| 00:00.10 | BabelO | Marex: i just enable debug support for IRQ, just to see |
| 00:00.33 | BabelO | good idea about mutex locks |
| 00:00.33 | exco | Marex: so I guess you don't like ubuntu (in general, or just Hardy) |
| 00:00.36 | Marex | I wont wonder if the irq handler just killed kernel with some dump to console |
| 00:00.50 | *** join/#htc-linux S^P (n=Masood@static-host119-30-93-223.link.net.pk) |
| 00:00.55 | Marex | BabelO, yea, I had my hell recently with them ;D |
| 00:00.56 | *** part/#htc-linux silven (n=zmc@adsl-72-50-84-84.prtc.net) |
| 00:01.09 | BabelO | Marex: actually i try to find why when i try to mouunt sdcard it freeze or when i use i2cdump too often it freeze too |
| 00:01.35 | Marex | exco, I dont care as far as they dont bother me ... also I like to play with ubuntu people :D |
| 00:02.01 | Marex | BabelO, i2c on omap is weird from 2.6.16 on ... |
| 00:02.01 | exco | nice hobby... |
| 00:02.09 | Marex | at least I had problems on omap310 |
| 00:02.23 | Marex | BabelO, sorry, my omap times are long gone :( |
| 00:02.35 | Marex | exco, dont you think :) |
| 00:02.47 | BabelO | Marex: ok, but my phone is recent ;) with omap850 |
| 00:03.10 | BabelO | Marex: once i fix this irq problem, i can finish my ts driver |
| 00:03.26 | *** join/#htc-linux amd64lover (n=miknix@81.193.82.69) |
| 00:03.54 | Marex | BabelO, try mutexes |
| 00:04.06 | Marex | BabelO, you probably enter the irq handler more than once |
| 00:04.49 | BabelO | Marex: ok, i ve to check all irq handler of each modules |
| 00:05.29 | BabelO | Marex: btw i just buy a hp photosmart c4380, and wworks very nice on linux |
| 00:06.01 | Marex | do you have something like .... |
| 00:06.25 | Marex | static DEFINE_SPINLOCK(spinlock); |
| 00:06.40 | BabelO | Marex: other question where is your patch for qemu ? ;) |
| 00:06.43 | BabelO | i ll look |
| 00:06.52 | Marex | static irqreturn_t irqhandler(int irq, void *id) { |
| 00:07.06 | Marex | spin_lock(&spinlock); // at the begining |
| 00:07.17 | Marex | spin_unlock(&spinlock); // at the end |
| 00:07.20 | Marex | } |
| 00:07.42 | BabelO | Marex: ok i double check each module i change, thanks |
| 00:07.43 | Marex | but beware ... spinlock is active wait so dont make the code in it too big |
| 00:08.23 | Marex | BabelO, about the patch, it's on my PC with shitload of weird changes |
| 00:08.53 | Marex | or rather my qemu tree looks like sewer |
| 00:09.04 | Marex | though I can use it now to debug bootloaders ;D |
| 00:09.46 | BabelO | Marex: yes, can be interesting, i hope i can do the suspend / resume on omap850 too |
| 00:10.31 | Marex | 850 isnt probably supported in my qemu tree ;) |
| 00:10.52 | Marex | you can write your own TracePC if you have support in qemu though |
| 00:11.00 | Marex | and arm_load_ram() is simple too |
| 00:12.02 | BabelO | Marex: there is no doc for 850 :( |
| 00:12.35 | BabelO | a omap730 doc can be more help |
| 00:13.28 | Marex | another custom made cpu from ti? shit really happes |
| 00:13.33 | Marex | *happens |
| 00:13.57 | BabelO | yes |
| 00:13.58 | Marex | I had 310, no docs as well, but it was mostly 1510 without dsp |
| 00:14.15 | Marex | err ... 5910 too |
| 00:14.23 | Marex | they were all the same |
| 00:14.31 | BabelO | yes 850 is mostly 730, with some addon like edge and stuff from 1610 |
| 00:15.23 | *** join/#htc-linux NiZoX_ (n=nizox@eros.ph0k.eu) |
| 00:15.28 | Marex | jeez |
| 00:19.03 | exco | Marex: btw... did someone pull that rm anything on you when you started linux? |
| 00:19.34 | Marex | I dont recall, it was quite some time ago |
| 00:20.29 | univac | lol |
| 00:20.37 | univac | ;] |
| 00:20.58 | exco | so busted |
| 00:21.13 | Marex | huh? |
| 00:21.20 | exco | :-P |
| 00:21.55 | Marex | exco, it was in times without ubuntu ;) |
| 00:22.15 | Marex | there were only hardcore distributions you dont even know ;) |
| 00:22.39 | _who | slackware? :-P |
| 00:22.51 | Marex | well there was some light of hope in those dark times though .... vim already existed ;D |
| 00:22.52 | exco | we'll... at least you made the right choice and stuck with it... I was like ... "ah... fuck that shit" |
| 00:23.16 | Marex | _who, yea, slack was nice :) |
| 00:23.16 | paulproteus | Marex, Heh (-: |
| 00:23.18 | univac | phi i using it since 13 years old i had ;d |
| 00:23.23 | paulproteus | Yggdrasil! |
| 00:23.27 | Marex | yea! |
| 00:23.35 | univac | pld! |
| 00:23.36 | univac | ;p |
| 00:23.38 | _who | Marex: well, at that time perhaps ;-) |
| 00:23.49 | paulproteus | I actually never used Yggdrasil, but at least I was around when it was still dying. (-; |
| 00:23.58 | Marex | univac, die, you polish .... *CENSORED* ;D |
| 00:24.08 | univac | ;) |
| 00:24.22 | Marex | paulproteus, :-) |
| 00:25.05 | Marex | well if this notebook didnt have such obscure hardware, Id have openbsd here by now |
| 00:25.23 | Marex | and hacking on x86 is no-go for me ... never ;D |
| 00:25.27 | univac | i never heard about yggdrasil before |
| 00:25.58 | Marex | oh my goddess :) |
| 00:26.18 | Marex | univac, check ibiblio, you can find images there probably |
| 00:26.24 | univac | the olders slackware peoples should have know ;p |
| 00:26.36 | univac | yup reading about it |
| 00:26.52 | univac | Listopad 1992 - The first commercial distribution, called Yggdrasil was created by Adam Richter: |
| 00:26.58 | Marex | yup |
| 00:27.11 | Marex | where did you got listopad ... dammit |
| 00:27.20 | univac | polish forum hehe |
| 00:27.26 | Marex | jeez |
| 00:28.01 | exco | paulproteus: root@magician:~$ apm -s (2:27) |
| 00:28.27 | paulproteus | exco, Cool, now we wait for you to wake up tomorrow and see if the phone does too? |
| 00:28.36 | exco | yeah |
| 00:28.42 | paulproteus | crosses his fingers. |
| 00:29.00 | exco | so I'll be back tomorrow |
| 00:29.07 | exco | Good night to you all |
| 00:29.12 | paulproteus | Night! |
| 00:30.28 | *** join/#htc-linux dcordes (n=user2@g227115046.adsl.alicedsl.de) |
| 00:30.41 | dcordes | hi |
| 00:31.07 | paulproteus | Hey now dcordes. |
| 00:31.16 | dcordes | hey paulproteus |
| 00:31.34 | dcordes | paulproteus: how is hermes? |
| 00:31.52 | paulproteus | I don't know, I have only Magician and Uni. |
| 00:32.09 | dcordes | how's magician then? |
| 00:32.30 | paulproteus | exco and I have been testing goxboxlive's images from a few months back, and they seem to have everything working. |
| 00:32.53 | paulproteus | So I'm going to build a fresh OM image soon, and maybe see about putting the Qtopia 4.3.x snapshot stuff on. |
| 00:33.08 | paulproteus | Everything including wake on GSM ring. |
| 00:33.12 | paulproteus | I didn't remember that working before. |
| 00:33.20 | paulproteus | exco is testing his battery life overnight tonight. |
| 00:37.11 | dcordes | ok |
| 00:37.11 | dcordes | I hope I will have pm on kaiser workig |
| 00:37.12 | dcordes | wonders where his fellow kaiser owners are |
| 00:37.12 | dcordes | ltg is back on |
| 00:39.59 | dcordes | paulproteus: what do you mean you will put the fresh qtopia stuff in the openmoko image? |
| 00:40.10 | paulproteus | We'll see... |
| 00:40.41 | dcordes | do I know you from #angstrom? |
| 00:40.49 | paulproteus | I just pasted stuff there, but I don't chat there much. |
| 00:42.18 | dcordes | chit chit chatter |
| 00:47.28 | dcordes | ~ping cr2 |
| 00:47.29 | apt | pong cr2 |
| 00:47.51 | Marex | dcordes, hi, if you find him, let me know ;) |
| 00:49.28 | dcordes | Marex: alright. I just want to send him my SRAM dump kaiser |
| 00:49.58 | dcordes | where are the gurus when you need them |
| 00:49.59 | BabelO | Marex: adding spin_lock in omap mmc driver is not good , now sdcard is not detected aat all |
| 00:50.06 | dcordes | hi BabelO |
| 00:50.09 | BabelO | hi dcordes |
| 00:50.27 | Marex | BabelO, you are debuging omap mmc driver? :O |
| 00:50.30 | Marex | what for ? |
| 00:50.38 | BabelO | Marex: yes ;) |
| 00:51.14 | dcordes | I don't even know how to debug mmc on msm at all! |
| 00:51.20 | *** join/#htc-linux patoche (n=root@lns-bzn-57-82-249-45-162.adsl.proxad.net) |
| 00:51.35 | dcordes | ah right. that remebers me I was to ping SanMehat when I had the timeout irqs |
| 00:51.50 | dcordes | ~ping SanMehat |
| 00:51.50 | apt | pong SanMehat |
| 00:52.03 | dcordes | and now I eve have power thanks to BabelO |
| 00:52.43 | BabelO | :) |
| 00:53.30 | BabelO | Marex: ok i think it does not work because of dma lock already here |
| 00:53.40 | Marex | :D |
| 00:53.53 | Marex | yea, spinlock and dma together doesnt make much good |
| 00:54.05 | Marex | BabelO, it uses sgdma, right ? |
| 00:54.06 | BabelO | i see :) |
| 00:54.37 | dcordes | guys what is spinlock? I think I have seen it in msm smd code before |
| 00:54.52 | BabelO | Marex: don't know, what is sgdma ? |
| 00:55.03 | Marex | scatter-gather DMA |
| 00:55.09 | BabelO | Marex: yes it use it |
| 00:55.34 | Marex | anyway ... why are you mucking around with mmc driver ? |
| 00:55.44 | Marex | is omap850 somehow different from others ?> |
| 00:55.57 | Marex | dcordes, active wait |
| 00:56.11 | dcordes | fi |
| 00:56.17 | BabelO | omap850 have a +2 bus shift that the only diff |
| 00:56.18 | dcordes | for what Marex |
| 00:56.34 | Marex | spinlock is active wait |
| 00:57.12 | BabelO | Marex: and second diff i found is that omap guy have disabled the sdcard status error detection |
| 00:57.21 | BabelO | i was failling on that since one week |
| 00:57.23 | BabelO | :( |
| 00:57.35 | BabelO | and i remove it, so sdcard is well detected now |
| 00:58.01 | BabelO | and i m in sync with some omap git |
| 00:58.05 | Marex | spinlock my_spinlock; function_called_in_threading_env() { spin_lock(&my_spinlock); do something really fast; spin_unlock(&my_spinlock);} |
| 00:58.14 | BabelO | but still don't understand why they remove that |
| 00:58.38 | Marex | heh |
| 01:00.49 | dcordes | <PROTECTED> |
| 01:00.52 | dcordes | bye |
| 01:01.05 | BabelO | Marex: oh it seem there is a pio access mode, i ll try |
| 01:03.30 | Marex | :) |
| 01:20.10 | *** join/#htc-linux S^P (n=Masood@static-host119-30-95-32.link.net.pk) |
| 01:43.53 | Marex | cr2, :) |
| 01:49.27 | BabelO | lol |
| 02:05.02 | BabelO | good night |
| 02:52.55 | *** join/#htc-linux zwox (i=zwox@par69-7-82-243-234-94.fbx.proxad.net) |
| 03:07.20 | *** join/#htc-linux S^P (n=Masood@static-host119-30-81-87.link.net.pk) |
| 03:32.46 | *** join/#htc-linux fdasfasdf (n=Masood@static-host119-30-81-248.link.net.pk) |
| 04:00.05 | *** join/#htc-linux rob_w (n=bob@M95c5.m.pppool.de) |
| 04:04.22 | *** join/#htc-linux fasdf (n=Masood@static-host119-30-82-90.link.net.pk) |
| 04:07.04 | *** join/#htc-linux dropkickpuppy (n=alienjef@24-151-89-189.dhcp.nwtn.ct.charter.com) |
| 04:36.21 | *** join/#htc-linux fdasfasdf (n=Masood@static-host119-30-83-144.link.net.pk) |
| 05:52.10 | *** join/#htc-linux skodde (n=skodde@unaffiliated/skodde) |
| 06:01.01 | *** join/#htc-linux S^P (n=Masood@static-host119-30-87-4.link.net.pk) |
| 06:24.07 | *** join/#htc-linux fdasfasdf (n=Masood@static-host119-30-87-172.link.net.pk) |
| 06:39.17 | *** join/#htc-linux pH5 (n=ph5@e178254169.adsl.alicedsl.de) |
| 06:57.30 | *** join/#htc-linux exco (n=excogita@e181077240.adsl.alicedsl.de) |
| 06:59.33 | exco | paulproteus: Magician woke up |
| 07:00.15 | exco | paulproteus: cat /proc/apm still shows 100% (don't know about that) |
| 07:08.10 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
| 07:08.38 | pH5 | exco: /sys/class/power_supply/ds2760-battery.0/charge_now shows the battery's remaining µAh |
| 07:13.56 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
| 07:17.06 | exco | pH5: I dont have the folder ds2760 |
| 07:17.06 | exco | cat /sys/class/power_supply/backup-battery/charge_now shows 100% |
| 07:19.19 | pH5 | the backup battery returns bogus values for charge, it only has a voltage reading |
| 07:19.40 | pH5 | do you have /sys/bus/w1/devices? |
