IRC log for #htc-linux on 20090819

00:00.47AstainHellbringwhats up tmzt
00:01.20glyphthe august 8 kaiser android build is the *bomb*
00:01.40glyphI used it all day today and it worked great for just about everything I use winmo for!
00:02.08glyphhaving a few minor issues though
00:02.39glyphI can't seem to mount the phone over USB and ... stranger, when someone calls me, the ringtone plays once, then stops
00:02.47glyphI can work around the latter thing by just using a really long ringtone
00:02.52glyphbut I'm kinda wondering if it's fixable :)
00:05.18*** join/#htc-linux darkstar62 (n=darkstar@97-126-104-157.tukw.qwest.net)
00:28.18*** join/#htc-linux marajin_ (n=marajin@87-194-102-189.bethere.co.uk)
00:41.29*** join/#htc-linux surge (i=surge@pool-98-118-154-21.bflony.fios.verizon.net)
00:49.49*** join/#htc-linux rashire2 (n=ed1112wa@pool-98-114-205-53.phlapa.fios.verizon.net)
02:15.19*** join/#htc-linux root2 (n=root@rgnb-5d87c058.pool.einsundeins.de)
02:16.05*** join/#htc-linux skodde (n=skodde@unaffiliated/skodde)
02:32.00*** join/#htc-linux rashire2 (n=ed1112wa@pool-98-114-211-161.phlapa.fios.verizon.net)
03:07.34*** join/#htc-linux AstainHellbring (n=AstainHe@unaffiliated/astainhellbring)
03:26.52*** join/#htc-linux Amaranth (n=travis@74.221.39.29)
03:37.36*** join/#htc-linux ImCoKeMaN (n=imcokema@pool-74-99-149-138.hrbgpa.fios.verizon.net)
03:42.02*** join/#htc-linux AstainKicksLTxda (n=AstainHe@unaffiliated/astainhellbring)
04:00.35*** join/#htc-linux Yuki (n=kaiba@m485736d0.tmodns.net)
04:05.23*** join/#htc-linux Shinto (n=John@f048053243.adsl.alicedsl.de)
04:14.47*** join/#htc-linux exception13 (n=exceptio@testdrive.kgts.ru)
04:14.53exception13all:
04:24.03*** join/#htc-linux stickboy (n=anonymou@74.68.145.150)
04:33.49*** join/#htc-linux goxboxlive (n=jrs@mail2.hjellnesconsult.no)
04:40.46*** join/#htc-linux droid0011 (n=mc@p4FDCEE35.dip.t-dialin.net)
04:41.08*** join/#htc-linux PoohbaLT (n=BlkPoohb@c-69-141-71-249.hsd1.nj.comcast.net)
04:47.51*** join/#htc-linux x29a_ (n=x29a@unaffiliated/x29a)
05:22.10*** part/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821)
05:42.33*** join/#htc-linux luc_ (n=luc@89-115-128-35.cl.ipv4ilink.net)
05:45.22*** join/#htc-linux Amaranth (n=travis@74-221-39-29.longlines.com)
05:46.40*** join/#htc-linux Amaranth_ (n=travis@ubuntu/member/Amaranth)
05:53.10*** join/#htc-linux skodde (n=skodde@unaffiliated/skodde)
05:55.11*** join/#htc-linux Amaranth_ (n=travis@74-221-39-29.longlines.com)
06:33.09*** join/#htc-linux Strid (n=Strid@ua-83-227-204-136.cust.bredbandsbolaget.se)
06:36.17*** join/#htc-linux Untouchab1e (n=Untoucha@111-36-8.connect.netcom.no)
06:37.15*** join/#htc-linux kiozen (n=oeichler@p54921421.dip0.t-ipconnect.de)
06:47.39*** join/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821)
06:55.00*** join/#htc-linux Hunterkll (n=kaiba@m4e5a36d0.tmodns.net)
06:56.38*** join/#htc-linux pleemans (n=toi@d51A49C45.access.telenet.be)
07:23.19*** join/#htc-linux kiozen (n=oeichler@p54921421.dip0.t-ipconnect.de)
07:25.36*** part/#htc-linux exception13 (n=exceptio@testdrive.kgts.ru)
08:03.27*** join/#htc-linux marajin (n=marajin@87-194-102-189.bethere.co.uk)
08:04.14*** join/#htc-linux ArteK (n=Artur@81.15.241.96)
10:07.33*** join/#htc-linux MethoS- (n=clemens@host-091-097-245-081.ewe-ip-backbone.de)
10:18.54*** join/#htc-linux Amaranth (n=travis@74-221-39-29.longlines.com)
10:22.49*** join/#htc-linux the_sys0p (n=the_sys0@cpe-67-49-192-228.bak.res.rr.com)
11:08.22*** join/#htc-linux Amaranth (n=travis@74-221-39-29.longlines.com)
11:36.26*** join/#htc-linux TigerTael (n=tr0n@196-210-186-232-wblv-esr-3.dynamic.isadsl.co.za)
11:36.28TigerTaelpmem pool /dev/pmem_camera ioctl(PMEM_GET_SIZE) is 8388608
11:36.32TigerTaelWhat's the equivalent usage of ioctl from the console?
