.45 | Turl | lundman: get an 802.11ac router :) |
| 10:48.51 | lundman | but yeah, i thought we were tlaking bandwidth to play 1080p |
| 10:49.10 | Turl | yeah |
| 10:49.26 | Turl | for the 1080p that I watch, 3MB is more than plenty |
| 10:49.31 | Turl | 3MB/s* |
| 10:49.33 | lundman | ok, movie time |
| 10:50.55 | TomNL | lundman: did you fix audio outs? |
| 10:51.16 | lundman | ton: i used the audiopatch floating around, which enables mkv/aac/ac3 audio |
| 10:51.18 | lundman | i dont think DTS |
| 10:52.27 | TomNL | ok cool :) |
| 11:07.02 | specing | 12:46 < specing> Also is there a way to have all the UARTs show up in android? |
| 11:08.01 | mnemoc | specing: configuring them in your script.bin |
| 11:09.03 | specing | looks |
| 11:10.16 | mnemoc | as long as the pins don't conflict with another feature obviusly |
| 11:11.16 | specing | mnemoc: Were you the one with that script (de)compiler? If so, where is it? |
| 11:11.43 | specing | ah found it |
| 11:11.50 | mnemoc | github/amery/sunxi-tools |
| 11:20.34 | specing | Could the 6 pad header near the A10 in fact be JTAG? |
| 11:21.21 | specing | https://ajdef.dynsite.net/A10-MID/hardware/board_upper.jpg right from the unbpopulated QFT |
| 11:21.46 | specing | I've probed it with a scope and S0 is not there |
| 11:22.12 | mnemoc | most pins in A10 are GPIO by default |
| 11:25.15 | RITRedbeard_ | mnemoc, but by default linux doesn't know about them or does anything with them, correct? |
| 11:26.41 | mnemoc | RITRedbeard_: nothing unless initialized for a supported function |
| 11:26.51 | mnemoc | which is the job of script.bin |
| 11:27.06 | specing | Is there a map of which uarts can be placed on which pins? |
| 11:28.16 | mnemoc | look at https://docs.google.com/open?id=0B-xbPiTMurpSTTlRXzlaQ2lZTVE |
| 11:28.46 | RITRedbeard_ | I haven't had a chance to go through the boot process, said my MK802 shipped but hasn't shown up yet. |
| 11:28.53 | RITRedbeard_ | I wonder if customs seized it? |
| 11:29.04 | RITRedbeard_ | or if Shenzen is actually behind and they just never told me |
| 11:29.23 | Turl | RITRedbeard_ did you wait a month since shipping? |
| 11:29.28 | RITRedbeard_ | Nearly. |
| 11:29.32 | Turl | that's usually what it takes since they ship it |
| 11:29.52 | RITRedbeard_ | I'm in no hurry. I have the bifferboard and AVR to toy with in the mean time. :) |
| 11:30.11 | Turl | mnemoc: https://github.com/allwinner-dev-team/linux-allwinner/commit/771f4a7ea01c9fed2a0198310aa17a9d7c65200b :) |
| 11:31.33 | mnemoc | Turl: what's that? |
| 11:31.49 | Turl | mnemoc: ramconsole ;) |
| 11:32.18 | Turl | mnemoc: if you never heard of it, it's a ram-based console |
| 11:32.33 | Turl | the console output is stored on a fixed ram region |
| 11:32.46 | Turl | and on boot you can read the previous session's console |
| 11:32.55 | Turl | ideal to catch kernel panics and the like |
| 11:33.36 | mnemoc | ic |
| 11:33.43 | RaYmAn | Turl: it's just a pity that not all devices (like mk802) have a way to force-reboot without disconnecting power |
| 11:34.16 | Turl | RaYmAn: if you're quick enough, ram contents hold decently :) |
| 11:34.32 | Turl | you'll see a bit of corruption but it'll still be there |
| 11:34.34 | RaYmAn | lol |
| 11:35.05 | Turl | RaYmAn: maybe enabling and bumping the ECC ramconsole values can help with your usecase :) |
| 11:35.58 | RaYmAn | heh |
| 11:36.11 | *** join/#arm-netbook CiHO (~CiHO@koln-5d818cb6.pool.mediaWays.net) |
