| 00:04.22 | specing | is idling too |
| 00:20.39 | lundman | :) |
| 00:35.39 | *** join/#arm-netbook vectory_ (~damon@unaffiliated/vectory) |
| 01:05.50 | *** join/#arm-netbook hipboi (~hipboi@221.4.213.95) |
| 01:18.37 | CIA-91 | rhombus-tech: Ashley master * rd273957f5db5 /allwinner_a10/orders/64738.mdwn: |
| 01:20.16 | CIA-91 | rhombus-tech: lkcl master * rbc092e0348ee /allwinner_a10/orders/ (4 files in 2 dirs): order processing cleanup |
| 02:07.03 | CapnKernel | mnemoc: Just digesting #arm-netbook now... |
| 02:40.00 | mikey_w | Built from sources and the login screen shows but after logging the desktop is empty with no home button, |
| 02:40.43 | mikey_w | The upper bar with time, etc. is displayed but I can't do anything. Any ideas? |
| 02:46.54 | ZaEarl | mikey_w, what are you building? |
| 02:48.28 | CapnKernel | I sent some PCBs by DHL from China to my customer in Australia last week - 1 day! http://www.dhl.com/content/g0/en/express/tracking.shtml?brand=DHL&AWB=8639520901 |
| 02:48.38 | CapnKernel | And just USD20 |
| 02:49.02 | CapnKernel | My customer == v. happy man |
| 02:56.50 | CapnKernel | mnemoc: QingPei can do that checking an order of magnitude faster than me. But if he doesn't, I will. The online diff is a win! |
| 02:57.20 | hp_ | weee my usb to tty plug just arrived |
| 03:58.24 | DonkeyHotei | points out to ZaEarl that hipboi is back online |
| 03:58.48 | mikey_w | armel-rootfs-201205132225.tgz |
| 04:17.06 | CIA-91 | rhombus-tech: Richard master * r5ba9acb6d85e /allwinner_a10/orders/fox.mdwn: |
| 04:30.30 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 05:25.29 | CIA-91 | rhombus-tech: alexandr master * r40401827b4bb /allwinner_a10/orders/alex2.mdwn: |
| 05:26.31 | CIA-91 | rhombus-tech: alexandr master * r03ee240d70c4 /allwinner_a10/orders/alex2.mdwn: |
| 05:47.53 | *** join/#arm-netbook arokux_h (~roman@HSI-KBW-046-005-103-006.hsi8.kabel-badenwuerttemberg.de) |
| 06:54.29 | mnemoc | CapnKernel: thanks :) |
| 06:54.53 | mnemoc | QingPei: ping |
| 06:58.14 | zenitraM | two minutes |
| 06:58.18 | zenitraM | for: http://gaga.aliexpress.com/100008408-553345085-detail.html |
| 06:58.28 | zenitraM | in case someone wants to try to score one |
| 07:00.35 | zenitraM | aand sold out. |
| 07:00.43 | mnemoc | zenitraM: don't forget the shipping cost. |
| 07:01.42 | zenitraM | yup, but even with that it ended at 52$ |
| 07:03.32 | mnemoc | for a craplet |
| 07:04.14 | mnemoc | such resolution is quite useless... imho |
| 07:04.59 | zenitraM | i know, i have my novo resting there without use.. that's why i didn't even try to get one |
| 07:14.54 | *** join/#arm-netbook hipboi_ (~hipboi@221.4.213.95) |
| 07:18.30 | mnemoc | hipboi_: ping |
| 07:18.42 | hipboi_ | pang |
| 07:18.46 | mnemoc | :) |
| 07:19.23 | mnemoc | hipboi_: do you use utf-8 locally? |
| 07:19.47 | hipboi_ | on ubuntu, yes |
| 07:20.31 | hipboi_ | my working pc is windows : ( |
| 07:20.36 | mnemoc | hipboi_: I made a reencoding branch in github (newer than the failed try where Turl found the problems) called wip/allwinner-v3.0-android-v2/utf-8 |
| 07:21.52 | mnemoc | hipboi_: before merging the changes we need to know the result is real chinese and not gibberish |
| 07:22.33 | hipboi_ | i think translating the Chinese comments to English comments is ok |
| 07:22.47 | mnemoc | hipboi_: do you have time to look at http://www.henriknordstrom.net/code/A10/utf-8.diff.txt ? |
| 07:23.20 | hipboi_ | yes? |
| 07:23.48 | mnemoc | translating them is a goal, but first I would like to solve the encoding problem :) |
| 07:24.00 | hipboi_ | ok |
| 07:24.31 | mnemoc | so the whole repo is either ascii or utf-8, and not 3 variants of GB and some Big5 :| |
| 07:26.15 | mnemoc | I'm starting to like how your glyphs look like :p |
| 07:27.09 | hipboi_ | well |
| 07:27.32 | hipboi_ | the Chinese characters comes from the shapes of things first |
| 07:28.18 | mnemoc | 检查传进的句柄的有效性 is beutiful.... but I can't find any shape in there :p |
| 07:28.34 | mnemoc | beautiful* |
| 07:29.28 | hipboi_ | well, you can't see any shape in these... |
| 07:30.36 | hipboi_ | years of chaning the characters has became not what they originally are |
| 07:30.55 | *** join/#arm-netbook Lerc (~Lerc@121-74-232-22.telstraclear.net) |
| 07:34.15 | hipboi_ | you can see the character of horse here of the changing |
| 07:34.18 | hipboi_ | http://zh.wikipedia.org/wiki/%E6%B1%89%E5%AD%97 |
| 07:34.26 | hipboi_ | 马 |
| 07:40.05 | *** join/#arm-netbook specing (~specing@unaffiliated/specing) |
| 07:42.22 | mnemoc | I suppose it's very hard to structure computing related sentences with characters modeled after objects available many thousand years ago |
| 08:06.57 | CapnKernel | can half-read it :-) |
| 08:07.13 | CapnKernel | hipboi_: 你好，你是哪国人？ |
| 08:08.21 | CapnKernel | Anyone who is interested in the mechanics behind Chinese characters will be interested in this: http://en.wikipedia.org/wiki/Chinese_character_classification |
| 08:08.40 | CapnKernel | hipboi_: 乒 |
| 08:09.22 | hipboi_ | it seems we are homemate : ) |
| 08:09.59 | CapnKernel | 什么？ |
| 08:10.07 | hipboi_ | i live in Zhuhai, China |
| 08:10.50 | mnemoc | you even use a variant of '?'... omg |
| 08:11.36 | CapnKernel | 我的老家是澳洲墨尔本，但是我平时住在深圳。 |
| 08:12.04 | CapnKernel | 品数如法就发了“？” |
| 08:12.12 | CapnKernel | 拼音输入法 |
| 08:12.49 | hipboi_ | yes, there are also variant of ！，。 |
| 08:13.04 | CapnKernel | 你是华人吗？我打了“乒”，你应该回答什么？;-) |
| 08:13.15 | hipboi_ | 乓 |
| 08:13.21 | CapnKernel | Haha, you pass the test :-) |
| 08:13.23 | hipboi_ | 囧 |
| 08:13.33 | CapnKernel | 哈哈 |
| 08:13.33 | mnemoc | hipboi_: to match the width and align properly? |
| 08:14.23 | hipboi_ | yes, full-width and half-width |
| 08:14.32 | CapnKernel | The western punctuation characters have equivalents in the chinese font, so a punctuation cell takes up the same room as a character cell. Otherwise a page would look terrible |
| 08:14.35 | zenitraM | understands the conversation decently through google translator. except.. <CapnKernel> hipboi_: Table Tennis |
| 08:14.52 | CapnKernel | What's another word for "table tennis"? |
| 08:15.01 | zenitraM | ping pong, i know :P |
| 08:15.57 | CapnKernel | 乒 is pronounced "ping" and 乓 is pronounced "pung". That's why hipboi_ replied "pang" to mnemoc when he asked "ping". |
| 08:16.21 | mnemoc | ^^ |
| 08:16.40 | CapnKernel | For example, if you are iterating over a group of items, you'll use 、as a separator, not ，. |
| 08:17.33 | mnemoc | zenitraM: http://mibbit.com/ (web based irc client) has built-in google translations |
| 08:18.38 | hipboi_ | mnemoc: you ar Mr.know-all |
| 08:18.53 | mnemoc | o_O |