11:56.30*** join/#htc-linux Strid (n=Strid@c-4f669719-74736162.cust.telenor.se)
12:03.12*** join/#htc-linux MethoS (n=clemens@host-091-097-246-044.ewe-ip-backbone.de)
12:56.38*** join/#htc-linux xray7224 (n=xray7224@unaffiliated/xray7224)
12:56.55*** part/#htc-linux xray7224 (n=xray7224@unaffiliated/xray7224)
13:32.33*** join/#htc-linux DJWillis (n=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk)
13:33.23*** join/#htc-linux madCoder` (n=madcoder@c-71-225-60-178.hsd1.pa.comcast.net)
14:06.01*** join/#htc-linux sdt555 (n=titus@147.145.40.44)
14:10.15*** part/#htc-linux TigerTael (n=tr0n@196-210-186-232-wblv-esr-3.dynamic.isadsl.co.za)
14:19.43*** part/#htc-linux sdt555 (n=titus@147.145.40.44)
14:25.50*** join/#htc-linux JuNu_KN (n=chatzill@p549ABA94.dip0.t-ipconnect.de)
14:35.14*** join/#htc-linux darkstar62 (n=darkstar@m670e36d0.tmodns.net)
15:00.34*** join/#htc-linux Gnutoo (n=gnutoo@ABordeaux-152-1-60-108.w82-125.abo.wanadoo.fr)
15:17.29Gnutoohi,does anyone knows angstrom,usb0 doesn't come up and I seem to have no keyboard on my htcdream,but all that work with SHR
15:25.27*** join/#htc-linux mugsie (n=mugsie@unaffiliated/mugsie)
15:34.20*** join/#htc-linux marex (n=marex@thor.hackndev.com)
15:45.34*** join/#htc-linux pH5 (n=ph5@e178195063.adsl.alicedsl.de)
16:00.20*** join/#htc-linux pleemans (n=toi@d54C2AAB7.access.telenet.be)
16:01.30*** join/#htc-linux luc_ (n=luc@89-115-128-35.cl.ipv4ilink.net)
16:42.26*** join/#htc-linux Zoolooc (n=fredsiba@p549526EC.dip0.t-ipconnect.de)
16:52.26*** join/#htc-linux rzk_ (n=rzk@daemonet.ru)
17:00.59*** join/#htc-linux x29a_ (n=x29a@unaffiliated/x29a)
17:35.10tmztGnutoo: well, it's not angstromspecific exactly
17:35.25Gnutoommm...I got usbnet but not dropbear
17:35.26tmztthe keyboad thing is confsuing though
17:36.00tmztok
17:36.16tmztyou added usb0 to /etc/netowrk/interfaces?
17:36.22Gnutoowhat's confsuing?
17:36.26Gnutooit's in it
17:36.29GnutooI'll debut it later
17:36.31tmztwhy it wouldn't work
17:36.38tmztand auto usb0?
17:36.56GnutooI'll look when I get the sd-card out of the phone
17:37.10GnutooI'll wait a bit for dropbear...it has to generate keys
17:43.29*** join/#htc-linux darkstar62 (n=darkstar@m670e36d0.tmodns.net)
17:45.01tmztsame kernel?
17:46.00Gnutooopenmoko-msm-2.6.29
17:46.15Gnutoobtw how's going the kenrel repo on linuxtogo
17:46.16Gnutoo?
17:47.50tmztah
17:47.55GnutooI'm tempted from copying openmoko's dropbear config
17:48.01Gnutoolike the generated keys etc...
17:48.07tmztcan you explain what the wireless merges are about?
17:48.33GnutooI didn't look at wireless at all
17:48.39GnutooI wanted touchscreen before?
17:48.44Gnutoos/?//
17:48.49tmztin openmoko-msm I mean
17:48.59tmztlook at the git log
17:49.11GnutooI didn't look at wireless in this kernel...I just disabled it because it didn't cross-compile
17:49.20Gnutoobut...I know they are merging the driver in mainline
17:49.25tmztit's full of commits from some linux-wireless tree
17:49.30Gnutoothat is to say using mac80211
17:49.36tmztyeah
17:49.47Gnutooyes that's the tree where all the wireless development occur
17:49.51tmztbut we need a simple/clean history
17:50.06Gnutoook
17:50.11tmztcan you give your gitweb?
17:50.21Gnutoo???
17:50.51GnutooI don't understand what you mean by giving my gitweb
17:51.10tmztit would be very hard to track down problems caused by an alpha upstream tree, you should be able to import just the 80211 stack
17:51.14tmzturl
17:51.45Gnutoobut the 80211 is already in mainline
17:51.58GnutooI don't know if wireless-dev one is needed
17:52.09Gnutoobut we don't have the driver in the mainline
17:52.15Gnutoothat is wl12xx
17:52.25tmztyes, but the g1 android git is not mainline (current)
17:52.29tmztright
17:52.31Gnutoook
17:52.42Gnutoobut mac80211 is old
17:52.51Gnutooit is definitely in android's kernel
17:52.52tmztyeah
17:52.59Gnutoos/it is/it should be
17:53.00tmztwhat is?
17:53.08GnutooI'll look
17:53.40GnutooCONFIG_MAC80211
17:53.51*** join/#htc-linux dcordes (n=luke@unaffiliated/dcordes)
17:53.52Gnutooit's in the old android one
17:53.56tmztyeah
17:54.02tmztbut it's old
17:54.04tmztright
17:54.05Gnutooyes
17:54.09tmztdcordes: hey
17:54.14Gnutoohi dcordes
17:54.24dcordeshi Gnutoo, tmzt others
17:55.24Gnutoommm....maybe sh> netcat ?