| 11:36.32 | RaYmAn | Turl: IEF found serial console (At least TX from device), so I just need to repeat that trick =P |
| 11:36.48 | Turl | heh :) |
| 11:37.09 | mnemoc | RaYmAn: got the courage to solder the wires? |
| 11:37.13 | Turl | hno said the watchdog works on uboot btw |
| 11:37.46 | RaYmAn | mnemoc: I've been building it up for the past few days :P But now I can't find my solder :( |
| 11:37.54 | mnemoc | doh |
| 11:38.43 | RaYmAn | indeed |
| 11:38.48 | specing | I see i could bring out uart though the camera port |
| 11:39.18 | mnemoc | some people has even used the twi of the vibrator motor iirc |
| 11:39.18 | RaYmAn | I'm kind of expecting that I'll have a dead mk802 once I do this, but meh :P I'll have an excuse to get a mele then. |
| 11:39.23 | specing | mnemoc: Do sunxi-tools check that only one function is being used on a pin? |
| 11:39.36 | mnemoc | specing: nope, no validation at all yet |
| 11:39.40 | specing | mnemoc: That was my idea, originally... |
| 11:39.53 | rm | I wonder will I be a total pervert if I use my MK802 as a file/http/NAS server :) |
| 11:40.00 | mnemoc | yes |
| 11:40.09 | rm | \o/ |
| 11:40.23 | RITRedbeard_ | Next thing you know you'll want XMBC on it! |
| 11:40.28 | RITRedbeard_ | To #raspberrypi you go! |
| 11:40.40 | rm | been there <s>done that</s> |
| 11:41.14 | RITRedbeard_ | A10/mk802 drawing 2 amps at load? |
| 11:41.34 | RITRedbeard_ | gonna need a better battery I think if this is true |
| 11:41.41 | rm | I'm no good with this electrics stuff |
| 11:41.56 | rm | I have my Mele A2000 connected via a watt meter |
| 11:42.04 | rm | it shows 4.6W at idle |
| 11:42.18 | Turl | RaYmAn: don't overheat it :) |
| 11:42.19 | rm | hth :P |
| 11:42.20 | RITRedbeard_ | oh god I'm a software guy oh jesus lord |
| 11:42.32 | RITRedbeard_ | I = R V? V R = I? |
| 11:42.40 | *** part/#arm-netbook antony (~Adium@2001:67c:2e8:13:95bf:5275:5fcb:cd57) |
| 11:42.46 | mnemoc | who cares? |
| 11:42.47 | RITRedbeard_ | sounds like nearly an amp at idle if supply wants 5VDC |
| 11:42.50 | RaYmAn | Turl: the pads are literally, RIGHT next to the a10 chip :/ |
| 11:42.54 | hno | Turl, well.. furter investigation on the watchdog shows why there is no driver. It's not usable as a watchdog, only temporary safeguard. |
| 11:42.59 | RITRedbeard_ | but sounds like a healthy current |
| 11:43.37 | rm | that reminds me, I am horribly understocked on Chinese-to-Euro electric plug thingies |
| 11:43.38 | hno | Turl, the small window I mentioned yesterday is not small at all. it's one second. |
| 11:43.43 | specing | mnemoc: it would be easy to implement the checking... in C++ |
| 11:43.44 | rm | even had to use that meter as one |
| 11:44.00 | Turl | and how big is the watch window hno? |
| 11:44.04 | specing | just throw them into a vector and iterate over it on every line |
| 11:44.33 | hno | Turl, takes one second to rearm the watchdog. Duringthat time you have no watchdog. |
| 11:45.31 | specing | mnemoc: No way to use the vibra motor as a tx line.... the only other function on that pin is IR0 |
| 11:45.35 | mnemoc | doesn't sound like a winner watchdog :< |
| 11:45.58 | hno | it's a dog that runs away and hides every time you say hi |
| 11:46.04 | mnemoc | :D |
| 11:46.17 | Turl | hno: I'd take a watchdog that works 4/5 over none though :P |
| 11:47.26 | hno | I guess that with a sufficiently long update frequency it might work. Max timeout is ca 2^6 seconds. |
| 11:48.02 | Turl | that's like a minute |
