IRC log for #htc-linux on 20130213

00:10.12*** join/#htc-linux BabelO (~fcr@2a01:e35:8adc:3150:21e:8cff:fe2d:7d3f)
00:10.12*** join/#htc-linux BabelO (~fcr@unaffiliated/babelo)
01:10.43*** join/#htc-linux d3tul3 (~detule@unaffiliated/d3tul3)
01:45.33jonpryd3tul3, looks like you have been having some fun on anthrax
01:46.42d3tul3no it's not me
01:46.56d3tul3folks are starting stuff in my name - ha i'm popular
01:47.15jonprymr goodman seems to know who you are
01:47.35jonprybig surprise
01:49.45d3tul3i mean it's a lost battle there....he's covered all possible angles....i am not going to give it much thought....i back to posting patches on linux-msm-d2 he can kang all he wants
01:50.03jonpryi'd sue him
01:50.06d3tul3it doesn't quite sit right with me that he's making money off my work, but i guess there's sheep out there...
01:50.13jonpryyour code, gpl, no compliance
01:50.53jonpryalthough that could get expensive
01:50.57d3tul3lol
01:51.19d3tul3if i got in a law suit over my computer hobby wife would divorce me
01:51.30jonpryhaha
01:54.10arrrghhhand as much as everyone wanted to sue tiad8 into oblivion
01:54.20arrrghhhit's just not worth the time, effort, or money.
01:54.49d3tul3arrrghhh, how goes the kernel
01:55.02arrrghhhkernel is alright
01:55.17arrrghhhsome guy started working on cm10, knows way more than me.
01:55.26arrrghhhalready has it booting, cam works, wifi, bt, etc.
01:55.38jonprywhat device?
01:55.44arrrghhhI am such a tard tho, I can't figure out the CDMA crap.  He uses the photon q as a GSM device only.
01:55.47arrrghhhQ 6^^
01:55.53d3tul3arrrghhh's got msm8960 too!
01:55.56d3tul3more hexagon!
01:55.56arrrghhhyay
01:56.02jonprytrustzone :(
01:56.06arrrghhhunlocked bootloader
01:56.11arrrghhhpour moi
01:56.48jonprysomehow it still has locked bootloader for hexagon
01:56.54jonpryat least on n4
01:58.20arrrghhhyea, moto released an official 'kiss your warranty goodbye' unlock
01:58.28arrrghhhfor my device.
01:58.39arrrghhhso now I have this lovely WARNING BOOTLOADER UNLOCKED screen everytime the phone boots up
01:58.47arrrghhhI can probably change it, but it's hilarious IMHO
01:59.01*** join/#htc-linux surge (surge@pool-98-118-154-23.bflony.fios.verizon.net)
01:59.08jonpryno kexec required?
02:00.01arrrghhhnope
02:00.15arrrghhhI rolled my own kernel and booted it within a day or so of getting it
02:00.31arrrghhhmoto provided all their kernel source too
02:00.42arrrghhhbuilds really easily, even for my dumbass :P
02:02.02jonprygotta stop using that krait stuff and run hexagon-android
02:03.02arrrghhhwat's the difference?
02:03.38jonpryhexagon is probably hella slower
02:04.01jonprybut it will be cool to run 2 linux on one phone
02:05.19d3tul3or three
02:05.20d3tul3or four
02:05.24arrrghhhyou're speakin greek
02:05.47arrrghhhall I know is, build.prop sucks.
02:07.12d3tul3hey i agree, it's all greek...trick is to nod
02:07.20d3tul3with the keyboard
02:07.22d3tul3"oh really"
02:07.26d3tul3"that's cool"
02:07.32d3tul3"nice job"
02:08.11arrrghhhhm
02:08.12arrrghhhso close
02:08.15arrrghhhyet so very far....
02:08.23jonpryhmm
02:41.46d3tul3the photon q must be one of the most unfortunate members of the entire 8960 family
02:43.32arrrghhhwhy do I always pick devices with so many damned radios
02:43.38arrrghhhgsm, lte, cdma
02:43.43arrrghhhblah
02:45.35d3tul3you know there' ssomething to be said about obscure devices....i don't think you need to worry about chad
02:49.24arrrghhhlol
02:49.30arrrghhhor ... anyone for that matter
03:55.23*** join/#htc-linux detule_ (~detule@pool-108-40-44-60.bltmmd.east.verizon.net)
03:55.39detule_hey jonpry this qseecom driver is no use?
