| 00:17.24 | *** join/#htc-linux TheXev (~user@18.sub-174-252-212.myvzw.com) |
| 00:18.09 | arrrghhh | dr1337, don't hold your breath |
| 00:19.06 | jONpRY | i'd be so on that if i could download/build it |
| 00:19.28 | jONpRY | need remote build server :p |
| 00:20.06 | arrrghhh | heh |
| 00:20.15 | arrrghhh | you could use mine :P |
| 00:21.22 | jONpRY | what kind of processor/ram? |
| 00:21.41 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
| 00:22.00 | arrrghhh | specs aren't great |
| 00:22.04 | jONpRY | they are saying 16gb required. sounds a little overkill to me |
| 00:22.06 | arrrghhh | my laptop actually has better specs, how sad. |
| 00:22.07 | arrrghhh | holy shit |
| 00:22.12 | arrrghhh | my lappy has 12gb... |
| 00:22.16 | arrrghhh | core i7 |
| 00:22.22 | arrrghhh | 2820qm or something |
| 00:22.48 | jONpRY | that would probably compile it in an hour if it was on an external and fast hard drive |
| 00:23.03 | arrrghhh | yeesh |
| 00:23.22 | arrrghhh | i grabbed the source and built in like 20 minutes, but all i wanted was the f@#$ fastboot binary |
| 00:23.35 | arrrghhh | well, built in 20 minutes |
| 00:23.36 | jONpRY | you has ics source? |
| 00:23.40 | arrrghhh | grabbing the source was longer... |
| 00:23.45 | arrrghhh | isn't it public? |
| 00:24.00 | jONpRY | yeah but not many people have actually finished the repo sync |
| 00:24.05 | arrrghhh | ah |
| 00:24.11 | arrrghhh | yea, i didn't actually sync for ics |
| 00:24.28 | arrrghhh | at least i don't think i did |
| 00:24.29 | arrrghhh | http://source.android.com/source/downloading.html |
| 00:24.32 | arrrghhh | i followed that |
| 00:24.41 | d3tul3 | jONpRY, them touchpad boys on it i am sure? |
| 00:24.47 | arrrghhh | although... d'oh |
| 00:24.56 | arrrghhh | i only checked out master. der. |
| 00:25.00 | jONpRY | repo init -u https://android.googlesource.com/platform/manifest -b android-4.0.1_r1 ? |
| 00:25.09 | arrrghhh | jONpRY, yea, i see my failure now |
| 00:25.16 | arrrghhh | i didn't check out ics. |
| 00:25.19 | arrrghhh | you want me to try? |
| 00:25.21 | jONpRY | no. nobody is working on it. |
| 00:25.35 | jONpRY | sure why not |
| 00:25.57 | jONpRY | a slower but more reliably connected box could be better |
| 00:26.17 | jONpRY | wistilt2 probably has something too |
| 00:27.18 | arrrghhh | hrm |
| 00:27.21 | jONpRY | i kind of want to use the CM manifest |
| 00:27.30 | arrrghhh | do i have to init master first? |
| 00:27.38 | arrrghhh | i get an error when doing that repo init command you gave jONpRY |
| 00:27.52 | jONpRY | what kind of error? |
| 00:27.53 | arrrghhh | looks like it's trying to pull from android.git.kernel.org, which is still down no? |
| 00:28.06 | arrrghhh | android.git.kernel.org[0: 149.20.4.77]: errno=Connection refused |
| 00:28.24 | jONpRY | your repo is probably too old |
| 00:28.36 | jONpRY | curl https://dl-ssl.google.com/dl/googlesource/git-repo/repo > ~/bin/repo |
| 00:28.38 | arrrghhh | the binary? |
| 00:28.42 | jONpRY | yeah |
| 00:28.43 | arrrghhh | damn i just updated that the other day. |
| 00:28.45 | arrrghhh | i'll do it again |
| 00:28.49 | jONpRY | well its just a boot strap script |
| 00:29.00 | jONpRY | it downloads the real repo |
| 00:29.19 | arrrghhh | hah |
| 00:29.27 | arrrghhh | i didn't think i would have to redo that step, workin no |
| 00:29.29 | arrrghhh | now* |
| 00:29.43 | arrrghhh | crap |
| 00:29.47 | arrrghhh | i should've done this in a screen session :/ |
| 00:30.17 | arrrghhh | think before you repo sync |
| 00:30.41 | d3tul3 | jONpRY, calming down that WARN_ON does make 3.0 play nice w scbs |
| 00:31.17 | arrrghhh | alrighty, it's poopin |
| 00:31.42 | arrrghhh | jONpRY, my connection isn't stupid fast like wistilt2, but it's alright. 30/5 |
| 00:31.54 | arrrghhh | er i stepped down to 20/3 recently. sorry :P |
| 00:32.04 | arrrghhh | still not bad |
| 00:32.37 | jONpRY | i think you should use a different repo anyways |
| 00:32.53 | arrrghhh | other than android.googlesource.com...? |
| 00:32.59 | jONpRY | yeah |
| 00:33.33 | jONpRY | repo init -u git://github.com/CyanogenMod/android.git -b ics |
| 00:33.50 | arrrghhh | can i have two running at the same time? lol |
| 00:34.09 | jONpRY | would have to be in different directories |
| 00:34.13 | arrrghhh | yea |
| 00:34.25 | jONpRY | sure why not |
| 00:35.02 | arrrghhh | k |
| 00:35.02 | jONpRY | there are lots of patches required to make it like build. and CM will get it done long before google does anything |
| 00:35.10 | arrrghhh | hopefully my hard drive isn't full by the time i get home. |
| 00:35.19 | arrrghhh | probably |
| 00:35.26 | arrrghhh | supposedly CM worked closely with google for ICS... |
| 00:36.21 | jONpRY | i dunno. last i heard the man in charge was having the same old trouble with build errors |
| 00:36.40 | jONpRY | and nobody has any idea how to get the GLES/EGL or anything to show signs of life |
| 00:37.49 | arrrghhh | great |
| 00:45.18 | jONpRY | d3tul3, any crashes yet? |
| 00:45.41 | d3tul3 | no, i am a little worried how well it runs |
| 00:46.03 | arrrghhh | lol |
| 00:46.09 | arrrghhh | it's not like a 2-stroke |
| 00:46.21 | jONpRY | its a 3.0? |
| 00:46.56 | d3tul3 | rpierce99 is using as well, so perhaps he'll have a more unbiased opinion |
| 00:47.07 | d3tul3 | i got 39 up to .4 |
| 00:50.45 | jONpRY | are there more? |
| 00:54.00 | d3tul3 | no that |
| 00:54.02 | d3tul3 | s it |
| 00:55.23 | jONpRY | does 3.0 have a higher patch level? |
| 00:56.35 | d3tul3 | is the first decimal patchlevel? |
| 00:56.39 | d3tul3 | there's 3.1 stable |
| 00:56.47 | jONpRY | 3.0.patch |
| 00:56.57 | d3tul3 | 3.0.9 is last from nov 11 |
| 00:57.20 | jONpRY | need to upgrade to 3.1 just to continue :) |
| 00:57.30 | d3tul3 | that's how my math works |
| 00:58.18 | jONpRY | hmm. well we need an initial checkin |
| 00:58.43 | d3tul3 | bleeding edge ineded, google hasn't patched ashmem in their 3.0 repo |
| 01:00.23 | d3tul3 | s/ineded/indeed/ |
| 01:00.52 | jONpRY | can't give up the lead now |
| 01:01.03 | d3tul3 | lol |
| 01:01.58 | d3tul3 | you said thinking about +39 on the touchpad? |
| 01:02.56 | jONpRY | theres actually a lot different in the mach folder |
| 01:03.15 | jONpRY | might be more tractable with a mach-msm2 |
| 01:04.26 | d3tul3 | what 3.1? |
| 01:05.04 | jONpRY | no touchpro/touchpad |
| 01:05.24 | jONpRY | 3.1's gonna be a breeze |
| 01:05.52 | arrrghhh | jONpRY, the first one from goog error'd out |
| 01:05.55 | arrrghhh | cm is still goin |
| 01:06.01 | arrrghhh | error: Exited sync due to fetch errors |
| 01:06.33 | jONpRY | oh btw. i dissaprove of this importing patch history from the mainline updates |
| 01:06.51 | jONpRY | it makes it like impossible to see the work we did to actually make the mainline kernels work |
| 01:07.11 | jONpRY | arrrghhh, sounds about right |
| 01:08.22 | d3tul3 | jONpRY, what's better? any way to supress the history yet keep the commits differentiated |
| 01:08.47 | jONpRY | i used git diff to generate the patches |
| 01:09.01 | jONpRY | so its just one big blob |
| 01:09.23 | d3tul3 | makes it kind of hard when tracking individual revisions |
| 01:09.53 | jONpRY | i don't think so |
| 01:10.19 | jONpRY | anything that was good and happened in those commits will be in newer kernels anyways |
| 01:10.43 | d3tul3 | k, at the same time the commit history is compressed on the repo page (vs. the branch page) |
| 01:11.03 | d3tul3 | so these 870 commits don't klog the screen |
| 01:11.07 | jONpRY | i use gitk |
| 01:13.16 | *** join/#htc-linux mitsutaka (~mitsutaka@p2004-ipbf2206marunouchi.tokyo.ocn.ne.jp) |
| 01:15.11 | arrrghhh | 1.8 MiB/s |
| 01:15.12 | arrrghhh | not bad |
| 01:15.14 | arrrghhh | still syncing |
| 01:15.33 | jONpRY | its like 20gb of source |
| 01:15.36 | jONpRY | 80gb to build |
| 01:15.40 | arrrghhh | zomg |
| 01:15.43 | arrrghhh | 80gb to build |
| 01:15.47 | arrrghhh | i only have a 50gb /home... |
| 01:16.08 | jONpRY | well thats not going to work |
| 01:16.17 | arrrghhh | hahahaa |
| 01:16.21 | arrrghhh | i didn't think it would take more than 50gb... |
| 01:16.23 | arrrghhh | that's insane |
| 01:17.00 | jONpRY | that assumes that it actually will build of course |
| 01:17.07 | arrrghhh | touche |
| 01:23.40 | jONpRY | d3tul3, can you commit a 3.1.1 to a new branch? |
| 01:27.05 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
| 01:28.46 | d3tul3 | jONpRY, sure |
| 01:30.15 | jONpRY | someday i'll be able to run the same kernel on tp2 tp laptop and prypad |
| 01:30.33 | *** join/#htc-linux hardwalker (~hardwalke@122-117-115-146.HINET-IP.hinet.net) |
| 01:32.31 | *** join/#htc-linux Ondalf (~ondalf@cable-roi-fffddd00-9.dhcp.inet.fi) |
| 01:34.59 | WisTilt2 | jONpRY: this idle_collapse isn't going to work until we fix some of the msm_irq 'sit looks like. trimmed down the sleep minimum and got it to trigger but looks like a few irq's don't like it. I guess they all need to be wake capable or something like that? |
| 01:38.33 | *** join/#htc-linux NYL (~NYL@109.96.36.166) |
| 01:45.13 | *** join/#htc-linux TheXev (~user@18.sub-174-252-212.myvzw.com) |
| 01:51.32 | *** join/#htc-linux Ondalf (~ondalf@unaffiliated/ondalf) |
| 01:52.35 | d3tul3 | any particular wish for a name of the branch |
| 01:53.09 | d3tul3 | htc-msm-3.1.y it is |
| 01:59.48 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
| 02:05.39 | jONpRY | sounds good to me |
| 02:08.37 | d3tul3 | rpierce99, i forgot to mention earlier, camera won't work |
| 02:10.57 | *** join/#htc-linux surge (surge@pool-72-88-82-28.bflony.fios.verizon.net) |
| 02:11.28 | jONpRY | in kernels we realize you have a choice, and you made the wrong one |
| 02:26.33 | *** join/#htc-linux TheXev (~user@18.sub-174-252-212.myvzw.com) |
| 02:28.35 | rpierce99 | yeah i think i've used the camera once or twice |
| 02:36.24 | d3tul3 | jONpRY, ugh, branch is up commit history went up with it |
| 02:37.55 | d3tul3 | hopefully it's ok for an initial commit, nothing ours on the other end |
| 02:49.40 | *** join/#htc-linux XirXes (~XirXes@67-2-15-211.slkc.qwest.net) |
| 02:51.44 | d3tul3 | rpierce99, before i get out of here, what's the verdict |
| 02:52.05 | rpierce99 | i'm loving how fast it is |
| 02:52.16 | rpierce99 | i haven't seen any of the old slowdowns |
| 02:52.27 | rpierce99 | i haven't tried adb |
| 02:52.42 | rpierce99 | wifi worked great |
| 02:52.49 | d3tul3 | great |
| 02:52.53 | rpierce99 | bluetooth isn't enabled in this one is it? |
| 02:53.18 | d3tul3 | no WisTilt2's looking into that |
| 02:53.24 | rpierce99 | that's what i thought |
| 02:54.01 | rpierce99 | overall i haven't seen anything negative that i didn't fully expect going in |
| 02:54.28 | d3tul3 | cool jONpRY says he used to be able to kill it by running birds a few times |
| 02:54.39 | rpierce99 | haven't installed the birds |
| 02:54.41 | d3tul3 | i dont even know where to get birds |
| 02:54.49 | rpierce99 | hm, might have just gotten sluggish on me |
| 02:55.11 | rpierce99 | yeah it's really bogging down right now |
| 02:55.40 | d3tul3 | damn you can't even pull up shell to do dmesg |
| 02:55.51 | rpierce99 | it's weird though |
| 02:55.56 | rpierce99 | like the lock screen is responsive |
| 02:56.09 | rpierce99 | but hitting power, whether on or off, takes like 10 seconds |
| 02:56.48 | rpierce99 | like the fade to black on screen off is so slow you can see every frame |
| 02:57.18 | d3tul3 | that sounds like the gp_timer issue from back when |
| 02:58.08 | d3tul3 | you need some sort of memory widget |
| 02:58.22 | d3tul3 | to at least verify the mem killers are working as they should |
| 02:58.26 | rpierce99 | no surprise, adb install locked up |
| 02:58.35 | rpierce99 | tried to push getlogs to it |
| 02:59.11 | rpierce99 | you have a reco on a widget |
| 02:59.32 | d3tul3 | i use go launcher ex which has a built in one |
| 02:59.49 | rpierce99 | i installed adw for the first time, wonder if it has one |
| 03:00.06 | d3tul3 | it usually tracks available memory pretty well in the sense i can see when the mem killers aren't working available memory drops down in the teens |
| 03:00.18 | d3tul3 | currently (when they are working) i can't bring it down underm 40MB |
| 03:00.51 | rpierce99 | it's really weird, it's like its segregated into two types of actions, one is working great, others are slow as hell |
| 03:01.15 | d3tul3 | whichever ones are tied to that timer |
| 03:01.17 | rpierce99 | i can't tap and hold the screen at all, just doesn't work, power button actions are slow, but the menu button brings the menu up with a quickness |
| 03:01.22 | d3tul3 | i think the buttons |
| 03:01.46 | d3tul3 | try tap and holding the screen for a looooooong time |
