| 00:00.01 | PITABoy_mobile | oh |
| 00:00.07 | cr2 | Kevin2: on omap850 the hw i2c port is used. |
| 00:00.23 | dcordes | is there something like RE for dummies? |
| 00:00.26 | cr2 | ok, but it may be bitbanging spi. |
| 00:01.07 | cr2 | dcordes: no, it's just experience. |
| 00:01.36 | cr2 | dcordes: the best thing you can do with haret is to watch the irq changes and memory/address space accesses. |
| 00:02.19 | cr2 | Kevin2: i've indentified all the 90c and 910 bits used by ATI SD driver. now i'd like to see them in action... |
| 00:02.21 | dcordes | I guess I don't understand the whole haret research thing. Is there any documentation about it? |
| 00:02.29 | dcordes | detailing the progress |
| 00:02.47 | Kevin2 | cr2: It closely follows the i2c spec. There are two gpios (sda and scl). The scl pin is always an output and it goes on/off in regular beats. The sda pin is either high or low, and is generally only changed when scl is low. After the address is sent, the last bit indicates whether or not the host wants to read/write. The transmission starts off with an sda change while scl is high (an i2c start event). etc. |
| 00:02.53 | dcordes | s/process /process / |
| 00:03.03 | dcordes | ~lart dcordes |
| 00:03.03 | apt | brandishes Excalibur! "With this sword, I vanquish thee, dcordes!" and lops off dcordes's head |
| 00:03.50 | cr2 | Kevin2: do you have the third gpio in this process lowered before the transmission, and set after it ? |
| 00:03.59 | Kevin2 | No - only two gpios. |
| 00:04.05 | cr2 | ok. |
| 00:13.05 | *** join/#htc-linux the_sys0p (i=the_sys0@gateway/tor/x-b86a292e69237d37) |
| 00:20.04 | *** join/#htc-linux AC_Mast (i=s@gateway/tor/x-b61b1ca00a170330) |
| 00:23.29 | *** join/#htc-linux PITABoy (i=Windoze@adsl-221-110-242.rmo.bellsouth.net) |
| 00:45.30 | pH5 | Kevin2: not sure, at least it uses the omap850's i2c controller to access it. |
| 01:44.54 | *** join/#htc-linux AC_Mast (i=s@gateway/tor/x-1493e91176c71df1) |
| 02:23.59 | *** join/#htc-linux PITABoy (i=Windoze@adsl-221-110-242.rmo.bellsouth.net) |
| 02:24.53 | PITABoy | QPST is cool |
| 02:26.06 | PITABoy | I wonder if I could change my number??? |
| 02:26.14 | PITABoy | i dont want to mess it up thou |
| 03:00.20 | *** part/#htc-linux st_lim (n=st_lim@cm119.epsilon73.maxonline.com.sg) |
| 03:49.32 | *** join/#htc-linux Zoolooc (n=fredsiba@p54953B1D.dip0.t-ipconnect.de) |
| 03:49.59 | *** join/#htc-linux st_lim (n=st_lim@cm119.epsilon73.maxonline.com.sg) |
| 05:34.47 | *** join/#htc-linux dblrbl (n=polyrhyt@c-24-18-133-36.hsd1.mn.comcast.net) |
| 05:35.11 | dblrbl | hello all |
| 05:36.28 | Kevin2 | Hi |
| 05:38.05 | dblrbl | How goes the Linux on Kaiser progress? The log is not yet posted from yesterday |
| 05:39.35 | Kevin2 | No new progress in the last few days (as far as I know). |
| 05:39.56 | dblrbl | Aha! The log is now posted, I will catch up |
| 05:43.35 | dblrbl | lol |
| 06:24.49 | dblrbl | is there a compiled zImage that you guys have been trying to boot on the Kaiser, or will I need to roll my own/ |
| 06:24.57 | dblrbl | *? |
| 06:25.52 | dblrbl | ahh I am too late -- it is not yet late for me on the west coast |
| 06:31.07 | Kevin2 | dblrbl: We're still a ways off before we'll having a kernel for msm. |
