| 00:23.34 | *** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) |
| 01:16.55 | *** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) |
| 01:30.14 | T0mW | GPSFan: anyway |
| 01:32.26 | GPSFan | hola.. |
| 01:33.01 | T0mW | GPSFan: filesystem for the zipit is at 77Meg, dunno why, I was about 40Meg before... |
| 01:33.14 | T0mW | GPSFan: lotta stuff in there! |
| 01:33.31 | T0mW | GPSFan: OH! I have mp3 files on it, that's why |
| 01:33.36 | GPSFan | 77Meg... wow whatcha got in there? |
| 01:33.58 | GPSFan | ah, that's probably it.. |
| 01:34.26 | T0mW | GPSFan: lotta stuff. heh. irssi (irc chat), pine, bash (real bash not busybox), sysvinit, initscripts, openvpn, openssh, ... |
| 01:35.21 | T0mW | sox, minicom, and the list goes on. |
| 01:35.41 | T0mW | links (ssl capable browser) |
| 01:35.58 | GPSFan | so this filesystem will live on an mmc, or in nfs? |
| 01:36.05 | T0mW | GPSFan: live on mmc |
| 01:36.21 | T0mW | GPSFan: no nfs at this point, the utils are there... |
| 01:36.43 | T0mW | GPSFan: basically, a mini-laptop but without a gui |
| 01:37.38 | T0mW | GPSFan: I'm just sharpening my skills with OE for now. Later, I plan to use OE to build up a system for a new design (in a couple of months from now). |
| 01:38.03 | T0mW | GPSFan: the zipit is a good place to try stuff out, and, I think it is somewhat useful? |
| 01:38.13 | GPSFan | the zipit is a good platform to do that on. almost a laptop, but not as bloated. |
| 01:39.06 | T0mW | GPSFan: I'm busy bloating... |
| 01:39.07 | T0mW | heh |
| 01:39.13 | GPSFan | hohoho... |
| 01:39.37 | T0mW | GPSFan: got 512Meg on the SD card to use for "stuff" |
| 01:39.38 | GPSFan | I bloat, you bloat, he-she-or it bloats... |
| 01:40.03 | T0mW | too bad there is so little RAM in this thing, another 16meg would have made it interesting. |
| 01:40.38 | GPSFan | a 32 meg zipit would have been awesome, or 64 meg.. |
| 01:41.01 | T0mW | I'm still not convinced that the CPU is actually running at 90MHz, I think it is running at 74MHz. |
| 01:41.14 | GPSFan | but 16meg keeps you on your toes. |
| 01:42.01 | T0mW | GPSFan: I can make it go faster, but then the serial baudrate gets fscked |
| 01:42.02 | GPSFan | hmm, check the baud rate divisors, if the 90meg ones are in use and the serial port works then it's at 90MHz. |
| 01:43.10 | GPSFan | been a while since I looked at the speed stuff, there is a register in the cpu that sets the speed IIRC. |
| 01:44.44 | T0mW | I think there is a mistake in my copy of the cirrus datasheet for the EP7212 regarding the "PLLW". |
| 01:46.10 | T0mW | GPSFan: that is the PLL multipler location (80002610). You can put any value into that and nothing happens. But, if you put values into DAI64FS (loc 80002600) it definately affects the CPU speed. |
| 01:47.08 | GPSFan | hold on a bit while i look at my datasheets. |
| 01:47.11 | T0mW | I should look to see if there is an errata someplace, or that I have the latest manual. |
| 01:50.51 | T0mW | GPSFan: this is the CPU clock init: http://rafb.net/p/Y9kL7d31.html |
| 01:52.19 | GPSFan | that is in the bootcode? |
| 01:52.33 | T0mW | ah |
| 01:53.01 | T0mW | SYSCON3 is only 16bit wide, PLLW is only 8 bits wide. I'm writing 32bits into each! |
| 01:53.08 | T0mW | doh |
| 01:53.52 | GPSFan | ah, that'l do it... |
| 01:55.35 | T0mW | yeah, I threw the OpenZipit sources for the zpm.S and loader.S away and wrote new sources from scratch. The old sources were just too confusing to follow. |