| 07:20.42 | exco | yes |
| 07:22.17 | pH5 | what's in there? |
| 07:22.23 | exco | voltage reading of the backup battery is 1048 |
| 07:22.42 | pH5 | I have `uname -r` = 2.6.21-hh20 |
| 07:22.56 | pH5 | and /sys/bus/w1/devices contains "30-0000115321ca w1_bus_master1" |
| 07:23.09 | pH5 | the first one is the ds2760 battery chip |
| 07:23.32 | exco | only the w1_bus_master1 folder |
| 07:24.08 | exco | atm I run 2.6.21-hh18 but trying to get hh20 to run |
| 07:25.01 | pH5 | strange, I'd still expect the battery to be found. what is your magician model. do you have a VPA by chance? |
| 07:25.59 | exco | it is an XDA Mini - but afaik there are only 2 Models (64MB vs 128MB) |
| 07:28.06 | pH5 | ok, so it's exactly the same as mine :/ |
| 07:28.40 | exco | so you're running hh20 with OpenMoko? |
| 07:29.28 | pH5 | there are some more models, at least the vodafone VPA is a bit different from the rest of the 64MB magicians. |
| 07:29.42 | exco | jumps up in exitement |
| 07:29.56 | pH5 | exco: theoretically |
| 07:30.03 | exco | well... I couldn't spot any differences http://pdadb.net/index.php?m=pdacomparer&id1=379&id2=296&id3=380&id4=550&id5=295&id6=363&id7=150&id8=136&id9=133&id10=134&id11=362&id12=135 |
| 07:30.05 | pH5 | I'm not doing much userspace work |
| 07:30.26 | exco | apart from 64/128MB RAM |
| 07:31.36 | pH5 | exco: yes, they are mostly identical. |
| 07:32.07 | pH5 | we know there are two different LCD screens |
| 07:32.29 | pH5 | I hope that this is the only difference between Xda/VPA, but who knows. |
| 07:33.05 | pH5 | anyway, as you have an Xda too, I can't blame your model for the missing battery :) |
| 07:34.51 | exco | yes, you're right... forgot about the lcd's |
| 07:47.41 | *** join/#htc-linux exc1 (n=excogita@e181077240.adsl.alicedsl.de) |
| 07:48.45 | exc1 | my system just totally froze on me after typing sudo apt-get update ... |
| 07:48.45 | exc1 | not nice at all |
| 07:48.45 | exc1 | not even ctrl+alt+backspace nor ctrl+alt+f.. wored.. |
| 07:48.45 | exc1 | totally uncool |
| 07:49.13 | exc1 | pH5: btw... I just booted the Magician to windows ... 92% battery |
| 07:50.19 | *** join/#htc-linux fasdf (n=Masood@static-host119-30-90-213.link.net.pk) |
| 08:27.10 | *** join/#htc-linux univac (n=univac@suita.chopin.edu.pl) |
| 08:35.41 | *** join/#htc-linux goxboxlive (n=goxboxli@208.84-48-176.nextgentel.com) |
| 08:46.18 | *** join/#htc-linux goxboxlive (n=goxboxli@208.84-48-176.nextgentel.com) |
| 08:54.41 | exc1 | hi goxboxlive |
| 08:54.48 | goxboxlive | hi |
| 08:54.58 | pH5 | hej goxboxlive |
| 08:55.08 | goxboxlive | hi |
| 09:02.13 | exco | goxboxlive: may I ask you how you built that Openmoko-Magician-Linux.exe? |
| 09:02.58 | *** join/#htc-linux S^P (n=Masood@static-host119-30-94-229.link.net.pk) |
| 09:03.12 | goxboxlive | exco: yes of course, it's a HOWTO in the haret source somewhere. Hold on i'll find it |
| 09:04.21 | goxboxlive | exco: here is the files, and the README: http://www.handhelds.org/~koconnor/initramfs-installer/ |
| 09:04.48 | exco | did you make the changes to get GSM working? |
| 09:06.03 | goxboxlive | exco: what image? openmoko? |
| 09:06.10 | goxboxlive | no, changes AFAIR |
| 09:09.01 | exco | that sounds great |
| 09:12.42 | goxboxlive | But just as you know i havent build a image for magician for months now, so i have no idea about the status. Maybe pH5 know. |
| 09:13.28 | pH5 | in current openmoko there is theme trouble and gsmd segfaults on startup |
| 09:13.55 | exco | goxboxlive: ... I played with your images you built in december for the last 2 days... thanks for putting those online |
| 09:14.08 | pH5 | I haven't investigated in detail, I'm still busy getting the kernel up to speed. |
| 09:15.25 | goxboxlive | pH5: have looked further in to the boot loader ? |
| 09:15.35 | goxboxlive | +you |
| 09:16.39 | pH5 | sporadically. there's still something going on with the timing that I don't understand, I get hangups that I can't trace to a certain code location. |
| 09:17.26 | goxboxlive | ok |
| 09:24.52 | *** join/#htc-linux JohnnyK (n=johnnyk@nfx-nat-229.pilsfree.net) |
| 09:26.39 | exco | goxboxlive: your OM image still boots fine with the latest kernel (GSM is gone) |
| 09:27.32 | goxboxlive | ok |
| 09:28.48 | exco | or I just don't know how to do it right :-) |
| 09:43.25 | *** join/#htc-linux S^P (n=Masood@static-host119-30-80-69.link.net.pk) |
| 10:01.31 | *** join/#htc-linux rob_w (n=bob@Mb23a.m.pppool.de) |
| 10:02.15 | exco | goxboxlive: I was mistaken. still works as before (hh20) |
| 10:09.28 | pH5 | exco: and the main battery is still gone? |
| 10:10.56 | exco | pH5: will check as soon as it's booted up again |
| 10:12.08 | *** join/#htc-linux marbalon (n=balon@aeft178.neoplus.adsl.tpnet.pl) |
| 10:13.20 | exco | still don't have that folder ... but I only replaced added the new hh20 kernel and corresponding modules... so that's not really a surprise |
| 10:13.28 | *** join/#htc-linux Marex (n=marex@85-132-216-250-eth3-gwfm10-user.802.cz) |
| 10:16.58 | Marex | morning |
| 10:17.13 | Marex | cr2, hello :) how's sound on n560 going ? O:-) |
| 10:17.18 | exco | good morning, Marex |
| 10:20.39 | *** join/#htc-linux patoche (n=root@lns-bzn-57-82-249-45-162.adsl.proxad.net) |
| 10:21.24 | *** join/#htc-linux JohnnyK (n=johnnyk@nfx-nat-229.pilsfree.net) |
| 10:30.46 | *** join/#htc-linux tsdogs_ (n=tsdogs@62.123.180.130) |
| 10:32.32 | *** join/#htc-linux fdasfasdf (n=Masood@static-host119-30-82-169.link.net.pk) |
| 10:32.52 | *** join/#htc-linux NiZoX (n=nizox@eros.ph0k.eu) |
| 11:07.12 | *** join/#htc-linux pH5 (n=ph5@e178205107.adsl.alicedsl.de) |
| 11:31.47 | *** part/#htc-linux exco (n=excogita@e181077240.adsl.alicedsl.de) |
| 11:35.53 | *** join/#htc-linux exco (n=excogita@e181077240.adsl.alicedsl.de) |
| 11:40.00 | Marex | univac, are you awake ? |
| 11:48.18 | *** join/#htc-linux Robwoerle (n=bob@X120f.x.pppool.de) |
| 12:00.23 | *** join/#htc-linux marbalon (n=balon@aegr221.neoplus.adsl.tpnet.pl) |
| 12:02.48 | *** join/#htc-linux Oktal (i=mat@cpc3-cmbg8-0-0-cust1018.cmbg.cable.ntl.com) |
| 12:05.41 | *** join/#htc-linux patoche_ (n=root@lns-bzn-27-82-248-48-182.adsl.proxad.net) |
| 12:08.30 | *** join/#htc-linux exco (n=excogita@e181077240.adsl.alicedsl.de) |
| 12:26.08 | *** join/#htc-linux revertive (n=revertiv@83.119.75.151) |
| 12:36.23 | *** join/#htc-linux BabelO (n=Fabrice@unaffiliated/babelo) |
| 13:16.38 | *** join/#htc-linux pH5 (n=ph5@e178205107.adsl.alicedsl.de) |
| 13:16.40 | marbalon | BabelO: hi |
| 13:19.34 | BabelO | hi marbalon |
| 13:19.58 | marbalon | continuing irq and egpio |
| 13:20.21 | exco | may I ask here what to do if I don't have a usb0 interface (yesterday I apparently still had it) |
| 13:21.19 | marbalon | why in htcapache.c irq resources for egpio start and end are the same ? |
| 13:22.01 | marbalon | BabelO: .start = IRQ_GPIO(GPIO_NR_HTCAPACHE_EGPIO_IRQ), .end = IRQ_GPIO(GPIO_NR_HTCAPACHE_EGPIO_IRQ) |
| 13:23.26 | BabelO | marbalon: because it is a slave irq to cascade with main chip i think |
| 13:23.50 | BabelO | marbalon: Kevin2 or pH5 know more about this part :) |
| 13:24.13 | pH5 | marbalon: as BabelO explained it's just one IRQ, but resources allow to give irq ranges, that's why start and end are the same |
| 13:24.15 | BabelO | marbalon: you need to find what it is on your device |
| 13:24.44 | pH5 | exco: ifconfig usb0 up? |
| 13:25.13 | pH5 | or if g_ether is not compiled into the kernel, but as a module, modprobe g_ether. |
| 13:25.22 | BabelO | hi pH5 ;) mmc is on the way ... |
| 13:25.48 | pH5 | hej. good stuff :) |
| 13:25.51 | exco | pH5: usb0: ERROR while getting interface flags: No such device |
| 13:25.57 | exco | thanks, btw |
| 13:26.11 | pH5 | exco: that means that either g_ether or pxa27x_udc is not compiled in / loaded. |
| 13:26.22 | exco | the thing is... yesterday all I did is |
| 13:26.22 | exco | modprobe usbnet, |
| 13:26.34 | marbalon | pH5: but it is physical IRQ or soft? |
| 13:26.35 | pH5 | exco: ah, you are talking about usb0 on the host? |
| 13:26.40 | exco | after that |
| 13:26.40 | exco | ifconfig usb0 192.168.3.1 netmask 255.255.255.0 broadcast 192.168.3.255 |
| 13:26.55 | pH5 | marbalon: the single GPIO_NR_HTCAPACHE_EGPIO_IRQ is 'physical' (as in cpu irq) |
| 13:26.58 | exco | and today it doesn't work (I could pull all my hair out) |
| 13:28.30 | marbalon | pH5: so it should be easy find it using haret |
| 13:30.06 | exco | pH5: also ... on the client (Magician) I get SIOCSIFFLAGS: No such device |
| 13:30.11 | pH5 | exco: did you change something? maybe the kernel on magician? |
| 13:30.30 | exco | :-) |
| 13:30.59 | exco | (that was the sth's wrong with me one) |
| 13:31.42 | pH5 | zcat /proc/config.gz and check whether there's CONFIG_USB_GADGET_PXA27X and CONFIG_USB_ETH enabled |
| 13:32.06 | exco | so the usb0 interface gets created on connect - if everything is configured right - on both ends? |
| 13:32.46 | pH5 | exco: on the magician side usb0 appears as soon as the ethernet gadget is loaded (CONFIG_USB_ETH/g_ether) |
| 13:33.02 | pH5 | on the host usb0 only appears after a usb ethernet gadget enumerates |
| 13:34.55 | exco | ok, so now I have to dig deeper... (do some reading I mean) because I don't know ... hell what do I even know |
| 13:36.03 | pH5 | :) |
| 13:36.55 | exco | pH5: what I did was - I took goxboxlive's image of OpenMoko, extracted the tar, included the modules for the latest kernel, and started it using hret and the latest kernel |
| 13:37.32 | exco | it works - but obviously misses the custom configurations |
| 13:37.39 | pH5 | what latest kernel? built yourself? |
| 13:37.55 | pH5 | maybe CONFIG_USB_GADGET_PXA27X and CONFIG_USB_ETH are configured as modules |
| 13:37.56 | exco | that would be cool... but no... |
| 13:38.04 | exco | got it from autobuild |
| 13:38.06 | pH5 | and the old OM userspace doesn't load them automatically |
| 13:38.36 | pH5 | try modprobe pxa27x_udc; modprobe g_ether on the phone |
| 13:38.59 | exco | I will (just fucked up the sd card because I forgot to unmount before unplugging) |
| 13:39.23 | Marex | pH5, hi |
| 13:39.27 | pH5 | hej Marex |
| 13:39.32 | Marex | pH5, mind commiting one patch for me? :) |
| 13:39.39 | Marex | looks like france died again :/ |
| 13:41.33 | Marex | ... and cr2 is blackmailing me with n560 sound ;D |
| 13:44.44 | exco | so if that Openmoko-Magician-Linux.exe from goxboxlive creates an image - that image contains the linux root and all the configurations he made? |
| 13:44.48 | cr2 | Marex: lol |
| 13:44.50 | cr2 | hi |
| 13:45.10 | Marex | cr2, :) hi |
| 13:45.10 | cr2 | marbalon: there are no irqs on kaiser cpld |
| 13:45.36 | cr2 | Marex: so, the n560 sound is working now ? :) |
| 13:46.26 | Marex | nah, but my red-black tree demo is ;-) wont you like that instead ? :) |
| 13:47.24 | exco | after modprobing g_ether on the client usb0 becomes available also usb0 comes available on the host after connecting... |
| 13:47.24 | exco | If we ever meet - I'll definitely buy you some beers, pH5 |
| 13:47.54 | pH5 | hi cr2 |
| 13:48.01 | pH5 | exco: great, will you be at linuxtag? :-P |
| 13:48.14 | pH5 | Marex: blackmail? |
| 13:48.47 | exco | pH5: I have to google that first ;-) |
| 13:48.48 | Marex | pH5, sure :) |
| 13:50.15 | *** join/#htc-linux kiozen (n=kiozen@rgnb-4db1df82.pool.einsundeins.de) |
| 13:50.22 | pH5 | Marex: well ok then, what about the patch :) |
| 13:50.32 | kiozen | hi |
| 13:50.44 | pH5 | hi kiozen |
| 13:50.56 | kiozen | cr2: is that assembly stuff still an issue? |
| 13:50.57 | Marex | pH5, I guess I will try blackmailing cr2 back ;D |
| 13:50.57 | exco | pH5: I will definitely put it in my calendar - but I can't tell you for certain - yet |
| 13:51.46 | cr2 | kiozen: yes. i want to have some portable solution that can be reused on artemis and ba6 |
