| 00:13.25 | MonMotha | swinchen: find a datasheet for a part that uses it extensively. They'll usually explain just about the whole thing |
| 00:13.34 | MonMotha | barring that, if you're a member of IEEE, it's a standard that you should have some access to |
| 00:13.41 | MonMotha | I don't think it's free otherwise |
| 00:14.16 | MonMotha | the basic premise though is that you create a freaking huge shift register comprised of all your parts, and you shift data through the chain |
| 00:14.45 | swinchen | hrrmmmm.... Do you have an example MonMotha? I am looking at the eZ430. It has a USB JTAG part and a TIMSP430 processor attached to it. |
| 00:18.47 | MonMotha | swinchen: USB JTAG? |
| 00:19.10 | MonMotha | the JTAG bus itself is 4 wires (well 5, with a ground reference): TDI, TDO, TMS, TCK |
| 00:19.30 | MonMotha | you clock the bus with TCK, shifting data in over TDI and receiving data over TDO. TMS is used to say when you want something to be commited |
| 00:20.25 | MonMotha | there are tons of different USB interfaces to JTAG, none of which speak anything resembling a standard protocol over USB AFAIK |
| 00:21.53 | prpplague | swinchen: http://www.elinux.org/wiki/JTAG |
| 00:22.13 | prpplague | swinchen: anything you can think of that is related to embedded dev is usually documented on the elinux wiki |
| 00:27.19 | MonMotha | I'm curious, how do most people pronounce JTAG? |
| 00:27.33 | MonMotha | Jay-Tag, Jay-Tee-Ehh-Gee? |
| 00:28.21 | prpplague | Jay-Tag |
| 00:30.07 | MonMotha | that's what I figured |
| 00:30.44 | prpplague | ~seen jbreven |
| 00:30.57 | purl | prpplague: i haven't seen 'jbreven' |
| 00:30.57 | agaffney | he's been around |
| 00:31.09 | agaffney | he said something ~2.5 hours ago |
| 00:31.36 | agaffney | he's also the one who said it would be boring if you disappeared :P |
| 00:31.58 | swinchen | This is the thing I am trying to access: http://www.ti.com/corp/docs/landing/ez430tool/index.htm?DCMP=430Day&HQS=Other+OT+ez430 |
| 00:32.47 | swinchen | And this is the datasheet (with full schematics) http://www-s.ti.com/sc/techlit/slau176 |
| 00:32.50 | swinchen | errr |
| 00:33.03 | swinchen | http://focus.ti.com/lit/ug/slau176a/slau176a.pdf |
| 00:33.21 | prpplague | swinchen: yea we are familiar with it |
| 00:33.27 | *** join/#edev jbevren (n=jbevren@dumt-02-164.dsl.netins.net) |
| 00:33.56 | prpplague | speak of the crack baby |
| 00:35.40 | swinchen | Ok, so basically the USB portion of the eZ430 uses a closed/proprietary protocol...? |
| 00:36.31 | MonMotha | probably, check the datasheet |
| 00:36.33 | prpplague | swinchen: i said we were familiar with it, hehe, not that we were dev'n for it, hehe |
| 00:36.45 | prpplague | swinchen: i've not looked at it in detail |
| 00:37.29 | swinchen | Ahh, ok. Thanks. I am not sure where a good starting point is. I have loaded the generic usbserial driver with the correct vendor and product id's... hrmm |
| 00:37.51 | MonMotha | isn't that for the UART things? |
| 00:38.10 | prpplague | swinchen: i'd start by doing some googles, i'm sure there are some linux people using it |
| 00:38.46 | prpplague | swinchen: if you get it working, please create a wiki page at elinux.org |
| 00:40.25 | MonMotha | if you do want a USB JTAG dongle, a company called Digilent makes a reasonably priced one that I believe they have specs out for |
| 00:40.28 | swinchen | MonMotha, I dont know... I tried using the ti_usb_3410_5052 driver (the eZ430 uses the 3410 USB Chip) but I could get that to create the device node. |
| 00:40.42 | MonMotha | oh, that's probably just a generic USB uC |
| 00:41.56 | swinchen | Yeah, I think it is. USB <-> UART so I tried opening it up in minicom and tried sending some crap basically to see if I could get it to say something. No luck. |
| 00:43.10 | MonMotha | it may not be a dedicated UART thingie. Look up the part |
| 00:43.22 | MonMotha | there's a bunch of USB enabled uCs laying around that can be programmed to do just about anything |
| 00:43.59 | prpplague | jbevren: ping |
| 00:46.26 | swinchen | "USB TO SERIAL PORT CONTROLLER" is what the datasheet title is... I will dig into it a little. |
| 00:47.27 | MonMotha | hum, dunno |
| 00:47.48 | MonMotha | you may just have to speak to whatever is hooked up to the serial port on there in a specific way before it will say anything back |
| 00:52.49 | swinchen | That is what I was thinking. :( I wonder if it is possible to "spy" on a usb port like you do with RS-232. I am sure it isn't as striaght forward. |
| 00:53.09 | MonMotha | you don't do it with hardware |
| 00:53.21 | MonMotha | there are USB drivers that will tell you everything |
| 00:53.35 | swinchen | Ahh, so I will need to find those in windows. |
| 00:53.56 | prpplague | i fairly sure there are some linux utils for that |
| 00:53.56 | swinchen | So when I use their development tools I can see everything. That would make this a lot easier! |
| 00:54.10 | MonMotha | IIRC there are windows stuff that does that, but I don't know of any free ones |
| 00:54.15 | swinchen | I can't do it in linux.... because the TI tools are windows only. |
| 00:54.19 | swinchen | Crap. |
| 00:59.09 | *** join/#edev aundry (n=jamesdoe@pool-71-252-243-155.dllstx.fios.verizon.net) |
| 01:04.09 | swinchen | Ahhh, it looks like the 3410 may be a general chip... it has a microcontroller and ports and all that jazz in it. |
| 01:23.17 | swinchen | hmm. Well, it certainly looks like it may be more work that I thought originally. |
| 02:07.52 | Chocobo | Dude look! A free USB snooper for Windows: http://usbsnoop.sourceforge.net/ |
| 02:08.41 | jbevren | Chocobo: http://mailman.uclinux.org/pipermail/uclinux-dev/2002-August/010335.html |
| 02:09.50 | Chocobo | I dont think I understand :/ |
| 02:09.59 | jbevren | you noted the samsung 3410 |
| 02:10.04 | jbevren | thats a post about uclinux on it :P |
| 02:11.11 | Chocobo | I wonder if it is similar to the TUSB3410 (The TI chip on my board) |
| 02:11.49 | Chocobo | That would be impressive if you could put uclinux on it. |
| 02:12.28 | jbevren | http://news.nationalgeographic.com/news/2006/06/060613-cat-bear.html |
