**** BEGIN LOGGING AT Mon Sep 19 02:59:56 2005 Sep 19 07:56:35 hello world Sep 19 07:56:52 anyone got a BDI-2000 here ? Sep 19 07:57:57 <[g2]> there may be 1 or 2 Sep 19 07:58:01 heh Sep 19 07:58:09 T0mW: you got yours in? Sep 19 07:58:26 I got mine a few weeks ago, today I'd like to get setup with gdb Sep 19 07:58:58 prpplague: I guess I'll follow what Marc did from his page Sep 19 08:01:10 ahh Sep 19 08:03:01 prpplague: been too busy to do anything but take it out of the box and look at it so far Sep 19 08:03:19 prpplague: freakin' customers want it NOW! Sep 19 08:03:22 heh Sep 19 08:03:43 hehe yea thats what i'm dealing with this morning Sep 19 08:05:36 prpplague: I'll hit you up for that MMC code some time when you can Sep 19 08:08:40 I had one lying around for a couple of months - pretty easy to use Sep 19 08:08:52 T0mW: i firgot which code you wanted Sep 19 08:09:07 T0mW: are you wanting code you can use with the kernel or bootloader/ Sep 19 08:12:29 prpplague: bootloader, I just need to do raw sector read / write stuff Sep 19 08:13:28 wish that the LPC family of processors had less Flash and a lot more RAM, then I'd just boot my code from the MMC Sep 19 08:14:42 T0mW: ok, you need it in c or asm? Sep 19 08:16:06 prpplague: C, if you got some Sep 19 08:16:37 prpplague: I won't be booting with it, just accessing the media from runtime code Sep 19 08:16:54 s/runtime/application/ Sep 19 08:17:28 hmm, crosstool is up to version 0.38 now Sep 19 08:18:35 T0mW: i don't have any fs code support just function calls for raw sector read.write Sep 19 08:20:05 cool, that is what I need: primatives Sep 19 08:20:31 prpplague: I can share back fs stuff if you want. I have to dev full file read / write. Sep 19 08:21:07 open / create file, write to file, close, delete file, etc. Sep 19 08:21:18 basically build my own msdos Sep 19 08:21:55 T0mW: its pretty simple stuff, i would expect so as thats the direction i was headed Sep 19 08:22:01 T0mW: custom bootloader? Sep 19 08:23:43 prpplague: no, the application I am writing will need to some data logging Sep 19 08:24:14 prpplague: also, I would like to build in the future ability to autorun stuff on the MMC. Sep 19 08:25:11 prpplague: like if you have to update Flash code in a number of units, you can prep an MMC such that when the system sees the autorun file, it will eval it and update the flash code store. Sep 19 08:25:49 T0mW: right, that was my goal in the bootloader code i have Sep 19 08:26:00 T0mW: right now its just reading raw code from the mmc/sd Sep 19 08:26:04 prpplague: this way, if you have 30..40 units to update / program, you can just do it from a single MMC rather than have to download from a laptop to each unit (PITA). Sep 19 08:26:24 T0mW: i.e. (dd if=zImage of=/dev/sda) Sep 19 08:26:30 prpplague: heh Sep 19 08:26:40 * prpplague is fs challenged Sep 19 08:26:59 prpplague: I've written disk operating system stuff before Sep 19 08:27:33 T0mW: i can;t seem to make the associations i need to understand FAT16 Sep 19 08:27:36 prpplague: although, not as ambitious as this project is turning out to be Sep 19 08:28:38 T0mW: i've been swamped with customer crap for almost 2 months Sep 19 08:28:50 T0mW: my time for dev has