| *** rpierce99 has quit IRC | 00:29 | |
| *** Yusuke has joined #xdandroid | 00:52 | |
| *** Yusuke is now known as Yusuke14 | 00:52 | |
| *** bzo has joined #xdandroid | 01:15 | |
| *** rpierce99 has joined #xdandroid | 01:16 | |
| *** rpierce99 has quit IRC | 01:38 | |
| *** babijoee has joined #xdandroid | 02:05 | |
| *** kalemas has joined #xdandroid | 02:29 | |
| *** bzo has quit IRC | 02:43 | |
| *** kalemas has left #xdandroid | 03:27 | |
| *** sythezn has joined #xdandroid | 06:03 | |
| sythezn | heya | 06:03 |
|---|---|---|
| sythezn | I'm getting a build error on trackMe in RefBase.cpp, passing void into bool... | 06:04 |
| sythezn | any clue where i start lookin' | 06:04 |
| *** p0g0 has joined #xdandroid | 06:26 | |
| *** kalemas has joined #xdandroid | 06:49 | |
| *** babijoee has quit IRC | 07:19 | |
| arrrghhh | sythezn, did you change something in that file...? | 08:50 |
| *** hardwalker has quit IRC | 09:22 | |
| arrrghhh | jonpry, did you see? ICS on the G1 lol. | 09:32 |
| arrrghhh | slow as all getout. | 09:32 |
| arrrghhh | seems to be stuck in landscape mode too, heh | 09:32 |
| *** emwe has joined #xdandroid | 10:29 | |
| *** rpierce99 has joined #xdandroid | 10:31 | |
| *** virsys has joined #xdandroid | 10:46 | |
| stinebd | it's super fast on the atrix | 11:09 |
| arrrghhh | i heard. still 3-4fps? | 11:09 |
| stinebd | 30-40 | 11:10 |
| stinebd | still not great | 11:10 |
| arrrghhh | not atrocious either | 11:10 |
| stinebd | but you can't see anything | 11:10 |
| arrrghhh | i guess with all that eye-candy, you need every bit of oomph to make it appear smooth. | 11:10 |
| arrrghhh | lol | 11:10 |
| stinebd | still waiting for my transformer prime to arrive | 11:11 |
| stinebd | angrily | 11:11 |
| arrrghhh | heh | 11:11 |
| stinebd | they took it off amazon | 11:12 |
| stinebd | wonder why | 11:12 |
| stinebd | any merkans other than me working today? | 11:15 |
| arrrghhh | i got called and had to login | 11:15 |
| arrrghhh | attempted to fix something, but couldn't. | 11:15 |
| rpierce99 | i fixed something this morning | 11:15 |
| arrrghhh | some physical cabling issue it seemed like | 11:15 |
| stinebd | i work in a building full of crazy people and crazy people like to go crazier on holidays | 11:16 |
| arrrghhh | heh | 11:16 |
| stinebd | therefore i'm boycotting work at noon | 11:16 |
| arrrghhh | sounds like a plan | 11:17 |
| arrrghhh | i've never seen any macy's thanksgiving parade before. | 11:17 |
| arrrghhh | someone should've told me they lipsync everything. i'm crushed. | 11:18 |
| stinebd | it's an enormous waste of time and money | 11:18 |
| rpierce99 | arrrghhh: have you seen the g1 ICS video? | 11:18 |
| arrrghhh | rpierce99, yea that's what i was talking about earlier | 11:19 |
| arrrghhh | oh you weren't here. | 11:19 |
| arrrghhh | good stuff | 11:19 |
| rpierce99 | yeah i'm not a log reader | 11:19 |
| rpierce99 | i was wondering if that's what yours was like | 11:19 |
| rpierce99 | or is it better | 11:19 |
| arrrghhh | worse | 11:19 |
| rpierce99 | which ones worse | 11:19 |
| arrrghhh | much, much worse | 11:19 |
| arrrghhh | ours | 11:19 |
| arrrghhh | er jonpry's | 11:19 |
| rpierce99 | got it | 11:19 |
| arrrghhh | i don't think anyone else has touched it lol | 11:19 |
| rpierce99 | so maybe there is something he can steal from jcarrz1 | 11:20 |
| arrrghhh | i couldn't really see much of anything, the launcher/screen was FUBAR | 11:20 |
| arrrghhh | yea i figured if the g1 can run it that's good news for us | 11:20 |
| arrrghhh | but that is a bag of ass. i know a lot is missing, but i'm not sure how much more they can optimize | 11:20 |
| arrrghhh | i saw jonpry talking about his fandangled new compositor | 11:20 |
