| 00:04.10 | jonpry | do phone calls work? |
| 00:05.54 | jonpry | you can disable sound temporarily by editing amss_para.c in amss_6125_para[] change AUDMGR_VERS to something besides 0 |
| 00:43.44 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
| 00:45.43 | detule | CptAJ, don't think your problems are wakelock related .... they are smd related....smd is buggy |
| 00:45.50 | detule | and seriously dated |
| 00:46.30 | detule | we occassinally see those smd related kernel panics on other devices as well |
| 00:47.47 | detule | ahigerd, that 3.3 kernel I gave you had stock (3.2) smd setup... |
| 00:56.37 | ahigerd | Doesn't seem relevant on further inspection |
| 00:56.44 | ahigerd | GBX has no data no matter how I configure it |
| 00:56.49 | ahigerd | Or at least no data in the main browser |
| 00:56.51 | ahigerd | Everything else works |
| 00:57.16 | detule | you get no data at all? |
| 00:58.04 | ahigerd | In the browser I get maybe a few seconds of data (enough to load the Google landing page) and then all subsequent pageviews just stall out. |
| 00:58.13 | ahigerd | In FRX or WinMo, it works fine |
| 00:58.19 | ahigerd | The strange part is it worked in GBX a couple weeks ago |
| 00:58.38 | ahigerd | Even when I revert to totally stock GBX kernel/rootfs with a fresh datafs |
| 01:00.10 | ahigerd | That is, it doesn't work even then |
| 01:00.27 | ahigerd | For a while I thought tmo was pushing bad configuration to me, except as I said it works in FRX/WinMo |
| 01:00.36 | ahigerd | And it works in every other app I've tested |
| 01:17.55 | jonpry | hey detule |
| 01:18.53 | detule | hi jonpry |
| 01:19.32 | detule | thanks for cherry-picking that into 3.2 |
| 01:20.10 | jonpry | yeah np |
| 01:20.24 | jonpry | we need that 3.3 autobuild |
| 01:20.48 | jonpry | and i guess they are up to 3.3.2 already |
| 01:23.19 | *** join/#htc-linux mitsutaka (~mitsutaka@221.219.116.16) |
| 01:28.14 | detule | yeah that autobuild would be ncie |
| 01:31.38 | jonpry | gotta keep the user up to date |
| 01:35.14 | detule | THE user ha |
| 01:35.27 | ahigerd | Who, me? XD |
| 01:35.52 | detule | rpierce99 too, though i don't think we'll be able to get him off this 51hr kernel |
| 01:35.58 | ahigerd | Heh |
| 01:36.08 | ahigerd | I'm good with 16 |
| 01:36.15 | ahigerd | It's enough |
| 01:38.30 | jonpry | 51hr? |
| 01:38.47 | ahigerd | Apparently he got over two days of battery life |
| 01:38.53 | jonpry | on what? |
| 01:38.57 | ahigerd | RHOD300 |
| 01:39.11 | *** join/#htc-linux mitsutaka (~mitsutaka@219.143.36.82) |
| 01:39.12 | jonpry | which kernel? |
| 01:39.14 | detule | 3.1 |
| 01:39.22 | jonpry | oh wow |
| 01:39.25 | detule | no data |
| 01:39.30 | ahigerd | Wifi on though |
| 01:39.35 | jonpry | the wizzard had a 300 |
| 01:39.48 | ahigerd | Though I suppose if he had background data turned off that wifi wouldn't be doing much |
| 01:40.12 | detule | currently our kernel doesn't suspend with wifi ON so that makes it even more impressive |
| 01:40.27 | ahigerd | ... oh REALLY. No wonder it dies so fast when I forget to turn it off. |
| 01:40.30 | jonpry | hrm |
| 01:40.34 | jonpry | idle pc? |
| 01:40.49 | ahigerd | PC wouldn't be happening if it's not suspending, no? |
| 01:41.01 | jonpry | idle pc will |
| 01:41.03 | ahigerd | Oh |
