| 00:00.22 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 00:08.24 | *** join/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca) |
| 00:13.52 | *** join/#edev linac (n=lin@199.90.76.222.broad.fz.fj.dynamic.163data.com.cn) |
| 00:20.48 | *** join/#edev dijenerate (n=dijenera@72.22.129.8) |
| 00:26.09 | *** join/#edev davidc__ (n=davidc__@c-24-8-151-128.hsd1.co.comcast.net) |
| 00:32.53 | *** join/#edev dijenerate (n=dijenera@72.22.129.8) |
| 01:06.08 | *** join/#edev dijenerate (n=dijenera@72.22.129.8) |
| 01:06.27 | *** part/#edev waldzell (n=HokieTux@c-24-127-46-231.hsd1.va.comcast.net) |
| 02:17.32 | *** join/#edev dijenerate (n=dijenera@72.22.129.8) |
| 03:16.40 | *** join/#edev diamat (n=cerne@2001:5c0:8fff:fffe:0:0:0:5b4f) |
| 04:15.35 | *** join/#edev thomasez (n=thomasez@castle.linpro.no) [NETSPLIT VICTIM] |
| 04:15.35 | *** join/#edev ds2 (i=noinf@netblock-66-245-251-24.dslextreme.com) [NETSPLIT VICTIM] |
| 04:27.53 | *** join/#edev thomasez (n=thomasez@castle.linpro.no) |
| 04:27.53 | *** join/#edev ds2 (i=noinf@netblock-66-245-251-24.dslextreme.com) |
| 04:40.30 | *** join/#edev CP|too (n=nobody@www.crouse-house.com) [NETSPLIT VICTIM] |
| 04:40.30 | *** join/#edev dfarnsworth (n=dfarnswo@65.200.49.156) [NETSPLIT VICTIM] |
| 04:40.30 | *** join/#edev ohsix (i=ohsix@bc111222.bendcable.com) [NETSPLIT VICTIM] |
| 04:40.30 | *** join/#edev cbrake_away (n=cbrake@oh-69-34-21-229.sta.embarqhsd.net) |
| 04:40.30 | *** join/#edev catastr0p (n=catastr0@d14-69-178-65.try.wideopenwest.com) [NETSPLIT VICTIM] |
| 04:40.30 | *** join/#edev valverde (n=valverde@ns1.clipsalportal.com) [NETSPLIT VICTIM] |
| 04:40.30 | *** mode/#edev [+o CP|too] by irc.freenode.net |
| 04:41.24 | *** join/#edev dfarnsworth (n=dfarnswo@65.200.49.156) |
| 04:42.34 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 04:55.53 | *** join/#edev thomasez (n=thomasez@castle.linpro.no) |
| 04:55.54 | *** join/#edev ds2 (i=noinf@netblock-66-245-251-24.dslextreme.com) |
| 05:01.34 | *** join/#edev CosmicPenguin (n=nobody@www.crouse-house.com) |
| 05:01.35 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 05:26.02 | *** join/#edev jiver (n=lax@c-69-142-29-26.hsd1.nj.comcast.net) |
| 05:31.38 | jiver | ok who is awake and has a clue about inetd and xeinetd |
| 05:33.13 | *** join/#edev redblue (i=star@ppp088.108-253-207.mtl.mt.videotron.ca) |
| 05:45.06 | *** join/#edev linac (n=lin@199.90.76.222.broad.fz.fj.dynamic.163data.com.cn) |
| 06:20.12 | davidc__ | jiver: know a bit about them |
| 06:51.09 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 07:14.54 | *** join/#edev adyer_ (n=adyer@adsl-69-223-13-78.dsl.chcgil.ameritech.net) |
| 08:07.15 | *** join/#edev Off_Namuh (i=GPS@gateway/tor/x-4c6d0de05fd1045d) |
| 08:33.41 | *** join/#edev Off_Namuh (i=GPS@gateway/tor/x-3eb0553b395fe2ea) |
| 09:32.03 | *** join/#edev howtoo (n=howtoo@195.159.62.66) |
| 10:08.44 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 11:23.11 | *** join/#edev redblue (i=star@ppp186.108-253-207.mtl.mt.videotron.ca) |
| 12:28.22 | *** join/#edev hw (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 12:49.53 | *** join/#edev Off_Namuh (i=GPS@gateway/tor/x-9fbdc2e95d44ffb5) |
| 12:55.41 | *** join/#edev linac (n=lin@121.41.40.202) |
| 13:12.24 | *** join/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net) |
| 13:14.13 | *** join/#edev roxfan[work] (n=dunno@78.218-64-87.adsl-dyn.isp.belgacom.be) |
| 13:48.19 | *** join/#edev GPSFan (n=kenm@12.10.255.246) |
| 13:48.19 | *** mode/#edev [+o GPSFan] by ChanServ |
| 13:59.29 | *** join/#edev prpplague (n=dave@mail.americanmicrosystems.com) |
| 14:28.13 | *** join/#edev dijenerate (n=dijenera@72.22.129.8) |
| 15:13.03 | T0mW | prpplague: You know... I am having the devil's own time of flashing the hammer. |
| 15:13.26 | prpplague | T0mW: oh? do tell |
| 15:13.33 | T0mW | prpplague: I follow the instructions on: http://www.elinux.org/Hammer_How_to_Reflash_Apex_and_kernel |
| 15:13.35 | prpplague | T0mW: you got a fresh apex build done first? |
