00:00.40 | *** join/#htc-linux Rajko (~Rajko@2001:470:1f0b:728:a0c1:ec16:3826:7162) |
00:01.12 | arrrghhh | WisTilt2, http://pastebin.com/jLQekHU6 <---- full dmesg |
00:02.01 | WisTilt2 | you have autobl off again also |
00:02.07 | arrrghhh | crap |
00:02.13 | arrrghhh | fresh boot, didn't check it |
00:02.43 | arrrghhh | WisTilt2, that is "fixed" tho now, right? |
00:03.02 | WisTilt2 | yeah you can run manual, just saying your log shows its off |
00:03.24 | WisTilt2 | notice you dont get those microp crashes in manual anymore |
00:03.28 | Alex[sp3dev] | setprop wlan.driver.status "ok" |
00:03.32 | Alex[sp3dev] | now that's a cheat |
00:05.02 | Alex[sp3dev] | E/WifiHW ( 1442): Unable to open connection to supplicant on "foo0": No such file or directory |
00:05.19 | arrrghhh | WisTilt2, heh no but that's good :D |
00:05.31 | WisTilt2 | i need to look further into that. it has to be some wake lock time that changes after some event. |
00:05.37 | arrrghhh | ok |
00:05.44 | arrrghhh | what else do you want us to test? |
00:05.50 | arrrghhh | i can't really simulate bad service. |
00:05.57 | arrrghhh | unless you were serious about the microwave thing |
00:06.32 | WisTilt2 | thats it, ril data was the main thing unless you can get bad signal data. yes serious about microwave, should knock signal down a bunch |
00:07.07 | arrrghhh | lol ok |
00:07.07 | arrrghhh | 20 mins on high? :P |
00:07.07 | WisTilt2 | and jonpry saying wrap in foil is actually good also |
00:13.48 | arrrghhh | alright gonna pull it and get you another radio log |
00:18.26 | *** join/#htc-linux Drakh_ (Drakh@190.42.40.222) |
00:22.23 | WisTilt2 | arrrghhh: got your signal down pretty low there but it stayed connected. good data though |
00:23.34 | jonpry | WisTilt2, i think i got a ramconsole from an sod. the kind that happen in airplane mode |
00:23.36 | arrrghhh | WisTilt2, awesome! let me know if you want moar. |
00:24.14 | WisTilt2 | jonpry: sounds juicy! must see |
00:24.36 | jonpry | does light turn orange before or after autobl? |
00:24.52 | WisTilt2 | on boot you mean? |
00:25.20 | jonpry | on wake |
00:25.33 | WisTilt2 | let me look |
00:25.35 | jonpry | think glemsom. green == sleep |
00:25.46 | arrrghhh | lol |
00:26.29 | jonpry | http://pastebin.com/1JGyABnz |
00:26.58 | WisTilt2 | actually, autobl is always on now if enabled. my last commit doesnt have the microp resume anymore i believe |
00:27.10 | jonpry | afaict the thing woke up |
00:27.28 | jonpry | userland and whole deal, but light was still green |
00:27.35 | arrrghhh | odd |
00:28.13 | jonpry | or not, i can't make heads or tails |
00:28.32 | jonpry | littered with garbage :p |
00:28.39 | arrrghhh | most SoD's are |
00:28.55 | arrrghhh | sometimes they're botched so bad there's nothing useful... |
00:28.59 | jonpry | nah debug printks |
00:29.36 | jonpry | but it looks like it woke up at 9719.670806 |
00:29.40 | WisTilt2 | looks like it began about 1.5 mins before it crashed |
00:29.44 | WisTilt2 | yeah |
00:29.57 | jonpry | i reset it eventually |
00:30.01 | jonpry | cause i thought it was dead |
00:30.24 | WisTilt2 | i wonder if there is something with clocks at times when a11 comes back up, that's kind of what this looks like |
00:30.41 | jonpry | the laps between 9719 and 9182 is strange |
00:31.02 | jonpry | should be logging every 5 seconds |
00:31.25 | WisTilt2 | oh, you're not correcting timestamps yet. that should have continued with the previous time |
00:31.48 | WisTilt2 | and whats with line 2694 |
00:31.52 | WisTilt2 | timestamp |
00:32.19 | jonpry | it looks like 9182 |
00:32.38 | WisTilt2 | yeah but it shouldnt be missing 3 chars [ 9 |
00:32.52 | jonpry | its probably missing a lot more, hence the jump in time |
00:32.58 | jonpry | like -1000 or something |
00:33.19 | WisTilt2 | are you suspending arm11 or full power collapse |
00:33.29 | jonpry | sleep_mode = 1 |
00:33.48 | WisTilt2 | autobuild code with no mods to pm? |
00:34.18 | jonpry | yeah |
00:34.56 | jonpry | there is quickwake in pm subsystem though. |
00:35.37 | WisTilt2 | this is the first actual dump from a sod. the fact that time is missing, along with characters, thats a start in where to look |
00:36.18 | WisTilt2 | quickwake isnt doing it i dont think. i bet this would be the same logs from most sod's, we've just never seen them |
00:36.36 | jonpry | its almost like its running, but something has gone terribly wrong |
00:36.40 | arrrghhh | dropbox isn't a good method? |
00:37.02 | WisTilt2 | dropbox doesnt get updated when sod happens |
00:37.41 | jonpry | WisTilt2, it seems obvious to me that the write pointer has been clobbered, or otherwise lost, ram refresh failure? |
00:37.55 | jonpry | look at garbage at beginning and end |
00:38.04 | jonpry | DBGC ? |
00:38.11 | *** join/#htc-linux arrrghhh_ (~arrrghhh@c-71-237-40-111.hsd1.co.comcast.net) |
00:38.12 | jonpry | thats new |
00:38.18 | *** join/#htc-linux arif-ali (~arif-ali@88-111-128-250.dynamic.dsl.as9105.com) |
00:38.22 | arrrghhh_ | blargh |
00:38.27 | jonpry | first couple bytes of ram console are the write pointer |
00:38.58 | jonpry | from there, ramconsole may actually be able to clobber the whole rest of kernel |
00:43.34 | WisTilt2 | moving ramconsole do any good? its right under the base of smi2 |
00:43.57 | jonpry | didn't you try that? |
00:44.09 | WisTilt2 | mine is running in smi2 now on my test kernel |
00:44.26 | jonpry | in smi2? |
00:44.27 | WisTilt2 | arrrghhh virtually never gets sod's and he's been running it a long time |
00:44.49 | WisTilt2 | yes, i found an area that doesnt get hit by anything in smi2 |
00:45.06 | jonpry | gimme |
00:45.30 | WisTilt2 | you want the address so you can stick it in your kernel? |
00:45.45 | jonpry | yeah |
00:45.49 | WisTilt2 | ok standby |
00:46.08 | arrrghhh | WisTilt2, how long is your list of commits now? :P |
00:46.17 | arrrghhh | you've been fixing sooo many things. |
00:47.18 | WisTilt2 | lol |
00:49.13 | arrrghhh | WisTilt2, so your new RIL... it seems like we need someone who had reproducible SoD's |
00:49.23 | arrrghhh | i have at least one dude that can produce them on command hehe |
00:50.05 | arrrghhh | or are you wanting to wait until you flesh out the code more from our logs? |
00:51.08 | *** join/#htc-linux Christie (Christine@75.1.6.172) |
00:53.07 | odz | would like to test |
00:54.01 | jonpry | mine is overclocked, is that correlated with sod? |
00:54.36 | arrrghhh | perhaps |
00:54.37 | arrrghhh | how much? |
00:54.45 | arrrghhh | how much you juicin |
00:54.49 | jonpry | 718 or something |
00:54.53 | arrrghhh | roid raging |
00:55.13 | arrrghhh | disable it, see what happens |
00:55.20 | arrrghhh | i don't OC anymore really |
00:55.52 | jonpry | why not? |
00:55.59 | jonpry | arrrghhh, are you sleep_mode=1? |
00:56.22 | arrrghhh | yes |
00:56.29 | arrrghhh | no OC tho |
00:56.43 | jonpry | and you got no sod while running scbs? |
00:56.57 | jonpry | and it was sleeping? |
00:56.58 | arrrghhh | not a one |
00:57.01 | arrrghhh | yup |
00:57.05 | arrrghhh | getting great battery life too |
00:57.16 | jonpry | hmm, then sod for you is impossible |
00:57.21 | arrrghhh | lol |
00:57.23 | arrrghhh | not necessairly |
00:57.26 | jonpry | it probably woke up 10,000 times |
00:57.29 | arrrghhh | got one driving thru glenwood canyon |
00:57.34 | arrrghhh | oh |
00:57.42 | arrrghhh | yea i ran it all night |
00:58.22 | jonpry | i can only get like 5 minutes of sleep before its caput |
00:59.00 | arrrghhh | damn no joke? that's weird. |
00:59.05 | arrrghhh | i didn't try it in airplane mode... |
00:59.10 | WisTilt2 | jonpry: sorry phone again. i have ramconsole running at smi2+22. that works on rhod 300-500 as we tested but your 210 might not so you should be safe at smi2+1 since it's small. |
00:59.38 | jonpry | ok, will try |
01:00.32 | WisTilt2 | jonpry: 718 is pushing it. 700 works ok on mine but probs at times so ive been running 614400 forever now with no issues. |
01:01.47 | WisTilt2 | arrrghhh: going to glean through the log data and wont put out a test ril until we get more stuff coded. |
01:01.57 | arrrghhh | WisTilt2, makes sense. |
01:02.08 | arrrghhh | i just can't replicate SoD due to service very well |
01:02.17 | arrrghhh | so... i'm not the best test case there. |
01:02.32 | arrrghhh | as much as i love testing :D |
01:03.11 | WisTilt2 | i actually thought a bit ago that all our fiber equipment is in our building's basement and i have virtually no service down there so going to try it down there. |
01:03.56 | arrrghhh | heh |
01:04.04 | arrrghhh | don't hang out down there much huh? |
01:04.18 | WisTilt2 | jonpry: that 22 is 22megs i assume you already know |
01:04.32 | WisTilt2 | didnt really make that clear |
01:04.44 | phh | MiB of course ? :D |
01:05.38 | jonpry | WisTilt2, i kind of figure it wasn't 2 bytes or anything |
01:06.56 | jonpry | i'm messing with lower clock atm. so far no sod |
01:07.29 | jonpry | i wonder if we need to ramp up clock on resume |
01:07.57 | jonpry | thats how smart reflex works on ti |
01:08.00 | phh | because they are ramped down on suspend ? |
01:08.28 | jonpry | no because pmic can't throttle the voltage yet |
01:08.55 | jonpry | they have some mechanism for raising the voltage when it appears that the device won't be able to work at the current clock rate |
01:09.14 | jonpry | but if you go jumping from off to 600mhz, that stuff isn't going to work |
01:12.50 | phh | well, what I thought at some point, is that resume might reset PLL from <cpu freq> to 1048 (don't remember the proper value) |
01:12.53 | phh | so cpu running at this speed |
01:12.55 | phh | and boom :D |
01:13.26 | arrrghhh | lol |
01:13.34 | arrrghhh | set it to 2048 |
01:13.37 | arrrghhh | and then come talk to me. |
01:13.58 | jonpry | phh: sounds like a good theory |
01:14.17 | phh | jonpry: that was an actual problem when porting O/C to android devices |
01:14.32 | WisTilt2 | didnt bzo's pll patch fix that possibility so it couldnt happen? |
01:14.40 | phh | I don't remember precisely, but I had to rewrite PLL frequency on every cpu frequency change or something |
01:15.27 | jonpry | argh, sod at 614mhz |
01:16.17 | phh | jonpry: try at 500MHz with overclock option to confirm. |
01:16.25 | phh | well freezing at 614MHz is definitely not normal :p |
01:17.22 | jonpry | phh: its like accelerated aging. wakes up every 5 seconds |
01:19.59 | jonpry | http://pastebin.com/0yPgE1er |
01:22.02 | jonpry | so in this log, its almost like it failed to go back to sleep. just processor randomly hung |
01:23.39 | jonpry | never gets to htc_pwrsink |
01:24.14 | jonpry | phh, is there an underclock? |
01:24.24 | phh | jonpry: what do you mean ? |
01:24.36 | jonpry | can i force it to run at 300mhz? |
01:24.46 | phh | with modded PLL or not ? |
01:25.06 | jonpry | simple is better |
01:25.27 | phh | so with normal clocks ? just write 300 to cpufreq_max_freq or something like that somewhere in /sys |
01:26.02 | jonpry | when it resumes will it use that? |
01:27.32 | jonpry | phh: you should go ahead and push the scbs patch :p |
01:27.48 | phh | the setting should be kept yes |
01:27.58 | phh | jonpry: tell acl :p |
01:28.06 | phh | i haven't even seen your patch actually |
01:28.11 | phh | ah or that's mostly userland stuff ? |
01:28.20 | jonpry | acl doesn't like to push anything. he says ask phh |
01:28.33 | jonpry | its and interface for userland stuff |
01:28.44 | jonpry | required some hacking of kernel |
01:28.55 | arrrghhh | jonpry, did you get it all automatic-like? |
01:29.20 | phh | jonpry: yeah well i'd like to see that userland stuff :p |
01:29.21 | jonpry | no, but that is a userland problem |
01:29.39 | jonpry | phh: http://gitorious.org/scbs/scbs |
01:30.34 | arrrghhh | oh lol |
01:31.18 | jonpry | only scaling_max_freq is rw |
01:32.15 | phh | jonpry: yeah that's that one |
01:32.43 | arrrghhh | bbl |
01:34.22 | jonpry | phh i have binaries |
01:34.55 | *** join/#htc-linux toastcfh (~toastcfh@29-37.200-68.tampabay.res.rr.com) |
01:35.08 | phh | ? |
01:35.09 | phh | luenberg is a friend of kalman ? |
01:35.18 | jonpry | basically |
01:36.01 | jonpry | it works on the good :p |
01:36.05 | *** join/#htc-linux bzo (~chatzilla@netblock-68-183-234-14.dslextreme.com) |
01:38.58 | jonpry | hi bzo |
01:39.20 | bzo | hi jonpry |
01:41.28 | phh | jonpry: could you link me again the kernel patch for scbs ? |
01:41.56 | jonpry | phh: i sent it to the ml, its not posted anywhere atm |
01:42.10 | phh | ah that was on the ml, ok |
01:45.10 | phh | jonpry: ah that patch is to get me is a slow-motion heart attack ? :D |
01:46.02 | phh | 1.enabled only for rhod |
01:46.10 | phh | 2.not an option (ok that's not critical) |
01:47.09 | phh | 3.deleting some mutex ? |
01:47.21 | emwe | uh... what is that makes boot so much slower when acpuclock and pm debug is enabled!? |
01:47.28 | stinebd | emwe go to bed! |
01:47.48 | emwe | lol |
01:47.51 | phh | 4.wtf is that change in pm.c ? isn't that one of our big problems ? :D |
01:49.03 | phh | 5.msmrtc_rewake.... really ? i'm pretty sure there is a proper api for that |