| 01:41.22 | detule | i have no idea what's broken with wifi+suspend perhaps i mucked up the embedded sdio patches, though i am pretty sure it does the same with the .39 from the autobuild |
| 01:42.56 | jonpry | wakelock? |
| 01:43.17 | detule | i actually have no clue let me try and see what errors out |
| 01:43.25 | detule | never cared enough to even check dmesg |
| 01:43.50 | jonpry | i can't believe rpierce99 is off of 27. thats like going to the darkside |
| 01:44.27 | detule | he's been off of .27 for a while afaik |
| 01:44.34 | detule | he just didn't have a rhod for a while |
| 01:45.08 | detule | 04-15 21:44:08.822 WARN : dhd_set_suspend: force extra Suspend setting |
| 01:45.08 | detule | 04-15 21:44:08.852 WARN : dhd_get_dtim_skip failed code -23 |
| 01:47.34 | detule | <PROTECTED> |
| 01:50.34 | jonpry | dhd_common? |
| 01:50.45 | detule | 04-15 21:44:09.872 WARN : stop_drawing_early_suspend: timeout waiting for userspace to stop drawing |
| 01:50.56 | detule | that might be the more important line |
| 01:52.11 | detule | i don't see it in dhd_common |
| 01:52.47 | jonpry | int |
| 01:52.47 | jonpry | dhd_get_dtim_skip(dhd_pub_t *dhd) |
| 01:54.34 | detule | derp |
| 01:56.36 | detule | i don't see a wake unlcok in the suspend path |
| 01:57.10 | detule | where _did_ you get this bcm4329 |
| 01:57.27 | detule | there might be an argument made to just pilfer it from google |
| 01:57.29 | detule | s .39 tree |
| 01:57.30 | jonpry | typically there isn't one. as the suspend path only gets run after all wakelocks have expired |
| 01:57.55 | jonpry | well iirc i got it from some nv tegra tree |
| 01:58.12 | jonpry | but it was horrible. i tried 10's of bcm drivers and all the different firmwares |
| 01:58.31 | jonpry | and different methods of doing sdio_embedded |
| 02:00.01 | jonpry | does the dmesg even say failed to suspend? it should spam dmesg like a bajillion times if its actually failing |
| 02:00.24 | detule | google has an extra wake unlcok in the earlysuspend path |
| 02:00.45 | detule | https://github.com/android/kernel_common/blob/android-2.6.39/drivers/net/wireless/bcm4329/dhd_linux.c#L643 |
| 02:01.32 | detule | derp we have something like that too |
| 02:01.59 | jonpry | you have a dmesg from this suspend? |
| 02:03.26 | detule | http://pastebin.com/bzVyqL43 |
| 02:03.55 | *** join/#htc-linux swc|666 (~goldbond@Aircrack-NG/Friend) |
| 02:04.23 | jonpry | it seems to have worked just fine |
| 02:04.32 | jonpry | slept for some 49 seconds |
| 02:05.32 | jonpry | failed suspend is *way* different than this |
| 02:05.57 | detule | hm let me get you a proper sequence for comparison |
| 02:06.34 | jonpry | sits there spamming trying to suspend every few milliseconds. and the kernel pm code says things like "driver that failed to suspend: bcm4329" |
| 02:06.59 | detule | i am using led debugging too |
| 02:07.10 | detule | and led stays orange it definitely doesn't call all the drivers' suspend calls |
| 02:09.22 | detule | here's a proper sequence http://pastebin.com/gtqJnMKV |
| 02:09.40 | detule | i only let it sleep for a few seconds |
| 02:11.03 | jonpry | ic |
| 02:11.21 | jonpry | maybe thats the key to power savings, just no pc |
| 02:11.26 | detule | lol |
| 02:11.47 | jonpry | back in the day when i got that bad pm.c from wistilt2. it did in fact use more power in collapse |
| 02:15.23 | detule | this stop_drawing_early_suspend that's in fbearlysuspend? |
