| 00:00.18 | BabelO | i do not understand what you say |
| 00:01.15 | BabelO | ok, i started new haret |
| 00:04.37 | Kevin2 | It is your kernel that blanked the screen. Haret handed control over to the kernel right after writing "Jumping to Kernel...". It was the kernel that blanked the screen. |
| 00:05.10 | Kevin2 | BTW, I thought you were able to boot a wizard or prophet kernel? |
| 00:05.24 | BabelO | yes, i can boot a wizard kernel |
| 00:05.53 | BabelO | but it is kernel from another repository, i just merge it with our current repository at handhelds.org |
| 00:08.56 | BabelO | maybe i can just draw a line at top of framebuffer to now where i am |
| 00:09.27 | Kevin2 | BabelO: I don't understand what you are looking to do. |
| 00:10.47 | BabelO | i just take code from wizard kernel source, merge it to handhelds cvs tree, because we are in revision 2.6.21 now |
| 00:12.57 | Kevin2 | Okay. So you are debugging your kernel. It is kind of a pain to do that. Do you have a serial port? How about irda? |
| 00:14.07 | BabelO | there is no irda and serial on artemis, just a miniusb connector :( |
| 00:15.12 | BabelO | ok maybe i have to replace our current clock.c by the one from wizard cvs, i merge code already, maybe i do a mistake with that |
| 00:16.21 | Kevin2 | Ughh. You could always try writing to frame buffer - but if you do this, you need to make sure you map in the framebuffer so that you can write to it. |
| 00:17.24 | Kevin2 | BTW, can you see gpio irqs with new haret? |
| 00:18.45 | BabelO | i see same trace as yesturday |
| 00:19.47 | Kevin2 | You don't see any GPIO-I reports? Even if you move mouse? |
| 00:21.06 | BabelO | i do this : |
| 00:21.23 | BabelO | ibit irqs 70 66 # because they occur often when doing nothing |
| 00:21.31 | BabelO | watch irqs 5 |
| 00:23.12 | Kevin2 | And do you see GPIO-I reports? |
| 00:23.56 | BabelO | no, this is output when i move mouse http://pastebin.ca/660458 |
| 00:24.19 | BabelO | first i see always same gpio 7 but when i monitor irqs, nothing more |
| 00:26.25 | *** join/#htc-linux Hawk||- (n=Hawk@p5B178F65.dip0.t-ipconnect.de) |
| 00:26.31 | Kevin2 | BabelO: Heh, it is "wi 10", not "watch irqs 10". :-) |
| 00:26.41 | BabelO | oups |
| 00:28.13 | BabelO | ok, better, sorry... |
| 00:28.14 | BabelO | 04.718 IRQS GPIO-I1: GPIO-I7(103)=1 |
| 00:33.10 | Kevin2 | So, GPIO7 is mouse down there? |
| 00:33.59 | Kevin2 | Do you always see "IH1-6" before a "GPIO-I1:" event? |
| 00:35.16 | BabelO | mouse up yes |
| 00:35.23 | BabelO | no not always |
| 00:36.18 | BabelO | Kevin2: this is a output http://pastebin.ca/660466 |
| 00:36.20 | Kevin2 | The register names are kinda confusing - I pushed http://handhelds.org/~koconnor/haret/memalias-20070816b.py with new names. |
| 00:37.32 | *** part/#htc-linux ketrox (n=ketrox@Xbf2e.x.pppool.de) |
| 00:37.40 | *** join/#htc-linux ketrox (n=ketrox@Xbf2e.x.pppool.de) |
| 00:38.07 | Kevin2 | Okay - so that capture always shows "IH1-17" before a "GPIO-I5:" event. |
| 00:38.43 | BabelO | yes, the gpio 128 occur often |
| 00:39.39 | BabelO | 000.000 IRQS GIRQ5: GPIO128(224)=1 |
| 00:40.50 | Kevin2 | What happens on mouse press? Do you see GPIO168 right before IH1-18? |
| 00:41.22 | Kevin2 | Err, right after, rather. |
| 00:43.26 | BabelO | 000.609 IRQS IH1: IH1-18(18)=1 |
| 00:43.26 | BabelO | 000.609 IRQS GIRQ6: GPIO168(264)=1 |
| 00:43.29 | BabelO | yes ;) |
| 00:45.34 | BabelO | why i can't mask gpio 128 with ibit gpios 128 ? |
| 00:45.44 | Kevin2 | What happens for mouse left? |
| 00:46.32 | Kevin2 | Make sure you run "ibit IRQS #", and make sure the # is from the inside of the parenthesis. |
| 00:46.56 | *** join/#htc-linux huhlig (n=UhligH@140.sub-70-210-224.myvzw.com) |
| 00:50.23 | BabelO | 000.776 IRQS IH1: IH1-6(6)=1 |
| 00:50.23 | BabelO | 000.776 IRQS GIRQ1: GPIO16(112)=1 |
| 00:50.41 | huhlig | hmm |
| 00:51.57 | Kevin2 | On http://pastebin.ca/660422 you show GPIO16 as down and GPIO136 as left. |
| 00:52.22 | BabelO | yes, because it is hard to move only in one way |
