Friday, 2011-11-04

*** raymonddull has joined #xdandroid00:11
*** ray|yar has quit IRC00:14
*** mayday_jay has joined #xdandroid00:32
*** rpierce99 has quit IRC00:44
*** raymonddull has quit IRC00:57
*** raymonddull has joined #xdandroid00:57
*** bzo has quit IRC01:36
*** ray|yar has joined #xdandroid02:36
*** raymonddull has quit IRC02:40
*** ryannathansSLEEP is now known as ryannathans03:09
*** AndersG has joined #xdandroid03:52
*** ray|yar has quit IRC03:54
*** ray|yar has joined #xdandroid03:54
*** kalemas has joined #xdandroid05:19
*** hardwalker has quit IRC06:24
*** mgross029 has joined #xdandroid07:48
*** ray|yar has quit IRC08:43
*** ray|yar has joined #xdandroid08:44
*** kalemas has left #xdandroid08:57
*** helicopter88 has joined #xdandroid10:07
*** dario_ has joined #xdandroid10:14
dario_hello...i'm trying to use cwm but it hanging up completely with "r5:00000000 r4:00000000"10:17
*** rpierce99 has joined #xdandroid10:18
dario_it freeze on this error: task rpcrouter:15 blocked10:26
*** virsys has quit IRC10:32
*** mgross029 has quit IRC10:45
*** ray|yar has quit IRC10:56
*** ray|yar has joined #xdandroid10:57
*** mgross029 has joined #xdandroid10:57
helicopter88dario_: do you have a data.img in the same folder as haret?10:59
dario_helicopter88 :no11:00
helicopter88weird11:01
mgross029helicopter88, I think he was using CWM, so dario_ you may want to ask over at #neopeek11:01
dario_mgross029 : thx for the suggestiong11:02
*** AndersG has left #xdandroid11:03
*** AndersG has joined #xdandroid11:04
*** programmer8922 has joined #xdandroid11:07
*** emwe has joined #xdandroid11:18
*** programmer8922 has quit IRC11:23
emwedetule: hey. you said you where kexec'ing like wild. i tend to say i have wasted 50% of my spare time waiting for reboots. what was the freakin' reason it's not in mainline userland, yet? stinebd?11:25
detuleemwe, you put kexec in the GB power menu and i'm GB-ing full time11:27
emweso kernel needs kexec support and the menu entry (frameworks base?) is missing, yap?11:27
emwethere was system_extras or _core which had a change for that, too, no?11:27
detuleright along with the binary and the kexec script but hyc has all those in system/xbin11:28
emwedetule: you are building gb yourself now?11:28
detuleemwe, if i was building it myself i would've probably had it in ther eby now :) i am apprehensive to start pulling because of all those manifest changes....plus my machine is 32-bit - isn't there something about it needing to eb 64 for gb?11:28
emweah sorry, i misread your first statement... it sounded liked you did put on powermenu and were done.... but that was if i *would* put it in...11:29
emweand regarding building yeah.. i dunno. i am 64bit and stock java6. no mods for like for froyo. just building...11:30
emweok, then i need to figure that stuff and talk to stinebd why it's not in. it would drop turn-around times insanely.11:31
detuleit makes life ridiculously easy adb push zImage/kexec11:31
emwedetule: perhaps you can provide me with a link or the required binaries/scripts via mail?11:31
emwedoes the zImage have to be placed at a specific location on sd?11:31
emweor how does it figure from where to load?11:32
detulei think it steals the commandline from the original boot, so wherever the kernel was originally placed when it booted11:32
emweaha11:32
emweah, i think i remember a kernel change for that as well. will go figure ehm... sometime soon i hope.11:33
detulei'll pull the binary/script from 06+ tonight....you ask me the ONE day when i am trying out a new SD card at work that has ONLY your gb build11:34
emwenah, don't worry. busy with acoustic still11:35
*** Himalayas has joined #xdandroid11:35
arrrghhh  lol11:35
emwethe incall-volume issue seems to stem from a kernelland change.11:35
emweCURRENT device was moved from id 256 to the one from IDLE .. so 0xd(5225) or 0x11(6125) .. putting it back to 256 on RAPH fixes it at least11:36
emwealso speaker incall vol.11:36
detule0x11 no good?11:36
emwejust made CURRENT 256 again on RAPH with .27-acoustic and incall earpiece and speaker vol-change is working again.11:37
emweneed to call VM again and test out enabling speaker. i think it turned on full vol and i had to change vol for it to actually react correctly.11:37
emwebut it's a start I think.11:38
detulewonder why alex changed it11:38
emwei think JB came up with it also11:38
emweor at least for experimental purposes.11:39
arrrghhhemwe, the current audio stuff does that i think11:39
emwe?11:39
ryannathansemwe: is that issue fixed yet?11:39
arrrghhhthe indicator does not match the actual volume11:39
emweuh wait waht. plz elaborate a bit more arrrghhh  :)11:39
emweryannathans: working on it. hold on...11:39
arrrghhhso you make a call11:39
arrrghhhflip speaker on11:39
emwei am limited with knowledge... :P11:39
arrrghhhtry to adjust the volume11:39
arrrghhhthe indicator on the phone will 'fill'11:40
arrrghhhhowever11:40
arrrghhhi can keep hitting vol up11:40
arrrghhhand the volume continues to increase 2-3 steps beyond "full"11:40
arrrghhhso the visual indicator does not match what is reality11:40
arrrghhhmake sense?  :P11:40
arrrghhhsorry, i'm not sure how to explain it.  kinda confusing.11:40
ryannathanssleepy tiems for me, night11:40
*** ryannathans is now known as ryannathansSLEEP11:40
emwearrrghhh: that is for incall speaker phone or the dialtone issue?11:41
emweah flip speaker on...11:41
emwetough week. sorry.11:41
arrrghhhnp11:42
