| 00:03:25 | TomW | Russ: Tim "under the knife" ? |
| 00:03:45 | file | surgery? |
| 00:03:54 | TomW | dunno |
| 00:04:57 | Russ|werk | cyst on his wrist or something |
| 00:05:05 | file | oh... |
| 00:05:41 | file | uh huh :| |
| 00:08:10 | file | anybody alive? |
| 00:09:32 | file | yellow? |
| 00:11:14 | jacques | ? |
| 00:12:07 | file | just checking |
| 00:19:39 | file | erm |
| 00:27:39 | as_ | Tom, GP23, what normally uses that line? I cant see IC802 in any of the schematics. |
| 00:27:54 | file | GP32? from Gamepark? handheld gaming device? |
| 00:28:25 | as_ | No, GPIO23 ;] |
| 00:28:48 | file | ahhhh |
| 00:30:20 | as_ | Its been wired as an OC line, which should mean devices could share it. Tim has cut the wire so I'm wondering what its connected too. |
| 00:31:16 | file | uh huh |
| 00:31:41 | as_ | Err, Tom not Tim. |
| 00:32:21 | file | confusing eh? |
| 00:32:30 | as_ | thinks, anyone asks I'll blame it on the time differential. |
| 00:32:58 | file | or... the disk or the processor is on fire? |
| 00:33:12 | file | Big to little endian conversion error? |
| 00:33:26 | file | Forced to support NT servers; sysadmins quit. |
| 00:33:40 | file | The Token fell out of the ring. Can you tell us where to find it? |
| 00:34:13 | as_ | I need to have my cupholder repaired. |
| 00:42:08 | jacques | cd/dvd drives are so fragile |
| 00:43:12 | as_ | They are very high performance though, and fit such a wide range of cup sizes. |
| 00:43:44 | file | your... unique... |
| 00:44:15 | TomW | as_: GP23 is attached to a 74125 gate which indicates that something is attached to J602. They must have used this in Production to signal some test software in the Flash that a test set of some kind was attached to the J602 connector. |
| 00:45:47 | file | TomW: You know too much... you must be killed :) |
| 00:46:06 | as_ | The 125 gate is wired as OC/OD with the input coming from IC802.9 What is that chip? |
| 00:46:52 | as_ | Also, you say you need to add a pullup, but since its wired OC it should allready have one... |
| 00:49:45 | TomW | as_: BUT, the 74125 has a drive spec of 50millamps of short circuit current, if you don't cut that pin loose, you will eventually burn ourn out the chip by overloading it by shorting it to ground. It is not good design practice to do what you suggested. |
| 00:50:01 | TomW | s/ourn/out/ |
| 00:51:04 | as2 | Re. |
| 00:51:14 | as2 | <as_> Also, you say you need to add a pullup, but since its wired OC it should allready have one... |
| 00:54:36 | TomW | ?? OC, explain. |
| 00:55:12 | as2 | Open Collector / Open Drain. |
| 00:55:22 | TomW | Open Collector means just that, no internal pullup resistor is inside the device. |
| 00:55:53 | as2 | It can only sink current, not source it. |
| 00:56:03 | as2 | Which is how the 125 is wired. |
| 00:56:35 | as2 | The only logical conclusion is the line is designed for multiple devices and allready has a passive pull up.. |
| 00:57:59 | TomW | Sorry Old Man, you better look carefully at your TTL data book again, the 74LVC125 is a Totem Pole driver, it sources & sinks 12ma of current. |
| 00:58:40 | as2 | The input from IC802 goes to the gate enable, right? |
| 00:58:52 | TomW | Correct. |
| 00:58:57 | as2 | And the input is grounded. |
| 00:59:31 | as2 | So the device cannot source current. |
| 00:59:52 | as2 | The two states for the device are now Ov and HighZ. |
| 01:00:15 | as2 | Controlled by the gate enable. |
| 01:00:16 | TomW | Ok, let me explain: yes, I could have just cut the gate control lead then grounded it, that involved a cut + a jumper. You are currect, it could have been done that way. |
| 01:01:07 | as2 | Not what I mean, is 802 actually using this line for anything? |
| 01:01:38 | as2 | I'm wondering if it is used for anything, if the driver could be rewritten to serve more than one device. |
| 01:02:08 | TomW | However, I is more economical (in terms of aggravation) to just cut the output itself so it didn't drive anything, this way no jumper was needed. Less likely hood of an inexperienced person shorting out pins soldering a jumper in place that is not necessary (when you only cut the output pin loose). |
| 01:02:53 | as2 | I'm wondering if it needs to be cut at all. |
| 01:03:19 | as2 | And why there isnt a pullup allready on that line. |
| 01:04:25 | as2 | Or is there and its connected to pin 3 before the track you cut? |
| 01:04:45 | TomW | There is a 100K pullup on the board, but that is not enough current to "snap" the line Hi after the 2N7000 pulls it to ground. Without a stiffer pullup on that line, it could be possible to miss an interrupt as the line starts to go Hi (after being pulled low), the IDE then issues another succesive interrupt repuest, but the ARM processor doesn't hear it as the line has not yet gone invalid (high). |
| 01:05:38 | as2 | Ahh, its too slow. |
| 01:05:55 | as2 | What is IC802 though? |
| 01:07:26 | TomW | as2: As GP24..GP27 are used by the PCMCIA circuitry to indicate that a card has been inserted, the only pin available was GP23 which is, apparently, an indicator (interrupt) line to tell the processor that something is attached to J602. Since J602 is not used in the Tux, the GP23 is meaningless. GP23 only goes to J602. |
| 01:08:01 | TomW | IC802 is irrelevant. |
| 01:08:11 | as2 | Why irrelevent? |
| 01:08:24 | as2 | Its wired to use GP23. |
| 01:08:27 | TomW | oh, wait, IC802 is buffering the signals to / from J602. |
