00:00.00 | jONpRY | 'm getting all hung up on webrtc |
00:02.25 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
00:04.07 | *** join/#htc-linux WisTilt2 (~wisgreg@wireless251.wirelesstcp.net) |
00:10.54 | *** join/#htc-linux NeoMatrixJR (~chatzilla@173-20-63-62.client.mchsi.com) |
00:17.14 | dr1337 | jONpRY gralloc or gralloc-qsd8k? |
00:17.43 | jONpRY | <PROTECTED> |
00:18.36 | dr1337 | do a search in framebuffer.cpp |
00:18.45 | dr1337 | look for the line int refreshRate |
00:19.08 | dr1337 | in libhardware modules gralloc |
00:19.30 | dr1337 | you'll see that they've commented a section about how refreshRate may be 0 under emulation |
00:19.42 | dr1337 | that's what's getting divided by 0 |
00:19.55 | dr1337 | which causes the segfault |
00:22.29 | jONpRY | don't they handle it? |
00:22.56 | WisTilt2 | jONpRY well she's working now, at least in idle suspend. i haven't changed it to collapse yet but it is now killing usb. im now in idle_suspend 87.7% of the time woohoo |
00:24.10 | *** join/#htc-linux toastcfh (~toastcfh@41-32.200-68.tampabay.res.rr.com) |
00:24.32 | *** join/#htc-linux toastcfh (~toastcfh@unaffiliated/toastcfh) |
00:24.34 | jONpRY | WisTilt2, very cool |
00:25.26 | WisTilt2 | usb blows up though so dont know how to fix that one |
00:25.29 | toastcfh | sup |
00:25.33 | dr1337 | they do in the gralloc |
00:25.39 | dr1337 | but not in the msm7k folder |
00:25.47 | jONpRY | probably just need to enable wake on some interruot |
00:26.05 | toastcfh | what u guys fgot going on |
00:28.16 | WisTilt2 | well im liking this a whole lot. 28ma draw with screen on at main page. jumps up to around 150ma when i change screens or open apps then back down around 28 |
00:29.17 | arrrghhh | WisTilt2, that sounds damned close to winmo |
00:29.39 | arrrghhh | i don't know what winmo spikes to... |
00:29.48 | WisTilt2 | i need to turn it off again and measure but im sure its a whole lot higher than that without |
00:30.20 | WisTilt2 | in full sleep still 20ma with scbs running so that didnt change as it shouldn't. |
00:31.03 | WisTilt2 | going to try full idle collapse and see if it blows up |
00:31.22 | arrrghhh | heh |
00:31.31 | arrrghhh | yea, would be nice to get sleep draw close to winmo |
00:31.39 | arrrghhh | s/nice/amazing/ |
00:34.04 | WisTilt2 | need to get this usb wake irq working though, this is a pain without adb being able to connect. having to see the results in the terminal |
00:34.31 | arrrghhh | heh |
00:34.36 | arrrghhh | get that monocle out WisTilt2 |
00:34.37 | jONpRY | you just have to call set_irq_wake, right after its request_irq thingy |
00:34.44 | arrrghhh | i want a monocle |
00:34.57 | jONpRY | whoa it built |
00:35.17 | WisTilt2 | for which irq though |
00:35.36 | jONpRY | whatever 72k_udc is requesting |
00:35.44 | WisTilt2 | ah initially during setup |
00:35.51 | arrrghhh | jONpRY, isn't htop purrrrrdy |
00:36.03 | jONpRY | yup. reeaal nice |
00:36.12 | arrrghhh | so you got an image to build? |
00:36.29 | jONpRY | maybe the vbus detect irq too. depends on when exactly this blow up is happening |
00:36.39 | jONpRY | yeah got a armv6 ics image |
00:38.06 | WisTilt2 | jONpRY this is a gadget correct? |
00:38.53 | arrrghhh | jONpRY, do i need to setup cron jobs to throw these into a web-server accessible directory? |
00:39.04 | arrrghhh | i guess i could just symlink the directory |
00:39.19 | jONpRY | i could scp |
00:39.24 | arrrghhh | yea, you can. |
00:39.27 | arrrghhh | :P |
00:40.01 | jONpRY | WisTilt2, i don't remember which one we are using anymore. might be the 72k_otg |
00:40.40 | WisTilt2 | yeah otg is what it is. i guess we need to remove wake also when it goes offline? |
00:40.54 | jONpRY | there still is some major work to get it into a haret bootable format |
00:41.08 | arrrghhh | i see |
00:41.11 | jONpRY | probably not. it won't be generating any interrupts if its offline |
00:41.15 | arrrghhh | you got it building tho. that didn't take long. |
00:41.16 | jONpRY | or they will be disabled |
00:41.20 | arrrghhh | perhaps i was slowing you down more than 1,000x :P |
00:41.38 | jONpRY | no clue if it works |
00:41.40 | WisTilt2 | ok building it with wake irq and collapse so we'll see shortly |
00:41.50 | jONpRY | probably just a big blob of garbage |
00:43.33 | WisTilt2 | jONpRY we set the irq wake on the otg->irg or the returned request irq? |
00:43.49 | jONpRY | otg->irq |
00:44.09 | WisTilt2 | ok thats what i did but started getting 2nd thoughts |
00:44.12 | WisTilt2 | building, brb |
00:57.30 | WisTilt2 | jONpRY idle collapse working great but its in collapse only 36% of the time. suspend looks better possibly due to the time collapse takes to shutdown and wake up maybe? usb didn't die when plugged in but as soon as i established adb it disconnected and locked up phone. |
00:58.59 | jONpRY | the newer pm's have threhold for suspend and collapse. so you can hit lower power levels depending on how long a particular sleep is |
00:59.28 | jONpRY | does usb work in suspend? |
00:59.41 | WisTilt2 | trying that now, tried it in collapse |
01:00.21 | WisTilt2 | also left the min sleep at defaults to probably need to tweak that value a bit to find a sweet spot |
01:01.26 | jONpRY | its weird that idle suspend has problem actually, afaik WFI will wake on all interrupts |
01:02.42 | WisTilt2 | suspend was fine. only problem was usb dead but building now with the irq wake and suspend again to test. should work ok |
01:03.13 | *** join/#htc-linux mitsutaka (~mitsutaka@g1-27-253-218-78.bmobile.ne.jp) |
01:03.16 | WisTilt2 | is the otg irq the only one we need wake on? |
01:03.21 | jONpRY | usb dead is strange though |
01:03.42 | jONpRY | all irq's in 72k_udc will need to wake |
01:03.52 | WisTilt2 | usb dead without irq wake after plugging in. showed usb connected then locked up. irq wake fixed that |
01:04.28 | jONpRY | so its 100% with suspend? |
01:04.37 | jONpRY | er all irq's in 72k_otg |
01:05.20 | WisTilt2 | i only see 2, one in usb_state_idle and one in offline |
01:05.48 | jONpRY | so only one request_irq()? |
01:05.52 | WisTilt2 | yes |
01:06.05 | jONpRY | there is probably a few things in pm.c that need to be fixed for collapse to work |
01:06.10 | WisTilt2 | 1 in idle 1 in offline |
01:06.32 | WisTilt2 | im still building all this with suspend and irq wake so let me see if i get same lockup like thiat |
01:06.34 | WisTilt2 | this |
