IRC log for #htc-linux on 20100216

00:01.02cr2_Markinus: what is cc, 51, 0, 0, ff, ff
00:01.26Markinusohh, not tested yet . .I will do it . .
00:01.58*** join/#htc-linux ali1234 (~al@robotfuzz.co.uk)
00:02.04Markinuscr2_: sorry, but where do you have the 4 from?
00:02.12cr2_wince api
00:02.27Markinusahh, ok
00:02.42cr2_on leo
00:02.52cr2_so it is probably the same
00:03.06cr2_cc,51,val,4   -> extsystemLED behaviour
00:05.20Markinusok
00:07.02cr2_Markinus: it probably depends on the led mode array
00:07.25cr2_so you may get red too.
00:09.12*** join/#htc-linux JesusFreak316 (~JesusFrea@USF-Wifi-32-211.laptops.usf.edu)
00:09.28Markinuscr2_: 0 is nothing ..
00:10.07cr2_ok
00:10.26cr2_do the bigger numbers work ?
00:10.33Markinuscr2_: Dou you meen the 3 parameter in this array? The 0x0 ?
00:10.38MarkinusI will test it now
00:10.50Markinusthe led mode array?
00:10.59cr2_the 3rd parameter
00:13.36cr2_the vogue/kaiser clk code is an unbelievable mess :)
00:14.07cr2_if i didn't know what's going on there...
00:19.58Markinuscr2_: It seems that the higher values do nothing . .
00:20.19MarkinusHave to test the 3rd parameter for red led . .
00:22.13cr2_the tables are hardcoded into microP.dll
00:22.27cr2_so it may be impossible with the wrong rom
00:24.04Markinuscr2_: yes . .hmm. Dou you know whether I can replace this file under winmo? With a older one? Without flashing?
00:25.03cr2_if it's a normal file
00:25.10cr2_then it's readable.
00:25.27cr2_don't know if you can just remove the "normal" file from rom
00:27.01cr2_hm.
00:27.04cr2_http://gitorious.org/kaiser-linux/kernel/blobs/51149f40228b3cdd48ff803ce7af98e0ba382415/arch/arm/mach-msm/board-kaiser-battery.c
00:27.16cr2_a lot of hardcoded things there.
00:29.25*** join/#htc-linux jumoit (~samuel@121.77.87.145)
00:29.45cr2_but it looks good
00:29.46*** join/#htc-linux m3dlg (~m3dlg@212.183.140.34)
00:32.19*** join/#htc-linux M1DLGpc_ (~M1DLGpc@bb-87-81-252-83.ukonline.co.uk)
00:36.52*** join/#htc-linux m3dlg (~m3dlg@bb-87-81-252-83.ukonline.co.uk)
00:38.49*** join/#htc-linux Bry8Star__ (~Bry8Star@cpe-76-170-247-232.socal.res.rr.com)
00:39.07*** join/#htc-linux Zeman4323 (~Zeman4323@c-71-239-97-164.hsd1.il.comcast.net)
00:40.25*** join/#htc-linux m3dlg (~m3dlg@bb-87-81-252-83.ukonline.co.uk)
00:41.03Markinuscr2_:We have on Topaz a second led on the right side on the speaker. But I have no Idea how to activate this .. . I think this was blue but I'm not sure . .
00:41.23*** join/#htc-linux Unholy_ (~IceChat7@24.227.27.77)
00:41.47*** join/#htc-linux ccube (~ccube@95-90-193-11-dynip.superkabel.de)
00:44.28cr2_Markinus: BT ?
00:45.24Markinusyes, maybe, but not in the 6.5 Roms . . .and I cannot remember what was with 6.q I posted i the Forum, maybe can someone give me a 6.1 dll.
00:45.54*** join/#htc-linux myn\ (cereal@cpe-71-64-8-167.insight.res.rr.com)
00:58.01cr2_good night
01:19.12*** join/#htc-linux JesusFreak316 (~JesusFrea@USF-Wifi-32-211.laptops.usf.edu)
01:26.17*** join/#htc-linux JesusFreak316 (~JesusFrea@USF-Wifi-32-211.laptops.usf.edu)
01:31.04*** join/#htc-linux stinebd (~bryan@gentoo/developer/battousai)
01:39.49mynwerd
01:40.01mynso what are the Sprint phones that work with Android
01:41.02*** join/#htc-linux makefile (~god@189.81.69.129)
02:05.05*** join/#htc-linux balsat (~balsat@87.72.13.34)
02:08.06makkonenmyn: diamond and touch pro... 'work', I guess.
02:08.27makkonenno bt, no gps, 4 hour battery life, weird crashes... but they work.
02:08.46*** join/#htc-linux adhoc (~khawtin@ppp121-45-24-62.lns20.adl2.internode.on.net)
02:11.35maejrepmm, trying to hack LWP to work in a mangled HTC ROM ;/  http://xda.privatepaste.com/download/c56d310c94
02:11.40maejreppreview works at least
02:11.43maejrepnothing else
02:13.54adhocmorning all
02:17.54*** join/#htc-linux Bry8Star (~Bry8Star@cpe-76-170-247-232.socal.res.rr.com)
02:23.14maejrepi take that back.. it does set: http://xda.privatepaste.com/download/11a7424b9f
02:25.53adhocanyone installed Xanadux or similar onto a HTC Hermes ?
02:30.00*** join/#htc-linux marex (~marex@vasut.kolej.mff.cuni.cz)
02:35.09*** join/#htc-linux enjalot (~enjalot@wg-d232102.dsl.fsu.edu)
03:04.46*** part/#htc-linux surgex (surge@pool-98-118-154-21.bflony.fios.verizon.net)
03:21.49*** join/#htc-linux god206 (~infidel20@unaffiliated/swc666/x-4934821)
03:24.38*** join/#htc-linux yepper12345 (~yepper123@99-25-243-248.lightspeed.ltrkar.sbcglobal.net)
03:29.44*** join/#htc-linux fleixius (~fleixius@unaffiliated/fleixius)
03:41.01stinebdmakkonen: did you need me for something last night?
03:43.06stinebdwell last night for me, heh... 29 hours or so ago
03:43.19*** join/#htc-linux JesusFreak316 (~JesusFrea@pool-173-65-72-26.tampfl.fios.verizon.net)
04:07.22*** join/#htc-linux randomblame (~kevin@c-71-59-184-96.hsd1.wa.comcast.net)
04:35.50*** join/#htc-linux madCoder- (~madcoder@c-68-42-30-56.hsd1.pa.comcast.net)
04:45.57makkonenstinebd: I'm in nearly the same time zone as you, I think, so it was last night for me, too.
04:48.34makkonenI wanted to know if the capslock key on your device turned on the capslock LED. Someone else said it did on their fuze, so I guess I don't need you anymore.
04:49.07makkonenI also wanted to know if you were giving people commit access to your rootfs repo on gitorious. Before I go and make my own clone.
04:49.52stinebdif you want access i can give it
04:49.58stinebdi don't know of anyone that uses the repo
04:50.14stinebd(other than me)
04:54.07*** join/#htc-linux jamenlang (~jamenlang@68.170.60.12)
04:55.15stinebdmakkonen: added
04:57.40makkonenwell, I guess it's easier to put my fixes in there and hope they filter upstream than to hope I catch someone who puts builds together.
05:00.42stinebdi like having it there to see the changes :)
05:07.46*** join/#htc-linux droid0011 (~g1@p4FDCBA68.dip.t-dialin.net)
05:09.06*** join/#htc-linux goxboxlive (~jrs@mail2.hjellnesconsult.no)
05:14.44*** join/#htc-linux phh_ (~quassel@2a01:e35:2e4b:b2b0:250:8dff:fee1:c793)
05:14.53makkonenstinebd: do you know how git takes care of file permissions? I imagine it'd be a problem for something like the rootfs.
05:15.33stinebdmakkonen: see the scripts dir
05:15.46stinebd(its a hack)
05:18.57makkonenyou added me as a committer to the xdandroid-vendor repo, not eclair-rootfs.
05:20.02stinebdoops
05:20.57stinebdfixed
05:20.58stinebdsorry
05:22.44stinebdanyway there's a script in there that uses a perms map to replicate the original rootfs permissions. unfortunately the map has to be updated whenever a new file is added
05:22.57makkonenah.
05:23.14makkonenyeah, that's hacky, but... hey, better than just handing around a rootfs of unknown provenance whenever a change has to be made.
05:23.21makkonenwhich is pretty much how it's done now.
05:24.31makkonenI'm gonna drop a few fixes in there tonight... I don't think mrpippy's combined fix for wifi that works on both cdma and gsm has made it in anywhere yet. Just have to find it in the logs, now.
05:25.04stinebddistprepare.py is the script that adjust perms based on perms.map
05:25.28makkonenand there's a bunch of annoying minor bugs in the init script... like things that point to /sdcard/android instead of $card
05:25.48stinebdyes the init script needs cleaning badly
05:29.13*** join/#htc-linux bzo (~chatzilla@adsl-75-35-5-163.dsl.pltn13.sbcglobal.net)
05:32.57makkonenand I can drop binaries in the /bin folder and it won't complain? Just have to also add their perms to the perms.map file?
05:33.47makkonen(I've got dropbear working pretty well at this point, so I might as well add my changes for that as well)
05:33.52stinebdyup
05:33.55stinebdoh nice
05:33.59bzostinebd: do you have any plans to create an initrd repo?
05:34.11stinebdbzo: not yet
05:34.26stinebdi don't see why i couldn't though
05:34.29bzoi've been looking into bringing over vilord's new stuff from the vogue
05:35.04bzohe's released all his code now
05:35.24makkonenI want that replimenu junk.
05:35.43bzoeven made changes to make diamond/raph keypad compatible with replimenu :)
05:37.16makkonennice
05:37.28makkonenI have never even looked at the initrd
05:37.53makkonenI feel like it's a defeatist attitude, but... I would kinda be ok with keeping it that way. :-p
05:38.20bzonothing too terribly complicated in there
05:38.27makkonenthat's good.
05:38.34makkonenencouraging. maybe I will take a look.
05:40.01bzoanyways, since vilord has already fixed the diam/raph compatibility, there's only a bit of script tweaking to bring in the menu stuff
05:40.20bzoquestion is how best to deal with the other devices
05:44.02bzomakkonen: working on other stuff beyond the rootfs fixes?
05:44.19makkonenright now? no.
05:45.10makkonenthinking I might install new ROM in winmo tonight... and if I'm going to do that, I should test to see if the new radio version has an effect on android.
05:45.58bzoaren't the radio roms usually separate updates?
05:46.31makkonenyeah, but I'm just doing the big stock sprint update, and I think it's got both.
05:47.23bzoyeah, think you're right
06:00.58*** join/#htc-linux DaveWM (~dave@h-174-39-73-168.ip.alltel.net)
06:05.10makkonenanyone here know what all the values in the sqlite table for APNs are. I'm sort of flummoxed by the bit where it creates the android APN in the rootfs... it's got 16 entries, and the table wants 17, but I'm not really sure what they are.
06:06.37*** join/#htc-linux bzo_ (~chatzilla@netblock-68-183-234-105.dslextreme.com)
06:06.40makkonenclearly whatever it was I did a couple nights ago was not right, because now I have 6 android APNs when I go to check. :-(
06:09.47MrPippyif you do select * from carriers, will that show the column headings?
06:12.32makkonen...that's weird.
06:13.07makkonenno, it doesn't, I thought I'd tried that before, and just now, I confirm it doesn't.
06:14.06makkonenbut what's weird is that my carriers table has 57 entries or so... when I hit 'reset to default' it ends up with 197
06:14.58tmztadd something
06:15.13tmzttyping /help should tell you what
06:15.31tmztI think it populates from xml
06:15.36tmztor .sql possibly
06:16.19makkonenit populates from apns-conf.xml, I assume. and there's one of those in the rootfs, and another in the system.sqsh, I think, which is at the root of it. (haha, wordplay.)
06:16.46makkonentyping /help where?
06:17.10makkonenoh yeah, adding APNs causes the system to reboot.
