| 00:01:40 | kernelwork | So long everyone, see you all monday |
| 00:08:06 | Tangent | kernelwork : Later |
| 01:40:03 | BZFlag | nite |
| 02:00:45 | dirkx | No: for each of them I use something like: .//configure --host=arm-uclibc --build=i686-pc-linux-gnu --with-prefix=arm-uclibc --with-profile=profiles/profile.tuxscreen --with-extra-ipath=/home/dirkx/tuxscrteen/jpeg-6b --with-extra-lpath=/home/dirkx/tuxscrteen/jpeg-6b --prefix=/home/dirkx/tuxscrteen/buildroot-tux/root/usr/local - and then wack it into the root. Same for wlan and same for the tuxphone. And then mount that root remotel |
| 02:00:51 | dirkx | So I skip the disk set you are doing,. |
| 02:29:39 | dirkx | msg harld No - i need to add a -D__arm__ in config.mk for the wlan package. I'll send you a diff. Hold on |
| 03:33:21 | SpaceCoaster | GoRK: hey, good job on linux-arm-kernel |
| 04:18:15 | sjm | test |
| 05:01:01 | Tangent | goes to bed... night allkernelwork : Later |
| 05:01:08 | Tangent | goes to bed... night all |
| 05:21:46 | SpaceCoaster | GoRK: you alive? |
| 09:40:02 | Sammy | cold today ... |
| 09:42:53 | | jacques: San Jose, San Jose International Airport, CA, United States; (KSJC) 37-21-33N 121-55-27W 25M; last updated: Nov 22, 2001 03:53 AM EST; Dew Point: 57.0 F (13.9 C); Precipitation last hour: 0.01 inches; Pressure (altimeter): 29.95 in. Hg (1014 hPa); Pressure tendency: 0.01 inches (0.3 hPa) higher than three hours ago; Relative Humidity: 89%; Sky conditions: overcast; Temperature: 60.1 F (15.6 C); Visibility: 10 mile(s); Weather: 77 F (25 C); Wind: from the SE (130 degrees) at 5 MPH (5 KT) |
| 09:42:53 | jacques | ibot1: weather ksjc |
| 09:48:22 | Sammy | bbl |
| 09:59:10 | vosque | That was fun! |
| 14:54:07 | SpaceCoaster | spse__: well done for defending the blkmtd code! |
| 14:54:43 | spse__ | ta ;) |
| 14:55:12 | spse__ | i had seen the crash before but i didnt know about the alt-sysrq-T trick |
| 14:55:35 | spse__ | btw, how is your atmel stuff coming along? |
| 14:56:35 | SpaceCoaster | No progress since we last talked, not been doing much of anything. |
| 14:56:46 | spse__ | ;) |
| 14:57:59 | SpaceCoaster | Currently trying to get the 2.4.13/15 kernels to boot, there is an intermittent problem, they mostly hang on bootup but sometimes not. |
| 14:58:38 | spse__ | some device initialisation going wrong? |
| 15:00:28 | SpaceCoaster | spse__: probably, I'll look at the UCB1200 code, that likes to hang, if the chip isn't reset properly. |
| 15:06:50 | SpaceCoaster | spse__: yup it's the ucb1200, fixed it twice already, wish I was invoicing for this :-) |
| 15:13:12 | jacques | SpaceCoaster: you already did the mtdblock fix on those kernels? |
| 15:13:29 | jacques | SpaceCoaster: where is it hanging? |
| 15:14:10 | jacques | I have been playing with 2.4.13 and 2.4.15 - main problem is they changed the touchscreen driver and microwindows does not like it at all |
| 15:14:35 | SpaceCoaster | jacques: not yet, hanging in mcp_sa1100_read, it's just after it initialises the RAMDISK. |
| 15:15:05 | jacques | <andersee> GoRK: The fix is to edit buildroot-tux/linux/drivers/mtd/mtdblock.c and remove the bogus call to invalidate_device() in mtdblock_release() |
| 15:15:13 | jacques | worked for me |
| 15:15:32 | jacques | fix is already in mtd cvs, but not in any kernel source |
| 15:15:56 | SpaceCoaster | jacques: thanks when I get it to boot that will come in useful. Saved me an hour or two or reintegrating MTD. |
| 15:16:29 | jacques | SpaceCoaster: yeah I tried usingthe mtd cvs, but I always got kernels which did not detect the partition table |
| 15:17:39 | jacques | anyone know about the touchscreen driver changes? |
| 15:17:48 | jacques | I would like to get microwindows working again |
| 15:18:19 | jacques | the ts major and minor changed, but fixing that did not totally fix the problem |