| 02:12.34 | jbevren | oh probably not |
| 02:12.41 | jbevren | I thought you were talking about the sc3410 arm cpu :) |
| 02:13.28 | Chocobo | no, sorry. Thanks though :) |
| 02:13.58 | Chocobo | Oh yeah... right here: http://benoit.papillault.free.fr/usbsnoop/ |
| 02:14.09 | Chocobo | That will crack my problem. Time to reverse engineer. |
| 02:20.41 | Chocobo | Well, goodnight all. |
| 02:33.06 | *** join/#edev kaenat_ (n=ringlej@pool-71-101-131-179.tampfl.dsl-w.verizon.net) |
| 02:38.27 | jbevren | http://news.nationalgeographic.com/news/2006/06/060613-cat-bear.html |
| 04:06.50 | *** join/#edev redblue (i=star@ppp165.108-253-207.mtl.mt.videotron.ca) |
| 09:26.17 | *** join/#edev redblue (i=star@ppp125.108-253-207.mtl.mt.videotron.ca) |
| 11:43.16 | *** join/#edev bd2 (n=ufo@81.222.177.146) |
| 12:47.41 | *** join/#edev tux1800 (n=eric@c66.110.147-33.clta.globetrotter.net) |
| 12:53.35 | *** join/#edev prpplague (n=billybob@72.22.134.23) |
| 13:07.35 | *** join/#edev Pix (i=pix@crazyfrogs.org) |
| 13:35.31 | *** mode/#edev [+o prpplague] by ChanServ |
| 14:06.22 | *** join/#edev kaenat_ (n=ringlej@m1.qa.keyvoice.com) |
| 14:06.46 | *** join/#edev tuxmaniac (n=aanjhan@60.254.67.17) |
| 14:07.09 | agaffney | mmm...nice |
| 14:07.19 | agaffney | I just picked up a 10.4" CTX LCD display |
| 14:07.25 | agaffney | 640x480 |
| 14:07.26 | agaffney | VGA |
| 14:07.29 | agaffney | *free* |
| 14:16.01 | sorphin | nod |
| 14:19.47 | prpplague | agaffney: interesting |
| 14:20.29 | agaffney | the web developer I work with got it when he worked at Dierberg's (large grocery chain) |
| 14:20.34 | agaffney | they were throwing a bunch of them out |
| 14:20.46 | agaffney | I just traded him an old dual tape deck I had at home for it |
| 14:20.46 | prpplague | ahh, man i'd like to have had some of those |
| 14:21.03 | agaffney | I've been meaning to get rid of the tape deck, since I never use it |
| 14:56.02 | *** join/#edev EvilDevil_ (n=miau@p54A6D5D6.dip.t-dialin.net) |
| 14:58.29 | agaffney | it would be nice if this thing was touch screen or something, but I suppose I should be happy with what I got :) |
| 15:23.21 | prpplague | T0mW: ping |
| 15:23.40 | T0mW | prpplague: hey |
| 15:23.55 | prpplague | T0mW: hey, got the url for some pics/info of your zipit upgrade board? |
| 15:24.04 | prpplague | T0mW: didn't see them on your cyberiansoftware page |
| 15:24.35 | T0mW | http://www.openhardware.net/?title=ZipIt%20Addon%20Board%20mod&dir=zipit&file=zipitBoard.html |
| 15:24.59 | prpplague | thanks |
| 15:25.57 | T0mW | prpplague: the backlight works great |
| 15:26.17 | T0mW | pita to install it behind the lcd glass though |
| 15:26.23 | prpplague | hehe |
| 15:26.41 | prpplague | hehe just noticed you had some of elinux pages listed |
| 15:26.42 | prpplague | hehe |
| 15:27.16 | T0mW | prpplague: yeah, I'm bi-partisan |
| 15:27.30 | sorphin | agaffney: you remember many many years ago |
| 15:27.43 | sorphin | When Schnucks had VideoCart ? |
| 15:28.46 | agaffney | dunno...what is VideoCart? :P |
| 15:32.09 | sorphin | they had a well, bout 10" LCD on the cart, that was a computer, that talked over RF to the store, and there were IR sensors on the ceiling so it knew where it was in teh store |
| 15:33.22 | agaffney | nope, don't remember it |
| 15:36.02 | jbevren | I saw a similar tech on some tv show |
| 15:37.44 | prpplague | jbevren: hey |
| 15:37.50 | jbevren | yawn |
| 15:37.52 | jbevren | its 'friday' |
| 15:37.53 | jbevren | heh |
| 15:37.54 | prpplague | jbevren: should have time this weekend to do some pixter hacking |
| 15:38.00 | jbevren | cool |
| 15:38.02 | jbevren | I'll be away though |
| 15:45.03 | prpplague | T0mW: hey |
| 15:45.09 | prpplague | T0mW: you do much z80 stuff? |
| 15:45.27 | T0mW | prpplague: oh God no! Please, not that! |
| 15:45.37 | prpplague | T0mW: hehe |
| 15:45.53 | T0mW | prpplague: I haven't code on Z80 in a few years |
| 15:46.04 | T0mW | s/code/coded |
| 15:46.18 | prpplague | T0mW: i'm tinkering with the idea of building a msx system |
| 15:46.41 | T0mW | prpplague: I've written a lot of Z80 code, at one time |
| 15:46.55 | prpplague | T0mW: ever looks at the msx stuff? |
| 15:47.04 | T0mW | no, url? |
| 15:48.00 | prpplague | http://en.wikipedia.org/wiki/MSX |
| 15:49.02 | T0mW | prpplague: LOL! "a standard for home computers in the 1980s" Hahahahahah |
| 15:49.13 | T0mW | prpplague: there was no such thing!!! |
| 15:49.35 | T0mW | prpplague: the only "standard" may have been CP/M |
| 15:50.37 | prpplague | T0mW: standard hardware |
| 15:50.53 | T0mW | yeah, somebody's wet dream |
| 15:50.56 | prpplague | T0mW: it really was far more standardized than anything else at the time |
| 15:51.26 | prpplague | T0mW: 30+ computers all from different manufacturers all using the same cartridge and peripherals, i'd say that was pretty big for the time |
| 15:51.54 | T0mW | yeah, probably was. I just find it halarious that someone would even venture to say that there was a standard offering in the pre-IBM PC days |
| 15:52.13 | prpplague | T0mW: well it really was, just not in the US |
| 15:52.45 | prpplague | T0mW: which goes to show, you really need to break into the US market for tech if you want to get it standarized |
| 15:52.47 | T0mW | Zenith was pretty big then and then there was another.. cannot recall the name right now. |
| 15:53.23 | T0mW | TI had their odd ball system for sale, the TI 9900 ?? |
| 15:53.24 | prpplague | T0mW: but the key here was that the computers were manufactured by indepent companies |
| 15:53.28 | prpplague | T0mW: yea |
| 15:53.45 | T0mW | Then Tandy was big with their Trash-80 |
| 15:54.14 | T0mW | IBM / Compaq / Osborne really did the world a big service |
| 15:54.54 | prpplague | T0mW: exactly all of those companies had their own designs, own software, own standards |
| 15:55.08 | prpplague | T0mW: all the companies that did msx all shared the same basic design |
| 15:55.19 | prpplague | T0mW: you can use a msx cartridge on any of the msx computers |
| 15:57.00 | T0mW | prpplague: I built my first "home" computer out of parts from an industrial computer (photo typesetter). It ran CP/M, had a Qume printer, and I got WordStar 0.9 running on it. |
