IRC log for #htc-linux on 20070407

00:00.15cr2where is that?
00:00.28Kevin2No idea.
00:00.44Kevin2CPU0: D VIVT write-back cache
00:00.44Kevin2CPU0: I cache: 16384 bytes, associativity 64, 32 byte lines, 8 sets
00:00.44Kevin2CPU0: D cache: 16384 bytes, associativity 64, 32 byte lines, 8 sets
00:00.44Kevin2Built 1 zonelists.  Total pages: 16256
00:00.44Kevin2<5>Kernel command line: root=/dev/ram0 ro console=ttySAC2,9600,8,1
00:00.44Kevin2irq: clearing pending ext status 00000040
00:00.47Kevin2irq: clearing pending status 00004000
00:00.48Kevin2irq: clearing subpending status 000001d0
00:00.50Kevin2irq: clearing subpending status 00000090
00:00.52Kevin2PID hash table entries: 256 (order: 8, 1024 bytes)
00:00.54Kevin2timer tcon=00510009, tcnt 733b, tcfg 00000203,00001000, usec 00002b64
00:01.23cr2console=ttySAC2
00:01.29cr2can you try tty0 ?
00:06.36Kevin2No difference.
00:07.19Kevin2http://pastebin.ca/427755
00:07.34Kevin2I pulled out all my custom printascii, and just have the main one in printk now.
00:08.41*** join/#htc-linux alsuren (n=alsuren@host86-148-18-74.range86-148.btcentralplus.com)
00:08.42cr2S3C244X: core 283.200 MHz, memory 70.800 MHz, peripheral 35.400 MHz
00:08.49cr2doesn't look right.
00:11.15Kevin2Last printk is from s3c2410_timer_setup
00:15.21cr2it may be helpful to port this code S3C244X: core 283.200 MHz, memory 70.800 MHz, peripheral 35.400 MH
00:15.24cr2to haret
00:23.42Kevin2cr2: The last printk is via time_init() in main.c -- the next block is console_init().  So, the code might be proceeding without IR output.
00:26.37Kevin2printk("pre irqs on\n");
00:26.37Kevin2early_boot_irqs_on();
00:26.37Kevin2local_irq_enable();
00:26.37Kevin2printk("pre console_init\n");
00:27.02Kevin2Only shows the first message..  Something is up with irqs.  (Or maybe printascii doesn't work with irqs on..)
00:29.05*** join/#htc-linux ellisway (n=ellis@host-87-74-241-174.bulldogdsl.com)
00:32.06*** join/#htc-linux TBBle (n=tbble@2001:388:f000:0:0:0:0:2d)
00:38.11Kevin2Well, disabling the printascii in printk doesn't improve things..
00:45.00cr2Kevin2: i don't like the 283MHz speed
00:45.07cr2imho it should be 400
00:54.52Kevin2cr2: Do you think it would make it crash?
00:57.12cr2maybe.
01:31.09cr2hehehe. it was a loose connector on himalaya :) the lcd is ok now.
01:38.01cr2Kevin2: still here ?
01:38.12cr216934400/12000000*283.2
01:38.32cr2=399652
01:39.06cr2set the clock to 0
01:39.18Kevin2cr2: I'm here
01:39.18cr2s3c24xx_init_clocks(16934400);
01:39.20cr2here
01:39.36cr2and try again.
01:39.47cr20 means 12000000
01:42.39Kevin2Okay.  I need to bring my apache back up for ir.
01:43.34cr2then the memclk will be ok too.
01:43.50cr2it can certainly lead the crashes :)
01:50.54Kevin2cr2: Still hangs.  I'll turn printascii back on and see what the last message is.
01:52.06cr2yes.
01:53.07Kevin2Dies in same place.  Last message is right before we enable interrupts.
01:53.30cr2nice. accembled himalaya. now with color and no jitter.
01:53.31cr2ok.
01:56.07cr2Kevin2: does it show the correct speed ?
01:56.58Kevin2S3C244X: core 399.651 MHz, memory 99.912 MHz, peripheral 49.956 MHz
01:57.32cr2ok. that's much better :)
02:09.50cr2Kevin2: is the vsfb init code correct ? since we don't have a real LCD driver...
02:10.54cr2the most close is ./mach-pxa/eseries/e330_lcd.c
02:20.45Kevin2cr2: We're looping on an interrupt.  Looks like IRQ_EINT4t7 is constantly triggering.
02:21.07cr2what is that?
02:21.25Kevin2One of the gpios.