01:07.12 | WisTilt2 | i want the value 1 in the set irq wake correct? |
01:07.13 | jONpRY | they already have enable_irq_wake(otg->irq); |
01:07.56 | WisTilt2 | hmm. i added irq_set_irq_wake also |
01:08.09 | WisTilt2 | which fixed some of it, now im confused |
01:09.37 | jONpRY | i only see one request_irq in otg |
01:10.14 | WisTilt2 | one in usb_state_idle and usb_state_offline |
01:10.24 | jONpRY | no thats in udc |
01:10.35 | WisTilt2 | oh, that's where i am |
01:10.49 | jONpRY | well i think that is not even being built |
01:10.57 | jONpRY | so not sure what good it would be doing |
01:11.13 | jONpRY | but this is not the right irq anyways |
01:11.34 | jONpRY | we don't care about the vbus stuff. that will wake it up. its data transfers in and out of the fifo's |
01:13.31 | WisTilt2 | so udc is only thing being built and their enable wake is there but what i added did get me further. about to boot this setup with suspend... |
01:18.01 | jONpRY | you have confused me |
01:22.00 | WisTilt2 | me too. well still same lockup on usb but it is working great in idle suspend. staying over 85% in idle |
01:22.39 | jONpRY | i rly don'y see the usb controller interrupt in there |
01:23.20 | WisTilt2 | im thinking another irq needs wake also yes |
01:23.51 | WisTilt2 | cant pull log once it locks up to see how far it got |
01:24.35 | jONpRY | no ramconsole? |
01:24.42 | WisTilt2 | this thing is just hanging right around 28ma with screen on doing nothing so its working |
01:25.03 | jONpRY | we needz usb tho |
01:25.06 | WisTilt2 | didnt try that. im getting out of here, ill pick this up when i get home in a bit |
01:25.22 | jONpRY | k |
01:25.28 | WisTilt2 | ill see if it dies before i get home. |
01:42.43 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
01:44.32 | jONpRY | anyone has a rhod startup.txt? |
01:45.42 | d3tul3 | still need it? |
01:47.16 | d3tul3 | jONpRY, http://pastebin.com/RP1bJcwe |
01:49.22 | jONpRY | thx |
01:50.13 | *** join/#htc-linux mgross029 (~mgross@cpe-107-10-11-218.neo.res.rr.com) |
01:50.15 | jONpRY | i think i'm gonna go initrd less |
01:50.30 | jONpRY | but i can't figure out whats up with these images |
01:51.40 | jONpRY | mount /home/jon/ics/system.img /media/loop0/ -o loop -t ext4 |
01:51.41 | jONpRY | mount: wrong fs type, bad option, bad superblock on /dev/loop0, |
01:53.17 | d3tul3 | ext4? |
01:53.37 | jONpRY | that appears to be what its using. but nothing works |
01:54.05 | jONpRY | make_ext4fs -s -l 268435456 -a system out/target/product/panda/obj/PACKAGING/systemimage_intermediates/system.img out/target/product/panda/system |
01:54.30 | d3tul3 | doing this on the TP? |
01:54.37 | jONpRY | tp2 |
01:56.52 | jONpRY | guess i'll just tarball |
01:58.48 | arrrghhh | jONpRY, odd. i thought they were ext2 images... |
01:58.57 | arrrghhh | ics is ext4 by default? |
01:59.16 | arrrghhh | have you tried without -t ext4 |
01:59.24 | arrrghhh | i thought it would work "auto" if you don't specify a type |
02:00.13 | arrrghhh | erm |
02:01.33 | arrrghhh | yea, doesn't work at all. |
02:01.34 | arrrghhh | hum |
02:02.34 | d3tul3 | i haven't been following this ics compilation effort, arrrghhh what are you guys going for here, trying to boot an immage on the rhod? |
02:02.37 | *** join/#htc-linux jonpry (~jon@unaffiliated/jonpry) |
02:02.50 | arrrghhh | d3tul3, i think that's the hope |
02:02.55 | jonpry | dur laptop crashed |
02:02.58 | arrrghhh | heh |
02:03.34 | jonpry | i think my tp2 battery might be trashed |
02:03.45 | arrrghhh | too much scbs? :P |
02:03.57 | jonpry | left it discharged for months |
02:04.06 | arrrghhh | ah |
02:04.25 | arrrghhh | yea, completely discharged for months, is bad. |
02:04.34 | arrrghhh | at 60-80%, that should be OK |
02:08.46 | jonpry | i might have another one around |
02:11.41 | *** join/#htc-linux Rob2222 (~Miranda@p4FFF19AA.dip.t-dialin.net) |
02:17.58 | jonpry | arrrghhh, i need access to repo |
02:18.41 | jonpry | or guess i could curl it. but figured it was in the path |
02:21.04 | arrrghhh | hrm |
02:21.07 | arrrghhh | it's in my path |
02:21.09 | arrrghhh | probably not in yours |
02:22.10 | *** join/#htc-linux mitsutak_ (~mitsutaka@g1-27-253-218-78.bmobile.ne.jp) |
02:25.32 | jonpry | dr1337, where is this hal work going on? |
02:25.56 | arrrghhh | jonpry, you good? |
02:26.02 | dr1337 | jonpry on my comp |
02:26.05 | dr1337 | lol |
02:26.10 | jonpry | no git? |
02:26.40 | dr1337 | no |
02:26.55 | dr1337 | massive pain to mirror 6gb of code |
02:26.57 | dr1337 | hmm |
02:27.04 | dr1337 | looks like i might be on to something... |
02:27.10 | arrrghhh | dr1337, lol |
02:27.15 | arrrghhh | i did it. twice. |
02:27.17 | arrrghhh | :D |
02:27.20 | jonpry | so its not just a hardware/msm7k kind of thing? |
02:27.30 | dr1337 | yeah |
02:27.54 | jonpry | i hope its not a fork of all the repos though |
02:28.04 | dr1337 | remember how i fixed gralloc only to be owned by the "unable to deque native buffer" error |
02:28.11 | jonpry | yeah |
02:28.43 | dr1337 | turns out that el google decided to rename every single android native buffer variable in the system |
02:28.49 | dr1337 | https://github.com/CyanogenMod/android_frameworks_base/commit/b2a153adc9860616acdb96d7cdd64494d5a2a0f2#diff-18 |
02:29.22 | dr1337 | which i hypothesise to be the cause of the that error message |
02:30.50 | jonpry | i don't see any dynamic linking kind of stuff in there |
02:32.22 | dr1337 | just a hypothesis |
02:32.28 | dr1337 | i could be wrong |
02:33.40 | dr1337 | man the cyanogenmod guys are so elitist |
02:34.23 | d3tul3 | what's your take on that jonpry :) |
02:37.16 | jonpry | they are willing to take your code. provided they don't have to give you much |
02:37.24 | arrrghhh | lol |
02:37.54 | jonpry | so like after they make a major release they will recruit people |
02:38.35 | dr1337 | so typical |
02:41.11 | *** join/#htc-linux WisTilt2 (~wisgreg@wireless248.wirelesstcp.net) |
02:41.41 | d3tul3 | WisTilt2, pushed the acoustic fix |
02:41.48 | d3tul3 | thanks for catching that |
02:42.13 | WisTilt2 | ah great. you were about to get a ping since we didnt get to chat about whatever the prob was. |
02:42.24 | WisTilt2 | easy fix hopefully? |
02:42.24 | d3tul3 | it was a typo |
02:42.32 | WisTilt2 | lol, at least you're not alone |
02:42.56 | jonpry | what did they do to you dr1337 |