| 01:09:28 | TomW | as2: Walk the logic through, you will see how it works. |
| 01:10:01 | as2 | Ive tried to find IC802, but I cant see it anywhere. What schematic is it on? |
| 01:10:54 | TomW | a LOW on J602.34 enables the databus I/O port via U403, all IC802 is is a buffer, it keeps static electricity from doing nasty things to the rest of the PCB. |
| 01:12:36 | TomW | as2: http://www.tuxscreen.net/download/schematics/PcmCTL.pdf |
| 01:14:08 | as2 | Looking. |
| 01:14:36 | TomW | as2: That schematic is incorrect though: U500.4 goes to U403.48 and U402.46 goes to U403.1 |
| 01:16:34 | as2 | I must be out of memory, damn things printed with no text. |
| 01:23:17 | as2 | This makes more sense. |
| 01:24:39 | as2 | I dont see any reason you need to make the cut on that line though. Presumably the 125 isnt sinking enough current so you used the transistor? |
| 01:26:02 | as2 | Forget that, I see the OE now. |
| 02:07:14 | GPSfan_ | TomW: http://www.tuxscreen.net/download/schematics/PcmCTL.pdf has been corrected to reflect your feedback. |
| 02:07:43 | TomW | GPSfan_: thanks, nice to have schematics to work off of. |
| 02:07:56 | GPSfan_ | TomW: actually it was the buffer schematic. |
| 02:08:03 | GPSfan_ | taht was fixed. |
| 02:08:31 | TomW | I'm still tinkering, trying to get an fdisk working. |
| 02:08:59 | TomW | I have the netwrinder sources, but they are asking for ncurses. |
| 02:09:14 | TomW | s/netwrinder/netwinder/ |
| 02:09:43 | GPSfan_ | TomW: have you done any tests with a PCMCIA card in place? with J602 34 grounded, the addresses to the sockets are always live. That might cause some hot insertion problems. |
| 02:10:11 | TomW | ... got to get a few more people set up with IDE drives in their tux's then maybe some hard drive utils will start appearing. |
| 02:10:20 | pmisfit | TomW: ncurses *sort of* works. the |
| 02:10:35 | pmisfit | the "make install" target is still fubar'ed. |
| 02:10:37 | TomW | GPSfan_: yes, I have my ethernet card inserted and it works fine. |
| 02:10:49 | TomW | ... both slots. |
| 02:11:07 | TomW | pmisfit: yes, I see that. also wondering about BUILD_CC |
| 02:11:48 | pmisfit | man, i went around in circles on that for days. must be something obvious, but i'm afraid if i touch it i'll lose a day again. :-/ |
| 02:11:49 | GPSfan_ | TomW: good. since the addresses are outputs from the socket, there shouldn't be any "open input" issues, |
| 02:12:29 | TomW | pmisfit: once you get the TARGET done, they work? |
| 02:13:05 | pmisfit | oh yes. especially frustrating since in my OLD build system it worked, in the new build system it doesn't, and i can't see anything different. weird. |
| 02:13:12 | pmisfit | but once you get it built it works swimmingly well. |
| 02:13:33 | TomW | GPSfan_: still cannot fathom why J602 was a READONLY data port. seems silly to me. |
| 02:14:30 | pmisfit | might be a namespace collision or some such thing. if you really need it, i can have another go at it. otherwise maybe try "slab" |
| 02:14:42 | GPSfan_ | TomW: I think that Lucent only needed a ReadOnly port for their FTE, and screwed up the logic, so didn't fix it when writes didn't work. |
| 02:15:36 | pmisfit | TomW: be aware tho that ncurses is portly. another reason to maybe try slab. |
| 02:16:06 | TomW | pmisfit: I know what you mean, I talked to myself for about 4 days on how the IDE interface would operate on the hardware & driver level before I even designed the board. Then spent tuo days doping out the ide driver. |
| 02:16:26 | TomW | GPSfan_: FTE? Field Test Equip? |
| 02:16:49 | pmisfit | is away: putting girlie to bed... |
| 02:17:35 | GPSfan_ | TomW: looking at the CPU board of the Tux, it doesn't look "finished" the I/O board has silkscreens for the R's & C's. Remember that Lucnet killed the project before it was done. There was at least 1 other rev of the board taht Tim talks about on the wiki. FTE=Factory Test Equipment. |
| 02:18:57 | TomW | GPSfan_: that would explain why GP23 goes to the CPU, once you enable the circuitry by grounding J602.34, the software would need to know if something was attached. |
| 02:20:03 | TomW | GPSfan_: nice of them to leave a bunch of Spares lying around and not use an FPGA. I would hate to think of telling people that the had to "dead bug" new parts onto the board! |
| 02:20:55 | GPSfan_ | TomW: maybe there was firmware in sboot to recognise that and run sort of special test based on what was connected. They also had JTAG to load the Flash, and maybe run some sort of boundrary scan tests on new boards. |
| 02:21:04 | TomW | GPSfan_: the whole reason I got a Tux was so that I could get some skills working with ARM processors. This whole project has turned out to be a lot of fun. |
| 02:21:46 | GPSfan_ | That's why I got mine, I started on the schematics in mid Sept. |
| 02:26:44 | GPSfan_ | TomW: lots of side tracks along the way, the original 5V PCMCIA mod. and Shanip were passsed down to us as being Lucent's mods that would make VOIP work. Both were poorly implemented. |
| 02:26:49 | as3 | Re. |
| 02:27:25 | as3 | Thats cool, thats why I got into it, I was after a StrongARM system to code/hack about with, and Alhaz told me about the tux. |
| 02:29:49 | TomW | GPSfan_: amazing how many mediocre engineers there are that are getting paid big bucks for doing crappy work, eh? |
| 02:31:55 | GPSfan_ | TomW: yep, also amazing how many engineers don't have any interest in this stuff off-hours, just do it 9-5. |