02:21.46cr2camera button ?
02:21.58cr2http://wiki.xda-developers.com/index.php?pagename=HermesGPIO
02:24.18Kevin2cr2: Probably wifi?
02:25.50Kevin2Yep.  If I turn off wifi.  Then I loop on the uart irq.
02:27.25cr2tur off uart ?
02:27.51Kevin2It's probably my own writes to the uart.
02:28.34Kevin2But, the linux code is definitely not initializing the irq registers fully.
02:29.10cr2ok. i'm really tired now. good night.
02:29.18Kevin2Good night.
02:35.00cr2Kevin2: ?
02:35.16cr2i booted and have an oops
02:35.26cr2in the framebuffer
02:35.36cr2and nice penguin :)
02:36.50Kevin2cr2: On Hermes?
02:36.55cr2yes.
02:37.06Kevin2Cool.  What did you do?
02:37.07cr2on himalaya i have other problems ;)
02:37.14Kevin2I just got to:
02:37.14cr2nothing.
02:37.16Kevin2timer tcon=00510109, tcnt a29d, tcfg 00000203,00001030, usec 00001ebf
02:37.16Kevin2pre irqs on
02:37.16Kevin2pre console_init
02:37.16Kevin2Console: colour dummy device 80x30
02:37.38cr2maybe i'll commit my config to hh.org
02:38.19Kevin2You made the clocks and kconfig change, but nothing else?
02:38.26cr2no, it scrolls too fast, i can't see the reason for oops.
02:39.20cr2Backtrace: ... (copy_process ... do_fork(...
02:39.27cr2really deep in the kernel.
02:39.51Kevin2Do you still have the IR stuff on?
02:40.15cr2no.
02:40.38Kevin2Which haret are you using?
02:40.48cr2yours.
02:41.15Kevin2Compiled from cvs?  Or from my account on the web?
02:41.26cr2taken from your account.
02:42.09cr2i'm doing a diff with the CVS first.
02:42.26cr2because i have a lot of changes in the himalaya/blueangel code too.
02:45.03cr2ok, committing.
02:46.18cr2done.
02:47.24cr2ok, now i really need to sleep :)
02:49.59*** join/#htc-linux cyrill62 (n=cyrill62@92.249.146.195.dynamic.adsl.abo.nordnet.fr)
02:58.20Kevin2cr2: It boots for me too now.  I don't see an oops.  Just unable to mount root fs on unknown-block(1,0)
02:58.39Kevin2But, I didn't specify an initrd or root partition.
02:59.26Kevin2I'm guessing it was the fact that you had uarts not config'ed that makes in boot.  Are configs look really similar otherwise.
03:00.26*** join/#htc-linux kelly_ (n=kelly@ppp206-236.lns3.mel4.internode.on.net)
03:07.16kelly_hey guys
03:09.06kelly_just trying to figure out what the status is on the hermes port? I have asked in xda-dev forum but no info. From the logs it looks like there is some work going on?
03:17.52Kevin2kelly_: cr2 and I just got Hermes to boot a few minutes ago.
03:18.19Kevin2kelly_: Lots more work to go before it is usable, but it is good progress.
03:19.30Kevin2Good night.
03:20.02kelly_I heard the the processor is supported in other devices so once it boots development might be quite quick?
03:20.17kelly_good night and well done. Very exciting stuff. Cant wait
03:45.30*** join/#htc-linux rob__w (n=bob@p213.54.166.130.tisdip.tiscali.de)
04:52.41*** join/#htc-linux tudenbart (n=willi@xdsl-81-173-171-155.netcologne.de)
05:24.27*** join/#htc-linux RoEn_PC (n=roen@p54a64fdd.dip.t-dialin.net)
06:42.41*** join/#htc-linux rmoravcik (n=rmoravci@adsl-dyn49.91-127-60.t-com.sk)
08:20.54*** join/#htc-linux psokolovsky_ (n=psokolov@82.193.98.4)
09:09.37*** join/#htc-linux WizMaui (n=WizMaui@62.112.90.250)
09:10.05*** part/#htc-linux WizMaui (n=WizMaui@62.112.90.250)
09:25.37*** join/#htc-linux cyrill62 (n=cyrill62@92.249.146.195.dynamic.adsl.abo.nordnet.fr)
09:37.06*** join/#htc-linux ||ugh (n=hugh@road.toad.com)
09:48.18*** join/#htc-linux goxboxlive (n=goxboxli@206.80-202-161.nextgentel.com)
10:32.25*** join/#htc-linux skodde (n=skodde@unaffiliated/skodde)
10:44.40cr2hi
10:46.35cr2Kevin2: i have removed the penguin, stack dump, etc.