| 02:16.17 | jonpry | not sure. but its in both the good and bad logs |
| 02:56.17 | detule | yeah def a wake lock issue |
| 02:56.17 | detule | 04-15 22:55:14.884 INFO : active wake lock wlan_wake |
| 03:05.28 | *** join/#htc-linux bukington (~bukington@fac34-2-82-228-151-145.fbx.proxad.net) |
| 03:18.14 | rpierce99 | detule: this thing is definitely killing wifi and sleeping, i just woke it up and watched wifi reconnect and then the indicator turn green |
| 03:19.56 | arrrghhh | i thought that was an ancient bug |
| 03:20.09 | arrrghhh | even when you would try to force wifi on, it would collapse in sleep |
| 03:20.52 | rpierce99 | that's what i remember too |
| 03:20.58 | rpierce99 | i never used wifi much before |
| 03:21.05 | arrrghhh | neither |
| 03:21.11 | arrrghhh | but i do remember people complaining about that |
| 03:21.27 | arrrghhh | users who would try to use the phone VoIP only, wifi only. |
| 03:21.35 | arrrghhh | painful |
| 03:24.02 | arrrghhh | oh well |
| 03:24.12 | arrrghhh | i hear wifi actually drains less @ idle than the cell radio |
| 03:24.22 | arrrghhh | which makes sense considering how much closer you are to a WAP |
| 03:30.49 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
| 03:31.39 | detule | rpierce99, can you easily pull dmesg ? |
| 03:31.53 | rpierce99 | i have getlogs installed :) |
| 03:32.06 | detule | it might be powering down the wifi radio but it's not releasing a wake lock so the rest of the system doesn't suspend |
| 03:32.46 | rpierce99 | give me the path and value for led regime and i'll swap it |
| 03:32.48 | detule | there's some stupid counting logic in the driver that counts how many times a wake lock has been calledÂ…but perhaps it should be unlocking the wakelock unconditionally on early suspend |
| 03:33.08 | detule | sys/class/board_htcrhodium_led/parameters |
| 03:33.10 | detule | led_regime |
| 03:33.13 | detule | echo "2" > |
| 03:33.39 | detule | er |
| 03:33.39 | detule | no |
| 03:33.49 | detule | sys/module/board_htcrhodium_led ? |
| 03:33.55 | rpierce99 | i'll explore |
| 03:33.59 | detule | find /sys -name led_regime |
| 03:44.08 | detule | actually with this counting logic in place i am not sure how this guy would ever suspend properly |
| 04:04.34 | rpierce99 | so it looks like this thing definitely doesn't do the quick sleep it is supposed to, it's still orange after a couple minutes, but if i remember right, after about 15 minutes wifi will die and the phone will sleep |
| 04:37.51 | *** join/#htc-linux MacDrunk (~marper@201.165.128.158) |
| 04:38.40 | *** part/#htc-linux MacDrunk (~marper@201.165.128.158) |
| 04:43.57 | rpierce99 | still orange, i'll let it sit overnight and see what happens |
| 06:16.04 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 06:31.42 | *** join/#htc-linux kiozen (~kiozen@p578a42db.dip0.t-ipconnect.de) |
| 06:45.58 | *** join/#htc-linux ogonzuhnina (~ogonzuhni@user-164-126-138-173.play-internet.pl) |
| 06:46.38 | *** join/#htc-linux ogonzuhnina (~ogonzuhni@user-164-126-138-173.play-internet.pl) |
| 08:04.26 | *** join/#htc-linux gauner1986 (~Miranda@ip-109-91-241-106.unitymediagroup.de) |
| 08:05.33 | *** join/#htc-linux ogonzuhnina (~ogonzuhni@user-164-126-138-173.play-internet.pl) |
| 08:22.38 | *** join/#htc-linux zeusk (77eb360e@gateway/web/freenode/ip.119.235.54.14) |
| 09:23.30 | *** join/#htc-linux lipp[a] (~lippa@101.160.12.182) |