been pretty small lately Sep 19 08:28:50 prpplague: yeah, the fat fs takes some study before it starts to make sense. it is similar to the CP/M file structure in some ways. CP/M used a bitmap of the allocations, fat uses a numerical table Sep 19 08:30:11 prpplague: lots of dev work headed my way lately. A job I quoted last year, and I quoted it really high to make it "go away", is back. They are still interested, not at the price I told them, but they still want it. Sep 19 08:31:09 prpplague: apparently they asked around and found nobody else intereseted in bailing them out of the fire (Municipal Contract with performance clause). Sep 19 08:31:47 ok. I'm going with crosstool-0.38 and hope for the best here. Sep 19 08:33:42 T0mW: i can't seem to get the conversion between block reads, and the c/h/s address Sep 19 08:35:05 prpplague: I know I did that with the compact flash bootloader, just don't recall the process right now. Sep 19 08:35:50 there is a physical block 0 and a logical block 0, fat is based upon the logical block 0 Sep 19 08:36:23 you have to offset to the logical block 0 against the reserved block amount Sep 19 08:36:39 ... to get to the logical block Sep 19 08:37:06 I don't recall where I picked up the data on the reserved sectors though Sep 19 08:37:24 it might have been from the partition table Sep 19 08:38:03 or from the drives' GET_GEOMETRY Sep 19 08:39:03 prpplague: I'll deal with that later, once I get the primative sector read / write working. Sep 19 08:42:01 prpplague: Dave, the fat file system can Sep 19 08:42:28 prpplague: Dave, the fat file system can't be that hard. After all, Bill Gates *cough*stole*cough* the code for it. Sep 19 09:02:08 prpplague: you have to remember that the FAT entries represent groupings of sectors: allocations. IIRC, the msdos boot block contains the mapping info of how many bytes/sectors to each allocation Sep 19 09:03:33 prpplague: STENCIL IS HERE!!! Sep 19 09:15:26 ep1220 Ping Sep 19 09:15:52 T0mW: hehe yea, thats what ppl keep saying, but i've yet to find a good example where someone walks you through the process Sep 19 09:16:19 ep1220: ping Sep 19 09:16:53 ka6sox: Destination unreachable Sep 19 09:16:55 heh Sep 19 09:17:42 rrrrright Sep 19 09:18:07 prpplague: I just wish I could find that old source to "Disk Doctor" that I wrote 18 years ago. That followed the chains to access files Sep 19 09:18:34 prpplague: moral? don't throw old code away. :-( Sep 19 09:19:03 prpplague: or, 5 inch media Sep 19 09:19:25 hehe yea Sep 19 09:20:15 * T0mW pulls out cdrom "Archive of Floppy Disks - volume 1" Sep 19 09:20:40 prpplague: heh, need TurboC v1.5? Sep 19 09:22:33 uh no Sep 19 09:24:38 heh "Clipper Summer '87' Sep 19 09:37:24 T0mW: ug, one of my wire transfers from the states is limbo as it is lost somewhere in new orleans Sep 19 09:45:10 prpplague: heh Sep 19 09:51:32 T0mW: dealing with the money transfers is the worst part of living her Sep 19 09:51:34 here Sep 19 13:37:41 <[g2]> beewoolie, ping Sep 19 13:37:48 Hey Sep 19 13:37:57 <[g2]> did you get those PMs ? Sep 19 13:39:10 <[g2]> beewoolie did you get the PMs ? Sep 19 13:39:20 By email? Sep 19 13:39:31 <[g2]> I've been messaging you Sep 19 13:39:44 I can see a