10:47.49cr2the kernel crashes after Console: switching to colour frame buffer ...
10:48.12cr2PC is at copy_process+...
10:48.58cr2LR is at __init_begin+...
10:53.29goxboxlivehiu cr2: I saw it got late for you this night :-)
10:57.37cr2yeah.
10:57.53cr2but the evil hermes boots now.
10:58.44goxboxliveyes i saw that. Great job :-)
10:59.03goxboxliveI am trying to let xqt run in Opie, but it doesnt look good
10:59.04cr2:)
10:59.18cr2what is xqt?
10:59.36goxboxliveIt lets x works under qt AFAIK
11:00.02cr2i have also disassembled and repaired himalaya yesterday at night.
11:00.05cr2ok.
11:00.05goxboxlivei was thinking about openmoko-dialer on Opie
11:00.27goxboxlivedid you fix hima too?
11:00.56cr2the LCD has colour now and does not jitter as before. hima is 100% functional now.
11:01.13cr2but it seems lkcl only booted it from nfs
11:01.32cr2i have the same readonly problems as on the blueangel now.
11:01.40goxboxliveok
11:01.52goxboxlivewhere is lkcl, havent seen him for a while
11:02.02cr2working hard.
11:02.09goxboxliveok
11:02.40cr2where is BabelOued...
11:02.43goxboxlivewell, are you going back to universal now, to finish it up since you have comed so far wuth the other devices :-)
11:02.49goxboxliveHe is working hard too
11:02.54cr2qtopia phone
11:03.47cr2i need to solve the readonly problem on himalaya.
11:10.53goxboxlivepsokolovsky_ Why did system sound stop working in Opie? It worked under 1.2.1 AFIAR
11:11.06psokolovsky_Hi!
11:11.46psokolovsky_goxboxlive: that's pain problem in OE. yes, ut worked in Familiar. no, I didn't hear it working in OE. Your help with resolving that is appreciated.
11:13.03goxboxliveYes that's right. I was using Familiar when system sounds worked. It worked building it with OE, but it was as you said Familiar, not Angstrom
11:13.29cr2ah. that's nice, it seems i've found it.
11:13.43goxboxlivecr2: found what?
11:14.26cr2readonly gpio on himalaya.
11:15.47goxboxliveso good, that's makes me happy :-)
11:15.54psokolovsky_cr2: cool. btw, any news on BA acx?
11:17.58psokolovsky_goxboxlive: btw, big changes come the opie way. in a few, I'll commit change to the default keyboard input handler.
11:18.34cr2it seems i need to load it as a module.
11:19.04goxboxliveok, i have just tried to build opie now, there where 3 packaqges that didnt build
11:19.05psokolovsky_goxboxlive: now the default will be handler which was used before on zauruses and ezx'es. this one reuses kernel keymaps and so configurable.
11:19.31goxboxlivethats great.
11:21.21cr2yes, it's write protect.
11:27.59*** join/#htc-linux law|ibook (n=_law_@mk084020170227.a1.net)
11:39.56*** join/#htc-linux RoEn_PC (n=roen@p54a64fdd.dip.t-dialin.net)
12:32.58*** join/#htc-linux apt_ (i=ibot@pdpc/supporter/active/TimRiker/bot/apt)
12:32.58*** topic/#htc-linux is HTC Linux Channel: Find logs at http://apt.rikers.org/%23htc-linux/ | please check http://handhelds.org/moin/moin.cgi/HTC_2dPhones | http://wiki.xda-developers.com/index.php?pagename=Xanadux | <cr2> let's define a common setup.
12:33.28goxboxlivecr2: The leds are still blinking but the keyboard and TS are freezed
12:33.49goxboxliveill check out usbnet
12:34.27goxboxliveno, dropbear is also down
12:35.49goxboxlivepsokolovsky: Should the latest changes to qte-mt make it use the kernel-keymap file without any changes for machine specific patches?
12:37.19psokolovsky_goxboxlive: yes, all machine-specific patches are going to /dev/null. it doesn't read any file, it reads exactly keymap out of kernel. one thing I'm note sure howexactly it uses it for all cases - see the source ;-)
12:39.13goxboxlivehmm ok, we have multiple keys [using a FN key to get in example / with [Fn]+[o]) and we are having a file kernel-keymap file in /etc. It is getting loaded at boot, because it works in the shell, but not in QT.