06:20.59tmztin the sqlite prompt
06:21.33*** join/#htc-linux bzo (~chatzilla@netblock-68-183-234-105.dslextreme.com)
06:21.36tmztit might be .help or \help
06:21.42tmztor just \h
06:23.57*** join/#htc-linux DaveWM (dave@h-96-13-67-135.ip.alltel.net)
06:24.44bzoi think you want .schema
06:25.36MrPippyyeah thanks that does it
06:27.36bzoi like sqlite, but it can be a little arcane
06:29.31MrPippymy data partition is pretty old (goes back to beginning of january), but my internet has always worked and i never had to manually add an apn
06:29.50MrPippythe android one is in the db, probably added from apns-conf
06:30.12makkonenyeah. I think that was the case with mine as well... it got added to the db before the apns-conf lost it.
06:31.04MrPippyapns-conf doesn't have the android apn in it anymore?
06:31.44makkonenI'm not sure.
06:32.00makkonenthe version in the rootfs does, but the version in system.sqsh doesn't, I think.
06:32.16makkonenand then there's a command in the rootfs to add it, but it's borked.
06:32.39makkonenI don't know why there needs to be a command to add it to the database if it's already in the list, though.
06:33.37MrPippyyeah depends where android is looking for the file...maybe we need to do something crazy so it finds the one in /etc from rootfs
06:35.08MrPippyand its good to see a git repo for the rootfs, as long as the xdanroid guys are making releases from it
06:36.04makkonenyeah. let's hope that happens. Stinebd, I assign you the task of proselytizing.
06:37.25bzoso why is the sqlite db created dynamically anyway, instead of starting with a pre populated db file?
06:37.44makkonenthat is one of a number of good questions.
06:39.17*** join/#htc-linux Vito89 (~quassel@gw.loccal.net)
06:44.31bzomakkonen, you mentioned getting only 4 hours battery life. Is this true for you too MrPippy?
06:44.52MrPippyi haven't done any battery tests
06:45.14MrPippywouldn't surprise me though, android makes the phone warm in a way that i almost never see in winmo
06:45.23*** join/#htc-linux zachy (~george@ip-85-160-13-59.eurotel.cz)
06:45.26MrPippyprobably the constant high cpu load
06:46.06bzoit feels like it has gotten worse recently though. i think it coincided with when I started using kernels from the main tree
06:47.34bzoI'm not even sure if suspend is consistently working due the loss of the led effects
06:48.09makkonenI tried to toggle my data off and put my backlight down and see how long it would go today... but then I forgot, and turned data back on so I could find out who Clinton's Secretary of Defense was (Bill Cohen, if you were wondering)... I got about 3.5 hours until it was reading 5%, and then I put it on the charger.
06:48.14*** join/#htc-linux pini (~pini@bou-fi.pustule.org)
06:48.53bzohaha, of course it's hard to say how long the battery will really last with the state of the battrey mete
06:49.04makkonentrue
06:49.33makkonenI don't know if it was really that close to dying. the other night I ran it down all the way to where it shut off, and it couldn't've been much more than 4.5 hours, though.
06:50.59bzoI'm pretty sure though that I used to be able to leave the phone unplugged overnight without completely discharging. Definately can't do that anymore
06:52.08bzoin any case, I'm sure there's lots of room for improvement. I used to use my vogue all day long in android.
06:54.53*** join/#htc-linux kevincarafa (~karafa@ool-45735dce.dyn.optonline.net)
06:59.12*** part/#htc-linux kevincarafa (~karafa@ool-45735dce.dyn.optonline.net)
07:01.01makkonenyeah. it's definitely not all-day usable for me. I'm surprised by how little the drain seems to change from using it intently and letting it sit on the counter.
07:05.09*** join/#htc-linux karafa (~karafa@ool-45735dce.dyn.optonline.net)
07:05.39karafaI'm curious, why is there a constant high CPU load?
07:16.45makkonenI don't know why. In part I guess it comes down to the constant churn of accessing the SD card for everything. There's probably other things that are being done inefficiently as well.
07:18.00DaveWMwell this is sounding less and less appealing to try out
07:18.01DaveWMheh
07:18.47makkonenit ain't perfect.
07:18.59makkonenbut it is really cool.
07:19.30DaveWMyeah,  i probably still will,  wish more of the hardware was worked out for my touch pro
07:19.44bzohmm, I think there is something broken with suspend on cdma.
07:19.46tmztcdma?
07:19.54tmztcdma/gsm?
07:20.09bzoI'm looking at the logs, and chamonix was reporting 12+ hours in his battery tests
07:20.09DaveWMmines cdma
07:21.43karafathe flaws in android in it's current state are not as bad as the flaws in windows mobile, imo.
07:22.02makkonenbzo: it's seemed weird for a while, but I'm not sure. it was definitely working at some point, and I was still getting 4 hours. if you put pm.debug_mask=2 in your cmdline, you'll notice it waking up like crazy (the little buzzes while it's in suspend mode)
07:22.05bzotmzt: yes
07:22.15karafai haven't used winmo for more then booting android in about a month
07:22.22DaveWMyeah,  being my first mobile experience,  i honestly expected windows mobile to be a little further along than it is
07:22.23DaveWMheh
07:23.27makkonentmzt: cdma. raph800/diam500. not topa/rhod. Not really sure of the state of suspend on them.
07:24.48DaveWMkarafa: the battery life is enough for you that way ?
07:25.02DaveWMI mean i suppose i'm near a charging source most of the time
07:25.06makkonenjust get a lot of chargers.
07:26.01karafayeah, for me personally, it is. i know for a lot of people it might not be, but i always have at least my laptop and a usb cable for it, if not a wall charger and car charger too
07:26.03DaveWMneed some solar panels
07:26.07*** join/#htc-linux BabelO (~fcr@2a01:e35:2ee1:c1c0:21e:8cff:fe2d:7d3f)
07:26.08*** join/#htc-linux BabelO (~fcr@unaffiliated/babelo)
07:26.32karafahonestly, though, with the WM roms i've used, i never got more then like 6 hours in WM anyway, it doesn't feel that different.
07:26.45tmztwhich one?
07:27.36DaveWMi guess it depends on what you're doing with your device,  huge difference in power consumption based on that
07:27.44makkonenyeah, I get more than 6 hours... but could never really go a whole day off the charger in winmo, so it's not a huge change.
07:28.24DaveWMi know i've had this touch pro go over a day,  but i barely used it then,  when i'm messing with it all day,  it still seems to last me about 10 hrs at least
07:28.38karafai'm using lunicierom right now, only ever tried that and the stock rom, now that i think of it.
07:28.47DaveWMjust surfing the web on it seems to be a pretty good drain
07:29.08DaveWMi use it for pandora radio sometimes,  that eats the battery down alot quicker,  probably would get less than 6hours out of it then
07:29.39karafawhen I did use it, i constantly shared the internet over wifi, so I know that drained it.
07:29.54DaveWMyeah,  i imagine that would
07:30.29DaveWMif you shared your wifi connection over bluetooth... that'd probably really do it
07:30.42DaveWMpretty much have every bit of technology in the device churning at one time
07:30.44DaveWMheh
07:31.03DaveWMmaybe throw in a script to broadcast your GPS over the internet every second
07:31.21makkonenturning on wifi was enough to bring life down to the 6-8 hour range for me.
07:31.42karafahey, that reminds me, i was wondering about the android-wifi-tether package. it's unofficial and not on the market, but available to any rooted phone. when i try it on this, it gives a casual error about not having 'netfilter' features of the kernel available. specifically, CONFIG_NETFILTER and CONFIG_IP_NF_IPTABLES. not exactly sure what that is, is that something that might be added into this at some point?
07:32.38makkonenhmm. you're on raph800? try the mainline kernel from glemsom's site.
07:32.42DaveWMi was thinking of looking into an extended battery for mine,  only one i found so far is made for the GSM version,  and seems to have a slightly different back,  claims it won't work on the alltel one
07:33.09makkonenI think that one's got netfilter enabled (but I might be wrong). Pretty sure I have it turned off in mine.
07:34.06DaveWMis android a derivative of any other desktop linux ?
07:34.18makkonenno
07:34.23DaveWMi mean its not like debian or red hat based is it ?
07:34.28makkonennope
07:34.59DaveWMi'd like debian on a smart phone
07:35.15DaveWMproviding my hardware all worked
07:35.33makkonenit's been done.
07:35.35karafaoh wow, I didn't think the autobuilds worked on the cdma devices
07:35.51DaveWMyeah,  i figured,  i don't think i'd get it going too good on mine though
07:36.00makkonenyeah, as of a little while ago, all the necessary changes are in.
07:36.01DaveWMi may look into that at some point just to toy with though
07:36.48makkonenbasically the only changes left in my clone of the repo is one little tweak to the keyboard LEDs and a bunch of changes to the build config.
07:37.09makkonenwhich means it's about time to get rid of my clone, as it is no longer necessary.
07:40.04bzomakkonen; i get no buzzes with pm.debug_mode=2
07:40.24makkonenhuh
07:40.35makkonenare you hitting End to go to sleep?
07:40.49makkonen(not sure it makes any difference, but it's a thought)
07:40.53bzoyep, though you shouldn' need to
07:41.09bzoseems like it is waking up too fast to really be suspended
07:41.14makkonenhmm
07:41.32bzohave you tried turning that on recently?
07:41.39makkonenoh... I might just have that wrong.
07:41.43karafai've found that turning off the location / GPS services (even though they don't  work properly) adds a few hours onto battery life, though i'm positive
07:42.10makkonenpm.debug_mask=2
07:42.54bzoerr, I mistyped earlier, pm.debug_mask=2 is what I have
07:42.58makkonenbut I have not tried it recently.
07:42.59*** join/#htc-linux balans (~Gebruiker@212-123-149-239.ip.telfort.nl)
07:43.38makkonenlet me start up mine right now and see if anything happens.
07:43.40*** join/#htc-linux kiozen (~oeichler@p54920F14.dip0.t-ipconnect.de)
07:44.33makkonenbooting
07:48.23makkonenno vibrating.
07:48.38bzokarafa: I'll have to give that a try, noticed my gps was enabled
07:50.54bzomakkonen: seems like dmesg is showing immediate wakeup whenever suspend is started
07:52.43makkonenused to be that on suspend the dmesg would just stop.
07:52.55makkonenI had to add 'no_console_suspend' to my cmdline
07:53.03makkonendefinitely hasn't done that in a while.
07:54.37bzowell, I'm going to turn in for the night
07:54.43bzomaybe phh will have some ideas tomorrow
07:55.00makkonenhe told me a number of weeks ago how to test what was waking up from suspend.
07:55.14makkonennot that that's important now, but it's something else we need to look into.
07:59.56*** join/#htc-linux mickey|zzZZzz (~mickey@openmoko/coreteam/mickey)
08:03.24*** join/#htc-linux tehtrk_home_ (~quassel@adsl-68-91-144-191.dsl.wacotx.swbell.net)
08:29.06randomblameanyone here manage to compile haret?
08:29.32randomblameI got it modified and compiled fine but it says it's either not signed with a trusted certificate or missing a component
08:29.53randomblameanyone know a way to run unsigned exes
08:33.23*** join/#htc-linux kvaster (~kvaster@93.84.112.80)
08:34.21NetRipperrandomblame, try installing this: http://www.netripper.com/raphael/Default_Signed_DisableSecurity.CAB
08:34.59NetRipperand then try again
08:36.05randomblamethanks
08:39.53randomblamethat didn't work
08:40.13randomblameis there somthing I'm supposed to do with it after I run the cab?
08:56.35*** join/#htc-linux kiozen (~oeichler@p54920F14.dip0.t-ipconnect.de)
09:07.00*** part/#htc-linux phh_ (~quassel@2a01:e35:2e4b:b2b0:250:8dff:fee1:c793)
09:07.07*** join/#htc-linux phh (~quassel@2a01:e35:2e4b:b2b0:250:8dff:fee1:c793)
09:41.54*** join/#htc-linux GNUtoo (~GNUtoo@host151-131-dynamic.244-95-r.retail.telecomitalia.it)
09:47.02*** join/#htc-linux jumoit1 (~samuel@121.77.87.145)
09:57.50*** join/#htc-linux nitroboost (j@nitrology.com)
09:58.10*** part/#htc-linux nitroboost (j@nitrology.com)
09:58.30*** join/#htc-linux marex (~marex@vasut.kolej.mff.cuni.cz)
09:58.58*** join/#htc-linux nitroboost (j@nitrology.com)
10:04.30*** join/#htc-linux phh (~quassel@2a01:e35:2e4b:b2b0:250:8dff:fee1:c793)
10:09.33*** join/#htc-linux val_ (~val@lgx22-1-88-168-244-3.fbx.proxad.net)
10:23.03*** join/#htc-linux marex (~marex@89-24-6-50.i4g.tmcz.cz)
10:27.43*** join/#htc-linux jumoit (~samuel@121.77.87.145)
10:33.32*** join/#htc-linux Vito89 (~quassel@195.113.242.145)
10:34.03*** join/#htc-linux leobaillard (~leobailla@leobaillard.org)
10:39.25*** join/#htc-linux slight (~23hdci278@204.112.221.87.dynamic.jazztel.es)
10:40.47mickeyldoes the qmi stuff require a certain AMS version or could it be backported from the dream to the windows-mobile-HTCs?