| 03:02.07 | rpierce99 | yep, that did it |
| 03:02.36 | rpierce99 | no fancy widgets in adw |
| 03:02.54 | d3tul3 | stupid bug i think that guy went away -> it must be tied to the pmem issues....something (acoustic?) writes somewhere where it's not supposed to |
| 03:03.16 | d3tul3 | s/think/thought/ |
| 03:03.32 | d3tul3 | check your desktop clock |
| 03:03.48 | d3tul3 | it should be stopped at the point it hit you |
| 03:04.52 | rpierce99 | the analog clock widget? |
| 03:05.00 | d3tul3 | no just the clock in the taskbar |
| 03:05.43 | rpierce99 | oh, it says 8:03, it's 9:05 |
| 03:06.51 | d3tul3 | well i think WisTilt2 or jONpRY or someone smarter than me should take a look at those logs if it hits you again |
| 03:07.06 | rpierce99 | i'm still trying to get getlogs over to it |
| 03:09.56 | *** join/#htc-linux TheXev (~user@18.sub-174-252-212.myvzw.com) |
| 03:10.09 | d3tul3 | email? |
| 03:11.04 | d3tul3 | it hasn't happened to me in over a month - though only yesterday or the day before some code went back into timer.c ... |
| 03:11.11 | rpierce99 | clock says 8:05 now :) |
| 03:11.28 | rpierce99 | so it moves about 1/3 of real time |
| 03:11.50 | d3tul3 | oh really, i timed the ticks once, and i got something like 1/20 |
| 03:11.53 | *** join/#htc-linux pw (~pw@188.40.47.220) |
| 03:12.40 | rpierce99 | well, my sample size is very small, 2->6, yours was in ticks so it was orders of magnitude more accurate |
| 03:13.58 | d3tul3 | you don't recall getting a txt msg or anything during that time that would make acoustic engage? |
| 03:14.37 | rpierce99 | well i don't know when it actually kicked into slowness, because the clock is still moving a bit |
| 03:14.59 | rpierce99 | but i set my ringtone earlier, i've received several emails |
| 03:15.16 | rpierce99 | i'm on silent though |
| 03:15.26 | rpierce99 | not sure that matters for setting your ringtone though |
| 03:15.29 | *** join/#htc-linux ccube (ccube@bnc.lukius.de) |
| 03:19.39 | rpierce99 | looks like that dmesg goes all the way back to 0 actually |
| 03:22.20 | d3tul3 | well at least the memory killers are working |
| 03:22.40 | *** join/#htc-linux WisTilt2 (~wisgreg@wireless248.wirelesstcp.net) |
| 03:23.25 | rpierce99 | yep, is that pmem spam normal? |
| 03:24.04 | WisTilt2 | there's that dirty word again |
| 03:24.08 | WisTilt2 | pmem |
| 03:24.36 | d3tul3 | rpierce99, shoot that log to WisTilt2 |
| 03:24.48 | d3tul3 | WisTilt2, rpierce99 is suffering from the gp_timer issue |
| 03:24.56 | d3tul3 | that i've tried to explain and failed miserably |
| 03:25.10 | WisTilt2 | damn. thought that bug was history |
| 03:25.22 | *** join/#htc-linux vinceweis (0c912122@gateway/web/freenode/ip.12.145.33.34) |
| 03:25.59 | d3tul3 | so did i |
| 03:26.01 | rpierce99 | yeah this thing was flying for a while and then now the power button and touchscreen are sloooowwww |
| 03:26.03 | WisTilt2 | d3tul3 is it on the current tree? i haven't had any slowdowns whatsoever forever |
| 03:26.24 | d3tul3 | WisTilt2, this is on 3.0.9 which is more or less in sync with 39 |
| 03:26.43 | rpierce99 | should i try a straight .39? |
| 03:26.52 | WisTilt2 | ok good. we dont want .39 polluted:) thanks rpierce99 ill take a look at that log |
| 03:26.57 | d3tul3 | meh the nature of the beast is such that it could be days before it happens again |
| 03:27.16 | rpierce99 | k gotta run and pick up the wife, gl |
| 03:29.22 | WisTilt2 | pmem protection errors blew it up pretty good there |
| 03:29.46 | d3tul3 | yes i doubt this is something that's contained in 3.0.9 |
| 03:29.59 | WisTilt2 | d3tul3 same identical memory locations as we saw in .39 |
| 03:30.10 | jONpRY | d3tul3, so we're back to the timer thing? |
| 03:30.42 | d3tul3 | i guess we are pm with the log incoming |
| 03:30.43 | WisTilt2 | these boundaries are not right, tons of overlapping addresses. guess ill look at that tonight instead of BT, which im tired of anyway. |
| 03:32.39 | jONpRY | i dunno how to detect the timer problem without calculating the rate of timer interrupts |
| 03:33.08 | arrrghhh | jONpRY, sync done |
| 03:33.17 | arrrghhh | should i just be ballsy and build it? |
| 03:33.20 | WisTilt2 | looks like something was sending sigkill to all the apps during this massive pmem dump also |
| 03:34.11 | d3tul3 | WisTilt2, that's the lowmemorykiller that's fine |
| 03:34.23 | d3tul3 | that's android freeing up memory |
| 03:34.48 | jONpRY | arrrghhh, sure you can build it. i think the panda config is the only one that works |
| 03:35.00 | WisTilt2 | im going to go through all the pmem addresses again. this access_prot stuff popping up all over isnt right. |
| 03:35.03 | jONpRY | but on cm repo you probably have to lunch and stuff |
| 03:35.23 | d3tul3 | slowdown occured before 495 in his dmesg (that's when he plugged in his usb to get adb going) *i*guess* |
| 03:39.04 | WisTilt2 | it looks like it may have started shortly after he turned off wifi |
| 03:39.38 | WisTilt2 | thats where things looks like they started lagging and all the pmem stuff started soon after |
| 03:41.55 | arrrghhh | jONpRY, panda? i'm looking at the building from source page for CM, which device should look at? |
| 03:42.52 | WisTilt2 | d3tul3: strangely, all the huge delays seem to be centered around keypad brightness also |
| 03:43.40 | jONpRY | arrrghhh, so vanilla ics source only builds for panda. CM support is surreal at best. there is a repo for a nook device configuration that will "build" |
| 03:44.11 | arrrghhh | heh |
| 03:44.11 | arrrghhh | ok |
| 03:44.31 | jONpRY | https://github.com/fat-tire/android_device_bn_encore |
| 03:44.36 | d3tul3 | WisTilt2, hm, i had this happen to me back in the day when the led code was .27 stock |
| 03:44.48 | jONpRY | zonk whatever is in devicie/bn/encore and clone it into that dir |
| 03:45.12 | jONpRY | the you have to run ./extract-files.sh |
| 03:45.20 | d3tul3 | WisTilt2, the keypad brightness, that's him pressing buttons desperately trying to get a response probably |
| 03:45.27 | jONpRY | then from ics root. source build/envsetup.sh |
| 03:45.37 | d3tul3 | er or not |
| 03:45.46 | jONpRY | then lunch cyanogen_encore-eng |
| 03:45.54 | jONpRY | then make |
| 03:45.57 | arrrghhh | hrm |
| 03:46.13 | WisTilt2 | yeah but notice right after wifi off most all keypad stuff takes forever. prior to that it was quick. is keypad brightness in a work queue now? |
| 03:46.46 | jONpRY | dur. d3tul3 you imported the whole linux-next history into our branch? |
| 03:47.02 | jONpRY | this is not what i had in mind |
| 03:47.17 | d3tul3 | jONpRY, sorry about that mentioend it above |
| 03:47.32 | d3tul3 | i don't know how else other than git clone/git push |
| 03:47.39 | jONpRY | download tarball |
| 03:47.43 | d3tul3 | ah |
| 03:47.49 | d3tul3 | k i'll delete thte branch tomorrow |
| 03:48.20 | jONpRY | k |
| 03:49.18 | jONpRY | i dunno if i can even download this thing |
| 03:50.05 | jONpRY | it would be cool if someone getting the slow down could confirm the gp timer interrupt rate |
| 03:51.05 | jONpRY | i'm willing to assume that is the problem for now though |
| 03:51.06 | d3tul3 | as in precisely what it is? |
| 03:51.23 | jONpRY | confirm the rate it higher than should be possible |
| 03:51.32 | d3tul3 | he said his taskbar clock was moving really slow |
| 03:51.37 | jONpRY | should be 100hz. or less on a NO HZ kernel |
| 03:51.42 | d3tul3 | jONpRY, i have powertop dumps during slowdown |
| 03:51.52 | jONpRY | but iirc you were measuring thousands of hz |
| 03:52.16 | d3tul3 | what happs is gp_timer for me was firing at a rate of about 1/20 of 100HZ |
| 03:52.27 | jONpRY | oic |
| 03:52.46 | d3tul3 | vsync fires at a normal rate |
| 03:52.50 | jONpRY | with NOHZ its probably hard to measure |
| 03:52.56 | d3tul3 | this was prior to NOHZ |
| 03:56.06 | jONpRY | i'm willing to operate under the assumption that timer.c is the problem |
| 03:56.52 | jONpRY | and looking at the code i can't tell exactly what it does |
| 03:57.09 | jONpRY | at the very least there is opportunity for overflows |
| 03:58.29 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
| 04:00.02 | d3tul3 | how far is our timer.c from .27 or .35 timer.c? |
| 04:01.05 | jONpRY | now that i look at it. there are potentially bugs. but none that could possibly effect anything other than scbs |
| 04:01.50 | jONpRY | was rpierce99 running scbs? |
| 04:01.57 | rpierce99 | not scbsd, no |
| 04:02.03 | rpierce99 | unless it's in emwes rootfs |
| 04:04.09 | jONpRY | d3tul3, http://pastebin.com/nrMhGcp9 |
| 04:04.11 | d3tul3 | its not |
| 04:04.34 | jONpRY | there are other changes but they are just name changes to api stuff's. this is the thing thats supposedly killing it |
| 04:04.58 | d3tul3 | this is just the diff from your patch right? |
| 04:05.13 | jONpRY | yeah |
| 04:05.25 | jONpRY | and all it does it write otherwise unused variables |
| 04:05.54 | d3tul3 | i am looking at the diff against emwe's 35 timer.c and it looks more elaborate |
| 04:06.25 | jONpRY | i think this is a 27 timer |
| 04:07.00 | jONpRY | either way this is the patch i just put in that apparently broke it |
| 04:08.05 | jONpRY | although i think we are calling msm_timer_pcom_time() from a potentially interesting context |
| 04:09.52 | d3tul3 | does this resume mode even get called if you are not running the daemon? |
| 04:11.05 | d3tul3 | jONpRY, yeah i don't see where in that patch you are messing with the existing code |
| 04:12.02 | jONpRY | the resume will get called on resume of the device, but it just modified stuff that doesn't get used by anyone |
| 04:13.31 | *** join/#htc-linux Rob2223 (~Miranda@p4FFF2BE0.dip.t-dialin.net) |
| 04:14.01 | jONpRY | arrrghhh, how's that build? |
| 04:15.12 | arrrghhh | haven't done much |
| 04:15.19 | arrrghhh | damn sister called and consumed some time |
| 04:15.39 | arrrghhh | that, and i'm not quite sure what you mean. but i don't know my ass from a hole in the ground when it comes to building. |
| 04:17.14 | jONpRY | i've never actually built it either |
| 04:17.33 | jONpRY | but i think trying to effect a armv6 build would probably be more fun |
| 04:17.51 | arrrghhh | heh |
| 04:19.47 | jONpRY | looks like no armv6 device paths in manifest |
| 04:19.55 | jONpRY | dur |
| 04:20.20 | jONpRY | try doing a source build/envsetup.sh; lunch |
| 04:21.42 | arrrghhh | full_panda-eng? |
| 04:21.57 | arrrghhh | or just full-eng |
| 04:22.02 | jONpRY | is there a cyanogen_panda-eng ? |
| 04:22.06 | arrrghhh | no |
| 04:22.13 | jONpRY | then full it is |
| 04:22.15 | arrrghhh | k |
| 04:22.39 | arrrghhh | make? |
| 04:22.43 | jONpRY | yup |
| 04:22.49 | jONpRY | -j whatev |
| 04:22.57 | arrrghhh | crap forgot that |
| 04:23.02 | arrrghhh | i should be doing this on my laptop too |
| 04:23.06 | arrrghhh | i can do -j32 or something |
| 04:23.22 | arrrghhh | eh screw it |
| 04:23.29 | jONpRY | thats a lot of thread no? |
| 04:23.29 | arrrghhh | already on the server |
| 04:23.32 | arrrghhh | might as well build on it |
| 04:23.39 | *** join/#htc-linux NeoMatrixJR (~chatzilla@173-20-63-62.client.mchsi.com) |
| 04:23.44 | jONpRY | there is also something called mmake or mmmake |
| 04:23.46 | arrrghhh | well stine said cores + 1 |
| 04:23.57 | arrrghhh | and somewhere else i read you should calculate it differently |
| 04:24.03 | arrrghhh | and they said for my proc -j32 |
| 04:24.09 | arrrghhh | i dunno, it worked. compiled really freaking fast. |
| 04:24.27 | arrrghhh | pegged out all cores too :D |
| 04:24.39 | arrrghhh | wtf |
| 04:24.42 | arrrghhh | all this froyo mess |
| 04:25.02 | arrrghhh | http://pastebin.com/7aDkVNKE |
| 04:25.07 | jONpRY | that does a good job of keeping processors busy unless every thread has a large memory load |
| 04:25.44 | arrrghhh | hrm. it didn't put everything in the Android_CM9 directory i made either |
| 04:25.50 | arrrghhh | i might have to resync. fawk. |
| 04:30.16 | jONpRY | dur |
| 04:30.57 | jONpRY | i usually run -j3 on the quad ht machine so it doesn't overheat too much |
| 04:31.03 | arrrghhh | heh |
| 04:31.21 | jONpRY | which is part of why i need build server |
| 04:31.41 | arrrghhh | definitely worth it |
| 04:31.42 | arrrghhh | for me |
| 04:31.54 | arrrghhh | for just building, offsite managed might be better |
| 04:31.57 | arrrghhh | dunno |
| 04:32.03 | jONpRY | mine will just turn off |
| 04:32.29 | jONpRY | even j3 is risky. i can't switch desktops, or open youtube |
| 04:32.42 | arrrghhh | damn |
| 04:33.14 | arrrghhh | alright, i should go eat. imma set this thing to sync again. |
| 04:33.21 | arrrghhh | hopefully i sync it correctly this time... |
| 04:33.28 | jONpRY | k |
| 04:33.43 | arrrghhh | you can live vicariously through my server :P |
| 04:35.02 | arrrghhh | where |