| 13:52.45 | kiozen | cr2: ok I try to have a look this evening (had some guests until now :) ) |
| 13:52.56 | cr2 | ok |
| 13:53.15 | Marex | pH5, sent ;) |
| 13:54.53 | kiozen | BabelO: you found a bug? |
| 13:54.55 | pH5 | Marex: ok, looks good. |
| 13:55.18 | Marex | thanks |
| 13:55.26 | Marex | cr2, here you have it, you blackmailer :p |
| 13:58.56 | BabelO | hi kiozen |
| 13:59.12 | kiozen | hi BabelO, you found a bug? |
| 13:59.47 | BabelO | kiozen: a small yes, i use grid tools, and i staty in wizard, if i did not reopen proj wizard, i give me error when starting "go on" |
| 14:01.30 | pH5 | Marex: committed |
| 14:02.02 | kiozen | BabelO: :) I do not understand |
| 14:02.24 | BabelO | kiozen: yes, i don't understand myself :) |
| 14:02.34 | kiozen | lol |
| 14:02.47 | BabelO | kiozen: i jusqt use tiff to geotiff wizard |
| 14:03.04 | BabelO | i set first time proj with wwizard |
| 14:03.14 | BabelO | then i use grid tool, then i push go on |
| 14:03.17 | Marex | pH5, thanks :) |
| 14:03.29 | BabelO | kiozen: then i reopen a new file |
| 14:03.53 | *** join/#htc-linux JEEB (n=kanakana@a88-112-94-153.elisa-laajakaista.fi) |
| 14:03.56 | BabelO | kiozen: and just clear reference point then do grid tool then go on, and it fail |
| 14:04.46 | kiozen | ok I will try |
| 14:04.59 | BabelO | kiozen: i pm you map i use |
| 14:05.09 | kiozen | yes I am downloading |
| 14:05.16 | BabelO | the first file contain the grid |
| 14:08.06 | kiozen | BabelO: it's UTM zone 31? |
| 14:08.24 | BabelO | kiozen: yes it seem :) |
| 14:14.04 | *** join/#htc-linux dcordes (n=lukas@f049169209.adsl.alicedsl.de) |
| 14:15.40 | dcordes | hi all |
| 14:16.14 | exco | hi |
| 14:17.21 | pH5 | hi dcordes |
| 14:18.11 | kiozen | BabelO: it works for me, what failed in detail? |
| 14:18.15 | dcordes | pH5, can you add the branch to ltg git we talked about? |
| 14:18.48 | BabelO | kiozen: the second time you do an image it works ? :( |
| 14:18.58 | kiozen | yes |
| 14:19.01 | BabelO | kiozen: so i miss something in my reproduce step |
| 14:19.14 | kiozen | I just missed to rotate the images :) |
| 14:19.38 | BabelO | kiozen: ahh images need to be rotated |
| 14:19.57 | BabelO | kiozen: it explain why i was not able to paste them together |
| 14:21.18 | dcordes | marbalon, I updated cpld gpio definitions: http://rafb.net/p/wB4UP282.html |
| 14:21.22 | pH5 | dcordes: yes, which branch is it in git.android.com? android-msm? |
| 14:21.30 | dcordes | right |
| 14:21.58 | cr2 | kiozen: the gcp projection handling needs some enhancements |
| 14:22.11 | dcordes | pH5, I fetch like this: git clone git://git.android.com/kernel.git kernel; cd kernel; git checkout -b android-msm origin/android-msm |
| 14:22.15 | dcordes | hi cr2 |
| 14:22.23 | cr2 | kiozen: sometimes i'm not sure what will the program do, or will do. |
| 14:22.27 | cr2 | hi dcordes |
| 14:22.41 | dcordes | cr2, can I send you SRAM dump? |
| 14:23.17 | kiozen | cr2: any details? |
| 14:24.11 | dcordes | marbalon, also added SMD support (not working yet, but it is initialized) from dzo's latest patch. may I send you my diff? |
| 14:24.41 | dcordes | what's your progress with egpio? |
| 14:25.02 | pH5 | dcordes: ok. give me some minutes, I'll wait until serenity is a bit less loaded |
| 14:25.32 | dcordes | pH5, I will fetch latest tree and merge my patch in it |
| 14:25.42 | dcordes | will have some rejects |
| 14:26.02 | cr2 | dcordes: the SPL at the beginning is not very interesting, do you have something behind the SPL ? |
| 14:26.18 | BabelO | kiozen: so i choose the standard projection for my map ? ; |
| 14:26.52 | cr2 | kiozen: reading and writing the gcps need some more strict standards |
| 14:27.30 | kiozen | BabelO: I used mercator WGS84 for the final map projection |
| 14:27.47 | kiozen | utm 31 wgs84 for the gcps |
| 14:28.03 | BabelO | kiozen: ok that is what i select |
| 14:28.34 | kiozen | I selected "clear all" after refrencing the 1st tile |
| 14:28.47 | dcordes | cr2, you told me to dump this |
| 14:28.50 | kiozen | loading the 2nd, using the grid tool again |
| 14:28.52 | dcordes | iirc |
| 14:29.06 | exco | does this sound right: iptables -t nat -A POSTROUTING -o eth1 -s 192.168.27.2/32 -j MASQUERADE |
| 14:29.18 | cr2 | kiozen: i don't quite follow the logic in some cases. |
| 14:29.24 | kiozen | that worked, both tiles have the correct relative position |
| 14:29.46 | kiozen | cr2: don't undertsand what you mean |
| 14:30.40 | cr2 | kiozen: what is the standard save format for gcps ? |
| 14:31.10 | kiozen | cr2: the one QLandkarte produces now |
| 14:31.38 | cr2 | kiozen: are there any docs ? |
| 14:32.00 | kiozen | cr2: of cource not :) who will read them? |
| 14:32.15 | cr2 | kiozen: i think GT can read "old .gcp", "V1.0 .gcp" and .tab ? |
| 14:32.45 | kiozen | cr2: yes I hope |
| 14:33.03 | cr2 | kiozen: if the "save" button makes some black magic without telling the user i need to read the docs :) |
| 14:33.03 | kiozen | haven't tested it lately |
| 14:33.29 | pH5 | dcordes: I added a branch android-msm that tracks git.android.com. I guess we should make another branch htc-msm for our actual commits. agreed? |
| 14:33.36 | kiozen | cr2: the save button will produce a file you can reload later on |
| 14:33.41 | kiozen | that's it |
| 14:34.00 | cr2 | kiozen: .tab is standard, the "old .gcp" and "V1.0 .gcp" need some docs for their grammar. |
| 14:34.04 | dcordes | pH5, you mean as a sort of pre-stage? |
| 14:34.20 | kiozen | cr2: let's forget about the old .gcp |
| 14:34.45 | cr2 | kiozen: ok. the new .gcp |
| 14:35.09 | kiozen | and the new .gcp is easy to understand, it's more or less gdal syntax |
| 14:35.14 | pH5 | dcordes: no, as working area. android-msm will then really only contain the git.android.com stuff (as a shortcut for android/android-msm basically) |
| 14:35.29 | dcordes | pH5, ok sounds great |
| 14:35.42 | cr2 | kiozen: i think the controversy comes from "gdal_translate compatible" .gcp and label support. |
| 14:36.25 | cr2 | kiozen: the strict gdal format 1. does not support label names. |
| 14:36.47 | dcordes | pH5, do you have any msm7x00x device? |
| 14:36.54 | kiozen | yes label support is done because of a sepcial person insited on them :) |
| 14:37.04 | cr2 | kiozen: 2. it specifies the target SRS ? |
| 14:37.23 | kiozen | cr2: no it specifies the SRS of the gcps |
| 14:37.28 | cr2 | kiozen: you mean "is removed" ? |
| 14:37.28 | *** join/#htc-linux ltxda (n=ltxda@c-98-196-125-205.hsd1.tx.comcast.net) |
| 14:37.47 | kiozen | cr2: don't know, do I support labels? |
| 14:37.57 | kiozen | have to look :D |
| 14:38.00 | cr2 | i think not. because they are not saved. |
| 14:38.11 | cr2 | into the .gcp format |
| 14:38.15 | kiozen | ok |
| 14:38.38 | dcordes | pH5, what's a quick way to show which files are in my working tree but not in my git index? |
| 14:39.16 | kiozen | cr2: yep no labels |
| 14:39.27 | *** join/#htc-linux p3t3r__ (n=p3t3r__@wlan1.net-you.de) |
| 14:40.29 | cr2 | kiozen: sure, because it break the gdal-compatibility |
| 14:41.00 | kiozen | cr2: yes I guess this was the last result of our discussion |
| 14:41.02 | pH5 | dcordes: no, I have not. git status shows changed/untracked files |
| 14:41.31 | cr2 | kiozen: i suggest to save into wgs84/latlong .tab |
| 14:41.56 | kiozen | cr2: the label? |
| 14:42.10 | cr2 | instead of recreating the gdal-incompatible .gcp format |
| 14:42.19 | pH5 | dcordes: ok, finished. you can rebase and push your stuff to htc-msm now :) |
| 14:42.34 | dcordes | ok that's great |
| 14:42.38 | kiozen | lol, no .tab saving support so far :) |
| 14:42.42 | dcordes | can you hint me on how to identify/push? |
| 14:43.20 | cr2 | kiozen: yes, .tab supports labels. to save yourself the projection troubles use CoordSys Earth Projection 1,104 only |
| 14:43.28 | cr2 | this is latlong/wgs84 |
| 14:44.02 | kiozen | cr2: gcp SRS will always be the final SRS of the map, and that is most likely merc/WGS84 |
| 14:44.11 | cr2 | kiozen: but then you'd preserver the "original" SRS for gcps internally |
| 14:44.29 | kiozen | cr2: nope |
| 14:44.52 | cr2 | kiozen: good luck saving merc/wgs84 in .tab :) |
| 14:45.13 | cr2 | i can't tell you the number right out of my head :) |
| 14:45.57 | *** join/#htc-linux exco (n=excogita@e181077240.adsl.alicedsl.de) |
| 14:46.20 | cr2 | kiozen: then you need to set the output projection before loading the gcps, and do not change it ? |
| 14:47.52 | kiozen | cr2: look at the code, it's better, gcps by the gridtool are longlat/wgs84 |
| 14:48.12 | kiozen | cr2: and they are converted to target SRS once you press "go on" |
| 14:48.22 | kiozen | cr2: thus everything is fine |
| 14:49.50 | cr2 | kiozen: hmm. so, can i load 2 different gcp sets with different srs ? |
| 14:50.30 | kiozen | cr2: no |
| 14:50.47 | cr2 | kiozen: i had a different problem recently. the default target setting was +proj=merc, but the gcps has a funny proj4 projection defined in .gcp |
| 14:51.24 | cr2 | i've loaded them, and saved |
| 14:52.05 | cr2 | the saved file had +proj=merc, but the values remained from the funny proj4. |
| 14:52.23 | cr2 | i don't think it's ok. |
| 14:52.57 | kiozen | cr2: you allways try to do things the hardest way possible |
| 14:53.03 | cr2 | maybe the save wizard will help. where one can choose the output proj. |
| 14:53.05 | cr2 | lol |
| 14:53.11 | BabelO | kiozen: i need to do some small scan to remove the blank part :) |
| 14:53.31 | kiozen | cr2: just do it the simple way |
| 14:53.38 | kiozen | use the grid tool, or |
| 14:53.59 | kiozen | give gcpcoord as longlat/wgs84 |
| 14:54.45 | cr2 | kiozen: i don't mind to write some code for it, but i need some better core support for such options. |
| 14:55.37 | kiozen | cr2: what you want is a projection string per gcp, to give each gcp another projection |
| 14:55.44 | cr2 | btw, it'd be trivial to save into .tab by trapping the output file extension. |
| 14:56.03 | kiozen | wouldn't it be easier to preprocess your gcps to longlat/wgs84 |
| 14:56.35 | kiozen | cr2: .tab is just not done, the method handler exists already |
| 14:56.38 | cr2 | kiozen: yes, to keep it preserved, and reproject to the target srs as needed. |
| 14:56.55 | cr2 | yes, i've seen , just didn't have time to add it. |
| 14:57.41 | cr2 | avoiding multiple transformations here and there is a good thing imho. |
| 14:58.20 | kiozen | cr2: yes, thus the rule is to supply gcps in long lat / wgs84 |
| 15:00.34 | cr2 | hehe. for the kadaster data ? |
| 15:01.00 | *** join/#htc-linux darkerror05 (n=darkerro@cpe-72-230-169-235.rochester.res.rr.com) |
| 15:01.05 | cr2 | kiozen: the fact is that one wants gcps in 2 coordinate systems. |
| 15:01.38 | cr2 | lonlat/wgs84 and some GK variant (or whatever IGN uses) |
| 15:02.04 | cr2 | i guess it covers 99.999% of all cases. |
| 15:02.14 | kiozen | cr2: why would I? Either I have a grid and use the gridtool |
| 15:02.28 | kiozen | or I have gcps and transform them into longlat |
| 15:02.43 | kiozen | most likely these are waypoints |
| 15:03.15 | cr2 | double converting the survey data to/from longlat/wgs84 is bad. |
| 15:03.22 | cr2 | imho of course. |
| 15:04.10 | cr2 | i don't mean you'd do it, but some minor core changes will be very helpful in future. |
| 15:04.44 | kiozen | how does your survey data look like? |
| 15:06.15 | cr2 | (double)x (double)y +proj4 |
| 15:07.07 | cr2 | + (double)z to be absolutely strict |
| 15:07.27 | cr2 | (double)x (double)y (double)z proj4 |
| 15:07.52 | kiozen | pixel information? |
| 15:08.32 | cr2 | x y 0 proj4 |
| 15:08.59 | cr2 | long lat 0 +proj=latlong +datum=WGS84 |
| 15:09.36 | cr2 | i want the .gcp/whatever to store the raw data. and preserve it. |
| 15:10.30 | kiozen | cr2: how is the original data stored? |
| 15:10.30 | cr2 | btw, it's the same what gdal uses internally |
| 15:10.53 | cr2 | CSV table for xyz |
| 15:11.20 | cr2 | x y z label |