| 06:31.45 | dblrbl | ok cool, I thought I had missed something |
| 06:31.53 | dblrbl | any way I can help in the meantime? |
| 06:38.31 | Kevin2 | dblrbl: It's hard to port linux without cpu docs. You could ask Qualcom for the docs, but I doubt that would work. |
| 06:38.43 | Kevin2 | The other option is to reverse engineer the info from wince. |
| 06:38.57 | Kevin2 | If I understand correctly, the irq registers are the biggest unknown right now. |
| 06:39.13 | *** join/#htc-linux GPFerror (n=GPF@76.187.41.132) |
| 06:39.15 | dblrbl | Ok, good to know |
| 06:39.27 | dblrbl | I know the docs have been asked for more than once |
| 06:40.03 | dblrbl | it drives me mad how slow my Kaiser runs compared to the clock speed |
| 06:41.08 | dblrbl | I'm hoping Linux will be able to help with that |
| 06:44.35 | dblrbl | maybe we should go old school and social-engineer the docs out of Qualcomm |
| 06:59.35 | Kevin2 | Good night. |
| 07:05.58 | dblrbl | good night |
| 07:59.51 | *** join/#htc-linux |dion| (n=dion@inhex.net) |
| 08:38.54 | *** join/#htc-linux pH5 (n=ph5@p5485FBAB.dip.t-dialin.net) |
| 10:15.31 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 10:19.27 | *** join/#htc-linux goxboxlive (n=goxboxli@81.80-202-132.nextgentel.com) |
| 10:22.11 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 10:26.54 | *** join/#htc-linux rob_w (n=bob@X0511.x.pppool.de) |
| 10:50.27 | pH5 | www.gmail.com |
| 11:15.46 | *** join/#htc-linux Keizer (n=keizer@c-68-54-193-215.hsd1.fl.comcast.net) |
| 11:18.30 | *** join/#htc-linux lieter (n=lieter@a62-251-100-248.adsl.xs4all.nl) |
| 11:19.41 | goxboxlive | Kevin2: Do i have to use the name "zImage" or could it be a autogenerated name from Openembedded? |
| 11:19.51 | goxboxlive | pH5: good morning. |
| 11:20.14 | pH5 | hey goxboxlive |
| 11:21.50 | goxboxlive | pH5: did you see Kevin2 work yesterday. It's quite easy way of booting/installing linux now. |
| 11:22.29 | goxboxlive | hmm Angstrom have moved over to gcc-cross-4.2.2. |
| 11:24.08 | pH5 | goxboxlive: no, it hasn't. |
| 11:24.16 | pH5 | 2008.1 is unstable testbed |
| 11:25.02 | goxboxlive | yes, it's 2008.1 who has moved over to 4.2.2. I am building a kernel now. And it started to build gcc-cross-initial-4.2.2 |
| 11:28.51 | pH5 | loop-mounted rootfs? nice idea, this also makes it easier for us, as we can continue just providing the tar.bz2 rootfs images |
| 11:30.17 | goxboxlive | pH5: right, i am making one for Magician now. I can upload it for you, but let me know when you have got it because my Qtopia image includes Helix which i am not allowed to share. Only for dev. |
| 11:30.43 | pH5 | heh, yes please. |
| 11:31.31 | goxboxlive | You only need a sd card with a vaft partition bigger than 128 MB :-) . |
| 11:31.45 | pH5 | 128mb is too small? |
| 11:32.46 | goxboxlive | No, it's enough |
| 11:32.56 | goxboxlive | damn gcc-cross inital failed building |
| 11:33.10 | pH5 | really? |
| 11:33.18 | pH5 | I'm using it since months ago. |
| 11:33.21 | pH5 | hm.. |
| 11:34.06 | goxboxlive | But i did a pull today |
| 11:34.12 | pH5 | goxboxlive: can you pastebin the log? |
| 11:34.14 | goxboxlive | and it where rebuilded |