| 08:19.19 | mnemoc | I don't know yet if http://www.henriknordstrom.net/code/A10/utf-8.diff.txt is properly reencoded :p and I really want to know :) |
| 08:19.36 | hipboi_ | yes, it is |
| 08:19.42 | mnemoc | \o/ |
| 08:19.55 | mnemoc | thanks :) I'll merge it them |
| 08:20.02 | CapnKernel | hipboi_: Is there any "nonsense" chinese in it? Stuff that should be utf-8 chinese but isn't? |
| 08:20.27 | mnemoc | holds |
| 08:20.52 | CapnKernel | Coz these westerners can only guess whether the funny looking symbols is really Chinese or not. |
| 08:21.30 | CapnKernel | （哈哈，我也是西方人） |
| 08:22.20 | hipboi_ | i see everything is alright in Chrome of that link |
| 08:22.29 | mnemoc | CapnKernel: are ausies "western"? |
| 08:23.04 | zenitraM | random google translator checks seem to make sense |
| 08:23.33 | mnemoc | thanks a lot, I'll merge it now :) |
| 08:23.59 | CapnKernel | Culturally yes. |
| 08:24.51 | CapnKernel | Although it's an immigrant nation, so most people have a second background from somewhere else, and that's ok. |
| 08:25.06 | zenitraM | related: http://xkcd.com/503/ |
| 08:25.08 | mnemoc | will have to watch Crocodile Dundee again and decide |
| 08:28.12 | mnemoc | zenitraM: and why is the north on top? |
| 08:28.42 | CapnKernel | http://www.satirewire.com/news/jan02/australia.shtml |
| 08:28.43 | mnemoc | (I'm from the south of chile) |
| 08:29.06 | CapnKernel | " AUSTRALIA GETS DRUNK, WAKES UP IN NORTH ATLANTIC: |
| 08:29.07 | CapnKernel | Tired of Being Isolated and Ignored, Continent Isn't Bloody Moving |
| 08:29.29 | mnemoc | :D |
| 08:35.14 | zenitraM | meh, i'm from the south-east of spain, so depending on the point of reference i'm either south-eastern or north-western |
| 09:36.01 | *** join/#arm-netbook techman2 (~gj@121.209.128.126) |
| 09:49.13 | *** join/#arm-netbook arokux_h (~roman@HSI-KBW-046-005-103-006.hsi8.kabel-badenwuerttemberg.de) |
| 10:01.02 | *** join/#arm-netbook Lerc (~Lerc@121-74-236-206.telstraclear.net) |
| 10:30.42 | *** join/#arm-netbook Lerc_ (~Lerc@121-74-236-206.telstraclear.net) |
| 12:19.29 | *** join/#arm-netbook rellla (~chatzilla@p5B0782B3.dip0.t-ipconnect.de) |
| 12:34.24 | *** join/#arm-netbook _hipboi_ (~hipboi@123.89.118.207) |
| 12:37.32 | CIA-91 | rhombus-tech: Tomasz master * rbd95b96f0c09 /allwinner_a10/hacking_the_mele_a1000.mdwn: |
| 13:17.50 | zenitraM | aand now we have the A2000... http://detail.tmall.com/item.htm?id=15051860754&prt=1329962623036 |
| 13:18.02 | zenitraM | ...and a cheaper A100 http://detail.tmall.com/item.htm?id=16964088540 |
| 13:19.26 | zenitraM | A100 seems to have no SATA |
| 13:20.39 | mnemoc | zenitraM: the A100 is bogus, a mistake on printing |
| 13:20.59 | mnemoc | zenitraM: hipboi_ gave a link to the press release from mele apologizing |
| 13:21.30 | specing | I cannot even imagine how a chinese keyboard looks like ... j/k |
| 13:22.54 | DonkeyHotei | mnemoc: what about the sata? |
| 13:23.05 | zenitraM | mnemoc: nope. it really exists |
| 13:23.14 | zenitraM | the mistake was some A2000 being labeled as A100 |
| 13:23.16 | zenitraM | see the link |
| 13:24.22 | CapnKernel | specing: in mainland china, keyboards are... QWERTY - ZOMG! |
| 13:24.48 | mnemoc | composing keys all day :p |
| 13:25.13 | specing | And selecting a character goes like this: CTRL-ALT-LSHIFT-RSHIT-FINGER-BREAK-key |
| 13:25.33 | DonkeyHotei | in mainland china, keyboards type on you (sorry, couldn't resist) |
| 13:28.25 | *** join/#arm-netbook j1nx (55df3ea3@gateway/web/freenode/ip.85.223.62.163) |
| 13:28.57 | mnemoc | DonkeyHotei: I haven't have time to debug the sata driver in v2 yet |
| 13:29.03 | Turl | don't be fooled, chinese don't use keyboards, they use speech recognition :P (j/k of course) |
| 13:30.53 | mnemoc | zenitraM: I still fail to understand why is the a2000 significatively more expensive than a1000 while it's just a redesigned case |
| 13:32.02 | mnemoc | well.. not that much more expensive, but still :p |
| 13:34.41 | DonkeyHotei | v2 of what? |
| 13:35.14 | mnemoc | allwinner-v3.0-android-v2 kernel branch, I assumed you were asking about that |
| 13:35.29 | DonkeyHotei | [Tue 2012-05-15 06:19:26 AM PDT] <zenitraM> A100 seems to have no SATA |
| 13:35.31 | DonkeyHotei | that |
| 13:35.38 | mnemoc | DonkeyHotei: ah, ok |
| 13:36.16 | mnemoc | DonkeyHotei: you catched me replying a mail about sata support :p |
| 13:36.40 | DonkeyHotei | ah |
| 13:40.44 | zenitraM | mnemoc: that still doesn't explain why the a2000 is more expensive than the *1000* if it even contains "less" features (as in, no HD box) |
| 13:41.34 | mnemoc | right, it should even be cheaper than the A1000 |
| 13:41.50 | DonkeyHotei | so the A2000 has no sata? |
| 13:42.11 | mnemoc | DonkeyHotei: A2000 is the same board than A1000 |
| 13:42.28 | mnemoc | DonkeyHotei: the only difference is that in one the uart connector is vertical and the other horizontal |
| 13:42.51 | DonkeyHotei | then what are the "less" features? |
| 13:43.16 | mnemoc | DonkeyHotei: the A1000 comes with a case for the sata disk |
| 13:43.19 | mnemoc | the A2000 doesn't |
| 13:43.38 | DonkeyHotei | so the bare drive sticks out? |
| 13:43.44 | mnemoc | right |
| 13:51.23 | zenitraM | unless you, well, buy a case. |
| 13:52.35 | _hipboi_ | A2000 is more expensive than A1000 |
| 13:52.42 | _hipboi_ | because it's 2000 |
| 13:53.02 | _hipboi_ | people usually think that bigger number is better |
| 13:53.10 | specing | I wonder what will happen when they go over 9000! |
| 13:55.31 | mnemoc | I wonder the same about sunxi and A1x, A10=sun4i, A12,A13=sun5i, .. no idea the chipnames of sun6i |
| 13:56.24 | _hipboi_ | sun6i is the coming soon one |
| 13:56.50 | zenitraM | "mmm, we have Cortex A8, Allwinner A10 or Allwinner A13?! definetly A13 is the best!" |
| 13:57.10 | mnemoc | _hipboi_: btw, can I buy a sun6i devkit on discount? :) |
| 13:59.41 | _hipboi_ | maybe |
| 13:59.57 | _hipboi_ | i can not talk much here about sun6i |
| 14:00.09 | _hipboi_ | but i will let you know when available |
| 14:00.27 | _hipboi_ | well |
| 14:00.34 | specing | hipboi_: you should be connecting to freenode through SASL |
| 14:00.47 | mnemoc | _hipboi_: /q ? |
| 14:01.10 | mnemoc | specing: ibot is logging and publishing |
| 14:01.34 | specing | Yes, but going through SASL and being clocked doesen't reveal who you are |
| 14:01.47 | specing | cloaked* |
| 14:02.18 | DonkeyHotei | yes it does |
| 14:02.30 | specing | It doesen't |
| 14:02.42 | specing | (Unless you explicitly state your identity) |
| 14:02.43 | DonkeyHotei | a cloak has your username |
| 14:02.59 | specing | Yes, but usernames can be arbitrary |
| 14:03.56 | DonkeyHotei | i don't think _hipboi_ is registered with NickServ anyway |
