IRC log for #openezx on 20070625

00:00.12Fry-kunwhat the hell.. can't download 42PCLEAN.tar.gz from motorola
00:00.16Fry-kunkeeps failing
00:03.20Fry-kunkinda sad how firefox can't even resume file transfers o_o
00:04.59stefan_schmidtdrWyrM: Would be great.
00:05.58stefan_schmidtdrWyrM: I also like the idea of a kernel/initrd/mtd-utils package for flashing a kernel. Already though about this myself sometime ago, but never come to the point of writing some specs down for it.
00:33.44drWyrMstefan_schmidt: yes, is a simple way to do it. altough it would be much slower than rsd.
01:28.22drWyrMgood night!
01:28.25*** part/#openezx drWyrM (n=wyrm@201.2.68.83)
02:28.36Fry-kunookay, back from a break
02:28.51Fry-kunso.. 2.4 & motorola changes
02:29.07Fry-kunthey don't have the entire kernel there, do they?
03:53.23*** part/#openezx Fry-kun (n=kostya@h-66-166-233-66.snvacaid.covad.net)
04:12.25*** join/#openezx dion (n=dion@inhex.net)
05:18.30*** join/#openezx goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
06:24.00Hopscotchgood morning
06:25.30*** join/#openezx pleemans (n=peter@leemanpe.ugent.be)
06:38.38*** join/#openezx pleemans (n=peter@leemanpe.ugent.be)
06:47.35*** join/#openezx koen (n=koen@dominion.kabel.utwente.nl)
07:36.07*** join/#openezx z72ka-ntb (n=hermanj@tux2.software602.cz)
07:59.51*** join/#openezx z72ka-ntb_ (n=hermanj@tux2.software602.cz)
08:04.28*** join/#openezx jaebird (n=jae@53-89.netblk-69-41-89.coolaccess.net)
08:06.12*** join/#openezx florian (n=fuchs@217.146.132.69)
08:07.37floriangood morning
08:11.50*** join/#openezx ade|desk (n=adavey@194.200.143.249)
08:21.24*** join/#openezx leoncamel (n=leoncame@210.225.239.78)
08:25.37*** join/#openezx z72ka-ntb (n=hermanj@tux2.software602.cz)
09:09.01*** join/#openezx z72ka-ntb_ (n=hermanj@tux2.software602.cz)
09:13.46*** join/#openezx z72ka-ntb (n=hermanj@tux2.software602.cz)
09:16.58*** join/#openezx z72ka-ntb (n=hermanj@tux2.software602.cz)
10:48.50*** join/#openezx dion (n=dion@inhex.net)
10:53.00*** join/#openezx AlexZhang (n=alex@121.34.49.146)
11:03.11*** join/#openezx z72ka-ntb (n=hermanj@tux2.software602.cz)
12:30.01*** part/#openezx AlexZhang (n=alex@121.34.49.146)
12:44.39*** join/#openezx drWyrM (i=c88e3026@gateway/web/cgi-irc/ircatwork.com/x-334be35d61e99e17)
12:45.24drWyrMgood morning.
12:58.26stefan_schmidtdrWyrM: hi
12:58.32stefan_schmidtdrWyrM: <XorA> thesing: I just registered an #alsa-soc channel
13:45.16*** join/#openezx drath_ (i=vmaster@p5B07F4E3.dip.t-dialin.net)
13:50.33drWyrMstefan_schmidt: looks like they are preparing some major ASoC core changes.. :)
13:51.39drWyrMstefan_schmidt: We will probably have a lot of work for .23, there will be a lot of changes on pxa2xx on arm-linux too.
13:51.41stefan_schmidtdrWyrM: Could be. Not really worked in ASoC.
13:52.31stefan_schmidtdrWyrM: Yeah, many changes it seems. But .22 should be a bit easier.
13:58.52*** part/#openezx ade|desk (n=adavey@194.200.143.249)
14:04.01*** join/#openezx goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
14:29.11*** join/#openezx jaebird (n=jae@53-89.netblk-69-41-89.coolaccess.net)
15:02.14*** join/#openezx jaebird (n=jae@53-89.netblk-69-41-89.coolaccess.net)
15:08.14*** join/#openezx LoneStar (n=xinzhen@218.81.28.68)
15:10.40*** join/#openezx ade|desk (n=adavey@194.200.143.249)
15:19.43LoneStaris the kernel source of Z6 released?
15:22.53stefan_schmidtLoneStar: I see nothing on opensource.mototola.com...
15:24.55LoneStarmaybe you should try motorola.com instead :-D
15:25.28stefan_schmidtLoneStar: heh
15:26.43LoneStarI see a lot of stuff there, but no kernel
15:27.28stefan_schmidtLoneStar: I see nothing for Z6 there. Only E6 and that is missing kernel, indeed.
