| 00:42.05 | *** join/#edev bigbootay (n=bigboota@CPE0004e28a9289-CM000a73a14ed4.cpe.net.cable.rogers.com) |
| 01:03.46 | *** join/#edev yaman (n=yaman@195.174.129.122) |
| 01:35.19 | *** join/#edev T0mW (n=Tom@24.238.68.20.res-cmts.sth.ptd.net) |
| 02:44.26 | *** join/#edev leoncamel (n=leoncame@219.238.107.107) |
| 03:01.01 | *** join/#edev kakaroto (n=cerne@2001:5c0:8fff:fffe:0:0:0:5b4f) |
| 04:09.46 | *** join/#edev Karo-sam (i=karosa@ob-889.adsl.datanet.hu) |
| 04:20.05 | jbevren | hm |
| 04:20.09 | jbevren | CF adapter for apple I |
| 04:20.12 | jbevren | (no, not apple2) |
| 05:04.30 | *** join/#edev kakaroto (n=cerne@2001:5c0:8fff:fffe:0:0:0:5b4f) |
| 05:32.56 | *** join/#edev redblue (i=star@ppp227.108-253-207.mtl.mt.videotron.ca) |
| 06:27.18 | *** join/#edev DoDo (n=joellind@S0106004005c017f8.ok.shawcable.net) |
| 06:30.56 | DoDo | Hi, recently go ncurses and dialog working on uclinux but I cant get the arrow keys to work properly-Can anyone help or point me to information on terminfo stuff? |
| 06:54.18 | jbevren | DoDo: try asking again in about 7 hours :) |
| 07:18.58 | *** join/#edev redblue (i=star@ppp035.108-253-207.mtl.mt.videotron.ca) |
| 07:23.45 | DoDo | k,thanks |
| 07:25.48 | *** join/#edev lkm` (n=c@gated.datecs.bg) |
| 07:26.35 | lkm` | T0mW: hi, prpplague said that you can help me with cs8900 |
| 07:27.12 | lkm` | T0mW: i cant find workable patch for smdk2410 for it |
| 07:27.40 | T0mW | wow, that has been a few years since I messed with that chip, but... |
| 07:28.22 | T0mW | lkm: that should be a standard chip in the network drivers, it has been in linux for years now. |
| 07:28.55 | T0mW | lkm: the only thing that you should have to do is to change the chip select and IRQ info in the driver. |
| 07:29.58 | lkm` | T0mW: i i sow driver is not platform indipendant, so it not present for smdk2410 |
| 07:30.34 | T0mW | lkm: look in drivers/net/cs89x0.c around line 166 |
| 07:31.33 | T0mW | lkm: I have it the cs8900 working for the LN2410SBC which is a s3c2410 based system. |
| 07:31.54 | T0mW | lkm: what board are you using? |
| 07:33.00 | lkm` | T0mW: firm that i'm working in produce some demo board, but it look very similar to smdk2410 scheme from samsung |
| 07:33.17 | lkm` | T0mW: ethernet is on cs2, irq9 |
| 07:34.28 | lkm` | T0mW: Acctualy i test several paches, but ass has the same problem - unable to handle kernel paging on virtual e0000300 |
| 07:34.59 | T0mW | lkm: you want a tarball of my patches against 2.6.17-8 to look at? |
| 07:35.36 | lkm` | T0mW: yes |
| 07:36.27 | T0mW | lkm: ok, let me do that now. I got pulled onto another aspect of this project, so, I don't have the USB or touchscreen working yet, but all else seems to be running. Ehternet is working fine. |
| 07:37.35 | T0mW | lkm`: email address? |
| 07:37.37 | lkm` | T0mW: ok, i need to run ethernet first |
| 07:37.45 | lkm` | T0mW: mslavchev@datecs.bg |
| 07:40.55 | T0mW | lkm`: ok, sent. patch up 2.6.17-8 and then grep on 'LN2410SBC' to see what I've done. Look in arch/arm/mach-s3c2410/ for my mach-ln2410sbc.c startup code. |
| 07:42.18 | T0mW | lkm`: you'll find the CS8900 base address being set in the map_desc struct |
| 07:43.21 | T0mW | lkm`: I've still got to do the design using the s3c2410 processor for this project. time, not enough time lately. :( |
| 07:46.20 | T0mW | lkm`: if you get the USB / USB Gadget working, please share the code with me? |
| 07:55.18 | lkm` | T0mW: ok, i sow email i shell test now. |
| 07:55.56 | lkm` | T0mW: i have no test usb yet, my problems are with framebuffer and cs8900 |
| 07:56.21 | T0mW | lkm`: what LCD are you using? TFT? |
| 07:57.01 | T0mW | lkm`: this is the board I have: http://www.littlechips.com/LN2410SBC_TFT.htm |
| 08:22.56 | lkm` | T0mW: LDC is PD064VT2, 640x480, hic colors |
| 12:19.39 | *** join/#edev prpplague (n=billybob@69.73.205.184) |
| 12:24.14 | *** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) |
| 12:28.45 | lkm` | prpplague: hi, i test sample. For what is lseek? It failed on it |
| 12:30.56 | prpplague | lkm`: one second |
| 12:31.12 | prpplague | lkm`: did you change the resolution settings? |
| 12:32.15 | lkm` | prpplague: yes |
| 12:34.03 | prpplague | lkm`: that basically just goes to the end of the fb device memory to begin the drawing |
| 12:34.16 | prpplague | lkm`: you say that failed? |
| 12:34.51 | lkm` | prpplague: actually i has seg fault and i check where it is accure. So i change to check lseek result |
| 12:34.58 | prpplague | lkm`: what is the resolution and pitch of your lcd? |
| 12:35.11 | lkm` | prpplague: and sow that actually lseek return (off_t)-1 |
| 12:35.35 | prpplague | lkm`: sounds like you fb settings might not be correct |
| 12:36.05 | prpplague | lkm`: what is the resolution and pitch of your lcd? |
| 12:38.12 | lkm` | in mach-smdk2410.c i add smdk2410_lcd_platdata, in it all resolutions are 640x480 |
| 12:38.16 | lkm` | prpplague: .bpp is 16 |
| 12:40.32 | prpplague | lkm`: so in the test application you set the pitch to 480/2 or 640/2 ? |
| 12:41.50 | lkm` | prpplague: oh, 160/2 |
| 12:42.03 | prpplague | lkm`: try with 640/2 |
| 12:54.51 | lkm` | prpplague: the same result, lseek give an error |
| 12:55.33 | lkm` | prpplague: mmap olso return -1 |
| 12:57.24 | prpplague | lkm`: well sounds like you got something broken |