| 11:34.16 | goxboxlive | yes |
| 11:34.39 | pH5 | if you just want to build the kernel, you can revert gcc-cross-initial back to 4.1.2 in angstrom-2008.1.conf. |
| 11:34.49 | pH5 | I'll try to rebuild and reproduce. |
| 11:35.05 | goxboxlive | pastebin.no/2796 |
| 11:37.46 | pH5 | oh dear, that ugly psignal issue. |
| 11:39.59 | goxboxlive | :-) |
| 11:40.39 | *** join/#htc-linux kiko_ (n=Kiko@st0192.nas931.soka.nttpc.ne.jp) |
| 11:49.34 | *** join/#htc-linux rob_w (n=bob@X035d.x.pppool.de) |
| 12:48.35 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 13:00.30 | *** join/#htc-linux rob_w (n=bob@X035d.x.pppool.de) |
| 13:15.36 | pH5 | goxboxlive: http://en.pastebin.ca/779796 <-- this change fixes the gcc-cross-initial build for me. |
| 13:16.37 | *** join/#htc-linux Marex-notebook (n=marex@vasut.kolej.mff.cuni.cz) |
| 13:17.31 | *** join/#htc-linux AC_Mast (i=s@gateway/tor/x-66faef9a8403ba05) |
| 13:20.40 | *** join/#htc-linux Mineque (n=Mineque@195.60.65.50) |
| 13:24.15 | *** join/#htc-linux Keizer (n=keizer@c-68-54-193-215.hsd1.fl.comcast.net) |
| 13:25.05 | *** join/#htc-linux rmoravcik (n=rmoravci@adsl-d100.84-47-105.t-com.sk) |
| 13:27.48 | *** join/#htc-linux TBBle_ (n=tbble@2001:388:f000:0:0:0:0:2d) |
| 13:38.02 | *** join/#htc-linux rmoravcik (n=rmoravci@adsl-d100.84-47-105.t-com.sk) |
| 14:18.32 | *** join/#htc-linux the_sys0p (i=the_sys0@gateway/tor/x-f2668533cddc6dda) |
| 14:19.25 | *** join/#htc-linux dcordes (n=user20@f048011150.adsl.alicedsl.de) |
| 14:19.27 | dcordes | hi |
| 14:19.31 | pH5 | hi dcordes |
| 14:22.04 | Zoolooc | hi ppl |
| 14:22.10 | dcordes | hey Zoolooc |
| 14:26.51 | cr2 | hi |
| 14:27.15 | cr2 | goxboxlive: i've tested qtopia yesterday, and have some questions. |
| 14:47.55 | dcordes | how do you guys usually decide on which alias to chose for a machine? eg in case of kaiser- why not calling it tytnII? |
| 14:53.32 | cr2 | dcordes: htc hardcodes the names into its bootloader |
| 14:54.04 | dcordes | I see. Is there any place where those names from the bootloader are listed? |
| 14:54.18 | cr2 | tytnII is something invented by brainless marketing people :) |
| 14:54.44 | dcordes | BWL0r |
| 14:54.48 | cr2 | the "short" names are reported by wince. |
| 14:56.43 | cr2 | yeah, by those who get their batchelor degree by tschibo ;) |
| 14:57.02 | dcordes | not neckermann? |
| 14:57.09 | cr2 | lol |
| 14:57.47 | cr2 | hmm. i'll look what kaiser resports as its name. |
| 14:58.04 | dcordes | you got one there? |
| 14:58.12 | cr2 | no. |
| 14:58.16 | cr2 | but the bootloader. |
| 14:58.23 | dcordes | ah ok |
| 14:58.54 | cr2 | it says "Kaiser SPL EVT" |
| 15:00.29 | cr2 | and there are 2 supported LCDs. toppoly and sony. |
| 15:00.42 | dcordes | propably SPL EVT defines windows version? |
| 15:01.23 | cr2 | no. |
| 15:01.38 | cr2 | i forgot what is EVT. |
| 15:02.22 | dcordes | and SPL? |
| 15:04.09 | cr2 | secondary program loader or something like that. |
| 15:04.19 | cr2 | it's the second level bootloader. |
| 15:06.35 | dcordes | k |
| 15:08.47 | goxboxlive | cr2: yes and whats the question? |
| 15:09.06 | dcordes | http://www.htc.com/product/03-products-11.htm#2 which one is this phone right here? |
| 15:09.09 | cr2 | goxboxlive: is there such distro for ba ? |