| arrrghhh | not sure where he got with it tho, still reading. | 11:20 |
| stinebd | fubar is how you can describe the atrix version right now | 11:26 |
| stinebd | artifacts everywhere, flickering, stuff disappearing | 11:26 |
| arrrghhh | stinebd, did you try booting it on the TP2? | 11:26 |
| stinebd | no | 11:26 |
| arrrghhh | or is yours still MIA | 11:26 |
| stinebd | one of them is | 11:26 |
| stinebd | heh | 11:26 |
| stinebd | no clue wtf happened to it | 11:26 |
| arrrghhh | damn | 11:26 |
| stinebd | i think one of the cats took off with it | 11:26 |
| arrrghhh | lol | 11:26 |
| stinebd | seriously it's gonna be behind furniture in another room or something | 11:27 |
| stinebd | that's what they do | 11:27 |
| jonpry | hey guys | 11:27 |
| arrrghhh | i was hoping they stole it for development purposes | 11:27 |
| arrrghhh | jonpry, wassup | 11:27 |
| stinebd | hey philip j. pry | 11:27 |
| arrrghhh | stine's been waiting to use that for weeks | 11:27 |
| jonpry | where's g1 video? is it fast? | 11:27 |
| arrrghhh | jonpry, not fast. 1 sec. | 11:27 |
| stinebd | it's not fast at all | 11:27 |
| stinebd | http://forum.xda-developers.com/showthread.php?p=19648827 | 11:28 |
| stinebd | i win | 11:28 |
| jonpry | i've almost got the hardware compositor going | 11:28 |
| arrrghhh | http://www.youtube.com/watch?feature=player_embedded&v=Ym4A82ft5pw | 11:28 |
| stinebd | youtube or it didn't happen | 11:28 |
| arrrghhh | stinebd, you do win, original thread even. | 11:28 |
| jonpry | should work on all devices | 11:28 |
| jonpry | http://pastebin.com/7XbZJVjP | 11:30 |
| jonpry | it might be worth while to search for msm 7230 blobs with different opengl extensions | 11:31 |
| arrrghhh | hrm | 11:31 |
| stinebd | i don't like the way this log looks | 11:33 |
| jonpry | need GL_TEXTURE_EXTERNAL_OES | 11:33 |
| jonpry | i have to rewrite something to not use it, then it will work | 11:33 |
| stinebd | will this work for atrix's old ass gingerbread blobs? | 11:34 |
| jonpry | the rewritten thing would | 11:35 |
| jonpry | i wonder what GL_QUALCOMM_direct_texture is | 11:35 |
| stinebd | some kinda direct texture operation | 11:36 |
| jonpry | sounds like what we need | 11:37 |
| stinebd | there's a qcom vbo extension too | 11:38 |
| jonpry | buts thats lame because 1.1 has vbo's as a non extension api | 11:39 |
| arrrghhh | so can our hardware support 1.0 or 1.1? | 11:40 |
| arrrghhh | i was under the impression that 1.1 works | 11:40 |
| stinebd | 1.1 yes | 11:40 |
| jonpry | see log. also notice that flinger is using the qcom renderer | 11:40 |
| jonpry | we need to get that g1 guy | 11:45 |
| arrrghhh | based on that xda-devs post, he didn't change much other than to make it compile for that arch... | 11:45 |
| jonpry | yeah which is already duplicated work | 11:46 |
| stinebd | should be in the hero forums somewhere | 11:46 |
| stinebd | the g1 guy says he got it from some other guy's heroc tree | 11:46 |
| jonpry | cause we did armv6 compile a week ago | 11:46 |
| arrrghhh | jonpry, i noes... | 11:46 |
| jonpry | we are 10x closer to having composition than CM is for msm8x60 | 11:49 |
| stinebd | we do armv6j for gingerbread | 11:49 |
| arrrghhh | lol | 11:49 |
| jonpry | does JIT work? | 11:50 |
| arrrghhh | ok, wtf google music. you say 20,000 songs yet choke at 14,841... wtf | 11:50 |
| stinebd | jonpry: yeah | 11:52 |
| stinebd | i'm always worried about it being unstable | 11:52 |
| stinebd | but it seems just about as good as fast | 11:52 |
| *** detule has joined #xdandroid | 11:53 | |
| *** helicopter88 has joined #xdandroid | 11:54 | |
| arrrghhh | i don't think our arch can really benefit fully (or at all) from JIT | 11:54 |