| 15:57.11 | prpplague | T0mW: hehe |
| 15:57.23 | prpplague | T0mW: i built a z80 system |
| 15:57.35 | T0mW | prpplague: along with Microsoft Basic and assembly language, then I discovered Small-C |
| 15:57.37 | prpplague | T0mW: mowed lawns all summer like a freakin mad mad |
| 15:57.52 | T0mW | prpplague: yeah, they weren't cheap back then! |
| 15:57.56 | prpplague | T0mW: hehe plugged it in a let the blue smoke out |
| 15:58.02 | T0mW | oh no! |
| 15:58.04 | prpplague | T0mW: had a cap in backwards |
| 15:58.07 | T0mW | LOL |
| 15:58.13 | T0mW | BOOOOM! |
| 15:58.15 | prpplague | T0mW: fortunately no major damage |
| 15:58.50 | T0mW | I've seen large caps go off like a small grenade and blow a hole in an enclosure |
| 15:58.53 | prpplague | T0mW: seriously though, i'm considering doing a home brew msx system |
| 15:59.02 | T0mW | hmmm |
| 15:59.11 | prpplague | T0mW: hehe you see those new, 10F caps on sparkfun? |
| 15:59.17 | T0mW | wow |
| 15:59.36 | T0mW | prpplague: it would run your wristwatch for a 'few' days, eh? |
| 15:59.41 | prpplague | T0mW: hehe yea |
| 15:59.53 | prpplague | T0mW: the memory mapper for the msx is totally fascinating |
| 16:00.05 | T0mW | banking logic |
| 16:00.05 | prpplague | T0mW: i'm thinking about trying it with some of the lpc series |
| 16:00.08 | prpplague | T0mW: yea |
| 16:00.20 | T0mW | I am so glad we now have MMUs |
| 16:00.23 | prpplague | hehe |
| 16:00.41 | T0mW | I have a board that I support that uses banked memory |
| 16:00.52 | prpplague | T0mW: yea well you really can't progress in your understanding of things till you have studied how we got where we are now |
| 16:01.27 | T0mW | 512K of SRAM organized as one 32K contiguous area and the rest is two 8K banked "windows" |
| 16:02.13 | prpplague | T0mW: cool |
| 16:02.29 | prpplague | T0mW: uzix uses that method to create a true multi-tasking os |
| 16:02.35 | prpplague | T0mW: totally fascinating |
| 16:02.37 | Pix | just like the good old EMS paging system under DOS ? |
| 16:02.45 | T0mW | prpplague: naw, it would be a good thing to play with. BTW, the 'new' Z80 chips from Zilog have the banked memory logic built into the CPU |
| 16:02.57 | T0mW | Z280 |
| 16:03.10 | prpplague | T0mW: yea thats what i've heard |
| 16:03.27 | prpplague | T0mW: i had ment to look and see what the configuration is |
| 16:03.36 | prpplague | T0mW: there is a version of uzi for the 280 |
| 16:04.12 | T0mW | prpplague: I had a develop system for the Z280 at one time, I prolly have it archived onto CD. It ran under msdos |
| 16:04.16 | T0mW | on the PC |
| 16:04.36 | T0mW | prpplague: you want me to look? |
| 16:05.02 | T0mW | I don't have the docs for it |
| 16:05.27 | prpplague | T0mW: naw, i'll grab it when i have the time |
| 16:09.06 | prpplague | T0mW: no the buildroot for the zipit is old old old |
| 16:09.25 | prpplague | T0mW: i'm on the buildroot dev team so if we want to get buildroot update for zipit we can |
| 16:10.10 | T0mW | prpplague: um, did it work at all? Maybe you can save me some time in setting my own up. |
| 16:10.30 | prpplague | T0mW: hehe |
| 16:10.31 | T0mW | prpplague: I plan to use an old old buildroot (TuxScreen) as the starting point. |
| 16:10.38 | prpplague | T0mW: no don;t do that |
| 16:10.42 | T0mW | no sense in using the new one |
| 16:10.48 | prpplague | T0mW: why? |
| 16:10.59 | T0mW | 2.6 kernel, later gcc, etc. |
| 16:11.07 | prpplague | T0mW: no |
| 16:11.12 | T0mW | I was planning on using 2.95 |
| 16:11.20 | T0mW | gcc 2.95 |
| 16:11.26 | prpplague | T0mW: why that version? |
| 16:12.05 | T0mW | Well, not sure, but it worked for Tux. something tells me that for the 2.4 kernel 2.95 was still relevant? |
| 16:12.52 | T0mW | as far as I'm concerned, gcc 3.x was problematic, 2.95 was pretty mature |
| 16:13.11 | prpplague | T0mW: yes some early 3.x was an issue |
| 16:13.20 | prpplague | T0mW: and 2.4 will compile with 3.x |
| 16:13.22 | T0mW | unless you went to gcc-3.4, but not sure if the 2.4 kernel could be compiled with that |
| 16:13.27 | T0mW | ok |
| 16:13.34 | prpplague | T0mW: no, 3.3.5 and 3.3.6 are very stable |
| 16:13.40 | T0mW | ok |
| 16:13.58 | prpplague | T0mW: personally i would recommend the latest buildroot |
| 16:14.02 | prpplague | T0mW: but its up to you |
| 16:14.19 | prpplague | T0mW: if you use the latest buildroot you can send patches to get zipit into the main build |
| 16:14.47 | T0mW | well, I would like to put the system together with a two stage boot (two kernels). One kernel acts as an initrd, the second located on the SD card would be loaded by the first then executed as the final linux running kernel |
| 16:15.16 | prpplague | T0mW: yea, easy enough |
| 16:15.18 | T0mW | prpplague: I know that is ambitious, but do-able |
| 16:16.00 | T0mW | prpplague: I cannot see building a system where you are going to base the filesystem on the SD drive and not allow it to have the runtime kernel on the drive. |
| 16:16.34 | T0mW | prpplague: using u-boot or whatever still means some kludge to do that, so, put the boot kernel into flash |
| 16:16.41 | T0mW | heh |
| 16:16.48 | T0mW | load once, boot many |
| 16:17.20 | T0mW | maybe, just maybe, have the boot kernel drop back to nfs if the drive is missing |
| 16:17.37 | prpplague | T0mW: yea, ya'll really need to get apex running on the zipit |
| 16:18.33 | T0mW | ~prpplague |
| 16:18.36 | purl | it has been said that prpplague is mailto:danders@amltd.com or Dave Anders |
| 16:22.36 | T0mW | prpplague: you have mail! |
| 16:23.04 | jbevren | prpplague: did you get a tiny kernel built for the pixter MM? |
| 16:23.20 | prpplague | jbevren: yea |
| 16:23.24 | jbevren | cool |
| 16:23.26 | prpplague | jbevren: but i was being quiet about it |
| 16:23.29 | jbevren | hehe |
| 16:23.55 | jbevren | shame I dont have an rs232 level shifter on my pixter yet |
| 16:23.56 | jbevren | <PROTECTED> |
| 16:23.57 | prpplague | jbevren: booted with only about 320k free |
| 16:24.03 | jbevren | haha doh! |
| 16:24.05 | T0mW | prpplague: you have more mail! |
| 16:24.11 | T0mW | heh |
| 16:24.16 | jbevren | ok my boss is retarded beyond reality |