| 14:04.38 | j1nx | The news page from the ¨Allwinner EOMA68¨ is a bit quite. Is there nothing to mention? |
| 14:04.51 | mnemoc | j1nx: no news, good news :) |
| 14:04.51 | zenitraM | it's not that like would take much effort for them to guess who is that random nickname that appeared on the channel only hipboi frequented and started talking about a product of them |
| 14:05.19 | specing | True :) |
| 14:05.36 | j1nx | would appriciate a kind of intermediate status update, even if it states; Nothing happened |
| 14:06.13 | j1nx | just to keep the topic alive on the internet |
| 14:06.29 | j1nx | the more people pre-order, the cheaper they get ;) |
| 14:09.42 | _hipboi_ | need to go to bed now |
| 14:09.50 | _hipboi_ | see you later |
| 14:10.15 | CIA-91 | rhombus-tech: nil master * r7c17eb48d25e /allwinner_a10/a10_boot_process.mdwn: corrected u-boot chainload instructions (from elinux.org) |
| 14:10.16 | specing | j1nx: http://memegenerator.net/instance/20445890 |
| 14:16.11 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 14:23.42 | j1nx | Hey guys: the leaked kernel sources includes the limadriver and they are getting compiled by the build script. |
| 14:24.21 | DonkeyHotei | is the nda attached? |
| 14:25.19 | j1nx | oh, no not lima, just the ARM sources for the mali |
| 14:25.39 | DonkeyHotei | is the nda attached? |
| 14:26.22 | j1nx | the what? |
| 14:26.24 | mnemoc | the kernel sources are GPL |
| 14:26.36 | mnemoc | the mali userspace driver is closed source |
| 14:27.22 | j1nx | yes, I was confused when I saw the compile scrolling by |
| 14:28.44 | j1nx | jep; just mali.ko and ump.ko |
| 14:29.09 | j1nx | srry; carry on with whatever you were doing ;) |
| 14:29.17 | mnemoc | ^^ |
| 14:36.36 | steev_ | uh |
| 14:36.57 | *** join/#arm-netbook vectory_ (~damon@unaffiliated/vectory) |
| 14:46.42 | zenitraM | mmm, do the linaro ubuntu images have mali support? |
| 14:47.25 | ZaEarl | my guess would be no. |
| 14:47.39 | mnemoc | zenitraM: igloo does |
| 14:48.06 | mnemoc | zenitraM: the problem is the license, we are not legally allowed to use it... |
| 14:50.37 | mnemoc | http://www.igloocommunity.org/support/BSP_Releases |
| 14:50.51 | mnemoc | and I just noticed they are using 3.3, not 3.2.... |
| 14:50.53 | mnemoc | .oO |
| 14:51.44 | traeak | yeah, 3.3.x nicer |
| 14:51.45 | traeak | ugh |
| 14:52.17 | zenitraM | ic.. well, i might have a bored friend on a third-world country where the license doesn't apply with a mele that might try to see if it works. |
| 14:52.43 | mnemoc | zenitraM: :) |
| 14:53.53 | traeak | how much source is released with the igloo one? |
| 14:54.17 | mnemoc | the availability of armhf mali driver should imo define what version to port our v3.0 tree to, as stepping stone toward upstreaming |
| 14:54.50 | mnemoc | traeak: binary drivers are still binary, the rest is just ubuntu 12.04 armhf |
| 14:55.54 | traeak | mnemoc: that's gross....so 3.0.8 is still "gold" i guess? |
| 14:56.05 | traeak | trying to sift through the mailing list |
| 14:57.35 | traeak | lima irc is pretty active this morning...discussion over the vertex shaders |
| 15:09.18 | cromo | mnemoc: could you please put the script.bin somewhere with working ethernet? I'm trying to get thing running without serial console |
| 15:09.38 | cromo | also does the -v2 have a working hdmi framebuffer console by any chance? |
| 15:11.26 | zenitraM | i think that it doesn't |
| 15:12.50 | zenitraM | i'm being told that there's mali_drv.la and mali_drv.so at /usr/lib/xorg/modules/drivers |
| 15:15.50 | zenitraM | aand there is a mali400 deb on https://launchpad.net/~igloocommunity-maintainers/+archive/snowball |
| 15:20.41 | zenitraM | ehr... there are sources on that repo for the mali x11 driver |
| 15:21.29 | zenitraM | and they all have this header: http://pastie.org/3915710 |
| 15:21.52 | traeak | from arm, as-is, good to go right? |
| 15:23.03 | zenitraM | there's other one, exactly the same but by Instituto Nokia de Tecnologia for other file |
| 15:23.35 | zenitraM | and the only unclear one is on a file that says: |
| 15:23.36 | zenitraM | <PROTECTED> |
| 15:23.36 | zenitraM | <PROTECTED> |
| 15:23.36 | zenitraM | <PROTECTED> |
| 15:23.36 | zenitraM | <PROTECTED> |
| 15:23.38 | zenitraM | <PROTECTED> |
| 15:23.39 | zenitraM | <PROTECTED> |
| 15:23.41 | zenitraM | <PROTECTED> |
| 15:23.43 | zenitraM | <PROTECTED> |
| 15:23.47 | zenitraM | (just next to the ARM one I pasted) |
| 15:24.30 | traeak | ahh, yeah, igloo is st-ericsson |
| 15:25.05 | traeak | the binary components are the pieces that are licensed differently |
| 15:25.39 | zenitraM | which are those? |
| 15:25.56 | zenitraM | this is on xf86-video-mali package. there is another mali400 one but the xf86 doesn't depend on it |
| 15:26.23 | traeak | i haven't pulled these |
| 15:27.25 | zenitraM | mali400 contains libGLESv2.so, libMali.so and libEGL.so |
| 15:27.42 | traeak | yeah see those |
| 15:28.20 | zenitraM | but i'm not really sure if they're needed for xf86-video-mali |
| 15:29.14 | traeak | mali binary stuff license wise is locked to the snowball kit |
| 15:30.16 | traeak | in fact that mali package is just a couple of binary files and some headers with it |
| 15:30.20 | traeak | so nothing too exciting |
| 15:30.34 | zenitraM | mmm don't we also have the .so's for the mele? |
| 15:30.53 | zenitraM | with the Mali demo that came with the ubuntu image |
| 15:31.09 | zenitraM | with unclear/no license |
| 15:31.35 | zenitraM | and... would xf86-video-mali be also license-locked? (the snowball license "overriding" the ARM one) |
| 15:33.03 | traeak | looks like the binary driver (the UMP part) is required for the xorg driver implementation |
| 15:33.28 | traeak | unified memory provider |
| 15:33.41 | zenitraM | libUMP.so |
| 15:35.15 | zenitraM | ..that we might have in binary form from outside snowball on the ubuntu image on the mali demo that was left on the user home, i don't remember if it came with it. |
| 15:35.41 | traeak | Copied from ubuntu precise in Linaro Overlay PPA |
| 15:36.40 | traeak | that's from the mali binary package description part (web page) |
| 15:37.20 | traeak | does that mean that snowball copied this stuff from linaro, re-licensed it and re-released it ? |
| 15:38.01 | mnemoc | cromo: you mean you want a copy of the stock script.bin for the mele? |
| 15:38.43 | mnemoc | cromo: the wemac driver in v2 works, but ignores the mac given by the bootloader |
| 15:39.08 | zenitraM | mali400 is here too...: https://launchpad.net/~linaro-maintainers/+archive/overlay/ |
| 15:39.28 | zenitraM | as well as xf86-video-mali |