dialog window. Sep 19 13:39:52 Are you talking about IM? Sep 19 13:39:57 <[g2]> IRC Sep 19 13:40:14 I may need to enable something...I'm not seeing anythingspecial. Sep 19 13:45:21 <[g2]> type /query [g2] Sep 19 13:45:24 <[g2]> type /query [g2] foo Sep 19 13:47:01 <[g2]> WOW kas11 is in the house Sep 19 13:47:17 hiya [g2] :) Sep 19 13:47:18 <[g2]> kas11 and beewoolie at the same time too special Sep 19 13:47:23 <[g2]> hey kas11 how goes ? Sep 19 13:47:31 Hey kas11! Sep 19 13:47:52 can't complain...nobody pays attn if I do anyhow ;) Sep 19 13:48:07 [g2]: Nothing from the query Sep 19 13:48:14 hiya beewoolie ;) Sep 19 13:48:34 <[g2]> you must have Private Messaging turn off Sep 19 13:48:35 Tries BE Debian yet? Sep 19 13:48:46 Let me see if I can find that... Sep 19 13:48:56 <[g2]> kas11 how are you doing ? what's up ? Sep 19 13:49:52 [g2] doing fine here...and nothing much going on at all Sep 19 13:50:13 <[g2]> well it's very nice of you to stop by Sep 19 13:50:19 [g2] got any new toys? Sep 19 13:50:25 <[g2]> OH yeah Sep 19 13:50:41 <[g2]> I've launched my hw company Sep 19 13:50:49 He's been trying to sharephotos with me, but now he complains that my IRC client is deficient. Sep 19 13:51:02 <[g2]> beewoolie can't PM Sep 19 13:51:09 CAN SO! Sep 19 13:51:11 I been pndering a X2 but they are still priced above my threshhold of pain Sep 19 13:51:17 <[g2]> then PM me Sep 19 13:51:22 grr Sep 19 13:51:40 <[g2]> X2 ? Sep 19 13:52:00 <[g2]> kas11 see http://www.giantshoulderinc.com Sep 19 13:52:00 Athlon 64 dualie Sep 19 13:52:04 <[g2]> AH... Sep 19 13:52:13 <[g2]> yeah they look sweet Sep 19 13:52:38 <[g2]> kas11 tell me how you like the pics Sep 19 13:53:59 beewoolie: Would it be possible to switch out ppcboot with apex on the ds101g+? Sep 19 13:54:23 <[g2]> NAiL APEX is an ARM boot loader not PPC Sep 19 13:54:25 NAiL: after some work. APEX has ARM code. Sep 19 13:54:31 oh Sep 19 13:54:40 <[g2]> just run DINK32 Sep 19 13:54:41 I don't know how hard it will be to support PPC. Sep 19 13:54:49 DINK32? Sep 19 13:54:57 ARM has some very slick features that make the bootloader easy to write. Sep 19 13:55:00 <[g2]> yeah the PPC boot loader from Moto Sep 19 13:55:19 aha Sep 19 13:58:03 Anyone got any experience with u-boot btw? Sep 19 13:58:09 I've used it. Sep 19 13:58:18 I wrote APEX because of it. Sep 19 13:58:38 heh Sep 19 13:58:52 doesn't exactly sound promising :P Sep 19 13:58:56 It *does* work. It's just not very...comfortable. Sep 19 14:00:31 heh, i've seen worse Sep 19 14:00:37 No doubt. Sep 19 14:01:06 <[g2]> beewoolie did you get the e-mail ? Sep 19 14:01:08 As I said, it does work. It can be challenging to figure out how to get it to do some things. don't really reember what Sep 19 14:01:11 ...checking Sep 19 14:01:47 It doesn't matter that much, I was mostly wondering because I was hoping to get ds101 and ds101g+ to be as close as possible firmware-wise Sep 19 14:01:56 I don't have a g+ (yet) Sep 19 14:02:18 and u-boot doesn't appear to support the NIC as of yet either Sep 19 14:02:56 Hey! I've got Debian on the slug! Most excellent! Sep 19 14:03:02 free Sep 19 14:03:06 ...oops? Sep 19 14:03:20 free gives you an oops? Sep 19 14:03:35 tee hee. Wrong window. Sep 19 14:04:00 