| 04:35.04 | arrrghhh | whew* |
| 04:35.09 | arrrghhh | 102gb free |
| 04:35.12 | arrrghhh | hopefully that'll be enough.. |
| 04:35.15 | jONpRY | lol |
| 04:36.46 | arrrghhh | ok, here's the damned problem |
| 04:36.52 | arrrghhh | repo initialized in /home/user |
| 04:37.00 | arrrghhh | but i'm in /home/user/Android_CM9 |
| 04:37.09 | arrrghhh | that's where i'm submitting the repo init command from. wtf? |
| 04:37.25 | jONpRY | make sure there is no .repo directory up a level |
| 04:37.29 | arrrghhh | ok |
| 04:37.31 | arrrghhh | ls |
| 04:37.33 | arrrghhh | oops |
| 04:37.51 | arrrghhh | there is... |
| 04:37.55 | arrrghhh | remove it? |
| 04:38.17 | jONpRY | moving it might potentially cut down on the amount of syncing |
| 04:38.25 | jONpRY | depending on wtf it is |
| 04:38.28 | arrrghhh | lol |
| 04:38.30 | arrrghhh | imma remove it |
| 04:38.39 | arrrghhh | don't want to mix stuff up |
| 04:39.03 | jONpRY | for sure |
| 04:39.11 | arrrghhh | boom! |
| 04:39.17 | arrrghhh | thanks dude. repo sync, food tiem. |
| 04:41.33 | d3tul3 | the lack of debugging info in this timer issue is crippling |
| 04:42.57 | jONpRY | or lack of concrete data. like was it really gone? |
| 04:44.50 | d3tul3 | ugh i swear when i was hunting for this, i once turned on all possible (somewhat relevant) debugging masks timer.c clocks.c pm.c etc...and then it didn't happen for a week |
| 04:46.49 | jONpRY | its some kind of collision |
| 04:46.57 | jONpRY | missing mutex? |
| 04:48.23 | d3tul3 | jONpRY, new branch is up |
| 04:48.42 | jONpRY | oh cool |
| 04:50.58 | jONpRY | presumably the kernel isn't broken and really has the thing programmed for 100hz |
| 04:51.56 | jONpRY | so i would say either timer.c is broken, the timers clock is being changed, or some interrupt handler is leaving interrupts disables for way over 10ms |
| 04:53.36 | jONpRY | something to verify any of that would be outstanding |
| 04:53.56 | d3tul3 | :) |
| 04:56.49 | jONpRY | i dunno what that sync_smem clock stuff is |
| 04:56.57 | jONpRY | i don't think we even have an smem clock |
| 04:57.33 | d3tul3 | is there any way to verify what uses 0x24E00000 ? |
| 04:57.48 | d3tul3 | and why pmem is so damn eager to map it |
| 04:58.01 | jONpRY | you can print the process information in the ioctl |
| 04:58.18 | jONpRY | but its always systemserver :p |
| 04:59.31 | jONpRY | what is 24e? |
| 04:59.44 | d3tul3 | the target addresses also cycle |
| 04:59.48 | jONpRY | i guess an sdram region? |
| 05:02.55 | WisTilt2 | d3tul3 you got a camera over here or what? that is exactly what i am working on and looking at right this moment! |
| 05:03.27 | WisTilt2 | that address is ebin yes, why it is always using that addr makes no sense |
| 05:05.32 | WisTilt2 | jonpry doesn't .39 use write combined for buffered? cached and buffered never happens, which is why we keep seeing that access prot spam |
| 05:06.15 | jONpRY | what pmem region is 24e? |
| 05:07.00 | WisTilt2 | its ebin sdram |
| 05:07.48 | WisTilt2 | if you're asking which region node i have no idea |
| 05:07.59 | jONpRY | yeah that what i am asking |
| 05:08.18 | jONpRY | there is nothing wrong with pmem_access_prot. just remove the printk |
| 05:08.38 | jONpRY | memory is default cached |
| 05:08.44 | d3tul3 | WisTilt2, looking at that prot message 0, 1, 0 corresponds to id,pmem[id].cached,pmem[id].buffered |
| 05:08.49 | jONpRY | so cached==0, means set no cached |
| 05:08.51 | d3tul3 | so i guess it's cached |
| 05:08.58 | WisTilt2 | yeah i see that now. so we should never see buffered or cached at all? |
| 05:09.00 | jONpRY | buffer==1 means write combine |
| 05:09.15 | d3tul3 | id = get_file(id), and that is also zero which seems somewhat strange |
| 05:09.23 | WisTilt2 | same here |
| 05:10.21 | WisTilt2 | it looks like 0x24exxx is part of ebin that we have camera adsp and other things in |
| 05:11.37 | jONpRY | 2*1025*1023??? |
| 05:12.22 | jONpRY | 24e is MSM_EBIN_BASE+128*1024*1024-50*1024*1024 |
| 05:13.42 | jONpRY | how does this work? CALC_PMEM(fb, pmem_camera, 2*1024*1024); |
| 05:13.47 | jONpRY | cause fb is not setup yet |
| 05:14.05 | jONpRY | maybe it is but not obviously from that point in the code anyways |
| 05:14.09 | WisTilt2 | yeah it is, uses 2mb right above the camera |
| 05:14.41 | jONpRY | ic |
| 05:14.52 | d3tul3 | so that's fb memory? |
| 05:14.54 | WisTilt2 | i dont like that logic myself |
| 05:15.08 | WisTilt2 | where are you seeing 2*1025*1023? |
| 05:15.33 | jONpRY | its for raphael |
| 05:15.54 | WisTilt2 | 24e would be the 78mb base addr of pmem |
| 05:15.59 | WisTilt2 | the 32mb block |
| 05:17.33 | jONpRY | i think that printk is misleading |
| 05:17.42 | jONpRY | cause it doesn't take into account the region data |
| 05:18.05 | WisTilt2 | yeah the prot one is for sure and can be removed. |
| 05:18.14 | jONpRY | no the other one |
| 05:18.25 | WisTilt2 | access_prot is pointless |
| 05:18.32 | WisTilt2 | it always gets there |
| 05:18.38 | jONpRY | %s: mmaping 0x%lX x 0x%lX to: 0x%lX\n is all wrong |
| 05:18.52 | jONpRY | cause it ignores region_node->region.offset, |
| 05:18.53 | jONpRY | <PROTECTED> |
| 05:19.00 | WisTilt2 | yeah that one too, but with it we wouldn't see this 24e issue |
| 05:19.13 | jONpRY | the issue might not exist |
| 05:19.42 | jONpRY | because it can ex: mmap pmem_base(0x24e*) + 0x31200 or something and only 4k bytes |
| 05:20.04 | jONpRY | but all the log messages will only show pmem_base and 32MB |
| 05:20.10 | WisTilt2 | what i dont get is the size 0x2xxx overlaps many of the vm mapped addresses at times and that's when camera crashes |
| 05:20.20 | jONpRY | cause that is not how much is being mapped |
| 05:20.42 | jONpRY | it only maps region_node->region.len |
| 05:20.55 | jONpRY | not pmem_len(id,data) |
| 05:22.09 | WisTilt2 | ok i can see that. |
| 05:22.36 | jONpRY | and also the base address can have the offset |
| 05:23.00 | jONpRY | so this stuff might be perfectly reasonable |
| 05:23.06 | jONpRY | might be like 4 different pages |
| 05:24.25 | WisTilt2 | while you're looking at this - you see where pmem_gpu1 is setup. do the math on it and tell me how the 8mb allocated for it is not also hitting the console start addr? this has bugged me for a long time. |
| 05:25.24 | jONpRY | cause console is in SMI1? |
| 05:25.35 | WisTilt2 | so is gpu1 |
| 05:25.39 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
| 05:25.45 | jONpRY | gpu1=MSM_EBI_BASE+99*1024*1024 |
| 05:26.20 | jONpRY | you mean gpu0? |
| 05:26.55 | WisTilt2 | <PROTECTED> |
| 05:26.59 | WisTilt2 | yeah thats smi |
| 05:27.28 | WisTilt2 | let me look at this again. this address was within the top of something also allocated here i thought |
| 05:27.42 | WisTilt2 | yeah gpu0 |
| 05:29.29 | jONpRY | well gpu0 = 7MB, size = 1MB, ram console is at 8.e MB |
| 05:32.03 | WisTilt2 | ok so console is the top 128k of the 8mb base got it. somehow a long time ago i was stuck on that being in the same area as gpu0. |
| 05:34.40 | WisTilt2 | so back on this mmap, we can remove those printks since they're kind of pointless. the other spam about clashes with overlapping memory will still pop up with the camera so we'll have to see whats causing that |
| 05:36.30 | jONpRY | the one printk should be modified to show the actual region being mmap'd. there still may be something interesting there |
| 05:37.01 | WisTilt2 | i agree. you want to do that or i can |
| 05:40.07 | jONpRY | try this http://pastebin.com/vn8nT1WS |
| 05:40.15 | jONpRY | my git is still checking out the new branch |
| 05:42.21 | WisTilt2 | whats this a complete replacement for the dream/pmem.c? |
| 05:43.43 | jONpRY | yes |
| 05:44.16 | WisTilt2 | will it pop right in, no header files etc? |
| 05:44.29 | jONpRY | no. i forgot to remove that one printk though |
| 05:46.43 | WisTilt2 | the access_prot ok. ill give it a run |
| 05:48.17 | arrrghhh | jONpRY, attempting a make again |
| 05:48.35 | arrrghhh | hrm, i wonder if alex's .27 acoustic tree is up-to-date |
| 05:50.28 | jONpRY | from there i guess we can clone the panda device and try to mod for armv6 |
| 05:50.40 | arrrghhh | we'll see if it builds |
| 05:50.57 | jONpRY | it will build but unless you have a pandaboard :p |
| 05:51.13 | jONpRY | i have a beagle which i think could work with minimal voodoo |
| 05:51.44 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-68-237.dynamic.mnet-online.de) |
| 05:51.51 | arrrghhh | heh |
| 05:51.55 | jONpRY | if it ever works on msm7k it will be with software rendering, ugh |
| 05:52.08 | arrrghhh | balls |
| 05:52.29 | arrrghhh | yea, with all the eyecandy in ics |
| 05:52.35 | arrrghhh | i don't have high hopes for it on the rhod... |
| 05:52.47 | arrrghhh | or any pre-adreno200 device for that matter |
| 05:53.46 | jONpRY | well candy is one thing. but i think we will lose stuff we have. like birds |
| 05:53.54 | jONpRY | no accel ever |
| 05:54.06 | arrrghhh | oh cuz of the new surfaceflinger? |
| 05:55.11 | jONpRY | yeah something like that |
| 05:55.15 | arrrghhh | lol |
| 05:55.56 | jONpRY | there is a lot of talk about what won't work. and people running around trying to upgrade this and that, but really nobody has tried much of anything |
| 05:56.06 | arrrghhh | epic |
| 05:56.24 | arrrghhh | gotta love speculation |
| 05:56.33 | arrrghhh | after all, our entire economy lives and dies by it... |
| 05:57.38 | jONpRY | heh |
| 05:57.58 | EdLin | arrrghhh by pre-adnero200 devices, you're excluding the HD2 as well, right? |
| 05:57.58 | arrrghhh | mostly dies |
| 05:58.06 | arrrghhh | EdLin, eh? |
| 05:58.06 | EdLin | adreno* |
| 05:58.16 | arrrghhh | doesn't that have the 205 or smth? |
| 05:58.23 | arrrghhh | doesn't know jack about the HD2 |
| 05:58.24 | EdLin | arrrghhh no, some ATI thingie. |
| 05:58.27 | arrrghhh | other than it's a snapdragon |
| 05:58.39 | arrrghhh | oh, well then you can really even compare :P |
| 05:58.50 | arrrghhh | apples & oranges. |
| 05:58.52 | *** join/#htc-linux XeKToReX (~xTx@CPE-124-188-18-211.hdbq1.win.bigpond.net.au) |
| 05:58.56 | arrrghhh | i'd hope the HD2 could run it |
| 05:59.02 | arrrghhh | just like the N1 |
| 05:59.05 | arrrghhh | should run it fine |
| 05:59.10 | EdLin | arrrghhh probably can, but the GPU is slower. |
| 05:59.12 | arrrghhh | but goog isn't supporting it for some reason. |
| 05:59.19 | arrrghhh | EdLin, can't be slower than the TP2 :P |
| 05:59.20 | jONpRY | anything gles2 should work |
| 05:59.27 | arrrghhh | ^^ |
| 05:59.34 | EdLin | it's got that. |
| 05:59.38 | arrrghhh | tp2 doesn't. |
| 05:59.52 | EdLin | tp2 has teh keyboard awesomesauce tho. |
| 06:00.01 | arrrghhh | why couldn't htc have made the tp2 & hd2 as sister devices |
| 06:00.05 | arrrghhh | that would've been tits |
| 06:00.14 | jONpRY | i'm guessing there is a software compositor |
| 06:00.15 | arrrghhh | they came out just a few months apart if i'm not mistaken |
| 06:00.26 | EdLin | arrrghhh yeah, they had to wait for WP7 to do something like that, and WP7 sucks keyboard-wise. |
| 06:00.45 | jONpRY | i was thinking of doing pcb interposers that would let you upgrade the cpu on a phone |
| 06:00.58 | arrrghhh | EdLin, lies, they're just cheap bastards |
| 06:01.08 | arrrghhh | recycled the tp1 hardware and upgraded the externals. voila, tp2. |
| 06:01.22 | EdLin | arrrghhh that's true too, they were using G1/tp1 class hardware for the longest time. |
| 06:01.27 | arrrghhh | yup |
| 06:01.45 | arrrghhh | hrm |
| 06:01.59 | arrrghhh | i wonder if i can still watch stuff while i'm building |
| 06:02.04 | arrrghhh | all 4 cores pegged, probably not. |
| 06:02.14 | jONpRY | is this on your laptop? |
| 06:02.16 | EdLin | arrrghhh building Ice Cream? |
| 06:02.17 | arrrghhh | nah |
| 06:02.21 | arrrghhh | laptop has 8 |
| 06:02.23 | arrrghhh | 4 logical |
| 06:02.25 | arrrghhh | EdLin, yea |
| 06:02.29 | arrrghhh | well |
| 06:02.30 | arrrghhh | cm9 |
| 06:02.45 | EdLin | that'll take time, what device do you have it planned for? |
| 06:03.01 | arrrghhh | eh, just messing around really |
| 06:03.10 | arrrghhh | jONpRY's gonna give me $200 if it builds |
| 06:03.14 | EdLin | arrrghhh watch out, that way lies Gentoo. |
| 06:03.20 | arrrghhh | haha |
| 06:03.26 | jONpRY | i'm vicariously building |
| 06:03.28 | WisTilt2 | jONpRY: nice! makes a hell of a lot more sense with the correct addresses and sizes. |
| 06:03.32 | arrrghhh | hey i can watch videos while its building |
| 06:03.49 | arrrghhh | i'm sure it couldn't transcode, but who cares. life's too short to transcode i've decided. |
| 06:04.19 | WisTilt2 | push it or would you like to do the honors |
| 06:05.01 | arrrghhh | jONpRY, the source didn't take nearly 20gb |