01:49.08 | jonpry | 3. the mutex's are all messed up in htc_battery. you pushed that garbage in the beginning :p |
01:49.33 | phh | i did some battery code ? |
01:49.40 | jonpry | i just removed some useless ones that were botching the whole deal |
01:49.49 | jonpry | no its camro's but you pushed it |
01:50.02 | phh | ah |
01:50.03 | jonpry | he's got mutexes guarding things like x = y |
01:50.07 | phh | that's what you cann "beginning" |
01:50.12 | phh | s/cann/call/ |
01:50.46 | phh | that was quite the end for me :D |
01:50.51 | phh | and yeah, I never understood that code |
01:51.00 | jonpry | and then there are mutexes that start in one subroutine, and get closed in a different one |
01:51.10 | phh | really ? |
01:51.11 | phh | ew |
01:51.16 | *** join/#htc-linux SG_away (~Samsunguy@cpe-76-186-50-40.tx.res.rr.com) |
01:51.27 | jonpry | yeah thats why the deletion look so weird |
01:51.44 | phh | a patch in a patch \o/ |
01:51.54 | stinebd | double diffing |
01:53.38 | phh | jonpry: well my only real concern is concerning the waking up thing, it seems to me you wake up every 5 seconds ? and you might conflict with android own control |
01:53.45 | jonpry | the change to pm.c may not be necessary, but seemed like a good idea |
01:54.04 | phh | it sounds so yes |
01:54.12 | jonpry | phh that only happens if scbs userland is running |
01:54.55 | jonpry | and i haven't experienced any troubles with. android doesn't seem to require that the phone be sleeping for any particular time. its just forces it to be awake |
01:55.14 | phh | the aim is that scbs always run no ? |
01:55.21 | phh | or i totally misunderstood your code ? |
01:56.30 | jonpry | yeah, that is the goal. but until userland is distributed, its not like its going to break the build |
01:56.44 | jonpry | you don't like msm_rtc_rewake? |
01:57.30 | jonpry | i thought it was a good idea because it is called when the phone is not actually awake or sleeping |
01:57.35 | *** join/#htc-linux DuperMan (~Duper@93-172-181-232.bb.netvision.net.il) |
01:57.50 | jonpry | as opposed to the other functions which may assume the phone is in some sane state |
01:58.03 | jonpry | i don't roll well with sane |
02:00.19 | *** join/#htc-linux DuperMan (~Duper@93-172-181-232.bb.netvision.net.il) |
02:02.02 | phh | why not use RTC's set_alarm function for instance ? |
02:02.05 | phh | i think there are even higher level functions |
02:02.34 | *** join/#htc-linux DuperMan (~Duper@93-172-181-232.bb.netvision.net.il) |
02:02.55 | *** join/#htc-linux Drakh_ (Drakh@190.42.40.222) |
02:03.02 | phh | haha. higher level functions are only available to userspace |
02:03.03 | phh | thanks android |
02:04.23 | jonpry | i guess the trouble was that i had to clobber alarm time to use msmrtc_suspend |
02:04.42 | jonpry | and suspend is actually only available as a pm function |
02:05.12 | *** join/#htc-linux DuperMan (~Duper@93-172-181-232.bb.netvision.net.il) |
02:05.51 | jonpry | how stupid is that :p |
02:05.56 | jonpry | i gtg |
02:06.21 | phh | jonpry: i wonder how this code can actually work |
02:06.30 | phh | it seems to me android overwrites the rtc alarm at suspend |
02:19.37 | *** join/#htc-linux Drakh_ (Drakh@190.42.40.222) |
02:23.15 | *** join/#htc-linux Drakh (Drakh@190.42.40.222) |
02:40.19 | *** join/#htc-linux SG (~Samsunguy@cpe-76-186-50-40.tx.res.rr.com) |
02:40.35 | *** join/#htc-linux arrrghhh (~arrrghhh@c-71-237-40-111.hsd1.co.comcast.net) |
02:50.24 | *** join/#htc-linux DuperGhost (~Duper@93-172-21-111.bb.netvision.net.il) |
03:12.58 | jonpry | phh: there? |
03:14.12 | jonpry | guess he does not like my patch |
03:16.52 | *** join/#htc-linux surge (surge@pool-98-118-157-221.bflony.fios.verizon.net) |
03:17.12 | dcordes | jonpry: send to mailing list ? |
03:17.37 | dcordes | [[Mailing_Lists]] |
03:17.37 | unilinky | http://htc-linux.org/wiki/index.php?title=Mailing_Lists |
03:21.10 | *** join/#htc-linux swc|666 (~x00@unaffiliated/swc666/x-4934821) |
03:22.18 | jonpry | dcordes, i sent it yesterday |
03:22.23 | jonpry | and nobody liked it then either |
03:23.10 | jonpry | then phh started ranting, and i had to cower and run. |
03:24.13 | jonpry | hrm, like a thousand sleeps at 300mhz |
03:24.59 | jonpry | arrrghhh, you hearing that :p |
03:25.19 | arrrghhh | dowha? |
03:25.40 | rpierce99 | he's not getting sods |
03:25.45 | jonpry | my sod problem is gone at 300mhz |
03:26.19 | arrrghhh | 300mhz |
03:26.21 | arrrghhh | lol wtf |
03:27.48 | rpierce99 | wishes he didn't have to click ok 18000 times to restore from backups |
03:27.52 | jonpry | er, scbs wasn't running |
03:28.05 | arrrghhh | actually paid for TB..... |
03:29.41 | jonpry | everybody wants 300mhz |
03:30.41 | jonpry | anyone try turning off proximity? this looks kind of bad. 5.002136] Unbalanced IRQ 154 wake disable |
03:30.49 | arrrghhh | i didn't realize you could force acpu down that low lol |
03:32.07 | jonpry | i did it with scaling_max_freq but i have no idea how/when/if its working |
03:32.19 | *** join/#htc-linux bzo (~chatzilla@c-76-126-175-200.hsd1.ca.comcast.net) |
03:32.44 | jonpry | maybe bzo likes my patch better |
03:33.25 | bzo | haven't really looked at it yet |
03:33.49 | jonpry | phh said it almost have him a heart attack |
03:34.18 | bzo | lol, I suppose all that waking looks a little scary |
03:36.38 | dcordes | jonpry: where di you send it ? |
03:36.51 | jonpry | xdandroid ml |
03:36.55 | dcordes | lol |
03:37.31 | arrrghhh | jb just sent his BT patches thar as well |
03:37.42 | arrrghhh | bzo's gonna be busy :D |
03:37.53 | arrrghhh | o lord of the commit |
03:38.00 | dcordes | I still don't see the reason for the isolated 7kA specific development processes |
03:38.05 | jonpry | its a little early, but i can't manage to get sod at 300mhz yet. or for that matter a slow or failed wake |
03:38.36 | dcordes | jonpry: how does it help excluding the reviews from people who work with other platforms ? |
03:38.52 | bzo | arrrghhh: I still have an unpushed front key backlight patch from jb as well |
03:38.59 | arrrghhh | Oo |
03:39.06 | arrrghhh | for all RHOD's? |
03:39.26 | bzo | should be, don't think it is sensitive to rhod variant |
03:39.32 | jonpry | dcordes: it doesn't help anything, but it allows our phone to move forward |
03:39.37 | bzo | only problem is that it doesn't work for the end key |
03:39.41 | arrrghhh | i thought wistilt2 ran into an issue |
03:39.45 | arrrghhh | lolwut |
03:40.27 | bzo | it's retarded, the first 3 keys are mapped like a keyboard key, the end key is a separate gpio button |
03:40.57 | dcordes | jonpry: yes and it allows peter to become lead guitarist and funding maintainer ... http://xandroid.com/wiki/Development_Team |
03:41.38 | *** join/#htc-linux Dinde (kayser@sur-internet.net) |
03:43.04 | jonpry | dcordes, that link is broken. and i don't catch your drift |
03:43.49 | jonpry | should i post to arm-llkml. the people who have refused all patches submitted in the past 10 years? |
03:44.23 | *** join/#htc-linux Drakh_ (Drakh@190.42.40.222) |
03:45.01 | bzo | jonpry: http://gitorious.org/linux-on-qualcomm-s-msm/linux-msm-home-work/commit/5da6216d8a41961bbda4d411532a4b58307f71b0 |
03:45.09 | arrrghhh | bzo, ah i remember hearing you (er seeing you) talk about that |
03:45.57 | bzo | jonpry: think that fixes that unbalanced irq thing you were talking about? |
03:46.19 | dcordes | jonpry: we have the mobile-linux discuss list which is also subscribed by people who don't have your msm7ka device |
03:47.12 | jonpry | oh. somebody signed me up for xdandroid, and then messages regarding patches started coming in, so i started making my own |
03:47.47 | dcordes | jonpry: it's obvious that most of our (all non-official msm/qsd kernels) don't have much potential for mainline.. it's not needed to upstream everything but I also think it makes no sense to go the other exterme and have the development in a very small circle only coping with one or few machines |
03:47.47 | *** join/#htc-linux surge (surge@pool-98-118-157-221.bflony.fios.verizon.net) |
03:48.30 | dcordes | jonpry: hence this channel, the wiki and all the other infrastructure without device specific restrictions |
03:49.10 | dcordes | it's not #htc-linux-msm7xxxA or wiki.htc-linux-msm7ka.org |
03:49.26 | dcordes | cause it wouldn't make no sense |
03:49.39 | jonpry | but xdandroid-dev it msm7k specific? |
03:50.00 | dcordes | these things existed.. #htcblueangel, vogue etc. but have less potential |
03:50.07 | dcordes | because they are restricting. |
03:50.58 | jonpry | soon everyone will be using scbs. it will ship on native android devices :p |
03:51.09 | arrrghhh | lol jonpry |
03:51.13 | dcordes | jonpry: I don't even know the website claims it is about 'a series' of devices but it always seemed to me it's raphael & friends only |
03:51.28 | dcordes | i.e. 7ka |
03:51.45 | fakker | up early, or late dcordes? |
03:51.46 | arrrghhh | dcordes, we support a boatload of devices, but you're right they're all msm7k |
03:53.18 | dcordes | fakker: early if you will. but I will go for further nap :P |
03:54.03 | jonpry | need to find an overachiever like camro to go figure out the scaling factor for all supported devices |
03:54.21 | arrrghhh | oic |
03:54.31 | arrrghhh | that's what phh meant when he said it'll only work for RHOD huh |
03:54.56 | jonpry | i dunno, it should work on all devices |
03:55.13 | dcordes | jonpry: anyway just wanted to point you to the mobile-linux-discuss [[Mailing_Lists]] if you like extra review |
03:55.13 | unilinky | http://htc-linux.org/wiki/index.php?title=Mailing_Lists |
03:55.23 | jonpry | you just have to run it with scbs -cf -vf |
03:55.27 | jonpry | dcordes, i am signed up |
03:56.29 | jonpry | i have been signed up apparently |
03:56.47 | arrrghhh | lol, against your will |
03:57.24 | jonpry | the trouble is that the last 2 messages were 4 months ago: * Einladung in mein Netzwerk auf LinkedIn Lukas-David Gorris via LinkedIn |
03:57.25 | jonpry | <PROTECTED> |
03:57.36 | bzo | I think I'm going to sign up jonpry to some lists of interest to me |
03:58.40 | jonpry | the good news is that sending it there would offer about the same chance of commit as xdandroid-dev |
03:58.56 | jonpry | j/k |
04:01.15 | bzo | I think due to the nature of the change, it's gonna take some time for people to get comfortable with it |
04:01.56 | jonpry | like most things, they will only get confortable with it, when they realize they have been running it for weeks |
04:02.23 | bzo | lol |
04:02.46 | bzo | seriously though, you ought to get it wistilt's kernel tests |
04:02.57 | bzo | errr, get it into... |
04:03.34 | jonpry | we need to get 300mhz out to the people. its awesome |
04:03.44 | jonpry | almost as good as turning off JIT |
04:04.01 | *** join/#htc-linux Rob2222 (~Miranda@p5DE4BDE2.dip.t-dialin.net) |
04:04.08 | jonpry | arrrghhh, can you contact the guy that SOD's on command and get him on 300mhz? |
04:04.21 | arrrghhh | uhm |
04:04.24 | arrrghhh | how did you do it again? |
04:04.33 | arrrghhh | he's actually in #xdandroid right now ironically |
04:04.37 | bzo | I am looking forward to getting the new battery stuff in, the results sound impressive |
04:05.56 | arrrghhh | he uses an airave, and older one and he said it SoD's it on command basically |
04:06.20 | jonpry | arrrghhh, echo 300000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_max_freq |
04:06.28 | dcordes | jonpry: the question is where you want to put the code |
04:06.34 | bzo | and for once we'll be getting some cutting edge stuff instead of some hand me downs from somewhere else |
04:06.38 | arrrghhh | jonpry, poop |
04:06.43 | arrrghhh | alrighty |
04:07.35 | jonpry | bzo: it seems to work pretty well. you can try it if your looking to waste some time |
04:08.35 | bzo | I do plan try it |
04:08.48 | jonpry | what is airave? |
04:09.05 | bzo | microcell |
04:09.13 | bzo | for sprint |
04:09.23 | arrrghhh | gives better service to people that have shitty service |
04:09.31 | arrrghhh | inside buildings etc |
04:09.48 | bzo | and another opportunity to charge you |
04:09.51 | arrrghhh | in theory it helps in rural areas as well |
04:09.59 | arrrghhh | bzo, oh yea. but in some cases they give 'em out for free |
04:10.14 | bzo | don't they count minutes/data the same as being on their network? |
04:10.14 | arrrghhh | if you bitch about service at your house, and you're supposed to be in the service area |
04:10.18 | arrrghhh | they'll hand 'em out |
04:10.27 | arrrghhh | it's just a booster for the existing network, so yes |
04:10.29 | *** join/#htc-linux Drakh_ (Drakh@190.42.70.170) |
04:10.44 | bzo | I thought it routed everything through your broadband |
04:10.48 | arrrghhh | just improves signal on your phone by grabbing signal and boosting it |
04:10.50 | arrrghhh | not sprint's |
04:10.52 | arrrghhh | that's tmo |
04:10.55 | arrrghhh | wifi calling |
04:11.03 | arrrghhh | that uses minutes too, for some god awful reason. |
04:11.11 | arrrghhh | different thingy tho ;) |
04:11.19 | bzo | oh, the sprint thing is a relay? |
04:11.44 | arrrghhh | yessir |
04:11.47 | arrrghhh | femtocell? |