| 15:18:53 | SpaceCoaster | jacques: I have written a translation module which will provide the old functionality but I haven't been able to test it coz I can't boot. |
| 15:19:48 | jacques | SpaceCoaster: have you tried the mtdblock.c fix? I'm confused - do you think your problem is something else? |
| 15:19:56 | SpaceCoaster | For integrating the functionality into microwindows it divides into two parts, struct changes (mostly trivial format change) and calibration. |
| 15:20:37 | jacques | ah, so it's more than trivial changes |
| 15:20:48 | jacques | looks like a total re-org of the ucb driver |
| 15:21:21 | SpaceCoaster | jacques: about 6 weeks ago I fixed a reset problem with the UCB1200 which caused 50% of TuxScreens to hang when the touchscreen was accessed. The ucb* reorg keeps moving this reset around and breaking it. |
| 15:22:07 | jacques | SpaceCoaster: ok, that's one hang, but are you sure that's what you're seeing now? |
| 15:22:22 | jacques | SpaceCoaster: is it hanging right after "mounting local file systems:" ? |
| 15:22:35 | jacques | SpaceCoaster: if so, it's the mtdblock thing |
| 15:22:40 | SpaceCoaster | jacques: do you have a million dollars to bet? |
| 15:23:04 | jacques | SpaceCoaster: ok, sigh. I am confused. that change worked for me on 2.4.13 and 2.4.15 |
| 15:24:10 | SpaceCoaster | jacques: I put a printk at the start of mcp_sa1100_read() and a printk before the end. The first one prints the second one doesn't. The code between contains a while loop looking for a bit to be set. |
| 15:25:37 | SpaceCoaster | the bit is never set because the UCB1200 is not functioning because it was never reset. Ask anyone, with the UCB1200 reset, SpaceCoaster is the "man" :-) |
| 15:26:11 | jacques | SpaceCoaster: ok, it just sounded like the problem I had |
| 15:28:32 | SpaceCoaster | jacques: take a look at the linux-arm-kernel mailing list for information on ucb1x00-ts, it is a really good techie punchup, ends with rmk taking his ball home and not playing. |
| 15:28:47 | jacques | ugh, not again |
| 15:29:07 | jacques | cool, I'll take a look |
| 15:29:51 | SpaceCoaster | I am new to this world, so I find it pretty funny. |
| 15:30:09 | jacques | well, it pretty much is |
| 15:30:20 | jacques | but can be counter-productive |
| 15:31:43 | SpaceCoaster | jacques: I think people should be more independent, rmk does good work but if he doesn't want to do something then fair enough, do it yourself, if you want to. |
| 15:32:09 | jacques | the problem is when two big guns want to do it in two incompatible ways |
| 15:32:40 | SpaceCoaster | eventually one will get bored or give in |
| 15:33:18 | jacques | true, but the carnage they leave in their wake can be scary |
| 15:34:47 | jacques | hmm, time to eat something |
| 15:35:15 | SpaceCoaster | jacques: me too |
| 16:00:31 | SpaceCoaster | Russ: The codec needs to be reset BEFORE calling mcp_reg_read() in ucb1x00-core.c, in fact why not reset it in shannon.c where the GPIO is configured. |
| 16:43:58 | andersee | GoRK: you around? |
| 16:52:27 | SpaceCoaster | andersee: hi, how are you? |
| 16:53:30 | andersee | doing fine. |
| 16:53:33 | andersee | sleepy though.... |
| 16:53:38 | andersee | And you? |
| 16:54:25 | SpaceCoaster | great, just worked out why 2.4.15 was hanging, twas the UCB1200 again |
| 16:54:43 | andersee | SpaceCoaster: btw, you may want to check out the latest uClibc... I think I got the ARM math lib all fixed. |
| 16:55:26 | SpaceCoaster | excellent, playing in kernel land today, might try later. |
| 16:55:33 | andersee | SpaceCoaster: I ripped out the math lib I had, grafted in the code from MacOs X's math lib, with GNU libc's math.h. :) |
| 16:56:07 | andersee | And then I fixed up the damage -- and it seems to work.... |
| 16:56:35 | SpaceCoaster | how can I download it/ |
| 16:57:01 | andersee | cd buildroot-tux/uClibc. cvs up |
| 16:57:21 | andersee | you may need to give it the passwd 'anoncvs' |