| 16:24.22 | jbevren | he knows precisely how a raid rebuild works |
| 16:24.30 | prpplague | what the hell did you send me? |
| 16:24.34 | jbevren | yet he believes wholeheartedly that a rebuild wont cause a server's load to go high |
| 16:25.01 | jbevren | of course, same boss thinks that 4100 domains on a single harddisk isnt too many |
| 16:25.29 | prpplague | holy @#$@#$ |
| 16:25.35 | prpplague | 4100 domains? |
| 16:25.39 | T0mW | prpplague: a really maligant virus |
| 16:25.46 | T0mW | prpplague: a really malignant virus |
| 16:25.47 | jbevren | root@bmw [~]# wc -l /etc/userdomains |
| 16:25.47 | jbevren | 4089 /etc/userdomains |
| 16:25.51 | jbevren | close enough to 4100 |
| 16:26.11 | prpplague | jbevren: jeeze |
| 16:26.34 | prpplague | T0mW: interesting |
| 16:26.48 | T0mW | prpplague: not what I was looking for, but |
| 16:26.50 | jbevren | prpplague: said server's array is in rebuild |
| 16:27.05 | jbevren | this server does everything: mysql, pop3, smtp, webservices, imap, webmail |
| 16:27.17 | jbevren | on a mirror raid (equiv: one sata drive) |
| 16:27.42 | prpplague | jbevren: wtf? |
| 16:27.51 | *** part/#edev sorphin (n=dans@mewp.captainrock.com) |
| 16:27.58 | jbevren | prpplague: and you wonder why I get so pissy |
| 16:28.00 | prpplague | jbevren: he wants to do a rebuild on a mission crittical server ? |
| 16:28.05 | prpplague | jeeze |
| 16:28.08 | jbevren | we dont have a choice |
| 16:28.11 | T0mW | ok girls, I've got to duck out and go butalize my body for an hour. |
| 16:28.13 | jbevren | the array does it on its own and you cant turn it off |
| 16:28.17 | prpplague | T0mW: later |
| 16:28.19 | jbevren | -every- time you reboot |
| 16:28.22 | jbevren | which happens a -lot- |
| 16:28.28 | jbevren | because the load randomly shoots past 200 |
| 16:28.30 | prpplague | jbevren: jeeze |
| 16:28.38 | jbevren | now multiply this by 250 |
| 16:28.47 | jbevren | (rough guess on how many servers there are in this farm) |
| 16:29.30 | jbevren | lovely aint it? |
| 16:29.57 | jbevren | the pop3 and imap daemons log every login/logout through syslog, so every time someone logs in or out, maillog gets fsync()ed |
| 16:30.14 | jbevren | mind you, I've changed that setting |
| 16:34.10 | jbevren | rebuild only took 14 hours |
| 16:34.16 | jbevren | <PROTECTED> |
| 16:35.35 | *** join/#edev Chocobo (n=swinchen@thecheat.eece.maine.edu) |
| 16:35.55 | Chocobo | Ok, so I got a USB snoop log of the eZ430. hrmmmm..... |
| 16:36.10 | jbevren | good. get to writing a driver :) |
| 16:37.10 | Chocobo | lol. I have no idea how. I don't really understand the USB snoop file to be honest. |
| 16:38.26 | prpplague | Chocobo: ok here is how to understand usb snoop data |
| 16:38.31 | jbevren | :) |
| 16:38.31 | prpplague | Chocobo: save the log |
| 16:38.35 | prpplague | Chocobo: close the window |
| 16:38.42 | prpplague | Chocobo: go drop some acid |
| 16:38.43 | Chocobo | ok |
| 16:38.44 | jbevren | haha |
| 16:38.46 | prpplague | Chocobo: drink a few beers |
| 16:38.46 | Chocobo | lol |
| 16:38.51 | prpplague | Chocobo: then open the log |
| 16:38.58 | prpplague | Chocobo: it will make much more sense now |
| 16:39.07 | jbevren | then look at all the pretty veins, eyeballs, and broken fangs. |
| 16:39.15 | Chocobo | oh so THAT is what a TransferBufferMDL is |
| 16:41.25 | Chocobo | hrmmm I would be happy as a first step just making something that got a response from the device... it doesn't need to be elaborate... just make sure I am talking with it. |
| 16:45.59 | prpplague | i am totally amazed at the world cup histeria that goes on |
| 16:46.08 | jbevren | oh thats right |
| 16:46.13 | jbevren | world cup |
| 16:46.29 | jbevren | wait whats the world cup (lol just kidding) |
| 16:46.37 | prpplague | my sons school has the game on a big screen tv for the kids |
| 16:46.40 | prpplague | jeeze |
| 16:46.53 | CosmicPenguin | Anyone have scores? |
| 16:46.55 | prpplague | no school from noon to 2 for the game |
| 16:47.03 | prpplague | CosmicPenguin: tnt just scored against england |
| 16:47.23 | prpplague | CosmicPenguin: oh my bad, sorry tnt just tried to score |
| 16:47.27 | prpplague | CosmicPenguin: still null |
| 16:48.06 | CosmicPenguin | null null, or just null for tnt? |
| 16:48.12 | prpplague | null all |
| 16:49.03 | prpplague | every radio channel, every television station |
| 16:49.05 | prpplague | its insane |
| 16:49.21 | prpplague | i've never seen anything like it |
| 17:05.28 | jbevren | what |
| 17:05.28 | jbevren | the |
| 17:05.29 | jbevren | fuck |
| 17:05.32 | jbevren | pardon my language |
| 17:05.40 | jbevren | but why exactly would firefox NOT know what to do with a HTML file?! |
| 17:07.04 | jbevren | it does this randomly with jpeg, gif, text, and html files |
| 17:07.10 | jbevren | while its displaying them at the same time |
| 17:08.42 | jbevren | (/vent) |
| 17:11.18 | prpplague | jbevren: yea i've noticed that on some recent stuff |
| 17:14.28 | jbevren | its retarded |
| 17:14.47 | prpplague | hehe yea well so is ins and dohs |
| 17:14.59 | jbevren | CONCUR |
| 17:15.00 | jbevren | hehe |
| 17:15.19 | *** join/#edev leoncamel (n=leoncame@219.142.130.97) |
| 17:17.42 | Chocobo | wow. writing a USB driver looks impossible... lol |
| 17:18.01 | prpplague | Chocobo: not impossible, just difficult |
| 17:18.08 | prpplague | Chocobo: don't worry about a driver |
| 17:18.13 | prpplague | Chocobo: just look at using libusb |
| 17:18.26 | prpplague | Chocobo: can make it easier to do it from user space |
| 17:19.57 | Chocobo | Oh, maybe I don't a kernel space driver? that would be handy. |
| 17:23.14 | Chocobo | Wow, each chapter in the libusb documentation is like 1 sentence long :/ |
| 17:26.01 | prpplague | very very easy |
| 17:28.10 | Chocobo | It looks like it. What is the catch? |
| 17:30.00 | jbevren | isnt usb actually a very simple protocol |
| 17:30.07 | jbevren | (thus the need for advanced software stacks) |
| 17:35.27 | jbevren | http://www.mozilla.com/add-ons/jogacompanion/ |
| 17:37.20 | prpplague | jbevren: uh in theory, but in actuality its extremely complexe due to all the "exceptions" |