04:12.30 | jonpry | and this airave is blowing his shit up? |
04:12.55 | arrrghhh | yup |
04:13.00 | arrrghhh | causes SoD's like crazy |
04:13.05 | arrrghhh | similar to bad cell service tho |
04:13.12 | arrrghhh | which isn't really your type of SoD it seems |
04:13.17 | jonpry | maybe the trouble is output power |
04:13.19 | arrrghhh | cuz you're in airplane mode the whole time right? |
04:13.23 | jonpry | right |
04:13.26 | arrrghhh | weird. |
04:13.32 | arrrghhh | these might not be the same SoD |
04:13.36 | arrrghhh | there could be multiple causes... |
04:13.45 | arrrghhh | wistilt2 in theory has fixed it in his new RIL |
04:13.55 | arrrghhh | which i also want this dude to test, but wistilt isn't wanting to release |
04:14.04 | jonpry | i'm thinking it turns up the output in bad areas. putting lots of rf in the power supply |
04:14.59 | jonpry | and you just can't run it that fast coming out of sleep. no power yet, no heat, all bad |
04:15.10 | arrrghhh | hm |
04:15.21 | arrrghhh | can't we wake it "slowly"? |
04:15.26 | arrrghhh | ramp up proc? |
04:15.30 | *** join/#htc-linux JesusFreak316 (~JesusFrea@pool-71-251-78-56.tampfl.fios.verizon.net) |
04:15.34 | jonpry | yeah if you can confirm fix, we can do something |
04:15.40 | arrrghhh | ok |
04:15.47 | arrrghhh | you seem to get a lot of 'em |
04:15.49 | arrrghhh | :P |
04:16.06 | arrrghhh | if fishingmedic can as well, that would be interesting. |
04:16.24 | jonpry | indeed |
04:16.32 | jonpry | could be celebration time |
04:16.50 | arrrghhh | salivating |
04:16.51 | arrrghhh | so many fixes |
04:17.11 | jonpry | i've notice that at 300mhz, i have to hold power for over 1 second to wake |
04:17.19 | jonpry | at 500, its an instant press |
04:17.32 | *** join/#htc-linux WisTilt2 (~wisgreg@wireless248.wirelesstcp.net) |
04:18.36 | WisTilt2 | arrrghhh are you really still in here or just a ghost? |
04:19.06 | odz | hes here |
04:19.10 | arrrghhh | shhhhh |
04:19.13 | arrrghhh | blow my cover |
04:19.17 | arrrghhh | :P |
04:19.21 | arrrghhh | WisTilt2, wassup |
04:19.39 | WisTilt2 | did my duties and took wife out for dinner so play time. are you currently running my ril setup? |
04:19.47 | arrrghhh | yup ypu |
04:19.48 | arrrghhh | er |
04:19.49 | arrrghhh | yea |
04:20.12 | jonpry | WisTilt2, i think i have fixed sod |
04:20.16 | WisTilt2 | can you do me a favor and logcat and grep for triggering: 75 and see how many you have? |
04:20.38 | WisTilt2 | fixed the weak signal sod? how |
04:20.56 | jonpry | 300mhz |
04:21.18 | WisTilt2 | ? locked cpu to 300mhz or still scaling? |
04:21.32 | jonpry | scaling_max |
04:21.38 | jonpry | i have some flashy delusional theory about why it happens on low signal |
04:21.53 | WisTilt2 | how would that stop sod? |
04:22.09 | WisTilt2 | unless cpu goes full on out of sleep i dont see how |
04:22.20 | arrrghhh | WisTilt2, i don't know what's in this thing, but it is SO freakin fast. |
04:22.40 | WisTilt2 | its my turbo kernel |
04:22.47 | jonpry | well its cold and power is not fully ramped, and pll isn't stable, so trying to hit the gates at 500mhz is not going to work |
04:23.15 | arrrghhh | WisTilt2, I never see Triggering:75.... |
04:23.26 | WisTilt2 | space after : i believe |
04:23.31 | arrrghhh | nope |
04:23.43 | arrrghhh | they're all Triggering:61, 57, 45, etc |
04:25.07 | arrrghhh | i even grepp'ed for 75, nothing that has triggering |
04:25.10 | WisTilt2 | jonpry: yeah i agree. doesnt cpu out of sleep start at bottom 2xxmhz whatever and ramp up? if its going full on that will cause problems, especially if sleeping long and nice and cooled down components |
04:25.25 | WisTilt2 | arrrghhh: any 27's? |
04:26.09 | arrrghhh | no 27's either |
04:26.20 | arrrghhh | lot's of 61's |
04:26.23 | arrrghhh | :P |
04:26.26 | WisTilt2 | how long you been running this ril since last boot? |
04:27.06 | jonpry | WisTilt2, i think it ramps but i am not sure how fast or really anything about how that works |
04:28.01 | bzo | there is some control over how quickly it ramps |
04:28.05 | WisTilt2 | need to add a delay out of sleep that limits scaling to a set point until X time has passed |
04:28.09 | arrrghhh | WisTilt2, a little over 4 and 1/2 hrs |
04:28.15 | bzo | in the userland init, it is set more aggressive than default |
04:28.54 | WisTilt2 | arrrghhh: let it run over night if possible and see if any 75's ever show up |
04:29.34 | WisTilt2 | bzo, out of sleep does it always start at lowest clock then ramp from there? |
04:29.59 | bzo | not sure, but it is controlled by the cpu governor |
04:30.09 | bzo | acpuclock just does what is requested |
04:30.18 | jonpry | apparently it is too fast :p |
04:31.50 | bzo | here's the line in init: |
04:31.52 | bzo | echo 40 >/sys/devices/system/cpu/cpu0/cpufreq/ondemand/up_threshold |
04:33.32 | WisTilt2 | another problem is in the ril caused by los or switching cells. there is an error if triggered stays in an eternal loop. our radio programmer confirmed this. im almost thinking this cpu issue could actually cause that condition. |
04:34.40 | WisTilt2 | bzo: what value does that expect, or what do the values mean |
04:35.42 | bzo | it just sets the sensitivity to how quickly the cpu ramp up, I think the numbers may be arbitrary |
04:35.56 | bzo | haven't looked in a while so don't remember what the default is |
04:36.11 | WisTilt2 | do you think they're based on number of clock cycles or time? |
04:39.08 | jonpry | arrrghhh, i'm confused, did his phone sleep or not? |
04:39.27 | arrrghhh | fishingmedic's? |
04:39.58 | jonpry | yeah |
04:40.14 | arrrghhh | sounded like he said it did |
04:40.18 | arrrghhh | took a while to wake |
04:40.29 | arrrghhh | and i guess he opened the market and it rebooted within android. |
04:41.37 | jonpry | so he gets sod on every sleep? |
04:43.16 | jonpry | WisTilt2, it seems to fix panel wakes too. any ideas on that? |
04:43.20 | jonpry | or just timing? |
04:43.51 | arrrghhh | jonpry, not every sleep i'd say |
04:44.00 | arrrghhh | his airave just makes sleep horrible. |
04:44.59 | bzo | WisTilt2: actually it looks like a percentage, the cpu needs to be at that utilization to ramp to the next higher freq |
04:45.11 | bzo | so 40 seems pretty aggressive to me |
04:45.35 | bzo | default is 80 |
04:45.48 | jonpry | that seems very strange |
04:45.54 | jonpry | and not good enough |
04:46.12 | jonpry | like when it comes out of sleep, nothing prevents it from being pegged |
04:46.19 | *** join/#htc-linux toastcfh (~toastcfh@29-37.200-68.tampabay.res.rr.com) |
04:47.18 | bzo | there is a sampling interval, so any ramping would not be instantaneous |
04:47.24 | bzo | and it should take a few steps to get up to max |
04:47.47 | *** join/#htc-linux Drakh_ (Drakh@190.42.27.201) |
04:47.54 | bzo | I suppose we ought to make sure that before we suspend, we're at the min rate though |
04:47.57 | jonpry | so the sampling interval is more the parameter we want to control on wake right? |
04:48.21 | bzo | well, both make will have an effect |
04:49.51 | jonpry | i've been at this for a bit now, and i just can't get an sod |
04:50.00 | jonpry | maybe i need an airave |
04:50.09 | WisTilt2 | jonpry: 300mhz fixed panel wakes? |
04:50.19 | jonpry | i think so |
04:50.25 | bzo | jonpry did you test running w/o OC at stock freq? |
04:50.46 | *** join/#htc-linux DuperMan (~Duper@93-172-21-111.bb.netvision.net.il) |
04:50.48 | jonpry | bzo: no i only made it down to 600mhz |
04:51.03 | jonpry | and decided to do something more drastic :p |
04:51.13 | bzo | because another factor when enabling OC is that we change the divider on the pll to derive the clock |
04:51.22 | bzo | that is probably less stable |
04:51.37 | jonpry | yeah, but i'm pretty sure that people have turned off OC and still get sod |
04:51.47 | WisTilt2 | ive been saying it was a timing issue and worked around it as best as could be done by bringing up panel quicker |
04:52.09 | WisTilt2 | wouldnt have thought cpu speed itself would have affected that though |
04:52.20 | bzo | jonpry: I think it's worth a test in your case to see what kind of difference it makes |
04:52.37 | bzo | you need to rename the oc parameter, setting it to 528 is not the same as stock |
04:52.46 | *** join/#htc-linux MassStash (~MassStash@c-67-175-41-173.hsd1.il.comcast.net) |
04:53.12 | jonpry | aight |
04:53.40 | bzo | I suspect however there is more than one cause for SOD |
04:54.03 | jonpry | i dunno. i started being able to get ramconsoles of it |
04:54.24 | jonpry | and its like the processor starts up, and hits the ground running, and then flys off the handle and stops |
04:55.23 | WisTilt2 | bzo, that brings up something interesting... does the divider on the pll change only if acpuclock is set in cmd line? if its not there and we're running at default freq is pll back to normal? |
04:55.32 | bzo | in any case, this seems like a good lead. If we're lucky once fixed, there won't be any more causes |
04:56.08 | bzo | WisTilt2: the presence of the parameter changes the divider, even if set to 528 |
04:56.27 | bzo | keep in mind there are 2 different levels of divider |
04:56.40 | bzo | the one I'm referring to is the master divider off the pll |
04:56.54 | bzo | from there, another divider is applied to get the desired frequency |
04:58.16 | WisTilt2 | jonpry: with your sod gone, you do have oc set or took it out of startup.txt? |
04:58.25 | jonpry | i had it set |
04:58.34 | jonpry | but now i am trying oc off, no under clock |
04:59.06 | *** join/#htc-linux programmer8922 (~Evan@156.110.78.2) |
04:59.30 | jonpry | my sod off, is like it being impossible. with the accelerated aging and all |
04:59.47 | WisTilt2 | so that means that the divider is in use right bzo? if jonpry is oc'd 600mhz and capping max to 300mhz, what is the divider equating to off that 600mhz? |
05:01.06 | bzo | would have to look at the stepping tables in acpuclock, I doubt there is a 300mhz setting, so it picking the next lowest |
05:01.17 | bzo | at only the table will tell you what pll is being used |
05:01.36 | bzo | the oc setting only changes the master divider of pll2 |
05:02.00 | bzo | the beginning of the dmesg will tell you what the table looks like (it will be different for any give OC setting) |
05:02.06 | WisTilt2 | im trying to correlate this to what i think is throwing off the vsync clock. |
05:02.39 | WisTilt2 | i need to actually see which clock vsync fires off of |
05:02.42 | bzo | pll2 is only used for the cpu, so I wouldn't think it would affect vsync |
05:04.38 | WisTilt2 | jonpry: with max freq locked to 300mhz is it taking longer to wake the panel or about the same? |
05:04.58 | jonpry | about the sam |
05:05.15 | jonpry | but my kernel has mod where it wakes it 30 times just for fun |
05:05.32 | bzo | btw would have to double check the table, but it looks like if you cap freq at 300mhz, it will only go up to 256mhz and never touch pll2 |
05:06.15 | WisTilt2 | hmm, not a good test then:) im betting it will take longer to come up and if it does, vsync is getting its clock off that somehow. |
05:07.12 | jonpry | WisTilt2, it takes longer, like 2 seconds or something. but it used to have only marginal improvement in failed wakes |
05:07.20 | *** join/#htc-linux Drakh_ (Drakh@190.42.27.201) |
05:07.25 | jonpry | got rid of 1 of 3 kinds of something |
05:07.50 | jonpry | but the 1 kind was a big problem with scbs because it used 10ms of time during wake, srewing the whole thing up |
05:08.20 | *** join/#htc-linux DuperMan (~Duper@93-172-21-111.bb.netvision.net.il) |
05:08.49 | bzo | ha, scbs is like the prime95 of suspend/wake testing |
05:09.32 | bzo | for those who don't know prime95 is the overclocker's tool of choice to test stability |
05:09.45 | jonpry | i thought it was superpi |
05:09.57 | bzo | that test is kinda useless, too short |
05:10.13 | bzo | running prime95 overnight is considered to be fully stable |
05:10.43 | jonpry | it helps to be able to reproduce the problem :p |
05:13.30 | jonpry | does sod happen on others? topaz, kaiser, etc? |
05:13.59 | arrrghhh | AFAIK yes |
05:14.33 | jonpry | i am having considerable trouble getting an sod with no oc |
05:14.55 | arrrghhh | did you try to run with no OC? |
05:14.58 | arrrghhh | i don't mean underclock |
05:15.01 | arrrghhh | 528mhz :P |
05:15.22 | jonpry | yeah i'm on stock clock now |
05:15.32 | arrrghhh | hrm |
05:15.35 | WisTilt2 | bzo: what is the turbo table in acpuclock? |
05:15.36 | jonpry | however i think this is way less cool than oc and 300mhz |
05:15.52 | bzo | pretty much all new htc devices |
05:15.53 | arrrghhh | wouldn't surprise me if there were several triggers for SoD |
05:16.07 | jonpry | cause if we can wake with oc enabled, then we can ramp, and everything is cool |
05:16.08 | bzo | they OC pll1, apparently to get a faster ebi clock |
05:16.29 | WisTilt2 | so we're using those turbo tables for he most part? |
05:16.37 | WisTilt2 | he=the |
05:17.11 | bzo | rhodium and probably topaz are turbo, others like diam,raph are non-turbo |
05:17.34 | *** join/#htc-linux Dinde (kayser@sur-internet.net) |