| 16:59:48 | SpaceCoaster | I will try a new buildroot-tux later today, Tim made a stack of changes and it's good to stay current. |
| 17:01:14 | andersee | heh |
| 17:01:36 | andersee | I need to spend some time updating my tree later today |
| 17:02:11 | SpaceCoaster | andersee: you seem to have been away from things for a while, been working? |
| 17:02:41 | sammy_wms | amm... |
| 17:03:20 | sammy_wms | hi andersee and SpaceCoaster :) |
| 17:03:29 | SpaceCoaster | hi sammy_wms |
| 17:03:40 | andersee | howdy |
| 17:03:59 | sammy_wms | some confuse ... |
| 17:03:59 | andersee | SpaceCoaster: yeah. I've been working pretty hard |
| 17:04:22 | SpaceCoaster | andersee: you do consultancy work, don't you? |
| 17:04:44 | sammy_wms | how to let the kernel go through smooth with ramdisk ? |
| 17:04:47 | andersee | SpaceCoaster: yup. And I've got a deliverable tonight, midnight.... |
| 17:05:03 | andersee | SpaceCoaster: So I've had about 8 hours sleep in the last 3 days... :( |
| 17:05:34 | SpaceCoaster | andersee: cool, I hope those hours are billable :-) |
| 17:06:15 | SpaceCoaster | sammy_wms: does it hang after "RAMDISK driver initialized: 16 RAM disks of 8192K size 1024 blocksize |
| 17:06:15 | SpaceCoaster | " |
| 17:06:38 | sammy_wms | yeap |
| 17:06:53 | sammy_wms | is that console question ? |
| 17:06:53 | andersee | nah. per project. But I should get paid today. And paid enough for it to be worth it. |
| 17:07:47 | sammy_wms | may I ask how is console normally set ? |
| 17:08:22 | sammy_wms | andersee : you sleep each day with 2 hours ? |
| 17:08:27 | SpaceCoaster | sammy_wms: hanging problem is because the UCB1200 has not been reset correctly. |
| 17:09:25 | andersee | sammy_wms: I slept 4 hours yesterday, 4 hours the day before... |
| 17:09:51 | SpaceCoaster | sammy_wms: in ucb1x00-core.c move /* reset the codec */ before mcp_reg_read(). |
| 17:10:37 | SpaceCoaster | sammy_wms: Do you want me to mail you ucb1x00-core.c? |
| 17:10:57 | SpaceCoaster | andersee: what are you delivering, is it exciting? |
| 17:11:23 | andersee | SpaceCoaster: not really. Just research... No actualy code. |
| 17:11:28 | andersee | SpaceCoaster: not really. Just research... No actualy code. |
| 17:11:36 | andersee | SpaceCoaster: That comes later.. |
| 17:11:41 | andersee | oops... |
| 17:12:46 | SpaceCoaster | andersee: probably best to stay at the computer, no long driving trips today, ok? :-) |
| 17:13:39 | andersee | heh. I'm going to go to sleep for a couple of hours. Then off to have some turkey dinner. But my wife will be driving. |
| 17:14:04 | andersee | SpaceCoaster: so what sort of progress have you been making? |
| 17:15:41 | sammy_wms | SpaceCoaster: Thank you very much |
| 17:16:59 | SpaceCoaster | not much, I hacked together a touchscreen coordinate module for the new "improved" touchscreen driver. Fixed a few things in the serial_flash code and have started some QPE stuff but nothing completed. Gonna give TuxScreens to my family for Xmas. |
| 17:19:29 | sammy_wms | SpaceCoaster: really, if it can work , also like a gift for me too, because some of my friend also write some App with QT embedded... |
| 17:20:50 | sammy_wms | so if Tux can run on QPE, that's good for me to run that App on it :) |
| 17:21:23 | Tangent | thinks that all the ISDN drivers should be removed from the kernel too to maintian the "I am not a sh*tbucket" approach |
| 17:22:12 | SpaceCoaster | sammy_wms: it can run QPE, performance is pretty good, memory usage is quite high though. |
| 17:22:42 | sammy_wms | SpaceCoaster: btw, I also need that too, but can you DCC to Sammy later ? |
| 17:23:27 | SpaceCoaster | sammy_wms: what do you need? |
| 17:24:22 | sammy_wms | the QPE stuff and that UCB1200 stuff ... |
| 17:26:36 | SpaceCoaster | sammy_wms: let me try the ucb1x00-core.c, never used DCC before |
| 17:27:25 | SpaceCoaster | might have firewall problems... |