02:42.58 | WisTilt2 | i got idle power collapse working nicely but need to fix usb that dies in that mode |
02:43.17 | d3tul3 | alright seems there's some confusion which usb drivers we are using, or no? |
02:43.35 | jonpry | my battery is slowly charging. up to 10% now. so maybe there is hope. and i have created *the* sd card |
02:43.53 | WisTilt2 | udc but we have another irq usb related not setup for wake |
02:44.04 | d3tul3 | i have 6 partitions going on here ranging from btrfs, nilfs2 to ext4 and vfat:) |
02:44.09 | jonpry | you mean otg? |
02:44.51 | WisTilt2 | otg is the one that's setup already isnt it? |
02:45.27 | jonpry | i wouldn't say either are setup exactly. setup for regular collapse maybe |
02:45.42 | jonpry | they seem to actively turn off wake when that may not be a good idea |
02:46.19 | jonpry | but this WFI failure is more worrisome because it mostly does not have to do with wake settings. something more broken |
02:47.08 | WisTilt2 | what wfi failure you seeing? dont follow |
02:47.21 | dr1337 | jonpry they just don't respond back when you ask for information |
02:49.11 | jonpry | they are trying to do msm ports, ignoring the omap devices which could probably work |
02:49.30 | jonpry | so i am doing the same thing |
02:49.34 | phh | d3tul3: and nilfs2 rox ? :D |
02:49.54 | jonpry | d3tul3, you try that 3.1? |
02:50.18 | d3tul3 | nilfs2 is nice as long as the daemon doesn't engage |
02:50.23 | phh | "omap could probably work" ? you mean omap that will work ? :p |
02:50.33 | d3tul3 | when the cleaner engages it pegs cpu 20-40% |
02:50.36 | phh | d3tul3: yeah well, someone needs to write some daemon to make it work during idle |
02:50.46 | jonpry | phh, not so clear cut on omap3 |
02:50.48 | phh | d3tul3: oh well if it's only cpu it's no big deal... here it just takes all the I/Os |
02:50.48 | d3tul3 | that would be clutch |
02:50.53 | phh | jonpry: ah yeah |
02:51.46 | phh | jonpry: *afaik* honeycomb works on omap3, so it should work with ICS too i think |
02:51.57 | d3tul3 | jonpry, about to give 3.1 a go |
02:52.03 | jonpry | it definitely can work |
02:52.19 | jonpry | just the HAL is maybe not working, i do see it in the source though |
02:59.19 | d3tul3 | jonpry, compiling past mach-msm |
02:59.57 | jonpry | it should build |
03:00.09 | phh | jonpry: ICS on milestone who be great |
03:00.46 | phh | ouch |
03:00.48 | phh | s/who/would/ |
03:03.11 | jonpry | i don't have one of those |
03:04.08 | phh | me neither, and ? :D |
03:04.12 | phh | would still be fun |
03:04.19 | phh | well i'd like to have ICS on droid 3 first indeed. |
03:05.05 | jonpry | need a zynq-7000 based phone |
03:06.14 | phh | "Dual 12bit 1Msps Analog-to-Digital converter" |
03:06.21 | phh | is looking at his 15€ arm board |
03:06.26 | phh | laugh |
03:06.39 | phh | jonpry: anyway you want it for the fpg a? |
03:07.17 | jonpry | its dual core a9, w/ 32bit 1.3MT ddr3 interface |
03:07.34 | jonpry | and massive fpga for whatever you can dream up |
03:08.08 | phh | jonpry: is there any dev board available ? |
03:08.10 | phh | or at least some price ? |
03:08.31 | phh | 512KB L2, 256KB of integrated memory |
03:08.41 | phh | L2 > integrated memory, lovely |
03:08.53 | jonpry | there is a SOM support to come out http://www.iveia.com/hardware-solutions/processing-modules/atlas-i-lpe/ |
03:09.20 | jonpry | er thats the wrong one |
03:09.21 | jonpry | http://www.iveia.com/hardware-solutions/processing-modules/atlas-i-z7e |
03:09.22 | phh | jonpry: sounds like a really good chipset for networking |
03:09.33 | *** part/#htc-linux marc1706 (~marc@phpbb/modifications/marc1706) |
03:10.07 | jonpry | beaglebone is only $60 i think |
03:10.41 | jonpry | although the beagleboard is actually quite a bit faster at some unique tasks |
03:11.00 | jonpry | i've got the dsp servicing 10k interrupts/sec without a sweat |
03:11.00 | phh | beaglebone doesn't have twop hardware ethernets |
03:11.34 | jonpry | the chip actually has at least one gigabit mac integrated |
03:11.43 | phh | yeah one is connected |
03:11.57 | phh | anyway, i mean that if someone wants to experiment hardware routing |
03:12.35 | jonpry | imho thats an fpga job. there is simply nowhere to put the data to buffer it |
03:12.49 | jonpry | gotta get it out as fast as it comes in |
03:13.23 | phh | that's what i'm saying. |
03:13.29 | phh | zynq 7000 is more suited to that |
03:13.47 | phh | than a beaglebone |
03:13.48 | jonpry | probably |
03:14.09 | jonpry | beaglebone still more fun that cortex |
03:14.24 | phh | muh ? |
03:14.34 | phh | beaglebone is cortex a8 |
03:14.40 | phh | or i'm totally lost |
03:14.42 | jonpry | cortex-M |
03:14.52 | phh | ah ok |
03:15.05 | phh | i don't know omap3's arch |
03:15.10 | phh | only cortex A8+DSP ? |
03:15.23 | phh | well i find omap4's arch really funny, 2 Cortex-M3, that's awesome :D |
03:15.28 | jonpry | yeah. c64x+ dsp |
03:16.02 | jonpry | i think its to cope with lack of interrupt handling capability |
03:16.18 | phh | and there is hardware video decoding or it's done by dsp ? |
03:16.27 | jonpry | dsp |
03:16.36 | jonpry | they call it the IVA2 subsystem |
03:16.56 | phh | hum, on omap4, IVA it's not the dsp |
03:17.21 | jonpry | dunno. no TRM |
03:17.48 | phh | really ? |
03:17.49 | phh | hum i thought so |
03:18.02 | d3tul3 | jonpry, 3.1 kexec? |
03:18.10 | jonpry | no idea |
03:18.26 | d3tul3 | we'll see in a se |
03:18.26 | d3tul3 | c |
03:18.33 | phh | that's the bad thing when you work for a company, you start to get easily docs :( |
03:18.49 | jonpry | still at archos? |
03:18.58 | phh | na |
03:19.05 | phh | i did only 2 months of internship |
03:19.27 | jonpry | dur. totally botched the startup.txt |
03:19.38 | phh | but that's the only time i took a look at omap4's doc |
03:20.14 | phh | soon i'll know how it is without docs :D |
03:20.16 | jonpry | i dunno why they won't sell it through catalog |
03:20.39 | d3tul3 | ugh no ext4 |
03:20.57 | phh | jonpry: well they are used to deliver them publicly |
03:21.12 | phh | so they might have not enough rights to publish freely that |
03:21.38 | jonpry | yeah it seems like all a9 devices are impossible to get |
03:21.50 | jonpry | like its anything more than two a8's |
03:22.08 | jonpry | or that anybody wants or reads the arm portion of the docs |
03:22.30 | jonpry | we have compilers, don't they know? |
03:25.20 | d3tul3 | jonpry, past flying text got up to bootani then blank screen |