arrrghhhi actually believe this to apply for most volume indicators11:42
arrrghhhif not all11:42
arrrghhhi guess i just notice it moar when i flip the speakerphone on11:42
arrrghhhsince i'm not watching the screen when using the earpiece...11:42
*** valyk has joined #xdandroid11:44
detulemgross029, thanks btw for the build.prop tweak11:44
mgross029emwe / detule I've noticed in couple of .java files there some comments that say 0 = Global with regards to default_network.  Was this something from a previous port possibly?11:45
mgross029detule, np I may have a better one. :p11:45
detulegit log the_file will show you the history of the file11:45
arrrghhhvalyk, why PM me?11:45
mgross029Thanks I will do that.11:46
valyk:-??11:46
valykdk11:46
arrrghhh...11:46
valykit seems quite faster to me11:46
valyklol11:46
valyk:D11:46
valyki have a lil' problem11:46
arrrghhhtype arr then tab11:46
arrrghhhin the room11:46
arrrghhhno need for PM11:46
arrrghhhkinda rude... that's all.11:47
valykok11:47
valyknext time i will11:47
valykso.. my phone doesn't recognise root even if it is rooted11:47
valyki have installed the froyob v7 rom11:47
detuleoh boy11:47
helicopter88valyk: wrong channel,here's xdandroid support11:47
helicopter88gbx and frx11:47
valykwhatever11:48
valyki will move to neopeek11:48
detuledon't do it11:48
detulewe need you here11:48
valyklol11:48
valykso will i get some help11:48
valykor not11:49
valyk?11:49
detulewell you are clearly a key ingredient to this project so under the threat of moving to neopeek I will gladly assist you with any build you may have issues with11:49
arrrghhhlol11:51
detuledamnit i am turning into arrrghhh11:52
detule:)11:52
detulebbl11:52
arrrghhhteehee11:52
mgross029emwe, I have found the ro.telephony.default_network is noted in the TelephonyManager.java and DatabaseHelper.java.  It grabs the value for the default_network from system.prop, which we currently have set to 011:58
mgross029After reading the comments over the function it states that 0 = Global as I stated above, but 0 = WCDMA Preferred in reality12:00
emwehm.12:00
emweprop name was ro.ril.default_network?12:00
mgross029correct12:00
mgross029After digging through some of the other .java files I found some definitions where Global is actually 712:01
mgross029I will have to go back and find out where I saw that though.  Did not write it down unfortuantely.12:02
valykdetule... can you help me with the root issue??12:11
emweNETWORK_TYPE_UNKNOWN = 0 here mgross029.12:11
arrrghhhtrolololololol12:12
mgross029emwe where did you see that?  I came across that one as well.  My VM host kind of took a dive and I lost all of my data.12:13
helicopter88NETWORK_TYPE_UNKNOWN = 0 same here12:13
emwemgross029: TelephonyManager.java public static final int NETWORK_TYPE_UNKNOWN = 0;12:13
mgross029Ahh that's right.  Thanks.12:14
emwemgross029: looks like the comment in DatabaseHelper.java is wrong. it actually sets PREFERRED_NETWORK_MODE which is int NETWORK_MODE_WCDMA_PREF     = 0; /* GSM/WCDMA (WCDMA preferred) */12:21
mgross029emwe, yeah NETWORK_TYPE is used differently that NETWORK_MODE I believe12:23
emweso now ... we need to distinquish between network_mode ..........you said it :)12:23
emwero.telephony.default_network is about the NETWORK_MODE if they didn't mess up all comments and souce :)12:25
*** kalemas has joined #xdandroid12:25
mgross029Normally with 0 or even 4 I see in the radio log it creates GSM Phone and the creates CDMA Phone.12:25
mgross029When I used 7, which as far as I can tell is Global, I only see it creating a CDMA Phone does not switch back and forth and I do not get a FC message on first boot even.12:26
emwelol, even in TelephonyManager they call it network type. thanks.12:26
mgross029Radio log appears much cleaner to me anyway12:26
mgross029Yeah I think those definitions are for another feature.  Haven't been able to determine yet though.12:27
emwethere's really not much changes in the files between froyo and gb so perhaps an behavioral difference between the system.s12:27
emwealso the RIL api version got increased in GB. our version 3 is still compatible though. perhaps we need some RIL love in order to get it setup correctly. but i haven't had much RIL contact and would have to read into that.12:28
mgross029That is exactly what I was thinking...  Froyo and GB seem to be different animals12:28
*** vinceweis has joined #xdandroid12:28
mgross029I know I saw some references to NETWORK_MODE or preferred network in RIL.java12:29
mgross029Radio log with the 7 option used http://pastebin.com/k3PWvSSj12:31
*** |Jeroen| has joined #xdandroid12:32
emwethis sounds good: World phone has no SIM card installed, just using CDMA12:32
mgross029I've seen that before with other setting, the big one is Network Mode set to 7 then Cdma Subscription set to 1 and right after that Connected to 'rild' socket12:33
mgross029No reference to GSM here12:34
*** austeregrim has joined #xdandroid12:35
mgross029Would be interesting to see how a GSM phone handles 712:35
emwei know you did, but i tend to have lost... do you have a complete radio log handy from a crash? (complete as from the very beginning)12:36
mgross029From a FC?12:36
emweyeah12:41
emwewhere default_network is apparently 012:42
emweperhaps we just need to set the property when we detect the device capabilities when the RIL is initializing.12:42
mgross029I'm checking my pastebin but I may just create a new will probably be faster.12:42
emwedon't wan't you to screw your setup...12:43