| 09:44.51 | *** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 09:49.45 | *** join/#htc-linux fon (~fon@79.173.145.154) |
| 09:49.58 | *** part/#htc-linux fon (~fon@79.173.145.154) |
| 10:28.15 | *** join/#htc-linux balans21 (~user@535466BE.cm-6-5b.dynamic.ziggo.nl) |
| 10:33.59 | *** join/#htc-linux zeusk (77eb360e@gateway/web/freenode/ip.119.235.54.14) |
| 10:42.17 | zeusk | htc-linux is listing leo with ds2746 on device page but in i2c addresses, it is listed as ds2745 :o |
| 10:48.49 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
| 10:49.12 | *** join/#htc-linux Bry8Star{EB (Bry8Star-@2607:f128:42:fb::2) |
| 10:51.11 | *** join/#htc-linux mitsutaka (~mitsutaka@221.219.116.16) |
| 11:05.55 | *** join/#htc-linux technikfreak (~ownpgroam@unaffiliated/technikfreak) |
| 11:06.21 | technikfreak | hi i have some problems to get the clockworkmod on the hdc dh2, magldr 1.13 is on there already |
| 11:06.48 | technikfreak | i have tried all the suggested things on the forum but it won't help |
| 11:07.32 | *** join/#htc-linux MethoS (~clemens@134.102.106.250) |
| 11:12.38 | Willd | Damn this "hdc dh2" sound cool |
| 11:25.59 | zeusk | but this is not support place, have you tried asking in thread ? |
| 11:47.58 | *** join/#htc-linux helicopter88 (~helicopte@95.235.115.175) |
| 12:04.32 | technikfreak | i have searched in all threads for a slution and i a proud user of irc |
| 12:06.57 | fakker | then you haven't seached hard enough |
| 12:07.09 | fakker | there are stickies where people have made apps; which include clk, or mag |
| 12:07.15 | fakker | with different partition sizes and recovery |
| 12:07.23 | fakker | heh |
| 12:08.15 | technikfreak | yes i ahve tried from 200-800 mb |
| 12:08.26 | technikfreak | use the newest magladr |
| 12:08.39 | technikfreak | also whipe a lot ect,... |
| 12:12.38 | *** join/#htc-linux Alex[sp3dev] (~alexander@178.76.204.13) |
| 12:22.59 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
| 12:36.28 | *** join/#htc-linux zeusj (77eb360e@gateway/web/freenode/ip.119.235.54.14) |
| 13:13.11 | *** join/#htc-linux Alex[sp3dev] (~alexander@178.76.204.13) |
| 13:45.01 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
| 13:48.48 | *** join/#htc-linux infernix (nix@unaffiliated/infernix) |
| 13:51.38 | *** join/#htc-linux Alex[sp3dev] (~alexander@178.76.204.13) |
| 13:53.15 | *** join/#htc-linux balans2 (~user@535466BE.cm-6-5b.dynamic.ziggo.nl) |
| 14:25.10 | *** part/#htc-linux ftoz (~root@9-151.gtt-net.cz) |
| 15:00.29 | *** join/#htc-linux apt (~apt@rikers.org) |
| 15:00.29 | *** topic/#htc-linux is Welcome to the HTC Linux project | Community portal & WiKi http://htc-linux.org | For IRC logs, HaRET & kernel mailing lists etc. see http://htc-linux.org/wiki/index.php?title=Contact | The htc-linux.org project is not affiliated with the HTC Corporation | This channel is for development purposes - Join #htc-linux-chat for offtopic |
| 15:12.31 | *** join/#htc-linux kiozen (~kiozen@ppp-88-217-12-182.dynamic.mnet-online.de) |
| 15:19.25 | *** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring) |
| 15:21.34 | *** join/#htc-linux zeusk (77eb360e@gateway/web/freenode/ip.119.235.54.14) |
| 15:41.30 | *** join/#htc-linux balans21 (~user@82-170-217-205.ip.telfort.nl) |
| 15:44.35 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-107-19.dhcp.stcd.mn.charter.com) |