| 15:09.41 | goxboxlive | cr2: what kind of distro? Angstrom, yes |
| 15:10.11 | cr2 | goxboxlive: battery level is not shown, gsm signal level is full, but i don't even have a sim card. how do i suspend ? |
| 15:10.18 | dcordes | cr2, there is a htcblueangel machine type in openembedded |
| 15:10.25 | cr2 | goxboxlive: qtopia for ba in the same format. |
| 15:10.37 | cr2 | dcordes: it's outdated afaik. |
| 15:10.43 | dcordes | I think so |
| 15:10.45 | cr2 | dcordes: because it was the first. |
| 15:10.55 | dcordes | pitty that nobody maintains it |
| 15:11.32 | cr2 | dcordes: it's apache ? |
| 15:12.02 | cr2 | i think it's a wrong idea to support _only_ blueangel. |
| 15:12.18 | dcordes | you mean in oe? there are several htc mach types |
| 15:13.55 | Kevin2 | goxboxlive: The make-bootbundle.py script does not care about file names at all. Choose any names you want. |
| 15:14.27 | Kevin2 | goxboxlive: I've pushed a new rootfs - a couple of minor changes - http://www.handhelds.org/~koconnor/initramfs-installer/rootfs.cgz |
| 15:14.45 | cr2 | yes. the kernel will be compiled per-device obviously. but the userspace (and its interface) programs should be the same for all. |
| 15:17.18 | *** join/#htc-linux kiozen (n=oeichler@rgnb-4db03010.pool.einsundeins.de) |
| 15:18.54 | kiozen | hi all |
| 15:19.09 | kiozen | cr2: any news on the n560 |
| 15:19.14 | kiozen | ? |
| 15:19.29 | dcordes | http://www.htc.com/product/03-products-11.htm#5 any clue which machine name this one has? |
| 15:19.40 | dcordes | HTC GPS PDA |
| 15:20.54 | cr2 | kiozen: no. probably i'm making some mistake there. my CPLD driver for athena crashes at boot too. |
| 15:22.01 | kiozen | ok, tell me if you need something from me |
| 15:22.39 | Kevin2 | cr2: I fixed the front keypad driver on hermes last night. We can now control the front leds, lcd power, and keypad on hermes. |
| 15:23.11 | cr2 | kiozen: ok. |
| 15:23.37 | cr2 | Kevin2: with the "i2c" driver too ? |
| 15:24.17 | Kevin2 | cr2: Apache and Hermes use the same bit-banging i2c-like device. (It isn't real i2c, because instead of an 8bit address, it only sends a 5bit address.) |
| 15:24.20 | cr2 | Kevin2: afair the LCD needs the magic upload to run. |
| 15:24.28 | cr2 | ok. |
| 15:24.28 | Kevin2 | But, yes, it's the same driver. |
| 15:24.49 | cr2 | Kevin2: does the TS work for you ? |
| 15:25.19 | Kevin2 | I haven't tested it. We're going to need a regular ADC driver though - battery monitoring is also done via S3C ADC. |
| 15:25.31 | cr2 | ok. |
| 15:25.49 | cr2 | i need to get the 910 and 90c live trace. |
| 15:26.01 | cr2 | then the SD driver will be ready for testing. |
| 15:26.51 | cr2 | maybe the bruteforce approach should be applied. |
| 15:27.28 | Kevin2 | Okay. On my list: write spi keyboard driver, write s3c adc driver, plug in touchscreen and battery monitoring, make backlight brightness adjustable. |
| 15:28.19 | cr2 | i know the code layout, disassembling the code chunk up from the str/ldr address will give the location of the local SDHC array which stores the virtual base. |
| 15:28.36 | cr2 | ok, we are making good progress :) |
| 15:30.05 | cr2 | Kevin2: look at the blueangel spi keyboard driver. it will be nice to reuse your driver there too. |