emweapart from that.. didn't i wan't to do acoustic? :P12:43
mgross029Oh I have room on my sd to create a new build folder.12:43
mgross029Ooops  Sorry for tearing you away from that!12:43
arrrghhhheh12:45
arrrghhhgotta fix 'em all emwe12:45
emwedetule: didn't you fixup some multiple cring stuff? just stumbled over the RIL_UNSOL_CALL_RING documentation in hardware/ril/include/telephony/ril.h; does that ring a bell regarding missed calls somehow? (apart from you saying you got it fixed up again)12:45
mgross029Radio log origininal system.ext2 with ro.telephony.default_network = 0 setting in build.prop http://pastebin.com/6CVCHS3Y12:59
*** turbolapin has joined #xdandroid13:03
*** fishhead2567 has quit IRC13:10
emwemgross029: why do your pastes always contain that additional blank lines? hard to read...13:20
rpierce99probably a bad text editor, windows most like, \r\n13:20
mgross029Good question. :p  Using Notepad++13:21
mgross029Ooops not I opend that one in Notepad.  bleh sorry13:21
emweok, it first creates a gsm phone and then it notices it's cdma and switches. i dunno how we can circumvent this. perhaps by setting the property when we find out what we really are...13:24
mgross029Well both notepad and notepad++ give me the same output (spaced)13:24
rpierce99so at least on froyo, the creating of gsm or cdma phone depending on values returned by the ril13:25
emwehm. perhaps your console producing them?13:25
mgross029Well the system.prop under device/xdandroid/msm can be modified in the source13:25
rpierce99if the RIL returned a specific value for check sim state (if i recall) it knew it was a cdma phone13:26
emwemgross029: yeah in the source, but this is a runtime thing. perhaps we can help it be setting the prop on ril init.13:26
emwerpierce99: yeah, but userland isn't told about it. it defaults to it's default mode which is 0 (gsm/wcdma) it looks.13:26
mgross029emwe, doesn't it pull from there to create the runtime and build.prop though?13:27
emweuh what? ;)13:27
emweit reads the prop, but it should ask the RIL imho. perhaps it does and I don't see it.13:28
emwesetting the prop is stupid idea of course. it's ro. ;)13:29
*** valyk has quit IRC13:29
mgross029int mode = SystemProperties.getInt("ro.telephony.default_network", -1);  That is what I got from this line in TelephonyManager.java anyway.  SystemProperties (ie system.prop)13:30
mgross029Understood, did not see where the variable for NETWORK_MODE was being set otherwise though13:30
emwemgross029: can you test out with: ro.telephony.default_network=-1 ?13:40
emwei whonder what happens then13:40
emweit returns a PHONE_TYPE_NONE then13:40
mgross029Sure13:40
mgross029./frameworks/base/telephony/java/com/android/internal/telephony/RILConstants.java  Found where the variables are set for each number.13:51
mgross029./frameworks/base/telephony/java/com/android/internal/telephony/RILConstants.java:    int PREFERRED_NETWORK_MODE      = NETWORK_MODE_WCDMA_PREF;13:52
emwethat is what i wrote earlier :) that's where i got it from.13:52
mgross029And that is where the PREFERRED_NETWORK_MODE is set13:52
mgross029Umm no that was network type from TelephonyManager.java right?13:53
emwe17:22 <emwe> mgross029: looks like the comment in DatabaseHelper.java is wrong. it actually sets PREFERRED_NETWORK_MODE which is int NETWORK_MODE_WCDMA_PREF     = 0; /* GSM/WCDMA (WCDMA preferred) */13:53
mgross029Ahh...  :p13:53
mgross029Forgot about that post sorry13:54
emweso yeah, the trial with ro.telephony.default_network=-1 might be worthless, but if you could give it a shot nevertheless, would be nice.13:54
emwenp. glad you found it :)13:54
rpierce99do you guys have this patch in your tree? http://highlandsun.com/hyc/proxy.dif.txt13:54
mgross029I'm just deleting my data.ext4 and rebooting...13:54
emwerpierce99: yes, but strangely only on the gingerbread branch. so ehm... wasn't this supposed to be in froyo initially?13:59
rpierce99it was a fix specifically for GB, http://xdandroid.com/irclogs/%23xdandroid.20110429.log.html14:00
rpierce99towards the end they discuss a GB phone.apk crash14:00
mgross029emwe, well it took the -1 11-04 14:01:52.083 I/PHONE   ( 1564): Network Mode set to -1, but still getting the FC and GSM messages in Radio Log14:05
mgross029Looks the same as if it were set to 014:05
emweok.14:06
emwegit a log?14:06
*** Tandyman100 has quit IRC14:06
mgross029Sure but I thought you didn't like my logs...  :p  http://pastebin.com/tCugVHfJ14:08
mgross029Pulled that one from my console screen instead of a piped txt file.14:08
*** fishhead2567 has joined #xdandroid14:12
rpierce99so from that log it sounds like the phone app and the ril need to be on the same page (gsm/cdma) in order for it to work, but sometimes they get seperated, does your regular logcat say phone app is in gsm or cdma?14:14
*** Tandyman100 has joined #xdandroid14:14
emwegsm from the logs14:15
emwe11-04 14:01:52.764 I/PHONE   ( 1564): Creating GSMPhone14:16
rpierce99gsm from radio log14:16
mgross029rpierce99, if I set the ro.telephony.default_network to 4 or 7 I do not see the switching in the log  Look ^^ there was a pastebin from earlier14:16
rpierce99i know, i'm not talking radio logs14:16
rpierce9904-29 19:52:55.770 D/PhoneApp( 1417): Radio technology switched. Now CDMA is active.14:16
mgross029[12:31] <mgross029> Radio log with the 7 option used http://pastebin.com/k3PWvSSj14:17