| 15:13.51 | T0mW | I put the original apex back in place (from cdrom) |
| 15:13.55 | chouimat | morning |
| 15:14.12 | T0mW | prpplague: then, flashed the kernel (from cdrom) |
| 15:14.24 | T0mW | prpplague: then, flashed the rootfs (from cdrom) |
| 15:14.33 | prpplague | T0mW: yea, those binaries have problems |
| 15:14.44 | prpplague | T0mW: all of that has been fixed |
| 15:14.50 | T0mW | and I keep getting either: Uncompressing Linux... |
| 15:14.50 | T0mW | bad gzip magic numbers |
| 15:14.55 | T0mW | or |
| 15:15.03 | prpplague | T0mW: first thing you need to do is build a new binary of apex |
| 15:15.10 | T0mW | No filesystem could mount root, tried: ext2 msdos vfat |
| 15:15.10 | T0mW | Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(1,0) |
| 15:15.20 | T0mW | prpplague: I tried that too |
| 15:15.39 | prpplague | T0mW: ok, let start there |
| 15:15.58 | T0mW | prpplague: I used my arm-elf- to build apex. It works, but, nothing I flashed would, so, I went back to the distro apex |
| 15:16.00 | prpplague | T0mW: you extracted the source, then what |
| 15:16.24 | prpplague | T0mW: you used the source for the wiki ? |
| 15:17.01 | T0mW | cp src/mach-s3c2410/hammer_config ./.config |
| 15:17.03 | T0mW | correct |
| 15:17.14 | T0mW | then I did a 'make |
| 15:17.15 | T0mW | ' |
| 15:17.43 | T0mW | err, make oldconfig ; make |
| 15:17.43 | prpplague | ok, that should be correct, go ahead and flash that binary |
| 15:19.25 | prpplague | T0mW: let me know when you got that done |
| 15:21.18 | T0mW | prpplague: ok, download, erase, flash, reset |
| 15:21.18 | T0mW | # copy $kernelsrc $bootaddr |
| 15:21.18 | T0mW | # copy nor:256k+768k 0x30008000 |
| 15:21.18 | T0mW | 786432 bytes transferred |
| 15:21.18 | T0mW | # copy $ramdisksrc $ramdiskaddr |
| 15:21.18 | T0mW | # copy nor:1024k+1024k 0x30800000 |
| 15:21.18 | prpplague | T0mW: ok, now issue the command "printenv" |
| 15:21.18 | T0mW | 1048576 bytes transferred |
| 15:21.29 | prpplague | hmm |
| 15:21.33 | prpplague | that isn't correct |
| 15:21.35 | T0mW | cmdline *= console=ttySAC0 root=/dev/ram0 |
| 15:21.35 | T0mW | startup *= copy $kernelsrc $bootaddr; copy $ramdisksrc $ramdiskaddr; |
| 15:21.35 | T0mW | ramdisksrc *= nor:1024k+1024k |
| 15:21.35 | T0mW | kernelsrc *= nor:256k+768k |
| 15:21.36 | T0mW | ramdiskaddr *= 0x30800000 |
| 15:21.36 | T0mW | bootaddr *= 0x30008000 |
| 15:21.40 | T0mW | that is printenv |
| 15:21.51 | prpplague | you said you used the source on the wiki? |
| 15:22.03 | T0mW | correct, let me do it again. |
| 15:23.58 | prpplague | T0mW: doh looks like the include config file is incorrect |
| 15:24.05 | T0mW | ok |
| 15:24.06 | prpplague | T0mW: one sec, let me get it for you |
| 15:25.30 | prpplague | T0mW: i almost always do a fresh config, so i guess that default one slipped past without getting updated |
| 15:27.33 | T0mW | prpplague: let me know when to download |
| 15:31.02 | prpplague | T0mW: http://www.elinux.org/Hammer_Board_Software |
| 15:31.04 | prpplague | T0mW: posted |
| 15:32.17 | T0mW | [tom@jtag HammerLCD]$ tar zxf Apex-1.5.6-hammer.tar.gz |
| 15:32.17 | T0mW | [tom@jtag HammerLCD]$ cd apex-1.5.6/ |
| 15:32.17 | T0mW | [tom@jtag apex-1.5.6]$ cp src/mach-s3c2410/hammer_config ./.config |
| 15:32.17 | T0mW | [tom@jtag apex-1.5.6]$ CROSS_COMPILE=../buildroot/build_arm/staging_dir/bin/arm-linux- make |
| 15:33.51 | T0mW | printenv: |
| 15:33.54 | T0mW | cmdline *= console=ttySAC0 root=/dev/ram0 |
| 15:33.54 | T0mW | startup *= copy $kernelsrc $bootaddr; copy $ramdisksrc $ramdiskaddr; |
| 15:33.54 | T0mW | ramdisksrc *= nor:1024k+1024k |
| 15:33.54 | T0mW | kernelsrc *= nor:256k+768k |
| 15:33.54 | T0mW | ramdiskaddr *= 0x30800000 |
| 15:33.57 | T0mW | bootaddr *= 0x30008000 |
| 15:34.01 | T0mW | heh |
| 15:35.51 | *** join/#edev CosmicPe1guin (n=nobody@163.181.251.103) |
| 15:37.12 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 15:37.29 | *** join/#edev CP|too (n=nobody@www.crouse-house.com) |
| 15:37.29 | *** mode/#edev [+o CP|too] by ChanServ |