| 12:57.37 | prpplague | lkm`: i suspect you have you fb driver settings incorrect |
| 12:58.43 | prpplague | lkm`: are you sure this is a 640x480 lcd ? or is it a 480x640 lcd? |
| 13:03.12 | lkm` | prpplague: i think it is 640x480. When i set console=tty, then all kernel output i can see on LCD too. |
| 13:04.08 | lkm` | prpplague: usually i use console=ttySAC0, to be able to write in console via serial port |
| 13:04.19 | prpplague | lkm`: yea thats not a problem |
| 13:04.35 | prpplague | lkm`: i still suspect you have something incorrectly set in your fb driver |
| 13:05.13 | lkm` | prpplague: you mean in smdk2410_lcd_platdata structure ? |
| 13:05.22 | prpplague | lkm`: yea |
| 13:05.28 | lkm` | prpplague: because actually it is the only thing i change |
| 13:07.03 | lkm` | prpplague: i talked with t0mW, and he send me some patches for hims board. Data for LCD there looks very similar to mine |
| 13:18.00 | *** join/#edev GPSFan (n=kenm@12.10.255.248) |
| 13:18.00 | *** mode/#edev [+o GPSFan] by ChanServ |
| 13:19.05 | *** join/#edev bigbootay_ (n=bigboota@CPE0004e28a9289-CM000a73a14ed4.cpe.net.cable.rogers.com) |
| 13:28.43 | prpplague | anyone else getting massive amounts of spam lately? |
| 13:30.47 | GPSFan | prpplague: yes, the level has risen to new highs. |
| 13:37.38 | prpplague | GPSFan: fun fun |
| 13:41.58 | lkm` | prpplague: by chance, do you work with platform builder ? |
| 13:42.16 | lkm` | prpplague: or you use only linux ? |
| 13:42.17 | prpplague | lkm`: i would guess not, since i have no clue what it is |
| 13:43.43 | *** join/#edev mbuf (n=mbuf@59.92.52.31) |
| 13:44.21 | prpplague | mbuf: hows the new job? |
| 13:44.38 | prpplague | lkm`: mbuf has also done some work with the littlechips s3c2410 board |
| 13:44.51 | mbuf | prpplague, not bad, gotta a conference call from US in another hour or so |
| 13:45.19 | lkm` | prpplague: pb is a wince build env. |
| 13:45.25 | mbuf | prpplague, looking at network security projects; so got to attend the interview and they are going to decide if i am fit for it |
| 13:46.03 | lkm` | prpplague: ok, t0mw give me hints about cs8900 and now i shell try first it |
| 13:46.38 | mbuf | lkm, isn't there a driver for the cs8900 already ? |
| 13:47.22 | prpplague | mbuf: yea but its buggy for the s3c2410 interface |
| 13:47.39 | prpplague | lkm`: i don't do any m$ development |
| 13:47.39 | mbuf | prpplague, i see |
| 13:47.52 | mbuf | same here |
| 14:04.47 | *** join/#edev prpplague^2 (n=billybob@72.22.146.236) |
| 14:05.48 | *** mode/#edev [+o prpplague] by ChanServ |
| 14:36.05 | *** join/#edev T0mW (n=Tom@24.238.68.20.res-cmts.sth.ptd.net) |
| 14:36.25 | T0mW | prpplague: hey |
| 14:36.35 | prpplague | T0mW: hey hey |
| 14:36.46 | T0mW | prpplague: call FedEx, the box is lost |
| 14:37.55 | prpplague | T0mW: sorry, no, it hasn't cleared customs, i meant to tell you, the custom offices were closed due to holiday, i spoke to someone yesterday evening, its being handed to fedex tuesday morning, and should be delivered to you by thursday |
| 14:38.31 | T0mW | prpplague: man! You are going to be sooo happy when you get back to civilzation! |
| 14:38.32 | prpplague | T0mW: sorry for the delay, i didn't realize it was gonna take so long to ship from here |
| 14:38.38 | prpplague | T0mW: NS |
| 14:39.03 | T0mW | prpplague: I just installed a bigger mailbox to hold some of the packages I've been getting lately |
| 14:39.08 | prpplague | T0mW: funny thing is, barbados was listed as the "least corrupt" island in the caribean |
| 14:39.34 | T0mW | prpplague: oh, it is! They don't do sh~t down there but watch the coconuts fall |
| 14:39.41 | prpplague | T0mW: hehe |
| 14:40.01 | T0mW | prpplague: hard to arrest someone for scalping coconuts |
| 14:40.06 | prpplague | T0mW: hehe |
| 14:40.17 | T0mW | mon |
| 14:40.22 | prpplague | T0mW: most of the major roads have coconut vendors |
| 14:40.30 | prpplague | T0mW: hehe, "mon" is jamaica |
| 14:40.40 | prpplague | T0mW: different accent here |
| 14:41.15 | T0mW | prpplague: I was just looking this schematic over for the s3c2410 chip. What LCD you use on that? single scan monochrome, 4bit interface? |
| 14:43.24 | prpplague | T0mW: yea |
| 14:43.45 | prpplague | T0mW: i think i included one in your package, but can't remember |
| 14:43.58 | T0mW | prpplague: can you use an MMC card on the SD interface, or do you have to put it on the SPI circuit? |
| 14:44.34 | T0mW | prpplague: turns out, I've got to design a CPU board for that system, they are not going to use the littlechips board. |
| 14:44.38 | prpplague | T0mW: yea you can use a mmc card with onboard controler, but we are actually just bitbanging the gpio's to do spi interface |
| 14:44.49 | prpplague | T0mW: lovely |
| 14:45.03 | prpplague | T0mW: they want to do their own board? |
| 14:45.41 | prpplague | T0mW: you and i might need to get together with my boss |
| 14:45.52 | prpplague | T0mW: might be advantageous for us to share the design |
| 14:45.52 | T0mW | prpplague: schematic only, I told them that I had no experience with laying out an FBGA multi-layer board. Plus, I wasn't going to put the thing together, they had to get someone else to build the board. |
| 14:46.36 | prpplague | T0mW: what kind of lcd support do you need? |
| 14:47.02 | T0mW | prpplague: form factor of the card will be 3.5" x 4.250" with card edge connector. |
