irclog2html for #htc-blueangel on 20051216

07:25.48*** join/#htc-blueangel JuNuKN (n=JuNuKN@K8147.k.pppool.de)
07:25.58JuNuKN~seen llange
07:26.00aptllange <n=llange@lange.nom.fr> was last seen on IRC in channel #htc-blueangel, 16d 8h 37m 2s ago, saying: 'no probs. =)'.
07:39.06*** join/#htc-blueangel rob_w-cgi (n=txuser@dd1010.kasserver.com)
07:39.06*** mode/#htc-blueangel [+o rob_w-cgi] by ChanServ
07:39.29rob_w-cgimorning
08:51.18*** join/#htc-blueangel |oxi| (n=oxi@212.106.207.206)
12:50.40*** part/#htc-blueangel geers_ (n=DevzMan@c51008F31.inet.catch.no)
14:02.03*** join/#htc-blueangel cp15 (i=joyrdo@p549A386F.dip0.t-ipconnect.de)
14:02.03*** mode/#htc-blueangel [+o cp15] by ChanServ
14:02.19cp15Good morning
14:29.22oximorning(noon here)
14:31.50cp15http://www.total-knowledge.com/~ilya/mips/ugt.html
14:31.58cp15Where are you from?
14:33.12oxiLOL
14:33.15oxispain
14:33.22oxigood morning then :D
14:36.06cp15Have you already looked into how to use haret?
14:36.36oxinot yet
14:36.46oxifirst im trying to port a basic kernel image
14:36.51oxiwith oe
14:37.37cp15Haret is the first step. Please look into it. A kernel image which doesn't work is useless
14:37.59cp15I will try to make a generic one which should run on every pxa machine
14:38.19oxiit wont even catch the lcd display?
14:39.42oxii wouldnt like to make you waste your time... i should do that work myself shouldnt i
14:42.18cp15The LCD display will work, since it is already set up by windows. It is using a generic framebuffer which needs start address and resolution
14:47.10oxiomg this is so complicated i wish i could contact cr2. If he was leading the porting project i guess i could contribute... but doing it myself alone i think thats beyond my potential
14:55.32cp15Climbing a mountain starts with a first step
15:18.56rob_whey dude
15:20.02cp15Hi rob_w
15:24.27oxihi
15:24.32oxitrue cp15 :)
15:33.12cp15Universal has WM 5.0, true?
15:40.45oxiyup
15:41.34cp15Might be a problem... At least on the BA haret doesn't work with WM 5.0
15:41.38oxithis is my first and only pda so i dont know whether its better than previous versions or not....
15:41.49oxidamn it....
15:43.03oximaybe i could downgrade to ppc2003?
15:47.53oxinot sure... but will KITL help doing the reverse engineering thing?
15:48.20oxii got this option on my boot menu, pulsing a combo when resetting
15:49.56oxipressing not pulsing*
15:52.00cp15Don't think it is possible to downgrade the Universal
15:52.06cp15But maybe we can fix haret
15:53.01cp15Universal has a pxa27x, if I remember correctly
15:53.08cp15Can you tell me your framebuffer base?
15:53.32oxiyes
15:53.48oxiuhh... how do i do this...
15:56.44cp15Install haret, telnet to it, and type print "%08x" VRAM
16:01.04Funklordhey =)
16:01.21FunklordI got my BA back =D
16:02.00Funklordsince I bought one, I've got two now
16:02.13cp15How did you get it back?
16:02.26FunklordI went to the club a couple of days later
16:02.37Funklordthey got it from someone
16:03.13Funklordas usual, the stylus wasn't in it =P
16:03.29cp15Hmm... But now you probably won't get the money from the insurance?
16:03.49FunklordI wasn't going to get any money from them anyways
16:04.05Funklordbastards had a clause for cell-phones and theft
16:04.17cp15OTOH: Now you have a development BA and can fully take part of the project
16:04.49Funklordapparently, if you tell them you *lost* it, the insurance will cover it
16:04.57Funklordbut not due to theft
16:05.26oxiawsome
16:06.09Funklordyeah I will definitely try to get hooked in development
16:08.34oxiHaRET(1)# print "%08x" VRAM
16:08.34oxi00000000
16:09.17cp15Try VRAMV instead of VRAM
16:09.23cp15This value doesn't make sense
16:09.47Funklordcp15: how do you charge it while it's running linux?
16:10.57oxiHaRET(5)# print "%08x" VRAMV
16:10.57oxiError: line 5: Unknown variable 'VRAMV' in expression
16:10.57oxi000005c8
16:11.10cp15Just a moment
16:11.38rob_wcu
16:11.47oxibye rob_w
16:12.29cp15VRAMV should exist
16:12.35cp15Hmm
16:13.05cp15Funklord: Just like with windows. The charger operates independently from the cpu
16:13.24oximay i paste the introductory message in telnet?
