| 00:32.36 | *** join/#gpe jg (~jg@h005018015b26.ne.client2.attbi.com) |
| 03:23.24 | *** join/#gpe chrisime (~chrisime@pD9E8CDC7.dip0.t-ipconnect.de) |
| 03:42.27 | *** join/#gpe glm2k (~greylensm@c-66-229-176-251.we.client2.attbi.com) |
| 05:52.33 | *** join/#gpe AM (~camrein@AMontsouris-108-1-19-57.w80-15.abo.wanadoo.fr) |
| 06:36.36 | *** part/#gpe greylensman (~greylensm@c-66-229-176-251.we.client2.attbi.com) |
| 07:06.48 | *** join/#gpe dc_ (~dctanner@r122226.ap.plala.or.jp) |
| 07:07.00 | dc_ | afternoon |
| 07:25.29 | *** join/#gpe chrisime_ (~chrisime@pD9E8CEE4.dip0.t-ipconnect.de) |
| 07:36.07 | *** join/#gpe florian (boor@141.99.42.43) |
| 07:36.12 | florian | morning |
| 07:36.47 | *** join/#gpe ChipX86 (chipx86@chipx86.radio.freenode) |
| 07:38.48 | *** join/#gpe dc___ (~dctanner@r122226.ap.plala.or.jp) |
| 08:00.09 | florian | bbl |
| 08:06.57 | *** join/#gpe AM (~camrein@AMontsouris-108-1-12-66.w217-128.abo.wanadoo.fr) |
| 08:09.35 | *** join/#gpe PaxAnima (~petter@b097c.studby.ntnu.no) |
| 08:15.41 | *** join/#gpe mallum (~mallum@82-35-112-58.cable.ubr07.dals.blueyonder.co.uk) |
| 08:21.12 | *** part/#gpe AM (~camrein@AMontsouris-108-1-12-66.w217-128.abo.wanadoo.fr) |
| 08:22.44 | *** join/#gpe nilsf (~nils@kc.celox-online.de) |
| 08:35.49 | mallum | nilsf: hey |
| 09:05.50 | nilsf | mallum: G'Morning :) |
| 09:07.25 | mallum | nilsf: I have a kernel question for you :-) |
| 09:08.13 | nilsf | mallum: Sure, go ahaead :) |
| 09:08.42 | mallum | nilsf: on the jornada, it speaks to things like the ts, battery etc by means of this mcu thing |
| 09:08.57 | mallum | nilsf: which is like a serial device you push and pop bytes from |
| 09:09.11 | nilsf | mallum: mcu = microcontroller ? |
| 09:09.19 | mallum | nilsf: yeah I think so |
| 09:09.30 | nilsf | mallum: OK. The problem? |
| 09:09.40 | mallum | nilsf: see http://lorien.handhelds.org/ftp.handhelds.org/linux/hp/720/jornada720.txt |
| 09:10.48 | mallum | nilsf: in the funcs which speak to the mcu there is no current support for atomic operations. So if you 'rub; the touchscreen and get info from the battery or backlight at the same time wierd things happen |
| 09:10.52 | nilsf | mallum: Cool! Documentation to hardware! Havn't seen something like this for a while :) |
| 09:11.21 | nilsf | mallum: Oh, I see... |
| 09:11.36 | mallum | nilsf: Im guessing the kernel has multiple threads right ( reall nwebie question here ) and there just interfering with each other |
| 09:12.11 | mallum | nilsf: Im thinking wrapping these mcu funcs with spinlock__irq(spinlock) things could fix it ? |
| 09:12.34 | nilsf | mallum: Hmm... I don't think so. |
| 09:12.50 | nilsf | mallum: The problem seems to be more the MCU unit not the kernel directly. |
| 09:13.18 | nilsf | mallum: I would think that if you start two commands on the MCU simultaniously then one is dropped. |
| 09:14.05 | mallum | nilsf: hmmmm |
| 09:14.27 | nilsf | mallum: Better would be to serialize the requests in kernel. |
| 09:15.04 | mallum | nilsf: Is that like what the 36xx code does with the mcio ? |
| 09:15.17 | nilsf | mallum: How is this interface constructed right now? Is there _one_ kernel driver for the MCU or is it accessed wherever needed? |
| 09:16.01 | mallum | nilsf: seems to be accessed where ever with some some macros and simple funcs |