| stinebd | you wouldn't think that | 11:54 |
| jonpry | hows that | 11:54 |
| detule | happy thanksgiving/holiday/whatever-you-are-celebrating | 11:55 |
| arrrghhh | detule, happy turkey day indeed | 11:55 |
| stinebd | don't dodge the topic | 11:56 |
| stinebd | explain your insubordination about jit | 11:56 |
| arrrghhh | lol | 11:56 |
| arrrghhh | well it blows things up | 11:56 |
| arrrghhh | and makes our devices slower (seemingly) | 11:57 |
| arrrghhh | so why would my 'insubordination' be unjustified? | 11:57 |
| stinebd | i'm gonna need to see some proof to those accusations | 11:57 |
| stinebd | it neither blows things up nor slows them down | 11:57 |
| arrrghhh | bugs.xdandroid.com | 11:57 |
| stinebd | gingerbread jit is solid | 11:57 |
| detule | jit? not sure about blowing things up but my device seems more responsive jit-less | 11:58 |
| arrrghhh | detule and rpierce99 said int:fast was faster than int:jit... | 11:58 |
| detule | bzo said something along the same lines as well | 11:58 |
| detule | as did mgross029 | 11:58 |
| stinebd | a couple of numbnuts on some irc channel doesn't convince me | 11:58 |
| stinebd | no offense | 11:58 |
| arrrghhh | detule, there was an ancient bug in froyo where if you spammed the volume rocker it would cause a mini-reset within Android. | 11:58 |
| stinebd | haha yeah that one was hilarious | 11:58 |
| stinebd | god damn froyo jit | 11:58 |
| detule | i am sure bzo wasn't offended | 11:59 |
| arrrghhh | i'm offended | 11:59 |
| stinebd | it was actually meant for rpierce99 | 11:59 |
| detule | backtracking | 11:59 |
| arrrghhh | lol | 11:59 |
| stinebd | time to go | 11:59 |
| stinebd | work is over | 11:59 |
| stinebd | 11:59, screw it | 11:59 |
| stinebd | i'm out early | 11:59 |
| arrrghhh | have a good 'un stinebd | 11:59 |
| sythezn | arrrghhh, nah I initially changed the makefile to remove the java version check, but I've set it back | 12:24 |
| arrrghhh | lol | 12:24 |
| sythezn | lol? | 12:25 |
| arrrghhh | "nah I initially changed the makefile to remove the java version check" | 12:25 |
| arrrghhh | brilliant | 12:26 |
| arrrghhh | let's remove all the checks from our system/kernel | 12:26 |
| arrrghhh | everything will work so much better!!!! | 12:26 |
| helicopter88 | i removed the check too | 12:26 |
| arrrghhh | fail. | 12:26 |
| helicopter88 | failed to build :( | 12:26 |
| sythezn | yeah, but i didn't change refbase | 12:26 |
| arrrghhh | why would it? | 12:26 |
| arrrghhh | you guys are insane. | 12:26 |
| sythezn | anyways, all i did was comment out the check | 12:26 |
| helicopter88 | arrrghhh,you have never messed with java then :) | 12:27 |
| arrrghhh | helicopter88, nope. just worked, following directions of course. | 12:27 |
| sythezn | then i selected the java version and uncommented | 12:27 |
| helicopter88 | well,i tried to follow direction,error 41 was there for me :( | 12:27 |
| arrrghhh | dunno | 12:27 |
| sythezn | frameworks/base/libs/utils/RefBase.cpp: In member function ‘void android::RefBase::weakref_type::trackMe(bool, bool)’: | 12:28 |
| sythezn | frameworks/base/libs/utils/RefBase.cpp:483:67: error: passing ‘const android::RefBase::weakref_impl’ as ‘this’ argument of ‘void android::RefBase::weakref_impl::trackMe(bool, bool)’ discards qualifiers [-fpermissive] | 12:28 |
| sythezn | make: *** [out/host/linux-x86/obj/STATIC_LIBRARIES/libutils_intermediates/RefBase.o] Error 1 | 12:28 |
| arrrghhh | ok | 12:28 |
| arrrghhh | sythezn, what are you trying to achieve exactly? | 12:28 |
| sythezn | I'm trying to build froyo | 12:29 |
| arrrghhh | why | 12:29 |
| helicopter88 | sythezn: this could be due to GCC 4.6,search the make error on google | 12:29 |
