| 00:48.05 | *** join/#htc-linux the-leviathan (~quassel@2001:470:26:484:6ef0:49ff:fee6:8dca) |
| 00:57.48 | *** join/#htc-linux BabelO (~fcr@2a01:e35:8adc:3150:21e:8cff:fe2d:7d3f) |
| 00:57.48 | *** join/#htc-linux BabelO (~fcr@unaffiliated/babelo) |
| 01:21.48 | *** join/#htc-linux the-leviathan (~quassel@c-82-192-226-27.customer.ggaweb.ch) |
| 03:58.14 | *** join/#htc-linux d3tul3 (~detule@unaffiliated/d3tul3) |
| 04:29.48 | *** join/#htc-linux ALoGeNo (~alogeno@81.32.131.110) |
| 04:29.48 | *** join/#htc-linux ALoGeNo (~alogeno@unaffiliated/alogeno) |
| 08:45.06 | *** join/#htc-linux eR^zeRa` (~zzeratul@ip-84-42-202-42.net.upcbroadband.cz) |
| 09:01.48 | *** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 09:43.30 | *** join/#htc-linux Cotulla (~myfakemai@109.205.253.11) |
| 09:57.32 | *** join/#htc-linux Alex[sp3dev] (~alexander@182.67.ipv4.fosdem.net) |
| 10:52.11 | *** join/#htc-linux BabelO (~fcr@unaffiliated/babelo) |
| 10:57.43 | *** join/#htc-linux |lippa|^ (~lippa@CPE-121-220-110-93.lnse3.win.bigpond.net.au) |
| 12:48.02 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-66-116.dynamic.mnet-online.de) |
| 13:42.06 | Cotulla | helo |
| 13:42.23 | Cotulla | Marc it should be this problem |
| 14:04.45 | Marc | Cotulla: what problem? |
| 14:06.24 | Cotulla | initrd |
| 14:07.10 | Marc | you mean the offset is the issue? |
| 14:07.18 | Cotulla | yes |
| 14:07.19 | Cotulla | it is |
| 14:07.21 | Cotulla | :( |
| 14:07.29 | Marc | oh, you confirmed it? |
| 14:07.34 | Cotulla | not really |
| 14:07.40 | Cotulla | but I got it in dream actually |
| 14:07.49 | Marc | hehe |
| 14:07.59 | Cotulla | there is zero inited data section |
| 14:08.04 | Cotulla | and it's large than 10m |
| 14:08.16 | Cotulla | and someone had that problem before |
| 14:08.20 | Cotulla | but I don't remember wh |
| 14:08.21 | Cotulla | o |
| 14:09.02 | Marc | easy to fix |
| 14:09.08 | Marc | just release a new magldr version :P |
| 14:09.19 | Cotulla | yea working under it |
| 14:09.35 | Cotulla | but no problem with boot.img should be |
| 14:09.39 | Marc | and then slap zeusk_ because it's obviously his fault |
| 14:09.40 | Cotulla | try it |
| 14:09.41 | Cotulla | maybe |
| 14:09.47 | Cotulla | no |
| 14:09.57 | Cotulla | "FEW YEARS AGO 10MB WAS ENOUGH" |
| 14:10.05 | Cotulla | it just can't get enough |
| 14:10.24 | Marc | that's like saying a few years ago 20 GB hdds were enough |
| 14:11.21 | Cotulla | yea |
| 14:11.28 | Cotulla | that time I will set 32 |
| 14:11.29 | Cotulla | LD |
| 14:11.31 | Cotulla | :D |
| 14:11.34 | Cotulla | no mercy |
| 14:12.16 | Marc | hehe |
| 14:14.04 | Marc | hm, primou has phys offset of 0x13F00000 |
| 14:14.19 | Marc | and this: initrd=0x11000000 |
| 14:14.27 | Cotulla | hm |
| 14:16.16 | Marc | weird stuff ^^ |
| 14:16.45 | Cotulla | it's called development |
| 14:16.51 | Cotulla | in linux world |
| 14:16.54 | *** join/#htc-linux d3tul3 (~detule@unaffiliated/d3tul3) |
| 14:17.04 | Cotulla | hm I think I know this guy |
| 14:17.10 | Cotulla | ~seen detule |
| 14:17.15 | apt | detule is currently on #htc-linux. Has said a total of 101 messages. Is idling for 19h 38m 57s, last said: 'cuss somebody out'. |
| 14:17.24 | Cotulla | ~seen d3tul3 |