12:39.23goxboxliveI'll try to take a lok.
12:39.40goxboxliveWhy isent it as easy as in X11?
12:41.00psokolovsky_well, it should be, help make it that way ;-)
12:41.42goxboxlivei'll try
12:42.27goxboxlivepsokolovsky_ : Do you know if it is much work to get xqt support multithreaded QT?
12:42.45goxboxliveWith xqt opie would be a killer.
12:42.48psokolovsky_nope
12:43.17psokolovsky_there's also kdrive-sdl which could be alternative, imho
12:43.49goxboxliveok
12:53.12cr2No init found.
12:54.56goxboxlivewhat?
12:55.02goxboxliveacx on BA?
12:56.13cr2no. cant boot with the same config on hima
12:56.29cr2it may be the 2gig card.
12:56.36goxboxliveo
12:56.37goxboxlivek
12:57.45cr2no. 128M does not boot too.
12:58.24cr2mounted root.
12:58.31cr2and then "Not init found"
12:58.42cr2s/"Not/"No/
13:02.12cr2-CONFIG_PXA_IPAQ=y
13:02.19cr2how is that enabled??
13:03.55goxboxlivedont know
13:04.47cr2config PXA_IPAQ
13:04.58cr2depends on
13:05.05cr2<PROTECTED>
13:05.11cr2it's some legacy junk
13:06.09psokolovsky_cr2: btw, let's rename himalya to htchimalaya, until it's too late?
13:07.08cr2psokolovsky_: the machine name HIMALAYA is hardcoded ;) we can rename the subdirectory, though.
13:07.23cr2but we have 'magician'
13:07.37cr2need to convince pH5
13:08.03psokolovsky_cr2: yeah, but we can rename defconfig, etc. and use that as machine name regardless what sloppy mach-types contains
13:08.23psokolovsky_cr2: oh yeah, magician is his pet, and he seems to ignore any hints ;-)
13:08.34cr2:)
13:09.45Kevin2Hi
13:10.06goxboxlivegood morning kevin2
13:11.30cr2Kevin2: hi. my hermes does oops.
13:12.01psokolovsky_Hi!
13:13.00Kevin2Good morning all.
13:13.25Kevin2cr2: I didn't see an oops on my screen.  I'll see if I can retry with some kind of logging.
13:16.50cr2Kevin2: which kernel are you using ?
13:19.49Kevin2http://pastebin.ca/428282
13:19.58Kevin2No oops for me.
13:20.21cr2gcc version 4.1.2 ?
13:20.35Kevin2cr2: hh.org head with default config.  I added the printascii to printk for this last run.
13:20.39Kevin2cr2: Yes.
13:21.22cr2i have gcc version 4.1.1 from oe
13:22.01cr2S3C244X: core 283.200 MHz, memory 70.800 MHz, peripheral 35.400 MHz
13:22.05cr2again ?
13:22.13Kevin2That is what is in hh.org head.
13:23.08cr2hmm.
13:23.11cr2Console: switching to colour frame buffer device 60x53
13:23.11cr2<PROTECTED>
13:23.11cr2<6>S3C2410 Watchdog Timer, (c) 2004 Simtec Electronics
13:23.29cr2i crash before the Watchdog Timer line
13:24.59Kevin2Hrmm.  I have a custom haret..  Let's me upload it.
13:25.36Kevin2Try:  haret-20070407.exe
13:25.44cr2http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/arch/arm/mach-s3c2410/htchermes/htchermes.c?rev=1.9&content-type=text/x-cvsweb-markup
13:25.49cr212MHz
13:25.54cr2ok.
13:28.17cr2NAND device: Manufacturer ID: 0x00, Chip ID: 0xec (Unknown NAND 1MiB 3,3V 8-bit)
13:28.22cr2that's interesting.
13:31.04cr2crashed at the same place.
13:31.13cr2can you upload your zimage?
13:33.25Kevin2Okay, I pushed it up.  Note I have wifi and phone disabled when I boot.
13:34.01cr2yes.
13:35.53cr2oops.
13:36.11Kevin2I have a HERM100 -- what do you have?
13:36.12cr2at the same place.
13:36.18cr2200 i think
13:37.40cr2yes, 200
13:38.52cr2is your IR on or off?
13:39.11Kevin2On.  (That's how I got the log I put on pastebin.ca)
13:39.38cr2tried both. same result.