| arrrghhh | it's already built, i don't think anything has changed on it in a while. | 12:29 |
| sythezn | yeah but i want to play with it to learn... | 12:29 |
| sythezn | you mean frx07? | 12:29 |
| arrrghhh | .1, yes | 12:29 |
| sythezn | uh... | 12:30 |
| * sythezn is lost | 12:30 | |
| arrrghhh | well the directions on the building the source page worked great for me. | 12:30 |
| arrrghhh | are you dense? | 12:30 |
| arrrghhh | FRX07.1 | 12:30 |
| arrrghhh | the newest version | 12:30 |
| arrrghhh | ? | 12:30 |
| sythezn | yeah I figured that | 12:30 |
| arrrghhh | so where are you lost. | 12:30 |
| arrrghhh | --_ | 12:30 |
| arrrghhh | er | 12:30 |
| arrrghhh | -_- | 12:30 |
| sythezn | the build instructions here failed | 12:31 |
| helicopter88 | are you running ubuntu? | 12:31 |
| sythezn | lol i was thinking froyo was still under active dev | 12:31 |
| sythezn | yeah | 12:31 |
| arrrghhh | froyo isn't | 12:31 |
| arrrghhh | .27 isn't | 12:31 |
| arrrghhh | it's probably not going to get any more updates. | 12:31 |
| sythezn | ah | 12:31 |
| sythezn | so gingerbread? | 12:32 |
| arrrghhh | unless there's some really compelling reason, dev focus is indeed on gb. | 12:32 |
| sythezn | i've synced both trees... | 12:32 |
| * sythezn tries a gb build | 12:32 | |
| arrrghhh | supposedly that's harder to build | 12:32 |
| arrrghhh | but i didn't have a problem with that either... | 12:33 |
| sythezn | ah | 12:33 |
| *** virsys has quit IRC | 12:52 | |
| *** detule has quit IRC | 13:04 | |
| *** virsys has joined #xdandroid | 13:05 | |
| *** sythezn has quit IRC | 14:11 | |
| *** virsys has quit IRC | 14:15 | |
| *** helicopter88 is now known as helicAWAY | 14:18 | |
| *** rpierce99 has quit IRC | 14:24 | |
| *** helicAWAY has quit IRC | 16:27 | |
| *** xdandroid has joined #xdandroid | 17:07 | |
| *** emwe has quit IRC | 17:09 | |
| *** sythezn has joined #xdandroid | 17:47 | |
| sythezn | does it make any difference if I'm building with a 64 or 32 bit system? | 17:48 |
| sythezn | nevermind... blonde moment | 17:49 |
| *** sythezn has quit IRC | 17:51 | |
| *** jonpry has quit IRC | 17:52 | |
| *** jonpry has joined #xdandroid | 18:05 | |
| *** jonpry has quit IRC | 18:09 | |
| *** ryannathans has quit IRC | 18:09 | |
| *** p0g0 has quit IRC | 18:09 | |
| *** Entropy512 has quit IRC | 18:09 | |
| *** ^Tempest has quit IRC | 18:09 | |
| *** arrrghhh has quit IRC | 18:09 | |
| *** lodtank has quit IRC | 18:09 | |
| *** purefusion has quit IRC | 18:09 | |
| *** phh has quit IRC | 18:09 | |
| *** stinebd has quit IRC | 18:09 | |
| *** larryone has quit IRC | 18:09 | |
| *** hyc has quit IRC | 18:09 | |
| *** Yusuke14 has quit IRC | 18:09 | |
| *** XirXes has quit IRC | 18:09 | |
| *** babijoee has joined #xdandroid | 18:13 | |
| *** jonpry has joined #xdandroid | 18:13 | |
| *** p0g0 has joined #xdandroid | 18:13 | |
| *** Yusuke14 has joined #xdandroid | 18:13 | |
| *** Entropy512 has joined #xdandroid | 18:13 | |
| *** phh has joined #xdandroid | 18:13 | |
| *** ryannathans has joined #xdandroid | 18:13 | |
| *** ^Tempest has joined #xdandroid | 18:13 | |
| *** XirXes has joined #xdandroid | 18:13 | |
| *** purefusion has joined #xdandroid | 18:13 | |
| *** hyc has joined #xdandroid | 18:13 | |
| *** stinebd has joined #xdandroid | 18:13 | |
| *** lodtank has joined #xdandroid | 18:13 | |
| *** arrrghhh has joined #xdandroid | 18:13 | |
| *** larryone has joined #xdandroid | 18:13 | |
| *** leguin.freenode.net sets mode: +o stinebd | 18:13 | |
| *** rpierce99 has joined #xdandroid | 18:16 | |
| *** kalemas has left #xdandroid | 18:17 | |
| *** babijoee_ has joined #xdandroid | 18:42 | |
| *** hardwalker has joined #xdandroid | 19:43 | |
| *** bzo has joined #xdandroid | 19:57 | |