| 11:48.03 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 11:48.08 | hno | not sure what the actual time unit it, but seems to be slightly shorter than a second. |
| 11:48.28 | hno | so about a minute. |
| 11:49.06 | hno | the timer width is 6 bits. |
| 11:49.29 | Turl | 49/50s of coverage sounds decent |
| 11:51.24 | hno | Turl, you are very welcome to write a driver for it. Linux watchdog drivers are very simple. Discovered register details is at http://rhombus-tech.net/allwinner_a10/A10_register_guide/A10_TIMER/?updated#index9h2 |
| 11:53.02 | Turl | hno: so 0b111 / 7 should reboot the system on a second right? |
| 11:53.35 | specing | UART3 is on the same pin as the G sensor |
| 11:53.46 | specing | How does the G sensor even look like? |
| 11:54.21 | specing | Is it the metalic can in top right corner of https://ajdef.dynsite.net/A10-MID/hardware/board_upper.jpg ? |
| 11:54.31 | Turl | specing: they look like a chip |
| 11:54.32 | Turl | https://lh3.ggpht.com/-hm2fX25dyC8/T6w6KjPtbCI/AAAAAAAAHYI/ObIVfgrWR80/s320/akm8975_02.jpg |
| 11:54.58 | CIA-121 | rhombus-tech: hno master * ra31246702e54 /allwinner_a10/A10_register_guide/A10_TIMER.mdwn: |
| 11:55.05 | hno | Turl, yes, 0x011 or 0x111 both resets within a second. |
| 11:55.29 | hno | s/0x/0b/g |
| 11:55.29 | Turl | bottom center maybe specing |
| 11:56.08 | Turl | hno: why 0b011 ? |
| 11:58.18 | specing | Turl: I found mine on freescale, its a 3*3 mm chip, no way to solder anything there |
| 11:58.35 | hno | Turl, 0b011 is what uboot and kernel uses. |
| 11:59.08 | hno | but I can not see any different delay with 0b111 |
| 12:00.15 | hno | 0b011 starts the counter at 0, so should reset "immediately". |
| 12:00.41 | hno | 0b111 starts counter at 1, which is apparently decremented immediately after load. |
| 12:02.01 | hno | have not found a way to read the current counter value. Only the interval register. |
| 12:02.24 | Turl | hno: is this sun4i only or general sunxi? |
| 12:02.41 | hno | most likely generic, but I have not looked at other versions. |
| 12:05.04 | hno | sun5i is the same. sun3i does not seem to have a reset method? |
| 12:05.28 | hno | so no idea if sun3i have the watchdog or not. |
| 12:05.45 | mnemoc | sad the sun3i guy didn't join us |
| 12:18.17 | *** join/#arm-netbook arete74 (~arete74@host160-66-static.225-95-b.business.telecomitalia.it) |
| 12:18.35 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 12:19.30 | *** join/#arm-netbook lkcl (~lkcl@host217-43-147-132.range217-43.btcentralplus.com) |
| 12:20.59 | *** join/#arm-netbook newbie (~kvirc@220-136-79-165.dynamic.hinet.net) |
| 12:21.37 | CIA-121 | rhombus-tech: lkcl master * r508ac43095ad /allwinner_a10/pcb.mdwn: add litkconn link |
| 12:21.47 | CIA-121 | rhombus-tech: lkcl master * r1241a9b1cb3b /: Merge branch 'master' of ssh://git.hands.com/rhombus |
| 12:23.17 | *** join/#arm-netbook hp_ (~kvirc@220-136-79-165.dynamic.hinet.net) |
| 12:32.47 | Turl | hno: so you need to stop it, wait for a clock tick, then start it, wait for another tick? |
| 12:32.59 | Turl | that's 2 seconds right? |
| 12:34.44 | *** join/#arm-netbook IEF (ief@shadowchild.nl) |
| 12:48.38 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 12:48.40 | *** join/#arm-netbook Salam (~chatzilla@31.57.77.26) |
| 12:52.17 | *** join/#arm-netbook eebrah (~chatzilla@41-139-199-130.safaricombusiness.co.ke) |
| 13:30.43 | Turl | hno: so, given this http://paste.ubuntu.com/1059058/ |
| 13:31.10 | Turl | writing WATCHDOG_ARM | WATCHDOG_REBOOT | WATCHDOG_SECONDS(50) should be ok right? |