13:47.02cr2need some food :)
13:47.19*** join/#htc-linux eway (n=pocketir@149.254.192.192)
14:12.03*** join/#htc-linux BabelO (n=kvirc@lun34-2-82-238-28-28.fbx.proxad.net)
14:58.51goxboxliveHi BabelO: How is OPie2 going?
14:59.06BabelOhi goxboxlive
14:59.45BabelOi have a new computer at home since few days so i try to compile new kernel
15:00.02BabelOi see lot of change like battery support
15:00.58*** join/#htc-linux psokolovsky_ (n=psokolov@82.193.98.4)
15:01.18BabelOi do a little app for opie2 phone
15:01.21BabelOthat's all
15:01.34goxboxliveok
15:02.04goxboxliveHave you considerd trying out gsmd (suplied wiuth openmoko) instead. It works with Universal and magician now.
15:02.55BabelOi see some channel log about this :) so maybe it works for blueangel too
15:03.16BabelOand Ph5 send a patch for tty multiplex
15:03.43goxboxliveok
15:04.21BabelOand you ? any camera support ? openmoko screenshot
15:05.57goxboxliveNo we have no cam support yet. Yes i am running openmoko, but it's still buggy. Openmoko-dialer is working bether in a GPE image
15:06.39*** join/#htc-linux cyrill62 (n=cyrill62@92.249.146.195.dynamic.adsl.abo.nordnet.fr)
15:07.23BabelOnice ;)
15:15.05goxboxlivehow do i is space written in source like in \Progam files\appname
15:15.21goxboxliveit's not \program%20files\ allready tested
15:15.35goxboxliveit is not \program\/files\
15:15.48goxboxliveand not \program/\files\
15:15.54goxboxliveThen what's left
15:18.44BabelO\progra~1\
15:18.57BabelOor between "
15:19.11goxboxliveright
15:20.54cr2hi BabelO
15:21.04BabelOhi cr2
15:21.07cr2the battery support on ba is broken now.
15:21.36cr2but you can try wifi :)
15:21.36BabelOi see :( and i can't compile with suspend
15:21.48cr2suspend too ?
15:21.54cr2i will need it for himalaya.
15:22.13BabelOgot this error
15:22.17BabelOarch/arm/mach-pxa/built-in.o: In function `blueangel_pm_suspend':
15:22.17BabelOssp.c:(.text+0x1cdc): undefined reference to `blueangel_lcd_hw_off'
15:22.38BabelOyou have a working wifi ?
15:23.39cr2don't know, but i have corrected the power and irq settings.
15:23.59BabelOok
15:24.00cr2it wants to load the firmware.
15:25.32cr2BabelO: which boardid do you have?
15:26.12BabelOboardid 5
15:26.26cr2ok, so you are not affected.
15:26.59cr2i've done major changes in the video modes, boardid 6 may be broken now.
15:28.00BabelOi got a message from asylumed about a working lcd suspend resume patch
15:29.49cr2when?
15:30.23BabelOone month ago i think, he send a mail to kernel list
15:30.47cr2well, you can test it anyway.
15:31.37cr2since i have a himalaya now, and it's a 95% blueangel, i do something in this area :)
15:33.24BabelOi don't understand error message in ssp.c
15:40.49cr2which one?
15:41.22BabelOssp.c:(.text+0x1cdc): undefined reference to `blueangel_lcd_hw_off'
15:42.26cr2in the latest CVS ?
15:42.33cr2it has nothing to do with ssp.c
15:42.37BabelOyes
15:43.24*** join/#htc-linux g3gg0_ (n=g3gg0@ppp-82-135-74-79.dynamic.mnet-online.de)
15:45.24cr2ok, let me check.
15:47.41goxboxlivehmm that create a cab file was quite easy. I am going to create one for each image, and name it Openmoko, Opie and so on. And let people install their opie/gpe... image to partition 1.
15:48.23goxboxliveCr2: To have the icon you where talking about we have to add it to Kevins Haret. It follows the exe application.
15:50.28cr2BabelO: i have put my defconfig (change the acx100 to "module" there).
15:50.47cr2don't see any problems with blueangel_lcd_hw_off() function
15:51.20cr2goxboxlive: are you using \Storage Card ?
15:51.22BabelOok i try again initializing my kernel tree
15:51.32cr2BabelO: cvs up -dfP
15:51.36goxboxlivecr2: No i am using Program files\
15:52.01cr2afaik, it is also not portable.
15:52.04goxboxliveAnd the haret shortcut is showiung up in programs like regular wince applications
15:52.20cr2there are some E1 and such things in wince.