04:36.27*** join/#htc-linux Bry8Star{T (~Bry8Star@gateway/tor-sasl/bry8star)
05:15.04*** join/#htc-linux GPFerror (~GPF@71-223-137-179.phnx.qwest.net)
05:15.45*** part/#htc-linux GPFerror (~GPF@71-223-137-179.phnx.qwest.net)
05:30.44*** join/#htc-linux ccxCZ (~ccxCZ@156.200.broadband11.iol.cz)
07:08.58*** join/#htc-linux kiozen (~kiozen@p578a42db.dip0.t-ipconnect.de)
07:49.46*** join/#htc-linux fakker (fakker@unaffiliated/fakker)
09:09.08*** join/#htc-linux lamikr (lamikr@nat/nokia/x-ayqttszqmnicwteq)
09:36.39*** join/#htc-linux noobhands (d03032a2@gateway/web/freenode/ip.208.48.50.162)
09:36.59noobhandswhere the hell is zeusk
09:37.11noobhandsi need help
10:22.46noobhandsscrew zeusk, i did it
11:07.37*** join/#htc-linux d3tul3 (~detule@pool-108-40-44-60.bltmmd.east.verizon.net)
11:11.56*** join/#htc-linux lilstevie (~null@2a00:dcc0:eda:3748:247:48:46:1)
11:25.08*** join/#htc-linux eR^zeRa` (~zzeratul@ip-94-112-113-32.net.upcbroadband.cz)
11:55.16jonpryw
12:03.23*** join/#htc-linux zeusk_ (653f129f@gateway/web/freenode/ip.101.63.18.159)
12:06.31zeusk_hi
12:06.38zeusk_Marc, how is 3.0 going ?
12:08.54Marcusb sucks :P
12:09.18Marcand I need it for adb
12:09.27Marcwhich I need in order to find out why it's soft rebooting ...
12:09.41zeusk_no message from kernel about reboot ?
12:10.20Marcwell a few services just die
12:10.22Marcbut that's about it
12:10.30zeusk_maybe put a boot service to log stuff for you and save to SD every few milli seconds
12:12.00Marcnot a bad idea actually
12:12.08MarcI'll have some free time starting tomorrow
12:12.14Marcso I'll see what I can accomplish :P
12:12.49zeusk_Hmm, BB10 isn't bad. I might just settle with one instead of getting another android device
12:13.07zeusk_you tried it yet ?
12:13.14Marcnah
12:15.17*** join/#htc-linux helicopter88 (~helicopte@host231-84-dynamic.37-79-r.retail.telecomitalia.it)
12:40.43*** join/#htc-linux fakker (fakker@unaffiliated/fakker)
12:46.17styblahm, is Nexus S hack-able?
12:47.17fakkermaybe, gnex is crackable
12:47.24fakkerjust gotta look at my screen
12:59.20*** join/#htc-linux helicopter88 (~helicopte@host231-84-dynamic.37-79-r.retail.telecomitalia.it)
13:16.29*** join/#htc-linux jonpry (~jon@2602:306:c417:8aa0:e849:a229:ccf1:8bf9)
13:26.31styblawell, somebody is selling Nexus S with cracked LCD for ~ $40 USD.
13:55.33*** join/#htc-linux Cotulla (~myfakemai@109.205.253.11)
13:56.17Cotullahello
13:56.28styblahey Cotulla
14:10.21jonpryi think this pil_svc thing supports other commands
14:13.11Cotullahow is it
14:13.15Cotullajonpry?
14:13.18Cotullahi
14:21.49jonpryhey
14:22.02jonprythere is like #define PAS_MEM_CMD             2
14:22.41Cotullawhere?
14:23.01jonpryit is defined in an old version of scm-pas.c
14:23.08jonprynever used though
14:23.32jonpryscm-io.c is confusing
14:23.34jonprysecure_writel
14:25.37Cotullahaha
14:25.43Cotullaat least not relaxed
14:27.47Cotullaso how is ur status?
14:28.01detulejonpry, this qseecom driver offers some sort of an interface for starting secure apps
14:28.11jonprymy status?
14:28.19jonprydetule, yeah i see that
14:28.36jonpryour code isn't signed though
14:28.37detulethough i am not sure how that's different than the trusted ops in pil
14:29.10jonpryalso it seems like these "apps" are modules of some sort
14:29.15Cotullayea
14:29.42CotullaELF files
14:30.57Cotullaor just binaries maybe
14:33.01jonpryyou think SCM_IO_READ just reads whatever it wants?