| 14:47.26 | T0mW | prpplague: simple 320x240 16 greylevel, with a touch screen interface. |
| 14:47.44 | prpplague | T0mW: ahh |
| 14:48.08 | T0mW | prpplague: the LCD will be for general status info, the heavy graphics lifting will be done by either a PC or WinCE. |
| 14:48.19 | prpplague | T0mW: what about using a sodimm cpu board that snaps into the 3.5"x4.25" board? |
| 14:48.29 | T0mW | prpplague: we could do that |
| 14:48.50 | T0mW | prpplague: it would have to lie flat onto the card. |
| 14:48.58 | prpplague | T0mW: yea |
| 14:49.02 | T0mW | prpplague: like Logic Products Dev does it |
| 14:49.07 | prpplague | T0mW: yea |
| 14:49.20 | prpplague | T0mW: you already have an lcd in mind? |
| 14:49.24 | T0mW | prpplague: I could sell that idea to them. |
| 14:49.26 | prpplague | T0mW: they want ts support? |
| 14:49.55 | T0mW | prpplague: definately "soft keys" ala touchscreen, no membrane keyboard for the unit. |
| 14:49.57 | prpplague | T0mW: i'm almost certain that management would be interested in selling just the sodimm as long as we don't have to do any support |
| 14:50.30 | prpplague | T0mW: give me a second to email you |
| 14:50.37 | T0mW | prpplague: run it by them, my customer has suggested using something like an SODIMM. |
| 14:51.53 | prpplague | T0mW: how much flash and ram do you need? |
| 14:52.48 | T0mW | prpplague: 2meg Flash to hold the kernel and get things going, then a transflash (micro-SD) to store data, and 64Meg of SDRAM. |
| 14:52.53 | T0mW | prpplague: sounds about right. |
| 14:53.21 | T0mW | prpplague: I would like the transflash as part of the card. the system will need local "drive" storage. |
| 14:53.25 | prpplague | T0mW: the other form factor we have is similiar to the board you are getting |
| 14:54.06 | prpplague | T0mW: what kind of qty per year you think you might have? |
| 14:54.10 | T0mW | prpplague: but, the transflash could be on the main CPU card. |
| 14:54.37 | T0mW | prpplague: I don't think that it will be that much quantity, I suspect that they sell less than 500 units per year. |
| 14:54.43 | T0mW | prpplague: but |
| 14:54.58 | T0mW | prpplague: they are working up a few OEM deals. |
| 14:55.34 | prpplague | ok that would work out well |
| 14:55.36 | T0mW | prpplague: "not invented here" crowd, if you want the business you private label someone else's stuff. |
| 14:56.21 | T0mW | prpplague: I'm not sure what you have in mind, share the development costs + the resulting design? |
| 14:56.49 | T0mW | prpplague: e.g. your company populates the boards, I design it? |
| 14:58.14 | lkm` | T0mW: i tried cs89x0 patch. Now it give me that can not find 89x0 controler. It failed when check CHIP_EISA_ID_SIG |
| 14:58.22 | prpplague | T0mW: i was thinking more that you use our CPU module and you just design the board you are needing it to snap into |
| 15:01.00 | prpplague | T0mW: you've got spam |
| 15:01.18 | T0mW | prpplague: bbiab, phone call |
| 15:01.22 | prpplague | T0mW: np |
| 15:03.05 | prpplague | lkm`: we can only help you so much with a device none of us have or have access to the schematic for |
| 15:08.14 | *** join/#edev redblue (i=star@ppp069.109-253-207.mtl.mt.videotron.ca) |
| 15:10.05 | lkm` | prpplague: oh, just now i check, on scheme ethernet is on nGCS3, IRQ9 |
| 15:11.39 | T0mW | prpplague: I don't know if that would work for them, they seem to be building up their proprietary design porfolio. |
| 15:11.53 | prpplague | T0mW: oh |
| 15:12.15 | T0mW | lkm`: did you make the change to the map_info structure so that the I/O select is allowed ? |
| 15:12.30 | prpplague | T0mW: well i was just thinking that if there were using our module, it would be easy for them, since we build in bulk, and we stock them |
| 15:12.42 | T0mW | lkm`: arm/arch/mach-s3c2410/mach-ln2410sbc.c |
| 15:13.18 | prpplague | T0mW: and there would be nothing stoping you from designing a sodimm cpu board with the same pinout |
| 15:13.31 | T0mW | prpplague: I will run it by them though, it would save us time and cost. However, we would need to U/L the thing. |
| 15:13.58 | prpplague | T0mW: yea i don't think that would be a problem |
| 15:14.15 | prpplague | T0mW: in fact i think we are already working on the for a couple of boards |
| 15:14.16 | T0mW | prpplague: got the email |
| 15:14.42 | T0mW | prpplague: gak! you mean I have to sign ANOTHER nda? |
| 15:14.44 | T0mW | heh |
| 15:14.48 | prpplague | T0mW: hehe |
| 15:14.58 | prpplague | T0mW: oh and of course the boards are RoHS compat |
| 15:15.41 | T0mW | prpplague: 90 degree mount? |
| 15:15.54 | T0mW | prpplague: 90 degree socket? |
| 15:16.14 | lkm` | T0mW: if you talk about ln2410sbc_iodesc, yes i put structure record in it |
| 15:16.20 | prpplague | T0mW: yea right angle with the board |
| 15:16.39 | T0mW | prpplague: this may fly |
| 15:16.50 | prpplague | T0mW: http://www.elinux.org/wiki/M5900 |
| 15:16.59 | prpplague | T0mW: basically identical to that dev board |
| 15:17.05 | T0mW | prpplague: definately 144pin, needs more I/O pin bits |
| 15:17.28 | prpplague | T0mW: yea, thats just an early dev board |
| 15:17.50 | T0mW | ah |
| 15:17.54 | T0mW | ok |
| 15:18.08 | prpplague | T0mW: but this way you'd get working jtag code, bootloader, kernel, and buildroot |