15:52.35goxboxlive??
15:52.43goxboxlivewhat do you mean by portable
15:53.20goxboxlivewhy should it be
15:55.45cr2to "national" wince
15:55.57cr2i think it was this picture http://linux4.be/pivot/index.php
15:57.05cr2read here http://www.ocpsoftware.com/products.php?nm=cecabmgr&pd=news
15:59.08cr2no.
15:59.11cr2http://www.kyz.uklinux.net/formats/wince_cab_format.html
15:59.17cr2this one.
15:59.39cr2Appendix B: a list of En% shortcuts
16:01.27goxboxliveok, so i can just use thoose instead of the path
16:04.41goxboxliveBut the appendix doesnt show start menu for handhelds, only the older devices
16:12.01*** join/#htc-linux WizMaui (n=WizMaui@62.112.90.250)
16:12.49*** part/#htc-linux WizMaui (n=WizMaui@62.112.90.250)
16:19.05*** join/#htc-linux Kevin2 (n=Kevin@207-237-52-30.c3-0.avec-ubr12.nyr-avec.ny.cable.rcn.com)
16:26.12Kevin2cr2: I sent a patch to kernel-discuss to fix the irq startup issue.  With that patch, one can boot hermes with wifi on.
16:28.23cr2ok, but i think it is a good idea to shutdown wifi/bt/lcd/backlight/phone/IR/whatever in haret on a per-machine basis.
16:29.08cr2i don't know how to do it right for the phone, but the rest is clear.
16:29.33cr2whatever may include the usb host and sound.
16:31.27cr2Kevin2: i don't know if fluffs reads kernel-discuss, maybe you should cc: him too.
16:34.14Kevin2Okay.
16:35.10Kevin2Do you know his email?
16:36.06Kevin2I found it.
16:36.45cr2the s3c24xx Kconfig system seems to be very fragile.
16:36.45cr2ok.
16:58.58Kevin2cr2: I'm telneted into Hermes.  :-)
16:59.29Kevin2http://pastebin.ca/428525
17:00.34Kevin2It's very slow..  Not sure why.
17:00.56cr2BogoMIPS        : 199.47
17:01.24cr2is it 400MHz ?
17:01.43cr2can you cat /proc/iomaps and /proc/interrupts ?
17:01.51Kevin2I made the s3c24xx_init_clocks(16934400) change
17:02.23cr2btw, can you create a haret+zimage+initramfs ?
17:02.37cr2what does your dmesg say ? are you logged over usb?
17:03.06Kevin2/ # cat /proc/interrupts
17:03.06Kevin2<PROTECTED>
17:03.06Kevin2<PROTECTED>
17:03.06Kevin2<PROTECTED>
17:03.06Kevin2<PROTECTED>
17:03.07Kevin2Err:          0
17:03.09*** join/#htc-linux tudenbart (n=willi@xdsl-213-196-251-216.netcologne.de)
17:03.40Kevin2Yes, I'm logged in via usb.  dmesg just shows what I had before.
17:04.37cr2283MHz clock ? but it was 400MHz with 12MHz xtal ?
17:05.43Kevin2There is no /proc/iomaps -- do you want /proc/iomem or /proc/ioports?
17:06.49Kevin2cr2: Did you want haret+kernel+initramfs for hermes specifically?
17:07.34cr2yes, but it does not help me right now. but it's a good idea in general.
17:08.07cr2a guy from xda-dev has tested your zimage on HERM300, but it seems that it crashed for him too.
17:08.27Kevin2http://pastebin.ca/428535
17:11.22cr2ok, what sub-device is missing now ?
17:11.30cr2adc ?
17:14.40cr2adc, pwm and spi is missing.
17:15.39*** join/#htc-linux rob_w (n=bob@p213.54.166.130.tisdip.tiscali.de)
17:25.37Kevin2cr2: http://www.handhelds.org/~koconnor/HTCHermes/hack-boot.exe
17:26.16Kevin2Note - this I'm using my Apache rootfs -- who knows what that will do to Hermes.  But it seems to get usb up and going.  (192.168.0.206)
17:27.20cr2Kevin2: excellent, i'll try it. we will refine the gui and ask the people for bug reports.
17:27.54cr2imho you should add it as a target to the haret Makefile
17:28.04Kevin2cr2: gui?
17:28.48Kevin2make linload KERNEL=../../kernel/herm-head/arch/arm/boot/zImage INITRD=../../kernel/initramfs-old/rootfs.cgz SCRIPT=dummy.txt
17:29.01Kevin2It is a target of the haret makefile.  Not sure I follow you.