| 00:53.00 | BabelO | wait i restart haret, when i use ibit command it is really slow |
| 00:53.18 | Kevin2 | Ahh, okay. I'm trying to see if certain GPIO IRQs are always related to some base IH-X irqs. |
| 00:53.50 | BabelO | ok, i can made test with the roller button |
| 00:53.52 | BabelO | better ;) |
| 00:54.04 | Kevin2 | It looks like they are, and they are following the same pattern defined for the omap730 as set in linux/include/asm-arm/arch-omap/irqs.h. |
| 00:54.17 | BabelO | have to boot my virtual windows machine now, i hope it is more stable over usb |
| 00:56.33 | ketrox | does one of u gys knows wich kernel was used to boot linux in the hw6515 ? |
| 00:56.37 | Kevin2 | So, if that is true, you should always see IH1-6 before GPIO0-31, IH1-7 before GPIO32-63, IH1-8 before GPIO64-95, IH2-10 before GPIO96-127, IH1-17 before GPIO128-159, and IH1-18 before GPIO160-191. |
| 00:57.03 | BabelO | Kevin2: ok i ll check |
| 00:59.20 | BabelO | stupid me :( i have no python under this windows machine :( |
| 01:00.11 | Kevin2 | ketrox: I don't. Take a look at the wiki on handhelds.org |
| 01:01.10 | ketrox | the only thing i could find there was kernel boots using haret |
| 01:11.36 | *** join/#htc-linux Kevin2 (n=Kevin@207-237-65-88.c3-0.avec-ubr12.nyr-avec.ny.cable.rcn.com) |
| 01:15.10 | BabelO | Kevin2: you are right ;) it is |
| 01:15.16 | BabelO | 000.000 IRQS IH1: IH1-18(18)=1 |
| 01:15.16 | BabelO | 000.000 IRQS GIRQ6: GPIO168(264)=1 |
| 01:15.37 | BabelO | 000.000 IRQS IH1: IH1-7(7)=1 |
| 01:15.37 | BabelO | 000.000 IRQS GIRQ2: GPIO39(135)=1 |
| 01:16.44 | BabelO | 000.937 IRQS IH1: IH1-17(17)=1 |
| 01:16.44 | BabelO | 000.937 IRQS GIRQ5: GPIO150(246)=1 |
| 01:16.53 | BabelO | so it is good |
| 01:17.31 | Kevin2 | Can you confirm all six? If so, I can setup haret to ignore them by default. (I'm leery of ignoring them without confirmation, because it could then mask an irq.) |
| 01:21.10 | BabelO | Kevin2: do you know why htchermes_lcd is comented out ? ( lcd is not working then ) |
| 01:21.49 | BabelO | Kevin2: i can't confirm for 96-127 |
| 01:22.18 | BabelO | 000.801 IRQS IH1: IH1-8(8)=1 |
| 01:22.18 | BabelO | 000.801 IRQS GIRQ3: GPIO76(172)=1 |
| 01:23.17 | BabelO | 000.663 IRQS IH1: IH1-6(6)=1 |
| 01:23.18 | BabelO | 000.663 IRQS GIRQ1: GPIO7(103)=1 |
| 01:23.28 | BabelO | i think all are ok |
| 01:23.39 | Kevin2 | Cool - thanks. |
| 01:23.50 | Kevin2 | You have hermes also? |
| 01:24.54 | BabelO | Kevin2: no i look for code that use vsfb, but all are comented out ( athena and hermes) |
| 01:26.50 | Kevin2 | Hrmm. LCD works on hermes, so I'm not sure. |
| 01:31.25 | Kevin2 | I pushed http://handhelds.org/~koconnor/haret/haret-20070816b.exe -- no real change - just ignoring those redundant irqs you identified. |
| 01:33.47 | BabelO | Kevin2: ok i download it |
| 01:34.49 | BabelO | for vsfb i see it line 182 :( |
| 01:38.07 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 01:44.50 | BabelO | good night |
| 01:50.41 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 02:54.20 | *** join/#htc-linux ketrox_ (n=ketrox@Xd8e0.x.pppool.de) |
| 03:47.46 | *** join/#htc-linux rob__w (n=bob@X01bf.x.pppool.de) |
| 04:12.24 | *** join/#htc-linux silven (n=zmc@adsl-64-237-236-57.prtc.net) |
| 04:12.30 | *** join/#htc-linux ProstheticS (n=Prosthet@58.175.185.38) |
| 04:16.09 | *** join/#htc-linux UhligH_ (n=UhligH@160.sub-75-210-57.myvzw.com) |
| 04:18.41 | *** join/#htc-linux ProstheticS (n=Prosthet@58.175.185.38) |
| 05:08.25 | *** join/#htc-linux sadeness_ (n=LamersIn@nat1.rlan.ru) |
| 05:35.47 | *** join/#htc-linux pleemans (n=peter@leemanpe.ugent.be) |
| 05:37.51 | *** part/#htc-linux rmoravcik (n=rmoravci@ip-89-103-152-149.karneval.cz) |
| 05:54.31 | *** join/#htc-linux __law__ (n=law@mail.stiftadmont.at) |
| 06:32.02 | *** join/#htc-linux anYc_ (i=mario@hadince17.hadiko.uni-karlsruhe.de) |
| 06:41.17 | *** join/#htc-linux rob_w (n=bob@M9c59.m.pppool.de) |