15:30.07drWyrMasking does not hurt: https://opensource.motorola.com/sf/discussion/do/listPosts/projects.sitewide/discussion.general_comments.topc1502
15:30.11drWyrM;)
15:32.20stefan_schmidtdrWyrM: Don't expect an answer. :(
15:32.41drWyrMyes, i know
15:32.59drWyrMthey havent replied the thread where i ask for E6 kernel source.
15:33.24drWyrMbut, placing OpenEZX URL on their forums is good anyway.
15:33.52*** part/#openezx ade|desk (n=adavey@194.200.143.249)
15:34.32stefan_schmidtheh
15:37.16*** join/#openezx agd5f (n=alex@c-69-143-17-173.hsd1.va.comcast.net)
15:38.56*** join/#openezx goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
15:45.46*** join/#openezx goxboxlive (n=goxboxli@176.84-48-210.nextgentel.com)
16:08.26*** join/#openezx dion (n=dion@inhex.net)
16:32.49*** join/#openezx CIA-24 (n=CIA@208.69.182.149) [NETSPLIT VICTIM]
16:36.57drWyrMhttp://bluez.cvs.sourceforge.net/bluez/utils/tools/hciattach.c?r1=1.46&r2=1.47
16:37.04drWyrMdone.
16:40.01stefan_schmidtdrWyrM: great
16:47.15LoneStarhttps://opensource.motorola.com/sf/frs/do/viewSummary/projects.rokrz6/frs , here is Z6
16:47.50*** join/#openezx TimRiker (n=timr@216.49.181.128)
16:48.29LoneStarI'm really curious about how a single-core phone works
16:48.46drWyrMLoneStar: kernel and everything.
16:48.52drWyrMit is single core?
16:49.11LoneStaraccording to linuxdevices, it is
16:50.02LoneStarbut i didn't see anything like a kernel there
16:50.14drWyrMhardhat*
16:50.28LoneStarisn't that the toolchain?
16:50.52drWyrMgcc* ?
16:51.23LoneStarwell...confused
16:52.05LoneStarhardhat==montavista IIRC
16:52.30drWyrMi am downloading it. painfully slow 15Kb/s
16:53.06drWyrMi tell you about as soon as it finishes.
16:53.17LoneStarOK ;-)
16:59.47koeneeeew, affix bluetooth
17:00.57koen"1. What's included in this source code distribution tarball?
17:00.58koen<PROTECTED>
17:00.58koen<PROTECTED>
17:00.58koen<PROTECTED>
17:06.50*** join/#openezx drWyrM (i=c88e3027@gateway/web/cgi-irc/ircatwork.com/x-163871aff03f6c98)
17:07.04koenso no kernel
17:07.13koenand crappy bluetooth
17:08.24drWyrMkoen: sorry. are you talking about z6?
17:10.48koenyes
17:10.49drWyrMz6 is 2.6 based.
17:11.02drWyrMhardhat* is the kernel, aint it?
17:12.00koenhardhat is an se-linux like sdk iirc
17:13.21koenaffix is so 2001
17:13.58drWyrMwyrm@p043628:~/Desktop$ cat hardhat-R60_G_80.31.15R.tgz.part | zcat | tar -xv | head -5hardhat/hardhat/kernel_includehardhat/linux-2.6.x/hardhat/linux-2.6.x/arch/hardhat/linux-2.6.x/arch/alpha/
17:14.50drWyrMit is a linux kernel source tree :)
17:14.58koennice
17:15.03*** join/#openezx ao2 (n=u@2001:1418:117:0:0:0:0:1)
17:15.11koenfunny how they call it 2.6.x
17:15.30koenI'd assume their engineer are competent enough to find out which kernel they use
17:15.34koenbut apparently not
17:18.03koenCC=$(COMPILERTOOLPATH)/arm_v6_vfp_le-gcc
17:18.09koenwhat kind of cpu is inthere?
17:19.40koenfreescale mxc it seems
17:20.30*** join/#openezx LuitvD (n=luitvd@beigetower/luitvd)
17:21.07LoneStarhow could you know?
17:21.19koenby looking at the defconfig
17:24.11koenCONFIG_ARCH_MXC=y
17:24.51koendrWyrM: http://www.freescale.com/webapp/sps/site/overview.jsp?nodeId=01m6cy2406
17:24.58LoneStar$ head linux-2.6.x/Makefile
17:24.58LoneStarVERSION = 2
17:24.58LoneStarPATCHLEVEL = 6
17:24.58LoneStarSUBLEVEL = 10
17:24.59LoneStarEXTRAVERSION =
17:24.59LoneStarNAME=Woozy Numbat
17:25.19koenthey did seem to patch in EABI
17:25.35koenbut I couldn't see how they'd handle vfp sanely without it
17:26.26LoneStarwhat is vfp?