| 15:39.52 | *** join/#arm-netbook octet8 (~octet8@58.48.198.231) |
| 15:40.25 | *** join/#arm-netbook rellla_ (~chatzilla@p5B0782B3.dip0.t-ipconnect.de) |
| 15:41.32 | traeak | these are licensed to the snowball kit only regardless |
| 15:41.33 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 15:41.43 | traeak | for the binary stuff i mean |
| 15:41.47 | mikey_w | Has anyone here successfully booted ubuntu on their mele A1000? |
| 15:42.31 | traeak | yes but not with video out |
| 15:42.40 | mnemoc | cromo: -v2 fbcon I doubt, but the display drivers (lcd, hdmi) compile just fine |
| 15:42.51 | zenitraM | yup. binary has a license file for ST-Ericcson |
| 15:43.20 | *** join/#arm-netbook ZaEarl (~malmrose@173-11-125-162-SFBA.hfc.comcastbusiness.net) |
| 15:45.16 | zenitraM | but the x11 driver has no restriction... and it's strange that it doesn't depend on the package mali400 if it needed it |
| 15:45.26 | traeak | it seems to require the UMP driver |
| 15:49.11 | steev_ | traeak: snowball development/support is done via linaro, so the linaro license is the snowball license |
| 15:49.31 | zenitraM | UMP is a kernel device driver accompanied with a userspace library. |
| 15:51.24 | Turl | UMP is open source right? |
| 15:51.36 | mikey_w | I see links about puppy linux, cm9, ics but no useful sd images. Where are the usable sd images? |
| 15:52.31 | zenitraM | i think it is not. the "less illegal" way to obtain them would be the libUMP.so from android or the ubuntu image |
| 15:52.52 | Turl | zenitraM: I saw libUMP code on their OS release I think |
| 15:52.54 | Turl | let me recheck |
| 15:53.06 | zenitraM | on which OS release? |
| 15:54.05 | Turl | http://www.malideveloper.com/developer-resources/drivers/open-source-mali-gpus-ump-user-space-drivers-source-code.php |
| 15:55.26 | zenitraM | ...the x11 drivers are also there |
| 15:56.34 | zenitraM | so, we have UMP, we have the X11 driver opensourced.. is anything missing? |
| 15:56.52 | Turl | libMali isn't OS right? |
| 15:56.58 | Turl | nor is the GL stuff |
| 15:58.27 | zenitraM | they are on the ubuntu image |
| 15:58.33 | zenitraM | released with the: |
| 15:58.38 | zenitraM | "Mali binary openGL related lib are in /lib, they are libEGL.so libGLESv1_CM.so libGLESv2.so libMali.so" |
| 15:58.41 | zenitraM | license, by Tom. |
| 15:59.14 | Turl | how is that a license? :P |
| 15:59.41 | zenitraM | that's the only thing he's ever said about them, so that's the nearest thing we have to a license :P |
| 16:00.28 | Turl | so you basically have no license :) |
| 16:01.54 | zenitraM | that makes trying to get x11 to work with those files, while not really opensource yet, at least unknowingly illegal |
| 16:02.45 | Turl | I haven't seen any android vendor sue over for using their GL libs though :) |
| 16:08.50 | zenitraM | and allwinner is definetly not going to be the first. |
| 16:09.25 | Turl | I'm actually surprised this vendor opensourced all what they did |
| 16:09.36 | Turl | they open sourced pretty much all the things |
| 16:09.43 | CapnKernel | Was the vendor entitled to open source it? |
| 16:10.00 | Turl | CapnKernel: I suppose so |
| 16:10.13 | Turl | otherwise they wouldn't have I guess |
| 16:10.13 | CapnKernel | Not the first time a vendor released stuff they shouldn't have. |
| 16:10.23 | CapnKernel | That logic doesn't follow. |
| 16:10.27 | Turl | CapnKernel: they haven't pulled it yet |
| 16:10.44 | CapnKernel | For example, TI ADSL modems. ATM module. Closed source, yet one of the modem vendors released the source. |
| 16:10.52 | ZaEarl | plenty of that stuff is GPL licensed, so they were supposed to release it |
| 16:11.21 | ZaEarl | it=the gpl licensed bits |
| 16:11.32 | CapnKernel | I'd like to believe it's all legitimate open source, but it's possible that the vendor slipped up on one or more files. |
| 16:11.41 | CapnKernel | I've seen this happen. |
| 16:11.47 | Turl | they open sourced all of the apache stuff too |
| 16:11.53 | Turl | most vendors don't |
| 16:11.59 | CapnKernel | One can't assume it's open source unless it explicitly says it in the header. |
| 16:12.05 | Turl | motorola does for a couple of stuffs |
| 16:12.15 | CapnKernel | Does one open source apache stuff? (puzzled) |
| 16:12.20 | zenitraM | they aren't really even a vendor, as in, manufacturer |
| 16:12.32 | zenitraM | if i'm not wrong they only distribute Ployer tablets |
| 16:12.35 | Turl | vendor as in sells stuff |
| 16:12.38 | ZaEarl | the bulk of the code released was apache and gpl licensed |
| 16:12.54 | CapnKernel | I'm just saying that it's wise not to assume that we have any particular rights to any particular piece of code |
| 16:12.57 | *** join/#arm-netbook lundman (~Accident@194.109.206.190) |
| 16:12.57 | *** join/#arm-netbook mnemoc (~amery@shell.opensde.net) |
| 16:13.19 | CapnKernel | You only need one source file to be unintentionally released, to cause a major stink. |
| 16:13.30 | Turl | CapnKernel: I've seen some code that has headers like copyright allwinner, to be used as per licensed or stuff like that too |
| 16:13.42 | CapnKernel | That doesn't make it open source |
| 16:13.51 | Turl | that isn't a license |
| 16:13.57 | CapnKernel | Also, they may have used code from someone else, like the core vendor. |
| 16:14.32 | Turl | CapnKernel: on the other hand, I've seen GPL code from nokia+samsung |
| 16:14.35 | Turl | blended with apache one |
| 16:14.37 | CapnKernel | The provenance of source code is important |
| 16:14.45 | Turl | and this "copyright" |
| 16:14.50 | Turl | so, not like they care much :) |
| 16:15.25 | zenitraM | like the "do whatever you want with it" license from tom for the video decoding libs |
| 16:15.51 | Turl | does tom have any right to make such a statement? |
| 16:16.11 | Turl | I can also say, do whatever you want with the windows source code and give you a copy |
| 16:16.18 | Turl | but that doesn't make it true :P |
| 16:16.32 | CapnKernel | PRECISELY |
| 16:17.04 | CapnKernel | Which goes back to my original question: Was the vendor entitled to open source it? |
| 16:17.11 | Turl | CapnKernel: but MS would sue me to death :) allwinner won't |
| 16:17.30 | zenitraM | we have no way of knowing it. |
| 16:17.31 | CapnKernel | Whether or not someone would sue you is independent of whether you're entitled to use it. |
| 16:17.32 | zenitraM | but that's the vendor fault |
| 16:17.46 | CapnKernel | It doesn't matter whether they'd sue you or not. |
| 16:17.57 | Turl | they'd need to sue the vendor anyway, not me :P |
| 16:18.15 | Turl | CapnKernel: and yeah, I agree, but for the purpose of hacking on the device, I wouldn't worry |