| 15:39.18 | T0mW | prpplague: that looks like it took: |
| 15:39.21 | T0mW | cmdline *= console=ttySAC0 root=/dev/ram0 |
| 15:39.22 | T0mW | startup *= copy $kernelsrc $bootaddr; copy $ramdisksrc $ramdiskaddr; wait 50 Type ^C key to cancet |
| 15:39.22 | T0mW | ramdisksrc *= nor:1280k+1024k |
| 15:39.22 | T0mW | kernelsrc *= nor:256k+1024k |
| 15:39.22 | T0mW | ramdiskaddr *= 0x30800000 |
| 15:39.22 | T0mW | bootaddr *= 0x30008000 |
| 15:39.47 | prpplague | T0mW: yep |
| 15:39.50 | prpplague | T0mW: thats correct |
| 15:41.27 | T0mW | building new kernel now |
| 15:47.48 | T0mW | new kernel boots ok, now for rootfs |
| 15:56.36 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 16:04.10 | prpplague | T0mW: the rootfs is pretty generic |
| 16:04.25 | prpplague | T0mW: just as long as the total size of the compressed ramdisk is less than 1024k |
| 16:04.59 | prpplague | T0mW: there are a couple minor fixes that you might want to do |
| 16:06.29 | prpplague | T0mW: in the target skeleton for the inittab, edit it and move the mount for proc up to be the first item that is mounted |
| 16:06.39 | T0mW | prpplague: yeah, just making sure of my starting point: that I can rebuilt a bootable system if I need to drop back to a known-good |
| 16:06.51 | T0mW | ok |
| 16:06.56 | prpplague | T0mW: yep |
| 16:15.15 | *** join/#edev ico2 (n=ico2@modem-294.chameleon.dialup.pol.co.uk) |
| 16:18.21 | T0mW | prpplague: DOH! borked apex, heh |
| 16:18.33 | prpplague | T0mW: doh, how did you do that? |
| 16:18.46 | prpplague | T0mW: not going to use the flyswatter? |
| 16:18.47 | CosmicPenguin | BDI \o/ |
| 16:18.55 | T0mW | LOL burned the rootfs over it. |
| 16:18.56 | T0mW | heh |
| 16:19.14 | prpplague | T0mW: doh, was it a fat finger or just wrong address? |
| 16:19.14 | T0mW | ~lart T0mW for being careless |
| 16:19.14 | purl | whacks T0mW with the cluebat for being careless |
| 16:19.56 | T0mW | prpplague: my mistake, I was pasting from the wrong area of the webpage |
| 16:20.03 | prpplague | ahh |
| 16:20.47 | T0mW | ok, I can build a good apex + kernel. Time to move to OE and get them built under that, for now. |
| 16:21.04 | prpplague | T0mW: we were suppose to go live on the web site last week, but the credit card stuff won't be complete until thursday |
| 16:21.23 | T0mW | prpplague: oh, good, what will be the cost on the Hammer? |
| 16:21.46 | prpplague | T0mW: http://www.tincantools.com/home.php |
| 16:22.12 | prpplague | T0mW: volume discounts ofcourse |
| 16:22.31 | T0mW | prpplague: that is reasonable. I just may have a volume sale for you. |
| 16:22.38 | prpplague | T0mW: good deal |
| 16:23.08 | prpplague | T0mW: btw, i do have a hack for the board to use 3.3v input and to move the led gpio to an external pin |
| 16:26.15 | T0mW | prpplague: I just had a customer asking about adding ethernet to an ARM7TDMI design of mine. IMO, it may be cheaper for them to use the Hammer as the hardware is engineered out and it may depend on their volume for their project. |
| 16:26.42 | T0mW | TCP stack implementation on an ARM7TDMI could be "interesting" |
| 16:26.58 | prpplague | T0mW: what kind of demands on the ethernet are they going to have? |
| 16:27.03 | prpplague | T0mW: i.e. 10baseT enough? |
| 16:27.08 | T0mW | yes |
| 16:27.30 | prpplague | T0mW: the enc28j60 would do nicely then |
| 16:28.32 | prpplague | T0mW: if you need to, you could use this with your arm7tdmi design - http://wiznet.co.kr/pro_iin_wiz810mj.htm |
| 16:28.43 | prpplague | T0mW: it has the tcp/ip stack hardcoded inside |
| 16:29.11 | prpplague | T0mW: its based on this chip - http://wiznet.co.kr/pro_iin_W5100.htm |
| 16:30.04 | prpplague | T0mW: of course i'd rather see you use the hammer board :P |
| 16:30.28 | prpplague | CosmicPenguin: the new battery is looking lovely |
| 16:30.43 | T0mW | prpplague: well, they have a point-to-point (single tcp socket) RS232-->10BaseT module that they currently use. |
| 16:31.02 | prpplague | T0mW: ahh |