| 06:05.17 | arrrghhh | took like, 6 maybe. i didn't look at it before i started building. |
| 06:05.25 | arrrghhh | not even 6 |
| 06:05.53 | arrrghhh | EdLin, to answer your question, i'm building for pandaboard |
| 06:06.32 | arrrghhh | i guess this means i need to go get one... |
| 06:06.36 | EdLin | arrrghhh lol |
| 06:07.08 | jONpRY | panda is cool |
| 06:07.21 | jONpRY | if not portable. gb of ram. dual core a9 omap4 |
| 06:07.28 | jONpRY | hdmi, wifi n |
| 06:07.33 | jONpRY | $99 or something |
| 06:07.37 | EdLin | I was kind of hoping someone was working on it for the HD2, development community seems kind of dead on XDA for it. |
| 06:08.01 | EdLin | my computer can't even build ICS at present. |
| 06:08.16 | arrrghhh | i don't think my server _should_ be building it |
| 06:08.19 | arrrghhh | but it is, so far lol |
| 06:08.25 | EdLin | lol |
| 06:08.42 | arrrghhh | is it sad that my laptop has way better specs than my server? |
| 06:08.50 | arrrghhh | oh well, next upgrade will go to the server. |
| 06:08.58 | EdLin | arrrghhh some day we'll be saying that about our phones. |
| 06:09.03 | arrrghhh | nj |
| 06:09.10 | arrrghhh | i remember my first smartphone |
| 06:09.16 | EdLin | arrrghhh unless that is, you intend to run the TP2 forever. :) |
| 06:09.25 | arrrghhh | i was rambling about how computers back in the day dreamt of 200mhz procs |
| 06:09.30 | arrrghhh | EdLin, i just might.... |
| 06:09.32 | jONpRY | i do |
| 06:09.33 | arrrghhh | is stubborn |
| 06:09.58 | arrrghhh | trying to think of what speed/proc my first phone had... |
| 06:10.00 | arrrghhh | googles |
| 06:10.04 | jONpRY | arrrghhh, how is your build server going to work :p. downloads from github and autobuild? |
| 06:10.35 | EdLin | arrrghhh yeah, HTC needs an Android phone with a keyboard like that. Microsoft has their keyboard team whipped. |
| 06:10.35 | arrrghhh | jONpRY, sure? |
| 06:11.00 | arrrghhh | EdLin, i think too many people want thin & light & cheap. IE no kb's. |
| 06:11.07 | jONpRY | thats going to be harder to setup than ssh, but might have its +'s |
| 06:11.08 | arrrghhh | cuz most phones, let's be honest, have awful keyboards. |
| 06:11.19 | arrrghhh | jONpRY, wat |
| 06:11.24 | arrrghhh | oh setting up an autobuild? |
| 06:11.27 | jONpRY | yeah |
| 06:11.38 | jONpRY | cron job i guess |
| 06:11.44 | arrrghhh | 'tis always on |
| 06:11.55 | jONpRY | then it would need a webserver or something to spit out the logs and products |
| 06:12.05 | arrrghhh | done |
| 06:12.12 | arrrghhh | already running that :P |
| 06:12.15 | arrrghhh | among other things |
| 06:12.18 | jONpRY | and really i try to avoid do a full rebuild |
| 06:12.28 | arrrghhh | cuz it takes days? |
| 06:12.34 | jONpRY | yup |
| 06:12.51 | jONpRY | but the build system is often too stupid to figure out what needs built |
| 06:12.57 | jONpRY | and needs a little coaxing |
| 06:13.05 | arrrghhh | doesn't sound automated |
| 06:13.31 | jONpRY | well if you could send it targets somehow |
| 06:13.43 | jONpRY | like make clean-libhardware |
| 06:14.51 | arrrghhh | 416 MHz PXA270 |
| 06:14.58 | jONpRY | actually that just isn't going to work |
| 06:15.11 | arrrghhh | clock speed of my first smartphone to my current isn't that wide. i do have an old smartphone, but not _that_ ol.d |
| 06:15.14 | arrrghhh | der |
| 06:15.18 | jONpRY | cause there will be framework changes and stuff |
| 06:15.52 | EdLin | arrrghhh there's more than just clock speed, there are some 800Mhz phones that outperform a 1st gen Snapdragon at 1GHz. |
| 06:15.53 | jONpRY | and atm i can't even look at the source cause there is no gitweb :( |
| 06:16.18 | arrrghhh | EdLin, true. lots of different innovations |
| 06:16.26 | arrrghhh | just funny to think about |
| 06:16.43 | arrrghhh | jONpRY, wth? do you just have to download it all? |
| 06:16.50 | jONpRY | yeah |
| 06:16.58 | arrrghhh | hah |
| 06:17.03 | EdLin | arrrghhh my first computer was 16Mhz. :) |
| 06:17.06 | arrrghhh | 16 |
| 06:17.09 | arrrghhh | holy poop |
| 06:17.13 | arrrghhh | mine was 133mhz |
| 06:17.24 | jONpRY | i had one of those, a 386sx |
| 06:17.26 | arrrghhh | although my dad still has an apple IIe from 1982 |
| 06:17.32 | EdLin | arrrghhh not counting the Apple II I hogged at my late step-grandfather's house, which was less than 2MHz. |
| 06:17.40 | arrrghhh | 1mhz w/a 8mhz "zip chip" |
| 06:17.41 | jONpRY | later i got some more ghetto shit. like kapro z80 |
| 06:18.09 | EdLin | jONpRY Kaypro came with a lot of software. |
| 06:18.32 | arrrghhh | whew, at least i'm remembering to do all this in a screen session. |
| 06:18.34 | jONpRY | my grandfather gave it to me, when i had a 486dx2 |
| 06:18.40 | jONpRY | had cp/m |
| 06:18.46 | EdLin | yep |
| 06:18.51 | jONpRY | 320kb flops |
| 06:18.56 | WisTilt2 | nothing beats the 4mhz 8086 or 8mhz Z80:) nite fellas catch ya tomorrow. |
| 06:18.59 | EdLin | Wordstar and a whole bunch of other stuff was included tho |
| 06:19.05 | EdLin | WisTilt2 lol :) |
| 06:19.34 | EdLin | too late |
| 06:20.13 | arrrghhh | yea that old stuff made programmers be frugal |
| 06:20.36 | *** join/#htc-linux mitsutaka (~mitsutaka@p2004-ipbf2206marunouchi.tokyo.ocn.ne.jp) |
| 06:23.38 | jONpRY | arrrghhh, still building? |
| 06:23.54 | arrrghhh | all 4 cores pegged |
| 06:24.03 | arrrghhh | while watching the new mythbusters |
| 06:24.03 | arrrghhh | win |
| 06:24.27 | arrrghhh | i see a lot of out/target's |
| 06:24.32 | arrrghhh | so i assume it's getting near the end |
| 06:24.41 | jONpRY | lol |
| 06:24.56 | arrrghhh | bad assumption? :P |
| 06:25.13 | jONpRY | its probably a good half way when it starts with java |
| 06:25.54 | arrrghhh | yea, i love it. absolutely no indication of progress |
| 06:26.22 | arrrghhh | my load is @ 6.00 |
| 06:26.22 | arrrghhh | lol |
| 06:39.52 | arrrghhh | i see java stuffs |
| 06:40.01 | arrrghhh | i'm guessing i'm a good half way thru wow... |
| 06:40.10 | arrrghhh | the load has dropped |
| 06:42.46 | arrrghhh | perhaps i should stop staring at it |
| 06:43.10 | jONpRY | lol |
| 06:57.29 | *** join/#htc-linux kiozen (~kiozen@p54BB7ADD.dip.t-dialin.net) |
| 07:00.33 | arrrghhh | jONpRY, Created filesystem with 1086/16384 inodes and 35441/65536 blocks |
| 07:00.33 | arrrghhh | Install system fs image: out/target/product/panda/system.img |
| 07:00.33 | arrrghhh | out/target/product/panda/system.img+ total size is 143017684 |
| 07:00.39 | arrrghhh | :D |
| 07:01.18 | arrrghhh | just over an hour |
| 07:02.08 | arrrghhh | ls |
| 07:02.11 | arrrghhh | damnit |
| 07:03.16 | arrrghhh | ~140mb? |
| 07:03.37 | arrrghhh | alright, this should boot on the rhod, right? :P |
| 07:04.32 | jONpRY | i wish |
| 07:04.41 | jONpRY | but cooel |
| 07:05.00 | arrrghhh | you want a linky to the system.img? |
| 07:05.29 | jONpRY | panda system.img is not terribly useful |
| 07:05.40 | arrrghhh | didn't think so |
| 07:05.52 | arrrghhh | just a experiment to see if it would work eh? :P |
| 07:06.53 | jONpRY | want to try something else? |
| 07:06.59 | arrrghhh | sure, what's up? |
| 07:07.19 | jONpRY | you have a device/ti/panda/BoardConfig.mk? |
| 07:07.56 | arrrghhh | yessir |
| 07:08.13 | jONpRY | probably best if you pastebin that, but look for TARGET_BOARD_PLATFORM |
| 07:08.23 | jONpRY | er. TARGET_ARCH_VARIANT |
| 07:08.33 | jONpRY | change to armv6j |
| 07:08.35 | arrrghhh | http://pastebin.com/4cz5DRFV |
| 07:09.16 | arrrghhh | TARGET_ARCH_VARIANT := armv7-a-neon |
| 07:09.44 | jONpRY | yeah change that |
| 07:09.52 | arrrghhh | TARGET_CPU_ABI := armeabi-v7a |
| 07:09.53 | jONpRY | and ABI := armeabi |
| 07:09.54 | arrrghhh | what about htat |
| 07:10.06 | jONpRY | yeah that |
| 07:10.14 | jONpRY | no USE_OPENGL_RENDERER |
| 07:10.15 | arrrghhh | cool |
| 07:10.22 | arrrghhh | for abi? |
| 07:10.44 | jONpRY | TARGET_CPU_ABI := armeabi |
| 07:11.26 | arrrghhh | k |
| 07:11.35 | jONpRY | make clean; make |
| 07:12.32 | jONpRY | also the system.img won't be enough when its done. need the rootfs components |
| 07:12.43 | arrrghhh | TARGET_BOARD_PLATFORM := omap4 |
| 07:12.45 | arrrghhh | what about hat |
| 07:12.51 | arrrghhh | that* |
| 07:13.56 | jONpRY | we're not prepared to change that |
| 07:14.00 | arrrghhh | k |
| 07:14.18 | arrrghhh | build/core/combo/TARGET_linux-arm.mk:39: *** Unknown ARM architecture version: armv6. Stop. |
| 07:14.30 | arrrghhh | they dropped support for it IIRC |
| 07:14.31 | jONpRY | armv6j ? |
| 07:14.51 | arrrghhh | same |
| 07:15.08 | jONpRY | can you show me build/core/combo/TARGET_linux-arm.mk |
| 07:15.14 | arrrghhh | 1 sec |
| 07:15.57 | arrrghhh | http://pastebin.com/R2tbVwRV |
| 07:16.04 | jONpRY | android is such a douche distro |
| 07:16.09 | arrrghhh | lol |
| 07:16.33 | arrrghhh | it does seem like a PITA |
| 07:17.08 | jONpRY | try armv5te |
| 07:17.49 | jONpRY | there is something called $(BUILD_COMBOS)/arch/<name>.mk |
| 07:18.01 | jONpRY | guessing that is build/arch/ |
| 07:20.36 | arrrghhh | that.... seemed to work |
| 07:20.43 | arrrghhh | "Entire build directory removed." |
| 07:20.50 | arrrghhh | make tiem |
| 07:20.51 | arrrghhh | crap |
| 07:21.15 | arrrghhh | forgot to -j5 |
| 07:21.17 | arrrghhh | heh |
| 07:21.23 | arrrghhh | will it take another hour+? |
| 07:21.31 | jONpRY | for sure |
| 07:21.43 | arrrghhh | damn. |
| 07:21.48 | arrrghhh | i have to go to bed... |
| 07:21.48 | jONpRY | unless it breaks. which i guess is like 99% |
| 07:21.55 | arrrghhh | lol |
| 07:22.02 | jONpRY | thats cool |
| 07:22.15 | jONpRY | just let it blow up. deal with it tomorrow |
| 07:22.17 | arrrghhh | this is fun tho |
| 07:22.19 | arrrghhh | indeed |
| 07:22.26 | arrrghhh | i have tomorrow + next week off too |
| 07:22.39 | arrrghhh | i had a bunch of PTO, like 15 days. |
| 07:22.40 | arrrghhh | had to use it |
| 07:22.45 | arrrghhh | so i've been taking every friday off |
| 07:22.50 | arrrghhh | and this whole next week |
| 07:22.59 | arrrghhh | and the last week in december |
| 07:23.06 | jONpRY | for cellphone hacking? |
| 07:23.18 | jONpRY | thats probably what i would be doing |
| 07:23.23 | arrrghhh | bionic/libc/bionic/pthread.c:953: warning: implicit declaration of function 'ARM_SMP_defined_but_no_DMB' |
| 07:23.40 | arrrghhh | well there is a lot of things i |
| 07:23.47 | arrrghhh | *'ve been meaning to test |
| 07:23.52 | jONpRY | pff warnings shmarnings |
| 07:24.02 | arrrghhh | oh that was just a warning |
| 07:24.08 | arrrghhh | hrm i dunno why it crashed |
| 07:24.11 | arrrghhh | oh there's the error |
| 07:24.21 | arrrghhh | /tmp/ccuduyLS.s:36: Error: selected processor does not support `mrc p15,0,r0,c13,c0,3' |
| 07:24.43 | arrrghhh | darn those mrc p15,0,r0,c13,c0,3 bastards |
| 07:24.52 | jONpRY | need more context |
| 07:25.00 | arrrghhh | 1 sec |
| 07:25.13 | arrrghhh | shit |
| 07:25.21 | arrrghhh | screen really screws up the scrollback |
| 07:25.23 | arrrghhh | i forget about htat |
| 07:25.25 | arrrghhh | that* |
| 07:25.40 | arrrghhh | http://pastebin.com/pJVfHNvn |
| 07:25.41 | jONpRY | the supported arch's are in ./build/core/combo/arch/arm/ |
| 07:26.30 | jONpRY | probably have to rebuild without -j. i can't tell what source caused it |
| 07:26.36 | arrrghhh | hrm |
| 07:26.36 | arrrghhh | ok |
| 07:26.39 | arrrghhh | i'll set that up |
| 07:26.43 | arrrghhh | bedtime in the meantime ;) |
| 07:26.54 | jONpRY | k |
| 07:27.19 | arrrghhh | thx man, ttyl |
| 07:30.24 | *** join/#htc-linux Dave888 (~Dave888@41.230.58.85) |
| 07:43.09 | *** join/#htc-linux Ondalf (~ondalf@cable-roi-fffddd00-9.dhcp.inet.fi) |
| 07:45.22 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-bzkdegriudcodxrg) |
| 07:46.00 | *** join/#htc-linux XeKToReX (~xTx@CPE-124-188-18-211.hdbq1.win.bigpond.net.au) |
| 07:50.37 | *** join/#htc-linux TheXev (~user@18.sub-174-252-212.myvzw.com) |
| 08:10.53 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 08:26.31 | *** join/#htc-linux BabelO (~fcr@unaffiliated/babelo) |
| 08:26.36 | *** join/#htc-linux rzk (~rzk@95-25-119-53.broadband.corbina.ru) |
| 08:32.47 | *** join/#htc-linux rob_w (~bob@ppp-188-174-6-194.dynamic.mnet-online.de) |
| 08:52.34 | *** join/#htc-linux gauner1986 (~Miranda@87.253.171.212) |
| 09:10.35 | *** join/#htc-linux rzk (~rzk@95-27-68-96.broadband.corbina.ru) |
| 09:15.04 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
| 09:23.23 | *** join/#htc-linux dobrin (~dobrin@85.91.150.26) |
| 09:43.10 | *** join/#htc-linux ychavan (ychavan@nat/redhat/x-njvepxdvexljmvpv) |
| 09:46.07 | *** join/#htc-linux TheXev (~user@18.sub-174-252-212.myvzw.com) |
| 10:06.14 | *** join/#htc-linux arif-ali (~arif-ali@ip-81-23-53-226.ask4internet.com) |
| 10:21.06 | *** join/#htc-linux rajkosto (~rajkosto@2001:470:d76b:da7a:98b4:a086:3053:ce58) |
| 10:24.39 | *** join/#htc-linux TheXev (user@211.sub-75-218-236.myvzw.com) |