| 02:32:20 | TomW | GPSfan_: those are the mediocre one. |
| 02:32:25 | TomW | s/one/ones/ |
| 02:32:33 | as3 | One more I hope when I go for a job. |
| 02:33:21 | as3 | Thats not always a bad thing GPS, take people that develop nuclear weapons 9 to 5... |
| 02:46:36 | prpplague | howdy howdy |
| 03:47:52 | prpplague | howdy TomW |
| 03:47:59 | TomW | hey Dave. |
| 03:49:09 | TomW | I just ran the xfer rate stuff that Erik gave me to test the IDE drive, it is only 1.6MB/s |
| 03:49:37 | prpplague | hmm |
| 03:49:50 | prpplague | not too fast |
| 03:50:09 | prpplague | but that still is pretty awesome in my boot |
| 03:50:12 | prpplague | book |
| 03:50:17 | TomW | So, was sitting here, puffing on my pipe (tobacco), thinking of how to optimize the insw() & outsw() functions I had to write to replace the system libs stuff. |
| 03:50:33 | TomW | Here, look at these: |
| 03:51:08 | TomW | #ifdef CONFIG_BLK_DEV_SHANNON |
| 03:51:08 | TomW | static inline void Shannon_insw (u8 * DriveData, u16 * SendTo, int count) |
| 03:51:08 | TomW | { |
| 03:51:08 | TomW | while (count--) { |
| 03:51:08 | TomW | *(SendTo++) = ((u8) DriveData[0]) | (((u8) DriveData[0x100]) << 8); |
| 03:51:09 | TomW | } |
| 03:51:11 | TomW | } |
| 03:51:13 | TomW | static inline void Shannon_outsw (u8 * DriveData, u16 * GetFrom, int count) |
| 03:51:15 | TomW | { |
| 03:51:17 | TomW | while (count--) { |
| 03:51:19 | TomW | DriveData[0x102] = *GetFrom >> 8; |
| 03:51:21 | TomW | DriveData[0] = *GetFrom; |
| 03:51:23 | TomW | GetFrom++; |
| 03:51:25 | TomW | } |
| 03:51:28 | TomW | } |
| 03:51:29 | TomW | #endif |
| 03:51:52 | TomW | I have the LSB of data at array[0] and the MSB at array[0x100]. Tough when you have an 8bit port. |
| 03:52:16 | prpplague | doh |
| 03:52:32 | TomW | The array[0x102] latches the MSB of the outgoing 16bit word for the drive. |
| 03:53:55 | prpplague | TomW: so have you fixed and re-tested? |
| 03:54:30 | TomW | In the arm code, they grab long words of data and xfer between the drive & buffer. Much faster since they have a 16bit buss into the IDE drive. |
| 03:55:03 | TomW | prpplague: still considering solutions... |
| 03:55:23 | TomW | Either way, I am stuck with an 8 bit port.. |
| 03:55:46 | prpplague | TomW: so how did the guys with LART do their ide io? |
| 03:55:55 | TomW | I will still have to do two fetches of bytes, shifting the MSB as necessary. |
| 03:56:01 | TomW | prpplague: as 16 bit. |
| 03:56:05 | prpplague | ahh |
| 03:57:18 | TomW | Without extensive rework of the J602 connector (cuts & jumpers) we are stuck using 8 bits. I did not want to do massive changes to the connector as people would be reluctant to undertake the IDE addition. |
| 03:57:36 | prpplague | TomW: yes i recall you mentioning that |
| 03:58:18 | TomW | Performance hackers can take the approach that I took and mod it into 16bits, but most people are not engineers. |
| 03:58:55 | TomW | IMO, most using the Tux are software people. There are quite a few hardware hackers, but we are a minority? |
| 03:59:12 | prpplague | true |
| 03:59:46 | prpplague | TomW: don't include me as a hardware hacker, i'm still a newbie |
| 04:00:12 | TomW | Still, the xfer rate is much higher than what you could get using a compact flash, the really expensive CF devices are only 20X (twenty time 150KB/s). |
| 04:01:34 | TomW | Well, probably comparable to CF, as the cost of the 12X stuff has dropped a lot lately. But, CF TYPE-1 only goes up to 320Meg, the IDE drive is much larger. |
| 04:01:35 | TomW | heh |
| 04:02:14 | prpplague | TomW: true and makes dev alot easier |
| 04:03:02 | TomW | yeah, no longer do we have to worry about running out of Flash or using up a PCMCIA slot for CF or an IDE adaptor. |
| 04:04:16 | TomW | I have been thinking about that ARM asm stuff, it almost make sense to me when I read it.. |
| 04:04:47 | prpplague | TomW: i've found a board house in az that will do the bga stuff |
| 04:04:59 | TomW | mount them for us? |
| 04:05:07 | prpplague | TomW: yep and get the parts too |
| 04:05:13 | TomW | on prototypes? |
| 04:05:17 | prpplague | yep |
| 04:05:31 | TomW | hmm, wonder how much that would cost. |
| 04:06:08 | prpplague | TomW: i'm writing up specs now to send for a quote, this is a place that Russ Dill has used before |
| 04:06:39 | TomW | The customer that would be interested in an ARM CPU board is having tough economic probs. They are almost 150 days past due on $9K that they owe me. |
| 04:07:23 | prpplague | TomW: this is sorta what i'm shooting for http://www.abcsinc.com/small-linux/images/hack-kit.jpg |
| 04:07:36 | TomW | meanwhile, I will keep hacking away at the tux and learning more about ARM. |
| 04:08:45 | TomW | what! no PCMCIA slots?!! ;-) |
| 04:09:06 | prpplague | TomW: a CF slot |
| 04:09:11 | prpplague | TomW: its not labeled |
| 04:09:27 | prpplague | TomW: ahh yes it is |
| 04:09:31 | TomW | You know, I want a job like Alan Cox has, get to stay home, collect a nice paycheck and play with linux. |
| 04:09:57 | prpplague | TomW: mee tooo |
| 04:10:01 | TomW | heh |
| 04:10:36 | prpplague | TomW: well i figured i'd need some extra logic for the pcmcia slot |
| 04:10:53 | prpplague | TomW: unless i wanted to support only 5v or 3.3v cards |
| 04:11:05 | TomW | Well, my last linux class will be over on Monday, I got a "paycheck" from the school district: I net'ed $378 to teach 12 days of class. |
| 05:32:51 | bmidgley | anyone compiled fbtuxcal? or found directions for setting up cross compiler setup? |