17:29.56cr2ok, i'll look.
17:30.07cr2btw, you may pack the .exe with upx.sf.net
17:31.38Kevin2cr2: I doubt it would help.  The exe is about 100K plus kernel/initrd.  So, it is basically already compressed.
17:33.15cr2ok.
17:33.35cr2it's just my microSD is almost full :(
17:33.54cr2scp: /media/8203562D07/hack-boot.exe: No space left on device
17:36.12cr2hey Kevin, it booted and did not crash :)
17:36.40Kevin2cr2: Did you set ramsize to 64M in your default.txt?
17:37.19cr2mounting home on /proc/bus/usb failed
17:37.56cr2g_ether.ko is from another kernel version.
17:37.56Kevin2I had usb plugged in during the boot, and then I unplugged it and plugged it back in.
17:38.17Kevin2cr2: Yes - I just grabbed my apache rootfs.  :-/  It did work though.
17:38.54cr2i'm impressed :D we only need to reposition the modal box, right now it's on the top.
17:40.16cr2PUEN is missing.
17:42.22cr2what is the "Bad eraseblock" ? there was some talk on xda-dev about the hermes-specific nand formatting.
17:42.56cr2Bus 002 Device 100: ID 0bb4:0b04 High Tech Computer Corp.
17:43.41Kevin2cr2: Were you able to telnet into it?
17:44.29cr2Bus 002 Device 101: ID 0525:a4a2 Netchip Technology, Inc. Linux-USB Ethernet/RNDIS Gadget
17:44.36cr2yes.
17:44.55cr2maybe you should change the address ?
17:45.24cr2g_ether: version magic '2.6.20-hh4 mod_unload ARMv5 ' should be '2.6.20-hh6 mod_unload ARMv4 '
17:45.37cr2s3c2410-rtc s3c2410-rtc: setting the system clock to 2026-08-01 06:12:48 (1785564768)
17:46.57cr2but it's unbelievably slow.
17:49.07Kevin2Hrmm.  I'm going to yank that hack-boot.exe.  I'm afraid of what the apache specific stuff might do to hermes.
17:50.03*** join/#htc-linux g3gg0__ (n=g3gg0@ppp-82-135-91-20.dynamic.mnet-online.de)
17:50.14rmoravcikcr2: start pinging herms ip, sometimes it help me
17:51.57cr2rmoravcik: hi. do you know something about nand ?
17:52.15rmoravcikcr2: hi about bad blocks?
17:52.38cr2yes.
17:52.47cr2Kevin2: my LCD suspended ;)
17:53.38cr2rmoravcik: 1s ping time.
17:53.50Kevin2cr2: Yeah - I got that too.  :-)
17:54.10rmoravcikcr2: i have the same problem with bad blocks, but windows is working fine
17:54.14cr2Bad eraseblock 254 at 0x000fe000
17:54.14cr2Bad eraseblock 255 at 0x000ff000
17:54.14cr2Creating 1 MTD partitions on "NAND 1MiB 3,3V 8-bit":
17:54.15cr20x00000000-0x00100000 : "Whole Flash"
17:54.36cr2do you know how the flash is partitioned in wince ?
17:54.48cr21Mib is strange.
17:55.41cr2Kevin2: we need to assign the devices to the subsystems.
17:56.03cr2maybe htc is evil, and put the wifi on SDIO. like on athena.
17:56.06rmoravciki only know, rx3000 use intel psm file system
17:56.21cr2i'll prefer to use wifi :)
17:57.15cr2rmoravcik: it may be the SPL, and i many people have bricked their hermes by reflashing it.
17:57.47cr2there are some magic 8byte fields after each 512byte data block.
17:58.10cr2some bits in the magic field mean "bad block".
17:58.52cr2but my expertise on nand is close to zero, so somebody more knowledgeable should have a look.
17:59.16rmoravcikmy too
18:00.08rmoravciki checked nand with bootloader and it seems ok
18:00.16cr2ok.
18:00.28goxboxlivecr2: Now it's more userfriendly: http://gnulinux.biz/files/Universal/
18:01.35cr2goxboxlive: nice. i should try.
18:02.39cr2Kevin2 has created a hack-haret.exe That was really impressive. Just run an executable,  and bye-bye wince :D
18:03.10cr2we only need to sign it.