| 14:17.24 | apt | d3tul3 is currently on #htc-linux (30s), last said: 'isn't hypervisor some sort of kernel patchset'. |
| 14:18.00 | Cotulla | lol yeah I win |
| 14:18.02 | Cotulla | it's same one |
| 14:34.50 | d3tul3 | what's up Cotulla |
| 14:35.07 | Cotulla | just checked if u same once as delute |
| 14:35.11 | Cotulla | *one |
| 14:35.32 | d3tul3 | yeah, that's my machine at work |
| 14:35.47 | Cotulla | hehe |
| 15:26.37 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
| 15:47.27 | *** join/#htc-linux eR^zeRa` (~zzeratul@ip-84-42-202-42.net.upcbroadband.cz) |
| 15:47.30 | *** join/#htc-linux d3tul3 (~detule@unaffiliated/d3tul3) |
| 15:55.07 | jonpry | d3tul3 i see they ravaged your thread |
| 16:04.21 | Cotulla | which thread? |
| 16:05.24 | jonpry | http://forum.xda-developers.com/showthread.php?t=2098956&page=2 |
| 16:05.40 | jonpry | i guess kernel porting is not sufficiently dev related |
| 16:05.41 | Marc | prepares for rampage |
| 16:06.57 | Cotulla | lol "like-minded individuals" |
| 16:08.27 | Cotulla | hehe |
| 16:16.35 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
| 16:28.19 | *** join/#htc-linux arif-ali (~arif-ali@81.144.163.60) |
| 17:07.37 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-66-116.dynamic.mnet-online.de) |
| 17:09.23 | *** join/#htc-linux stroughtonsmith (~steven@86-42-129-30-dynamic.b-ras1.bbh.dublin.eircom.net) |
| 17:32.18 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 17:35.13 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 17:40.07 | d3tul3 | jonpry, it's a testament to the state of development for S3 |
| 17:40.51 | d3tul3 | pretty amazing really.....perhaps the fact that kondik has it, scared all other developers away |
| 17:43.10 | *** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 17:44.31 | arrrghhh | d3tul3, I can't believe your thread got moved to general... wtf |
| 17:45.11 | Cotulla | lol arrrghhh here |
| 17:45.21 | *** join/#htc-linux bartman (~bart@2607:f2c0:f00e:700::dd) |
| 17:45.29 | Cotulla | and batman |
| 17:45.30 | Cotulla | lol |
| 17:45.30 | arrrghhh | lol Cotulla thar? |
| 17:45.35 | arrrghhh | bartman |
| 17:45.39 | arrrghhh | ^^ |
| 17:45.41 | Cotulla | lol |
| 17:46.26 | Cotulla | funny nicks |
| 17:46.29 | Cotulla | on this channel |
| 18:07.23 | jonpry | Cotulla, do anything with qdsp? |
| 18:07.39 | Cotulla | not yet |
| 18:07.46 | Cotulla | u? |
| 18:07.53 | jonpry | nah |
| 18:08.22 | Cotulla | but looked a bit to 80-NB419-3_A_hexagon_virtual_machine_spec.pdf |
| 18:08.24 | jonpry | need to get out the apq8060 |
| 18:09.13 | Cotulla | it has rather good description |
| 18:09.19 | Cotulla | for example memory subsystem |
| 18:09.57 | jonpry | i read through it |
| 18:12.58 | Cotulla | funny thing |
| 18:13.20 | Cotulla | that they are talking about "qct camera drivers can not be included into mainline" |
| 18:13.27 | Cotulla | while included that hexagon stuff |
| 18:13.58 | Cotulla | into mainline |
| 18:16.56 | Cotulla | so u think better to implement |
| 18:17.04 | Cotulla | HVM by own code? |
| 18:17.08 | Cotulla | or change kernel |
| 18:17.50 | jonpry | will linus accept the patches? |
| 18:19.38 | jonpry | if not, i say build HVM |
| 18:21.09 | d3tul3 | i got my apq too! |
| 18:21.43 | jonpry | going to run linux on hexagon? |