| 01:56.17 | GPSFan | what rev is your 7312 user manual? or date? |
| 01:56.39 | T0mW | so, there is probably some mistakes there. I wonder if that would also explain why, after uploading new code into flash, the LCD will "sparkle" like mad on the first bootup, then settle down? |
| 01:57.09 | T0mW | GPSFan: March 2002, DS508UM4 |
| 01:57.39 | GPSFan | ah it's way out of date , my DS508UM4 is jan 2004 |
| 01:59.14 | GPSFan | latest one is http://www.cirrus.com/en/pubs/manual/ep73xx_um-f.pdf |
| 02:00.46 | T0mW | I don't think the kernel has code to set the cpu clock, do you? |
| 02:01.29 | GPSFan | I don't think it's recommended for the 7312, once the kernel is running. |
| 02:02.38 | GPSFan | you ever use qemu? |
| 02:02.41 | T0mW | no |
| 02:02.47 | T0mW | what is that? |
| 02:03.06 | GPSFan | multi-architecture emulator. |
| 02:03.20 | T0mW | oh, no |
| 02:03.21 | GPSFan | arm, mips,ppc... |
| 02:03.53 | T0mW | I haven't seen / heard of any of those things that actually are worth the trouble setting up to use them. |
| 02:04.55 | T0mW | I did find one that was interesting, but the documenation was absolutely abismal! Lots of code and little documentation. Don't recall what one that was, the server was in Japan. |
| 02:05.06 | GPSFan | I'm trying to find a way to run some mipsel binaries. I don't have access to the actual hardware. |
| 02:05.17 | T0mW | nod |
| 02:06.07 | GPSFan | the OPENWRT folks have used it to help reverse engineer some of the WRT54G binary blobs. |
| 02:06.08 | T0mW | only thing that I've done along those lines was taking C code for an 8051 and compiling / testing it on a PC (linux). Then feeding it data via serial ports. |
| 02:06.42 | T0mW | GPSFan: reverse engineering ARM code must be brutal |
| 02:06.58 | GPSFan | ah, maybe dave has, I'll ping on him tomorrow, or maybe tuesday. |
| 02:07.19 | GPSFan | WRT54G is MIPS |
| 02:07.48 | T0mW | GPSFan: well, I guess if you want something bad enough. I once disassembled the tdremote of borlandc so I could get it working inside an embedded 80188 system... |
| 02:08.23 | T0mW | lots of paper man, lots of paper! I was printing out a lot of disassembled code so that I could follow the program flow |
| 02:08.43 | T0mW | I was younger then... |
| 02:08.53 | GPSFan | I used to use Soft-ICE on DOS many moons ago, it and Sourcer were great for R-E. |
| 02:10.38 | T0mW | well, I reverse engineered some 8051 + 8042 code at one time. Needed a keyboard decoder for the AT keyboards and reverse engineered an 8042 once, I had to write the disassembler and the assembler (company was too cheap to spend the $3K to purchase the tools). |
| 02:11.11 | GPSFan | man, that's hard core.. |
| 02:11.11 | T0mW | wasn't too hard considering it was only 2K of binary. |
| 02:11.57 | T0mW | hand written assembler is very easy to take apart, it has structure to it. The problem is taking apart C sourced code, you have to know the compiler really well to get away with that. |
| 02:12.21 | GPSFan | nod |
| 02:13.15 | T0mW | with assembler, you only have to get inside the head of the programmer and learn how he writes code. On that note, I did disassemble a ROM for a Ham Radio Repeater Controller one time to add a female voice to it. It was the CAT-5 controller. |