17:28.27stefan_schmidtWow, Z6 2.6 kernel sources would be great.
17:28.49stefan_schmidtSeems it is time to buy a new phone then. ;)
17:28.51drWyrMgrep MACHINE_START arch/arm/mach-mxc91231/*
17:29.26LoneStardiscover how GSM signaling is processed for us, stefan_schmidt :)
17:29.40stefan_schmidtLoneStar: heh
17:31.54stefan_schmidtThis naming confuses me again. ROKR Z6 == ROZR Z6?
17:32.12LoneStarRIZR
17:32.19stefan_schmidteh, yes
17:32.30LoneStarthey changed its name in the last second
17:32.45stefan_schmidtSo the new one is ROKR E6?
17:33.33LoneStaryes, the new music phone
17:33.37stefan_schmidtk
17:33.45stefan_schmidtAnyway, food now.
17:34.59LoneStarE6 is another ezx phone with touchscreen
17:36.34LoneStardoesn't differ a lot with A1200
17:48.43drWyrMMotorola is evil.
17:49.34drWyrMaudio.c inside power_ic/ is a big ioctl switch(). everything comes from userspace.
18:06.25koenno, it's an ljap :)
18:47.27*** join/#openezx Fry-kun (n=kostya@h-67-100-89-239.snfccasy.covad.net)
18:50.18*** join/#openezx florian (n=fuchs@f049168091.adsl.alicedsl.de)
18:53.01florianre
18:58.07Fry-kunwb
19:06.16*** join/#openezx z72ka (n=hermanj@r5af183.net.upc.cz)
19:11.09*** join/#openezx drWyrM (n=wyrm@201.2.68.83)
19:38.12Fry-kunhi drWyrM
19:38.31drWyrMhi Fry-kun.
19:39.24Fry-kuni kinda ran off yesterday, got distracted by visitors
19:39.43Fry-kungot a question: what's the best way to get started
19:40.03Fry-kuni downloaded the motorola firmware, and a copy of vanilla 2.4.17
19:40.11Fry-kundiff shows a LOT of changes
19:40.22Fry-kunand i'm not sure where to start
19:40.43drWyrMyes, mostly are montavista changes.
19:40.45Fry-kundon't even know where's the entry point
19:41.10drWyrMthe TF specific bits are.. (let me look, just a min)
19:41.41Fry-kunwhy does montavista differ so much from vanilla, btw?
19:43.58drWyrMdrivers/mmc/* arch/arm/mach-ezx/hainan.c include/asm/arch/ezx.h
19:44.49drWyrMand driver/power_ic/ (for the 'power on card' register)
19:47.35Fry-kundrivers/mmc/ and arch/arm/mach-ezx/ are not available at all in vanilla...
19:47.55Fry-kundon't see include/asm/arch/ezx.h in either version...
19:48.07drWyrMi think that a1200 uses the same register as a780 for powering on the card. I think that there is something wrong with the gpio configuration.
19:48.16Fry-kundriver/power_ic/ also missing from vanilla
19:48.22*** join/#openezx TimRiker (n=timr@pdpc/supporter/bronze/TimRiker)
19:48.45Fry-kunoh, so it's implemented but broken?
19:48.48Fry-kuni see
19:48.54drWyrMinclude/asm-arm/arch-pxa/
19:49.03drWyrMif you havent configure it yet.
19:49.28Fry-kunah there it is :)
19:49.32drWyrMIn fact it is just a copy of the a780 mci code.
19:49.37Fry-kuni see
19:49.55Fry-kunso what's this gpio and how do i fix it?
19:50.39Fry-kunsort of like serial i/o interface?
19:51.21drWyrMdownload the pxa developers manual. It is a general purpose io.
19:52.13drWyrMyou use alternate functions to route the pins to the pxa internal mci controller.
19:52.22Fry-kunmany results in search, do you have a url?
19:52.38Fry-kunokay
19:53.22drWyrMlet me search for the url.
19:53.28Fry-kunthx
19:55.11drWyrMlook for filetype:pdf +clk_pio +chout on google.
19:55.38drWyrMit is a >10M file.
19:56.13Fry-kunIntelĀ® PXA270 Processor <- this one?
19:56.22Fry-kunballoonboard.org/hardware/300/ds/PXA270.pdf
19:56.42drWyrMxscale-freak
19:57.05Fry-kunwww.xscale-freak.com/XSDoc/PXA27X/28000304.pdf <- 2.3MB
19:57.17Fry-kunmissing something?