| 16:31.15 | prpplague | CosmicPenguin: 5100mAh at 3.7v with fuel gauge |
| 16:32.07 | prpplague | CosmicPenguin: (techincally its a 5200mAh, but give it a little slack) |
| 16:35.12 | *** join/#edev ico2_ (n=ico2@modem-1382.chimpanzee.dialup.pol.co.uk) |
| 16:35.27 | adyer_ | for your arm7 tcp/ip check out http://www.sics.se/~adam/lwip/ |
| 16:36.34 | adyer_ | or even http://www.sics.se/~adam/uip/index.php/Main_Page |
| 16:37.53 | T0mW | adyer_: thanks |
| 16:38.23 | CosmicPenguin | prpplague: nice |
| 17:03.27 | prpplague | interesting |
| 17:03.44 | prpplague | i wonder how hard it would be to use one of those as the primary stack under linux |
| 17:03.58 | prpplague | as the linux tcp/ip is way bloated these days |
| 17:04.18 | adyer_ | dunno. I think eCos uses lwip as one of the options for it's tcp/ip |
| 17:05.15 | prpplague | interesting, looks like it could be done pretty easy |
| 17:06.20 | adyer_ | http://rtl-lwip.sourceforge.net/ |
| 17:07.42 | adyer_ | fair number of other hits googling lwip for linux |
| 17:07.51 | prpplague | thanks i'll have a look, but i doubt i'll have time to play, i'm totally swampped |
| 17:08.55 | prpplague | adyer_: yea, looks like it's supported as part of rtlinux |
| 17:10.44 | prpplague | interesting |
| 17:10.57 | prpplague | i'll have to see if i can patch up my hammer kernel build for a test |
| 17:13.12 | prpplague | interesting |
| 17:13.29 | prpplague | would be nice if that was part of the -tiny tree |
| 17:13.37 | CosmicPenguin | patches welcome |
| 17:14.28 | prpplague | CosmicPenguin: you working with the -tiny tree? |
| 17:14.40 | CosmicPenguin | not really |
| 17:14.48 | prpplague | oh |
| 17:15.22 | prpplague | CosmicPenguin: hehe, you familiar with the openmoko neo1973 phone? |
| 17:16.19 | CosmicPenguin | not at all |
| 17:16.25 | prpplague | CosmicPenguin: ahh ok |
| 17:30.37 | *** join/#edev linac (n=lin@121.41.38.171) |
| 17:36.01 | CosmicPenguin | I haven't noticed |
| 17:37.17 | prpplague | CosmicPenguin: i didn't figure there would be since we haven't officially annouced the hammer |
| 17:37.23 | prpplague | CosmicPenguin: but i was just curious |
| 17:37.45 | prpplague | CosmicPenguin: i'm really anxious for them to go live with the website |
| 17:51.26 | *** join/#edev adyer_ (n=adyer@adsl-69-223-13-78.dsl.chcgil.ameritech.net) [NETSPLIT VICTIM] |
| 17:51.26 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) [NETSPLIT VICTIM] |
| 18:34.44 | *** join/#edev ico2__ (n=ico2@modem-841.cougar.dialup.pol.co.uk) |
| 18:36.55 | ds2 | morning |
| 18:43.00 | chouimat | grrr |
| 18:43.23 | prpplague | chouimat: already gotten fired? |
| 18:43.41 | *** join/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net) |
| 18:43.48 | chouimat | prpplague: no next tuesday a new moorcock is released |
| 18:44.43 | CosmicPenguin | a what? |
| 18:44.51 | chouimat | new moorcock novel |
| 18:45.16 | prpplague | CosmicPenguin: an author |
| 18:45.41 | chouimat | which will bring the number of unread books in my collection to around 150 ... |
| 18:46.03 | prpplague | CosmicPenguin: http://en.wikipedia.org/wiki/Michael_Moorcock |
| 18:48.38 | *** join/#edev beadyallen (n=ben@82-46-232-54.cable.ubr07.newy.blueyonder.co.uk) |
| 18:51.26 | *** join/#edev Off_Namuh (i=GPS@gateway/tor/x-166e58f4fa13be4b) |
| 18:51.32 | prpplague | ho ho hum |
| 18:57.01 | T0mW | prpplague: S3C2410 is armv4t ?? and arm922t ? |
| 18:57.26 | prpplague | i keep meaning to post a elinux page on that |
| 18:57.33 | T0mW | or is it arm5t |
| 18:58.03 | chouimat | i think it's a am920t |
| 18:58.04 | prpplague | T0mW: arm920t and arm922t are armv4t |
| 18:58.23 | prpplague | T0mW: arm926t is armv5tej |
| 18:59.03 | T0mW | ah |
| 18:59.10 | *** join/#edev Helios_ (n=marius@p57A88DB4.dip0.t-ipconnect.de) |
| 18:59.31 | T0mW | I knew it should be arm922t, wasn't sure about opcode set though |
| 19:00.37 | prpplague | ?? |
| 19:00.55 | prpplague | T0mW: the s3c2410 is an arm920t |
| 19:02.50 | *** join/#edev ico2 (n=ico2@modem-841.cougar.dialup.pol.co.uk) |