| 18:22.41 | jonpry | put whole debian on it. then just like switch the framebuffer over to "real" linux |
| 18:22.45 | Cotulla | back |
| 18:22.49 | Cotulla | we can compile android |
| 18:22.54 | Cotulla | be first with 6 cores :D |
| 18:22.59 | Cotulla | 6 cores android |
| 18:24.04 | Cotulla | well, it's not clear why they included it to mainline - QCT doesn't provide any ability to run custom kernels on QDSP, as well HVM instances for support. ever for me not clear how they are going to port it to the new versions and test. |
| 18:24.46 | jonpry | i don't think they understand that this HVM doesn't exist |
| 18:24.58 | Cotulla | lets write to them :D |
| 18:25.51 | jonpry | please tear out arch/hexagon. it is a ruse :) |
| 18:26.06 | Cotulla | yes, something like that. |
| 18:26.30 | Cotulla | or if they provided hardware to mainline team - it also doesn't sound fair |
| 18:26.37 | jonpry | maybe we can just patch to remove HVM support |
| 18:26.45 | Cotulla | yes it's possible |
| 18:27.13 | jonpry | like [PATCH] 1/1 convert hexagon arch to run on real hardware instead of nonexistant virtual machine |
| 18:27.30 | Cotulla | yes |
| 18:27.41 | jonpry | qcom would love that |
| 18:27.43 | Cotulla | :D |
| 18:28.08 | Cotulla | I just don't understand sense to include that inside mainline |
| 18:28.21 | Cotulla | ever don't understand reason why it was included |
| 18:28.43 | jonpry | me neither |
| 18:29.07 | Cotulla | what is good email for question? |
| 18:29.12 | Cotulla | I wanna troll them a bit |
| 18:29.21 | jonpry | lkml? |
| 18:29.24 | Cotulla | yes |
| 18:29.49 | d3tul3 | so what's the deal, this hypervisor virtual machine does not exist? |
| 18:29.58 | jonpry | linux-kernel@vger.kernel.org |
| 18:30.06 | Cotulla | it exists, but not avaiable |
| 18:30.21 | Cotulla | ever I had never seen it |
| 18:30.28 | d3tul3 | who is it available to? |
| 18:30.47 | jonpry | there is something called the qdsp access program that may provide it |
| 18:30.49 | Cotulla | for customers I think |
| 18:30.51 | Cotulla | yes |
| 18:30.53 | jonpry | under NDA |
| 18:30.57 | Cotulla | like built in fw |
| 18:31.35 | d3tul3 | i guess it got accepted to korg because it's completely contained in arch/hexagon? |
| 18:31.51 | Cotulla | but it can not work without HVM fw |
| 18:32.23 | jonpry | i think they did not understand that the targeted machine is only in qcom's mind |
| 18:32.40 | Cotulla | targeted machine is closed yes |
| 18:33.20 | Cotulla | is linux-kernel@vger.kernel.org maillist? |
| 18:34.09 | jonpry | yes |
| 18:34.42 | Cotulla | ok] |
| 18:35.28 | noobhands | troll it good |
| 18:37.12 | Cotulla | hm should it be List: linux-arch; ? |
| 18:37.49 | Cotulla | or who will sort it |
| 18:37.49 | Cotulla | pff |
| 18:39.16 | jonpry | i think linux-kernel so that you can get a rise out of linus |
| 18:40.48 | jonpry | too bad he is out for the week |
| 18:42.35 | Cotulla | hm |
| 18:42.40 | Cotulla | it looks hard to answer :D |
| 18:43.28 | Cotulla | okk |
| 18:43.30 | Cotulla | letme write |
| 18:44.04 | d3tul3 | make sure you share maybe we can all pile on your thread |
| 18:45.00 | Cotulla | hm it's not clear for me how to send message to "linux-kernel" via common email majordomo@vger.kernel.org |