| 02:13.48 | T0mW | heh, the guy that wrote the code for the CAT-5 controller was NOT an embedded programmer! |
| 02:13.56 | T0mW | sloppy sloppy code |
| 02:14.24 | T0mW | looked more like a BASIC programmer struggling to deal with assembler |
| 02:15.07 | T0mW | I use state machines a lot with my code. My code never can stop and wait for things to happen: test & move on. |
| 02:15.43 | T0mW | the CAT-5 guy would actually put the controller into a test & wait loop on situations |
| 02:16.21 | T0mW | the standard joke was "how do you crash a CAT controller, answer: tell it to do two things at once" |
| 02:16.32 | T0mW | heh |
| 02:16.57 | T0mW | GPSFan: how goes the weather your way, btw, just where are you? |
| 02:17.28 | GPSFan | well, today it rained most of the day, mid 40's. |
| 02:18.19 | GPSFan | the past week was mostly nice, 50's - 60's daytime. |
| 02:18.50 | T0mW | GPSFan: where do you live? |
| 02:19.01 | GPSFan | western colorado. |
| 02:19.06 | T0mW | ah |
| 02:19.07 | T0mW | ok |
| 02:19.21 | T0mW | yeah, I thought it was UTAH |
| 02:19.28 | GPSFan | ~75 air miles from Aspen. |
| 02:19.57 | GPSFan | Utah... not me thanks... |
| 02:19.58 | T0mW | GPSFan: sheep farmer? |
| 02:20.24 | GPSFan | sort of, we raise sheep, have horses, and an orchard. |
| 02:21.07 | T0mW | ok, interesting. Probably pay a lot less in taxes than you would here in the east |
| 02:21.36 | GPSFan | yep, property taxes on ag land are very low here. |
| 02:23.40 | T0mW | regardless of how I set PLLW, the bogo mips remain the same (44.95) |
| 02:24.20 | ds3 | does target online still have the ZipIt? |
| 02:24.22 | T0mW | the serial port baudrate screws up though if just leave it at the 90MHz divsion rate and don't write into PLLw |
| 02:24.34 | T0mW | ds3: dunno, I have 5 of them... |
| 02:25.03 | T0mW | ds3: I wanted to make sure I have some to destroy if need be, heh |
| 02:25.32 | T0mW | GPSFan: shouldn't bogomips be a relative indication of performance?? |
| 02:25.50 | ds3 | they seem very tempting |
| 02:25.55 | T0mW | I would think that if I alter between 90MHz adn 74MHz that the bogomips would differ |
| 02:26.29 | T0mW | ds3: I've put a filesystem on the MMC card and am booting the kernel from it. That removes the limitation of the tiny Flash filesystem. |
| 02:26.50 | ds3 | T0mW: it has a b&w display, right? |
| 02:27.06 | T0mW | what I'd like to try next is to put the kernel itself onto the mmc card, then use apex to suck the kernel in off the mmc card, then boot it. |
| 02:27.28 | T0mW | ds3: 240x320 monochrome (8 grey level) |
| 02:27.40 | ds3 | think that was the part I did not like |
| 02:27.43 | T0mW | or maybe it is 16 level grey, not sure |
| 02:27.47 | T0mW | ds3: yeah |
| 02:27.51 | GPSFan | T0mW: should be, but I forget how it's calculated in the zipit kernel. |
| 02:28.14 | T0mW | GPSFan: it should still be a loop in kernel/ dir |
| 02:29.25 | T0mW | linux-2.4.21/init/main.c: printk("%lu.%02lu BogoMIPS\n" |
| 02:29.50 | GPSFan | 320x240 mono may not be cool, but it's dynamite on power consumption. |
| 02:30.06 | T0mW | GPSFan: what I'm thinking is that we're not actually at 90MHz, maybe we are but I'd like to prove it. |
| 02:31.06 | T0mW | I guess the baudrate screwing up does prove that |
| 02:31.11 | GPSFan | T0mW: so would I, that would help with mp3 decoding if we are NOT at 90M. |