17:55.33Gnutooif dropbear doesn't work
17:55.45tmztmaybe
17:55.59tmztwhy shouldn't it work thoygh?
17:56.20tmztwhat's your cmdline?
17:56.21tmztkernel vmdline
17:56.25tmztcmdline
17:57.30GnutooI'll look but it's a dropbear problem
17:57.34Gnutoonot a CMDLINE one
17:59.22tmztyeah
17:59.45tmztbut this is a lot easier if keyboard works
18:00.12tmztah
18:00.32tmztso these commit are from android 2.6.29?
18:01.34tmztwhat are the chaneges in asm?
18:02.04tmztthat's very risky to do, very very hard to debug without serial port
18:02.20Gnutoorisky no
18:02.20Gnutoobut hard to debug yes
18:02.27Gnutoowhat commits
18:02.28Gnutoo?
18:02.33Gnutoothe ones for the wifi?
18:02.36Gnutooif so yes
18:02.50Gnutooin asm the mach name changed
18:04.47tmztit looks like you upgraded the kernel locally and committed it though
18:05.49Gnutoo???
18:06.02GnutooI only have commit access in oe
18:06.09Gnutooand I didn't commit a kernel
18:06.19tmzton your gitweb
18:06.25Gnutoowhat gitweb?
18:06.25tmztopenmoko-msm
18:06.33GnutooI've not access to this repo
18:06.33tmztthat's not you
18:06.36tmzt?
18:06.36Gnutoono
18:06.40GnutooI'll look
18:06.43tmztoh, sorry
18:06.45GnutooI didn't do git push
18:07.12tmztso what kernel are you asking about on ltg?
18:07.23tmzthtc-msm-2.6.29
18:07.37tmztdcordes: whose tree is openmoko-msm?
18:07.49Gnutooit's leviathan's
18:07.51Gnutoonot mine
18:08.05GnutooI've no public personal msm tree
18:08.41tmztok
18:08.55tmztyeah, leviathan
18:10.07Gnutooah ok Exited: couldn't open random device
18:10.11Gnutoonetcat now works
18:10.28*** join/#htc-linux darkstar62 (n=darkstar@m670e36d0.tmodns.net)
18:10.33Gnutoomaybe it used mdev without the fix
18:10.39GnutooI should commit a workarround fix
18:11.06Gnutoobecause the discussion about the fix didn't automagically solved the problems
18:11.09tmztyeah
18:11.16tmztthat makes sense
18:11.30tmztwhat fix?
18:12.27Gnutoothere was an empty conf file
18:12.29Gnutooin oe
18:13.22Gnutoo~/oe/org.openembedded.dev/recipes/mdev/mdev-1.2.1/mdev.conf
18:13.31Gnutoobut it's not empty in my tree
18:13.36Gnutooso I added it
18:13.44GnutooI must commit that and some others fixes
18:14.21dcordes~seen pH5
18:14.25aptph5 is currently on #maemo (2h 28m 51s) #htc-linux (2h 28m 51s) #oe (2h 28m 51s) #openmoko (2h 28m 51s) #openezx (2h 28m 51s). Has said a total of 4 messages. Is idling for 2h 10m 34s, last said: 'zecke: unfortunately, no'.
18:14.36dcordespH5, hi
18:24.28Gnutoommm so what should we do for the kernel?
18:25.07Gnutoowas indeed udev...didn't start corectly
18:27.37Gnutoommm....strange: http://www.pastebin.ca/1535381
18:30.13Gnutoowas udev again
18:30.52Gnutooah ok devpts.sh
18:40.41Gnutoommm http://www.pastebin.ca/1535404
18:40.45Gnutoodoesn't crash anymore
18:41.06Gnutoobut still black screen
18:41.11GnutooI'll add some driver
18:41.12Gnutoos
18:41.47*** join/#htc-linux stickboy (n=anonymou@198.105.45.141)
18:54.56*** join/#htc-linux Echo31 (n=olivier@mir31-4-82-240-194-54.fbx.proxad.net)
19:16.39*** join/#htc-linux kiozen (n=kiozen@rgnb-5d87c480.pool.einsundeins.de)
19:20.25*** join/#htc-linux marajin_ (n=marajin@87-194-102-189.bethere.co.uk)
19:24.46*** join/#htc-linux pleemans (n=toi@d54C2AAB7.access.telenet.be)
19:33.18*** join/#htc-linux Strid (n=Strid@ua-83-227-204-136.cust.bredbandsbolaget.se)
19:51.16*** join/#htc-linux szsoftware (n=sven@e178212127.adsl.alicedsl.de)
19:54.39*** join/#htc-linux cr2 (n=cr2@ip-77-25-233-54.web.vodafone.de)
19:55.35cr2hi
19:55.54cr2back at home now :)
20:00.02luc_wow
20:00.06luc_welcome
20:00.13tmzthey
20:00.15luc_from nasa you are back??
20:02.24cr2i was traveling last week, and have even reached Canada
20:03.36cr2+       if(is_msm7200()) { // Looks like msm7200 module table is different so do some conversion here!
20:03.38cr2+               if(module==3) module=4; // PPTASK
20:03.39cr2+               if(module==12) module=0x13; // RECTASK
20:03.53cr2this is in the vogue tree. looks familiar ;)
20:04.15*** join/#htc-linux MethoS- (n=clemens@host-091-097-246-204.ewe-ip-backbone.de)
20:05.26*** join/#htc-linux BabelO (n=fcr@lun34-2-82-238-28-28.fbx.proxad.net)
20:05.30Echo31cr2: hi
20:05.56cr2hi Echo31
20:06.14cr2Echo31: got any progrees with cpld1 init values ?