| 19:02.56 | T0mW | doh |
| 19:36.26 | T0mW | where are people coming up with these binutils versions?: "binutils-2.16.91.0.6" "binutils-2.17.50.0.12" |
| 19:36.39 | T0mW | what do they do? smoke crack and make them up? |
| 19:36.45 | T0mW | heh |
| 19:37.07 | prpplague | http://www.elinux.org/ARM_Processor |
| 19:37.27 | prpplague | T0mW: that should help clear up some of the confusion on the arm isa |
| 19:39.49 | T0mW | well, transmission is bad in the car: $2800 |
| 19:40.06 | chouimat | ouch |
| 19:45.03 | T0mW | chouimat: that is for a factory rebuilt unit, 100K warranty |
| 19:45.14 | chouimat | nice |
| 19:45.30 | T0mW | 100K miles, not 100K kilometers, heh |
| 19:45.50 | chouimat | ok the 160K one :) |
| 19:46.24 | T0mW | chouimat: well, I cannot trade the car in. I drilled holes in the trunk lid to mount antennas on the car. :P |
| 19:46.42 | steliosk | prpplague : Do you maintain the elinux wiki ? |
| 19:46.50 | *** join/#edev Off_Namuh (i=GPS@gateway/tor/x-0bbccb55853d100f) |
| 19:46.55 | steliosk | I have a problem trying to download http://elinux.org/index.php?title=Special:Upload&wpDestFile=Ppc405_gpio.tar.gz |
| 19:47.08 | steliosk | it thinks i am trying to upload it.... |
| 19:47.24 | prpplague | steliosk: ok, one sec, let me look |
| 19:50.17 | prpplague | steliosk: looks like that file didn't get transfered from the old wiki |
| 19:50.21 | prpplague | steliosk: let me look |
| 19:50.49 | *** join/#edev joe_bleau (n=joe_blea@adsl-69-152-195-79.dsl.fyvlar.swbell.net) |
| 19:51.52 | prpplague | steliosk: yea, looks like that was one of the file flameman destroyed |
| 19:56.16 | prpplague | steliosk: what page were you trying to down it from? |
| 19:56.39 | steliosk | prpplague : got it now. Thanks ! |
| 19:56.42 | steliosk | page is http://elinux.org/DHT-Walnut-Flameman#JTAG_.26_OCD_Commander |
| 19:57.11 | prpplague | steliosk: yea flameman totally trashed the DHT-Walnut pages |
| 19:58.31 | ds2 | prpplague: nice table with the ARM bits, thanks! |
| 19:58.38 | prpplague | ds2: np |
| 19:59.32 | steliosk | prpplague : looks like it. |
| 20:00.07 | prpplague | steliosk: he came in one day and just deleted the entire DHT-walnut pages and placed an "under contruction note" |
| 20:01.26 | steliosk | There are some very interesting bits of info for the ppc there. The wiggler stuff especially |
| 20:02.38 | joe_bleau | GPSFan did a lot of the wiggler work on that project. |
| 20:05.03 | prpplague | T0mW: hey don't suppose you know a distributor that sells chrontel parts? |
| 20:13.05 | T0mW | no |
| 20:13.21 | T0mW | they may only direct sell? |
| 20:13.26 | prpplague | hmm, they have a lovely digital lcd input to vga chipset |
| 20:14.54 | T0mW | prpplague: yeah, I saw that one too |
| 20:15.04 | adyer_ | looks like sager carries them |
| 20:15.27 | prpplague | adyer_: sager? |
| 20:15.29 | prpplague | url? |
| 20:15.44 | adyer_ | www.sager.com |
| 20:16.20 | adyer_ | hmm dont see it on their line card |
| 20:19.34 | adyer_ | if you don't need scaler functionality, lot of time you can hook up a video dac and fiddle the timing to make it work. something like ADV7120 |
| 20:20.44 | prpplague | just looking for very basic composit and vga support |
| 20:22.11 | adyer_ | most dacs do r/g/b video or composite, not so many do both |
| 20:25.07 | prpplague | yea, that looks like a nice chip, but i need something a little more robust like the ch7013 |
| 20:26.54 | prpplague | adyer_: looks like they are non-stock everywhere |
| 20:31.43 | adyer_ | yeah chrontel is a bit of a pita |
| 20:34.25 | adyer_ | I think conexant and focus make similar parts, AD, TI and NXP make some less fancy parts |
| 20:36.17 | prpplague | adyer_: hmm, the chrontel rep is going to call me back with some pricing and some samples |
| 20:36.23 | prpplague | adyer_: but i'll have a look at nxp |
| 20:36.31 | prpplague | adyer_: we do alot of biz with nxp |
| 20:37.36 | *** join/#edev ico2_ (n=ico2@modem-1299.chimpanzee.dialup.pol.co.uk) |