emwerpierce99: is this piece of log around the same time as this from radio log? > 11-04 12:57:19.110 D/PHONE   ( 1561): [PhoneProxy] Switching phone from GSMPhone to CDMAPhone14:17
rpierce99so the phone app is on cdmaphone but the ril started up a gsmphone?14:18
rpierce99that doesn't look like the same thing actually, that sounds like it's talking about the ril gsmphone14:18
emwecurrently it looks the phoneapp goes gsm and then switches over.14:19
rpierce99which would be fine if it didn't crash while attempting the switch?14:20
mgross029I think I found something.  Look at these two.  One for CDMA https://gitorious.org/xdandroid/frameworks_base/blobs/gingerbread/telephony/java/com/android/internal/telephony/cdma/CdmaServiceStateTracker.java14:21
mgross029One for GSM https://gitorious.org/xdandroid/frameworks_base/blobs/gingerbread/telephony/java/com/android/internal/telephony/gsm/GsmServiceStateTracker.java14:21
*** f00bar80 has joined #xdandroid14:21
rpierce99i hope there is more text coming14:21
mgross029Do a find for network_mode in both14:21
rpierce99i think i found something, look at these 2 giant source files! :)14:21
mgross029:p14:22
f00bar80is there a data packets transmission log, if yes wher can i find it?14:22
mgross029network_mode seems missing from cdmaservicestatetracker.java14:22
rpierce99there are packet capture apps on the market14:22
f00bar80rpierce99, yea i know that, but i want to know the history of transmission from a month14:23
rpierce99you think your phone kept all of the data packets sent and received for the last month?14:23
mgross029This is showing the gsmstatetracer mPreferredNetworkType = RILConstants.NETWORK_MODE_GLOBAL14:23
f00bar80rpierce99, i thought it can be like any other OS symbian , winmo ..etc, all of them do that14:24
mgross029Which NETWORK_MODE_GLOBAL = 7 in RILConstants.java14:24
rpierce99mgross029: the cdma version doesn't even seem to support the network_mode events14:25
rpierce99err network_type14:25
*** |Jeroen| has quit IRC14:25
rpierce99EVENT_GET_PREFERRED_NETWORK_TYPE, EVENT_SET_PREFERRED_NETWORK_TYPE14:25
f00bar80rpierce99, also maybe i was not clear in my query, i meant a packets counter log14:28
rpierce99oh, how many, not the content?14:28
mgross029rpierce99, so shouldn't they both have those statements in their code?14:29
f00bar80rpierce99, yea14:30
detuleemwe, CALL_RING is working fine, i hardened the ril against missed calls when booting with an active connection in winmo - however this can be avoided by having winmo in airplane mode14:30
emwemgross029: try this one. contains the 2nd patch from hyc referenced by rpierce99. dunno if it helps. http://dl.dropbox.com/u/12183048/htctopaz/packages/Phone.apk14:33
emwedetule: with stock ril people reported missed calls, right? i whondered about that comment. wasn't there some multiple CRING reporting removed which caused missed calls?14:34
emwedetule: from the looks, Phone.apk thinks our ril will send multiple and not internally generate some. 11-04 14:01:57.629 D/PHONE   ( 1564): mDoesRilSendMultipleCallRing=true14:35
mgross029emwe, ok...  I will load it up.  Any thoughts on why these <type>StateTracker.java files differ so much?  ^^^14:35
emwedetule: stumbled over the comment in ril.h and whondered...14:35
f00bar80rpierce99, so if it's a how many , is there a way to know how many packets phone kept for the last month?14:36
emwemgross029: i'd say why not... but i dunno all the stuff happening there. it's like asking me why apple trees carry 5 fruits per branch and plum trees only 414:37
*** rpierce99_ has joined #xdandroid14:37
detuleemwe, yeah our ril sends multiple CRING, it was an issue of how those were handled -> now we just send CALL_RING back to system, before we were doing something else which was not as elegant and caused a storm of ril traffic14:37
emwedetule: ah ok. thanks.14:38
emwedetule: and that is in mainline?14:38
detuleemwe, it is, however like I was saying, it's not hardened against booting with an active winmo connection -> you will miss calls 10 times out of 10 if you booth with winmo not in airplane mode14:39
mgross029emwe, :p  Sorry just thought they should be similar in regards to the main meat of the code but the items would either differ from gsm or cdma statement14:39
emwemgross029: if network mode=0 gsmservicestatetracker will fire up anyway and will detect network and likely switch to cdma14:39
*** rpierce99 has quit IRC14:40
*** rpierce99_ is now known as rpierce9914:40
emweit's unlikely for cdmaservicestatetracker being up first. just an assumption.14:40
mgross029That makes sense14:40
emwedetule: i always boot in non-airplane mode. so it sounds this patch needs to go public, eh?14:40
detuleemwe, hyc calls booting with winmo in non-airplane mode user error14:41
detule:)14:41
emwei call it more burden on supporters and less donations if that patch stays private :P14:42
emweif it has no implications, it should go out.14:43
detuleyeah it's pretty minor with no other implications i just have to find it14:44
emwethe ML will be happy with some traffic.14:45
mgross029emwe, booting now with fresh data.  bind mounting the Phone.apk...  I'll let you know the result14:45
emweyou bindmounters!!!! just remount /system read-write and push to /system/app/14:45
mgross029Yeah I could do that to but it takes longer. :p14:46
emwethat takes longer?!14:46
mgross029Yes by about 2 seconds.  I timed it. :p  j/k14:46
emwebrb514:47