20:07.07Echo31cr2: yes, cpld1 works now
20:07.15cr2well, i don't see anything spectacular in the vogue tree too
20:07.22cr2Echo31: cpld2 ?
20:07.40cr2Echo31: which initial values do you have for cpld1 ?
20:07.41Echo31cr2: it is ok
20:09.01cr2Echo31: is your patch merged into pH5 tree ?
20:09.30Echo31cr2: the value of register 4 must be 0x4   ( the value does not find from wiki)
20:09.46cr2tmzt: i've got some new DEX ids
20:09.46Echo31cr2: not yet
20:09.58cr2Echo31: edit the wiki ?
20:10.01tmztok, for?
20:10.26cr2tmzt: for quering the single battery values
20:10.58cr2tmzt: without messing with the 2/4 bytes and different locations for the "unified" DEX call
20:11.07cr2but we need to test them first
20:11.11tmztcr2: it seems the para values in smem are not the actual configuration
20:11.28cr2tmzt: where ?
20:11.29tmztoh, haven't done anything with bary
20:11.43tmztauidoparas I mean
20:11.51tmztbattery
20:12.06tmztaudioparams
20:12.11cr2audio paras are a hell of a new stuff for us :)
20:12.24tmztso the call is to push them to dsp
20:12.33Echo31cr2: the keyboard and the touchscreen work by SPI
20:12.42tmztwhat do you mean?
20:13.06cr2tmzt: ok.
20:13.09cr2tmzt: we don
20:13.29cr2t have support infrastructure for the IIR & friends in the kernel
20:13.44tmztiir?
20:13.53cr2Echo31: but these are different spi channels
20:14.00cr2tmzt: digital filter
20:14.07tmztok
20:14.16tmztwhat do we need for that?
20:14.21tmztwhat kernel
20:14.23cr2tmzt: for audio
20:14.29Echo31cr2: Yes the first and the second
20:14.32cr27201A
20:14.33tmztmagic, hero are 2.6.29
20:14.54cr2tmzt: can we switch to .29 ?
20:15.08tmztwe were talking about getting a tree with working raph, etc. and trout
20:15.19cr2tmzt: i don't see any difference for "my" work, but why not
20:15.26cr2ok
20:15.39tmztthen gnutoo would do baseline comparisons on g1
20:15.52cr2Echo31: does the kbd work for you ?
20:15.53tmztwe also need some code review in place
20:16.11tmztlast commits on htc-msm broke a few devices
20:16.35cr2tmzt: i don't mind if anybody does the code cleanup
20:16.48cr2hmm. i saw the BT clock revert too
20:17.00tmztit's more than that, patches need to go to a list
20:17.13cr2but these changes make a lot of sense. maybe i'd just fix them.
20:17.13Echo31cr2: yes, the keyboard works well. i implemented the hook for uart (GPS,BT) but i don't find from the wiki, the proper values
20:17.15tmztyeah, cleanup :)
20:17.34tmztcan we get a TODO in wiki or somewhere for that?
20:17.51cr2tmzt: yes, why not
20:18.31tmztwhat do our changes look like versus google's stuff?
20:18.45tmztwhat is shared between their devices and ours?
20:18.46Echo31cr2: I implemented the DPRAM but also I don't find the proper values specific for 3 pairs channels
20:19.14tmztI know we did everything on halibut originally
20:19.21cr2tmzt: the drivers, and some silly goggle stuff
20:19.30tmztI think that's been cleaned up
20:19.35tmztI mean boards
20:19.44cr2tmzt: we need to push some block changes there.
20:19.51tmztthe drivers we can share with them
20:19.53tmzt??
20:19.59cr2tmzt: the gpio ALT/MFD whatever you call it
20:20.07tmztah yeah
20:20.27cr2and the whole clock-wince stuff
20:20.33tmztwe need to work on an email for lkml/Pavel maybe
20:21.01cr2he has different AMSS
20:21.04Echo31cr2: i  am going to attack the gsensor by i2c (from modification of htc shift gsensor)
20:21.24tmztso we can move in the right direction, test our changes on 29
20:21.27cr2Echo31: what you don't understand for uarts ?
20:21.27tmztyeah
20:21.43cr2Echo31: and what is unknown for hte DPRAM ?
20:21.50tmztI mean gpio_ api abstraction for both amss's
20:22.26cr2Echo31: also look at the moko2 gsensor driver.
20:22.43cr2tmzt: is the g1 gpio area protected by mpu ?
20:23.30cr2tmzt: i mean if g1 can manipulate the (arm9) gpios directly without any weird google rpc api ?
20:24.18cr2the gpio setup looks pretty static, and i think we are not going to hurt amss in any way
20:25.25cr2tmzt: IMVHO it's better to push Pavel into this direction : -> direct hw control instead of google innovations
20:26.37Echo31cr2: i took and ported the the old patch (hh.org) of pH5 for hooking the BTuart FFuart (GPS)
20:28.29*** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl)
20:30.03BabelO<PROTECTED>
20:30.12BabelOcr2: this page is no more here :( http://wiki.xda-developers.com/index.php?pagename=ArtemisGPIO
20:31.17tmztcr2: I don't know about mpu
20:31.29tmztI guess we should try it
20:32.00tmztsame for clocks?