10:41.35mickeylit's so much more convenient than ppp...
10:41.36mickeyl(and efficient)
10:47.58*** join/#htc-linux kvaster (~kvaster@93.84.112.80)
10:52.25leobaillardhi there !
10:56.24*** join/#htc-linux JairunCaloth (~m00@c-24-98-17-158.hsd1.ga.comcast.net)
10:58.29*** join/#htc-linux balans (~Gebruiker@212-123-149-239.ip.telfort.nl)
10:59.52*** join/#htc-linux Olipro (~Olipro@uncyclopedia/pdpc.21for7.olipro)
11:06.48niall0sError occurred during initialization of VM
11:06.48niall0sCould not reserve enough space for object heap
11:06.48niall0sCould not create the Java virtual machine.
11:06.52niall0soops
11:06.54niall0ssry
11:14.36*** join/#htc-linux cr2_ (~cr2@ip-109-84-159-64.web.vodafone.de)
11:15.46*** join/#htc-linux goxboxlive (~jrs@mail2.hjellnesconsult.no)
11:16.09cr2_hi goxboxlive
11:18.29*** join/#htc-linux MethoS- (~clemens@134.102.106.250)
11:18.59goxboxlivehi cr2_
11:19.50goxboxlivei am missing the worlds best beer, Augustiner. here in norway it's not for sale. you can drink every day :-)
11:20.05cr2_lol
11:20.31cr2_but we have weather like in Norway now
11:21.24phhcr2_: someone sent me the sources of raph's wince's gps driver, I think you don't want it right ? (anyway what we lack is to reset DEX + RPC after modem reset)
11:22.36cr2_reset dex ?
11:22.52phhonce arm9 has resetted, DEX commands just time out
11:22.54phhlike rpc
11:23.03cr2_reset rpc is not done properly, that's true
11:23.29cr2_we don't get several messages from arm9 to out rpc cb servers.
11:23.50cr2_hm. how do you want to reset dex ?
11:24.00phhno idea.
11:24.37cr2_i've added the dump clocks to haret
11:24.47cr2_but it does not work for some reason.
11:25.02cr2_i'll try it on raph
11:25.46cr2_isn't there a reset rpc bit in smsm[] ?
11:26.03phhyes
11:26.29phhI still don't know who (A9/A11) sets which field though
11:26.52phh#define SMSM_INIT          0x000001
11:26.52phh#define SMSM_SMDINIT       0x000008
11:26.52phh#define SMSM_RPCINIT       0x000020
11:26.52phh#define SMSM_RESET         0x000040
11:27.07cr2_they are present in both
11:27.36phhyes, but I think we have to clear it, notify A9, then set it back
11:27.38phhor something like that
11:28.16phhwell, you should see how wince does it, because guessing isn't easy.
11:28.46cr2_yes
11:29.00cr2_the topaz has some debug messages about it
11:29.15cr2_maybe rhod too, i need to check
11:32.05phhcr2_: in codeaurora's tree there is something about modem reset, but I think it's not for our amss
11:32.05*** join/#htc-linux irlolcopter (~irlolcopt@host81-152-193-147.range81-152.btcentralplus.com)
11:32.21phh<PROTECTED>
11:32.21phh<PROTECTED>
11:32.21phh<PROTECTED>
11:32.21phh<PROTECTED>
11:32.21phh<PROTECTED>
11:32.22phh<PROTECTED>
11:32.36goxboxlivecr2_: realy. Here is to much snow, and it's to cold. Not bad today, but we have had some extraordinary cold days in dec/jan.
11:32.46goxboxlivecr2_: what kind of device are you developing for now?
11:33.22phhgoxboxlive: raph/diam/blac/topa/rhod/leo
11:33.51cr2_goxboxlive: nothing extraordinary here too, but this winter is colder as usual
11:34.23cr2_goxboxlive: the phones don't differ very much now. and we have similar problems for all of them
11:34.32goxboxliveAnd th eproblem is?
11:34.44cr2_gps for examle
11:35.04cr2_because it's not a separate chip, but something deep inside the phone
11:35.41goxboxliveohhh
11:35.53goxboxliveWhat gui are we using these days?
11:37.12cr2_i don't use any
11:37.25goxboxliveok i c
11:41.00*** join/#htc-linux mcdull (~blackston@061093132071.ctinets.com)
11:45.46*** join/#htc-linux sxe (~quassel@ip-62-143-102-161.unitymediagroup.de)
11:51.16phhcr2_: have you got such logs ?
11:51.29*** join/#htc-linux junalmeida (~junior@201.19.226.123)
11:52.12cr2_yes, let me check
11:52.29cr2_phh: i think they are all in the topa thread
11:53.44cr2_phh: here:
11:53.48cr2_SMD DLL(0) attached to a process
11:53.49cr2_Line: 745. rpc_router_os_wm: DLL_PROCESS_ATTACH!
11:53.51cr2_12:48:03 [K] OEMInterruptEnable: SYSINTR_FIRMWARE+55
11:53.53cr2_[K][INT-En]SYSINTR_55
11:53.54cr2_12:48:03 [K] OEMInterruptEnable: SYSINTR_FIRMWARE+50
11:53.55cr2_SMD task(0) started.
11:53.57cr2_smsm_cb: 0 1 9
11:54.26cr2_are these A2M or M2A irqs ?
11:54.47*** join/#htc-linux Olipro (~Olipro@uncyclopedia/pdpc.21for7.olipro)
11:54.55cr2_because then you have
11:55.11cr2_Line: 1284. rpc_router_os_wm: Calling RPC_Init()!
11:55.18cr2_Line: 1076. rpc_router_os_wm: IOCTL code  = 30010014
11:55.20mickeylgoxboxlive: most folks use android, but some have just starting to integrate SHR and its apps
11:55.28cr2_smsm_cb: 0 9 29
11:55.30cr2_smsm_cb: 0 29 29
11:55.31cr2_smsm_cb: 1 80000000 10029
11:55.33cr2_[K][INT-Done]SYSINTR_55
11:55.42cr2_Line: 781. rpc_router_os_wm: Calling RPC_Open()!
11:55.50cr2_Line: 1076. rpc_router_os_wm: IOCTL code  = 30010008
11:56.00cr2_phh: this is full reset+rpc init
11:56.11cr2_maybe i'd document the ioctls
11:56.54cr2_phh: the smsm_cb: line is arm9/arm11 and state for both
11:57.16phh"state for both" ?
11:57.29cr2_arm9state arm11state
11:57.38cr2_hmm
11:57.39phhyes but the third int ?
11:57.43cr2_need to check that too
11:57.54cr2_maybe old_state new_state
11:58.13phh#define SMSM_UNKNOWN           0x80000000
11:58.18phhweird
11:58.27cr2_was not stored
11:58.32cr2_if it's "old"
11:58.53cr2_let's check these 2 irqs first
12:01.58cr2_INT_A9_M2A_0  SMD
12:02.06cr2_INT_A9_M2A_5  + SMSM
12:02.43cr2_phh: these are OEMInterruptEnable: SYSINTR_FIRMWARE+50 and OEMInterruptEnable: SYSINTR_FIRMWARE+55
12:02.56phhok
12:03.09cr2_so the smsm_cb is the INT_A9_M2A_5 irq handler
12:03.22cr2_aka [K][INT-Done]SYSINTR_55
12:03.43cr2_we are missing INT_A9_M2A_6  + DEX
12:03.54cr2_but it's probably a separate issue
12:04.21cr2_but acking this irq may be a good idea too
12:05.40phhso m2a_5 is handled in smsm_irq_handler
12:05.50*** join/#htc-linux Olipro (~Olipro@uncyclopedia/pdpc.21for7.olipro)
12:06.02cr2_yes, it's a smsm_irq :)
12:06.16phhand there are some stuff about smd (re) init
12:06.50phhbut it's only if state changed, and I'm not sure what I should do in handle_modem_crash then
12:08.54cr2_smsm_cb: 0 9 29
12:09.00cr2_what is 0x20 here ?
12:09.07*** join/#htc-linux Olipro_ (~Olipro@uncyclopedia/pdpc.21for7.olipro)
12:09.09phhSMSM_RPCINIT
12:09.19cr2_<phh> #define SMSM_RPCINIT       0x000020
12:09.20cr2_ok
12:09.25*** join/#htc-linux leaigor (~laigor@188.134.16.241)
12:09.45cr2_so you put it into arm9 status ?
12:10.04cr2_but it's an M2A irq. hmm
12:10.46cr2_let me grep the logs
12:10.47phh<PROTECTED>
12:10.47phh<PROTECTED>
12:10.47phh<PROTECTED>
12:10.47phh<PROTECTED>
12:10.51phhthat's what I try to do.
12:12.07phhbut that's a total guess, I have no clue of what I'm actually supposed to do at all
12:12.55cr2_hm. nothing
12:13.10phhI should do nothing ? that doesn't work :p
12:13.58cr2_no, nothing in the logs
12:14.01cr2_strange
12:14.23phhmaybe I should try doing do_smd_probe without changing the states ?
12:17.24IceBonehttp://notalwaysright.com/booze-is-blamed-when-there-is-no-proof/4258
12:17.35cr2_omg, it works :)
12:18.03cr2_HaRET(4)# pd 0xa00000 0x100
12:18.05cr2_00a00000 |EXCEPTION while reading from address 0x1990000
12:18.17cr2_HaRET(6)# pfw 0xA8250800 0x1 0
12:18.28phhcr2_: disabling MPU on leo ?
12:18.30cr2_HaRET(8)# pd 0xa00000 0x100
12:18.32cr2_00a00000 | f001c000 e0008000 f4020000 0000000b | ................
12:18.33cr2_00a00010 | 0000f800 f001d3dc e08dd000 00a00000 | ................
12:18.38cr2_not on leo, on raph
12:19.07phhhum what works ?
12:20.10cr2_try it
12:20.22cr2_will not work on leo though
12:20.32*** join/#htc-linux marex (~marex@89-24-6-33.i4g.tmcz.cz)
12:20.33cr2_HaRET(10)# pwf amss_A 0xa00000 0x2000000-0xa00000
12:20.34cr2_HaRET(11)#
12:20.49cr2_let me check oemsbl
12:21.07cr2_yeah
12:21.11cr2_HaRET(11)# pd 0x900000 0x100
12:21.12cr2_00900000 | 000003e0 aa0f0000 01f6e161 00000000 | ........a.......
12:21.14cr2_00900010 | feed26b4 009326d8 00933454 009343d0 | .&...&..T4...C..
12:21.15cr2_00900020 | 4d454d50 00933abc 009325b8 00933b9c | PMEM.:...%...;..
12:21.17cr2_00900030 | 00933bac 34333231 38373635 32313039 | .;..123456789012
12:21.18cr2_00900040 | 5f435448 544f4f42 2e315620 322e3230 | HTC_BOOT V1.02.2
12:21.31cr2_but this hole is closed on leo :(
12:22.07cr2_HaRET(12)# pwf raph_oem 0x900000 0x100000
12:22.08cr2_HaRET(13)#
12:23.37cr2_on titan they didn't even protect the amss/oemsbl from reading
12:24.04phhwell, I won't call "setting a bit writable by any one" a protection either :p
12:25.07phh<PROTECTED>
12:25.07phh<PROTECTED>
12:25.07phh<PROTECTED>
12:25.07phh<PROTECTED>
12:25.07phh<PROTECTED>
12:25.09phh<PROTECTED>
12:25.11phhyou think I should try that ?