| 05:33:39 | TomW | http://www.openhardware.net/TuxScreen/Development.html |
| 07:28:19 | lyakh_ | Morning!:-) I still can't figure out - to order a tux, do I just use a link on the site to calculate postage cost, add to the tux' price, add insurance (optional) and send this all to BZFlag, or should there be some sort of invoice from him (at least a confirmation of the amount calculations)? |
| 07:31:02 | TomW | don't know. don't you have a paypal account? |
| 07:33:47 | lyakh_ | I am in the process (waiting for my balance / accessing the online one...), then what? You anyway have to at least specify an amount, don't you? |
| 07:36:59 | TomW | ok, I am compiling now, now I can chat, for a few minutes |
| 07:39:11 | TomW | under PayPal you choose Send Money, the next form will ask you to Whom and How Much, you give the form the PayPal account name (tim's account is: Sales@tuxscreen.net). Got it so far? |
| 07:39:44 | lyakh_ | Good, btw, what time is it there where you are? Why don't the US just switch to GMT?:-) |
| 07:39:49 | TomW | Also fill in the amount of the paymnet on that form. |
| 07:40:09 | lyakh_ | Ok, not doing it at the moment - but that's the question - how much? |
| 07:40:14 | TomW | it is 3:43AM, here on the East Coast of the U.S. |
| 07:41:02 | lyakh_ | My parents, sister, other relatives live in NY:-) |
| 07:41:42 | TomW | Send Tim an email and ask him what the International Rate would be.. Or, get on the Tux-Devel mailing list and ask there, there are a number of people in Germany that have Tux's, they must have figured out how much money to send for Postage somehow. |
| 07:43:03 | lyakh_ | Yeah, I've sent a couple of emails to Tim - but he hasn't come round to answering them yet... I am on the list - will ask, thanks, good idea, not exactly tux-devel though:-) |
| 07:44:09 | TomW | I suspect that Tim may be in the Hospital, it was something that Russ had said tonite, but he wouldn't say more. |
| 07:45:10 | lyakh_ | Another thing - just noticed on tux-linux-devel site - make pulls in > 70 Meg first time... What the heck 70MB may you need?... |
| 07:45:41 | lyakh_ | Yep, I noticed about Tim, hope he'll be better soon, no prob with waiting a bit. |
| 07:46:16 | TomW | humph, wait until you launch the "make", it pulls in another 50 Meg, or so. |
| 07:47:11 | TomW | You need something more than a dialup modem to do serious development work, under linux or anything else for that matter. I have an 800Kb/s cable modem here. |
| 07:47:16 | Russ | no, he's not in the hospital |
| 07:49:38 | lyakh_ | Sorry, I've been doing that for quite some time now, and I've never seen anything like that before... The whole kernel sources? gcc? glibc? I'm sure one could provide all what's needed and just specify paths. |
| 07:51:20 | lyakh_ | Will have to hack the Makefile?... At least you should have an option to specify an alternative source. I can download stuff at work, burn a CD, take it home... |
| 07:53:39 | TomW | I cannot vouch for how the "new" makefile system is working. I am using a copy of it before they went to the jffs2 kernel arrangement. Normally, you just get the CVS copy, then type "make" and it builds everything, ready to install into the Flash of the Tux. A real no-brainer. |
| 07:54:15 | TomW | lyakh_: how much linux development work have you done? |
| 07:56:44 | lyakh_ | Ok, I haven't been here at the time of 0.0.1, but 4 years - should be enough to get a feeling of what a usual make process involves?:-) |
| 07:59:50 | TomW | didn't ask you how long you been using linux, how much development work have you done? Kernel hacking, etc. |
| 08:02:37 | lyakh_ | Specificly kernel - well, first time I opened a kernel source file in an editor was about 2 years ago, when I couldn't get DMA to work on my Pentium-75:-) And it turned out that no kernel after 2.0.40 was able to do that, so, I had to invent a local hack (pretty dirty, but even Andre Hedrick couldn't help me...) |
| 08:03:57 | lyakh_ | Since then I've been working mainly with RTAI, which involves kernel-modules writing, and now for more than 7 months I'm working for a company, where I have to take care of the Trizeps-kernel patch, some drivers, RTAI... |
| 08:04:30 | lyakh_ | I meant IDE-DMA (bus-mastering) |
| 08:08:56 | TomW | ok, then you should be comfortable with developing code for the Tux. |
| 08:10:54 | lyakh_ | Thanks:-) Why I want it - I am one of the maintainers of the RTAI ARM-port, so far my only test/development-platform was our Trizeps board. So, I want an alternative (SA) board for comparisons... |
| 08:17:26 | TomW | Why not? The reason I got a Tux was so I could learn about ARM processors. I had done uClinux on an M68K board that I designed. ARM looked very interesting. |
| 08:19:02 | lyakh_ | Exactly - why not?:-) So, if everything goes fine, we'll have real time soon on Tux:-) |
| 08:25:19 | TomW | Maybe, FSM labs sent everyone here in the U.S. running for cover with that stupid patent. Our patent system is totally out of control! |
| 08:26:20 | TomW | I was using that technique about 12 years ago, and I understand that a lot of other developers used it too. But we cannot challenge the patent unless FSM Labs sues us for Patent Infringement. |
| 08:28:17 | lyakh_ | Oh, no, not again!:-)) Well, as of 24.1.9 rtai should be fully protected from the patent claims. If you are interested - there have been A LOT of discussions about it in various related and not too related places... Anyway, most people are sure - FSM Labs will NEVER take RTAI to court... |