20:32.25Echo31cr2: for the implementation of DPRAM ( from blueAngel GPRS), i don't find the IRQ_ATHENA_DPRAM_RX_FULL and  IRQ_ATHENA_DPRAM_TX_EMPTY
20:39.17Echo31cr2: for gsensor, the source modified are written by  Pau Oliva Fora. The device is correctly attached. The communication i2c is ok. But the accelometer events polling don't work (!!?)
20:44.09cr2Echo31: can you trace the i2c gsensor traffic ?
20:44.55cr2tmzt: the clocks are a more complex issue. we need somebody with g1 who can compile his own kernel with our clock modifications
20:45.03cr2hi BabelO
20:45.26cr2BabelO: gogel cach ?
20:46.01cr2Echo31: i think pH5 does the uart power with rfkill now
20:46.27cr2Echo31: so you only need to put the right gpios in the right rfkill places
20:46.47BabelOcr2: no, no google cache for this one :(
20:47.32cr2yeah, i see it
20:47.47cr2BabelO: but most of the gpios were from the spl
20:48.04BabelOcr2: yes i ve to redo
20:48.12cr2BabelO: and the alt settings can be checked agains the .pdf
20:48.19BabelOyes
20:48.23cr2BabelO: htc-linux.org wiki ?
20:48.39BabelOcr2: the linwizard gsm driver works on artemis
20:48.49cr2with sound ?
20:48.56BabelOcr2: so i want to fix the syren stuff to set bt and other for sound
20:49.02BabelOmic works
20:49.07BabelObut not speaker
20:49.08dcordescr2, welcome back
20:49.13BabelOi ve to found the init
20:49.31cr2BabelO: ok
20:49.34cr2hi dcordes
20:49.46cr2BabelO: it must be in the spl
20:49.56BabelOcr2: other thing :) you have to download M with the installer :)
20:50.05BabelOcr2: yes i see them in spl
20:50.11cr2BabelO: but afair it, the syren setup is big and complex
20:50.12BabelOjust have to do the code :)
20:50.16cr2ok
20:50.37cr2BabelO: didn't see kiozen here for a long time
20:50.48BabelOcr2: on M channel :)
20:50.54BabelOcr2: on qlandkarte channel
20:51.00cr2BabelO: which devices would you like me to check
20:51.09cr2ok
20:51.10BabelOcr2: both :)
20:51.16BabelO640x480 and 320x240
20:51.21cr2i have more
20:51.29BabelOcr2: user guide here http://qlandkarte.org/index.php?option=com_content&view=article&id=13&Itemid=14
20:51.34cr2240x240 too :)
20:51.55cr2and many wince versions
20:52.18BabelOcr2: http://sourceforge.net/projects/qlandkartegt/files/
20:52.30BabelOcr2: in the qlandkartem section file you have the bioth download
20:52.51cr2BabelO: ok, i'll try :)
20:53.18Echo31cr2: i found the old driver (2.6.14.6) of Epson S1R72V05 near the S1R72V03 of Athena
20:53.34cr2Echo31: afair it's very different
20:53.45cr2Echo31: and we have the docs for the V03
20:54.22cr2V05 too, but i think the chip is very different
20:54.22Echo31Tcr2: the ide part is same
20:54.25cr2Echo31: let's fix the BT and GPS first. i don't see any reason for them not to work
20:54.40cr2Echo31: are you sure ?
20:54.48cr2Echo31: about ide
20:55.52Echo31cr2: the comments into the source code works  for whole familly. ( ATA IDE controller)
20:56.54cr2Echo31: try it.
20:57.03cr2Echo31: but i have some doubts
20:57.36cr2Echo31: the SD may be easier
20:58.47Echo31cr2: the calls of ide_hwif_t strucure are different than new kernel
21:00.12cr2Echo31: ok, i can't really say anything well-informed about it. unlike the SD .
21:00.52cr2Echo31: the BA DPRAM has only 1 channel, on athena you have many channels
21:01.03Echo31cr2: what is your mail address to send to you this driver code?
21:01.53cr2Echo31: the RX gpio irqs are documented, and for TX you need to flip some other gpios. i think it's all documented in the wiki
21:01.55Echo31cr2: yes, athena has 3 dual channels
21:02.03cr2Echo31: @htc-linux.org
21:02.26*** join/#htc-linux AstainHellbring (n=AstainHe@unaffiliated/astainhellbring)
21:02.42cr2Echo31: 1 channel is more tricky, because it writes the magic value "CE" into some weird register, but i don't know if we really need it
21:03.10cr2Echo31: i'd be happy with the AT channel first. then we can extend it to the UMTS channel
21:03.15glyphoh hello!  people!
21:04.05glyphI see the topic exhorts me to go to #android for android-centered discussion, but I have some problems that are rather specific to android on my Kaiser, which G1 owners would look at me funny about :)
21:04.17glyphshould I discuss issues with androidonhtc.com builds here or there?
21:05.01cr2glyph: we are talking more about kernel issues here
21:05.22cr2glyph: i've even never tried android myself :)
21:05.24Echo31cr2: the DPRAM driver mounts the ttyS8. normally, now, i need only  both irqs
21:06.43cr2dcordes: why did you revert the BT clock patch ? i've carefully read the discussion in the logs, and this patch makes a lot of sense to me
21:07.21ali1234glyph: if it is the same issue from the other day you should report it to androidonhtc.com (which is badly named btw)
21:07.32cr2Echo31: i think that you don't need the TX overrun irq
21:07.38cr2Echo31: on athena
21:08.15AstainHellbringhi
21:08.50cr2hi AstainHellbring
21:09.09AstainHellbringwelcome back
21:09.20dcordescr2, it broke the raphael build the way it was commited.