| 17:37.34 | prpplague | jbevren: if you read the specs it generally goes like this |
| 17:37.53 | Chocobo | struct usb_bus *usb_get_busses(void); <--- That stucture needs to be freed at the end of the program right? |
| 17:38.00 | jbevren | prpplague: nod |
| 17:38.01 | jbevren | so |
| 17:38.03 | prpplague | jbevren: "page 1. here is the way this procedure is required to operate" |
| 17:38.05 | jbevren | if malloc() is a void |
| 17:38.07 | jbevren | how do you get the pointer back? |
| 17:38.19 | jbevren | oh, c=*malloc()? |
| 17:38.24 | prpplague | jbevren: "page 2. if you don't like page1 you can basically do it the way you want to" |
| 17:39.36 | Chocobo | yeah, the function probably has a malloc in there. |
| 17:39.36 | jbevren | ok that gives me the same error |
| 17:41.00 | MonMotha | Chocobo: IIRC usb_get_busses returns a pointer to a global |
| 17:41.16 | jbevren | malloc() returns a pointer but youre not supposed to get it? |
| 17:41.17 | MonMotha | you can access the same struct using a global variable in unix, but that doesn't work on all OSes so they added usb_get_busses |
| 17:41.25 | jbevren | malloc.c:6: error: void value not ignored as it ought to be |
| 17:41.29 | Chocobo | hmm, I can free it without crashing the program... |
| 17:41.41 | MonMotha | Chocobo: you can free anything without immediately crashing many programs :) |
| 17:41.57 | MonMotha | is this just at program exit? |
| 17:42.14 | MonMotha | jbevren: malloc returnsp ointer to void |
| 17:42.29 | jbevren | so |
| 17:42.29 | MonMotha | jbevren: in other words, it returns a pointer to "a hunk of memory that you can do whatever you want with" |
| 17:42.34 | jbevren | right |
| 17:42.41 | jbevren | so how do I save this pointer? |
| 17:42.42 | MonMotha | the normal thing to do is immediately (or implicitly) typecast it to pointer to whatever type |
| 17:42.54 | jbevren | hmm |
| 17:42.57 | jbevren | so *c=malloc()? |
| 17:43.18 | MonMotha | so you have something ilke "struct foo *fooptr; fooptr = (struct foo *)malloc(sizeof(struct foo));" |
| 17:43.34 | prpplague | CosmicPenguin: england just scored |
| 17:43.35 | MonMotha | you can then access elements of the struct foo by using fooptr->fooelement |
| 17:43.41 | prpplague | CosmicPenguin: england 1 tnt0 |
| 17:44.35 | MonMotha | alternately, if you wanted a primitive array but didn't want it on the stack (you didn't know the size beforehand, it's huge, etc.): "char *array; array = (char *)malloc(n);" |
| 17:44.59 | MonMotha | where n is the size of the array [note: if you aren't making an array of char, you need sizeof(type*n)] |
| 17:45.10 | MonMotha | you can then access elements of the array with "array[i]" |
| 17:45.44 | MonMotha | if you want to be really awful, you can leave off the typecast from the return of malloc, it'll get implicitly typecasted for you (and you'll probably get a warning) |
| 17:46.30 | MonMotha | void * is C's method of handling the passing around of generic things, like unknown structs, arrays of an unknown type, etc. |
| 17:46.49 | MonMotha | you just have to make absolutely sure it gets typecasted back into the right thing, since there's no checking |
| 17:47.21 | MonMotha | you can (ab)use that to do some silly things, like treat 64bit integers as an 8 byte array |
| 17:50.04 | jbevren | why is it that I have a fixed swap cache of 200M when I have swap turned off? |
| 17:56.30 | Chocobo | I think that is my biggest complaint with C. When a function returns a pointer to an datatype, I never know if I should free it when I am done. |
| 17:56.32 | Chocobo | :( |
| 17:56.49 | jbevren | not just that |
| 17:56.57 | jbevren | but i still dont get how to save the pointer to a simple malloc() to a variable heh |
| 17:57.06 | jbevren | no wonder firefox has so many leaks ;-) |
| 17:57.21 | jbevren | brb |
| 17:59.22 | MonMotha | Chocobo: unfortunately, there's no real convention on that |
| 17:59.41 | MonMotha | usually, you either need to free it or call some other cleanup routine that will free it for you |
| 18:00.00 | MonMotha | of course, if you're at program exit anyway, the OS will clean up for you usually |
| 18:00.20 | Chocobo | But it could also return a pointer to a global stucture correct? |
| 18:00.27 | Chocobo | that is true. |
| 18:00.28 | MonMotha | the issue is that some functions return pointers to globals |
| 18:00.36 | MonMotha | which I believe is what usb_get_busses does |
| 18:00.49 | MonMotha | freeing those is usually a bad idea |
| 18:03.03 | Chocobo | Indeed. Now, I wonder if the global is dynamic or static? freeing someting that is static... I would think that would crash. |
| 18:03.46 | MonMotha | that's the issue. It may be created on the heap as a result of calling the libusb init routines |
| 18:04.04 | MonMotha | however, if that's the case, you probably are NOT required to free it |
| 18:04.11 | MonMotha | since the libusb deinit routines should clean all that up |
| 18:04.19 | MonMotha | freeing it yourself would result in a double free |
| 18:04.57 | MonMotha | either way, I'm guessing you would not be required to free it. Do the libusb docs say anything? |
| 18:06.36 | Chocobo | Nope. I guess that is a good indication that I do not need to free it. |
| 18:06.59 | MonMotha | given that I know that interface is also implemented as a global, I'm guessing you should not free it |
| 18:07.10 | MonMotha | you could go use the source, I guess |
| 18:11.49 | Chocobo | that is true |
| 18:12.09 | CosmicPenguin | you know what annoy's me? |
| 18:12.19 | CosmicPenguin | nice grammar, Jordan |
| 18:13.06 | CosmicPenguin | anyway, I don't like the capital C in "uClibc" |
| 18:13.07 | prpplague | CosmicPenguin: oh? |
| 18:13.07 | prpplague | CosmicPenguin: why? |
| 18:13.16 | prpplague | CosmicPenguin: actually it was originally "uCLIBC" |
| 18:13.26 | prpplague | CosmicPenguin: where the small "u" to represent micro |
| 18:13.37 | jbevren | micro controller (uC) |
| 18:13.53 | jbevren | !np |
| 18:14.25 | prpplague | jbevren: ?? |
| 18:15.05 | Chocobo | I think it should be uCclib |
| 18:15.20 | Chocobo | micro controller c library |