16:13.58Funklordcp15: the last times I've tried it's in the cradle, but still runs out of power
16:16.40cp15Hmm... That would be strange
16:17.18FunklordI keep it connected for shell acces
16:18.05Funklordbut it still runs out of power and refuses to work until I get the emergency charge routine going
16:19.23Funklordor, maybe some battery meter goes nuts
16:19.35Funklord(I haven't measured the battery voltage)
16:20.07cp15Hmm... In the newest release the battery monitor is working (partly...) Maybe you can check it
16:23.05Funklordok, yeah, I'll have to unravel this mystery ;)
16:27.47Funklordttyl
16:27.48oxicp15: http://pastebin.com/466716
16:33.07cp15Hmm...
16:33.19cp15Maybe we have to dump the pxa registers to get its value
16:34.47cp15The pxa lcd is quite complicated
16:39.27oxiis that gpio?
16:41.21oxihttp://pastebin.com/466731
16:48.00cp15No... We have to determine the framebuffer address from the pxa registers, since haret doesn't know it
16:48.41cp15do a pd 0x44000204 4
16:49.34oxi44000204 | a3fe77a0                            | .w..
16:51.16cp15Try again to see if the value is constant
16:52.12oxinope
16:52.23oxiwhere did you get that address?
16:52.30cp15From the data sheet
16:52.49oxihmm
16:52.53cp15Try 10 times and then tell me the minimum you get
16:53.03oxiokay
16:53.10cp15It probably gives the current display address, not its base address
16:53.57cp15Apart from that it looks good: It is at the top of the memory
16:54.34cp15When done, try print "%08x" RAMSIZE to see if the memory is right
16:54.46oxi5c006400
16:55.02cp15What is that?
16:55.09oxithe min value i got
16:55.21oxiHaRET(19)# print "%08x" RAMSIZE
16:55.22oxi06000000
16:56.03cp15The min value doesn't look right, I assume it is 0xa3fe6400
16:56.10cp15Lot of guesswork...
16:56.14cp15Lets see the ram size
16:56.25oxisome of them start with a3, others start with 5c
16:57.21cp155c is the internal memory... Hmmm
16:57.54cp15But it is too small to hold the picture
16:58.03cp15But we can try to paint some black pixels
16:58.17oxiheheh cool
16:59.51cp15Try pfw 0xa3fe77a0 10 0
17:00.01cp15Should get you some black pixels somewhere
17:00.56oxinothing :(
17:01.07cp15Try pfw 0xa3fe77a0 10 0xffffffff
17:01.19cp15To get some white one (maybe it was already in a black area)
17:01.52oxinothing...
17:02.28cp15Try pd 0xa3fe77a0 after each of the above command to see if something has changed
17:02.52cp15pd 0xa3fe77a0 4 I mean
17:03.07oxiyes its full of f's
17:03.25cp15After the pfw with the fff?
17:03.41cp15Then try the one with 0 again and see if you get 0s
17:03.57oxiyes
17:04.19cp15And nothing changes on the screen?
17:04.29cp15:-(
17:04.42cp15Try to replace 10 with 100 to get a bigger area
17:05.32oxi10/100 is for 10/100 words?
17:05.46cp15Yes
17:05.55oxiYES
17:05.58oxii see a line!
17:06.03oxiyay!
17:06.29oxiits a bit above by the center
17:06.31cp15Is it horizontal or vertical
17:06.39oxihorizontal
17:06.50cp15In landscape format?
17:07.01oxii got the screen in vertical mode, to see the whole haret window
17:07.28cp15So the the height is bigger than the width?
17:07.38oxiyes
17:07.56cp15Ok, now lets get the base address
17:08.20cp15Try to lower the address in the pfw until the line starts in the upper left corner
17:08.21oxihmm i wonder if it'd write a horizontal line in landscape format this wat
17:08.26oxiway*
17:08.34oxiokay!
17:08.38cp15Try it if you can, would be interesting
17:09.26oxiuh oh
17:09.37oxii think i went beyond the limit
17:09.44oxisuddenly the screen turn white...
17:09.53cp15Reset time
17:09.56oxilol
17:15.48oxialmost
17:16.22cp15The address is probably a round one, so if you are close it might be enough
17:19.38oxi0xa3faa787 this is the closest i got, but not sure if got to the limit
17:20.36cp15It needs a 0 at its end
17:20.36oxi0xa3faa788 seems to paint the upper-leftmost dot, but not sure
17:20.51oxihmm
17:20.57cp15Hmm... Maybe not...