| 11:08.48 | *** join/#htc-linux TheXev (nat32@211.sub-75-218-236.myvzw.com) |
| 11:26.11 | *** join/#htc-linux Bry8Star_ (~Bry8Star@gateway/tor-sasl/bry8star) |
| 11:45.51 | *** join/#htc-linux TheXev (nat32@211.sub-75-218-236.myvzw.com) |
| 11:49.13 | *** join/#htc-linux helicopter88 (~helicopte@host116-118-dynamic.23-79-r.retail.telecomitalia.it) |
| 12:27.18 | *** join/#htc-linux Khooban (~Khooban@46.100.191.166) |
| 12:28.11 | Khooban | hello "apt" |
| 12:28.38 | Khooban | salam "ali1234 |
| 12:28.40 | Khooban | " |
| 12:29.43 | Khooban | is there anyone ? |
| 12:29.56 | Khooban | somebody help me!!!!!!!!!!!!!!!!!!!!!!!!!! |
| 12:30.42 | helicopter88 | Khooban: ask something |
| 12:30.56 | helicopter88 | someone will maybe help you |
| 12:31.05 | Khooban | i cant lunch the "haret.exe" |
| 12:31.17 | Khooban | on my hd2 |
| 12:33.32 | Khooban | when i tap that , it becoms to an erorr with 2 choosable keys : |
| 12:33.52 | Khooban | "send" and "don't send" |
| 12:34.33 | Khooban | can U help me " helicopter88 " ? |
| 12:34.49 | helicopter88 | I have a kovsky,and i can run haret fine. |
| 12:36.04 | Khooban | did U see "cotulla" today ? |
| 12:36.17 | Khooban | in this chat room |
| 12:36.19 | Khooban | ? |
| 12:36.20 | helicopter88 | Not today,but yesterday |
| 12:36.26 | helicopter88 | yes |
| 12:36.46 | Khooban | do u know when he comes here ? |
| 12:36.57 | helicopter88 | I'm not cotulla,so i don't know |
| 12:37.08 | Khooban | i mean usually |
| 12:37.25 | helicopter88 | In the afternoons |
| 12:38.03 | Khooban | Is there anyone who has HD2 ? |
| 13:33.20 | *** join/#htc-linux NYL (~NYL@109.96.36.166) |
| 13:44.32 | Blista | I do |
| 13:45.10 | Blista | ah he's gone |
| 13:46.21 | Blista | screams while glancing madly at the sky: Why am I always late ... Whyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy!!!! |
| 13:58.05 | *** join/#htc-linux MethoS- (~clemens@134.102.106.250) |
| 14:30.25 | *** join/#htc-linux helicopter88_1 (~helicopte@host156-115-dynamic.10-87-r.retail.telecomitalia.it) |
| 14:32.04 | *** join/#htc-linux Cotulla (~myfakemai@nat100-255-205-109.tvoe.tv) |
| 14:33.30 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 14:43.17 | Cotulla | hi |
| 14:46.05 | jONpRY | hello |
| 14:46.44 | Blista | cheers |
| 14:46.48 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
| 14:48.21 | Cotulla | checks memory stats in Blista's brain |
| 14:52.56 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
| 14:59.30 | *** join/#htc-linux Ondalf (~ondalf@unaffiliated/ondalf) |
| 15:00.58 | *** join/#htc-linux NYL_ (~NYL@109.96.36.166) |
| 15:06.18 | *** join/#htc-linux mgross029 (c0234f46@gateway/web/freenode/ip.192.35.79.70) |
| 15:11.34 | *** join/#htc-linux emwe (~mweirauch@cable-86-56-10-252.cust.telecolumbus.net) |
| 15:11.39 | *** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring) |
| 15:21.44 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 15:22.32 | Blista | total used free shared buffers |
| 15:22.36 | Blista | Mem: 423124 363904 59220 0 3140 |
| 15:22.38 | Blista | -/+ buffers: 360764 62360 |
| 15:22.41 | Blista | Swap: 0 0 0 |
| 15:24.05 | Blista | yes I do have a leo ... |
| 15:24.33 | Blista | hum |
| 15:24.38 | Blista | no swap ... |
| 15:31.32 | *** join/#htc-linux helicopter88 (~helicopte@host156-115-dynamic.10-87-r.retail.telecomitalia.it) |
| 15:36.49 | *** join/#htc-linux marc1706 (~marc@phpbb/modifications/marc1706) |
| 15:47.30 | jONpRY | i needz ics |
| 15:47.54 | *** join/#htc-linux |Jeroen| (~jeroen@d5152B25B.access.telenet.be) |
| 15:48.01 | Cotulla | it's since mango already |
| 15:48.04 | Cotulla | lol |
| 15:48.19 | jONpRY | ? |
| 15:48.47 | Cotulla | ics since mango |
| 15:49.04 | jONpRY | like the fruit? |
| 15:49.29 | Cotulla | ya |
| 15:50.15 | jONpRY | i really don't understand |
| 15:50.59 | Cotulla | ICS for me - Internet Connection Sharing. Was added in Windows Phone 7.5 "MANGO". |
| 15:51.00 | Cotulla | :D |
| 15:51.21 | jONpRY | oh. ice cream sandwich |
| 15:52.45 | jONpRY | windows phone is an oxymoron |
| 15:56.10 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-166-124.netcologne.de) |
| 16:15.50 | *** join/#htc-linux vinceweis (43346ffa@gateway/web/freenode/ip.67.52.111.250) |
| 16:16.40 | detule | jONpRY, any damage from pmem_registering pmem_gpu0 and hw3d devices on the same physical space? |
| 16:17.43 | jONpRY | pmem_gpu0 is how hw3d is setup |
| 16:17.51 | jONpRY | did you push that 3.0 stuff? |
| 16:18.08 | detule | what 3.0 stuff? |
| 16:18.49 | detule | jONpRY, i see that in registering hw3d we are using start and size from pmem_gpu0, however we are also registering device pmem_gpu0 |
| 16:18.51 | jONpRY | upgrade to 3.0.9, ashmem |
| 16:18.59 | detule | jONpRY, yeah it's all there |
| 16:19.12 | jONpRY | hmm. i'm having trouble checking it out for some reason |
| 16:19.50 | jONpRY | i dunno hw3d works |
| 16:19.59 | jONpRY | different things need to know where it is |
| 16:20.42 | detule | so i guess pmem.c is ok with registering multiple devices on the same address space |
| 16:21.12 | detule | https://gitorious.org/linux-msm-rhod/linux-msm-rhod/commits/htc-msm-3.0.0-rc1 looks like it's all there |
| 16:21.25 | jONpRY | yeah i probably have something botched |
| 16:24.17 | arrrghhh | jONpRY, failed again |
| 16:24.18 | arrrghhh | http://pastebin.com/tfhQz4hs |
| 16:24.20 | arrrghhh | looks the same... |
| 16:24.44 | arrrghhh | btw, do you sleep? :P |
| 16:25.00 | jONpRY | sometimes |
| 16:25.04 | arrrghhh | lol |
| 16:25.12 | jONpRY | so i have a plan on how to get it to compile for armv6 |
| 16:25.27 | jONpRY | which might fix this. else small patch required |
| 16:25.38 | arrrghhh | o, |
| 16:25.39 | arrrghhh | ok |
| 16:26.04 | jONpRY | need to see this build/core/combo/arch/arm/armv5te.mk |
| 16:26.14 | arrrghhh | 1 sec |
| 16:26.30 | arrrghhh | i probably should just give you access to my server, lol |
| 16:26.56 | jONpRY | thats one way |
| 16:26.58 | arrrghhh | lol well this is small |
| 16:27.09 | arrrghhh | http://pastebin.com/nxib5kNL |
| 16:27.48 | jONpRY | make new file in that directory called armv6j.mk with this in it http://pastebin.com/kZrn3TdS |
| 16:28.19 | jONpRY | then edit the boardconfig and try again |
| 16:28.23 | arrrghhh | k |
| 16:29.17 | arrrghhh | make clean worked, w00t |
| 16:29.35 | arrrghhh | and so the fun begins :D |
| 16:29.41 | *** join/#htc-linux WisTilt2 (~wisgreg@wireless251.wirelesstcp.net) |
| 16:29.42 | arrrghhh | now i need to figure out how the hell to test gb on .27 |
| 16:29.44 | arrrghhh | WisTilt2, wassup |
| 16:29.45 | jONpRY | its sad that we can't even build bionic |
| 16:29.53 | WisTilt2 | Morning |
| 16:31.26 | jONpRY | hi WisTilt2 |
| 16:31.57 | WisTilt2 | howdy. |
| 16:32.34 | WisTilt2 | jONpRY did you want me to push that pmem.c change or did you already do it, i didn't look yet |
| 16:32.50 | WisTilt2 | your fancy printk change |
| 16:32.52 | jONpRY | please push |
| 16:33.00 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 16:33.11 | WisTilt2 | that new one should help narrow down the camera memory clashes easier |
| 16:33.13 | emwe | jONpRY: stinebd pulled in armv6j support to xdandroid from cyanogen. you guys might wan't to do the same. |
| 16:33.34 | jONpRY | thats where i just grifted it from |
| 16:33.59 | emwe | ai |
| 16:34.06 | jONpRY | and we are trying to build cyanogen |
| 16:34.17 | emwe | ics still i assume |
| 16:34.23 | jONpRY | yup |
| 16:34.25 | emwe | was funny reading the logs from last night. |
| 16:34.51 | jONpRY | oh yeah? |
| 16:35.29 | arrrghhh | hrm |
| 16:35.34 | arrrghhh | seemed to fail for a different reason this time |
| 16:36.02 | jONpRY | you know if it is before or after errno.c? |
| 16:36.15 | arrrghhh | http://pastebin.com/dqnNtsrW |
| 16:36.29 | detule | them google boys *must* have a script "make ics"...we need to make friends |
| 16:36.31 | arrrghhh | i can't tell to be honest |
| 16:36.47 | arrrghhh | detule, yea, sure, for the galaxy nexus :P |
| 16:37.31 | detule | script "make ARCH=blah blah ics" |
| 16:37.38 | jONpRY | do you actually have dalvik/vm/mterp/out/InterpAsm-armv5te.S? |
| 16:37.40 | detule | i bet they have them too |
| 16:37.46 | arrrghhh | jONpRY, lemme look |
| 16:38.08 | arrrghhh | yes i do |
| 16:39.03 | *** join/#htc-linux avinashhm (~avinash-h@192.91.66.189) |
| 16:39.15 | jONpRY | in dalvik/vm/mterp is there a config-armv6j? |
| 16:41.11 | jONpRY | if not. probably need some of this: http://pastebin.com/5TtUpZUu |
| 16:41.44 | jONpRY | then go in that dir. and do ./gen-mterp.py armv6j out |
| 16:41.48 | arrrghhh | there is no config-armv6j |
| 16:41.54 | arrrghhh | ok |
| 16:43.29 | arrrghhh | Failed: handler-style must be first command |
| 16:43.50 | arrrghhh | yet it created the InterpAsm-armv6j.S file? |
| 16:44.33 | jONpRY | run it with no arguments. must be different than the gb command |
| 16:45.29 | jONpRY | or post up the config-armv5te |
| 16:45.37 | jONpRY | something different |
| 16:45.39 | *** join/#htc-linux swc|666 (~gecko@unaffiliated/swc666/x-4934821) |
| 16:45.45 | mgross029 | arrrghhh: just give jONpRY access. :p |
| 16:45.48 | mgross029 | heh |
| 16:46.24 | jONpRY | dalvik is serious progress though. it must have actually built bionic |
| 16:46.27 | arrrghhh | mgross029, i probably should |
| 16:46.47 | arrrghhh | Usage: ./gen-mterp.py target-arch output-dir |
| 16:47.03 | arrrghhh | and it did create the file |
| 16:47.11 | jONpRY | yeah i think the contents of config-armv6 needs to be modified |
| 16:47.19 | jONpRY | there is 2 files though |
| 16:47.50 | jONpRY | Generating out/InterpC-armv6j.c, out/InterpAsm-armv6j.S |
| 16:48.35 | arrrghhh | yea, InterpAsm-armv6j.S is pretty much empty |
| 16:49.15 | jONpRY | show me config-armv5te |
| 16:49.21 | arrrghhh | k |
| 16:49.45 | arrrghhh | http://pastebin.com/qcBC68pS |
| 16:53.07 | jONpRY | how about this http://pastebin.com/Fuhha25r |
| 16:53.17 | jONpRY | dur |
| 16:53.41 | *** join/#htc-linux rob_w (~bob@ppp-188-174-6-194.dynamic.mnet-online.de) |
| 16:53.44 | jONpRY | http://pastebin.com/bJEZPhAm |
| 16:54.34 | jONpRY | i think there is also a directory of 3 files you will need |
| 16:55.26 | arrrghhh | you want me to save that over the config-armv5te...? |
| 16:55.42 | jONpRY | no armv6j |
| 16:55.50 | arrrghhh | k |
| 16:56.26 | arrrghhh | Failed: unrecognized command 'Copyright' |
| 16:56.27 | arrrghhh | lolwut |
| 16:56.38 | arrrghhh | der |
| 16:56.42 | arrrghhh | is retarded |
| 16:56.49 | arrrghhh | first line wasn't commented out, lets try that again... |
| 16:57.01 | arrrghhh | jONpRY, worked, no errors |
| 16:57.23 | arrrghhh | that InterpAsm-armv6j.S seems to have some meaningful stuff in it now too. |
| 16:57.26 | arrrghhh | make clean; make? |
| 16:57.41 | jONpRY | sure. i think it will barf without this dir of files |
| 16:57.52 | jONpRY | probably don't even need a clean |
| 16:58.10 | arrrghhh | ok |
| 16:58.35 | jONpRY | just see what it does i guess |
| 16:58.46 | jONpRY | download this in case http://www.box.com/s/o6sznjl2r056vdd52d2s |
| 16:59.34 | arrrghhh | yea it failed |
| 16:59.38 | arrrghhh | where do i put those files? |
| 16:59.45 | *** join/#htc-linux Dave888 (~Dave888@41.230.58.85) |
| 17:00.00 | arrrghhh | `dalvik/vm/compiler/codegen/arm/armv6j/CallingConvention.S' |
| 17:00.06 | jONpRY | dalvik/vm/mterp/armv6/ |
| 17:00.35 | *** part/#htc-linux Dave888 (~Dave888@41.230.58.85) |
| 17:02.35 | jONpRY | ^^^ no j |
| 17:02.38 | arrrghhh | if i do make -j5 > make.txt, i should have all the output from the make in that file right? |
| 17:02.46 | arrrghhh | ah poop |
| 17:03.21 | jONpRY | isn't it like &> or something |
| 17:03.34 | jONpRY | in general the logs are all but useless with j > 1 |
| 17:03.53 | arrrghhh | ok |
| 17:04.41 | arrrghhh | well i see a lot of issues, but it hasn't stopped yet |
| 17:05.34 | jONpRY | no deal breakers? |
| 17:05.44 | arrrghhh | evidently not |
| 17:05.47 | arrrghhh | cuz it's still chugging |
| 17:07.02 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-68-237.dynamic.mnet-online.de) |
| 17:09.39 | mgross029 | Did anyone make a determination to why I2C_BOARD_INFO("audience_A1010", 0xf4>>1) is disabled in the board? Just curious. |
| 17:10.04 | mgross029 | Saw detule bring it up the other day. |
| 17:10.43 | WisTilt2 | it breaks mic in call in speakerphone mode |
| 17:10.53 | jONpRY | arrrghhh, this is gonna be sweet |
| 17:11.09 | mgross029 | Hmm.. Interesting. Thanks |
| 17:11.13 | WisTilt2 | jONpRY: do you still want to leave these printks in with that devices_and_enter stuff? |
| 17:11.32 | rpierce99 | this thing is gradually fading again :( The quick vibe for email received is lasting like 20 seconds, and the inactivity screen off is going on a couple minutes now |