21:10.28cr2dcordes: can you paste the error log ?
21:11.11dcordescr2, yes I need to reconstruct
21:11.24cr2dcordes: the patch just changes the clock source bitmask ( &7 instead of &3 ) and adds some more clock entries. i don't see what can go wrong ther.
21:12.58cr2dcordes: BT still does not work for me with this patch too, but at least i have understood the rationale behind it
21:13.01dcordescr2, must be some syntax error then
21:14.09cr2AstainHellbring: yeah, i need to attack the rpc/adsp problem...
21:15.31Echo31cr2: gpio_to_irq(GPIO51_HTCATHENA_DPR_IRQ_AT) does give it, the RX irq ?
21:15.34*** join/#htc-linux BHSPitLappy (n=BHSPitLa@unaffiliated/bhspitmonkey)
21:15.51cr2Echo31: i'll check
21:17.19cr2Echo31: yes
21:17.21*** join/#htc-linux swc|666 (n=infidel2@unaffiliated/swc666/x-4934821)
21:17.31cr2GPIO 51   00080000   DPR-AT   I   FE   1
21:18.50cr2if you get this irq, you just need to memcpy the head,tail-head into rx buffer
21:18.51glyphali1234: I'm well aware that it's badly named, since the only android phones anybody actually uses are HTC anyway :)
21:18.51glyphcr2: what sort of linuxy environment do you run on your HTC devices?
21:18.59ali1234glyph: not only that, but androidonhtc.com only supports msm phones
21:19.34ali1234glyph: there is another totally unrelated project for android on htc omap-based phones
21:19.56cr2glyph: a barebones angstrom ramdisk
21:20.19dcordesglyph, are you the androidonhtc person?
21:21.44glyphdcordes: I'm just a guy trying to run android, and therefore linux, on my phone :)
21:21.48tmztthat's Untouchable isn't it?
21:21.51cr2tmzt: it may be interesting to discuss some things with androids, but i expect a polite "fuck off, it's all NDA area here"
21:22.28tmztI usually get "read the source"
21:23.42cr2tmzt: the source has some magic values hardcoded. take the LCD mddi speed as an example
21:24.01tmztyeah, that
21:24.02*** part/#htc-linux ArteK (n=Artur@81.15.241.96)
21:24.06tmztI thought you meant qdsp
21:24.29tmztsomebody there asked which qdsp version we had or something like that, but I haven't heard from him in a while
21:24.51Echo31cr2: for i2S sound , why i cannot use the AK4641 driver with the same way that  sound/soc/pxa/hx4700?
21:24.57tmztI imagine they were just trying to get it to work
21:25.11cr2tmzt: our wince devices use a different PLL to source the clock, and a different speed. and we more or less know how to set it. but i think that only some $qualified_qualcomm_engineer can tell us why it is so
21:25.17*** join/#htc-linux Strid (n=Strid@ua-83-227-204-136.cust.bredbandsbolaget.se)
21:25.17tmztEcho31: you need eac I think
21:25.35tmzt#linwizard and #wing-linux people have been trying to get that to work
21:25.56cr2Echo31: the ak4641 i2s is not used afair
21:25.59tmztcr2: what, the mddi bus speed?
21:26.04tmztnot pixclock right?
21:26.19cr2tmzt: the mddi bus speed
21:26.24tmztyeah
21:26.25tmztok
21:26.38tmztwhich clock is sourced differently? pclk?
21:27.01ali1234best i can tell EAC means nothing. I2S is a real "standard" though
21:27.06cr2Echo31: the i2s datastream is sourced from ATI DSP (I2S)
21:27.55cr2tmzt: there are (at least) 3 PLLs where you can source the clocks
21:28.17tmztI guess parent is used for the clk source in linux, I didn't know that
21:28.27tmztso we do have clk tree api and google does use it
21:28.38tmztbut they don't expose all the details
21:29.12cr275 static struct clkctl_acpu_speed  acpu_freq_tbl[] = {
21:29.13tmztyeah, they can't talk about clock internals, swetland even commented they might not have the documentation
21:29.14cr276         { 19200, ACPU_PLL_TCXO, 0, 0, 19200, 0, VDD_0, 0, 0, 8 },
21:29.43tmztwhat's the issue with mddi?
21:29.48tmztjust trying to make it more generic?
21:30.09cr2we have gathered enough data about the clocks
21:30.40cr2no, this table is too CPU specific
21:31.08cr2and it includes the voltages, where we don't care about
21:31.43cr2but the core code logic is the same:
21:32.14tmztwe have source for msm7200A'st no w
21:32.24tmztnow
21:32.44cr2you have the TCX0=19.2MHz and PLL0, PLL1 and PLL2 which are controlled by a multiplier +M7
21:32.49cr2M/N settings
21:33.35cr2the SD/wifi, MDDI and BT clocks (at least) are derived from these.
21:33.56cr2and this part is more or less understood.
21:34.40cr2the ARM11 cpu also can source these 4 sources, as seen in the table above
21:35.10cr2but then we get the g1 vs. wince differences
21:35.16Echo31cr2: i have got an old code for ch7013_I2C.c  and for tps65023_i2c (OMAP)
21:35.20tmztthose are a-clocks?