| 09:17.00 | nilsf | mallum: I would strongly guess that the system on the h36xx is similar, yes. Tose MCUs are simple devices with not much software in them. So you cannot expect much smartness from them. |
| 09:17.31 | nilsf | mallum: Especially RAM is almost none on those MCUs which is why they cannot serialize requests themselves, I guess... |
| 09:17.46 | mallum | http://cvs.handhelds.org/cgi-bin/viewcvs.cgi/linux/kernel/arch/arm/mach-sa1100/jornada720.c?rev=1.20&content-type=text/vnd.viewcvs-markup |
| 09:17.54 | mallum | see the mcu_start etc funcs |
| 09:18.52 | mallum | jornada720_battery() is an example of how it accesses info from it |
| 09:20.10 | nilsf | Hmm... adding locks would be the straight forward idea, yes. But which functions do call those? I fear that you might cause a freeze or stall... |
| 09:21.06 | mallum | at the moment nothing in hh src, I have a version Im hacking on with binds stuff like ts control to an ioctl on /dev/ts ( just like ipaq ) |
| 09:21.08 | *** join/#gpe pb_ (~pb@dsl-62-3-66-201.zen.co.uk) |
| 09:23.17 | nilsf | Oh well ... this whole stuff should be more unified... I hacked the SIMpad patch into the hh.org kernel and there again you have different methods for things on the SIMpad than on the 720 than on the iPAQ :( |
| 09:24.14 | mallum | right indeed |
| 09:24.52 | mallum | I dont think with the ipaq way you can query what the display can have adjusted - eg contrast, brighness etc |
| 09:26.03 | mallum | nilsf: so do you think spinlocks maybe the easiest way to fix problems ? |
| 09:27.25 | nilsf | mallum: Huhu... difficult. You could try. It's the quickest possible hack, yes. But to say if it really works or is the right way I would have to read the driver more thoroughly. |
| 09:28.02 | nilsf | I am not convinced that /dev/ts is the right place to query brightness or contrast settings... |
| 09:28.29 | mallum | nilsf: okey, one more quick que ... does the spinlock instance passed to the spinlock code have to be global ? or can it be local to the calling function |
| 09:28.53 | mallum | nilsf: and did you hear anything from ATI ? |
| 09:29.05 | pb_ | nilsf: no, indeed. The H3600 just stuck all that stuff in /dev/h3600_ts because it was a pre-existing device node. A better design would be to have a dedicated /dev entry for those functions. |
| 09:29.05 | nilsf | Some /proc interface like /proc/hal/ seems a little more adequate to me. |
| 09:29.58 | nilsf | mallum: No news from ATI but also no bad news at least :) But I got a negative response from Pretec concerning their very nice 1.3mpixel CF camera :( |
| 09:30.11 | mallum | nilsf: :( |
| 09:30.27 | nilsf | mallum: The lock structs are in general global to the module, yes. |
| 09:30.37 | mallum | nilsf: it would be cool if there was something like soft ice for pocketpc ... |
| 09:31.04 | nilsf | mallum: He? Ah! You mean JTAG hardware debugger? |
| 09:31.07 | mallum | nilsf: just to the module ? or global to anything which may need to be locked out from accessing ? |
| 09:31.32 | mallum | nilsf: With softive for windows, you can spy on anything so very easy to reverse engineer stuff |
| 09:32.05 | mallum | nilsf: thats how this guy figured out the TS protocol on my fujitsu point510 |
| 09:32.07 | nilsf | mallum: Cool :) Would be nice, yes! I even already have the PPC SDK for it... |
| 09:32.43 | nilsf | mallum: With some work one should be able to disassemble the DLLs... |