| 17:11.58 | jONpRY | WisTilt2, are they helpful at all? |
| 17:12.22 | WisTilt2 | might be with the slowdowns rpierce99 is starting to see |
| 17:14.55 | mgross029 | WisTilt2: I enabled it in .35 and it's working for me with Speakerphone |
| 17:15.40 | arrrghhh | jONpRY, bah it failed out |
| 17:15.52 | WisTilt2 | kills it on my 300 in .39 for sure. handset mic still works in normal call though but speakerphone mic is muted or just dead |
| 17:16.27 | arrrghhh | yea, i don't think that got the errors. someone gave me a tip, lemme build again |
| 17:16.38 | mgross029 | WisTilt2: I did change 0xf4>>1 to 0xf4/2 not sure if that is the difference |
| 17:17.06 | *** join/#htc-linux Markinus (~Adium@212.255.36.215) |
| 17:17.07 | mgross029 | Saw an old post from crt and phh from 1/15/2010 where they were looking at it. |
| 17:19.19 | arrrghhh | thanks helicopter88 --> make -jX >> logfile 2>&1 |
| 17:19.33 | helicopter88 | you're welcome :) |
| 17:20.27 | WisTilt2 | mgross029: you mean 0xf4>>2 not 0xf4/2 correct? |
| 17:20.57 | mgross029 | No 0xf4/2 |
| 17:21.22 | WisTilt2 | thats kind of the same result |
| 17:21.48 | WisTilt2 | so 0xf2>>1 should be working also |
| 17:21.51 | WisTilt2 | 0xf4 |
| 17:22.26 | mgross029 | Ok. Just found this post and was going off of that to see what made the difference. http://irclog.netripper.com/htc-linux/2010/1/15/ |
| 17:22.49 | mgross029 | Look at around 22:40 time frame |
| 17:23.02 | mgross029 | post form crt |
| 17:23.38 | mgross029 | I will recompile with >>1 to see if that is working for me too then. |
| 17:24.04 | WisTilt2 | yeah both are the same. he's just referring to the /2 make it easier to understand for people that don't know what shifting is:) |
| 17:24.47 | WisTilt2 | it will work. ex: 4>>1 = 2, 4/2 = 2 its the same thing |
| 17:24.49 | rpierce99 | >> isn't bit shifting, 0xf2 is just a LOT bigger than 1 :P |
| 17:25.22 | WisTilt2 | ok |
| 17:25.56 | WisTilt2 | detule: have you had any issues with losing data with .39 lately? |
| 17:25.59 | rpierce99 | <PROTECTED> |
| 17:26.04 | WisTilt2 | i know:) |
| 17:27.29 | mgross029 | WisTilt2: Ok. Understood. I just started looking into this because on my first initial incoming call in-call mic doesn't respond unless I flip speakphone on/off, so was seeing if this would fix the problem. Thanks for the info. |
| 17:28.04 | rpierce99 | i'm going to reboot to resolve this slowness, before I do I wanted to ask if I should pull logs or switch to another kernel, seems my phone is a good test case for the gp_timer thing |
| 17:28.14 | WisTilt2 | that only seems to be happening on cdma devices. i never have that on my 300 on .39 for past few months |
| 17:28.45 | *** join/#htc-linux arif-ali (~arif-ali@ip-81-23-53-226.ask4internet.com) |
| 17:28.54 | WisTilt2 | probably something weird in the cdma part of the ril cuz it's not a problem in the kernel afaict |
| 17:29.26 | mgross029 | WisTilt2: Isn't it always a cdma device thing? :p |
| 17:29.27 | WisTilt2 | rpierce99 you might want logs. we can compare from the last night one |
| 17:29.37 | WisTilt2 | yeah, gsm rules:) |
| 17:30.34 | arrrghhh | jONpRY, http://pastebin.com/ZMmJQPRk |
| 17:31.36 | jONpRY | thats the same error we had yesterday? |
| 17:32.20 | arrrghhh | indeed |
| 17:32.26 | rpierce99 | http://pastebin.com/qDiiS5Kj <- gp_timer slowness issue, not as severe as last night |
| 17:32.30 | arrrghhh | just took a lot longer to come about lol |
| 17:34.12 | jONpRY | need bionic/libc/private/bionic_tls.h |
| 17:34.49 | arrrghhh | http://pastebin.com/DE9DU6v5 |
| 17:35.00 | *** join/#htc-linux gnutoo_ (~GNUtoo@host246-111-dynamic.48-82-r.retail.telecomitalia.it) |
| 17:35.20 | jONpRY | something is setting HAVE_ARM_TLS_REGISTER |
| 17:35.28 | jONpRY | grep through /device/ti |
| 17:36.52 | arrrghhh | ok, grep noob here. |
| 17:36.54 | jONpRY | or everything. atm i don't know where it is |
| 17:37.00 | arrrghhh | grep pattern file right? |
| 17:37.03 | jONpRY | find | xargs grep HAVE_ARM_TLS |
| 17:37.07 | arrrghhh | heh |
| 17:37.09 | arrrghhh | thanks |
| 17:38.40 | jONpRY | it may actually be ARM_HAVE_TLS_REGISTER |
| 17:39.07 | arrrghhh | its found some things |
| 17:39.16 | jONpRY | in device? |
| 17:39.16 | arrrghhh | lots of no such file or directory tho, odd |
| 17:39.31 | arrrghhh | so far they've all been in bionic |
| 17:39.45 | WisTilt2 | rpierce99 you had that thing running a long time before it started slowing down. if you can, next time this happens cat the gp_timer over 10-15 secs or so and see if it is wildly going up or dragging. |
| 17:39.52 | jONpRY | yeah just search in device for ARM_HAVE |
| 17:40.10 | jONpRY | i think it should be in ti/panda/BoardConfigCommon.mk |
| 17:40.14 | rpierce99 | yeah it had been running since about 2am, but it was all overnight, nothing was going on |
| 17:40.17 | mgross029 | arrrghhh: grep -s surpresses the no directory |
| 17:40.23 | arrrghhh | cool |
| 17:40.23 | rpierce99 | it had been slowing down slowly for an hour or two |
| 17:40.39 | WisTilt2 | you reboot it yet or still in that condition? |
| 17:40.44 | jONpRY | ex ./htc/leo/BoardConfig.mk:ARCH_ARM_HAVE_TLS_REGISTER := true |
| 17:40.54 | rpierce99 | froze at "powering off" |
| 17:42.22 | arrrghhh | jONpRY, |
| 17:42.22 | arrrghhh | ./ti/panda/BoardConfig.mk:ARCH_ARM_HAVE_TLS_REGISTER := true |
| 17:42.45 | jONpRY | comment out or otherwise remove that line |
| 17:42.48 | arrrghhh | k |
| 17:43.17 | WisTilt2 | sounds like battery pulling time. well see what gp_timer is doing next time it happens. |
| 17:43.40 | rpierce99 | will do |
| 17:43.50 | rpierce99 | what's the path for that |
| 17:44.02 | mgross029 | rpierce99: After you reboot back in to Android try to pull the last_kmsg file |
| 17:44.02 | WisTilt2 | jONpRY did detule put that memory killer or whatever it was in 3.xx into .39 tree also? |
| 17:44.22 | rpierce99 | there won't be one since i pulled the batt, memory was cleared |
| 17:44.23 | WisTilt2 | that wont have the gp_timer interrupts though |
| 17:44.32 | jONpRY | WisTilt2, i think its already there |
| 17:47.03 | *** join/#htc-linux paulk_ (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 17:52.34 | WisTilt2 | rpierce99: are you seeing the same slow downs on .39? |
| 17:52.52 | rpierce99 | i haven't run a vanilla .39 in a while, that's why i asked if i should try another kernel |
| 17:53.31 | WisTilt2 | yeah i'd try the current .39 and see. this thing is flying on my phone. |
| 17:53.45 | rpierce99 | .39 glemsom autobuild? |
| 17:54.05 | WisTilt2 | if its current. should be gc58e59c |
| 17:54.25 | jONpRY | arrrghhh, no boom? |
| 17:54.34 | arrrghhh | still mak'ing |
| 17:54.36 | WisTilt2 | my autobuild should be current unless it decided not to build again |
| 17:54.46 | rpierce99 | a package hasn't build on glemsom since 11/8 |
| 17:55.15 | rpierce99 | they are failing |
| 17:55.33 | WisTilt2 | mine is current and built after i pushed this morning |
| 17:55.52 | rpierce99 | do i have a link to yours? |
| 17:56.01 | WisTilt2 | autobuild.wirelesstcp.net |
| 17:56.17 | WisTilt2 | both .39 and 3.x stuff there |
| 17:56.46 | WisTilt2 | not as fancy as what glemson has with logs etc but has the kernel packs |
| 17:56.50 | arrrghhh | hrm |
| 17:56.53 | arrrghhh | jONpRY, different error 1 sec |
| 17:56.54 | rpierce99 | does kexec work from a kernel that supports it to a kernel that doesn't, I'd guess no |
| 17:57.08 | arrrghhh | http://pastebin.com/4ZDgFdFp |
| 17:57.13 | WisTilt2 | .39 has kexec built in so should work |
| 17:57.19 | rpierce99 | oh goody |
| 17:57.25 | WisTilt2 | at least i think its still enabled |
| 17:57.38 | jONpRY | rpierce99, yes |
| 17:59.28 | WisTilt2 | jONpRY btw, i measured drain again with/without scbs daemon - average over 1hr without was 18ma, with it 20ma so around 10% more. if we changed polling to 1min that would probably be practically no difference over the 18ma |
| 18:00.05 | jONpRY | excellent |
| 18:00.19 | WisTilt2 | if i can ever get this idle_collapse working it probably wont matter anyway since we should be under 10ma |
| 18:01.06 | WisTilt2 | im showing idle percentage around 95% so we're burning up a lot of power we could be saving |
| 18:03.00 | jONpRY | idle collapse should effect the drain during sleep |
| 18:03.09 | arrrghhh | jONpRY, did you see that new pastebin? |
| 18:03.19 | jONpRY | no, thanks |
| 18:05.02 | jONpRY | there might be something in the BoardConfig about SMP? |
| 18:05.08 | jONpRY | cause panda is a dual core |
| 18:05.49 | arrrghhh | hrm |
| 18:05.54 | arrrghhh | the natives are getting restless |
| 18:06.09 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-vartndwikgnphdhp) |
| 18:10.05 | *** join/#htc-linux Mavy-bnc (mavfree@91.196.169.2) |
| 18:10.22 | WisTilt2 | arrrghhh: in GB what exactly is the difference with the signal bars when they are green or grey? looks like when they turn grey i have no data even though full bars and H still showing up. |
| 18:10.26 | arrrghhh | yup |
| 18:10.39 | arrrghhh | green = connection to google's servers (internet access basically) |
| 18:10.46 | arrrghhh | gray = no connectivity to the outside world |
| 18:11.03 | arrrghhh | g2g tho, ttyl guys |
| 18:11.25 | WisTilt2 | ok thanks and later |
| 18:11.44 | WisTilt2 | nice to have such a good dictionary to go to:) |
| 18:12.17 | jONpRY | guess the ics port is on hold for a minute. looks like bionic compiled if it didn't link yet |
| 18:13.00 | jONpRY | dur and its compiling smartreflex |
| 18:13.14 | rpierce99 | didn't stinebd get it running, he said something about 3fps |
| 18:13.34 | jONpRY | orly? |
| 18:13.55 | jONpRY | i thought we were on the bleeding edge |
| 18:14.17 | rpierce99 | you know him, he's a giant liar |
| 18:16.47 | helicopter88 | jONpRY: wouldn't be easier to add a new device instead of messing with panda? |
| 18:17.22 | jONpRY | no i don't think so |
| 18:18.24 | rpierce99 | WisTilt2: does this autobuild .39 have the BT stuff you've been testing |
| 18:18.50 | WisTilt2 | no, built off our .39 git tree |
| 18:31.50 | *** join/#htc-linux [acl] (~abel@96.246.167.90) |
| 18:34.53 | [acl] | emwe: sup bro.. how's life |
| 18:36.07 | [acl] | dang.. that good huh |
| 18:36.15 | [acl] | i hear ya |
| 18:36.27 | WisTilt2 | hey [acl] |
| 18:36.58 | [acl] | WisTilt2: sup dood.. the left you all alone over here? |
| 18:37.16 | WisTilt2 | i work better alone where it's quiet:) |
| 18:37.26 | [acl] | lol |
| 18:37.33 | WisTilt2 | hey we got u-boot to boot up .39 |
| 18:37.46 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
| 18:37.50 | [acl] | thats awesome. you figure out how nand is layed out ? |
| 18:37.56 | [acl] | for gsm |
| 18:38.11 | WisTilt2 | no winmo in flash at all now but lots of mapping for gsm to figure out so we can init everything winmo did |
| 18:38.30 | WisTilt2 | i have one of my guys doing that, i got better things to do |
| 18:38.38 | [acl] | yeah, thats the tough part. Let em handle it |
| 18:38.59 | WisTilt2 | there is so much free flash memory we could put about anything on it |
| 18:39.01 | [acl] | i was able to boot, but i thought i knew the layout but it turns out i was wrong |
| 18:39.06 | [acl] | i cant write to certain sections |
| 18:39.10 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-102-103.dhcp.stcd.mn.charter.com) |
| 18:39.42 | WisTilt2 | you have to unlock the whole flash. we have enough phones that i figured if we trashed a few no biggy but got it on 1st attempt |
| 18:39.54 | [acl] | ahh, that makes sense |
| 18:39.59 | [acl] | so in theory you can upgrade the radio too |
| 18:40.03 | WisTilt2 | yep |
| 18:40.27 | [acl] | i wonder if now jonpry can do what he wanted to do in the past and run linux in arm9 |
| 18:40.28 | [acl] | lol |
| 18:40.50 | WisTilt2 | im thinking that's entirely possible now |
| 18:41.06 | [acl] | yup, should be. Fuck you could do so much now |
| 18:41.11 | emwe | [acl]: busy. i haven't answered your mail, i know. perhaps we can talk later... |
| 18:41.16 | [acl] | hell you can even try the saphire radio like cotulla |
| 18:41.31 | jONpRY | is that a NUPA |
| 18:41.36 | [acl] | emwe: its all good bro.. im traveling right now so i wont be home for a few days as well |
| 18:43.29 | jONpRY | why would you mess with the flash? isn't hot patching arm9 code way easier? |
| 18:43.31 | [acl] | WisTilt2: if you have a few more to trash, try the saphire radio :-p .. actually if you can uboot, you should be able to just try anything now without breaking |
| 18:43.54 | WisTilt2 | where do i get it |
| 18:44.07 | [acl] | WisTilt2: not sure :-( .. |
| 18:44.26 | [acl] | im sure someone can find it |
| 18:44.26 | [acl] | lol |