18:03.39rmoravciknext week i will try boot linux on my friend's herm
18:03.45Kevin2cr2: It looks like the apache junk is what slows down usb.  It works fine once I cleared up the rootfs.
18:04.00cr2should ask pof how they did it for their stuff.
18:04.21cr2Kevin2: can you recreate an .exe ? then we will get a report for HERM300
18:04.36Kevin2Hrmm.  Maybe not..  It seems to get progressively slower.
18:04.45cr2rmoravcik: the real hacking is on the ATI chip now.
18:04.49cr2memory leak ?
18:05.05Kevin2Let me clear out my printascii hacks too..
18:15.05goxboxliveHow is hack-haret working. What exactly does it?
18:15.41cr2haret+zimage+default.txt in one exe.
18:15.55cr2+initramfs+busybox
18:16.33cr2in principle, you can partition an sd card and fill it with this image.
18:17.03cr2we only need to find the PUEN gpio to autoreset the UDC.
18:18.32Kevin2Okay I've upped a version that is a little better:  http://www.handhelds.org/~koconnor/HTCHermes/herm-linux-20070407.exe
18:21.54goxboxliveok sounds goog
18:22.02goxboxlives/goos/good/
18:36.50cr2Kevin2: your new file crashes on HERM300. the older one works.
18:37.04cr2i have not tried your new file on HERM200
18:38.44Kevin2cr2: Is the person with herm300 using windows as a host?
18:39.39cr2don't know.
18:39.59Kevin2I turned off RNDIS in hopes it would get rid of stalls.  It might prevent windows logins, however.
18:40.07cr2i dislike rndis in general.
18:41.46cr2cool.
18:42.00cr2athena has ak4641
18:42.21cr2it's very close to hermes.
18:42.30cr2hrmm. also wifi ?
18:49.02cr2Kevin2: yes, he is using windows on the host.
18:52.49Kevin2cr2: Yeah, I saw - I placed a comment on the same forum.
18:53.29Kevin2Not much I can say about it - I don't know what the process is from Windows or if it will work at all.
18:57.01cr2ok.
19:47.13cr2Kevin2: is it possible to watch the i2c datastream on hermes ?
19:48.09goxboxliveI amde a cab file for the ramdisk image to, so it's easier for people to make a linux partition.
19:51.48cr2ok.
20:42.40goxboxlivebut anyway, they still need a vfat partition on the SD card.
20:42.57goxboxliveThey have a place to save the GUI image
20:43.34cr2who ?
20:44.16cr2goxboxlive: http://wiki.xda-developers.com/index.php?pagename=AthenaResearch
20:45.22goxboxlivenewbies
20:46.51cr2does the rndis really work ?
20:47.13goxboxlivehow are you able to gathered all this info without the device
20:47.19goxboxlivedont know
20:48.29cr2i've got some pictures.
20:50.26goxboxlive:)
21:26.29cr2BabelO: did you compile the kernel ?
21:31.51Kevin2cr2: I'm toying with the idea of adding a generic memory tracing capability to HaRET.  Other than that, I don't know of any way to trace i2c.
21:33.38cr2what is "generic memory tracing" ?
21:33.45cr2just a busy loop ?
21:34.28Kevin2Alter the mmu table so that writes to areas of memory cause a fault.  Then catch the fault, log what happened, emulate the instruction, and return to userspace.
21:34.56cr2not bad :)
21:35.31Kevin2So, for i2c, I'd alter the virtual mapping of 0x5a000000, and try to catch the writes that way.
21:42.47cr2i have already forgot the details, but ADC is used for the touchscreen.
21:42.57cr2pwm for the backlight
21:43.29*** part/#htc-linux rmoravcik (n=rmoravci@adsl-dyn49.91-127-60.t-com.sk)
21:43.34cr2both spi ports: one for the keyboard, the second unknown.
21:43.59cr2which SD is used for what is also unclear
21:44.12cr2where is wifi is uncertain.
21:53.25cr2the ext2fs read code in uboot is from grub
21:54.26cr2it should work on the raw block device. the code in psdread must have all the needed code.
22:05.40cr2good night.
22:09.06Kevin2Good night.
22:31.11*** join/#htc-linux goxboxlive (n=goxboxli@206.80-202-161.nextgentel.com)
22:32.15goxboxliveevening
22:54.14*** join/#htc-linux cyrill62 (n=cyrill62@arf62-1-82-237-248-201.fbx.proxad.net)
23:58.30*** join/#htc-linux kelly (n=kelly@ppp206-236.lns3.mel4.internode.on.net)

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