beewoolie: but i'm getting oopses in the other window :) Sep 19 14:04:06 <[g2]> beewoolie congrats ! Sep 19 14:04:28 The LED is blinking.. Interesting feature. Sep 19 14:05:26 <[g2]> beewoolie didn't read the wiki :) Sep 19 14:05:38 No. I am reading the wiki Sep 19 14:05:50 I'm still intrigued. Sep 19 14:05:59 <[g2]> well it tells how to turn it off :) Sep 19 14:06:01 BTW, it says "play with leds". Sep 19 14:06:15 ...experimenting... Sep 19 14:06:20 this could be interesting: http://groups.yahoo.com/group/nslu2-developers/message/39 Sep 19 14:06:45 ByronT: I'm testing the patch now Sep 19 14:07:06 I want to see how it handles hubs first Sep 19 14:07:13 Nice. Sep 19 14:07:14 well I saw the LED discussion and I thought I'd post that URL Sep 19 14:07:33 Someone told me there is a 2.0 IAL. Anyone looked at it? Sep 19 14:07:33 yeah, it's something I've been waiting for ;) Sep 19 14:07:41 beewoolie: #nslu2-linux Sep 19 14:07:50 beewoolie: It's been the main topic for a day or two Sep 19 14:08:02 Oh. Been gone for a couple of weeks. Sep 19 14:08:04 (If I've understood the discussion) Sep 19 15:02:15 Arr! Sep 19 15:02:44 heh Sep 19 15:02:48 morning dyoung Sep 19 15:04:33 arrr! Sep 19 23:22:44 ka6sox: ping Sep 19 23:23:11 <[g2]> ep1220 hey! Sep 19 23:23:27 [g2]: hey! Sep 19 23:23:41 a late night for You ? :-) Sep 19 23:23:47 <[g2]> yeah :) Sep 19 23:23:53 <[g2]> getting ready for some sleep Sep 19 23:24:07 <[g2]> I got my first proto board last friday Sep 19 23:24:19 Found in the log Yu started Your HW company Sep 19 23:24:32 congrats Sep 19 23:24:34 <[g2]> yep Sep 19 23:24:36 <[g2]> thx Sep 19 23:24:39 Proto working OK ? Sep 19 23:24:48 <[g2]> it's rock'n! Sep 19 23:24:53 great Sep 19 23:25:15 I am doing a PCB for the FTDI Sep 19 23:25:23 <[g2]> I was booting to Debian on CF with a supplied kernel on the first day Sep 19 23:25:51 so did not even need JTAG ? Sep 19 23:25:51 <[g2]> the unofficial sarge be port Sep 19 23:26:22 <[g2]> well when I start playing with boot loaders Sep 19 23:26:38 <[g2]> JTAG would be nice for debugging Sep 19 23:28:00 <[g2]> but the board has been wrung out pretty well to start with Sep 19 23:28:33 <[g2]> Serial, both ethers, MiniPCI, CF all are working Sep 19 23:28:43 <[g2]> I haven't tried POE Sep 19 23:28:55 sounds very good Sep 19 23:29:18 <[g2]> My 2.6 kernel port is working, but I've got to forward port the CF/IDE driver from 2.4 Sep 19 23:29:38 <[g2]> however, I've booted to an Openslug jffs2 Sep 19 23:30:18 <[g2]> so tell me about the PCB ? :) Sep 19 23:30:48 It will be for the DLP2232M module. So it is easy to solder Sep 19 23:31:11 using levelshifters 1.2 to 3.3V Sep 19 23:31:11 <[g2]> ah.. those are like $50 ? Sep 19 23:31:17 39$ Sep 19 23:31:34 plus serial at 3.3V Sep 19 23:31:40 <[g2]> cool... Sep 19 23:31:53 <[g2]> not 5v ? Sep 19 23:32:02 <[g2]> or RS232 Sep 19 23:32:09 JTAG at 5V is also possible, but would need a transistor Sep 19 23:32:23 <[g2]> I think 3.3 for JTAG is fine Sep 19 23:32:35 For ARM, yes Sep 19 23:32:35 <[g2]> 1.2 to 3.3 range ? Sep 19 23:32:39 yes Sep 19 23:32:54 but PPC and some DSPs also use lower voltages Sep 19 23:33:01 <[g2]> excellent Sep 19 23:33:12 <[g2]> when will protos be ready ? Sep 19 23:33:44 Currently