| 18:15.35 | jbevren | the !np is for another channel, ignore it |
| 18:15.36 | jbevren | hehe |
| 18:15.44 | Chocobo | even libc seems backwards |
| 18:15.59 | MonMotha | it's the traditional way of naming libraries on unix, though |
| 18:16.08 | MonMotha | libfoo would be the foo library |
| 18:16.17 | CosmicPenguin | right, I know what uC means |
| 18:16.18 | MonMotha | libc is therefore the c library |
| 18:16.22 | CosmicPenguin | its just counter unix |
| 18:16.38 | MonMotha | well, they just took the unix naming scheme and stuck uC on the front |
| 18:16.48 | MonMotha | I agree that naming the files that way is counter unix, but for a project name it's fine |
| 18:19.15 | MonMotha | you know, a good shampoo and conditioner will prevent split hairs |
| 18:20.59 | Chocobo | lol. I know, and I am fine with it... libc just sounds funny... like tipsy. "Last night I drank too much guiness and got libc" |
| 18:22.02 | T0mW | must be a dose of glibc |
| 18:30.01 | CosmicPenguin | ~cart ctor and dtor |
| 18:30.03 | purl | ACTION ties ctor and dtor into a shopping cart then leaves ctor and dtor on an interstate off ramp |
| 18:30.04 | *** join/#edev redblue (i=star@ppp158.108-253-207.mtl.mt.videotron.ca) |
| 18:30.21 | CosmicPenguin | heh |
| 18:30.27 | CosmicPenguin | thats funny |
| 18:30.33 | CosmicPenguin | ~cart me |
| 18:30.35 | purl | ACTION ties cosmicpenguin into a shopping cart then leaves cosmicpenguin on an interstate off ramp |
| 18:30.45 | CosmicPenguin | still - nice call, Timmy |
| 18:30.48 | CosmicPenguin | ~botsnack |
| 18:30.49 | purl | CosmicPenguin: aw, gee |
| 18:31.13 | CosmicPenguin | anyway, ctor and dtor can kiss my ass |
| 18:39.17 | flyback | man I love usb so much |
| 18:39.39 | flyback | for the hell of it I attached a working dvd-rom drive to my usb to ide cable and plugged it into a usb 1.1 port on a old PII 300 laptop |
| 18:39.47 | flyback | it's not perfect but it's pretty damn impressive |
| 18:39.57 | flyback | compared to the built in dvd-rom drive that doesn't allow dma and loses data at random |
| 18:39.57 | prpplague | flyback: why would it not work? |
| 18:40.05 | flyback | I think usb 1.1 is a bit too slow |
| 18:40.28 | flyback | also shit tends to defy laws of physics when it's my own stuff to blow up in my face |
| 18:40.41 | flyback | oops lang |
| 18:43.01 | MonMotha | prpplague: it's not a hard drive? Some of those USB to IDE adapters won't speak ATAPI |
| 18:43.10 | MonMotha | the new ones all seem to, at least |
| 18:43.36 | prpplague | MonMotha: ahh would think if you were to implement and ide adapter that i'd implement atapi as well |
| 18:43.42 | prpplague | just me iguess |
| 18:44.25 | jbevren | yawn heh |
| 18:44.30 | MonMotha | prpplague: you would though that people would do that...you would think |
| 18:44.34 | MonMotha | but you know what happens when you think |
| 18:44.47 | flyback | MonMotha |
| 18:44.52 | flyback | mine is a genesys chipset |
| 18:44.57 | flyback | appears to do full atapi |
| 18:45.10 | flyback | I tested a junk burner and it even canucked up the same as it does on ide :) |
| 18:45.26 | flyback | when you do c2 testing it finds bad sectors at random on each scan |
| 18:45.27 | flyback | :P |
| 18:45.35 | flyback | btc/top-g/iomagic trash |
| 18:46.37 | jbevren | gak |
| 18:46.46 | jbevren | if I dont waste bits, it maps 16Gbytes |
| 18:46.48 | jbevren | heh |
| 18:49.52 | flyback | MonMotha I will give you the spec file later |
| 18:54.25 | jbevren | for 256 bits more I could load an illegal opcode detect |
| 18:54.26 | jbevren | heh |
| 18:54.37 | jbevren | I'm currently using 128kbit per TLB |
| 18:54.45 | jbevren | wasteful yes/no? |
| 18:55.23 | jbevren | oh, 92kbit. I trimmed junk out |
| 18:56.07 | jbevren | 4096 entries, 24 bits per entry = 96kbit (there. corrected my math) |
| 18:56.16 | MonMotha | jbevren: and I thought my 22bit address space on a Z-80 was overkill |
| 18:56.20 | jbevren | hehe |
| 18:56.21 | jbevren | well, |
| 18:56.27 | jbevren | I wanted to map at least a 32bit space |
| 18:56.30 | jbevren | the 816 has a 24bit space |
| 18:56.42 | jbevren | if I only mapped 32 bits, I'd waste namespace |
| 18:56.54 | jbevren | the concept is to divide the cpu view into 4k segments a-la intel |
| 18:57.04 | jbevren | and page that into a physical memory space of at least 32 bits |
| 18:57.24 | jbevren | that gives me 4096 pages to map outward, and by filling bits, I get 22+12 bits of virtual space |
| 18:57.28 | MonMotha | I divided the CPU view into pages of 16kiB |
| 18:57.30 | jbevren | er 23+12 |
| 18:57.32 | MonMotha | so you only get 4 pages |
| 18:57.35 | jbevren | yeah |
| 18:57.44 | jbevren | I wanted a finer level of control |
| 18:57.46 | MonMotha | but it's simple enough to be implemented with a few 74 series |
| 18:57.54 | jbevren | since there's already an opensource multitasking OS for the 816 |
| 18:58.03 | MonMotha | ah |
| 18:58.12 | flyback | quick question folks |
| 18:58.15 | jbevren | sure flyback |
| 18:58.16 | flyback | sorry to bug about trival stuff |
| 18:58.24 | jbevren | just ask :P |
| 18:58.28 | MonMotha | my attempt here was mostly to be able to change interrupt handlers once the system was booted |
| 18:58.30 | flyback | you know bounce buffers to deal with isa dma limits |
| 18:58.47 | jbevren | MonMotha: that should be fairly trivial. look at how cp/m systems did it |
| 18:58.49 | flyback | that essentially destroys any cpu freeing function of dma does it not? |
| 18:59.03 | jbevren | flyback: it has some cpu utilization, |
| 18:59.06 | jbevren | but. |
| 18:59.18 | flyback | i thought it completely eliminates dma |
| 18:59.18 | jbevren | you can memcopy() the data instead of copying the data out of the IO device byte by byte |
| 18:59.23 | flyback | because the cpu has to intercept it all |
| 18:59.30 | jbevren | so basically, the device dma's data to the dma'able memory region |
| 18:59.38 | jbevren | then you use a cpu block move operation |
| 18:59.43 | flyback | oh it can still dma to a region |
| 18:59.49 | flyback | I didn't think it could even do that |
| 18:59.51 | jbevren | now compare this to a software loop to manually increment pointers and write to high memory |