17:21.26oxigonna try white dots
17:22.29cp15It is probably 0xa3faa000
17:23.00cp15You are off 4 lines and 4 pixels
17:23.17oxihmm
17:23.35cp15Lets do some calculation
17:24.05oxihmm yes i was missing some line
17:24.22cp15Buahh... It doesn't make sense
17:24.47oxiwhy?
17:25.04cp15640*480*2=0x96000
17:25.11cp15Resolution*Bytes per pixel
17:25.27oxionly 2 bytes per pixel?
17:25.41oxioh yes
17:26.06cp15A3FAA000+96000 = A4040000
17:26.14cp15Which is off limit for the ram
17:26.57oxihow do you know the ram limit?
17:27.00oxiwhere is it
17:27.10cp15Try pfw with count 240
17:27.16cp15Does it fill a complete line?
17:27.53oxiyes
17:28.02cp15http://wiki.xda-developers.com/index.php?pagename=UniversalMemoryMap
17:29.20cp15Maybe the screen is split. Part of it is in the external ram, part of it in the internal
17:31.50cp15Ok, we do a trick. We reduce the resolution to 360*480, so the linux screen will be only a part of the whole screen for now, but it fits in the external ram
17:33.05cp15Wanna test the first kernel?
17:33.24oxisure :)))
17:33.37oxitheres something i dont get about this:
17:33.37cp15Ok, can you do dcc?
17:34.12oxii think so
17:34.38oxiforget it...
17:39.11oxithank you :)
17:39.15cp15put startup.txt and zImage in the same directory as haret and run haret. But make a backup first if you have valuable data
17:39.37cp15When you are lucky you will se a kernel panic
17:39.56oxiokay
17:39.58cp15When you have less luck, you will see something unrecognizeable
17:40.18cp15If you have even lesser luck, you will see a penguin with red eyes
17:40.56cp15If you don't even see the red eyes, haret needs to be fixed
17:42.10oxiLOL
17:42.34oxii see microscopic letters in the upper side of screen
17:42.43oxikernel panic :)
17:42.55cp15A huge success I would say
17:43.07oxii agree :)
17:43.58oxii guess its taking the upper part of screen cause it can only write there, right?
17:44.17cp15Yeah, we have to find out where the lower half is
17:44.53oxiwhy did they design it this way? isnt it too... stupid
17:45.32cp15Well, maybe not... There is 256KB memory integrated into the pxa. This one is probably used for the other half
17:45.53cp15It frees memory bandwith and is a good use for the internal memory
17:46.04oxihmm
17:46.33cp15Ok, I will prepare a diff to the kernel sources
17:59.23cp15Ok, this is what you need to do next:
17:59.33cp15Get a arm toolchain
17:59.40cp15ftp://ftp.handhelds.org/projects/toolchain/arm-linux-gcc-3.4.1.tar.bz2
17:59.51cp15might work, but I am using gcc 3.4.4
18:00.04oxihmm got oe installed, doesnt that have one?
18:00.19cp15Ok, if you have oe installed, you don't need the external one
18:00.59cp15You need to set your PATH like this:
18:01.02cp15PATH=$BUILD/tmp/cross/bin:$BUILD/tmp/staging/i686-linux/bin:$PATH
18:01.18cp15where $BUILD is your oe build directory
18:01.49cp15Check out the BA kernel sources:
18:01.50oxiokay
18:01.54cp15cvs -z3 -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/xanadux co -ko -r K2-6-12-hh2-xda0-ba0 linux-2.6-xda
18:02.13cp15apply the diff I sent you on these sources
18:03.02cp15In linux-2.6-xda, do a cp arch/arm/configs/arm_defconfig .config
18:03.05cp15make
18:03.35cp15Try out the kernel arch/arm/boot/zImage
18:04.15cp15If it works, cp -rp arch/arm/mach-pxa/generic arch/arm/mach-pxa/universal
18:04.33cp15Edit the appropriate Files to use universal instead of generic
18:04.48cp15Test the kernel again
18:05.28cp15Maybe set a larger font, so you don't need a microscope
18:05.49cp15Try to get usb working... Not easy, but would be even harder if you don't have a screen
18:06.20cp15Create a initrd which sets up a usb connection
18:06.40cp15Remove the vsfb framebuffer and use pxafb
18:06.48cp15Enough for now :-)
18:07.09oxiphew...
18:11.54cp15Oops, one of my instructions was wrong
18:11.56cp15Replace
18:12.00oxiso much homework :)
18:12.07cp15In linux-2.6-xda, do a cp arch/arm/configs/arm_defconfig .config
18:12.10cp15with
18:12.23cp15In linux-2.6-xda, do a cp arch/arm/configs/generic_defconfig .config
18:12.32oxiokay
18:12.39oxithank you a lot :)))
18:13.11cp15You are welcome
18:13.26cp15If you are stuck, don't hesitate to ask
18:13.39cp15But I am afraid you have to find out much by yourself
18:13.48cp15Have you already tried to contact cr2?