| 15:11.39 | kiozen | and the projection string? |
| 15:12.33 | cr2 | de-facto yes. |
| 15:12.44 | cr2 | per CSV table of course. |
| 15:13.36 | cr2 | i guess they use autocad or something like that originally, so they don't care about the projection itself. |
| 15:15.11 | dcordes | cr2, what is "something behind the SPL"? |
| 15:15.18 | kiozen | can't you convert that into a .gcp? |
| 15:15.30 | exco | can I use ifconfig to set a gateway |
| 15:16.54 | cr2 | kiozen: yes, that's what i've done. but the labels were eaten, and after i've moved them around in GT the projection as +proj=merc |
| 15:19.01 | kiozen | cr2: I see, that is because a_srs is the map's target projection, not the on of the gcps |
| 15:19.53 | kiozen | the moment GT detects longlat it assumes them as WGS84 and converts them into a_srs |
| 15:20.30 | cr2 | [16:37] <kiozen> cr2: no it specifies the SRS of the gcps |
| 15:21.22 | kiozen | cr2: hm, I was wrong :) |
| 15:21.25 | cr2 | you detect longlat by the NE°''' syntax ? |
| 15:21.30 | cr2 | ok. |
| 15:21.41 | cr2 | that's why i was confused too. |
| 15:21.49 | kiozen | and double values < 180 |
| 15:23.14 | cr2 | unless you are in the central afrika, or very close to the GK zone border :) |
| 15:23.22 | dcordes | pH5, ok rebase done |
| 15:23.31 | dcordes | and working |
| 15:24.14 | kiozen | cr2: yes I know, let's wait for that bug report :P |
| 15:24.35 | cr2 | LOL |
| 15:24.44 | pH5 | dcordes: perfect |
| 15:25.03 | dcordes | pH5, how can I commit the tree now? |
| 15:25.28 | pH5 | git push ssh://lgorris@linuxtogo.org/home/groups/mobile-linux/kernel.git htc-msm |
| 15:26.43 | dcordes | pH5, src refspec htc-msm does not match any. |
| 15:27.41 | pH5 | dcordes: don't you have a local htc-msm branch tracking linuxtogo.org/htc-msm? |
| 15:28.17 | dcordes | pH5, no I rebased with androids hmmm |
| 15:28.35 | *** join/#htc-linux DigitalElf (n=neil@S01060016b6264068.cg.shawcable.net) |
| 15:29.02 | kiozen | cr2: ADD R3, R4, R3,LSR#29 |
| 15:29.18 | kiozen | does this apply the shift befor or after the add |
| 15:29.38 | pH5 | dcordes: do you have origin/htc-msm in the "git branch -a" output? |
| 15:31.13 | dcordes | pH5, no I have the usual tree I used before. |
| 15:31.32 | dcordes | pH5, that only shows the available heads on git.android.com |
| 15:32.43 | cr2 | kiozen: before the add |
| 15:33.07 | pH5 | ok. git remote add -t htc-msm -f mobile-linux git://linuxtogo.org/home/groups/mobile-linux/kernel.git |
| 15:33.09 | cr2 | r3=r4+(r3>>29) |
| 15:33.30 | dcordes | pH5, on top of my androids rootdir? |
| 15:33.43 | pH5 | dcordes: why not, if that is your workspace |
| 15:33.47 | dcordes | it is |
| 15:34.01 | pH5 | of course you can always create a completely separate tree |
| 15:34.20 | pH5 | when you added that remote you should have the mobile-linux/htc-msm remote branch |
| 15:34.44 | pH5 | then you can "git branch --track htc-msm mobile-linux/htc-msm" to make a local tracking branch |
| 15:35.07 | pH5 | then rebase all your patches onto local htc-msm and push htc-msm like above |
| 15:35.12 | dcordes | pH5, now I have mobile-linux/htc-msm in the available branches but android-msm is still selected |
| 15:35.55 | dcordes | pH5, wouldn't it be more clean to checkout a new tree from our tracking tree, apply my changes on this and then push? |
| 15:36.06 | *** part/#htc-linux DigitalElf (n=neil@S01060016b6264068.cg.shawcable.net) |
| 15:36.55 | pH5 | dcordes: sure. if you take your current tree and remove all android.com remote branches you'll get the same result. |
| 15:38.17 | dcordes | git-branch -d or -D ? |
| 15:39.24 | pH5 | -D to remove dangling branches, -d only removes merged ones. and then git-gc some time to remove the unused objects |
| 15:41.56 | dcordes | pH5, hmm I fear I'll be messing the tree up now |
| 15:42.36 | pH5 | dcordes: then make a backup or move it out of the way and start over if you're more comfortable that way :) |
| 15:42.56 | pH5 | just ask if there are any problems. |
| 15:44.10 | dcordes | pH5, ok. git clone git://linuxtogo.org/home/groups/mobile-linux kernel ? |
| 15:44.23 | BabelO | kiozen: look nice once i remove blank border ;) i just miss a good srtm data source |
| 15:45.09 | kiozen | BabelO: hm, don't know one, my SRTM data was on the DVD with the maps |
| 15:47.37 | pH5 | git clone git://git.linuxtogo.org/home/groups/mobile-linux/kernel.git |
| 15:47.42 | kiozen | cr2: http://pastebin.ca/1007062 |
| 15:48.57 | dcordes | pH5, will this clone the whole stuff in mobile-linux including hh stuffs? |
| 15:49.07 | pH5 | yes |
| 15:49.27 | dcordes | that will be useful to easily compare things |
| 15:49.45 | kiozen | cr2: but I do not think it's 0 for negative values |
| 15:49.52 | pH5 | dcordes: you'll have all the linuxtogo branches in origin/* |
| 15:50.19 | dcordes | wow that seemed fast |
| 15:51.11 | cr2 | kiozen: r0 is the bit count. 0-7 for 8bit and 0-15 for 16bit |
| 15:52.07 | dcordes | wonders if his patch from android rebase will still apply |
| 15:52.26 | cr2 | kiozen: i had a feeling that it's just some dumb macro (for 32bit transfers maybe, or even longer) but wanted an independent confirmation. |
| 15:52.37 | pH5 | should, today linuxtogo's htc-msm is exactly the same as android's android-msm |
| 15:53.17 | dcordes | pH5, will it be the same patch -p1 < diff.diff in rootdir process? |
| 15:53.41 | cr2 | kiozen: the next command is 'bne' ? so it's always fals.e |
| 15:54.17 | kiozen | cr2: what if r0 = 0x0F? |
| 15:55.09 | kiozen | that would result into r1 = 0x0f - 0x80 |
| 15:55.13 | kiozen | grr |
| 15:55.25 | kiozen | r1 = 0x0f - 0x08 |
| 15:57.02 | dcordes | cr2, can't I just send you a complete memory dump? |
| 15:57.24 | pH5 | dcordes: I think so. Or if you have patches generated from git commits via git-format-patch, you can just apply them with git-am. |
| 15:58.18 | dcordes | yes I did it with git-diff origin/android-msm |
| 15:58.51 | dcordes | and git-add all files before |
| 16:00.57 | dcordes | pH5, clone done. do I make a "git checkout -b htc-msm origin/htc-msm" ? |
| 16:01.14 | cr2 | dcordes: hmm. found a local copy. |
| 16:01.39 | pH5 | dcordes: exactly |
| 16:01.45 | dcordes | cr2, ok. would be nice if you could look for fifo |
| 16:01.46 | cr2 | kiozen: confusing... |
| 16:02.25 | kiozen | cr2 http://pastebin.ca/1007084 |
| 16:02.38 | kiozen | if have tried that with gcc and kdbg |
| 16:02.49 | kiozen | r3 will be 0x07 at the end |
| 16:03.31 | kiozen | the second line will copy r4 to r3 |
| 16:03.33 | exco | where do I put g_ether so it loads on startup? |
| 16:04.21 | kiozen | the 3rd line will shift r3 to 0x01 |
| 16:04.40 | exco | to /etc/modutils and create a new file? |
| 16:05.12 | kiozen | cr2: and the 4th will subtract 0x08 from 0x0f -> 0x07 |
| 16:06.44 | dcordes | pH5, I accidently did a "git-config my name" instead of "git-config user.name my name" . did that break any config? |
| 16:06.48 | cr2 | kiozen: so it'll be true after each byte ? i need to draw the diagram ;) |
| 16:07.41 | pH5 | dcordes: no. git-config --list shows all configuration variables |
| 16:07.42 | cr2 | kiozen: this code allows to jump around the rising edge |
| 16:07.53 | pH5 | I guess you can just --unset "my" |
| 16:07.57 | kiozen | aaaaaaaaah! |
| 16:09.24 | cr2 | kiozen: for r0=0; r0<16; r0++ |
| 16:10.09 | kiozen | cr2: http://pastebin.ca/1007098 |
| 16:11.08 | dcordes | pH5, ok. I just patched and noticed I again have all updated/new files to the index. is there a function in git to automat thist? git-add file1 file2.. is nasty |
| 16:11.25 | dcordes | s/have all/have to add/ |
| 16:12.31 | pH5 | dcordes: if you give a directory, all included files will be added recursively. also there's git add -i |
| 16:14.32 | cr2 | kiozen: hm. it's the original |
| 16:15.06 | dcordes | pH5, did a git-add . in kernel followed by the ssh commit command you showed me above. and it says everything up to date |
| 16:16.06 | pH5 | because you didn't commit anything yet. git-add only adds stuff to the index |
| 16:16.21 | dcordes | I did a git push ssh://lgorris@linuxtogo.org/home/groups/mobile-linux/kernel.git htc-msm |
| 16:17.57 | cr2 | kiozen: the asm looks different... |
| 16:19.04 | pH5 | dcordes: that's right. but... |
| 16:19.09 | pH5 | do you want to apply all your changes in a single commit? |
| 16:19.45 | pH5 | or are there logical separations (kaiser base, driver1, driver2) etc. that would make more sense? |
| 16:19.51 | kiozen | cr2: have to cook, bbl |
| 16:19.53 | dcordes | pH5, it would be hard to reconstruct the development progress |
| 16:20.24 | dcordes | and I'm not entirely inside the initial patching |
| 16:20.48 | exco | the smile is back |
| 16:21.09 | pH5 | dcordes: I know. but if you hacked completely separate parts of the tree, it might make sense to commit only base kaiser support in one go and maybe changes that you made to other subsystems as separate commits. |
| 16:21.25 | pH5 | if you just want to dump everything in one single big commit, "git commit -a" is what you want |
| 16:21.26 | *** join/#htc-linux diogene31 (n=rj@mur31-2-82-243-122-54.fbx.proxad.net) |
| 16:21.37 | pH5 | (commit everything you just added to the index with git-add) |
| 16:22.11 | dcordes | pH5, I think that really is what I want for now. Most of the stuff is in arch/arm/mach-msm/ anywawy |
| 16:22.44 | dcordes | pH5, besides there is htc-egpio and I think the other stuff already was added to the android git |
| 16:24.21 | pH5 | dcordes: ok, right now it's your branch so you decide how to do it best :) |
| 16:27.24 | dcordes | pH5, to do a seperation I would just git-commit file1 file2 file3 write a description, commit, git-commit file4 file5 file6 ...? |
| 16:27.32 | pH5 | exactly |
| 16:28.49 | dcordes | pH5, I did git-commit files now I'm a bit clueless what to type? |
| 16:29.00 | dcordes | Please enter the commit message for your changes. |
| 16:29.13 | dcordes | I'm typing but nothing happens |
| 16:29.47 | pH5 | it opens your $EDITOR |
| 16:29.59 | pH5 | probably nano or vi, depending on your distro/setup |
| 16:30.53 | pH5 | if it's vi, just type "i", enter your commit message, press Esc and then type ":wq" and press Enter |
| 16:32.02 | dcordes | pH5, ok. I'm about to add egpio. Just noticed it has also changes in board.h and .c |
| 16:32.17 | dcordes | but that will be provided in a different commit also |
| 16:32.26 | dcordes | what does the usual kernel hacker do in that situation? |
| 16:33.03 | pH5 | I usually would make three commits. a) board.h/c without egpio, b) egpio driver, c) egpio related changes to board.h/c |
| 16:33.46 | dcordes | that means I would have to rip off the egpio stuff again from the board. I fear I'm not brave enough for that |
| 16:34.07 | pH5 | dcordes: then just drop the egpio driver first and put the whole board stuff on top of that |
| 16:34.12 | dcordes | ok |
| 16:34.25 | pH5 | did you use htc-egpio from mainline or from hh.org? |
| 16:34.37 | dcordes | hh.org |
| 16:34.39 | dcordes | mainline didn't work |
| 16:34.54 | cr2 | pH5: i've corrected (a bit) the alpine memory map. IMHO it can't be supported by htc-egpio, because the irq area is separate from gpio |
| 16:35.02 | cr2 | pH5: http://wiki.xda-developers.com/index.php?pagename=AlpineMemoryMap |
| 16:35.11 | dcordes | I know you told BabelO and me to use the mainline but I couldn't fiddle that in |
| 16:35.33 | cr2 | pH5: and the pasic3 values are written verbatim as they are entered in wiki (at init) |
| 16:35.55 | cr2 | pH5: if you suspend, then pasic3 gets some different values fed in. |
| 16:36.07 | pH5 | then you'll have to update anyway once the mainline driver comes in with 2.6.26, hh.org is ok for now if it works |
| 16:36.30 | dcordes | yea works fine |
| 16:36.45 | pH5 | cr2: true, that'd either need two platform devices for gpio/irq or another resource input for htc-egpio to handle separate chip selects for gpio/irq |