19:58.19drWyrMdamn. i am connected trough GPRS, to expensive to send it to you ;)
19:58.39Fry-kunoh, sorry
19:58.48Fry-kundidn't know.. try tinyurl.com ;)
19:59.28drWyrMno, i dont have the url. i am searching for it on google using keywords i know that are on the doc :)
19:59.48drWyrMsending you the 10M would be expensive hehehe
20:01.27drWyrMtry this: http://balloonboard.org/hardware/300/ds/PXA270-dev-manual.pdf
20:01.29Fry-kunyeah, no kidding.. i'll try to send fewer messages to save you the expense.  google returns 3 URLs, one of them is xscale-freak, but the pdf file is 2.3MB, not 10MB. its title is "IntelĀ® PXA27x Processor Family"
20:01.46*** join/#openezx pleemans (n=peter@d51A5E76A.access.telenet.be)
20:02.14Fry-kunyeah, this one is much larger (8.6MB) - and it didn't show up in my search results. should be fine now, though, thanks!
20:02.20ao2http://pubs.gumstix.org/documents/PXA%20Documentation/PXA270/
20:02.51Fry-kunthanks, ao2 :)
20:03.04drWyrMhi ao2.
20:03.16ao2I'm giving a quick look to the Specification update too.
20:03.21ao2hi drWyrM
20:04.15koenI wonder why moto went with freescale and stopped using TI and PXA
20:04.44koen(a good move, since PXA sucks, but a curious one)
20:04.56drWyrMkoen: One processor to do it all?
20:05.37Fry-kun..and in the darkness bind them :D
20:06.22ao2wow: http://bluez.cvs.sourceforge.net/bluez/utils/tools/hciattach.c?r1=1.46&r2=1.47
20:06.32drWyrMthe readability of their code improved a lot. We have to admit that! :)
20:06.40drWyrMao2: finally. ;)
20:06.47ao2and bluez-utils 3.12 has been released :)
20:08.58koenwith the bc patch?
20:09.32ao2downloading
20:11.12ao2yes, koen, let's ping OE bluez packagers :)
20:11.28koenPackaged contents of bluez-libs into /data/build/koen/OE/build/tmp/angstrom/deploy/glibc/ipk/armv5te/libbluetooth2_3.12-r0_armv5te.ipk
20:12.21ao2(ok, I'd better check first, next time...)
20:12.33koenI committed it 5 seconds ago :)
20:13.26ao2:) sweet, maybe you can also drop bcm2035-tool, drWyrM ?
20:13.51drWyrMdroping bcm2035-tool seems fine for me.
20:14.43koenit should be in the angstrom feeds later today
20:15.01koenso 'ipkg update ; ipkg upgrade' should do it for testers
20:15.06ao2thanks
20:28.40koenbluez 3.12 uploaded
20:30.52Fry-kunalright, so back to TF stuff... do i really have to read through that 1200-page document to get started?
20:31.15drWyrMno.
20:31.16Fry-kunthere's gotta be a simpler way
20:31.52drWyrMjust to familiarize with gpio, as you asked :)
20:32.29drWyrMcompare the mci bits on ezx.h on the 2.4 kernel with the ones on OpenEZX
20:32.47drWyrMi know that the detect card gpio is wrong.
20:32.57drWyrMit is probably the best starting point.
20:33.04stefan_schmidtdrath: Are you around tomorrow? I'll prepare some JTAG testing.
20:33.33Fry-kundrWyrM: okay, can do =)
20:36.21drWyrMthe detect card gpio is configured as a interrupt. Currently it is hardcoded on a1200-mci.patch
20:36.30Fry-kunstupid question: where in the OE tree is that file?
20:37.18ao2Fry-kun, OE is openembedded, take care ;)
20:37.36Fry-kunoops, right
20:37.48drWyrMFry-kun: work on the separate kernel tree that you set-up yesterday.
20:38.31drWyrMFry-kun: arch/arm/mach-pxa/ezx-a1200.c and include/asm/arch/ezx.h
20:38.42Fry-kungot it
20:43.50stefan_schmidtdrWyrM: Yeah, would be great.
20:44.11stefan_schmidtdrWyrM: But even single stepping through the kernel is really helpfull
20:46.57drWyrMstefan_schmidt: yes, there are still some irq ordering issues on the pcap driver on some rare conditions. And using it on the 2.4 kernel will help on the asoc driver. :)
20:47.39stefan_schmidtdrWyrM: Jup
20:48.30stefan_schmidtdrWyrM: _If_ I get it working a add the infos to the wiki. Testpad naming, OpenOCD config, ...
20:49.15Fry-kundrWyrM: something to do with "#define GPIO_MMC_DETECT         GPIO_TFLASH_INT"?  not yet sure what's what
20:49.18koenyou guys know about the asoc irc channel, right?