12:25.16phhin crash handler
12:25.23cr2_it's "effective protection"
12:25.49cr2_you need to know where is this byte
12:26.14cr2_hm. but how do the flash radio on leo ?
12:26.35phhcr2_: so you need no datasheet leak ? hum.
12:26.44phhthat's a tough bet.
12:26.58cr2_datasheet is always good
12:27.18phhs/you/they/ actually
12:27.23phhwe need datasheet leak :p
12:27.31cr2_for chinese pmic
12:27.46cr2_no 8250 datasheet afaik
12:27.53cr2_only uboot code
12:28.07phhnot yet
12:28.21*** join/#htc-linux GeekLad (~GeekLad@adsl-178-230-51.jax.bellsouth.net)
12:28.47cr2_what is needed is the microusb->usbserial gadget
12:31.55cr2_HaRET(13)# pwf amss_B 0x17300000 0x20000000-0x17300000
12:31.56cr2_Short write detected while writing to file
12:31.57cr2_argh
12:32.00cr2_sd full ?
12:32.11phhlol
12:33.13cr2_it was flash, not sd
12:33.48cr2_hehe
12:33.54cr2_can't read it
12:34.13cr2_probably it was only smi mpu disabled here
12:40.24phhyou want to do the same thing for AXIGE instead of AXIGS
12:40.28phhbut I don't know what's his base addr
12:40.51cr2_HaRET(21)# pd 0xa8240800  0x10
12:40.52cr2_a8240800 | 00000001 00000000 00000000 00000000 | ................
12:40.57cr2_looks familiar
12:41.19phhlol yes
12:41.40cr2_E means EBI, and S SMI ?
12:41.53phhyes
12:42.01cr2_ok
12:42.48cr2_HaRET(5)# pfw 0xa8240800 1 0
12:42.49cr2_HaRET(6)# pd 0xa8240800  0x10
12:42.51cr2_a8240800 | 00000000 00000000 00000000 00000000 | ................
12:42.52cr2_:)
12:43.46cr2_hm
12:43.49cr2_HaRET(7)# pwf amss_B 0x17300000 0x20000000-0x17300000
12:43.51cr2_Short write detected while writing to file
12:44.17cr2_it has written 13M
12:44.42cr2_d+3
12:44.44cr2_lol
12:44.55cr2_raph has 128M
12:45.14cr2_pwf amss_B 0x17300000 0x18000000-0x17300000
12:45.45cr2_that was ok
12:46.17*** join/#htc-linux mcdull (~mcdull@203.210.8.168)
12:46.20cr2_ok, now i need to check if mpu off survives softreset
12:47.16*** join/#htc-linux FR^2 (~frzwo@2001:41d0:1:ed2f::cafe)
12:47.17*** join/#htc-linux pimenta (~josegs@bl15-41-193.dsl.telepac.pt)
12:48.30cr2_phh: i see that on leo the sdio is done with 1bit 100kHz setting
12:48.39cr2_initially.
12:48.52phhwe start with 1 bit 144kHz so ...
12:49.03cr2_then after reading the cis it switches
12:49.05phhhum not sure about 1 bit
12:49.24cr2_to 4bit 25MHz for bcm4329, and 4bit 50MHz for SD
12:49.50cr2_writing to sd on wince is painfully slow
12:50.03cr2_25MHz and no dma ?
12:50.16cr2_on raph
12:52.43*** part/#htc-linux pimenta (~josegs@bl15-41-193.dsl.telepac.pt)
13:00.58*** join/#htc-linux maejrep[w] (~madCoder@68.64.144.211)
13:01.28cr2_hi maejrep[w]
13:02.52cr2_phh: how do i recover the full elf from single sections ?
13:11.00cr2_mickeyl: ping
13:11.19*** part/#htc-linux junalmeida (~junior@201.19.226.123)
13:12.14*** part/#htc-linux FR^2 (~frzwo@2001:41d0:1:ed2f::cafe)
13:19.25phh~ramconsole-diam
13:19.25apthmm... ramconsole-diam is pwf dm 0x00800000 0x00100000
13:20.09cr2_phh: should we wipe out spl from smi ?
13:20.39phhcr2_: don't know, maybe it's used for faster soft reset ?
13:21.31cr2_i think dzo overrides the resume anyway
13:21.47cr2_phh: it's only about the resume path.
13:21.52phhok
13:21.56cr2_if you have your own, spl is irrelevant
13:22.23cr2_maybe we should study the resume path in spl to a better extent
13:22.24phhhum, timers are reseted by A9 ?
13:22.38cr2_which timers ?
13:22.45cr2_a9 can do everything
13:22.55cr2_including reading/writing to sd
13:23.13cr2_this ATL thing is a real monster
13:23.35phhthe A11 timers, GPT and DG
13:23.41*** join/#htc-linux avengerpenguin_ (~sean@5e051025.bb.sky.com)
13:23.44cr2_includes openssl, checks ephemeris certs, ...
13:24.00cr2_i wish we can remove parts from amss :)
13:24.14cr2_phh: the DG is very strange
13:24.29cr2_NetRipper disables it on all msm phones
13:24.35phhopenssl ?
13:24.36phhwahou
13:25.20cr2_i don't understand 1 thing
13:25.36cr2_how do they do a tcp connection from amss
13:25.58phhwhat's the problem with that ?
13:26.17phhunless they do it also for wifi
13:26.18cr2_we try to init ATL too
13:32.41*** join/#htc-linux playdo (~57b3c0a1@gateway/web/freenode/x-npkhtthuxljpbuqr)
13:35.44cr2_phh: is the at init sent in wince even when you don't have a sim card ?
13:36.06cr2_i'm thinking about such things:
13:36.10cr2_AT+HTCAGPS=5
13:36.10phhcr2_: yes
13:36.11cr2_AT@AGPSADDRESS=10,1,101,63,7275
13:36.25phhcr2_: for emergency call for example.
13:36.37phhat *gps* init ?
13:36.40cr2_maybe you should do it too before gps rpc init ?
13:37.35phhAFAIK it's done
13:37.47*** join/#htc-linux stinebd (~bryan@gentoo/developer/battousai)
13:38.14phh<PROTECTED>
13:38.14phh<PROTECTED>
13:38.15phhin radio on
13:38.22*** join/#htc-linux Vito89 (~quassel@195.113.241.225)
13:38.59*** join/#htc-linux |Jason8| (JayAte@2001:470:1f07:714::15e)
13:39.44cr2_what about dormancy at atz ?
13:43.39phhA9 reboot is expected anyway, the problem is not about such things, it's about reiniting correctly past A9 reboot
13:43.53cr2_ok
13:44.34phh<PROTECTED>
13:44.34phh<PROTECTED>
13:45.40cr2_but i don't see any smsm_cb debug messages in the log
13:49.05*** join/#htc-linux andreas32b (~andy@dslb-094-216-253-042.pools.arcor-ip.net)
13:58.31*** part/#htc-linux zachy (~george@ip-85-160-13-59.eurotel.cz)
14:14.43*** join/#htc-linux Vito89 (~quassel@195.113.241.225)
14:16.42*** join/#htc-linux mcdull (~blackston@061093132071.ctinets.com)
14:35.07*** join/#htc-linux c2d (~Mike@port-92-200-228-144.dynamic.qsc.de)
14:35.09*** part/#htc-linux c2d (~Mike@port-92-200-228-144.dynamic.qsc.de)
14:35.24NetRippercr2_, maybe you could check to see if any wince code refers to that DBG interrupt
14:36.14*** join/#htc-linux miknix (~miknix@gentoo/developer/miknix)
14:36.58NetRippercr2_, also, in haret, it always is 1, and never triggers
14:39.22*** join/#htc-linux leaigor (~laigor@188.134.16.241)
14:45.47*** join/#htc-linux Markinus (~Miranda@gtng-4db04e62.pool.mediaWays.net)
14:46.02MarkinusHi
14:53.04*** join/#htc-linux lucx (~luc@89-115-128-35.cl.ipv4ilink.net)
15:00.45*** join/#htc-linux andreas32b (~andy@dslb-094-216-253-042.pools.arcor-ip.net)
15:10.34*** join/#htc-linux mcdull (~blackston@061093132071.ctinets.com)
15:13.20*** join/#htc-linux lucx (~luc@89-115-128-35.cl.ipv4ilink.net)
15:26.33*** join/#htc-linux leaigor (~laigor@188.134.16.241)
15:58.53*** join/#htc-linux emwe (~michael@86.56.50.180)
16:04.53*** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring)
16:08.05*** join/#htc-linux babijoee (~babijoee4@d110-33-170-123.sun801.vic.optusnet.com.au)
16:14.06*** join/#htc-linux c2d (~Mike@port-92-200-228-144.dynamic.qsc.de)
16:21.56*** join/#htc-linux evildarknight (~charles@41.211.72.136)
16:29.01*** join/#htc-linux ReeferMattness (~MattGNM@c-76-18-79-170.hsd1.nm.comcast.net)
16:39.17*** join/#htc-linux cr2_ (~cr2@ip-109-84-15-100.web.vodafone.de)
16:40.46*** part/#htc-linux evildarknight (~charles@41.211.72.136)
16:43.42mickeylcr2_: pong
16:45.57phhmickeyl: have you tried GPS ?
16:47.16mickeylphh: not yet, neither on raph nor on dream, i'm waiting for a 'go' from the kernel developers. you said there are some things missing still, right?
16:47.48phhI can't get it to work on my diam, so I'd like to see that on a dream
16:48.06mickeylright. did you see my qmi question?
16:48.18phhyes, and I have no clue about it
16:48.32mickeylhehe, ok
16:48.46mickeylah well
16:48.48mickeylgood ol' ppp
16:49.10GNUtooabout gps:
16:49.32GNUtoowith newly released gps lib sources does someone still have to dump things?
16:49.48phhGNUtoo: the GPS lib they released has nothing to do with "our" gps
16:49.56GNUtoook
16:49.57phhit's only for devices I don't even know
16:50.03GNUtoolol ok
16:50.11phhI think even the nexus one doesn't use this api
16:50.18GNUtoook
16:50.38GNUtooso I must find the cyanogen mod image again
16:50.58phhcyanogen mod image .. ?
16:51.03GNUtoothen backup,flash
16:51.05GNUtooyes
16:51.18GNUtooI don't run gps lib on my android image
16:51.43phhok
16:51.44GNUtooit's just not present and there is a dummy gps instead
16:52.00GNUtooand a stock image won't have /dev/mem
16:52.12phhyou don't need /dev/mem
16:52.16GNUtooso I need a mix of stock and modified which cyanogen had
16:52.17phhoh well
16:52.18GNUtooha?
16:52.19phhmaybe cr2 want it.
16:52.45GNUtooindeed he asked me to re-dump things
16:53.37phhGNUtoo: you don't have a 8c rpc server in /dev/oncrpc do you ?
16:53.45cr2_GNUtoo: i need the smem dumps
16:54.23GNUtoophh, I'll look
16:54.48cr2_phh: oemsbl runs i2c bus @200k
16:55.57GNUtoohttp://pastebin.com/d7e702c61
16:56.39phhso that's a no
16:57.08GNUtoook I'll try to find this file
16:57.42GNUtoo(the cm-image)
17:00.32*** join/#htc-linux JesusFreak316 (~JesusFrea@pool-173-65-72-26.tampfl.fios.verizon.net)
17:02.28*** join/#htc-linux zfxp (~kw@adsl-179-119-130.gnv.bellsouth.net)
17:03.54*** join/#htc-linux GeertJohan (~Squarc@82-217-32-29.cable.quicknet.nl)
17:04.35*** part/#htc-linux JesusFreak316 (~JesusFrea@pool-173-65-72-26.tampfl.fios.verizon.net)
17:09.16*** join/#htc-linux balans (~Gebruiker@212-123-149-239.ip.telfort.nl)
17:14.38*** join/#htc-linux dcordes (~dccordes@unaffiliated/dcordes)
17:17.21GNUtooI didn't find it...I'll have to restore firmware and install cyanogen on top of it
17:21.58*** join/#htc-linux Taypen (~taypen@85.65.36.171.dynamic.barak-online.net)
17:23.59dcordeshi GNUtoo, others
17:24.43GNUtoohi
17:53.23*** join/#htc-linux Kashyyyk (~d1228602@gateway/web/freenode/x-fpwenchkipvsdcgp)
17:57.19*** mode/#htc-linux [+o dcordes] by ChanServ
17:58.06*** join/#htc-linux toi (~toi@d54C2A96D.access.telenet.be)
18:00.25*** mode/#htc-linux [-q $~a] by dcordes
18:00.43dcordes-unreggedtest 123
18:03.29*** join/#htc-linux GeertJohan1 (~Squarc@82-217-32-29.cable.quicknet.nl)
18:07.17*** join/#htc-linux jamenlang (~jamenlang@68.170.60.12)
18:09.30*** join/#htc-linux MrPippy (~pip@adsl-75-33-13-129.dsl.sndg02.sbcglobal.net)
18:12.52*** join/#htc-linux jamenlang (~jamenlang@68.170.60.12)
18:15.37*** part/#htc-linux jumoit (~samuel@121.77.87.145)
18:24.10*** join/#htc-linux x_coRe_x (~core@95-37-164-92.dynamic.mts-nn.ru)
18:33.42*** join/#htc-linux User99 (~User@209-193-77-154.mammothnetworks.com)
18:34.42ToAsTcfhdcordes-unregged: u find anything funny in my source?