rpierce99f00bar80: I don't see any way to go back that far unless your interface has been up that whole time, it doesn't appear to store it except it /proc and /sys entries, netstat and ifconfig print it out, or you can look in /sys/class/net/<interface>/statistics or /proc/net/tcp14:53
mgross029emwe, so first boot with new Phone.apk still got the FC on com.android.phone14:55
rpierce99what's the stack trace on that crash14:56
emwerpierce99: i think there's been none as stupid as it sounds.14:57
emwemgross029: can you spot some stack trace in logcat?14:57
mgross029I will see what I can find...  Forgot about the SOD usb bug and lost my connection to the phone.14:58
emwePhone.apk will be a bit more chatty actually14:59
rpierce99i have a version of getlogs that is supposed to be GB compatible, don't know if arrrghhh ever tested it though14:59
detulethe funny thing is the stack trace is all about ToneGenerator blah blah14:59
emweso leave both running, logcat and logcat -b radio14:59
emweah i remmeber detule ... hm.14:59
emweah damnit. this stuff drags me off from acoustic....14:59
detule:) perhaps they are related15:00
detulethough i recall acl had these on .27 + gb15:00
detulewith identical stack traces15:00
detuledid he ever share the secret on how he got over those?15:00
rpierce99when hyc originally wrote the RIL he was getting FCs on phone.apk that would go through a bluetooth stack15:01
detule(i should say the stack traces i saw from mgross029 back in the day were about ToneGenerator, don't know if the problem has evolved since then)15:01
detuleacl patched somethiing about the bluetooth stack15:01
detuleright around the time his FCs stoped15:01
detuleemwe,  you know the patch i am talking about15:02
rpierce99hyc's patch was to PhoneProxy to fix that15:02
rpierce99i just wanted to see a stacktrace for this crash to see if it's the same thing and the patch doesn't fix it anymore, or if its a completely different crash15:02
emwedetule: a bt patch? i think i forgot about it. was it a JNI issue with the boolean?15:03
detulehttps://github.com/acl/device_htc_rhodium/commit/f6d1dfa87a78b77226dca49af552ad454134ffb115:03
emwedetule: that one?! that made the phone.apk crash go away for him?15:03
emwei once built an image with that stuff in.15:03
detuleno idea15:03
rpierce99ah, yes it did i believe15:03
emwehm, so my testers haven't tested right?! :P15:04
rpierce99well, no guarantee that's the same crash15:04
detulei am just spurting things off the top of my head here15:04
emwebut why should it crash for cdma folks only?15:04
rpierce99hycs original crash was something tried to use the gsmphone after it had been disposed, or something along those lines, so he spun up a new phone prior to killing the old phone, or something like that15:05
detuleno idea ToneGenerator has some CDMA specific values it uses15:05
rpierce99gsm folks wouldn't be affected in that case because there's no transition15:05
rpierce99it could be the bluetooth stack that tries to access the old phone15:05
detulewe need mgross029 to come back with that stack trace to see if it's the same i saw in acl's omgb thread15:06
detulei should say if it's still the same15:07
emwerpierce99: sounds "logical" to me.15:10
*** turbolapin has quit IRC15:12
detuleemwe, here's a pastebin of a crash mgross029 sent me a long time ago v15:13
detulehttp://pastebin.com/0PfKHjbT15:13
rpierce99method called after release()15:13
detuleyeah i remember looking at it a long time ago but java is pretty confusing :)15:14
rpierce99why is it trying to notify of a call?15:15
detuleit was on a incomming phone call i believe15:15
mgross029Sorry rl kicked in here15:16
rpierce99i thought it crashed once at startup15:16
detulehere's the radio log to go with that logcat http://pastebin.com/a7Ajbppy ( i think )15:16
detulemgross029, hope you don't mind i am posting your old pastebins15:16
emwetimestamps don't match up :/15:17
mgross029detule, not at all...  Do we need updated?15:17
rpierce99yes, do you want a copy of getlogs so you can pull both at the same time?15:17
*** ray|yar has quit IRC15:17
*** ray|yar has joined #xdandroid15:17
emwei am building an image with abels patch15:17
mgross029I'm pulling lc and radio15:20
detulerpierce99, feel free to send GB getlogs my way15:21
rpierce99detule: uploading to dropbox right now, I haven't been able to test it on gb but i fixed the crash that arrrghhh was getting, so hopefully it works15:21
detuledo you have any "su -" commands in there?15:21
detuleit might not work with current emwe immages15:22
emwebecause?! :P15:22
detule(not sure about how superuser is being built currently)15:22
emweit's fine :P15:22
detuleok15:22
detule:)15:22
rpierce99it calls su to get root15:22
rpierce99http://dl.dropbox.com/u/50423/GetLogs.apk15:22
emwedetule: it was only one image in between which was broken.15:22
detuledon't worry i won't badmouth your images any more :)15:23
rpierce99if you get any errors just throw me a stack trace and i'll fix it up quick15:23
detuleif you could email me a copy this would facilitate install15:24
rpierce99if you could pm me your email it would facilitate facilitation15:24
emweworks fine on the RAPH with GB15:25
detulerpierce99, seems to be working fine15:27
rpierce99good, there was an issue with some of the filters where one of the filtered strings was returning null, i had to add some null checking15:28
rpierce99hope that data that is missing isn't causing phone.apk to crash15:28
*** helicopter88 is now known as helicAWAY15:29
emwefeedback from abel/acl. that bt patch has not solved it. he'll get back to me later when he is home. it was another patch he pushed towards omgb people.15:30