20:49.36drWyrMstefan_schmidt: i never tried JTAG on my phone (in fact i never disassembled it). I am waiting for the e680 to plug a scope and try jtag.
20:49.55drWyrM: ircname  : Daniel Ribeiro
20:49.56drWyrM| channels : #openezx #alsa-soc
20:50.00drWyrM;)
20:50.39stefan_schmidtkoen: Already in. :)
20:50.50koennice
20:51.07stefan_schmidtdrWyrM: No need for a scope if I got it running, but always good to have on in reach. ;)
20:51.47drWyrMthe scope is to assure that i have the right clock on the sspclk3
20:52.08stefan_schmidtah, ok
20:52.32drWyrMeverything is connected together (network mode) on ssp3 of the pxa. bt, pxa, pcap and bp, all connected together.
20:52.53stefan_schmidtdrWyrM: wrt E680. Last time I checked it still waits for shipping in .de
20:52.55stefan_schmidt:(
20:53.42stefan_schmidtPlan would have been better, but the prices had driven me crazy.
20:55.00Fry-kundrWyrM: when is the latest time I can catch you online?
20:56.49drWyrMFry-kun: i am always online from 09:00 to 15:00 (GMT-3). Other than that i cant promess. :)
20:57.00*** join/#openezx flo_lap (n=fuchs@f049168091.adsl.alicedsl.de)
20:57.18Fry-kunI'm in the PST timezone (GMT-8) and usually leave around midnight
20:58.22drWyrMbut you have my email already. I always reply as fast as possible :)
21:00.09Fry-kunah okay, gotcha. it's just that i need to go do some work. you'll probably be sleeping by the time i get home (7-8 hrs from now)
21:00.31drWyrMyes, probably :)
21:00.35Fry-kunguess i'll just keep reading the code and poking it here and there
21:01.17drWyrMtrial and error usually works.. ;)
21:01.21Fry-kunspeaking of poking, how will i know when i've got something right? I need a 2nd partition on my TF?
21:02.29Fry-kuncan you give me a quick example on how to display info on screen when the kernel starts up? and how to read the buffer (since it goes off screen very fast)
21:03.10drWyrMwhen you see something like:
21:03.14drWyrM<6>mmcblk0: mmc0:bffc SU02G 1985024KiB
21:03.40Fry-kunthat would probably mean TF was recognized
21:03.57drWyrMit depends on your card, but if you see "mmcblk0: *" and * is not an error, then it is ok.
21:04.18drWyrMthere is no way to scroll back the kernel messages on the frame buffer.
21:05.02drWyrMyou could use a serial console, but then you would need to build an rs232<->ttl voltage converter.
21:05.15drWyrMand open your phone.
21:05.57Fry-kund'oh
21:06.34Fry-kuner... i read on the wiki that the mini-USB port on a1200 doubles up as a serial port
21:06.50Fry-kunbut would still need to build the converter, right?
21:07.02drWyrMyes, but we dont support EOC yet :)
21:07.12drWyrMyes, you would still need the converter.
21:07.20Fry-kunbah...
21:07.45drWyrMEOC = Emu One Chip, it is a new chip to control EMU, on a780 it is controlled by pcap and some analog switches.
21:08.47Fry-kunthen it's pretty much useless unless i build that converter... since kernel panic message fills up the screen with register info
21:09.01drWyrMhumm...
21:09.04drWyrMit shouldnt.
21:09.05drWyrM:)
21:09.27Fry-kunwell, that's what it does with my current build
21:09.36drWyrMit should panic because cant mount rootfs, on a 2 lines message.
21:09.53drWyrMcan you take a picture for me?
21:10.16Fry-kunsure, wait a minute or two
21:14.12drWyrMFry-kun: do you know that there is a bounty on motorolafans.com to extend a1200 support to the same level of a780?
21:15.00drWyrMit is currently on US$80, an extra reason to keep you motivated ;)
21:15.33Fry-kunguess i can't take a picture now - coworkers' phones don't have closeup mode
21:16.00Fry-kunand it's too small for far shots.. ironic, since a1200 DOES have the closeup mode
21:16.13Fry-kunoh nice
21:16.24Fry-kun^ about the bounty
21:16.39drWyrMno probs, send the pic when you can.
21:16.49drWyrMyou can try also to quilt pop -a
21:17.16drWyrMand then quilt push each patch and try to boot the kernel, to determine where the buggy code is.
21:17.38Fry-kunso vanilla kernel shouldn't crash?
21:18.07drWyrMyou have to apply at least the ezx-core.patch.