| 15:45.48 | *** join/#htc-linux zeusk_ (77eb360e@gateway/web/freenode/ip.119.235.54.14) |
| 15:56.50 | *** join/#htc-linux ahigerd (~ahigerd@173.218.156.102) |
| 15:56.50 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
| 16:11.10 | *** join/#htc-linux DuperMan (~Duper@85-250-29-177.bb.netvision.net.il) |
| 16:14.05 | *** join/#htc-linux zeusk (~androirc@117.195.142.103) |
| 16:14.40 | zeusk | Great, my first crash on windows 8 from the audio driver... |
| 16:15.04 | zeusk | oops wrong channel |
| 16:32.54 | *** join/#htc-linux zeusk (77eb360e@gateway/web/freenode/ip.119.235.54.14) |
| 16:33.39 | *** join/#htc-linux |Jeroen| (~jeroen@d5153E8CC.access.telenet.be) |
| 16:35.14 | ahigerd | stares at phone |
| 16:35.22 | ahigerd | ... Google Play has gone missing |
| 16:36.41 | zeusk | are you connected to internet ? (can contact google servers) ? |
| 16:36.49 | ahigerd | No, I mean, the app is gone |
| 16:37.00 | ahigerd | Missing from my menu, generic icon on home screen |
| 16:39.20 | ahigerd | Well, I did an "Uninstall Updates" and it showed up so I can re-update |
| 16:45.35 | CptAJ | shutting down audio from amss_para had no effect |
| 16:45.49 | CptAJ | (on the crashes that is. Calls have no sound now XD) |
| 16:46.35 | ahigerd | Heh, fun! |
| 16:46.41 | ahigerd | Well, at least we know that's not it :P |
| 16:47.01 | CptAJ | yup. its SOME progress ;) |
| 16:48.10 | CptAJ | http://www.droid-developers.org/wiki/Main_Page this project has such a good wiki |
| 16:48.21 | CptAJ | at least on the surface |
| 16:48.33 | CptAJ | they even have flowcharts XD |
| 16:48.50 | CptAJ | how have they not cracked that thing yet? >_> |
| 16:48.50 | ahigerd | Flowcharts make everything better! |
| 16:49.04 | CptAJ | flowcharts show you mean business |
| 16:50.10 | detule | CptAJ, still having those smd panics? |
| 16:50.52 | CptAJ | the ARM 9 crashed ones |
| 16:51.08 | CptAJ | not the wakelock traceback ones |
| 16:51.57 | CptAJ | http://paste2.org/p/1982053 these ones |
| 16:55.37 | detule | if you disable data any better? |
| 16:56.26 | CptAJ | if I go into airplane mode its perfect. I'll test with data off |
| 16:56.55 | CptAJ | (I'm pretty sure it crashes with data off but I don't remember 100% I'll test again) |
| 17:03.08 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-hjndrxvdmgxdzczj) |
| 17:06.33 | CptAJ | http://paste2.org/p/1982066 just got this one too. (Browsing with data on) |
| 17:06.41 | CptAJ | different than the one before |
| 17:06.46 | CptAJ | testing with data off now |
| 17:08.12 | detule | the close-open sequence in our smd is pretty antiquated and has been known to cause a panic or two |
| 17:08.58 | detule | for some reason your data channel cycles through these close-open sequences quite a lot |
| 17:16.42 | CptAJ | well, data reliability is incredibly bad here. could that be the reason for the cycling? |
| 17:17.35 | detule | could be |
| 17:17.57 | *** join/#htc-linux rana_ (6a4d808e@gateway/web/freenode/ip.106.77.128.142) |
| 17:18.13 | CptAJ | no way around that though. does 35 have a different smd? |
| 17:18.25 | detule | no...but i do :) |
| 17:18.36 | CptAJ | show me dem XD |
| 17:18.55 | detule | see first if you can verify that it doesn't die on you without data |
| 17:19.02 | CptAJ | doing that just now |
| 17:19.14 | detule | then we can cherry pick what you need you likely won't need this full blown codeaurora smd |