| 06:44.25 | *** join/#htc-linux anYc (i=mario@hadince17.hadiko.uni-karlsruhe.de) |
| 06:58.43 | *** join/#htc-linux jott_ (n=j@unaffiliated/jott) |
| 07:30.07 | *** join/#htc-linux ketrox (n=ketrox@x463.vpn.hrz.tu-darmstadt.de) |
| 08:29.38 | *** join/#htc-linux DTMonterrey (n=evandro@cornelius.estig.ipb.pt) |
| 08:33.50 | *** join/#htc-linux sameo (n=samuel@AAnnecy-257-1-84-79.w90-36.abo.wanadoo.fr) |
| 08:49.35 | *** join/#htc-linux rob_w|laptop (n=rob_w@p549B9828.dip0.t-ipconnect.de) |
| 10:03.33 | *** join/#htc-linux jeanseb (n=jeanseb@gazypan.dyndns.org) |
| 10:06.20 | *** join/#htc-linux ketrox_ (n=ketrox@x274.vpn.hrz.tu-darmstadt.de) |
| 10:24.46 | *** join/#htc-linux corvus (n=corvus@26.Red-88-26-184.staticIP.rima-tde.net) |
| 10:52.28 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 11:06.33 | *** join/#htc-linux robotician (n=dale@usr016.bb275-03.udp.im.wakwak.ne.jp) |
| 11:18.03 | *** join/#htc-linux rob_w|laptop (n=rob_w@p549B9828.dip0.t-ipconnect.de) |
| 11:25.04 | *** join/#htc-linux the_sys0p (n=the_sys0@cpe-66-75-196-173.bak.res.rr.com) |
| 11:30.50 | *** join/#htc-linux rob_w|laptop (n=rob_w@p549B9828.dip0.t-ipconnect.de) |
| 12:28.41 | *** join/#htc-linux ketrox (n=ketrox@x48.vpn.hrz.tu-darmstadt.de) |
| 12:52.05 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-66-42-88.dynamic.dsl.as9105.com) |
| 13:19.48 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-74-85-237.dynamic.dsl.as9105.com) |
| 13:28.46 | *** join/#htc-linux silven (n=zmc@32.134.186.246) |
| 13:39.15 | *** join/#htc-linux ketrox_ (n=ketrox@x9.vpn.hrz.tu-darmstadt.de) |
| 13:59.11 | *** join/#htc-linux rob_w (n=bob@M9c59.m.pppool.de) |
| 14:28.47 | *** join/#htc-linux DTMonterrey (n=evandro@cornelius.estig.ipb.pt) |
| 15:43.48 | *** join/#htc-linux dion (n=dion@inhex.net) |
| 15:47.53 | *** join/#htc-linux okias (n=okias@esthe.cust.centro-net.cz) |
| 16:03.19 | *** join/#htc-linux bodski (n=root@84-51-139-57.johncr545.adsl.metronet.co.uk) |
| 16:04.14 | bodski | Hi there |
| 16:04.54 | bodski | I have just set up an sd card ready for linux on my blueangel |
| 16:05.47 | bodski | i was wondering what distro would people recomend using ie gpe gpe2 opie etc |
| 16:06.35 | bodski | I would like to help with testing and can try to get bt headset working |
| 16:14.14 | *** part/#htc-linux okias (n=okias@esthe.cust.centro-net.cz) |
| 16:27.12 | *** part/#htc-linux DTMonterrey (n=evandro@cornelius.estig.ipb.pt) |
| 16:39.44 | *** join/#htc-linux pH5 (n=ph5@p5485E0DE.dip.t-dialin.net) |
| 16:46.12 | bodski | hello anyone? |
| 16:47.19 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-103-152-149.karneval.cz) |
| 16:47.26 | pH5 | hello bodski |
| 16:47.42 | bodski | hi ph5 |
| 16:48.20 | bodski | i have a blueangel and was wondering if there is an recent angstrom image i can download |
| 16:48.51 | bodski | i am not a developer but want to try out / help with testing |
| 16:49.25 | bodski | situation is confusing - latest images i can find are from lkcl in 2006 |
| 16:51.08 | pH5 | bodski: Babel0 is the blueangel expert here. |
| 16:51.40 | bodski | ahh ok, cheers |
| 16:51.43 | pH5 | There are some images at http://www.linuxtogo.org/~htcpxa and on goxboxlive's homepage, maybe it's linked from the handhelds.org wiki |
| 16:52.06 | bodski | ok will try that :) |
| 16:54.53 | bodski | nice theres some stuff to try there, btw there is no link to this page on handhelds.org blueangel pages |
| 16:56.50 | bodski | ph5 <<< goxboxlive's homepage, you know the url? |
| 16:58.06 | bodski | bbl - left holding the baby... |
| 16:58.35 | pH5 | bodski: google to the rescue, http://www.gnulinux.biz/files/blueangel/ |
| 17:02.57 | bodski | cheers ph5! |
| 17:07.33 | ketrox | bodski i don't have any idea about blueangel but if u have a linux machine i will sugest to build your own image following (adapting) the instruction in this link http://wiki.xda-developers.com/index.php?pagename=HowToCompile4Himalaya |
| 17:08.52 | bodski | yes i would like to build my own it just seems there are more caveats/complications with blueangel than there are for other phones |