| 13:32.44 | Turl | doesn't seem to do anything |
| 13:33.33 | *** join/#arm-netbook test (~test@WimaxUser36145-155.wateen.net) |
| 13:49.20 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 14:10.27 | *** join/#arm-netbook NinjAPT (~NinjAPT@bl17-21-142.dsl.telepac.pt) |
| 14:50.46 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 14:57.46 | Turl | hno: my bad, I forgot to add the platform device |
| 15:09.34 | Turl | hno: doesn't seem to like big numbers |
| 15:09.43 | Turl | 60 seems to reboot it instantly |
| 15:27.47 | traeak | reminds me, probably need to figure out the kernel command line to set the proper video mode |
| 15:27.49 | traeak | hmm |
| 15:30.09 | *** join/#arm-netbook rell_ (~chatzilla@p5B0785C5.dip0.t-ipconnect.de) |
| 15:32.17 | *** join/#arm-netbook rzk (~rzk@2002:2590:28d2:1:222:4dff:fe7c:450f) |
| 15:49.14 | *** join/#arm-netbook rzk (~rzk@2002:5f1c:829b:1:222:4dff:fe7c:450f) |
| 15:51.51 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 16:03.21 | *** join/#arm-netbook rzk (~rzk@2002:5f19:4714:1:222:4dff:fe7c:450f) |
| 16:11.41 | *** join/#arm-netbook traeak (~bolsen@75-151-91-165-Colorado.hfc.comcastbusiness.net) |
| 16:12.02 | zenitraM | http://www.gp32spain.com/foros/showthread.php?91412-ZEOMA-Especificaciones-T%E9cnicas-Hardware-Antiguo-proyecto-VICIOSA |
| 16:12.36 | zenitraM | ^ some people drafting a possible EOMA + mini-tablet + gaming controls dock + keyboard dock project |
| 16:13.13 | mnemoc | oh oh |
| 16:13.43 | mnemoc | anyone we know? |
| 16:15.20 | zenitraM | i'm on that forum... and suggested EOMA on another post there some weeks ago |
| 16:15.31 | mnemoc | nice |
| 16:15.37 | zenitraM | it seems they liked my suggestion... |
| 16:15.58 | mnemoc | hope we have actual boards on time... |
| 16:16.16 | zenitraM | although they are now just exercising imagination.. doubt they can raise money to make that reality easier |
| 16:16.25 | zenitraM | s/easier/soon/ |
| 16:17.02 | Turl | zenitraM: with kickstarter, they can do so on a day or so :P |
| 16:17.16 | mnemoc | kickstarter is US-only |
| 16:17.25 | *** join/#arm-netbook von_fritz (~fritz@host57-254-dynamic.211-62-r.retail.telecomitalia.it) |
| 16:17.27 | Turl | didn't know that |
| 16:17.45 | mnemoc | but i suppose there are alternatives |
| 16:18.51 | Turl | who wants to play with the watchdog? |
| 16:19.29 | Turl | http://paste.debian.net/176258/ |
| 16:19.59 | Turl | doesn't seem to actually work |
| 16:20.15 | Turl | at least now how one would want it to :) |
| 16:20.19 | Turl | now how* |
| 16:20.24 | Turl | not how* fail |
| 16:24.15 | *** join/#arm-netbook NinjAPT (~NinjAPT@233.163.108.93.rev.vodafone.pt) |
| 16:38.17 | specing | Anyone have any boot.scr's around? I assume this shows up on the screen? |
| 16:50.52 | CIA-121 | rhombus-tech: Sepp master * r5d4c8fee2f22 /allwinner_a10/orders/kackmail.mdwn: |
| 16:51.20 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 16:52.31 | specing | What is the difference between bImage (stock kernel) and uImage that hno's uboot expects? |
| 16:52.55 | Turl | uImage is for uboot |
| 16:53.05 | Turl | bImage is the same as "Image" |
| 16:53.22 | Turl | afaik you can use zImage too instead of bImage |
| 16:53.36 | specing | Then how does the internal uboot boot the bImage? |
| 16:53.46 | specing | Will the new uboot boot the old bImage? |
| 16:54.05 | Turl | the stock uboot loads a android boot.img afaik |