| 18:44.52 | jONpRY | doesn't uboot come after radio? |
| 18:45.40 | [acl] | jONpRY: good question for WisTilt2 |
| 18:45.41 | WisTilt2 | not as we have it |
| 18:45.42 | [acl] | lol |
| 18:45.52 | WisTilt2 | normally it would yes |
| 18:46.07 | jONpRY | so you boot the radio? |
| 18:46.10 | WisTilt2 | yep |
| 18:46.32 | WisTilt2 | doing it just like our own device, or similarly i should say. |
| 18:46.48 | *** join/#htc-linux mastermerlin (~Adium@p4FEE594F.dip.t-dialin.net) |
| 18:46.51 | jONpRY | and the htc radio is fine with this? |
| 18:47.50 | WisTilt2 | not yet. we have a lot of initializing to do still. just basically got the kernel booting to see if it was possible this way. |
| 18:47.57 | phh | huhu |
| 18:48.11 | phh | and u-boot is run on A11 ? |
| 18:48.16 | Cotulla | hi phh |
| 18:48.20 | Cotulla | how is ur uM? |
| 18:48.33 | phh | uM? |
| 18:48.37 | [acl] | dang all the heavy hitters are here |
| 18:48.41 | phh | [acl]: :p |
| 18:48.45 | [acl] | like the dead work up |
| 18:48.48 | phh | missing cr2. |
| 18:48.50 | [acl] | damn |
| 18:49.04 | Cotulla | which THUMB2 only |
| 18:49.10 | Cotulla | hi [acl] dood |
| 18:49.20 | [acl] | Cotulla: :-p |
| 18:49.20 | phh | Cotulla: yeah well i definitely suck at it |
| 18:49.21 | jONpRY | cortex m4f |
| 18:50.09 | Cotulla | at which? |
| 18:50.33 | phh | jONpRY: mostly at building sources people made on linux ... |
| 18:51.09 | jONpRY | not putting linux on the m4? |
| 18:51.16 | phh | 192kB of RAM. |
| 18:51.21 | [acl] | WisTilt2: anyway.. thats cool .. you can pretty much do anything you want now. |
| 18:51.26 | phh | i don't really feel like putting an external ram |
| 18:51.44 | jONpRY | does it have an sdram controller? |
| 18:52.01 | phh | it's an SRAM controller, even though i don't know the difference |
| 18:52.33 | phh | (isn't it just N address pins + 2**M data pins + WR bit !?!) |
| 18:52.46 | jONpRY | sdram's are very complicated |
| 18:52.59 | jONpRY | with banks, fifos, precharges |
| 18:53.58 | jONpRY | sram they don't really make anymore. micron makes something called PSRAM, or PSSRAM |
| 18:54.31 | [acl] | damn.. way over my head |
| 18:54.49 | Cotulla | I think if there controller |
| 18:54.55 | Cotulla | it's enough to setup it right |
| 18:57.55 | phh | bah, in worst case i could do a software ram controler |
| 18:58.05 | phh | (i'm sure that would be amazingly fast :D) |
| 18:58.24 | *** join/#htc-linux MN-- (~yaaic@5e0dd1f9.bb.sky.com) |
| 18:58.37 | jONpRY | in vhdl maybe |
| 18:58.46 | phh | i said soft. |
| 18:59.02 | jONpRY | on fpga |
| 18:59.15 | jONpRY | or is that too firm |
| 18:59.18 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 19:06.42 | phh | Cotulla: anyway i think you're right, i'll do a bootloader from linux |
| 19:07.03 | phh | i'm wondering if i should do full user-land way though |
| 19:07.53 | phh | would be funny, but so useless |
| 19:09.35 | Cotulla | on 192kb ram? |
| 19:10.26 | phh | -_-' |
| 19:10.30 | phh | i'm speaking of droid3 now |
| 19:11.04 | Cotulla | hm |
| 19:11.34 | Cotulla | linux is bad for bootloaders |
| 19:11.43 | Cotulla | as any other OS with processes and complex memory access :) |
| 19:11.45 | phh | userland bootloaders* |
| 19:11.53 | rob_w | hehe |
| 19:11.54 | phh | but just make kexec a module |
| 19:12.00 | Cotulla | ok |
| 19:12.14 | phh | (afaik modules are available on droid3) |
| 19:12.19 | phh | (that kills all the fun.) |
| 19:12.23 | phh | Cotulla: anyway there is access to /dev/mem |
| 19:12.37 | phh | it just requires to change the memory of some specific kernel's function and call it |
| 19:12.46 | Cotulla | yes |
| 19:18.53 | jONpRY | kexec on cortex a9 is a disaster |
| 19:19.07 | Cotulla | why? |
| 19:19.23 | [acl] | aite fellas.. got a bus to catch.. ttyl |
| 19:19.49 | jONpRY | it requires you get to a single core mmu off state. which is not even possible on some architectures |
| 19:20.05 | Cotulla | on which? |
| 19:20.20 | Cotulla | u need shutdown others |
| 19:20.24 | Cotulla | for next OS |
| 19:20.29 | jONpRY | somebody made it work on realview |
| 19:20.41 | phh | jONpRY: :( |
| 19:20.50 | Cotulla | all OS which I saw start up another cores during boot |
| 19:20.57 | jONpRY | like on snapdragon there doesn't seem to be a way to turn off the second core power domain |
| 19:21.04 | jONpRY | it always has to be spinning somewhere |
| 19:21.09 | jONpRY | even if its WFI |
| 19:21.24 | phh | jONpRY: meh ? |
| 19:21.25 | Cotulla | maybe u just don't know such way ;) |
| 19:21.42 | phh | jONpRY: aSMP is only about clocks ? |
| 19:21.43 | jONpRY | given the documentation |
| 19:21.52 | Cotulla | but I think there must be off function |
| 19:21.58 | phh | well droid3 is omap4, so at least i've got doc available. |
| 19:22.04 | phh | Cotulla: reboot ? :p |
| 19:22.10 | Cotulla | no |
| 19:22.17 | Cotulla | reboot |
| 19:22.20 | Cotulla | reboot can be too different |
| 19:22.53 | Cotulla | dunno, I am still with single core QSD8250 :) |
| 19:23.04 | phh | i'm still with single core msm7200 -_-' |
| 19:23.16 | jONpRY | thats the best |
| 19:23.18 | Cotulla | no |
| 19:23.27 | phh | not really. |
| 19:23.29 | Cotulla | :P |
| 19:23.32 | Cotulla | it's so slow |
| 19:23.37 | Cotulla | there no L2 |
| 19:23.42 | Cotulla | even |
| 19:23.45 | phh | wait, you mean there is L1 ? |
| 19:23.56 | Cotulla | it's there, but not L2? |
| 19:24.03 | jONpRY | there is l2 it just doesn't have or need the l2 cache controller |
| 19:24.11 | Cotulla | u sure? |
| 19:24.33 | jONpRY | yeah. there is no need for cache coherency with single core. so no controller, and no driver. ie l2x0 |
| 19:24.35 | Cotulla | so why they go to additional brainache with controller |
| 19:24.58 | Cotulla | driver used for 7627 7630 too |
| 19:25.16 | Cotulla | and it was always multicore lol |
| 19:25.16 | Cotulla | :D |
| 19:25.24 | Cotulla | dunno... all this is strange |
| 19:25.35 | Cotulla | I hate V7 d cache clear |
| 19:25.37 | Cotulla | it's so long |
| 19:25.41 | Cotulla | :( |
| 19:25.42 | phh | perhaps it's a L2 on AHB bus ? |
| 19:25.55 | phh | (for the multi core problem) |
| 19:26.12 | Cotulla | it's important how cache is tagged |
| 19:26.22 | Cotulla | also |
| 19:26.30 | phh | hum ahb bus is not specially fast, that's stupid. |
| 19:26.38 | Cotulla | is it AXI? |
| 19:26.43 | phh | yeah |
| 19:27.12 | Cotulla | dunno, I am not so good in HW |
| 19:27.38 | phh | pfff :p |
| 19:27.42 | Cotulla | it's hard to research something there and find truth anyway |
| 19:28.19 | detule | WisTilt2, the memory shrinker that was broken in 3.0 is something that is already working in .39 (and all prior android kernels for that matter) ... it's not a feature, it's a must |
| 19:29.49 | jONpRY | i've almost got a 3.1 kernel |
| 19:29.50 | WisTilt2 | detule: have you had any of these problems with data stopping in .39 lately? have to toggle data off/on to get it to come back up. |
| 19:30.18 | Cotulla | 3.1? |
| 19:30.36 | jONpRY | yeah linux 3.1.1 |
| 19:30.45 | Cotulla | they are fast |
| 19:30.53 | detule | GB is not 100% ok with restarting data on our devices, at least on my cdma....it's happened to me a few times that i've had to go to airplane mode and back....though not recently |
| 19:31.34 | WisTilt2 | detule: ok not going to chase it then. just didn't notice it before but i was on froyo still thanks. |
| 19:32.15 | detule | jONpRY, if we go high enough the gp_timer issue might disappear |
| 19:32.21 | detule | perhaps it's fixed in 3.2 |
| 19:32.22 | detule | :) |
| 19:32.26 | mgross029 | WisTilt2: I think I found a workaround for no-mic on first call and no ringer too. |
| 19:32.43 | jONpRY | no. gotta disable that smem_sync stuff |
| 19:34.11 | mgross029 | My thought was that ToneGenerator must not be getting initialized on boot, so I down loaded Boot Pride, which plays audio at boot up. No so far two tests and all is good. |
| 19:36.08 | WisTilt2 | mgross029: is there any way you could boot without that boot pride, get a logcat after you make a call without toggling speakerphone, then do same thing with boot pride? |
| 19:36.47 | mgross029 | Sure... I was just thinking of doing that. |
| 19:37.32 | detule | jONpRY, disable sync_smem? |
| 19:38.02 | jONpRY | in timer.c, all that stuff about smem |
| 19:40.52 | detule | sorry to be dense here, you saying we need to do this, or you had to do that in 3.1 to get it to work |
| 19:41.09 | WisTilt2 | mgross029: logcat and dmesg for each if you can |
| 19:42.33 | mgross029 | WisTilt2: I will work on that. |
| 19:43.55 | jONpRY | detule, i haven't done it. something i want to try on 3.0 |
| 19:45.55 | detule | ~30hrs on 3.0 here |
| 20:08.24 | jONpRY | getting it to build is the half the battle |
| 20:09.39 | detule | well phone is not blowing up *without* pmem_gpu0 and pmem_gpu1 |
| 20:09.52 | jONpRY | is 3d working? |
| 20:10.21 | detule | sure hw3d is still being registered on the same addresses as pmem_gpu* |
| 20:11.03 | *** join/#htc-linux Ondalf (~ondalf@cable-roi-fffddd00-9.dhcp.inet.fi) |
| 20:11.57 | detule | i *think* open_gl only looks for /dev/hw3d |
| 20:13.21 | WisTilt2 | detule: did you remove the gpu0/1 stuff from mach-msm/pmem.c or what? |
| 20:13.44 | detule | i removed the code where it registers the devices in devices_htc.c |
| 20:14.07 | detule | it registers gpu0 and gpu1 and over the same physical region it also registers hw3d |
| 20:14.46 | WisTilt2 | so just removed the platform_device_register for both those? |
| 20:14.51 | detule | yeah |
| 20:15.30 | WisTilt2 | ok. guess the will still work if hw3d does it. i wonder if we gained memory up in smi and ebi doing this |
| 20:16.13 | *** part/#htc-linux Cotulla (~myfakemai@nat100-255-205-109.tvoe.tv) |
| 20:16.21 | detule | need to run quadrant or something to verify 3d still works |
| 20:16.21 | WisTilt2 | should gain 8mb in ebi and 7 in smi |
| 20:16.40 | detule | well hw3d is still registered over the same region |
| 20:16.56 | WisTilt2 | i dont see that |
| 20:17.33 | *** join/#htc-linux XeKToReX (~xTx@115-64-132-224.static.tpgi.com.au) |
| 20:17.38 | detule | line 190 and below in devices_htc.c |
| 20:17.42 | detule | or perhaps i am reading it wrong |
| 20:18.08 | WisTilt2 | yes but its not using same region as gpu's were |
| 20:18.23 | detule | no? |
| 20:18.32 | WisTilt2 | thats directly to the hw address now |
| 20:21.38 | detule | as soon as my phone is done analyzing this scbs model i'll try quadrant |
| 20:21.41 | mgross029 | WisTilt2: I moved the gb image and kernel to another 35 build folder and tested. (so without boot pride installed) I still seem to not get ringer but mic worked on first call. Going to reboot and recapture. |
| 20:22.23 | mgross029 | Somthing to note I also change the A1010 from /2 to >>1 which as you said is the same in bit shifting terms. |
| 20:23.07 | WisTilt2 | mgross029: you're testing this on .39 right? |
| 20:23.33 | mgross029 | No still on .35.. Haven't adjusted it on .39 yet |
| 20:26.43 | mgross029 | WisTilt2: I still need to get back to the .39 bt testing too. Just having an incomming call (from my wife) and the mic or ringer not working is not good for my health. :p |
| 20:27.01 | WisTilt2 | detule: looking further yes hw3d still mapping same region so dont see why gpu0/1 were also registered in addition |
| 20:27.23 | detule | right it seems like a pretty obvious overlap there |
| 20:27.35 | WisTilt2 | mgross029: tell her its all in the name of science:) |
| 20:27.41 | arrrghhh | jONpRY, ping |
| 20:28.24 | mgross029 | WisTilt2: Yeah right. LOL :p |
| 20:28.30 | WisTilt2 | detule: you still left lines 192-199 right? |
| 20:29.01 | detule | WisTilt2, yeah didn't mess with hw3d |
| 20:29.54 | WisTilt2 | should all still work unless something is still trying to user gpu0/1 then |
| 20:30.00 | WisTilt2 | user=use |
| 20:31.27 | detule | ok, but thus region overlap, say something is using both gpu* and hw3d |
| 20:31.36 | detule | wouldn't this cause issues |
| 20:32.27 | WisTilt2 | i'd think so. you might try camera again after this and see if those clashes still show up |
| 20:35.37 | WisTilt2 | lunch has finally arrived so ill bbl |
| 20:37.27 | rpierce99 | if something is using gpu0, that would be broken now though, but if nothing was using gpu0 before, then this won't fix anything right? |
| 20:38.11 | jONpRY | hi arrrghhh |
| 20:38.25 | arrrghhh | yo |
| 20:39.31 | detule | rpierce99, there could be code in userland that checks for existence of gpu0 and then decides to do something else if not present |
| 20:41.42 | detule | at any rate it seems suspect that there would be two devices drawing on the same physical region, the way i read staging/pmem.c for each device it creates a bitmap for the entire region it's on with flags if a particular block is in use or not......it would seem with multiple devices on the same region there would be multiple bitmaps marking whether the same piece of memory is available or not |