emwestopping the image build :P15:30
detule:)15:31
emwei checked our ISP and only crappy offers for faster internet.15:31
emwewe have a tv, phone and internet package. they figured it was a big of a loss for them. now they separated tv again. bastards.15:32
rpierce99as much crap as we get here in the US for having slow internet, at least it is cheap15:32
rpierce99well, mobile data is relatively cheap, landline data is still stupid expensive15:32
emwei just need more upload bandwidth. those 4kbit down are ok.15:32
emweerm, meant 4000kbit down.15:33
rpierce994k? are you serious right now?15:33
emwethose 256kbit down are a pita15:33
rpierce99ok that's better15:33
emweup15:33
emweman... bah. i should get a rest.15:33
rpierce99currently 12 down 2 up on my wireless network, while streaming netflix15:34
mgross029Home is 30 down and 5 up for me. :p15:36
emweshowoffs :P15:36
mgross029My logs have extra spacing in them.  Working to clean them up so emwe can read them easier. :p15:37
emwelol, yeah. and saves some of your 5mbps upload bandwidth.15:38
mgross029Nah I'm at work.  300Mbps Internet pipe.15:39
mgross029Of course that is shared with over 50,000 employees plus15:39
emwei am going to share my food with myself now.15:40
emwelaters.15:40
mgross029enjoy...15:41
mgross02915:25:58...  Look at that point in time  LC http://pastebin.com/zNCmBR4F Radio http://pastebin.com/a9yLqDh215:46
*** helicopter88 has joined #xdandroid15:46
rpierce99same stack trace, method called after release() in  tonegenerator15:47
*** helicAWAY has quit IRC15:48
mgross029Yep.15:52
*** SytheZN has joined #xdandroid15:56
*** SytheZN has quit IRC15:57
rpierce99so do you think that stack trace is saying that stopTone() shouldn't have been called, because the ToneGenerator has been released, or do you think something in the stopTone() native method called something that was released15:57
*** SytheZN has joined #xdandroid15:57
mgross029Well if we update this ./frameworks/base/telephony/java/com/android/internal/telephony/RILConstants.java:    int PREFERRED_NETWORK_MODE      = NETWORK_MODE_WCDMA_PREF; to = NETWORK_MODE_GLOBAL it should work for every phone.  Just my thoughts15:58
SytheZNGreetings .etc15:58
rpierce99wouldn't that be simpler to change the preferred mode in build.prop15:58
* SytheZN lights a smoke16:00
mgross029That would not be a proper fix though...16:01
rpierce99changing the framework is?16:01
mgross029If you wanted to do that you could always make the change in system.prop before you build the system16:01
mgross029Well if that is where the value is getting written from.  I would say yes.  Start at the origin of code and not at the result.  Right?16:02
mgross029Like I said just my thoughts.16:03
detulerpierce99, here's the release on radiotechnology change http://gitorious.org/xdandroid/packages_apps_phone/blobs/bc7ffda9dfa6b95f313acc203ecfee81e5a86f4d/src/com/android/phone/CallNotifier.java#line81516:03
*** SytheZN has quit IRC16:04
detulenot sure which instance afterwards gets called16:04
rpierce99i was just going to go look for that :)16:04
rpierce99and it happens on radio technology change too16:05
rpierce99that's excellent16:05
rpierce99the signalinfotoneplayer runs on it's own thread16:06
detuleyeah and that's the bad guy that calls stopTone16:08
detuleok i think just set mSignalInfoblahblah to null16:08
detuleafter release16:08
rpierce99seems logical, why would you release something and let it hang around16:09
mgross029detule, so why would changing the ro.telephony.default_network have an effect on the ToneGenerator?  Just curious16:10
rpierce99mgross029: i'm not saying we couldn't fix it that way, all i'm saying is there's no reason for the app to crash when switching modes16:10
detulemgross029, this business with the tone generator occurs on radio technology changes16:10
detulethese are affected by those build.prop settings16:10
mgross029I see.16:11
rpierce99if you set the preferred mode to cdma, it never has to switch tech16:11
rpierce99but in world phone mode, it looks like it starts gsm no matter what, and then if that doesn't work it switches to cdma, that tech switching isn't being handled properly16:11
*** SytheZN has joined #xdandroid16:12
mgross029Well if I use Auto Mode or Global it seems to make the decision on its own.  NETWORK_MODE_GLOBAL         = 7; /* GSM/WCDMA, CDMA, and EvDo (auto mode, according to PRL)16:12
mgross029Not saying this ToneGenerator code doesn't need to be fixed but why would we want to force all phones to WCDMA preferred?16:13
SytheZNMeh this network's driving me nuts16:13
rpierce99i don't think that's what we want, that is "use cdma if available (home), if i roam to europe and gsm is all i have, use that" i think16:13
rpierce99the setting we have now is if someone has their sim card in, assume they might be using their world phone as a gsm phone16:14
detulemgross029, try your hand at patching userland add right after line 815 a line "mSignalInfoToneGenerator=null;"16:14
rpierce99otherwise fall back to cdma16:14
detulethen rebuild the Phone.apk using "mm"16:14
rpierce99are his release keys the same as emwes16:14
detulehe's probably using his own system16:15
rpierce99ah i thought he was on emwes system16:15
mgross029rpierce99: but according to that code line comments it would depend on the PRL you are using.  So if you were on a US PRL it should use CDMA if a CDMA phone / GSM if  GSM phone and in Europe it would update the PRL as it connects to the towers16:17