| 17:28:46 | sammy_wms | SpaceCoaster: ok , try this sam@mail.cnct.com.tw |
| 17:29:04 | andersee | g'night all. |
| 17:29:19 | andersee | have a happy thanksgiving |
| 17:29:26 | sammy_wms | night andersee :) |
| 17:29:31 | andersee | (for those in the US) |
| 17:38:06 | GPSFanL | is away: I'm busy |
| 17:38:06 | GPSFanL | is back (gone 00:00:07) |
| 17:39:05 | sammy_wms | ( brb ) |
| 19:25:30 | dirkx | q |
| 19:25:59 | dirkx | /msg NickServ IDENTIFY bliepje |
| 19:27:30 | dirkx | q |
| 19:33:41 | dirkx | msg jorn (cd /usr/netbsd/src; ./build.sh -m hpcarm -a arm -D/usr/outdir word) |
| 19:41:40 | Russ | psst dirkx |
| 19:41:46 | Russ | might want to change your password |
| 19:42:32 | SpaceCoaster | Russ: I found out why 2.4.1[35] were hanging for me. |
| 19:43:40 | dirkx | Russ - yes - I just did :-) thanks. I missed a space in my .ircrc script - and the silly thing then sends it out as is :-) |
| 19:44:52 | Russ | SpaceCoaster: hows that? |
| 19:44:58 | dirkx | The mount problem you where describing - any chance that I am seeing the same; using the tuxen on an NFS mount - the machine freezes up after 3-4 hours. Could that be related ? |
| 19:45:08 | Russ | dirkx: no |
| 19:45:29 | SpaceCoaster | Russ: the old reset the ucb1200 before reading the register trick! |
| 19:45:31 | dirkx | russ: thanks - digging deeper then. |
| 19:45:47 | Russ | SpaceCoaster: thats in the latest kernels |
| 19:47:04 | SpaceCoaster | 2.4.15-pre6-rmk1-tux3 doesn't work, what is latest? |
| 19:47:30 | Russ | that is latest |
| 19:48:21 | SpaceCoaster | have a look in ucb1x00-core.c, mcp_reg_read() is called to get the UCB_ID before the codec is reset. |
| 19:48:50 | jacques | Russ: what about the mtd deadlock fix? |
| 19:49:54 | Russ | jacques: its in latest mtd CVS |
| 19:50:02 | Russ | SpaceCoaster: ah, oops |
| 19:50:40 | SpaceCoaster | Russ: how about moving it to shannon.c along with the other GPIO initialisation? |
| 19:51:24 | Russ | not a bad idea... |
| 19:54:50 | GPSFan | jacques: did you ever get your LCC's soldered on? |
| 19:55:25 | jacques | GPSFan: not yet - they are tiny - I need glasses - soldering the wires for the flash unlock was bad enough |
| 19:56:10 | jacques | i cant focus on anything less than a foot away :-\ |
| 19:56:19 | GPSFan | jacques: I got my sample from Memsic, and they sent a proto board with it, I just put it on, and it wasn't really too bad. ftp://ftp.desertweyr.com/pub/Image01.jpg |
| 19:57:57 | jacques | GPSFan: i need something like that |
| 19:58:14 | jacques | GPSFan: what did you use to solder it on? IR? |
| 19:58:31 | dirkx | jacques - that looks a lot like the acceleration meters I used to work with. What is it ? |
| 19:58:48 | jacques | dirkx: got it in one |
| 19:59:06 | GPSFan | jacques: I used a fine tim controlled temp iron, and a stereo zoom microscope. here is another shot of the leads (or not-leads) ;>) ftp://ftp.desertweyr.com/pub/Image02.jpg |
| 19:59:24 | dirkx | Hmm - that is *small*. The one I used for my spectrometer was about 5 times that size. Is the extra pin for the Z axis - or is it true 2 acis and no angular rotation ? |
| 19:59:58 | GPSFan | dirkx: http://www.memsic.com/product_data.htm |
| 20:00:13 | dirkx | GPSFan: thanks - fetching. What is the chip id ? |
| 20:00:20 | dirkx | Ah got it 205. Cool. |
| 20:01:18 | GPSFan | dirkx: it's a 2 axis and has the same basic outputs as the AD parts, but the guts are totally different. |
| 20:01:43 | jacques | yeah, I got the AD version |
| 20:02:02 | GoRK | what are you guys doing? accelerometer? |
| 20:02:06 | GPSFan | dirkx: works good too, only looking at the outputs on a scope though, but it operates down to 2.7V |
| 20:02:34 | jacques | GoRK: I want to test vehicle performance |
| 20:02:43 | dirkx | Very neat. And very cheap. I spend nearly a thousand times more for an underwater instrument :-) would have been nice to have a few cheap vervsions to play with first (as flooding the intstrument is a real risk in the beginning). |