| 16:54.19 | specing | Im trying to reuse working stuff to find a serial port |
| 16:54.53 | specing | I've enabled all but one serial ports in script.bin and Im now setting up uboot on a spare SD card |
| 16:55.13 | specing | Is there a way/tool to convert the bImage to an uImage? |
| 16:58.07 | Turl | specing: mkimage or something like that, on the uboot package |
| 16:58.34 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
| 17:03.45 | *** join/#arm-netbook revident (~scott@fw.iplink.net) |
| 17:12.29 | specing | Hmm, mkimage expect a ton of parameters |
| 17:13.55 | TomNL | gimli: how are things going with xbmc, need any help? |
| 17:15.00 | *** join/#arm-netbook madmalkav (~Hamlet@188.86.253.69) |
| 17:17.37 | gimli | TomNL: currently i hang with alpha blending of the gui and the decoder is a bit whacky. |
| 17:18.26 | TomNL | gimli: you run without X i assume? |
| 17:18.31 | gimli | yes |
| 17:19.46 | gimli | runns most of the time headless :D |
| 17:20.13 | *** join/#arm-netbook NinjAPT (~NinjAPT@233.163.108.93.rev.vodafone.pt) |
| 17:20.42 | *** join/#arm-netbook rell01 (~chatzilla@p4FE562A9.dip0.t-ipconnect.de) |
| 17:20.48 | TomNL | hehe even developping headless ;-) |
| 17:21.01 | gimli | sure ;) |
| 17:21.58 | specing | Damn now I have forgot where I put my modified script.bin :( |
| 17:22.13 | gimli | hehe, so much to headless :D |
| 17:22.20 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 17:23.43 | Turl | gimli: did you take over work on the directfb patches? |
| 17:24.15 | gimli | nope. we runn directly on the framebuffer with mali EGL rendering |
| 17:26.24 | TomNL | gimli: how is the decoder whacky? performance whise? |
| 17:28.01 | gimli | yes, and i also susbect the linux libarary doesn't return the correct frame order. |
| 17:31.54 | specing | Doesen't boot... |
| 17:32.53 | specing | mnemoc: fex2bin segfaults on bin2fex's output on my stock script.bin... |
| 17:34.22 | specing | err hno |
| 17:34.37 | specing | or mnemoc |
| 17:34.38 | specing | or both |
| 17:43.06 | *** join/#arm-netbook omgmog_ (~textual@host109-148-65-161.range109-148.btcentralplus.com) |
| 17:44.42 | specing | fel write 0x43000000 /dev/shm/dir/script.bin |
| 17:44.42 | specing | usb send: Input/output error |
| 17:48.15 | mnemoc | specing: can you upload both (.bin and .fex) somewhere and open a ticket? |
| 17:49.02 | mnemoc | s/open a ticket/file an issue/ |
| 18:06.09 | *** join/#arm-netbook madmalkav (~Hamlet@188.86.253.69) |
| 18:07.27 | specing | mnemoc: https://ajdef.dynsite.net/A10-MID/backup/nanda-boot/ |
| 18:08.51 | mnemoc | specing: I can't look at it now, please open a ticket on github |
| 18:09.17 | mnemoc | link both files, the .bin and the .fex |
| 18:09.20 | specing | D:D:D:D: opening tickets |
| 18:09.42 | specing | Its not critical though |
| 18:10.04 | traeak | wow, those files have been there a very long time |
| 18:10.08 | specing | The fex can be fixed (line 802) with an editor |
| 18:10.22 | specing | traeak: :D |
| 18:10.45 | traeak | correct time tags are very helpful though |
| 18:10.58 | specing | Can't help it, stock rom :| |
| 18:23.12 | mnemoc | specing: and my ticket? |
| 18:23.21 | specing | lazy :S |
| 18:25.09 | specing | mnemoc: same as this: https://github.com/amery/sunxi-tools/issues/1 |
| 18:25.33 | specing | the :pmu thing |
| 18:25.37 | mnemoc | you said "fex2bin segfaults on bin2fex's output" |
| 18:26.34 | specing | It does, on that line :P |
| 18:27.12 | mnemoc | #1 doesn't involve any segmentation fault. please file a new ticket |