03:25.33 | jonpry | oh sweet |
03:25.40 | d3tul3 | adb works |
03:25.43 | jonpry | its a miracle it even entered |
03:25.45 | d3tul3 | brought up desktop |
03:25.52 | d3tul3 | even though no bootani ?hw3d? |
03:26.08 | d3tul3 | ok so no touch screen i can't unlock |
03:26.18 | jonpry | but you has adb? |
03:26.30 | jonpry | then we are good to go. nobody can stop it now |
03:26.55 | d3tul3 | i am saying even the desktop came up |
03:27.01 | d3tul3 | i mean the lockscreen |
03:27.11 | d3tul3 | dmesg looks ok |
03:27.14 | jonpry | yeah awesome |
03:27.19 | jonpry | i can fix touch no problem |
03:27.44 | d3tul3 | i can do it too - just add tssc_manager |
03:27.45 | d3tul3 | ? |
03:28.33 | jonpry | probably everything in driver/input/touchscreen. that way you get the makefile and kconfig |
03:28.51 | jonpry | don't use the other stuff so it doesn't matter |
03:29.19 | jonpry | there were some patches to drivers/input/evdev.c that will need to be merged |
03:29.57 | jonpry | sometimes i do a diff from head to initial. then apply that |
03:30.38 | jonpry | dur rootwait isn't working |
03:34.10 | d3tul3 | well ". kexec.sh" proved to be not the right idea |
03:34.32 | jonpry | this sounds good |
03:35.20 | jonpry | how's that for a kernel port though |
03:36.22 | d3tul3 | booting w tssc_manager |
03:37.59 | d3tul3 | alright |
03:38.01 | d3tul3 | got it going |
03:41.43 | jonpry | whoa. the android is on the screen |
03:41.53 | jonpry | and its freaking |
03:42.49 | d3tul3 | when was the last time you fired up that rhod |
03:42.57 | jonpry | months |
03:43.02 | jonpry | but this is ICS |
03:43.09 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
03:43.34 | d3tul3 | ok why no bootani? where do i go looking here |
03:43.37 | jonpry | i think it would not be freaked out if i knew how to turn off console |
03:43.55 | jonpry | logcat first i think |
03:44.11 | jonpry | check config options to see if hw3d is actually being built in |
03:44.26 | jonpry | dmesg for failures |
03:44.51 | jonpry | how cans i turn off the fb_console? |
03:46.22 | d3tul3 | in config FBCON something or another? |
03:47.00 | d3tul3 | CONFIG_FRAMEBUFFER_CONSOLE=y |
03:47.20 | d3tul3 | no errors in dmesg |
03:47.47 | d3tul3 | E/libEGL ( 1419): No EGLDisplay for hardware EGL! |
03:49.21 | jonpry | you see anything about HW3D in dmesg? |
03:49.36 | jonpry | i want to turn off fbconsole after it boots |
03:49.42 | jonpry | but its running. i can see it :p |
03:49.46 | jonpry | sometimes |
03:50.54 | d3tul3 | [ 86.313201] hw3d_release: in release for pid=20 tid=20 |
03:51.05 | d3tul3 | [ 202.433837] hw3d_resume: resuming |
03:52.01 | jonpry | hrm |
03:52.45 | d3tul3 | msm_hw3d is present in /dev |
03:53.20 | jonpry | most likely the userland driver didn't like what it got back from memory. there was trouble in 39 with setting up the regions |
03:53.40 | jonpry | hopefully pmem is actually built in and running |
03:55.25 | d3tul3 | it must be because there's some hw3d_mmap activity |
03:55.53 | d3tul3 | http://pastebin.com/xN6VkNv0 |
03:56.29 | jonpry | probably check the patches around the enabling of hw3d on 3.0/39 and make sure they are applied |
03:58.37 | d3tul3 | will do, i am looking at may logs |
04:02.29 | jonpry | i fix :p |
04:04.03 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
04:07.14 | jonpry | sweet. i haz ics on tp2 |
04:07.41 | *** join/#htc-linux Rob2223 (~Miranda@p4FFF2EEC.dip.t-dialin.net) |
04:08.19 | d3tul3 | congratulations |
04:08.56 | jonpry | its same thing as stine's atrix, but not worse |
04:13.08 | arrrghhh | jonpry, it booted?!? |
04:13.22 | jonpry | hells yeah |
04:13.48 | arrrghhh | nice |
04:13.51 | jonpry | i unlocked it eventually. and loaded up the browser |
04:13.55 | arrrghhh | i feel slightly responsible |
04:14.01 | arrrghhh | like 0.000000001% :D |
04:14.05 | arrrghhh | nice |
04:14.09 | arrrghhh | slow as all getout? :P |
04:14.22 | jonpry | same as atrix i think |
04:14.29 | jonpry | 3 maybe 4 fps |
04:14.32 | arrrghhh | ouch |
04:15.17 | jonpry | its pretty impressive for an omap4 hal |
04:22.19 | arrrghhh | impressive it works at all |
04:24.11 | jonpry | using the shjitter |
04:34.12 | *** join/#htc-linux bzo (~chatzilla@c-174-62-79-238.hsd1.ca.comcast.net) |
04:38.11 | d3tul3 | jonpry, what's this got for clocks? |
04:40.19 | jonpry | mach folder is just a copy of 3.0 |
04:45.34 | jonpry | dr1337, it seems like the msm7k hal might be older than the gingerbread stuff |
04:47.49 | bzo | congrats jonpry, tp2 the first with a .39/3.0/3.1 kernel and now one of the first to boot ics, lol |
04:48.20 | jonpry | well the ics may never work right |
04:49.45 | bzo | well, I did say boot and not port :P |
04:49.51 | dr1337 | jonpry yes it is |
04:49.56 | arrrghhh | lol |
04:50.04 | arrrghhh | yea, who cares if it works right |
04:50.06 | arrrghhh | it boots! lol |
04:50.08 | dr1337 | jonpry a lot of it is from froyo |
04:50.34 | dr1337 | and well lets just say the reason we had to wait so long for gingerbread on n1 was because much of the drivers are still proprietary |
04:51.53 | arrrghhh | stupid. |
04:52.09 | arrrghhh | ffffffuuuuuuu htc |
04:53.07 | jonpry | dr1337, you have errors about libaudiointerface? |
04:53.35 | jonpry | or did you just use the gingerbread hal? |
04:53.54 | dr1337 | gingerbread hal doesn't work because of incompatible libs |
04:58.10 | jonpry | if there is no more libaudiointerface, then how is the ics hal any different |
04:59.28 | arrrghhh | lazy bastards couldn't have made anything legacy-compatible |
04:59.36 | arrrghhh | i'm honestly surprised they're already turning their back on the N1 |
05:00.26 | jonpry | dur. maybe its all alsa all the time |
05:02.29 | dr1337 | no i mean you can't use the binaries because it's built against gingerbread deps |
05:07.13 | jonpry | we don't have binaries. except winmo ones. and nobody like the idea of running those in android |
05:18.02 | jonpry | we probably need vids of ICS |
05:18.48 | jonpry | to attract attention |
05:20.02 | bzo | who's going to be xdandroid's greentheonly |
05:20.48 | jonpry | i dunno |
05:21.06 | jonpry | he's such a hard worker with those videos. takes hours to plan, film and edit |
05:21.27 | bzo | and also who's wacky enough to fill those shoes? |
05:22.17 | jonpry | probably need the video first to attract that kind of talent |