| 18:45.09 | Cotulla | oh via linux-kernel@vger.kernel.org |
| 18:45.12 | Cotulla | seems |
| 18:45.15 | jonpry | yeah |
| 18:48.53 | d3tul3 | arrrghhhAWAY, can you pm whatever relevant mod is and have that thread closed |
| 18:53.13 | Cotulla | :D |
| 19:01.45 | Cotulla | okay sent |
| 19:03.38 | Cotulla | but it doesn't appear in the list |
| 19:07.35 | jonpry | it takes a while |
| 19:07.38 | Cotulla | ok |
| 19:07.45 | Cotulla | lets port it until message appear |
| 19:07.47 | jonpry | what is the subject? |
| 19:07.47 | Cotulla | :D |
| 19:07.58 | Cotulla | "Hexagon Architecture inside kernel" |
| 19:08.12 | jonpry | 3.7 kernel? |
| 19:08.25 | Cotulla | I checked it's ever in 3.8 |
| 19:08.36 | Cotulla | http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=blob;f=arch/hexagon/mm/vm_tlb.c;h=9647d00cb761002c3d4370df41785b740f6c980c;hb=88b62b915b0b7e25870eb0604ed9a92ba4bfc9f7 |
| 19:08.40 | Cotulla | this is 3.8 right? |
| 19:08.40 | jonpry | i don't porting rc kernels though |
| 19:08.50 | Cotulla | oh I compiled 3.4 |
| 19:08.51 | jonpry | *like porting |
| 19:08.51 | Cotulla | :D |
| 19:12.37 | jonpry | where did you get hexagon gcc? |
| 19:12.46 | Cotulla | from CA I think |
| 19:17.49 | jonpry | what is the difference between qdsp6 and hexagon? |
| 19:18.02 | Cotulla | should be same |
| 19:18.07 | Cotulla | just different names |
| 19:18.18 | jonpry | 2 different compilers |
| 19:18.24 | Cotulla | they renamed it |
| 19:18.31 | Cotulla | since some moment in the past |
| 19:18.45 | Cotulla | no idea whyy |
| 19:20.38 | Cotulla | lol cool description |
| 19:20.39 | Cotulla | "The HVM environment provides a trust model which is distinct from, but interoperable |
| 19:20.40 | Cotulla | with the trust model of other cores in the system." |
| 19:20.40 | Cotulla | :D |
| 19:27.45 | *** join/#htc-linux d3tul3 (~detule@unaffiliated/d3tul3) |
| 19:35.45 | jonpry | thinks Coutulla was preemptively banned from lkml |
| 19:36.31 | Cotulla | maybe or my service provider or they tired by noobs from yandex.ru :) |
| 19:36.35 | Cotulla | :D |
| 19:37.21 | Cotulla | anyway as usually I don't wait a lot of from them |
| 19:37.48 | Cotulla | they are playing own games |
| 19:41.59 | phh | hexagon sounds cooler. |
| 19:42.04 | Cotulla | heh |
| 19:42.07 | Cotulla | phh |
| 19:42.10 | Cotulla | old good phh |
| 19:42.11 | Cotulla | :D |
| 19:42.21 | phh | i'm not that old... oh right. |
| 19:43.09 | Cotulla | hexagon = 6 cores |
| 19:43.25 | Cotulla | actually it's like Hyper Threading |
| 19:43.28 | Cotulla | 6 hardware threads |
| 19:43.38 | Cotulla | but ALU and other blocks are shared |
| 19:44.06 | jonpry | i thought it was just qdsp6 = hexagon, qdsp5 = pentagon |
| 19:44.25 | Cotulla | qdsp5? pentagon? never heard about it |
| 19:44.28 | phh | +1 jonpry |
| 19:44.50 | Cotulla | hm intersting |
| 19:44.55 | Cotulla | so u mean 6 in name |
| 19:45.00 | Cotulla | not 6 HW threads |
| 19:45.02 | Cotulla | :P |
| 19:45.04 | phh | yeah :p |
| 19:45.08 | phh | Cotulla: it's really 6 threads ? |
| 19:45.11 | phh | that sounds weird |
| 19:45.14 | Cotulla | yea |
| 19:45.22 | Cotulla | I was ever able to start another |
| 19:45.51 | phh | a DSP for which there is less than a 6th time used for computation |