| 16:36.58 | cr2 | pH5: ok, sounds good. |
| 16:37.39 | cr2 | pH5: i think 2 devices is ok. i'm using 2 devices on athena. |
| 16:39.14 | pH5 | cr2: ok |
| 16:39.38 | pH5 | cr2: do you have any idea what those 0x20..0x23 registers could be about? |
| 16:39.52 | pH5 | they have to do something with the blinking, but I can't wrap my head around what they actually do. |
| 16:40.14 | exco | if you run ipkg update ; ipkg upgrade on the mobile .. do you expect to break it ? |
| 16:40.57 | pH5 | probably. there is no angstrom machine mentor for magician who regularly takes care that these thinks work. |
| 16:41.07 | pH5 | s/inks/ings/ |
| 16:42.52 | dcordes | pH5, will git be able to show me what I didn't commit yet after the commit is done? |
| 16:43.34 | pH5 | yep. git status |
| 16:43.39 | dcordes | k |
| 16:43.56 | pH5 | or "git diff HEAD" |
| 16:44.28 | exco | apt: where was the difference beteen your and pH5's post |
| 16:44.32 | dcordes | phew I hope this stuff will be any usable after it's commited :D |
| 16:45.16 | pH5 | exco: apt is a bot. the difference was a corrected typo |
| 16:45.20 | dcordes | pH5, Created commit 2a82a48: basic support for htc kaiser and htc vogue |
| 16:45.26 | pH5 | dcordes: well, otherwise you'll have to fix it ;) |
| 16:45.43 | pH5 | dcordes: yay. push. |
| 16:45.44 | dcordes | pH5, now when I do the ssh push stuff 2a82a48 will be pushed? |
| 16:45.57 | exco | live and learn :-) |
| 16:46.28 | pH5 | yes, because with each commit on htc-msm, the htc-msm branch head moves. when you push htc-msm, all your commits are included. |
| 16:46.47 | dcordes | ok I will do all commits then before I push |
| 16:48.04 | exco | I guess I'm too unexperienced to help out with that machine mentor thing ? |
| 16:48.45 | cr2 | pH5: what are their values for magician ? |
| 16:51.19 | cr2 | pH5: 0x10- 0x14 are also cryptic |
| 16:52.58 | *** join/#htc-linux rob_w (n=bob@X120f.x.pppool.de) |
| 16:54.40 | dcordes | how to remove a line in vi :x? |
| 16:59.19 | *** join/#htc-linux p3t3r__ (n=p3t3r__@gate1.net-you.de) |
| 16:59.38 | dcordes | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=shortlog;h=refs/heads/htc-msm :-) |
| 17:02.05 | *** join/#htc-linux LunohoD_ (n=alex@e180074141.adsl.alicedsl.de) |
| 17:04.51 | *** join/#htc-linux S^P (n=Masood@static-host119-30-82-178.link.net.pk) |
| 17:08.52 | dcordes | cr2, did you look at kaiser dump? |
| 17:09.59 | cr2 | pH5: where is the hh kernel ? i've lost the link to Kevin2's kbd-spi driver again. |
| 17:10.16 | cr2 | dcordes: not everything at once. |
| 17:10.36 | dcordes | you looking for hh on ltg? |
| 17:11.13 | cr2 | yes. |
| 17:11.15 | pH5 | cr2: I added magician initial pasic3 values to the wiki |
| 17:11.24 | cr2 | pH5: ok. |
| 17:11.49 | cr2 | i'm wondering if -msm and hh clash too much |
| 17:11.54 | pH5 | cr2: gitweb is at git.linuxtogo.org |
| 17:12.07 | pH5 | cr2: we're not trying to merge them (yet?) |
| 17:12.14 | dcordes | http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=tree;h=refs/heads/hh.org;hb=refs/heads/hh.org |
| 17:12.16 | cr2 | pH5: heads/work ? |
| 17:13.29 | cr2 | pH5: i can see only 1 major trouble: asic3_mmc. everything else is bearable |
| 17:14.41 | exco | if OM quits loading like 1mm bevor the bar touches the right side... where do I look first? |
| 17:16.49 | exco | before |
| 17:18.20 | pH5 | I'd boot with psplash=false, if you boot from haret you can add it to cmdline in default.txt |
| 17:18.29 | pH5 | usb0 didn't come up yet? |
| 17:21.57 | exco | yes, it did |
| 17:22.09 | exco | I just had to modprobe g_ether like you said |
| 17:22.25 | cr2 | dcordes: can you dump the wince kernel from ram ? have you ever tried testwm5 ? |
| 17:22.37 | exco | now I steup interfaces to have the right gateway... works like a charm now |
| 17:23.07 | pH5 | so can you log in via ssh when OM loading stops? |
| 17:24.09 | exco | I think it is to early... but I'll try if you want (I just reverted back to before ipkg upgrade) |
| 17:25.09 | pH5 | exco: never mind, it's just what I do when something hangs. a) turn of splash to see boot messages and b) try to log in somehow (edit boot scripts to load g_ether and ifup usb0 early, if needed) |
| 17:28.53 | exco | that's what I would do ... if I could ... I'll get there |
| 17:29.33 | *** join/#htc-linux Robwoerle (n=bob@Mb72f.m.pppool.de) |
| 17:31.09 | exco | that's odd.. now ipkg tells me there's nothing to be upgraded after I run ipkg update ipkg upgrade... |
| 17:33.22 | cr2 | pH5: ds2746 ? |
| 17:33.25 | dcordes | cr2, how to dump the kernel? no never tried testwm5 |
| 17:33.38 | cr2 | dcordes: try testwm5 first |
| 17:33.45 | dcordes | ok |
| 17:33.48 | dcordes | where can I get it? |
| 17:34.01 | cr2 | http://hands.com/~lkcl |
| 17:35.26 | dcordes | lol it has button1 |
| 17:35.27 | cr2 | dcordes: lucky you. ds2746 is an i2c chip |
| 17:35.44 | marbalon | back |
| 17:36.06 | cr2 | marbalon: did you run i2c tools on kaiser ? |
| 17:36.20 | cr2 | i suspect a ds2746 there. |
| 17:36.37 | pH5 | cr2: kaiser use ds2746 instead of ds2760? |
| 17:36.38 | dcordes | cr2, isn't i2c the touchscreen stuff? we have perfectly working ts |
| 17:36.52 | marbalon | cr2: yes, but but get some errors while use i2cdetect |
| 17:38.17 | cr2 | pH5: i think so. |
| 17:38.36 | dcordes | cr2, should testwm give me any fancy output files? |
| 17:38.49 | cr2 | dcordes: there may be multiple devices on i2c bus |
| 17:38.57 | exco | what seems to be the issue here http://pastebin.ca/1007185 ? |
| 17:38.57 | exco | I'm trying to install nano using: ipkg install nano |
| 17:38.58 | cr2 | dcordes: you have \Storage Card ? |
| 17:39.09 | dcordes | nopes. german rom |
| 17:39.13 | cr2 | hmm. |
| 17:39.26 | cr2 | dcordes: dump the nk.exe instead |
| 17:39.32 | dcordes | why do they all hardcode that. also in linload |
| 17:40.51 | dcordes | cr2, with pmemdump.exe? |
| 17:40.52 | cr2 | it's easier to hardcode than to read from registry ;) |
| 17:41.20 | pH5 | exco: that repository is gone |
| 17:41.34 | exco | I just used it like 15minutes ago |
| 17:42.04 | pH5 | look for Packages.gz files below http://buildhost.openmoko.org/snapshots/ and update /etc/ipkg.conf (I think) |
| 17:42.08 | pH5 | really? |
| 17:42.25 | exco | yes... I told you I made an upgrade ... :-) |
| 17:42.30 | cr2 | dcordes: pwf kaisernk 0x10300000 0x300000 |
| 17:42.36 | pH5 | ah, I'm sorry |
| 17:42.42 | pH5 | only the neo1973 directory is gone |
| 17:43.02 | pH5 | so that result probably means there is no nano in the openmoko feeds |
| 17:43.46 | cr2 | dcordes: the source for testwm5 is available. it even works to some extent if compiled with cegcc. but there are some header bugs that need to be fixed first |
| 17:43.52 | exco | we'll.. it was - 15 minutes ago - but you're right - now it's gone |
| 17:44.22 | dcordes | cr2, I will look at this or flash an english rom tomorrow |
| 17:45.36 | pH5 | exco: try the angstrom feeds |
| 17:45.47 | pH5 | http://www.angstrom-distribution.org/feeds/2007/ipk/glibc/armv5te/base/nano_2.0.6-r0_armv5te.ipk |
| 17:46.45 | cr2 | dcordes: ok. |
| 17:47.04 | cr2 | This application has requested the Runtime to terminate it in an unusual way. |
| 17:47.05 | cr2 | Please contact the application's support team for more information. |
| 17:47.30 | cr2 | hehe :) |
| 17:48.15 | dcordes | cr2, dump'd. anything else to dump? I'm off in a min |
| 17:48.42 | exco | so |
| 17:48.44 | cr2 | no. i've found it too ;) sorry for troubling you |
| 17:49.18 | exco | /etc/ipkg.confdest root / |
| 17:49.18 | exco | lists_dir ext /var/lib/ipkg |
| 17:49.21 | exco | sorry |
| 17:49.59 | exco | ipkg.conf points to var/lib/ipkg - so I have to uncomment the link that is now gone and add the armstrong repos there? |
| 17:50.53 | pH5 | the feeds are in /etc/ipkg/some-feed.conf |
| 17:51.19 | exco | thanks |
| 17:52.20 | dcordes | cr2, you found the kaisernk? |
| 17:53.03 | cr2 | yes |
| 17:53.37 | dcordes | ok |
| 17:53.45 | dcordes | pH5, did you see my commit arrangement? |
| 17:54.02 | dcordes | I hope this is any useful |
| 17:56.52 | dcordes | bye |
| 17:59.36 | exco | I uncommented the snapshot-neo1973 now I get ruturn error 7 instead of 4 |
| 18:05.19 | *** join/#htc-linux LunohoD_ (n=alex@e180072169.adsl.alicedsl.de) |
| 18:08.46 | *** join/#htc-linux S^P (n=Masood@static-host119-30-85-103.link.net.pk) |
| 18:12.17 | *** join/#htc-linux CVirus (n=GoD@196.205.192.134) |
| 18:13.11 | *** join/#htc-linux miknix (n=miknix@bl4-82-69.dsl.telepac.pt) |
| 18:13.47 | exco | thanks for nano |
| 18:26.46 | paulproteus | Morning, everyone. |
| 18:27.27 | kiozen | cr2: back again, so what is next? |
| 18:27.47 | exco | hey paulproteus |
| 18:28.05 | exco | well... the test was successful |
| 18:28.33 | cr2 | kiozen: i'll recursively convert the asm to c |
| 18:28.55 | kiozen | anything left I can do with that last issue? |
| 18:29.19 | paulproteus | exco, Awesome - for me, it seems to have lost the GSM connection overnight. |
| 18:29.20 | cr2 | kiozen: the obvious functions like set_cs(gpio_cs) and fe(gpio_clk) first |
| 18:29.37 | paulproteus | pH5 was talking about Magician machine mentor for Angstrom. |
| 18:30.00 | cr2 | kiozen: doing it right now. |
| 18:30.02 | exco | the GSM part doesn't seem to stable |
| 18:30.13 | paulproteus | I got some dependency errors last night when I tried to build openmoko-image with Angstrom/OE. |
| 18:30.27 | kiozen | ok, so I will fall back to task 'loading geotiffs into GT directly' :) |
| 18:30.32 | exco | paulproteus: he said there's none at the moment |
| 18:31.59 | exco | paulproteus: I experimented a bit today... you can just unpack openmoko.linux.rootfs.tar.bz2 |
| 18:32.22 | exco | add the hh20 kernel modules and put the hh20 kernel on the fat partiton and boot it using hret |
| 18:33.15 | marbalon | pH5: continuing our conversation about egpio and irqs |
| 18:33.32 | marbalon | pH5: kaiser don't have irq related to egpio http://wiki.xda-developers.com/index.php?pagename=Kaiser_IRQ |
| 18:33.42 | exco | paulproteus: GSM still "works" |
| 18:34.01 | exco | so now I'm using hh20 with OM |
| 18:34.41 | CVirus | hh20 ? |
| 18:34.47 | miknix | BabelO, ping |
| 18:34.50 | CVirus | exco: what phone is that ? |
| 18:34.53 | exco | also I was able to run ipkg update & upgrade (didn't boot all the way after that) |
| 18:35.33 | marbalon | pH5: I thought it should be a CPU irq - correct me if I'm wrong |
| 18:35.35 | exco | CVirus: I guess it's a kernel revision? |
| 18:35.51 | CVirus | what phone ? |
| 18:35.54 | exco | Magician |
| 18:36.04 | CVirus | ah ok |
| 18:36.32 | paulproteus | exco, Oh, that's good at least re: hh20 + extract. |
| 18:36.53 | pH5 | marbalon: I wasn't aware that you were talking about kaiser |
| 18:37.01 | *** join/#htc-linux tsdogs (n=tsdogs@62.123.180.130) |
| 18:37.01 | pH5 | sometimes forgets who is working on what |
| 18:37.19 | exco | paulproteus: yes, exactly |
| 18:37.30 | BabelO | ~pong miknix |
| 18:37.31 | apt | miknix: PONG!!!!!!!!!!!!!!!!!! |
| 18:37.35 | pH5 | marbalon: if the kaiser cpld does not have irqs (gpio only), then you don't need to supply it with an irq resource at all |
| 18:37.48 | miknix | BabelO, working on your artemis? |
| 18:37.55 | BabelO | miknix: yes |
| 18:38.36 | miknix | BabelO, the kernel version I was working yesterday had very upstream changes. they removed a lot of code |
| 18:39.07 | darkerror05 | hello |
| 18:39.18 | miknix | BabelO, so I'm trying to get i2c working on a more stable kernel version |
| 18:39.26 | exco | paulproteus: you loose the changes goxboxlive made ... but we could ask him what he did (so far I've only seen that you need to load g_ether and setup interfaces if you want to use ssh over USB) |
| 18:39.34 | darkerror05 | does anyone know how to stablize wifi usage on a htcuniversal under linux? |