05:19.29 | bzo | :q |
05:19.38 | bzo | oops, wrong window for vi |
05:19.45 | jonpry | underclocking seems to invoke a very negative perception |
05:20.38 | bzo | I just noticed that the power collapse frequency is set to 19.2mhz, one that is not normally used in wake |
05:20.46 | bzo | that may cause more transition than is necessary |
05:22.18 | bzo | WisTilt2: maybe running this slow coming out of wake is screwing up the panel timing |
05:22.23 | *** join/#htc-linux Drakh_ (Drakh@190.42.27.201) |
05:23.19 | jonpry | its so strange that phone that works fine at 720, can't wake itself up at 600 |
05:23.23 | *** join/#htc-linux Kasjopaja23 (~Tina@p578FFEF3.dip.t-dialin.net) |
05:23.48 | jonpry | how sure are we that ramping is actually happening? |
05:24.13 | bzo | turn on the debug code in acpuclock, and you will be flooded with messages ramping up and down |
05:24.25 | jonpry | but non oc 528 looks pretty good. |
05:24.59 | jonpry | it must be ramping awfully fast then, cause it is crashing before the light turns orange |
05:25.27 | arrrghhh | oh WisTilt2 now that you're here, can you describe the issue you had with the 4-button lights to bzo ? I don't know the details, but i thought you got it to work on some RHOD's but not others...? |
05:25.34 | bzo | it would think it would crash immediately if ramping onto an unstable pll |
05:26.35 | jonpry | in that case oc 528 should crash as well right? |
05:26.54 | WisTilt2 | arrrghhh: button lights work fine on all phones, it was the kbd backlight we had probs on. i just havent got back to finishing cleaning up that button code to commit it. |
05:27.14 | arrrghhh | oh, i thought there was some RHOD that you had problems with. |
05:27.14 | WisTilt2 | we did have some issues in the beginning with those but that was long time ago |
05:27.21 | bzo | jonpry: not necessarily, but it would narrow down the cause - i.e. whether changing master divider causes instability |
05:27.28 | arrrghhh | well i guess jb made a commit, and they're having issues with the end key? |
05:28.28 | bzo | there's no mystery around it, just a vexing thing to decide how to fix |
05:28.47 | arrrghhh | well WisTilt2 fixed it AFAIK |
05:29.17 | arrrghhh | just tryin to help, without duplicating efforts ;) |
05:29.31 | WisTilt2 | jb got button lights done? good, then i dont have to finish that cleanup now:) end key shouldnt be a problem, its one led lighting all 4 buttons, unless there is independent leds i didnt find. mine turned all on and off together if you remember when testing. |
05:29.42 | bzo | the cause is well understood, the options to fix just annoying |
05:30.00 | arrrghhh | WisTilt2, i do... |
05:30.12 | bzo | WisTilt2: the issue is that 3 keys are mapped as keyboard keys, and the end as a gpio button |
05:30.30 | bzo | there's no easy way to get notified when the end button is pressed |
05:30.33 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
05:30.46 | bzo | without some hacking of the gpio driver |
05:30.47 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
05:30.47 | WisTilt2 | ah, to get them to come on yes. end uses a gpio |
05:31.04 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
05:31.18 | bzo | right, microp knows about he first 3 keys but not the end |
05:31.18 | jonpry | gpio keypad? |
05:31.36 | bzo | yeah |
05:31.55 | bzo | it has no provision for a call back, it just sends it to the input system |
05:32.00 | arrrghhh | oic, for triggering. i was thinking just turning it on, d'oh. |
05:32.34 | WisTilt2 | bzo: yep. i kludged it via irq but thats certainly not the right way to do it. |
05:32.39 | bzo | I suppose you could register a listener to the input system, but that seems overkill to monitor one key |
05:34.21 | bzo | anyways, I'm leaning towards just letting userland control it |
05:34.52 | jonpry | will it pass the phh test? |
05:35.21 | bzo | I would think so, that is the proper place for it |
05:35.27 | WisTilt2 | back on the ramping... when we suspend, is the cpu clock frozen to that level and resumes there when wake or does wake start things back up at the bottom? |
05:36.44 | jonpry | things could be different in theory and practice |
05:37.01 | bzo | the answer is in pm.c, it is aware of acpuclock |
05:37.18 | WisTilt2 | im watching cpu freq in system panel and it looks like its resuming where it was at suspend. |
05:37.32 | bzo | yeah, that's what the code looks like to me |
05:37.32 | WisTilt2 | that could just be delay to update though |
05:37.50 | bzo | if (sleep_mode <= MSM_PM_SLEEP_MODE_RAMP_DOWN_AND_WAIT_FOR_INTERRUPT) |
05:37.59 | bzo | acpuclk_set_rate(pm_saved_acpu_clk_rate, SETRATE_PC) |
05:38.07 | jonpry | ok, so i got an sod with 528mhz oc |
05:38.19 | arrrghhh | jonpry, you removed the acpu statement or...? |
05:38.25 | jonpry | but not with no oc. confidence interval = 50% |
05:38.41 | bzo | damn, guess there's a reason qualcomm runs the pll clock at 2x |
05:39.01 | jonpry | but that doesn't really mean anything |
05:39.19 | jonpry | i mean, since your change to pll, i had to go down from 798 to 720mhz |
05:39.32 | jonpry | but whatev. its a little jitter, that costs you hertz |
05:39.44 | bzo | which change to pll? |
05:40.01 | jonpry | i think it was the divider |
05:40.10 | jonpry | was in december or something |
05:40.19 | bzo | right the wake fix for OC |
05:40.24 | jonpry | right |
05:40.49 | jonpry | so like a 450mhz with this divider should be the same as 528 with the other or something. they needn't be the same |
05:40.55 | bzo | hmm, that could be the cause of instability |
05:41.04 | bzo | maybe not waiting long enough after changing the divider? |
05:41.10 | WisTilt2 | it was happening before then |
05:41.10 | bzo | that would happen right after wakeup |
05:41.39 | bzo | that's what makes me think there may be more than one type of SOD |
05:42.24 | jonpry | i dunno fishingmedic has the ultimate to be caused by radio sod, and he can't get one on 300mhz |
05:42.39 | jonpry | he can turn on that box and crash it, lol |
05:44.30 | bzo | there's a 50ms delay for that oc wake fix, maybe that is too high, lol |
05:44.37 | *** join/#htc-linux massstash (~massstash@c-67-175-41-173.hsd1.il.comcast.net) |
05:44.38 | bzo | if it's hung there you could see how it could lose a vsync |
05:44.44 | jonpry | unstable pll, is not the same as high jitter pll |
05:45.15 | bzo | I'm talking about the lost vsync problem WisTilt2 was asking about |
05:45.33 | bzo | I'm assuming that nothing else can run while in that mdelay? |
05:45.47 | WisTilt2 | looks like vsync gets its clock from ebi clock, does that sound right? |
05:46.05 | jonpry | yeah, or if the stupid thing is preserving clock rate across sleeps. then wake timing will be way different. you know, like 6 different scenarios |
05:46.14 | bzo | not sure but probably, most things are clocked off pll1 |
05:46.59 | bzo | WisTilt2: for the lost vsync panel wake issue, the kernel is still alive right? just the panel is stuck? |
05:47.16 | WisTilt2 | yes |
05:47.31 | bzo | ok, so it is a totally different SOD |
05:47.48 | bzo | (SOD to the user's point of view) |
05:48.02 | jonpry | it will wake up in a few |
05:48.11 | WisTilt2 | yes and with that one every single time we get the msm fb vsync errors timeouts |
05:48.23 | bzo | it recovers from the lost vsync? |
05:48.27 | jonpry | yeah |
05:48.30 | *** join/#htc-linux programmer8922 (~Evan@67.219.164.162) |
05:48.40 | jonpry | but it has to timeout first or something |
05:48.45 | bzo | ah, not a SOD then |
05:48.54 | WisTilt2 | it does recover but sometimes its almost a minute |
05:49.37 | bzo | WisTilt2: I think it's worth testing a smaller delay, 50ms is excessive, maybe it should be more like 50us |
05:49.49 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
05:50.17 | WisTilt2 | you talking about a delay in acpuclock? |
05:50.33 | bzo | the delay I put in my oc wakeup fix |
05:50.39 | bzo | yes, acpuclock |
05:51.37 | *** join/#htc-linux GNUtoo|laptop (~gnutoo@host226-134-dynamic.25-79-r.retail.telecomitalia.it) |
05:52.16 | WisTilt2 | you have it at 50ms right now? yeah that might be way too much. what im seeing is what seems to be panel unblank being called before clocks are fully up. |
05:52.31 | bzo | yeah, most of the other delays to wait for pll to settle in acpuclock are udelay(50) |
05:52.51 | WisTilt2 | btw, the last cpu clk rate is what it resumes with according to the current pm code. |
05:53.05 | bzo | yes, that's how it seemed to me |
05:53.26 | jonpry | WisTilt2, i think if you change that all will be well |
05:53.28 | WisTilt2 | that could be easily changed to drop it to the bottom rate |
05:53.57 | bzo | wtf is the cpu_relax() function |
05:54.13 | bzo | I could use a me_relax() function |
05:54.24 | arrrghhh | lol |
05:54.41 | arrrghhh | bzo, it's the scotch function |
05:54.54 | WisTilt2 | whats this bring_me_another_beer(int quick) function? |
05:55.00 | arrrghhh | lol |
05:55.07 | bzo | arrrghhh: right, a wrapper to drink_scotch() function |
05:55.08 | arrrghhh | wife: because beer is heavy |
05:55.41 | bzo | user customizable though, for jonpry it would be configured for drink_rum() |
05:55.58 | arrrghhh | hehe |
05:56.00 | jonpry | and it exports arrrghhh |
05:56.26 | jonpry | you know piracy and all |
05:56.30 | arrrghhh | :D |
05:56.40 | arrrghhh | no... i have no idea what you're talking about... |
05:56.49 | arrrghhh | wink wink nudge nudge saynomore saynomore |
05:57.28 | WisTilt2 | bzo: dumb question here but shouldn't the scaled freqs be multiples of 19200? |
05:57.35 | arrrghhh | they are |
05:57.42 | WisTilt2 | they're not |
05:57.49 | arrrghhh | no? |
05:57.51 | WisTilt2 | in the tables anyway |
05:57.56 | arrrghhh | they're supposed to be then ;) |
05:58.12 | bzo | only the master clocks needs to be a mult of 19200 |
05:58.13 | WisTilt2 | unless my math is way off |
05:58.23 | bzo | if you apply a divider to the clock, it won't necessarily be a multiple |
05:58.57 | WisTilt2 | so the pll's do not need to be? |
05:59.12 | bzo | by master clocks, I mean pll |
05:59.26 | bzo | the freqs in that table are some pll freq / something |
06:00.32 | jonpry | is cpu_relax something it can do that won't be affected by poor quality clock? |
06:01.16 | bzo | I would think anything executing with an unstable clock is liable to fail |
06:01.52 | *** join/#htc-linux DuperMan (~Duper@93-172-21-111.bb.netvision.net.il) |
06:01.58 | jonpry | maybe some reset vector halt stuff |
06:02.57 | bzo | jonpry: know what a compiler barrier is? that's what cpu_relax() is supposed to be |
06:03.17 | jonpry | know i don't |
06:03.20 | jonpry | er |
06:03.35 | jonpry | avoids reordering? |
06:03.56 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
06:04.17 | *** join/#htc-linux Kasjopaja23 (~Tina@p578FFEF3.dip.t-dialin.net) |
06:04.18 | bzo | not sure, has something to do with yielding, also some talk of a memory barrier |
06:04.24 | WisTilt2 | bzo: looks like all the other delays are already udelay except that one in PLL_2 |
06:04.44 | bzo | sounds like we found the panel resume culprit |
06:04.49 | WisTilt2 | 50ms is a long time |
06:04.57 | bzo | yeah, an eternity |
06:05.39 | jonpry | so like line 312 in pm.c, if we just get rid of that, it will be cool right? |
06:05.41 | WisTilt2 | thats your patch, you want to patch that:) |
06:06.09 | WisTilt2 | need delay. udelay(50) |
06:06.15 | bzo | WisTilt2: I think fixing that, and forcing resume to start at 128mhz would be a good next test kernel for you |
06:06.24 | WisTilt2 | yes me too |
06:06.38 | odz | agrees |
06:06.45 | WisTilt2 | i might do that right now and get arrrghhh to put things in motion |
06:06.58 | arrrghhh | Oo |
06:07.00 | odz | awesome |
06:07.03 | arrrghhh | springs into motn |
06:07.10 | WisTilt2 | if someone would push jonpry's batt stuff i could have that in there also:) |
06:07.10 | arrrghhh | s/motn/motion/ |
06:07.19 | arrrghhh | heh can't correct that can you apt |
06:07.25 | arrrghhh | WisTilt2, lol |
06:07.26 | jonpry | WisTilt2, or you could apply |
06:07.30 | arrrghhh | bzo, do eeeeet |
06:07.31 | arrrghhh | :P |
06:07.54 | WisTilt2 | jonpry: is that ready to run without any user intervention? |
06:08.02 | arrrghhh | no |
06:08.04 | bzo | WisTilt2: you'll have to apply jonpry's stuff locally to your tree, doesn't sound like phh is ready to accept yet |
06:08.12 | arrrghhh | userland magic needs to be done |
06:08.23 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
06:08.38 | jonpry | WisTilt2, i think the kernel stuff is done. but userland needs work. putting the kernel stuff in doesn't do any harm when there is no userland |
06:08.58 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
06:09.06 | WisTilt2 | so if i stick it in the kernel, users will still need to push scbs over and do the whole chmod and mknod thing? |
06:09.09 | jonpry | bzo, i think phh will crack. he just needs to realize there is no better way |
06:09.16 | jonpry | correct |
06:09.40 | arrrghhh | and run scbs manually |
06:09.42 | jonpry | but only if they want to. just works like it used to if no scbs is done |
06:09.45 | arrrghhh | and crunch data |