| 19:46.29 | Cotulla | start is very easy |
| 19:46.37 | Cotulla | just command start (r0) |
| 19:46.42 | Cotulla | r0 is address where to start lol |
| 19:46.56 | jonpry | they both ran? |
| 19:47.01 | Cotulla | yea |
| 19:47.09 | jonpry | oh wow |
| 19:47.49 | Cotulla | fuck yea |
| 19:51.37 | jonpry | how does it decide which thread gets interrupts? |
| 19:54.05 | *** join/#htc-linux d3tul3 (~detule@unaffiliated/d3tul3) |
| 19:55.59 | Cotulla | I didn't got that yet |
| 19:56.04 | Cotulla | but may in future |
| 19:56.21 | Cotulla | I think a mask |
| 19:58.53 | Cotulla | but more important to implement SW MMU |
| 19:59.03 | Cotulla | I looked already how it was done for MIPS/SH CPUs in Windows CE |
| 20:12.20 | jonpry | you know the format of hexagon tlb entries? |
| 20:12.26 | Cotulla | yes |
| 20:12.52 | *** join/#htc-linux d3tul3 (~detule@unaffiliated/d3tul3) |
| 20:13.34 | jonpry | maybe i am missing something, but software mmu doesn't seem very hard |
| 20:14.08 | Cotulla | I missed a place where I said that software mmu is bad :O |
| 20:14.14 | Cotulla | :D |
| 20:14.38 | Cotulla | but it should be implemented |
| 20:14.52 | jonpry | yes |
| 20:15.44 | Cotulla | like FILO |
| 20:16.07 | jonpry | except some entries |
| 20:16.23 | jonpry | like exception handler |
| 20:16.39 | Cotulla | actually maybe I & D separate |
| 20:16.43 | Cotulla | yes |
| 20:16.48 | Cotulla | and one entry for common code |
| 20:17.09 | Cotulla | map 16K region to that |
| 20:19.51 | d3tul3 | you boys are starting a github project for this? |
| 20:19.57 | Cotulla | not yet |
| 20:20.03 | Cotulla | just talking |
| 20:20.26 | d3tul3 | need some framework here - fork 3.7 stable and create a repo? |
| 20:20.43 | Cotulla | I actually compiled old 3.4 |
| 20:21.17 | jonpry | so the kernel sends entire pagetable to HVM? which creates the TLB? |
| 20:21.38 | jonpry | and how does that work with multiple threads? |
| 20:22.42 | Cotulla | they share same TLB |
| 20:22.45 | Cotulla | always |
| 20:22.59 | Cotulla | I think this call just setup pointer |
| 20:23.08 | Cotulla | or maybe copied entries to internal table |
| 20:23.12 | Cotulla | and then SW MMU put them to TLK |
| 20:23.13 | Cotulla | TLB |
| 20:23.18 | Cotulla | inside exception handlers |
| 20:23.24 | jonpry | how can they share TLB |
| 20:23.34 | Cotulla | they are |
| 20:23.35 | jonpry | like 6 threads might have different code at the same VA |
| 20:23.39 | Cotulla | no |
| 20:23.44 | Cotulla | they can not by design seems |
| 20:24.10 | jonpry | i don't understand |
| 20:24.15 | Cotulla | ? |
| 20:24.20 | Cotulla | just TLB entries are always common |
| 20:24.29 | Cotulla | all HW threads are using them to |
| 20:24.33 | Cotulla | convert VA -> PA |
| 20:24.37 | Cotulla | same set |
| 20:24.43 | jonpry | how can that work? |
| 20:24.49 | Cotulla | why it can not? |
| 20:24.59 | Cotulla | oh |
| 20:25.01 | jonpry | different code at same address |
| 20:25.03 | Cotulla | u mean from OS position? |
| 20:25.09 | Cotulla | hm |
| 20:25.19 | Cotulla | yea really OS like Linux can have threads in different places |
| 20:25.22 | Cotulla | hm |
| 20:25.33 | jonpry | TLB entries must have thread id or something |
| 20:25.41 | Cotulla | there is ASID AFAIK |