| 15:31.29 | Kevin2 | cr2: If you know the code address - why don't you just set a breakpoint and inspect what address it writes to? |
| 15:33.10 | cr2 | the driver is not dumped from my hermes, so the addresses may differ. |
| 15:33.31 | cr2 | i have had such problem already. |
| 15:34.04 | cr2 | maybe testwm5-style dumper should be added to haret after all. |
| 15:36.00 | Kevin2 | I'm not sure if it is legal here in the states - DMCA type stuff.. |
| 15:36.13 | cr2 | ok. |
| 15:42.08 | pH5 | cr2: have you figured out anything about the sd booting? |
| 15:42.51 | pH5 | I was digging around in the bootloader, and the function to check the security level in the sd download part is the same that is used to show the current level in task 32 (which shows as 0), so I have no idea what is going wrong. |
| 15:43.24 | cr2 | pH5: on magician ? |
| 15:43.30 | pH5 | yes |
| 15:43.47 | cr2 | because i get ff on the uni. |
| 15:44.01 | *** part/#htc-linux rmoravcik (n=rmoravci@adsl-d100.84-47-105.t-com.sk) |
| 15:44.03 | cr2 | and the function is different afair. |
| 15:44.39 | cr2 | goxboxlive: are you still here ? |
| 15:57.39 | goxboxlive | cr2: Battery should work, dont know why. I have to check it later. GSM is showing because i have disabled the modem for you in that image. |
| 15:57.55 | goxboxlive | cr2: isent BabelO creating images for BA |
| 15:58.17 | goxboxlive | Kevin2: ok, that's great so i dont have to rename. |
| 15:58.24 | goxboxlive | Kevin2: i'll try it later. |
| 15:58.54 | cr2 | goxboxlive: what about suspend ? |
| 15:59.17 | goxboxlive | Should work. you have to set it up in GUI mode. |
| 15:59.26 | cr2 | i press the power button, and it just hangs. |
| 15:59.27 | goxboxlive | Settings-> powersetteing |
| 15:59.34 | cr2 | ok. |
| 15:59.37 | goxboxlive | cr2: maybe 100MB isent enuogh |
| 16:00.05 | cr2 | for a real install ? |
| 16:00.06 | goxboxlive | Maybe i should try to edit Kevin2 script and test with a bigger image.. |
| 16:00.09 | cr2 | maybe. |
| 16:00.10 | goxboxlive | yes |
| 16:00.18 | cr2 | but it's more a demo install. |
| 16:00.27 | goxboxlive | The image is huge now, because it's a common for A780/UNI and amg |
| 16:00.29 | goxboxlive | mag |
| 16:00.33 | goxboxlive | yes |
| 16:00.35 | cr2 | we should not push everything into the demo. |
| 16:00.56 | goxboxlive | of course not, but this way i only need to build one qtopia and not three |
| 16:01.04 | cr2 | but i guess the phone and suspend should work. |
| 16:01.05 | cr2 | ok. |
| 16:01.12 | Kevin2 | 100Meg is a lot - why not just run "df" and see how much space is free? |
| 16:01.40 | goxboxlive | Kevin2: beacuse it hangs |
| 16:01.46 | Kevin2 | df hangs? |
| 16:02.09 | goxboxlive | But this is just on my 128MB sd card. On my 2GB card i dont have that problem |
| 16:02.18 | goxboxlive | linux/devie |
| 16:02.22 | goxboxlive | device |
| 16:02.47 | Kevin2 | Hrmm. The latest rootfs that I posted does a much better job of checking for errors. You may want to retry with it. |
| 16:13.33 | *** join/#htc-linux TBBle (n=tbble@2001:388:f000:0:0:0:0:2d) |