| 20:03:37 | GoRK | btw i am "around" now to all who were asking |
| 20:03:56 | jacques | GPSFan: how does the memsic work? I saw some that measured temperature differentials based on air movement |
| 20:05:01 | GPSFan | jacques: they have a cool animation on their web site, the only moving part is a"bubble of heated air". |
| 20:05:31 | GPSFan | jacques: Memsic is a spin-off of AD. |
| 20:05:42 | jacques | GPSFan: ah, so it's internal then? |
| 20:05:56 | GPSFan | jacques: the bubble-yes. |
| 20:06:09 | jacques | GPSFan: sounds like it could take a lot more g-forces than the micromachines AD uses |
| 20:06:23 | GPSFan | jacques: 50,000g's |
| 20:06:41 | jacques | yeah, I think the AD is only 1000 |
| 20:07:07 | SpaceCoaster | GoRK: welcome back |
| 20:07:08 | GPSFan | rumor has it that MMD's wear out too. |
| 20:07:35 | jacques | yeah I could see that |
| 20:07:52 | GoRK | SpaceCoaster: thanks. i have been sick. i am still getting over it a bit but better |
| 20:08:20 | dirkx | Hmm - but the Si is in the oder 0.004 per sqrt(hz). That is not that neat considering the frequency response. |
| 20:08:39 | SpaceCoaster | GoRK: do some hacking, that always makes me feel better. |
| 20:09:49 | GPSFan | dirkx: that's true, but there is a temp output so you can compensate. |
| 20:10:18 | dirkx | Hmm - A/D has nicer ways of doing it :-) i.e. on substrate H bar vibrators. This sort would never give you things like angular momentums. Nice though - for those prices ! wow ! you could put them in a joystick or in a PDA and nevr thing twice of it. |
| 20:10:46 | dirkx | Just ordered one :-) I've been cominging a trimble GPS with www.activewireinc.com - and this makes a nice addition. |
| 20:10:59 | dirkx | Thanks for the hint. |
| 20:11:05 | GPSFan | dirkx: that's where they see their market. |
| 20:11:40 | dirkx | The sample they send you - is that one of each ? or simply just one ? (aside from the mount board). |
| 20:12:08 | jacques | dirkx: ever hear of the MS sidewinder pro gamepad? |
| 20:12:19 | dirkx | http://www.memsic.com/AppNote5.pdf has the Spice model. :-) |
| 20:12:31 | dirkx | jacques - no - I take it is a games joy stick ? |
| 20:12:34 | GPSFan | dirkx: ActiveWire USB is cool, I wish I had mor use for it. the sample was only one- MXD205 and the proto board. they believe that you need to be serious to solder the LCC on. |
| 20:12:37 | jacques | dirkx: it has an adxl202 in it |
| 20:13:39 | jacques | dirkx: they are tilt game-pads |
| 20:13:49 | jacques | didn't sell very well but then the games sucked |
| 20:15:26 | jacques | anyone know if laptop drives use 5V ? or 3.3? |
| 20:15:40 | jacques | never mind |
| 20:15:45 | jacques | i just looked at mine |
| 20:15:49 | jacques | 5V |
| 20:16:09 | dirkx | jacques: depends on the type: my sony uses 3.3 - but the IBM thinkpad 5. I think sony is the exception as in the larger sonies they use 5v too. |
| 20:16:25 | jacques | dirkx: ah interesting |
| 20:22:47 | GoRK | whoops i guess i got disconnected.. |
| 20:22:50 | GoRK | heh |
| 20:22:53 | GoRK | andersee: ill test out libm today |
| 21:24:06 | GoRK | andersee: perl 'make test' is looking very very good |
| 21:39:40 | GoRK | Failed 3 test scripts out of 249, 98.80% okay. |
| 21:40:05 | GoRK | Really it only failed 3 subtests out of 12503 |
| 21:41:31 | GoRK | andersee: i mailed you the output. I will look into those other non-critical failures this evening to try and see if there are more little uclibc bugs hiding there |
| 21:41:53 | GoRK | one is in ipc one is in groups |
| 22:16:32 | RedBaron | Howdy howdy |
| 22:16:44 | andersee | greeting |
| 22:16:46 | andersee | s |
| 22:18:23 | RedBaron | Seems kind of quiet here. |
| 22:20:49 | Tangent | Nod... I'm quiet becasue I'm hacking away at pyton |
| 22:22:08 | SpaceCoaster | RedBaron: I'm quiet because I'm full of turkey. |