| 08:30:47 | TomW | FINALLY! got rid of all the typecasting warnings to the IDE driver modifications. |
| 08:31:43 | TomW | heh, if they take someone to court over that patent, I am sure that a lot of people will show up to show Prior Art. |
| 08:33:15 | TomW | three hours of work to find out that re-coding my C functions in ARM assembler did not affect the performance of the IDE drive. :-( |
| 08:33:33 | lyakh_ | Ok, so, we'll count on your support then:-) |
| 08:33:39 | TomW | Well, I did get to write some ARM asm code for the first time.. |
| 08:34:57 | mallum | hi; |
| 08:35:15 | mallum | TomW: did you get your netwinder disk speeds ? |
| 08:35:31 | lyakh_ | Great! (ARM-asm), it also happened to me some time ago - in fact, it turned out, that an asm-version of a C-function, that is doing _exactly_ (to the best of my understanding - with a neglegible difference that is) the same is its C-analog works, whereas the C-version - dosn't... |
| 08:35:34 | mallum | TomW: I can time my CATS ( netwinder like ) machine if you want |
| 08:36:13 | TomW | mallum: no, I didn't. Does your CATS use a "standard" IDE interface, or is it PCMCIA? |
| 08:36:17 | lyakh_ | s/dosn't/doesn't/ |
| 08:37:24 | mallum | TomW: standard IDE |
| 08:37:33 | TomW | mallum: I would appreciate a: "test dd if=/dev/hda of=/dev/null bs=1M count=100" on a fresh reboot. Then give me the times it reports. |
| 08:37:44 | mallum | TomW: sure I've just booted it |
| 08:37:44 | TomW | s/test/time/ ! |
| 08:38:02 | TomW | time dd if=/dev/hda of=/dev/null bs=1M count=100 |
| 08:38:14 | TomW | cool |
| 08:38:53 | TomW | I am getting an elapsed time of 1minute 1.2 seconds. |
| 08:39:21 | mallum | TomW: it doing it |
| 08:39:36 | TomW | sits on edge of seat |
| 08:39:37 | mallum | TomW: it has a _really_ crap old 2gig ide drive in it |
| 08:39:45 | mallum | doodoo:/home/mallum# time dd if=/dev/hda of=/dev/null bs=1M count=100 |
| 08:39:45 | mallum | 100+0 records in |
| 08:39:45 | mallum | 100+0 records out |
| 08:39:45 | mallum | real 0m33.845s |
| 08:39:45 | mallum | user 0m0.010s |
| 08:39:46 | mallum | sys 0m15.870s |
| 08:40:25 | TomW | Ok, you are double the throughput, I expected that as you are grabbing 16bits at a time and I grab 8bits. |
| 08:41:08 | TomW | I am running at 1.63MB/s and you are at 2.95MB/s. |
| 08:41:30 | TomW | mallum: thanks! I was afraid you would be at something > 10MB/s |
| 08:42:47 | mallum | TomW: np |
| 08:42:58 | mallum | TomW: http://www.simtec.co.uk/sa110eval.php is the board its based on |
| 08:43:21 | mallum | TomW: basically one of them boards and all the crap peices of hardware I had lying around |
| 08:44:19 | TomW | mallum: I am still like a little kid with a new toy. I still get a kick out of booting the Tux just to see it identify the hard drive. :-D |
| 08:44:39 | mallum | TomW: :) |
| 08:44:50 | TomW | mallum: do they have sources for the linux stuff? I need an fdisk. |
| 08:45:26 | mallum | TomW: get it from debian arm distro |
| 08:45:49 | mallum | TomW: you could probably install full debian on the tux now |
| 08:46:12 | TomW | ?? I haven't looked at debian-arm. |
| 08:46:26 | mallum | thats what I run on my CATS box |
| 08:47:13 | TomW | yeah, but then... Eh, I guess I could use the Tux kernel source and the debian-arm sources for all the rest of the stuff? |
| 08:47:55 | mallum | TomW: yep |
| 08:47:58 | TomW | hmmm, would be interesting to have rpm running on the tux. |
| 08:47:59 | mallum | TomW: http://www.debian.org/ports/arm/ |
| 08:48:38 | mallum | TomW: apt / dpkg is much better than rpm trust me ;) |
| 08:48:40 | TomW | mallum: bookmarked, thanks! |
| 08:49:28 | mallum | TomW: one thing I do notice with my cats machine is if theres a lot of disk and network io ( eg I scp a big file to it ) it crashes |
| 08:49:41 | TomW | mallum: I have three people that have expressed a desire in getting a total of 8 IDE interface boards. Hopefully, we can get more and drive the cost down. |
| 08:50:38 | TomW | well guys, it has been a long night. time for bed... |
| 08:51:42 | lyakh_ | TomW: good night:-) |
| 11:29:41 | lyakh | Hm, with all those discussions wrt IDE for tux (it's good and cool, but as just yet another possibility) - has anybody tried NFS? |
| 11:50:48 | TomW | That's interesting, 200Meg of swapspace on a Tux.. |
| 11:51:48 | mallum | that was a quick sleep |
| 11:52:16 | TomW | couldn't, so I played some more.. |
| 11:52:28 | TomW | mind won't quiet down. |
| 11:54:41 | TomW | Was tweaking the timings on the CS3 to the drive. I got it so that the kernel recognized the IDE drive on every boot and the xfer rate is still at 1.6MB/s. Cannot go any faster or the drive disappears on boot. |
| 12:04:00 | lyakh | TomW: wow, no sleep at all... I just asked - did anybody try NFS on tux? |
| 12:04:35 | TomW | It works. |
| 12:05:23 | lyakh | Good. |
| 12:06:04 | TomW | the Tux is very close to being a full computer. |
| 12:07:15 | lyakh | What's still missing?:-) |
| 12:07:55 | TomW | software. |
| 12:08:47 | TomW | although, maybe not, I have to give debian-arm a try. Looks like they have a lot of packages there. |
| 12:09:20 | lyakh | Yep, give it a try. This is what we are using on Trizeps here. It works:-) |
| 12:09:50 | TomW | I did, once, I couldn't get it to pivot-root. |
| 12:11:34 | TomW | I would like to launch lilo instead. I already have software to boot from CF drives, it loads the kernel into RAM then launches it. |