| 02:32.08 | T0mW | that's what I'm thinking. A number of people have commented on that, that @ 90MHz the zipit should be fast enough to do good software decode. Maybe it is the MMC card that is the problem |
| 02:32.33 | ds3 | what's the power source on them? |
| 02:32.53 | T0mW | in any event, I have to take my changes to OE and merge them into their source tree. I'd like to see if I can get commit privs and publish what I have. |
| 02:32.59 | T0mW | ds3: lithum batt |
| 02:33.23 | T0mW | IIRC, it is a 800ma cell |
| 02:33.56 | T0mW | I computed it one time, but, don't recall. They have no fscking part numbers on most of their stuff. |
| 02:34.30 | T0mW | I do have one here that I put another lithium pack into, that pack is a 1000ma cell |
| 02:34.45 | T0mW | runs a bit over 3 hours with the wifi active |
| 02:35.00 | ds3 | that's pretty decent |
| 02:35.52 | T0mW | A fully charged Zipit, with backlight turned on, will run for 130 minutes before the battery is exhausted |
| 02:35.55 | ds3 | wonder if there is a spare serial port (besides the console) to put a BT module in it |
| 02:36.16 | T0mW | A unit without the backlight will run for 169 minutes |
| 02:36.40 | T0mW | ds3: naw, you only can put a three wire serial port on it: TXD, RXD, and GND |
| 02:37.04 | ds3 | T0mW: is that in addition to the console? |
| 02:37.18 | T0mW | ds3: the idiots that designed this thing deliberately took pains to NOT leave anything like expansion lying around on the board. |
| 02:37.29 | ds3 | oh |
| 02:37.32 | T0mW | ds3: 802.11b |
| 02:38.07 | T0mW | ds3: some people have figured out how to put an SD card in it (bit bang SPI bus). |
| 02:38.19 | T0mW | ds3: http://www.openzipit.org/Addon_Board/ |
| 02:38.41 | GPSFan | if you shut of the radio, the zipit only draws quite a bit less. |
| 02:38.50 | T0mW | ds3: goto the bottom of that page |
| 02:38.55 | GPSFan | ~ 1/2 watt. |
| 02:39.03 | T0mW | GPSFan: yeah |
| 02:39.24 | T0mW | GPSFan: but, where is the fun in that? |
| 02:39.27 | ds3 | ah |
| 02:39.43 | T0mW | GPSFan: ah, well, I guess a pocket MP3 player... heh |
| 02:39.55 | T0mW | ds3: that is my board |
| 02:40.02 | GPSFan | T0mW: well if you are solar powered, a long way from power lines, every bit helps. |
| 02:40.06 | T0mW | jej |
| 02:40.08 | T0mW | heh |
| 02:40.22 | ds3 | T0mW: are you the ISP owner who started all this? (first heard about this from the LH Board) |
| 02:40.42 | GPSFan | http://www.paonia.com/~kenm/Pix/?display=images/Zipit/Weather/pict1442.jpg&width=800&height=600 |
| 02:40.54 | T0mW | ds3: dunno what "started" means. No, I'm just one of the camp-followers |
| 02:42.05 | T0mW | ds3: I just took a lot of the hardware hacks and pulled them together into something more centralized. |
| 02:46.05 | T0mW | now, I'm wondering why the same gcc & binutils on two machines give different results. gdb does not like the local vars debug info on the 64bit O/S while it is fine with the 32bit system. Both machines are AMD X2 systems... |
| 02:46.10 | ds3 | n/m... think the person I was thinking has a elkgrovewireless.com address |
| 02:46.42 | T0mW | GPSFan: screw it, 44.95 is fast enough |
| 02:47.07 | T0mW | this 2.4 kernel really is not cool, wish this was 2.6 |
| 02:47.35 | T0mW | I keep bumping into probs with OE over NPTL threads |
| 02:47.47 | GPSFan | T0mW: It's been about 2 years since I came to much the same conclusion... ;>) |
| 02:48.01 | T0mW | GPSFan: heh, I'm a slow learned |
| 02:48.03 | T0mW | GPSFan: heh, I'm a slow learner |