05:28.23 | jonpry | d3tul3, there is a flashy patch somewhere that puts special tags on pieces of memory used in ioremap |
05:28.27 | jonpry | i think we need that |
05:32.09 | d3tul3 | i just put it in |
05:32.26 | jonpry | oh cool. where is it? |
05:32.42 | d3tul3 | it's in a3b518b8a93b718c7e0c5406ba96600206ca07b4 in 3.0 |
05:33.05 | d3tul3 | i just added the io_pfn_remap or whatever to defconfig |
05:33.59 | jonpry | it wasn't in there? |
05:34.51 | jonpry | iirc its one of those options you can't exactly set as is caused as a side effect of some other option |
05:35.12 | jonpry | but that is the patch that will make it work :p |
05:35.42 | d3tul3 | this guy CONFIG_HAS_ARCH_IO_REMAP_PFN_RANGE=y |
05:35.57 | d3tul3 | i just added that and brought the Kconfig from arch/arm to match |
05:36.16 | jonpry | yeah Kconfig |
05:36.33 | d3tul3 | not sure if there was more to the commit |
05:36.43 | jonpry | no |
05:36.56 | d3tul3 | k it's building |
05:37.16 | d3tul3 | need to talk to WisTilt2 about a 3.1 autobuild |
05:37.33 | jonpry | and ics |
05:37.39 | d3tul3 | jonpry, i read something about you unsetting some RD only flags to HIGHMEM |
05:37.42 | d3tul3 | in .39 |
05:38.06 | d3tul3 | in the logs as i was sifting for hw3d |
05:38.14 | jonpry | thats tls stuffs |
05:38.15 | WisTilt2 | can do but can't make it happen til Monday. |
05:38.43 | WisTilt2 | you got ics booted on tp2??? |
05:38.47 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
05:40.47 | d3tul3 | well that RD_ONLY flag is still set in 3.` |
05:40.50 | d3tul3 | 3.1 |
05:41.13 | jonpry | and i think it didn't pan out quite like i planned. problem ended up being something else |
05:41.18 | jonpry | you bet |
05:41.21 | jonpry | if that was wrong it wouldn't boot at all |
05:41.32 | d3tul3 | but it's unset in 39 |
05:41.35 | jonpry | busybox and adb would be the only working things |
05:41.44 | jonpry | yeah i had a theory |
05:41.49 | jonpry | it was wrong :p |
05:42.05 | d3tul3 | ok i am just wonderinf if it could be causing some of the issues |
05:42.24 | jonpry | nah. it will blow up hard |
05:44.01 | jonpry | life is better on single partition system |
05:44.20 | jonpry | no initrd. so rootfs, no data. just one big system |
05:51.14 | d3tul3 | jonpry, this defconfig -> where did you pull it out from |
05:52.08 | jonpry | well its what happens if you start with 3.0 defconfig. build. then enable v4l2, msm_camera, and the sensor |
05:52.55 | d3tul3 | bootani |
05:55.26 | d3tul3 | k want those pushed out? |
05:56.16 | jonpry | yes |
05:59.31 | d3tul3 | done |
05:59.53 | d3tul3 | good job, that has to be the fastest kernel port in history < 24hrs |
06:00.41 | d3tul3 | i'll spend some time tomorrow catching it up with patchwork |
06:01.44 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
06:01.56 | jonpry | everything works like sound and phone? |
06:02.42 | d3tul3 | i used it as a phone so yeah |
06:03.25 | jonpry | the todo that i know of: no wifi, evdev needs patching, oom proc thing no in there, sdio core needs patching |
06:03.41 | d3tul3 | scbs spam is choking resume |
06:03.57 | d3tul3 | probably bring the defconfig from 3.0? |
06:04.11 | jonpry | i did |
06:04.23 | d3tul3 | hm there's some CPU_FREQ stuff that's different at first glance |
06:04.38 | jonpry | but things get removed if they are not possible because of missing Kconfig, differing Kconfig dependencies and stuff |
06:05.04 | d3tul3 | i see ok |
06:07.53 | d3tul3 | CONFIG_MSM_CPU_FREQ_SCREEN_ON=384000 does this mean it's underclocked? |
06:09.38 | jonpry | probably |
06:18.32 | d3tul3 | i am out good night |
06:51.32 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-95-95.dynamic.mnet-online.de) |
06:54.32 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
08:05.34 | *** join/#htc-linux paulk (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
08:24.34 | *** join/#htc-linux |lippa| (~lippa@ppp118-209-41-138.lns20.mel4.internode.on.net) |
08:38.06 | *** join/#htc-linux dobrin (~dobrin@85.91.150.26) |
08:41.58 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
08:55.30 | *** join/#htc-linux |Jeroen| (~jeroen@d5152B25B.access.telenet.be) |
09:06.36 | *** join/#htc-linux TheXev (user@86.sub-75-242-212.myvzw.com) |
09:24.28 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
09:28.53 | *** join/#htc-linux paulk (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
09:33.26 | *** join/#htc-linux rob_w (~bob@host-188-174-146-225.customer.m-online.net) |
09:48.27 | *** join/#htc-linux kiozen (~kiozen@mnch-5d85abe1.pool.mediaWays.net) |
10:35.10 | *** join/#htc-linux jonpry (~jon@unaffiliated/jonpry) |
10:42.30 | *** join/#htc-linux BabelO_ (~fcr@unaffiliated/babelo) |
10:52.55 | *** join/#htc-linux TheXev (~user@155.sub-174-252-214.myvzw.com) |
10:54.29 | *** join/#htc-linux rzk (~rzk@95-27-68-96.broadband.corbina.ru) |
10:58.41 | *** join/#htc-linux mastermerlin (~Adium@p4FEE548D.dip.t-dialin.net) |
11:21.45 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
11:23.27 | *** join/#htc-linux Cotulla (~myfakemai@nat100-255-205-109.tvoe.tv) |
11:39.56 | *** join/#htc-linux marc1706 (~marc@phpbb/modifications/marc1706) |
11:48.24 | *** join/#htc-linux rajkosto (~rajkosto@2001:470:d76b:da7a:bc46:b8a8:5c6:621e) |
12:05.58 | *** join/#htc-linux TheXev (~user@155.sub-174-252-214.myvzw.com) |
12:07.22 | *** join/#htc-linux toer (~tore@179.81-166-86.customer.lyse.net) |
12:41.01 | *** join/#htc-linux helicopter88 (~helicopte@host187-27-dynamic.49-79-r.retail.telecomitalia.it) |
12:58.30 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-166-124.netcologne.de) |
13:33.50 | *** join/#htc-linux TheXev (user@248.sub-75-214-35.myvzw.com) |
13:49.14 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
13:50.19 | *** join/#htc-linux TheXev (nat32@248.sub-75-214-35.myvzw.com) |
14:02.54 | *** join/#htc-linux gnutoo_ (~GNUtoo@host246-111-dynamic.48-82-r.retail.telecomitalia.it) |
14:03.24 | *** join/#htc-linux mastermerlin1 (~Adium@p4FEE548D.dip.t-dialin.net) |
14:39.45 | *** join/#htc-linux gauner1986 (~Adium@ip-178-201-214-71.unitymediagroup.de) |
14:44.50 | *** join/#htc-linux emwe (~mweirauch@cable-86-56-10-252.cust.telecolumbus.net) |
15:05.28 | *** join/#htc-linux rzk (~rzk@95-27-68-96.broadband.corbina.ru) |
15:21.28 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
15:31.03 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