| 15:18.22 | T0mW | prpplague: let me pass it by them, you pass it by your people, then we can have my people contact your people |
| 15:18.24 | T0mW | LOL |
| 15:18.26 | lkm` | T0mW: if there any requirements for virtual address, because it is something 0xf1300000 |
| 15:18.56 | prpplague | T0mW: yea, let me know |
| 15:19.12 | T0mW | prpplague: uh, I need ethernet |
| 15:19.22 | prpplague | T0mW: enc28j60 |
| 15:19.33 | T0mW | prpplague: ok, 2.6.x driver? |
| 15:19.51 | prpplague | T0mW: working, but not accpetable yet to netdev standards |
| 15:20.08 | T0mW | prpplague: wtf is "netdev" ? heh |
| 15:20.26 | prpplague | T0mW: network device group, the ones that handle network drivers for the kernel |
| 15:20.57 | T0mW | prpplague: oh, you mean the netdev people, those a**h*les, mmm |
| 15:21.06 | T0mW | :P |
| 15:21.06 | prpplague | hehe |
| 15:21.12 | prpplague | T0mW: so you do know them! |
| 15:21.32 | T0mW | prpplague: their reputation preceeds them. |
| 15:21.40 | prpplague | hehe |
| 15:22.03 | T0mW | prpplague: ok, I'll let you get back to work, I'm going out to purchase some more foooooooD! |
| 15:22.09 | prpplague | okie dokie |
| 15:22.14 | prpplague | T0mW: just let me know |
| 15:22.37 | T0mW | I'm down to those few cans of beans and tuna (and, I have no Miracle Whip). |
| 15:22.53 | T0mW | no bread either |
| 15:23.11 | T0mW | prpplague: BUT, I do know that today is Friday |
| 15:23.19 | T0mW | isn't it? |
| 15:25.15 | chouimat | morning |
| 15:30.49 | *** join/#edev CosmicPenguin (i=nobody@nat/amd/x-5d22d893f9b7d7a1) |
| 15:30.49 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 15:33.19 | T0mW | prpplague: <sigh> more freakin' fires breaking out! "we need this 'real soon'" |
| 15:34.28 | T0mW | last year I basically sat around and watched the grass grow, this year I have to much work. |
| 15:34.51 | prpplague | hehe |
| 15:34.52 | T0mW | Why was it we voted the Republicans out ? The Economy? |
| 15:35.05 | prpplague | T0mW: yea something like that |
| 15:35.07 | T0mW | Doh! Rumsfeld, okay, I got it. |
| 15:35.20 | agaffney | congressional pages? :P |
| 15:35.59 | T0mW | agaffney: I guess the pages look good as every hooker in DC will "tell all" ? |
| 15:37.17 | prpplague | T0mW: yea, i expect things to be a little rough for next 4-8 years |
| 15:38.03 | *** join/#edev chouimat (n=dieu@r2351064.cidc.net) |
| 15:38.46 | T0mW | prpplague: eat up that overtime pay boys, the lean times are a'commin |
| 15:39.47 | prpplague | T0mW: oh yea |
| 15:40.25 | CosmicPenguin | I'm sorry - who was in charge while computer engineers went from obscurity to rock-star status? |
| 15:40.46 | prpplague | CosmicPenguin: ?? |
| 15:41.11 | chouimat | CosmicPenguin: if you find out tell me ... I have a nice baseball bat for him :) |
| 15:42.15 | T0mW | CosmicPenguin: I dunno, the republicans have held both houses of congress since 1994 |
| 15:42.27 | prpplague | CosmicPenguin: i thought all of us computer geeks were still in obscurity |
| 15:43.21 | CosmicPenguin | blah blah blah country sucks, I'm going to move to blah blah blah |
| 15:43.26 | CosmicPenguin | geeks and politics don't mix |
| 15:43.34 | prpplague | CosmicPenguin: hehe |
| 15:43.37 | prpplague | CosmicPenguin: yea |
| 15:43.54 | wmat | CosmicPenguin: in any form |
| 15:44.08 | chouimat | hi prpplague T0mW wmat CosmicPenguin |
| 15:44.19 | prpplague | CosmicPenguin: trustme though, after living overseas, even is there are lots of problems with the USA, its far better than any of the other places i've been |
| 15:44.20 | wmat | hello |
| 15:44.27 | prpplague | wmat: greetings |
| 15:44.40 | wmat | prpplague: i'm struggling to stomach even writing policy for the new wiki ;( |
| 15:44.46 | prpplague | hehe |
| 15:53.57 | prpplague | interesting, dem's seem to favor reinstatement of the section 245(i) clause |
| 15:54.42 | chouimat | prpplague: I doubt they will have time to change anything before the 2008 election :) |
| 15:55.02 | prpplague | chouimat: why's that? |
| 15:55.17 | chouimat | prpplague: not enough time :) |
| 15:55.28 | prpplague | chouimat: whys it not enough time? |
| 15:55.41 | prpplague | chouimat: thats two full years away |
| 15:56.37 | chouimat | prpplague: my guess is nothing will be done because so far it look like your gov will be in a nice deadlock ... but I can be wrong ... |
| 15:57.07 | prpplague | chouimat: ahh, not that they don't have time, just will not be able to pass anything |
| 15:59.13 | prpplague | amazing how neither side actually states what they want to actually do |
| 16:03.12 | prpplague | man does this speak volumes or what - http://republican.senate.gov/public/index.cfm?FuseAction=Issues.Detail&Issue_id=42 |
| 16:24.27 | lkm` | T0mW: i test a lot of variants, only in (CS3 + 0x01000000) it return me 0x0e for chip ID. In all other cases it returns 0x00 |
| 16:26.10 | lkm` | T0mW: or chip is not there, or not well initialize. |
| 16:31.11 | T0mW | lkm`: ok, you set up the map_desc structure to map the chip select into virtual memory, correct? |
| 16:31.21 | *** join/#edev CosmicPenguin (i=nobody@nat/amd/x-5d22d893f9b7d7a1) |
| 16:31.21 | *** join/#edev prpplague (n=billybob@72.22.146.236) |
| 16:31.21 | *** join/#edev dfarnsworth (n=dfarnswo@65.200.49.156) [NETSPLIT VICTIM] |
| 16:31.21 | *** mode/#edev [+oo CosmicPenguin prpplague] by irc.freenode.net |