21:19.03Fry-kunmake -f patches/Makefile.OpenEZX CROSS_COMPILE=../../build/tmp/cross/bin/arm-angstrom-linux-gnueabi- zImage-a1200-vanilla
21:19.07drWyrMi mean, apply only core.patch, try to boot, apply the next, try to boot again.. until you determine which one adds the panic :)
21:20.03Fry-kunquilt push ezx-core.patch
21:20.05drWyrMquilt pop -a; while true; do quilt push; make ...; boot_usb ...; done
21:20.26Fry-kunyeah, got that - just checking that i'm issuing right commands
21:21.14drWyrMjust quilt push. it is a stack like patching system.
21:21.59Fry-kun`Now at patch patches/patch-2.6.21.4`
21:22.28drWyrMpush again. at least core.patch is needed.
21:23.24Fry-kunoooh okay
21:23.34Fry-kungot it now :)
21:24.00Fry-kunmake -f patches/Makefile.OpenEZX CROSS_COMPILE=../../build/tmp/cross/bin/arm-angstrom-linux-gnueabi- zImage-a1200
21:24.00Fry-kunmake: `zImage-a1200' is up to date.
21:24.15drWyrMrm -f zImage-a1200
21:24.16Fry-kundoesn't it see that files changed? o_O
21:24.21drWyrM:)
21:24.50Fry-kunalright, it started compiling now
21:24.51drWyrMno, it doesnt. you can rm -f zImage-a1200 or touch patches/defconfig-a1200
21:25.22Fry-kuni thought make was supposed to always check the freshness.. oh well
21:26.07drWyrMit is not the standard kernel build system. zImage-a1200 depends only on defconfig-a1200 :)
21:26.27Fry-kunaaah, i see
21:26.57drWyrMbasicaly it is a .config switcher, to compile images for all the phones with a single command.
21:27.16Fry-kunyeah, that's convenient
21:29.47Fry-kunok, better now
21:30.01Fry-kunstill says kernel panic but no register info
21:30.22drWyrMcant mount rootfs?
21:30.31Fry-kunVFS: Cannot open root device "3e00" or unknown-block(62,0)
21:30.38drWyrMyes.
21:30.38Fry-kunPlease append a correct "root=" boo topion
21:30.56Fry-kunKernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(62,0)
21:31.06drWyrMpush the next patch, and determine on which patch it dumps registers.
21:31.28Fry-kunokay
21:32.26drWyrMif you apply successfully until a1200-mci.patch you can stop pushing. a1200-mci is where you need to work. :) any later patch is not useful until you fix mci.patch.
21:36.04Fry-kunso far so good
21:36.18Fry-kunhow many patches are there?
21:36.39drWyrMcat patches/series
21:36.40Fry-kunsame as # of .patch files?
21:37.08Fry-kungotcha
21:38.45drWyrMstefan_schmidt: remember when i told you that leaving an OpenEZX URL on opensource.motorola was useful?
21:39.10drWyrMstefan_schmidt: Fry-kun discovered OpenEZX there ;)
21:39.27stefan_schmidtdrWyrM: heh, nice
21:39.53stefan_schmidtdrWyrM: So helpful for other devs but uninteresting for Moto
21:40.32drathstefan_schmidt: i'll be available from around 10am to 10pm
21:41.15stefan_schmidtdrath: Great. Plan is to fetch all the needed stuff around 11:00, prepare setup and start tests.
21:41.42stefan_schmidtdrath: Should be finished before lunch. ;)
21:42.13drWyrMor you will have lunch on dinner time ? ;)
21:42.13drathstefan_schmidt: everything's supposed to work now ;)
21:42.15stefan_schmidtdrath: cbrake suggested that it only works with the binary lib for him on pxa27x
21:42.40drathstefan_schmidt: yeah, libftdi chickens out on larger queues
21:42.58stefan_schmidtdrath: So I will test JTAGKey + binary lib first
21:42.59drathstefan_schmidt: but that's only a minor issue - i guess the driver is supposed to split larger buffers
21:43.19stefan_schmidtdrWyrM: heh
21:43.21drathstefan_schmidt: you'll only experience problems when trying to flash
21:43.37drathstefan_schmidt: because the OpenOCD tries to build buffers larger than what libftdi accepts
21:44.11drathanyway - too much beer here, so i guess it's better to stop now
21:44.15stefan_schmidtdrath: If I get fast a working solution with these components I'll test some more combinations with tiny, OpenMoko debug board, wiggler and libftdi
21:44.36stefan_schmidtdrath: ok, see you tomorrow
21:44.44drathstefan_schmidt: right, bye stefan_schmidt
21:46.15*** join/#openezx pmarques (n=pmarques@bl7-3-20.dsl.telepac.pt)
21:46.27drWyrMhi pmarques.