15:40.24 | *** join/#htc-linux crawling (~crawling@a94-132-161-77.cpe.netcabo.pt) |
15:46.05 | *** join/#htc-linux d3tul3 (~detule@pool-96-234-141-27.bltmmd.east.verizon.net) |
15:49.30 | *** join/#htc-linux marc1706 (~marc@phpbb/modifications/marc1706) |
15:54.55 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-166-124.netcologne.de) |
16:04.36 | *** join/#htc-linux MN-- (~yaaic@5e0dd1f9.bb.sky.com) |
16:08.52 | *** join/#htc-linux GNUtoo (~gnutoo@host246-111-dynamic.48-82-r.retail.telecomitalia.it) |
16:09.43 | *** join/#htc-linux gauner19861 (~Adium@c230168.adsl.hansenet.de) |
16:16.45 | *** join/#htc-linux TheXev (~user@62.sub-174-255-52.myvzw.com) |
16:20.53 | *** join/#htc-linux TheXev (~user@62.sub-174-255-52.myvzw.com) |
16:21.00 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
16:24.27 | *** join/#htc-linux TheXev (~user@62.sub-174-255-52.myvzw.com) |
16:29.29 | *** join/#htc-linux kiozen (~kiozen@mnch-5d85abe1.pool.mediaWays.net) |
16:30.23 | *** join/#htc-linux marc1706 (~marc@phpbb/modifications/marc1706) |
16:40.31 | *** join/#htc-linux gauner1986 (~Adium@ip-178-201-214-71.unitymediagroup.de) |
16:50.41 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
16:51.41 | *** join/#htc-linux MN-- (~yaaic@5e0dd1f9.bb.sky.com) |
17:17.36 | phh | my TP2 stayed alive for something like two days and a half, on android, with push-mail working, and some calls. |
17:17.37 | phh | wtf. |
17:18.02 | arrrghhh | lol |
17:18.03 | arrrghhh | what did you do |
17:18.52 | phh | wel the thing is that now it stoppedd, so i can't dump anything to check -_-' |
17:18.57 | phh | but that's still good old FR7 |
17:19.09 | phh | with the same conf i already got 6 hours of battery life doing nothing |
17:19.37 | jonpry | no omgb? |
17:19.47 | arrrghhh | weird |
17:19.50 | arrrghhh | jonpry, not ready for GSM folk yet |
17:20.10 | jonpry | the gb aosp is good though right? |
17:20.48 | arrrghhh | what's on the git repo is great |
17:20.56 | arrrghhh | what's publicly released, is ancient and clunky |
17:21.07 | arrrghhh | last public gb release we did was march... |
17:21.14 | arrrghhh | i guess i should say 'only' public gb release lol |
17:21.49 | jonpry | should hook phh up with the 39/gb combo |
17:22.07 | emwe | given our constantly stagnating userbase my private gb builds are public arrrghhh :P |
17:22.09 | jonpry | or ics |
17:22.12 | arrrghhh | emwe, lol |
17:22.17 | arrrghhh | jonpry, do eeeeeeet |
17:22.20 | phh | i'll have a droid 3 like on monday, so don't worry about me. |
17:22.29 | emwe | phh: need the old tp2 still? |
17:22.33 | arrrghhh | took the moto plunge eh? |
17:22.42 | arrrghhh | emwe, his isn't the 100 |
17:22.44 | phh | arrrghhh: yeah, that will be a lot of fun to hack :p |
17:23.15 | emwe | phh: what rhod you got? |
17:23.15 | *** join/#htc-linux kiozen (~kiozen@mnch-5d85abe1.pool.mediaWays.net) |
17:23.27 | emwe | arrrghhh: think anything up to 300 will do as well :) |
17:23.29 | phh | at&t's |
17:23.42 | emwe | was that a 300? |
17:23.42 | arrrghhh | i thought you wanted the front cam emwe :P |
17:23.44 | arrrghhh | yes |
17:23.48 | arrrghhh | att rhod = 300 |
17:24.01 | emwe | well, i ideally wan't a working prox sensor :P |
17:24.06 | emwe | the cam is a side effect. |
17:24.11 | arrrghhh | yea |
17:24.15 | arrrghhh | working prox would be nice :P |
17:24.27 | emwe | phh: so if you ever wan't to get rid of it, drop me a line. |
17:24.41 | phh | ok |
17:24.45 | emwe | my gf will kill me. |
17:24.58 | arrrghhh | lol |
17:25.07 | phh | i think i'm too attached to it but well we'll see |
17:25.23 | emwe | yeah, i can understand phh. kinda the same here. can't let loose topa. |
17:25.25 | emwe | ;) |
17:26.40 | phh | and i think it can be a great SSH-er |
17:26.41 | jonpry | is there like rsync over ssh? |
17:26.48 | phh | just need to find out what OS to put for that. |
17:26.54 | phh | jonpry: well rsync by default uses ssh |
17:27.08 | phh | unless you specifiy explicitly rsync protocol |
17:27.14 | arrrghhh | jonpry, yes. |
17:27.49 | jonpry | i'm thinking that is the easiest way to get partial rebuild products |
17:37.34 | *** join/#htc-linux TheXev (~user@62.sub-174-255-52.myvzw.com) |
18:18.15 | *** join/#htc-linux NYL (~NYL@109.96.36.166) |
18:35.01 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
18:42.33 | *** join/#htc-linux GNUtoo (~gnutoo@host246-111-dynamic.48-82-r.retail.telecomitalia.it) |
18:44.06 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
18:45.34 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-166-124.netcologne.de) |
19:04.43 | *** join/#htc-linux x1o (57f54144@gateway/web/freenode/ip.87.245.65.68) |
19:05.16 | *** part/#htc-linux x1o (57f54144@gateway/web/freenode/ip.87.245.65.68) |
19:05.39 | *** join/#htc-linux leviathan (~quassel@2001:470:26:484:6ef0:49ff:fee6:8dca) |
19:10.28 | jonpry | phh, you know anything about the gralloc? |
19:10.35 | phh | a bit |
19:10.50 | phh | iirc it's fairly easy on msm7k |
19:11.14 | jonpry | we are using a binary one on rhod right? |
19:11.23 | phh | no |
19:11.39 | phh | we've got hacks in it |
19:11.59 | jonpry | so the cm repo gralloc probably won't work on wvga? |
19:12.06 | phh | yup |
19:12.28 | phh | just 3D won't work thouogh |
19:12.40 | jonpry | 3d is what i'm trying to get |
19:12.57 | phh | afaik we increased the size of the "buffers" |
19:13.25 | jonpry | is the gralloc different on gb/froyo? |
19:13.51 | phh | different between gb and froyo you mean ? |
19:13.54 | jonpry | yes |
19:14.02 | phh | i don't know, i haven't ever played with gb |
19:15.38 | emwe | our froyo and gb libgralloc are identical. |
19:15.53 | emwe | supper time. laters. |
19:32.30 | *** join/#htc-linux Alex[sp3dev] (~alexander@86.110.163.19) |
19:34.31 | *** join/#htc-linux WisTilt2 (~wisgreg@wireless248.wirelesstcp.net) |
19:35.29 | WisTilt2 | jonpry/detule: i added 3.1 to my autobuild and it looks like it built fine 1st attempt. There ya go. |
19:38.26 | *** join/#htc-linux mastermerlin (~Adium@p4FEE548D.dip.t-dialin.net) |
19:44.58 | mgross029 | WisTilt2, Very cool |
19:45.13 | WisTilt2 | emwe: when you get a chance can you tell me where in sysfs that usb plugged in gets updated? |
19:46.04 | WisTilt2 | mgross029 did you happen to try the .39 again after the last update to see if mic is still dead without toggling speakerphone? |
19:46.25 | WisTilt2 | mine works every time both in and out of speakerphone |