| 16:33.14 | *** join/#edev lkm` (n=c@gated.datecs.bg) [NETSPLIT VICTIM] |
| 16:33.14 | *** join/#edev Karo-sama (i=karosa@ob-889.adsl.datanet.hu) [NETSPLIT VICTIM] |
| 16:33.14 | *** join/#edev lkm (n=lkm@85-130-39-16.1702037.ddns.cablebg.net) [NETSPLIT VICTIM] |
| 16:33.14 | *** join/#edev wmat (n=wmat@CPE000f6641f21a-CM00111adf5f4a.cpe.net.cable.rogers.com) [NETSPLIT VICTIM] |
| 16:34.20 | *** join/#edev sunrunner20 (n=jamesdoe@pool-71-252-248-115.dllstx.fios.verizon.net) [NETSPLIT VICTIM] |
| 16:34.30 | *** join/#edev chouimat (n=dieu@kde/developer/chouinard) [NETSPLIT VICTIM] |
| 16:34.30 | *** join/#edev silverpower (n=shrike@adsl-153-145-124.cha.bellsouth.net) [NETSPLIT VICTIM] |
| 16:34.30 | *** join/#edev FireEgl (i=FireEgl@Sebastian.Atlantica.US) [NETSPLIT VICTIM] |
| 16:34.30 | *** join/#edev isenbeck (n=clint@rrcs-71-40-86-22.sw.biz.rr.com) [NETSPLIT VICTIM] |
| 16:34.30 | *** join/#edev youam (n=youam@ciara.youam.de) [NETSPLIT VICTIM] |
| 16:34.30 | *** join/#edev JoseJX (n=jjezak@c-67-171-66-14.hsd1.pa.comcast.net) [NETSPLIT VICTIM] |
| 16:34.30 | *** join/#edev ccjoe (i=clrsrv@www.clearcore.com) [NETSPLIT VICTIM] |
| 16:36.32 | *** join/#edev redblue (i=star@ppp069.109-253-207.mtl.mt.videotron.ca) [NETSPLIT VICTIM] |
| 16:36.32 | *** join/#edev rgb (n=rgb@OTWAON23-1168100517.sdsl.bell.ca) [NETSPLIT VICTIM] |
| 16:36.32 | *** join/#edev ohsix (i=ohsix@bc111222.bendcable.com) [NETSPLIT VICTIM] |
| 16:36.39 | T0mW | lkm`: ok, you set up the map_desc structure to map the physical address of the chip select into virtual memory, correct? |
| 16:36.39 | lkm` | T0mW: yes, i use the same scheme as in your patch |
| 16:36.41 | T0mW | lkm`: s3c24xx_init_io() adds it to the table |
| 16:36.42 | T0mW | lkm`: you're running as little-endian? |
| 16:36.42 | lkm` | T0mW: yes, map_desc is ok |
| 16:36.42 | lkm` | T0mW: yes, ETH address is base virtual adress + 300 |
| 16:36.43 | lkm` | T0mW: virtual address is 0xf4000000, hadware address is cs3 + 0x01000000 |
| 16:36.44 | T0mW | lkm`: the s3c2410 configured to run as little-endian, correct? |
| 16:36.44 | lkm` | T0mW: It is in little endiag 100%, jffs root fs is build with that flag |
| 16:37.56 | T0mW | ok |
| 16:37.57 | T0mW | lkm`: things get "real interesting" with the CS8900 in big-endian. |
| 16:37.57 | lkm` | T0mW: kernel also use le, i'm sure because eary i test that with #define<>#warning |
| 16:38.37 | lkm` | T0mW: actually this board was copied from some firm partners, that provide and wce image for it |
| 16:40.37 | lkm` | T0mW: And with it this chip works |
| 16:40.38 | T0mW | lkm`: the thing you have to keep in mind is that the patches I sent you is for the LN2410SBC board, the .config file defines the machine with 'CONFIG_MACH_LN2410SBC=y', you may have to edit my '#ifdef CONFIG_MACH_LN2410SBC' to make it match your machine definition. |
| 16:41.20 | lkm` | T0mW: Actually i search into yours first 4 paches and just change code into 2.6.18 |
| 16:41.37 | *** join/#edev silverpower_ (n=shrike@adsl-234-211-250.cha.bellsouth.net) |
| 16:41.48 | lkm` | T0mW: A change all configs to be usable for me |
| 16:42.06 | T0mW | lkm`: and, you changed LN2410SBC_PA_CS8900_BASE to S3C2410_CS3 ? |
| 16:42.23 | lkm` | T0mW: Yes, |
| 16:43.27 | lkm` | T0mW: Also IRQ to 9, but tracing the code i sow that in point driver failed, no irq need. It expected from some port some data, that not mach |
| 16:44.01 | T0mW | lkm`: ok, now, with the MMU, the S3C2410 virtual memory is broken out into various addresses. These are defined in include/asm/arch/map.h |
| 16:45.11 | T0mW | lkm`: you probably won't need to change anything in map.h, just consider what is in that file. |
| 16:46.27 | lkm` | T0mW: what i must check in it ? |
| 16:47.37 | lkm` | T0mW: I'm not sure what virtual addres i must use to map physical |
| 16:47.53 | lkm` | T0mW: So i use +0x04000000 |
| 16:48.29 | lkm` | T0mW: also, why eth address is +0x300 offset ? |
| 16:49.47 | T0mW | lkm`: yeah, that is what I'm looking at now, how I determined that the CS8900 on chip select #1 was placed at 0xF1300300 |
| 16:52.36 | lkm` | T0mW: i'm on chip select 3 |
| 16:54.23 | T0mW | lkm`: 0x300 is the default address of a CS8900, when you first power the cs8900 up, it will only use that address IF, and I say IF, you do NOT have an EEPROM on the cs8900. Without the EEPROM to tell it differently, it defaults to 0x300 |
| 16:54.50 | lkm` | T0mW: hmm, i was thinking that you can use whatever Virtual Address you want, just you must map it |
| 16:54.57 | T0mW | NO |
| 16:55.36 | lkm` | T0mW: so in every case with and without eeprom it must answear on 300 ? |
| 16:55.36 | T0mW | lkm`: well, yes, but you have to determine the physical address first of where the chip really is. |
| 16:56.01 | T0mW | lkm`: correct, it will always answer at 0x300 without the EEPROM |
| 16:56.18 | lkm` | T0mW: as in hardware show me, it resides on cs3, irq9 |
| 16:56.57 | T0mW | lkm`: so, you need the Chip Select active and be accessing the chip at 0x300..0x3xx |
| 16:57.10 | lkm` | T0mW: and i'm a little curious because all schemes in internet for smdk2410 are with eth on cs3+ 0x01000000 |
| 16:58.19 | lkm` | T0mW: if it is active .. who know .. |
| 16:58.42 | T0mW | lkm`: ok, I see what I did |
| 16:59.03 | T0mW | lkm`: look at include/asm/arch/map.h |
| 16:59.18 | lkm` | yes |