| 17:19.31 | detule | just need to spruce up the close-open sequence to wait for ack from the server |
| 17:20.20 | detule | here's the branch btw https://gitorious.org/~detule/linux-msm-rhod/detules-linux-msm-rhod/commits/3.3-smd_update |
| 17:21.20 | *** join/#htc-linux gauner1986 (~Miranda@ip-109-91-241-106.unitymediagroup.de) |
| 17:26.40 | CptAJ | doing my best to crash it |
| 17:26.47 | CptAJ | calls, music, games, filemanager |
| 17:26.49 | CptAJ | still going |
| 17:26.53 | CptAJ | but this has happened before |
| 17:27.27 | detule | yeah false positives are a curse |
| 17:36.11 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 18:07.59 | *** join/#htc-linux kiozen (~kiozen@ppp-88-217-7-86.dynamic.mnet-online.de) |
| 18:09.40 | CptAJ | Still running. I called from and to it a few times. Left the music on while having lunch. and called some more. |
| 18:13.06 | CptAJ | gotta go do some work now. we'll talk about that smd when I'm back ;) |
| 18:14.37 | *** join/#htc-linux rob_w (~bob@unaffiliated/rob-w/x-1112029) |
| 18:42.48 | *** join/#htc-linux ahigerd (~ahigerd@173.218.156.102) |
| 18:43.06 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
| 18:50.48 | hyc | hm. What is CptAJ testing, GSM or CDMA? |
| 18:51.20 | hyc | How did codeaurora get their info on MSM? |
| 18:53.26 | detule | qualcomm has their msm trees on codeaurora |
| 18:55.05 | hyc | oh? and they're public? |
| 18:55.27 | detule | sure https://www.codeaurora.org/gitweb/quic/la/?p=kernel/msm.git;a=summary |
| 18:55.57 | hyc | cool |
| 18:56.10 | hyc | so does this new smd fix the gsm errors? |
| 18:56.52 | detule | no, i think we tracked down the gsm errors to something else...the patchset is in the 3.2 autobuild if you want to give them a try |
| 18:57.21 | detule | ahigerd, has been on those patches for roughly a week now, and even without a hacked ril he hasn't reported any "AT< WCDMA" weirdness |
| 18:57.22 | hyc | really? dang, how did I miss that being solved... |
| 18:57.31 | hyc | cool |
| 18:59.23 | hyc | April 4 commit, fixed Mutex handling by jonpry? |
| 19:10.26 | hyc | Mmmm. Nice. Looks like it's a win here too. |
| 19:12.03 | hyc | hmmm. need to look at this more closely. was the actual fix, dropping the high speed hack on non-PPP channel, or was the actual fix fixing the mutex? |
| 19:12.25 | hyc | I seem to recall running that March 27 patch before with no positive effect |
| 19:12.41 | hyc | if that's the case then we can go back to "high speed hack" on all channels, yes? |
| 19:15.46 | ahigerd | We turned off idle PC in the latest 3.2 autobuild too |
| 19:15.49 | ahigerd | I suspect that may be the real win |
| 19:17.02 | hyc | hmmm you're talking about the last 4 commits by detule? |
| 19:17.13 | hyc | Wasn't sure they were in the autobuild since that said Fri April 13 |
| 19:17.24 | hyc | and gitorious says Sat Apr 14 |
| 19:17.32 | hyc | but I guess that's timezone diff |
| 19:18.02 | detule | ahigerd, had been running a homebrewed test kernel with those patches in before jonpry pushed them out |
| 19:18.10 | hyc | ok |
| 19:18.22 | hyc | now if only the battery guage was closer to reality |
| 19:18.34 | ahigerd | Mine's acting surprisingly reliable |
| 19:18.38 | hyc | it unfortunately says 100% way too soon |
| 19:18.39 | ahigerd | I may have finally kicked the calibration into gear |