19:49.36 | mgross029 | WisTilt2, Have you updated your super private link with anything new? |
19:50.22 | WisTilt2 | nope. that is just BT stuff that i put on back burner for now. fine tuning idle suspend and collapse right now. |
19:51.57 | mgross029 | Ok. I'll throw the one I compiled into my build folder then. Also, with A1010 enabled and a new userland image I get ringer and mic on first call. |
19:52.39 | WisTilt2 | nice |
19:53.00 | mgross029 | The A1010 fires like crazy in the logs too. |
19:53.10 | WisTilt2 | A1010 enabled in board file you mean? |
19:53.16 | mgross029 | Yes |
19:53.45 | WisTilt2 | ok i need to enable that again and try. it killed speakerphone mic before but i bet that as without the last fix. |
19:54.14 | WisTilt2 | you're not seeing any i2c errors at all then in dmesg? |
19:54.51 | mgross029 | Now that emwe enable the gsensor those errors seem to go away. |
19:55.18 | mgross029 | I do see A1010 i2c messages but they aren't errors |
19:56.27 | *** join/#htc-linux helicopter88 (~helicopte@host187-27-dynamic.49-79-r.retail.telecomitalia.it) |
19:56.29 | mgross029 | Here is one of the messages in logcat I/AudioHardwareMSM72XX_wince( 1290): A1010: do ioctl(A1010_SET_CONFIG) to 0 |
19:57.16 | WisTilt2 | that is normal, as long as you don't see i2c write errors in dmesg |
19:58.02 | mgross029 | Exactly... I'm also seeing these in dmesg now and I don't remember them before ADC3001[0x34] = 0xfc |
19:58.25 | WisTilt2 | which kernel is that on? |
20:00.19 | mgross029 | .35 |
20:00.53 | mgross029 | ADC3001 is right below audience_A1010 in the board file and it is already enabled |
20:01.33 | mgross029 | I have to clean up my logs because of extra spaces before I can post them. |
20:02.20 | mgross029 | I enable a1010 on my local .39 repo too and I think that is the last one I compiled. I'll have to check. If not I will recompile again. |
20:03.44 | *** join/#htc-linux marc1706 (~marc@phpbb/modifications/marc1706) |
20:13.19 | *** join/#htc-linux Andreyxxl[HD2EU] (~Andreyxxl@94.52.236.39) |
20:17.52 | jonpry | something happens with h3d. seems to show up on the screen but the colors are wrong, and everything is stretched |
20:19.00 | mgross029 | jonpry, could that be the video problem stinebd was talking about? |
20:19.22 | jonpry | i doubt it |
20:19.44 | jonpry | by the time it gets to the unlock screen its back in software mode |
20:20.23 | mgross029 | Do you have the force_hw3d=1 in startup? Just trying to think things that could effect it. |
20:20.55 | jonpry | pretty sure hw3d is working |
20:21.11 | mgross029 | ok |
20:21.26 | jonpry | if i had to guess somebody is using it as 16bit display and something else 32 bit |
20:22.13 | mgross029 | hmmm. That would make sense. Kind of throwing each other off |
20:22.54 | phh | jonpry: well, to know if hw3d is working: is the bootlogo slower than usual ? |
20:23.03 | phh | if not, then it's not using hw3d |
20:23.39 | jonpry | yes much slower now |
20:24.03 | jonpry | and the hw3d interrupts are flying |
20:24.12 | phh | stretched would mean it renders as 32bit and displays as 16bits |
20:24.12 | jonpry | the only strange thing is that there is no EGL info from libGLES_qcom |
20:24.20 | jonpry | just says it loaded it |
20:24.24 | phh | display as 16 bit is well known, but 3D in 32bit ? wat ? |
20:24.51 | phh | jonpry: either way, i'd say the bug is in gralloc |
20:24.53 | jonpry | well gotta fix it somehow |
20:24.54 | phh | (but i could be wrong) |
20:25.07 | jonpry | hrm |
20:25.17 | phh | well |
20:25.22 | phh | there are two problems |
20:25.30 | phh | 1. android renders as 32bits, 16bits is waaaaaaaaaaay enough for us |
20:25.41 | phh | 2.bliting 32 => 16 doesn't work |
20:25.48 | phh | none of those are in gralloc actually -_-' |
20:38.58 | *** join/#htc-linux x1o (57f54144@gateway/web/freenode/ip.87.245.65.68) |
20:39.04 | emwe | WisTilt2: i dunno offhand where that happens, sorry. |
20:39.20 | *** part/#htc-linux x1o (57f54144@gateway/web/freenode/ip.87.245.65.68) |
20:39.28 | emwe | mgross029: i2c errors, got none. haven't you "messed" with some i2c board setup there? |
20:39.42 | emwe | mgross029: a1010 is disbaled officially still here. |
20:39.58 | *** join/#htc-linux marc1706 (~marc@phpbb/modifications/marc1706) |
20:44.21 | mgross029 | emwe, correct. No errors here either all good messages so far |
20:44.38 | emwe | i haven't had any i2c errors before though |
20:44.54 | emwe | you enabled a1010? |
20:45.06 | mgross029 | I would get one every now and then from the gsensor |
20:46.01 | mgross029 | Yes... And it seems to be working pretty good. On .35 with a fresh userland I get ringer and in-call mic on first call received. |
20:46.05 | emwe | ah, from the bma150? ok. |
20:46.11 | emwe | hm |
20:46.29 | *** join/#htc-linux jonpry (~jon@unaffiliated/jonpry) |
20:46.30 | emwe | have you had incall issues on .39? |
20:47.06 | emwe | .39 also has a1010 still disabled. |
20:48.41 | mgross029 | I did before yes. Trying now with a1010 enabled. |
20:49.32 | emwe | perhaps the failing ioctls - allthough they shouldn't delay much - have an effect on userland. dunno. |
20:51.03 | mgross029 | Yeah first call mic and ringer problem still exist even with A1010 enabled in .39 |
20:53.56 | emwe | missing a1010 shouldn't do any "delay" harm. probes it's presence on libaudio init. so no runtime issue. |
20:54.13 | emwe | apart from that i haven't had issues so far. |
20:57.08 | mgross029 | Well I have notice my in-call panel pops up a little quicker with it enabled. Unless that is just me, but on .35 it seems to work really well and fires a lot of messages in the log. |
20:59.38 | jonpry | so i think we can get acceleration of ics by patching libhwui |
21:01.00 | mgross029 | cool |
21:01.18 | phh | on msm7k ? |
21:01.24 | *** join/#htc-linux d3tul3 (~detule@pool-96-234-141-27.bltmmd.east.verizon.net) |
21:01.25 | jonpry | it would be a serious effort |
21:01.35 | jonpry | sure |
21:02.07 | jonpry | it is just a graphics library that uses gles2 |
21:02.46 | jonpry | but gles1 implements 75% of gles2 that we need at least. and some things are just eye candy and can be left out, or implemented a different way |
21:03.02 | phh | how would you handle composition ? |
21:03.37 | jonpry | isn't composition just rendering to a texture? |
21:04.25 | phh | yes it is |