| 16:59.38 | T0mW | lkm`: starting at line 45 are the virtual addresses of the various things |
| 16:59.51 | T0mW | various parts of the CPU internals |
| 17:00.26 | T0mW | scroll down that list to line #153, the definition for the SDI |
| 17:00.40 | lkm` | they ends with 0x00D00000 |
| 17:01.06 | T0mW | lkm`: oh?! |
| 17:01.08 | lkm` | SDI 0x01200000 |
| 17:01.21 | T0mW | lkm`: yes, that one |
| 17:01.42 | T0mW | lkm`: I simply went further and used 0x1300000 |
| 17:01.56 | T0mW | lkm`: I simply went further and used 0x1300000 for the cs8900, I used a "spare" |
| 17:02.38 | T0mW | lkm`: you see that the SDI Virtual Address is at 0x1200000 and the length is SZ_1M ? |
| 17:03.05 | lkm` | yes |
| 17:03.09 | T0mW | lkm`: that would be the range of 0x1200000..0x12fffff |
| 17:03.30 | T0mW | so, 0x1300000 is the next available virtual address |
| 17:04.02 | lkm` | aham, i use 0x04000000, i shell change that but dont believe that this is some problem |
| 17:04.03 | T0mW | so, I mapped CS1 (physical address) into 0x1300000 virtual |
| 17:04.43 | T0mW | lkm`: well, I'm just explaining to you how I did it, YMMV |
| 17:04.46 | lkm` | into 0xf1300000 ? |
| 17:04.51 | T0mW | correct |
| 17:04.54 | lkm` | ok |
| 17:06.03 | T0mW | 0xf0000000 is the virtual base address for the s3c2410 "registers", the chipselects being one of the "registers". RAM is mapped lower into 0xC0000000 |
| 17:07.05 | T0mW | lkm`: but, what they do on ARM is compress the 128Meg address range of the other chip selects into virtually smaller spaces. Like, why does a cs8900 need a whole 128Meg address space?? |
| 17:07.35 | T0mW | lkm`: so, we only give it SZ_1M of virtual room instead of the whole 128Meg. |
| 17:07.40 | lkm` | T0mW: yes, in map_desc i use only 1 meg |
| 17:07.47 | T0mW | that is ok |
| 17:07.57 | T0mW | lkm`: just don't overlap them |
| 17:08.39 | T0mW | things could get real interesting if you screw up the virtual mapping and try to activate more than one thing at a time (fault?) |
| 17:09.36 | lkm` | T0mW: before i was forget to add in map_desc entry, and kernel raise fault |
| 17:10.20 | lkm` | T0mW: in netcard_portlist i add Virtual address + 0x300 |
| 17:10.45 | T0mW | lkm`: ok, once I established where I wanted to map the CS1, I then modified the map_desc __init structure, then made sure it was processed by an __init call to s3c24xx_init_io so the CS1 got virtually mapped into the MMU. |
| 17:11.15 | T0mW | lkm`: you understand all this so far? |
| 17:12.32 | lkm` | T0mW: yes, just check if it it 100% right in my code |
| 17:13.03 | lkm` | T0mW: i have an entry into map_desc |
| 17:13.09 | T0mW | lkm`: when we access 0xf1300300, the CS1 goes active and an address of 0xf1300300 is placed on the address bus. Since my cs8900 is only wired to the lower three bits of the address bus (A1,A2,A3), the cs8900 responds. |
| 17:14.29 | lkm` | T0mW: hyh, here i'm not sure ... So CS3 enables in time you map it ? |
| 17:14.52 | T0mW | lkm`: say what you just said, but differently |
| 17:15.18 | T0mW | lkm`: I didn't understand that last question |
| 17:16.05 | T0mW | lkm`: do you have an oscilloscope? |
| 17:17.32 | T0mW | lkm`: do you have an oscilloscope? |
| 17:18.23 | lkm` | T0mW: so because address wire is common for many device, with cs you anable what device you should use |
| 17:18.32 | lkm` | right ? |
| 17:18.45 | lkm` | no, a have no osciloscope here |
| 17:19.57 | T0mW | lkm`: do you have a logic probe? |
| 17:21.03 | lkm` | T0mW: i can get some multimeter |
| 17:21.10 | T0mW | lkm`: no good |
| 17:22.13 | T0mW | lkm`: to do this type of work, you really need an oscilloscope. A 100MHz scope is real cheap today, you can get one on ebay for less than 140 euros |
| 17:22.42 | lkm` | T0mW:so i ask that, if you want to access some physical address, ARM CPU raise chip select signal automaticaly ? |
| 17:23.17 | lkm` | T0mW: here we have a lot of oscilloscopes, but in hardware department. |
| 17:23.48 | lkm` | T0mW: and just now i'm the last standing .. |
| 17:23.57 | T0mW | lkm`: OH, so you DO HAVE one, just not in your department. |
| 17:24.16 | T0mW | lkm`: ever deadlock the linux kernel? |
| 17:24.30 | T0mW | lkm`: what I mean is this |
| 17:24.37 | lkm` | T0mW: i'm in software team :) |
| 17:26.19 | T0mW | lkm`: in the init code of the kernel, IIRC, kernel/init.c, you can make a pointer to the virtual address of the cs8900. then do a "char * ptr = 0xf1300000; while (1) (*ptr)++;" |
| 17:26.55 | T0mW | lkm`: that will bang hell out of the chip select line and you can then look at the chip select with a scope to verify it is wiggling. |
| 17:27.33 | lkm` | T0mW: actually i try to this linux work in free time or after work time. Here bosses are not very entusiastic for linux |
| 17:28.55 | T0mW | lkm`: well, this comes down to hardware. You can printk() your way around the code, but eventually you will have to look at the hardware to see if it is doing what you think it should. |
| 17:28.56 | lkm` | T0mW: So for redure questions and eplanations why i do that i arrange with limitted resources |
| 17:30.08 | lkm` | T0mW: So for redure questions and eplanations why i do that i |
| 17:30.08 | lkm` | <PROTECTED> |
| 17:30.14 | T0mW | lkm`: I would suggest that you get a cheap 30 euro Logic Probe. A logic probe will tell you if a chip select is wiggling. Logic probes are pretty small and people won't notice if you have one. |