I am thinking which JTAG connectors to add Sep 19 23:33:57 ka6sox suggeste 14 and 20 pin ARM + Digilent Sep 19 23:34:13 <[g2]> sounds great to me :) Sep 19 23:34:25 personally I do not like having too many connectors on the board Sep 19 23:34:31 <[g2]> will the pins be configurable at all ? Sep 19 23:34:45 Not by SW. Sep 19 23:34:58 <[g2]> by hw ? Sep 19 23:35:21 Currently I think using a 20pin "universal" connector plus external PCBs for adapters. Sep 19 23:35:55 I want to support ARM,MIPS,PPC,FPGA, TI And AD DSPs, evt. AVR Sep 19 23:36:10 <[g2]> ep1220, JTAG king! Sep 19 23:36:36 the proto should be ready sometimes next week. Sep 19 23:36:51 <[g2]> so you going to put macraigor outta business ? :) Sep 19 23:37:37 <[g2]> pls keep me posted on the status Sep 19 23:37:41 not so likely, unless some OS friends write all the needes SW :-) Sep 19 23:37:49 needed Sep 19 23:38:13 <[g2]> I'm quite interested in Xscale :) Sep 19 23:38:19 I'll do this Sep 19 23:39:21 if you want to pick up some work there, let me know ;-) Sep 19 23:39:38 <[g2]> I missed that Sep 19 23:40:40 A fast FLASH progger for the XSCALE is on top of the TODO list Sep 19 23:41:18 <[g2]> ah.. a fast flash programmer Sep 19 23:41:57 <[g2]> the work was on the fast flash programmer ? Sep 19 23:42:05 yes Sep 19 23:43:01 <[g2]> actually, I think the fastest is writting to memory and doing it in sw Sep 19 23:43:16 What we said some time ago: Put a stub into the minicache, send data to it, let the XScale do the programming Sep 19 23:43:25 <[g2]> right. Sep 19 23:43:26 Did not mean a HW programmer Sep 19 23:43:56 <[g2]> no the hw programmer feeds the mini/I/cache and starts the program Sep 19 23:44:49 such a soft would also be helpfull for people who can afford wiggler type JTAG only Sep 19 23:44:50 <[g2]> I think we're saying the same thing Sep 19 23:44:54 yes Sep 19 23:45:19 <[g2]> and possibly have serial available Sep 19 23:45:50 serial for what ? Sep 19 23:45:53 <[g2]> I'll be really interested in the debugging and performance monitoring abilities Sep 19 23:46:03 <[g2]> input/output feedback Sep 19 23:46:15 I see Sep 19 23:46:20 <[g2]> although maybe that can be brought back through the JTAG Sep 19 23:46:38 <[g2]> it depends on the host appliation Sep 19 23:47:01 rigth and likely at higher speed (on JTAG) Sep 19 23:47:28 on the other side: once You have RAM running Sep 19 23:47:34 and EThernet Sep 19 23:47:55 debugging on the system itself should be more comfortable Sep 19 23:48:04 esp. for profiling Sep 19 23:48:13 <[g2]> it's to capture the hangs Sep 19 23:48:33 <[g2]> break-in and see what was happening or where things are looping Sep 19 23:48:56 yeah, there JTAG has a big advantage Sep 19 23:49:43 IIRC your Loft has a 14pin JTAG header ? Sep 19 23:49:58 <[g2]> nod and there's some extra pins there Sep 19 23:50:45 <[g2]> that's why I asked about the configuration earlier Sep 19 23:51:10 <[g2]> ok it's 14-pins just some are used for something else :) Sep 19 23:51:15 <[g2]> no "extra" pins Sep 19 23:51:24 <[g2]> cause then it wouldn't be 14 Sep 19 23:51:34 <[g2]> it'd be > 14 :) Sep 19 23:51:37 that was already puzzling me :-) Sep 20 00:01:21 <[g2]> nite **** ENDING LOGGING AT Tue Sep 20 02:59:56 2005