06:09.48 | arrrghhh | etc :P |
06:09.57 | arrrghhh | jonpry, it is quite elegant. |
06:10.02 | bzo | I think getting this into WisTilt's kernels will push things forward on the scbs front. That will alleviate concerns about the frequent wakes |
06:10.08 | arrrghhh | not sure how well it actually works in reality. hoping to run it moar. |
06:10.28 | jonpry | you just have forgotten how bad the old meter was already :p |
06:10.36 | arrrghhh | lol |
06:10.43 | arrrghhh | i honestly haven't run it much :/ |
06:10.49 | bzo | hey, it's accurate to at least +/-30% :P |
06:10.51 | arrrghhh | other than that one night. |
06:10.54 | arrrghhh | lmao |
06:11.28 | jonpry | bzo the really cool part is when it swings 60% during a plug in |
06:11.51 | bzo | say, will improving the meter fix the incomplete charging problem? |
06:11.55 | arrrghhh | that's a feature |
06:12.03 | WisTilt2 | jonpry: last scbs.c and .h i have is 03/09 |
06:12.14 | jonpry | i don't know that there is an incomplete charging problem |
06:12.27 | jonpry | WisTilt2, do you get xdandroid-dev? |
06:12.40 | bzo | doesn't charge as full as winmo fo rme, maybe 90-95%? |
06:12.47 | WisTilt2 | oh yeah, i saw you had the patch on the ml |
06:12.47 | bzo | turns off charging current prematurely |
06:13.21 | WisTilt2 | ok let me make all these clock, batt, panel, whatever changes before my brain fails |
06:13.32 | arrrghhh | WisTilt2, lol |
06:14.06 | WisTilt2 | jonpry: thanks for the email:) |
06:14.13 | jonpry | oh sure |
06:14.31 | jonpry | bzo, afaik arm9/pmic takes care of charge termination |
06:14.53 | jonpry | but it can take a long long time. and android might tell you its done, when its not |
06:15.20 | arrrghhh | jonpry, so assuming i do get a testing kernel, can i publish the necessary info on scbs so people who do want do use it, can? |
06:16.04 | jonpry | i want to make a rootfs. so you don't need to add the dev entries and stuff |
06:16.14 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
06:16.41 | arrrghhh | okie |
06:16.48 | bzo | jonpry: looks like htc_batt can set charging to off/slow/fast, i suspect it is getting set to slow prematurely |
06:16.51 | jonpry | feel free to publish whenever |
06:16.52 | *** join/#htc-linux programmer8922 (~Evan@67.219.164.162) |
06:17.09 | jonpry | bzo, but that is the rpc |
06:17.26 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
06:17.41 | bzo | nah, gpios |
06:17.49 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
06:18.00 | *** join/#htc-linux Drakh_ (Drakh@190.42.27.201) |
06:18.01 | bzo | look at battery_charging_ctrl() |
06:18.23 | jonpry | ic |
06:18.35 | jonpry | i guess if that stuff works its possible |
06:18.47 | jonpry | i wouldn't put too much stock in it though |
06:19.38 | jonpry | i mean we have keyboard buttons and bluetooth interrupts switched. how would charge enable get figured out |
06:20.37 | bzo | pretty sure it does, we had a problem early on because slow charge was used and not charging the batt |
06:20.45 | bzo | fast enough |
06:21.04 | jonpry | what about charge_enable? |
06:21.12 | *** join/#htc-linux Kasjopaja23 (~Tina@p578FFEF3.dip.t-dialin.net) |
06:21.28 | jonpry | we we can definitely figure out something, i had thought it necessary |
06:21.35 | jonpry | er hadn't |
06:22.13 | bzo | which charge_enable? |
06:22.29 | jonpry | .gpio_charger_enable = RHODIUM_CHARGE_EN_N, |
06:22.33 | *** join/#htc-linux Kasjopaja23 (~Tina@p578FFEF3.dip.t-dialin.net) |
06:23.15 | bzo | that just defines the gpio pin that is used by that function, no? |
06:23.23 | jonpry | yeah |
06:23.34 | jonpry | you asked which one? |
06:23.48 | *** part/#htc-linux Kasjopaja23 (~Tina@p578FFEF3.dip.t-dialin.net) |
06:24.00 | bzo | I didn't know what you were asking about: "what about charge_enable?" |
06:24.18 | jonpry | oh, you think it works :p |
06:24.34 | bzo | yes |
06:25.01 | bzo | on the phone side the battery charging is very simplistic |
06:25.29 | bzo | the chip in the battery takes care of the sophisticated stuff like termination |
06:26.21 | jonpry | i dunno. i'm pretty sure the electronics in the battery actually just prevent over discharge. pmic is doing most everything else |
06:26.50 | *** join/#htc-linux Kasjopaja23 (~Tina@p578FFEF3.dip.t-dialin.net) |
06:27.36 | bzo | to me those things seem the same (charging logic = prevent overcharge) |
06:27.58 | *** join/#htc-linux AgentBlair (~AgentBlai@adsl-67-124-200-190.dsl.sndg02.pacbell.net) |
06:28.34 | jonpry | early charge termination is usually used to prevent too many cycles on the battery. like you try to enforce some kind of thing where you won't try again until its down to at least 90% or so |
06:29.25 | bzo | yeah, i wouldn't expect the battery ic to do that. Just prevent over charge/discharge |
06:32.30 | bzo | jonpry: btw, you think udelay/mdelay are busy waits? |
06:33.03 | jonpry | spin loop no? |
06:33.22 | bzo | for micro sec resolution, you would think so |
06:33.52 | bzo | point being, you can see how spinning for 50ms could cause some instability right after a clock change |
06:34.22 | jonpry | not really |
06:34.56 | jonpry | why would that be worse than doing anything else after a clock change |
06:35.34 | bzo | most other things probably are blocking, preventing maxing out the cpu |
06:35.55 | jonpry | there no such thing as a block |
06:36.13 | jonpry | theres just reducing clock speed, or collapsing power |
06:36.49 | bzo | sure the cpu runs at the clock freq, but there's a big diff in noop vs doing work |
06:37.13 | jonpry | on some arch's there is |
06:37.33 | bzo | it must be that way on arm, given the focus on power consumption |
06:37.38 | jonpry | but like on x86 that is fairly recent, like pentium III i think |
06:37.55 | bioterror | booting Rhobuntu 9.04 19_09_2010 on htc kovsky is really trivial |
06:38.02 | bioterror | someties it boots, sometimes it doesnt |
06:38.13 | jonpry | i dunno, to keep power consumption down you just want minimize the number of function units in action |
06:38.38 | jonpry | so an op that decodes to like r0=r0, doesn't really fire up much |
06:39.53 | jonpry | but anyways, on arch's with flashy nop, udelay can be implemented with a bunch of nops per loop |
06:40.46 | bzo | yeah, I suppose these busy waits may actually be pretty well optimized |
06:45.00 | WisTilt2 | jonpry: your quickwake kinda took me by surprise when compiling. thought i forgot to set the environment there for a minute |
06:45.45 | jonpry | yeah needs a new config |
06:45.58 | arrrghhh | WisTilt2, going to bed soon. do you think you'll have a new kernel or should i just go with that RIL test kernel and see if i get any of those 75's? |
06:45.59 | WisTilt2 | going to boot this kernel up and make sure it works with all these goodies in it then arrrghhh cant stop drooling |
06:46.07 | arrrghhh | lol |
06:46.11 | jonpry | it builds? |
06:46.11 | arrrghhh | just tired mang :P |
06:46.28 | jonpry | i thought applying the patch may be hard on your branch |
06:46.28 | WisTilt2 | go with this one, if you havent gotten 75's by now then its not going to happen, no biggie |
06:46.42 | arrrghhh | heh ok. what are you lookin for? |
06:46.49 | arrrghhh | with these mysterious 75's |
06:47.00 | WisTilt2 | built fine no errors to speak of other than lots of whitespaces during the apply |
06:47.16 | arrrghhh | kernel pack for masses? |
06:47.16 | *** join/#htc-linux |Jeroen| (~jeroen@d5152B25B.access.telenet.be) |
06:47.43 | WisTilt2 | those are a type of beacon that i guess cdma towers dont send |
06:47.50 | arrrghhh | hrm |
06:47.50 | arrrghhh | ok |
06:47.59 | WisTilt2 | yeah pack shortly after i make sure this all boots up |
06:48.05 | arrrghhh | okie |
06:48.07 | bzo | WisTilt2: does a wakeup happen on radio changes, like switching/losing towers? |
06:49.03 | WisTilt2 | depends, switching normally no but if los and re-registration happens then yes |
06:49.36 | jonpry | that would explain the low signal correlation |
06:49.39 | bzo | just wondering if maybe that condition is like lite version of scbs, and causing lots of wakes and potentials SODs |
06:50.27 | WisTilt2 | scbs doesnt fully wake everything does it jonpry? btw, kernel booting so far so good |
06:50.57 | jonpry | WisTilt2, it used to. but it wakes enough for acpu and to break |
06:51.14 | WisTilt2 | does it make it to microp resume at all? |
06:51.23 | jonpry | no resumes |
06:53.56 | WisTilt2 | ok at lock screen just like normal:) i have wake cpu clk at 128mhz and fixed the udelay. over a dozen sleep/wakes and this puppy is turning on screen instantly now every time so far. this is with my pm though with the fast sleep but looking good! |
06:54.12 | arrrghhh | :D |
06:54.23 | arrrghhh | WisTilt2, will this be RIL-test friendly as well? |
06:54.28 | arrrghhh | not sure what you changed in that kernel TBH |
06:54.34 | WisTilt2 | damn, that clock did it. the screen is on almost exact same time orange comes on now |
06:54.47 | arrrghhh | sweet! |
06:54.55 | arrrghhh | drooling uncontrollably |
06:55.12 | WisTilt2 | arrrghhh you can run this kernel with my ril test system but dont know what it will break |
06:55.12 | odz | O___________O |
06:55.12 | arrrghhh | some might say frothing at the mouth |
06:55.23 | arrrghhh | WisTilt2, lol |
06:55.23 | *** join/#htc-linux kvaster (~kvaster@vpn-e0.bas-net.by) |
06:55.40 | arrrghhh | WisTilt2, so which would you rather me run overnight? |
06:55.46 | arrrghhh | i guess that is the question. |
06:55.59 | WisTilt2 | ok kernel pack is uploaded. run this kernel and ril image and see how it works |
06:56.05 | arrrghhh | kk |
06:56.23 | WisTilt2 | oh yeah..... this is definitely much much better bzo |
06:56.38 | arrrghhh | WisTilt2, a278f11c95f2e3a4bdc8f24800a2c025 |
06:56.47 | WisTilt2 | panel comes on out of sleep just as fast as if it was just blanked and still powered on |
06:56.47 | arrrghhh | gotta be new |
06:56.51 | arrrghhh | based on timestamps |
06:56.59 | WisTilt2 | yep thats it |
06:58.10 | jonpry | WisTilt2, i think i may have included the wake it up 30 times in that patch |
06:58.37 | WisTilt2 | what file you put it in? |
06:58.37 | jonpry | you may want to check for a function called power_on_panel2 in msm_fb.c |
06:59.07 | WisTilt2 | i would have gotten a conflict on that file applying i think. i have heavily modded that file |
06:59.17 | WisTilt2 | or would it still apply it? |
06:59.21 | jonpry | its in the patch |
07:00.20 | bzo | well, not necessarily a bad thing for this test. That will be some serious stress testing |
07:00.28 | WisTilt2 | yeah its in there. loop 10 times |
07:01.11 | jonpry | i find it to help. but maybe its all fixed now |
07:01.14 | *** part/#htc-linux Kasjopaja23 (~Tina@p578FFEF3.dip.t-dialin.net) |
07:01.44 | WisTilt2 | so this sets up work to start the panel up sequence? that means all my logging is going to be in dmesg 10 times each wake lol |
07:02.10 | jonpry | no its started as delayed work |
07:02.16 | jonpry | then it hammers the panel |
07:02.27 | arrrghhh | WisTilt2, market hammering the RIL with multiple dls. let's see if it reboots within android... |
07:02.32 | bzo | hmm, maybe that's not so useful for this test then |
07:03.07 | jonpry | theory was that vsync was not being started somehow. like mddi_remotewrote was getting botched |
07:03.28 | WisTilt2 | arrrghhh: it might still crash if you run too many. i only have it handling 10 streams now but that will be a good test. |
07:03.29 | jonpry | so i figure if it works sometimes, then if we do it 10 times, it will definitely work :p |
07:03.58 | bzo | only problem is then we don't know exactly what fixes the problem in this test |
07:03.58 | WisTilt2 | where you calling that from jonpry? |
07:04.07 | WisTilt2 | bzo agree |
07:04.24 | WisTilt2 | i can remove it and test locally real quick |
07:04.33 | jonpry | INIT_WORK(&msmfb->resume_work, power_on_panel2); |
07:04.51 | jonpry | you can remove the 2 from that line |
07:04.58 | jonpry | and it will be back to norml |
07:05.52 | WisTilt2 | isnt there a panel on already though with same name? |
07:06.04 | jonpry | it doesn't have the 2 |
07:06.19 | WisTilt2 | yeah if i remove the 2 we'd have 2 funcs with same name |
07:06.27 | jonpry | ugh, no in the call |
07:06.35 | jonpry | INIT_WORK(&msmfb->resume_work, power_on_panel); |
07:06.36 | WisTilt2 | ah got it |
07:06.47 | jonpry | 2 calls 1 10 times |
07:07.24 | bzo | LOL, just saw what scbs is abbrev for |
07:07.26 | WisTilt2 | arrrghhh test it but dont upload it yet, might have a modded one |
07:07.50 | arrrghhh | WisTilt2, sure. |
07:08.01 | jonpry | :) |
07:08.07 | arrrghhh | panel has trouble waking with all these downloads |
07:08.12 | arrrghhh | but it hasn't rebooted |
07:08.17 | arrrghhh | would've a long time ago usually. |
07:08.41 | jonpry | spontaneous reboot? that is a current bug? |
07:09.24 | arrrghhh | jonpry, i could consistently recreate it by deleting my data.img and going to the market. it would download 5+ updates, and the phone would slow to a crawl and eventually just choke and reboot to bootani |
07:09.34 | arrrghhh | after that it was fine lol :P |