| 20:25.53 | Cotulla | hm |
| 20:28.56 | d3tul3 | i thought tlb was percpu |
| 20:29.03 | Cotulla | I hope it's not like uLinux |
| 20:29.38 | d3tul3 | oh i see tlb *entries* having thread id |
| 20:29.52 | Cotulla | where |
| 20:30.36 | d3tul3 | no no i am just saying I (think) I understand what you guys are talking about.... |
| 20:31.27 | Cotulla | I hope there is a way to specify thread id |
| 20:31.31 | Cotulla | but not sure |
| 20:31.50 | jonpry | is there even a way to get the id of the current thread? |
| 20:31.56 | Cotulla | yes |
| 20:32.12 | Cotulla | on V2 it's done like this: |
| 20:32.13 | jonpry | thats half the battle |
| 20:32.13 | Cotulla | <PROTECTED> |
| 20:32.13 | Cotulla | get_hwthread_num: |
| 20:32.13 | Cotulla | <PROTECTED> |
| 20:32.13 | Cotulla | <PROTECTED> |
| 20:32.13 | Cotulla | <PROTECTED> |
| 20:37.42 | jonpry | do we need a small test program to try different bits in the TLB? |
| 20:38.05 | Cotulla | in qMAGLDR I just made static mapping |
| 20:38.10 | Cotulla | for devices and RAM |
| 20:38.31 | jonpry | and the second thread is able to use those mappings? |
| 20:38.41 | Cotulla | looks like yes |
| 20:38.44 | Cotulla | I can check agian |
| 20:38.46 | Cotulla | again |
| 20:38.53 | Cotulla | but there is global bit in TLB |
| 20:39.11 | jonpry | probably don't want that |
| 20:45.29 | jonpry | ASID = address space identifier |
| 20:45.35 | Cotulla | yes |
| 20:45.41 | Cotulla | usually it is |
| 20:46.05 | jonpry | that sounds like the right thing tm |
| 20:46.12 | Cotulla | hm |
| 20:46.26 | Cotulla | in theory if setup different ASIDs to every thread |
| 20:46.27 | Cotulla | yes |
| 20:48.04 | jonpry | The prototype Hexagon Virtual Machine |
| 20:48.04 | jonpry | <PROTECTED> |
| 20:48.17 | Cotulla | saw it |
| 20:57.17 | jonpry | i think that might be misleading |
| 20:57.26 | jonpry | afaict linux is not a usermode app |
| 20:57.50 | jonpry | guest mode or something |
| 20:58.04 | Cotulla | what exactly |
| 20:58.05 | Cotulla | ? |
| 20:58.19 | jonpry | single ASID |
| 20:58.30 | Cotulla | it looks like 2^6 ASIDs |
| 20:58.40 | Cotulla | maybe they mean single asid per thread? |
| 20:58.58 | jonpry | is an ASID the same as a thread ID? |
| 20:59.07 | Cotulla | I think no |
| 21:00.26 | Cotulla | we can check linux config |
| 21:02.38 | jonpry | like how would as ASID be set for a thread? |
| 21:11.08 | Cotulla | looks via ssr |
| 21:11.10 | Cotulla | but I am not sure |
| 21:12.53 | Cotulla | yea it can be used I think |
| 21:13.00 | Cotulla | for separation |
| 21:13.06 | Cotulla | in that case in tlbmiss |
| 21:13.08 | Cotulla | we need |
| 21:13.16 | Cotulla | hm |
| 21:13.19 | jonpry | gdb seems to have some tools for reading asid |
| 21:13.20 | Cotulla | read thread id |
| 21:14.04 | Cotulla | and select right table |
| 21:14.08 | Cotulla | and put entry with right asid |
| 21:14.19 | jonpry | yeah |
| 21:15.51 | Cotulla | and 6 separate page tables |
| 21:15.55 | Cotulla | :O |
| 21:16.20 | d3tul3 | no more than 6 threads at a time? |
| 21:16.29 | Cotulla | no |
| 21:16.39 | Cotulla | at least in mine |
| 21:16.41 | d3tul3 | is that some sort of a hardware limitation/ |
| 21:16.43 | Cotulla | v2 :P |
| 21:16.45 | Cotulla | yes |
| 21:16.55 | Cotulla | it's like hyperthreading |