| 16:39.19 | goxboxlive | Kevin2: i will, i'll try later. I visitors. I also think we should add verbose when extracting the rootfs. Because then people will see that something happends, so they dont think it's hanging. |
| 16:40.00 | Kevin2 | Yeah. I had it that way originally, but then the screen scrolls by and you don't see past status messages. |
| 16:40.36 | Kevin2 | I think it may be possible to convert every 50 files into a single "." character. But it would need some shell magic to do it. |
| 16:46.42 | dcordes | are there any special haret versions for msm7x00 or are features for devices with this kind of chipset included in mainline? |
| 16:47.03 | Kevin2 | dcordes: mainline. |
| 16:47.07 | dcordes | ok |
| 16:47.43 | goxboxlive | Kevin2: yes a series of '.' would be perfect. |
| 16:51.08 | pH5 | tar jfvx | 'sed -n 1~50i.' |
| 16:51.18 | pH5 | and some way to get rid of the newlines |
| 16:51.46 | pH5 | I guess "| while read line; do echo -n $line; done" only works in char buffered terminals? |
| 16:52.01 | Kevin2 | pH5: Couldn't we replace the newlines with "." somehow? |
| 17:03.08 | cr2 | Kevin2: there are 48,4a,4c and 4e mappings for the ATI control regs. do you know which mapping is used by which subsystem ? |
| 17:03.58 | Kevin2 | cr2: No, I haven't kept track of that. |
| 17:04.54 | cr2 | ok. |
| 17:06.15 | cr2 | in 48 i see 200 and 204 |
| 17:06.36 | cr2 | GEN_INT_CNTL |
| 17:06.54 | cr2 | and GEN_INT_STATUS |
| 17:08.43 | cr2 | 4a has a lof of different regs. |
| 17:08.57 | pH5 | Kevin2: I don't know how to do it with sed. how about tar jfvx | awk '{ if ((NR % 10) == 0) printf "." }' |
| 17:10.43 | cr2 | d0,d4,d8,dc |
| 17:11.10 | cr2 | these are i2s clock regs. |
| 17:12.14 | cr2 | 148 RB soft reset |
| 17:12.25 | *** join/#htc-linux wireddd (n=wired@97-81-78-105.dhcp.athn.ga.charter.com) |
| 17:15.54 | cr2 | in 4e 140 21c |
| 17:16.18 | cr2 | RB again |
| 17:17.42 | cr2 | yes :) |
| 17:17.47 | cr2 | the 910 and 90c are in 4c |
| 17:19.16 | cr2 | Kevin2: what is IRQS cpldirq: CPLD1(65)=1 ? |
| 17:19.57 | cr2 | i get # a=4c010204 v=02000000 |
| 17:20.06 | cr2 | immediately after it |
| 17:20.14 | cr2 | and it's the GEN_INT_STATUS |
| 17:25.33 | cr2 | hm. why do i get the asm decoded in the log, but not on the screen ? |
| 17:29.40 | cr2 | so 2 is "vline irq", and is handled in w100_vsync() |
| 17:30.22 | cr2 | 2000000 is the sd irq. |
| 17:32.03 | cr2 | write to status means clear irq. |
| 17:33.55 | Kevin2 | cr2: I don't know what CPLD1 is - is this hermes? Are you using usb? |
| 17:35.17 | Kevin2 | The decoded value in the log is a minimal decoding that haret does. Can you do a cvs up and try the latest haretconsole? It will look in /opt/mingw32ce/bin for objdump now. |
| 17:35.40 | cr2 | i've linked the objdump. |
| 17:35.44 | cr2 | mem b4700004=00000002 (00000002) |
| 17:36.29 | cr2 | looks like SD irq to me |
| 17:36.53 | cr2 | because it's always followed by |
| 17:36.58 | cr2 | mmutrace 0924f05c: e583a204(str) 4c010204 02000000 (00000000) |
| 17:37.10 | Kevin2 | Oh, I get CPLD1 irqs also - looks like an ATI interrupt. |
| 17:37.13 | cr2 | which is irq ack |
| 17:37.23 | cr2 | Kevin2: yes, i'm on hermes. |
| 17:37.59 | Kevin2 | cr2: You can decode the logs too - for example: ../console/dis.py < haretlog-20071116_193056.log | less |