07:11.09 | jonpry | arrrghhh, if the phone was not sleeping then i don't really see how anything would have changed |
07:12.02 | WisTilt2 | whoot! still waking like it was never powered off |
07:12.13 | arrrghhh | jonpry, eh? |
07:12.43 | arrrghhh | market is downloading a lot of apps. it seems like my button presses aren't registered, or are delyaed |
07:12.49 | arrrghhh | once they do register, panel wakes quickly in fact. |
07:13.23 | bzo | well, that's satisfying progress for one evening. doubly so if the SOD is also fixed |
07:13.48 | arrrghhh | no joke |
07:14.00 | arrrghhh | bzo, have you looked at the BT commits? :D |
07:14.03 | WisTilt2 | this almost makes me think its not going into sleep but logs show arm11 collapsed and everything really is down, nice! |
07:14.16 | jonpry | arrrghhh, yeah i've been seeing the 1 second press requirement |
07:14.25 | bzo | arrrghhh: yes, but they won't be useful until jb is ready with the rest of the stuff |
07:14.32 | arrrghhh | userland? |
07:14.33 | arrrghhh | yea |
07:14.42 | bzo | yep |
07:14.57 | WisTilt2 | arrrghhh: uploaded the new one to post for the masses. 55420aa6d8f32078baf859910b8ee2c6 |
07:15.16 | arrrghhh | WisTilt2, what's the difference? just doesn't have that loop? |
07:15.31 | WisTilt2 | yeah |
07:15.40 | arrrghhh | cool |
07:16.01 | bzo | ok, I'm out, gn all |
07:16.08 | WisTilt2 | nite bzo |
07:16.08 | arrrghhh | gn |
07:16.11 | jonpry | night bzo |
07:17.30 | *** join/#htc-linux Timber (467ef4e3@gateway/web/freenode/ip.70.126.244.227) |
07:17.50 | WisTilt2 | arrrghhh: this kernel has the full new pm also, which the 3/1 did not. its also running fb in smi but i think the previous one was also, dont remember |
07:18.03 | arrrghhh | ok |
07:18.09 | WisTilt2 | and, as you already noticed, its a bit faster |
07:18.17 | arrrghhh | heh |
07:18.29 | Timber | hello |
07:18.39 | Timber | anyone using the touchpro2 or tilt2? |
07:18.49 | jonpry | yeah some of us |
07:19.49 | Timber | does the tv out function work on it for rhobuntu? |
07:19.50 | *** join/#htc-linux MacDrunk (~marper@201.165.118.194) |
07:20.05 | jonpry | don't you have to like solder something on for that? |
07:20.05 | Timber | or the linux version for the it? |
07:20.27 | Timber | not to my knowledge if using it out of the boc it supports tv out |
07:20.39 | Timber | but dont know if the tv out works when runnign linux |
07:21.02 | arrrghhh | kernel definitely doesn't have the code to support it |
07:21.10 | jonpry | actually i think it does |
07:21.29 | jonpry | its part of qcomm msm_fb |
07:21.36 | WisTilt2 | arrrghhh have you posted that pack yet? |
07:21.38 | jonpry | but i dunno how you turn it on |
07:21.43 | arrrghhh | WisTilt2, yes...? |
07:21.55 | WisTilt2 | change it to the first one if you have it |
07:22.02 | arrrghhh | del'd it |
07:22.05 | arrrghhh | crap. |
07:22.09 | arrrghhh | why? |
07:22.13 | WisTilt2 | you still have the prev one? |
07:22.23 | jonpry | WisTilt2, you think the loop is a winner? |
07:22.25 | arrrghhh | no |
07:22.49 | arrrghhh | it's...gone. |
07:22.53 | Timber | im lost here one person says it works another says it doesnt? |
07:22.58 | WisTilt2 | yeah i do. sod is probably fixed with the clock deal but wake fail still happens now and then |
07:23.08 | WisTilt2 | jonpry: timing problem for sure |
07:23.14 | Timber | just a heads up, i just bought th tilt2 today so Idont have it to try for myself |
07:23.30 | jonpry | Timber, that doesn't make any sense |
07:24.03 | Timber | im just nasking if I can use the tvout function if I put linux on my phone |
07:24.05 | WisTilt2 | arrrghhh: e4cdc304f381b38c337ae451b5089eac |
07:24.11 | *** join/#htc-linux Ondalf (~ondalf@cable-roi-fededd00-86.dhcp.inet.fi) |
07:24.13 | Timber | the tilt 2 |
07:24.24 | jonpry | believe it or not, nobody has tried |
07:24.26 | Timber | which supprts tv out out of the box |
07:24.35 | arrrghhh | Timber, in winmo |
07:24.38 | arrrghhh | with the extUSB |
07:24.47 | WisTilt2 | it doesnt work ive tried it |
07:25.07 | WisTilt2 | get no sig on tv |
07:25.10 | jonpry | isn't there some msm_fb option to turn it on? |
07:25.21 | jonpry | and maybe tlmm_config |
07:25.27 | WisTilt2 | probably, i didnt know much about this stuff back when i tried it |
07:25.34 | Timber | ok so ill need to do it in winmo then.. how well does it work with winmo I was hoping ot stream movie (files not internet streams) to my tv |
07:25.59 | WisTilt2 | Timber, works great in winmo. i also have a tilt2 btw |
07:26.06 | jonpry | lol |
07:26.33 | Timber | o, guess ill just leave my movie files to the winmo then. most of them are .avi anyway |
07:27.08 | WisTilt2 | jonpry: panel_on triggers the whole mddi init seq right? |
07:27.16 | jonpry | yeah |
07:28.06 | WisTilt2 | so with this loop banging it, that points to timing. its like unblank gets called before clocks are stable or fb has data |
07:28.50 | jonpry | could be. seems like all results are inconclusive |
07:28.56 | Timber | is it true that ubuntu has alot of lag on the tilt 2? |
07:29.14 | jonpry | lag in what |
07:29.53 | arrrghhh | it runs it as well as you can expect a 528mhz single core arm proc with 288mb of ram can handel |
07:29.59 | arrrghhh | s/handel/handle/ |
07:30.10 | jonpry | with no floating point |
07:30.23 | arrrghhh | lol |
07:30.40 | jonpry | the msm xserver is not good either |
07:32.10 | Timber | like if im watching a video in the vmcplayer or something inside linux on it is it going to be useless or will it make do? |
07:32.40 | Timber | I assume presentations and word proecessing woudlnt be an issue. |
07:33.00 | arrrghhh | how to present them? |
07:33.07 | jonpry | don't even go there. i spent a few days trying to get mplayer to work. and like it could only play if you use 1/4 of the screen |
07:33.19 | Ondalf | lemme guess - no gpu support for normal xserver |
07:33.26 | arrrghhh | yea there's a screen res limiatation it seems |
07:33.30 | arrrghhh | limitation |
07:33.33 | arrrghhh | i'm freakin tired. |
07:33.52 | jonpry | arrrghhh, we need to know if sod is fixed |
07:33.59 | jonpry | :p |
07:34.01 | arrrghhh | WisTilt2, as always, thanks for the updates |
07:34.10 | arrrghhh | jonpry, i'll update you tomorrow |
07:34.16 | WisTilt2 | of course, and thanks for, well being you |
07:34.17 | arrrghhh | i have a few guys that get a lot of them |
07:34.21 | arrrghhh | lol |
07:34.23 | arrrghhh | anytime? |
07:34.29 | arrrghhh | updated gingerbread bundle |
07:34.33 | WisTilt2 | chief tester |
07:34.44 | arrrghhh | that got some nice updates. plus new dialer. gonna get that on froyo too |
07:34.54 | arrrghhh | <3 the real freakin dialer screen |
07:35.00 | arrrghhh | i've always hated our blank dialer. |
07:35.51 | arrrghhh | man this sleeps fast |
07:35.55 | arrrghhh | let me make a phone call |
07:36.01 | arrrghhh | see if i can muck it up :P |
07:36.04 | WisTilt2 | under a second still should be |
07:36.42 | WisTilt2 | you still using the ril image also? |
07:37.08 | arrrghhh | yessir |
07:37.20 | arrrghhh | heh, back to the long sleep after a phone call. |
07:37.28 | arrrghhh | you want a dmesg after it finally sleeps? |
07:37.37 | WisTilt2 | only after you made a call? |
07:37.50 | arrrghhh | yea it was sleeping fine until i made a call |
07:38.30 | WisTilt2 | see if powermanager wake lock is still holding it |
07:39.10 | arrrghhh | hrm |
07:39.15 | arrrghhh | looks like evdev |
07:39.21 | arrrghhh | let me see if they long sleeps continue |
07:39.44 | arrrghhh | s/they/the/ |
07:40.13 | arrrghhh | indeed they are |
07:40.14 | WisTilt2 | i have all those clamped down pretty far so none of them should have anything like 30s |
07:40.19 | arrrghhh | i'll go thru a couple |
07:42.40 | arrrghhh | yup, more powermanager service |
07:42.41 | arrrghhh | i2x |
07:43.08 | WisTilt2 | if its not too much trouble can you paste that |
07:43.48 | arrrghhh | no trouble 1 sec |
07:44.07 | arrrghhh | http://pastebin.com/EQRpMn6m |
07:45.03 | Timber | thank oyu for the assistnacce |
07:45.07 | Timber | imheading ot bed now! |
07:45.12 | Timber | night! |
07:45.16 | WisTilt2 | ok ill go thru this you go to bed. hopefully we can get some input by tomorrow on this kernel |
07:45.40 | Timber | i cant do any testing |
07:46.52 | arrrghhh | Timber, you're so helpful |
07:46.56 | arrrghhh | aw |
07:47.10 | arrrghhh | WisTilt2, sounds good. i'll let you know what the feedback is, as usual ;) |
07:47.32 | WisTilt2 | ok, ill be around sometime tomorrow |
07:47.54 | odz | running smooth with the latest gb build so far |
07:48.16 | WisTilt2 | wow arrrghhh, you had a kind of system_server issue on that log |
07:48.24 | arrrghhh | orly? |
07:48.27 | arrrghhh | didn't notice it |
07:48.38 | jonpry | no vysnc messages |
07:48.41 | arrrghhh | i mean i was downloading a lot of crap |
07:48.47 | arrrghhh | pegging proc doing that |
07:48.49 | WisTilt2 | yeah, starting at line 946, thats what we saw the started the whole slow down |
07:49.03 | *** join/#htc-linux dekar_ (~dekar@drms-590c59ab.pool.mediaWays.net) |
07:49.27 | WisTilt2 | oh thats while you were dling. still shouldnt be getting binder errors like that |
07:49.34 | arrrghhh | hrm |
07:49.52 | WisTilt2 | jonpry you looking at his log? |
07:49.55 | jonpry | yeah |
07:50.13 | WisTilt2 | lines 890-896 whats that all about |
07:50.27 | WisTilt2 | first time ive seen that in any logs like that |
07:51.06 | jonpry | what is that |
07:51.37 | WisTilt2 | the suspend is normal but not like that. |
07:51.56 | WisTilt2 | line 872 is what it should look like before and after |
07:52.21 | WisTilt2 | it was after he put it so sleep but it didnt make it all the way |
07:53.03 | jonpry | yeah thats the stupid timer thing no? |
07:53.47 | WisTilt2 | which one? |
07:54.28 | jonpry | in arch_idle is does these very strange checks to see if sleep can happen |
07:54.45 | jonpry | and this: sleep_time = msm_timer_enter_idle(); |
07:54.59 | jonpry | like how would msm_timer know anything about sleep |
07:55.28 | arrrghhh | alright, sleep time for reals. talk to you guys later! |
07:57.49 | jonpry | i gotta sleep as well. good night all |
07:58.00 | WisTilt2 | nite, catch ya tomorrow |
08:00.34 | *** join/#htc-linux kernelzilla (~patrick@63-225-214-129.phnx.qwest.net) |
08:01.09 | *** join/#htc-linux [acl] (~abel@cpe-69-203-141-229.si.res.rr.com) |
08:01.48 | *** join/#htc-linux camro (~camro@89-104-29-101.customer.bnet.at) |
08:05.12 | *** join/#htc-linux Rob2222 (~Miranda@p5DE4BDE2.dip.t-dialin.net) |
08:05.41 | *** join/#htc-linux Rob2222 (~Miranda@p5DE4BDE2.dip.t-dialin.net) |
08:16.09 | *** join/#htc-linux Ceesheim (~ceesheim@p508DEB22.dip.t-dialin.net) |
08:23.30 | *** join/#htc-linux Drakh_ (Drakh@190.42.66.148) |
08:26.26 | *** join/#htc-linux Drakh (Drakh@190.42.26.179) |
08:41.12 | *** join/#htc-linux Drakh_ (Drakh@190.42.26.179) |
08:52.00 | *** join/#htc-linux _twitch (~burning_a@64.112.96.58.static.exetel.com.au) |
08:56.21 | *** join/#htc-linux Drakh_ (Drakh@190.42.26.179) |
08:56.33 | *** join/#htc-linux Ondalf_ (~ondalf@cable-roi-fededd00-86.dhcp.inet.fi) |
09:06.15 | *** join/#htc-linux rob_w (~bob@ppp-188-174-85-12.dynamic.mnet-online.de) |
09:07.47 | *** join/#htc-linux Erikson (~Erik@p54B77A74.dip.t-dialin.net) |
09:07.59 | *** join/#htc-linux Ondalf_ (~ondalf@cable-roi-fededd00-86.dhcp.inet.fi) |
09:13.55 | *** join/#htc-linux kernelzilla (~patrick@63-225-214-129.phnx.qwest.net) |
09:23.33 | *** join/#htc-linux Drakh_ (Drakh@190.42.26.179) |
09:26.45 | *** join/#htc-linux Drakh (Drakh@190.42.66.148) |
09:39.13 | *** join/#htc-linux relaxed (~relaxed@unaffiliated/relaxed) |
09:53.33 | *** join/#htc-linux Drakh_ (Drakh@190.42.27.142) |
09:59.28 | *** join/#htc-linux Drakh_ (Drakh@190.42.66.148) |
10:06.14 | *** join/#htc-linux Andreyxxl[HD2EU] (~Andreyxxl@89.32.146.153) |
10:13.42 | *** join/#htc-linux kvaster (~kvaster@vpn-e0.bas-net.by) |
10:18.13 | *** join/#htc-linux JustLiving (~Quilck@90.183.82.215) |
10:21.25 | *** join/#htc-linux kernelzilla (~patrick@174-26-107-146.phnx.qwest.net) |
10:26.30 | *** join/#htc-linux Drakh_ (Drakh@190.42.26.88) |
10:30.20 | *** join/#htc-linux Andreyxxl[HD2EU] (Andreyxxl@89.32.146.153) |
10:31.03 | *** join/#htc-linux MassStash (~MassStash@c-67-175-41-173.hsd1.il.comcast.net) |
10:42.16 | *** join/#htc-linux Kasjopaja (~Tina@p578FFEF3.dip.t-dialin.net) |
10:50.46 | *** join/#htc-linux Drakh_ (Drakh@190.42.26.88) |
11:01.15 | *** join/#htc-linux Drakh_ (Drakh@190.42.66.148) |
11:26.55 | *** join/#htc-linux JustLiving (~Quilck@90.183.82.215) |
11:36.09 | *** join/#htc-linux netson-ubuntu_ (~netson-ub@110.138.157.233) |
11:43.16 | *** join/#htc-linux skodde (~skodde@unaffiliated/skodde) |
11:47.08 | *** join/#htc-linux MethoS- (~clemens@134.102.106.250) |
11:55.01 | *** join/#htc-linux emwe (~emwe@cable-86-56-10-158.cust.telecolumbus.net) |
12:07.42 | *** join/#htc-linux mitsutaka (~mitsutaka@p1227-ipbf3907marunouchi.tokyo.ocn.ne.jp) |
12:13.18 | *** join/#htc-linux ftoz (~root@ip-85-160-132-32.eurotel.cz) |
12:23.45 | *** join/#htc-linux tyween (~tyween@pool-71-163-117-57.washdc.fios.verizon.net) |