| 18:59.55 | jbevren | yeah |
| 18:59.57 | jbevren | think um |
| 19:00.04 | jbevren | on an ISA system you can only DMA to 16megs |
| 19:00.08 | flyback | yes I kno :) |
| 19:00.13 | jbevren | above 16megs you'll need a bounce buffer as you describe |
| 19:00.13 | MonMotha | jbevren: basically it boots off of a page that happens to reside in ROM, then you can remap things so that RAM resides at the interrupt table |
| 19:00.22 | jbevren | so your hardware dma's to this buffer, and you copy in a tight loop to upper ram |
| 19:00.41 | jbevren | it's not as fast as direct dma, but its faster than straight PIO in many cases |
| 19:00.46 | jbevren | MonMotha: right |
| 19:00.54 | jbevren | MonMotha: many systems face that challenge.. m68k, z80, arm |
| 19:01.12 | jbevren | does the z80 indicate when its pulling a vector? |
| 19:01.26 | MonMotha | it can, depending on the IRQ mode |
| 19:01.32 | jbevren | say, |
| 19:01.35 | jbevren | on power-up reset |
| 19:01.35 | jbevren | :) |
| 19:01.40 | jbevren | or |
| 19:01.44 | jbevren | hmm you could use a counter |
| 19:01.51 | MonMotha | oh, well, I just have my MMU reset so that the interrupt vector just happens to be in ROM |
| 19:01.55 | jbevren | and map the rom to address 0 for the first 16 cycles or so |
| 19:02.23 | MonMotha | then you can reconfigure the page tables later to put whatever page you want wherever you want |
| 19:02.30 | MonMotha | so you can chuck a page that resides in RAM down at 0 |
| 19:02.33 | flyback | jbevren |
| 19:02.44 | flyback | so basically bounce buffers work liek a nic buffer or serial port buffer |
| 19:02.46 | MonMotha | just do it with interrupts off :) |
| 19:02.51 | jbevren | MonMotha: anyway, my project is to provide the level of virtualization you get on x86 |
| 19:02.52 | flyback | you still get soem cpu freeing time |
| 19:02.57 | flyback | but the cpu has to flush the buffer once in a while |
| 19:03.04 | jbevren | so that you dont have to worry about having 8M of free ram, but only in chunks of 16k or smaller :) |
| 19:03.14 | jbevren | flyback: yeah, sort of |
| 19:03.41 | jbevren | in the caqse of an ISA eqipped 486, |
| 19:03.42 | jbevren | you can dma to the bounce buffer from the isa card |
| 19:03.48 | jbevren | and then use the DMA controller to copy the data to upper memory |
| 19:03.54 | jbevren | cause the DMAC can access all ram but isa cannot |
| 19:03.55 | jbevren | <PROTECTED> |
| 19:04.08 | jbevren | hmm |
| 19:04.16 | MonMotha | mine is much more simple. I've got 64kB of physical DRAM, 64kB of physical ROM, and like 2kB of NVSRAM. The 16 bit address space of the z80 is extended to 22 via 4 8-bit latches accessable from IO space and selected via the upper 4 bits of the z80's address bus. |
| 19:04.26 | jbevren | seems silly since the DMAC handles it :) |
| 19:04.36 | MonMotha | so you write whatever you want to in the page register and that forms the upper 8 bits of 22 of the physical address |
| 19:04.39 | jbevren | MonMotha: indeed |
| 19:04.51 | jbevren | I tossed page faulting in for fun |
| 19:05.03 | MonMotha | it's simple, can be implemented with about 6 74 series chips (which I'm going for old school here) and pretty effective |
| 19:05.07 | jbevren | so you can map invalid ram and use virtual memory (a bit indicates whether a page is mapped to physmem or not) |
| 19:05.09 | MonMotha | ah, this won't do that |
| 19:05.10 | flyback | jbe |
| 19:05.17 | MonMotha | yeah, no mmap for you |
| 19:05.17 | flyback | isa only has 24 bit addressing |
| 19:05.27 | jbevren | flyback: you misunderstand ISA |
| 19:05.29 | jbevren | or rather, |
| 19:05.31 | jbevren | DMA on a PC |
| 19:05.44 | jbevren | when you do a DMA transfer, the isa card does -not- provide the memory address ;-) |
| 19:05.49 | jbevren | the dma controller does |
| 19:05.54 | jbevren | you only provide data |
| 19:06.08 | flyback | WHOA |
| 19:06.30 | jbevren | the dma controller clocks data out of your card, one word at a time, and writes it according to its internal registers |
| 19:06.50 | jbevren | for busmastering, youre limited to 16M |
| 19:06.53 | jbevren | for dma, this isn't true |
| 19:06.58 | flyback | uhh isn't bussmastering dma |
| 19:07.09 | jbevren | nope. Busmastering is a complete takeover of the system's ISA bus |
| 19:07.15 | jbevren | it disables the cpu and you do your work |
| 19:07.36 | jbevren | DMA however is handled by the chipset, where teh DMAC pulls data out of your card |
| 19:07.58 | flyback | jmp! |
| 19:07.58 | flyback | jmp! |
| 19:07.59 | jbevren | there are cases (example: amiga 600) where you can only dma 16megs, even with the dmac |
| 19:08.01 | flyback | asm will make ya |
| 19:08.02 | flyback | jmp! |
| 19:08.05 | flyback | dmac will make ya |
| 19:08.06 | flyback | jmp! |
| 19:08.13 | jbevren | hehe ok stop flooding :P |
| 19:08.19 | flyback | sorry it just kinda came to me |
| 19:08.46 | jbevren | hehe |
| 19:08.52 | jbevren | lda 0xBABE |
| 19:08.56 | jbevren | and 0xD00D |
| 19:09.01 | jbevren | jmp 0x2BED |
| 19:09.04 | jbevren | anyway |
| 19:09.12 | jbevren | olllld c64 asm joke |
| 19:09.18 | jbevren | probably older than the c64 :) |
| 19:09.22 | flyback | MonMotha |
| 19:09.24 | flyback | this is the one I have |
| 19:09.25 | flyback | http://www.geeks.com/details.asp?invtid=USB2IDE-N&cpc=SCH&srm=0 |
| 19:09.50 | jbevren | flyback: I want to get one of those some time |
| 19:10.15 | flyback | dude they rock |
| 19:10.20 | flyback | that one time you have a old hd on the shelf |
| 19:10.27 | flyback | and even usb 1.1 speed so you don't have to reboot to check it |
| 19:10.27 | jbevren | woah they have imacs for $75 |
| 19:10.31 | flyback | it's the $10 right there |
| 19:10.33 | flyback | worth it |
| 19:10.35 | flyback | paid in full |
| 19:10.47 | jbevren | imacs make nice RAS terminals |
| 19:10.50 | flyback | dude this PII 300 laptop is playing a dvd |
| 19:10.52 | jbevren | and packet throwers |
| 19:10.58 | flyback | usb 1.1 thru to dvd-rom |
| 19:11.03 | flyback | it's not exactely smooth but it works |
| 19:11.09 | jbevren | flyback: it likely has hardware acceleration :) |