| 12:12:55 | TomW | would be interesting to load lilo into ram, launch it and have it access an initrd to boot the system up. Then you don't need a 4 meg flash, just a 32K boot prom + an IDE / CF drive. |
| 12:14:33 | lyakh | lilo on ARM?... Never heard if it was ported. blob is, redboot - the latter one should also be able to boot from a file. |
| 12:15:16 | | redboot is, like, at http://www.redhat.com/embedded/technologies/redboot/ or at http://sources.redhat.com/ecos/boards/ipaq.html |
| 12:15:16 | TomW | redboot? |
| 12:16:39 | lyakh | The configuration we are using at the moment, which I consider almost acceptable - with blob you boot from flash, then, if you want to replace the kernel, just write it with jffs... Of course, you still don't have a choise at boot-time. |
| 12:17:48 | lyakh | Yes, somebody even already ported redboot to trizeps, but we didn't try it yet. |
| 12:17:52 | TomW | yeah, tried that, you need to have the filesystem in DRAM / IDE before you can write to /dev/mtd1 on the Tux. |
| 12:18:55 | TomW | Maybe that is the solution then. blob + kernel in Flash, pivot-root to the IDE, then when I need to update the kernel, just write to /dev/mtd1. |
| 12:19:31 | lyakh | Why? Can't you copy it from any mounted filesystem? |
| 12:20:30 | TomW | I don't know what Flash prices are like today, but two years ago there was a terrific shortage of Flash memory, *everyone* was on allocation! If you needed 10..20 chips, you were SOL (Simply Out of Luck). |
| 12:22:01 | TomW | That is when I went to using a 32K EPROM to bootstrap the system, it came up an pulled the kernel from the CF disk, then booted it. It was not only cheaper to do it that way, but I could go down to the local camera store and purchase the CF media! |
| 12:23:23 | TomW | I'll try that Familiar stuff again, maybe this time it'll work... |
| 12:23:26 | TomW | sighs |
| 14:17:53 | | there is not here |
| 14:17:53 | TomW | lyakh: still there? |
| 14:19:12 | TomW | I have pivot_root working now, I can pivot to the hard drive. I am getting the debian utils working, slowly, I have fdisk, mkswap & mke2fs working. Starting to partition the dirve. |
| 14:20:31 | lyakh | Cool!!! Super-cool! |
| 14:21:53 | lyakh | I am using kvIRC, and my preferred mode is full-screen, so, I put it on a separate vt and switch here from time to time... So, delays may vary:-) |
| 14:25:03 | TomW | heh, I have a 200Meg swap file.. |
| 14:26:40 | TomW | this is way cool! |
| 14:30:01 | lyakh | nice, nice:-) will anybody ever tell me how much should I pay for this toy? Or I am afraid, if you continue your improvements with this speed for another week the price will shoot up or all left tux's will be sold in 2 hours:-) |
| 14:30:42 | TomW | gotta remember to add the sticky bit to /tmp |
| 14:42:25 | TomW | lyakh: you ready for this? |
| 14:43:01 | TomW | lyakh: check this out: |
| 14:43:04 | TomW | ~ # df -h |
| 14:43:04 | TomW | Filesystem Size Used Available Use% Mounted on |
| 14:43:04 | TomW | /dev/hda1 3.0M 1.3M 1.7M 43% /mnt/initrd |
| 14:43:04 | TomW | /dev/hda1 290.7M 5.3M 270.4M 2% / |
| 14:43:04 | TomW | /dev/hda5 984.3M 20.0k 934.3M 0% /tmp |
| 14:43:05 | TomW | /dev/hda6 387.3M 3.0k 367.3M 0% /var |
| 14:43:07 | TomW | /dev/hda7 2.3G 312.0k 2.2G 0% /usr |
| 14:43:09 | TomW | /dev/hda8 5.1G 20.0k 4.8G 0% /home |
| 14:44:35 | TomW | Something in pivotroot /' uclibc / busybox is making some odd entries for df.. Maybe I'll try the debian version. |
| 14:56:02 | lyakh | :-) not bad, not bad at all, but yes, hda1 looks a bit strange to me:-) |
| 14:57:03 | lyakh | What does mount || /proc/mounts say? |
| 14:57:26 | TomW | lyakh: now I got it: |
| 14:57:29 | TomW | Filesystem Size Used Available Use% Mounted on |
| 14:57:29 | TomW | /dev/hda1 290.7M 5.3M 270.4M 2% / |
| 14:57:29 | TomW | /dev/hda5 984.3M 20.0k 934.3M 0% /tmp |
| 14:57:29 | TomW | /dev/hda6 387.3M 3.0k 367.3M 0% /var |
| 14:57:29 | TomW | /dev/hda7 2.3G 312.0k 2.2G 0% /usr |
| 14:57:30 | TomW | /dev/hda8 5.1G 20.0k 4.8G 0% /home |
| 14:57:32 | TomW | /dev/mtdblock2 3.0M 1.3M 1.7M 43% /mnt/initrd |
| 14:58:05 | TomW | The flash memory block for the old filesystem was not being placed where I wanted it. |
| 14:59:10 | TomW | lyakh: how is this? : |
| 14:59:13 | TomW | total: used: free: shared: buffers: cached: |
| 14:59:13 | TomW | Mem: 64077824 4534272 59543552 0 241664 983040 |
| 14:59:13 | TomW | Swap: 210042880 0 210042880 |
| 14:59:26 | lyakh | Yep, that's better! Now you can start loading all those GBs with games:-) Or what else do people do with them?:-) |
| 14:59:29 | TomW | heh, 200Meg swapfile! This is so cool! |
| 14:59:44 | TomW | heh, yeah |
| 15:01:10 | TomW | It now boots off the Flash, does a pivot root to the hard drive. If I screw something up, I can recover by pivot-ing to an nfs mount. I'll leave that nfs mount in place in the linuxrc file, just in case. |
| 15:08:06 | lyakh | Cool! Bye! |
| 15:08:36 | TomW | prpplague: I have it booting off the hard drive now. |
| 15:09:12 | prpplague | TomW: wahoo, thats too cool |
| 15:09:31 | prpplague | TomW: i wish i had the time to work on my sa-11xx stuff |
| 15:09:53 | TomW | prpplague: well, technically it is booting off the Flash then pivoting to the hard drive. It is using the Tux system with a few debian utils (and libs) thrown in.. |
| 15:12:17 | prpplague | TomW: none the less, a large step forward |
| 15:13:01 | TomW | prpplague: hmmm, wonder if debian has an sshd package..? |