| 18:39.42 | marbalon | pH5: hmm a read gpio driver for kaiser and it uses IRQ 32 and 33 for normal gpio, this numeration is above cpu irq ??? |
| 18:39.55 | darkerror05 | if i try to download something for more then 3 minutes the wifi locks up |
| 18:40.50 | pH5 | marbalon: this is invalid. there are no irqs that correspond to the CPLD's GPIOs if it doesn't have a separate irq register. |
| 18:40.53 | exco | paulproteus: so what next? |
| 18:41.17 | miknix | : o |
| 18:41.23 | miknix | ~ping Kevin2 |
| 18:41.24 | apt | pong Kevin2 |
| 18:42.39 | paulproteus | exco, Well, I think we should fix Angstrom. |
| 18:42.49 | pH5 | but I'm talking without knowing much about the cpld. are there input gpios on it? keys? if yes, there has to be a corresponding mux gpio, but it should then be easy to find with haret. |
| 18:42.50 | exco | you mean GSM? |
| 18:42.52 | paulproteus | We should try to build a fresher OpenMoko for the phone as well as a modern Qtopia image. |
| 18:43.06 | paulproteus | exco, My hope is that it's just a silly userspace issue as far as GSM reliability. |
| 18:43.19 | pH5 | paulproteus, exco: yay, I see a userspace team forming |
| 18:43.19 | paulproteus | The OpenMoko software has become*way* more stable as of a few weeks ago. |
| 18:43.42 | paulproteus | pH5, It's why I bought a second SIM to use on the Magician. |
| 18:43.51 | paulproteus | Now that I think about it, maybe I should just have two Magicians, one with my real SIM. (-: |
| 18:44.05 | paulproteus | But anyway - yes, userspace on Magician. |
| 18:44.07 | exco | ;-) |
| 18:44.42 | darkerror05 | is Xanadux obsolite? |
| 18:45.10 | exco | paulproteus: did you check for n_tihtc in the Armstrong image? |
| 18:47.18 | *** join/#htc-linux tsdogs (n=tsdogs@62.123.180.130) |
| 18:47.44 | paulproteus | exco, I haven't checked for it yet, no. |
| 18:48.19 | paulproteus | pH5, I get dependency errors trying to build an openmoko-image. |
| 18:48.37 | paulproteus | Is "bitbake openmoko-image" the right way to build a fresh image of the OpenMoko stack? |
| 18:49.05 | exco | the image we're running now has it enabled - hence the "working" GSM |
| 18:49.14 | paulproteus | exco, It's true (-: |
| 18:49.18 | paulproteus | I can reboot into it and see. |
| 18:49.56 | paulproteus | Still, I want to know the right way so that our work is sustainable. |
| 18:50.10 | exco | the right way regarding GSM? |
| 18:50.16 | paulproteus | Regarding building images. |
| 18:50.25 | exco | yes, of course |
| 18:51.04 | pH5 | paulproteus: I always do bitbake openmoko-devel-image with DISTRO=angstrom-2008.1, but I guess openmoko-image should work the same. you just don't have to build all the native toolchain and debugging cruft. |
| 18:51.23 | paulproteus | I'm trying -devel-image now. |
| 18:51.34 | paulproteus | NOTE: Runtime target 'openmoko-icon-theme-standard2-qvga' is unbuildable, removing... |
| 18:51.34 | paulproteus | Missing or unbuildable dependency chain was: ['openmoko-icon-theme-standard2-qvga', 'librsvg-native'] |
| 18:51.34 | paulproteus | NOTE: Runtime target 'task-openmoko-ui' is unbuildable, removing... |
| 18:51.34 | paulproteus | Missing or unbuildable dependency chain was: ['task-openmoko-ui', 'openmoko-icon-theme-standard2-qvga', 'librsvg-native'] |
| 18:51.34 | paulproteus | ERROR: Required build target 'openmoko-devel-image' has no buildable providers. |
| 18:51.36 | paulproteus | Missing or unbuildable dependency chain was: ['openmoko-devel-image', 'task-openmoko-ui', 'openmoko-icon-theme-standard2-qvga', 'librsvg-native'] |
| 18:51.42 | *** join/#htc-linux LunohoD_ (n=alex@e180078190.adsl.alicedsl.de) |
| 18:52.32 | paulproteus | exco, Yes - n_tihtc is loaded in the Angstrom image. |
| 18:52.57 | exco | hmm... so we have to find out why it's working with OM? |
| 18:53.14 | paulproteus | Well, I think we should start by getting an OpenMoko stack from this year. (-: |
| 18:53.25 | pH5 | oh sh*t, for openmoko-icon-theme-standard2-qvga you have to add ASSUME_PROVIDED += "imagemagick-native librsvg-native" to your local.conf (and install imagemagick and librsvg on the host) |
| 18:53.35 | exco | paulproteus: or if it's working with your hopefully successful new OM image stick to that? |
| 18:53.53 | pH5 | I think the openmoko-icon-theme-standard2 *should* have the small icons by now so we can drop the -qvga packages, but I'm not sure. |
| 18:55.40 | paulproteus | pH5, I'll try that right now, thanks. |
| 18:58.29 | paulproteus | pH5, What handheld are you using? |
| 18:58.29 | *** join/#htc-linux kiozen_ (n=kiozen@rgnb-4db1d8d4.pool.einsundeins.de) |
| 18:58.33 | paulproteus | And are you using Qtopia at all? |
| 18:58.52 | paulproteus | pH5, Thanks, it's starting to build openmoko-devel-image now btw. |
| 18:59.10 | cr2 | kiozen: done. |
| 18:59.39 | kiozen_ | ok |
| 18:59.40 | cr2 | kiozen: looks like a start bit for each byte to me. |
| 18:59.47 | cr2 | i'll pastebin. |
| 19:00.00 | pH5 | paulproteus: O2 Xda mini (rev 3 magician with toppoly display and 64MB RAM) |
| 19:00.02 | paulproteus | exco, Yeah, that's what I'm thinking - if it works with the new OM image, let's just build on top of that. |
| 19:00.07 | paulproteus | pH5, Oh, nice! |
| 19:00.13 | cr2 | if your math is correct, it can be done a bit easier :) |
| 19:00.20 | paulproteus | I didn't realize there were other Magician users floating around. |
| 19:00.28 | paulproteus | Using it as a phone, ph5? |
| 19:01.04 | kiozen_ | cr2: hope it is, I have never done anything with pxa assembly |
| 19:01.14 | *** part/#htc-linux p3t3r__ (n=p3t3r__@gate1.net-you.de) |
| 19:03.14 | pH5 | paulproteus: not really, I mostly play kernel hacker. |
| 19:03.19 | cr2 | kiozen: http://pastebin.ca/1007285 |
| 19:03.20 | paulproteus | Cool. |
| 19:04.03 | cr2 | pH5: does this code make sense to you http://pastebin.ca/1007285 ? |
| 19:04.18 | cr2 | pH5: we are on the way to blow up the LCD :) |
| 19:07.49 | cr2 | i don't understand only 1 thing. why do they send as fast as possible. on BA6 there is a 10ms delay before _tx and _fe |
| 19:08.27 | pH5 | cr2: what is set_fe doing exactly? |
| 19:09.12 | pH5 | and I don't really understand why there is a start bit 1 sent before byte1 and not before byte0 |
| 19:09.14 | kiozen_ | cr2: "if (count%8 == 0)" there is no "count" |
| 19:09.29 | *** join/#htc-linux Marex (n=marex@vasut.kolej.mff.cuni.cz) |
| 19:10.16 | pH5 | cr2: is the ba clocked faster during boot or something similarly trivial? |
| 19:11.10 | cr2 | kiozen: count1 or course |
| 19:11.21 | cr2 | pH5: create falling edge |
| 19:11.39 | cr2 | pH5: i doubt it. |
| 19:12.22 | pH5 | so it's just pulling the gpio down and then up again - is there a delay there? |
| 19:12.25 | tsdogs | darkerror05: afair wifi is only stable at 1Mbps |
| 19:12.36 | cr2 | only pulling, no delay. |
| 19:13.54 | cr2 | tsdogs: i've just looked at the acx100 patch yesterday. patching firmware for hx4700 looks really nasty. but afair we have the same firmware, but this patching does not work. |
| 19:14.50 | tsdogs | hmm, well I never used wifi on uni, as I only have wpa APs |
| 19:19.02 | exco | paulproteus: how's it going with that image? |
| 19:26.38 | paulproteus | exco, dbus is starting to configure. |
| 19:26.42 | paulproteus | Task 587 of 1916 |
| 19:26.58 | *** join/#htc-linux kiozen (n=kiozen@rgnb-4db1d8c3.pool.einsundeins.de) |
| 19:27.27 | paulproteus | I'm also repacking the Qtopia git repository in the hopes it'll be smaller. |
| 19:27.40 | cr2 | marbalon: are you here ? |
| 19:27.51 | marbalon | cr2: yes |
| 19:28.33 | cr2 | marbalon: can you check this address on kaiser |
| 19:29.02 | cr2 | pd 0x90000 0x20 |
| 19:29.18 | marbalon | cr2: under haret ? |
| 19:29.21 | cr2 | yes |
| 19:29.30 | marbalon | cr2: restart phone.. |
| 19:32.12 | marbalon | cr2: 00090000 | 00000063 ffffffff ffffffff ffffffff | c............... |
| 19:32.23 | marbalon | cr2: 00090010 | 00000000 00000001 ffffffff 00000000 | ................ |
| 19:32.41 | exco | I have to try building an image ... sometime soon |
| 19:32.48 | marbalon | cr2: what do you try to find ? |
| 19:32.49 | paulproteus | exco, It's not too hard really. |
| 19:33.14 | cr2 | marbalon: the phone is on ? |
| 19:33.34 | exco | also I will build a custom kernel for this Ubuntu once I know exactly what I need |
| 19:33.50 | marbalon | cr2: no |
| 19:34.00 | cr2 | switch it on. |
| 19:34.02 | marbalon | cr2: repeat when phone is on ? |
| 19:34.05 | exco | do you need OE to build that image or can you also do that with Eclipse? |
| 19:34.05 | cr2 | yes |
| 19:34.30 | paulproteus | exco, You will need to use OE. |
| 19:34.39 | paulproteus | But if you read the Getting Started page, you'll find it's not too hard. |
| 19:34.46 | paulproteus | And I can help as necessary. |
| 19:34.59 | marbalon | cr2: the same values |
| 19:35.06 | exco | do you know what you're getting yourself into? |
| 19:35.37 | paulproteus | exco, Er, who, me? |
| 19:35.50 | exco | yes, paulproteus |
| 19:36.00 | cr2 | marbalon: hmm. |
| 19:36.06 | paulproteus | I think I'm getting myself into spending a few dozen compute-hours building an image for this phone.... |
| 19:36.10 | paulproteus | Am I getting myself into something else? |
| 19:36.28 | cr2 | marbalon: do you have \Storage Card ? |
| 19:36.51 | exco | "paulproteus: And I can help as necessary." |
| 19:36.57 | paulproteus | exco, Oh, well. (-: |
| 19:37.04 | paulproteus | So far you've been a pretty good student. |
| 19:37.40 | marbalon | cr2: yes |
| 19:37.41 | exco | normally I'm getting the hang of stuff pretty fast - thank god |
| 19:40.51 | *** join/#htc-linux patoche (n=root@lns-bzn-27-82-248-48-182.adsl.proxad.net) |
| 19:40.56 | *** join/#htc-linux patoche (n=root@lns-bzn-27-82-248-48-182.adsl.proxad.net) |
| 19:41.18 | marbalon | cr2: do you thinking about testwm5 ? |
| 19:41.39 | *** join/#htc-linux CrAsH_x64 (n=crash@84.78.99.40) |
| 19:42.09 | exco | paulproteus: as for building the image - can you point me where to start |
| 19:42.38 | cr2 | majorburn: yes, you can look an the wince kernel |
| 19:43.32 | marbalon | cr2: majorburn -> me ? |
| 19:43.49 | *** join/#htc-linux pikapika (n=pikapika@mar75-8-88-164-227-147.fbx.proxad.net) |
| 19:44.04 | pikapika | hi |
| 19:44.04 | paulproteus | exco, http://oe.linuxtogo.org/node/20 ! |
| 19:47.09 | cr2 | hm. yes. |
| 19:48.06 | marbalon | cr2: link to this tool... |
| 19:48.20 | marbalon | cr2: and mini how to ;) |
| 19:49.22 | cr2 | marbalon: http://hands.com/~lkcl |
| 19:49.54 | cr2 | marbalon: put on the sd card, and run. you'd have 128+MB free space on the sd. |
| 19:52.27 | marbalon | cr2: testwm5 working..... |
| 19:53.38 | marbalon | cr2: what is output ? maybe useful for me... |
| 19:54.04 | CrAsH_x64 | hello, anybody knows where i can get updated image for htc universal? the lastest image i have found was built last year |
| 20:03.48 | marbalon | cr2: how long this take ? working and working... ;) |
| 20:04.13 | paulproteus | exco, That's a lot to chew on, but there's no reason I should be the bottleneck for us two using new images. (-: |
| 20:05.02 | marbalon | cr2: just finish, you need all files ? |
| 20:09.52 | tsdogs | CrAsH_x64: http://www.linuxtogo.org/~htcpxa/htcuniversal/index.html those are the latest built. |
| 20:12.17 | *** join/#htc-linux Marex (n=marex@vasut.kolej.mff.cuni.cz) |
| 20:13.13 | cr2 | marbalon: no, only nk.exe |
| 20:14.06 | CrAsH_x64 | thanks tsdogs |
| 20:22.42 | marbalon | cr2: http://baliniak.pl/android/nk.exe |
| 20:23.54 | marbalon | full dump should be available soon |
| 20:24.05 | cr2 | marbalon: forbidded |
| 20:25.38 | marbalon | cr2: moment... |
| 20:26.09 | marbalon | cr2: try now |
| 20:26.39 | cr2 | ok. |
| 20:26.51 | marbalon | cr2: what for is nk module ? |
| 20:29.26 | cr2 | it's the wince kernel |
| 20:34.31 | cr2 | marbalon: i've fixed some memory wiki bugs, and added BT to gpio. interesting.... |