| 21:16.58 | Cotulla | in Intel |
| 21:17.06 | d3tul3 | perhaps hence the name hexagon |
| 21:17.13 | Cotulla | yea |
| 21:17.53 | jonpry | more than 6 threads can be scheduled. but only 6 can run concurrently |
| 21:18.19 | Cotulla | no |
| 21:18.25 | Cotulla | only 6 |
| 21:18.29 | jonpry | huh? |
| 21:18.36 | Cotulla | 6 HW threads . . . |
| 21:18.45 | Cotulla | it's hardware limitation |
| 21:19.00 | Cotulla | but maybe urs with 4 |
| 21:19.02 | Cotulla | or 5 :P |
| 21:19.02 | jonpry | yeah i think d3tul3 is asking a somewhat different question |
| 21:19.08 | jonpry | lost in translation |
| 21:19.16 | Cotulla | yes maybe |
| 21:19.22 | Cotulla | if he about OS threads |
| 21:19.30 | jonpry | i think so |
| 21:19.31 | Cotulla | limited only by memory |
| 21:19.35 | Cotulla | should be |
| 21:19.45 | Cotulla | but we can hard code a rule for 99 specially for him :P |
| 21:19.46 | Cotulla | :D |
| 21:19.51 | jonpry | lol |
| 21:20.04 | Cotulla | if (atomic_inc(threads_count) > 99) BUG(); |
| 21:20.44 | Cotulla | should work |
| 21:21.09 | jonpry | hex core |
| 21:21.21 | jonpry | and if you can take over modem, then it's like 18 core |
| 21:22.40 | Cotulla | no modem in APQ |
| 21:22.45 | Cotulla | :P |
| 21:22.57 | jonpry | d3tul3 has modem |
| 21:23.04 | Marc | isn't the modem on a seperate chip? |
| 21:23.04 | Cotulla | well |
| 21:23.13 | Cotulla | for APQ MDM is used |
| 21:23.21 | Cotulla | also 2 + 1 |
| 21:23.35 | Cotulla | hm |
| 21:23.39 | Cotulla | 6 * 3 = 18 |
| 21:23.46 | jonpry | i guess the l2 cache needs to be configured |
| 21:23.46 | Cotulla | but no coherency |
| 21:24.00 | jonpry | screw coherency |
| 21:24.02 | Cotulla | it won't work |
| 21:24.17 | Cotulla | not sure |
| 21:24.19 | Cotulla | maybe possible |
| 21:24.22 | Cotulla | maybe not |
| 21:24.30 | jonpry | 3 different linux's with MPI |
| 21:25.15 | jonpry | you see how l2 can be setup as cache or TCM? |
| 21:29.08 | jonpry | i guess it will at least run with/without cache |
| 21:29.48 | Cotulla | ? |
| 21:30.29 | jonpry | no idea about cache enable instructions |
| 21:31.09 | jonpry | and it needs to be setup because l2 can be any size up to 256kb. with rest being sram |
| 21:31.10 | d3tul3 | lol |
| 21:31.26 | Cotulla | yes |
| 21:31.41 | d3tul3 | there's some guy who claims to have 3.4 kernel on s3....no source or anything mind you "I ran into some issues w/ antikanging and html stuff |
| 21:31.41 | d3tul3 | so one of the sr. developers from qualcomm will be coming over 2mrw to assist me |
| 21:31.41 | d3tul3 | (see - need help sometimes - and have no problem admitting this)" |
| 21:32.09 | jonpry | lol |
| 21:32.09 | Cotulla | "html stuff" wtf? |
| 21:32.11 | Cotulla | :D |
| 21:32.30 | Cotulla | in 3.4 drivers are in HTML ? |
| 21:32.31 | jonpry | antikanging |
| 21:32.36 | jonpry | new technology |
| 21:32.50 | d3tul3 | this is too good |
| 21:33.01 | jonpry | is that your thread? |
| 21:33.12 | d3tul3 | yeah that's what everyone suggested i lookup |
| 21:33.19 | d3tul3 | before i had the thread closed |
| 21:35.29 | jonpry | sounds like somebody changed include/version |
| 21:36.04 | Cotulla | lol |
| 21:36.14 | Cotulla | I want HTML5 drivers |
| 21:36.20 | noobhands | has them |