| 17:09.03 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 17:09.06 | bodski | cheers for the link ketrox :) |
| 17:09.48 | ketrox | instead of make himalaya_defconfig |
| 17:10.13 | ketrox | look in arch/arm/config or sthg like that for the file for blueangel |
| 17:45.59 | *** join/#htc-linux bodski (n=root@84-51-139-57.johncr545.adsl.metronet.co.uk) |
| 17:48.13 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 17:58.50 | *** join/#htc-linux jeanseb (n=jeanseb@gazypan.dyndns.org) |
| 17:59.00 | *** join/#htc-linux punkgeek (n=jeanseb@gazypan.dyndns.org) |
| 18:01.34 | bodski | wow linux booting now on my blueangel (1st time ever!) :) |
| 18:03.05 | *** join/#htc-linux dion (n=dion@inhex.net) |
| 18:04.46 | bodski | gpe2 looks promising |
| 18:14.27 | *** join/#htc-linux BabelO (n=Fabrice@lun34-2-82-238-28-28.fbx.proxad.net) |
| 18:18.17 | bodski | hi babel0 |
| 18:21.12 | bodski | have just got my blueangel running angstrom/gpe2 and was wondering what distro/rootfs/kernel images you recommend using currently |
| 18:22.13 | bodski | i am up for testing stuff if it helps |
| 18:24.17 | pH5 | bodski: great. that's only the beginning :) |
| 18:24.18 | pH5 | hi BabelO |
| 18:24.49 | BabelO | hi bodski |
| 18:24.52 | BabelO | hi pH5 |
| 18:25.30 | BabelO | bodski: you can test x11 image |
| 18:26.19 | BabelO | it is best image for PDA use, and you can also add openmokodialer with qvga theme :) |
| 18:27.37 | bodski | ok great is the latest image the one at http://www.linuxtogo.org/~htcpxa/htcblueangel/ |
| 18:27.46 | BabelO | yes |
| 18:27.54 | BabelO | i use OE to build it |
| 18:28.20 | BabelO | i hope i can build openmoko v2 soon ;) |
| 18:28.35 | bodski | ok great I will install x11-gpe now |
| 18:28.55 | bodski | are there instructions on installing the dialer available? |
| 18:29.27 | BabelO | no justssh to ba with an usb connection |
| 18:29.32 | BabelO | then use ipkg command |
| 18:29.44 | bodski | ok, btw has anyone tried to trace the jtag points on the ba? |
| 18:30.29 | bodski | i have a spare mainboard (still working) and wondered whether it was worth it |
| 18:31.11 | bodski | it would kill the board (using a normal oven to remove chips) but if it would help people out i am up for doing it |
| 18:32.39 | *** join/#htc-linux blackace (n=blackace@gentoo/developer/blackace) |
| 18:32.42 | *** join/#htc-linux goxboxlive (n=goxboxli@141.80-202-161.nextgentel.com) |
| 18:34.26 | BabelO | bodski: so you have a bt headset ? |
| 18:34.36 | bodski | yes |
| 18:34.39 | BabelO | i want to buy one but a ad2p |
| 18:35.06 | bodski | ahh yeh v.nice, too expensive for me tho |
| 18:35.11 | *** join/#htc-linux NickBee (n=NickBee@122.162.81.100) |
| 18:35.19 | bodski | i have a jabra jx-10 |
| 18:35.48 | NickBee | Hi pH5 |
| 18:35.53 | pH5 | hi NickBee |
| 18:35.55 | NickBee | hello BabelO |
| 18:36.10 | NickBee | how is everything there ? |
| 18:36.21 | BabelO | pH5: i think we have same issue, never test it on BA, i just copy magician code for uda1380 |
| 18:36.27 | bodski | ph5 << sco headset=normal bt headset? |
| 18:36.47 | NickBee | was Corvus online today? |
| 18:36.50 | pH5 | bodski: yes, normal mono bt headset |
| 18:37.01 | BabelO | hi NickBee |
| 18:37.40 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-74-27-16.dynamic.dsl.as9105.com) |
| 18:37.42 | NickBee | BabelO, you can reach me at djuneja - at - loginworks.com whenever you need me for artemis port tests |
| 18:37.53 | pH5 | BabelO: it involves some fiddling with an alternate channel on the uart link to the gsm chip (that I can trace with haret), but I have no idea how to configure the bluecore chip correctly. |
| 18:38.28 | bodski | BabelO: has anyone done/tried the jtag points for bluaeangel? |
| 18:38.58 | bodski | would it be worth butchering my spare ba mainboard to find them? ;) |
| 18:39.12 | BabelO | bodski: no i don t think |
| 18:40.23 | bodski | thought it might help with bootloader development |