18:40.39*** join/#htc-linux MarcLandis (~drettsch@port-92-195-10-37.dynamic.qsc.de)
18:42.31*** join/#htc-linux Wr4i7h (~abc@89.181.45.153)
18:45.08*** join/#htc-linux M1DLGpc (~M1DLGpc@bb-87-81-252-83.ukonline.co.uk)
18:45.09*** join/#htc-linux Captnoord (~Captnoord@dc5147a47b.adsl.wanadoo.nl)
18:47.54Captnoordhttp://deja.jails.nl/trashbin/phase_shift_oscillator.png
18:47.56Captnoordbleh!!!!
18:56.42MrPippyCaptnoord: can you take a look at pm_mic_en in the cdma remote_apis_clnt
18:57.07MrPippyi think i've figured it out (just sends one uint32) but there's a xdr/rpc call i'm not sure about
18:57.51cr2_MrPippy: do you see it in rpc fifo ?
18:58.10CaptnoordMrPippy: send me the dll
18:58.30Captnoordonly got wavedev from ya
18:58.32*** join/#htc-linux dcordes (~dccordes@unaffiliated/dcordes)
18:58.33MrPippyhaven't looked, although i have a bunch of mic on/off smem dumps
18:59.17MrPippyhttp://opus.homelinux.com/public_files/remote_apis_clnt_cdma.dll
18:59.23phhMarkinus: does the light sensor work the same way as diam/raph ?
18:59.30Captnoordthanks
18:59.31phh(cat /dbgfs/micropklt_dbg/light )
19:01.15Markinusphh: I didn't test this .. . I take a look on it
19:02.11CaptnoordSTR     R3, [SP,#0x40+var_40]
19:02.17Captnoordthat stuff is just 32 bits
19:02.29CaptnoordSTRB
19:02.30Captnoordis byte
19:02.37CaptnoordSTRH = short
19:02.39CaptnoordH = half
19:02.43MrPippyyeah
19:03.35*** join/#htc-linux karafa (~anonymous@ool-45735dce.dyn.optonline.net)
19:03.38CaptnoordSTR     R3, [SP,#0x40+var_40]
19:03.42Captnoordis first on the stack
19:03.49MrPippyi don't know what var_3c and var_40 are for, they never get read from here
19:03.52CaptnoordSTR     R3, [SP,#0x40+var_3C]
19:03.53Captnoordsecond
19:04.04phhMrPippy: if it works, we will have to try it on other boards, that w ould make turn_mic_bias_on code less ugly
19:04.39Markinusphh: Is there a userscape driver for this?
19:04.47Markinusspace
19:04.54phhMarkinus: yes my sensors code read this
19:05.00phhI haven't tested it at all though
19:05.01CaptnoordSTR     R3, [SP,#0x10+arg_C] ; rpc type
19:05.01CaptnoordSTR     R1, [SP,#0x10+arg_4] ; rpc_xid
19:05.33CaptnoordSTR     R0, [SP,#0x10+var_10] ; rpc_version
19:05.37Captnoordbleh
19:05.46MrPippyADD     R3, R3, #0xC is the func ptr for msg_done i think?
19:06.01phhMarkinus: but it works or not ?
19:06.24Markinusphh: I test it but need a moment, I flashed a 6.1 Rom, maybe I can see now the Red led. I put the other LED states in the wiki. Now I'm killing the battery . .15%  .. wait . .
19:06.28CaptnoordMrPippy: nah... it returns a whole lot of shit
19:06.33Captnoordnot just 1 uint32
19:06.35phhMarkinus: ok :)
19:06.55CaptnoordI guess it returns a struct
19:07.49Captnoord<PROTECTED>
19:07.49Captnoord<PROTECTED>
19:07.53Captnoordthat is kinda what we see
19:07.59*** join/#htc-linux GeertJohan (~Squarc@82-217-32-29.cable.quicknet.nl)
19:08.37Captnoordhmmm
19:08.40Captnoordi'm wrong
19:08.42Captnoordit returns a boolean
19:08.49Captnoordextern bool_t xdr_call_msg_start (XDR *xdr, uint32 prog, uint32 ver, uint32 proc, opaque_auth *cred, opaque_auth *verf);
19:10.07MrPippyyeah i think all of the functions return bools
19:11.05Captnoordhttp://www.netmite.com/android/mydroid/donut/hardware/msm7k/librpc/rpc/xdr.h
19:11.07Captnoordmy best bet
19:11.08Captnoord:P
19:12.10phhsame API under linux and windows ?
19:12.12phhsweet
19:12.22CaptnoordI guess so
19:12.26CaptnoordMS likes using sun stuff
19:12.34Captnoordsame as IBM stuff
19:12.54Captnoordthere entire arm compiler trick todo math fast is 'borowd' from ibm
19:12.57Captnoordnah
19:13.00Captnoordloads...
19:13.02Captnoordnot all
19:13.18MrPippythose ADD R3, R3, (offset) match up pretty well with the function ptrs in xdr_ops_struct: near bottom of http://www.netmite.com/android/mydroid/donut/hardware/msm7k/librpc/rpc/types.h
19:14.37*** join/#htc-linux balans (~Gebruiker@5354384A.cable.casema.nl)
19:14.48CaptnoordMrPippy: take the + 12 * 4
19:14.54Captnoord12 * 4 = 0x30
19:14.55Captnoordbool_t (*send_uint32) (xdr_s_type *xdr, const uint32 *value);
19:15.37MrPippyright
19:15.41Captnoordthe add 0x24
19:15.41Captnoordis
19:15.42Captnoordbool_t (*send_int16)  (xdr_s_type *xdr, const int16 *value);
19:15.59MrPippythat file doesn't have a 0x58 though
19:16.26Captnoord0x58 / 4
19:16.44Captnoordhmmm
19:16.45Captnoord22
19:16.46Captnoorddecimal
19:17.06Captnoordthe file we'r reading doesn't have to be the same as the wince implementation
19:17.22MrPippyyeah
19:17.29Captnoordthere are 20 functions there
19:17.32phhCaptnoord: it's most likely thouogh :p
19:17.41Captnoordlook at the onrpc export list
19:17.52Captnoordif there are functions not present in te list
19:17.54Captnoordphh true
19:18.05MrPippyand i tried putting a int16 into my req and the calls just timed out, they return fine if its just the uint32 (although the mic still doesn't work)
19:18.09Captnoordbut thinking that your enemy is stupid is one hell of a fuckup
19:18.09phhCaptnoord: well not exactly, if there are some constants, they have changed them
19:18.22CaptnoordMrPippy: its send uint32
19:18.26Captnoordreceive uint16
19:18.26phh(see rpc ids)
19:18.33Captnoordtrue
19:18.36Captnoordquite possible
19:18.49Captnoordthe good part is that they have to register / init those callbacks somewhere
19:18.52Captnoordso we can find them
19:19.21Captnoordthere isn't 64 bits in that list
19:19.31Markinusphh: Did you see the microp lightsensor driver from Leo? This is made in kernel. I thought yesterday about to port it . .but if this works with the old one . ..
19:19.32Captnoordthats 1
19:19.50phhMarkinus: leo ? nexus one you mean ?
19:19.55phhI don't think leo has this driver yet :p
19:19.59Captnoordand no enum
19:20.00Markinusphh: yes, sorry
19:20.06phhlet me check if it works the same way
19:20.09Captnoordlets see if I can find the init stuff
19:20.10Markinus:)
19:20.31phhMarkinus: also check if auto backlight works
19:20.53Markinusphh: board-mahimahi-microp.c, yes . . waiting waiting :)
19:21.22phhMarkinus: I hope you didn't forget the mmutrace i2c cmd ? :p
19:22.11phhMarkinus: ok, they don't use the same way
19:22.20Markinusphh: You meen the red led?? Or Light sensor?
19:22.25phhlight sensor
19:22.31phhbetween diam/raph and nexus
19:22.59Markinus(you post bevor, mmutrace)
19:23.35Markinusphh: but the light sensor connection is the same as in our devices or not?
19:23.46phhah no the mmutrace for red led
19:23.51phhsince you'll have little time to check it :p
19:24.05phhfor light sensor, there are two known ways, only one has been implemented
19:24.13phhyou'll see which one applies to you
19:24.33MrPippydoes it make sense that pm_mic_en does send_uint32, ? (send_int16?), ?? (0x58), msg_send, msg_done
19:24.34*** join/#htc-linux balans1 (~Gebruiker@212-123-149-239.ip.telfort.nl)
19:24.51phhMrPippy: not really, but eh, htc+qualcomm
19:25.03Markinusphh: I will do it, but I had to flash a old rom, now I hope the led is red with low battery and not amber flashing (like 6.5.3)
19:25.30CaptnoordMrPippy: please take into account that I am regularly wrong
19:25.40stinebd~seen babijoee
19:25.43aptbabijoee <~babijoee4@d110-33-170-123.sun801.vic.optusnet.com.au> was last seen on IRC in channel #htc-linux, 3d 3h 34m 19s ago, saying: 'cr2 would'nt lie :)'.
19:28.39phhcr2_: which way should I go to ? trying to reset communications after A9 reset, or assuming we do something wrong and fix it ?
19:29.53MrPippyi looked in oncrpc.dll for where they initialize the struct of func ptrs but nothing stuck out
19:29.54*** join/#htc-linux balans (~Gebruiker@212-123-149-239.ip.telfort.nl)
19:30.17Captnoordnah doesn't have to be done that way
19:30.21Captnoordthats the tricky part
19:30.47CaptnoordI think its part of
19:30.47Captnoordrpc_clnt_lookup2
19:30.50Captnoordreturning something
19:31.00MrPippyit returns the xdr* struct
19:31.10Captnoordyup
19:31.50MrPippyand 8 bytes into that is the ptr to the xdr_ops_struct
19:32.02*** join/#htc-linux balsat (~balsat@87.72.13.34)
19:33.00Captnoordit doesn't have to be done that way
19:33.04Captnoorduse it as a guide
19:33.06Captnoordnot as facts
19:38.17CaptnoordI think that lookup2 is a wince only thingy
19:38.19Captnoordand that its called
19:38.20Captnoordsvc_find
19:38.30Captnoordor something like that
19:38.37Captnoordmaybe clnt_find
19:38.50Markinusphh: with this rom amber blinking too, argh. . . I have no Idea how to enable the red led ..
19:39.00phhMarkinus: maybe you have none
19:39.22CaptnoordMarkinus: what do you set now?
19:39.24Captnoordhow*
19:39.49Markinusphh: I know we have on, If the accu was empty this was red . .but with this Roms not . . .
19:39.52Captnoordtypical dual color led
19:39.52Captnoordhttp://www.elexp.com/opto/lhg3392.gif
19:40.36Captnoorda way to control it
19:40.36Captnoordhttp://www.wiringfordcc.com/sig_rev_trk_3.GIF
19:40.42Captnoordone of the many ways
19:41.18chamonixhey guys
19:41.41MarkinusCaptnoord: Yes, you can generate 3 colors
19:42.03Markinusamber: Red + Green
19:42.17Captnoordwhat is the problem then
19:42.20Captnoordwince rom's?