| 17:30.15 | lkm` | hyh |
| 17:31.26 | T0mW | lkm`: if you understood everything I showed you about the MMU setup for the chip select, you just need to configure your kernel #defines to match your hardware. |
| 17:32.21 | T0mW | lkm`: then, make a two line change in the cs8900 where the #defines are for the chip select address and IRQ line. |
| 17:32.51 | T0mW | lkm`: the IRQ will be whatever is connected to the cs8900 IRQ pin. |
| 17:33.17 | T0mW | lkm`: just look at the schematic and follow the wires to see where they go. |
| 17:33.24 | lkm` | T0mW: I think that this is clear. But i think to map all gcs3 address space, and in cycle to read all adresses |
| 17:33.40 | T0mW | lkm`: no |
| 17:33.45 | lkm` | T0mW: so if chip is there i shell be able to fing it ?!? |
| 17:33.55 | lkm` | find it .. |
| 17:34.53 | lkm` | T0mW: on the scheme it is on irq9, |
| 17:35.28 | lkm` | T0mW: but irq is used later in code, so first i need to find the exact location of chip |
| 17:35.50 | T0mW | lkm`: unless it has an eeprom on pins 3,4,5 and 6 of the cs8900, it WILL be at 0x300. Provided that the A8 + A9 of the cs8900 pins are wired to +3v and all the rest of the address line go to GND (0v), with the exception of A1,A2 and A3, they should be wired to the CPU Adress Bus. |
| 17:36.42 | T0mW | lkm`: you have any friends in the hardware department that can verify that the cs8900 is on CS3 at 0x300 ? |
| 17:37.33 | T0mW | lkm`: I have to leave now |
| 17:47.00 | *** join/#edev ][-DoDo (n=joellind@208.181.82.132) |
| 17:48.45 | ][-DoDo | Hi - I have gotten dialog and ncurses working in uClinux, but I am having troubles getting the arrow keys to work properly. I have been messing around with my inputrc file, but that doesn't seem to help much. I can't find much help on the web, though I am sure it is there but I dont know where to look |
| 17:50.06 | prpplague | didn't think there was much work that you needed to do for dialog and ncurses, they are already packages in buildroot |
| 17:50.26 | ][-DoDo | prpplague - I am using uClinux |
| 17:50.36 | prpplague | and? |
| 17:50.39 | ][-DoDo | prpplague - snapgear |
| 17:50.43 | prpplague | and? |
| 17:50.59 | ][-DoDo | prpplague - no dialog and ncurses didn't compile without a lot of help |
| 17:51.22 | prpplague | odd, i compile it all the time from buildroot with no problems |
| 17:51.41 | ][-DoDo | prpplague - but you are probably much better than me with linux |
| 17:51.58 | prpplague | enable it in the config and type make |
| 17:52.41 | ][-DoDo | prpplague - I wish it was that easy for me. I have to make my own makefiles etc and then I usually get a bunch of errors |
| 17:53.18 | ][-DoDo | prpplague - the main thing is its working now, except the arrow keys/keypad |
| 17:56.11 | prpplague | ][-DoDo: 1) then use buildroot, 2) if the keys aren't working, then its not working |
| 17:57.24 | ][-DoDo | prpplague - all the regular keys work, when I say keypad I mean the numeric keypad |
| 17:57.38 | prpplague | ][-DoDo: ok is this on a serial console? |
| 17:58.00 | ][-DoDo | prpplague - both on a serial console and when logging in via ssh |
| 17:58.26 | prpplague | ][-DoDo: is your terminal emulation correct? |
| 17:58.46 | prpplague | not all terminal interfaces support numberic key pads |
| 17:59.03 | ][-DoDo | prpplague - I have tried several. My problem being that I am not that familiar with terminal emulation |
| 17:59.15 | ][-DoDo | prpplague - but I have tried linux, xterm, vt100 |
| 18:00.09 | prpplague | ][-DoDo: sounds like you need to do some reading |
| 18:00.22 | prpplague | ][-DoDo: are you using minicom to connect on the serial console? |
| 18:00.29 | CosmicPenguin | minicom sucks |
| 18:00.36 | ][-DoDo | prpplague - I agree wholeheartedly. I am having trouble finding good resource |
| 18:01.21 | prpplague | CosmicPenguin: what do you use? |
| 18:01.48 | prpplague | ][-DoDo: http://tldp.org/HOWTO/Text-Terminal-HOWTO.html |
| 18:02.05 | ][-DoDo | prpplague - connecting from windows with tera term |
| 18:02.29 | prpplague | ][-DoDo: and what emulation do you have set on tera term? |
| 18:03.08 | ][-DoDo | prpplague - vt100 |
| 18:03.36 | prpplague | ][-DoDo: and after you log in to the client what is the TERM variable set to? |
| 18:03.53 | wmat | ][-DoDo: have you tried Putty |
| 18:04.09 | prpplague | wmat: hehe don't confuse him,hehe |
| 18:04.13 | prpplague | wmat: putty is nice btw |
| 18:04.18 | ][-DoDo | prpplague - term is xterm, but I think I have also tried vt100 |
| 18:04.37 | ][-DoDo | wmat - I use putty for ssh, couldnt make it work for serial |
| 18:04.39 | prpplague | ][-DoDo: ok well you have some problems then |
| 18:05.02 | prpplague | ][-DoDo: you terminal type in tera term should match whats set as the TERM variable |
| 18:05.15 | prpplague | ][-DoDo: do you have a termcap and terminfo entries available for vt100? |
| 18:05.33 | ][-DoDo | prpplague - this is where you start to lose me |
| 18:05.46 | wmat | prpplague: putty is the only ssh client i use |
| 18:06.05 | ][-DoDo | prpplague - I actually took the terminfo database from the buildroot that I was previously using |
| 18:06.17 | prpplague | ][-DoDo: and termcap? |
| 18:06.32 | prpplague | ][-DoDo: i'd suggest you go do some reading |