| 09:33.01 | mallum | hmm, if I google for pocketpc softice I get a load of warez sites :/ |
| 09:33.53 | nilsf | mallum: Concerning the locks, the critical functions are mcu_start, mcu_byte and mcu_end. Simply put a lock in start and unlock in end and you should be almost fine (I think). |
| 09:34.17 | mallum | nilsf: ah cool |
| 09:35.23 | nilsf | mallum: Worst that can happen then is that your kernel freezes ;) |
| 09:35.35 | mallum | nilsf: but I dont wanna spinlock a whole func like j720_battery() that makes a seres of mcu calls ? |
| 09:36.57 | nilsf | mallum: No no, just inside the critical funcs! |
| 09:38.28 | nilsf | mallum: You want to prevent that mcu_start can be entered more than one time. So put a lock inside the mcu_start and unlock in mcu_end. |
| 09:38.45 | *** join/#gpe jg (~jg@h005018015b26.ne.client2.attbi.com) |
| 09:38.49 | mallum | nilsf: right |
| 09:39.25 | mallum | nilsf: I've learn alot from playing with the j720 kernel source :-) I'll fix the 54xx touchscreen later today ;-) |
| 09:40.46 | nilsf | mallum: Perfect! |
| 09:41.33 | mallum | pb_: do you have this tetris problem with newer mb ? |
| 09:41.43 | pb_ | mallum: dunno, I haven't tried lately. |
| 09:42.02 | pb_ | I don't have any working ipaqs here at the moment so I can't test it right away. |
| 09:42.17 | mallum | pb_: last time I tested with a 0.6 mb it was fine |
| 09:42.48 | pb_ | ok |
| 09:44.33 | mallum | hmm, I should probably uplaod some new mb packages |
| 09:45.32 | mallum | pb_: btw I made some cairo packages last night which are in unstable |
| 09:45.52 | pb_ | yep, so I saw. that's cool. |
| 09:46.05 | mallum | pb_: and there are arm test binaries at hh.org/~pb/cairo |
| 09:46.06 | pb_ | we should get Carl to package his Ricochet Robots game as well :-) |
| 09:46.14 | pb_ | there are?! |
| 09:46.22 | mallum | ~mallum I mean |
| 09:46.24 | mallum | hehe |
| 09:46.30 | pb_ | aha. |
| 09:46.53 | mallum | yeah I'll try and do the rr game later today |
| 09:47.15 | mallum | I have wedding tasks to do later though .. bluewater visit for bridemaids dresses |
| 09:47.59 | pb_ | ah right. enjoy :-) |
| 09:49.12 | mallum | I get to visit the lego shop though which is a bonus |
| 09:49.28 | pb_ | heh |
| 09:54.53 | *** join/#gpe florian (~fuchs@kc.celox-online.de) |
| 09:55.15 | florian | re |
| 09:55.25 | pb_ | florian: re |
| 09:56.04 | mallum | hey florian |
| 10:05.53 | pb_ | florian: I'm now at a machine with a working web browser, so I added you to geeklog's GPE group. |
| 10:06.38 | florian | pb_: ah, cool - thanks |
| 10:08.46 | pb_ | cool! |
| 10:09.00 | pb_ | oh, btw, I have a Yopy here on loan for a while. |
| 10:11.21 | mallum | pb_: whats it like ? and how did you get it ?? |
| 10:12.40 | nilsf | pb_: And which Yopy? 3000, 3500 or 3700? |
| 10:15.49 | pb_ | It's a 3700. No idea what it's like yet, I haven't even taken it out of the box yet. |
| 10:19.39 | nilsf | pb_: Poor boy :) No time for your toys? |
| 10:20.34 | pb_ | nilsf: Something like that, yeah. Even my SL-C750 remains untouched. |
| 10:21.00 | florian | nilsf: Eh... i could remind you to some things to do ;-) |
| 10:27.44 | mallum | I uplaoded the first round of split mb packages to unstable |
| 10:27.49 | mallum | expect everything to break ;-) |
| 10:28.20 | florian | ah cool... |