| 20:48.21 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 20:51.01 | detule | well camera works |
| 20:51.25 | detule | though bzo was in a weird spot where it would work for him at times and not at other |
| 20:54.10 | detule | sure enough 2nd photo failed |
| 21:02.29 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
| 21:05.36 | *** join/#htc-linux avinashhm (~avinash-h@192.94.92.11) |
| 21:07.20 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
| 21:07.58 | mgross029 | Later guys. g2g |
| 21:20.46 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 21:22.02 | jONpRY | alright first build of 3.1 |
| 21:27.57 | jONpRY | anxiously awaits anyone booting it |
| 21:29.03 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-68-237.dynamic.mnet-online.de) |
| 21:30.09 | helicopter88 | jONpRY: is it rhod only? |
| 21:30.36 | jONpRY | yes |
| 21:30.44 | helicopter88 | oh,then i can't help you |
| 21:30.50 | helicopter88 | *test it |
| 21:30.56 | jONpRY | and unless you know how to use haret and ram console. it probably will appear to do nothing |
| 21:31.25 | helicopter88 | i neither know how to use ram console.. |
| 21:36.06 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
| 21:36.17 | dr1337 | 3.1? |
| 21:37.17 | arrrghhh | lol |
| 21:37.19 | arrrghhh | yes 3.1 |
| 21:39.59 | *** join/#htc-linux dr1337_ (78931828@gateway/web/freenode/ip.120.147.24.40) |
| 21:40.00 | jONpRY | why does it build all this annoying stuff before libc |
| 21:40.14 | arrrghhh | i... don't know. |
| 21:40.14 | dr1337_ | what are you guys building? |
| 21:40.15 | dr1337_ | agreed |
| 21:40.18 | arrrghhh | cm9 |
| 21:40.42 | dr1337_ | i thought everything depended on libc.... |
| 21:40.55 | jONpRY | me too |
| 21:41.22 | jONpRY | but its build out/target/common or something |
| 21:41.27 | dr1337_ | isn't cm9 pretty much just vanilla ics source rightnow? |
| 21:41.40 | jONpRY | yes |
| 21:42.13 | dr1337_ | which devices are you guys working on? |
| 21:42.21 | jONpRY | rhodium |
| 21:43.38 | dr1337_ | ah ok so you're working on the msm7k device drivers too |
| 21:44.46 | jONpRY | we are not that far |
| 21:49.18 | dr1337_ | k ping me when you get there. i'm working on them now |
| 21:49.29 | arrrghhh | dr1337_, what are you on? |
| 21:49.31 | arrrghhh | device-wise |
| 21:49.39 | dr1337_ | nexus one |
| 21:50.49 | jONpRY | so your doing an ics hal? |
| 21:51.09 | dr1337_ | yes |
| 21:51.26 | dr1337_ | hey can you guys explain a little about LK bootloaders? |
| 21:51.49 | jONpRY | nexus one is not really msm7k although your hal will probably work |
| 21:52.05 | dr1337_ | yeah i know |
| 21:52.07 | dr1337_ | it's qsd8k |
| 21:52.30 | dr1337_ | libaudio is b0rked in ICS |
| 21:52.48 | dr1337_ | but i'm thinking of using this http://htc-linux.org/wiki/index.php?title=SnapdragoonAlsaDriver |
| 21:53.15 | dr1337_ | i think ICS has alsa support now |
| 21:53.24 | dr1337_ | or at least that's how the acer tablets did them in honeycom |
| 21:53.49 | jONpRY | yeah i wrote one of those for touchpad |
| 21:54.07 | jONpRY | cause the msm alsa driver is not exactly alsa compliant |
| 21:55.11 | dr1337_ | you mean the link i provided? |
| 21:55.37 | jONpRY | no |
| 21:56.22 | jONpRY | well the kernel thing might be similar. but i had to make a new libaudio for android |
| 22:09.04 | GNUtoo-dekstop | dr1337_, hi |
| 22:09.22 | GNUtoo-dekstop | you want alsa on qsd? |
| 22:12.26 | *** join/#htc-linux jONpRY (~jon@unaffiliated/jonpry) |
| 22:16.32 | dr1337_ | GNUtoo-desktop yeah that would be great |
| 22:16.37 | dr1337_ | brb updating my comp |
| 22:32.21 | *** join/#htc-linux rockosays (~rockosays@unaffiliated/rockosays) |
| 22:34.39 | *** join/#htc-linux dr1337 (78931828@gateway/web/freenode/ip.120.147.24.40) |
| 22:34.49 | dr1337 | back |
| 22:42.45 | *** join/#htc-linux d3tul3 (~detule@pool-96-234-141-27.bltmmd.east.verizon.net) |
| 22:44.25 | jONpRY | it seems pretty insistent on compiling stuff in hardware/ti/omap4 and the stuff uses invalid assembler instructions |
| 22:51.51 | jONpRY | hi d3tul3 |
| 22:51.56 | d3tul3 | hey |
| 22:52.02 | jONpRY | you see 3.1? |
| 22:52.07 | d3tul3 | i just compiled something with all smem_syncs gone |
| 22:52.24 | d3tul3 | yeah i need to checkout the branch locally so i can see the diff and hopefully learn something |
| 22:52.45 | jONpRY | its not totally finished yet. but i think it would boot |
| 22:52.58 | d3tul3 | where is it in relation say to the 39 tree |
| 22:53.08 | d3tul3 | june-ish? |
| 22:53.33 | jONpRY | the files that have been done are up to date |
| 22:53.53 | d3tul3 | btw you reunited with your rhod or doing this blind? |
| 22:54.02 | jONpRY | its missing touch screen, and i think the usb .config is not working right |
| 22:54.07 | jONpRY | going blind |
| 22:54.44 | d3tul3 | i remember there was some clutch irq commit in tssc_manager in .39 |
| 22:55.24 | jONpRY | i just didn't mess with drivers/input because it lacks any of those circular dependencies with the rest of it |
| 23:04.40 | d3tul3 | if it ever finish checking out over here i'll give it a go |
| 23:04.55 | d3tul3 | so what are all those smem_sync calls for on enter/exit sleep? |
| 23:05.03 | jONpRY | no idea |
| 23:05.11 | d3tul3 | seems to work fine without it |
| 23:05.32 | jONpRY | hopefully it works better |
| 23:07.24 | d3tul3 | jONpRY, what no binaries? |
| 23:08.42 | *** join/#htc-linux ImCoKeMaN (~imcokeman@pool-98-111-117-171.hrbgpa.fios.verizon.net) |
| 23:09.00 | jONpRY | guess there was a .gitignore |
| 23:09.11 | jONpRY | i dunno if i checked in the defconfig |
| 23:09.16 | jONpRY | and its tricky |
| 23:10.10 | jONpRY | i think this thing is actually going to build |
| 23:10.48 | arrrghhh | jONpRY, you can do more than -j2 if you want |
| 23:10.51 | arrrghhh | up to you |
| 23:10.58 | arrrghhh | i was doing -j5 on it yesterday w/o issue |
| 23:11.09 | d3tul3 | arch/arm/mach-msm/board-htcrhodium.c:1204:2: error: implicit declaration of function 'cam_gpio_init' |
| 23:11.30 | jONpRY | well i was running j5, then it died, so i started with lower j to find out where |
| 23:11.49 | arrrghhh | i see |
| 23:11.49 | Willd | arrrghhh: Why would -j5 make anything go wrong? |
| 23:11.55 | jONpRY | d3tul3, yeah thats config |
| 23:11.55 | arrrghhh | Willd, i dont' know |
| 23:12.02 | arrrghhh | i'm telling him he can use it :P |
| 23:12.07 | Willd | Heh |
| 23:12.08 | jONpRY | it doesn't make it go wrong, it just makes the build log unreadable |
| 23:12.20 | Willd | jONpRY: Oh, makes sense.. |
| 23:12.35 | arrrghhh | back to cleaning, bbl |
| 23:13.29 | WisTilt2 | detule got a second? |
| 23:14.02 | WisTilt2 | d3tul3^^ cant tell which one of you i should talk to |
| 23:15.57 | d3tul3 | what's up |
| 23:16.25 | WisTilt2 | you have microp-klt totally gone now correct? |
| 23:16.50 | d3tul3 | .c is not compiling, .h is still needed for the g-sensor |
| 23:17.19 | jONpRY | d3tul3, i just pushed a new config |
| 23:17.27 | d3tul3 | the dualmic functions are in board-htcrhodium-audio if those are the ones you are after |
| 23:18.11 | WisTilt2 | something in that transition broke part of acoustic that still relies on that. in speakerphone mode adc3001 is not enabling and getting all kind of i2c write errors. |
| 23:18.36 | d3tul3 | ugh, let me take a look |
| 23:19.12 | WisTilt2 | adc3001 doesnt look like its getting turned on now during the dualmic area so i2c goes nowhere |
| 23:19.41 | jONpRY | or arm9 has taken over the i2c |
| 23:20.13 | WisTilt2 | possibly. didn't happen until -klt was totally removed though |
| 23:20.49 | WisTilt2 | jONpRY you need a break for a minute so check this out: |
| 23:21.07 | jONpRY | ok |
| 23:21.16 | d3tul3 | WisTilt2, i have an idea what i broke |
| 23:21.25 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-epmkxszmgsybtrni) |
| 23:21.33 | WisTilt2 | i did some power tests, suspend sleep which is the default draws 42ma, collapse is 18ma. |
| 23:21.55 | WisTilt2 | we need to get idle collapse working. i bet we get to 8ma like winmo can do |
| 23:22.05 | WisTilt2 | d3tul3 fire away |
| 23:22.18 | jONpRY | but that should not effect the drain during sleep |
| 23:22.32 | emwe | WisTilt2: you aren't power collapsing on .39 or are you saying where are not PCing anywhere? |
| 23:22.42 | WisTilt2 | true... we'll gain a ton while awake though |
| 23:22.51 | WisTilt2 | emwe i am |
| 23:23.00 | WisTilt2 | its not by default though |
| 23:23.06 | WisTilt2 | its in suspend |
| 23:23.56 | WisTilt2 | the autobuild .39 will be in suspend so 42ma or thereabouts is whats drawing. im seeing 18ma on mine in collapse |
| 23:24.49 | emwe | so what are you effectively saying? msm_pm_collapse() not doing what we expect from it? |
| 23:24.50 | WisTilt2 | jONpRY my normal use stats are showing idle requests in the 95% range so we really will gain a lot of saving while awake with idle collapse wouldn't you think? |
| 23:25.15 | WisTilt2 | emwe it is but the default mode is suspend so collapse only puts it in suspend not full power collapse |
| 23:25.20 | jONpRY | yeah. like i got those 24ma numbers in winmo with panel on |
| 23:25.34 | WisTilt2 | yep exactly |
| 23:26.42 | WisTilt2 | problem with idle collapse looks like we need to do some work on the msm irqs though so it can happen. right now it always fails the test to allow any sleep at idle due to those irq's |
| 23:27.01 | jONpRY | what test? |
| 23:27.03 | emwe | you guys got that SM_NO_INT_INFO error still right? |
| 23:27.13 | WisTilt2 | if allowed to sleep fails 100% of the time |
| 23:27.27 | emwe | you haven't massaged a structure size. |
| 23:27.32 | WisTilt2 | emwe no, we commented it out:) |
| 23:28.15 | jONpRY | msm_irq_idle_sleep_allowed() ? |
| 23:28.55 | emwe | idle sleep will not fail due to missing smem interrupt info, no? just asking ... |
| 23:29.01 | WisTilt2 | jONpRY: msm_irq_idle_sleep_allowed(); always returns 0 |
| 23:30.25 | WisTilt2 | emwe yes i believe we're talking the same thing |
| 23:30.58 | emwe | was just a thought. you are more into the PM stuff then me. |
| 23:31.20 | WisTilt2 | looks like the msm_irq masks never get in the state to allow idle sleep for whatever reason |
| 23:31.26 | jONpRY | so this code basically depends all interrupts are disabled? |
| 23:31.40 | WisTilt2 | yes or have set wakes of some sort |
| 23:32.09 | jONpRY | i don't see anything about wakes |
| 23:32.31 | WisTilt2 | mach-msm/irq.c |
| 23:33.10 | WisTilt2 | smsm_irq is the culprit i believe. |
| 23:33.32 | jONpRY | looks like this voodoo only applies if msm_irq_to_smsm[irq] == 0 |
| 23:33.41 | WisTilt2 | if you follow that allowed code you'll see that it always returns 0 |
| 23:33.48 | WisTilt2 | yes |
| 23:34.21 | WisTilt2 | i just haven't followed it all the way to see how any other condition can happen |
| 23:35.06 | WisTilt2 | but im going to get this idle collapse going somehow so im working on that since you said you'd be fixing the rest of BT:) |
| 23:35.11 | WisTilt2 | im sick of it for now |
| 23:35.44 | jONpRY | this code is super simple |
| 23:35.48 | emwe | !!(null) |
| 23:36.00 | emwe | it will always return 0 if smsm_int_info isn't set, no? |
| 23:36.55 | emwe | plz correct me if i am wrong. |
| 23:37.18 | WisTilt2 | yeah, guess i looked right past that |
| 23:37.36 | WisTilt2 | thats how its been forever so its never been able to work |
| 23:39.03 | emwe | +#if defined(CONFIG_MSM_AMSS_VERSION_WINCE) |
| 23:39.03 | emwe | + uint32_t padding; |
| 23:39.03 | emwe | + uint32_t reserved[8]; |
| 23:39.03 | emwe | +#endif |
| 23:39.08 | emwe | but that will break 5225. |
| 23:39.32 | WisTilt2 | what device is 5225? |
| 23:39.39 | emwe | in smd_private.h smsm_interrupt_info |
| 23:39.44 | emwe | all but RHOD and TOPA |
| 23:39.52 | WisTilt2 | so we dont care:) |
| 23:39.53 | emwe | .27 has it setup dynamically |
| 23:43.01 | WisTilt2 | .39 only runs on rhod now anyway |
| 23:43.23 | WisTilt2 | thanks emwe |
| 23:43.57 | emwe | your welcome. was waiting for repo sync ... so had nothing else to do |
| 23:44.35 | dr1337 | yeah finally fixed gralloc issues |
| 23:44.40 | WisTilt2 | on a friday night even |
| 23:44.59 | dr1337 | now the proprietary adreno drivers are causing grief |
| 23:45.11 | WisTilt2 | im still at the office another couple hours working on this so my friday night hasn't begun yet |
| 23:46.31 | emwe | it's already saturday here :) |
| 23:49.28 | dr1337 | anyone encountered the "unable to deque native buffers" error? |
| 23:56.56 | jONpRY | dr1337, how did you fix gralloc? |
| 23:59.58 | jONpRY | i |