mgross029AFAIK16:17
rpierce99i'm not going by code, i'm going by what is happening, when we are on 0, we see it create a gsmphone, fail, and then switch to cdma, which is where we get the crash, on 7 it apparently starts straight to cdma, which wouldn't work for those folks that want to use their rhod400 on a gsm network16:18
rpierce99i don't know, maybe it falls gracefully into GSM after it can't register properly on the cdma network16:19
mgross029detule, I will see if I can work on this over the weekend.  My VM Host went down and I can't get to my build system right now.16:19
rpierce99i know there was a reason hyc did it this way, but i'm not smart enough to know what it is16:19
mgross029rpierce99, I'm not saying I'm right just my thoughts of how the logic should work.  We would have to have someone with a Rhod400 on gsm, emwe, test that out to see if it works. :p16:21
mgross029If it goes by the PRL then it should put the phone into gsm mode when the radio attachs to the tower.16:22
SytheZNhow would I go about key remapping without rebuilding?16:26
detulerpierce99, this is almost funny -> compare the radiotechnology function in the GB CallNotifier I sent you, with the radiotechnology function in our current froyo tree http://gitorious.org/xdandroid/packages_apps_phone/blobs/3c4fdc50bc848f36d0d6a9fda442f7f501dde4a2/src/com/android/phone/CallNotifier.java#line71416:28
detulenotice currently in froyo we re-aquire mSignalInfoToneGenerator after we release it16:28
detuleand we have a much more robust updateonradiotechnology function16:28
detulemuch much more16:28
rpierce99it makes sense that we would need to reinstantiate the tone generator, but they could have moved that somewhere else16:29
rpierce99SytheZN: the keymaps are in the rootfs16:30
SytheZNrpierce: ta, but isn't rootfs read-only while running?16:32
detulei don't see where they moved it, they only create it once in the constructor.....i guess there could be multiple instances of CallNotifier16:32
mgross029It is unless you change ro.secure in default.prop16:32
rpierce99mount -o remount,rw /16:32
mgross029Or   ^^ :p16:33
*** SytheZN has quit IRC16:36
*** SytheZN has joined #xdandroid16:36
*** Kraln has quit IRC16:37
rpierce99VolumePanel.java is smart enough to null out its tone generators after releasing them16:37
SytheZNOi :'(16:37
*** Kraln has joined #xdandroid16:37
rpierce99did you see my answer SytheZN16:38
SytheZNYeah I did thanks rpierce9916:38
rpierce99what are you trying to do to your keymap16:38
SytheZNWell I'm trying to add in a handful of chars which are missing on the raph100 keymap to fn keys16:39
rpierce99is it something that has been done before so you can just steal the keymap?16:40
*** kalemas has left #xdandroid16:40
SytheZNNot as far as I know, but it may have been16:41
SytheZNKeys such as left & right angle brackets, curly braces etc16:42
SytheZN? Not root, no sudo... any thoughts?16:43
SytheZNlol nm16:44
* SytheZN kicks himself for asking before thinking16:45
* SytheZN repeats ^^ repeatedly16:46
emwedetule: rpierce99: seems you figured, eh? going to look.16:46
detuleemwe, i'll write you a patch really quickly :)16:46
SytheZNWhat're the chances of me getting dpkg or another package manager running on here?16:55
*** SytheZN has quit IRC17:06
emwemgross029: http://dl.dropbox.com/u/12183048/htctopaz/packages/Phone.apk one again17:10
mgross029emwe, thanks pulling down now17:12
detulei wonder if this somehow affected other issues with acoustic + phone.apk17:14
detuleprobably not17:14
emwehm, now does this crash or does this crash for me? ;)17:23
detule?crashes for you?17:23
rpierce99shouldn't affect non-cdma phones17:23
detuledid you add that if phone type switch17:23
emwenope, because google dropped it in gb17:25
emwenow it worked after it had stop/start hickups17:25
emwejust going with a clean boot17:25
mgross029Ok..  Pushed the Phone.apk to the phone17:27
mgross029Success!  Made a call to the phone and did not get a FC message17:28
emwedo it again, please :P17:29
emwewas this a fresh image?17:29
mgross029btw that was with my setting still at -1 on ro.telephony. :)17:29
mgross029Calling out now17:29
mgross029Prox works. :p17:29
mgross029No FC...17:29
mgross029No fresh image but it was the -117:29
emweok. assume it's fixed.17:30
mgross029I would yes.  I will do some further testing over the weekend.17:31
detulemgross029, just for fun try the volume rocker17:31
mgross029emwe not sure if you saw my message in htc-linux about testing your Rhod400 with 7 to see if you are able to get gsm in Europe.  Would be intersting to know if that works.17:32
mgross029detule, ok. sec.17:32
emwemgross029: that depends on my free time.17:32
emwenow peeps do your tests, wan't the patch off :P17:32
mgross029detule, no problems...17:33
mgross029emwe, understood.  Thanks...17:33
emwei currently do have the SIM in the raph so i can screw that install for acoustic testing...17:33
mgross029Ok I'm out guys...  Have a good one!  Good find btw!17:34
mgross029emwe, If you're bored. :p17:34
emwenever :)17:34
mgross029:)  Laters!17:35
*** mgross029 has quit IRC17:35
rpierce99at least now you can focus on acoustic emwe :)17:39
emwenow i only need to be succesful there...17:39
rpierce99sorry, i will be of much less help there17:39
detulei am starting to think my phone is a freak outlier with this acoustic business17:40