| 10:52.01 | *** join/#gpe spung (~rewt@a213-22-98-205.netcabo.pt) |
| 10:54.19 | mallum | hmm, somewhere along the lines ipkg thinks 0.6.2.xxx matchbox is newer than 0.6.3 :( |
| 11:02.45 | florian | :-/ |
| 11:03.39 | mallum | florian: best thing to is ipkg remove matchbox, then ipkg install libmb1 matchbox-common matchbox-wm matchbox-desktop matchbox-panel |
| 11:03.54 | florian | ok |
| 11:03.57 | mallum | florian: the matchbox-wm has some extra files it dont need Im just gonna fix that |
| 11:11.26 | nilsf | pb_: We just got the 750 for one week two days ago... nice thing! |
| 11:11.51 | mallum | nilsf: OZ now has support for th c7xx's apparently |
| 11:12.02 | mallum | nilsf: if you want I can build you an X based image |
| 11:12.24 | mallum | nilsf: apparently the newer c7xx kernel also work on the c700's so I may build a new image for my self |
| 11:12.36 | pb_ | nilsf: Yep, certainly is! |
| 11:15.15 | nilsf | mallum: Thanks for the offer but we have the device only for one week (have to send it back on monday) and it has to stay in a resellable state. |
| 11:15.35 | mallum | nilsf: okey |
| 11:20.27 | nilsf | I must have missed something but what is ibot talking about with this RO.*T thing? |
| 11:21.13 | florian | nilsf: looks like he loves spung :-) |
| 11:23.46 | pb_ | nilsf: ibot says that whenever somebody whose username is "root" (or a close variant) joins the channel. |
| 11:24.42 | florian | yeiks! millions of opie screenshots! |
| 11:25.14 | pb_ | I think the Opie dudes are gearing up for their 1.0 release. |
| 11:26.03 | florian | :-/ |
| 11:26.18 | florian | We need more coders...# |
| 11:26.23 | pb_ | That's why they were asking for a Familiar 0.7.1 release yesterday. |
| 11:26.32 | pb_ | florian: always. |
| 11:31.35 | mallum | porting GPE to Z would probably get more coders |
| 11:32.18 | florian | Yep, this is maybe a good idea... |
| 11:32.49 | florian | We'll talk about this later... brb, lunch! |
| 11:32.55 | nilsf | pb_: Ah, makes sense, yes. tnx |
| 11:33.05 | florian | s/lunch/breakfast |
| 11:33.11 | pb_ | heh |
| 11:33.18 | florian | nilsf: food?! |
| 11:33.31 | nilsf | florian: Already? |
| 11:33.44 | florian | nilsf: right NOW! |
| 11:34.04 | pigeon | gar... i need more food too... |
| 11:37.52 | mallum | mmm, food |
| 11:58.26 | *** join/#gpe Vinny`work` (xdcc@cbl-216.191.25.115.galaxy-data.com) |
| 12:18.13 | florian | re |
| 12:42.34 | *** join/#gpe jsmanrique (jsmanrique@cmr-213-141-46-198.telecable.es) |
| 12:42.37 | jsmanrique | hi! |
| 12:42.42 | pb_ | hi jsmanrique |
| 12:42.50 | jsmanrique | someone has fill hh.org/scap site.. |
| 12:43.03 | jsmanrique | how many opie scaps!!! |
| 12:45.03 | pb_ | heh, indeed. |
| 12:45.37 | jsmanrique | drawpad looks good, but it is not what I am looking for.. |
| 12:45.49 | jsmanrique | a replacement for notes on wince |
| 12:54.06 | mallum | jsmanrique: the matchbox with the spanish locale is in unstable btw |
| 13:01.29 | *** join/#gpe mallum_ (~mallum@82-35-112-58.cable.ubr07.dals.blueyonder.co.uk) |
| 13:07.25 | jsmanrique | mallum: thanks, I'll go to unstalbe as soon as posilbe |
| 13:08.42 | mallum_ | jsmanrique: okey |
| 13:17.33 | *** join/#gpe Vinny`work` (xdcc@cbl-216.191.25.115.galaxy-data.com) |
| 13:42.20 | *** join/#gpe chris (~chris@grasmoor.elec.aca.mmu.ac.uk) |
| 13:46.14 | *** join/#gpe oldham_1 (~chris@grasmoor.elec.aca.mmu.ac.uk) |