| 15:13:14 | prpplague | TomW: i'm sure it does |
| 15:20:03 | TomW | prpplague: debian has a lot of packages! |
| 15:20:30 | TomW | hello morn |
| 15:23:23 | TomW | prpplague: oooooh! rsh-client, I cna back up my Tux hard drive to the work station, eh? |
| 15:27:59 | prpplague | TomW: ya you could rsync back |
| 15:29:06 | TomW | I think that I'll get another IDE drive for the 2nd 64Meg Tux and set that one up as a web server / gateway, just for giggles. |
| 15:29:37 | Russ | morning prpplague |
| 15:30:00 | prpplague | Russ: morning |
| 15:30:05 | prpplague | Russ: get your package? |
| 15:30:22 | Russ | yup yup yup |
| 15:33:01 | prpplague | Russ: cool, whatcha think of the camion unit? |
| 15:37:19 | Russ | I really like it, going to take it to work today |
| 15:37:53 | Russ | probably let my gf play with it a bit before I start putting new stuff on there, ker family is from korea, and she's been there 2-3 times |
| 15:38:22 | prpplague | Russ: cool, make she can contact ciit and get someone to respond |
| 15:38:44 | Russ | I can't even find anything on the web on it |
| 15:39:10 | prpplague | Russ: it was like, "hey i got your unit, i really like it, can i get the source?" "sure, click" |
| 15:39:23 | prpplague | Russ: that was the last time i talked to them |
| 15:40:12 | Russ | what if you talk to the people at linuette |
| 15:41:14 | prpplague | Russ: i did, they said they only provided the basic rootfs, that the ppl at ciit did all the kernel driver work |
| 15:42:35 | Russ | right, but if they changed the souce of the linuette stuff, then linuette has a vested interest as well |
| 15:43:17 | prpplague | Russ: true |
| 15:45:08 | Russ | you open one up yet to get a general idea of the layout and which chips are used? |
| 15:47:02 | Russ | anyway, I'm off to work |
| 15:47:48 | prpplague | Russ: ya, i'll drop you an email later with the specs |
| 15:50:26 | prpplague | Morn: morn |
| 15:50:31 | | bonjour, Morn |
| 15:50:31 | Morn | hi |
| 15:50:33 | Morn | I'm bouncing |
| 15:50:52 | Morn | I got a business class cable modem connection, and they are letting me host my server here |
| 15:51:34 | Morn | I just need to remember how to make solaris do dhcp instead of static ip |
| 15:57:40 | mallum | matchbox on a psion -> http://screenshots.tader.net/image.php?image=20020515030916-216.194.5.64_1.png&page=1 |
| 16:09:02 | fontenot | tweet |
| 16:18:08 | fontenot | Analysts have estimated that at the initial retail price of $299, Microsoft has been losing anywhere between $76 and $105 on every Xbox sold. |
| 16:18:16 | fontenot | and now they are at 199 |
| 16:18:21 | fontenot | i think I will go buy a bunch |
| 16:24:05 | prpplague | Russ|werk: howdy howdy |
| 16:25:47 | fontenot | anyone know why spazdot is down? |
| 16:25:54 | fontenot | was down last night too |
| 16:26:33 | fontenot | oh it's just extremely slow now |
| 16:44:18 | CosmicPenguin | fontenot: as soon as somebody figures out how to stick Linux on them, I'll buy a couple.... :) |
| 16:45:22 | fontenot | CosmicPenguin: yeah, honestly I have little/no use for them otherwise |
| 16:45:50 | CosmicPenguin | Or even BSD... :) |
| 16:46:43 | fontenot | yes that too would be acceptable |
| 16:47:44 | CosmicPenguin | hates it when he has 971 unread LKML messages |
| 16:47:54 | CosmicPenguin | anything super important happen in the kernel world in the last 2 days? |
| 16:51:37 | fontenot | you mean other than linus changing VM systems again? |
| 16:51:41 | prpplague | CosmicPenguin: we were visited by aliens who gave us technology to make really cool handheld peez dispensers |
| 16:52:33 | prpplague | Russ|werk: i sent you a chip list to your asu.edu email |
| 16:52:42 | prpplague | Russ|werk: i'll have some mb pics tomorrow |
| 16:53:00 | fontenot | (just kidding about the VM thing BTW) |
| 16:55:16 | prpplague | Russ|werk: i also sent another round of emails to mizi to see what kind of assitance i can get from them regarding the cmaion |
| 16:58:32 | Russ|werk | thanks |
| 17:02:54 | | andersee was last seen on #tuxscreen 3 days, 21 hours, 14 minutes and 44 seconds ago, saying: oh. [Sat May 11 20:48:10 2002] |
| 17:02:54 | Russ|werk | ibot: seen andersee |
| 17:03:03 | prpplague | argh! this makes me mad - http://www.linuxandmain.com/modules.php?name=News&file=article&sid=60 |
| 17:06:15 | mallum | well at least m$ dont own your entire countries goverment web presence ... |
| 17:14:21 | fontenot | :-\ |
| 17:15:13 | prpplague | my big thing is why so many scifi web site are non-linux friendly |
| 17:15:28 | prpplague | after all who do you think the primary visitors of scifi sites are |
| 17:16:33 | prpplague | the main star trek page, the official babylon5 page, the scifi channels farscape, the starwars.atomfilms.com page |
| 17:30:31 | lyakh | Hi, anybody has an idea if BZFlag is going to appear here any time soon? |
| 17:36:04 | Russ|werk | I dunno, he's having trouble typing right now |
| 17:37:18 | lyakh | oops, I see, well, the only thing I want to know is 6 keystrokes - xxx.xx - price for a tux to Germany:-) |
| 18:13:46 | | hola, Raymond_ |
| 18:13:46 | Raymond_ | Hi |
| 18:13:54 | | niihau, Raymond_ |
| 18:13:54 | Raymond_ | hola |
| 18:14:06 | Raymond_ | uh? |
| 18:14:57 | Raymond_ | Russ, can you dcc me that file pls? |
| 18:21:33 | Russ|werk | ya, lemme snag it |
| 18:23:26 | Raymond_ | Thanks |
| 18:27:28 | Raymond_ | Bye |
| 18:53:38 | Russ|werk | arghhh |