| 19:19.01 | ahigerd | The top end is a little finicky but the bottom end is pretty accurate now |
| 19:19.01 | hyc | the only way I can get a good charge is turning off, or switched into winmo |
| 19:19.16 | hyc | otherwise the phone dies while android says it still has power |
| 19:19.29 | hyc | hmmm |
| 19:19.39 | detule | i am sure it's placebo but i am seeing better battery life with this new timer....also in a few test runs i get better battery life when charging in android over night versus charging with device off, or charging in winmo |
| 19:20.07 | hyc | ok, will give this new kernel a bit of time to sink in |
| 19:20.26 | hyc | but it still just charged from 72% to 90% in like 1 minute. not convincing. :P |
| 19:20.43 | detule | at least one thing i am fairly certain is not placebo -> it is snappier....not trying to power collapse 20-30 times a second is probably doing its thing |
| 19:20.53 | hyc | cool |
| 19:21.03 | ahigerd | I don't reproduce detule's improved life ;) |
| 19:21.12 | ahigerd | I'm right about at parity to before |
| 19:21.24 | detule | detule's life hasn't been improved |
| 19:21.25 | ahigerd | At best maybe 30% better |
| 19:21.30 | detule | certainly not since he got introduced to xdandroid |
| 19:21.33 | ahigerd | XD |
| 19:21.49 | hyc | :P |
| 19:23.30 | ahigerd | (And I say "maybe 30% better" because I'm getting ~16 hours on a charge and I had never pushed it beyond 12 before) |
| 19:24.19 | hyc | well that sounds pretty good. |
| 19:24.39 | ahigerd | But I never tested it formally |
| 19:24.40 | hyc | I'm more accustomed to getting at least 24 hours. i.e. only needing to charge once a night. |
| 19:24.51 | hyc | these days it hasn't lasted thru the night. |
| 19:24.56 | ahigerd | I always unplugged my phone around 8AM or so and plugged it back in before bed |
| 19:25.02 | hyc | hopefully this will improve now |
| 19:25.07 | ahigerd | So I guess that would have been 14 hours or so |
| 19:25.22 | ahigerd | So that's why I feel it's basically unchanged |
| 19:25.31 | hyc | ok |
| 19:25.50 | hyc | I definitely had better life with data off... |
| 19:26.02 | hyc | still too much crap going on in background |
| 19:26.19 | ahigerd | Well yeah :P |
| 19:26.27 | ahigerd | This is with 3G on and background data enabled |
| 19:26.46 | ahigerd | I could probably pull 24 hours if I turned data off when I wasn't using it |
| 19:34.14 | detule | whatever i don't trust even my own "feels like battery life..."....on the other hand sounds like acl is working on the battery driver so perhaps he'll come up with something we can then cherry pick |
| 19:34.28 | ahigerd | I was thinking the same |
| 19:34.29 | hyc | cool |
| 19:35.19 | jonpry | no high speed hack on all channels. it breaks my phone |
| 19:36.43 | detule | btw there's something in the codeaurora tty about copying directly from the smd buffer to the userland buffer completely circumventing the tty line discipline |
| 19:36.59 | *** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 19:37.07 | jonpry | they found our patches? |
| 19:37.34 | detule | lol nah your and hyc speed up hacks still used the line discipline |
| 19:38.05 | jonpry | but the ldisc no longer has its own buffer |
| 19:38.35 | hyc | hm, my hack copied directly to userland |
| 19:38.52 | ahigerd | Why not take a look at the patch and compare? :P |
| 19:38.54 | hyc | unless I misremember. which could easily be true |