21:35.29cr2the MDDI on wince uses a different PLL and has faster speed
21:36.00tmztis that different because of 18-bit instead of 16-bit?
21:36.07cr2tmzt: yes. the non-A people need to do their own research here, but it's not going to be very different
21:36.39cr2no, the 18bit vs 16bit should be the mddi client issue.
21:36.45cr2hmm. maybe.
21:37.06cr2tmzt: do you  have the CDMA raph ?
21:37.22*** join/#htc-linux maejrep (n=madCoder@68.64.144.211)
21:37.23cr2can you post your CLK dump into wiki ?
21:37.31tmztyeah
21:37.33tmzthow?
21:37.34cr2hey maejrep
21:37.46tmzthe's mostly working
21:37.49cr2tmzt: pd 0xa8600000 0x400
21:37.53tmztok
21:38.05cr2tmzt: well, i work now too.
21:39.07cr2tmzt: the raph100 dump is here http://www.htc-linux.org/wiki/index.php?title=MSM_CLK
21:40.54cr2a8600300 | 00000007 0000000c 00000004 00000005 | ................
21:40.55cr2a8600310 | 00003000 00010480 00000005 00000007 | .0..............
21:40.57cr2a8600320 | 00000028 00000000 00000001 00003000 | (............0..
21:40.58cr2a8600330 | 00010580 00000008 00000007 00000037 | ............7...
21:41.00cr2a8600340 | 00000000 00000001 00002000 00010580 | ......... ......
21:41.01cr2a8600350 | 0000000b 00000000 0000002d 00000000 | ........-.......
21:41.03cr2a8600360 | 00000001 00000000 00010580 00000000 | ................
21:41.05cr2these are raph100 PLL settings
21:41.54cr2Echo31: the ch7013 is the TVout ?
21:42.12cr2Echo31: the tps needs to be i2c-traced too.
21:43.13tmzthttp://pastebin.com/m261e113a
21:43.13cr2Echo31: i think the TVout chip is not the one documented in wiki, and it uses gpio i2c bitbang to operate. it's going to be a PITA to impleemnt fully
21:43.38tmztjust the encoder?
21:43.46tmztmost of the chrontel chips are documented
21:43.55cr2tmzt: which device is that ?
21:44.02cr2tmzt: raph500 ?
21:44.05tmzt21:56 < cr2> Echo31: the ch7013 is the TVout ?
21:44.07tmztyeah
21:44.46cr2tmzt: i've checked the driver disassembly with the ch7013 .pdf, and it has a different i2c id.
21:48.19cr2tmzt: afair the raph800 mddi clock was different
21:48.37cr2a8600300 | 00000007 0000000a 00000006 00000019 | ................
21:48.38cr2a8600310 | 00003000 00010480 00000004 00000007 | .0..............
21:48.40cr2a8600320 | 00000028 00000000 00000001 00003000 | (............0..
21:48.41cr2a8600330 | 00010580 00000008 00000007 00000037 | ............7...
21:48.43cr2a8600340 | 00000000 00000001 00002000 00010580 | ......... ......
21:48.44cr2a8600350 | 0000000b 00000000 0000002d 00000000 | ........-.......
21:48.46cr2a8600360 | 00000001 00000000 00010580 00000000 | ................
21:49.16cr2[23:40] <cr2> a8600310 | 00003000 00010480 00000005 00000007 | .0..............
21:49.25cr2[23:48] <cr2> a8600310 | 00003000 00010480 00000004 00000007 | .0..............
21:49.36cr24 vs. 5 here for the PLLs
21:50.08cr2need to check the logs what does it mean
21:50.15tmztok
21:50.22tmztraph800 kernel works fine though, even video
21:50.24cr2tmzt: it's probably something for the haretconsole
21:50.31tmztwhat do you mean?
21:50.40tmztI can pwf
21:50.46cr2hehe. if you don't touch the wince setup , that is.
21:50.58tmztah
21:51.01tmztwell that could be
21:51.14cr2haretconsole can parse these values, and report the CLK speeds and divisors
21:51.29cr2and multipliers for the PLLs
21:51.47cr2i can't just keep all this stuff in my head. and in the logs :)
21:52.01tmzttrue
21:52.07tmztharetconsole has that support now?
21:52.25Echo31tmzt: unfortunately, the ch7013 is the TVout.
21:52.33cr2you'd know better. i've not heard from Kevin2 for some time
21:53.03tmztEcho31: yeah, but that's a chrontel chip right?
21:53.09cr2tmzt: but it can be also done with "default.txt". just less usef-friendly
21:55.33cr2tmzt: btw, is there any project to add the 7200A support to uboot ?