| 18:27.26 | specing | Wait |
| 18:27.30 | specing | It doesent |
| 18:27.37 | specing | Something else segfaultet |
| 18:27.40 | specing | Hmm |
| 18:27.45 | specing | 's head is confused |
| 18:29.48 | mnemoc | in #1 both bin2fex and fex2bin do their job fine. problem is the closed source compiler doesn't validate the field names while fex2bin does |
| 18:30.53 | mnemoc | so the stock bin include properties with semantically invalid names which fex2bin refuses to accept |
| 18:31.44 | mnemoc | the planned fix will comment out the invalid fields |
| 18:32.25 | mnemoc | generating ; invalid field = value instead of invalid field = value |
| 18:40.56 | specing | Still not booting :/ |
| 18:59.13 | *** join/#arm-netbook j1nx (~IceChat9@163-62-223.ftth.xms.internl.net) |
| 18:59.39 | j1nx | Good evening fellows |
| 19:02.35 | TomNL | good evening j1nx |
| 19:03.06 | specing | Still not booting D: |
| 19:03.48 | specing | What would happen if I moved my modified script.bin to nanda and it would not boot anymore? Is the fel tool in a "usable" state? |
| 19:04.59 | Turl | specing: can't you use the 'livesuite' program on your device? |
| 19:05.20 | specing | Well idk |
| 19:05.41 | specing | Would it work on a prehistoric XP box? |
| 19:06.06 | DonkeyHotei | if it has usb, quite likely yes |
| 19:06.25 | specing | Well Im moving the script to nanda |
| 19:06.59 | Turl | specing: yeah, you'd need a firmware .img though |
| 19:07.30 | specing | Umm... |
| 19:07.37 | specing | that scared me |
| 19:07.52 | *** join/#arm-netbook NinjAPT (~NinjAPT@bl15-88-159.dsl.telepac.pt) |
| 19:08.57 | specing | I dont even know who the manufacturer of this tablet is |
| 19:55.01 | j1nx | hmm, those 4 pins on the uHost are NOT UART |
| 19:55.48 | j1nx | They are for an extension board. Looks like a USB/PS2 type of device |
| 19:56.22 | j1nx | http://ovalelephant.com/flymouse1.png |
| 19:56.31 | j1nx | http://ovalelephant.com/flymouse2.png |
| 20:03.38 | RITRedbeard | silly question about MK802: is the full size USB port is responsible for powering the device, correct? |
| 20:03.58 | j1nx | Do you thing we can reroute the UART to those pins by changing the script.bin? |
| 20:08.55 | mnemoc | j1nx: you would need to identify the pins and then check the datasheet if you can assign them to one of the 8 uarts |
| 20:15.12 | *** join/#arm-netbook madmalkav (~Hamlet@188.86.253.69) |
| 20:32.23 | j1nx | mnemoc: is this up to date; http://rhombus-tech.net/allwinner_a10/pinouts/ |
| 20:33.22 | mnemoc | no idea, prefer the datasheet |
| 20:34.13 | mnemoc | https://docs.google.com/open?id=0B-xbPiTMurpSTTlRXzlaQ2lZTVE |
| 20:45.04 | j1nx | * reading |
| 20:59.04 | lkcl | jlnx: i wrote that page a long time ago, right at the beginning of the project, from the datasheet, which has information missing and sometimes incorrect. |
| 21:00.18 | j1nx | Oh, no worries. That device will not be shipped before the 3rd of July, so plenty of time to read through the document. Am already glad it is in english ;) |
| 21:00.40 | lkcl | jlnx: you want this - http://git.hands.com/?p=eoma.git |
| 21:01.18 | lkcl | then fire that up in kicad (or just look at the allwinner.lib file directly) |
| 21:02.57 | lkcl | jlnx: i created the allwinner.lib kicad part directly from the wits-tech EVB schematics, which is *the* authoritative available source of accurate information regarding the A10 CPU. so i *may* have introduced errors but i did ask people to help verify the schematics some months back |