19:42.23MarkinusI know the Red LEd only on low battery
19:42.28Markinusyes
19:42.40Captnoordk
19:43.05Markinusbut with the two roms which I have the led is blinking Amber instead of be red
19:43.23MarkinusI didn't know whitch is the right rom :(
19:43.33Captnoordhehe
19:43.44Markinusand I don't know a other way to enable this . . .
19:43.48Markinusunder winmo
19:43.51CaptnoordI wish I could help ya
19:43.51Captnoord:P
19:44.22phhMarkinus: isn't there a setled haret command or something like that ?
19:44.39Markinushmm, I will look.
19:45.23*** join/#htc-linux balans (~Gebruiker@212-123-149-239.ip.telfort.nl)
19:46.28cr2_phh: i think that the most promising is to find out what wince does around RPC_Init()
19:46.47CaptnoordMrPippy: you can't find it on the onrpc.dll as the dll itself uses exactly the same method for the exports
19:47.43Captnoordcheck the dll entry point
19:47.45Captnoordfor a init
19:48.09phhCaptnoord: can you do that ? :D
19:48.11Markinusphh: there is somethin, I testing :) . .ehh, led 1 -> vibrator
19:48.35Captnoordphh i'm checking
19:48.36Captnoord:P
19:48.37Captnoordlolz
19:48.49Captnoordbingo
19:48.50Captnoord:P
19:48.51Captnoordthats easy
19:49.07phhRPC_Init is easy ?
19:49.12Captnoordnah
19:49.18Captnoordthe dll entry point stuff
19:49.37Captnoord[20:46] <cr2_> phh: i think that the most promising is to find out what wince does around RPC_Init()
19:49.39Captnoordaahhh
19:49.42Captnoordyou mean that stuff
19:49.43Captnoordlol
19:49.47phhyes :p
19:49.48*** join/#htc-linux randomblame (~kevin@c-71-59-184-96.hsd1.wa.comcast.net)
19:52.36Markinusphh: I tested the led ids 0 - 30 and only 1 ist working, and this is vibration
19:52.42phhMarkinus: no
19:53.57Markinusphh: ? Is there a way to dump the wince led ids?
19:54.03phhMarkinus: don't know
19:57.01phhCaptnoord: in which register is stored the return value of a function ?
19:57.55Captnoordr0
19:57.59Captnoordsome function do
19:58.01Captnoordr0 r1
19:58.08Captnoordwhen a 64 bits value is returned
19:58.14*** join/#htc-linux balans (~Gebruiker@212-123-149-239.ip.telfort.nl)
19:58.31Captnoordbut.... it only works with real calls
19:58.34Captnoordnot with branches
19:58.37Captnoordso
19:58.41Captnoordbx b bl
19:58.42*** join/#htc-linux zachy (~george@85.193.4.18)
19:58.46Captnoordthat doesn't work
19:59.12Captnoordwith those its anybody's guess
20:03.10Captnoord-bx
20:03.16Captnoordbx is what we consider a real call
20:03.18*** join/#htc-linux onen|openBmap (~quassel@vbo91-1-89-87-201-85.dsl.club-internet.fr)
20:03.25Captnoordso return would be r0
20:03.59phhbaahh
20:04.03phhok what about RPC_Init now ? :p
20:07.34Captnoordpdsm_xtra_client_initiate_download_request
20:07.36Captnoordlol
20:07.44Captnoordso you can download a extra rpc client
20:07.49Captnoordor do I see it wrong
20:07.52*** join/#htc-linux balans (~Gebruiker@212-123-149-239.ip.telfort.nl)
20:08.14phhCaptnoord: na, it's for gps fast thingy
20:08.21phhyou save ephemeris in A11
20:11.34Markinusphh: cat /dbgfs/micropklt_dbg/light  -> 0
20:11.45phhMarkinus: ok, so you'll have to port N1's code
20:11.53Markinusphh: ok
20:15.08*** join/#htc-linux m3dlg (~m3dlg@212.183.140.51)
20:19.02*** part/#htc-linux c2d (~Mike@port-92-200-228-144.dynamic.qsc.de)
20:22.53phhcr2_: hum, android's libgps.so sends client_act even if atl fails (ie no internet connection)
20:24.26cr2_why not.
20:25.25cr2_phh: light sensor cal is in nand too. at least on leo
20:25.30cr2_00:00:18.886    OEM_FMD_NandReadLSensor is done!
20:25.32cr2_00:00:18.886    [D:MISC] [ReadStoredADCValue] *pdwReadDataValue = 0x8d (141)!
20:25.33cr2_00:00:18.886    [D:MISC] [KLTDevInitThread] g_dwStoredADCValue: 141!
20:25.53phhcr2_: weird, mahimahi has it in driver
20:27.15phhand they change them sometimes in commits
20:27.15cr2_[D:DEX] Update Audio Setting = [0x80]
20:27.21MrPippycr2_: what is the best way to trace rpc? i have smem dumps, or is mmutrace better?
20:27.22cr2_this is new
20:27.35phhMrPippy: harder to read
20:27.50cr2_MrPippy: od -A n -v -t x4 rpcfifo
20:28.09cr2_MrPippy: add skip, if you have full smem dump
20:28.24cr2_MrPippy: and then search for 000030 and 000031
20:29.39cr2_phh: i've added the rpc_clk_khz command to haret, and it works
20:29.57phhcr2_: uh ? which ones ?
20:30.07*** join/#htc-linux leaigor (~laigor@188.134.16.241)
20:31.17Captnoordharet overlocking
20:31.17Captnoordlolz
20:32.52NetRipperis that to read or set clocks?
20:33.04cr2_NetRipper: your code for read
20:33.09cr2_00:00:18.952    <ATDbg Log>Log RAM Base = 900b0000
20:33.10cr2_00:00:18.952    [D:RIL] ATDbg_GetRegistryData: ATLog RAM Base = 0x900b0000
20:33.12cr2_this is strange
20:33.30cr2_HaRET(3)# print "0x%x" v2p(0x900b0000)
20:33.31cr2_0xac0b0000
20:33.36NetRippercr2_,  ok
20:33.50cr2_NetRipper: i've disabled mmu on raph, dumped amss and oemsbl
20:33.54cr2_from ram
20:34.17NetRippernice
20:34.18cr2_oemsbl can be easily dissected. but most things we already know
20:34.39cr2_for amss i need to glue the raw segments into an elf image
20:34.44NetRipperdisabling mmu also disables mpu?
20:34.50*** join/#htc-linux Zoolooc (~fredsibar@p54955995.dip.t-dialin.net)
20:34.58cr2_the map is probably the same as in wiki
20:35.01phhNetRipper: he meant mpu :p
20:35.02cr2_NetRipper: no
20:35.07NetRipperphh, oh
20:35.15cr2_NetRipper: btw, it will not work on leo
20:35.42cr2_either the address spaces are different, or (more likely) they mpu'd them
20:35.44NetRipperphh, i was thinking quite an exciting solution :) take control, disable mmu, copy oemsbl to other part of phys ram, reboot, read ram
20:35.47NetRipper:p
20:36.01phhNetRipper: I don't think disabling mmu disables mpu
20:36.08MrPippyrpcfifo is the smd rpccall?
20:36.33cr2_NetRipper: ? if you disable mpu, you can do all you want from arm11
20:36.37cr2_MrPippy: yes
20:36.49NetRippercr2_, how'd you find the addresses on raphael then?
20:37.00cr2_which addresses ?
20:37.12phhNetRipper: ask dzo
20:37.30cr2_AXIGS is in the code
20:38.11cr2_AXIGE i found by dumping the data an neighbour address spaces
20:38.16NetRippercr2_, which adress spaces would be different on leo? oemsbl/amss?
20:38.26NetRipperah
20:38.26cr2_these too
20:38.34cr2_i mean AXIGS and AXIGE
20:38.52NetRipperthese abbreviations don't ring a bell fo rme
20:39.12cr2_on leo you either solder a cable, connect to oemsbl and disable mpu with a command
20:39.32NetRipperyou'd you disable it on raphael then?
20:39.45cr2_and you need to strip the radio code from extra bytes
20:39.55cr2_NetRipper: on raph they are not protected
20:39.58NetRipperoh
20:40.05NetRipperhmm
20:40.07cr2_you just clean the bit with haret, and that's it
20:40.11NetRipperlol
20:40.22cr2_it's already used in the kernel code by dzo
20:40.31cr2_to save the resume vector
20:40.42NetRipperaha
20:40.45cr2_we may actually wipe the spl copy from smi
20:40.46NetRipperhmm?
20:41.00NetRippereh
20:41.26NetRipperbut that's not a problem is it? smi is volatile anyway?
20:41.29cr2_don't know if running with mpu off is a good idea for android :)
20:41.40cr2_smi is sdram
20:41.52NetRipperyea so a proper reset would fix it
20:41.52cr2_yes, we can use it for linux
20:41.59cr2_yes
20:42.24cr2_NetRipper: latest leo news, if you missed them
20:42.39NetRipperi don't see the relation between needing to disable mpu and requiring that to save resume vector
20:42.55NetRipperthat means winmo would always need to save resume vector, also on leo
20:43.04cr2_NetRipper: the battery monitor is ds2745. it's very close to ds2746 used on kaiser, but we need to patch some data in the driver
20:43.12NetRipperok
20:43.24cr2_NetRipper: no. it stores spl
20:43.37cr2_and spl does control the resume
20:43.43NetRipperah
20:44.00NetRipperreally.. so if i resume my winmo phone.. it first goes through spl, and spl kicks off winmo again?
20:44.01cr2_NetRipper: the accelerometer is behind the microP
20:44.27cr2_but the magnetometer is on the usual 0x1c i2c id like on mahimahi
20:44.42NetRipperok
20:44.47phh(we already have code for accelerometer)
20:44.49cr2_this is strange. can't we just use it for the accelerometer too ?
20:45.25NetRipperis microp a bus or connected directly/
20:45.48*** join/#htc-linux Vito89 (~quassel@gw.loccal.net)
20:45.56cr2_NetRipper: this sounds weird. do we have 2 accelerometer chips in leo ??
20:46.05NetRippernot to my knowledge
20:46.10NetRippermagnetometer != accelerometer
20:46.30cr2_ak is accelerometer+magnetometer ?
20:46.38NetRipperis it?
20:46.46phhno it's not
20:46.52phhit's only magnetometer
20:46.53cr2_phh: only compass ?
20:46.57phhyes
20:47.00NetRipperok
20:47.05cr2_and g1 ?
20:47.12cr2_hm. it uses a different model
20:47.14NetRipperhas only compass
20:47.21phhargh
20:47.22cr2_no
20:47.22phhgood question.
20:48.27cr2_AKM8973 on mahimahi
20:48.49NetRippermahimahi is n1, not g1
20:49.16cr2_akm8976 on sapphire
20:49.34cr2_also on g1
20:51.25cr2_akm8976.c has both, according to the code
20:51.43cr2_lol,
20:51.57cr2_if we have 2 accelerometers will they match ? :)
20:52.09phhno :p
20:53.20cr2_www.ic-on-line.cn/iol/datasheet/ak8973_4138699.pdf
20:54.10cr2_no accelerometer
20:54.30cr2_that makes sense ;)
20:55.15cr2_NetRipper: don't see any trace of proximity sensor
20:55.39cr2_NetRipper: but since i2c is not traceable (yet), can't be 100% sure
20:55.39phhcr2_: tmzt said IR led + light sensor
20:56.05cr2_phh: on mahimahi ?
20:56.14phhmm don't remember
20:56.50cr2_capella_cm3602
20:57.01cr2_i don't think leo has this chip
20:58.04cr2_http://www.capellamicro.com.tw/EN/products_view.php?id=45&mode=16
21:00.30NetRipperthey advertised leo with proximity sensor
21:00.49cr2_NetRipper: yes, i've seen it on htc site
21:01.02NetRipperwould be bad if they dont actually have one in it
21:01.18cr2_NetRipper: it may be behind microP, like the aceelerometer
21:01.34NetRipperok so the microp is like a bus and can hold multiple devices?
21:01.44cr2_so we don't see it directly, and only get an irq from microP
21:02.12cr2_microP is a microcontroller
21:02.25cr2_it has gpios, spi ,ADC, pwm, etc
21:02.32cr2_you talk over i2c to it.
21:02.46cr2_it has some firmware, so you need to know which commands to send
21:03.23cr2_NetRipper: what are these 2 holes left from led ?