| 19:39.01 | hyc | sounds like a good idea |
| 19:39.04 | detule | hyc i thought your hack called jonpry's callback which used the ldisc |
| 19:39.10 | detule | here's what i am talking about https://www.codeaurora.org/gitweb/quic/le/?p=kernel/msm.git;a=commitdiff;h=ab46a8d513817169d8f3f4e1e3fb838e100836d5;hp=e1b157dd460311921c6262f6add301f784e08994 |
| 19:39.30 | detule | though not sure how this could be used with stream channels |
| 19:39.31 | jonpry | it calls ldisc apis but more or less bypasses the ldisc |
| 19:40.23 | detule | oh thanks for the clarification |
| 19:44.20 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
| 19:48.28 | hyc | hmm ok yeah. our hacks just bypassed that ridiculous flip buffer |
| 19:49.58 | hyc | fwiw, I did have measurably faster throughput on speedtest.net... |
| 19:50.20 | ahigerd | We're still using your hack on data, right? |
| 19:50.25 | hyc | I think so |
| 19:50.31 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-164-174.netcologne.de) |
| 19:50.59 | detule | which tty_ldisc_ops implementation are we using anyways? |
| 19:51.14 | hyc | raw, IIRC |
| 19:51.21 | hyc | there is a specific ppp ldisc too |
| 19:51.44 | hyc | my memory is too fuzzy |
| 19:53.34 | detule | i suspect the one in n_tty.c |
| 19:55.12 | detule | anyways there too receive_buf does a memcpy before copy to user |
| 19:56.23 | detule | but it might all be a moot point, as it seems to me this codeaurora user buffer business is almost geared towards a char device not a tty device....as in need to keep track of the user buffer somewhere in the smd layer |
| 19:56.58 | hyc | yeah, sounds suspect |
| 19:57.38 | hyc | drivers/net/ppp_async.c |
| 19:59.52 | *** join/#htc-linux yuuki_ (~yuuki@dsbg-4d0536a8.pool.mediaWays.net) |
| 20:14.09 | *** join/#htc-linux Ondalf (~ondalf@cable-roi-fe98dc00-121.dhcp.inet.fi) |
| 20:14.38 | *** join/#htc-linux ahigerd (~ahigerd@173.218.156.102) |
| 20:14.39 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
| 20:15.14 | ahigerd | glares annoyedly at phone |
| 20:15.24 | ahigerd | I'm having major problems with data right now... |
| 20:15.29 | ahigerd | I thought it was bad bfore |
| 20:15.30 | ahigerd | before* |
| 20:15.41 | ahigerd | It's getting worse x.x |
| 20:36.47 | jonpry | lol |
| 20:42.13 | jonpry | maybe your isp is having some kind of peering issue |
| 20:42.44 | ahigerd | It works in WinMo |
| 20:42.54 | ahigerd | And in FRX |
| 21:21.22 | detule | the android browser has always been buggy for me (similar to what you are describing) ... i've never had issues with opera mini hanging on loading pages |
| 22:04.49 | *** join/#htc-linux T-Junk (atari@i22172.upc-i.chello.nl) |
| 22:29.52 | *** join/#htc-linux ahigerd (~ahigerd@173.218.156.102) |
| 22:29.52 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
| 23:00.19 | *** join/#htc-linux surge (surge@pool-98-118-183-214.bflony.fios.verizon.net) |
| 23:05.01 | *** join/#htc-linux rajkosto (~rajkosto@de.iandoh.info) |
| 23:07.50 | *** join/#htc-linux rajkosto (~rajkosto@de.iandoh.info) |
| 23:08.00 | *** join/#htc-linux ccube (ccube@nx.ccube.de) |
| 23:11.00 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
| 23:28.57 | ahigerd | ohai, 3.3 autobuilds are up |
| 23:29.52 | jonpry | way cool |
| 23:45.30 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-mjppjxoemiyfqbdj) |