| 21:09.32 | j1nx | lkcl: To be honest with you. That kicad stuff is out of my comfort zone. (More a software man) |
| 21:10.31 | mnemoc | j1nx: https://github.com/amery/sunxi-tools/blob/pins/pins_a10.c <--- so something like this is better? |
| 21:11.22 | mnemoc | (still incomplete) |
| 21:12.51 | mnemoc | the goal of that branch was to add pin listing and validation to fex2bin |
| 21:14.24 | j1nx | So PF2 and PF4 are the UART we need to attach to two of those pins |
| 21:14.48 | j1nx | So now we have to search for the PS2/USB pins |
| 21:15.11 | j1nx | (if I am right the extension board is nothing more then a PS/USB dingle |
| 21:15.21 | mnemoc | checkout that branch, there is a tool (`pins`) that lists them in a grep friendly way |
| 21:15.56 | j1nx | Will do, if I ahve my hands on it |
| 21:20.16 | j1nx | Who is working on any A13 tablet? |
| 21:26.47 | mnemoc | i believe no one yet |
| 21:27.53 | mnemoc | j1nx: but if you want to write the pins table for it this is the datasheet (60p) I found https://docs.google.com/open?id=0B-xbPiTMurpSRU41aVRlcXlqdlU |
| 21:30.41 | j1nx | mnemoc: Other project. An A13 might be on his way to start playing around with. |
| 21:31.04 | j1nx | Will search for UART on that thing first ;) |
| 21:33.19 | mnemoc | we also have a kernel branch with sun5i changes pending to merge into 3.0-v2 |
| 21:33.36 | j1nx | mnemoc: I know ;) |
| 21:34.27 | mnemoc | well... my point was "may be you can help us to get that stuff merged" :) |
| 21:36.00 | j1nx | Exactly my point as well |
| 21:36.08 | mnemoc | great :) |
| 21:36.16 | j1nx | Anyway, I call it a day |
| 21:36.21 | j1nx | goodnight |
| 21:36.38 | mnemoc | good night j1nx |
| 21:42.38 | RITRedbeard | mk802 has standard sized HDMI female, no? |
| 21:44.03 | mnemoc | yes |
| 21:44.14 | mnemoc | .oO |
| 21:45.37 | mnemoc | nope |
| 21:45.48 | RITRedbeard | mini? |
| 21:48.50 | mnemoc | seems micro to me |
| 21:49.15 | mnemoc | scratches his head |
| 21:50.22 | mnemoc | i should better go to sleep too |
| 21:51.00 | mnemoc | RITRedbeard: yes, mini |
| 21:52.06 | mnemoc | http://liliputing.wpengine.netdna-cdn.com/wp-content/uploads/2012/06/mk802_006.jpg |
| 21:52.17 | mnemoc | good night |
| 21:54.08 | RITRedbeard | night |
| 21:54.15 | Turl | night mnemoc |
| 22:24.53 | hno | specing, fel can not write to DRAM unless you initialized it first.. and I don't have a working code snippet doing that in a fel compatible manner yet. |
| 22:25.41 | hno | fel is very very low level. |
| 22:26.20 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
| 22:38.09 | lundman | we must go lower |
| 22:38.45 | lundman | felception |
| 22:39.09 | DonkeyHotei | divides fel by zero |
| 23:03.00 | hno | only jtag is lower from a software perspective. |
| 23:04.04 | hno | wonder what the other fel commands do.. found a bunch of them in BROM/FEL. |
| 23:07.04 | hno | Confusing with several different ARM instructions for doing the same thing. |
| 23:17.33 | *** join/#arm-netbook Pichet (~Pichet@91.180.34.136) |
| 23:38.40 | *** join/#arm-netbook ssvb (~ssvb@a88-114-220-213.elisa-laajakaista.fi) |
| 23:41.04 | *** join/#arm-netbook eFfeM_work (~frans@D4B268BA.static.ziggozakelijk.nl) |
| 23:41.49 | *** join/#arm-netbook gsilvis (~almostsix@50.12.163.241) |
| 23:44.20 | *** join/#arm-netbook Gumboot (~gumboot@rev.bovine.muck.net.nz) |
| 23:56.56 | *** join/#arm-netbook nibb (~jyuyu@66.227.218.87.dynamic.jazztel.es) |