| 16:18.19 | CapnKernel | They can sue anyone who distributes it. |
| 16:18.27 | CapnKernel | Just be careful, is all. |
| 16:18.52 | CapnKernel | One of the criticisms of the GPL (by corporate types who don't "get it"), is that it's viral. |
| 16:19.02 | CapnKernel | Well for us, proprietary code is also viral. |
| 16:19.52 | CapnKernel | Even if the vendor gave it to you, unless it explicitly says open source, there's a non-zero chance that even looking at it will lead to future claims of stolen intellectual property. |
| 16:20.22 | Turl | CapnKernel: between using a prebuilt found on your device, and building your own from this source; and distributing it, imo, doesn't make any difference on the risk |
| 16:20.52 | CapnKernel | SCO accused Linus of stealing errno.h. How pathetic is that? |
| 16:22.07 | Turl | fuu customs :P they got my tab still :( |
| 16:22.36 | CapnKernel | You're pretty safe if you build it yourself, for your own use. However, if you were to distribute it, or if it came down to a question of what you knew ("Your honour, Turl couldn't have written the driver without looking at proprietary code, and we know he had access to proprietary code"), you could be dragged into a lawsuit concerning lost profits. |
| 16:23.08 | CapnKernel | Proprietary code is the new viral. |
| 16:23.11 | CapnKernel | Just don't look at it. |
| 16:23.27 | Turl | I agree, but my point remains |
| 16:23.43 | Turl | besides, if anything, it'll increase their profits |
| 16:24.01 | CapnKernel | Well as it may, that's not going to work in court. You're assuming you're dealing with rational people. |
| 16:24.22 | *** join/#arm-netbook specing (~specing@unaffiliated/specing) |
| 16:25.00 | Turl | they'd need to extradition me to judge me too right? :) |
| 16:25.31 | CapnKernel | Who knows, do you really want to find out? |
| 16:25.48 | CapnKernel | And besides, you could put a whole project under a cloud. |
| 16:26.32 | CapnKernel | Someone takes that "free" code, ships it in a product, 100s of 1000s get shipped, then the IP owner can claim that number of instances of distribution of proprietary knowledge. |
| 16:26.54 | CapnKernel | You just don't fuck with these people. |
| 16:27.25 | CapnKernel | Believe me, I've been involved in several court cases, some involving IP issues and Linux. |
| 16:27.58 | Turl | any involving chinese vendors? |
| 16:28.43 | CapnKernel | I) I can't talk about it and ii) whether it did or it didn't, it doesn't change what I'm saying now. Just leave proprietary code alone. |
| 16:28.48 | j1nx | * * This confidential and proprietary software may be used only as * authorised by a licensing agreement from ARM Limited * (C) COPYRIGHT 2008-2010 ARM Limited * ALL RIGHTS RESERVED * The entire notice above must be reproduced on all authorised * copies and copies may only be made to the extent permitted * by a licensing agreement from ARM Limited. */ |
| 16:29.25 | CapnKernel | If it's something you must have, speak to whoever you got it from. If it's theirs to give, they'll be happy to give you something saying "yep it's open source, and we're entitled to claim that". |
| 16:29.34 | j1nx | ^ Files with that license are within the kernel src from QWare |
| 16:29.48 | Turl | CapnKernel: I understand your point, just wondering |
| 16:29.54 | CapnKernel | Which is the scariest thing I've seen all day :-( |
| 16:29.54 | j1nx | Are we allowed to use the ARM files |
| 16:29.56 | CapnKernel | Ticking time bomb |
| 16:30.09 | CapnKernel | With that message on it? Absolutely not. |
| 16:30.09 | Turl | j1nx: if it's on the kernel and on a shipping device, it must be GPL |
| 16:30.13 | CapnKernel | NO |
| 16:30.37 | CapnKernel | There are several examples where proprietary code has crept into the kernel, and it has been a nightmare to go through the process of excising it. |
| 16:30.56 | CapnKernel | That's one reason the kernel maintainers now require that EVERY SINGLE FILE has a GPL declaration at the top. |
| 16:30.56 | Turl | CapnKernel: it's illegal to ship a device with proprietary code on the kernel |
| 16:31.05 | Turl | because you need to ship the code too |
| 16:31.07 | CapnKernel | Not necessarily. |
| 16:31.13 | CapnKernel | No that's not true. |
| 16:31.20 | CapnKernel | Why is it illegal? |
| 16:31.53 | j1nx | It is not within the kernel, there are seperate folders containing the ARM source code downloaded from here; |
| 16:31.54 | j1nx | http://www.malideveloper.com/developer-resources/drivers/open-source-mali-gpus-linux-kernel-device-drivers.php |
| 16:32.22 | j1nx | There is a license on that page which I have not seen/accepted (or whatever) |
| 16:32.25 | Turl | CapnKernel: b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. |
| 16:32.54 | Turl | j1nx: that's a GPL license |
| 16:33.29 | zenitraM | "used only as * authorised by a licensing agreement from ARM Limited" |
| 16:33.46 | zenitraM | that licensing agreement happens to be the GPL license in this case. |
| 16:35.11 | j1nx | OK, because I am in progress of updating the included Mali drivers with the new release from the 8th |
| 16:36.15 | j1nx | BTW in the new release there is now also a; x11 folder with a mali_drm module? |
| 16:36.18 | CapnKernel | Turl: Not necessarily. Here's a counter-example. Let's say I have a proprietary library. I can make my GPL code call that library. My library is not derived from the GPLed code. I only have to distribute the library if it is a derived work. |
| 16:37.32 | j1nx | the mali drm module needs a kernel reference? But what is it for? |
| 16:38.24 | CapnKernel | Is there a file that has that "used only as authorised by a licensing agreement from ARM Limited" text in it, which is explicitly open source? (It's not clear from the info you've given) |
| 16:38.27 | mnemoc | userspace drivers can't access hardware directly |
| 16:38.49 | mnemoc | the kernel driver provides them with mmap and ioctrl interfaces to the hardware |
| 16:39.19 | mnemoc | but the kernel driver doesn't know to use the hardware, only to allow the userspace driver the capability to reach it |
| 16:40.03 | Turl | CapnKernel: yeah, but you cannot build your app statically as it's a derived work then |
| 16:40.14 | Turl | would be equivalent to built in kernel code |
| 16:40.16 | mnemoc | with the a10 we have two of those, mali (GPU) and cedarx (VPU) |
| 16:40.44 | Turl | mnemoc: most cedarx code got released |
| 16:40.50 | mnemoc | Turl: ? |
| 16:41.00 | Turl | mnemoc: the userspace code |
| 16:41.32 | mnemoc | Turl: I've only seen client code using binary libraries in a 1MB .a |
| 16:41.41 | traeak | unfortunately i think the binary side is the one that has to have licenses for h264, mpeg4, mpeg2, etc |