| 21:36.24 | Cotulla | HAHAHAHAH |
| 21:36.25 | Cotulla | :D |
| 21:36.41 | Cotulla | but hands are they so it should be have not has? |
| 21:36.48 | Cotulla | cuz it's not one instance of hands |
| 21:36.52 | Cotulla | *hand |
| 21:37.00 | Cotulla | but 2 instances |
| 21:37.04 | noobhands | i dunno, i will have to wait for qualcomm guy to come over my place |
| 21:37.07 | noobhands | i can't do everything |
| 21:37.07 | Cotulla | which are both nooby? |
| 21:37.28 | noobhands | i lost my fucking key tho - so maybe next week |
| 21:40.22 | d3tul3 | i got three hexagons in here and however many more in the apq8060 - freaking supercenter |
| 21:40.31 | d3tul3 | thread bonanza |
| 21:40.44 | jonpry | distcc |
| 21:42.41 | jonpry | d3tul3, mebbe u r just famous |
| 21:42.43 | jonpry | http://rootzwiki.com/topic/37216-romunofficial-cm101-for-d2vzw-2013-01-23-34-based-kernel/ |
| 21:45.54 | d3tul3 | that's pretty good....but damn this guy chad goodman takes the cake |
| 21:46.09 | arrrghhh | chad goodman is a winnar |
| 21:46.48 | d3tul3 | wish I could link his posts but he was apparently banned from xda and now posts on this anthrax forum you need some sort of three tier registration to (quote above was taken from there)... |
| 21:47.00 | arrrghhh | weird |
| 21:47.11 | arrrghhh | yea, chad has been banned since his evo3d asshattery |
| 21:47.37 | jonpry | i like this whole notion of private and not gpl violating |
| 21:47.42 | d3tul3 | anyway apparently chad's been working on some 3.4 kernel for the last 5 months...and everything is working he's just doing some antikanging and html5 mods |
| 21:47.51 | arrrghhh | yea, s because the forum is "private" you're not violating the GPL? |
| 21:47.53 | arrrghhh | dafaq? |
| 21:48.32 | d3tul3 | if he fixed the ioctl mismatch that jonpry found i'll tip my hat to him |
| 21:48.42 | Cotulla | are u english still? |
| 21:50.02 | jonpry | i just don't believe it |
| 21:50.21 | Cotulla | just can't get enough? |
| 21:51.30 | jonpry | is qMAGLDR going to run linux? |
| 21:52.19 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
| 21:53.00 | Cotulla | how? |
| 21:53.11 | Cotulla | I will load image from aMAGLDR as usual |
| 21:53.19 | Cotulla | this qImage |
| 21:53.21 | Cotulla | to ram |
| 21:53.26 | Cotulla | and then run it |
| 21:54.45 | Cotulla | so |
| 22:00.48 | d3tul3 | Cotulla, q/magdlr still closed? |
| 22:03.04 | Cotulla | yes |
| 22:03.34 | Cotulla | it should be placed to hall of shame actually :D |
| 22:03.40 | arrrghhh | lol |
| 22:05.00 | Cotulla | it was fork of 1.07 |
| 22:05.03 | Cotulla | very dirty code |
| 22:05.05 | *** join/#htc-linux teemodk (~teemodk@87-104-122-81-dynamic-customer.profibernet.dk) |
| 22:05.19 | Cotulla | as well I ported code from CL to GCC |
| 22:05.22 | Cotulla | also hacks |
| 22:07.00 | arrrghhh | HAX |
| 22:07.11 | Marc | LAX |
| 22:07.54 | Cotulla | BUS STOP |
| 22:33.18 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
| 22:52.57 | *** join/#htc-linux the-leviathan (~quassel@c-82-192-226-27.customer.ggaweb.ch) |
| 22:56.30 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 23:14.21 | *** join/#htc-linux Mavy (mavfree@unaffiliated/mavy) |
| 23:22.29 | *** join/#htc-linux jonpry (~jon@user-0c6tmn0.cable.mindspring.com) |