21:55.43tmztnot that I know of
21:55.50tmztI think raster was thinking about, not sure
21:56.10tmztI don't think uboot is the best end user bootloader
21:56.23tmztvery hard to configure
21:56.55cr2yeah
22:00.07cr2[Di Jul 7 2009] [11:50:56] <cr2>        tmzt: the PLL1 has mode=7 L=0x28 M=0 N=1
22:00.09cr2[Di Jul 7 2009] [11:51:26] <cr2>        tmzt: since PLL1=768MHz it looks like PLL1=L*TCX0
22:00.10cr2[Di Jul 7 2009] [11:51:56] <cr2>        so we can derive the SD clocks from the "first principles"
22:00.12cr2[Di Jul 7 2009] [11:52:24] <tuxhero>    i think we need to reset arm9 for asdp
22:00.13cr2[Di Jul 7 2009] [11:52:42] <cr2>        PLL0 is more difficult: mode=7 L=0xc M=4 N=5
22:01.28cr2[Fr Jul 17 2009] [22:01:21] <q_tpro_>   Captnoord: the msm7201A clocks are a mess, but i've understood some things about it
22:01.29cr2[Fr Jul 17 2009] [22:01:46] <q_tpro_>   there are 3 PLLs and TCX0
22:01.30cr2[Fr Jul 17 2009] [22:02:05] <q_tpro_>   PLL0, PLL1, PLL2 and TCX0
22:01.32cr2[Fr Jul 17 2009] [22:02:35] <q_tpro_>   PLLx are derived from TCX0 (MSM_CLK+0x3xx)
22:01.56tmzt22:14 < cr2> [Di Jul 7 2009] [11:52:24] <tuxhero>    i think we need to reset  arm9 for asdp
22:02.07tmztwhat did he mean?
22:02.26cr2it's impossible
22:02.42tmztyeah
22:03.15cr2we need to notify the arm9 that the arm11 wince has a total failure, and we need to reinit the rpc endpoints
22:03.58cr2otherwise we still get the arm9 messages to dead wince endpoints (from my logs)
22:04.43tmztyeah
22:04.49tmztI think that's right
22:05.04tmztbut I don't know how to restart it
22:05.06cr2dcordes: [Fr Jul 17 2009] [22:06:21] <q_tpro_>   btw, the current PWR_OFF DEX call should be moved into pm.c
22:05.54cr2tmzt: i'd like to know what is this "second" endpoint is doing
22:06.11tmztwhich endpoint?
22:06.15cr2tmzt: the first one was the rpc route, and we reinit it already
22:06.24cr2s/route/router/
22:07.56cr2[Fr Jul 17 2009] [21:58:25] <q_tpro>    if we will use the 0xa19 value (on raph100 and blac100) it's not going to be PLL0...
22:08.20cr2ok, this one refers to the MDDI clock differences from raph100 vs. raph800
22:10.02Echo31cr2: what is the channel used (spi, sdio) of the wl1250 ?  with  the porting from  wl1251_spi driver
22:10.04cr2[Di Jul 7 2009] [08:33:16] <cr2>        [08:23] <cr2> Line: 1422. rpc_router_core rpc_router_open'ing: Processor:0x00000001, Process:0xfffffffe Handle:0x00000000.M
22:10.05cr2[Di Jul 7 2009] [08:33:22] <cr2>        [08:24] <cr2> Line: 1422. rpc_router_core rpc_router_open'ing: Processor:0x00000001, Process:0x0feb8f4a Handle:0x00000000.M
22:10.41tmztEcho31: there's already a sdio port of wl1251
22:10.47cr2tmzt: "Processor:0x00000001, Process:0xfffffffe"  is the rpc router on arm11
22:10.50tmztEcho31: which does your device use?
22:10.55tmztyeah, I think so
22:11.06tmztit's mplemented in the kernel
22:11.47cr2Echo31: the sdio part is used only for the chip config, eeprom and wifi hw address. suh things
22:12.16cr2tmzt: but what is "Process:0x0feb8f4a" on arm11
22:12.39cr2tmzt: the arm9 still sends some messages to it, even after wince is already dead.
22:13.29tmztah, I think we figured that one out before
22:13.37tmztnot how to stop it, but what it is
22:13.51cr2?
22:13.59Echo31cr2: ok , the sdio + SPI are needed
22:13.59tmzttrying to remember :)
22:14.24cr2Echo31: it's the pxa mmc port, not spi
22:14.42tmztso ti's sdio
22:14.42cr2Echo31: so you have the driver already.
22:14.46tmztit's
22:15.33cr2Echo31: wifi on athena will use pxamci
22:16.46cr2tmzt: if i'll know what is the purpose of this *ept Process:0x0feb8f4a, we may think how to reinit it.
22:20.39*** join/#htc-linux madCoder- (n=madcoder@c-71-225-60-178.hsd1.pa.comcast.net)
22:21.49*** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl)
22:21.58Kevin2cr2: I'm still around.  I just haven't had time to do anything with mobile devices.
22:22.04*** join/#htc-linux lvh_ (n=lvh@83.101.78.37)
22:22.06lvh_hello :-)
22:23.47tmzthttp://tibook.jb.b4m.com/j0b0/raph100-info/raph100-wince-dmesg.txt
22:24.06lvh_I've managed to get an old HTC Prophet (Qtek S200)
22:24.08*** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl)
22:24.31*** join/#htc-linux Squarc (n=Squarc@82-217-32-29.cable.quicknet.nl)
22:24.37lvh_I'm developing mobile software for it (the platform, not necessarily winmo of course)
22:24.44lvh_I can't find information about using the builtin GPRS modem
22:25.32tmztomap850?
22:25.37tmztor that's xda
22:25.40tmztpxa
22:28.16dcordestmzt, omap850 linwizard kernel boots on it
22:28.40tmztok, yeah
22:28.58tmztlvh_: #linwizard should be supporting it on candidate-android now, gprs also works
22:33.15lvh_awesome!
22:33.18lvh_it's omap
22:33.26lvh_Thank you very much! :-)
23:22.27*** join/#htc-linux stickboy (n=anonymou@ool-45767959.dyn.optonline.net)

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