| 16:42.07 | Turl | mnemoc: there's all the android bindings from what I was able to skim over it |
| 16:42.13 | Turl | but yeah it has .a files on there |
| 16:42.35 | mnemoc | the .a is the real library, the rest is example application code to use the library |
| 16:43.56 | traeak | in the x86 world we just software decode everything (and bypass licensing) |
| 16:44.13 | Turl | traeak: good luck doing so on the A10 :) |
| 16:44.54 | traeak | Turl: yeah, i noticed it's not a real powerhouse, not even close to my atom netbook |
| 16:45.40 | Turl | it's an A8 right? |
| 16:46.08 | Turl | must be around the same computing power as a Nexus One |
| 16:46.23 | Turl | https://github.com/Quarx2k/android_external_cedarx/blob/b1e9e102fd8af0d179c1293e0270b708f0194be8/CedarA/CedarAPlayer.cpp the fact that we get this kind of stuff is nice |
| 16:46.31 | Turl | (don't worry, it's Apache licensed) |
| 16:47.56 | zenitraM | -- This software is confidential and proprietary and may be used -- -- only as expressly authorized by a licensing agreement from -- -- Softwinner Products. -- |
| 17:23.41 | *** part/#arm-netbook QingPei (~qingpei@117.81.195.4) |
| 17:49.29 | CIA-91 | rhombus-tech: Giampaolo master * rbdb731044551 /allwinner_a10/orders/Giampy88na.mdwn: |
| 18:04.07 | zenitraM | in a probably not good idea dist-upgrades from lucid to precise the ubuntu image |
| 18:17.42 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 18:26.20 | *** join/#arm-netbook arokux_h (~roman@HSI-KBW-046-005-103-006.hsi8.kabel-badenwuerttemberg.de) |
| 19:00.58 | *** join/#arm-netbook ibot (~ibot@rikers.org) |
| 19:00.58 | *** topic/#arm-netbook is EOMA: Embedded Open Modular Architecture - Don't ask to ask. Just ask! - http://elinux.org/Embedded_Open_Modular_Architecture/EOMA-68 - ML arm-netbook@lists.phcomp.co.uk - Logs http://ibot.rikers.org/%23arm-netbook - http://rhombus-tech.net/ |
| 19:01.42 | cromo | mnemoc: you there? is the v2 preferred now over old 3.0 branch? |
| 19:06.25 | vgrade | we don't have all the bits of the gpu puzzle until we can get libmali.so built for libc and not andriod bionic |
| 19:09.33 | mnemoc | meh |
| 19:10.04 | mnemoc | vgrade: that's where linaro for igloo help |
| 19:10.05 | mnemoc | s |
| 19:14.11 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 19:14.37 | cromo | vgrade: sorry, got disconnected. I am running headless currently, so not interested in mali yet |
| 19:14.54 | zenitraM | vgrade, mnemoc: libmali.so is also built for libc on hipboi's ubuntu image. |
| 19:17.05 | zenitraM | less license breaking |
| 19:17.58 | zenitraM | and all libraries are also there: |
| 19:17.59 | zenitraM | root@ubuntu:/lib# ls -l |grep ubuntu |
| 19:17.59 | zenitraM | -rwxr-xr-x 1 ubuntu ubuntu 121241 Mar 6 11:10 libEGL.so |
| 19:17.59 | zenitraM | -rwxr-xr-x 1 ubuntu ubuntu 287334 Mar 6 11:10 libGLESv1_CM.so |
| 19:17.59 | zenitraM | -rwxr-xr-x 1 ubuntu ubuntu 239682 Mar 6 11:10 libGLESv2.so |
| 19:18.00 | zenitraM | -rwxr-xr-x 1 ubuntu ubuntu 772639 Mar 6 11:10 libMali.so |
| 19:18.01 | zenitraM | -rwxr-xr-x 1 ubuntu ubuntu 30620 Mar 6 11:10 libUMP.so |
| 19:19.38 | mnemoc | so for softfp we are done? |
| 19:19.47 | zenitraM | seems so. |
| 19:19.55 | mnemoc | great^3 |
| 19:20.12 | zenitraM | snowball ones are for bloat softfp and hardfp |
| 19:20.36 | zenitraM | s/bloat // |
| 19:20.48 | mnemoc | \o/ |
| 19:21.47 | mnemoc | btw, can anyone not-me reply Tom's mail saying the sourcedrop closes the chapter about officially releasing 3.0.8? |
| 19:22.10 | mnemoc | for mainlining we need a proper release, not a leak |
| 19:22.35 | mnemoc | and a mail to reference telling everything there is GPL |
| 19:26.27 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 19:26.47 | mikey_w | Does anyone know the passwords need to log into lucid ubuntu? |
| 19:27.17 | zenitraM | other user -> root, pass ubuntu |
| 19:27.26 | mnemoc | cromo: that doesn't have yet an absolute answer yet, because v1 has been more tested and has sata and better ethernet than v2 |
| 19:28.09 | mnemoc | cromo: but v2 is preferred for development because that branch is expected to deprecate the other soon |
| 19:29.09 | mnemoc | cromo: it's better than v1 in everything except sata and ethernet, so v2 will be improved and v1 abandoned |
| 19:29.43 | mikey_w | That works. Thanks. |
| 19:30.13 | j1nx | guys, a quick question as all allwinner files are not how they should |
| 19:30.20 | j1nx | Notes on integrating the Mali DRM module: The Mali DRM is a platform device, meaning that you have to add an entry for it in your kernel architecture speci fication. Example: (arch/arm/mach-<platform>/mach-<platform>.c) #ifdef CONFIG_DRM_MALI static struct platform_device <platform>_device_mali_drm = { .name = "mali_drm", .id = -1, }; #endif static struct platform_device *<platform>_devices[] __initdata = { .. |
| 19:31.11 | j1nx | There is no arch/arm/mach-sun4i/mach-sun4i.c file right? |
| 19:31.38 | mnemoc | j1nx: feel free to file an issue in https://github.com/amery/linux-allwinner |
| 19:31.49 | j1nx | Where can I include the above at the moment |
| 19:32.12 | mnemoc | there is a tab that says "Issues" |
| 19:32.37 | vgrade | zenitraM, how do you know those libs are libc and not bionic? |
| 19:32.42 | mnemoc | where you can report the issue |
| 19:32.46 | j1nx | Oh I will, but there are alreadu sun4i kernel achtitecture parameters set, so where are they |
| 19:33.23 | mnemoc | core.c |
| 19:33.32 | zenitraM | vgrade: they are used by the mali demo |
| 19:33.38 | zenitraM | that was left on /home/ubuntu |
| 19:33.47 | mnemoc | there are only 3 .c files in mach-sun*i, not that hard to guess |
| 19:33.56 | vgrade | mnemoc, it is an offical release from the nl crowd |
| 19:34.25 | mnemoc | vgrade: the allwinner-v3.0-android-v2 branch is based on qware's source drop, yes |
| 19:34.52 | j1nx | DOH! |
| 19:35.19 | vgrade | zenitraM, ok, I did not try and run that demo, is that a fullscreen demo |
| 19:36.04 | mnemoc | vgrade: but qware doesn't own the copyright of any file there |
| 19:36.04 | zenitraM | nope, it only shows on a part of the screen |
| 19:36.24 | mnemoc | vgrade: so it's not enough for been accepted upstream |
| 19:36.29 | vgrade | in a properly decorated X window? |
| 19:36.40 | zenitraM | on top of the framebuffer |
| 19:37.16 | vgrade | zenitraM, nod |
| 19:39.52 | zenitraM | has a mess of SD cards so he might accidentaly end up putting the one with the snowball rootfs on the mele. |
| 19:39.54 | hno | nor sure if the libs in hipbois image is built for Linux glibc or Android. |
| 19:40.09 | vgrade | zenitraM, happens |
| 19:40.23 | zenitraM | i have them here. any way to check? |