| 14:20.11 | *** join/#gpe TheMasterMind1 (~aman@h-69-3-84-221.MCLNVA23.covad.net) |
| 14:33.23 | *** join/#gpe _fishy (~fishy@9027386f6c64.ssl.walledcity.de) |
| 15:09.29 | *** join/#gpe looker_ (~looker@223-200-ADSL.red.retevision.es) |
| 15:09.31 | looker_ | hi! |
| 15:11.07 | looker_ | is anybody here¿? |
| 15:15.37 | looker_ | well, I want to buy a pda to install gpe in, but I dont know wich one is a good buy |
| 15:16.49 | pb_ | iPAQ 3000 series is currently the best supported. |
| 15:23.15 | looker_ | ok |
| 15:23.17 | looker_ | thanks |
| 15:46.22 | looker_ | what about the 3900 series¿ |
| 15:47.38 | pb_ | works except for MMC. |
| 15:57.17 | looker_ | MMC? |
| 15:58.59 | looker_ | WTF is mmc? |
| 16:00.49 | pb_ | ibot, mmc? |
| 16:00.50 | | extra, extra, read all about it, mmc is MultiMediaCard mentioned at http://www.handhelds.org/z/wiki/IpaqHardwareCompatibilityStorage or at http://www.zauruszone.farplanet.net/wiki/index.php?MMC%20or%20SD%20card, or http://hacking.the.cia.com/~knight/ZaurusIndex.php |
| 17:37.32 | florian | bbl |
| 18:31.15 | *** join/#gpe jason (~jason@bgp527150bgs.ebrnsw01.nj.comcast.net) |
| 18:34.47 | *** part/#gpe jason (~jason@bgp527150bgs.ebrnsw01.nj.comcast.net) |
| 18:46.34 | *** join/#gpe pb_ (~pb@pc2-cmbg4-3-cust239.cmbg.cable.ntl.com) |
| 18:46.54 | *** join/#gpe ade|desk (~adrian@modem-1056.gazelle.dialup.pol.co.uk) |
| 18:50.45 | *** join/#gpe florian (~fuchs@pD9E5FF29.dip.t-dialin.net) |
| 18:50.48 | florian | re |
| 18:50.58 | PaxAnima | hi, florian |
| 18:51.07 | nilsf | florian: Welcome home! |
| 18:51.23 | *** join/#gpe ade|home (~adrian@modem-278.jaguar.dialup.pol.co.uk) |
| 18:52.17 | *** join/#gpe BBrox (~ulmer@bbrox.org) |
| 18:52.22 | florian | yep... lets do something good... |
| 18:53.08 | nilsf | florian: Exactly, also go home... darn! Still one hour to go 8) |
| 18:56.28 | *** join/#gpe jg (~jg@dhcp-r23.crl.dec.com) |
| 19:20.45 | *** join/#gpe ade|desk (~adrian@modem-3930.jaguar.dialup.pol.co.uk) |
| 20:46.46 | *** join/#gpe looker (~looker@223-200-ADSL.red.retevision.es) |
| 20:46.51 | looker | hi again" |
| 20:49.16 | looker | can you install amsn in gpe? it uses tcl/tk |
| 20:51.09 | pb_ | can't think of any reason why it wouldn't work. |
| 20:51.13 | pb_ | tcl/tk is in debian. |
| 20:52.07 | looker | ? |
| 20:52.30 | looker | but tcl/tk comes with familiar and gpe |
| 20:52.33 | looker | ? |
| 20:54.56 | florian | looker: it's definitely not part of the bootgpe image, but i think there is at least one in unstable feed |
| 20:55.36 | looker | ok |
| 20:55.41 | looker | thanks |
| 21:19.17 | *** join/#gpe ade|home (~adrian@modem-311.gazelle.dialup.pol.co.uk) |
| 21:37.26 | *** join/#gpe spung (~rewt@213.22.98.205) [NETSPLIT VICTIM] |
| 21:37.35 | *** join/#gpe TheMasterMind1 (~aman@h-69-3-84-221.MCLNVA23.covad.net) |
| 21:38.18 | *** join/#gpe stargirl (hmw26@student.cusu.cam.ac.uk) [NETSPLIT VICTIM] |
| 21:38.24 | *** join/#gpe ade|desk (~adrian@modem-311.gazelle.dialup.pol.co.uk) |
| 22:17.17 | florian | night |
| 22:17.29 | pb_ | florian: night |
| 22:34.28 | *** join/#gpe toto (~camrein@AMontsouris-108-1-33-194.w81-53.abo.wanadoo.fr) |
| 23:34.52 | pigeon | Hello everyone. |
| 23:35.27 | pb_ | Hi pigeon |
| 23:38.26 | *** join/#gpe paul_ (~paul@hibernia.jakma.org) |
| 23:39.43 | *** join/#gpe ade|home (~adrian@modem-2529.gorilla.dialup.pol.co.uk) |