| 20:36.05 | cr2 | marbalon: can you check the MSM_DMOV, where it's located in the android kernel ? |
| 20:36.42 | marbalon | cr2: I fixed some cpld val yesterday, find keyboard leds and backlight ;) |
| 20:37.28 | marbalon | cr2: in kernel source ? |
| 20:37.58 | cr2 | yes |
| 20:38.26 | cr2 | marbalon: there will be some extra cpld bits. for arm9 reset and such. |
| 20:38.47 | cr2 | i think we will need 1 extra cpld gpio bank. |
| 20:38.52 | cr2 | neunth |
| 20:39.07 | marbalon | cr2: but still can't find egpio irq :( |
| 20:40.05 | marbalon | cr2: #define MSM_DMOV_BASE 0xE0002000 |
| 20:40.17 | marbalon | cr2: #define MSM_DMOV_PHYS 0xA9700000 |
| 20:40.38 | cr2 | a97. ok, so it's one more bug. |
| 20:41.10 | cr2 | marbalon: there is no egpio irq. |
| 20:41.17 | exco | paulproteus: the installation instructions on linuxtogo (OEandYourDistro) are not up to date, I'm trying |
| 20:41.33 | paulproteus | exco, /me nods |
| 20:41.39 | paulproteus | Try to update it with the correct stuff if possible. |
| 20:41.42 | paulproteus | It is a wiki after all. |
| 20:42.50 | exco | create a sub section for Ubuntu 8.04? |
| 20:43.07 | marbalon | cr2: msm have 2 irq vectors, 2x32 irq |
| 20:43.46 | paulproteus | Sure, that sounds reasonable. |
| 20:45.41 | cr2 | hm. i'm blind. |
| 20:45.43 | marbalon | cr2: first of them partialy documented in wiki, second one partialy in sources |
| 20:46.44 | marbalon | cr2: http://pastebin.ca/1007392 |
| 20:48.11 | cr2 | marbalon: DMA regs fixed now. |
| 20:48.55 | cr2 | marbalon: DMA +UART1DM is used by BT |
| 20:49.31 | marbalon | cr2: ok, I will check it tomorrow. good night |
| 20:49.39 | cr2 | marbalon: this list is for 7200A, there may be some differences |
| 20:52.40 | *** join/#htc-linux dzo (n=dzo@121.98.128.127) |
| 20:53.54 | cr2 | hi dzo |
| 20:54.05 | *** join/#htc-linux patoche (n=root@lns-bzn-27-82-248-48-182.adsl.proxad.net) |
| 21:00.07 | paulproteus | pH5, When bitbaking openmoko-devel-image, I get: |
| 21:00.07 | paulproteus | NOTE: package linux-handhelds-2.6-2.6.21-hh20-r20: task do_configure: started |
| 21:00.07 | paulproteus | ERROR: function do_configure failed |
| 21:00.07 | paulproteus | ERROR: log data follows (/home/paulproteus/openembedded/stuff/tmp/work/magician-angstrom-linux-gnueabi/linux-handhelds-2.6-2.6.21-hh20-r20/temp/log.do_configure.9432) |
| 21:00.08 | paulproteus | | /home/paulproteus/openembedded/stuff/tmp/deploy/uclibc/images/magician// does not exist, you may need to bitbake it separately |
| 21:00.11 | paulproteus | NOTE: Task failed: /home/paulproteus/openembedded/stuff/tmp/work/magician-angstrom-linux-gnueabi/linux-handhelds-2.6-2.6.21-hh20-r20/temp/log.do_configure.9432 |
| 21:00.16 | paulproteus | (And then it's all over.) |
| 21:03.11 | exco | :-( |
| 21:04.00 | exco | well... I'll start trying once you succeed, paulproteus |
| 21:42.36 | *** join/#htc-linux tuff (n=phil@tuffnatty.dialup.corbina.ru) |
| 21:45.48 | pH5 | paulproteus: you have to "ANGSTROM_MODE=uclibc bitbake initramfs-bootmenu-image" first so the kernel will be built with an included initramfs image. |
| 21:46.24 | pH5 | the idea is nice, although it .. complicates the build process a bit. |
| 21:47.16 | paulproteus | Okay, trying to build that now. |
| 21:48.22 | pH5 | .zZ |
| 21:49.42 | exco | paulproteus: do you have to take care of n_tihtc by compiling a module? |
| 21:50.30 | exco | I did try an upgrade earlier using ipkg... guess that wasn't really a good idea ? |
| 21:50.59 | paulproteus | It's always a little scary. (-: |
| 21:52.43 | exco | that would be really cool if you could just dist-upgrade or sth |
| 21:54.58 | exco | did you find a way to read your battery level |
| 21:56.52 | paulproteus | The battery level just shows up in the upper right corner. |
| 21:57.06 | paulproteus | I agree, dist-upgrade-like would be great. (-: |
| 21:57.43 | tsdogs | cr2: look in pm |
| 21:57.47 | exco | yes, but is it doesn't show percentage or how long it will last or anything.... |
| 21:58.15 | exco | and mine always shows green |
| 21:58.19 | paulproteus | exco, Yes, that's true, I haven't considered that part very carefully yet. |
| 21:58.58 | exco | well the 6 hours mine was in standby - it dropped less than 8% (I reset to windows to check) |
| 22:01.22 | exco | paulproteus: the only thing that gives some interesting reading is |
| 22:01.22 | exco | cat /sys/class/power_supply/backup-battery/voltage_now |
| 22:01.22 | exco | charge_now is bogus according to pH5 |
| 22:01.51 | exco | And I don't know if mine is charging while on USB |
| 22:02.36 | tsdogs | exco: do you have /sys/class/leds or something like that? |
| 22:04.46 | exco | yes, tsdogs |
| 22:05.20 | tsdogs | exco maybe simply echoing some value on the trigger for one led whould do the trick |
| 22:05.38 | tsdogs | what image are we talking about? |
| 22:05.46 | tsdogs | I mean gui? |
| 22:05.55 | exco | OpenMoko |
| 22:06.55 | tsdogs | hmm, ok have no idea about OM, but the led would be nice too |
| 22:07.36 | tsdogs | I don't exactly remember what the values can be, but you can get them with "cat trigger" in one of the leds dirs |
| 22:08.03 | cr2 | tsdogs: looking :) |
| 22:08.12 | exco | magician:blue$ cat trigger |
| 22:08.12 | exco | [none] ac-online usb-online timer hwtimer backlight backup-battery-charging-or-full backup-battery-charging backup-battery-full |
| 22:08.36 | tsdogs | btw does magician have 2 batteries? |
| 22:08.42 | BabelO | cr2: is there a way in gt to keep only the visual maps ? |
| 22:08.57 | tsdogs | ok then "echo backup-battery-charging > trigger" should enable the led when it's charging |
| 22:09.00 | exco | tsdogs: yes, it has a backup battery |
| 22:09.18 | tsdogs | exco: so you are missing the main battery there? |
| 22:09.46 | exco | it seems so |
| 22:10.08 | cr2 | BabelO: visual ? |
| 22:10.25 | exco | echoing doesn't enable the led |
| 22:10.51 | tsdogs | exco: it has to be triggered, so probably u need to disconnect and reconnect USB |
| 22:11.12 | tsdogs | but then that is not what you want as you are missing main battery |
| 22:11.28 | *** join/#htc-linux dcordes (n=lukas@f049185239.adsl.alicedsl.de) |
| 22:11.39 | dcordes | hi |
| 22:11.48 | cr2 | tsdogs: i think the program was written for qt1 originally :) |
| 22:11.52 | BabelO | cr2: i ve done my whole ign map in gt now, buti have some picture in bottom of others but only partial.. so i want only to keep picture on top |
| 22:11.58 | exco | well reconnecting didn't turn it on either |
| 22:12.33 | tsdogs | cr2: it been pretty hard to port, probably an interface rewrite would be better :) |
| 22:13.30 | exco | tsdogs: the echo command works (the echoed command shows brackets in nano) |
| 22:13.51 | cr2 | tsdogs: thanks :) it's always sad when the old programs disappear. |
| 22:14.36 | tsdogs | cr2: it still has bugs, and don't quite know how to display toolbars the good way, but it's a start. |
| 22:15.18 | cr2 | tsdogs: segfaulted :) |
| 22:15.22 | tsdogs | lol |
| 22:15.43 | tsdogs | I was afraid you would say that as a first run :) |
| 22:15.43 | cr2 | Program received signal SIGSEGV, Segmentation fault. |
| 22:15.43 | cr2 | [Switching to Thread 0x2b1905154b60 (LWP 17052)] |
| 22:15.43 | cr2 | 0x000000000041965a in MainWindow::doPlot () |
| 22:16.00 | cr2 | i'll recompile with -g |
| 22:16.13 | cr2 | i tried to load a ~300MB file |
| 22:16.22 | tsdogs | cr2: can u send me some data u are using ? lol |
| 22:16.52 | tsdogs | ok no 300MB file, but a test (don't think a 1 2 3 xyz value is good enough) |
| 22:18.51 | cr2 | 0x000000000041965a in MainWindow::doPlot (this=0x69e730, p=0x7fff7cc5c3b0) at mainwindow.cc:219 |
| 22:18.51 | cr2 | 219 if ( (Daten->getNDim (Daten->plotId(currentVar())) > 1) && (range_maximize->isChecked()) ) { |
| 22:19.07 | cr2 | hm. i think i know what's the problem |
| 22:19.08 | dcordes | cr2, did wince kernel reveal anything about kaiser's shared mem registers? |
| 22:19.23 | cr2 | dcordes: no, it's too difficult. |
| 22:19.57 | dcordes | makes me wonder how we will find them out |
| 22:20.00 | cr2 | dcordes: but i've added the BT gpio settings, and fixed (my guess) the DMA registers which look a bit different from 7200A |
| 22:20.26 | exco | tsdogs: do I need to load any module to get the triggers working |
| 22:20.38 | cr2 | dcordes: ask dzo. i guess he was seeking through the RAM and looking for the "cached" AT commands |
| 22:20.54 | dcordes | cr2, I'm looking for the fifo registers in shared mem right? |
| 22:21.08 | cr2 | dcordes: btw, you may dump the 0x1f* area and check for them. |
| 22:21.34 | tsdogs | exco: nope, though they are connected to the battery module so maybe they are not working as expected. but they need to be initialized the first time u boot to respect the real status afair |
| 22:21.43 | dcordes | cr2, didn't I do that yet (the thing where no change became obvious) |
| 22:23.33 | tsdogs | exco: it's pretty strange u don't have a ds2760-battery.0 anyway, but pH5 knows better. |
| 22:23.41 | cr2 | dcordes: pwf smem 0x01f00000 0x100000 |
| 22:24.21 | cr2 | tsdogs: here |
| 22:24.24 | cr2 | 0x00000000004393be in tData::loadncc2A (this=0x6a42d0, parent=<value optimized out>, name=<value optimized out>) at tdataitem.h:52 |
| 22:24.24 | cr2 | 52 void setRef (int x=-1, int y=-1, int z=-1) { varX = x; varY = y; varZ = z; } |
| 22:24.53 | tsdogs | cr2: ? |
| 22:25.29 | tsdogs | it's failing to load the data? |
| 22:25.35 | exco | well he wondered about that too :-) |
| 22:26.26 | cr2 | tsdogs: i'll take a small .sel file |
| 22:26.59 | cr2 | hm. now it hangs in read() |
| 22:27.09 | dcordes | cr2, can I email you the output file? |
| 22:27.21 | tsdogs | cr2: I almost havent touched the tdata and tdataitem |
| 22:28.50 | tsdogs | cr2: it has really strange behavier some times... it would loop on reading. I think that some checking should be needed, there are plenty of warnings about result codes not beeing handled |
| 22:29.09 | cr2 | tsdogs: i' ve created a test.xyz file |
| 22:29.11 | cr2 | 1 2 3 |
| 22:29.13 | cr2 | 4 5 6 |
| 22:29.27 | cr2 | it hangs at the same place. i'll run it in gdb |
| 22:30.00 | cr2 | heh |
| 22:30.02 | cr2 | #3 0x00002b5c2e6489c9 in fread () from /lib64/libc.so.6 |
| 22:30.03 | cr2 | #4 0x0000000000410407 in drawWidget::drawCoastline (this=0x834300) at drawwidget.cc:188 |
| 22:30.14 | cr2 | it didn't find the coastline data probably |
| 22:30.14 | tsdogs | cr2: did u remove the qWarning ? |
| 22:30.19 | cr2 | no |
| 22:30.53 | tsdogs | I had the same and adding a qWarning in that function works, seems some strange timing ... |
| 22:31.36 | tsdogs | cr2: ok it does that to me too... |
| 22:31.40 | cr2 | opens |
| 22:31.53 | cr2 | open("/tmp/viewer-2.3.4/data/continents.dat", O_RDONLY) = 10 |
| 22:32.05 | cr2 | i'll try on 32bit |
| 22:32.18 | cr2 | but it was working on alpha before. |
| 22:32.22 | tsdogs | cr2: I'm developing on x86_64 |
| 22:33.04 | tsdogs | though I have the same with qt3.3 version |
| 22:33.07 | cr2 | 4.3.0 on p4 |
| 22:33.34 | cr2 | hm. netcdf not installed there. |
| 22:34.05 | cr2 | btw, i'm seriously thinking about using netcdf for .rdm |
| 22:34.35 | tsdogs | cr2: it's a timing issue defenately |
| 22:35.06 | tsdogs | just put a qWarning()<<"whatever" at line 183 of drawwidget.cc |
| 22:35.07 | cr2 | compiling netcdf |
| 22:35.07 | exco | how do you sync dirs? (I miss totalcommander in linux) |
| 22:35.20 | tsdogs | exco: rsync |
| 22:37.18 | exco | I just saw video playback inside the alt-tab window for the first time... how cool |
| 22:40.25 | tsdogs | fyi: http://www.nature.com/nature/journal/v453/n7191/abs/nature06932.html;jsessionid=02B6089B784A1B85707AF6D90470BA49 |
| 22:43.16 | dcordes | cr2, http://rapidshare.com/files/112586612/smem.html |
| 22:43.21 | dcordes | exco, what distro :) ? |
| 22:47.26 | cr2 | dcordes: you are 31337 |
| 22:50.02 | cr2 | dcordes: +0x41f0 for AT, what does dzo have ? |