| 19:40.35 | vgrade | dump the symbols |
| 19:40.38 | zenitraM | how? |
| 19:41.12 | vgrade | nm -g |
| 19:41.29 | zenitraM | <PROTECTED> |
| 19:41.32 | zenitraM | and GLIBC everywhere |
| 19:41.53 | vgrade | zenitraM, nice |
| 19:41.55 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 19:42.15 | vgrade | so looks like we have all the bits |
| 19:42.35 | zenitraM | X11 should be easy, if it doesn't work already |
| 19:43.20 | vgrade | its just where these things came from |
| 19:43.38 | zenitraM | the x11-video-mali driver on linaro's PPA is under the snowball license but compiled from the GPL sources from ARM |
| 19:44.00 | vgrade | the X11 stuff on arm mali site is where to start |
| 19:44.03 | zenitraM | the .so has the tom's "heh, those files were forgotten there" license |
| 19:44.17 | zenitraM | and the kernel mali driver is gpl |
| 19:44.22 | vgrade | yup |
| 19:45.01 | MaDMaLKaV | just got home. what have I missed? |
| 19:45.06 | DonkeyHotei | if linaro has a ppa for userspace, and the kernel driver is gpl, what's missing that's actually needed? |
| 19:45.35 | MaDMaLKaV | mnemoc , as I understand it, Allwinner only have to offer the sources to their customers, the tablet seller is the one responisble for provinding the gpl code to their customers |
| 19:45.38 | MaDMaLKaV | is somewhat a chain |
| 19:45.41 | zenitraM | MaDMaLKaV: we just figured out we have all the pieces for X11 GPU accelerated |
| 19:45.49 | vgrade | nothing now just a bit of magic |
| 19:45.56 | MaDMaLKaV | zenitraM , fucking cool, yeah |
| 19:46.16 | MaDMaLKaV | where did be got the libc compiled files? |
| 19:46.28 | zenitraM | they are on the ubuntu image uploaded by tom |
| 19:46.45 | DonkeyHotei | zenitraM: so nothing is missing? |
| 19:46.45 | DonkeyHotei | [Tue 2012-05-15 12:45:06 PM PDT] <DonkeyHotei> if linaro has a ppa for userspace, and the kernel driver is gpl, what's missing that's actually needed? |
| 19:46.58 | zenitraM | DonkeyHotei: nothing except licenses. |
| 19:47.15 | zenitraM | the packages on linaro ppa have their use "forbidden" on devices outside the snowball platform |
| 19:47.24 | DonkeyHotei | ahhh |
| 19:47.31 | zenitraM | (and the libraries they have are the ones we have on the ubuntu image) |
| 19:47.41 | DonkeyHotei | maybe linaro can negotiate more? |
| 19:48.04 | zenitraM | we don't really need them |
| 19:49.30 | MaDMaLKaV | not hf thought, right? |
| 19:49.46 | MaDMaLKaV | I think that last sentence doesn't make any sense... too tired to write in english :/ |
| 19:50.00 | zenitraM | the hf ones are on that ppa |
| 19:50.04 | zenitraM | but behind that license |
| 19:52.30 | MaDMaLKaV | and the tom ones are softfloat and we can use it, am I right? |
| 19:53.45 | zenitraM | yes |
| 19:54.01 | zenitraM | well, we can use them both. but it's a matter of.. you know, legality. |
| 19:55.00 | MaDMaLKaV | yeah, when I mean "we can use", I want to mean "we can legally use" |
| 19:55.31 | MaDMaLKaV | if I even suggest to use the files we can't use, I will say it in a suspicious manner, probably hiding behind a newspaper or somethin' like that, spy type |
| 19:56.45 | MaDMaLKaV | ok, so we have it for softfloat -will be ok until we can get the hardfloat ones legally somehow-, and we miss the licensing on the vpu stuff |
| 19:57.05 | zenitraM | we don't *really* have it for softfloat |
| 19:57.18 | zenitraM | as in: the files are in the "Huh, i forgot those files there" license from Tom. |
| 19:57.26 | DonkeyHotei | is there something complete that can be legally used? if the a10 if hf, it can use either |
| 19:57.41 | zenitraM | probably nothing will happen by using them (half of Android community does modding with GL libs) |
| 19:57.48 | zenitraM | but.. you know. |
| 19:58.23 | zenitraM | what was the wifi module for the mele? |
| 19:58.29 | mikey_w | How do I gain the internet with the ludid ubuntu demo. |
| 19:58.45 | zenitraM | depmod -a, reboot, you'll have wifi |
| 19:58.50 | MaDMaLKaV | ARGH |
| 19:59.00 | mikey_w | ok |
| 19:59.08 | MaDMaLKaV | why I couldn't go the easy way and just get a pandaboard? |
| 20:00.15 | mikey_w | Great. |
| 20:02.05 | zenitraM | okay, snowball image with x11 mali bits disabled boots.. but no gnome-panel |
| 20:02.10 | zenitraM | more news after dinner. |
| 20:02.36 | MaDMaLKaV | dinner, good idea |
| 20:02.52 | MaDMaLKaV | why you disabled the x11 mali bits on the snowball image? |
| 20:03.30 | MaDMaLKaV | in fact, what I really want to ask is, does the mali hf bits in snowball image works ok? I'm not going to make an image based on that, of course, it is just curiosity coff coff |
| 20:04.01 | traeak | MaDMaLKaV: gpu is still totally closed on panda ... and you enjoy the cutting edge stuff, rihgt? |
| 20:05.24 | mnemoc | MaDMaLKaV: I know Allwinner is not obligated with us, but Tom was pushing marketing people to release the 3.0.8... but after the notification of the source drop he replied the list telling that thanks to this we didn't need the official release anymore |
| 20:05.48 | mnemoc | MaDMaLKaV: but we still do |
| 20:07.25 | MaDMaLKaV | yeah |
| 20:07.54 | MaDMaLKaV | it will be nice to see that kind of move by their part, they are not really being as supportive as they seemed to me initially |
| 20:07.59 | MaDMaLKaV | well, gonna dinner, seeya |
| 20:31.20 | zenitraM | MaDMaLKaV: the snowball image is sf not hf |
| 20:31.35 | zenitraM | disabled mali just in case it didn't work, as i don't know where is my serial dongle |
| 20:31.42 | zenitraM | so i could get a shell to enable ssh |
| 20:34.05 | mnemoc | zenitraM: 12.04 is supposed to be hf... |
| 20:36.55 | zenitraM | mmm how can i tell from inside? |
| 20:37.09 | zenitraM | uname says: armv7l armv7l armv7l |
| 20:41.16 | vgrade | readelf -A *.so |
| 20:41.54 | vgrade | looking for "Tag_ABI_VFP_args: VFP registers" |
| 20:41.58 | vgrade | for hardfp |
| 20:42.58 | zenitraM | yup. it's hardfp then |
| 20:43.25 | zenitraM | idk where i'd read armel instead of armhf |
| 20:45.15 | zenitraM | FATAL: Module mali_drm not found. |
| 20:45.15 | zenitraM | (EE) MALI(0): mali_drm_open_master Unable to open DRM: No such file or directory |
| 20:45.16 | zenitraM | (EE) MALI(0): Failed to be master of DRM! |
| 20:45.58 | vgrade | so you need the drm code in the kernel |
| 20:46.00 | zenitraM | mmm wasn't there a drm.ko on 2.6.36? |
| 20:46.19 | zenitraM | i don't have that module here on 3.0.8+ |
| 20:46.33 | vgrade | its not a module |
| 20:48.57 | vgrade | http://www.igloocommunity.org/gitweb/?p=kernel/igloo-kernel.git;a=blob;f=drivers/gpu/mali/mali400ko/x11/mali_drm/README.txt;h=47009fe7686bf63794d82e301e591dd7d1fe5709;hb=refs/heads/integration-ubuntu-ux500 |