14:36.52CotullaI think they are checking ranges
14:36.57Cotullainside
14:37.24jonpryi don't understand why they would mark things as secure only and then create this service
14:37.34Cotullawhich service?
14:37.47Cotullathey can check access
14:38.02Cotullaallow access only to right registers
14:38.07jonprySCM_IO_SVC
14:38.16Cotullaever check bits
14:38.29jonpryhmm
14:41.56CotullaI will continue kernel work soon
14:42.22jonpryi'm going to try and run it on the v3 soon
14:42.31jonprybut really i want v4
14:42.58jonprywhen are you getting at 8960?
14:43.00Cotullawell join party
14:43.12jonpry*an
14:43.15Cotulladunno
14:43.22Cotullamaybe got some WP8 phone
14:43.28Cotullaif could hack it
14:43.39jonprymy skills seem insufficient
14:44.10Cotulla:(
14:44.14Cotullav3 sucks
14:44.25jonpryyeah
14:45.03jonpryi think every processor except the arm runs in the higher security domain
14:46.30Cotullawhat u mean under security domain?
14:46.52jonpryin AXI3
14:46.58jonprythere is a 33rd bit
14:47.09CotullaI think it's just XPU
14:47.23jonprythat says if the transaction is from a hypervisor or the lower domain
14:47.54Cotullawell all that configured in XPU
14:48.04CotullaI am sure in that
14:48.23Cotullaofcourse some things are not "exported"
14:48.27jonpryi think qcom abandoned all the old stuff and just used ARM TZ stuffs
14:48.27arrrghhhI was the 34th bit on the grassy knoll
14:48.33Cotullalike QDSP6 timers or ARM timers
14:51.28Cotullabut maybe possible to config ADM
14:51.30Cotullato make a bridge
14:51.40jonpryADM?
14:51.47CotullaAPSS DataMover
14:51.53Cotullano
14:51.56CotullaApplication DataMover
14:51.59Cotullashould be more right
14:52.14jonprythat thing has domains though
14:52.17Cotullayes
14:53.36Cotullabut in ur case it can be all well configured
14:53.39Cotullatry maybe GPUI
14:55.40jonprysecure_readl didn't work
14:55.50detulethat's stupid
14:55.59Cotullayea
14:56.02detulethey should've named it "maybe_secure_readl"
14:56.07jonprylol
14:56.08Cotulla:(
14:57.19jonpryi was wondering if all of q6 images are signed
14:57.25jonprylike if the elfhdr isn't
14:57.36jonprythen i can fudge the entry point and get it to work
14:57.36Cotullaoh
14:57.39Cotullayea
14:57.44Cotullanot bad idea
14:58.22Cotullabut it also can have secure settings
14:58.25Cotullafor QDSP6 access
14:58.54jonpryi think q6 can access everything
14:59.01jonprycan't prove it though
14:59.02Cotullawhy
14:59.08Cotullacan be also rules
14:59.17jonpryarm is the only thing virtualized by TZ
14:59.36Cotullawell, TZ doesn't really important
14:59.50CotullaMPU/XPU is important
14:59.55jonpryyou don't have 8960
15:00.10Cotullayes
15:00.12jonpryi've seen no evidence XPU even exists
15:00.41Cotullano evidences == doens't mean that it doesn't exist :)
15:00.43detuleneed motherlode
15:00.51jonpryyeah
15:01.04jonprylike those AMSS security advisories
15:01.38Cotullaand I am not talking from air usually :)
15:03.35jonpryi know
15:03.51jonpryi just think 8960 has totally different security hardware
15:03.59Cotullaso I may know a lot of ever without ever touching 8960 somehow
15:04.12jonpryand it's more stock ARM stuff instead of qcom specific
15:04.24Cotullahehe I don't think so :)
15:04.44jonpryneed those docs i guess
15:05.23Cotullawhcih ones?
15:05.35jonprythat xda post with 140 pdf's
15:06.09Cotullaoh
15:12.20Cotullaso I need modify linker script
15:12.25Cotullato put section at FFF00000
15:12.26Cotulla?
15:20.45Cotullathis code is extremely crapped
15:20.50Cotullano logic
15:22.08*** join/#htc-linux gassed (faxed@c-76-30-161-228.hsd1.tx.comcast.net)
15:24.58jonpryi don't understand why there is fixed code
15:26.59Cotulla?