| 02:48.28 | T0mW | well, I just think of all that ARM experience that I'm getting.. |
| 02:49.32 | T0mW | GPSFan: I can now write ARM asm without hardly ever having to look at the Quick Reference Card anymore. |
| 02:50.19 | GPSFan | ds3: there were several sites long ago that had zipit related stuff, but only after zipitpet & I got a useable kernel running and I wrote/re-wrote the lcd, keyboard and sound drivers did things take off. |
| 02:50.22 | GPSFan | http://www.aibohack.com/zipit/ |
| 02:50.39 | T0mW | yeah, then I "fixed" what they did |
| 02:50.43 | T0mW | hehhehee |
| 02:51.05 | T0mW | GPSFan: seriously though, you two did a lot of nice work |
| 02:51.07 | ds3 | nice to get some live info |
| 02:52.00 | T0mW | GPSFan: time to put all this into OE so it can suffer from bit-rot |
| 02:52.03 | T0mW | ;-) |
| 02:52.14 | GPSFan | T0mW: well there is still a lot more to do, one day I'll have to publish the new wireless driver, and port to kernel 2.4.27. I think I finally got a legal way for the new wireless driver to be built. |
| 02:52.50 | T0mW | GPSFan: whenever you are ready, I can do the integration |
| 02:53.17 | T0mW | it has taken me about 5 months to learn OE to the point where I don't feel overwhelmed by it |
| 02:54.21 | GPSFan | I was almost to the point of publishing it in Jan. then I got sidetracked with a UPS problem and had to build a new server. haven't got back to it... |
| 02:55.44 | T0mW | GPSFan: what is the driver code based off off? orinico driver (spelling)? |
| 02:55.45 | GPSFan | nightall... |
| 02:55.51 | T0mW | GPSFan: nite |
| 04:47.53 | *** part/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca) |
| 04:48.04 | *** join/#edev LyosNorezel (n=lyosnore@unaffiliated/lyosnorezel) |
| 04:59.15 | *** join/#edev ico2 (n=ico2@80-42-140-38.dynamic.dsl.as9105.com) |
| 06:23.40 | *** join/#edev DrZimmerman (n=theo@84-74-134-243.dclient.hispeed.ch) |
| 06:51.14 | *** join/#edev redblue (i=star@ppp040.108-253-207.mtl.mt.videotron.ca) |
| 07:51.48 | *** join/#edev redblue (i=star@ppp201.108-253-207.mtl.mt.videotron.ca) |
| 08:19.01 | *** join/#edev redblue (i=star@ppp169.108-253-207.mtl.mt.videotron.ca) |
| 10:17.01 | *** join/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net) |
| 12:43.32 | *** join/#edev swinchen (n=swinchen@strongbad.eece.maine.edu) |
| 12:47.58 | *** join/#edev Chocobo (n=swinchen@strongbad.eece.maine.edu) |
| 13:27.52 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:29.18 | prpplague | T0mW: you looking for me over the weekend? |
| 13:29.32 | *** mode/#edev [+o prpplague] by ChanServ |
| 13:30.35 | T0mW | prpplague: yeah, question about apex |
| 13:31.16 | T0mW | prpplague: I'd like to store the kernel image on the MMC card ext3 filesystem and only have apex in Flash |
| 13:31.35 | T0mW | prpplague: then have apex suck the kernel off the mmc and put it into RAM |
| 13:31.45 | T0mW | prpplague: do you know if apex can do that? |
| 13:31.49 | prpplague | T0mW: easy enough to do |
| 13:31.54 | T0mW | good |
| 13:32.32 | prpplague | T0mW: ext2, has mmc framework, and you can script it do what you need as far as copying the kernel image from A to B |
| 13:32.41 | T0mW | prpplague: I'll do that with the zipit then. The other project could benefit from not having to program the Flash memory in the field to update the kernel image. |