| 20:49.29 | *** join/#arm-netbook cromo (~cromo@176-232-251-94.net.stream.pl) |
| 20:50.51 | vgrade | as I said needs a bit of magic |
| 20:55.25 | mnemoc | wants an elegant drivers/gpu/mali/mali400ko instead of a modules/mali/modules/mali/DX910-SW-99002-r2p3-01rel0 |
| 21:00.54 | j1nx | I am half way to have the mali.ko ump.ko and mali_drm from the 8th of may release compiled together with the 3.0.31 kernel |
| 21:01.51 | vgrade | just a warning you need to match the mali.ko with the userspace libs as there is a version check |
| 21:02.57 | mnemoc | vgrade: then, what version do we need to use igloo's driver? |
| 21:04.56 | mnemoc | j1nx: what tree/branch are you working on? |
| 21:05.09 | j1nx | 3-v2 |
| 21:05.28 | mnemoc | \o/ |
| 21:05.54 | j1nx | updated the mali device driver and ump driver from ARM (8th release) |
| 21:06.22 | j1nx | and am now in progess to have the DRI support added and teh mali_drm module compiled |
| 21:07.11 | mnemoc | j1nx: please diff the real DX910-SW-99002-r2p3-01rel0 with ours as allwinner may have modified it |
| 21:07.23 | j1nx | but i don know shit about DRI, so it goes sssllllooooooww |
| 21:07.33 | mnemoc | and we will need to do the same changes to any new version |
| 21:07.51 | j1nx | mnemoc: did not think of with, but doubt they changed something |
| 21:08.10 | j1nx | They included the cofiguration files for the hardware, but i think thatś it |
| 21:09.03 | mnemoc | mali was modified 2 times after the update to r2p3 in the source drop |
| 21:09.30 | mnemoc | but you have looked deeper than i |
| 21:09.32 | j1nx | I will look if I can download the old r2p3-01rel0 aswell |
| 21:10.40 | mnemoc | git diff e098dbde33a1edaf76254fb03470f2fe13bfca4c -- modules/mali/ shows some evilness |
| 21:11.00 | *** join/#arm-netbook arokux_h (~roman@HSI-KBW-046-005-103-006.hsi8.kabel-badenwuerttemberg.de) |
| 21:12.40 | vgrade | mnemoc, im not sure of the versions we have in userspace |
| 21:13.08 | vgrade | dmesg should show the version when the gles demos are ran |
| 21:13.21 | vgrade | zenitraM, can you check the dmesg |
| 21:13.48 | mnemoc | nice to have a mali-team now forming up :p |
| 21:15.08 | vgrade | the api version is found in the mali.ko code in this header, http://www.igloocommunity.org/gitweb/?p=kernel/igloo-kernel.git;a=blob;f=drivers/gpu/mali/mali400ko/driver/src/devicedrv/mali/common/mali_uk_types.h;h=9a1583658c9b08c3faecd2fd81ea1c29b79fb3c3;hb=refs/heads/integration-ubuntu-ux500#l736 |
| 21:15.46 | vgrade | I've been working this on the vivaldi(zenithnk) tablet so know my way around |
| 21:16.22 | vgrade | thing is I don't have a userspace which has an x11 backend |
| 21:17.23 | vgrade | version 10 is the latest API version I've seen |
| 21:17.48 | vgrade | each version relates to one of the r1p2p3 drops from ARM |
| 21:19.32 | *** join/#arm-netbook burndtjamb (~rapefruit@CPE-72-131-120-166.wi.res.rr.com) |
| 21:31.24 | mnemoc | wonders if the xi in sunxi (platform name for sun3i, sun4i, sun5i, sun6i, ...) can be pronounced chi to match the meaning of Qi.... |
| 21:32.05 | mnemoc | sun-chi instead of sun-x-i |
| 21:46.22 | j1nx | hmm drivers/gpu/drm/mali/mali_drv.o does not compile |
| 21:46.48 | j1nx | pffff, to late now. Will continue tomorrow |
| 21:47.28 | j1nx | have mali.ko, ump.ko, drm.ko, now only need mali_drm.ko |
| 21:47.38 | j1nx | then on to the X11 drivers |
| 21:48.53 | mnemoc | j1nx: before the X11 drivers, the patch to make a wip branch for this :) |
| 21:49.06 | mnemoc | so others can enjoy/test it too |
| 21:49.33 | mnemoc | a git fork is ok too |
| 21:50.40 | j1nx | yes offcourse, first the patch |
| 21:51.00 | j1nx | but we still need the closed source Mali DDK |
| 21:51.33 | j1nx | But hopefully Tom can leak/release it as Allwinner is the only one who has the ARM license agreement |
| 21:52.59 | mnemoc | this crap with the license agreements to use a driver for which you already paid the royalties by buying the device/chip seriously pisses me off |
| 21:53.48 | j1nx | Me too, hopefully LIMA is filling up that part rather sooner then later |
| 21:56.17 | mnemoc | I can understand they don't want to open the source of the drivers, even if that driver only works on hardware that is already paying them royalty. but restricting on what devices people can actually use the driver is beyond absurdity |
| 22:02.49 | j1nx | I think the X11/xf86-video-mali driver also needs a refernece to the mali ddk |
| 22:14.04 | *** join/#arm-netbook A0Sheds (~ly@unaffiliated/l84supper) |
| 22:36.46 | arokux_h | so there is a part of mali which sits in the kernel and is open and a closed userspace library? |
| 22:55.54 | hno | hipboi_, why is the size of the first 512MB memory bank reduced by 64MB in the kernel magic memory detection? |
| 22:57.31 | hno | By that I mean mi->bank[0].size = SZ_1M * (size - 64); in arch/arm/mach-sun4i/core.c |
| 22:58.46 | *** join/#arm-netbook mpthompson (~IceChat77@c-50-131-39-81.hsd1.ca.comcast.net) |
| 22:59.34 | hno | arokux, yes. |
| 22:59.59 | hno | actually a large part of the userspace is open as well, just not the 3D parts. |
| 23:15.04 | *** join/#arm-netbook arokux_h (~roman@HSI-KBW-046-005-103-006.hsi8.kabel-badenwuerttemberg.de) |
| 23:16.36 | *** join/#arm-netbook hp_|2 (~kvirc@220-136-72-229.dynamic.hinet.net) |
| 23:36.23 | hp_ | this new cotton candy looks really greate: http://blog.laptopmag.com/fxi-cotton-candy-usb-computer-to-start-shipping-this-month |
| 23:49.31 | ZaEarl | cute, but $200? |
| 23:49.43 | ManoftheSea | "somewhere under $200"? |
| 23:49.58 | ZaEarl | yeah, that means 199.99 |
| 23:50.30 | DonkeyHotei | and nine tenths |
| 23:51.11 | ManoftheSea | I mean, yeah, it's nice that it has sd, hdmi, wifi, bt, and usb... |
| 23:51.23 | ManoftheSea | and I can lose it. |
| 23:53.56 | hp_ | but its pretty awsome.. although yeah $200 |
| 23:54.19 | hp_ | prolly the allwinner will beat that for sure |
| 23:55.44 | ManoftheSea | "the allwinner"? |
| 23:55.52 | ManoftheSea | If it really is the all-winner, it would have to win. |
| 23:56.06 | ZaEarl | allwinner always wins |
| 23:56.40 | hp_ | :) |
| 23:56.53 | DonkeyHotei | it's full of allwin |
| 23:56.59 | DonkeyHotei | dows |
| 23:57.09 | hno | The cotton candy is a fair bit more powerful than A10 however. |
| 23:57.32 | hp_ | yep with the flagship SOC of Samsung in its belly.. |
| 23:57.42 | ManoftheSea | does it run on USB power? |
| 23:57.43 | hno | dual core A9 @1.2GHz, MALI 400 quad-core. |
| 23:57.49 | hno | compared do |
| 23:59.04 | burndtjamb | sounds tasty |
| 23:59.32 | ManoftheSea | because if that's running on 2.5 W... that's impressive |