12:29.53 | *** join/#htc-linux Jhinta (~Jhinta@s529c116e.adsl.wanadoo.nl) |
12:43.13 | *** join/#htc-linux Jhinta (~Jhinta@s529c116e.adsl.wanadoo.nl) |
12:46.53 | *** join/#htc-linux LordDeath (~Lord|Lapt@cable-81-173-166-52.netcologne.de) |
12:47.47 | *** join/#htc-linux LordDeath (~Lord|Lapt@cable-81-173-166-52.netcologne.de) |
13:00.39 | *** join/#htc-linux Bry8Star (debian-tor@gateway/tor-sasl/bry8star) |
13:15.10 | *** join/#htc-linux phh (~quassel@tatiana.rezel.enst.fr) |
13:24.40 | *** join/#htc-linux gauner1986 (~Miranda@p508C7A17.dip.t-dialin.net) |
13:25.49 | *** join/#htc-linux daevoq (~dae@host239-95-dynamic.52-79-r.retail.telecomitalia.it) |
13:37.27 | *** join/#htc-linux mastermerlin (~Adium@p4FEE4EB5.dip.t-dialin.net) |
13:54.07 | *** join/#htc-linux gauner1986 (~Miranda@p5B3825D1.dip.t-dialin.net) |
14:18.19 | *** join/#htc-linux rpierce99 (~rpierce99@71-82-139-28.dhcp.roch.mn.charter.com) |
14:33.53 | *** join/#htc-linux JustLiving (~Quilck@90.183.82.215) |
14:41.17 | *** join/#htc-linux kiozen (~kiozen@rgnb-4d0d6580.pool.mediaWays.net) |
14:57.16 | *** join/#htc-linux Rajko (~Rajko@2001:470:1f0b:728:817:ef9b:c061:eb77) |
15:07.33 | *** join/#htc-linux akirax (~androirc@80.31.199.60) |
15:11.18 | *** join/#htc-linux Noellenchris (~kvirc@pool-173-61-114-241.cmdnnj.fios.verizon.net) |
15:15.14 | *** join/#htc-linux Daevoq (~IceChat7@host243-0-dynamic.18-79-r.retail.telecomitalia.it) |
15:37.12 | *** join/#htc-linux Markinus (~Miranda@212.255.18.217) |
15:48.49 | *** join/#htc-linux |Jeroen| (~jeroen@d5152B25B.access.telenet.be) |
15:49.08 | *** join/#htc-linux GlemSom (~glemsom@0x5da34bca.cpe.ge-1-1-0-1105.sdnqu1.customer.tele.dk) |
15:55.11 | *** join/#htc-linux kvaster (~kvaster@vpn-e0.bas-net.by) |
16:02.58 | *** join/#htc-linux rpierce99 (~rpierce99@71-82-139-28.dhcp.roch.mn.charter.com) |
16:27.15 | *** join/#htc-linux santuzzo (523b2f8f@gateway/web/freenode/ip.82.59.47.143) |
16:31.58 | *** join/#htc-linux mastermerlin (~Adium@p4FEE4EB5.dip.t-dialin.net) |
16:34.53 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
16:36.06 | *** join/#htc-linux FlawlesStyle (~LOL@208.43.77.85-static.reverse.softlayer.com) |
16:36.45 | *** join/#htc-linux Andreyxxl[HD2EU] (~Andreyxxl@89.32.146.153) |
16:43.34 | *** join/#htc-linux joshkoss (~android@c-76-127-147-21.hsd1.ct.comcast.net) |
16:43.44 | *** join/#htc-linux elvis (b4b0395d@gateway/web/freenode/ip.180.176.57.93) |
16:53.28 | *** join/#htc-linux arrrghhh (~arrrghhh@c-71-237-40-111.hsd1.co.comcast.net) |
16:57.42 | *** join/#htc-linux leviathan (~quassel@2001:470:26:484:6ef0:49ff:fee6:8dca) |
17:00.53 | *** join/#htc-linux mastermerlin (~Adium@p4FEE4EB5.dip.t-dialin.net) |
17:02.15 | *** join/#htc-linux mahrz (~harder@147.197.165.178) |
17:11.19 | *** join/#htc-linux AndIrc_ (~android@c-76-127-147-21.hsd1.ct.comcast.net) |
17:11.33 | *** part/#htc-linux joshkoss (~android@c-76-127-147-21.hsd1.ct.comcast.net) |
17:12.46 | *** join/#htc-linux mitsutaka (~mitsutaka@p1227-ipbf3907marunouchi.tokyo.ocn.ne.jp) |
17:34.04 | *** join/#htc-linux arrrghhh_ (~arrrghhh@c-71-237-40-111.hsd1.co.comcast.net) |
17:48.16 | *** join/#htc-linux BHSPitLappy (~BHSPitLap@unaffiliated/bhspitmonkey) |
17:49.59 | *** join/#htc-linux Segnale007 (~Segnale00@151.62.134.201) |
17:52.46 | *** join/#htc-linux Markinus (~Miranda@212.255.17.191) |
17:53.06 | *** join/#htc-linux Cotulla (~opera@nat004-252-205-109.tvoe.tv) |
17:53.40 | Cotulla | hey |
17:54.17 | arrrghhh | what's up Cotulla |
17:56.01 | jonpry | fishingmedic says he got crashing. i wonder if that was on 300mhz or the new ramping kernel |
17:56.07 | jonpry | er sod |
17:56.26 | arrrghhh | with his airave? |
17:56.32 | arrrghhh | these might be different SoD's jonpry |
17:56.38 | gauner1986 | hey Cotulla |
17:57.14 | Cotulla | something strange |
17:57.20 | Cotulla | hey Gau |
17:58.59 | Cotulla | arrrghhh, u? |
17:59.23 | arrrghhh | Cotulla, not much. trying to see why things break ;) |
18:04.50 | *** join/#htc-linux mastermerlin (~Adium@p4FEE4EB5.dip.t-dialin.net) |
18:05.47 | *** join/#htc-linux FlawlesStyle (~LOL@unaffiliated/flawlesstyle) |
18:26.42 | *** join/#htc-linux arrrghhh (~arrrghhh@c-71-237-40-111.hsd1.co.comcast.net) |
18:35.27 | *** join/#htc-linux Alex[sp3dev] (~alexander@86.110.163.19) |
18:35.38 | Alex[sp3dev] | jonpry: hi |
18:38.13 | *** join/#htc-linux DuperMan (~Duper@46-116-16-100.bb.netvision.net.il) |
18:38.28 | arrrghhh | Alex[sp3dev], jonpry went into hiding. |
18:38.47 | arrrghhh | jonpry, just wanted to mention, one of the users that can reproduce the SoD's due to service still have that issue. |
18:39.01 | arrrghhh | with WisTilt2's newest kernel. |
18:39.01 | Cotulla | tell us his new nick :P |
18:39.18 | Alex[sp3dev] | Cotulla: it's not YOUR kind of hiding i guess |
18:39.29 | arrrghhh | lol |
18:39.32 | Cotulla | we not talking about me now |
18:39.39 | arrrghhh | hiding in plain sight you see |
18:40.04 | Alex[sp3dev] | Cotulla: got collapse working on 32 yet? |
18:40.22 | Cotulla | no |
18:40.36 | Cotulla | I was busy with otehr things |
18:40.55 | Alex[sp3dev] | yeah.. exams are coming. but until then, i'll do some hacking on 35 |
18:41.54 | Cotulla | nahh |
18:42.00 | Alex[sp3dev] | i've disables SMI mpu (to allow writing to 0x0 to set resume vectors) - modes 0 and 1 seem to work, but drain power. if i insert dex call (well, a raw write to command and counter) before collapse(), it dies. if i insert after, seems to work but wake up often. but that may be unrelated |
18:42.26 | Cotulla | there must be wake up source somewhere |
18:42.26 | Alex[sp3dev] | Cotulla: ok, i know exams are not even remotely a reason to worry for you |
18:42.39 | arrrghhh | lol |
18:42.40 | Alex[sp3dev] | Cotulla: playing captain obvious? sorry |
18:43.05 | Cotulla | arrrghhh, is it obvious for u? |
18:43.36 | arrrghhh | that you're not concerned about exams/ |
18:43.43 | arrrghhh | i don't even know if you're in school so, no. |
18:43.46 | Cotulla | I mean about wake up source |
18:43.49 | Cotulla | wtf |
18:43.56 | arrrghhh | lol |
18:44.08 | arrrghhh | no idea on the wake up source |
18:44.16 | arrrghhh | wistilt2 went romping thru PM |
18:44.27 | Alex[sp3dev] | wistilt2 is on 27 though |
18:44.29 | Cotulla | it's just looks like Alex have exams every month |
18:44.31 | arrrghhh | but i he hasn't committed it to .27 tree yet |
18:44.40 | Alex[sp3dev] | Cotulla: every 2 month to be more precise |
18:44.41 | arrrghhh | Alex[sp3dev], i would think his PM changes could help .35? |
18:45.22 | Cotulla | yeah |
18:45.30 | Cotulla | it's why u talking so much about them |
18:45.39 | Alex[sp3dev] | yeah, they suck |
18:46.26 | Cotulla | so how other it wake ups |
18:46.27 | Cotulla | ? |
18:46.40 | Alex[sp3dev] | Cotulla: i think, once in 2 seconds or so |
18:48.06 | arrrghhh | emwe was having the problem too AFAIK |
18:48.17 | Cotulla | and is it big problem? |
18:48.20 | Alex[sp3dev] | yes, but i didn't see his pm.c code |
18:48.22 | Cotulla | it can go to sleep again? |
18:48.42 | Alex[sp3dev] | Cotulla: well, i'll reboot and make a dmesg. it may be failing to enter, actually |
18:52.19 | *** join/#htc-linux AndIrc__ (~android@c-76-127-147-21.hsd1.ct.comcast.net) |
18:52.39 | *** part/#htc-linux joshkoss (~android@c-76-127-147-21.hsd1.ct.comcast.net) |
18:53.05 | Cotulla | yes looks like making PM working is nice idea |
18:53.09 | Cotulla | now it have big drain\ |
18:53.14 | Cotulla | about 230 with screen off |
18:55.26 | *** join/#htc-linux balans1 (~Gebruiker@82-170-217-205.ip.telfort.nl) |
19:14.07 | fakker | privet |
19:15.05 | Cotulla | poka |
19:16.23 | *** join/#htc-linux Andreyxxl[HD2EU] (Andreyxxl@89.32.146.153) |
19:22.33 | *** join/#htc-linux pepermint (~pepermint@host68-88-dynamic.8-87-r.retail.telecomitalia.it) |
19:23.25 | *** join/#htc-linux GNUtoo|laptop (~gnutoo@host226-134-dynamic.25-79-r.retail.telecomitalia.it) |
19:24.41 | *** join/#htc-linux tre9 (d0360433@gateway/web/freenode/ip.208.54.4.51) |
19:26.54 | *** join/#htc-linux crawling (~crawling@a94-132-172-247.cpe.netcabo.pt) |
19:49.56 | *** join/#htc-linux kvaster (~kvaster@vpn-e0.bas-net.by) |
19:55.45 | *** join/#htc-linux arrrghhh_ (~arrrghhh@c-71-237-40-111.hsd1.co.comcast.net) |
19:59.10 | *** join/#htc-linux Andreyxxl[HD2EU] (Andreyxxl@89.32.146.153) |
20:03.30 | fakker | poka |
20:14.19 | *** join/#htc-linux LordDeath (~Lord|Lapt@cable-81-173-166-52.netcologne.de) |
20:22.27 | *** join/#htc-linux skodde (~skodde@unaffiliated/skodde) |
20:41.06 | *** join/#htc-linux Drakh_ (Drakh@190.42.30.123) |
20:48.06 | *** join/#htc-linux crawling (~crawling@a94-132-172-247.cpe.netcabo.pt) |
20:49.25 | *** join/#htc-linux dcordes (~dcordes@ip-80-226-0-1.vodafone-net.de) |
20:49.25 | *** join/#htc-linux dcordes (~dcordes@unaffiliated/dcordes) |
20:50.35 | *** join/#htc-linux skodde (~skodde@unaffiliated/skodde) |
21:05.36 | jonpry | phh: you reread the rtc patch? |
21:05.46 | phh | not yet |
21:24.32 | *** join/#htc-linux crawling (crawling@a94-132-172-247.cpe.netcabo.pt) |
21:33.08 | *** join/#htc-linux gauner1986 (~Adium@p5B3825D1.dip.t-dialin.net) |
21:43.33 | jonpry | i'm not hiding :p |
21:44.06 | Cotulla | but u reading logs |
21:44.10 | Cotulla | :P |
21:44.19 | jonpry | yeah |
21:45.18 | arrrghhh | heh |
21:45.35 | jonpry | Cotulla, how is your .32? |
21:45.43 | Cotulla | I tell u truth: it was arrrghhh words |
21:45.44 | Cotulla | :P |
21:45.47 | gauner1986 | it's boring.. where are the leo updates? |
21:45.48 | gauner1986 | :P |
21:45.56 | arrrghhh | leo is done |
21:46.02 | arrrghhh | you guys have everything :P |
21:46.10 | gauner1986 | yeah |
21:46.13 | gauner1986 | thats the problem |
21:46.15 | Cotulla | much more than everything |
21:46.16 | gauner1986 | :D |
21:46.18 | arrrghhh | let us get some lovin |
21:46.25 | gauner1986 | no |
21:46.28 | arrrghhh | fu |
21:46.29 | gauner1986 | missing 720p |
21:46.31 | gauner1986 | :P |
21:46.40 | arrrghhh | fix it then :P |
21:46.55 | gauner1986 | teach me some assemblerskills |
21:46.59 | arrrghhh | heh |
21:47.01 | arrrghhh | you first |
21:47.03 | gauner1986 | maybe i can fix agps then also |
21:47.08 | arrrghhh | agps? |
21:47.12 | arrrghhh | doesn't work? |
21:47.16 | gauner1986 | no |
21:47.19 | arrrghhh | oh right quickgps |
21:47.24 | arrrghhh | i was thinking cell tower netloc |
21:47.25 | gauner1986 | or xtra |
21:47.27 | arrrghhh | yea |
21:47.28 | gauner1986 | like quickgps does |
21:47.32 | arrrghhh | indeed |
21:47.37 | arrrghhh | i'd like that to work on RHOD as well |
21:47.39 | arrrghhh | :D |
21:47.56 | gauner1986 | needs some additional rpc calls |
21:48.02 | gauner1986 | that probably are in quickgps dll or exe |
21:48.24 | arrrghhh | yea |
21:48.30 | *** join/#htc-linux Total_Meltdown_ (~Total_Mel@cpe-72-224-201-130.maine.res.rr.com) |
21:48.54 | jonpry | i think we have some of it |
21:49.27 | jonpry | afaict its part of the rpc that turns on gps. you have to tell modem proc some network info, and then it downloads the crap all by itself |
21:50.14 | arrrghhh | entropy was trying to fix gps |
21:50.16 | Cotulla | u too like word crap :P |
21:50.17 | arrrghhh | that 0,0 bug |
21:50.25 | *** join/#htc-linux aZuZu[Doma] (~aZuZuHome@93-141-105-65.adsl.net.t-com.hr) |
21:50.29 | aZuZu[Doma] | hi to all! |
21:50.48 | aZuZu[Doma] | question |
21:51.03 | aZuZu[Doma] | i'm porting honeycomb to x10mini.. |
21:51.10 | gauner1986 | hm |
21:51.17 | dcordes | [[x10mini]] |
21:51.17 | unilinky | http://htc-linux.org/wiki/index.php?title=x10mini |
21:51.30 | *** join/#htc-linux RaiderX (~aaron@91.109.22.100) |
21:51.31 | Cotulla | dcordes appear from air |
21:51.31 | dcordes | aZuZu[Doma], no such device |
21:51.34 | Cotulla | or maybe it's bot |
21:51.39 | Cotulla | new channel bot? :P |
21:51.41 | gauner1986 | did any of you guys try to use the loc_api driver? |
21:51.45 | RaiderX | hey |
21:52.04 | Cotulla | which parse word "porting to" :D |
21:52.12 | gauner1986 | lol |
21:52.17 | dcordes | [[:D]] |
21:52.17 | unilinky | http://htc-linux.org/wiki/index.php?title=:D |
21:52.22 | arrrghhh | lol |
21:52.24 | Cotulla | hm |
21:52.24 | aZuZu[Doma] | hmh |
21:52.29 | Cotulla | looks like bot |
21:52.33 | Cotulla | :D |
21:52.46 | dcordes | what's up ? |
21:52.51 | Cotulla | u not bot? |
21:52.58 | jonpry | [[a][b]] |
21:52.58 | unilinky | http://htc-linux.org/wiki/index.php?title=a%5D%5Bb |
21:53.10 | Cotulla | [[[[[a]]]]]] |
21:53.10 | unilinky | http://htc-linux.org/wiki/index.php?title=%5B%5B%5Ba |
21:53.12 | *** join/#htc-linux imilka (~kvirc@93-80-229-76.broadband.corbina.ru) |
21:53.12 | dcordes | Cotulla, no I'm the true dcordes |
21:53.15 | Cotulla | okay |
21:53.17 | aZuZu[Doma] | so can i ask question or not? |
21:53.34 | gauner1986 | you're trying to trick unilinky to recursion? |
21:53.35 | gauner1986 | :D |
21:53.36 | dcordes | aZuZu[Doma], if you tell us what device you're talking about. I don't know x10mini |
21:53.50 | aZuZu[Doma] | sony-ericsson x10 mini |
21:53.54 | jonpry | its one of those home automation devices |
21:54.02 | Cotulla | 7627 I guess |
21:54.13 | arrrghhh | home automation lol |
21:54.25 | *** join/#htc-linux programmer8922 (~Evan@67.219.164.162) |