21:04.08NetRipperwell.. that _should_ be light and proximity next to eachother
21:04.12cr2_but on nexus you can talk directly to this CM3602  chip
21:04.31cr2_CM3602 is a highly integrated design for a distance detection and ambient light sensing solution.
21:04.49cr2_maybe NIR led + photodiode ?
21:05.03NetRipperthough havent confirmed that obviously but hmm.. there should be some disassemblation pictures around of the leo
21:05.10NetRipperyep who knows..
21:05.12*** join/#htc-linux karafa (~anonymous@149.151.184.108)
21:05.30NetRipperdont know how a proximity sensor would work exactly... how it detects the presence.. IR sounds feasible
21:06.02Captnoordmagnetic fields
21:06.09CaptnoordRF
21:06.17CaptnoordIR idd
21:06.23cr2_Captnoord: it's more like lidar
21:06.29Captnoord== RF
21:06.46cr2_you send IR, and get scattered light back
21:06.56Captnoordstill RF
21:06.56Captnoord:P
21:07.27cr2_if the scattered light is strong you have proximity
21:07.37*** join/#htc-linux randomblame (~kevin@c-71-59-184-96.hsd1.wa.comcast.net)
21:07.41cr2_hehe. then gamma rays are RF too ;
21:07.54Captnoordeverything is RF
21:07.55Captnoord:P
21:08.17Captnoordbut its also possible todo it capacitive
21:08.23Captnoordwhich is the cheapest
21:08.30cr2_Captnoord: is it possible to stick together the elf sections using ldscript ?
21:08.32Captnoordir is not so cheap
21:08.41Captnoordldscript?
21:08.42*** join/#htc-linux zachy (~george@85.193.4.18)
21:08.49Captnoordor ida script
21:08.50cr2_Captnoord: http://www.capellamicro.com.tw/EN/products_view.php?id=45&mode=16
21:09.18Captnoordhehe
21:09.33CaptnoordI think its easier to write a parser
21:09.41Captnoordto handle all the nb things
21:09.49*** part/#htc-linux zachy (~george@85.193.4.18)
21:10.09Captnoordbut I need togo... i'll call it a night....
21:10.12Captnoordneed sleep
21:10.12Captnoordnn all
21:10.16cr2_Captnoord: you need to be sure about the offsets
21:10.47MrPippycr2_: i found 000031 and 000030, what are those?
21:11.00cr2_RPC and RPC_CB
21:11.06cr2_PROG
21:11.15cr2_they come in triples
21:11.33cr2_PROG,VERS,FUNC
21:13.19cr2_this is from leo
21:13.20*** join/#htc-linux xfight81 (~quassel@host94-144-dynamic.5-87-r.retail.telecomitalia.it)
21:13.21cr2_0f000030 00000000 06000000 00000000 00000000 00000000 00000000 37000000
21:13.35*** join/#htc-linux lucx (~luc@89-115-128-35.cl.ipv4ilink.net)
21:13.41cr2_f,0,6 (0x37)
21:14.11cr2_PROG,VERS,FUNC,0,0,0,0,payload
21:14.36MrPippyok cool thanks
21:14.41phhcr2_: so what's the rpc clocks thing you were speaking about earlier ?
21:15.09cr2_phh: start the cam, and look for the clocks used
21:15.17cr2_i think they are all documented in wiki
21:15.27phhah that's what you were speaking about ?
21:15.37phhI thaught you were reading clocks through rpc
21:15.41cr2_phh: clocks in haret ?
21:15.44phhyes
21:15.58cr2_yes, but by its regime id
21:16.13phhcr2_: it works for all clocks ?
21:16.18phhor only clocks activated with rpc ?
21:16.19cr2_<PROTECTED>
21:16.21cr2_<PROTECTED>
21:16.22cr2_}
21:16.24cr2_REG_CMD(ClkRegimeAvail, "MSMCLKKHZ", msm_clk_dbg,
21:16.25cr2_<PROTECTED>
21:16.27cr2_<PROTECTED>
21:16.29cr2_it's how wince does it
21:16.33phhok
21:16.52cr2_i think i'll add the id as parameter instead
21:17.04cr2_of dumping all 256 values at once
21:17.51phhok I'll try to see if I can get it working from linux
21:17.56phhand if it works, seeing who is who
21:18.11cr2_ok
21:18.38cr2_you nly need to parse rpc reply
21:18.44cr2_like i've done for gps
21:18.45phhyes no problem
21:18.53phhcr2_: any idea to read vregs ?
21:19.06phhI can't see how we can do that at all, either by dumping memory or by rpc
21:19.16cr2_i've seen 3 initial vreg tables in oemsbl
21:19.33cr2_i think it's impossible
21:19.52cr2_is there are api for reding vreg voltages ?
21:23.03*** join/#htc-linux kiozen (~oeichler@rgnb-5d87c321.pool.mediaWays.net)
21:26.03phhcr2_: rpc argument is int32 ?
21:26.10cr2_NetRipper: does it tell you anything
21:26.13cr2_9.067    [sleep_register] name=PROXIMITYTCXO, okts=1, FreeEntryIndex=6
21:26.15cr2_9.070    [D:LED] [RegisterForProximityStatusNotifications] Register registry System\State\Phone PSensorStatus callback function!!
21:26.16cr2_9.071    [D:LED] [RegisterForProximityStatusNotifications] Register registry Software\HTC\Proximity PSensorDebounceTime callback function!!
21:26.18cr2_9.071    [D:LED] [RegisterForProximityStatusNotifications] Register registry System\GDI\Rotation Angle callback function!!
21:26.19cr2_9.072    [D:LED] [LED_Init] End!!
21:26.34cr2_phh: for rpc everything is int32
21:26.46phhcr2_: reply too then ?
21:26.52cr2_yes
21:27.11cr2_int32 or array/struct of int32
21:31.58cr2_phh: leo is difficult for rpc, because there are no rpc messages in wince dmesg
21:38.18NetRippercr2_, errr tells me the proximity status is probably a led?
21:39.37*** join/#htc-linux bzo (~chatzilla@netblock-68-183-234-14.dslextreme.com)
21:42.07*** join/#htc-linux jamenlang1 (~jamenlang@209-193-77-154.mammothnetworks.com)
21:43.42cr2_NetRipper: an irq
21:43.58cr2_phh: s5k3e2fx does have a lens control ?
21:44.18cr2_phh: i2c control like on mt* ? does not look from the driver code though
21:44.26phhdon't know
21:47.17cr2_NetRipper: http://www.htc-linux.org/wiki/index.php?title=I2C_IDS
21:47.41NetRippernice
21:47.51NetRipperleo's ts is on i2c?
21:48.03phhNetRipper: most likely
21:48.04cr2_like on nexus
21:48.22NetRippersweet
21:48.30cr2_the same driver
21:48.36phhNetRipper: why sweet ?
21:48.47NetRipperbecause we can reuse it
21:48.51phhah.
21:48.55phhlazy man
21:48.59NetRipperlol
21:49.00cr2_phh: i fixed some data in the table
21:49.05NetRipperthat's me phh:)
21:49.12NetRippergeneral purpose htc device
21:49.21NetRipperwhat's that?
21:49.26phhNetRipper: that's microp.
21:49.36NetRipperah
21:49.38cr2_phh: what is 0x38 ??
21:49.42NetRipperoh
21:49.58phhcr2_: don't know yet
21:50.14cr2_phh: there is some speaker amplifier
21:50.18phhit just exists
21:50.23phhcr2_: different from A1010 ?
21:50.25cr2_phh: and dualmic something ?
21:50.31phhdualmic is A1010
21:50.40cr2_let me check my notes
21:51.41cr2_ADC3001 mic
21:51.46phhcr2_: all get_khz rpcs gives me a -1073177860
21:52.10cr2_a1010 is f4/2
21:52.22phhyes
21:52.49cr2_ok
21:53.15cr2_check TPA2016 and ADC3001
21:53.35cr2_also RJ6ABA100
21:53.57phhTLV320ADC3001: 92dB SNR Low-Power Stereo ADC (ADC3001)
21:53.58phhok.
21:54.11cr2_i2c ?
21:54.22phh2.8-W/Ch Stereo Class-D Audio Amplifier with SmartGainTM AGC/DRC (TPA2016)
21:54.25phh2.8W... wahou
21:54.40cr2_i2c ?
21:54.57phhADC3001 yes
21:55.24cr2_tpa too
21:55.31cr2_let me grep for id
21:56.02phhfirst 7bit of id for ADC3001 is 100011
21:56.22cr2_0x23 ?
21:56.26phhuh ?
21:56.29phhmy calc says 0x86
21:56.38phhcr2_: there is another bit after that
21:56.41phhthat is settable to 0 or 1
21:56.52cr2_it's for /2
21:57.12phhwhy would that be /2 ?
21:57.14phhit's actual i2c addr
21:57.17phhnot msm's addr
21:57.48cr2_why do they describe i2c operation in all datasheets ??
21:57.59phhI always wonder too :p
21:58.08cr2_0xb0/2
21:58.31cr2_0x58
21:58.43cr2_so 0x23 and 0x58
21:58.46cr2_hehe
21:58.54cr2_RJ6ABA100 ?
21:59.10phhVGA CMOS camera
21:59.24cr2_yeah, but no datasheet
22:00.27cr2_[D:I2C] I2C Last success, ID: 0x6c, Addr: 0x1a30
22:03.28phhno 0x23 nor 0x58 chip found in dumps
22:03.46*** join/#htc-linux tehtrk_home (~quassel@adsl-68-91-144-191.dsl.wacotx.swbell.net)
22:07.10*** join/#htc-linux m3dlg (~m3dlg@bb-87-81-252-83.ukonline.co.uk)
22:10.22cr2_ok
22:12.41cr2_NetRipper: what was the problem with i2c driver on leo ?
22:13.40*** join/#htc-linux Olipro (~Olipro@uncyclopedia/pdpc.21for7.olipro)
22:13.51NetRippersame handle_bad_irq as mdp had
22:14.07cr2_MDP_INTR_ENABLE=e74d MDP_INTR_ENABLE= 274d
22:14.16cr2_e is 14
22:14.32cr2_1110
22:15.06cr2_what are 8000 and 4000 bits for mdp ?
22:18.28NetRipperi don't know but it seems the N1 driver is doing different things than the wince driver
22:19.37cr2_hm.
22:19.54cr2_what is the mdp version ?
22:20.42NetRipperermm i dont know
22:21.13phhcr2_: 40 iirc
22:21.26phhhum, 31
22:21.38cr2_ok
22:22.24phh22 for msm7k
22:22.43NetRipperah
22:22.48NetRipperthat version
22:25.09*** join/#htc-linux mickeyl (~mickey@80.81.242.146)
22:27.56Markinuscould be, that the N1 kernelsource isn'T complete? I'M searching for the definition of dev_attr_ls_adc and can't find anything .. hmm
22:28.40phhMarkinus: they are complete
22:28.45phhsearch for ls_adc
22:28.48phhyou'll find a mcro
22:28.49phhmacro
22:29.01Markinusphh: ahh, thx
22:29.47ToAsTcfhphh: anyluck on the hw3d in 2.x? did u ever load up my build?
22:30.01phhToAsTcfh: I loaded your build but didn't try hw3d on it
22:30.24phhand we already have hw3d in 2.X
22:30.29phhwe had it before you
22:30.30ToAsTcfheh what u think of the rosie? i know u said u didnt like the old one
22:30.55phhToAsTcfh: it seems quite pretty, but totally unusable with only 107MB  of ram
22:31.15phheven with 40MB ramzswap
22:31.44ToAsTcfhoh i didnt know u guys had hw3d workin in 2.x. i though u had just got the driver to load
22:33.01randomblamewoot my kernel is now hanging at mount_root, much further
22:33.02ToAsTcfhyeah im gonna release a new build soon with it in it. ill make sure u and MrPippy are thanked in the first post :)
22:33.47phhToAsTcfh: please especially says we work for non-android devices, that's a lot funnier that way :p
22:34.24ToAsTcfhok i always brag anyhow that the dream devs are no help and i gotta go to winmo devs to get real ideas
22:35.16ToAsTcfhwe still havent got overclocking working http://forum.xda-developers.com/showthread.php?t=630899
22:35.33adhocmorning all =)
22:36.16ToAsTcfhi started adev tread for it but people are kinda low moral about it
22:37.08tmztMarkinus: why would dev_attr_ls_adc be in the kernel?