| 20:38.07 | adyer_ | mostly the nxp stuff takes in 4:2:2 or ITU-601 streams. don't know about straight RGB. If you wanted I bet you could do it with some R2R ladder dacs an op-amp or 2 and a few caps ;-) |
| 20:39.29 | prpplague | adyer_: 4:2:2 would be nice |
| 20:39.52 | adyer_ | nxp's web site is second in suckness only to ST's |
| 20:41.39 | adyer_ | try SAA7121, been around for a long time |
| 20:45.03 | prpplague | hmm, nice |
| 20:45.09 | prpplague | adyer_: know one that is rgb? |
| 20:59.55 | *** join/#edev ico2__ (n=ico2@modem-1299.chimpanzee.dialup.pol.co.uk) |
| 21:01.20 | ds2 | does anyone know what name the part of motorola that did the GPS boards before is call now? |
| 21:01.33 | ds2 | should've saved off the datasheets fromthose bgmicro GPS boards :( |
| 21:10.15 | adyer_ | http://focus.ti.com/docs/prod/folders/print/ths8200.html |
| 21:16.36 | adyer_ | http://www.focussemi.com/product_family/video_convergence.html |
| 21:22.28 | prpplague | interesting |
| 21:22.42 | prpplague | adyer_: now that i'm looking, i'm not sure what i have in mind will work |
| 21:28.17 | adyer_ | what kind of video are you starting with? 6/6/6 TTL RGB + syncs? |
| 21:29.36 | prpplague | 8-bit cstn interface |
| 21:30.10 | adyer_ | eww. |
| 21:30.29 | adyer_ | for dual scan or single scan panel? |
| 21:30.44 | prpplague | single scan |
| 21:30.56 | adyer_ | at least that's better |
| 21:31.29 | adyer_ | VGA should be doable, composite a bit trickier |
| 21:31.50 | adyer_ | interlace and the 525 line thing kill you |
| 21:33.00 | prpplague | mainly i'd like a nice clean way to do composite, doesn't nessarily have to be via the cstn interface |
| 21:34.38 | adyer_ | http://www.freescale.com/files/32bit/doc/app_note/AN2639.pdf?fsrch=1 |
| 21:34.50 | adyer_ | an example of adapting a 15-bit lcd interface to tv-out |
| 21:36.58 | prpplague | thanks i'll have a look, but i'm limited to 8-bits |
| 21:37.52 | ds2 | prpplague: isn't composite officially obsolete in about 1year? |
| 21:38.06 | prpplague | ds2: no |
| 21:38.47 | adyer_ | ds2 no analog broadcast is, composite will live on for a long time |
| 21:39.05 | adyer_ | ds2 cable analog goes away a year or two later |
| 21:39.29 | ds2 | Hmmmm then I don't see what's the big deal about having to put warnings on consumer NTSC items :/ |
| 21:39.55 | adyer_ | NTSC is broadcast, hance the warnings |
| 21:40.16 | ds2 | thought NTSC == Composite? |
| 21:40.27 | adyer_ | once you demodulate it, then it's composite |
| 21:40.40 | ds2 | Oh I see |
| 21:41.20 | T0mW | prpplague: maybe you could use an S-Video type output instead? |
| 21:41.27 | adyer_ | at least in the common uses of the terms |
| 21:41.49 | prpplague | T0mW: not sure how that would make a difference |
| 21:42.01 | adyer_ | S-video isn't really that different, it just separates the Y/C from luma (IIRC) |
| 21:42.18 | adyer_ | still interleaved, still 525 lines, etc |
| 21:44.01 | prpplague | http://www.rickard.gunee.com/projects/video/sx/howto.php |
| 21:44.10 | prpplague | maybe i should just use that |
| 21:45.36 | adyer_ | that's pretty much what i was talking about with the joke about R-2R ladder dacs |
| 21:46.17 | prpplague | hmm, |
| 21:48.35 | adyer_ | http://www.siliconmotion.com.tw/en/en2/products_sm502.html :-) |
| 21:49.47 | prpplague | hmm, i'd really like to try that composite video method |
| 21:49.54 | adyer_ | sm501 code is already in tree ;-) |
| 21:50.46 | prpplague | interesting |
| 21:50.54 | prpplague | not as much fun though, hehe |
| 21:52.07 | adyer_ | if you do the composite thing you might be better off doing s-video style stuff, alternate pixels drive Y/CrCb and Luma DACs |
| 21:52.39 | prpplague | adyer_: my main concern is that i don't want to tax the cpu too heavily |
| 21:53.04 | CosmicPenguin | taxes are bad |
| 21:53.09 | prpplague | adyer_: this guy does it with a pic@50mhz, so i figure that the s3c2410@200 should be ok |
| 21:53.21 | prpplague | CosmicPenguin: indeed |
| 21:53.31 | prpplague | the main problem is going to be latency |