| 18:44.12 | BabelO | pH5: i m looking for someone that have done a script to look at i2c dialog with haret dump |
| 18:49.25 | pH5 | BabelO: I only have some ugly ad-hoc scripts for magician |
| 18:49.43 | BabelO | ok |
| 18:49.55 | pH5 | http://en.pastebin.ca/661330 |
| 18:50.06 | pH5 | dunno if it's even good enough as a starting point :) |
| 18:51.23 | BabelO | pH5: thanks, yes good starting point for a newbie in python like me :) |
| 18:53.46 | BabelO | now i have to write to artemis frame buffer to now where i am in boot process |
| 18:54.00 | pH5 | BabelO: rock! |
| 18:54.40 | BabelO | :( i merger whole wizard source kernel with handhelds... i miss something, but don't know where :( |
| 18:55.54 | *** part/#htc-linux rmoravcik (n=rmoravci@ip-89-103-152-149.karneval.cz) |
| 18:56.07 | pH5 | ah, the joys of source code management with cvs |
| 18:57.06 | BabelO | pH5 : no ;) i take there repository with a HG command line, there kernel are 2.6.16, so old now |
| 18:57.45 | pH5 | ouch |
| 19:04.49 | *** join/#htc-linux dcordes (n=tsingtao@e179099054.adsl.alicedsl.de) |
| 19:09.39 | dcordes | hey |
| 19:09.45 | pH5 | hej dcordes |
| 19:13.21 | dcordes | !seen Babel0 |
| 19:13.32 | BabelO | hi dcordes |
| 19:13.35 | dcordes | how does taht work correctly? |
| 19:13.36 | BabelO | it is |
| 19:13.37 | dcordes | ah ^^ |
| 19:13.48 | BabelO | ~seen BabelO |
| 19:13.51 | apt | babelo is currently on #htc-linux (59m 24s). Has said a total of 23 messages. Is idling for 3s, last said: '~seen BabelO'. |
| 19:14.01 | dcordes | ok thanks |
| 19:14.09 | dcordes | how you guys doing? |
| 19:14.21 | dcordes | BabelO: your omap fine? |
| 19:14.21 | BabelO | good :) |
| 19:14.46 | BabelO | yes, i think i found why it doen't boot with hh kernel |
| 19:16.04 | dcordes | so what's the reason? |
| 19:16.21 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-74-9-16.dynamic.dsl.as9105.com) |
| 19:16.35 | BabelO | clock settings, as usuall ;) |
| 19:18.08 | dcordes | clock like processor related or time? |
| 19:18.24 | dcordes | (I have no idea about coding and kernels) |
| 19:18.47 | BabelO | to boot wizard, they keep state of some clock procesor internal |
| 19:18.57 | BabelO | and i remove this |
| 19:23.38 | pH5 | BabelO: to help with the omap850 effort, I'll lend stefan_schmidt's htc herald for a while. |
| 19:26.20 | dcordes | BabelO: did you talk to Silven from linwizard? |
| 19:26.44 | dcordes | we were thinking about the chances of a common htc distro for openembedded |
| 19:26.57 | dcordes | getting organized a bit more on htc development |
| 19:28.25 | *** join/#htc-linux silven (n=zmc@69.89.36.38) |
| 19:28.36 | BabelO | pH5: Ok, i see stefan_schmidt speaking on openmoko chennel yesturday ;) |
| 19:29.18 | BabelO | dcordes: once i have same status on hh kernel than with linwizard, we can start common devel |
| 19:30.21 | dcordes | so you would like to carry on with linzwizard? |
| 19:30.35 | BabelO | yes |
| 19:31.21 | dcordes | I think there are better build systems than the cross compilers silven is using |
| 19:31.53 | dcordes | also thinking about openmoko, openembedded would be perfect for a htc project |
| 19:32.21 | *** join/#htc-linux ketrox (n=ketrox@Xd8e0.x.pppool.de) |
| 19:32.36 | silven | dcordes : Those have to use crosscompilers at some level. If you're compiling code on an x86 and running it on an arm, you are using a cross compiler. |
| 19:33.01 | dcordes | I'm talking about openembedded dude |
| 19:33.27 | pH5 | hi silven, you are working on wizard? |
| 19:33.36 | silven | pH5: yep. |
| 19:33.53 | silven | dcordes : So you download precompiled stuff, or you compile it right on the handheld? |
| 19:35.15 | dcordes | nono sorry I put that completely wrong |
| 19:35.33 | dcordes | I'm aware of the need to cross compile |
| 19:36.10 | dcordes | just that it would be better to start using openembedded right from the beginning |
| 19:36.13 | silven | I think that's why I'm confused. I'm not using any build system at the moment. (unless you count make) |