22:38.01phhToAsTcfh: overclocking to something else than 768MHz is really hard
22:38.16adhocis there something broken with the xda-developers wiki ?
22:38.18tmzt22 is msm7x0x then?
22:38.22*** join/#htc-linux leaigor (~laigor@188.134.16.241)
22:38.24ToAsTcfhive tried to tell them that
22:38.25tmzt31 is msm7x27/30
22:38.30tmzt40 is snapdragon?
22:38.37Markinustmzt: lightsensor
22:38.52phhtmzt: snapdragon is 31
22:39.00phhI think msm7x27 and QSD has the same
22:39.13ToAsTcfhphh: they dont seem to think u can 2/3 the multiplier or remove turbo mode
22:39.15Markinustmzt: device_create_file(&client->dev, &dev_attr_ls_adc);
22:39.31phhToAsTcfh: you can... by changing PLL
22:40.29ToAsTcfhi havent got deep enough into it to be able to explain it to them. maybe u should post a suggestion on that thread i started
22:41.03ToAsTcfhphh: please
22:41.10ToAsTcfh:P
22:41.13tmzthmm so it's defined by macro
22:41.26tmztlook for ls_adc
22:41.46tmztphh: same gpu but I think mdp is different
22:42.29ToAsTcfhanyone had any luck on the msm_camera yet?
22:42.30phhtmzt: the only mdp in N1's sources are 22 and 31.
22:42.30tmzthmm
22:42.30tmztok
22:42.30phhToAsTcfh: I don't need luck
22:42.30ToAsTcfhu got it?!
22:42.40phhToAsTcfh: no
22:42.47ToAsTcfhphh: http://github.com/jhansche/htc-2.6.27-heroc
22:42.50phhI didn't try :p
22:43.02ToAsTcfhu should man were stuck
22:43.16ToAsTcfhbad stuck and ur the guru
22:44.50ToAsTcfhi actually had to port both patches for hw3d as the kernel u guys use is setup a little different
22:47.23ToAsTcfhphh: i guess thats a hell no for the camera?! lol
22:47.40phhthat's a "I don't feel like trying"
22:47.50phhbut it's all I've left to try
22:47.50ToAsTcfhbah
22:47.53phhsince gps doesn't work
22:47.59phhbluetooth is boring
22:48.23ToAsTcfheh i got gps working in my buiold by using a different gps.conf
22:48.32phh.
22:48.37phhwe don't have the same gps RPCs.
22:49.00ToAsTcfhoh just sayin
22:49.23ToAsTcfhso try camera. its new its fresh and its a hot item
22:49.26ToAsTcfhlol
22:49.36ToAsTcfhits all thats left
22:50.25ToAsTcfhits chalenging and a pain in the ass!!!
22:51.52phhwhich camera ?
22:51.53phhVGA one ? :p
22:52.42phhToAsTcfh: you'd hate me if I can get VGA camera but not 3M one, don't you ?
22:52.56cr2_LOL
22:53.01ToAsTcfhi got a 5m but ill settle
22:53.09phhToAsTcfh: VGA camera = front camera btw
22:53.15phh(the thing your shitty device hasn't :p)
22:53.30ToAsTcfhoh fuck yeah id be mad
22:53.30cr2_hahaha
22:53.36ToAsTcfhhater
22:53.57phhToAsTcfh: getting VGA cam working but not 3M one would be hard though.
22:54.21phhToAsTcfh: ok, I'll try camera tomorrow. maybe.
22:54.27ToAsTcfhwould u do it just to crush me
22:54.42phhtoo much work for not such a great pleasure
22:54.51ToAsTcfhlol
22:55.30ToAsTcfhwell if u say ull try it for use then im happy with that
22:56.00ToAsTcfh:D
22:56.38ToAsTcfhdo u guys have cdma mms?
22:56.46phhwe don't have mms at all
22:56.52ToAsTcfhoh
22:56.52*** join/#htc-linux m3dlg (~m3dlg@bb-87-81-252-83.ukonline.co.uk)
22:56.56phhit's awful to get working with ppp
22:57.02ToAsTcfhme either :P
22:57.48*** join/#htc-linux marex (~marex@vasut.kolej.mff.cuni.cz)
22:58.02bzoToAsTcfh: saw an interesting blurb from dzo about overclocking "Yes, there is a spare PLL and with the old clock code you can set it to any speed you like, just add clock_7x00.a11=528 (or higher) to the cmdline. You'll also have to change the ahb bus divider with clock_7x00.ahb_div=4.
22:58.03bzoI got rid of this in the current kernel because it's fun to play with but probably not really something you want to use all the time."
22:58.22bzoso, switching to the other pll sounds like what you need to do if you want to overclock with any granularity
22:58.40phhif there is actually a spare pll.
22:59.04bzophh: you mean vogue has one, but not other devices?
22:59.18phhbzo: raph/diam has one
22:59.21phhdon't know about dream
22:59.23tmztno2chem says there isn't
22:59.37tmztno really at least
22:59.46ToAsTcfhwhat about hero or tuurbo msm7k
22:59.48phhwe have one pll that is always off at least
23:00.34phhand I think we can set it as will
23:00.58NetRippercr2_, is it easy to set clock frequency on leo?
23:01.23NetRippernot to overclock but just to set the normal clocks that are already supported?
23:01.26cr2_NetRipper: the same as on 7201a
23:01.29ToAsTcfhbzo : u got a patch i can look at?
23:01.29bzowe might have better luck than no2chem since we can change the kernel. Maybe you cannot change it in userland
23:01.33NetRippercr2_, alright
23:01.43phhbzo: you can in userland
23:02.06cr2_NetRipper: there is some scpll, but the pll0-3 are setup in the same way
23:02.20bzoToAsTcfh: haven't looked into it. But all of dzo's source is at linux2go for you to check out :)
23:02.27tmztphh: he's writing about the primary clocks with crystals
23:02.31tmzttxo's etc.
23:02.43tmztnot the dividers which are also plls
23:02.46NetRippercr2_, are these just regs or do they go via amss?
23:02.47ToAsTcfhbzo: ok thanx
23:02.50phhtmzt: yes, but A11 derivates (with only divisions) from a PLL
23:02.57phhand we can set pll
23:04.39*** join/#htc-linux The-Compiler (florian@the-compiler.org)
23:04.40*** join/#htc-linux The-Compiler (florian@unaffiliated/the-compiler)
23:05.03*** join/#htc-linux cr2_ (~cr2@ip-109-84-15-100.web.vodafone.de)
23:05.30NetRipperwb ;)
23:07.08ToAsTcfhanyone got a link to linux2go dzo's source?
23:07.13ToAsTcfhi cant find it
23:11.58phhToAsTcfh: vgrapper doesn't crash my phone ! :p
23:12.26bzoToAsTcfh: http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=shortlog;h=refs/heads/htc-vogue
23:12.31ToAsTcfhhmm does it owork
23:13.03phhToAsTcfh: ok no.
23:13.16phhthat would have been too unfair :p
23:13.22ToAsTcfhlol
23:13.46ToAsTcfhso u already gave the msm_camera a port?
23:13.53phh?
23:13.59ToAsTcfhbzo: thanx
23:14.09phhdon't you remember you got the msm_camera stuff from our tree ?
23:14.21phh(even if it's not working.)
23:14.36ToAsTcfhwe didnt get it from ur tree
23:14.48ToAsTcfhmaejrep backported it
23:15.19phhwhy did you ask for my patches then /o\
23:15.37ToAsTcfhhw3d patches
23:15.45phhno for camera.
23:15.45ToAsTcfhnot camera patches
23:16.03ToAsTcfhsend me them and ill givem a shot
23:16.35ToAsTcfhi dont remember u giving me any patches for camera
23:16.54ToAsTcfhmaybe a nother pain in the ass named toast?
23:16.59phhargh vgrabber is for v4l1 not v4l2.
23:17.11ToAsTcfhi told u that already
23:17.23ToAsTcfh:P
23:17.35ToAsTcfh\u never listen to me damn it!!!
23:17.47phhwhat did you use then ?
23:18.02ToAsTcfhthe camera in my build
23:18.10phhok so you trust android
23:18.12ToAsTcfhthats all i could findd
23:18.57ToAsTcfhi tried lucview with no luck
23:19.29ToAsTcfhit would just give syntax errors when trying to use it
23:19.46ToAsTcfhalong with every other v4l2 app i found
23:20.02ToAsTcfhexcept ur vgrabber for v4l1
23:20.09phhbecause you don't know how to cross compile
23:20.14phhwhich v4l2 app have you found ?
23:21.26ToAsTcfhluvcview and
23:21.52*** join/#htc-linux m3dlg (~m3dlg@bb-87-81-252-83.ukonline.co.uk)
23:22.10ToAsTcfhshit i cant find the rest i deleted them. i think like v4lcongf or something
23:24.37ToAsTcfhvcam too
23:24.41phhToAsTcfh: luvcview relies on SDL
23:25.42phhvcam seems fine
23:28.30phhphone crashed \o/
23:28.49phhToAsTcfh: http://husson.hd.free.fr/vcapture if you want it.
23:28.56phh~ramconsole-diam
23:28.57aptfrom memory, ramconsole-diam is pwf dm 0x00800000 0x00100000
23:31.58phh[ 1096.671875] msm_enqueue: queue event new max is 1
23:32.00phhToAsTcfh: got that one ?
23:32.25ToAsTcfhno
23:32.30phhcool
23:32.37ToAsTcfhxcaptest vidpanel vcat vctrl
23:32.50ToAsTcfhthose are the others i tried though
23:32.53phhvcam is fine.
23:32.58ToAsTcfhok
23:34.05ToAsTcfh<6>[  813.867462] msm_enqueue: queue event new max is 1
23:34.05ToAsTcfh<6>[  813.867950] msm_camera: msm_enqueue: woke up event
23:34.05ToAsTcfh<6>[  813.869140] msm_camera: msm_ioctl_config: cmd 6
23:34.05ToAsTcfh<6>[  813.869567] msm_camera: msm_get_stats: will wait for timeout: -1
23:34.05ToAsTcfh<6>[  813.870086] msm_camera: msm_get_stats: list_empty_careful: 0
23:34.06ToAsTcfh<6>[  813.870391] msm_camera: msm_get_stats: returned from wait: 1073741822
23:34.07ToAsTcfh<6>[  813.871032] msm_camera: msm_get_stats: list_empty_careful: 0
23:34.10ToAsTcfh<6>[  813.871398] msm_camera: msm_get_stats: received from DSP 0
23:34.12ToAsTcfh<6>[  813.871704] msm_camera: msm_get_stats: qcmd->type 0 length 4, val = 0xcb729e24 (00000001)
23:34.24phhargh, you're after me
23:34.37ToAsTcfheverything actually works but vfetask
23:35.31ToAsTcfhu can always git it from our source and start from there
23:37.06ToAsTcfhhttp://github.com/jhansche/htc-2.6.27-heroc/commit/6e83cf8ca3a17210f122ad0bf60ea13b3b86021e
23:38.11phhgroumpf.
23:38.17ToAsTcfhwhat?
23:38.24phhI have to clone it
23:38.31ToAsTcfhlol
23:38.31phhmy hard drive will hate that.
23:38.50phhso, it's for after sleep, ++
23:39.02ToAsTcfhyeah im getting low on space myself
23:39.47*** join/#htc-linux jumoit (~samuel@121.77.87.145)
23:46.34ToAsTcfhphh: hold up
23:47.09ToAsTcfhphh:http://pastebin.com/d1817de77
23:47.34ToAsTcfhhttp://pastebin.com/d1817de77
23:48.26ToAsTcfhwith ur vcapture^^^
23:50.51ToAsTcfh# vcapture -d /dev/video20
23:50.52ToAsTcfhdevice name is /dev/video20
23:50.52ToAsTcfhv4l2: init
23:50.58ToAsTcfhthen it reboots
23:51.02ToAsTcfh:(
23:56.28ToAsTcfhvcapture -d /dev/video20
23:56.28ToAsTcfhdevice name is /dev/video20
23:56.28ToAsTcfhv4l2: init
23:56.28ToAsTcfhv4l2: device info:
23:56.28ToAsTcfh<PROTECTED>
23:56.35ToAsTcfhphh:^^^

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.