| 22:51.17 | cr2 | ends at +0x6208 |
| 22:51.22 | exco | on my laptop I'm running Ubuntu 8.04... but I'm not shure if I'm keeping it |
| 22:51.40 | cr2 | dcordes: looks like 0x2000 buffer for me + some header |
| 22:51.50 | exco | (you gotta start somewhere) |
| 22:52.41 | cr2 | from 0x4208 to 0x6208 |
| 22:52.51 | paulproteus | exco, Ubuntu's reasonably good. I'm running it on this here desktop. |
| 22:52.54 | dcordes | cr2, where would I look that up what dzo has? in http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=arch/arm/mach-msm/dma.c;h=bb28fdcb2c547751331a5b652d05336dc41f140b;hb=refs/heads/htc-msm ? |
| 22:53.18 | cr2 | smd-vogue |
| 22:54.02 | cr2 | tsdogs: works on p4 |
| 22:54.55 | exco | the CPU runs too hot for my taste (firefox crashes too often with flash, or ajax on my google startpage), hibernate works 2out of 3 times ... |
| 22:54.59 | tsdogs | hmm, it must be something strange on freads |
| 22:55.27 | exco | phc patching wasn't too hard though |
| 22:55.47 | cr2 | tsdogs: autoscale has a tooltip "plot zonal mean values" and does not work. |
| 22:56.04 | dcordes | cr2, http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=blob;f=arch/arm/mach-msm/kaiser-smd.c;h=462d9d5f5c41244f1c3163ce43716a0910142aba;hb=refs/heads/htc-msm (that's just a copy of the vogue file) |
| 22:56.09 | dcordes | forgot to push the vogue file |
| 22:56.56 | cr2 | dcordes: you have 0x2000, he has 0x200 |
| 22:57.01 | tsdogs | cr2: yes I commented out the code there, I need to find a way to replace that code with something else. |
| 22:57.19 | cr2 | ok, but you have HSDPA |
| 22:57.29 | cr2 | tsdogs: ok. |
| 22:58.09 | tsdogs | lol, long segmengs; fread(&segments, 4, 1, stream); |
| 22:59.19 | cr2 | tsdogs: yeah. but how did it work on alpha... |
| 22:59.30 | tsdogs | cr2: no idea :) |
| 22:59.46 | cr2 | i've even run it on a 16cpu itanium-2 :) |
| 23:00.26 | cr2 | tsdogs: it's a very buggy, but small and fast viewer. |
| 23:00.52 | tsdogs | :) I really still don't understand what it is used for |
| 23:01.07 | dcordes | cr2, pushed the missing vogue-smd.c |
| 23:01.36 | dcordes | I hope the whole tree does still work with vogue to any extent |
| 23:01.56 | cr2 | tsdogs: usually for 2D lat/long profiles, and their altitude dependence. |
| 23:02.40 | cr2 | tsdogs: you could have clicked a point and see the altitude profile. it was all done for some profile quality assurance. |
| 23:03.40 | cr2 | instead of some monstrosity that used IDL, and where you never have enough licenses ;) in 2000 it was all alpha stuff. |
| 23:03.59 | tsdogs | :) ok |
| 23:04.30 | dcordes | cr2, I changed 0x200 to 0x2000 in the code |
| 23:04.47 | cr2 | tsdogs: on the 300MB file |
| 23:04.50 | cr2 | terminate called after throwing an instance of 'std::bad_alloc' |
| 23:04.50 | cr2 | <PROTECTED> |
| 23:05.06 | cr2 | dcordes: it's not enough i guess. |
| 23:05.07 | dcordes | cr2, will this make anything work? |
| 23:05.12 | tsdogs | cr2: btw changing long to int works on x86_64 |
| 23:05.45 | cr2 | tsdogs: shoud use size_t ? |
| 23:05.59 | cr2 | dcordes: look at ch->send.start=(char *)(MSM_SHARED_RAM_BASE+0xf0200); |
| 23:06.58 | tsdogs | cr2: I'm not sure as the data is read from the file, maybe initializing it to 0 is enough... (but on alpha, hmm) |
| 23:07.40 | cr2 | dcordes: there is nothing at f0200 |
| 23:08.04 | cr2 | dcordes: and you need to check the channel numbers |
| 23:08.13 | tsdogs | cr2: initializing the value to 0 seems to be working, not sure how it would work on alpha... |
| 23:08.13 | dcordes | is the MSM_SHARED_RAM_BASE value set correctly for kaiser? |
| 23:09.36 | cr2 | dcordes: yes, it's the 0x1f0000 |
| 23:09.38 | dcordes | cr2, 0xf0200 should be 0xf02000 right? |
| 23:10.04 | cr2 | dcordes: hmm. |
| 23:10.22 | dcordes | :( |
| 23:10.43 | cr2 | dcordes: check the offset where you have some nonzero data |
| 23:10.55 | cr2 | 404c0 for example. |
| 23:11.47 | cr2 | btw, 4208 it's the send fifo ? |
| 23:12.22 | tsdogs | time for bed, good night |
| 23:12.32 | dcordes | sorry where shall I look for the offset? |
| 23:12.36 | cr2 | tsdogs: good night |
| 23:12.42 | dcordes | night tsdogs |
| 23:12.54 | cr2 | dcordes: the next nonzero is 0x8220 |
| 23:13.23 | cr2 | probably to 0xa220 |
| 23:14.10 | dcordes | can't follow at all |
| 23:14.10 | cr2 | yes, then you have rx buffer |
| 23:14.25 | cr2 | 0x32430 |
| 23:15.00 | miknix | Kevin2, ping? |
| 23:15.02 | cr2 | goes to 0x34430 |
| 23:15.11 | dcordes | *_* |
| 23:15.16 | cr2 | dcordes: so it's a 0x2000 recv fifo |
| 23:15.31 | cr2 | circular buffer probably, like on the DPRAM |
| 23:16.00 | cr2 | interesting. ba had 0x200 (256byte) bufs. |
| 23:16.27 | cr2 | byte=2byte :) |
| 23:16.38 | cr2 | don't remember about athena. |
| 23:17.01 | cr2 | dcordes: time to edit kaisermemorymap it seems. |
| 23:17.41 | cr2 | yes, athena also has 0x200 |
| 23:17.43 | cr2 | http://wiki.xda-developers.com/index.php?pagename=AthenaDPRAM |
| 23:18.02 | dcordes | I didn't understand one byte |
| 23:28.21 | cr2 | dcordes: http://wiki.xda-developers.com/index.php?pagename=KaiserMemoryMap |
| 23:28.38 | dcordes | I'm just there |
| 23:28.46 | dcordes | looking at your diff |
| 23:29.27 | cr2 | compare to http://wiki.xda-developers.com/index.php?pagename=TitanMemoryMap |
| 23:29.30 | BabelO | cri just finish my whole map :) |
| 23:29.32 | cr2 | looks very different. |
| 23:29.37 | BabelO | cr2 so happy :) |
| 23:29.45 | cr2 | BabelO: lol :) |
| 23:30.39 | BabelO | cr2: i pm you the link, with qmap file, now i want to ask kiozen how he manage it when uploaded to the phone |
| 23:30.55 | cr2 | BabelO: \xb0 in the strings. utf8 bug for kiozen :) |
| 23:31.09 | cr2 | BabelO: i've put these files in /home/root |
| 23:31.18 | cr2 | and then just loaded. |
| 23:31.42 | BabelO | cr2: yes me too, but there is some data lost, as i told you one hour ago ;) |
| 23:31.56 | dcordes | cr2, looks like titans have some work left? |
| 23:31.56 | BabelO | cr2: some maps override part of other |
| 23:32.02 | cr2 | BabelO: hmm. should i download it ? |
| 23:32.25 | BabelO | crif you can, yes just to look at the quality, and tell me how to manage it |
| 23:32.31 | cr2 | BabelO: now you need to make gps work on artemis :) |
| 23:33.08 | BabelO | cr2: i think it work, bluetooth, and mmc too, but i still have this issues with irqs |
| 23:33.26 | BabelO | cr2: it freeze my device only when i m outiste boot process |
| 23:33.36 | BabelO | cr2: when i call from ssh session |
| 23:33.37 | BabelO | :( |
| 23:33.53 | BabelO | i try like Marex told me yestuday |
| 23:34.04 | BabelO | to add spinlock, but without success |
| 23:34.45 | cr2 | BabelO: strange. |
| 23:34.55 | Marex | was the spinlock correct idea_ |
| 23:34.58 | Marex | ? |
| 23:35.28 | cr2 | Marex: the correct idea is to make sound on n560 work :) |
| 23:35.31 | BabelO | Marex: hi |
| 23:35.31 | dcordes | cr2, I'm trying to stuff this into the code |
| 23:35.37 | BabelO | Marex: no |
| 23:36.14 | Marex | cr2, you damned blackmailer ;D |
| 23:36.28 | BabelO | Marex: i m trying to recheck... but did not have more idea about that |
| 23:36.29 | cr2 | hehe. |
| 23:36.31 | Marex | BabelO, ah yea ... dma locks, sorry |
| 23:36.39 | Kevin2 | miknix: pong |
| 23:37.01 | miknix | hi Kevin2 |
| 23:37.11 | Marex | cr2, Im going to try palmld usb2 chip ;) |
| 23:37.18 | BabelO | Marex: maybe if i found how to run mmc without dma, i see there is pio mode, but it result with kernel panic |
| 23:37.21 | Marex | cr2, it's probably mmaped and wont be that hard |
| 23:37.26 | Marex | cypress sx2 |
| 23:37.32 | dcordes | cr2, what about size and start? |
| 23:37.43 | cr2 | Marex: n560 is the device of kiozen, i just want to finish it. |
| 23:37.56 | Marex | cr2, palmld is the device of me ;D |
| 23:38.00 | cr2 | dcordes: we need to talk to dzo |
| 23:38.17 | Marex | cr2, who do you think I care more about ? :-) |
| 23:38.28 | miknix | Kevin2, I own a htc wizard. I'm interested in helping to make changes on the kernel tree to get it running on the wizard |
| 23:38.30 | Marex | ... well actually I dont care, but ... |
| 23:38.46 | miknix | Kevin2, I'm always here and mainly work with cr2 and BabelO |
| 23:38.47 | Kevin2 | miknix: cool. |
| 23:39.00 | dcordes | cr2, ah wait isn't size the value you mentioned before? 0x2000? |
| 23:39.05 | miknix | Kevin2, I'm asking if its possible to have push permissions to the git repo |
| 23:39.14 | cr2 | dcordes: the buffer size |
| 23:39.45 | miknix | Kevin2, I'm in the linwizard team, we have our own git repo. But since the nature of both projects is the same. I would like to keep the two repos in "sync" |
| 23:40.05 | dcordes | cr2, ch->send/recv.size=0x2000? |
| 23:40.15 | cr2 | dcordes: i think so |
| 23:40.20 | Kevin2 | miknix: Did you put in a request to "join" the group via the linuxtogo web site? |
| 23:40.28 | miknix | Kevin2, no |
| 23:40.38 | cr2 | Kevin2: the difference between them are htc-egpio vs. htc-i2c-egpio |
| 23:40.38 | dcordes | ~ping dzo |
| 23:40.44 | apt | pong dzo |
| 23:40.44 | miknix | Kevin2, I'm doing it then :) |
| 23:41.30 | Kevin2 | cr2: huh? |
| 23:41.51 | cr2 | dcordes: SMD1 send/recv, it's for the data ? |
| 23:42.23 | dcordes | shrugs |
| 23:42.27 | cr2 | Kevin2: the wizard et al. use a htc gpio extender controller over i2c, not mmaped |
| 23:42.33 | dcordes | cr2, would have been my next question |
| 23:43.09 | cr2 | dcordes: i think so, because these are the only 2 things that work on vogue afaik |
| 23:44.17 | dcordes | cr2, at and data traffic? |
| 23:44.26 | cr2 | dcordes: btw, the AT send/recv buffer pointers differ only by 4 bytes, so it's some struct. |
| 23:44.29 | miknix | Kevin2, I'm registered |
| 23:45.29 | cr2 | dcordes: weird sizes for CDMA data btw. |
| 23:45.41 | dcordes | cr2, why? |
| 23:45.45 | BabelO | crthe map is hard to move with zoom x4 |
| 23:46.21 | BabelO | miknix: better is to start with 2.6.25 from the git ;) |
| 23:47.25 | cr2 | BabelO: you probably need to add tiff overviews for deep zooming. |
| 23:47.46 | cr2 | BabelO: the idea of kiozen was to use different maps for different zoom levels. |
| 23:48.12 | cr2 | BabelO: which makes senso for non-sat imagery. |
| 23:48.19 | BabelO | yes |
| 23:48.58 | cr2 | dcordes: you have Ba6 ? |
| 23:49.09 | miknix | BabelO, the repo is the one in mobile linux project? |
| 23:49.34 | dcordes | cr2, yep |
| 23:49.46 | cr2 | dcordes: prepare to destroy it soon :) |
| 23:50.09 | BabelO | miknix: yzq onz branch |
| 23:50.16 | dcordes | cr2, are you alluding to bootloader development? |
| 23:50.55 | dcordes | or bitbanging driver |
| 23:51.00 | cr2 | dcordes: no, i've jjust written the LCD spi init driver for n560. it should be easy to adapt it for ba6. i've already looked at it. |
| 23:51.09 | cr2 | bitbanging driver |
| 23:51.18 | dcordes | yea ok let me know if there are testing needs |
| 23:51.31 | cr2 | ok. |
| 23:51.35 | miknix | is cloning |
| 23:51.39 | dcordes | I'd like to make ba more usable anyway for propaganda use |
| 23:51.52 | miknix | BabelO, yzq onz branch? wtf! |
| 23:52.19 | cr2 | dcordes: ba is a nice phone because it has 128MB ram, wifi and 2D accel support. |
| 23:52.59 | dcordes | and you can use it verey well as a close combat weapon |
| 23:53.49 | dcordes | cr2, do you know if there are klinke adapters for ba headphone jack? |
| 23:54.16 | cr2 | maybe, but i've never seen them in real life. |
| 23:55.33 | dcordes | a quick search reveals ebay has td028steb1 for ba |
| 23:55.54 | dcordes | apache tytn ba have the same display? |
| 23:55.58 | BabelO | miknix: my cat was speaking, sorry |
| 23:56.11 | BabelO | miknix: yes it is a snapshot branch in mobile linux repo |
| 23:56.37 | miknix | BabelO, lol, the branch name is quite unusual |
| 23:56.51 | BabelO | cr2: it is impressive, gt is on lot of mirror website |
| 23:59.51 | cr2 | BabelO: which one ? |