detuleit's a little depressing how much time i've spent looking at both the kernel and userland side17:40
emwehttps://gitorious.org/xdandroid/packages_apps_phone/commit/3db92dd32ea8f8090989e3e5e4e55207da2a971f i took the chance and credited acl as well17:42
detulemgross029 gets the nod for testing like a champ17:44
detulepunching out folks talk to you later17:44
rpierce99cya17:44
emweouch yeah, forgot mgross029...17:45
emwelaters.17:45
*** programmer8922 has joined #xdandroid17:55
*** programmer8922 has quit IRC18:02
*** helicopter88 has quit IRC18:04
*** AndersG has quit IRC18:06
*** ray|yar has quit IRC18:43
*** ilkinansr92 has joined #xdandroid18:53
*** ilkinansr92 has quit IRC18:55
*** turbolapin has joined #xdandroid19:07
*** d3tul3 has joined #xdandroid19:08
*** vinceweis has quit IRC19:08
*** dario_ has quit IRC19:13
turbolapinguys why when i download the source of .27 kernel and compile it, the size is different from the one i download from here: http://glemsom.users.anapnea.net/android2/htc-msm-linux/ ?19:24
emwedepends on the toolchain you use to compile. also if you strip modules or not, given the case you build them as well.19:30
turbolapinthanks. may i ask which toolchain you use? my sizes are 7mb for modules instead of 800kb and only a few files are different19:32
*** d3tul3 has quit IRC19:33
emweyou need to strip the modules19:36
emwei don't recall if there was a wiki page now with build instructions or not.19:37
emweusing 2010-09, still19:38
emwedon't try the 2011 one. won't work without patches.19:38
emwehere's a quite outdated tutorial: http://www.facebook.com/topic.php?uid=394317262428&topic=1401219:39
emwetoo lazy to look in xdandroid.com if arrrghhh setup an uptodate instructions page there.19:40
rpierce99the best build instructions were on ppcg, one of acls nand helpers put em up19:44
turbolapinthank you both19:52
turbolapinone more lazy question.19:53
turbolapinbefore initrd which is the script that runs?19:53
emwebefore initrd? nothing IMHO.19:55
emwerunning. gnight.19:55
turbolapini see that before the init script starts something else is runnning. probably the kernel is loading but which script in the kernel?19:57
turbolapinbtw i don't know if it makes any sense but i'm trying nand on raph10019:57
turbolapinsome more info: i'm getting kernel panic and reboot. so i need to check why19:58
rpierce99doesn't look like the ppcg thread talks about stripping modules19:59
*** emwe has quit IRC20:00
rpierce99by far the most common problem with the self-built kernels is the size limit, generally you have to be running the new haret with the updated initrd_offset to make it work20:00
turbolapinwrong offset may brick my phone?20:01
rpierce99no, it's an offset that gives the kernel enough room to fit in memory20:02
rpierce99just won't boot if it's wrong20:03
turbolapinis it this line?20:04
turbolapin.equ KERNEL_OFFSET,  0x800020:04
rpierce99no, it's in startup.txt20:04
rpierce99and it requires you to be running a compatible haret.exe20:04
turbolapini'm talking about nand :S20:04
rpierce99oh, well then forget it20:04
rpierce99i have no idea how nand layout works20:04
turbolapinok never mind. thank you for trying20:05
turbolapinbb all20:13
*** turbolapin has quit IRC20:16
*** ryannathansSLEEP is now known as ryannathans20:30
ryannathansmoaning21:36
ryannathansmorning*21:36
ryannathanshmm i wonder if repo is fixed21:37
rpierce99moaning is more accurate methinks21:39
rpierce99seeing as how it is after 8pm21:39
arrrghhhheh21:39
arrrghhhi doubt it's the same in australia21:39
rpierce99not because i'm doing stuff21:40
*** rpierce99 has quit IRC21:59
*** ringer has joined #xdandroid22:13
*** ringer has quit IRC22:23
*** ringer has joined #xdandroid22:23
Tandyman100arrrghhh: Here's some good news for you; I finally got a naitive Android device.22:33
Tandyman100Some more good news: As of Monday I might be able to donate a Rhod 210 to xdandroid for nand development.22:34
arrrghhhlol22:36
arrrghhhdon't tell me you got that G!22:36
arrrghhher22:36
arrrghhhg122:36
*** Himalayas has left #xdandroid22:38
*** ringer has quit IRC22:44
Tandyman100arrrghhh: Yes. I got a G1. But not the one you think22:49
Tandyman100It has a cracked screen, but it was $20.22:49
ryannathanso.022:49
Tandyman100$20 for a device with the same specs as my mom's $150 phone.22:49
Tandyman100Which I can root, and have installed Cyanogenmod 6 on.22:50
arrrghhhok22:53
arrrghhhso long as you realize it's pretty much dad22:53
arrrghhhdead*22:53
arrrghhhand CM6 is probably the best you're going to be able to run on it.22:53
arrrghhhvery limited on RAM that thing.22:53
Tandyman100yeah22:55
Tandyman100but it does what I need to22:55
Tandyman100just fine22:55
arrrghhhk22:56
arrrghhhalso, ACL might be getting a RHOD300 donor22:56
arrrghhhbut we'll see22:56
Tandyman100Ok23:00
*** elesbb has joined #xdandroid23:03
elesbbhey arrrghhh is it easier to build Android in a MAC or in Linux ?23:04
arrrghhhlinux23:04
elesbbokay , i got debian running in Virtual box .23:04
arrrghhhk23:05
elesbblol thanks , i wasnt sure which would be easier didnt mean to bother you .23:05
arrrghhhnp23:06
*** elesbb has quit IRC23:07
*** [7] has quit IRC23:07
*** TheSeven has joined #xdandroid23:07
*** bzo has joined #xdandroid23:40
*** virsys has joined #xdandroid23:52
arrrghhhbzo, how goes it23:59
bzohey arrrghhh not bad23:59
bzohaving an exciting friday night as you can see23:59
arrrghhhlol23:59
arrrghhhditto23:59

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!