| 19:37.34 | silven | From a distribution/useland standpoint, I think I agree with you, but you still have to compile/patch the kernel. Frankly when I'm debugging kernel drivers a userspace make framework is irrelevant to me. When it's done there will be a changeset or patch that can be used universally. |
| 19:37.53 | pH5 | dcordes: openembedded makes sense once you build images. for kernel development I use the OE-built toolchain, but that's about it. I really don't want to build packages every time I change something in the kernel. |
| 19:38.12 | silven | :D |
| 19:39.10 | dcordes | ok stopping oe propaganda |
| 19:39.57 | pH5 | dcordes: no need, I'm running two bitbake image build as I'm speaking :) |
| 19:40.39 | silven | dcordes : from what I've seen OE is a great framwork, and I do believe we should work with them and share code and the like. But when you arn't even to the point of installing a userland beyond busybox, there isn't much point yet. |
| 19:40.41 | dcordes | see silven use the openembedded toolchain, telling ya! |
| 19:41.02 | dcordes | :) |
| 19:41.13 | silven | dcordes : what's the difference, afaic it's arm binaries when they're done. |
| 19:41.39 | BabelO | pH5: two at same time, you are able to lose 6 hours ;) |
| 19:41.56 | silven | dcordes : Are there stability / performance differences? What is the underlying compiler? |
| 19:45.36 | pH5 | BabelO: don't matter, I lost too many hours of my life to OE already :) |
| 19:45.51 | BabelO | :) |
| 19:46.36 | pH5 | silven: gcc-4.1.2/eabi is default for angstrom |
| 19:46.50 | BabelO | pH5: maybe if we can build a big mosix cluster with our computer over vpn thenwe have powerfull environment ;) |
| 19:48.00 | silven | I'm using arm-iwmmxt-linux-gnu-4.1.1, I imagine, the differences are trivial. Do you know the differences between these two varients of gcc? |
| 19:51.31 | pH5 | silven: no, probably just some bug fixes. see http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.1.2 |
| 19:51.58 | *** join/#htc-linux dion (n=dion@inhex.net) |
| 19:52.36 | dcordes | silven: as I already told you all that compilling and kernel stuff is newland for me and I have no idea about coding :) |
| 19:53.23 | silven | pH5 : and between eabi and iwmmxt? |
| 19:55.22 | pH5 | silven: eabi is the new arm abi, in contrast to oabi, the old one. they most notably differ in calling conventions and floating point handling. iwmmxt has nothing to do with that, that's just support for some pxa27x-specific instructions (that omap850 doesn't have) |
| 19:56.10 | pH5 | (iwmmxt denotes an instruction set, like armv4t, armv5te, armv6, etc.) - omap850 is armv4t cpu afaik |
| 19:57.09 | silven | pH5 : Okay, thaks, so eabi would actuallt make more sense, (unless iwmmxt is a superset of eabi, in that case there would be no difference, correct?) |
| 19:58.10 | pH5 | silven: oabi/eabi and armv4t/armv5te/iwmmxt... are orthogonal. you can build armv4t/oabi or armv4t/eabi builds for omap850, but armv5t or iwmmxt builds won't work. |
| 19:59.36 | silven | pH5 : That doesn't make sense. I was doing kernel devel and testing on the Wizard with iwmmxt and everything worked fine. |
| 19:59.36 | pH5 | obviously the iwmmxt gcc build doesn't use the intel extensions by default, it just supports them. that's what the -march, -mcpu, -mtune options are for |
| 19:59.45 | silven | ahh okay. |
| 20:00.24 | silven | So as long as my march and mcpu flags are sane, it won't use the extended instructions and give me binaries that are compatable with the 850. Correct? |
| 20:00.32 | pH5 | correct :) |
| 20:01.08 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 20:01.15 | silven | pH5, Okay. I'm not going to worry about it then. Recompiling the cross compiler was annoying and I'd prefer not to do it again unless I'm bored. :) |
| 20:04.23 | pH5 | agreed |
| 20:05.17 | *** join/#htc-linux ltxdaMBL (n=ltxdaMBL@mef1336d0.tmodns.net) |