| 18:06.39 | ][-DoDo | prpplague - This is why reading is required - dont know what termcap is |
| 18:06.44 | prpplague | ][-DoDo: there is probably nothing wrong with your ncurses or dialog |
| 18:06.50 | ][-DoDo | prpplague but want to make sure I am reading about the right stuff |
| 18:07.07 | prpplague | ][-DoDo: well go read that text-term howto |
| 18:07.21 | ][-DoDo | prpplague - will do merci |
| 18:07.47 | wmat | http://docs.hp.com/en/5991-6413/ch02s02.html |
| 18:08.01 | wmat | there's a few steps there for a serial console using putty on windows |
| 18:08.26 | ][-DoDo | wmat - thanks if I can use putty for serial that would be nice |
| 18:08.34 | wmat | http://faq.1and1.com/dedicated_servers/windows_server/serial_console_sercon/2.html |
| 18:08.36 | wmat | or there |
| 18:09.22 | wmat | oh wait, that's SSH |
| 18:09.24 | wmat | forget ti |
| 18:10.51 | ][-DoDo | wmat - the problem I have had with putty is I cant even find where to enter the com information |
| 19:40.46 | lkm` | quit |
| 20:08.01 | prpplague | jbevren: http://tinyurl.com/y8qw85 |
| 20:08.07 | prpplague | GPSFan: http://tinyurl.com/y8qw85 |
| 20:08.13 | prpplague | GPSFan: hehe i want to hack that |
| 20:10.48 | CosmicPenguin | wow |
| 20:10.52 | CosmicPenguin | lookit that price |
| 20:11.29 | prpplague | CosmicPenguin: good or bad? |
| 20:11.34 | CosmicPenguin | decent |
| 20:11.35 | GPSFan | prpplague: soon vtech will have a real laptop |
| 20:11.36 | CosmicPenguin | not much storage space |
| 20:11.51 | prpplague | CosmicPenguin: no, it only has about 4k from what i gather |
| 20:12.13 | prpplague | CosmicPenguin: but i really just want it for the plastics and lcd |
| 20:12.26 | prpplague | CosmicPenguin: i'll gut the actual cpu board in favor of an arm board |
| 20:13.21 | prpplague | CosmicPenguin: s3c2410 probably |
| 20:13.44 | prpplague | CosmicPenguin: with just 4 AA's i'll probably run it at 103mhz |
| 20:14.17 | prpplague | CosmicPenguin: i might be surprised and find an arm7tdmi in there |
| 20:15.43 | prpplague | GPSFan: yea, if they'd do something like the zipit in that package i'd buy it in an instant |
| 20:19.11 | GPSFan | prpplague: you and vtech should collaborate on a color version of the zipit. could have 2 form factors, one zipit-sized, one laptop-sized. same guts, arm proc, sd/mmc |
| 20:19.36 | prpplague | GPSFan: yea, i wish i could get a gig like that |
| 20:19.47 | prpplague | GPSFan: or even with oregon scientific |
| 20:20.01 | prpplague | GPSFan: OS has some really nice plastic |
| 20:20.53 | prpplague | GPSFan: http://www2.oregonscientific.com/shop/browse.asp?cid=8&scid=17 |
| 20:21.34 | prpplague | GPSFan: i really like the x18 case - http://www2.oregonscientific.com/assets/product/photos/X18_lg.jpg |
| 20:21.59 | GPSFan | prpplague: indeed, the plastic's/case/"thing the user touches" is the key to widespread acceptance of the product. Of course if it works like crap you will only sell the first few. |
| 20:22.29 | prpplague | GPSFan: yea, that seems to be the rule right now with these |
| 20:22.52 | GPSFan | prpplague: think of those cases with good guts... |
| 20:22.56 | prpplague | GPSFan: most of the companies are spending lots of money on the case but very little on the actual electronics |
| 20:23.01 | prpplague | GPSFan: yea |
| 20:23.07 | prpplague | GPSFan: thats what i'm thinking |
| 20:23.54 | prpplague | GPSFan: i figure i could put a small arm board in there with some serious features for about $70 |
| 20:24.28 | prpplague | GPSFan: actually you could yank the main board from a pixter-mm and shove it into that easy |
| 20:25.28 | prpplague | GPSFan: the educational toy company that "gets it" about linux and the ability to put your own content on their device is gonna be the big winner |
| 20:25.57 | prpplague | GPSFan: parents will pay more for a device that they know that they don't have to purchase a bunch of extra stuff for |
| 20:26.38 | GPSFan | prpplague: if anyone like that ever "gets it" you will have "OLPC-like" devices that sell for 79.95 and will run rings around the current olpc. |
| 20:27.07 | prpplague | GPSFan: well i doubt that it will be that good |
| 20:27.28 | prpplague | GPSFan: i do believe it won't ever been x86 based |
| 20:27.52 | GPSFan | never x86, can't do the mips/watt. |
| 20:28.17 | prpplague | GPSFan: but i can run my s3c2410 board on 4xAA |
| 20:28.25 | prpplague | GPSFan: never tested how long though |
| 20:29.06 | prpplague | GPSFan: we use a 7.2V 1800mAh battery in production stuff |
| 20:29.10 | GPSFan | prpplague: how wide a voltage range can it tolerate? does it have a regulator to take the 4AA to VCC? |
| 20:29.40 | prpplague | GPSFan: well the vcc for the s3c2410 is 3.3v |
| 20:29.45 | prpplague | GPSFan: so yea |
| 20:29.59 | prpplague | GPSFan: but generally you'd probably need some other devices running 5v |
| 20:30.03 | prpplague | GPSFan: i.e. usb host |
| 20:30.42 | GPSFan | you could do a small boost reg. to supply the usb vcc, just watch what you plug into it. |
| 20:30.58 | prpplague | GPSFan: yea |
| 20:33.37 | prpplague | GPSFan: gotta head home, talk at you later |
| 21:39.26 | *** join/#edev silverpower (n=shrike@adsl-234-211-250.cha.bellsouth.net) |
| 23:31.21 | *** join/#edev prpplague (n=dave@69.73.217.80) |
| 23:34.22 | *** join/#edev redblue (i=star@ppp132.108-253-207.mtl.mt.videotron.ca) |