18:20.38*** join/#htc-blueangel rob_w (n=bob@p85.212.151.171.tisdip.tiscali.de)
18:20.38*** mode/#htc-blueangel [+o rob_w] by ChanServ
18:20.39oxinope i couldnt find an email in that webpage
18:20.50oxihi rob_w
18:20.58cp15http://forum.xda-developers.com/profile.php?mode=viewprofile&u=16735
18:21.15rob_whey<
18:21.18cp15You have to register with xda-developers to send a mail
18:21.34oxithanks again :)
18:28.23rob_wanybody interessted in dtmf tones playing when dialing ?
18:30.21*** join/#htc-blueangel pH5 (n=ph5@e178248116.adsl.alicedsl.de)
18:30.44pH5cp15: what is going on in the blueangel world?
18:34.29rob_whey pH5
18:34.34pH5hey rob_w
18:35.21rob_wcp15 did already a incredible job .. also many others are working together now .. did get a nice little project
18:37.28rob_whow about you pH5
18:38.49pH5regarding my magician, I didn't achieve anything yet
18:39.30rob_whis price dropped now to the 1 € +contract
18:45.01cp15Tell me whether it worked or not
19:23.31*** join/#htc-blueangel BrianR___ (i=brianr@c-24-61-206-174.hsd1.ma.comcast.net)
19:23.52*** part/#htc-blueangel BrianR___ (i=brianr@c-24-61-206-174.hsd1.ma.comcast.net)
19:45.37pH5cp15: the zImage only gives me my haret pixels and than stays at bigeye-tux
19:45.41pH5trying zImage now
19:45.49pH5s/zImage/Image/
19:52.29cp15Hmmm... Then I guess there is still some memory corruption
19:52.31cp15Strange
19:53.23cp15I will prepare a diff so you can compile your own kernel
19:54.47pH5cp15: thanks. great work anyway. I'm curious what the difference will be to my last tries
21:02.29*** join/#htc-blueangel apt (i=ibot@pdpc/supporter/active/TimRiker/bot/apt)
21:02.29*** topic/#htc-blueangel is HTC BlueAngel is still under development ! Feel free to ask and check also -> http://www.handhelds.org/moin/moin.cgi/BlueAngel | http://wiki.xda-developers.com/index.php?pagename=BlueangelResearch | http://gnulinux.biz/files/blueangel/sd/linux/ channel chat logs at http://ibot.rikers.org/%23htc-blueangel/ * http://www.userfriendly.org/ *
21:31.48immolo:o MS are taking the graphic interface out of the kernel :o
21:33.08immoloThat's like the main speed boost windows gets over X11 gone
21:33.11cp15Yeah? Where do they put it? In an X-Server?
21:33.20immolouserspace
21:33.29immolothat's all they are saying
21:34.18immolothey say 90% of crashes are do with that
21:34.49immoloso now vista is pretty much looking like linux
21:35.03bhimaWindows XP, I believe, does some nifty stuff with auto-reloading of device drivers that crash.
21:35.17bhimaA friend of mine had his video driver crash.
21:35.32immoloso I guess blackcomb will probley just be the linux kernel with some ms tools
21:35.33bhimaThe monitor flickered, and then it auto-loaded a default VGA driver for his card.
21:35.48immoloheh, that's pretty cool
21:35.57bhimaimmolo: there's a _lot_ more to OS design than how much is in the kernel and how much is in userspace...
21:36.19immolobhima, I never got much further then the boot loader :P
21:36.30bhimaRemoving the video stuff from kernelspace is really only a small simililarity.
21:36.58immoloI was talking about the user permissions
21:37.14immoloit's why you can only install NTFS on vista
21:37.23immolono more admin account
21:49.09immolohow is gomunictor looking these days?
21:49.59rob_w~larts immolo for having a typo in THAT name
21:50.09rob_w~lart immolo for having a typo in THAT name
21:50.57immolowhich is pretty bad seeing english is my native langauge and you can spell better then me :P
21:51.34rob_wi just thought about rearranging the visuals vice versa , so then we put a volume slider in the free space and connect the physical up/down to the mixers
21:52.34immolonice
21:53.00immoloso then we could turn the volume off for the phone but keep it on for the pda
21:54.28immolothere are so many things I wish I could code for this device
22:01.31cp15rob_w: Did you get the doc driver?
22:03.07immoloyou can read the onboard storage now?
22:03.33rob_wthey didnt had one ;-)
22:04.17rob_wgomunicator breaks on not knowing PATH_MAX ?
23:13.32cp15immolo: I am working on a doc "shim" that allows a doc driver for an older motorola kernel to be loaded
23:14.08cp15I can already get the partitions and with a userspace test program I can even read sectors

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.