| 19:11.20 | flyback | also the usb to ide cpu usage seems to be next to nill |
| 19:11.21 | flyback | I am surprised |
| 19:13.07 | flyback | there was a problem I think with this usb to ide bridge with > 32kb packets but linux has fixes in higher revisions of 2.4 and 2.6 |
| 19:13.08 | flyback | so np |
| 19:13.34 | MonMotha | I want hardware h.264 decoding |
| 19:13.38 | MonMotha | 1080p |
| 19:14.40 | flyback | wow they have usb to pcmcia/pci |
| 19:15.12 | jbevren | hehe |
| 19:15.13 | jbevren | that'd rock |
| 19:15.20 | jbevren | flyback: careful, |
| 19:15.25 | jbevren | your laptop might not have cardbus support |
| 19:15.29 | jbevren | make sure it does before getting one |
| 19:15.39 | jbevren | cardbus is basically pci over pcmcia |
| 19:15.41 | flyback | no dude the other way |
| 19:15.47 | jbevren | older cards are isa over pcmcia |
| 19:15.47 | flyback | attach pci to usb |
| 19:15.51 | flyback | I know |
| 19:15.55 | Chocobo | ~botsnack |
| 19:15.55 | purl | Chocobo: :) |
| 19:16.04 | jbevren | bah |
| 19:16.06 | jbevren | ~botsnack |
| 19:16.06 | purl | jbevren: thanks |
| 19:16.09 | jbevren | :) |
| 19:16.13 | jbevren | stupid bot |
| 19:16.16 | jbevren | whine when I give it a snack |
| 19:16.18 | jbevren | ~lart prul |
| 19:16.19 | MonMotha | ~botsmack |
| 19:16.20 | purl | OWW! |
| 19:16.33 | MonMotha | good bot :) |
| 19:16.38 | Chocobo | Who is prul? |
| 19:17.00 | jbevren | ~purl |
| 19:17.01 | purl | i heard purl is not a bot |
| 19:17.01 | jbevren | prl is a program |
| 19:17.06 | jbevren | =purl |
| 19:17.13 | Chocobo | ~botsmack |
| 19:17.14 | purl | OWW! |
| 19:17.30 | Chocobo | ahhh |
| 19:17.33 | jbevren | ok so back to useful chat |
| 19:17.36 | jbevren | or relatively so |
| 19:17.37 | MonMotha | purl: would ou do it for TWO Scooby snacks? |
| 19:17.46 | jbevren | 16G is a huge space |
| 19:17.49 | MonMotha | awww, no MegaHAL? :) |
| 19:17.53 | jbevren | but I then get the freedom to map large framebuffer spaces :) |
| 19:18.33 | MonMotha | yeah, I need to design a video device for my z80 as well |
| 19:18.48 | MonMotha | I've got some pages designated as memory mapped IO for such a thing |
| 19:19.21 | jbevren | nod |
| 19:20.54 | MonMotha | my objective is to build this thing as "old-school" as possible |
| 19:21.03 | MonMotha | all DIP parts, tons of 74 series holding it together |
| 19:21.30 | MonMotha | I think the most modern part in it is the 16550 UART :) |
| 19:25.43 | Chocobo | One of my cats is named perl. |
| 19:26.15 | jbevren | shrug |
| 19:26.28 | jbevren | the wdesignc implementation of the 65816 runs stable at 20mhz, 5v |
| 19:26.29 | Chocobo | shake |
| 19:26.30 | jbevren | heh |
| 19:26.37 | jbevren | its rated for 14mhz |
| 19:26.49 | jbevren | anyway, it also runs at 1.8, 2.5, and 3.3v |
| 19:27.00 | jbevren | so I'm just going to stick a pair on a 3.3v PLD and go |
| 19:27.43 | jbevren | stick the mmu core, basic IO, and a sdram interface in the PLD |
| 19:27.44 | jbevren | :) |
| 19:27.52 | jbevren | oh, and a bus arbiter since there'll be two cpu's |
| 19:28.27 | Chocobo | FPGA? |
| 19:28.36 | jbevren | most likely yeah |
| 19:28.59 | Chocobo | Are you using windows or linux? |
| 19:29.11 | jbevren | <PROTECTED> |
| 19:29.13 | jbevren | for? |
| 19:29.21 | jbevren | my ISE runs on windows |
| 19:29.25 | MonMotha | jbevren: building a DRAM controller out of 74 series is fun... :) |
| 19:29.25 | Chocobo | development platform. |
| 19:29.27 | jbevren | cause the linux installer is .. at best, horrible. |
| 19:29.36 | jbevren | for software though I prefer linux |
| 19:29.39 | MonMotha | jbevren: oh, you've tried to use the Xilinx linux software? |
| 19:29.48 | MonMotha | you know, the stuff linked against glibc from 1998 |
| 19:29.58 | jbevren | MonMotha: yeah, the cdrom software |
| 19:30.02 | jbevren | it attempts to write to the cd |
| 19:30.06 | MonMotha | haha |
| 19:30.07 | jbevren | and then bails because it cant |
| 19:30.11 | Chocobo | ISE? |
| 19:30.20 | MonMotha | I couldn't even get it to start loading because it was linked against a truely ancient version of glibc |
| 19:30.30 | MonMotha | to the point that DEBIAN didn't have a compatibility package, and I wasn't inclined to build one |
| 19:30.43 | jbevren | my cd's installer ran without problems |
| 19:30.49 | MonMotha | mine barfed with bad libraries |
| 19:30.52 | jbevren | but when I -finally- got something to install, it was broken |
| 19:30.59 | jbevren | and the ISE couldnt find its own tools |
| 19:31.13 | MonMotha | I know I had to install Maple 9 manually by extracting all the archives myself |
| 19:31.22 | MonMotha | their Java installer loaded, ran, and didn't do anything |
| 19:31.35 | MonMotha | I haven't tried MATLAB yet |
| 19:33.23 | jbevren | heh |
| 19:33.24 | Chocobo | mmmmmm Matlab. |
| 19:34.03 | Chocobo | Two of my professors are the authors of the "Mastering Matlab" book. |
| 19:35.17 | Chocobo | Guess that was a boring story. I think maybe it is time for me head home. |
| 19:37.26 | Chocobo | See you all later. |
| 19:37.37 | Chocobo | ~botsnack ~botsnack |
| 19:37.37 | purl | :), Chocobo |
| 19:57.57 | flyback | http://www.geeks.com/details.asp?invtid=BABYBLUE-TINIO-JP |
| 19:57.58 | flyback | damn tempting |
| 20:02.25 | flyback | guys |
| 20:02.26 | flyback | http://www.geeks.com/details.asp?invtid=G20911A |
| 20:04.33 | flyback | those can be hooked to pc with tv out |
| 20:04.43 | flyback | some can even be opened and accept vga with a few resisters |
| 20:13.43 | flyback | <flyback> http://www.qbik.ch/usb/devices/search_res.php?pattern=genesys |
| 20:13.43 | flyback | [16:12] <flyback> since the usb id is the same |
| 20:13.43 | flyback | [16:12] <flyback> how the hell do I tell which chip I have |
| 20:13.43 | flyback | [16:12] <flyback> the 611 the 702 or the 811 |
| 21:44.04 | *** join/#edev kaenat_ (n=ringlej@webadmin.keyvoice.com) |
| 21:54.38 | *** join/#edev GPSFan (n=kenm@12.10.255.248) |
| 21:54.38 | *** mode/#edev [+o GPSFan] by ChanServ |
| 22:55.19 | *** join/#edev FireEgl (i=Atlantic@Atlantica.US) [NETSPLIT VICTIM] |
| 23:59.34 | *** join/#edev dijenerate (n=dijenera@72.22.159.138) |