| 17:38.06 | cr2 | Kevin2: where is the ATI SD block mapped ? |
| 17:38.07 | cr2 | ok. |
| 17:39.26 | Kevin2 | cr2: I don't know. I see accesses to 0x4c... and 0x48... |
| 17:40.09 | cr2 | ok, i'll check. 4c sounds logical. |
| 17:40.12 | Kevin2 | I use the following trace commands: |
| 17:40.14 | Kevin2 | addlist mmutrace 0x48000000 4*1024*1024 |
| 17:40.14 | Kevin2 | addlist mmutrace 0x4a000000 4*1024*1024 |
| 17:40.14 | Kevin2 | addlist mmutrace 0x4c000000 4*1024*1024 |
| 17:40.14 | Kevin2 | addlist mmutrace 0x4e000000 4*1024*1024 |
| 17:40.14 | Kevin2 | addlist mmutrace 0x94800000 2*1024*1024 |
| 17:40.15 | Kevin2 | addlist mmutrace 0xb4800000 2*1024*1024 |
| 17:40.17 | Kevin2 | set permissivemmutrace 1 |
| 17:40.19 | Kevin2 | ibit irqs 38 |
| 17:40.59 | cr2 | Kevin2: it's too much. |
| 17:41.05 | cr2 | 4a is used by sound. |
| 17:41.28 | cr2 | didn't see anything for the cam, but it must use the ATI i2c regs |
| 17:45.23 | cr2 | yes, sd is at 4c |
| 17:46.45 | cr2 | addlist mmutrace 0x4c202000 0x1000 |
| 17:46.46 | cr2 | and |
| 17:46.49 | cr2 | addlist mmutrace 0x4c010000 0x10000 |
| 17:48.29 | *** join/#htc-linux kiozen (n=oeichler@rgnb-4db03010.pool.einsundeins.de) |
| 17:48.49 | cr2 | Kevin2: is it possible to 'sed' the mmutrace 0924b058: e19330b1(ldrh) 4c202044 00000300 (00000000) |
| 17:49.05 | cr2 | to show IntMaskBuffer instead of 4c202044 ? |
| 17:51.02 | Kevin2 | What is IntMaskBuffer? |
| 17:52.02 | cr2 | http://wiki.xda-developers.com/index.php?pagename=ATIWxxxx |
| 17:52.06 | cr2 | at the bottom |
| 17:52.29 | cr2 | a register name |
| 17:52.49 | Kevin2 | Oh, you mean decode the addresses to register names. Yeah, it is possible, but it'll be a pain. |
| 17:53.47 | cr2 | ok, i#ll write a sed file. |
| 17:53.50 | Kevin2 | With "watch", we know exactly what we're going to get. With mmutrace, we might get one byte reads to 4 byte registers.. etc. |
| 17:55.55 | Kevin2 | Oh, if you want a quick and dirty way... Change dis.py so that where it prints m.group('addr') have it do a lookup on memalias.ArchRegs. |
| 17:56.15 | kiozen | cr2: dump question: how do I tell make modules_install to use another path than /lib/modules ? |
| 17:56.43 | cr2 | kiozen: it's easy. i'll look in the Makefile |
| 17:57.02 | cr2 | Kevin2: ok, i'll try. |
| 17:58.12 | cr2 | kiozen: INSTALL_MOD_PATH is the root of 'lib/modules' |
| 17:58.13 | Kevin2 | Hrmm. ArchRegs uses physical addresses - so, you'll also need to map the virtual to physical. |
| 17:58.20 | kiozen | cr2: ok got it |
| 17:58.39 | cr2 | Kevin2: ok, it's easy. |
| 18:00.16 | Kevin2 | bbl |
| 18:02.41 | goxboxlive | Kevin2: i am testing with an bigger partition on Magian now. |
| 18:04.45 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 18:07.01 | goxboxlive | cr2: do we use htc-egpio? |
| 18:11.31 | cr2 | goxboxlive: not yet. |
| 18:11.51 | cr2 | but i want to switch. |
| 18:12.09 | cr2 | goxboxlive: does the phone work on magician ? |
| 18:12.54 | goxboxlive | not with Qtopia, only openmoko |
| 18:14.11 | cr2 | hmm. |
| 18:27.04 | goxboxlive | cr2: how come? |
| 18:29.14 | goxboxlive | I have problems with Magician and installer. Magician freezes |
| 18:29.15 | cr2 | then qtopia is not usable on ba. |