21:04.28 | *** join/#htc-linux LordDeath_ (~LordDeath@cable-81-173-166-124.netcologne.de) |
21:04.46 | jonpry | well i think we can render to a texture. hope so at least |
21:07.41 | jonpry | anyone have the list of egl extensions? |
21:08.10 | jonpry | there is OES_framebuffer_object, anything with pbuffers, or fallback to glCopyTexImage2D |
21:09.16 | mgross029 | WisTilt2, Here are some logs with A1010 enabled but still did not get the in-call mic to work on first call or a ringer. http://dl.dropbox.com/u/45928762/39_logs.7z |
21:09.40 | mgross029 | Going to reboot and delete my userland image and retest |
21:09.52 | emwe | i am out guys. have fun. |
21:10.02 | mgross029 | Later emwe gnight |
21:10.26 | jonpry | any work on this renderer has to take place after i get the boot ani fixed though |
21:12.18 | *** join/#htc-linux TheXev (~user@62.sub-174-255-52.myvzw.com) |
21:14.17 | WisTilt2 | mgross029 ok thanks. im deep atm on finalizing this idle collapse stuff while its fresh in my head |
21:14.57 | mgross029 | NP. Just passing along more info to shove in your head :p |
21:21.10 | *** join/#htc-linux TheXev (~user@62.sub-174-255-52.myvzw.com) |
21:30.44 | *** join/#htc-linux TheXev (~user@62.sub-174-255-52.myvzw.com) |
21:37.27 | jonpry | dur. tried to run neocore. got a bunch of this: I/QXEngine( 2160): not a valid QX file |
21:37.29 | jonpry | W/asset ( 2160): unexpected whence 1976320 |
21:37.29 | jonpry | I/QXEngine( 2160): ASSERT ../../QXEngine/QXRender/QXResourceMgr.c(612): FALSE |
21:38.05 | d3tul3 | neocore doesn't rended all textures in GB on my rhod at least not sure if what you are seeing is related |
21:38.38 | jonpry | on to birds i guess |
21:39.09 | *** part/#htc-linux Cotulla (~myfakemai@nat100-255-205-109.tvoe.tv) |
21:39.16 | d3tul3 | which birds do you benchmark on - something from the market? |
21:40.24 | jonpry | something i got off rapidshare |
21:40.27 | jonpry | adfree |
21:44.39 | jonpry | i don't see any life from the birds |
21:44.57 | jonpry | but no failure. seems to bring up gpu |
21:48.53 | jonpry | figured out the touchscreen problem. this actually works pretty well |
21:50.07 | d3tul3 | whoa whoa, when i peeked a few hours ago we were at 2-3fps |
21:50.34 | jonpry | turns out that is mostly on the home screen |
21:50.53 | jonpry | and also it only noticable because of the cursor |
21:51.05 | jonpry | other wise its about as responsive as anything else |
21:51.47 | d3tul3 | what;s the deal w the homescreen |
21:52.18 | jonpry | it must put serious load on the compositor |
21:52.46 | jonpry | when your in an app, it is a lot easier, and if the cursor would go away, it would be normal speed |
21:52.54 | d3tul3 | i dont know much about ICS is it that much different |
21:52.59 | d3tul3 | i mean the homescreen |
21:54.01 | jonpry | it looks different. but i think its done is such a way that it has constant cpu load. no matter how many 3d apps you have on card kind of things or however it works |
21:56.01 | d3tul3 | well birds works |
21:56.20 | d3tul3 | works being broadly defined |
21:56.46 | jonpry | oh yeah? |
21:57.21 | d3tul3 | i have no reference point, need to check 3.0 with birds |
21:58.19 | jonpry | only ever worked right on 39 for me |
21:58.40 | d3tul3 | as in fluidly? skips a lot over here |
21:59.14 | d3tul3 | but then again we are missing half the things in config |
22:00.25 | jonpry | i dunno about skips. it only gets a few fps though |
22:01.20 | jonpry | which is ok. we just don't have the juice. bugs manifest themselves as missing texture, reboots, that kind of thing |
22:01.44 | d3tul3 | no works fine, few fps but i can play it |
22:01.55 | jonpry | perfect then |
22:02.18 | d3tul3 | ashmem was already patched in 3.1 |
22:02.41 | d3tul3 | patched lowmemory killer and got the cpu freq up |
22:04.27 | jonpry | yeah i did the ashmem |
22:07.10 | jonpry | this ics thing is a hassle. i think the renderer would be fun. but its so much work to even get that started |
22:11.20 | jonpry | there is always touchpad i guess |
22:17.45 | jonpry | anywhoz, i can't pull all the weight myself |
22:19.48 | jonpry | probably need a release |
22:38.34 | jonpry | yo |
22:38.41 | jonpry | whoops |
22:58.44 | *** join/#htc-linux rob_w (~bob@host-188-174-146-225.customer.m-online.net) |
23:03.44 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
23:14.49 | *** join/#htc-linux BabelO_ (~fcr@AMontpellier-553-1-44-52.w90-48.abo.wanadoo.fr) |
23:14.49 | *** join/#htc-linux BabelO_ (~fcr@unaffiliated/babelo) |
23:15.57 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
23:16.32 | dr1337 | hey guys, does anyone know how to debug a device before adb logcat gets up and running? |
23:34.46 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
23:35.22 | *** join/#htc-linux d3tul3 (~detule@pool-96-234-141-27.bltmmd.east.verizon.net) |
23:39.05 | jonpry | hi dr1337 |
23:39.29 | jonpry | i was able to get our gralloc to do something with the patch you suggested |
23:40.07 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
23:41.40 | jonpry | get your rhodium ics here: http://www.box.com/s/f1qc8pibutcpzl593kuc with instructions http://pastebin.com/247Twezc |
23:43.52 | d3tul3 | Boot animation will be in strange colors and be there for significant time. All the while doing this terrible flashing thing. |
23:45.02 | jonpry | the strange colors go away if you turn off the gles stuff |
23:45.21 | jonpry | flashing thing is fighting with the fb console |
23:48.19 | dr1337 | hey jonpry |
23:48.32 | dr1337 | nice one |
23:48.52 | dr1337 | there's a bit of an error with my n1 build |
23:49.00 | jonpry | i really want to fix the color distortion |
23:49.11 | jonpry | so i can make a new renderer |
23:49.19 | dr1337 | the system seems to kill itself with my egl libs |
23:50.16 | jonpry | at lockscreen? |
23:50.44 | dr1337 | nah it goes into bootloop |
23:50.48 | dr1337 | even before I can do a logcat |
23:51.20 | *** join/#htc-linux mgross029 (~mgross@cpe-107-10-11-218.neo.res.rr.com) |
23:51.48 | jonpry | last_kmsg will probably tell you what crashed the kernel |
23:52.39 | dr1337 | that should be in the cache partition right? |
23:52.52 | jonpry | <PROTECTED> |
23:54.15 | dr1337 | can't access that |
23:54.26 | dr1337 | system goes into bootloop before i can even run adb shell |
23:54.46 | dr1337 | and when i boot into recovery it does't work because it's loading its own kernel into /proc/kmesg |