| 13:32.50 | prpplague | T0mW: the only thing thats not in the main tree is the mmc commands for the s3c24xx |
| 13:33.00 | T0mW | ok |
| 13:33.09 | prpplague | T0mW: but thats very easy to add |
| 13:33.16 | T0mW | prpplague: that could be easily handcoded in assembler |
| 13:33.32 | T0mW | or C |
| 13:33.36 | prpplague | T0mW: no need, there is a frame work there already in apex |
| 13:33.42 | prpplague | T0mW: just need to fill in the blanks |
| 13:33.48 | T0mW | prpplague: ok, thanks. |
| 13:33.54 | prpplague | np |
| 13:34.00 | T0mW | prpplague: any idea when you'll be sending that other unit out to me? |
| 13:34.59 | prpplague | T0mW: what? you should already have it! |
| 13:35.09 | T0mW | prpplague: no |
| 13:35.14 | prpplague | T0mW: let me get the tracking number from shipping and trace it |
| 13:35.24 | prpplague | T0mW: i'll let you know in an hour or so |
| 13:35.47 | prpplague | T0mW: i gave it to our shipping girl over a month ago |
| 13:36.30 | T0mW | prpplague: ok, great |
| 13:40.00 | *** join/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca) |
| 13:56.36 | *** join/#edev Chocobo (n=swinchen@strongbad.eece.maine.edu) |
| 14:01.10 | chouimat | morning |
| 14:05.10 | *** join/#edev mattastrophe (n=bigboota@209.250.156.214.tor.pathcom.com) |
| 14:14.04 | prpplague | anyone have experience with SAP ? |
| 15:03.50 | T0mW | prpplague: well, when I was a kid, my cousin and I would tap the sugar maples on the farm... |
| 15:04.10 | *** join/#edev redblue (i=star@ppp058.110-253-207.mtl.mt.videotron.ca) |
| 15:16.49 | *** join/#edev Crofton (n=balister@hc65216ee.dhcp.vt.edu) |
| 15:20.27 | Crofton__ | I did some of that also :) |
| 15:25.33 | *** join/#edev dfarnsworth_ (n=dfarnswo@65.200.49.156) |
| 15:38.08 | prpplague | Crofton__: friday will be 30 days for scox |
| 15:38.37 | prpplague | Crofton__: today if they count the one day that scox closed at $1 |
| 15:39.18 | Crofton | cool |
| 15:39.23 | Crofton | What is SAP? |
| 15:39.39 | prpplague | <PROTECTED> |
| 15:39.47 | prpplague | accounting #$%#$%#$%!@#!@#!23 |
| 15:42.28 | Crofton | Sorry no experience with them |
| 15:46.03 | chouimat | prpplague: I was supposed to have some training on it but I quit after 2 days ... that thing scared me |
| 15:50.28 | *** join/#edev josejx2 (n=jjezak@c-71-199-122-57.hsd1.pa.comcast.net) |
| 15:52.50 | prpplague | chouimat: yea its a real nightmare |
| 15:52.59 | prpplague | chouimat: total legacy design |
| 15:55.31 | chouimat | prpplague: and the place where I was backend their business process was a fscking nightmare too |
| 15:55.40 | chouimat | s/backend/back then |
| 15:55.49 | chouimat | can't type today ... need more coffee |
| 16:05.25 | *** join/#edev GPSFan (n=kenm@12.10.255.248) |
| 16:05.25 | *** mode/#edev [+o GPSFan] by ChanServ |
| 18:25.37 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-31-110.westend.de) |
| 19:21.43 | *** join/#edev Chocobo (n=swinchen@strongbad.eece.maine.edu) |
| 19:42.16 | *** join/#edev Crofton (n=balister@hc65216ee.dhcp.vt.edu) |
| 20:36.22 | *** join/#edev purl (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 20:36.22 | *** topic/#edev is a channel for the discussion of hacking small consumer electronic devices, mostly ARM based |
| 21:31.24 | *** join/#edev ico2 (n=ico2@modem-1.chameleon.dialup.pol.co.uk) |
| 22:29.41 | *** join/#edev redblue (i=star@ppp118.108-253-207.mtl.mt.videotron.ca) |