| 20:05.26 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-74-81-188.dynamic.dsl.as9105.com) |
| 20:10.16 | BabelO | silven: if you want you can take crosscompiler at codesourcery.com |
| 20:11.25 | silven | <PROTECTED> |
| 20:11.27 | silven | Sourcery G++? |
| 20:13.36 | BabelO | silven: yes htis one is eabi http://www.codesourcery.com/gnu_toolchains/arm/download.html |
| 20:13.39 | silven | well I found this: http://www.codesourcery.com/gnu_toolchains/arm/download.html, I have my own gcc-based crosscompliler that works. |
| 20:13.42 | silven | :) |
| 20:13.51 | BabelO | ;) |
| 20:14.06 | silven | Yea. Unless there is a compelling reason to change muy compiler / toolchain, I'm not likely to do it. |
| 20:14.34 | BabelO | i change mine to build with eabi support, just for that |
| 20:15.14 | silven | What I have makes binaries that work in predictable ways, and does it reasonably quickly. |
| 20:15.15 | silven | What advantage does eabi offer? |
| 20:15.40 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 20:15.42 | BabelO | silven: all new image from eabi are with eabi, advantage are power consumption as far i read news |
| 20:15.47 | pH5 | stop discussing compilers and do some kernel hacking ;) |
| 20:15.50 | BabelO | maybe pH5 can tell more about |
| 20:16.35 | pH5 | silven: eabi has a unified function calling convention and standardized floating point (softfloat) handling, so all eabi binaries work on all eabi systems (if the instruction set is supported). |
| 20:16.38 | silven | pH5 : Not happening today, I'm at work and leaving soon. Maybe in the evening when I can actually devote some real time to it. |
| 20:17.03 | pH5 | for oabi there were a bazillion different combinations of possible calling conventions and soft/hardfloat configurations |
| 20:17.25 | pH5 | (as I understand) |
| 20:17.51 | silven | That was the problem that okias was having. |
| 20:18.00 | pH5 | I know there were always problems of incompatible binaries compiled for different softfloat vs. different hardfloat methods. |
| 20:18.22 | silven | I don't see it, but he runs a 64bit userland, and I dont. |
| 20:20.16 | pH5 | silven: but gcc 4.1.1 has eabi support, and if you enable EABI in the kernel configuration you should be just fine. |
| 20:20.39 | pH5 | (i.e. have a kernel that can run newly built eabi userspace images) |
| 20:21.12 | silven | I'm using crossdev to generate my toolchain, so it should be trivial to replace. I'll start it off. |
| 20:21.54 | pH5 | silven: I think there isn't even need to do that (unless you want to build eabi glibc for userspace stuff) |
| 20:22.38 | silven | I probably should. Eventually it'll go in, and if I don't do it now it will be forgotten. |
| 20:23.46 | silven | eabi or gnueabi ? |
| 20:24.41 | BabelO | silven: i have the gnueabi |
| 20:24.54 | pH5 | yes, gnueabi is the right one |
| 20:25.01 | silven | Thanks |
| 20:25.15 | silven | crossdev --target arm-unknown-linux-gnueabi -- sound sane? |
| 20:26.53 | pH5 | I use armv5te-angstrom-linux-gnueabi/iwmmxt-angstrom-linux-gnueabi for pxa27x and armv4t-angstrom-linux-gnueabi for s3c2410, so that sounds sane to me. |
| 20:27.07 | silven | Beautiful. Thanks. |
| 20:27.17 | silven | I'll be in by the time I get home from work :D |
| 20:27.47 | pH5 | cool, I was always hoping we'd get some omap850 manpower into this channel, eventually :) |
| 20:28.41 | silven | I was working with canned-spaghetti a while back on it, we wrote the first (bad / limping) version of the keyboard driver for the wizard. |
| 20:29.42 | silven | After like 2 days of work someone came and said "oh yea, you have to let this hardware settle. Just change this waitstate, you didn't have to rewrite the whole thing." D'oh! |
| 20:31.50 | pH5 | hehe. is there a good place to look at for omap850 documentation? |
| 20:32.33 | silven | I have some pdf's for the processor, and a map of the interconnects, but aside from that, igh |
| 20:32.44 | silven | All the gpio stuff is in mysteryland. |
| 20:33.34 | silven | At least it was in november. |