| 21:53.41 | adyer_ | well I think modulating the color info on top of the luma in the CPU is more taxing than doing it externally in analog |
| 21:53.49 | prpplague | if there is a way i could use the lcd controller to do the data dump, that might work |
| 21:54.03 | prpplague | adyer_: yea |
| 21:54.29 | prpplague | adyer_: if i get one of those 8-bit capable d2a's that would help |
| 21:54.44 | adyer_ | if you have a decent dma controller you could do it with an external FIFO and some dacs. maybe an fpga |
| 21:56.17 | prpplague | adyer_: i'd like to steer clear of fpga's |
| 21:59.10 | prpplague | ds2: ntsc is a standard of data transmited via composite |
| 21:59.33 | adyer_ | I love them. it's so liberating to be able to make something work the way you want :-) |
| 21:59.41 | prpplague | ds2: you can transmit pal or just about anything else via composite |
| 21:59.56 | prpplague | adyer_: indeed, just that they are out of reach for most hobbists |
| 22:00.33 | adyer_ | ? spartan-3E in tqfp is only like $10 in small qty. |
| 22:01.24 | adyer_ | i agree there's a learning curve |
| 22:02.25 | adyer_ | and to get really good takes a while |
| 22:02.51 | prpplague | adyer_: another aspect is , as i understand it, not a whole lot of linux dev tools for fpgas |
| 22:03.31 | prpplague | http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=296-3001-ND |
| 22:03.34 | prpplague | pretty pricy |
| 22:04.17 | adyer_ | Xilinx webpack supports linux, and the big parts are pricey, but you can do a lot in the 'little' ones |
| 22:04.37 | adyer_ | i believe altera also supports linux |
| 22:04.59 | prpplague | adyer_: my pricy comment was about the TI video chip |
| 22:05.22 | adyer_ | ahh yes. DACs are expensive, especially fast ones |
| 22:05.22 | prpplague | adyer_: the xilinx tools support linux? |
| 22:05.41 | prpplague | adyer_: i could do a custom frame buffer |
| 22:06.29 | adyer_ | http://www.xilinx.com/ise/ossupport/index.htm |
| 22:06.29 | prpplague | adyer_: use it in 8-bit grey scale and code the data for Y-Pb-Pr |
| 22:06.51 | adyer_ | you'll need a frame buffer store, not enough mem inside the fpga for all that |
| 22:06.53 | prpplague | adyer_: hmm, only listing is for RHEL |
| 22:07.02 | prpplague | adyer_: yea, the s3c2410 has one |
| 22:07.26 | adyer_ | yes, but it works under others, that's just the official support list |
| 22:07.43 | prpplague | ahh |
| 22:08.51 | adyer_ | if you put frame buffer control in an fpga, you'll have to dma through the samsung part to get at the ram |
| 22:09.28 | prpplague | adyer_: naw, just use the standard frame buffer in linux just code it differently |
| 22:18.42 | adyer_ | are you using someone else's board? |
| 22:21.41 | prpplague | http://www.elinux.org/Hammer_Board |
| 22:31.51 | adyer_ | i looked at 2410, it will do 8/16/24 bits per pixel in tft mode, but you have to demux them. If you feed the 8 bit data out into 3 cascaded 8-bit flip-flops you can take 24 bits at a time into an RGB dac and do VGA |
| 22:32.13 | adyer_ | (assuming the bandwidth is OK) |
| 22:33.27 | prpplague | adyer_: indeed it will do 8/16/24, however on the hammer board we only brought out 8-bit data for the lcd |
| 22:34.04 | adyer_ | if you look it puts 24 bits out as 3 bytes of 8 bits wide |
| 22:34.45 | prpplague | adyer_: on 24 data lines |
| 22:34.54 | adyer_ | no, on 8 |
| 22:35.02 | prpplague | i beg to differ |
| 22:35.23 | *** join/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca) |
| 22:35.50 | MonMotha | prpplague: as far as Linux FPGA dev tools, Xilinx ISE runs on Linux at this point |
| 22:35.57 | *** part/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca) |
| 22:36.12 | MonMotha | there's a couple pieces of functionality provided by 3rd party apps that you miss out on (mostly simulation related), but ISE, including XST runs |
| 22:36.18 | prpplague | adyer_: when using tft mode for 24-bpp, you use all 24 data lines |
| 22:36.26 | MonMotha | so you can run the IDE, synthesize, and there is a cable driver to program thing thing |