21:46.44pmarqueshi Daniel
21:47.32stefan_schmidtwb pmarques
21:48.26pmarquesthanks... I've been kind of busy lately, but I hope to have a little more spare time now
21:50.30drWyrMpmarques: good. Fry-kun started working on a1200. We are all waiting for some feedback from you about the fb driver.
21:51.12pmarquesI think the fb driver already works, but fbcon doesn't support 18bpp.... I was thinking about testing it by using a small executable running from initramfs
21:51.35pmarqueswhat's the max size of a kernel booted using boot_usb?
21:51.49drWyrMthere is no limit.
21:52.10pmarquesthere is at least a 48Mb limit ;)
21:52.31drWyrM32Mb. :)
21:52.38stefan_schmidtdrWyrM: There is a limit inside blob for upload into RAM. 1 MB IIRC
21:53.09drWyrMstefan_schmidt: this limit only applies when copying from flash.
21:53.36stefan_schmidtdrWyrM: Are you really sure? I don't think so
21:53.47drWyrMi booted a kernel (1.1M) plus a mtd-ram image >5M
21:53.54drWyrMyes, i am :)
21:54.15stefan_schmidtdrWyrM: So you uploaded bigger kernels already?
21:54.38pmarqueswell, I've already booted an image that is 1162924 bytes long, so that is at least a little over 1Mb
21:54.57drWyrMyes. I have tested it already. i booted a 1.4Mb kernel from flash using blob2 too.
21:55.33stefan_schmidthmm, so why do I still think this limit? :)
21:55.54drWyrMthe right limit is 896Kb when booting from the original blob from flash. No limits when booting from ram.
21:57.08pmarquesbut you might be wrong about th 32Mb limit, though... the kernel must decompress itself and the initramfs so it needs space to do so
21:57.52pmarquesalthough I wasn't thinking of trying to load something thaaat big.... 5Mb seem more than enough for me :)
21:58.39stefan_schmidtpmarques: I think there is somethink like fbtest around. Not sure if it works and can really help you.
21:58.52drWyrMpmarques: yes, you are right. You need at least zImage size + vmlinux size. But when decompressing you can make use of the 16mb discontiguous memory.
21:59.22drWyrMbut i dont know if a1200 has this extra 16mb.
21:59.46drWyrMon a780 i think it cames from the DoC, and iirc a1200 doesnt have a DoC
22:01.05pmarquesthat would be weird... RAM should be much faster than a DoC
22:01.18drWyrMahh, pmarques, how much user available memory do e6 have?
22:01.44pmarques48Mb
22:01.48drWyrMpmarques: it is like a combo. 64Mb DoC + 16Mb RAM on the same package.
22:02.06drWyrMi mean flash memory for user data.
22:02.21pmarquesahhh... that makes more sense... although I think the E6 has a lot less flash than that
22:03.21pmarques(pmarques finds a fbtest for Solaris :P )
22:04.01drWyrMa780 has more than 40Mb flash for user data. a1200 has only 8Mb. Thats why i think that moto removed the DoC on newer phones.
22:04.41stefan_schmidtpmarques: http://linux-fbdev.sourceforge.net/ should be in cvs
22:08.58stefan_schmidtpmarques: http://linux-fbdev.cvs.sourceforge.net/linux-fbdev/fbtest/
22:09.33stefan_schmidtpmarques: But it looks pretty old and unmaintained. Not sure if it will really help you.
22:09.37pmarqueslooks like the E6 has 16Mb of flash
22:10.34pmarquesstefan: it looks pretty empty, too :)
22:10.41drWyrMpmarques: 16Mb available to the user?
22:10.53pmarquesnop, 16Mb total
22:11.22drWyrMAre you sure? i think it is 32Mb strataflash.
22:11.22pmarques"df" shows: /dev/mtdblock8            6144       980      5164  16% /ezx_user
22:11.24stefan_schmidtpmarques: ups, no idea where it is gone. Perhaps they have moved to another repo or merged with another project. No idea, sorry.
22:11.27pmarques/dev/mtdblock9            8064      2632      5432  33% /ezxlocal
22:12.17drWyrMit is like the a1200.
22:12.29pmarquesdon't worry... I was thinking about writing an extremely simple program that just writes sequentially to /dev/fb so that I could see how the bits are getting mapped
22:13.23pmarquesfrom motorolafans, I got the impression that most applications that worked for the A1200 also worked for the E6, so the hardware must be similar at least
22:13.35stefan_schmidtpmarques: Sounds good. Just test order and colors sequenzes.
22:13.59drWyrMpmarques: dont forget about the other partitions.. bootloader+kernel+rootfs+itunescrap+FOTA+usermemory...
22:15.03pmarquesit's a shame that motorola didn't try to open up even a little bit more... everything about the phone seems to be sub-par... kernel 2.4... old ABI instead of EABI... Qt 2.x
22:15.03LuitvDfota?