15:29.07*** join/#htc-linux gassedup (faxed@c-76-30-161-228.hsd1.tx.comcast.net)
15:50.17jonprycode fixed at 0xfff0
15:51.05CotullaI need
15:53.29jonpryi can't imagine why
15:54.26CotullaI tell u before
15:54.30Cotullafor "core sys"
15:54.37CotullaSW MMU and etc
15:54.54Cotullaexceptions
15:55.03Cotullaall inside one TLB entry
15:55.39jonprywhole kernel can be inside one TLB entry
15:56.01Cotullait can not
15:56.05Cotulla4M and 16M
15:56.10jonpryso
15:56.11Cotullakernel is around 10-11M
15:56.19jonpryalias
15:56.27CotullaI don't want to waste memory
15:56.31jonpryno waste
15:56.38jonpryvirtual aliases
15:56.38Cotulla?
15:56.53Cotullahow it will work?
15:56.58jonprylike map 16MB 0xc000 -> 0x8000 or whatever
15:57.08jonprykernel is only 11mb
15:57.09Cotullacode must be relocated to right address to run
15:57.24jonpryso 0xc000 + 12MB points to some garbage
15:57.33jonprybut that memory can be mapped to something else
15:58.18Cotullawell kernel assume kernel virtual space
15:58.22Cotullafor kalloc and etc
15:58.27jonpryyes
15:58.41Cotullaso it assumes it can map memory directly after kernel end
15:59.02Cotullabut there is already this 16M TLB
15:59.18jonpryi think you can move VMALLOC_START
15:59.30Cotullawell
15:59.40Cotullawe can just move "core sys" to FFF0 0000 and be happy
16:03.54Cotullabut I am afraid a problem about relocations
16:10.43jonpryseems that elfhdr is signed
16:12.24Cotulla:(
16:35.03Cotullalooks like kernel was builded for native before
16:44.54jonpryhow so?
16:49.22Cotullacomments tell about that
16:49.25Cotullaand some macros
16:56.08jonpryany idea how this gpu works?
16:57.41Cotullacommand stream
16:57.44Cotullawith commands?
16:57.57Cotullashould be write mem there one
16:59.12jonprylike MEM_TO_REG REG_TO_MEM
16:59.33Cotullano
16:59.35Cotullaanother one
16:59.49Cotullavalue in packet to mem
17:00.17Cotullaor event write
17:01.31jonprythere is no way to know if write worked
17:01.53Cotullayea
17:01.58Cotullabut maybe gpu halt
17:02.06Cotullaaxi error
17:02.09CotullaBUS STOP
17:02.46jonpryi don't see any read commands
17:02.56*** join/#htc-linux lilstevie (~null@2a00:dcc0:eda:3748:247:48:46:1)
17:03.16Cotullawhere is it?
17:03.25jonpryadreno_pm4types.h
17:05.33CotullaEVENT_WRITE
17:05.48CotullaMEM_WRITE
17:06.27jonpryyeah i see mem_write
17:07.22jonpryand it has some kind of mmu
17:07.33Cotullaalso IOMMU
17:09.17*** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net)
17:23.32*** join/#htc-linux EdLin (~EdLin@midway.packetsense.net)
17:24.53*** join/#htc-linux eR^zeRa` (~zzeratul@ip-94-112-113-32.net.upcbroadband.cz)
17:53.26*** join/#htc-linux GPFerror_ (~GPF@71-223-137-179.phnx.qwest.net)
17:53.49*** part/#htc-linux GPFerror_ (~GPF@71-223-137-179.phnx.qwest.net)
19:22.39*** join/#htc-linux kiozen (~kiozen@ppp-93-104-85-99.dynamic.mnet-online.de)
20:51.09*** part/#htc-linux Cotulla (~myfakemai@109.205.253.11)
20:59.42*** join/#htc-linux BabelO (~fcr@unaffiliated/babelo)
23:19.27*** join/#htc-linux Bry8Star{T (~Bry8Star@gateway/tor-sasl/bry8star)
23:19.31*** join/#htc-linux d3tul3 (~detule@pool-108-40-44-60.bltmmd.east.verizon.net)
23:50.37*** join/#htc-linux eR^zeRa` (~zzeratul@ip-94-112-113-32.net.upcbroadband.cz)
23:55.46d3tul3jonpry, trying to get the lode

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