| 20:33.39 | silven | Maybe that's changes. |
| 20:33.43 | silven | s/es/ed |
| 20:35.20 | pH5 | I'd like to get the processor docs, I'm going to lend an omap850 phone and jump in to help. |
| 20:35.26 | pH5 | s/lend/borrow/ |
| 20:38.35 | silven | I have them somewhere. Lemme look |
| 20:47.35 | BabelO | silven: do you have leds support on wizard ? |
| 20:49.15 | silven | BabelO : Yes, but last I knew it wasn't working. I haven't really touched the code since like november when I left the project. I'm just coming back as of this week,. |
| 20:50.04 | BabelO | silven, ok i just try to have mine booting with hh kernel :( but it hand somewhere in main |
| 20:50.46 | silven | do you have mercuial or git? |
| 20:51.10 | BabelO | i have mercurial i think |
| 20:51.32 | silven | try pulling the one that maui put up, see if that's any better. |
| 20:51.51 | BabelO | where is it ? |
| 20:51.54 | silven | If not I can put my git repo up temporarilty |
| 20:51.57 | silven | lemme check |
| 20:52.27 | silven | This weekend I was going to pour through all the crap I have and put up something quazi stable |
| 20:52.33 | silven | hg clone http://linwizard.maui.at/linux-wizard |
| 20:53.38 | *** join/#htc-linux nilton (n=nilton@loco-gw.ic.unicamp.br) |
| 20:54.15 | BabelO | silven: thx, i ll cdiff it with my current hh kernel |
| 20:55.03 | silven | I'm looking for what used to be my current-devel |
| 20:55.52 | BabelO | silven: it is better if you can use the hh kernel, there is lot of driver and facilities |
| 20:56.35 | silven | I'm planning on pulling in hh/montevista with the changes i had. |
| 20:56.50 | pH5 | and psokolovsky is a really nice driving force towards unification of different ports |
| 20:57.13 | pH5 | now we just have to improve our cleanup and pushing to mainline |
| 20:59.02 | silven | That should be okay. I work with git here at work and have experience with change sets / squash-merge commits and such. |
| 21:02.03 | BabelO | silven: ok, it is same repo i have, try to build this one and boot it on artemis just to see if all is here ;) |
| 21:02.46 | silven | I'll have my changes and something sane on soureforge hopefully by the weekend. |
| 21:03.50 | BabelO | silven: ok, that sound really good, i identify artemis gpio and also some touchscreen related event |
| 21:04.07 | silven | Wow! That's great! |
| 21:04.37 | silven | iff if this helps: |
| 21:04.44 | silven | walrus.centrixpr.com |
| 21:05.03 | BabelO | silven: with haret we look look lot of thing, i just need ida now to disaassemble something ;) |
| 21:05.33 | silven | hold one one sec. apache problems. |
| 21:07.34 | BabelO | silven: i have that from yesturday http://pastebin.ca/660422 |
| 21:07.55 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-74-81-188.dynamic.dsl.as9105.com) |
| 21:19.46 | BabelO | ok silven kernel boot ok ( compiled from source) |
| 21:54.50 | *** join/#htc-linux CheSS^lP (n=CheSS_lP@124.82.152.105) |
| 21:58.37 | *** join/#htc-linux giorgio (n=giorgio@adsl-ull-151-65.47-151.net24.it) |
| 22:31.30 | giorgio | tbble, are you here? i have a problem that you have already had (Maybe cable is bad?)... perhaps you can help me |
| 23:03.29 | *** join/#htc-linux dion (n=dion@inhex.net) |
| 23:20.25 | *** part/#htc-linux bodski_away (n=root@84-51-139-57.johncr545.adsl.metronet.co.uk) |
| 23:39.01 | *** join/#htc-linux scatterp- (n=zsirc@196.40.38.117) |
| 23:41.08 | scatterp- | hi i have a htc blueangel and a 4gb chip the phone wont recognise the chip i found an application that can make the phone act as a usb drive and then tryed to partition the chip and that did not work currently the chip seems to think it has around 50 partitions all messed up and when i try to connect to the phone to get it working as a usb drive it does not work due to all the partitions is there a way to get linux working on the phone as far as fdisk s |
| 23:43.33 | *** join/#htc-linux bodski (n=root@84-51-139-57.johncr545.adsl.metronet.co.uk) |