| bzo | I can shed some light on the 12c000 | 19:57 |
| jonpry | hi bzo | 19:59 |
| bzo | hey jonpry | 19:59 |
| jonpry | please do, and fix my 3d's | 19:59 |
| bzo | I can explain the 12c000, but am clueless about much more :P | 19:59 |
| jonpry | i hacked it to check for larger amounts, but there is still nothing on the screen | 19:59 |
| bzo | basically, the hardcoded buffer size in the libgles_qcom so and gralloc have to match | 20:00 |
| jonpry | clueless is about where i am at | 20:00 |
| bzo | arm assembly is kinda weird in how it represents numbers | 20:00 |
| jonpry | i'm pretty sure they do. but that is 2f0000 | 20:00 |
| bzo | instead of just a 32 bit value, it is a 2 part thing with a barrel shift | 20:00 |
| bzo | the value specified in gralloc is the closest we could get | 20:01 |
| bzo | with doing a simple patch to the lib | 20:01 |
| bzo | representing 3000<<10 takes more than 32 bits | 20:01 |
| bzo | err I meant explaining the 2F0000, not the 12c000 | 20:02 |
| jonpry | its one of those immediate instructions that takes an 8 bit with a shift or something? | 20:03 |
| bzo | yeah | 20:03 |
| bzo | don't remember the specifics but it is damn confusing | 20:03 |
| bzo | maybe it's not 32 bits either, been a long time since I looked at it | 20:03 |
| jonpry | i suspect the 0x2f0000 is working ok because i'm using the xdandroid blob and gralloc with xdandroid constants | 20:04 |
| jonpry | also hw3d sort of seems to be working. i am blasting it with gl commands and it at least pretends to be rendering something | 20:04 |
| bzo | there would be a lot of chatter in logcat if you ran out of gralloc memory | 20:05 |
| bzo | that's how we knew to expand it when we were fixing up the camera | 20:05 |
| jonpry | no errors like that | 20:06 |
| jonpry | only E/EGL.oem ( 52): invalid buffer offset (0x000bb800 + 0x000bb800 > 0x0012c000) | 20:06 |
| bzo | hmm, seems like that might cause it not to display anything | 20:07 |
| jonpry | i hacked it to accept larger values as [acl] reported that fixed something | 20:07 |
| jonpry | but i wonder why this error appears. same blob under GB does not have that error | 20:08 |
| bzo | was that the old buf size constant before we changed it to 2f0000? | 20:08 |
| jonpry | not sure | 20:08 |
| bzo | ah yes, it is 12C000=480*320*4*2 | 20:09 |
| bzo | so I would assume any constant of that value needs to be patched to the new one | 20:09 |
| jonpry | it must be using a different code path. like the image buffer has different flags, or is in different memory than on GB | 20:10 |
| jonpry | the whole system pretty much blows my mind | 20:14 |
| bzo | all this indirection from having to deal with the closed libs sucks | 20:14 |
| bzo | it's amazing that jb was able to reverse all the camera stuff | 20:15 |
| jonpry | there is all this madness with surface flinger doing all the allocations and then having other processes use its memory | 20:15 |
| jonpry | and they have all these opaque pointers that are then recast and used by other modules, including the blob | 20:15 |
| jonpry | the only sane place to have a 3d driver is in the kernel | 20:17 |
| bzo | I understand the position of the hard liners in the kernel now about "tainted" stuff now | 20:25 |
| jonpry | source would make this much easier | 20:43 |
| *** XirXes has quit IRC | 20:45 | |
| *** XirXes has joined #xdandroid | 20:48 | |
| *** detule has joined #xdandroid | 20:59 | |
| *** phh has quit IRC | 22:24 | |
| *** phh has joined #xdandroid | 22:26 | |
| *** TheSeven has quit IRC | 22:32 | |
| *** [7] has joined #xdandroid | 22:32 | |
| *** ryannathans has quit IRC | 23:32 | |
| *** detule has quit IRC | 23:40 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!