| 22:36.33 | prpplague | MonMotha: ahh, i need to test it under linux |
| 22:36.40 | prpplague | MonMotha: we've been using a winblows box |
| 22:37.56 | prpplague | adyer_: for stn mode you can do 8-bit with up to 12bpp |
| 22:38.34 | MonMotha | prpplague: I've at least verified that it does install and start up |
| 22:38.49 | MonMotha | there's also no EDK for Linux yet (which is odd, since it's mostly TCL/Bash scripts and GNU tools running under Cygwin) |
| 22:38.50 | prpplague | MonMotha: ahh cool |
| 22:39.09 | prpplague | MonMotha: is the programming protocol closed? |
| 22:39.22 | MonMotha | but unless you want to use one of their premade embedded procs (microblaze/picoblaze or the embedded PPC on the VIrtex series), you don't need that |
| 22:39.31 | prpplague | MonMotha: ahh |
| 22:39.34 | MonMotha | no, they tell you everything you need to download the bitstream via JTAG |
| 22:39.44 | adyer_ | ahh, you're right. could still use the same trick though, just make the framebuffer 3x wider ;-) |
| 22:39.50 | prpplague | MonMotha: ahh cool, i probably could use my flyswatter board then |
| 22:40.09 | MonMotha | I don't think Digilent has released API specs on their USB JTAG cable, though |
| 22:40.23 | MonMotha | and I don't think Xilinx releases specs on their USB JTAG cable, either, though they do provide a driver (which requires root) |
| 22:40.38 | prpplague | adyer_: again, the only problem is i only have LVD7:0 |
| 22:41.13 | adyer_ | clock it through 3 levels of flip flops, then take it broadside as 24-bits |
| 22:41.23 | prpplague | adyer_: ahh right |
| 22:41.50 | prpplague | adyer_: hmm, |
| 22:41.54 | adyer_ | xilinx usb cable has been reverse engineered by laurent pinchart |
| 22:42.10 | adyer_ | and xilinx parallel cable works ok |
| 22:42.20 | prpplague | adyer_: lets see, i'd probably have to run the lcd at 90hz refresh to get all that data in there properly |
| 22:43.31 | prpplague | adyer_: nice idea |
| 22:43.37 | prpplague | adyer_: i'll have a look |
| 22:43.59 | adyer_ | 'round 80MHz pixclk |
| 22:44.17 | adyer_ | for 640x480x60hz ;-) |
| 22:44.45 | prpplague | adyer_: gonna have to a little higher to compensate for the 3 clock cycles for data latching |
| 22:44.45 | adyer_ | only 50Mhz for 16 bit |
| 22:45.36 | prpplague | adyer_: that'd work ok for vga, but i was planning smaller for the composit |
| 22:46.26 | adyer_ | no, think of it as 8 parallel 3-bit shift regs. running at 3x (or 2x) what clock you really want. |
| 22:48.07 | prpplague | adyer_: yea, i was thinking 30frames/sec x 3 = 90 |
| 22:48.44 | adyer_ | vga runs ~ 60 frames/sec |
| 22:49.04 | adyer_ | i suppose you could slow it down and use 320x200 mode instead of 640x480 |
| 22:49.11 | prpplague | composite = 30 frames |
| 22:49.13 | prpplague | :) |
| 22:49.53 | adyer_ | and interleaved with a weird color space |
| 22:50.01 | prpplague | yea |
| 22:50.11 | prpplague | adyer_: just rough guys at the moment |
| 22:50.26 | prpplague | adyer_: i'll see about ordering parts and do an initial trial with a b&W |
| 22:53.01 | adyer_ | here's an example of what I meant |
| 22:53.20 | prpplague | adyer_: schematic example? |
| 22:53.47 | adyer_ | yes, video comes in as 8 bits on top left, flows through 2 regs and broad sides into a dac |
| 22:54.28 | adyer_ | 8 bit video is clocked at 3x eventual pixel rate to the dac |
| 22:55.13 | adyer_ | been in production for a while... |
| 22:57.27 | *** join/#edev Off_Namuh (i=GPS@gateway/tor/x-efd4a5600b56b416) |
| 23:05.55 | prpplague | adyer_: url? |
| 23:12.40 | *** join/#edev redblue (i=star@ppp011.108-253-207.mtl.mt.videotron.ca) |
| 23:16.51 | *** join/#edev CosmicPe1guin (n=nobody@163.181.251.103) |
| 23:17.16 | *** mode/#edev [+o CosmicPenguin] by ChanServ |
| 23:25.54 | ds2 | T0mW (if you are around), this should work with your zipit board, right: |
| 23:25.56 | ds2 | http://www.bgmicro.com/index.asp?PageAction=VIEWPROD&ProdID=10556 |
| 23:36.46 | adyer_ | sorry, not on web, tried to dcc you and it failed |