| 18:53:44 | Russ|werk | busybox is doing strange things |
| 18:53:46 | Russ|werk | where's andersee |
| 19:10:57 | as_ | Evnin all. |
| 19:15:49 | as_ | Hiya prp. |
| 19:16:07 | | prpplague was last seen on #tuxscreen 0 seconds ago, saying: ibot: seen prpplague [Wed May 15 20:16:07 2002] |
| 19:16:07 | prpplague | ibot: seen prpplague |
| 19:16:15 | prpplague | as_: howdu |
| 19:16:17 | prpplague | y |
| 20:17:49 | sjhill | anyone every seen 'insmod: kernel: QM_SYMBOLS' when using busybox when inserting a module? |
| 20:18:08 | prpplague | Russ|werk: think you'll be able to do something with the camion unit? |
| 20:29:37 | Russ|werk | ya, it'll be a lot of fun |
| 20:30:58 | Russ|werk | the main difficulty will probably be the CPLD |
| 20:32:25 | as_ | CPLD on what? |
| 20:33:11 | as_ | Whats a camion? |
| 20:37:23 | Russ|werk | http://www.linuxdevices.com/news/NS6770596476.html |
| 20:47:58 | as_ | Ahh, interesting. |
| 20:48:29 | prpplague | as_: we got a couple in for eval, but couldn't get any source from the company |
| 20:48:47 | prpplague | as_: now they won't even return my emails, faxes, or phone calls |
| 20:49:33 | as_ | They probably scan them for high exposive and anthrax. |
| 20:49:40 | as_ | You'll have to do better than that. |
| 20:52:46 | Fab|o | hi there |
| 20:52:58 | sjhill | Fab|o: hi |
| 20:53:22 | Fab|o | [] |
| 20:53:35 | Fab|o | ya are american? |
| 20:53:50 | Fab|o | this is as american server right? |
| 20:53:53 | sjhill | yup...i'm a yankee...not that matters |
| 20:54:03 | as_ | Most of them are but dont hold it against them. |
| 20:54:10 | sjhill | Fab|o: yes...but who cares? i don't care where you're from |
| 20:54:53 | Fab|o | lol |
| 20:54:55 | Fab|o | yap |
| 20:54:59 | as_ | You'll have to be careful what you say them, there looking for another country to 'liberate'. |
| 20:54:59 | Fab|o | and?? |
| 20:55:15 | as_ | hides. |
| 20:55:19 | Fab|o | as_ i'm from portugal |
| 20:55:33 | sjhill | as_: watch it or i'll nuke your ass :) |
| 20:56:10 | as_ | Youve got real problems then Fab, thats practically a third world country by USA standards. |
| 20:56:37 | Fab|o | lol |
| 20:56:48 | Fab|o | if you say as_ ... |
| 20:56:49 | Fab|o | LOL |
| 20:57:10 | as_ | Take a hint, dont tell them you have oil. |
| 20:57:19 | sjhill | lol |
| 20:57:45 | as_ | whitles nonchalently while applying for a change of identity. |
| 21:13:21 | kergoth | hey tim |
| 21:56:05 | | andersee was last seen on #tuxscreen 4 days, 2 hours, 7 minutes and 55 seconds ago, saying: oh. [Sat May 11 20:48:10 2002] |
| 21:56:05 | jacques_gone | ibot: seen andersee |
| 22:11:36 | Russ|werk | hey andersee |
| 22:11:40 | andersee | BZFlag: what is it with the topic on this channel? |
| 22:11:46 | andersee | Russ|werk: howdy |
| 22:11:46 | Russ|werk | init.c never closes the inittab file handle |
| 22:11:59 | andersee | Russ|werk: Oh. thats bad. :( |
| 22:12:09 | andersee | goes and looks |
| 22:12:37 | CosmicPenguin | Bah - who needs to close file handles anyway... :) |
| 22:12:40 | kergoth | hah |
| 22:17:26 | jacques | andersee: is it really you? |
| 22:17:49 | andersee | yup |
| 22:18:07 | andersee | Back home from the hospital today with a new baby girl. :-) |
| 22:18:15 | jacques | andersee: did your wife have the kid? |
| 22:18:17 | jacques | Cool |
| 22:18:27 | jacques | how much for the little girl? |
| 22:18:31 | jacques | :-D |
| 22:19:01 | CosmicPenguin | That should go into the topic |
| 22:19:08 | andersee | jacques: dunno yet. But my wife talked em into a 10% discount. :) |
| 22:19:12 | andersee | http://www.codepoet.org/gallery/albums.php |
| 22:19:17 | jacques | :-) |
| 22:20:32 | CosmicPenguin | Kewl |
| 22:29:52 | jacques | andersee: congrats! :-) |
| 22:30:25 | Russ|werk | andersee comes back fram the hospital: "there's a bug in busybox" |
| 22:33:55 | andersee | jacques: thanks |
| 22:34:13 | andersee | Russ|werk: fixed now in CVS (the baby is sleeping) :-) |
| 22:34:53 | CosmicPenguin | Those brand new processes demand large timeslices |
| 22:35:51 | andersee | indeed |
| 22:35:58 | Russ|werk | congrats andersee |
| 22:36:08 | Russ|werk | (about the baby, not the CVS commit) |
| 22:36:36 | andersee | CosmicPenguin: I will try to get the preempt code merged in, but that will take _years_ at the current rate. ;-) |
| 22:36:48 | andersee | Russ|werk: thanks |
| 22:37:45 | jacques | anybody know about the PCMCIA bug in current buildroot? |
| 22:37:55 | jacques | several ppl have told me PCMCIA dies not work |
| 22:37:59 | jacques | does |
| 22:42:54 | CosmicPenguin | jacques: specifics? |
| 22:45:02 | BZFlag | andersee: topic? |
| 22:47:31 | andersee | BZFlag: every time I have changed the topic to eliminate the silly stuff like "Yes, you will be", it is always back the next time I join |
| 22:47:50 | andersee | I thought you had over 200 left, for example |
| 22:48:08 | andersee | And pico and micro work now, don't they? |
| 22:49:30 | jacques | chanserv keeps setting the topic to a really old version |
| 22:49:43 | jacques | CosmicPenguin: is it working for you? |
| 22:49:54 | jacques | ppl are telling me their pcmcia network cards no longer work |
| 22:49:57 | jacques | with new buildroot |
| 22:50:22 | jacques | i guess I will hve to try it myself but I was waiting for new blob to get fixed |
| 22:50:42 | jacques | oh, and I need to build a new toolchain :-\ |
| 23:18:37 | CosmicPenguin | is away: Uh, oh.. the weasel got out again... |
| 23:28:53 | bmidgley | 'llo |