22:15.29drWyrMpmarques: Moto just released kernel for Z6.
22:15.43drWyrMits 2.6 based, much better coding style.
22:15.47stefan_schmidtLuitvD: Firmware Over The Air
22:16.05LuitvDstefan_schmidt: okay, that's worth crying over
22:16.33LuitvDbut I guess it's not as bad as ISP-forced ads
22:17.28drWyrMLuitvD: your ISP cant brick your computer.
22:17.28*** join/#openezx Fry-ku1 (n=kostya@h-67-100-89-239.snfccasy.covad.net)
22:17.38stefan_schmidtLuitvD: Getting a new firmware for *my* phone from the provider behind my back is worse enough
22:17.43pmarqueshow close is it to mainline? is it still monta-crap based?
22:18.14stefan_schmidtpmarques: freescale crap based this time. With MV patches of course. :)
22:18.19stefan_schmidtpmarques: 2.6.10...
22:18.36*** join/#openezx Fry-kun (n=kostya@h-67-100-89-239.snfccasy.covad.net)
22:18.42drWyrMyes, montavista based. But much more readable than the 2.4 series i think.
22:18.46stefan_schmidtpmarques: But at least no need to port to the big chances introduced in 2.5
22:19.18pmarquesif it is done right, we don't even have to compile a new kernel for it... ;)
22:19.37LuitvDdream on ;)
22:19.58drWyrMpmarques: i repeat LuitvD. dream on.
22:19.59stefan_schmidtpmarques: heh, I just run phones with a selfcompiled kernel. :)
22:20.29LuitvDdoes anyone know if the Z8 is EZX-based?
22:20.47stefan_schmidtTime to sleep. See you guys tomorrow.
22:20.53pmarquessee ya
22:20.58drWyrMbye stefan, good night.
22:21.12LuitvDI heared it has an OMAP processor... guess that makes it non-EZX, right?
22:21.18LuitvD'night stefan_schmidt
22:21.47drWyrMLuitvD: dont know. Z6 uses a freescale mx processor.
22:21.54LuitvDanyone? rizr z8? could it be?
22:22.13pmarquesOMAP is also a ARM with extra stuff... so it might be the same software stack on top
22:22.14LuitvDI do want to buy one of those z8's
22:22.29LuitvDpmarques: no, it's Symbian + UIQ
22:22.52LuitvD(versions 9.2 and 3.1 respectively)
22:22.59pmarquesok
22:23.03LuitvDbrb, brushing teeth :)
22:23.48pmarquesI stopped looking at other phones when after I bought mine, so that I wouldn't feel bad if a better one appeared in the meanwhile :)
22:25.38drWyrMi like my a780 :) If we manage to completely replace the FW, i will only replace it when it dies.
22:26.21drWyrMi love having a GPS on the phone. Shame that copilot doesnt have maps for Brasil.
22:27.34ao2bye
22:27.48drWyrMbye ao2. good night.
22:28.18pmarquesbye
22:29.36drWyrMFry-kun: any progress?
22:34.28drWyrMbb in 20 mins.
22:34.51*** join/#openezx Fry_kun (n=Fry-kun@h-67-100-89-239.snfccasy.covad.net)
22:39.17*** join/#openezx Fry_kun (n=Fry-kun@h-67-100-89-239.snfccasy.covad.net)
22:39.29*** part/#openezx Fry_kun (n=Fry-kun@h-67-100-89-239.snfccasy.covad.net)
22:39.38*** join/#openezx Fry_kun (n=Fry-kun@h-67-100-89-239.snfccasy.covad.net)
22:40.54*** join/#openezx Fry-kun (n=kostya@h-67-100-89-239.snfccasy.covad.net)
22:52.52*** part/#openezx Kun (n=kun@203.144.27.9)
23:00.24*** part/#openezx Fry-kun (n=kostya@h-67-100-89-239.snfccasy.covad.net)
23:21.07drWyrMib.
23:28.07drWyrMthere is a positive point in newer phones not having a DoC.
23:28.29drWyrMwe will not need to bother supporting it ;)
23:32.29pmarques:)
23:33.09pmarquesI have to go now... I'll let you guys know when I've made some progress
23:33.27drWyrMpmarques: bye, thank you.
23:37.26drWyrMpurl: i think that you are the only one left awake. can i talk to you?
23:37.27purlYou think that you are the only one left awake. can i talk to you?
23:38.13drWyrMpurl: good night.
23:38.14purlGood night., good! Have a cookie... oh no! The cookie jar is dry!
23:38.35drWyrMbye all. :)
23:38.43*** part/#openezx drWyrM (n=wyrm@201.2.68.83)

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