21:54.26 | dcordes | aZuZu[Doma], does it have a codename? what SoC does it have? where are the kernel sources you are using on it? |
21:54.40 | aZuZu[Doma] | but device model it self is not important |
21:54.57 | aZuZu[Doma] | it is msm7227 / adreno200 |
21:55.08 | jonpry | oO vfp |
21:55.34 | jonpry | doesn't it already run android? |
21:55.45 | dcordes | aZuZu[Doma], kernel ? |
21:57.45 | aZuZu[Doma] | 2.6.29 i think.. |
21:58.28 | jonpry | but honeycomb will not run without es 2.0 |
22:00.35 | dcordes | aZuZu[Doma], http://developer.sonyericsson.com/cws/download/1/910/966/1294910745/x10_x10mini_X10minipro_x8_eclair_2.1.A.0.435.tar.gz =# |
22:00.43 | dcordes | aZuZu[Doma], you think this might be it ? |
22:01.14 | aZuZu[Doma] | http://dl.dropbox.com/u/10158456/Other/device.png << this is screen shot with msm7k gralloc |
22:02.14 | aZuZu[Doma] | http://dl.dropbox.com/u/10158456/Other/ScreenShot.png this i w/o pure libgles_android.so.. and sdk gralloc. |
22:02.45 | jonpry | the second one just has wrong stride |
22:03.09 | aZuZu[Doma] | stride? |
22:03.16 | jonpry | pixels per line |
22:03.21 | dcordes | aZuZu[Doma], have you ever booted an adreno kernel you compiled ? |
22:04.03 | aZuZu[Doma] | i didn't compile kernel |
22:04.21 | aZuZu[Doma] | it is sdk port |
22:04.28 | arrrghhh | sdk ports seem pointless |
22:04.34 | arrrghhh | wait for AOSP code to drop |
22:04.42 | aZuZu[Doma] | CM? |
22:04.50 | arrrghhh | AOSP |
22:05.12 | aZuZu[Doma] | i was right... |
22:05.13 | aZuZu[Doma] | android = new WM.. |
22:05.36 | arrrghhh | lolwut |
22:06.03 | jonpry | xdandroid is sdk port |
22:06.12 | arrrghhh | we use AOSP |
22:06.16 | Cotulla | it won't ever replace is WM |
22:06.16 | *** join/#htc-linux Ceesheim (~ceesheim@p508DEB22.dip.t-dialin.net) |
22:06.18 | arrrghhh | we don't port from SDK |
22:06.19 | aZuZu[Doma] | wait till someone does all .. you just copy.. |
22:06.37 | arrrghhh | aZuZu[Doma], google is a pretty good maintainer |
22:06.44 | arrrghhh | i don't mind them maintaining upstream |
22:06.50 | jonpry | there's no honeycomb port in AOSP? |
22:06.57 | arrrghhh | jonpry, code hasn't been released |
22:07.09 | arrrghhh | only SDK |
22:07.33 | *** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring) |
22:07.38 | jonpry | well, i think his problem is that screen res is set wrong |
22:07.46 | jonpry | maybe it should be 256 instead of 240 |
22:07.54 | aZuZu[Doma] | ? |
22:08.06 | aZuZu[Doma] | it is qvga device lol |
22:08.20 | jonpry | yeah, but the math is faster if you waste a little framebuffer |
22:08.34 | jonpry | its obviously wrong |
22:08.56 | aZuZu[Doma] | tho is did test to max 230.. |
22:09.20 | aZuZu[Doma] | mayby is sould test upper.. values |
22:09.30 | jonpry | it 240x320 yeah? |
22:09.38 | aZuZu[Doma] | yap |
22:09.49 | jonpry | but you only tried 230? |
22:10.05 | aZuZu[Doma] | yes |
22:10.45 | jonpry | ok. i don't get it, but mess with the width and it will get unwarped |
22:12.13 | aZuZu[Doma] | that is the point with sdk gralloc and libgles w/o anything else.. in hw and egl... |
22:12.34 | AstainHellbring | whats new jonpry |
22:12.49 | aZuZu[Doma] | libcat says device reconfigured new size 156x1246 or something.. |
22:13.06 | jonpry | AstainHellbring, battery meter |
22:13.21 | aZuZu[Doma] | libcat = logcat |
22:13.41 | AstainHellbring | tp2? |
22:14.11 | jonpry | yeah |
22:15.48 | *** join/#htc-linux curious_ (8bb3cfc7@gateway/web/freenode/ip.139.179.207.199) |
22:16.04 | *** part/#htc-linux curious_ (8bb3cfc7@gateway/web/freenode/ip.139.179.207.199) |
22:19.20 | *** join/#htc-linux Pr0z (Pr0z@bzq-84-109-110-248.red.bezeqint.net) |
22:22.24 | dcordes | aZuZu[Doma], you mean like user space port ? |
22:22.39 | dcordes | aZuZu[Doma], I always thought android is heavily depending on kernel version |
22:22.46 | *** join/#htc-linux manekineko (~manekinek@newyork.homeserver.com) |
22:23.24 | arrrghhh | dcordes, we run gb on .27 |
22:23.27 | arrrghhh | lol |
22:23.35 | dcordes | ok |
22:23.46 | dcordes | aZuZu[Doma], the link above does in fact include the kernel sources |
22:23.50 | arrrghhh | not sure if it's a good idea or not. |
22:23.53 | arrrghhh | but it does work |
22:24.08 | stinebd | uh |
22:24.23 | stinebd | not sure if running gingerbread on a bunch of old winmo devices is a very great idea either |
22:24.29 | arrrghhh | lmao |
22:24.31 | arrrghhh | that is a good point |
22:24.32 | arrrghhh | but we do it |
22:25.48 | dcordes | LOL it includes grub source code ? |
22:26.23 | dcordes | aZuZu[Doma], adreno device doesn't come with grub installed, does it ? |
22:30.17 | dcordes | [[adreno]] |
22:30.17 | unilinky | http://htc-linux.org/wiki/index.php?title=adreno |
22:30.21 | *** join/#htc-linux toastcfh (~toastcfh@29-37.200-68.tampabay.res.rr.com) |
22:32.28 | aZuZu[Doma] | wonders how to compile anything without sys folder in source.. |
22:34.25 | *** join/#htc-linux mitsutaka_ (~mitsutaka@p1227-ipbf3907marunouchi.tokyo.ocn.ne.jp) |
22:38.34 | NetRipper | dcordes, http://irclog.netripper.com/htc-linux/dcordes |
22:38.35 | NetRipper | ;) |
22:39.13 | *** join/#htc-linux mitsutaka_ (~mitsutaka@p1227-ipbf3907marunouchi.tokyo.ocn.ne.jp) |
22:41.39 | *** join/#htc-linux programmer8922 (~Evan@67.219.164.162) |
22:42.47 | Cotulla | olololo |
22:42.51 | Cotulla | NetRipper |
22:42.55 | Cotulla | <PROTECTED> |
22:43.10 | NetRipper | :p |
22:43.21 | Cotulla | :P |
22:43.43 | Cotulla | now we can find ur first words |
22:43.45 | Cotulla | ?^^ |
22:43.55 | NetRipper | my words? |
22:44.10 | dcordes | NetRipper, awesome! you're da man |
22:44.12 | Cotulla | yeah |
22:44.15 | Cotulla | or somebody else |
22:44.18 | NetRipper | yes |
22:44.25 | Cotulla | but filter works only for today |
22:44.29 | Cotulla | it reset day month |
22:44.30 | NetRipper | no |
22:44.31 | Cotulla | year |
22:44.41 | NetRipper | the search is for all period |
22:44.43 | NetRipper | not only today |
22:44.47 | Cotulla | dunno |
22:44.50 | NetRipper | it just displays today if you click it |
22:44.52 | Cotulla | If I select other |
22:44.55 | Cotulla | day |
22:44.59 | NetRipper | yes doh, then it shows just thatd ay |
22:45.01 | Cotulla | and type something again |
22:45.03 | NetRipper | it doesn't search in that day :) |
22:45.11 | Cotulla | it go to 13 march again |
22:45.13 | NetRipper | yes |
22:45.15 | NetRipper | correct |
22:45.17 | *** join/#htc-linux Alex[sp3dev] (~alexander@86.110.163.19) |
22:45.35 | Alex[sp3dev] | stinebd: ping |
22:45.42 | stinebd | Alex[sp3dev]: pong |
22:45.45 | NetRipper | Cotulla, the search is cross-date |
22:45.57 | Alex[sp3dev] | stinebd: building gingerbread fails for me |
22:46.03 | Alex[sp3dev] | packages/apps/Calculator/src/com/android/calculator2/HistoryAdapter.java:65: cannot access org.xmlpull.v1.XmlPullParser |
22:46.04 | Alex[sp3dev] | bad class file: org/xmlpull/v1/XmlPullParser.class(org/xmlpull/v1:XmlPullParser.class) |
22:46.04 | Alex[sp3dev] | unable to access file: corrupted zip file |
22:46.05 | NetRipper | i.e. the date selection does not matter... you always get the 'latest' |
22:46.20 | *** join/#htc-linux mitsutaka_ (~mitsutaka@p1227-ipbf3907marunouchi.tokyo.ocn.ne.jp) |
22:46.40 | stinebd | Alex[sp3dev]: that looks like it's likely something on your end |
22:46.52 | Alex[sp3dev] | NetRipper: according to search results, [acl] and me are the top users of the words fuck and poop |
22:47.04 | arrrghhh | lol |
22:47.11 | dcordes | :> |
22:47.14 | NetRipper | lol |
22:47.31 | Cotulla | hm search search everthing? |
22:48.18 | jonpry | lucky bastards |
22:48.40 | NetRipper | yes |
22:49.09 | Cotulla | but slow :P |
22:49.20 | NetRipper | only if you search something that is uncommon |
22:49.35 | NetRipper | but if you search something common, like dcordes, it is fast |
22:49.36 | Alex[sp3dev] | google indexes that all anyway |
22:49.37 | Cotulla | u hardcode list of words there |
22:49.39 | Cotulla | right? |
22:49.39 | Cotulla | :D |
22:49.42 | NetRipper | no |
22:50.10 | dcordes | NetRipper, I noticed the general log viewing got much faster |
22:50.42 | *** part/#htc-linux Total_Meltdown_ (~Total_Mel@cpe-72-224-201-130.maine.res.rr.com) |
22:50.51 | jonpry | irc log search is awesome. absolutely none of it makes sense |
22:51.10 | jonpry | Unholy yea but i wouldent care much snince im alredy dranning my battry of at 700mhz |
22:51.24 | *** join/#htc-linux Drakh_ (Drakh@190.42.30.123) |
22:52.07 | NetRipper | dcordes, also works for looking up urls like you wanted to: http://irclog.netripper.com/htc-linux/http%253A%252F%252F |
22:52.23 | dcordes | jonpry, now we can spy ourselves better |
22:52.42 | dcordes | awesome |
22:52.50 | NetRipper | dcordes, yes it got a lot better since i reworked the table structure |
22:53.05 | *** join/#htc-linux MassStash (~MassStash@c-67-175-41-173.hsd1.il.comcast.net) |
22:54.04 | dcordes | I prepared meatsy foodstuffs and now I smell like garlic 10km against the wind direction |
22:54.18 | NetRipper | when did you not smell? |
22:54.20 | NetRipper | ;) |
22:54.28 | dcordes | :( |
22:54.32 | dcordes | http://irclog.netripper.com/htc-linux/garlic |
22:55.32 | NetRipper | :) |
22:55.50 | dcordes | I'm happy to announce marajin prepared salmon with garlic end of July 200 and eight |
22:58.18 | stinebd | Alex[sp3dev]: dont know man, i'd say try syncing again and make a make clean |
22:59.15 | Alex[sp3dev] | stinebd: i'll try to disable building that app and see if it breaks again |
22:59.40 | Alex[sp3dev] | but android build system is such crap |
23:00.00 | Alex[sp3dev] | they could've used cmake and it would be faster.. and easier to do partial builds |
23:00.29 | dcordes | Alex[sp3dev], port it to openembedded |
23:00.40 | Alex[sp3dev] | dcordes: why spoil oe with android? |
23:01.39 | dcordes | http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-April/019132.html |
23:04.07 | *** join/#htc-linux Drakh_ (Drakh@190.42.30.123) |
23:05.07 | Alex[sp3dev] | stinebd: which machine do you have for building android btw? |
23:05.23 | *** join/#htc-linux Ceesheim (~ceesheim@p508DEB22.dip.t-dialin.net) |
23:05.28 | stinebd | Alex[sp3dev]: quad-core phenom running gentoo |
23:05.53 | Alex[sp3dev] | yeah. gentoo guys need top hardware to survive |
23:10.41 | Cotulla | ^^ |
23:14.49 | *** join/#htc-linux Drakh_ (Drakh@190.42.30.123) |
23:21.38 | *** part/#htc-linux aZuZu[Doma] (~aZuZuHome@93-141-105-65.adsl.net.t-com.hr) |
23:22.53 | MassStash | who should i ask just a random Q bout why the HD2 doesn't put CWM on a partition instead of running off the sdcard? |
23:23.18 | Cotulla | u have TP2? |
23:23.31 | MassStash | yea, but work on my friends HD2 from time to time |
23:23.51 | MassStash | we were messing with it yesterday n i learned that u need the sdcard still in to run CWM roms right? |
23:24.08 | MassStash | that's sorta a degression from your MAGLDR |
23:24.10 | Cotulla | dunno how u learned this... ppl flashing CWN to partition as well |
23:24.20 | MassStash | oohhhh k |
23:24.27 | MassStash | perhaps i stopped looking into too soon |
23:24.30 | Cotulla | but some start to think that it can be on SD also |
23:25.15 | MassStash | do you see any need for CWM then really? |
23:25.50 | MassStash | its sorta nerd prefference at that point? |
23:25.52 | MassStash | haha |
23:26.10 | Cotulla | different people have different opinions :) |
23:26.16 | MassStash | riiight right |
23:26.19 | MassStash | koo thanks |
23:26.37 | MassStash | you see acl's recovery pic? pretty sweet |
23:26.39 | Cotulla | my opinion is oldscholl - every time when u flash rom it must clear all device, to prevent random bugs. |
23:26.47 | MassStash | werd to that |
23:26.53 | MassStash | same mentality here |
23:27.01 | Cotulla | I am against "updates" |
23:27.17 | MassStash | got cha |
23:31.20 | *** join/#htc-linux Erikson (~Erik@p54B73A97.dip.t-dialin.net) |
23:33.17 | MassStash | how's things coming along for you then Cotulla? |
23:33.51 | dcordes | gnight |
23:34.17 | RaiderX | is there a WP7 development channel? |
23:34.32 | *** join/#htc-linux daevoq (~dae@host243-0-dynamic.18-79-r.retail.telecomitalia.it) |
23:37.18 | *** join/#htc-linux manekineko (~manekinek@newyork.homeserver.com) |
23:37.52 | Cotulla | no |
23:38.09 | *** join/#htc-linux parmaster (par@dipole.idlepattern.com) |
23:40.59 | jonpry | htc-wp7? |
23:41.55 | Cotulla | there was some channel |
23:42.00 | MassStash | htc-WP7 Pro? |
23:42.03 | Cotulla | htc-hd2-wp |
23:42.05 | Cotulla | maybe |
23:42.07 | Cotulla | or maybe nto |
23:44.27 | fakker | htc-wp-chat |
23:44.36 | fakker | haha |
23:44.45 | jonpry | what is easiest config file format to parse? |
23:47.01 | Cotulla | INI |
23:47.04 | Cotulla | or just plain text |
23:47.10 | Cotulla | one line - one item |
23:47.23 | Cotulla | but I ask this question too :P |
23:47.39 | jonpry | yeah INI could work |
23:47.51 | Cotulla | but for INI u need parser |
23:48.14 | jonpry | i know, but it needs to be human readable |
23:48.21 | jonpry | bunch of floats will suck |
23:48.44 | jonpry | i don't know about categories. maybe just foo=bar |
23:49.22 | Cotulla | floats will suck anyway |