| 18:29.26 | goxboxlive | no, BabelO has the same problem |
| 18:29.58 | goxboxlive | We have asked ljp to talk to the comm gyus at TT, maybe he can help us a little. |
| 18:37.50 | goxboxlive | hmm magician freezes totaly when i load Qtopia. Wonder what that might be. |
| 18:38.18 | goxboxlive | It runs well in console mode. |
| 18:39.10 | cr2 | goxboxlive: can i load the qtopia by BabelO in the same way ? |
| 18:39.17 | goxboxlive | And now i am running with a 1GB vfat card so the size shouldnt matter. I have tested it with Universal and 100MB should be enough also for Magian then, since i am using the same Qtopia build |
| 18:39.49 | goxboxlive | cr2: yes, but you will need BabelOs zImage. I guess it is suplied in the /boot folder. |
| 18:41.25 | cr2 | ok, so the modules are not in the initramfs ? |
| 18:44.54 | goxboxlive | cr2: Now, it seems for me that the initramfs is only ceating a live image, and then we dont use it anymore. |
| 18:45.50 | goxboxlive | pH5: Could you add vfat, loop and ext2 to the magian defconfig, they are now aplied as modules, but with the new method we need tohave them in the kernel zImage |
| 18:56.33 | goxboxlive | Now i have magian and qtopia working. It was my fault |
| 18:57.59 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 19:26.03 | *** join/#htc-linux the_sys0p (i=the_sys0@gateway/tor/x-01ad69a5ed81e1cc) |
| 19:33.51 | *** join/#htc-linux AC_Mast (i=s@gateway/tor/x-bf2ad6a7e98cebd5) |
| 19:49.56 | *** join/#htc-linux AC_Mast (i=s@gateway/tor/x-7d56507abab711b9) |
| 19:58.07 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 21:22.36 | *** join/#htc-linux the_sys0p (i=the_sys0@gateway/tor/x-56809f21a75b516a) |
| 21:38.55 | *** join/#htc-linux dblrbl_ (n=polyrhyt@c-24-18-133-36.hsd1.wa.comcast.net) |
| 21:43.16 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-103-152-149.karneval.cz) |
| 22:14.28 | BabelO | hello |
| 22:40.14 | cr2 | hi BabelO |
| 22:41.44 | BabelO | hi cr2 |
| 22:41.46 | cr2 | BabelO: have you seen the new installer by Kevin2 ? |
| 22:42.10 | BabelO | cr2: yes i read the log, need to compile one for BA ;) reaaly good tool |
| 22:42.24 | cr2 | yes. |
| 22:42.37 | BabelO | cr2: and for battery display on BA i found the bug, it is just two line to add in a config file under /etc |
| 22:43.07 | BabelO | cr2: a question, how can i convert unint timestamp -> string ? funciton ctime ? |
| 22:44.29 | Zoolooc | hello |
| 22:44.47 | dblrbl | hello |
| 22:45.23 | BabelO | hello Zoolooc ,dblrbl |
| 22:47.50 | cr2 | BabelO: yes. |
| 22:48.05 | BabelO | ok |
| 22:48.31 | cr2 | BabelO: battery does not show up for me on the universal too. an the red led makes me nervous :) |
| 22:49.13 | cr2 | i don't know the purpose of $MACHINE in the led name, but how do we unify the led control ? |
| 22:52.40 | BabelO | cr2: don't know about that, i ve done a filter for my app and led, i just take name after ":" |
| 22:54.13 | BabelO | cr2: for battery, just check /opt/Qtopia/etc/Tasks.cfg, you need two line in section [prestartup] HtcphoneBattery and HtcphoneHardware |
| 22:54.39 | BabelO | battery management change a little between 4.2 and 4.3 :) |
| 22:56.12 | cr2 | ok. |
| 23:09.41 | dblrbl | rofl |
| 23:09.45 | dblrbl | oops |